Text Book Full Report of Word Economy PDF

Wor l d Eco no m i c a nd F i na nci a l S ur v e y s
World Economic Outlook

April 2018
Cyclical Upswing, Structural Change
I N T E R N A T I O N A L M O N E T A R Y F U N D
©2018 International Monetary Fund
Cover and Design: Luisa Menjivar and Jorge Salazar

Composition: AGS, An RR Donnelley Company
Cataloging-in-Publication Data
Joint Bank-Fund Library
Names: International Monetary Fund.

Title: World economic outlook (International Monetary Fund)
Other titles: WEO | Occasional paper (International Monetary Fund) | World economic and
financial surveys.
Description: Washington, DC : International Monetary Fund, 1980- | Semiannual | Some
issues also have thematic titles. | Began with issue for May 1980. | 1981-1984: Occasional
paper / International Monetary Fund, 0251-6365 | 1986-: World economic and financial
surveys, 0256-6877.
Identifiers: ISSN 0256-6877 (print) | ISSN 1564-5215 (online)
Subjects: LCSH: Economic development—Periodicals. | International economic relations—
Periodicals. | Debts, External—Periodicals. | Balance of payments—Periodicals. |
International finance—Periodicals. | Economic forecasting—Periodicals.
Classification: LCC HC10.W79
HC10.80
ISBN 978-1-48433-827-8 (paper)

978-1-48434-971-7 (Web PDF)
978-1-48434-973-1 (ePub)
978-1-48434-974-8 (Mobi)
The World Economic Outlook (WEO) is a survey by the IMF staff published twice a
year, in the spring and fall. The WEO is prepared by the IMF staff and has benefited
from comments and suggestions by Executive Directors following their discussion of the
report on April 2, 2018. The views expressed in this publication are those of the IMF
staff and do not necessarily represent the views of the IMF’s Executive Directors or their
national authorities.
Recommended citation: International Monetary Fund. 2018. World Economic Outlook:

Cyclical Upswing, Structural Change. Washington, DC, April.
Publication orders may be placed online, by fax, or through the mail:

International Monetary Fund, Publication Services
P.O. Box 92780, Washington, DC 20090, USA
Tel.: (202) 623-7430 Fax: (202) 623-7201
E-mail: publications@imf.org
www.imfbookstore.org
www.elibrary.imf.org
CONTENTS
Assumptions and Conventions x
Further Information xi
Data xii
Preface xiii
Foreword xiv
Executive Summary xvi
Chapter 1. Global Prospects and Policies 1

Recent Developments and Prospects 1
The Forecast 12
Risks 21
Policy Priorities 24
Scenario Box 1: Impact of Tighter Global Financial Conditions 32
Box 1.1. Smartphones and Global Trade 34
Box 1.2. What Has Held Core Inflation Back in Advanced Economies? 37
Box 1.3. Recent Dynamics of Potential Growth 40
Box 1.4. Has Mismeasurement of the Digital Economy Affected Productivity Statistics? 44
Box 1.5. Macroeconomic Impact of Corporate Tax Policy Changes 46
Box 1.6. A Multidimensional Approach to Trade Policy Indicators 48
Box 1.7. Growth Outlook—Advanced Economies 51
Box 1.8. Growth Outlook—Emerging Market and Developing Economies 52
Box 1.9. Inflation Outlook—Regions and Countries 54
Special Feature: Commodity Market Developments and Forecasts 55
Box 1.SF.1. The Role of Metals in the Economics of Electric Vehicles 60
References 69
Chapter 2. Labor Force Participation in Advanced Economies: Drivers and Prospects 71

Introduction 71
Patterns of Labor Force Participation in Advanced Economies 74
Understanding Trends in Participation Rates 80
Prospects for Labor Force Participation 92
Conclusions and Policy Implications 95
Box 2.1. Youth Labor Force Participation in Emerging Market and Developing Economies versus
Advanced Economies 97
Box 2.2. Permanently Displaced? Labor Force Participation in US States and Metropolitan Areas 100
Box 2.3. Still Attached? Labor Force Participation Trends in European Regions 103
Box 2.4. Storm Clouds Ahead? Migration and Labor Force Participation Rates 106
Annex 2.1. Data Sources and Country Coverage 110
Annex 2.2. Additional Stylized Facts 111
Annex 2.3. The Role of Aging and Cyclical Factors 113
Annex 2.4. The Role of Policies and Other Factors: Aggregate Cross-Country Analysis 114
International Monetary Fund | April 2018 iii

WORLD ECONOMIC OUTLOOK: CYCLICAL UPSWING, STRUCTURAL CHANGE
Annex 2.5. The Role of Individual and Household Characteristics: Micro-Level Analysis 122
Annex 2.6. Prospects for Labor Force Participation: Cohort-Based Analysis 122
References 123
Chapter 3. Manufacturing Jobs: Implications for Productivity and Inequality 129

Introduction 129
Structural Transformation: Key Trends and Drivers 132
Growth and Development beyond Manufacturing 139
Implications for Income Inequality 147
Box 3.1. The Changing Service Content of Manufactures 152
Box 3.2. The Rise of Services Trade 155
Box 3.3. Are Manufacturing Jobs Better Paid? Worker-Level Evidence from Brazil 157
Annex 3.1. Data Sources and Country Coverage 159
Annex 3.2. Value-Added Decomposition 160
Annex 3.3. Sectoral Productivity, Aggregate Growth, and Convergence 161
Annex 3.4. Manufacturing and Inequality 167
References 168
Chapter 4. Is Productivity Growth Shared in a Globalized Economy? 173

Introduction 173
Conceptual Framework 175
Measuring Innovation 176
The Innovation Landscape 177
Determinants of Knowledge Flows 180
The Impact on Innovation and Productivity 181
The Impact of Global Value Chains on Patenting: A Firm-Level Analysis 185
The Role of Greater International Competition 187
Box 4.1. Patent Data and Concepts 191
Box 4.2. International Technology Sourcing and Knowledge Spillovers 193
Box 4.3. The Role of Foreign Aid in Improving Productivity in Low-Income Developing Countries 195
Box 4.4. Relationship between Competition, Concentration, and Innovation 198
Annex 4.1. Data, Sample, and Variable Definition 199
Annex 4.2. Determinants of Knowledge Flows: Additional Results 201
Annex 4.3. Impact of Foreign Knowledge on Domestic Innovation and Productivity: Additional
Results for Panel Estimation of Long-Term Relationships 204
Annex 4.4. Methodology for Local Projection Method Estimation 208
Annex 4.5. Impact of Global Value Chains on Firm-Level Patenting: Methodology and Robustness 209
References 212
Statistical Appendix 215

Assumptions 215
What’s New 216
Data and Conventions 216
Country Notes 217
Classification of Countries 218
General Features and Composition of Groups in the World Economic Outlook Classification 218
Table A. Classification by World Economic Outlook Groups and Their Shares in Aggregate GDP,
Exports of Goods and Services, and Population, 2017 219
iv International Monetary Fund | April 2018

CONTENTS
Table B. Advanced Economies by Subgroup 220

Table C. European Union 220
Table D. Emerging Market and Developing Economies by Region and Main Source of Export Earnings 221
Table E. Emerging Market and Developing Economies by Region, Net External Position, and
Status as Heavily Indebted Poor Countries and Low-Income Developing Countries 222
Table F. Economies with Exceptional Reporting Periods 224
Table G. Key Data Documentation 225
Box A1. Economic Policy Assumptions Underlying the Projections for Selected Economies 235
List of Tables 239
Output (Tables A1–A4) 240
Inflation (Tables A5–A7) 247
Financial Policies (Table A8) 252
Foreign Trade (Table A9) 253
Current Account Transactions (Tables A10–A12) 255
Balance of Payments and External Financing (Table A13) 262
Flow of Funds (Table A14) 266
Medium-Term Baseline Scenario (Table A15) 269
World Economic Outlook, Selected Topics 271
IMF Executive Board Discussion of the Outlook, April 2018 281
Tables
Table 1.1. Overview of the World Economic Outlook Projections 14
Annex Table 1.1.1. European Economies: Real GDP, Consumer Prices,
Current Account Balance, and Unemployment 62
Annex Table 1.1.2. Asian and Pacific Economies: Real GDP, Consumer Prices,
Annex Table 1.1.3. Western Hemisphere Economies: Real GDP, Consumer Prices,
Annex Table 1.1.4. Commonwealth of Independent States Economies: Real GDP,
Consumer Prices, Current Account Balance, and Unemployment 65
Annex Table 1.1.5. Middle East, North African Economies, Afghanistan, and Pakistan:
Real GDP, Consumer Prices, Current Account Balance, and Unemployment 66
Annex Table 1.1.6. Sub-Saharan African Economies: Real GDP, Consumer Prices,
Annex Table 1.1.7. Summary of World Real per Capita Output 68
Table 2.1. Drivers of Labor Force Participation Rates 85
Table 2.2.1. Drivers of Labor Force Participation Rates in US Metropolitan Areas 102
Table 2.3.1. Drivers of Labor Force Participation Rates in European Regions 105
Annex Table 2.1.1. Data Sources 110
Annex Table 2.1.2. Country Coverage 111
Annex Table 2.4.1. Drivers of Youth (Ages 15–24) Labor Force Participation Rates: Robustness 117
Annex Table 2.4.2. Drivers of Prime-Age Male (Ages 25–54) Labor Force Participation Rates: Robustness 118
Annex Table 2.4.3. Drivers of Prime-Age Female (Ages 25–54) Labor Force Participation Rates:
Robustness 119
Annex Table 2.4.4. Drivers of Older Workers’ (Ages 55 and over) Labor Force Participation Rates:
Robustness 120
Annex Table 2.4.5. Drivers of Aggregate Labor Force Participation Rates: Robustness 121
Annex Table 2.5.1. Determinants of Being in the Labor Force 123
Annex Table 3.1.1. Data Sources 159
International Monetary Fund | April 2018 v

Annex Table 3.1.2. Sample of Economies Included in the Analytical Exercises 159
Annex Table 3.1.3. Sectors, Individual Industries, and Abbreviations Used in Chapter 160
Annex Table 3.3.1. Estimation Results, Beta-Convergence—Extended Sample (9 Sectors) 164
Annex Table 3.3.2. Estimation Results, Beta-Convergence—Reduced Sample (26 Sectors) 165
Table 4.1. Impact of Foreign Knowledge on Domestic Innovation and Productivity 182
Table 4.2. Impact of Global Value Chain Participation on Average Firm Patenting and Employment 187
Annex Table 4.1.1. List of Variables, Variable Definitions, and Sources 199
Annex Table 4.1.2. List of Sectors in Estimation Samples 200
Annex Table 4.1.3. List of Countries in Estimation Samples 200
Annex Table 4.2.1. Gravity Model of Knowledge Diffusion: Baseline Results for Different Time Periods 202
Annex Table 4.2.2. Gravity Model of Knowledge Diffusion: Including Cross-Sectoral Pairs 203
Annex Table 4.3.1. Impact of Foreign Knowledge on Domestic Innovation: Robustness 205
Annex Table 4.3.2. Impact of Foreign Knowledge on Domestic Labor Productivity: Robustness 205
Annex Table 4.3.3. Impact of Competition on Innovation 207
Annex Table 4.5.1. Impact of Global Value Chain Participation on Firm-Level Innovation: Robustness 211
Annex Table 4.5.2. Relationship between Country-Year Fixed Effects and Selected Policy Variables 211
Online Tables
Table B1. Advanced Economies: Unemployment, Employment, and Real GDP per Capita
Table B2. Emerging Market and Developing Economies: Real GDP
Table B3. Advanced Economies: Hourly Earnings, Productivity, and Unit Labor Costs in Manufacturing
Table B4. Emerging Market and Developing Economies: Consumer Prices
Table B5. Summary of Fiscal and Financial Indicators
Table B6. Advanced Economies: General and Central Government Net Lending/Borrowing and General
Government Net Lending/Borrowing Excluding Social Security Schemes
Table B7. Advanced Economies: General Government Structural Balances
Table B8. Emerging Market and Developing Economies: General Government Net Lending/Borrowing
and Overall Fiscal Balance
Table B9. Emerging Market and Developing Economies: General Government Net Lending/Borrowing
Table B10. Selected Advanced Economies: Exchange Rates
Table B11. Emerging Market and Developing Economies: Broad Money Aggregates
Table B12. Advanced Economies: Export Volumes, Import Volumes, and Terms of Trade in Goods and
Services
Table B13. Emerging Market and Developing Economies by Region: Total Trade in Goods
Table B14. Emerging Market and Developing Economies by Source of Export Earnings: Total Trade in Goods
Table B15. Summary of Current Account Transactions
Table B16. Emerging Market and Developing Economies: Summary of External Debt and Debt Service
Table B17. Emerging Market and Developing Economies by Region: External Debt by Maturity
Table B18. Emerging Market and Developing Economies by Analytical Criteria: External Debt by Maturity
Table B19. Emerging Market and Developing Economies: Ratio of External Debt to GDP
Table B20. Emerging Market and Developing Economies: Debt-Service Ratios
Table B21. Emerging Market and Developing Economies, Medium-Term Baseline Scenario: Selected
Economic Indicators
Figures
Figure 1.1. Global Activity Indicators 2
Figure 1.2. Contributions to the Change in Real GDP Growth, 2016–17 2
Figure 1.3. Global Investment and Trade 3
Figure 1.4. Contributions to Trade Growth 4
Figure 1.5. Commodity and Oil Prices 4
Figure 1.6. Global Inflation 5
vi International Monetary Fund | April 2018

CONTENTS
Figure 1.7. Advanced Economies: Monetary and Financial Market Conditions 7

Figure 1.8. Real Effective Exchange Rate Changes, August 2017–March 2018 7
Figure 1.9. Emerging Market Economies: Equity Markets and Credit 8
Figure 1.10. Emerging Market Economies: Interest Rates 8
Figure 1.11. Emerging Market Economies: Capital Flows 9
Figure 1.12. Terms-of-Trade Windfall Gains and Losses 10
Figure 1.13. GDP Growth, 1999–2023 11
Figure 1.14. Per Capita Real GDP Growth 12
Figure 1.15. Fiscal Indicators 13
Figure 1.16. Global Current Account Balance 18
Figure 1.17. Real Exchange Rates and Current Account Balances in Relation to Economic Fundamentals 19
Figure 1.18. Net International Investment Position 19
Figure 1.19. Growth for Creditors and Debtors 20
Figure 1.20. Geopolitical Risk Index 23
Figure 1.21. Risks to the Global Outlook 23
Figure 1.22. Recession and Deflation Risks 24
Scenario Figure 1. Inflation Surprise and Term Premium Shocks in the United States 32
Figure 1.1.1. Global Sales of Personal Computers and Smartphones 34
Figure 1.1.2. Global iPhone Sales 35
Figure 1.1.3. China: Smartphone Export Cycle 35
Figure 1.2.1. Advanced Economy Core Consumer Price Inflation 37
Figure 1.2.2. Advanced Economy Core Consumer Goods and Core Services Consumer Price Inflation 38
Figure 1.2.3. Cross-Country Distribution of Changes in Core Goods and Core Services Inflation,
2011–17 versus 2002–08 39
Figure 1.2.4. Changes in Sectoral Inflation, 2011–17 versus 2002–08 39
Figure 1.3.1. Different Measures of Growth 41
Figure 1.3.2. Production Function Decomposition: Selected Large Advanced Economies 41
Figure 1.3.3. Investment in Selected Advanced Economies 42
Figure 1.3.4. Total Factor Productivity Growth 42
Figure 1.4.1. Difference between ICT Price Indices and General Non-ICT Price Index, Selected
OECD Countries, 1994 = 1 44
Figure 1.5.1. Impact of a Temporary Increase in the Investment Expense Allowance and a Temporary
Decrease in the Corporate Income Tax Rate 47
Figure 1.6.1. Selected Trade Policy Indicators: Example with Group of Twenty Member Countries 49
Figure 1.6.2. Free Trade Normalization: Alternative Normalization 50
Figure 1.SF.1. Commodity Market Developments 55
Figure 1.SF.2. Detrended IMF Crude Oil Index and Economic and Market Conditions 56
Figure 1.SF.1.1. One Hundred Years of Cobalt Mining 60
Figure 2.1. Demographic Transition: Recent Trends and Projections 72
Figure 2.2. Change in Labor Force Participation Rates, 2008–16 73
Figure 2.3. Labor Force Participation Rates by Gender and Age 75
Figure 2.4. Labor Participation and School Enrollment of the Young 76
Figure 2.5. Labor Force Participation Rates of Prime-Age Men and Women by Demographic
Characteristics, 2000 and 2016 77
Figure 2.6. Subgroups of the Inactive, 2000 and 2016 78
Figure 2.7. The Role of Exposure to Routinization 79
Figure 2.8. Average Annual Changes in Labor Force Participation Rates 80
Figure 2.9. Decomposition of Labor Market Shifts 81
Figure 2.10. Changes in Participation Rates, 2008–16 83
Figure 2.11. Drivers of Participation Rates: Policies 86
Figure 2.12. Drivers of Participation Rates: Additional Policies 87
International Monetary Fund | April 2018 vii

Figure 2.13. Changes in Participation Rates, Actual versus Predicted, 1995–2011 89

Figure 2.14. Average Contributions to Changes in Participation Rates, 1995–2011 89
Figure 2.15. Change in the Odds of Being Active 91
Figure 2.16. Policies and the Effect of Routine Exposure on Labor Force Participation 92
Figure 2.17. Age and Cohort Effects of Labor Force Participation 93
Figure 2.18. Projected Changes in Participation Rates under Alternative Scenarios 94
Figure 2.1.1. Labor Force Participation by Age Group 97
Figure 2.1.2. Median Secondary Enrollment by Geographic Region 98
Figure 2.1.3. Youth Labor Force Participation by Gender 98
Figure 2.1.4. Implied 10-Year Improvement in Country Gender Gaps for Youth 99
Figure 2.2.1. Labor Force Participation and Change in Labor Force Participation by State 100
Figure 2.2.2. Change in Labor Force Participation Rate by Metropolitan Area 101
Figure 2.2.3. Decomposition of Labor Market Changes in Metropolitan Areas 101
Figure 2.2.4. Routine and Offshoring Exposure by State 102
Figure 2.3.1. Change in Labor Force Participation by Region, 2000–16 103
Figure 2.3.2. Change in Labor Force Participation Rate by Region 104
Figure 2.3.3. Decomposition of Labor Market Changes 104
Figure 2.3.4. Initial Routine Exposure by Region, 2000 105
Figure 2.3.5. Initial Offshoring Exposure by Region, 2000 105
Figure 2.4.1. Contributions of Natural Population Growth and Net Migration to Total Population Growth 106
Figure 2.4.2. Projected Evolution of Labor Force Participation Rates 107
Figure 2.4.3. Labor Force Participation Rates of Prime-Age Natives and Migrants, 2000–16 108
Figure 2.4.4. Change in the Odds of Being Active 108
Annex Figure 2.2.1. Changes in Labor Force Participation Rates, 1985–2016 111
Annex Figure 2.2.2. Share of Households by Employment Composition, 2000–16 112
Annex Figure 2.2.3. Changes in Labor Force Participation Rates of Prime-Age Men and Women,
1985–2016 112
Annex Figure 2.2.4. Average Annual Changes in Labor Force Participation Rates 112
Annex Figure 2.2.5. Flows into Employment, Unemployment, and Inactivity 113
Figure 3.1. Share of Manufacturing in Aggregate Employment 130
Figure 3.2. Sectoral Employment Shares 130
Figure 3.3. Change in Manufacturing Employment Share and Inequality, 1980–2010 131
Figure 3.4. Share of Manufacturing in Aggregate Employment and Output 132
Figure 3.5. Estimated Trends in Manufacturing Employment and Output Shares, 1960–2015 133
Figure 3.6. Cross-Country Distribution of Estimated Trends in Manufacturing Shares, 1970–2015 134
Figure 3.7. Peak of Manufacturing Employment Share 134
Figure 3.8. Share of Manufacturing in Final Consumption versus Income per Capita, 1980–2011 135
Figure 3.9. Relative Price of Manufacturing, 1970–2015 136
Figure 3.10. Estimated Change in Manufacturing Shares and Relative Prices, 1960–2015 136
Figure 3.11. Manufacturing Gross Output and Final Expenditure on Manufacturing Goods, 1995–2011 138
Figure 3.12. Change in Services Employment Share, 1970–2015 139
Figure 3.13. Difference in Labor Productivity Growth between Manufacturing and Services
before and after 2000 141
Figure 3.14. Distribution of Labor Productivity Growth of Individual Industries 142
Figure 3.15. Sectoral Labor Productivity, 2005 142
Figure 3.16. Structural Transformation and Aggregate Labor Productivity Growth, 2000–10 144
Figure 3.17. Estimation Results, Beta-Convergence 145
Figure 3.18. Sigma-Convergence 146
Figure 3.19. Productivity Gap in 2005 146
Figure 3.20. Average Gross Wages in Industry and Services in the 2000s 148
Figure 3.21. Labor Income Inequality in the 2000s 148
viii International Monetary Fund | April 2018

CONTENTS
Figure 3.22. Contribution to Change in Overall Labor Income Inequality between the 1980s and 2000s 149
Figure 3.1.1. Nonmanufacturing Value-Added Content in Gross Manufacturing Output, 1995–2011 152
Figure 3.1.2. Services Value-Added Content in Gross Manufacturing Output, 1995 and 2011 153
Figure 3.1.3. Change in Services Value-Added Content in Manufacturing Gross Output, 1995–2011 153
Figure 3.1.4. Change in Services Nominal and Real Value-Added Content in Manufacturing
Gross Output, 1995–2009 154
Figure 3.2.1. Exports of Services 155
Figure 3.2.2. Increase in Service Trade, 1980–2014 155
Figure 3.2.3. Services Exports by Industry, 1990–2014 156
Figure 3.3.1. Sectoral Employment Shares and Wage Inequality 157
Figure 3.3.2. Wage Gap between Manufacturing and Services 157
Figure 3.3.3. Wages of Workers Switching to Manufacturing Jobs 158
Figure 3.3.4. Inequality in Manufacturing and Services 158
Annex Figure 3.3.1. Distribution of Total Factor Productivity Growth of Individual Industries 161
Annex Figure 3.3.2. Sectoral Labor Productivity Growth, 2000–10 162
Annex Figure 3.3.3. Skill Composition of Workers by Sector, 2000–07 162
Annex Figure 3.3.4. Sectoral Labor Productivity, 2010 163
Annex Figure 3.3.5. Sigma-Convergence 166
Figure 4.1. International Patent Families by Publication Year 174
Figure 4.2. Technology Diffusion 175
Figure 4.3. Patenting and Research and Development at the Frontier 177
Figure 4.4. Countries at the Technology Frontier 178
Figure 4.5. Slowing Patenting and Productivity 178
Figure 4.6. The Evolution of Cross-Patent Citations within and across Regions 179
Figure 4.7. Knowledge Diffusion across Barriers over Time 180
Figure 4.8. Contribution of Foreign Knowledge to Labor Productivity Growth 183
Figure 4.9. The Dynamics of Technology Diffusion 184
Figure 4.10. Patenting and Global Value Chain Participation 186
Figure 4.11. The Effects of Global Value Chain Participation and Policy Variables 188
Figure 4.12. International Competition and Global Concentration 189
Figure 4.13. The Effect of Competition on Innovation and Technology Diffusion 189
Figure 4.2.1. Innovation Intensity 193
Figure 4.2.2. Foreign Patents by Source Country, 2013 194
Figure 4.3.1. Sub-Saharan Africa: Net Foreign Direct Investment and Aid Inflows 195
Figure 4.3.2. Official Development Assistance Commitment by Sector 195
Figure 4.3.3. Aid Commitment to Energy Generation 196
Annex Figure 4.2.1. Diffusion of Knowledge from G5 with Expanded Emerging Market Economy
Sample 202
Annex Figure 4.2.2. Reduction of Knowledge Flow with Additional Barriers: Including
Cross-Sectoral Citations 203
Annex Figure 4.2.3. Reduction of Knowledge Flow with Additional Barriers: Unrestricted Cited Sample 204
Annex Figure 4.2.4. Reduction of Knowledge Flow with Additional Barriers: Excluding China
from Baseline 204
International Monetary Fund | April 2018 ix

WORLD ECONOMIC OUTLOOK: TENSIONS FROM THE TWO-SPEED RECOVERY
ASSUMPTIONS AND CONVENTIONS
A number of assumptions have been adopted for the projections presented in the World Economic Outlook
(WEO). It has been assumed that real effective exchange rates remained constant at their average levels during
January 26 to February 23, 2018, except for those for the currencies participating in the European exchange rate
mechanism II, which are assumed to have remained constant in nominal terms relative to the euro; that established
policies of national authorities will be maintained (for specific assumptions about fiscal and monetary policies for
selected economies, see Box A1 in the Statistical Appendix); that the average price of oil will be $62.31 a barrel
in 2018 and $58.24 a barrel in 2019 and will remain unchanged in real terms over the medium term; that the
six-month London interbank offered rate on US dollar deposits will average 2.4 percent in 2018 and 3.4 percent
in 2019; that the three-month euro deposit rate will average –0.3 percent in 2018 and 0.0 in 2019; and that the
six-month Japanese yen deposit rate will yield on average 0.0 percent in 2018 and 0.1 percent in 2019. These are,
of course, working hypotheses rather than forecasts, and the uncertainties surrounding them add to the margin of
error that would in any event be involved in the projections. The estimates and projections are based on statistical
information available through April 2, 2018.
The following conventions are used throughout the WEO:
. . . to indicate that data are not available or not applicable;
– between years or months (for example, 2017–18 or January–June) to indicate the years or months cov-
ered, including the beginning and ending years or months; and
/ between years or months (for example, 2017/18) to indicate a fiscal or financial year.
“Billion” means a thousand million; “trillion” means a thousand billion.
“Basis points” refers to hundredths of 1 percentage point (for example, 25 basis points are equivalent to ¼ of
1 percentage point).
Data refer to calendar years, except in the case of a few countries that use fiscal years. Table F in the Statistical
Appendix lists the economies with exceptional reporting periods for national accounts and government finance
data for each country.
For some countries, the figures for 2017 and earlier are based on estimates rather than actual outturns.
Table G in the Statistical Appendix lists the latest actual outturns for the indicators in the national accounts,
prices, government finance, and balance of payments indicators for each country.
What is new in this publication:
• No changes have been introduced for the April 2018 WEO database.
In the tables and figures, the following conventions apply:
• If no source is listed on tables and figures, data are drawn from the WEO database.
• When countries are not listed alphabetically, they are ordered on the basis of economic size.
• Minor discrepancies between sums of constituent figures and totals shown reflect rounding.
As used in this report, the terms “country” and “economy” do not in all cases refer to a territorial entity that is
a state as understood by international law and practice. As used here, the term also covers some territorial entities
that are not states but for which statistical data are maintained on a separate and independent basis.
Composite data are provided for various groups of countries organized according to economic characteristics or
region. Unless noted otherwise, country group composites represent calculations based on 90 percent or more of
the weighted group data.
The boundaries, colors, denominations, and any other information shown on the maps do not imply, on the
part of the International Monetary Fund, any judgment on the legal status of any territory or any endorsement or
acceptance of such boundaries.
x International Monetary Fund | April 2018

FURTHER INFORMATION
Corrections and Revisions

The data and analysis appearing in the World Economic Outlook (WEO) are compiled by the IMF staff at the
time of publication. Every effort is made to ensure their timeliness, accuracy, and completeness. When errors are
discovered, corrections and revisions are incorporated into the digital editions available from the IMF website and
on the IMF eLibrary (see below). All substantive changes are listed in the online tables of contents.
Print and Digital Editions

Print copies of this World Economic Outlook can be ordered at https://www.bookstore.imf.org/books/title/
world-economic-outlook-april-2018.
The WEO is featured on the IMF website at http://www.imf.org/publications/WEO. This site includes a PDF
of the report and data sets for each of the charts therein.
The IMF eLibrary hosts multiple digital editions of the World Economic Outlook, including ePub, enhanced
PDF, Mobi, and HTML: http://elibrary.imf.org/AR18WEO.
Copyright and Reuse

Information on the terms and conditions for reusing the contents of this publication are at http://www.
imf.org/external/terms.htm.
International Monetary Fund | April 2018 xi

DATA
This version of the World Economic Outlook (WEO) is available in full through the IMF eLibrary (www.elibrary.
imf.org) and the IMF website (www.imf.org). Accompanying the publication on the IMF website is a larger compila-
tion of data from the WEO database than is included in the report itself, including files containing the series most
frequently requested by readers. These files may be downloaded for use in a variety of software packages.
The data appearing in the WEO are compiled by the IMF staff at the time of the WEO exercises. The histori-
cal data and projections are based on the information gathered by the IMF country desk officers in the context
of their missions to IMF member countries and through their ongoing analysis of the evolving situation in each
country. Historical data are updated on a continual basis as more information becomes available, and structural
breaks in data are often adjusted to produce smooth series with the use of splicing and other techniques. IMF
staff estimates continue to serve as proxies for historical series when complete information is unavailable. As a
result, WEO data can differ from those in other sources with official data, including the IMF’s International
Financial Statistics.
The WEO data and metadata provided are “as is” and “as available,” and every effort is made to ensure their
timeliness, accuracy, and completeness, but these cannot be guaranteed. When errors are discovered, there is a
concerted effort to correct them as appropriate and feasible. Corrections and revisions made after publication are
incorporated into the electronic editions available from the IMF eLibrary (www.elibrary.imf.org) and on the IMF
website (www.imf.org). All substantive changes are listed in detail in the online tables of contents.
For details on the terms and conditions for usage of the WEO database, please refer to the IMF Copyright and
Usage website (www.imf.org/external/terms.htm).
Inquiries about the content of the WEO and the WEO database should be sent by mail, fax, or online forum
(telephone inquiries cannot be accepted):
World Economic Studies Division
Research Department
International Monetary Fund
700 19th Street, NW
Washington, DC 20431, USA
Fax: (202) 623-6343
Online Forum: www.imf.org/weoforum
xii International Monetary Fund | April 2018

PREFACE
The analysis and projections contained in the World Economic Outlook are integral elements of the IMF’s
surveillance of economic developments and policies in its member countries, of developments in international
financial markets, and of the global economic system. The survey of prospects and policies is the product of a
comprehensive interdepartmental review of world economic developments, which draws primarily on information
the IMF staff gathers through its consultations with member countries. These consultations are carried out
in particular by the IMF’s area departments—namely, the African Department, Asia and Pacific Department,
European Department, Middle East and Central Asia Department, and Western Hemisphere Department—
together with the Strategy, Policy, and Review Department; the Monetary and Capital Markets Department; and
the Fiscal Affairs Department.
The analysis in this report was coordinated in the Research Department under the general direction of Maurice
Obstfeld, Economic Counsellor and Director of Research. The project was directed by Gian Maria Milesi-Ferretti,
Deputy Director, Research Department; Oya Celasun, Division Chief, Research Department; and Helge Berger,
Assistant Director, Research Department and Head of the IMF’s Spillover Task Force.
The primary contributors to this report were Aqib Aslam, Christian Bogmans, Wenjie Chen, Federica Coelli,
Johannes Eugster, Francesco Grigoli, Bertrand Gruss, Giang Ho, Florence Jaumotte, Zsóka Kóczán, Toh Kuan,
Nan Li, Weicheng Lian, Akito Matsumoto, Malhar Nabar, Natalija Novta, Carolina Osorio Buitron, Roberto
Piazza, Yu Shi, Petia Topalova, and Rachel Yuting Fan.
Other contributors include Jorge Alvarez, Gavin Asdorian, Felicia Belostecinic, Olivier Bizimana, Patrick Blagrave,
John Bluedorn, Luisa Calixto, Benjamin Carton, Diego Cerdeiro, Sophia Chen, Pankhuri Dutt, Angela Espiritu,
Sung Eun Jung, Emilio Fernandez Corugedo, Chanpheng Fizzarotti, Gregg Forte, Khondoker Haider, Meron Haile,
Mandy Hemmati, Benjamin Hilgenstock, Ava Hong, Keiko Honjo, Benjamin Hunt, Hao Jiang, Christopher Johns,
Rahel Kidane, Lama Kiyasseh, Jungjin Lee, Yiqun Li, Davide Malacrino, Joannes Mongardini, Mico Mrkaic, Daniela
Muhaj, Susanna Mursula, Rachel J. Nam, Cynthia Nyanchama Nyakeri, Emory Oakes, Ilse Peirtsegaele, Evgenia
Pugacheva, Marshall Reinsdorf, Daniel Rivera Greenwood, Kadir Tanyeri, Nicholas Tong, Menexenia Tsaroucha,
Ke Wang, Shan Wang, Jilun Xing, Yuan Zeng, Fan Zhang, Qiaoqiao Zhang, and Huiyuan Zhao.
Joseph Procopio from the Communications Department led the editorial and production team for the report
with support from Linda Kean, Christine Ebrahimzadeh, James Unwin, Lucy Scott Morales, Angela White, and
Vector Talent Resources.
The analysis has benefited from comments and suggestions by staff members from other IMF departments, as
well as by Executive Directors following their discussion of the report on April 2, 2018. However, both projections
and policy considerations are those of the IMF staff and should not be attributed to Executive Directors or to their
national authorities.
International Monetary Fund | April 2018 xiii

FOREWORD
T
he global economic upswing that began spread its benefits lower in the income distribution, and
around mid-2016 has become broader and build resilience to the hazards that lie ahead.
stronger. This new World Economic Outlook Future growth prospects look challenging indeed
report projects that advanced economies as for advanced economies and many commodity export-
a group will continue to expand above their potential ers. In advanced economies, aging populations and
growth rates this year and next before decelerating, lower projected advances in total factor productivity
while growth in emerging market and developing will make it hard to return to the precrisis pace for
economies will rise before leveling off. For most the average household’s income growth. Substantially
countries, current favorable growth rates will not last. raising middle and lower incomes looks even tougher.
Policymakers should seize this opportunity to bolster Moreover, growth rates will inevitably bend toward
growth, make it more durable, and equip their gov- their weaker longer-term levels. Policy support will
ernments better to counter the next downturn. fade in the United States and China—a necessity in
Global growth seems on track to reach 3.9 percent view of those countries’ macroeconomic imbalances.
this year and next, substantially above our October And countries that currently can grow more quickly
forecast. Helping to drive this output acceleration is by putting underutilized labor and capital back to
faster growth in the euro area, Japan, China, and the work will reach full capacity. The need for a forward-
United States, all of which grew above expectations last looking policy perspective is therefore urgent—to
year, along with some recovery in commodity export- limit risks as well as enhance growth.
ers. Along with China, several other emerging market As usual, Chapter 1 of this report sets out the
and developing economies will also do better this year risks to the forecast. These are balanced over the next
than in our past projections—that group includes several quarters, with the possibility of more buoy-
Brazil, Mexico, and emerging Europe. The aggregate ant growth than forecast balancing out unfavorable
gains for this country group are, however, weighed contingencies. But as time passes, the likelihood of
down by sharp downward revisions for a few countries negative shifts in the forecast rises.
in the grip of civil strife, notably Libya, Venezuela, and Monetary policy might tighten sooner than
Yemen. Growing trade and investment continue as expected if excess demand emerges, a notable possibil-
notable factors powering the global upswing. ity in the United States, where fiscal policy has turned
Growth this broad based and strong has not been much more expansive even as the economy has neared
seen since the world’s initial sharp 2010 bounce back full employment. Financial tightening, in turn, would
from the financial crisis of 2008–09. The synchronized stress highly indebted countries, firms, and house-
expansion will help to dispel some remaining legacies holds, including in emerging market economies.
of the crisis by speeding the exit from unconventional An escalating cycle of trade restrictions and retalia-
monetary policies in advanced economies, encouraging tion is another risk. The first shots in a potential trade
investment, and healing labor market scars. war have now been fired. Conflict could intensify
Other aftereffects of the crisis seem more durable, if fiscal policies in the United States drive its trade
however, including higher debt levels worldwide deficit higher without action in Europe and Asia to
and widespread public skepticism about policymak- reduce surpluses. The multilateral rules-based trade
ers’ capacity and willingness to generate robust and system that evolved after World War II and that nur-
inclusive growth. That skepticism will only be rein- tured unprecedented growth in the world economy
forced—with negative political consequences down the needs strengthening. Instead, it is in danger of being
road—if economic policy does not rise to the challenge torn apart.
of enacting reforms and building fiscal buffers. Success The renewed popularity of nationalistic poli-
in such efforts would strengthen medium-term growth, cies is another aftereffect of the financial crisis and
xiv International Monetary Fund | April 2018

FOREWORD
its prolonged aftermath. Diminished prospects for to overrule market forces and subsidize manufactur-
household income growth in advanced economies, ing—possibly a zero-sum game globally—but instead
coupled with trends of higher polarization in jobs to aim to raise productivity across the economy. The
and incomes, have fueled a widespread political latter effort requires structural reforms, including
backlash hostile to traditional political modalities. If lower barriers to services trade, along with many of
policymakers are complacent and do not tackle the the same investments in people that will enhance
challenge of strengthening long-term growth, political labor force attachment, as described in Chapter 2.
risks could intensify, possibly reversing some of the Finally, Chapter 4 studies the process through
progress that economic reforms and integration have which innovative activity and technological know-
achieved to date. how spread across national borders. Cross-border
The three analytical chapters in this World Economic knowledge flows from technological leaders to poorer
Outlook are unified by their focus on central determi- countries have historically been significant driv-
nants of long-term economic growth. ers of income convergence. Now, the emergence of
Population growth, age distribution, and other China and Korea as leaders in some sectors offers the
structural employment trends are critical for under- promise of positive repercussions for others, including
standing growth, investment, and productivity. Chap- the long-established high-income countries. Interna-
ter 2 focuses on labor force participation in advanced tional trade and competition, this chapter suggests,
economies, where population aging and, for many promote global knowledge diffusion and thus provide
countries, declining overall participation rates are an important channel through which all countries can
substantial headwinds to growth. Especially worrisome benefit from globalization. From this perspective, poli-
is the widespread decline in participation of young cies that restrict trade to prop up politically favored
and prime-age men. The chapter shows how a range sectors of the economy will ultimately harm produc-
of policies—for example, educational investments and tivity growth.
tax policies—can mitigate these effects. But participa- Global growth is on an upswing, but favorable con-
tion will continue to decline even under best-practice ditions will not last forever, and now is the moment
approaches. to get ready for leaner times. Readiness requires not
Chapter 3 focuses on the declining share of manu- only cautious and forward-looking management of
facturing employment globally and, most dramatically, monetary and fiscal policies, but also careful atten-
in advanced economies. This structural transformation to financial stability. Also necessary are structural
tion, driven by technology advances as well as global- and tax policies that raise potential output, including
ization, has sparked popular concern about greater by investing in people and ensuring that the fruits of
earnings inequality as “good jobs” disappear. Another growth are widely shared. While there is much each
worry is that currently poor countries may be trapped country can do on its own, multilateral cooperation
far from the global income frontier if they never pass on a range of issues—stretching from trade to reduc-
through a developmental stage of substantial manu- ing global imbalances to cybersecurity to climate—
facturing employment. The chapter, however, suggests remains essential.
that services can offer considerable scope for produc- Maurice Obstfeld
tivity gain. Therefore, the best policy response is not Economic Counsellor
International Monetary Fund | April 2018 xv

EXECUTIVE SUMMARY
The upswing in global investment and trade continued above 4 percent, the strongest since the second half
in the second half of 2017. At 3.8 percent, global growth of 2010, supported by a recovery in investment.
in 2017 was the fastest since 2011. With financial Outcomes exceeded the October 2017 World Eco-
conditions still supportive, global growth is expected to nomic Outlook forecasts in the euro area, Japan, the
tick up to a 3.9 percent rate in both 2018 and 2019. United States, and China, and continued to improve
Advanced economies will grow faster than potential this gradually in commodity exporters. Financial condi-
year and next; euro area economies are set to narrow tions remain supportive, despite the recent volatility
excess capacity with support from accommodative mon- in equity markets and increases in bond yields fol-
etary policy, and expansionary fiscal policy will drive the lowing signs of firming inflation in advanced econo-
US economy above full employment. Aggregate growth in mies. With broad-based momentum and expectations
emerging market and developing economies is projected to of a sizable fiscal expansion in the United States over
firm further, with continued strong growth in emerging this year and the next, global growth is now pro-
Asia and Europe and a modest upswing in commodity jected at 3.9 percent for 2018–19, a 0.2 percentage
exporters after three years of weak performance. point upgrade for both years relative to the October
Global growth is projected to soften beyond the next 2017 forecast.
couple of years. Once their output gaps close, most This positive momentum will eventually slow,
advanced economies are poised to return to potential however, leaving many countries with a challenging
growth rates well below precrisis averages, held back by medium-term outlook. Some cyclical forces will wane:
aging populations and lackluster productivity. US growth financial conditions are expected to tighten naturally
will slow below potential as the expansionary impact of with the closing of output gaps and monetary policy
recent fiscal policy changes goes into reverse. Growth is normalization; US tax reform will subtract momen-
projected to remain subpar in several emerging market tum starting in 2020, and then more strongly as full
and developing economies, including in some commodity investment expensing is phased out starting in 2023;
exporters that continue to face substantial fiscal consoli- and China’s transition to lower growth is expected to
dation needs. resume as credit growth and fiscal stimulus dimin-
While upside and downside risks to the short-term ish. At the same time, while the expected recovery in
outlook are broadly balanced, risks beyond the next investment will help raise potential output, weak pro-
several quarters clearly lean to the downside. Downside ductivity trends and reduced labor force growth due
concerns include a possibly sharp tightening of financial to population aging constrain medium-term prospects
conditions, waning popular support for global economic in advanced economies. (Chapter 2 examines the driv-
integration, growing trade tensions and risks of a shift ers of labor force participation in advanced econo-
toward protectionist policies, and geopolitical strains. mies.) The outlook is mixed across emerging market
The current recovery offers a window of opportunity and developing economies. Prospects remain favorable
to advance policies and reforms that secure the current in emerging Asia and Europe, but are challenging
upswing and raise medium-term growth to the benefit in Latin America, the Middle East and sub-Saharan
of all. Such policies should focus on strengthening the Africa, where—despite some recovery—the medium-
potential for higher and more inclusive growth, building term outlook for commodity exporters remains
buffers to deal more effectively with the next downturn, generally subdued, with a need for further economic
improving financial resilience to contain market risks and diversification and adjustment to lower commodity
stability concerns, and fostering international cooperation. prices. More than one-quarter of emerging market
Economic activity in 2017 ended on a high and developing economies are projected to grow by
note—growth in the second half of the year was less than advanced economies in per capita terms over
xvi International Monetary Fund | April 2018

Executive Summary
the next five years, and hence fall further behind in The current juncture offers a window of opportu-
terms of living standards. nity to advance policies and reforms that safeguard the
Risks around the short-term outlook are broadly upswing and raise medium-term growth to the benefit
balanced, but risks beyond the next several quarters of all.
are clearly to the downside. On the upside, the growth • Strengthen the potential for higher and more inclu-
spurt in advanced economies may turn out to be stron- sive growth. All countries have room for structural
ger and more durable than in the baseline, as slack reforms and fiscal policies that raise productiv-
in labor markets can be larger than currently assessed ity and enhance inclusiveness—for instance, by
(Chapter 2 of the October 2017 WEO). Furthermore, encouraging experimentation and diffusion of new
the ongoing recovery in investment could foster a technologies, increasing labor force participation,
rebound in productivity, implying higher potential supporting those displaced by structural change,
growth going forward. On the downside, financial and investing in the young to enhance their job
conditions—which remain easy despite the onset of opportunities. The analysis, in Chapter 3, of one
monetary policy normalization—could tighten sharply aspect of structural change—the decline in the
and expose vulnerabilities that have accumulated share of manufacturing jobs in overall employment
over the years, with adverse repercussions for growth. and its implications for productivity growth and
Indeed, as discussed in the April 2018 Global Finan- inequality—highlights the importance of facilitating
cial Stability Report, Growth-at-Risk analysis suggests the reallocation of labor to the most dynamic sec-
that risks to medium-term growth, stemming from tors through workforce skills development, lowering
easy financial conditions, remain well above historical job search costs, and reducing barriers to entry and
norms. In the United States, financial conditions could trade in services.
tighten faster than expected, triggered, for example, • Complete the recovery and build buffers. Monetary
by an adjustment in market pricing of the future path accommodation needs to continue where inflation
of monetary policy, higher realized or expected wage is weak, but a well-communicated, data-dependent
and price inflation, and/or a sudden decompression normalization should follow in countries where
of term premiums. Tighter financial conditions in the inflation looks set to return to the central bank’s
United States would have spillovers to other econo- target. Fiscal policies should start rebuilding buffers
mies, including through a reduction in capital flows to where needed, incorporate supply-side measures to
emerging markets. Very expansionary fiscal policy in bolster potential output, and promote inclusive-
the United States, at a time when the current account ness. In countries at or close to full employment,
deficit is already larger than justified by fundamentals, with an excess current account deficit and an
combined with persistent excess current account sur- unsustainable fiscal position (notably the United
pluses in other countries, is projected to widen global States), there is a need to stabilize and eventually
imbalances. Anxiety about technological change and reduce the debt and reverse the procyclical stimulus
globalization is on the rise and, when combined with that is already in place. This will require ensuring
wider trade imbalances, could foster a shift toward higher future revenues and gradually containing
inward-looking policies, disrupting trade and invest- the growth of public spending, while changing
ment. Recent import restrictions announced by the its composition toward improving infrastructure,
United States, announced retaliatory actions by China, boosting labor force participation, and reduc-
and potential retaliation by other countries raise ing poverty. Countries with both excess current
concerns in this regard and threaten to damage global account surpluses and fiscal space (for example,
and domestic activity and sentiment. Similarly, changes Germany) should increase public investment that
in US tax policies are expected to exacerbate income boosts potential growth and demand.
polarization, which could affect the political climate • Improve financial resilience. Macro- and micropru-
for policy choices in the future. Climate change, dential policies can curb rising leverage and contain
geopolitical tensions, and cybersecurity breaches pose financial market risks. In some advanced economies,
additional threats to the subdued medium-term global balance sheet repair needs to continue. Emerging
outlook. market economies should keep monitoring exposures
International Monetary Fund | April 2018 xvii

to foreign currency debt. Building on recent efforts, diversification and improve the capacity to cope
China should continue to rein in credit growth and with climate shocks where needed.
address financial risks. • Foster cooperation. Maintaining financial and regula-
• Improve convergence prospects for low-income develop- tory reform momentum and preserving an open, mul-
ing countries. Continued progress toward the 2030 tilateral trade system should take priority. As Chapter
United Nations Sustainable Development Goals 4 documents, global integration has helped increase
will require low-income developing countries to cross-border knowledge flows, the diffusion of innova-
implement policies that strengthen their fiscal tion, and productivity growth across countries—a key
positions, boost financial resilience, reduce poverty, driver of improvements in living standards and welfare
and make growth more inclusive. Investment in over time. It is also crucial that countries collaborate
workforce skills, improving access to credit, and to address shared problems, such as excess external
reducing infrastructure gaps can promote economic imbalances, cybersecurity, and climate change.
xviii International Monetary Fund | April 2018

1
CHAPTER
GLOBAL PROSPECTS AND POLICIES
World growth strengthened in 2017 to 3.8 percent, The current favorable juncture offers a window to
with a notable rebound in global trade. It was driven enact policies and reforms that protect the upswing
by an investment recovery in advanced economies, and raise medium-term growth to the benefit of
continued strong growth in emerging Asia, a nota- all—strengthening the potential for higher and more
ble upswing in emerging Europe, and signs of recov- inclusive growth, building buffers that will help deal
ery in several commodity exporters. Global growth is more effectively with the next downturn, improv-
expected to tick up to 3.9 percent this year and next, ing financial resilience to contain financial market
supported by strong momentum, favorable market risks, and fostering international cooperation.
sentiment, accommodative financial conditions, and
the domestic and international repercussions of expan-
sionary fiscal policy in the United States. The partial Recent Developments and Prospects
recovery in commodity prices should allow conditions
An Investment-Led Pickup in Growth
in commodity exporters to gradually improve.
Over the medium term, global growth is projected to At 3.8 percent, global growth last year was
decline to about 3.7 percent. Once the cyclical upswing ½ percentage point faster than in 2016 and the stron-
and US fiscal stimulus have run their course, pros- gest since 2011. Two-thirds of countries accounting for
pects for advanced economies remain subdued, given about three-fourths of global output experienced faster
their slow potential growth. In emerging market and growth in 2017 than in the previous year (the highest
developing economies, in contrast, growth will remain share of countries experiencing a year-over-year growth
close to its 2018–19 level as the gradual recovery in pickup since 2010). The preliminary outcome for
commodity exporters and a projected increase in India’s global growth in 2017 was 0.2 percentage point stron-
growth provide some offset to China’s gradual slowdown ger than forecast in the October 2017 World Economic
and emerging Europe’s return to its lower-trend growth Outlook (WEO), with upside surprises in the second
rate. Nevertheless, 40 emerging market and developing half of 2017 in advanced as well as emerging market
economies are projected to grow more slowly in per capita and developing economies.
terms than advanced economies, failing to narrow income Resurgent investment spending in advanced econ-
gaps vis-à-vis the group of more prosperous countries. omies and an end to the investment decline in some
Despite strong aggregate figures in the baseline forecast commodity‑exporting emerging market and develop-
and buoyant market sentiment, the current momentum ing economies were important drivers of the uptick
is not assured. Upside and downside risks are broadly in global GDP growth and manufacturing activity
balanced over the next several quarters, but risks far- (Figures 1.1–1.3).
ther down the road are skewed to the downside. With •• Across advanced economies, the 0.6 percentage
still-easy financial conditions and persistently low point pickup in 2017 growth relative to 2016 is
inflation that has required protracted monetary policy explained almost entirely by investment spending,
accommodation, a potential further buildup of finan- which remained weak since the 2008–09 global
cial vulnerabilities could give way to rapid tightening financial crisis and was particularly subdued in 2016
of global financial conditions, denting confidence and (Figure 1.2, left column). Both stronger gross fixed
growth. The support to growth that comes from procyclical capital formation and an acceleration in stock build-
policies, including in the United States, will eventually ing contributed to the pickup in investment, with
need to be reversed. Other risks include a shift toward accommodative monetary policy, stronger balance
inward-looking policies that harm international trade sheets, and an improved outlook helping release
and a worsening of geopolitical tensions and strife. pent-up demand for capital goods.
International Monetary Fund | April 2018 1

WORLD ECONOMIC OUTLOOK: Cyclical Upswing, Structural Change
Figure 1.1. Global Activity Indicators Figure 1.2. Contributions to the Change in Real GDP Growth,
2016–17
Global growth surprised on the upside in the second half of 2017 amid (Percentage points)
strengthening industrial production and trade.
Stronger investment spending in advanced economies and an end to fixed
14 1. World Trade and Industrial Production investment contractions in commodity exporters were important contributors
12 (Three-month moving average; annualized percent change) to the pickup in global growth.
10 Industrial production
8 World trade volumes Fixed investment Net exports Inventories
6 Public consumption Private consumption Real GDP
4
2 2.0
0
–2
1.5 Fixed investment: commodity
–4
exporters and Brazil, Colombia, Peru
–6
2011 12 13 14 15 16 17 Feb. 1.0
18
14 2. Manufacturing PMI 3. Consumer Confidence 130
0.5
12 (Three-month moving (Index, 2010 = 100)
125
average; deviations Advanced economies1 Fixed investment: other EMDEs
10 from 50) World 120
Emerging market 0.0
8 Advanced economies1 economies2 115
6 Emerging market World
economies2 110 –0.5 Fixed investment: China and India
4
105
2
0 100 –1.0
–2 95
–4 90 –1.5
2012 13 14 15 16 Feb. 2012 13 14 15 16 Mar. Advanced economies Emerging market and
18 18 developing economies
GDP Growth
(Annualized semiannual percent change) Source: IMF staff calculations.
Note: EMDEs = emerging market and developing economies.
October 2017 WEO April 2018 WEO
4 4. Advanced Economies 5. Emerging Market and 9

Developing Economies •• Across emerging market and developing econo-
8
3 mies, the 0.4 percentage point pickup in 2017
7 growth came primarily from an acceleration in
2 6 private consumption (Figure 1.2, right column).
But the picture is mixed within the group. Growth
5
1
in China and India last year was supported by
4 resurgent net exports and strong private consump-
tion, respectively, while investment growth slowed.
0 3
2011: 13: 15: 17: 19: 2011: 13: 15: 17: 19: An end to fixed investment contractions in
H1 H1 H1 H1 H2 H1 H1 H1 H1 H2
commodity-exporting countries that were severely
affected by the commodity price downturn during
Sources: CPB Netherlands Bureau for Economic Policy Analysis; Haver Analytics;
Markit Economics; and IMF staff estimates. 2015–16 (notably Brazil and Russia, but also
Note: CC = consumer confidence; PMI = purchasing managers’ index; Angola, Ecuador, and Nigeria) instead played an
WEO = World Economic Outlook.
1
Australia, Canada (PMI only), Czech Republic, Denmark, euro area, Hong Kong important role in their growth pickup in 2017.
SAR (CC only), Israel, Japan, Korea, New Zealand (PMI only), Norway (CC only), Higher fixed investment growth (2.3 percentage
Singapore (PMI only), Sweden (CC only), Switzerland, Taiwan Province of China,
United Kingdom, United States.
points above its 2016 level) also supported the
2
Argentina (CC only), Brazil, China, Colombia (CC only), Hungary, India (PMI only), growth performance of other emerging market and
Indonesia, Latvia (CC only), Malaysia (PMI only), Mexico (PMI only), Philippines (CC developing economies, alongside stronger private
only), Poland, Russia, South Africa, Thailand (CC only), Turkey, Ukraine (CC only).
consumption.
2 International Monetary Fund | April 2018

CHAPTER 1 Global Prospects and Policies
A Cyclical Rebound in Global Trade Figure 1.3. Global Investment and Trade
(Percent change)
Global trade—which tends to be highly correlated
with global investment (see Figure 1.3 and Chapter Global trade recovered strongly in 2017 after two years of weakness as
2 of the October 2016 WEO)—recovered strongly investment spending picked up.
in 2017 after two years of weakness, to an estimated
real growth rate of 4.9 percent. The upsurge was Real investment Real GDP at market prices Real imports
more pronounced in emerging market and developing
15 1. Advanced Economies
economies (with trade growth rising from 2.2 percent
10
in 2016 to 6.4 percent in 2017), reflecting improved
5
investment growth rates in formerly stressed commod-
0
ity exporters as well as the recovery in advanced econ-
–5
omy investment and domestic demand more generally.
–10
Among advanced economies, large exporters, such as
–15
Germany, Japan, the United Kingdom, and the United
–20
States, contributed strongly to the recovery in exports 1998 2000 02 04 06 08 10 12 14 16 17
(Figure 1.4, panel 1), while the recovery in imports
was broad based, except in the United Kingdom (Fig- 35 2. China
ure 1.4, panel 2). 30
Among emerging market and developing economies, 25
20
as shown in Figure 1.4, panel 3, the rebound in export
15
growth was particularly strong in emerging Asia, 10
especially China.1 In contrast, the rebound in imports 5
largely reflects an import recovery among commod- 0
ity exporters—countries that had earlier experienced –5
1998 2000 02 04 06 08 10 12 14 16 17
sharp investment and import contractions during the
2015–16 commodity price downturn. This is shown in 30 3. Selected Commodity
Figure 1.4, panel 4: the blue bars represent commodity Exporters1
20
exporters that had a particularly pronounced cycle in
10
imports (Angola, Brazil, Ecuador, Nigeria, Russia); the
green bars represent remaining commodity exporters, 0
which account for an important part of the import –10
demand cycle among other emerging market and –20
developing economies. –30
1998 2000 02 04 06 08 10 12 14 16 17
Rising Commodity Prices 25 4. Emerging Market and Developing Economies Excluding

20 China and Selected Commodity Exporters1
The IMF’s Primary Commodities Price Index rose
15
16.9 percent between August 2017 and February 10
2018—that is, between the reference periods for the 5
October 2017 WEO and the current report (Fig- 0
ure 1.5). As described in the Commodities Special Fea- –5
ture, the increase was driven primarily by rising oil and –10
–15
natural gas prices. Among the other subindices, metals 1998 2000 02 04 06 08 10 12 14 16 17
and agricultural commodity prices also rose, although
less rapidly than energy prices. Source: IMF staff calculations.
1
•• Oil prices increased to more than $65 a barrel in Selected commodity exporters = Angola, Brazil, Ecuador, Nigeria, Russia.
January, the highest level since 2015, following
1Box 1.1 discusses the role of the so‑called tech cycle in explaining
the rebound in trade in Asian economies and elsewhere.

Figure 1.4. Contributions to Trade Growth Figure 1.5. Commodity and Oil Prices
(Percent) (Deflated using US consumer price index; index, 2014 = 100)
The trade recovery was particularly pronounced in emerging market and Commodity prices, notably of oil and natural gas, have risen since the fall, but the
developing economies. medium-term outlook remains subdued.
7 1. Advanced Economies: Export Growth 180

6 Germany Japan Average petroleum spot price
5 United Kingdom United States 160 Food
Other AEs1 Total1 Metals
4
3 140
2
120
1
0 100
–1
2011 12 13 14 15 16 17
80
7 2. Advanced Economies: Import Growth 60

6 Germany Japan
United Kingdom United States 40
5
Other AEs1 Total1
4
20
3
2
0
1 2011 12 13 14 15 16 17 18 19
0
–1 Sources: IMF, Primary Commodity Price System; and IMF staff estimates.
2011 12 13 14 15 16 17
10 3. Emerging Market and Developing Economies: Export Growth

Selected commodity exporters China unplanned outages on the US Gulf Coast and in
8
Emerging Asia excluding China Other EMDEs
6 Total Libya, the North Sea, and Venezuela; an extension
4
to the end of 2018 of the Organization of the Petro-
leum Exporting Countries agreement on production
2
targets; and stronger global economic growth. Prices
0
moderated to $63 a barrel in February, 27 percent
–2 above their August level.
2011 12 13 14 15 16 17
•• The natural gas price index—an average for Europe,
16 4. Emerging Market and Developing Economies: Import Growth Japan, and the United States—rose sharply, by
Selected commodity exporters China 45 percent from August 2017 to February 2018,
12 Emerging Asia excluding China Other EMDEs reflecting seasonal factors. Strong demand for
Total
8 liquefied natural gas (LNG) in China, where the
government has restricted the use of coal to mitigate
4
air pollution, helped drive the spot LNG price to
0 its highest level in three years. Higher oil prices also
–4 added upward pressure in countries where oil‑linked
2011 12 13 14 15 16 17
pricing is more common.
•• Metal prices increased 8.3 percent from August to
Source: IMF staff calculations.
Note: Trade growth reflects export and import volumes from external sector data. February, in line with stronger growth in all major
AEs = advanced economies; EMDEs = emerging market and developing economies. Demand for base metals—especially
economies; selected commodity exporters = Angola, Brazil, Ecuador, Nigeria, aluminum—was strong, while supply was limited in
Russia.
1
Excludes Ireland. part due to China’s production capacity cuts. Iron
ore prices rose 4.1 percent from August to February,
rallying recently thanks to strong steel prices and
rising coal costs.

•• The IMF’s agricultural price index rose 4.1 percent Figure 1.6. Global Inflation
(Three-month moving average; annualized percent change, unless noted
from August 2017 to February 2018, as unfavorable
otherwise)
weather conditions in recent months are expected Headline inflation has picked up, reflecting stronger fuel prices, but core inflation
to reduce this year’s harvests of many grains and remains soft.
oilseeds. The subindices of food and agricultural Consumer price inflation Core consumer price inflation
raw materials rose 4.1 percent and 6.0 percent, 3 1. Advanced Economies 2. Emerging Market and 7
respectively. Developing Economies
2 6
1 5
Headline Inflation Has Picked up, but Core Inflation 0 4
Remains Sluggish –1 3
–2 2
With the upturn in oil prices since September, 2013 14 15 16 Feb. 2013 14 15 16 Feb.
headline consumer price inflation has picked up 18 18
again (Figure 1.6). Core inflation—inflation rates
20 3. Producer Price Inflation1 2.2 4. Consumer Price 3.8
when fuel and food prices are excluded—generally 2.1 Inflation Expectation
15 World AEs
remains soft. It has begun to show signs of recovery 2.0 (Percent) 3.6
10 EMDEs 1.9
in advanced economies and appears to have bottomed 5 1.8 3.4
out in emerging market and developing economies. As 0 1.7
2
1.6 AEs 3.2
illustrated in Box 1.2, the continued weakness of infla- –5 1.5 EMDEs (right scale)
tion in advanced economies relative to precrisis years –10 1.4 3.0
2013 14 15 16 Feb. Jan. Jul. Jan. Jul. Mar.
reflects primarily nontraded consumer services, such as 18 16 16 17 17 18
medical services and education. Traded goods inflation
5. Core Consumer Price 6. US Labor3
has remained low but has not declined. 5 Inflation (Percent of labor force) 14
•• In most advanced economies, core inflation remains 4 (Percent change from Unemployment rate 12
3 a year ago) Involuntary part-time 10
below target but appears to be edging up in response employment
2 8
to stronger demand. In the United States, where 1
Hourly wage growth, annual
6
unemployment is close to its lowest level since the average
0 4
EA United States
late 1960s, core personal consumer expenditure –1 Japan United Kingdom 2
inflation (the Federal Reserve’s preferred measure) –2 0
Jan. Apr. Jul. Oct. Jan. Apr. Jul. Oct. Feb. 1999–2000 2006–07 2016–17
has begun to firm. In February, it stood at about 16 16 16 16 17 17 17 17 18
1.6 percent when measured on a 12-month basis,
5 7. Unemployment Rate and Wage Growth in AEs4 3
but slightly above 2 percent (the Federal Reserve’s (Percent)
Unemployment rate (right inverted scale) 4
medium-term target), measured on a three-month 4
Wage rate (two-quarter moving average; 5
(annualized) basis. Twelve-month core inflation 3
percent change from a year ago)
6
notched up to 1.1 percent in the euro area in 7
2
February (just above its average for the past cou- 8
ple of years), while in Japan it has remained on a 1 9
2005 06 07 08 09 10 11 12 13 14 15 16 Dec.
gentle upward trajectory in recent months, reaching 17
0.4 percent in January. The United Kingdom is an
Sources: Consensus Economics; Haver Analytics; Organisation for Economic
exception to the pattern of below-target inflation. At Co-operation and Development; US Bureau of Labor Statistics; and IMF staff
2.4 percent in February, UK core inflation is below calculations.
Note: AEs = advanced economies (AUT, BEL, CAN, CHE, CZE, DEU, DNK, ESP, EST,
the peak it reached in 2017 in the aftermath of the FIN, FRA, GBR, GRC, HKG, IRL, ISR, ITA, JPN, KOR, LTU, LUX, LVA, NLD, NOR, PRT,
June 2016 Brexit referendum pound depreciation, SGP, SVK, SVN, SWE, TWN, USA); EA = euro area; EMDEs = emerging market and
developing economies (BGR, BRA, CHL, CHN, COL, HUN, IDN, IND, MEX, MYS, PER,
but remains above the Bank of England’s target of 2 PHL, POL, ROU, RUS, THA, TUR, ZAF). Panel 6 is equalized to 100 in 2007 by
percent. shifting the level. Country list uses International Organization for Standardization
(ISO) country codes.
•• Wage growth also remains tepid in most advanced 1
AEs excludes HKG, ISR, and TWN. EMDEs includes UKR; excludes IDN, IND, PER,
economies, moving broadly in line with labor and PHL.
2
productivity when measured in real terms (hence AEs includes AUS; excludes LUX.
3
Hourly wage growth refers to the growth of production and nonsupervisory
implying a limited increase in unit labor costs). As workers in private industries.
4
documented in Chapter 2 of the October 2017 Blue line includes AUS and NZL; excludes BEL. Red line includes AUS and MLT;
excludes HKG, SGP, and TWN.

WEO, the sluggishness in wages partly reflects pace of its asset purchase program from €60 billion to
continued slack in labor markets, especially a €30 billion, with purchases intended to continue until
still-elevated share of workers involuntarily work- the end of September 2018, or beyond if necessary.
ing part-time. Changes in the composition of the Among other advanced economies, the United Kingdom
workforce—new entrants earning relatively lower raised its bank rate to 50 basis points in November and
wages than retiring workers—may also have played Canada raised its policy rate to 1.25 percent in January.
a role. The January uptick in US hourly earnings With strengthening economic activity and expecta-
growth was a welcome sign of a firming labor mar- tions of more rapid increases in the policy rate in the
ket after a period of strong payroll gains. A sustained United States, nominal yields on 10-year US Treasury
acceleration of labor earnings will be needed to push bonds have risen by over 50 basis points since August
real wage growth above labor productivity gains, (as of end March 2018). This increase reflects primarily
raise cost pressures for firms, and support the return a steeper expected path for short-term interest rates.
of core inflation toward the medium-term target. Over the same period, long-term bond yields have
•• In many emerging market and developing econ- risen by some 10 basis points in Germany and 25
omies, recent currency stability or appreciations basis points in the United Kingdom, while they have
against the US dollar have helped keep a lid on remained around zero in Japan. Long term bond yields
core inflation. Core inflation is around historical have remained broadly unchanged in Italy and Spain,
lows in Brazil and Russia, where demand has been as their spreads over German bunds have compressed
recovering from the deep contractions of 2015–16, with the increase in German yields.
while it has picked up in India after falling sharply Despite the early February turbulence and declines
in the second quarter of 2017 due to one-off factors. in March following the announcements of intended
In China, core inflation remains broadly stable at US tariff actions on steel and aluminum and a range
about 2 percent. In contrast, other countries—in of Chinese products, as well as the announcement
sub‑Saharan Africa; the Commonwealth of Inde- by China of retaliatory tariffs on imports from the
pendent States; and the Middle East, North Africa, US, equity market valuations remain stronger than in
Afghanistan, and Pakistan region—continue to August (Figure 1.7, panel 5). Volatility has subsided
grapple with high inflation stemming from the but remains higher than the pre-February episode
pass-through of earlier exchange rate depreciations. lows, with spillovers beyond equity markets generally
contained. Corporate credit spreads are tighter or little
changed relative to August (Figure 1.7 panel 6).
Financial Conditions—Still Loose Despite widening interest rate differentials, the US
Despite equity market turbulence in early Feb- dollar weakened modestly in real effective terms, by
ruary, equity market declines in March, and some about 1½ percent between August 2017 and end-
increases in bond yields in response to firmer March 2018, and is about 4 ½ percent weaker than its
growth and inflation, market sentiment generally 2017 average (Figure 1.8). The euro has appreciated by
appears stronger than in August. Confidence in the around 1 percent and stands about 4 percent stronger
strength of the global outlook has gained ground, than its 2017 average. Among other currencies, the
and financial conditions remain accommodative Japanese yen has remained broadly stable, while the
and supportive of the recovery, as discussed in the British pound appreciated 5 ½ percent after the Bank
April 2018 Global Financial Stability Report (GFSR). of England raised interest rates in November and as
Central bank monetary policy moves have been expectations of a Brexit deal rose.
well telegraphed and absorbed smoothly by markets. In emerging market economies, financial conditions
Withdrawal of monetary support in the United States since August have generally remained supportive of a
has continued, with increases in short-term interest rates pickup in economic activity. Monetary policy was eased
in December and March amid a firmer labor market further in Brazil and Russia, while it was tightened in
and emerging signs of strengthening inflation. Markets Mexico. Equity markets have strengthened (Figure 1.9)
are currently pricing in two additional interest rate and spreads on the J.P. Morgan Global Emerging Mar-
increases in 2018—a more rapid pace of normalization kets Bond Index have declined (Figure 1.10). Long-term
than expected a few months ago (Figure 1.7). In January interest rates on local currency bonds have increased
2018, the European Central Bank reduced the monthly modestly in countries growing rapidly, such as in emerg-

Figure 1.7. Advanced Economies: Monetary and Financial Figure 1.8. Real Effective Exchange Rate Changes,
Market Conditions August 2017–March 2018
(Percent, unless noted otherwise) (Percent)
With stronger domestic demand, a steeper path of expected policy rates has lifted Exchange rate movements since the fall have been modest across advanced
US long-term yields since the fall. Yields have risen to a lesser extent among other economies and for most emerging market and developing economies.
advanced economies.
Latest relative to February 2018

3.0 1. US Policy Rate 2. Policy Rate Expectations1 7.0 February 2018 relative to August 2017
Expectations1 (Percent; dashed lines
6.0
2.5 are from the October
2017 WEO) 5.0 8 1. Advanced Economies1
2.0 United States 6
4.0
Euro area
1.5 United Kingdom 3.0 4
2.0 2
1.0
Apr. 3, 2017 1.0 0
0.5 Sep. 15, 2017
Mar. 21, 2018 0.0 –2
0.0 –1.0 –4
2017 18 19 20 Mar. 2017 18 19 20 Mar.
21 21 –6
–8
6 3. Key Interest Rates2 4. Credit Spreads3 1,000 USA EA JPN GBR SWE CHE KOR TWN SGP CAN NOR AUS NZL
US average 30-year (Basis points)
5 1
fixed-rate mortgage US high yield 800 12 2. Emerging Market Economies
4
9
3 600
United States 6
2 400 3
1 Euro high yield 0
Germany
Japan US high grade 200
0 –3
Euro high grade –6
–1 0
2013 14 15 16 17 Mar. 2013 14 15 16 17 Mar. –9
18 18
–12
ZAF CHN IND IDN MYS PHL THA HUN POL RUS TUR BRA CHL COL MEX PER
200 5. Equity Markets 6. Price-to-Earnings Ratios3 35
180 (Index, 2007 = 100)
United States Source: IMF staff calculations.
160 S&P 500 Japan 30 Note: EA = euro area. Data labels use International Organization for Standardization
140 Germany (ISO) country codes.
120 Italy 25 1
Latest data available are for March 30, 2018.
100
80 20
60
MSCI Emerging Market
40 15
Euro Stoxx ing Asia and emerging Europe, while they eased further
20 TOPIX
0 10 in Latin America and in Russia.
2013 14 15 16 17 Mar. 2013 14 15 16 17 Mar. Among emerging market currencies, the Chinese
18 18
renminbi appreciated 3½ percent in real effective terms
Sources: Bloomberg Finance L.P.; Thomson Reuters Datastream; and IMF staff
between August 2017 and end-March 2018 and by a
calculations. similar amount relative to its average value in 2017.
Note: MSCI = Morgan Stanley Capital International; S&P = Standard & Poor’s; The South African rand rebounded by 10 percent on
TOPIX = Tokyo Stock Price Index.
1
Expectations are based on the federal funds rate futures for the United States, the reduced political uncertainty and the Malaysian ringgit
sterling overnight interbank average rate for the United Kingdom, and the euro by over 8 percent on an improved growth outlook and
interbank offered forward rate for the euro area; updated March 21, 2018.
2
Interest rates are 10-year government bond yields, unless noted otherwise. Data stronger commodity prices. In contrast, the Turkish
are through March 23, 2018. lira depreciated by more than 10 percent on higher
3
Data are through March 23, 2018.
inflation readings.
Financial flows to emerging market economies mod-
erated in the second half of 2017 after surging in the

Figure 1.9. Emerging Market Economies: Equity Markets and Figure 1.10. Emerging Market Economies: Interest Rates
Credit
Emerging market bond spreads have declined, while yields on local-currency
long-term bonds have increased modestly in some fast-growing economies.
Financial conditions in emerging market economies generally remain supportive of
a pickup in economic activity. Emerging Europe China
Emerging Asia excluding China Latin America
220 1. Equity Markets
(Index, 2007 = 100) Emerging Asia Latin America 13 1. Policy Rate
200 excluding China (Percent)
12
180 11
160 10
140 9
120 8
7
100
6
80 5
60 Emerging Europe China 4
2012 13 14 15 16 17 Mar.
40
2012 13 14 15 16 17 Mar. 18
18 10 2. Real Policy Rates1
1 (Percent) August 2016
Real Credit Growth 8 August 2016 average
(Year-over-year percent change)
6 March 2018
March 2018 average
25 2. BRA CHN 3. COL IDN 40 4
20 IND MEX MYS RUS 2
TUR 30
15 0
10 20 –2
5 BRA CHL CHN COL IDN IND MEX MYS PER PHL POL RUS THA TUR ZAF
0 10
14 3. Ten-Year Government Bond Yields2
–5 (Percent)
0 12
–10
10
–15 –10
2012 13 14 15 16 Jan. 2012 13 14 15 16 Jan. 8
18 18
6
Credit-to-GDP Ratio1
(Percent) 4
2
85 4. 230 5. 30 2013 14 15 16 17 Mar.
BRA COL MEX (right scale) 18
75 IDN IND 210 CHN 25 600 4. EMBI Sovereign Spreads2
65 RUS TUR 190 MYS (Basis points)
20 500
55 170
15 400
45 150 300
10
35 130 200
25 110 5
100
15 90 0 0
2006 08 10 12 14 16 17: 2006 08 10 12 14 16 17: 2013 14 15 16 17 Mar.
Q4 Q4 18
Sources: Bloomberg Finance L.P.; Haver Analytics; IMF, International Financial
Statistics (IFS) database; and IMF staff calculations. Sources: Bloomberg Finance L.P.; Haver Analytics; IMF, International Financial
Note: Data labels use International Organization for Standardization (ISO) country Statistics; and IMF staff calculations.
codes. Note: Emerging Asia excluding China comprises India, Indonesia, Malaysia, the
1
Credit is other depository corporations’ claims on the private sector (from IFS), Philippines, and Thailand; emerging Europe comprises Poland, Romania, Russia,
except in the case of Brazil, for which private sector credit is from the Monetary and Turkey; Latin America comprises Brazil, Chile, Colombia, Mexico, and Peru.
Policy and Financial System Credit Operations published by Banco Central do EMBI = J.P. Morgan Emerging Markets Bond Index. Data labels use International
Brasil, and China, for which credit is total social financing after adjusting for local Organization for Standardization (ISO) country codes.
government debt swaps. 1
Deflated by two-year-ahead World Economic Outlook inflation projections.
2
Data are through March 30, 2018.

first half of the year but remained robust. Following a Figure 1.11. Emerging Market Economies: Capital Flows
strong start to 2018, portfolio flows to emerging mar-
ket economies softened in the immediate aftermath of Portfolio flows to emerging market economies softened immediately after the
global equity market turbulence of early February, but have recovered since.
the global equity market turbulence of early February
but have recovered since (Figure 1.11). 40 1. Net Flows in Emerging Market Funds Bond US presidential
30 (Billions of US dollars) Equity election
Taper
20 EM-VXY
tantrum
Key Forces Shaping the Outlook 10
0
Advanced Economies: Output Gaps Closing amid
–10
Structurally Stronger Growth China
–20 Greek Irish 1st ECB
equity market
Since 2014 advanced economies have experienced a –30 crisis crisis LTROs
sell-off
continued, if at times halting, recovery from the reces- –40
2010 11 12 13 14 15 16 17 Mar.
sions in the aftermath of the 2008–09 global financial 18
crisis and the 2011–12 euro area sovereign debt crisis. 15 2. Capital Inflows Emerging Europe
Accommodative monetary policy and the gradual fad- 12 (Percent of GDP) Emerging Asia excluding China
ing of crisis-related drags have been pivotal in helping 9
Latin America
advanced economies attain above‑potential growth 6
and reduce unemployment. Measures of potential 3
growth and output gaps are inherently very uncertain, 0
especially in the aftermath of a deep crisis with lasting China Total
–3 Saudi Arabia
macroeconomic legacies. Nonetheless, potential growth –6
for advanced economies is also estimated to have 2007 08 09 10 11 12 13 14 15 16 17:
Q4
recovered in recent years.2
15 3. Capital Outflows Excluding Change in Reserves
The faster-than-expected pace of activity in advanced (Percent of GDP)
12
economies since mid-2016 has not only sped up the Emerging Europe
9 Emerging Asia excluding China
closing of output gaps, it has also led to a reassessment Latin America
6
of medium-term output.
3
•• Some 40 percent of the 0.6 percentage point
0
cumulative growth surprise for 2016–17 relative to China Total
–3
the October 2016 WEO projections is attributed Saudi Arabia
–6
to a faster-than-expected closing of output gaps 2007 08 09 10 11 12 13 14 15 16 17:
(a cyclical recovery in demand), while the rest has Q4
been matched by an upward revision to estimated 15 4. Change in Reserves
(Percent of GDP)
potential growth (implying a structurally stron- 12 Emerging Europe
Emerging Asia excluding China
ger recovery). 9
Latin America
•• Likewise, about 40 percent of the 1.7 percentage 6
point revision to cumulative growth in advanced 3
economies during 2016–21 (relative to the Octo- 0
China Total
ber 2016 WEO projections) is attributed to faster –3
Saudi Arabia
closing of output gaps; the rest is attributed to faster –6
2007 08 09 10 11 12 13 14 15 16 17:
potential growth. Higher potential output relative Q4
to earlier projections implies that employment is Sources: Bloomberg Finance L.P.; EPFR Global; Haver Analytics; IMF, International
Financial Statistics; and IMF staff calculations.
Note: Capital inflows are net purchases of domestic assets by nonresidents.
2Box 1.3 updates the potential growth projections in Chapter 3
Capital outflows are net purchases of foreign assets by domestic residents.
of the April 2015 WEO. The analysis—based on multivariate Emerging Asia excluding China comprises India, Indonesia, Malaysia, the
filtering techniques—suggests a pickup in potential growth of about Philippines, and Thailand; emerging Europe comprises Poland, Romania, Russia,
0.4 percentage point between 2011 and 2017 in a selected group of and Turkey; Latin America comprises Brazil, Chile, Colombia, Mexico, and Peru.
advanced economies. The estimated change in potential growth is ECB = European Central Bank; EM-VXY = J.P. Morgan Emerging Market Volatility
Index; LTROs = longer-term refinancing operations.
almost identical to the pickup for the aggregated group of advanced
economies over the same period in the current WEO projections,
which also incorporate country-specific factors.

Figure 1.12. Terms-of-Trade Windfall Gains and Losses expected to push output above potential, is projected
to see a later, but sharper, return to potential growth
Despite the projected short-term increase in commodity prices, terms-of-trade than most other advanced economies. Box 1.5 presents
windfall gains and losses are expected to be modest over 2018–19 compared with
2015–17. a stylized scenario analysis of the elements of the US
tax reform to shed light on why the US economy is
1
15 1. Oil Exporters projected to grow considerably faster than potential for
10 (Percent of GDP)
a few years. The simulations illustrate that the tempo-
5
rary allowance for full expensing of investment has a
0
–5 particularly large short-term impact on activity because
–10 it provides strong incentives to firms to advance and
–15 2015–16 (cumulative) complete investment projects while the allowance is
–20 2017
2018–19 (average) in place. As a result, the US tax reform will reduce
–25
growth momentum starting in 2020, and then more
–30
–35 strongly when full investment expensing begins to be
SAU KAZ IRN COL CAN MEX BRA phased out in 2023.
DZA NGA RUS MYS AUS ARG IDN
The medium-term per capita growth rates of
advanced economies are expected to be lower—not
8 2. Oil Importers1
7 (Percent of GDP) only than they currently are, but also below those
6 2015–16 (cumulative) registered in the precrisis decades. The main reason is
5 2017 the slowdown in labor force growth as populations of
4 2018–19 (average)
3 advanced economies continue to age (as discussed in
2 Chapter 2), a drag that is expected to be offset only
1 partially by some recovery in the growth of total factor
0
–1 productivity (to rates that are well below those regis-
–2 tered in the precrisis years; Box 1.4 discusses produc-
–3 tivity measurement in the digital age).
USA TUR POL ITA ESP IND THA
EGY FRA DEU CHN JPN PAK KOR
Emerging Market and Developing Economies: Effects
Source: IMF staff estimates. of Recent Commodity Price Increases
Note: Data labels use International Organization for Standardization (ISO) country
codes. The declines in metal prices since 2011 and the
1
Gains (losses) for 2018–19 are simple averages of annual incremental gains plunge in oil prices in 2014 drove a wedge between the
(losses) for 2018 and 2019. The windfall is an estimate of the change in
disposable income arising from commodity price changes. The windfall gain in economic performance of commodity-importing and
year t for a country exporting x US dollars of commodity A and importing m commodity-exporting emerging market and develop-
US dollars of commodity B in year t – 1 is defined as (ΔptAxt – 1 – ΔptBmt – 1) / Yt – 1,
in which ΔptA and ΔptB are the percentage changes in the prices of A and B ing economies (Figures 1.12 and 1.13). The growth
between year t – 1 and year t, and Y is GDP in year t – 1 in US dollars. See also rates of the two groups were broadly similar before
Gruss (2014).
2014 (excluding faster-growing China) but have since
diverged, with importers continuing to grow fast and
exporters seeing their growth slow to about half of its
expected to be sustained at a higher level as well.3 average 2000–14 pace. With idiosyncratic problems
The continued decline in headline unemployment exacerbating the loss in commodity revenues, some
rates, with limited signs of wage and price accelera- larger exporters—such as Brazil and Russia—experienced
tion, is consistent with this interpretation. deep recessions in 2015–16, while Venezuela has suf-
fered an intensifying economic and humanitarian crisis
Once the gaps close (estimated to occur by the end since 2014. Likewise, Saudi Arabia and some other oil
of 2018 for the advanced economy group), growth exporters in the Middle East and sub-Saharan Africa
is expected to start declining toward potential. The have experienced recessions and/or substantial growth
United States, where recent fiscal policy changes are slowdowns in recent years as they started adjusting fiscal
3Advanced economy employment projections for 2021 have
policy to the permanent loss of commodity revenues.
been raised by about 1.4 million relative to those in the Octo- Output, and especially domestic demand, decel-
ber 2016 WEO. erated sharply in oil exporters in the aftermath of

terms-of-trade losses, which gave rise to large fiscal Figure 1.13. GDP Growth, 1999–2023
(Percent)
and external adjustment needs and tighter financial
conditions. The extent of macroeconomic stress asso-
Growth in commodity exporters is projected to stabilize close to current levels over
ciated with the large decline in oil prices has become the medium term, well below the past average. Diversified economies are
more apparent over time, with projected growth in expected to maintain relatively robust growth rates.
oil exporters’ GDP, and especially domestic demand,
revised down through 2017 even as oil prices firmed 20 1. Growth in Emerging Market and Developing Economies
Commodity exporters
somewhat. Looking ahead, the increase in commodity 16 Noncommodity exporters in EMDEs excluding China
prices in the second half of 2017 creates space for oil China
12
exporters to consolidate fiscal balances more gradu-
ally but is only a very partial reversal of their initial 8
terms-of-trade losses during 2014–16. In some cases, 4
the price increase also reflects production restraints
that directly weigh on real GDP. In addition, domestic 0
political discord and strife continue to weigh heavily –4

1999 2001 03 05 07 09 11 13 15 17 19 21 23
on economic activity in several oil exporters. As a
result of these offsetting forces, the recovery in growth
Advanced economies Noncommodity exporters excluding China,
in oil exporters since the 2015 trough has been very India India, and Brazil
gradual, and growth projections for the next five years China Commodity exporters
are broadly unchanged since October 2017.
For oil importers, when oil prices fall, the windfall 5 2. Contributions to World 3. Contributions to World GDP 5
GDP Growth Growth
gains as a share of income tend to be smaller than (PPP weights) (Market weights)
4 4
the corresponding losses for oil exporters, given that
the oil import bills of the former group are generally 3 3
lower as a share of overall income than the oil export
receipts of the second, smaller, group. The boost to 2 2
domestic demand in oil importers stemming from
1 1
the oil price decline of 2014 was, in many cases,
partially offset by a reduction in energy subsidies, 0 0
2017 2023 2017 2023
which implies an incomplete pass-through of the
windfall to final users. To the extent that the recent
18 4. Growth in Low-Income Developing Countries
oil price increases are passed on to final users, they 16 Nigeria
may temper domestic demand. The negative effect, in 14 Others
Commodity exporters excluding Nigeria
many cases, is not large enough to trigger downward 12
and Yemen
growth revisions, however, given offsetting improve- 10
ments in external conditions, in particular stronger 8
6
external demand. 4
2
Prospects for Income Convergence—A Glass One- 0
Quarter Empty –2
1999 2001 03 05 07 09 11 13 15 17 19 21 23
The record of income convergence between
advanced economies and emerging market and Source: IMF staff estimates.
developing economies has not been favorable over Note: EMDEs = emerging market and developing economies; PPP = purchasing
power parity. Commodity exporters includes fuel and nonfuel primary products
the past five decades (as discussed in Chapter 2 of the exporters, as indicated in Table D of the Statistical Appendix, plus Brazil and Peru.
April 2017 WEO). Over the next five years, the glass
will be one-quarter empty: 40 emerging market and
developing economies (about 27 percent of the total)
are not expected to narrow their per capita income
gaps relative to advanced economies. In fact, per cap-
ita incomes in 12 of those economies are expected to

Figure 1.14. Per Capita Real GDP Growth one-quarter (14 countries), but these represent a
(Percent)
larger share of the total population and GDP for the
country group (some 30 percent).
Prospects for emerging market and developing economies to narrow their per
capita income gaps relative to advanced economies vary across regions. Convergence prospects vary across regions. Income
convergence is projected to continue in China, India,
1995–2005 2006–17 2018–23 and east Asia more broadly, as well as in emerging
Europe and parts of the Commonwealth of Indepen-
10 1. By Country Group dent States. By contrast, per capita growth in sub-
8 Saharan Africa, Latin America and the Caribbean, and
6 the Middle East, North Africa, Pakistan, and Afghani-
4 stan region is projected to fall short of or barely exceed
2 that in advanced economies over the next few years,
0 reflecting the weak performance of the many commod-
–2 ity exporters in these regions.
–4
–6
AEs EMDEs China Fuel exporters Nonfuel The Forecast
exporters
excluding China
Policy Assumptions
8 2. Emerging Market and Developing Economies by Region The aggregate fiscal policy stance for advanced econ-
omies is projected to remain expansionary in 2018 and
6 especially in 2019, while it is projected to turn broadly
neutral in emerging market and developing economies
4
(Figure 1.15). Relative to the October 2017 WEO
2 assumptions, the forecast assumes a looser fiscal policy
stance in 2018 and 2019, which reflects, to a large
0 extent, expected weaker US structural fiscal balances in
light of the recently legislated overhaul of the tax code.
–2
LAC MENAP EMDE Asia EMDE SSA CIS Fiscal policy is expected to be mildly contractionary
excluding Europe in advanced economies for 2020–22 and more clearly
China
contractionary in 2023, when the investment expens-
Source: IMF staff estimates. ing provisions of US tax reform begin to expire.
Note: Bars denote PPP GDP-weighted averages, red markers indicate the medians, On monetary policy, the forecast assumes faster
and black markers denote the top and bottom deciles of per capita GDP growth in
the country groups. Country groups are defined in Chapter 3 of the April 2015 normalization of the policy interest rate in the
World Economic Outlook. The fuel and nonfuel exporter subgroups are defined in United States than projected in the October 2017
Table D of the Statistical Appendix and cover EMDEs only. AEs = advanced
economies; CIS = Commonwealth of Independent States; EMDE = emerging
WEO, reflecting stronger demand and inflation
market and developing economy; LAC = Latin America and the Caribbean; MENAP pressure under more expansionary fiscal policy. The
= Middle East, North Africa, Afghanistan, and Pakistan; PPP = purchasing power US policy interest rate target is projected to rise to
parity; SSA = sub-Saharan Africa.
about 2.5 percent by the end of 2018 and about
3.5 percent by the end of 2019, declining back to
a long-term equilibrium rate of slightly less than
decline over the five-year forecast horizon. Most econ- 3 percent in 2022. In the euro area and Japan, the
omies with per capita growth below that of advanced forecast assumes that monetary policy will remain
economies are either commodity (mostly oil) export- very accommodative. Short-term rates are projected
ers or small states (Figure 1.14)—they account for a to remain negative in the euro area until mid‑2019
smaller share of the total population and GDP of all and close to zero in Japan over the five-year forecast
emerging market and developing economies (about horizon. The assumed monetary policy stances across
11 percent). If the sample is limited to low-income emerging market economies and the revisions relative
developing countries, the share of the countries not to October 2017 vary, reflecting these economies’
expected to narrow their per capita income gap is diverse cyclical positions.

Figure 1.15. Fiscal Indicators Assumptions on Financial Conditions and

(Percent of GDP, unless noted otherwise)
Commodity Prices
The fiscal policy stance is projected to remain expansionary in advanced Global financial conditions are assumed to remain
economies in 2018 and especially 2019, while it is projected to turn broadly generally accommodative during 2018–19. Contin-
neutral in emerging market and developing economies.
ued easing of lending conditions, notably in the euro
1.0 1. Change in the Structural Primary Fiscal Balance
area, is expected to offset the anticipated gradual rise
(Percentage points) in long-term interest rates, while the normalization of
0.5 monetary policy in the United States and the United
0.0 Kingdom is expected to proceed without trigger-
ing large or protracted increases in financial market
–0.5
2014 2015 2016 volatility. Except for some vulnerable economies, most
–1.0 2017 2018 2019 emerging markets are expected to face accommodative
October 2017 WEO
–1.5 financial conditions under the baseline forecast, with
Advanced Emerging market and higher policy rates but sustained risk appetite (continu-
economies developing economies
ing the recent record of generally contained sovereign
bond spreads and strong equity market performance in
3 2. Change in the Structural Primary Fiscal Balance
(Percentage points) most cases).
2014 2015 2016
2 2017 2018 2019 The IMF’s commodity price index is expected to
October 2017 WEO rise about 11.9 percent in 2018 relative to its 2017
1 average (bringing the cumulative increase from 2016
to about 28.9 percent) and then to fall about 3.7 per-
0
cent in 2019. Oil prices are expected to average $62.3
–1
a barrel in 2018 (up from $52.8 in 2017 and well
United States Japan1 France, Germany, Greece, Ireland, above the projection of $50.2 a barrel in the Octo-
United Kingdom Italy, Portugal, Spain
ber 2017 WEO). As supply recovers, oil prices are
expected to decline to $58.2 a barrel in 2019, and
2 3. Fiscal Balance
further to about $53.6 a barrel in 2023. Metal prices
0
are expected to strengthen by 13 percent in 2018,
–2 following a 22.2 percent increase in 2017 spurred by
–4 stronger global demand, and remain broadly sta-
–6 World ble thereafter.
Advanced Emerging market and
–8 economies developing economies
–10 Global Growth Outlook: Short-Term Strengthening,
2001 03 05 07 09 11 13 15 17 19 21 23
Medium-Term Moderation
180 4. Gross Public Debt Global growth is projected to strengthen from
160 World Advanced economies2 3.8 percent in 2017 to 3.9 percent in 2018 and
140 Major advanced Emerging and Latin America and 2019, driven by a projected pickup in growth in
economies2,3 developing Asia the Caribbean
120 emerging market and developing economies and
Other emerging market
100 and developing economies resilient growth in advanced economies (Table 1.1).
80
60
The forecast for 2018 and 2019 is stronger than in
40 the October 2017 WEO by 0.2 percentage point for
20 each year, with positive revisions compared with the
1950 60 70 80 90 2000 10 20 23
October 2017 WEO for emerging market and devel-
oping economies and especially for advanced econ-
Source: IMF staff estimates.
Note: WEO = World Economic Outlook. omies. The global effects of US fiscal policy changes
1
Japan’s latest figures reflect comprehensive methodological revisions adopted in account for almost half of the global growth upgrade
December 2016.
2
Data through 2000 exclude the United States. for 2018–19 compared with October. Beyond 2019,
3
Canada, France, Germany, Italy, Japan, United Kingdom, United States. global growth is projected to gradually decline to
3.7 percent by the end of the forecast horizon. The

Table 1.1. Overview of the World Economic Outlook Projections

(Percent change, unless noted otherwise)
Difference from January Difference from October
Projections 2018 WEO Update1 2017 WEO1
2017 2018 2019 2018 2019 2018 2019
World Output 3.8 3.9 3.9 0.0 0.0 0.2 0.2
Advanced Economies 2.3 2.5 2.2 0.2 0.0 0.5 0.4
United States 2.3 2.9 2.7 0.2 0.2 0.6 0.8
Euro Area 2.3 2.4 2.0 0.2 0.0 0.5 0.3
Germany 2.5 2.5 2.0 0.2 0.0 0.7 0.5
France 1.8 2.1 2.0 0.2 0.1 0.3 0.1
Italy 1.5 1.5 1.1 0.1 0.0 0.4 0.2
Spain 3.1 2.8 2.2 0.4 0.1 0.3 0.2
Japan 1.7 1.2 0.9 0.0 0.0 0.5 0.1
United Kingdom 1.8 1.6 1.5 0.1 0.0 0.1 –0.1
Canada 3.0 2.1 2.0 –0.2 0.0 0.0 0.3
Other Advanced Economies2 2.7 2.7 2.6 0.1 0.0 0.2 0.1
Emerging Market and Developing Economies 4.8 4.9 5.1 0.0 0.1 0.0 0.1
Commonwealth of Independent States 2.1 2.2 2.1 0.0 0.0 0.1 0.0
Russia 1.5 1.7 1.5 0.0 0.0 0.1 0.0
Excluding Russia 3.6 3.5 3.6 0.1 0.1 0.2 0.1
Emerging and Developing Asia 6.5 6.5 6.6 0.0 0.0 0.0 0.1
China 6.9 6.6 6.4 0.0 0.0 0.1 0.1
India3 6.7 7.4 7.8 0.0 0.0 0.0 0.0
ASEAN-54 5.3 5.3 5.4 0.0 0.1 0.1 0.1
Emerging and Developing Europe 5.8 4.3 3.7 0.3 –0.1 0.8 0.4
Latin America and the Caribbean 1.3 2.0 2.8 0.1 0.2 0.1 0.4
Brazil 1.0 2.3 2.5 0.4 0.4 0.8 0.5
Mexico 2.0 2.3 3.0 0.0 0.0 0.4 0.7
Middle East, North Africa, Afghanistan, and Pakistan 2.6 3.4 3.7 –0.2 0.2 –0.1 0.2
Saudi Arabia –0.7 1.7 1.9 0.1 –0.3 0.6 0.3
Sub-Saharan Africa 2.8 3.4 3.7 0.1 0.2 0.0 0.3
Nigeria 0.8 2.1 1.9 0.0 0.0 0.2 0.2
South Africa 1.3 1.5 1.7 0.6 0.8 0.4 0.1
Memorandum
European Union 2.7 2.5 2.1 0.2 0.0 0.4 0.3
Low-Income Developing Countries 4.7 5.0 5.3 –0.2 0.0 –0.2 0.1
Middle East and North Africa 2.2 3.2 3.6 –0.2 0.3 0.0 0.4
World Growth Based on Market Exchange Rates 3.2 3.4 3.3 0.1 0.1 0.3 0.3
World Trade Volume (goods and services) 4.9 5.1 4.7 0.5 0.3 1.1 0.8
Imports
Advanced Economies 4.0 5.1 4.5 0.7 0.0 1.3 0.9
Exports
Advanced Economies 4.2 4.5 3.9 0.3 –0.1 0.9 0.5
Commodity Prices (US dollars)
Oil5 23.3 18.0 –6.5 6.3 –2.2 18.2 –7.2
Nonfuel (average based on world commodity export
weights) 6.8 5.6 0.5 6.1 –0.5 5.1 1.0
Consumer Prices
Advanced Economies 1.7 2.0 1.9 0.1 –0.2 0.3 –0.1
Emerging Market and Developing Economies6 4.0 4.6 4.3 0.1 0.0 0.2 0.2
London Interbank Offered Rate (percent)
On US Dollar Deposits (six month) 1.5 2.4 3.4 0.1 0.0 0.5 0.5
On Euro Deposits (three month) –0.3 –0.3 0.0 0.0 0.1 0.0 0.0
On Japanese Yen Deposits (six month) 0.0 0.0 0.1 0.0 0.0 –0.2 –0.1
Note: Real effective exchange rates are assumed to remain constant at the levels prevailing during January 26–February 23, 2018. Economies are listed on the
basis of economic size. The aggregated quarterly data are seasonally adjusted.
1Difference based on rounded figures for the current, January 2018 World Economic Outlook Update, and October 2017 World Economic Outlook forecasts.
2Excludes the Group of Seven (Canada, France, Germany, Italy, Japan, United Kingdom, United States) and euro area countries.
3For India, data and forecasts are presented on a fiscal year basis and GDP from 2011 onward is based on GDP at market prices with fiscal year 2011/12 as a
base year.
4Indonesia, Malaysia, Philippines, Thailand, Vietnam.

Table 1.1 (continued)
Year over Year Q4 over Q47
Projections Projections
2016 2017 2018 2019 2016 2017 2018 2019
World Output 3.2 3.8 3.9 3.9 3.2 4.0 3.9 3.8
Advanced Economies 1.7 2.3 2.5 2.2 2.0 2.6 2.4 2.0
United States 1.5 2.3 2.9 2.7 1.8 2.6 3.0 2.3
Euro Area 1.8 2.3 2.4 2.0 2.0 2.7 2.2 2.0
Germany 1.9 2.5 2.5 2.0 1.9 2.9 2.5 1.9
France 1.2 1.8 2.1 2.0 1.2 2.5 1.8 2.0
Italy 0.9 1.5 1.5 1.1 1.1 1.6 1.3 1.1
Spain 3.3 3.1 2.8 2.2 3.0 3.1 2.5 2.1
Japan 0.9 1.7 1.2 0.9 1.5 2.1 0.8 –0.1
United Kingdom 1.9 1.8 1.6 1.5 2.0 1.4 1.6 1.6
Canada 1.4 3.0 2.1 2.0 2.0 2.9 2.1 1.9
Other Advanced Economies2 2.3 2.7 2.7 2.6 2.5 2.9 2.7 2.8
Emerging Market and Developing Economies 4.4 4.8 4.9 5.1 4.3 5.2 5.2 5.2
Commonwealth of Independent States 0.4 2.1 2.2 2.1 0.8 1.9 2.3 1.6
Russia –0.2 1.5 1.7 1.5 0.6 1.5 2.1 1.3
Excluding Russia 1.9 3.6 3.5 3.6 ... ... ... ...
Emerging and Developing Asia 6.5 6.5 6.5 6.6 6.2 6.7 6.5 6.6
China 6.7 6.9 6.6 6.4 6.8 6.8 6.5 6.4
India3 7.1 6.7 7.4 7.8 6.0 7.5 7.4 7.8
ASEAN-54 5.0 5.3 5.3 5.4 4.8 5.4 5.4 5.5
Emerging and Developing Europe 3.2 5.8 4.3 3.7 3.7 5.9 3.5 3.7
Latin America and the Caribbean –0.6 1.3 2.0 2.8 –0.8 1.7 2.3 2.4
Brazil –3.5 1.0 2.3 2.5 –2.4 2.2 3.1 2.3
Mexico 2.9 2.0 2.3 3.0 3.2 1.5 3.0 2.8
Middle East, North Africa, Afghanistan, and Pakistan 4.9 2.6 3.4 3.7 ... ... ... ...
Saudi Arabia 1.7 –0.7 1.7 1.9 2.2 –1.2 2.3 2.1
Sub-Saharan Africa 1.4 2.8 3.4 3.7 ... ... ... ...
Nigeria –1.6 0.8 2.1 1.9 ... ... ... ...
South Africa 0.6 1.3 1.5 1.7 1.0 1.9 0.7 2.3
Memorandum
European Union 2.0 2.7 2.5 2.1 2.1 2.9 2.3 2.0
Low-Income Developing Countries 3.5 4.7 5.0 5.3 ... ... ... ...
Middle East and North Africa 4.9 2.2 3.2 3.6 ... ... ... ...
World Growth Based on Market Exchange Rates 2.5 3.2 3.4 3.3 2.6 3.4 3.3 3.0
World Trade Volume (goods and services) 2.3 4.9 5.1 4.7 ... ... ... ...
Imports
Advanced Economies 2.7 4.0 5.1 4.5 ... ... ... ...
Emerging Market and Developing Economies 1.8 6.4 6.0 5.6 ... ... ... ...
Exports
Advanced Economies 2.0 4.2 4.5 3.9 ... ... ... ...
Emerging Market and Developing Economies 2.6 6.4 5.1 5.3 ... ... ... ...
Commodity Prices (US dollars)
Oil5 –15.7 23.3 18.0 –6.5 16.2 19.6 3.2 –5.9
Nonfuel (average based on world commodity export
weights) –1.5 6.8 5.6 0.5 10.3 1.9 7.0 0.3
Consumer Prices
Emerging Market and Developing Economies6 4.3 4.0 4.6 4.3 3.6 3.6 3.9 3.9
London Interbank Offered Rate (percent)
On US Dollar Deposits (six month) 1.1 1.5 2.4 3.4 ... ... ... ...
On Euro Deposits (three month) –0.3 –0.3 –0.3 0.0 ... ... ... ...
On Japanese Yen Deposits (six month) 0.0 0.0 0.0 0.1 ... ... ... ...
5Simple average of prices of UK Brent, Dubai Fateh, and West Texas Intermediate crude oil. The average price of oil in US dollars a barrel was $52.81 in
2017; the assumed price based on futures markets is $62.30 in 2018 and $58.20 in 2019.
6Excludes Argentina and Venezuela. See country-specific notes for Argentina and Venezuela in the “Country Notes” section of the Statistical Appendix.
7For World Output, the quarterly estimates and projections account for approximately 90 percent of annual world output at purchasing-power-parity weights.
For Emerging Market and Developing Economies, the quarterly estimates and projections account for approximately 80 percent of annual emerging market
and developing economies’ output at purchasing-power-parity weights.
slowdown is entirely because of advanced economies, few years from 2022 onward, offsetting some of the
where growth is projected to moderate in line with earlier growth gains.
their modest potential growth; growth across emerg- The above-trend growth rates of the euro area and
ing market and developing economies is expected to Japan—important contributors to the long-awaited
stabilize close to the current level. strengthening of economic activity in advanced
economies—are expected to continue during 2018–19.
Advanced Economies The recovery in the euro area is projected to pick up
Advanced economies are projected to grow at slightly from 2.3 percent in 2017 to 2.4 percent this
2.5 percent in 2018—0.2 percentage point higher than year, before moderating to 2 percent in 2019. The
in 2017—and 2.2 percent in 2019. For both years, forecast is higher than in the October WEO by 0.5
this forecast is considerably stronger than the October and 0.3 percentage point for 2018 and 2019, respec-
WEO forecast (0.5 and 0.4 percentage point higher tively, reflecting stronger-than-expected domestic
for 2018 and 2019, respectively). Positive revisions are demand across the currency area, supportive mone-
broad based, reflecting stronger prospects for the euro tary policy, and improved external demand prospects.
area and Japan and especially the projected domestic Medium-term growth in the euro area is projected at
and spillover effects of expansionary fiscal policy in 1.4 percent, held back by low productivity amid weak
the United States. Growth is projected to decline to reform efforts and unfavorable demographics. Japan’s
1.5 percent over the medium term, broadly in line growth is projected to moderate to 1.2 percent in
with modest potential growth. The reversal of some of 2018 (from a strong above‑trend outturn of 1.7 per-
the positive short-term output effects of US tax reform cent in 2017) before slowing further to 0.9 percent
beyond 2020 contributes to this decline.4 Despite this in 2019. The upward revision of 0.5 percentage point
slowdown, GDP is projected to remain above potential in 2018 and 0.1 percentage point in 2019 relative to
in 2023 in many advanced economies, including the the October WEO reflects more favorable external
United States and the euro area.5 demand prospects, rising private investment, and the
In the United States, growth is expected to rise from supplementary budget for 2018. Japan’s medium-term
2.3 percent in 2017 to 2.9 percent in 2018, before prospects, however, remain weak, owing largely to a
moderating slightly to 2.7 percent in 2019 (0.6 and shrinking labor force.
0.8 percentage point stronger than projected for 2018
and 2019, respectively, in the October WEO). The Emerging Market and Developing Economies
upward revision reflects stronger-than-expected activity Growth in emerging market and developing econ-
in 2017, firmer external demand, and the expected omies is expected to increase further—from 4.8 per-
macroeconomic impact of the December 2017 tax cent in 2017 to 4.9 percent in 2018 and 5.1 percent
reform—particularly lower corporate tax rates and the in 2019 (Table 1.1). Although the high growth rate
temporary allowance for full expensing of investment, reflects primarily continued strong economic per-
which is anticipated to stimulate short-term activity. formance in emerging Asia, the projected pickup in
The revision also reflects higher public spending fol- growth reflects improved prospects for commodity
lowing the February 2018 bipartisan budget agree- exporters after three years of very weak economic
ment. Fiscal policy changes are projected to add to activity. Growth forecast revisions were positive for
growth through 2020, so that US real GDP is 1.2 per- 2019: 0.1 percentage point for the aggregate, with
cent higher by 2020 than in a projection without the the largest positive revisions for emerging Europe and
tax policy changes. Given the increased fiscal deficit, Latin America. Beyond 2019, growth in emerging
which will require adjustment down the road, and market and developing economies is projected to
the temporary nature of some provisions, growth is stabilize at about 5 percent over the medium term.
expected to be lower than in previous forecasts for a This reflects some modest further strengthening in
economic growth in commodity exporters, though
to rates much more modest than over the past two
4The
temporary full expensing of investment implies more decades; a steady decline in China’s growth rate to a
investment up front, but less investment down the road; see level that is still well above the emerging market and
Box 1.5.
5Box 1.7 discusses in more detail the outlook for individual developing economy average; a gradual increase in
advanced economies. India’s growth rate as structural reforms raise poten-

tial output; and continued strong growth in other projected to increase 1.7 percent this year, before
commodity importers.6 moderating slightly to 1.5 percent through the rest of
Emerging Asia, which is forecast to continue grow- the projection horizon, weighed down by structural
ing at about 6½ percent during 2018–19, remains the headwinds and the effect of sanctions on investment.
most important engine of global growth. In China, Growth in the Middle East, North Africa, Afghan-
growth is projected to soften slightly from 6.9 per- istan, and Pakistan region is also expected to pick up
cent in 2017 to 6.6 percent in 2018 and 6.4 percent in 2018 and 2019, but remains subdued at about 3½
in 2019. The forecast is higher (by 0.1 percentage percent. While stronger oil prices are helping a recov-
point in both 2018 and 2019) relative to the Octo- ery in domestic demand in oil exporters, including
ber WEO, reflecting an improved external demand Saudi Arabia, the fiscal adjustment that is still needed
outlook. Over the medium term, the economy is is projected to weigh on growth prospects.
projected to continue rebalancing away from invest- Growth in sub-Saharan Africa is also projected to
ment toward private consumption and from industry rise gradually during 2018–19 to 3.4 percent and
to services, but nonfinancial debt is expected to con- 3.7 percent, respectively, as the challenging outlook in
tinue rising as a share of GDP, and the accumulation commodity exporters gradually improves. Growth in
of vulnerabilities clouds the medium-term outlook. South Africa is expected to strengthen from 1.3 percent
Growth in India is projected to increase from 6.7 per- in 2017 to 1.5 percent in 2018 and 1.7 percent in
cent in 2017 to 7.4 percent in 2018 and 7.8 percent 2019 (stronger than in the October WEO by 0.4 and
in 2019 (unchanged from the October WEO), lifted 0.1 percentage point, respectively, for 2018 and 2019).
by strong private consumption as well as fading tran- Business confidence is likely to gradually firm up with
sitory effects of the currency exchange initiative and the change in the political leadership, but growth pros-
implementation of the national goods and services pects remain weighed down by structural bottlenecks.
tax. Over the medium term, growth is expected to The medium-term outlook is subdued, with growth
gradually rise with continued implementation of expected to stabilize at 1.8 percent over 2020–23.
structural reforms that raise productivity and incentiv-
ize private investment.
Growth in emerging and developing Europe, now Inflation Outlook
estimated at close to 6 percent in 2017, is projected to With supply effects and stronger demand put-
moderate to 4.3 percent in 2018 and 3.7 percent in ting upward pressure on commodity prices—and a
2019, supported by a favorable external environment strengthening global outlook narrowing output gaps—
with easy financial conditions and stronger export headline inflation is picking up, and core inflation
demand from the euro area and, for Turkey, an accom- is expected to rise gradually as wage dynamics start
modative policy stance. reflecting tighter labor markets.7
A gradual growth recovery continues in Latin Amer- As shown in Table 1.1, headline inflation rates in
ica and the Caribbean, a region severely affected by the advanced economies are projected to pick up to about
2014–16 decline in commodity prices; growth is fore- 2 percent in 2018–19 (0.3 percentage point higher for
cast to rise to 2.0 percent in 2018 and 2.8 percent in 2018 than in the October WEO) from 1.7 percent
2019. Following a deep recession in 2015–16, Brazil ’s in 2017, mostly as above-trend growth and closing
economy returned to growth in 2017 (1.0 percent) output gaps add to price pressures.
and is expected to improve to 2.3 percent in 2018 Core consumer price inflation (CPI)—excluding
and 2.5 percent in 2019, buoyed by stronger private fuel and food prices—is expected to vary across
consumption and investment. Medium-term growth the advanced economy group. In the United States,
is set to moderate to 2.2 percent, weighed down by where output is set to rise above potential follow-
population aging and stagnant productivity. ing the expected sizable fiscal expansion, core CPI
Improved oil export revenue, stronger business is projected to increase from 1.8 percent in 2017 to
confidence, and looser monetary policy helped Russia’s 2 percent in 2018 and 2.5 percent in 2019, before
economy return to growth in 2017. Real GDP is declining to 2.3 percent over the medium term.
6Box 1.8 discusses in more detail the outlook for individual 7See Box 1.9 for details of the inflation outlook for individ-
emerging market and developing economies. ual countries.

Figure 1.16. Global Current Account Balance Excluding Venezuela (where inflation this year and
(Percent of world GDP)
next is expected to exceed 10,000 percent), headline
inflation in emerging market and developing econo-
Current account balances are expected to remain broadly at their 2017–18 levels
over the medium term. mies is expected to increase to 4.6 percent this year,
from 4.0 percent in 2017. The projection for 2018
4 is stronger by 0.2 percentage point relative to the
Afr. and ME Japan China
Eur. creditors Adv. Asia Oil exporters October WEO. In 2019 and beyond, inflation is
3 expected to moderate to about 4.0 percent as energy
prices stabilize and output gaps close. Compared
2
with advanced economies, there is considerable
1 diversity in inflation rates among emerging market
and developing economies, reflecting heterogeneity
0 in cyclical positions, central bank credibility, and
inflation targets.
–1
–2 External Sector Outlook

United States Other adv. Em. Asia Discrepancy
Current Account Positions
–3
Eur. debtors Lat. Am. CEE Current account balances in 2017, on the whole,
–4 have remained broadly stable compared with their
2002 04 06 08 10 12 14 16 18 20 22 23
2016 levels (Figure 1.16). The most notable change
Source: IMF staff estimates. has been an improvement in the current account
Note: Adv. Asia = advanced Asia (Hong Kong SAR, Korea, Singapore, Taiwan balance of oil exporters (close to 3 percent of
Province of China); Afr. and ME = Africa and the Middle East (Democratic Republic
of the Congo, Egypt, Ethiopia, Ghana, Jordan, Kenya, Lebanon, Morocco, South
their GDP), reflecting a partial recovery in their
Africa, Sudan, Tanzania, Tunisia); CEE = central and eastern Europe (Belarus, export prices.
Bulgaria, Croatia, Czech Republic, Hungary, Poland, Romania, Slovak Republic, Forecasts for 2018 and 2019 indicate some further
Turkey, Ukraine); Em. Asia = emerging Asia (India, Indonesia, Pakistan,
Philippines, Thailand, Vietnam); Eur. creditors = European creditors (Austria, improvement in the current account balances of oil
Belgium, Denmark, Finland, Germany, Luxembourg, Netherlands, Norway, exporters (as average oil prices are projected to exceed
Sweden, Switzerland); Eur. debtors = European debtors (Cyprus, Greece, Ireland,
Italy, Portugal, Spain, Slovenia); Lat. Am. = Latin America (Argentina, Brazil, Chile, those in 2017), as well as a widening of the US current
Colombia, Mexico, Peru, Uruguay); Oil exporters = Algeria, Azerbaijan, Iran, account deficit, driven by expansionary fiscal policy
Kazakhstan, Kuwait, Nigeria, Oman, Qatar, Russia, Saudi Arabia, United Arab
Emirates, Venezuela; Other adv. = other advanced economies (Australia, Canada, (partially offset by stronger external demand). Over the
France, Iceland, New Zealand, United Kingdom). medium term, current account balances are projected
to remain broadly stable at their 2017–18 levels, with
some narrowing of the US current account deficit as
The Federal Reserve’s preferred measure—core per- the expansionary effects of fiscal policy fade, mirrored
sonal consumption expenditure price inflation—is by some narrowing of surpluses in China and to a
projected to increase to 1.7 percent in 2018 and lesser extent in Europe.
2.2 percent in 2019 (from 1.5 percent in 2017). In As highlighted in the IMF’s 2017 External Sector
the euro area, with growth projected at above-trend Report, current account imbalances in 2016 were too
rates over 2018–19, core CPI is expected to increase large in relation to country-specific norms consistent
from 1.1 percent in 2017 to 1.2 percent in 2018 and with underlying fundamentals and desirable policies.
1.7 percent in 2019. Core CPI is projected to grad- As shown in the first panel of Figure 1.17, current
ually increase to 2 percent by 2021 as output gaps account balances in 2017 moved in a direction con-
narrow across the currency area and inflation expec- sistent with some reduction in those excess imbal-
tations strengthen. In the United Kingdom, core CPI ances, with medium-term current account projections
is expected to increase from 2.4 percent in 2017 to suggesting a further reduction. However, the projected
2.5 percent this year, before moderating to 2.2 percent changes in current account balances for some of the
in 2019 (and further to 2 percent over the medium world’s largest economies suggest only a modest nar-
term) as interest rate hikes and the withdrawal of mon- rowing of imbalances (for example, Germany) or some
etary support proceeds. widening (for example, the United States).

Figure 1.17. Real Exchange Rates and Current Account Figure 1.18. Net International Investment Position
Balances in Relation to Economic Fundamentals
Creditor and debtor net international investment positions are projected to widen
In 2017, current account balances moved modestly in directions consistent with slightly over the medium term.
reducing 2016 excess imbalances. Relative to 2016, real effective exchange rates
have also moved slightly in a direction consistent with narrowing 2016 exchange
rate gaps. 1. Global International Investment Position
(Percent of world GDP)
40
Afr. and ME Japan China
1. 2016 Current Account Gaps and Change in Current Account 30 Eur. creditors Adv. Asia Oil exporters
7 Balances, 2016–17
20
6 SAU
Change in current-account-to-
10
5 BRA
GDP ratio, 2016–17
4 AUS MEX JPN 0

GBR RUS CHE
3 EA –10
ZAF
2 POL NLD –20
1 CAN United States Other adv. Em. Asia
ITA MYS SGP –30
0 Eur. debtors Lat. Am. CEE
BEL IDN –40
–1 USA DEU
FRA HKG SWE THA 2005 07 09 11 13 15 17 19 21 23
–2 ESP IND CHN
TUR KOR
–3
–8 –6 –4 –2 0 2 4 6 8 10 2. Current Account Balance and Change in Net IIP, 2017
Current account gap, 2016 8 (Percent of GDP)
Eur. creditors
6
Change in IIP, 2016–17

United States
4
30 2. Real Effective Exchange Rate Changes and Gaps Japan
(Percent) 2 Em. Asia
25
REER gap 0 Lat. Am.
20 Eur. debtors
REER change, 2016–Feb 2018 –2 China
15 Afr. and ME CEE
–4
10 Other adv. Oil exporters
–6
5 Adv. Asia
–8
0
–10
–5 –6 –4 –2 0 2 4 6 8 10
–10 Current account balance, 2017
–15
–20 40 3. Net IIP, 2017, and Projected Changes, 2017–23
SAU TUR ITA ESP CAN RUS AUS CHN EA POL SWE THA MEX SGP
Projected change in IIP, 2017–23
(Percent of GDP)
USA FRA GBR BEL BRA ZAF IND IDN CHE JPN MYS NLD KOR DEU 30
Eur. debtors
Source: IMF staff calculations. 20 Japan
Adv. Asia
Note: Data labels use International Organization for Standardization (ISO) country
codes. EA = euro area; REER = real effective exchange rate. 10 Eur. creditors
CEE Em. Asia
0 Lat. Am.
Afr. and ME China Oil exporters
–10 United States Other adv.
Panel 2 of Figure 1.17 shows changes in real
effective exchange rates between 2016 and their levels –20
–100 –75 –50 –25 0 25 50 75 100 125
as of February 2018, together with the exchange Net IIP, 2017
rate gaps for 2016 identified in the 2017 External
Sector Report. Real effective exchange rates have also, Source: IMF staff estimates.
on average, moved modestly in a direction consis- Note: Adv. Asia = advanced Asia (Hong Kong SAR, Korea, Singapore, Taiwan
Province of China); Afr. and ME = Africa and the Middle East (Democratic Republic
tent with a narrowing of the 2016 gaps. Of course, of the Congo, Egypt, Ethiopia, Ghana, Jordan, Kenya, Lebanon, Morocco, South
changes in macroeconomic fundamentals since 2016 Africa, Sudan, Tanzania, Tunisia); CEE = central and eastern Europe (Belarus,
Bulgaria, Croatia, Czech Republic, Hungary, Poland, Romania, Slovak Republic,
have affected not only real exchange rates and current Turkey, Ukraine); Em. Asia = emerging Asia (India, Indonesia, Pakistan,
account balances, but also their equilibrium value. Philippines, Thailand, Vietnam); Eur. creditors = European creditors (Austria,
Belgium, Denmark, Finland, Germany, Luxembourg, Netherlands, Norway,
An example is the strengthening of the terms of trade Sweden, Switzerland); Eur. debtors = European debtors (Cyprus, Greece, Ireland,
for most commodity exporters, which is reflected in Italy, Portugal, Spain, Slovenia); IIP = international investment
position; Lat. Am. = Latin America (Argentina, Brazil, Chile, Colombia, Mexico,
their real appreciations depicted in panel 2. The 2018 Peru, Uruguay); Oil exporters = Algeria, Azerbaijan, Iran, Kazakhstan, Kuwait,
External Sector Report will discuss how changes in Nigeria, Oman, Qatar, Russia, Saudi Arabia, United Arab Emirates, Venezuela;
fundamentals and desirable policies have affected the Other adv. = Other advanced economies (Australia, Canada, France, Iceland, New
Zealand, United Kingdom).

Figure 1.19. Growth for Creditors and Debtors assessment of excess current account imbalances and
(Percent) exchange rate gaps for 2017.
Growth in domestic demand was faster in creditor countries than in debtor International Investment Positions
countries in 2017, but the contribution of net external demand remained positive
in creditor countries and negative in debtor countries. Changes in international investment positions reflect
both net financial flows and valuation changes arising
Net external contribution to Domestic demand contribution to from fluctuations in exchange rates and asset prices. As
growth growth
panel 1 in Figure 1.18 shows, over the next five years,
Total
creditor and debtor positions as a share of world GDP
8 1. Growth for Creditors are projected to widen slightly. On the creditor side,
this widening is explained primarily by the growing
6
creditor positions of a group of European advanced
4 economies—a result of large projected current account
2
surpluses. On the debtor side, this reflects some
increase in the debtor position of the United States and
0 other advanced economies (a group including Canada,
–2 France, and the United Kingdom, among others), par-
tially offset by a decline in the debtor position of euro
–4
area debtor countries.8
2015–16
2017
2015–16
2017
2015–16
2017
2015–16
2017
2015–16
2017
2015–16
2017
Similar trends are highlighted in panel 3 of Fig-

All creditors Eur. China Japan Adv. Asia Oil exporters
ure 1.18, which shows the projected changes in net
creditors international investment positions in percent of
domestic GDP across countries and regions between
8 2. Growth for Debtors 2017 and 2023 (the last year of the WEO projection
horizon). The creditor positions in advanced European
6
economies and Japan are projected at or above 80 per-
4 cent of their GDP, while the debtor position of the
United States is projected to reach 50 percent of GDP.
2 One notable change is the reduction in net interna-
tional investment position liabilities of a group of euro
0
area debtor countries, including Italy and Spain, which
–2 are expected to fall by more than 20 percentage points
of their GDP.
2015–16
2017
2015–16
2017
2015–16
2017
2015–16
2017
2015–16
2017
2015–16
2017
2015–16
2017
Domestic and External Contributions

All United Eur. Other adv. Latin Em. Asia CEE
Debtors States debtors America to GDP Growth
Another way to look at the prospects for global
Note: Adv. Asia = advanced Asia (Hong Kong SAR, Korea, Singapore, Taiwan rebalancing is to examine the domestic and exter-
Province of China); CEE = central and eastern Europe (Belarus, Bulgaria, Croatia, nal contributions to GDP growth in creditor and
Czech Republic, Hungary, Poland, Romania, Slovak Republic, Turkey, Ukraine);
Em. Asia = emerging Asia (India, Indonesia, Pakistan, Philippines, Thailand, debtor countries.
Vietnam); Eur. creditors = European creditors (Austria, Belgium, Denmark, Finland, Growth in domestic demand was faster in credi-
Germany, Luxembourg, Netherlands, Norway, Sweden, Switzerland);
Eur. debtors = European debtors (Cyprus, Greece, Ireland, Italy, Portugal, Spain, tor countries than in debtor countries in 2017, as in
Slovenia); Latin America = Argentina, Brazil, Chile, Colombia, Mexico, Peru, previous years, primarily reflecting high growth in
Uruguay; Other adv. = other advanced economies (Australia, Canada, France,
Iceland, New Zealand, United Kingdom); Oil exporters = Algeria, Azerbaijan, Iran,
Kazakhstan, Kuwait, Nigeria, Oman, Qatar, Russia, Saudi Arabia, United Arab 8Valuation changes can affect the evolution of these positions.
Emirates, Venezuela.
For instance, between the end of 2016 and the end of 2017,
the US net international investment position improved despite
the US current account deficit, given the depreciation of the US
dollar over this period, which increased the domestic currency
value of foreign currency assets held by US residents (Fig-
ure 1.18, panel 2).

China (Figure 1.19). At the same time, the net external the period ahead. In turn, an acceleration in poten-
contribution to growth was again positive for creditors, tial output would expand the scope for demand to
driven this time by positive contributions from China, rise before it hits capacity constraints and generates
creditor Europe, and Japan. In contrast to the two pre- inflation pressure.
vious years, the net external contribution to growth in On the downside, the early February 2018 market
oil exporters was instead negative, reflecting subdued turbulence and the equity market correction in March
export volumes and a recovery in imports after two following the US tariff announcement on steel and alu-
years of severe contraction. Among debtor countries, minum and a range of Chinese products, as well as the
the net external contribution to growth was strong and announcement by China of retaliatory tariffs on imports
positive for euro area debtor countries but remained from the US, serve as a cautionary reminder that asset
slightly negative for the United States in 2017 and is prices can correct rapidly and trigger potentially disrup-
projected to become more negative in 2018 because of tive portfolio adjustments. Although volatility is slightly
expansionary fiscal policy. higher than the pre-February episode lows, and term
Predicted changes in global macroeconomic policies, premiums are not as tightly compressed as they were
together with their potential exchange rate repercus- in the fall, global financial conditions remain highly
sions, could lead flow imbalances to widen again— supportive. A more severe version of the early February
even further than currently anticipated (should, for episode—financial conditions tighten suddenly, trig-
instance, the dollar appreciate sharply on expectations gered, for instance, by a faster pickup in inflation in the
of faster tightening of US monetary policy).9 Stronger United States—remains a possibility. Depending on the
reliance on demand growth in some creditor countries, magnitude of the repricing and the extent to which vol-
especially those with policy space to support it, such atility is affected, this could temper the pickup in global
as Germany, would help facilitate domestic and global demand (Scenario Box 1). In this context, a worsening
rebalancing while sustaining world growth over the of trade tensions and the imposition of broader barriers
medium term. In the US economy, which is already to cross-border trade would not only take a direct toll
close to full employment, a medium‑term plan to on economic activity (as shown in Scenario Box 1 of the
reverse the rising ratio of public debt, accompanied by October 2016 WEO) but would also weaken confi-
fiscal measures to gradually boost domestic capacity dence, with further adverse repercussions.
along with demand, would help ensure more sustain- Beyond the next few quarters, risks to the growth
able growth dynamics while helping contain exter- outlook are skewed to the downside. Concerns include
nal imbalances. a possible buildup of financial vulnerabilities as financial
conditions remain easy; an erosion of support for global
economic integration that could spur an inward shift
Risks in policies; and a host of noneconomic risks, includ-
The balance of risks to the near-term forecasts ing geopolitical strains, political discord, and climate
remains two-sided and broadly balanced. The poten- shocks. The risks are interlinked: if one materializes, it
tial for upside growth surprises remains. Business could trigger the others. For example, a shift toward
and consumer confidence stayed strong through inward-looking policy approaches to cross‑border flows
mid-February, and high-frequency indicators suggest of goods, capital, and labor can add to geopolitical ten-
that growth is likely to maintain a solid pace in the sions and global risk aversion, and noneconomic shocks
months ahead. Expectations of stronger business can weigh on short-term economic activity and on
profitability could lead firms to expand their invest- confidence in the longer-term outlook, limiting appetite
ment and hiring plans, as slack in labor markets may for investment. The resulting negative impact on growth
be larger than currently assessed (Chapter 2 of the could be severe, considering that there would be less
October 2017 WEO). Furthermore, the ongoing room to cut interest rates or increase public spending to
recovery in investment could foster a rebound in combat downturns than in the past.
productivity, implying higher potential growth in
Financial Vulnerabilities
9The WEO assumes that real effective exchange rates remain
broadly stable at the level of the reference period (in this case, The recent bout of turbulence in financial mar-
February 2018). kets does not eliminate the possibility that financial

conditions will remain accommodative into the could undermine international payment systems and
medium term, with vulnerabilities building amid disrupt the flow of goods and services.
a search for yield. As discussed in the April 2018
GFSR, financial conditions are broadly unchanged
relative to the fall, even as the US Federal Reserve has Waning Support for Global Integration
raised the policy interest rate and continued to allow The Comprehensive and Progressive Agreement
a gradual contraction of its bond holdings. As noted for Trans-Pacific Partnership—covering 11 countries
in the October 2017 and April 2018 GFSR, inves- accounting for approximately 15 percent of global
tors have moved into riskier asset classes to coun- trade—and the announced signing of the agreement
teract the low returns of more traditional securities. to establish the Continental Free Trade Area, which
At the same time, the share of companies with low would cover all of Africa, represent encouraging prog-
investment-grade ratings in advanced economy bond ress on plurilateral trade integration. At the same time,
indices has increased significantly. Corporate debt support for globalization appears to have weakened
remains high in some emerging markets—in some in some advanced economies. Free-trade agreements
cases with a high reliance on funding sources outside such as NAFTA and the economic arrangements
traditional banking relationships. Tighter regula- between the United Kingdom and rest of the Euro-
tion of nonbank intermediation in China, where pean Union are being renegotiated. The United States
nonfinancial corporate sector debt is still rising, is a recently increased tariffs on imported solar panels and
welcome start of a needed policy response to contain washing machines, and announced tariff actions on
the accumulation of vulnerabilities. steel and aluminum and a range of Chinese products,
Credit risk may be contained while global growth while China announced retaliatory tariffs on imports
momentum is strong and borrowing rates are low, from the US. An increase in tariffs and nontariff trade
but it could come to the fore over the medium barriers could harm market sentiment, disrupt global
term, exposing financial fragility. An eventual global supply chains, and slow the spread of new technolo-
repricing of risk could be triggered by various shifts, gies, reducing global productivity and investment (Box
including a broad-based pickup in inflation. The 1.6 documents a rise in trade-restricting measures in
US economy operating above potential output amid G20 economies in recent years). Greater protectionism
temporary tax cuts could require faster-than-expected would also lower consumer welfare by making tradable
tightening of US monetary policy, which could lead consumer goods more expensive. Scenario analysis
to a rise in term premiums and debt service costs. (IMF 2016a, Box 1) indicates that rising protection-
Depending on its timing, the drag from such tighten- ism in all countries—leading to a 10 percent increase
ing of financial conditions could coincide with softer in import prices everywhere—lowers global output
US demand following the reversal of tax cuts, which and consumption by about 1 ¾ percent after 5 years
would amplify its negative international spillovers. and close to 2 percent in the long term, while global
Even as the health of banking systems continues investment and trade fall by even more. Moreover,
to improve, policies still have a key role to play in curbs on immigration would prevent aging societies
managing risks in both the bank and the nonbank from effectively counteracting trend declines in the
financial sectors. Against this backdrop, a broad labor force growth rates. Widening external imbalances
rollback of stronger financial regulation and over- in some countries, including the United States—where
sight since the global financial crisis—both nation- the current account deficit is poised to increase given
ally and internationally—could facilitate excessive the projected impact of fiscal stimulus on domes-
risk taking, with negative repercussions for global tic demand—could add to protectionist pressure.
financial stability. Increased trade tensions also make it more difficult
Finally, among emerging potential sources of for countries to deal cooperatively with international
financial tension, if the recent rapid growth of crypto disruptions or shocks.
assets is maintained and draws in larger institutional
investors, the linkages with the broader financial
system are likely to expand and may create new sources Noneconomic Factors
of financial stability risk. More broadly, cybersecurity The medium-term global outlook remains clouded
breaches and cyberattacks on financial architecture by geopolitical tensions (Figure 1.20), notably in east

Figure 1.20. Geopolitical Risk Index Figure 1.21. Risks to the Global Outlook
(Index)
The risks around the central growth forecast are two-sided and broadly even over
Geopolitical risks remain elevated. 2018–19.
6 1. Prospects for World GDP Growth1

600 (Percent change)
Iraq Russian 5
invasion actions
in Crimea Paris 4
500 London WEO baseline
attacks
bombing ISIS 3
Syrian 50 percent confidence interval
400 Madrid civil war escalation 70 percent confidence interval
2
9/11 bombing Iran nuclear escalation 90 percent confidence interval
tensions Arab Spring: 1 90 percent confidence interval from April 2017 WEO
300 Syrian and
Libyan wars 0
2015 16 17 18 19
200
1.5 2. Balance of Risks Associated with Selected Risk Factors2
(Coefficient of skewness expressed in units of the underlying
1.0 variables)
100
0.5
0 0.0
2000 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 Mar.
18 –0.5 Balance of risks for
Current year
–1.0 Next year
Source: Caldara and Iacoviello (2017).
Note: ISIS = Islamic State. –1.5
Term spread S&P 500 Inflation risk Oil market risks
Dispersion of Forecasts and Implied Volatility3

80 3. 1.2 125 4. 0.5
Asia and the Middle East. For many countries already GDP (right scale) Term spread
70 (right scale)
VIX (left scale) 1.0 100
severely affected by conflict or its spillovers, the central 60 Oil (left scale) 0.4
forecast assumes a gradual easing of strains; more pro- 50 0.8 75
tracted resolution of tensions would delay recovery in 40 0.3
these economies.10 30 0.6 50
Political uncertainty also gives rise to reform imple- 20 0.2
mentation risks or the possibility of reoriented policy 0.4 25
10
agendas, including in the context of upcoming elec-
0 0.2 0 0.1
tions or their immediate aftermath in several countries 2006 08 10 12 14 16 Feb. 2006 08 10 12 14 16 Feb.
18 18
(such as Brazil, Colombia, Italy, and Mexico). Weak
governance and large-scale corruption can also under- Sources: Bloomberg Finance L.P.; Chicago Board Options Exchange (CBOE);
Consensus Economics; Haver Analytics; and IMF staff estimates.
mine confidence and popular support for reforms, 1
The fan chart shows the uncertainty around the April 2018 World Economic
taking a toll on economic activity. Outlook (WEO) central forecast with 50, 70, and 90 percent confidence intervals. As
shown, the 70 percent confidence interval includes the 50 percent interval, and the
Finally, recent extreme weather developments point 90 percent confidence interval includes the 50 and 70 percent intervals. See
to the risk of recurrent severe climate events that Appendix 1.2 of the April 2009 WEO for details. The 90 percent intervals for the
current-year and one-year-ahead forecasts from the April 2017 WEO are shown.
impose devastating humanitarian costs and eco- 2
The bars depict the coefficient of skewness expressed in units of the underlying
nomic losses on the affected regions. They may also variables. The values for inflation risks and oil market risks enter with the opposite
add to migration flows that could destabilize recipi- sign since they represent downside risks to growth.
3
GDP measures the purchasing-power-parity-weighted average dispersion of GDP
ent countries. growth forecasts for the Group of Seven economies (Canada, France, Germany,
Italy, Japan, United Kingdom, United States), Brazil, China, India, and Mexico. VIX is
the CBOE Standard & Poor’s (S&P) 500 Implied Volatility Index. Term spread
measures the average dispersion of term spreads implicit in interest rate forecasts
10Recent research shows that higher geopolitical tensions for Germany, Japan, the United Kingdom, and the United States. Oil is the CBOE
crude oil volatility index. Forecasts are from Consensus Economics surveys.
can weigh on global activity. See, for instance, Caldara and
Dashed lines represent the average values from 2000 to the present.
Iacoviello (2017).

Figure 1.22. Recession and Deflation Risks views about its future behavior. The wider spread of
(Percent) oil price forecasts seems to reflect, in part, differing
views on the causes and likely persistence of the recent
For most regions, recession and deflation risks over a four-quarter horizon have
declined since last fall.
pickup in prices.
With stronger growth, the probability of a recession
70 1. Probability of Recession, 2018:Q2–2019:Q1 over a four‑quarter horizon (2018:Q2–2019:Q1) has
60 October 2017 WEO: 2017:Q4–2018:Q3 declined in most regions relative to the probability
computed in the October 2017 WEO (Figure 1.22).
50
At the same time, medium-term risks to growth
40
remain salient. As discussed in the April 2018 GFSR,
30 Growth-at-Risk analysis suggests that easy financial
20 conditions imply some upside risk to short-term
10 growth but pose risks to medium-term growth that are
0
well above historical norms. Deflation risks—as mea-
United Euro area Japan East Asia Latin Rest of the sured by the four-quarter-ahead probability of deflation
States America 5 world
in the second quarter of 2019, occurring together with
a negative output gap—have generally declined. In the
30 2. Probability of Deflation, 2019:Q21
euro area, the joint probability of four‑quarter head-
October 2017 WEO: 2018:Q4
25 line inflation turning negative in the second quarter of
20 2019 and a negative output gap in the same quarter,
which is just above 10 percent, has risen modestly
15
because of the base effect of a peak in oil prices in early
10 2018 and their subsequent decline.
5
0
United Euro area Japan East Asia Latin Rest of the
Policy Priorities
States America 5 world As discussed in the “Recent Developments and
Prospects” section, the current recovery is the broad-
Note: East Asia comprises China, Hong Kong SAR, India, Indonesia, Korea,
est synchronized upsurge in global activity in close
Malaysia, the Philippines, Singapore, Taiwan Province of China, and Thailand; to a decade. Domestic and multilateral policies have
Latin America 5 comprises Brazil, Chile, Colombia, Mexico, and Peru; Rest of the a vital role to play in ensuring that the momentum
world comprises Argentina, Australia, Bulgaria, Canada, Czech Republic, Denmark,
Israel, New Zealand, Norway, Russia, South Africa, Sweden, Switzerland, Turkey, is sustained, remaining output gaps close, and infla-
the United Kingdom, and Venezuela. October 2017 WEO data refer to simulations tion expectations are well anchored. The strength of
run in September 2017. WEO = World Economic Outlook.
1
Deflation risk is measured by the four-quarter-ahead probability of deflation short-term economic activity provides an opportunity
occurring together with a negative output gap. to start rebuilding fiscal buffers where needed and
allows for more policy focus on other medium- and
long-term priorities: boosting potential growth, reduc-
ing inequality, strengthening financial resilience, and
coping with climate change.
Fan Chart Analysis
A fan chart analysis—based on equity and commod-
ity market data, as well as the dispersion of inflation Policies—Advanced Economies
and term spread projections of private forecasters—
Monetary Policy: Divergence Warranted by
shows that uncertainty around the central growth
Differences in the State of the Cycle
forecast is broadly even, but wider than a year ago
(Figure 1.21). The increase is chiefly due to greater dis- The upswing in activity across advanced economies
persion of views about future inflation and oil prices. has lifted job creation, lowered unemployment rates,
Continued subdued inflation despite stronger demand and narrowed output gaps. In most advanced econo-
appears to have contributed to a divergence in analysts’ mies, however, nominal wage growth and core inflation

remain subdued, and market expectations of future debt ratios remain sustainable, rebuild fiscal policy
inflation point to a slow convergence path back to space to counter future downturns, and leave these
central bank targets. Set against the backdrop of many economies better positioned to address long-term fiscal
years of subpar growth and low inflation, macroeco- challenges stemming from aging-related health and
nomic conditions in advanced economies generally call pension outlays. The pace of consolidation should be
for continued monetary accommodation.11 However, calibrated to the strength of the recovery and avoid
if output is close to potential and inflation is rising sharp drags on growth.
toward target, a gradual withdrawal of monetary sup- Countries with fiscal space should raise potential
port is warranted. output and productivity by enhancing workforce skills,
Continued monetary support is needed in the including in the area of digital literacy. These countries
euro area and Japan until inflation durably increases should improve infrastructure where needed and—
toward central bank targets. The unemployment where aging is expected to exert a significant drain on
rate in the United Kingdom is close to historic lows; labor supply—should boost labor force participation
further declines could add to inflation pressure by through stronger family-friendly policies, reconsid-
triggering faster wage growth in a context of infla- eration of labor taxation, actuarially fair pension
tion that is already above target following currency systems, and labor market matching enhanced by more
depreciation after the June 2016 Brexit referendum. efficient active labor market programs (as discussed
Gradual monetary tightening is therefore needed to in Chapter 2).
ensure that inflation returns to target and expecta- In the euro area, several countries have exhausted
tions remain anchored. Similarly, unemployment their fiscal space and should gradually consolidate in as
rates in the United States have, over the past year, growth-friendly and evenly phased a manner as possi-
approached lows last seen in the 1990s, and there ble to rebuild buffers. In Italy and Spain, for example,
are nascent signs of a pickup in wages. With the high sovereign debt ratios together with unfavorable
economy already likely at potential, the December demographic trends call for an improvement in the
2017 tax code overhaul and the February 2018 structural primary balance to put debt firmly on a
budget agreement could significantly stimulate downward path. By contrast, Germany has fiscal space
activity and stoke wage and price pressures—in that should be used to increase public investment
which case a faster withdrawal of monetary support in areas that will lift potential growth by improving
may be needed. Overall, this highlights the need for productivity and increasing the labor force participa-
data-dependent monetary policy normalization and tion of women and recent immigrants. These areas
the continued crucial role of communications in include enhancing digital infrastructure, child care and
ensuring a smooth adjustment. after-school programs, and the training and integra-
tion of refugees into the workforce. An important
Fiscal Policy: Rebuild Buffers and Focus on by-product of more public investment in Germany
Medium-Term Objectives would be higher imports from the rest of the euro area,
The cyclical recovery affords an opportunity to which would facilitate rebalancing of demand within
orient fiscal policy more firmly toward medium-term the common currency area.
goals (see also Chapter 1 of the April 2018 Fiscal In Japan, a premature drop in the level of fiscal
Monitor). In countries with little fiscal space, where support should be avoided so as to sustain growth and
a gradual strengthening of fiscal buffers is warranted, promote structural reforms. The debt trajectory needs
consolidation should proceed hand-in-hand with a to be anchored by a credible medium-term fiscal con-
shift in budget composition toward areas that lift solidation plan, which should include a streamlining of
potential output growth, while also remaining mindful health, pension, and long-term care benefits together
of reducing inequality and improving the welfare of with gradual and steady increases in the consumption
the most vulnerable. Doing so would help sovereign tax rate starting in 2019.
The recently legislated tax code overhaul and biparti-
11As discussed in Chapter 2 of the October 2017 WEO there
san agreement on the federal budget in the United
may be greater slack in labor markets than is captured by headline States will further add to rising fiscal deficits and
unemployment rates. unsustainable debt dynamics over the next five years. It

is therefore imperative to ensure higher future revenues and protect consumers remain essential for preserving
and take measures to gradually curb the dynamics financial stability.
of public spending while shifting its composition
toward much-needed improvements in infrastructure, Structural Policies: Boost Potential Growth and
poverty-alleviating measures, and policies to strengthen Ensure that Benefits Are Shared Widely
labor force participation. Once output gaps close and advanced economies
complete their cyclical recovery, the pace of expansion
is set to moderate toward subdued potential growth
Financial Sector Policies: Complete Balance Sheet over the medium term. Rising inequality and income
Cleanup, Increase Resilience to Shocks polarization also threaten medium-term growth pros-
As discussed in the “Risks” section, a range of pects by fueling support for inward-looking policies
triggers could ignite financial tensions in global and could harm health and education outcomes among
markets and undermine global growth prospects. In the affected groups.
advanced economies, the postcrisis financial regula- In the United States, policy measures that can help
tory reform and balance sheet cleanup has improved lift potential output growth include public investment
institution‑specific and system-wide resilience in to augment infrastructure and maintain the quality of
financial sectors, but a few pockets of weakness remain. the existing stock; improvements in the efficiency of
Fortifying these segments and, more broadly, avoiding education spending; and more support for vocational
a rollback of the regulatory reforms are essential for apprenticeships, reskilling, and lifelong learning pro-
containing financial vulnerabilities. grams. According to the US Congressional Joint Com-
In the euro area, continued progress on reducing mittee on Taxation, the tax code overhaul is projected
nonperforming loans is essential for shedding crisis to reduce the average tax rate on upper-income US
legacies and lifting an important constraint on credit households relative to those in the middle and lower
intermediation (notably in Greece, Italy, and Portu- segments, especially over the medium term (when
gal). More generally, there is a need to improve banks’ some provisions benefiting lower- and middle-income
cost efficiency and profitability, which will require taxpayers expire), thus increasing income polariza-
proactive supervision and consolidation in over- tion.12 Measures that can raise labor force participa-
banked economies. Appropriate and predictable use tion and arrest income polarization include a larger
of creditor bail-ins and precautionary recapitalizations Earned Income Tax Credit, expanded child tax credits,
will be vital for reducing uncertainty and counter- means‑tested tax relief for lower-income working fam-
party risk in situations of financial stress as well as ilies for childcare-related expenses, and reform of the
for limiting the burden placed on taxpayers. For the disability insurance program to encourage part-time
whole currency area, completing the banking union work over disengagement from the labor force.
remains a priority for placing the financial system on Relatively low total factor productivity growth
a stronger footing. and a trend decline in the labor force are key factors
In Japan, the prolonged low-interest-rate environ- weighing on potential output growth in Japan. Raising
ment and demographic headwinds have gradually productivity will require reforming the labor market
weakened the profitability of financial institutions, to increase efficiency (for instance, with contracts
particularly among regional banks. Increasing fee-based that strike a better balance between job security and
income and diversifying revenue sources, together with flexibility while promoting worker mobility across
consolidation and rationalization, should help boost firms); lowering entry barriers to draw in more private
profitability. investment (for example, in telecommunications
In the United States, recent simplifications of and professional services); and furthering corporate
regulations on medium-sized banks are warranted governance reform. Offsetting the trend decline in the
and unlikely to increase systemic risk. Broad-based size of the labor force will require further increasing
deregulation that would loosen constraints on larger female and older worker labor force participation and
banks should, however, be avoided because it could allowing more use of foreign workers.
once again encourage excessive risk taking and leave
the financial system vulnerable to disruptive correc- 12Box 1.2 of the April 2018 Fiscal Monitor discusses the distribu-
tions. Continued efforts to improve financial literacy tional implications of the US tax overhaul.

Structural reform priorities to boost productivity targets and open up space for further reduction of the
and innovation and reduce competitiveness disparities tax burden. In Brazil, legislating social security reform
across the euro area vary, depending on country-specific remains a priority to ensure that spending is consistent
bottlenecks. For instance, Spain should try to further with the constitutional fiscal rule and to guarantee
reduce labor market duality and employment protec- long-term fiscal sustainability. Making use of the recent
tion gaps between permanent and temporary workers, strengthening of activity to improve the primary bal-
and target training and active labor market policies ance over the short term would complement the over-
to boost employment prospects for young people and all consolidation strategy. In China, fiscal policy has
the long-term unemployed. In Italy, reforming wage played a vital part in shoring up short-term growth at
bargaining arrangements to allow more firm-level the expense of eroding valuable policy space. Gradual
flexibility should help align wages with productivity. In consolidation, together with a shift of spending back
Germany, deregulating services would foster more com- onto the budget and away from off‑budget channels,
petition and efficiency gains, and expanding the avail- would help improve sustainability. India’s high public
ability of venture capital could promote innovation. debt and recent failure to achieve the budget’s defi-
cit target call for continued fiscal consolidation into
the medium term to further strengthen fiscal policy
Policies—Emerging Market Economies credibility.
Policy priorities in emerging market economies
differ across countries within the group, depending on Strengthening Financial Resilience
their cyclical positions and country-specific vulnera- Balance sheet vulnerabilities pose a downside risk
bilities. Common objectives across the group include to medium-term growth prospects in many emerging
strengthening financial resilience so that income gaps market economies, requiring policy action. The corpo-
relative to advanced economies can continue to narrow rate debt overhang and associated banking sector credit
sustainably and ensuring that opportunities and bene- quality concerns exert a drag on investment in India.
fits associated with higher per capita income are shared The recapitalization plan for major public sector banks
broadly across the population. announced in 2017 will help replenish capital buffers
and improve the banking sector’s ability to support
Cyclical Policies: Manage Trade-Offs growth. However, recapitalization should be part of a
In several emerging market economies, inflation is broader package of financial reforms to improve the
relatively subdued compared with historical averages. governance of public sector banks, and banks’ debt
Improvements to monetary policy frameworks also recovery mechanisms should be further enhanced. In
appear to have lowered inflation expectations, includ- Turkey, limiting balance sheet currency mismatches and
ing in Brazil and India. These developments have the high exposure to foreign exchange risk are urgent
created room for monetary policy to support activity priorities, especially with monetary policy normaliza-
should downside risks to growth materialize. How- tion under way in the United States and the United
ever, in a few countries, such as Argentina and Turkey, Kingdom (and the resulting possibility of a shift of
inflation remains above central bank targets, requiring capital flows away from emerging market economies).
a tight monetary stance to keep expectations anchored. Moreover, given that sudden repricing of term premi-
Fiscal policy is generally more constrained by the ums remains a distinct possibility (as discussed in the
need to strengthen buffers and ensure sustainabil- “Risks” section) and that portfolio shifts could occur,
ity of social insurance programs—particularly in it is important to mitigate rollover risk by avoiding
commodity-exporting emerging market economies excessive reliance on short-term borrowing. Regulators
faced with subdued medium-term prospects for com- in China have taken important measures to rein in
modity prices, but also more broadly. shadow banking and bring financial activity back onto
In Argentina, fiscal reforms approved at the end of bank balance sheets, where capital and provisioning
2017 provide improved guidance on fiscal discipline requirements provide greater loss absorption capacity
and will help address the country’s large pension than in opaque off-balance-sheet channels. Neverthe-
imbalances and begin a gradual reduction of high and less, total credit growth remains high. Early recognition
distortionary taxes. However, further cuts to primary of nonperforming assets, a reduction of forbearance,
spending will be needed to achieve the primary deficit and gradually unwinding of the system of implicit guar-

antees to better align borrowing costs with risk‑adjusted main priorities for lifting constraints on job creation
returns remain essential for improving credit allocation and ensuring that the demographic dividend is not
and containing the accumulation of vulnerabilities. wasted are to ease labor market rigidities, reduce
More broadly across emerging market economies, infrastructure bottlenecks, and improve educa-
as shown in Chapter 2 of the October 2017 WEO, tional outcomes.
medium-term growth outcomes are improved by In Brazil, reducing tariff and nontariff barriers to
avoiding credit booms that lead to excessive risk taking trade will help improve efficiency and raise productiv-
and by permitting exchange rate flexibility to minimize ity growth, and enhancing the appeal of the infrastruc-
the distortion of relative price signals and associated ture concessions program to investors would help draw
resource misallocation. in private investment and fill important infrastructure
shortfalls. In Mexico, implementation of certain aspects
Boosting Potential Output Growth and Enhancing of the wide-ranging reform agenda approved five
Inclusiveness years ago has progressed well, including in the energy,
Strong growth over long stretches of the post-2000 financial, and telecommunications sectors. Building on
period has allowed several emerging market economies these areas, priority should be accorded to initiatives
to narrow income gaps relative to advanced economies that will lift key constraints on investment and boost
and has enabled millions in these countries to climb growth over the medium term, including implement-
out of poverty. As discussed earlier, the medium-term ing judicial reforms that target corruption and promote
outlook for many emerging market economies is the rule of law, as well as labor market reforms that
relatively subdued compared with the growth rates help reduce informality.
achieved since 2000. Country-specific constraints are, More generally across emerging markets, there
in many cases, important contributing factors that is room to make growth more inclusive and reduce
weigh on medium-term growth, limit employment inequality by increasing the coverage of personal
opportunities for the working-age population, and income taxes, lowering the burden of indirect taxes,
prevent the benefits of growth from spreading widely. and increasing the share of transfers to the lowest
In South Africa, the election of new political leader- income groups through improved targeting. Condi-
ship reduces some of the policy uncertainty. However, tional cash transfers—adopted for example in Brazil
advancement of the outstanding reforms is critical for and Mexico—that are linked to school enrollment or
reinvigorating economic growth and making it more attendance at health clinics can lower current inequal-
inclusive. Improving infrastructure; reducing barri- ity and, by improving education and health outcomes,
ers to entry in key sectors, including transportation future income inequality (see the October 2017
and telecommunications; improving the efficiency of Fiscal Monitor).
government spending; and reducing policy uncertainty
remain central to attracting private investment, raising
productivity across the economy, and promoting job Policies—Low-Income Developing Countries
creation. The proposal to introduce a national min- Low-income developing countries face multiple
imum wage has the potential to hurt firms’ compet- challenges in their effort to progress toward their 2030
itiveness and employment prospects in the formal Sustainable Development Goals. Fiscal positions have
sector, but it could improve working conditions and worsened across several countries in this group, poverty
reduce poverty. For a sustained rise in living standards and inequality remains high, and financial vulnerabili-
and inclusiveness, however, broad-based efforts are ties appear to be on the rise in some cases. Commodity
needed to raise the quality of education and improve exporters and those particularly exposed to natural disas-
access to opportunities for all segments of society. ters face additional complex challenges of diversifying
India has made progress on structural reforms their economies—a long‑standing goal that has acquired
in the recent past, including through the imple- renewed urgency with the subdued medium-term out-
mentation of the goods and services tax, which look for commodity prices and recurrent climate-related
will help reduce internal barriers to trade, increase events as global temperatures rise (Chapter 3 of the
efficiency, and improve tax compliance. While the October 2017 WEO). Many of the policy priorities dis-
medium-term growth outlook for India is strong, an cussed below are interlinked, are mutually reinforcing,
important challenge is to enhance inclusiveness. The and can achieve multiple objectives.

A Widespread Need to Strengthen Fiscal Positions vocational skills training, and improved infrastructure
Fiscal positions have deteriorated in recent years would help support new firm entry and boost opportu-
across most low-income countries—encompassing both nities for gainful employment of larger numbers.
commodity exporters (those countries that gener- Enhancing Financial Resilience
ate at least 50 percent of their export revenue from
commodities) and more diversified economies. While As discussed in IMF 2018a, some low-income
lower commodity prices since 2014 have dragged on countries (Mozambique, Nigeria) have experienced
revenue in commodity exporters, the broader pat- financial stress or deteriorating loan quality in recent
tern across low-income countries of worsening fiscal years as growth has moderated and corporate balance
positions suggests that domestic revenue mobilization sheets have weakened. In some countries—including
efforts have generally fallen short of rising expenditure Chad and Zambia—worsening fiscal positions have led
requirements. Current spending—including rising debt governments to build up arrears to private contractors
service costs—appears to have contributed more than and have made it difficult for them to stay current on
has public investment to the increase in total spending their loans. Further deterioration in loan quality would
(IMF 2018a). impair credit intermediation and the ability of the
Continued efforts to broaden the tax base, enhance banking sector to support growth in these countries and
compliance, and reduce wasteful, poorly targeted would raise the risk of costly recapitalization, which
subsidies would create essential resources for meeting would severely burden already-strained public finances.
critical social and developmental needs—including Proactive supervision, ensuring adequate provi-
in the areas of health, sanitation and water deliv- sioning for losses by banks, reducing forbearance,
ery, electricity generation, roads, and education and and improving resolution frameworks to minimize
training facilities. Fiscal consolidation efforts that expensive public bailouts are essential for strengthen-
focus on cutting current and recurrent expenditures ing financial resilience. Fiscal adjustments that place
generally appear to have smaller negative effects on public finances on a sustainable path would addition-
economic activity than an equivalent reduction in ally help curb budgetary arrears, allowing debt service
public investment (see the October 2017 Regional to proceed on schedule and curtailing the buildup of
Economic Outlook: Sub-Saharan Africa). Undertaking nonperforming loans.
these efforts now—while growth is on the mend and Furthermore, for economies that are not part of a
the ongoing increase in commodity prices offers some currency union, allowing exchange rate flexibility while
respite—would help prevent a more painful adjust- using reserves to smooth excess volatility can help buf-
ment farther down the road. fer external shocks and, over time, prevent sustained
departures from fundamental valuation (which lower
the overall efficiency of economic activity).
Promoting Inclusive Growth
As documented in the October 2017 Fiscal Monitor, Diversification and Coping with Climate Shocks
inequality has declined since 2000 across sub-Saharan Economic diversification away from excessive depen-
Africa, Asia, and Latin America—regions where several dence on commodities, or on a few sectors such as
low-income economies are situated. Nevertheless, it agriculture or tourism, is an overarching imperative for
remains high. Ensuring that poverty and inequality commodity exporters and those countries that are par-
continue to fall is imperative from a welfare perspective ticularly exposed to natural disasters. While there is no
and to secure support for needed structural reforms, unique template for all circumstances, general policy
avoid debilitating political conflict and civil strife, and attributes that facilitate diversification or help countries
make growth sustainable. cope with climate shocks include sound macro man-
Priority policy areas to foster inclusive growth agement and judicious use of policy buffers to smooth
include universal health coverage of essential services to fluctuations, investment in education and training to
reduce infant and maternal mortality, targeted efforts improve workforce skills, increased access to credit,
to improve the delivery and take-up of early childhood and a reduction in infrastructure gaps (see Chapter 3
education, initiatives to close primary and secondary of the October 2017 WEO and the October 2017
enrollment gaps, and greater availability of clean water Regional Economic Outlook for sub-Saharan Africa).
and sanitation. Increased access to credit, expanded More broadly, governance reforms—for instance,

strengthening incentives to improve the efficiency of financial regulatory reform agenda to make the
public administration, reducing the risk of expropria- financial system safer, including through stronger
tion, enhancing transparency in project selection, and bank capital buffers, a better bank asset liquidity
expediting business dispute resolutions according to profile, and more stable funding. Key remaining
established legal principles—would help lift private areas for action to complete the regulatory reform
investment, create jobs, and expand the range of activ- agenda and strengthen global financial stability
ity beyond primary, resource-based sectors. include devising effective resolution frameworks
for globally important financial institutions,
bolstering central counterparty clearing for
Multilateral Policies derivatives, and filling data gaps and enhancing
Sustaining global improvements in living stan- supervision and regulation of nonbank financial
dards and delivering greater economic security to institutions. Continued close cooperation is also
a rising share of the world’s population requires a needed on combating cross-border money launder-
well-functioning multilateral framework that can ing, financing of terrorism, and fortifying financial
facilitate a cooperative approach to addressing shared infrastructure against cybersecurity breaches. At
challenges and resolving disagreements. Multilateral the same time, regulators must ensure that corre-
cooperation in a range of areas can help amplify the spondent banking relationships—through which
benefits of the country-level actions discussed in the globally active banks provide deposit‑taking and
preceding section while minimizing any adverse spill- remittance services to smaller banks in low-income
overs they may generate. This is particularly relevant at countries—stay intact to ensure that these coun-
a time when unilateral tariff actions threaten to weaken tries have access to vital international payments.
the rules-based global trading system that has helped In addition, an adequately financed global safety
lift millions out of poverty and raised consumer welfare net remains critical for countries to have quick and
by lowering the price of tradable goods over the past predictable access to international liquidity if they
several decades. are unable to tap existing mechanisms, includ-
•• Trade: Trade openness and global economic integra- ing their own reserves, bilateral swap lines, and
tion under a rules-based, multilateral trading system regional financing agreements. Finally, as discussed
have been crucial for diffusing innovation, lifting in the “External Sector Outlook” section, both
productivity, and expanding the variety of goods deficit and surplus economies must implement
and services available globally in recent decades (see, measures that rebalance the composition of global
for example, Baldwin 2016). Reducing barriers in demand and prevent a further buildup of excess
high-tariff sectors such as agriculture; fully imple- global imbalances.
menting commitments under the February 2017 •• Taxation: Differences across jurisdictions in the
Trade Facilitation Agreement; and adapting the rules tax treatment of corporate profits and personal
to cover areas of growing relevance, such as digital income encourage profit shifting and can enable
trade and e-commerce, can help further lower trade tax evasion. Such erosion of tax bases may reduce
costs and contribute to global growth. While agree- national governments’ revenues while some of
ments at the global level, which cover the bulk of the more aggressive preferential tax regimes bring
cross-border trade flows, are optimal in this regard, limited substantive economic benefits or knowledge
broad regional and plurilateral arrangements—such spillovers to the destination locations. Multilateral
as the revised Trans‑Pacific Partnership—can also cooperation on taxation is a long-standing impera-
help forge cross-country consensus on best practices. tive. It has acquired renewed urgency at a time when
Trade openness, as is the case with other forces of high inequality and a stronger sense that global
structural transformation, can hurt certain groups as integration favors large corporations and wealthy
activity shifts to locations with comparatively lower individuals have combined to increase the appeal of
overall operating costs. Measures should be adopted inward-looking policy platforms that could under-
to help those adversely affected by greater economic mine the global recovery.
integration. •• Noneconomic issues: As described in the “Risks”
•• Global financial stability: Cooperative global efforts section, a range of noneconomic factors threaten
have been instrumental in advancing the postcrisis the sustainability of global growth. Cross-border

cooperation remains vital for mitigating green- October 2017 WEO). And by adding to migrant
house gas emissions and for containing the flows, climate-related events compound an already
associated detrimental consequences of rising complex situation of displaced individuals and
global temperatures and devastating climate refugees fleeing conflict areas, often to countries
events, such as droughts, tropical cyclones, and already under severe strain. Multilateral effort
wildfires. These events disproportionately hurt remains indispensable for alleviating these pres-
low-income countries, which have contributed the sures through financial resources directed to the
least to emissions and have relatively low capac- recipient countries and for ensuring unimpeded
ity to tackle their fallout (see Chapter 3 of the aid flows to source locations.

Scenario Box 1. Impact of Tighter Global Financial Conditions

The IMF’s Group of Twenty Model is used to
explore the downside risks of tighter global financial Scenario Figure 1. Inflation Surprise and
conditions. To illustrate the potential implications, it is Term Premium Shocks in the United States
(Percent deviation from baseline for real GDP;
assumed that an inflation surprise in the United States percentage point difference from baseline for CPI
(more likely to occur under expansionary fiscal policy) inflation and interest rates)
leads to faster-than-expected decompression of the US
Inflation surprise and term premium shocks
term premium, which rises 50 basis points in 2018 With sovereign and corporate risk spreads
and an additional 50 basis points in 2019 (blue line
0.2 1. US Real GDP 2. US Core CPI 0.8
in Scenario Figure 1). Thereafter, the term premium Inflation
gradually returns to the baseline. The increase in the 0.0 0.6
US term premium is passed on to all other countries in –0.2 0.4
accordance with the spillover relationship established –0.4 0.2
by the empirical work in the 2014 Spillover Report.
–0.6 0.0
In addition, the tightening in financial conditions
is assumed to heighten risk aversion, with associated –0.8 –0.2
increases in sovereign and corporate risk premiums –1.0 –0.4
2017 19 21 22 2017 19 21 22
(red line in Scenario Figure 1). In line with the time
1.0 3. US Long-Term 4. US Real 0.8
profile of the rise in the term premium, selected sov-
Interest Rate Corporate
ereign and corporate risk premiums increase in 2018 0.8 Interest Rate
0.6
and 2019, decreasing thereafter at the same speed as
0.6
the increase in the term premium. Risk spreads are
0.4
calculated and categorized according to the IMF’s 0.4
assessment of countries’ vulnerabilities stemming from 0.2
0.2
financial, fiscal, and external risks, as well as from
cross-sectoral and cross-border spillovers. In addition, 0.0 0.0
capital flow pressures are assumed to constrain emerg- 2017 19 21 22 2017 19 21 22
ing market central banks so that they cannot fully off- 0.4 5. Advanced 6. Advanced 1.4
0.2 Economies Economies 1.2
set the tightening in financial conditions by loosening Real GDP1 Real Corporate
monetary policy. Conventional monetary policy is also 0.0 Interest Rate1 1.0
assumed to be constrained in the euro area and Japan –0.2 0.8
by the path for short-term policy rates in the World –0.4 0.6
Economic Outlook (WEO) baseline. –0.6 0.4
Higher real effective interest rates in the United –0.8 0.2
States, owing to the faster decompression of the term –1.0 0.0
premium and the need to lean against unexpectedly 2017 19 21 22 2017 19 21 22
higher inflation, dampen aggregate demand, bringing 0.2 7. Emerging 8. Emerging 1.2
Markets Real Markets Real
US real GDP roughly ¾ percentage point below the GDP Corporate 1.0
0.0
WEO baseline by 2020.1 Weaker US demand and, Interest Rate 0.8
more important, the impact of tighter global financial
–0.2 0.6
conditions (red line) cause output to decline by about
½ percent relative to the baseline level by 2020 in 0.4
–0.4
emerging market economies, and more than ¾ percent 0.2
in advanced economies (excluding the United States). –0.6 0.0
In the latter, roughly half of the impact on activity 2017 19 21 22 2017 19 21 22

Note: CPI = consumer price inflation.
1In the outer years, the US real GDP path is a bit higher than 1
Excluding the United States.
in the first layer of the scenario. This is because US monetary
policy is a little more accommodative as the more appreciated
currency puts downward pressure on inflation, and the resulting
stronger domestic demand in the United States more than offsets
weaker foreign demand.

Scenario Box 1. Impact of Tighter Global Financial Conditions (continued)

comes from the faster normalization of the term pre- markets, the overall impact from the term premium
mium (blue line) and half from increased risk aversion increase is relatively small as monetary policy responds
(red line). The assumption of limited conventional and offsets a large part of the impact on real interest
monetary policy space in the euro area and Japan rates, mitigating the impact on real activity (blue line).
exacerbates the impact of the higher term and risk However, when risk aversion increases and capital
premiums on real interest rates and, thus, activity. outflow pressures intensify, the scope for monetary
However, the resulting impact could be mitigated in policy to respond is limited and activity slows more
the euro area and Japan if unconventional monetary significantly and persistently relative to the baseline
policy measures were implemented. In emerging (red line).

Box 1.1. Smartphones and Global Trade

In 2017, global smartphone sales reached close Figure 1.1.1. Global Sales of Personal
to 1.5 billion units—one for every fifth person on Computers and Smartphones
earth (Figure 1.1.1). Demand has been driven by the (Millions of units)
increasing use of smartphones as the main computing
1,600
platform across the world, substituting in part for per-
sonal computers. Mobile technology and services are
1,400 Personal computers
estimated to have contributed $3.6 trillion (4.5 per-
Smartphones
cent) to 2017 global GDP (GSM Association 2018). 1,200
The enormous global demand for smartphones in
recent years has created highly complex and evolving 1,000
supply chains across Asia. In 2017, China exported
$128 billion worth of smartphones to the rest of the 800
world, equivalent to 5.7 percent of its total exports.
In Korea (the main supplier of smartphone compo- 600
nents) semiconductor exports alone accounted for
17.1 percent of total exports. Similarly, components 400
for smartphone production at the peak (October
2017) accounted for more than one-third of exports 200
from Taiwan Province of China, 17.4 percent from
0
Malaysia, and 15.9 percent from Singapore. 2009 10 11 12 13 14 15 16 17
Smartphones contributed about one-sixth the
estimated growth rate of global trade in 2017.1 This Source: Gartner; and IDC.
growth was driven mainly by an increase in value
added per unit, rather than units sold, which declined
for the first time on record. As a result, the average
to note that the income generated from smartphone
sale price of an iPhone increased from $618 in 2016
sales does not fully contribute to the Irish economy.
to $798 in 2017, according to Apple Inc. quarterly
The acquisition of foreign-owned intellectual property
financial statements. In the five main Asian economies
assets leaves domestic employment mostly unchanged.
involved in the tech cycle (China, Korea, Malaysia,
(See Box 1.2 of the April 2017 World Economic Outlook
Singapore, Taiwan Province of China), total exports
for further details on issues relating to the measurement
grew by 6.7 percent in 2017. Even though tech
of Ireland’s GDP.) In Korea, the production chain of
exports accounted for less than 10 percent of total
smartphone-related components is estimated to have
exports in the region, smartphone-related exports
contributed about one-third of real GDP growth rate
contributed about one-third the growth rate of
in 2017. In Taiwan Province of China, the contribu-
total exports.
tion is estimated at about 40 percent. In contrast, for
Ireland, Korea, and Taiwan Province of China are
China the contribution is estimated to be much smaller,
estimated to be the main beneficiaries of the new
reflecting a larger and more diverse economy.
tech cycle in value-added terms. In Ireland, where the
intellectual property of Apple Inc. resides, staff estimate A Rising Tech Cycle
the contribution in value-added terms of iPhone exports
Demand for smartphones is highly cyclical and
to account for one-fourth of the country’s economic
related to the release dates of new smartphone models
expansion in 2017.2 At the same time, it is important
by global producers. Thus, production and trade in
several Asian countries have become highly correlated,
The authors of this box are Benjamin Carton, Yiqun Li, and shaping a new tech cycle, which differs from the earlier
Joannes Mongardini. tech cycle associated with personal computers.
1The contribution is calculated as the net change in real
exports of smartphone components as a share of the net change

in total real exports.
2These estimates are based on iPhone sales as stated in Apple hardware costs, research and development costs, and distribution
Inc. quarterly financial statements and staff assumptions about margins.

Box 1.1 (continued)

Figure 1.1.2. Global iPhone Sales Figure 1.1.3. China: Smartphone Export Cycle
(Millions of units, quarterly) (Millions of units)
80 Mobile phone exports Estimated trend

Tech cycle Lunar new year in
70 China effect
60 160 iPhone release dates

(vertical lines)
150
50
140
40
130
30 120
110
20
100
10 90
0 80
2012 13 14 15 16 17
70
Source: Apple Inc. quarterly financial statements (quarters 60

shown are for calendar years).
50
Sep. Sep. Sep. Sep. Sep. Sep. Dec.
2011 12 13 14 15 16 17
In a recent paper (Carton, Mongardini, and Li

Sources: Haver Analytics; TDM Data; and IMF staff
2018), the new tech cycle is shown as being captured calculations.
by nonseasonal factors. It critically depends on the
release dates of iPhones as Apple Inc. flagship models
drive global demand. In fact, iPhones topped global
sales in the fourth quarter of 2017, surpassing Sam- smartphones. The second is the postrelease cycle,
sung Electronics phones. with shipments of smartphones from China to the
Apple Inc.’s iPhone releases are the key determinant rest of the world. Both pre- and postrelease cycles
of the new tech cycle. Reflecting booming global have a strong impact on growth and trade patterns in
demand, iPhone sales surged from 35.1 million units Asia and beyond.
in the first quarter of 2012 to 78.3 million in the Has the Global Market for Smartphones
fourth quarter of 2016 (Figure 1.1.2). While a clear Become Saturated?
quarterly pattern is emerging—in which second- and
third-quarter sales are usually weaker, reflecting the Global sales of smartphones may have plateaued
expectations of another release in the fourth quarter— in late 2015. By decomposing the cycle from trend
the amplitude of this quarterly pattern has only really for Chinese exports of smartphones, regression results
been established since the release of the iPhone 6/6 show that the trend is nonlinear and may have reached
Plus in September 2014. Moreover, there are clear its peak in September 2015, suggesting that future
spillovers from the fourth quarter of the previous year global demand for smartphones may grow more
onto the first quarter of the following year, ahead of slowly (driven more by replacement demand than new
the Lunar New Year in China. acquisitions). This is confirmed by updated regression
The new tech cycle can be subdivided into two results on Chinese export data up to December 2017
components. The first is the prerelease cycle, which (see Figure 1.1.3). In fact, global shipments of smart-
comprises the export of all components from several phones declined in 2017 for the first time on record
Asian countries to China—the final producer of most (IDC 2018).

Box 1.1 (continued)

However, Asia continues to gain market share in six years, the enormous global demand for smart-
other consumer electronics, including embedded phones has changed the export and growth perfor-
automobile computers, smart appliances, and wearable mance of several Asian countries through complex and
devices. This is evident in the rising demand for Korean evolving supply chains that involve several countries in
semiconductors and, to a lesser extent, in Taiwan Prov- the region. While the global market for smartphones
ince of China’s electronic export orders. In fact, trend may become saturated, demand for other electronics
demand for Korean semiconductor exports continues to products continues to boost production of semicon-
accelerate, despite the slowdown in global smartphone ductors, particularly in Korea. Therefore, the influence
sales, while in Taiwan Province of China electronic of the tech sector on Asia’s export patterns and growth
export orders continue to grow at a healthy pace. is unlikely to fade soon.
Overall, the new tech cycle has become an import-
ant new feature of the global economy. Over the past

Box 1.2. What Has Held Core Inflation Back in Advanced Economies?
Core consumer price inflation in advanced econo- Figure 1.2.1. Advanced Economy Core
mies declined a couple of years after the global finan- Consumer Price Inflation
cial crisis and has not recovered meaningfully since (Percent, year over year; dashed lines indicate
(Figure 1.2.1). Wage growth in advanced economies 2001–08 and 2011–17 averages)
has also remained remarkably sluggish, with wages
2.50
growing 1.5 percentage points less in 2017 than in the
years leading up to the crisis. The absence of stronger
wage and price pressures has been particularly puzzling 2.25
in the past two years given the acceleration in demand
and decline in unemployment in many countries
2.00
(October 2017 World Economic Outlook [WEO] and
Chapter 2 of the October 2016 WEO).
Several explanations have been put forth for the 1.75
seemingly widespread disconnect between inflation
and domestic activity. Some possible forces behind
1.50
sluggish inflation could be domestic in origin,
but may have operated in a synchronized manner
across countries: 1.25
1. Underestimation of slack: Growth in productive
capacity (potential output) may have been under-
1.00
estimated, and excess capacity may not have been 2001 03 05 07 09 11 13 15 17
declining as fast as the acceleration in activity or
the decline in unemployment would suggest.1 Source: IMF staff calculations.
2. Expectations: Even if output is accelerating and
labor markets are tightening, firms may be reluc-
tant to bid up wages and raise prices if they doubt
lower inflation from their trading partners.3 The
the sustainability of the recovery. Another possibil-
widespread use of digital technologies may have
ity is that the inflation expectations of firms and
lowered trading costs, intensifying the competition
workers may have drifted down in a context of per-
for home-produced goods and putting downward
sistent undershooting of inflation targets, long-term
pressure on their prices.4
unemployment, and a perceived narrowing in mon-
4. Enhanced tradability: More generally, enhanced
etary policy space. Some foreign factors may also
tradability and the threat of production relocation
have weighed on core inflation. With an increasing
may have made inflation less sensitive to domestic
range of products, services, and tasks traded across
factors and more responsive to foreign factors,
countries, competition from abroad may have put
including foreign demand and slack.
a lid on the relative prices and inflation rates of
Which of these factors have been more import-
tradable products.2
ant in restraining inflation?5 Disaggregated inflation
3. Drag from import prices and foreign competition:
data could shed light on the relative contributions of
With about half of advanced economy imports
in 2016 originating in economies where out- 3Chapter 2 of the October 2016 WEO documents that excess
put was below potential, sluggish inflation in industrial capacity in major economies, especially China, exerted
advanced economies may in part reflect imported downward pressure on producer price inflation in 2015–16
through lower import prices.
4The decline in the prices of goods relative to services reflects
The authors of this box are Oya Celasun, Weicheng Lian, faster efficiency gains in the production of goods in the past and
and Ava Hong. the continued integration of countries with lower production
1Indeed, wage inflation has been more sluggish where the costs into value chains and trade.
share of workers who are involuntarily working part time has 5In the traditional Phillips curve framework, which relates
remained high (Chapter 2 of the October 2017 WEO). inflation rates to domestic slack, the channels in (1) and (2)
2See also Chapter 3 of the April 2006 WEO and Carney (2017) would result in persistently negative error terms; channel (4)
for a conceptual discussion of the effect of global factors on inflation. would also imply a flattening of the Phillips curve.

Box 1.2 (continued)

domestic versus foreign factors. If foreign factors are Figure 1.2.2. Advanced Economy Core
behind the weakness in inflation, it would suggest a Consumer Goods and Core Services
larger decline in the inflation of tradable goods relative Consumer Price Inflation
to nontradables, such as most services. Conversely, a (Percent, year over year; dashed lines indicate
broad-based decline in inflation rates across compo- 2002–08 and 2011–17 averages)
nents within a country would suggest a greater role of
2.5
domestic factors.
Separating core consumer price inflation in
advanced economies into “core goods” and “core 2.0 Core services
services” components reveals that disinflation since the
global financial crisis—and the additional weakening 1.5
over the past two years—was overwhelmingly the
result of weaker services inflation (Figure 1.2.2).6 By 1.0
contrast, the weighted average of core goods inflation
across 15 advanced economies shows no system- 0.5
atic weakening since the global financial crisis (but
rather continued high volatility around a low level). 0.0
While the changes in core goods inflation have been
Core goods
heterogeneous across countries (with some countries
–0.5
actually witnessing higher core goods inflation), the
decline in services inflation has been remarkably broad
–1.0
(Figure 1.2.3). 2003 05 07 09 11 13 15 17
An examination of core inflation by sector is also
instructive (Figure 1.2.4). The sectors in which inflation Sources: Haver Analytics; and IMF staff calculations.
has weakened the most relative to 2000–07 include Note: The sample comprises 16 advanced economies:
Australia, Austria, Canada, Denmark, Finland, France,
medical services, education, and transportation services. Germany, Japan, Italy, Netherlands, Norway, Portugal,
By contrast, inflation has been higher for traded Spain, Sweden, the United Kingdom, and the United States.
goods such as vehicles, medical products, and apparel,
conflicting with arguments that the rise of online retail
platforms has reduced profit margins and dragged goods
not, however, seem more tradable than a few years
prices down. Regression analysis suggests that declines
ago. It is, instead, more likely that government policies
in core inflation in specific sectors since the global
have contributed to the decline in the rate of price
financial crisis are more strongly related to country-spe-
increases in these sectors, given that prices of medical
cific factors than sector-specific factors. This suggests
services and education are administered or regulated in
that global forces that affect inflation in specific sectors
many countries.
in a similar fashion across countries are unlikely to have
All in all, disaggregated inflation trends suggest
contributed significantly to the decline in core inflation
that enhanced tradability and global competition are
since the crisis.7
unlikely to have been the main culprits behind the
One possibility to consider is that tradability may
sluggishness in inflation in recent years.8 The weakness
have increased for services only, with no change in the
in the inflation of consumer service prices points to
tradability of goods. The sectors with especially weak
domestic factors—including government policies—as
inflation—medical care services and education—do
the more important restraints. A fruitful direction
for future research would be to study the impact of
6Food and all subcomponents relating to fuels are excluded enhanced tradability of service products and changes
from the core goods and core services series. in factor mobility and labor market contestability
7In a regression of the change in sector-level inflation between on inflation.
2002–08 and 2011–17, restricted to tradable sectors, country
dummies explain 29 percent of the variation and sector dummies
only 5 percent. In a similar analysis for nontradables, country 8Cross-border trade in services has increased markedly in
dummies explain 21 percent of the variation and sector dum- recent years, aided by improvements in information and commu-
mies 17 percent. nication technologies.

Box 1.2 (continued)

Figure 1.2.3. Cross-Country Distribution of Figure 1.2.4. Changes in Sectoral Inflation,
Changes in Core Goods and Core Services 2011–17 versus 2002–08
Inflation, 2011–17 versus 2002–08
(Percent, year over year) Change in inflation
10 percent confidence interval
2.0 90 percent confidence interval
Interquartile ranges
Median
1.5 2.5
2.0
1.0
1.5
0.5
1.0
0.0 0.5
0.0
–0.5
–0.5
–1.0 –1.0
–1.5 –1.5
–2.0
–2.0
Core goods Core services –2.5
–3.0
Medical services
Education
Transportation services
Housing
Other services
Household operation
Medical products
Vehicles
Communications
Entertainment
Apparel
Note: The sample comprises 16 advanced economies:
Australia, Austria, Canada, Denmark, Finland, France,
Germany, Japan, Italy, Netherlands, Norway, Portugal,
Spain, Sweden, the United Kingdom, and the United States.
Sources: Haver Analytics; and IMF staff calculations.

Note: Dummies from a regression of changes in sectoral
inflation between 2002–08 and 2011–17 over both country
and sector dummies. The sample comprises 16 advanced
economies: Australia, Austria, Canada, Denmark, Finland,
France, Germany, Japan, Italy, Netherlands, Norway,
Portugal, Spain, Sweden, the United Kingdom, and the
United States.

Box 1.3. Recent Dynamics of Potential Growth

Global economic activity has gathered momentum pin down estimates of the output gap, and thus the
over the past year, thanks in part to a resurgence of evolution of potential growth over time. Applying this
investment in advanced economies. Whether this approach suggests that potential growth picked up, on
momentum can be sustained, and what it implies for average, by 0.4 percentage point between 2011 and
the calibration of macroeconomic policies, depends 2017 for 10 large advanced economies, compared with
partly on whether the faster expansion is mostly an average uptick in actual growth of 0.6 percentage
cyclical (that is, reflecting an acceleration in demand) point over this period. By contrast, a group of five
or also a reflection of faster growth in potential output emerging markets (excluding China) has seen poten-
(that is, an acceleration in supply capacity). A recovery tial growth decline about 0.7 percentage point since
based on stronger potential growth is more likely to be 2011, compared with an actual growth slowdown of
sustained than one driven purely by demand. 1.9 percentage points—more recently, however, there
Potential growth is estimated to have declined in are signs of a turnaround (Figure 1.3.1).
both advanced and emerging market economies in
the wake of the global financial and euro area crises What Is Driving the Recovery?
(April 2015 World Economic Outlook [WEO]), reflect- To shed light on the drivers of potential growth
ing weaker growth in labor, capital, and total factor in advanced economies, the estimates of potential
productivity. In the aftermath of these crises, potential output can be decomposed in line with a standard
growth was projected to rise at a relatively limited pace Cobb-Douglas production function:
through 2020. This box updates the estimates in the
Y  ¯ t  L 
  ¯  t =  A   ¯  t  α Kt  1 − α
  ,
April 2015 WEO and finds that potential growth has
indeed increased somewhat in recent years—mainly in which  Y  ¯  tis potential output as estimated by
due to a recovery in total factor productivity growth— the multivariate filter, K t is the stock of capital, L    ¯  t
but remains well below precrisis rates. The box also is potential employment, and A   ¯
   t is potential total
discusses incorporating information on financial cycles factor productivity, which is treated as the residual
into the calculation of potential output—the concept in our approach.1 For the analysis, the share of labor
of “sustainable growth.” for each country (α) is proxied using estimates from
Chapter 3 of the April 2015 WEO.
To What Extent Has Potential Growth Recovered? ¯  t ) are derived
Estimates of potential employment ( L 
The behavior of inflation in relation to unemploy- from estimates of the nonaccelerating inflation rate of
ment and output contains valuable information on unemployment (NAIRU, U    ¯  t), the working-age popu-
the underlying dynamics of potential growth. When lation (Wt  ) , and the trend labor force participation rate
output outstrips potential output and labor markets ¯  t 
( LFRP  ) as follows:2
tighten, inflation pressure is expected to strengthen;
conversely, inflation is expected to weaken when   ¯  t = (1  −  U 
L  ¯  t) Wt   LFRP 
¯  t 
.
demand falls short of supply. The puzzlingly weak Based on this exercise, the modest increase in
response of inflation to the pickup in output and estimates of potential growth is attributed predomi-
declining unemployment over the past one and a half nantly to a pickup in total factor productivity (TFP)
years suggests that potential output may have risen growth—the residual in the framework (Figure 1.3.2).
alongside actual output. The rebound in TFP growth can be partly explained
Multivariate filtering techniques (as described for by cyclical factors, as some of the headwinds from
example in Blagrave and others 2015) make use of a
simple model that incorporates information on the
relationship between the degree of slack in the econ- 1This residual includes utilization of the inputs of produc-
omy on one hand and inflation and unemployment on tion (labor and capital), labor quality (that is, human capital
the other. Specifically, the Phillips curve (for inflation) accumulation), and possible measurement errors in the inputs of
production. Data on capital stock are from the Organisation for
and Okun’s law (for unemployment) are used to Economic Co-operation and Development.
2Baseline estimates of the trend participation rate are con-
The authors of this box are Olivier Bizimana, Patrick structed using WEO data, whereas estimates of the NAIRU
Blagrave, Mico Mrkaic, and Fan Zhang, with support from are produced by the multivariate filter during the estimation of
Sung Eun Jung. potential output.

Box 1.3 (continued)

Figure 1.3.1. Different Measures of Growth Figure 1.3.2. Production Function
(Percent) Decomposition: Selected Large Advanced
Economies
Real GDP growth (WEO) (Percent)
Potential growth (MVF estimate)
Sustainable growth (MVF estimate)
3.0
4 1. Selected Large AEs Potential growth
3 2.5 Capital
Labor
2 TFP
1 2.0
0
–1 1.5
–2
–3 1.0
–4
2001 03 05 07 09 11 13 15 17 0.5
12 2. Selected Large EMs

0.0
10
8 –0.5
2001 03 05 07 09 11 13 15 17
6
4 Note: Advanced economies = Australia, Canada, France,
Germany, Italy, Japan, Spain, United Kingdom, United
2 States; TFP = total factor productivity.
0
2001 03 05 07 09 11 13 15 17
12 3. Selected Large EMs Excluding China the global financial crisis and euro area sovereign debt
crisis have subsided. In particular, the meaningful
10 easing of financial conditions since 2014 is likely to
8 have facilitated investment in productivity-enhancing
innovation, such as research and development and
6 intangible capital, which can boost total factor pro-
4 ductivity (Figure 1.3.3, panel 1). However, there is
heterogeneity among advanced economies, with invest-
2 ment in intangible assets showing a strong rebound
0 in some (for example, the United States and Japan),
2001 03 05 07 09 11 13 15 17 while it contracted in others (for example, Canada
and Australia). In addition, capacity utilization rates in
Note: AEs = advanced economies (Australia, Canada, France, most major advanced economies have bounced back to
Germany, Italy, Japan, Korea, Spain, United Kingdom, United more normal levels. The recent uptick in estimates of
States); EMs = emerging market economies (Brazil, China, trend TFP growth are closely aligned with estimates of
India, Mexico, Russia, Turkey); MVF = multivariate filter;
WEO = World Economic Outlook. TFP growth using actual GDP, capital stock, and labor
force data (Figure 1.3.4).
Interestingly, despite the recent recovery of invest-
ment growth in major advanced economies, the
contribution of capital-stock growth to potential
remains weak, and well below its precrisis average. This

Box 1.3 (continued)

Figure 1.3.3. Investment in Selected Figure 1.3.4. Total Factor Productivity
Advanced Economies Growth
(Percent)
10 1. Intangible Asset Investment
(Percent, year over year) 3 1. Selected Large AEs
8
6 2
4
1
2
0 0
–2 –1
–4
1996 98 2000 02 04 06 08 10 12 14 16 –2
2001 03 05 07 09 11 13 15 17
11 2. Nonresidential Gross Fixed Investment1
(Percent of GDP) 5 2. Selected Large EMs
10 4
3
9 2
1
8
0
–1
7
2001 03 05 07 09 11 13 15 17 –2
2001 03 05 07 09 11 13 15 17
Note: Advanced economies = Australia, Canada, France, 5 3. Selected Large EMs Excluding China
Germany, Italy, Japan, Korea, Spain, United Kingdom, United
States. 4
1
Gross fixed capital formation data are used for Japan and
Korea. 3
2
1
is because the level of investment (as a share of output) 0
remains depressed, as shown in (Figure 1.3.3, panel
–1
2), implying that growth in the capital stock remains
subdued. Also of note is that, despite a slight rebound –2
2001 03 05 07 09 11 13 15 17
in the contribution of labor inputs in advanced
economies, it remains weak overall because of tepid Source: IMF staff calculations.
working-age population growth in many countries, Note: AEs = advanced economies (Australia, Canada,
which counteracts the impact of a recent slight decline France, Germany, Italy, Japan, Korea, Spain, United
Kingdom, United States); EMs = emerging market
in the NAIRU on potential employment growth. economies (Brazil, China, India, Mexico, Russia, Turkey).
Estimates of Sustainable Growth

A second relevant concept—“sustainable growth”—
aims to estimate an economy’s growth in the absence
of imbalances associated with financial cycles. Similar
to the estimation of potential GDP, sustainable GDP
growth rates are estimated by means of a mul-

Box 1.3 (continued)

tivariate filter (see Berger and others 2015). The filter period imply that recent GDP growth is at least
estimates sustainable growth by controlling for devi- partly fueled by a financial acceleration; conse-
ations in credit, house, and stock prices and inflation quently, the estimate of underlying sustainable out-
from their own longer-term trends and removing their put growth is corrected downward. For emerging
cyclical influence on output from the estimates. For
markets, estimates of sustainable output growth are
example, if wide swings in output are accompanied by
modestly weaker than those for potential growth,
wide swings in credit, the filter interprets such joint
movements as unsustainable and adjusts the rate of
with financial factors playing a similar role.
sustainable growth accordingly.3 Summary
Sustainable growth estimates are similar to those
Estimates of potential growth have increased
for potential growth in advanced economies but slightly in recent years as temporary crisis-related
show a slightly more modest increase in recent effects on total factor productivity growth have
years. The acceleration of credit activity and the unwound. Still, there is not yet any signal that con-
growth of property and equity prices in the recent tributions from labor and capital inputs are on a fast
upswing. This finding indicates that policy measures
3The methodology for estimating sustainable growth is based
to address structural weakness—including investment
on the work of Borio, Disyatat, and Juselius (2013). Related
methods of estimating potential or sustainable output—
in infrastructure and labor market initiatives to offset
including those that incorporate estimates of equilibrium interest the economic effects of aging—are needed to boost
rates—are discussed in Alichi and others (forthcoming). medium-term growth prospects.

Box 1.4. Has Mismeasurement of the Digital Economy Affected Productivity Statistics?
Slow productivity growth has led to questions
about whether productivity is being underestimated.
Figure 1.4.1. Difference between ICT Price
Overstated deflators for information and communi-
Indices and General Non-ICT Price Index,
cations technology (ICT) products are a likely source
Selected OECD Countries
of underestimation. Research on deflators in the US (Percent difference in average annual growth,
national accounts suggests an underestimation of 2010–15)
about 0.3 percentage point, compared with a produc-
3 1. ICT Equipment
tivity slowdown of about 1.5–2.0 percentage points. (Differential growth between IT price index
2
Profit shifting to tax havens may also have depressed and non-ICT price index)
1
measured US productivity growth before 2008.
0
Productivity grows when output increases faster
–1
than labor inputs or, in the case of total factor produc-
–2
tivity, combined labor and capital inputs. A worldwide
–3
slowdown in productivity growth beginning near the
–4
time of the global financial crisis is seen in the data
–5
for most of the world’s economies, with productivity
–6
growth 1–2 percentage points lower than its previous AUS DEU FRA AUT USA CAN NLD ESP BEL ITA GBR
trend in many advanced economies (Adler and others 3 2. Computer Software and Databases
2017). However, advances in digital technology and 2 (Differential growth between software
their diffusion throughout economies seem more rapid price index and non-ICT price index)
1
than ever, leading some to suggest that productivity 0
growth is being underestimated. –1
Research on errors in measuring ICT products –2
before and after the productivity slowdown shows that –3
they play at most a small role, in that productivity –4
was also underestimated before the slowdown began –5
(Byrne, Fernald, and Reinsdorf 2016). Nevertheless, –6
the underestimation is more important compared with AUS USA GBR DEU ITA NLD AUT FRA CAN BEL ESP
measured productivity at today’s lower rates of produc- 3 3. Communication Services
tivity growth (often less than 1 percent a year). 2 (Differential growth between telecom services
price index and CPI)
Accuracy of the deflators used to calculate real 1
output growth is a key question in measuring 0
productivity. Adjusting prices for quality change –1
is often challenging: underadjustment for quality –2
changes could mean price changes are overstated for –3
ICT equipment and software that embody improved –4
technology. Price samples may also underrepresent new –5
products and suppliers that have become important in –6
FRA ITA ESP NLD AUS DEU USA BEL AUT GBR CAN
buyers’ purchasing patterns. Deflators for ICT prod-
ucts vary widely across Organisation for Economic Source: Ahmad, Ribarsky, and Reinsdorf 2017.
Co-operation and Development (OECD) countries, Note: Data reported for Spain for ICT equipment and
computer software and database correspond to 2010–14.
which may reflect differences in quality adjustment Data reported for Austria for communications services
procedures and item samples (Figure 1.4.1). For exam- correspond to 2011–15. Data labels use International
ple, research in the United Kingdom suggests that the Organization for Standardization (ISO) country codes.
CPI = consumer price index; ICT = information and
rate of change in telecommunications service prices communications technology.
was overstated by 7 percentage points during 2010–15
(Abdirahman and others 2017).
The impact of mismeasurement of ICT prices on
aggregate measures of productivity depends on the

Box 1.4 (continued)

weight of the affected items. Quality-adjusted prices for Many of the free digital replacements could be
ICT equipment and software in the United States in captured in productivity statistics by including them
Byrne, Fernald, and Reinsdorf (2016) and Byrne and in the deflator calculations as quality improvements in
Corrado (2017) show substantially lower growth than a priced digital product. Based on the weights of items
the official deflators, but the implied adjustment to US with free or low-cost digital replacements in an average
labor productivity growth is just 24 basis points during consumption basket for OECD countries in 2005,
2004–14. Adjusting for unmeasured improvements the average impact on productivity growth could be
in telecommunications services and unmeasured price 0.1–0.2 percent a year during the years of significant
savings from e-commerce could add another 8–10 basis digital replacement, with a smaller effect today (Reins-
points.1 The size of these effects in other economies dorf and Schreyer 2017).
depends on their price measurement methods and on Research on techniques for inclusion of viewers’
the importance of these items in domestic production. consumption of free media funded by advertising gen-
A different kind of measurement issue is raised by erally finds only a tiny effect on productivity growth.
globalization. Productivity statistics cover production However, a more extreme proposal, by Nakamura,
within a nation’s economic territory, but digitalization Samuels, and Soloveichik (2017), to count all free
has facilitated fragmentation of production across information supplied for marketing purposes in
global supply chains, as well as multinational com- viewers’ consumption—and to define the deflator for
panies’ relocation for tax arbitrage purposes of their online media in a way that implies a rapid decline—
headquarters, intellectual property and other assets, would increase the estimated rate of US productivity
and operations. Multinational companies engaged in growth by 0.1 percentage point. The effect of expan-
tax-driven relocation may misreport the location of sion of the definition of investment to include collec-
production of their output. To investigate this possi- tion of user data has not yet been investigated.
bility, Guvenen and others (2017) use indicators to Questions about nonmarket production or welfare
apportion the worldwide output of multinational com- (that are beyond the scope of productivity statistics) on
panies with headquarters or operations in the United market producers may be addressed in future research
States. This apportionment increases the estimated rate on other indicators. Production of open-source soft-
of US productivity growth by 0.25 percentage point ware by volunteers is one such question. Furthermore,
for 2004–08 but has no effect thereafter. by expanding access to information and variety, and
The scope of productivity statistics is limited to enabling new kinds of services, digital platforms have
output sold at market prices, raising questions about raised households’ productivity in the use of their time
the possible omission of welfare gains from free digital for nonmarket production for their own consump-
products. These products fall into three categories: tion. Tasks previously part of market production
free replacements for nondigital products, such as have shifted to nonmarket production (for example,
video calls over the Internet, online bill paying, and households now act as their own travel agents), and
the camera and GPS capabilities of a smartphone; free low-productivity uses of households’ time have shifted
media, funded by advertising and data collection; and to market production (for example, e-commerce has
products produced by volunteers. replaced driving to the store and finding items on
the shelf ). Research on these questions could provide
1Mismeasurement of quality changes in medical care could important context for productivity statistics.
also be significant but estimates of the possible impact are
unavailable.

Box 1.5. Macroeconomic Impact of Corporate Tax Policy Changes

This box uses the IMF’s Global Integrated Monetary Results
and Fiscal Model (GIMF) to compare the macro- The propagation of both tax policies is broadly
economic impact in the United States of temporarily similar because they both increase the return on capital
reducing the corporate income tax rate with the expenditure. However, the quantitative difference in
impact of temporarily increasing the investment their impact is striking. The impact of introducing
expensing allowance.1 The results illustrate that a a temporary increase in the investment expensing
temporary increase in investment expensing can have allowance (red line in Figure 1.5.1) has a significant
a much larger short-term impact on activity than a impact on investment and output. Under the increase
temporary cut in the corporate income tax that causes in the investment expensing allowance scenario, firms
an identical decline in fiscal revenue. receive a tax benefit only from their investment expen-
Assumptions diture, which sharply increases the incentive to invest
while the increased allowance is in place.3 The boom
Under the corporate income tax reduction sim- in investment supports employment and real wages,
ulation, it is assumed that revenue from domestic which also boosts consumption and raises the price of
and foreign sales is subject to tax and that all costs domestically produced goods.
can be deducted from revenue to calculate the profit In response, the monetary authority tightens policy.
that forms the tax base (to approximate the corpo- The resulting higher real interest rate offsets some of
rate income tax system in the United States). These the reduction in the cost of capital, which dampens
costs include labor, rent, capital depreciation, interest private investment and partly offsets the impact of
expenses, and intermediate inputs, including those higher household incomes on consumption expendi-
that are imported. ture. In addition, the increase in the real interest rate
The simulation that reduces the corporate income leads to an appreciation of the real effective exchange
tax rate reduces the tax payments on corporate rate, lowering import prices and export competitive-
profits.2 The simulation that increases the investment ness. Lower import prices together with additional
expensing allowance component assumes that the cor- domestic demand increase the demand for imports,
porate tax rate remains unchanged, but that some of while higher export prices decrease exports. Net
firms’ investment expenditure can now be considered exports contribute negatively to GDP growth, and the
expenses and, hence, deducted from corporate revenue. current account worsens.
In both simulations, it is assumed that the fiscal Once the investment allowance expires and the
authority credibly changes the corporate tax system return on capital declines, firms reduce their invest-
(the corporate income tax rate or the share of investment expenditure and let the capital stock return to
ment expenditure that can be deducted) such that the baseline. This lowers employment and real wages
the government revenue from taxing the corporate and, hence, consumption—all of which eventually
sector is reduced by half a percent of GDP for five return to the baseline. The gradual reduction in
years. Thereafter, the corporate tax rate and investment domestic demand leads policymakers to ease mone-
allowances both return to their baseline. After five tary policy, which eventually reduces the real interest
years, transfer payments to households are reduced to rate and leads the real exchange rate to depreciate.
bring government debt back to its baseline level over As a result, imports decline, whereas exports get a
the long term. boost from the temporary increase in output and the
exchange rate decline.
The impact of a temporary corporate income tax
reduction (blue line in Figure 1.5.1) increases the
The authors of this box are Benjamin Carton, Emilio Fernan- returns from the use of all factors of production,
dez, and Benjamin Hunt.
1See Carton, Fernandez-Corugedo, and Hunt (2017) for including already-installed capital, thereby stimulat-
a detailed description of the version of GIMF used in these
simulations.
2That is, it reduces both the tax payments stemming from 3The initial increase in investment is dampened by the
corporate revenue and the amount that can be deducted from presence of investment adjustment costs, which capture the fact
the cost of using all inputs, including capital depreciation and that firms do not have additional investment plans ready to be
interest expensing. launched and that it can be costly to install capital.

Box 1.5 (continued)

Figure 1.5.1. Impact of a Temporary Increase ing investment, employment, and real wages over the
in the Investment Expense Allowance and a duration of the cut. Because the return on all factors
Temporary Decrease in the Corporate Income of production rises and because much of that increase
Tax Rate in return is on existing capital, the corporate income
(Percentage point difference from baseline, unless tax cut affects the incentive to invest much less than
noted otherwise)
does the increase in the investment expensing allow-
United States reduces CIT ance. Hence, the temporary tax cut has less impact on
United States increases IEA investment, which also results in lower employment,
2.0 1. Real GDP 2. Real Investment 10 wages, consumption, and domestic demand.
1.5 (Percent (Percent 8
difference) difference) 6
1.0
4
0.5
2
0.0 0
–0.5 –2
0 2 4 6 8 10 0 2 4 6 8 10
1.5 3. Real 4. Primary 0.8
Consumption Deficit-to-GDP
1.0 (Percent Ratio 0.4
difference)
0.5 0.0
0.0 –0.4
–0.5 –0.8
0 2 4 6 8 10 0 2 4 6 8 10
3 5. Real Exports 6. Real Imports 3
(Percent (Percent
2 difference) difference) 2
1 1
0 0
–1 –1
0 2 4 6 8 10 0 2 4 6 8 10
0.4 7. Inflation 8. Nominal Interest 0.8
0.3 Rate 0.6
0.2 0.4
0.1 0.2
0.0 0.0
–0.1 –0.2
–0.2 –0.4
0 2 4 6 8 10 0 2 4 6 8 10
0.2 9. Current Account 10. Physical Return 0.6
0.1 Balance-to-GDP to Capital 0.4
Ratio 0.2
0.0 0.0
–0.1 –0.2
–0.4
–0.2 –0.6
–0.3 –0.8
0 2 4 6 8 10 0 2 4 6 8 10
Note: Years are on the x-axes. CIT = corporate income tax
rate; IEA = investment expense allowance.

Box 1.6. A Multidimensional Approach to Trade Policy Indicators

Recent events underscore the importance of trade There are at least two salient features in the figure.
and trade policy to the global economy. A move First, on average, G20 advanced economies appear
toward inward-looking policies can risk undermin- to have more open trade regimes than their emerging
ing the economic recovery under way. In contrast, market counterparts, with the exception of agricul-
addressing trade distortions can also raise produc- tural support, which remains relatively large in some
tivity and growth, and a generalized move to more advanced economies. Emerging market economies
open trade can facilitate adjustment in countries should not necessarily be expected to be as open as
facing greater import competition (IMF-WTO-WB advanced economies, which began to open to trade
2017). Trade policy discussions stand to benefit from much earlier. At the same time, emerging market
having a strong factual basis covering the multiple economies have liberalized faster over the past two
dimensions that are nowadays relevant to assess trade decades, particularly from the mid-1990s to the
policy. To this end, this box describes and discusses mid-2000s, after which liberalization slowed across all
a set of indicators of trade regimes which can be a countries. Second, the gap between advanced econo-
helpful tool to guide policy discussions (see Cerdeiro mies and emerging market economies is particularly
and Nam 2018). pronounced for one of the services trade measures
Barriers to trade can take many different forms, and trade facilitation. This lends partial support to
ranging from import tariffs to regulatory barriers, the view that liberalization efforts have been some-
restrictions on services trade, and controls on foreign what asymmetric—not just across countries, but also
investment. Because of this diversity, no single indica- across sectors.
tor can provide a complete characterization of a coun- In light of concerns that the period since the 2008
try’s trade regime. The indicators discussed here relate global financial crisis has seen an increase in differ-
to three areas of trade policy—trade in goods, trade in ent forms of protectionism, despite pledges to avoid
services, and foreign direct investment. It is important this outcome, indicators that reflect the evolution
to note that none of the indicators described aims to of a more granular set of trade policies since 2008
benchmark countries’ performance against commit- are presented. While there is considerable hetero-
ments they may have, either under the World Trade geneity within groups of countries, overall, emerg-
Organization (WTO) or vis-à-vis any other forum ing market G20 members appear to have adopted
or agreement. more trade-restricting measures since the crisis
Figure 1.6.1 illustrates the results for the Group of (Figure 1.6.1, panel 2).
Twenty (G20) as well as its advanced economy and All countries, however, including advanced econo-
emerging market members. In panel 1, four of the mies, remain relatively far from free trade. Figure 1.6.2
indicators aim to measure barriers to goods trade: shows the resulting indicators of countries’ trade
average tariffs, the fraction of imports covered by regimes if the edge of the figure represents free trade,
nonautomatic licensing procedures, an index on trade rather than the most open country within the G20.
facilitation, and the level of agricultural support. In The distance from free trade is largest for services
addition, two indicators measure restrictions to ser- restrictions, investment restrictions, trade facilitation,
vices trade, and two aim to capture barriers to foreign and, interestingly, in further most-favored-nation tariff
direct investment. All indicators are normalized, such reductions.
that being closer to the edge of the figure in panel 1 Because of the limitations inherent to any sum-
should be read as being more open.1 mary indicator, and given the lack of quantitative
information for some important policy areas, these
indicators are best used in conjunction with qualita-
tive sources, including WTO Trade Policy Reviews.
The authors of this box are Diego Cerdeiro and
Rachel J. Nam. Information about countries’ specific context is also
1Given that the different indicators are not expressed in essential for discussions about the scope, sequenc-
comparable units of measure, every indicator is normalized ing, and pace of trade reforms (see, for example,
with respect to a reference set of countries (G20 members in
this box), where 0 corresponds to the country that is least open
and 1 to the country that is most open for that indicator. It is policy dimensions that this normalization allows is only in a
important to bear in mind that the comparability across different distance-to-frontier sense.

Box 1.6 (continued)

Figure 1.6.1. Selected Trade Policy Indicators: Example with Group of Twenty Member
Countries
(0 = least open country in G20; 1 = most open country in G20)
G20 average G20 AE average G20 EM average
1. Overall Trade Regime

Nonautomatic licensing
Goods 2010–16
Services 1.0
Average applied OECD average trade
FDI 0.8
MFN tariff facilitation performance
0.6
2015–16 2017
0.4
0.2
0.0
WB-IFC ease of Agricultural support: producer
starting a foreign business support equivalent (OECD)
2008–11 2015–16
OECD FDI Regulatory OECD Services Trade

Restrictiveness Index Restrictiveness Index
2016 2017
WB STRI for services exports
under commercial presence
2008–11
2. Trade Restrictive Measures Adopted since 20081

(Measures in effect as of end-January 2018)
Import duties and taxation
1.0
0.8
0.6
Number of FDI Nontariff import
measures 0.4 restrictions
0.2
0.0
Localization Trade defense

requirements
Export restrictions
Sources: Global Trade Alert; Organisation for Economic Co-operation and Development (OECD); United Nations
COMTRADE database; United Nations Conference on Trade and Development TRAINS; World Bank STRI; World Trade
Organization (WTO), World Tariff Profiles; and IMF staff calculations.
Note: The indicators reflect no judgment as to WTO compliance of underlying measures, nor whether certain measures
(such as trade defense) are an appropriate response to the actions of other countries. The “ease of starting a business”
indicator is based on perceptions as part of an established International Finance Corporation survey process.
AE = advanced economy; AM = advanced economies; EM = emerging market; FDI = foreign direct investment;
G20 = Group of Twenty; IFC = International Finance Corporation; MFN = most-favored nation; STRI = Services Trade
Restrictiveness Index; WB = World Bank.
1
Import (export) coverage ratio, except for the case of FDI (number of measures).
IMF 2010). It would also be useful to quantify government procurement, and intellectual property.
other aspects of countries’ trade regimes—including Better data, both across countries and in terms of
behind-the-border regulations that can hinder trade, policy areas that significantly affect trade, would help
state support (subsidies, state-owned enterprises), to better inform policy discussions.

Box 1.6 (continued)

Figure 1.6.2. Free Trade Normalization: Alternative Normalization
(0 = G20 most closed; 1 = free trade)
G20 average G20 AE average G20 EM average
Nonautomatic licensing
Goods 2010–16
Services 1.0
FDI Average applied OECD average trade
0.8
MFN tariff facilitation performance
0.6
2015–16 2017
0.4
0.2
0.0
WB-IFC ease of Agricultural support: producer
starting a foreign business support equivalent (OECD)
2008–11 2015–16
OECD FDI Regulatory OECD Services Trade

Restrictiveness Index Restrictiveness Index
2016 WB STRI for services exports 2017
under commercial presence
2008–11
Sources: Organisation for Economic Co-operation and Development (OECD); United Nations COMTRADE database; World
Bank STRI; World Trade Organization (WTO), World Tariff Profiles; and IMF staff calculations.
Note: The indicators reflect no judgment as to WTO compliance of underlying measures, nor whether certain measures
(such as trade defense) are an appropriate response to the actions of other countries. The “ease of starting a business”
indicator is based on perceptions as part of an established International Finance Corporation survey process.
AE = advanced economy; AM = advanced economies; EM = emerging market; FDI = foreign direct investment;
G20 = Group of Twenty; IFC = International Finance Corporation; MFN = most favored nation;
STRI = Services Trade Restrictiveness Index; WB = World Bank.

Box 1.7. Growth Outlook—Advanced Economies

Advanced economies are projected to grow at 2.5 in 2018 and moderate to 2.0 percent in 2019. Ita-
percent in 2018—0.2 percentage point higher than in ly’s economy is also set to grow at a stable rate
2017—and 2.2 percent in 2019 (Table 1.1). For both of 1.5 percent this year, softening to 1.1 percent
years, this forecast is considerably stronger than the in 2019. In Spain, growth is projected to decline
October World Economic Outlook (WEO) forecast (0.5 from 3.1 percent in 2017 to 2.8 percent in 2018
and 0.4 percentage point higher for 2018 and 2019, and 2.2 percent in 2019. Medium-term growth in
respectively). Growth in advanced economies is pro- the euro area is projected at 1.4 percent, held back
jected to decline to 1.5 percent over the medium term, by low productivity amid weak reform efforts and
broadly in line with modest potential growth. unfavorable demographics.
•• The growth forecast for the United States has been •• In the United Kingdom, growth is projected to slow
revised up given stronger-than-expected activity from 1.8 percent in 2017 to 1.6 percent in 2018
in 2017, higher projected external demand, and and 1.5 percent in 2019, with business investment
the expected macroeconomic impact of the recent expected to remain weak in light of heightened
changes in fiscal policy. As a by-product, stronger uncertainty about post-Brexit arrangements. The
domestic demand is projected to increase imports forecasts are broadly unchanged relative to the
and widen the current account deficit. The US October WEO. The medium-term growth forecast
growth forecast has been raised from 2.3 to 2.9 per- is also broadly unchanged at 1.6 percent, reflect-
cent in 2018 and from 1.9 to 2.7 percent in 2019. ing the anticipated higher barriers to trade and
Growth is expected to be lower than in previous lower foreign direct investment following Brexit.
forecasts for a few years from 2022 onward, given Assumptions regarding the Brexit outcome remain
the temporary nature of some tax provisions. broadly unchanged relative to the October WEO.
•• The recovery in the euro area is expected to (The UK is assumed to exit the customs union and
strengthen from 2.3 percent in 2017 to 2.4 percent the single market, but tariffs on goods trade with
in 2018, before moderating to 2.0 percent in 2019. the European Union remain at zero, and non-tariff
The forecast is higher than in the October WEO costs increase only moderately.)
by 0.5 and 0.3 percentage point for 2018 and •• Japan’s growth is projected to moderate to 1.2 per-
2019, respectively, reflecting stronger-than-expected cent in 2018 (from a strong above‑trend outturn
domestic demand across the euro area, supportive of 1.7 percent in 2017) before slowing further
monetary policy, and improved external demand to 0.9 percent in 2019. The upward revision of
prospects. Growth forecasts for 2018–19 for 0.5 percentage point in 2018 and 0.1 percentage
all major economies in the euro area have been point in 2019 relative to the October WEO reflects
revised up relative to the October WEO. In France, more favorable external demand prospects, rising
growth is expected to firm up from 1.8 percent private investment, and the Supplementary Budget
in 2017 to 2.1 percent this year, before soften- for 2018. Japan’s medium-term prospects, however,
ing slightly to 2.0 percent in 2019. In Germany, remain weak, mainly due to unfavorable demo-
growth is expected to remain stable at 2.5 percent graphics and a trend decline in the labor force.

Box 1.8. Growth Outlook—Emerging Market and Developing Economies

Growth in emerging market and developing econo- WEO), benefiting from stronger US growth.
mies is expected to increase from 4.8 percent in 2017 Complete implementation of the structural
to 4.9 percent in 2018 and 5.1 percent in 2019 (0.1 reform agenda is projected to maintain growth
percentage point higher for 2019 than in the October close to 3 percent over the medium term.
World Economic Outlook (WEO); Table 1.1). Beyond oo Following a deep recession in 2015–16, Brazil’s
2019, growth in emerging market and developing economy returned to growth in 2017 (1.0 per-
economies is projected to stabilize around 5 percent cent) and is expected to improve to 2.3 percent
over the medium term. in 2018 and 2.5 percent in 2019 on the back
•• In China, growth is projected to moderate from of stronger private consumption and invest-
6.9 percent in 2017 to 6.6 percent in 2018 and ment. The growth forecast is higher than in the
6.4 percent in 2019. The forecast is higher (by October WEO by 0.8 percentage point for 2018
0.1 percentage point in both 2018 and 2019) rela- and 0.5 in 2019. Medium-term growth is set to
tive to the October WEO, reflecting an improved moderate to 2.2 percent.
external demand outlook. Over the medium term, oo In Argentina, growth is expected to moderate
growth is expected to gradually slow to 5.5 per- from 2.9 percent in 2017 to 2.0 percent in 2018
cent with continued rebalancing from investment (0.5 percentage point lower than in the October
to consumption as policy support through fiscal WEO forecast) due to the effect of the drought
and credit channels is gradually reduced, the social on agricultural production, as well as the needed
safety net is strengthened, and precautionary saving fiscal and monetary adjustment to improve the
declines. The economy is also assumed to maintain sustainability of public finances and reduce high
progress on rebalancing from industry to services. inflation. Thereafter, growth is set to recover
However, rising nonfinancial debt as a share of gradually to 3.3 percent over the medium term.
GDP and the accumulation of vulnerabilities weigh oo In Venezuela, real GDP is forecast to fall by
on the medium-term outlook. about 15 percent in 2018 and a further 6 percent
•• Growth elsewhere in emerging and developing in 2019—a significant downward revision com-
Asia is expected to remain strong. India’s economy pared with the declines projected in the October
is projected to grow at 7.4 percent in 2018 and WEO (9.0 percent and 4.0 percent, respectively,
7.8 percent in 2019, up from 6.7 percent in 2017. for 2018 and 2019)—as the collapse in oil pro-
The forecast is unchanged from the October WEO, duction and exports intensifies the crisis that has
with the short-term firming of growth driven by a led to output contraction since 2014.
recovery from the transitory effects of the currency •• The outlook for the Commonwealth of Independent
exchange initiative and implementation of the States is broadly unchanged since the October 2017
national goods and services tax, and supported by WEO, with growth for the region expected to
strong private consumption growth. Among the inch up from 2.1 percent in 2017 to 2.2 percent
ASEAN-5 economies (Indonesia, Malaysia, Philip- in 2018 and stabilize around that level thereafter.
pines, Thailand, Vietnam), broadly stable growth Russia’s return to growth in 2017 was supported
is projected for the group, at 5.3 percent in 2018 by improved oil export revenue, stronger business
and 5.4 percent in 2019 (compared with 5.3 per- confidence, and looser monetary policy. The Rus-
cent in 2017). sian economy is expected to grow by 1.7 percent
•• Recovery in Latin America and the Caribbean is this year, before softening to 1.5 percent over the
strengthening, with growth for the region projected medium term as structural headwinds and sanctions
to increase from 1.3 percent in 2017 to 2.0 percent weigh on activity. Russia’s emergence from recession
in 2018 and 2.8 percent in 2019 (an upward revi- has helped other economies in the region through
sion of 0.1 and 0.4 percentage point, respectively, trade and remittance flows. Growth projections for
for 2018 and 2019 relative to the October WEO). 2018 have been revised up for Azerbaijan to 2.0
oo Mexico’s economy is projected to accelerate from percent (0.7 percentage point higher than in the
2.0 percent in 2017 to 2.3 percent in 2018 and October WEO) on higher public investment, and
3.0 percent in 2019 (0.4 and 0.7 percentage Kazakhstan to 3.2 percent (higher by 0.4 percentage
point higher than projected in the October point relative to the October WEO) reflecting

Box 1.8 (continued)

stronger oil production, but medium-term prospects to the October WEO) as the firming of oil prices
remain subdued. lifts disposable income and business sentiment
•• Growth in emerging and developing Europe is pro- improves. Growth in South Africa is also expected to
jected to moderate from 5.8 percent in 2017 to a strengthen from 1.3 percent in 2017 to 1.5 percent
still-robust 4.3 percent in 2018 and soften further to in 2018 and 1.7 percent in 2019, (stronger than
3.7 percent in 2019 (0.8 and 0.4 percentage point in the October WEO by 0.4 and 0.1 percentage
higher, respectively, than projected in the October point, respectively, for 2018 and 2019). Business
WEO). Stronger external demand, generated by the confidence is likely to gradually firm up as polit-
improved momentum in euro area economic activity, ical uncertainty diminishes, but growth prospects
has generally lifted near-term growth prospects across remain weighed down by structural bottlenecks.
the group. In Poland, strong domestic consumption, •• In the Middle East, North Africa, Afghanistan, and
faster absorption of EU funds, and supportive macro Pakistan region, growth is projected to increase
policies are expected to lift activity above potential from 2.6 percent in 2017 to 3.4 percent in 2018
this year. Growth is projected at 4.1 percent in 2018, and 3.7 percent in 2019. Growth is expected to
moderating to 3.5 percent in 2019—stronger by 0.8 stabilize thereafter at about 3.6 percent through the
and 0.5 percentage point, respectively, than projected medium term. The need for fiscal consolidation as
in the October WEO. Turkey’s economy is also pro- a result of structurally lower oil revenues, security
jected to grow above potential, buoyed by improved challenges, and structural impediments weigh on
external demand conditions and supportive policies the medium-term prospects for many economies in
on multiple fronts—expansionary fiscal policy, state the region. Relative to the forecasts in the Octo-
loan guarantees, procyclical macroprudential policy, ber WEO, with the pickup in oil prices, prospects
and an accommodative monetary stance. Growth is for oil exporters have improved somewhat (with
projected at 4.4 percent in 2018 and 4.0 percent in a small downward revision to 2018 growth and
2019, an upward revision of 0.9 percentage point for a more-than-offsetting positive revision to 2019
2018 and 0.5 percentage point for 2019 relative to growth), while those of oil importers have soft-
the October WEO. ened slightly. In Saudi Arabia, growth is projected
•• Growth in sub-Saharan Africa is projected to rise to resume this year, rising to 1.7 percent from a
to 3.4 percent in 2018 (from 2.8 percent in 2017) contraction of 0.7 percent in 2017. Growth in
and improve slightly thereafter through the medium 2019 is expected to rise slightly to 1.9 percent as oil
term to about 4.0 percent. While the headline num- output increases, with the assumed expiration of the
bers suggest a broadly unchanged picture relative to Organization of the Petroleum Exporting Coun-
the October WEO, revisions to growth projections tries Plus production cut agreement. The forecast
for key large economies point to underlying differ- has been revised up from the October WEO by
ences in prospects across the region. In Nigeria, the 0.6 and 0.3 percentage point for 2018 and 2019,
economy is projected to grow 2.1 percent in 2018 respectively. Growth in Egypt is projected to rise to
and 1.9 percent in 2019 (up from 0.8 percent 5.2 percent in 2018 and 5.5 percent in 2019 (0.7
in 2017), reflecting improved oil prices, revenue, and 0.2 percentage point higher, respectively, than in
and production and recently introduced foreign the October WEO), reflecting stronger momentum
exchange measures that contribute to better foreign in domestic demand and the effect of structural
exchange availability. The forecast is 0.2 percentage reforms. Pakistan’s economy is expected to expand at
point stronger in each year relative to the October a robust pace of 5.6 percent this year (up from 5.3
WEO forecast. Similarly, for the region’s other large in 2017), before moderating to 4.7 percent in 2019.
oil dependent economy, Angola, growth is projected While the forecast for 2018 is unchanged relative
to rise from 0.7 percent in 2017 to 2.2 percent in to the October WEO, for 2019, it has been revised
2018 and 2.4 percent in 2019 (upward revisions of down by 1.3 percentage points, partly reflecting an
0.6 and 1.0 percentage point, respectively, relative increase in macroeconomic vulnerabilities.

Box 1.9. Inflation Outlook—Regions and Countries

As shown in Table 1.1, inflation rates in advanced wages and firms’ operating costs and a very gradual
economies are projected to pick up to 2.0 percent adjustment in inflation expectations.
in 2018, from 1.7 percent in 2017. Inflation in •• Diminishing slack in the economy, together with
emerging market and developing economies exclud- the pass-through effects of depreciation of the
ing Venezuela is expected to increase to 4.6 percent pound, are projected to keep inflation in the United
this year, from 4.0 percent in 2017. The group Kingdom above the Bank of England’s target in
aggregates mask notable differences across individ- 2018. Headline inflation is projected at 2.7 percent
ual countries. in 2018, the same as in 2017. Core CPI (exclud-
ing energy, food, alcoholic beverages and tobacco
Advanced Economies prices) is expected to increase from 2.4 percent in
•• In the United States, headline consumer price 2017 to 2.5 percent this year, before moderating to
inflation is expected to increase from 2.1 percent 2.2 percent in 2019 (and further to 2 percent over
in 2017 to 2.5 percent in 2018, before softening to the medium term) as interest rate hikes and the
2.4 percent in 2019. Core consumer price inflation withdrawal of monetary support proceed.
(CPI)—excluding fuel and food prices—is projected
to increase from 1.8 percent in 2017 to 2.0 percent Emerging Market and Developing Economies
in 2018 and 2.5 percent in 2019 as output is set to •• Headline inflation in China is expected to pick up
rise above potential following the expected sizable to 2.5 percent this year and to about 3 percent over
fiscal expansion. Core personal consumption expen- the medium term as food and energy prices rise and
diture price inflation, the Federal Reserve’s preferred core inflation inches up as a result of diminished
measure, is projected to increase from 1.5 percent excess capacity in the industrial sector, continued
in 2017 to 1.7 percent in 2018 and 2.2 percent robust demand for services, and growing pressure
in 2019. Inflation rates are projected to moderate on wages as the labor force declines.
over the medium term, reflecting a monetary policy •• In Brazil and Russia, headline inflation is expected
response that will keep expectations and actual to remain subdued in the range of 3–4 percent
inflation well anchored. in 2018 as output gaps gradually close, with
•• Headline inflation in the euro area is expected growth continuing to recover from the recession
to remain at 1.5 percent in 2018 and inch up to of 2015–16. Inflation is expected to rise over the
1.6 percent in 2019. With the recovery boosting medium term, with firmer core inflation and the
growth above potential for 2018–19, core CPI is projected modest pickup in commodity prices, but
expected to increase from 1.1 percent in 2017 to to remain at levels well below the average of the
1.2 percent in 2018 and 1.7 percent in 2019. Core past decade. In Mexico, the average inflation rate is
CPI is projected to gradually increase to 2 percent projected to decline to about 4.4 percent in 2018
by 2021 as growth remains above trend for the next (from 6.0 percent in 2017) as the effects of tem-
couple of years and inflation expectations strengthen. porary factors such as fuel price liberalization fade,
•• In Japan, headline inflation is expected to increase dropping further to about 3.0 percent in 2019.
to 1.1 percent in 2018–19 (from 0.5 percent in •• Inflation in sub-Saharan Africa is projected to
2017) due to higher energy and food prices and moderate slightly in 2018 and 2019 but is expected
strong domestic demand. Core inflation is projected to remain in double digits in key large economies,
to rise from 0.1 percent in 2017 to 0.5 percent in reflecting the pass-through effects of currency
2018 and increase further to 0.8 percent in 2019. depreciation and their impact on inflation expec-
Inflation is likely to remain below the Bank of tations (Angola), supply factors, and assumed
Japan’s target over the forecast horizon, reflecting monetary policy accommodation to support fiscal
generally slow pass-through of strong demand to policy (Nigeria).

Special Feature Commodity Market Developments and Forecasts
Special Feature: Commodity Market Developments and Forecasts
Commodity prices have increased since the release of Figure 1.SF.1. Commodity Market Developments
the October 2017 World Economic Outlook (WEO).
Supply outages, the extension of the production agree- 300 1. Commodity Price Indices
All commodities Energy
260 (2005 = 100)
ment by the Organization of the Petroleum Exporting Food Metals
Countries (OPEC), and stronger-than-expected global 220
economic growth all pushed oil prices higher. Metal prices 180
also increased following better-than-expected growth 140
in all major economies and production cuts in China. 100

60
Agricultural prices rose markedly less than those of other
20
commodities, but they have been catching up following 2005 06 07 08 09 10 11 12 13 14 15 16 Feb. 18
unfavorable weather, especially in the Western Hemisphere.
80 2. Brent Futures Curves1
The IMF’s Primary Commodities Price Index rose 70
(US dollars a barrel; expiration dates on x-axis)
16.9 percent between August 2017 and February
60
2018, the reference periods for the October 2017 and
50
current WEO (Figure 1.SF.1, panel 1). Energy prices October 2016 WEO
and metal prices increased substantially, 26.9 per- 40 April 2017 WEO
October 2017 WEO
cent and 8.3 percent, respectively, while agricultural 30 April 2018 WEO
prices increased markedly less, by 4.1 percent. Oil 20
Dec. 16 Dec. 17 Dec. 18 Dec. 19 Dec. 20 Dec. 21 Dec. 22 Dec. 23
prices increased to above $65 a barrel (as of Jan-
uary), attaining their highest level since 2015, in
200 3. Brent Price Prospects2
response to unplanned outages and stronger global (US dollars a barrel)
Futures
economic growth. Since then, prices have receded 160
following stronger-than-expected US production. 120
Natural gas prices increased sharply as a result of
winter heating use and strong demand from China. 80
Coal prices increased, but by less than other energy 40

prices, because a shift from coal to gas is under way in
0
many countries. 2013 14 15 16 17 18 19 20 21
180 4. Metal Price Indices

Oil Prices Highest since 2015 160 (Jan. 2, 2014 = 100)
Aluminum Copper
140
Among key influences on oil prices, on November Iron ore Nickel
120
30, 2017, OPEC agreed to extend to the end of 2018
100
the production target in place since January 2017. This 80
extension was the second (following the April 2017 60
agreement that had extended the November 2016 40
agreement). The agreement entails a cut of 1.2 mil- 20
Jan. 2014 Jan. 15 Jan. 16 Jan. 17 Jan. 18
lion barrels a day (mbd) relative to October 2016
production. Russia and other non-OPEC countries Sources: Bloomberg Finance L.P.; IMF, Primary Commodity Price System;
agreed to stick to current production levels, implying Thomson Reuters Datastream; and IMF staff estimates.
Note: WEO = World Economic Outlook.
additional cuts of about 0.6 mbd relative to the Octo- 1
WEO futures prices are baseline assumptions for each WEO and are derived from
ber 2016 level. futures prices. April 2018 WEO prices are based on February 22, 2018, closing.
2
Derived from prices of futures options on February 22, 2018.
The authors of this feature are Christian Bogmans (team coleader),

Akito Matsumoto (team coleader), and Rachel Yuting Fan, with
research assistance from Lama Kiyasseh.

Figure 1.SF.2. Detrended IMF Crude Oil Index and Economic mostly supply-driven oil price rally. The main reason is
and Market Conditions that a typical income elasticity of demand would imply
at most a 0.2 percent increase in oil demand as a result
Economic and market conditions IMF crude oil index of the 0.2 percentage point upward revision of global
growth for 2018 in the current WEO. Based on a fixed
0.8
supply curve, with price elasticity of supply between 0.03
0.6 and 0.1, the 0.2 percent increase in oil demand would
imply a 2–6 percent increase in prices—that is, a $1 to
0.4 $3 price increase over an initial level of $50 a barrel.
The biggest supply surprise is the faster-than-
0.2
expected deterioration in Venezuelan output. Venezu-
0.0
ela produced 2.38 mbd of crude oil in 2016 and 2.10
mbd in the third quarter of 2017. The latest produc-
–0.2 tion figure stands at 1.62 mbd in December 2017, and
many expect that it will decline to close to 1.0 mbd by
–0.4 the end of 2018. An additional decline in production,
some of which is probably already priced in, would
–0.6
push prices even higher.
–0.8 To sum up, if the supply forecast for 2018 is revised
1999 2001 03 05 07 09 11 13 15 17 down by 0.8 mbd, and the oil demand elasticity is
identical to the oil supply elasticity, it implies that
Sources: Haver Analytics; IMF, Primary Commodity Price System; and IMF staff
calculations. roughly 80 percent of the recent price increase was
caused by a deterioration in supply conditions.
An alternative method to infer the role of demand
In addition to the OPEC extension, unplanned and supply factors in driving price changes uses
outages, including on the US Gulf Coast, in Venezu- regression analysis. Figure 1.SF.2 plots a proxy
ela, and in other locations, cut supply unexpectedly. for global demand; that is, economic and market
Although 2017 non-OPEC supply was slightly stron- conditions—a weighted index based on the pur-
ger than expected, the sharp decline in production in chasing managers’ index, industrial production, and
Venezuela—following further deterioration in its mac- equity prices against the detrended movement in oil
roeconomic and financial conditions—more than offset prices (obtained by using the Hodrick-Prescott filter).
the increase in non-OPEC production. While Libya’s The purchasing managers’ index and equity prices
production increased dramatically during 2017, a recent proxy for market sentiment and financial factors,
outage there together with one in the North Sea further respectively—the latter relates to speculative demand
reduced global oil supply. Hurricane damage to infra- for oil. Figure 1.SF.1 shows that global demand
structure slowed the US production response to rising fluctuations explain oil price movements well over the
oil prices. (The rig count returned to its August 2017 past couple of decades, especially earlier in the sample
level only in February 2018, even though oil prices were period, when demand from China and the financial
rising from their trough below $50 a barrel since June crisis of 2008 and its recovery were key drivers of oil
2017.) However, the stronger-than-expected increase in prices. More recently, however, fluctuations in global
US oil production in early 2018 eventually helped pull demand have been muted, compared with the large
oil prices down from the January high. These events swings in prices, suggesting that demand shocks have
were concentrated between late 2017 and early 2018, so lost much of their explanatory power. Specifically,
spot prices moved much more than futures. the price collapse of 2014 and the notable subse-
quent swings seem only weakly related to movements
in global demand. A regression-based calculation
Oil Price Rally: Largely Supply Driven suggests that only 20 percent of oil price fluctuations
Despite the increase in global aggregate demand, since August 2017 can be attributed to changes in
recent revisions to oil market expectations point to a global demand.

SPECIAL FEATURE Commodity Market Developments and Forecasts
Oil Futures Metal Prices Increasing

Oil futures contracts point to a decline in prices to Metal prices increased by 8.3 percent between
about $53.6 a barrel in 2023 (Figure 1.SF.1, panel 2). August 2017 and February 2018, in line with
Baseline assumptions for the IMF’s average petroleum better-than-expected growth in all major economies.
spot prices, based on futures prices, suggest average Purchasing managers’ indices for major economies
annual prices of $62.3 a barrel in 2018—an increase of have been well above the 50-point mark that separates
18.0 percent from the 2017 average—and $58.2 a bar- growth from contraction, led by the United States
rel in 2019 (Figure 1.SF.1, panel 3). The decline is due and the euro area, and were about 60 as of February
to an expected increase of US supply and the eventual 2018. The World Bureau of Metal Statistics reported a
end of the OPEC deal. wider demand-supply gap for all base metals, especially
Uncertainty remains around the baseline assump- aluminum, as solid economic growth led to higher
tions for oil prices, although risks are balanced. Upside demand, while supply was limited, partially owing
risks include further declines in Venezuelan production to China’s production cuts. Depreciation of the US
and unplanned outages elsewhere. At the same time, dollar has also supported dollar-denominated metal
stronger-than-expected US and Canadian production commodities.
could push prices down sooner than predicted.1 How- Iron ore returned to trading at about $78 a ton
ever, the long end of the futures curve is expected to in February, rising 4.1 percent from its August price
stay at about $55, given current technology trends. of $74.6 a ton. The force behind the recovery comes
from higher steel prices and state-mandated curbs
on steel mills in China, which have lowered output
Natural Gas and Coal despite strong demand. Rising coal prices due to
The natural gas price index—an average for China’s import restrictions further amplified the effect
Europe, Japan, and the United States—rose sharply, during the traditional restocking season for iron ore,
by 45.0 percent, between August 2017 and Febru- adding more demand to this raw material for steel
ary 2018, reflecting seasonal factors, including an production. However, markets are expecting a decline
extremely cold winter in Europe. Strong demand over the medium term, linked to expected lower
for liquefied natural gas (LNG) in China, where the steel production.
government has reduced the use of coal to mitigate Aluminum and copper hit multiyear highs following
air pollution, helped drive the spot LNG price to its production cuts in China (which contributes more
highest level in three years. India’s LNG demand also than half of both global production and consump-
grew strongly in the second half of 2017. Higher oil tion of aluminum) to reduce air pollution during
prices add extra upward pressure to natural gas prices the winter. In turn, this has led to a larger global
in countries where oil-linked pricing is more common. supply-demand deficit and pushed aluminum prices to
The coal price index—an average of Australian close February 7.5 percent higher than August. Like-
and South African prices—increased by 8.4 percent wise, copper prices gained 8 percent during the same
from August 2017 to February 2018. Following the period, boosted by solid demand in China. Futures
introduction of coal import restrictions in July 2017, markets suggest further price increases of both metals
China’s coal imports declined in the second half of over the medium term, in line with improved global
2017 compared with the previous year, although macroeconomic prospects.
total imports were higher than in 2016 as a result of The price of nickel, a key ingredient in stainless
increases in the first half of the year. More recently, steel and batteries in electric vehicles, reached multi-
however, Chinese import restrictions were temporarily year highs in February, up 24.8 percent over August
lifted to accommodate strong winter heating demand. 2017. Owing to strong demand from China and tight
supplies, nickel inventories at London Metal Exchange
warehouses fell since October to a 14-month low in
1The US Energy Information Administration expects US crude oil January. Cobalt, another raw material for batteries,
production in 2018 to reach 10.3 mbd, exceeding the previous high has experienced sharp price increases since 2016,
of 9.6 mbd recorded in 1970, and to increase further in 2019. Can-
fueled by tight supply and rising demand from electric
ada’s oil production, which has been growing strongly, is expected to
grow further. vehicle manufacturers. Hitting a nine-year high in late

January, cobalt prices were up 38.2 percent in February Palm oil prices rose by 3.4 percent from August
2018 relative to their August 2017 average.2 2017 to February 2018. Prices trended down through-
Uranium was hovering at about $20 since August, out 2017 as production growth in Indonesia and
but rallied in early November following the announce- Malaysia continued to outpace demand growth and
ment of production cuts by two of the world’s biggest stocks recovered. But prices increased in early 2018 as
producers. The price has receded since early Decem- higher oil prices stimulated biodiesel demand in Indo-
ber and fell 11.2 percent between August 2017 and nesia. Another major support for palm oil prices is the
February 2018. reduction in supplies of rival oilseeds, such as soybeans,
caused by bad weather.
Cotton prices increased by 11.3 percent between
Adverse Weather Driving Food Prices Higher August 2017 and February 2018. The recent price
The IMF’s agricultural price index rose 4.1 per- increase follows worries over pest damage to India’s
cent from August 2017 to February 2018, given that crop, resulting in lower stocks available for export, as
unfavorable weather conditions in recent months well as setbacks to the latest US harvest during the
are expected to reduce this year’s harvests of many hurricane season. Looking ahead, the recent increase
grains and oilseeds. The subindices of food and in oil prices is likely to provide support for cotton
agricultural raw materials rose by 4.1 and 6.0 per- prices, because it makes artificial fibers more expensive.
cent, respectively, and the beverages index declined Falling stocks in China are also likely to contribute
by 3.6 percent. The drop in beverage prices can be upward pressure on prices.
attributed to a substantial decline in the price of Pork prices declined by 11.2 percent from August
coffee (by 12.7 percent) while the gain in the index 2017 to February 2018 due to seasonal factors. While
of raw agricultural materials follows a rally in the supplies are expected to increase in 2018, especially in
price of cotton. the United States, strong demand from China, Japan,
Wheat prices increased by 23.9 percent between Mexico, and the United States implies that markets
August 2017 and February 2018. Following the are again expected to clear at higher year-over-year
Northern Hemisphere harvests and continued stock prices. Beef prices rose by 3.1 percent because supply
building in most of the world, except China, wheat growth in the United States, a major producer and
prices remained under significant pressure until exporter, was offset by strong export demand. More-
November. Since then, prices have rallied—winter over, drought in the United States reduced the number
wheat crops in the key southern Plains region of the of cattle placed on feedlots.
United States were likely significantly damaged by cold Following dry weather in west Africa at the begin-
and dry winter weather. ning of 2018, output of cocoa is expected to fall in all
Soybean prices trended up from August 2017 to producer countries, including the top producer, Côte
February 2018, increasing by 7.5 percent, following d’Ivoire, although the world is still projected to run
concerns over weather in South America. A deterio- a production surplus in 2017–18. The reduction in
ration in the next Argentine soybean crop because of expected supply comes at a time of strong demand.
hot and dry conditions has stimulated early buying, These developments led to an increase in the price
providing price support for the soybean complex. The of cocoa of 6.8 percent between August 2017 and
outlook is bullish as continued feed demand growth February 2018.
and supportive global biodiesel policies counter histori- The price of Arabica coffee declined by 7.6 percent
cally large global stocks. between August 2017 and February 2018, reflecting
Maize prices have also increased since August, weaker-than-expected demand for exports at the begin-
rising by 10.1 percent, following the upward trend of ning of the 2017–18 season.
soybean prices. While dry weather in Argentina has The price of sugar decreased by 6.7 percent between
already reduced yields of the partially harvested corn August 2017 and February 2018, reflecting upward
crop, in Brazil, rainfall is hampering planting, poten- revisions to an expected 2017–18 surplus global
tially reducing future yields. production. In India, most notably, output may exceed
that of the previous season by as much as 40 percent.
2Box 1.SF.1 studies the role of cobalt and lithium as important Strong supplies from Brazil and Europe in 2018–19
raw materials in the production of electric vehicle batteries. are likely to lead to another surplus year.

SPECIAL FEATURE Commodity Market Developments and Forecasts
The prices of most major agricultural commodities several commodities. The recent worries over Argentina’s
have been revised up slightly, reflecting diminishing soybean crop, as well as the reported setback to winter
excess supply. Overall, food prices are projected to wheat crops in the key southern Plains region of the
increase by 2.6 percent in 2018 and 1.8 percent in United States—both caused by cold and dry winter
2019, mostly on account of rising cereal and oilseed weather—are consistent with historical patterns of the
prices (compared with the previously projected decrease weather phenomenon. Changes in trade policies may
of 0.7 percent and increase of 2.6 percent, respectively) be another upside risk factor, especially for agricultural
and are expected to decline again thereafter. importers. A depreciating US dollar helped stimulate
Weather disruptions and variability are an upside exports in 2017, but a partial reversal in 2018 could
risk to the forecast for agricultural prices. The ongoing put upward pressure on prices for importing countries.
weak-to-moderate La Niña weather pattern has peaked Uncertainty over global corn acreage, as production
and is expected to weaken further over the spring. It has margins for farmers remain low, could put upward pres-
proved to be a significant source of price volatility for sure on corn prices by the end of this year.

Box 1.SF.1. The Role of Metals in the Economics of Electric Vehicles

The emergence of electric vehicle markets is supported by Figure 1.SF.1.1. Hundred Years of Cobalt
the falling costs of lithium-ion batteries, the most common Mining
and industry-preferred battery for such vehicles. Con-
versely, the emergence of electric vehicles has helped reduce Natural log of world production (tons, right scale)
the production costs of these batteries through economies of Cobalt unit value ($98/ton, left scale)
scale. A lithium-ion battery consists of an anode, typically
graphitic carbon, and a cathode, separated by a liquid 140,000 15
organic electrolyte. The cathode typically uses lithium and
some combination of copper, nickel, manganese, alumi- 120,000
num, and cobalt. 13
100,000
Expenditures on metals for cathode construction
make up a large share of total lithium-ion and electric 80,000
11
vehicle costs. Substitution with other materials is
difficult. Lithium is an important ingredient because
60,000
it is an element that is easily ionized or “charged.” 9
In addition, it allows for high energy density and, as
40,000
such, yields batteries that dominate in the automotive
area and in portable electronics. Cobalt is important 7
for similar reasons but, at historically high prices, its 20,000
cost share significantly exceeds that of lithium.
As supplies of lithium and cobalt have been unable 0 5
1915 30 40 50 60 70 80 90 2000 10 17
to keep up with the surge in demand following the
rapid growth of electric car sales in recent years, prices
Sources: US Geological Survey; and IMF staff calculations.
have been rising. The Chinese spot price of lithium Note: Gray lines indicate boom years.
carbonate increased by more than 30 percent in 2017.
Even more noteworthy is the price path of cobalt:
after more than doubling between September 2016
and April 2017, prices rose an additional 25 percent Unlike lithium, cobalt supply is likely to remain
between November 2017 and January 2018. The relatively tight, at least over the next 5 to 10 years. In
question now is how production of these metals will 2016 more than 50 percent of global supply orig-
change. To answer this question, this box analyzes inated in the Democratic Republic of the Congo.
global supply conditions. China (6.3 percent), Canada (5.9 percent), and Russia
(5.0 percent) are other important, but much smaller,
Supply Conditions of Lithium and Cobalt players. There is also an unofficial “artisanal” stream of
Australia and Chile are by far the biggest produc- production, some of it under the control of insur-
ers of lithium, together accounting for more than gent militias and relying on child labor. Geopolitical
three-quarters of world production; Argentina is a instability in the Democratic Republic of the Congo
distant third. According to the US Geological Survey, has the potential to disrupt supply, as it did at the end
world reserves stood at 600 times global output in of the 1970s when political unrest led to a price boom
2015. Production is thus not limited by physical (Figure 1.SF.1.1). Furthermore, the refining of cobalt
resource scarcity. But, although recent production is also geographically concentrated, with China by far
deficits and rising prices have encouraged new pro- being the biggest producer.
ductive capacity, this new capacity has not so far kept The specificities of the cobalt production process are
prices in check. perhaps the weakest link in the supply chain. Cobalt
is mostly produced as a by-product of mining of other
metals, nickel (50 percent), and copper (35 percent);
only 6 percent of world production originates from
The authors of this box are Christian Bogmans and primary production (see Olivetti and others 2017).
Lama Kiyasseh. For nickel-cobalt mines, most of the revenues come

Box 1.SF.1 (continued)

from nickel. This implies that the supply of cobalt be matched by investments in productive capacity, but
from nickel-cobalt mines is inelastic with respect to there could still be supply constraints: new mining
the price of cobalt. projects have long lead times, and concerns about the
The situation is different, however, for copper: given local environmental impact of mining in Latin Amer-
last year’s prices, a copper‑cobalt mine could have ica and elsewhere could slow the issuance of permits.
obtained more than half its revenue from cobalt. Most When it comes to cobalt, the situation seems to
cobalt-copper ore and reserves are in the Democratic be more pressing. Based on a modest forecast of
Republic of the Congo, which implies that the rising 10 million electric vehicle sales in 2025, Olivetti and
price of cobalt will generate new supplies primarily others (2017) suggest a demand for cobalt exceed-
from that country, further concentrating cobalt pro- ing 330 kilotons by 2025—almost three times the
duction. Last year, mining companies from the West current world production. Such demand would require
and China invested heavily in copper mines in the average annual growth of more than 11 percent for
Democratic Republic of the Congo. the next decade, well beyond that of the past 50 years.
Since 1915 there have been four price boom Historical evidence from the 20th century suggests
episodes—defined as a sequence of years during which that most commodity price booms peak within two
real prices are in the upper 10 percent of a normal years of their onset (Jacks 2013) as they give way to
distribution’s right tail. Those during 1978–81 and permanent changes in productive capacity and new
1995–96 elicited sharp responses: world production productivity-enhancing investment. But occasion-
grew by 54.1 and 36.1 percent in 1983 and 1995, ally they last longer. The required growth in cobalt
respectively, significantly higher than the 50-year production—historically unprecedented—is a risk to
average of 4.8 percent. As of January 2018, prices of the electrification of the transportation sector.
15-month cobalt futures suggest that 2018 will be the Several developments could limit price volatility.
first boom year since the 1995–96 episode. These include increased recycling of cobalt and new
primary production mining techniques. Perhaps most
Outlook important, progress in battery technology could bring
Future demand for cobalt and lithium will depend the surge in cobalt prices to a halt. One of the lead-
on the growth of their end-use products—including ing alternatives to the lithium-ion battery concept—
electronics and automobiles—which in turn depend the solid-state battery—would mean smaller and
on oil prices, economic growth, and battery tech- more-energy-dense batteries that would not need cobalt
nology, among other factors. Based on a forecast as an input. Widescale adoption of a mature solid-state
of global lithium-ion battery consumption, global battery concept would reduce the demand for cobalt.
lithium demand is expected to increase from 181 Continued research in this area will prevent resource
kilotons of lithium carbon equivalent to 535 kilotons constraints from delaying or altogether halting progress
by 2025 (Deutsche Bank 2017). This demand could in electric vehicles and portable electronics.

Annex Table 1.1.1. European Economies: Real GDP, Consumer Prices, Current Account Balance, and Unemployment
(Annual percent change, unless noted otherwise)
Real GDP Consumer Prices1 Current Account Balance2 Unemployment3
Projections Projections Projections Projections
2017 2018 2019 2017 2018 2019 2017 2018 2019 2017 2018 2019
Europe 3.0 2.7 2.3 2.6 2.7 2.6 2.3 2.4 2.4 ... ... ...
Advanced Europe 2.4 2.3 2.0 1.7 1.7 1.7 2.9 2.9 3.0 7.9 7.4 7.1
Euro Area4,5 2.3 2.4 2.0 1.5 1.5 1.6 3.5 3.2 3.2 9.1 8.4 8.1
Germany 2.5 2.5 2.0 1.7 1.6 1.7 8.0 8.2 8.2 3.8 3.6 3.5
France 1.8 2.1 2.0 1.2 1.5 1.6 –1.4 –1.3 –0.9 9.4 8.8 8.4
Italy 1.5 1.5 1.1 1.3 1.1 1.3 2.9 2.6 2.2 11.3 10.9 10.6
Spain 3.1 2.8 2.2 2.0 1.7 1.6 1.7 1.6 1.7 17.2 15.5 14.8
Netherlands 3.1 3.2 2.4 1.3 2.0 2.2 9.8 9.6 8.9 5.1 4.9 4.8
Belgium 1.7 1.9 1.7 2.2 1.6 1.8 0.1 0.3 0.2 7.2 7.0 6.8
Austria 2.9 2.6 1.9 2.2 2.2 2.2 2.1 2.5 2.0 5.5 5.2 5.1
Greece 1.4 2.0 1.8 1.1 0.7 1.1 –0.8 –0.8 –0.6 21.5 19.8 18.0
Portugal 2.7 2.4 1.8 1.6 1.6 1.6 0.5 0.2 –0.1 8.9 7.3 6.7
Ireland 7.8 4.5 4.0 0.3 0.9 1.3 12.5 9.8 8.7 6.7 5.5 5.2
Finland 3.0 2.6 2.0 0.8 1.2 1.7 0.7 1.4 1.9 8.7 8.0 7.5
Slovak Republic 3.4 4.0 4.2 1.3 1.9 1.9 –1.5 –0.3 0.5 8.3 7.5 7.4
Lithuania 3.8 3.2 3.0 3.7 2.2 2.2 1.0 –0.1 –0.6 7.1 6.9 6.8
Slovenia 5.0 4.0 3.2 1.4 1.7 2.0 6.5 5.7 5.2 6.8 5.9 5.5
Luxembourg 3.5 4.3 3.7 2.1 1.4 1.8 5.5 5.4 5.3 5.8 5.5 5.2
Latvia 4.5 4.0 3.5 2.9 3.0 2.5 –0.8 –1.9 –2.2 8.7 8.2 8.1
Estonia 4.9 3.9 3.2 3.7 3.0 2.5 3.2 2.0 0.7 5.8 6.3 6.7
Cyprus 3.9 3.6 3.0 0.7 0.4 1.6 –4.7 –4.1 –4.6 11.3 10.0 9.1
Malta 6.6 5.7 4.6 1.3 1.6 1.8 10.2 9.9 9.5 4.0 4.2 4.4
United Kingdom 1.8 1.6 1.5 2.7 2.7 2.2 –4.1 –3.7 –3.4 4.4 4.4 4.5
Switzerland 1.1 2.3 2.0 0.5 0.7 1.0 9.3 9.7 9.4 3.2 3.0 3.0
Sweden 2.4 2.6 2.2 1.9 1.5 1.6 3.2 3.1 3.1 6.7 6.3 6.3
Norway 1.8 2.1 2.1 1.9 1.9 2.0 5.1 6.1 6.5 4.2 3.9 3.7
Czech Republic 4.3 3.5 3.0 2.4 2.3 2.0 1.1 0.3 0.4 2.9 3.0 3.2
Denmark 2.1 2.0 1.9 1.1 1.4 1.7 7.6 7.6 7.2 5.8 5.7 5.6
Iceland 3.6 3.2 3.0 1.8 2.4 2.3 3.6 3.3 2.6 2.8 3.1 3.3
San Marino 1.5 1.3 1.3 0.9 1.0 1.1 ... ... ... 8.0 7.4 6.8
Emerging and Developing Europe6 5.8 4.3 3.7 6.2 6.8 6.3 –2.6 –2.9 –2.7 ... ... ...
Turkey 7.0 4.4 4.0 11.1 11.4 10.5 –5.5 –5.4 –4.8 11.0 10.7 10.7
Poland 4.6 4.1 3.5 2.0 2.5 2.5 0.0 –0.9 –1.2 4.9 4.1 4.0
Romania 7.0 5.1 3.5 1.3 4.7 3.1 –3.5 –3.7 –3.7 5.0 4.6 4.6
Hungary 4.0 3.8 3.0 2.4 2.7 3.3 3.6 2.5 2.4 4.0 3.8 3.5
Bulgaria5 3.6 3.8 3.1 1.2 2.0 2.1 4.5 3.0 2.3 6.2 6.0 5.8
Serbia 1.8 3.5 3.5 3.1 2.7 3.0 –4.6 –4.5 –4.1 14.6 14.3 14.0
Croatia 2.8 2.8 2.6 1.1 1.5 1.5 3.7 3.0 2.1 12.2 12.0 11.2
Note: Data for some countries are based on fiscal years. Refer to Table F in the Statistical Appendix for a list of economies with exceptional reporting periods.
¹Movements in consumer prices are shown as annual averages. Year-end to year-end changes can be found in Tables A6 and A7 in the Statistical Appendix.
2Percent of GDP.
3Percent. National definitions of unemployment may differ.
4Current account position corrected for reporting discrepancies in intra-area transactions.
5Based on Eurostat’s harmonized index of consumer prices except for Slovenia.
6Includes Albania, Bosnia and Herzegovina, Kosovo, FYR Macedonia, and Montenegro.

Annex Table 1.1.2. Asian and Pacific Economies: Real GDP, Consumer Prices, Current Account Balance, and Unemployment
2017 2018 2019 2017 2018 2019 2017 2018 2019 2017 2018 2019
Asia 5.7 5.6 5.6 2.1 2.9 2.9 2.1 1.8 1.8 ... ... ...
Advanced Asia 2.4 2.1 1.9 1.0 1.4 1.5 4.3 4.3 4.3 3.4 3.4 3.3
Japan 1.7 1.2 0.9 0.5 1.1 1.1 4.0 3.8 3.7 2.9 2.9 2.9
Korea 3.1 3.0 2.9 1.9 1.7 1.9 5.1 5.5 5.8 3.7 3.6 3.3
Australia 2.3 3.0 3.1 2.0 2.2 2.4 –2.3 –1.9 –2.3 5.6 5.3 5.2
Taiwan Province of China 2.8 1.9 2.0 0.6 1.3 1.3 13.8 13.6 13.5 3.8 3.8 3.7
Singapore 3.6 2.9 2.7 0.6 1.2 1.0 18.8 18.9 18.7 2.2 2.1 2.1
Hong Kong SAR 3.8 3.6 3.2 1.5 2.2 2.1 3.0 3.1 3.2 3.1 3.1 3.1
New Zealand 3.0 2.9 2.9 1.9 1.7 2.1 –2.7 –2.6 –3.0 4.7 4.5 4.6
Macao SAR 9.3 7.0 6.1 1.2 2.2 2.4 30.4 32.1 33.1 2.0 2.0 2.0
Emerging and Developing Asia 6.5 6.5 6.6 2.4 3.3 3.3 0.9 0.6 0.6 ... ... ...
China 6.9 6.6 6.4 1.6 2.5 2.6 1.4 1.2 1.2 3.9 4.0 4.0
India4 6.7 7.4 7.8 3.6 5.0 5.0 –2.0 –2.3 –2.1 ... ... ...
ASEAN-5 5.3 5.3 5.4 3.1 3.2 2.9 2.1 1.5 1.3 ... ... ...
Indonesia 5.1 5.3 5.5 3.8 3.5 3.4 –1.7 –1.9 –1.9 5.4 5.2 5.0
Thailand 3.9 3.9 3.8 0.7 1.4 0.7 10.8 9.3 8.6 0.7 0.7 0.7
Malaysia 5.9 5.3 5.0 3.8 3.2 2.4 3.0 2.4 2.2 3.4 3.2 3.0
Philippines 6.7 6.7 6.8 3.2 4.2 3.8 –0.4 –0.5 –0.6 5.7 5.5 5.5
Vietnam 6.8 6.6 6.5 3.5 3.8 4.0 4.1 3.0 2.4 2.2 2.2 2.2
Other Emerging and Developing
Asia5 6.0 6.1 6.3 5.2 5.2 5.3 –2.2 –2.7 –2.6 ... ... ...
Memorandum
Emerging Asia6 6.6 6.5 6.6 2.3 3.2 3.2 1.0 0.7 0.7 ... ... ...
1Movements in consumer prices are shown as annual averages. Year-end to year-end changes can be found in Tables A6 and A7 in the Statistical Appendix.
2Percent of GDP.
4See country-specific notes for India in the “Country Notes” section of the Statistical Appendix.
5Other Emerging and Developing Asia comprises Bangladesh, Bhutan, Brunei Darussalam, Cambodia, Fiji, Kiribati, Lao P.D.R., Maldives, Marshall Islands, Micronesia,
Mongolia, Myanmar, Nauru, Nepal, Palau, Papua New Guinea, Samoa, Solomon Islands, Sri Lanka, Timor-Leste, Tonga, Tuvalu, and Vanuatu.
6Emerging Asia comprises the ASEAN-5 (Indonesia, Malaysia, Philippines, Thailand, Vietnam) economies, China, and India.

Annex Table 1.1.3. Western Hemisphere Economies: Real GDP, Consumer Prices, Current Account Balance, and Unemployment
2017 2018 2019 2017 2018 2019 2017 2018 2019 2017 2018 2019
North America 2.3 2.8 2.6 2.5 2.7 2.5 –2.4 –3.0 –3.3 ... ... ...
United States 2.3 2.9 2.7 2.1 2.5 2.4 –2.4 –3.0 –3.4 4.4 3.9 3.5
Canada 3.0 2.1 2.0 1.6 2.2 2.2 –3.0 –3.2 –2.5 6.3 6.2 6.2
Mexico 2.0 2.3 3.0 6.0 4.4 3.1 –1.6 –1.9 –2.2 3.4 3.5 3.4
Puerto Rico4 –7.7 –3.6 –1.2 1.9 2.2 0.8 ... ... ... 12.5 12.0 11.3
South America5 0.7 1.7 2.5 ... ... ... –1.4 –2.0 –2.2 ... ... ...
Brazil 1.0 2.3 2.5 3.4 3.5 4.2 –0.5 –1.6 –1.8 12.8 11.6 10.5
Argentina 2.9 2.0 3.2 25.7 22.7 15.4 –4.8 –5.1 –5.5 8.4 8.0 7.5
Colombia 1.8 2.7 3.3 4.3 3.5 3.4 –3.4 –2.6 –2.6 9.3 9.2 9.1
Venezuela –14.0 –15.0 –6.0 1,087.5 13,864.6 12,874.6 2.0 2.4 3.6 27.1 33.3 37.4
Chile 1.5 3.4 3.3 2.2 2.4 3.0 –1.5 –1.8 –1.9 6.7 6.2 5.8
Peru 2.5 3.7 4.0 2.8 1.6 2.0 –1.3 –0.7 –1.1 6.7 6.7 6.7
Ecuador 2.7 2.5 2.2 0.4 1.0 1.4 –0.4 –0.1 0.3 4.6 4.3 4.3
Bolivia 4.2 4.0 3.8 2.8 3.5 4.5 –5.8 –5.4 –5.2 4.0 4.0 4.0
Uruguay 3.1 3.4 3.1 6.2 7.0 6.1 1.6 0.6 –0.1 7.4 7.1 7.1
Paraguay 4.3 4.5 4.1 3.6 4.2 4.0 –1.8 –2.0 –1.2 5.7 5.7 5.7
Central America6 3.7 3.9 4.0 2.6 3.4 3.4 –2.5 –2.9 –2.6 ... ... ...
Caribbean7 2.7 3.8 3.7 3.8 4.5 3.5 –3.2 –3.2 –2.9 ... ... ...
Memorandum
Latin America and the Caribbean8 1.3 2.0 2.8 4.1 3.6 3.5 –1.6 –2.1 –2.3 ... ... ...
East Caribbean Currency Union9 1.8 1.8 3.6 1.1 1.2 1.8 –9.2 –12.0 –8.5 ... ... ...
2Percent of GDP.
4Puerto Rico is a territory of the United States but its statistical data are maintained on a separate and independent basis.
5Includes Guyana and Suriname. Data for Argentina’s and Venezuela’s consumer prices are excluded. See country-specific notes for Argentina and Venezuela in the “Country
Notes” section of the Statistical Appendix.

6Central America comprises Belize, Costa Rica, El Salvador, Guatemala, Honduras, Nicaragua, and Panama.
7The Caribbean comprises Antigua and Barbuda, The Bahamas, Barbados, Dominica, Dominican Republic, Grenada, Haiti, Jamaica, St. Kitts and Nevis, St. Lucia, St. Vincent
and the Grenadines, and Trinidad and Tobago.

8Latin America and the Caribbean comprises Mexico and economies from the Caribbean, Central America, and South America. Data for Argentina’s and Venezuela’s con-
sumer prices are excluded. See country-specific notes for Argentina and Venezuela in the “Country Notes” section of the Statistical Appendix.
9Eastern Caribbean Currency Union comprises Antigua and Barbuda, Dominica, Grenada, St. Kitts and Nevis, St. Lucia, and St. Vincent and the Grenadines as well as
Anguilla and Montserrat, which are not IMF members.

Annex Table 1.1.4. Commonwealth of Independent States Economies: Real GDP, Consumer Prices, Current Account
Balance, and Unemployment
2017 2018 2019 2017 2018 2019 2017 2018 2019 2017 2018 2019
Commonwealth of Independent States4 2.1 2.2 2.1 5.5 4.6 4.8 1.3 2.8 2.3 ... ... ...
Net Energy Exporters 2.0 2.1 2.0 4.8 4.1 4.5 1.9 3.6 3.1 ... ... ...
Russia 1.5 1.7 1.5 3.7 2.8 3.7 2.6 4.5 3.8 5.2 5.5 5.5
Kazakhstan 4.0 3.2 2.8 7.4 6.4 5.6 –2.9 –1.4 –1.3 5.0 5.0 5.0
Uzbekistan 5.3 5.0 5.0 12.5 19.5 12.9 3.7 0.2 –1.1 ... ... ...
Azerbaijan 0.1 2.0 3.9 13.0 7.0 6.0 3.5 5.6 7.0 5.0 5.0 5.0
Turkmenistan 6.5 6.2 5.6 8.0 9.4 8.2 –11.5 –9.0 –7.8 ... ... ...
Net Energy Importers 3.1 3.2 3.3 10.2 8.3 6.7 –3.7 –4.2 –4.1 ... ... ...
Ukraine 2.5 3.2 3.3 14.4 11.0 8.0 –3.7 –3.7 –3.5 9.4 9.2 8.8
Belarus 2.4 2.8 2.4 6.0 6.0 6.0 –1.8 –2.5 –2.7 1.0 1.0 1.0
Georgia 4.8 4.5 4.8 6.0 3.6 3.0 –9.3 –10.5 –9.5 ... ... ...
Armenia 7.5 3.4 3.5 0.9 3.5 4.0 –2.6 –2.8 –2.8 18.9 18.9 18.6
Tajikistan 7.1 4.0 4.0 7.3 6.3 6.0 –2.6 –5.2 –4.7 ... ... ...
Kyrgyz Republic 4.5 3.3 4.9 3.2 4.5 5.0 –7.8 –13.6 –12.2 7.1 7.0 7.0
Moldova 4.0 3.5 3.8 6.6 4.7 5.1 –4.7 –3.7 –4.7 4.2 4.2 4.1
Memorandum
Caucasus and Central Asia5 4.1 3.7 3.9 9.0 9.1 7.2 –2.5 –2.0 –1.7 ... ... ...
Low-Income CIS Countries6 5.4 4.6 4.7 9.5 13.5 9.6 –1.1 –4.0 –4.2 ... ... ...
Net Energy Exporters Excluding Russia 3.9 3.7 3.8 9.6 9.9 7.6 –2.0 –1.2 –1.0 ... ... ...
1Movements in consumer prices are shown as annual averages. Year-end to year-end changes can be found in Table A7 in the Statistical Appendix.
2Percent of GDP.
4Georgia, Turkmenistan, and Ukraine, which are not members of the Commonwealth of Independent States (CIS), are included in this group for reasons of geography and
similarity in economic structure.

5Caucasus and Central Asia comprises Armenia, Azerbaijan, Georgia, Kazakhstan, the Kyrgyz Republic, Tajikistan, Turkmenistan, and Uzbekistan.
6Low-Income CIS countries comprise Armenia, Georgia, the Kyrgyz Republic, Moldova, Tajikistan, and Uzbekistan.

Annex Table 1.1.5. Middle East, North African Economies, Afghanistan, and Pakistan: Real GDP, Consumer Prices, Current
Account Balance, and Unemployment
2017 2018 2019 2017 2018 2019 2017 2018 2019 2017 2018 2019
Middle East, North Africa, Afghanistan,
and Pakistan 2.6 3.4 3.7 6.3 8.2 6.8 –0.9 0.5 –0.3 ... ... ...
Oil Exporters4 1.7 2.8 3.3 3.4 6.3 5.5 1.2 3.0 1.8 ... ... ...
Saudi Arabia –0.7 1.7 1.9 –0.9 3.7 2.0 2.7 5.4 3.6 ... ... ...
Iran 4.3 4.0 4.0 9.9 12.1 11.5 4.3 7.0 6.3 11.8 11.7 11.6
United Arab Emirates 0.5 2.0 3.0 2.0 4.2 2.5 4.7 5.3 5.1 ... ... ...
Algeria 2.0 3.0 2.7 5.6 7.4 7.6 –12.3 –9.3 –9.7 11.7 11.2 11.8
Iraq –0.8 3.1 4.9 0.1 2.0 2.0 0.7 0.2 –1.6 ... ... ...
Qatar 2.1 2.6 2.7 0.4 3.9 3.5 1.3 2.5 1.8 ... ... ...
Kuwait –2.5 1.3 3.8 1.5 2.5 3.7 2.0 5.8 3.6 1.1 1.1 1.1
Oil Importers5 4.2 4.7 4.6 12.4 12.2 9.5 –6.5 –6.2 –5.7 ... ... ...
Egypt 4.2 5.2 5.5 23.5 20.1 13.0 –6.5 –4.4 –3.9 12.2 11.1 9.7
Pakistan 5.3 5.6 4.7 4.1 5.0 5.2 –4.1 –4.8 –4.4 6.0 6.1 6.1
Morocco 4.2 3.1 4.0 0.8 1.4 2.0 –3.8 –3.6 –3.5 10.2 9.5 9.2
Sudan 3.2 3.7 3.5 32.4 43.5 39.5 –5.5 –6.2 –6.8 19.6 18.6 17.6
Tunisia 1.9 2.4 2.9 5.3 7.0 6.1 –10.1 –9.2 –7.8 15.3 15.0 14.8
Lebanon 1.2 1.5 1.8 4.5 4.3 3.0 –25.0 –25.8 –25.2 ... ... ...
Jordan 2.3 2.5 2.7 3.3 1.5 2.5 –8.7 –8.5 –7.9 ... ... ...
Memorandum
Middle East and North Africa 2.2 3.2 3.6 6.6 8.7 7.1 –0.6 1.1 0.2 ... ... ...
Israel6 3.3 3.3 3.5 0.2 0.7 1.3 3.0 2.6 2.7 4.2 4.2 4.2
Maghreb7 6.4 3.8 3.0 5.4 6.7 6.3 –8.2 –7.8 –7.8 ... ... ...
Mashreq8 3.9 4.8 5.1 20.8 17.8 11.7 –9.7 –8.2 –7.6 ... ... ...
2Percent of GDP.
4Includes Bahrain, Libya, Oman, and Yemen.
5Includes Afghanistan, Djibouti, Mauritania, and Somalia. Excludes Syria because of the uncertain political situation.
6Israel, which is not a member of the economic region, is included for reasons of geography but is not included in the regional aggregates.
7The Maghreb comprises Algeria, Libya, Mauritania, Morocco, and Tunisia.
8The Mashreq comprises Egypt, Jordan, and Lebanon. Syria is excluded because of the uncertain political situation.

Annex Table 1.1.6. Sub-Saharan African Economies: Real GDP, Consumer Prices, Current Account Balance, and Unemployment
2017 2018 2019 2017 2018 2019 2017 2018 2019 2017 2018 2019
Sub-Saharan Africa 2.8 3.4 3.7 11.0 9.5 8.9 –2.6 –2.9 –3.1 ... ... ...
Oil Exporters4 0.4 1.9 2.0 18.3 15.5 14.8 0.2 –0.2 0.0 ... ... ...
Nigeria 0.8 2.1 1.9 16.5 14.0 14.8 2.5 0.5 0.4 16.5 ... ...
Angola 0.7 2.2 2.4 31.7 27.9 17.0 –4.5 –2.2 –0.1 ... ... ...
Gabon 0.8 2.7 3.7 3.0 2.8 2.5 –4.8 –1.5 –1.9 ... ... ...
Chad –3.1 3.5 2.8 –0.9 2.1 2.6 –5.2 –4.3 –5.5 ... ... ...
Republic of Congo –4.6 0.7 4.6 0.5 1.5 1.6 –12.7 3.0 4.8 ... ... ...
Middle-Income Countries5 3.0 3.1 3.5 5.2 5.0 5.0 –2.3 –2.7 –2.9 ... ... ...
South Africa 1.3 1.5 1.7 5.3 5.3 5.3 –2.3 –2.9 –3.1 27.5 27.9 28.3
Ghana 8.4 6.3 7.6 12.4 8.7 8.0 –4.5 –4.1 –4.0 ... ... ...
Côte d’Ivoire 7.8 7.4 7.1 0.8 1.7 2.0 –1.2 –1.5 –1.3 ... ... ...
Cameroon 3.2 4.0 4.5 0.6 1.1 1.3 –2.5 –2.5 –2.4 ... ... ...
Zambia 3.6 4.0 4.5 6.6 8.2 8.0 –3.3 –2.6 –1.9 ... ... ...
Senegal 7.2 7.0 7.0 1.4 1.5 1.5 –9.4 –7.9 –7.5 ... ... ...
Low-Income Countries6 6.0 5.8 6.1 8.9 7.4 6.2 –6.8 –6.7 –7.5 ... ... ...
Ethiopia 10.9 8.5 8.3 9.9 11.2 8.6 –8.1 –6.5 –6.3 ... ... ...
Kenya 4.8 5.5 6.0 8.0 4.8 5.0 –6.4 –6.2 –5.7 ... ... ...
Tanzania 6.0 6.4 6.6 5.3 4.8 5.0 –3.8 –5.4 –6.0 ... ... ...
Uganda 4.5 5.2 5.8 5.6 3.6 4.3 –4.5 –6.9 –9.5 ... ... ...
Madagascar 4.1 5.1 5.6 8.1 7.8 6.8 –3.4 –4.0 –4.8 ... ... ...
Democratic Republic of the Congo 3.4 3.8 4.0 41.5 25.8 13.7 –0.5 0.3 –0.9 ... ... ...
Memorandum
Sub-Saharan Africa Excluding
South Sudan 2.9 3.4 3.7 10.5 9.2 8.6 –2.6 –2.9 –3.1 ... ... ...
1Movements in consumer prices are shown as annual averages. Year-end to year-end changes can be found in Table A7 in the Statistical Appendix.
2Percent of GDP.
4Includes Equatorial Guinea and South Sudan.
5Includes Botswana, Cabo Verde, Lesotho, Mauritius, Namibia, Seychelles, and Swaziland.
6Includes Benin, Burkina Faso, Burundi, the Central African Republic, Comoros, Eritrea, The Gambia, Guinea, Guinea-Bissau, Liberia, Malawi, Mali, Mozambique, Niger,
Rwanda, São Tomé and Príncipe, Sierra Leone, Togo, and Zimbabwe.

Annex Table 1.1.7. Summary of World Real per Capita Output

(Annual percent change; in international currency at purchasing power parity)
Average Projections
2000–09 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2023
World 2.4 4.0 3.0 2.0 2.2 2.3 2.1 1.9 2.4 2.7 2.7 2.5
Advanced Economies 1.1 2.5 1.1 0.7 0.9 1.6 1.7 1.1 1.9 2.0 1.8 1.1
United States 0.8 1.7 0.9 1.5 1.0 1.8 2.1 0.7 1.5 2.1 1.8 0.6
Euro area1 1.0 1.8 1.3 –1.1 –0.5 1.1 1.7 1.5 2.3 2.2 1.9 1.4
Germany 0.9 4.2 3.7 0.5 0.3 1.5 0.6 1.0 2.1 2.4 1.9 1.3
France 0.7 1.5 1.6 –0.3 0.1 0.4 0.6 0.8 1.5 1.6 1.6 1.1
Italy 0.1 1.2 0.2 –3.2 –2.3 –0.3 0.9 1.1 1.6 1.3 1.2 0.8
Spain 1.3 –0.4 –1.4 –3.0 –1.3 1.7 3.5 3.3 3.2 2.9 2.3 1.8
Japan 0.4 4.2 –0.3 1.7 2.2 0.5 1.5 1.0 1.9 1.4 1.2 1.0
United Kingdom 1.1 0.9 0.6 0.8 1.4 2.3 1.5 1.1 1.2 1.0 0.9 1.2
Canada 1.0 1.9 2.1 0.6 1.3 1.7 0.1 0.3 1.7 0.8 1.1 0.7
Other Advanced Economies2 2.6 5.0 2.5 1.3 1.6 2.1 1.3 1.5 2.0 1.9 1.9 1.6
Emerging Market and Developing
Economies 4.4 5.9 4.9 3.7 3.6 3.2 2.8 2.8 3.3 3.6 3.7 3.7
Commonwealth of Independent
States 5.9 4.3 4.9 2.8 2.0 1.4 –2.5 0.0 1.8 1.9 1.8 2.0
Russia 5.7 4.5 5.0 3.6 1.7 0.6 –2.6 –0.3 1.5 1.7 1.5 1.7
CIS excluding Russia 7.0 4.4 5.2 1.9 3.4 2.6 –1.7 1.0 2.9 2.8 2.9 3.2
Emerging and Developing Asia 6.9 8.5 6.7 5.9 5.9 5.8 5.8 5.4 5.5 5.5 5.6 5.4
China 9.6 10.1 9.0 7.4 7.3 6.7 6.4 6.1 6.3 6.0 5.9 5.4
India3 5.2 8.7 5.2 4.1 5.0 6.0 6.8 5.7 5.4 6.0 6.4 6.8
ASEAN-54 3.6 5.5 3.2 4.7 3.7 3.3 3.6 3.7 4.0 4.1 4.1 4.2
Emerging and Developing Europe 3.5 3.7 6.2 2.1 4.3 3.4 4.3 2.8 5.3 3.8 3.2 2.8
Latin America and the Caribbean 1.6 4.7 3.4 1.7 1.8 0.2 –0.9 –1.9 0.1 0.9 1.7 1.8
Brazil 2.1 6.5 3.0 1.0 2.1 –0.4 –4.3 –4.2 0.2 1.5 1.8 1.6
Mexico 0.2 3.8 2.4 2.4 0.2 1.7 2.2 1.8 1.0 1.3 2.1 2.0
Middle East, North Africa,
Afghanistan, and Pakistan 1.9 2.5 3.9 1.0 0.1 –0.1 0.2 2.3 –0.1 1.4 1.7 1.7
Saudi Arabia 0.5 1.3 7.1 2.5 –0.1 1.1 3.3 –0.7 –2.7 –0.3 –0.1 0.3
Sub-Saharan Africa 2.7 4.2 2.4 1.3 2.6 2.4 0.7 –1.2 0.1 0.8 1.0 1.4
Nigeria 5.4 8.3 2.1 1.5 2.6 3.5 –0.1 –4.2 –1.9 –0.6 –0.8 –0.8
South Africa 2.3 1.6 1.8 0.7 1.0 0.3 –0.3 –1.0 –0.3 –0.1 0.1 0.2
Memorandum
European Union 1.4 1.8 1.5 –0.6 0.1 1.6 2.0 1.7 2.4 2.3 2.0 1.6
Low-Income Developing Countries 3.7 5.3 3.5 1.9 3.7 3.7 1.9 0.9 2.3 2.8 3.1 3.2
Note: Data for some countries are based on fiscal years. Please refer to Table F in the Statistical Appendix for a list of economies with exceptional reporting periods.
1Data calculated as the sum of individual euro area countries.
2Excludes the G7 (Canada, France, Germany, Italy, Japan, United Kingdom, United States) and euro area countries.
3See country-specific notes for India in the “Country Notes” section of the Statistical Appendix.

References sentation: Adding Corporate Income and Cash Flow Taxes to

Abdirahman, Mo, Diane Coyle, Richard Heys, and Will Stewart. GIMF.” IMF Working Paper 17/259, International Monetary
2017. “A Comparison of Approaches to Deflating Telecoms Fund, Washington, DC.
Services Output.” Presented at the 5th IMF Statistical Forum. Carton, Benjamin, Joannes Mongardini, and Yiqun Li. 2018. “A
http://www.imf.org/~/media/Files/Conferences/2017-stats New Smartphone for Every Fifth Person on Earth: Quan-
-forum/session-6-heys.ashx?la=en. tifying the New Tech Cycle.” IMF Working Paper 18/22,
Adler, Gustavo, Romain Duval, Davide Furceri, Sinem Kiliç International Monetary Fund, Washington, DC.
Çelik, Ksenia Koloskova, and Marcos Poplawski-Ribeiro. Cerdeiro, D. A., and R. J. Nam. 2018. “A Multidimensional
2017. “Gone with the Headwinds: Global Productivity.” Approach to Trade Policy Indicators.” IMF Working Paper
IMF Staff Discussion Note 17/04. https://www.imf.org/en/ 18/32, International Monetary Fund, Washington, DC.
Publications/Staff-Discussion-Notes/Issues/2017/04/03/Gone Deutsche Bank. 2017. “Lithium 101.” http://www.belmontresources
-with-the-Headwinds-Global-Productivity-44758. .com/LithiumReport.pdf.
Alichi, Ali, and others. Forthcoming. “Estimates of Poten- Gartner, Inc. 2009–18. Press Release (various). https://www.
tial Output and the Neutral Rate for the US Economy.” gartner.com/newsroom/archive/.
IMF Working Paper, International Monetary Fund, Gruss, Bertrand. 2014. “After the Boom—Commodity Prices
Washington, DC. and Economic Growth in Latin America and the Caribbean.”
Baldwin, Richard. 2016. “The Great Convergence: Information IMF Working Paper 14/154, International Monetary Fund,
Technology and the New Globalization.” Cambridge, MA, Washington, DC.
Belknap Press. GSM Association. 2018. “The Mobile Economy 2018.” https://
Berger, Helge, Thomas Dowling, Sergi Lanau, Mico Mrkaic, Pau www.gsma.com/mobileeconomy .
Rabanal, and Marzie Taheri Sanjani. 2015. “Steady as She Guvenen, Fatih, Raymond J. Mataloni Jr., Dylan G. Rassier, and
Goes—Estimating Potential Output during Financial ‘Booms Kim J. Ruhl. 2017. “Offshore Profit Shifting and Domestic
and Busts.’” IMF Working Paper 15/233, International Mon- Productivity Measurement.” NBER Working Paper 23324.
etary Fund, Washington, DC. www.nber.org/papers/w23324.
Blagrave, Patrick, Roberto Garcia-Saltos, Douglas Laxton, and IDC Research, Inc. 2009–18. Press Release (various). https://
Fan Zhang. 2015. “A Simple Multivariate Filter for Estimat- www.idc.com/.
ing Potential Output.” IMF Working Paper 15/79, Interna- International Data Corporation (IDC). 2018. “Apple Passes
tional Monetary Fund, Washington, DC. Samsung to Capture the Top Position in the Worldwide
Borio, Claudio, Piti Disyatat, and Mikael Juselius. 2013. Smartphone Market while Overall Shipments Decline 6.3%
“Rethinking Potential Output: Embedding Information in the Fourth Quarter, According to IDC.” February 1.
about the Financial Cycle.” BIS Working Paper 404, Bank for https://www.idc.com/getdoc.jsp?containerId=p
rUS43548018.
International Settlements, Basel. International Monetary Fund (IMF). 2010. “Reference Note
Byrne, David, and Carol Corrado. 2017. “ICT Prices and ICT on Trade Policy, Preferential Trade Agreements, and WTO
Services: What Do They Tell Us about Productivity and Consistency.” Washington, DC, September 2010.
Technology?” Finance and Economics Discussion Series Jacks, David S. 2013. “From Boom to Bust: A Typology of Real
2017–15, Board of Governors of the Federal Reserve System, Commodity Prices in the Long Run.” NBER Working Paper
Washington, DC. https://doi.org/10.17016/FEDS.2017.015. 18874, National Bureau of Economic Research, Cambridge, MA.
Byrne, David, John Fernald, and Marshall Reinsdorf. 2016. Nakamura, Leonard, Jon Samuels, and Rachel Soloveichik.
“Does the United States Have a Productivity Slowdown or a 2017. “Measuring the ‘Free’ Digital Economy within the
Measurement Problem?” Brookings Papers on Economic Activity GDP and Productivity Accounts.” http://www.imf.org/
(Spring): 109–57. https://www.brookings.edu/wp-content/ ~/media/Files/Conferences/2017-stats-forum/session-1
uploads/2016/03/byrnetextspring16bpea.pdf. -oloveichik.ashx?la=en.
Caldara, Dario, and Matteo Iacoviello. 2017. “Measuring Olivetti, Elsa A., Gerbrand Ceder, Gabrielle G. Gaustad,
Geopolitical Risk.” Working Paper, Board of Governors of the and Xinkai Fu. 2017. “Lithium-Ion Battery Supply Chain
Federal Reserve Board, Washington, DC. Considerations: Analysis of Potential Bottlenecks in Critical
Carney, Mark. 2017. “[De]globalization and Inflation.” Metals.” Joule 1 (2): 229–43.
2017 IMF Michel Camdessus Central Banking Lecture, Reinsdorf, Marshall, and Paul Schreyer. 2017. “Measuring Con-
September 18. sumer Inflation in a Digital Economy.” http://www.imf.org/~/
Carton, Benjamin, Emilio Fernandez-Corugedo, and Benjamin media/Files/Conferences/2017-stats-forum/session-1-schreyer
L. Hunt. 2017. “No Business Taxation without Model Repre- -and-reinsdorf.ashx?la=en.

2
CHAPTER
LABOR FORCE PARTICIPATION IN ADVANCED ECONOMIES:
DRIVERS AND PROSPECTS
Despite the acceleration in population aging in almost Introduction

all advanced economies over the past decade, aggregate Population growth in advanced economies is
labor force participation rates show divergent trajectories. slowing, life expectancy is rising, and the number of
Headline numbers also hide strikingly different shifts in elderly people is soaring. As these trends gather steam,
the labor force attachment of different groups of workers: the United Nations projects that by the middle of this
participation has increased among prime-age women and, century, total population will be shrinking in almost
more recently, older workers, but it has fallen among the half of advanced economies (Figure 2.1). The burden
young and among prime-age men. This chapter finds that will fall on those currently considered to be of working
aging and the drag from the global financial crisis can age, who in a few decades will support close to double
explain a significant share of the decline in the participa- the number of elderly people they do now. Unless
tion of men during the past decade. However, the rising more people participate in labor markets, aging could
participation of women underscores the importance of slow advanced economies’ growth and, in many cases,
other factors in shaping participation decisions. The anal- undermine the sustainability of their social security
ysis suggests that labor market policies and institutions, systems (Clements and others 2015). Increases in labor
together with structural changes and gains in educational supply accounted for a significant share of advanced
attainment, account for the bulk of the dramatic increase economies’ potential growth during 1985–2000, but
in the labor force attachment of prime-age women and their contribution has since fallen (Chapter 3 of the
older workers in the past three decades. At the same April 2015 World Economic Outlook [WEO]).
time, technological advances such as automation, while Even though population aging is already exerting
beneficial for the economy as a whole, have weighed pressure on labor supply, considerable differences in
moderately on participation rates. These findings highlight the evolution of overall labor force participation are
the considerable scope for policies to counteract the forces evident throughout the world’s advanced economies
of aging by enabling those who are willing to work to do (Figure 2.2).1 In half of those economies, the aggre-
so. Investing in education and training, reforming the tax gate participation rate has actually increased since the
system, and reducing incentives to retire early—along with global financial crisis of a decade ago, which coincided
stronger policies that improve the job-matching process with an acceleration of the demographic transition.
and help workers combine family and work life—can Headline numbers also hide stark differences in the
encourage people to join and remain in the workforce. participation rates of different groups of workers. For
Ultimately, however, the dramatic shifts in demographic example, male participation has declined almost every-
structure could overwhelm the ability of policies to where, while female participation has increased (see
mitigate the effects of aging on labor force participa- Box 1.1 of the October 2017 WEO).
tion, which underscores the need to rethink migration What underlies these strikingly divergent trajecto-
policies to boost labor supply in advanced economies. ries across countries and for different workers? Various
forces are likely at play. Differences in the exact timing
and pace of the demographic transition may explain
some of the divergence. However, the disparity in
participation trends across specific groups of workers
The main authors of this chapter are Francesco Grigoli, Zsóka
suggests a potentially important role for policies and
Kóczán, and Petia Topalova (lead), with support from Benjamin
Hilgenstock, Christopher Johns, and Jungjin Lee and contributions institutions that influence people’s decisions to join,
from John Bluedorn, Benjamin Hilgenstock, and Davide Malacrino.
We are grateful to Mitali Das, Romain Duval, and Davide Furceri 1The labor force participation rate is the fraction of the adult
for sharing their data on routinization and labor market policies. population (ages 15 and over) either working or looking for work. In
The chapter benefited from comments and suggestions by Stepha- this chapter, labor force participation and workforce attachment are
nie Aaronson. used interchangeably.

Figure 2.1. Demographic Transition: Recent Trends and ysis in this chapter suggests that strengthening specific
Projections groups of workers’ attachment to the labor force has
helped many countries defuse the effects of an aging
Population growth is slowing in both advanced and emerging market and population on aggregate labor force participation.
developing economies. In advanced economies, the number of elderly is rising
precipitously relative to the working-age population, a process that accelerated Accordingly, this chapter addresses the follow-
significantly after 2008. ing questions:
•• How have labor force participation rates evolved
Total Population Growth across advanced economies? Do the dynamics differ
(Percent)
systematically by worker characteristics? Have trends
in participation changed in the aftermath of the
4.0 1. Advanced Economies 2. Emerging Market and 4.0
Developing Economies global financial crisis?
3.5 3.5
3.0 3.0
•• What are the key drivers of changes in aggregate par-
2.5 2.5 ticipation rates and the attachment of various groups
2.0 2.0 of workers to the labor force? More specifically:
1.5 1.5 oo How much of the changes seen in aggregate rates
1.0 1.0 in the past decade can be attributed to the accel-
0.5 0.5 eration in demographic shifts and cyclical effects,
0.0 0.0 including the severe recessions associated with the
–0.5 –0.5 global financial and European debt crises?
1960 80 2000 20 40 50 1960 80 2000 20 40 50
oo Historically, what has been the role of policies
Population Ages 65 and Older and institutions that shape individuals’ decisions
(Percent of population ages 20–64) to work, compared with forces that may have
shifted the demand for certain types of workers,
80 3. Advanced Economies 4. Emerging Market and 80 such as automation and structural transformation?
Developing Economies
70 70 •• What are the prospects for labor force participation?
60 60
50 50 The chapter starts by taking stock of the changes
40 40 in the labor force participation of different groups of
30 30 workers in advanced economies over the past three
20 20 decades. Several considerations justify the focus
10 10 on participation. First, the availability of factors of
0 0 production, including labor, is an important deter-
1960 80 2000 20 40 50 1960 80 2000 20 40 50 minant of actual and potential growth in the long
term. The participation rate, together with popula-
Sources: United Nations; and IMF staff estimates.
Note: Solid lines show median; and blue shaded areas show interquartile range. tion growth, is the key determinant of labor supply.
Gray shaded areas indicate projections. Dashed vertical lines in panels 3–4 show Second, labor force participation data have good
year 2008.
coverage geographically and temporally, by gender
and age group, and capture the pent-up supply of
remain in, or reenter the labor force. Differences people who work part time but want to work full
in exposure and resilience to global forces such as time and those currently unemployed but willing to
technological advances and trade may have depressed work (Burniaux, Duval, and Jaumotte 2004). Finally,
long-term demand for workers with certain skill sets.2 economic theory provides clear guidance for the
Identifying and ranking the key drivers of participa- evolution of people’s labor force participation over
tion across population groups is necessary in designing the course of their lives. The chapter then uses com-
policies that could enable those willing to work to do plementary analytical approaches to assess the forces
so and counteract the forces of aging. Indeed, the anal- shaping participation trends. The first part quantifies
the change in country-level participation rates that
2See, for example, Acemoglu and Autor (2011); Autor and Dorn
would be consistent with demographic shifts since the
(2013); Goos, Manning, and Salomons (2014); Autor, Dorn, and
Hanson (2016); Chapter 3 of the April 2016 WEO; and Chapter 2 mid-2000s, when aging accelerated significantly in
of the October 2017 WEO. many advanced economies. The second part assesses

CHAPTER 2 Labor Force Participation in Advanced Economies: Drivers and Prospects
Figure 2.2. Change in Labor Force Participation Rates, 2008–16

(Percentage points)
Changes in aggregate participation rates between 2008 and 2016 show considerable heterogeneity across advanced economies, with participation rates of men
(women) generally decreasing (increasing).
1. Aggregate 2. Men 3. Women

DEU SWE ISR
CZE KOR DEU
ISR CZE ESP
KOR DEU GRC
AUT ISR CZE
SWE AUT AUT
ITA CHE KOR
CHE ITA JPN
GBR FRA ITA
JPN GBR CHE
FRA AUS GBR
BEL BEL SWE
AUS JPN BEL
GRC CAN AUS
ESP NLD FRA
NLD DNK NLD
CAN USA CAN
DNK GRC DNK
USA ESP PRT
PRT PRT USA
–4 –3 –2 –1 0 1 2 –5 –4 –3 –2 –1 0 1 –3 –2 –1 0 1 2 3 4 5
Sources: Organisation for Economic Co-operation and Development; and IMF staff calculations.
Note: The panels show the 20 largest advanced economies by 2017 total population. Data labels in the figure use International Organization for Standardization (ISO)
country codes.
in more detail the drivers of participation among •• Aging and the drag from the global financial crisis
specific groups of workers by examining differences can explain a significant share of the decline in the
in these trends across countries and over time. The aggregate participation rate of men during the past
third part hones in on the predictors of individuals’ decade. However, the rise in the participation rate
participation decisions, shedding light on the role of of women, even as women’s average ages increased
such characteristics as education, family composition, and despite unfavorable cyclical developments,
and exposure to technological advances. Finally, the underscores the important role of policies and other
long-term prospects for labor force participation are factors in shaping labor supply decisions and miti-
evaluated using a cohort‑based model. gating the effect of aging.
These are the chapter’s main findings: •• The analysis suggests that policies and institutions,
•• Although aggregate labor force participation rates such as the tax-benefit system, public spending on
in advanced economies show divergent trajectories, active labor market programs, and policies targeted
surprisingly similar trends emerge across countries to encourage specific groups to participate, together
for specific groups of workers. The magnitude of with structural changes and gains in educational
change varies from country to country, but partic- attainment, account for the bulk of the dramatic
ipation by women has increased dramatically since increase in the labor force attachment of prime-age
the mid-1980s. More recently, participation has women and older workers in the past three decades.
picked up considerably among older workers and •• On the other hand, technological advances, such as
has fallen among the young. In almost all advanced routinization—the automation of tasks for which
economies, prime-age men (ages 25–54), particu- labor can be easily substituted by capital—have
larly those with lower educational attainment, have weighed on the participation rates of most groups
become increasingly detached from the labor force of workers. The decrease in the relative price of
over the past 35 years, although participation rates investment is associated with larger declines in
are still high and vary little across countries. participation in countries that are more exposed to

routinization because of the mix of their workers’ in aggregate participation. Unless technology delivers
occupations, which may partially explain lower offsetting productivity gains, these findings highlight the
prime-age male participation. need for many advanced economies to rethink immi-
•• While analysis of micro data confirms the signifi- gration policies to boost their labor supply, alongside
cant impact of exposure to routinization on people’s policies to encourage older workers to postpone retire-
detachment from the labor force, it also suggests ment. Although receiving migrants can pose challenges,
that policy efforts aimed at enhancing connective potentially prompting a political backlash, it can also be
networks in labor markets can partially offset this a boon for host countries. The chapter analysis suggests
effect. Higher spending on active labor market that net migration accounts for roughly half of the
programs and education is associated with a lower population growth in advanced economies over the past
likelihood that a person previously employed in a three decades. Migration can relieve the strain of pop-
routinizable occupation will drop out of the labor ulation aging and contribute to other long-term gains,
force. This likelihood is also significantly lower in such as higher growth and productivity, documented
urban areas, pointing to the importance of access to elsewhere (see Chapter 4 of the October 2016 WEO).
diverse pools of jobs. It is important to emphasize from the outset that this
chapter seeks to identify patterns and correlations rather
The findings in the chapter suggest that many coun- than to establish causality between various policies,
tries have so far successfully counteracted the negative structural characteristics, and individual characteris-
forces of aging on aggregate labor force participation tics on one hand and labor force participation on the
by strengthening the attachment of specific groups of other. Many of the variables explored when looking at
workers to the labor force. Policies that reduce disin- individuals—including choices about education, mar-
centives for joining or remaining in the labor force and riage, and fertility—coincide with decisions about par-
policies that help workers combine family and work ticipating in the labor force. Changes in national labor
life can broaden these gains by enabling people who market policies and institutions may also reflect the
are willing to work to do so.3 Further investment in evolution of societal and cultural attitudes toward work
education, training, and activation policies can not only that influence observed trends in labor supply beyond
encourage individuals to be active in the labor market their impact on policies.4 Sorting out these possibili-
but also make the workforce more resilient to global ties is beyond the scope of this chapter, which aims to
developments, such as technological progress or global- present a rich description of the patterns of labor force
ization, that may obviate the need for certain skills. participation across countries and over time and their
Ultimately, however, dramatic shifts in demographic association with a broad set of drivers, thus offering
structure projected in advanced economies could over- valuable guidance on potential areas for policy action.
whelm the ability of policies to offset the forces of aging.
The chapter’s illustrative simulations suggest that aggre-
gate participation will eventually decline—even if gender Patterns of Labor Force Participation in
gaps are fully closed—and that the participation of Advanced Economies
older workers must rise significantly to stem the decline An investigation into the long-term trends of aggre-
gate labor force participation and the workforce attach-
3Beyond the obvious contribution to potential output from an ment of individual groups of workers in advanced
increase in the labor supply, higher female labor force participation economies reveals several striking patterns.5
has been shown to bring about other macroeconomic benefits, such
as greater economic diversification (Kazandjian and others 2016),
lower inequality (Gonzales and others 2015b), and growth that 4For example, the evolution of social norms toward more egali-
is less sensitive to inequality (Grigoli and Robles 2017). Greater tarian gender roles may induce both family legislation and higher
representation of women in senior corporate positions is associated female labor force participation. Female labor supply shifts may also
with higher firm profitability (Christiansen and others 2016a), while create political support for more family-friendly policies, leading
appointing more women to bank supervisory boards is correlated simultaneously to higher female employment and greater parental
with enhanced bank stability and financial sector resilience (Sahay leave rights (Olivetti and Petrongolo 2017).
and others 2017). Evidence also suggests that later-life employment 5The discussion of the long-term trends is based on the analysis of
improves nonfinancial outcomes, such as a person’s sense of identity, participation rates in 21 advanced economies for which 1985–2016
social integration, and support (Erikson, Erikson, and Kivnick 1986; data are available to ensure sample consistency. The patterns
Cohen 2004), as well as emotional and physical well-being (for described are qualitatively identical if all advanced economies are
example, Cohen 2004; and Calvo 2006). included in the analysis.

Aggregate Participation Rates Figure 2.3. Labor Force Participation Rates by Gender and Age
(Percent, unless noted otherwise)
Over the past 30 years, the aggregate average labor
force participation rate in advanced economies as a Trends in participation rates in advanced economies also differ dramatically by
group has barely changed (Figure 2.3, panel 1). How- gender and age groups.
ever, the group aggregate masks significant differences
80 1. Aggregate 2. Distribution of Changes in 9
in the experience of individual countries. While in Aggregate Participation
a large share of advanced economies aggregate labor 75 Rates, 1985–2016
force participation in 2016 was within a couple of per- 70 (Number) 6
centage points of what it was in 1985, several countries 65
saw very significant increases in the workforce attach- 60 3
ment of their populations, with aggregate participation
55
rates gaining more than 5 percentage points in such
50 0
countries as Germany, Korea, the Netherlands, and 1985 90 95 2000 05 10 16 < = –4 (–2; 0] (2; 4] (6; 8]
Spain (Figure 2.3, panel 2). Moreover, there has been a (–4; –2] (0; 2] (4; 6] >8
remarkable narrowing of the distribution of participa-
80 3. Men 4. Women 80
tion rates across advanced economies.
70 70
Participation of Specific Groups of Workers 60 60
Even more striking is the divergence in the trends in 50 50

labor force participation of different groups of workers 40 40
(Figure 2.3, panels 3–8). Across advanced economies,
30 30
the share of women who are employed or actively 1985 90 95 2000 05 10 16 1985 90 95 2000 05 10 16
looking for work has increased dramatically. For the
median advanced economy, the female labor force 70 5. All, Ages 15–24 6. All, Ages 25–54 100
participation rate has increased by close to 10 per- 90
centage points. Gains in female participation were 60
80
substantially larger in countries where women were
50 70
historically less likely to be part of the workforce (see
60
Annex Figure 2.2.1), a convergence that has signifi- 40
cantly narrowed the dispersion in women’s participa- 50
tion across advanced economies since 1985. The rise 30 40

1985 90 95 2000 05 10 16 1985 90 95 2000 05 10 16
in women’s labor force participation is also consistent
with the increasing share of two‑earner households (see 80 7. All, Ages 55–64 8. All, Ages 65 and Older 30
Annex Figure 2.2.2). Conversely, participation rates
70 25
of men, which are significantly higher and tend to be
60 20
much more similar across countries, have come down
almost across the board. For the median advanced 50 15
economy, the participation rate among men was more 40 10

than 4 percentage points lower in 2016 than in 1985. 30 5
These divergent trends have narrowed gender gaps. 20 0
Significant differences also exist in how participation 1985 90 95 2000 05 10 16 1985 90 95 2000 05 10 16
rates have evolved across individuals of different ages

Sources: Organisation for Economic Co-operation and Development; and IMF staff
(Figure 2.3, panels 5–8). The young (between ages 15 calculations.
and 24) were significantly less likely to be part of the Note: In panels 1, 3, and 4, lines show median; shaded areas show interquartile
range; and dotted lines denote population-weighted average. In panels 5–8, black
labor force in 2016 than in 1985, with similar trends lines show median; gray shaded areas show interquartile range; and black dotted
observed for men and women (see Box 2.1 for trends lines show population-weighted average for total age group, respectively. Blue and
in youth labor force participation across advanced red lines show median for men and women, respectively. In panel 6, dotted blue
and red lines show interquartile range for men and women, respectively. Figure is
and emerging market and developing economies). To based on a balanced panel of 21 advanced economies.
a significant extent, declining labor force attachment

Figure 2.4. Labor Force Participation and School Enrollment reflects the secular trend toward greater investment
of the Young in human capital and higher school enrollment rates
(Figure 2.4).6 In fact, the share of “idle” youth—defined
Labor force participation of the young (ages 15–24) in advanced economies is as those who are neither employed, unemployed, nor
falling, while their school enrollment is rising.
enrolled in school—is quite small and has been stable
25 1. Changes in School Enrollment and Labor Force since the early 2000s.7 Given the increase in the returns
20 Participation of the Young, 2000–12 to schooling in many advanced economies, the decline
(Percentage points)
15 in labor force participation among the young could in
Change in labor force
y = –0.58**x – 0.81
participation rate
10 part reflect an expected response to economic incentives

5 (Krueger 2017). Indeed, there is a negative correlation
0 between changes in youth labor force participation and
–5
changes in returns to tertiary education relative to the
–10
returns to primary education across countries.8
–15
At the same time, participation rates of older men
–20
–5 0 5 10 15 20 25 and women (ages 55 and older) have increased sig-
Change in school enrollment rate nificantly since the mid-1990s, following decades of
steady decline.9 The increase is particularly pronounced
5 2. Changes in Returns to Education and Labor Force
Participation of the Young, 1987–2013
for the 55–64 age group, but in the past decade, even
4
individuals older than 65 have been remaining in the
labor force participation rate
(Percentage points)
Annual average change in
3
2 labor force longer.10 The gains in participation among
1 older workers should be viewed in the context of
0 significantly longer lives. Life expectancy at birth has
–1
increased by about seven years, and at age 50 by more
–2
–3 than five years, since 1985, prompting many countries
y = –2.15*x – 0.40**
–4 to adopt policies to encourage longer working lives
–5 through later retirement.11
–0.5 –0.4 –0.3 –0.2 –0.1 0.0 0.1 0.2 0.3 0.4 0.5
Annual average change in returns to education Among prime-age workers, the most notable pattern
is diverging trends of the labor force attachment of
3. Distribution of the Young by Activity Status, 2000–16
(Percent) 6While some in this age group are in school and in the labor
Labor force Idle Student force, there is a significant association between increasing enrollment
100 rates and declining participation rates across countries. See Canon,
90 Kudlyak, and Liu (2015) for evidence from the United States.
80 7The concept of idle youth is distinct from that of NEETs
70 (defined as those not in employment, education, or training), given

60 that the latter includes unemployed individuals. Youth unemploy-
50 ment increased and remains high since the global financial crisis in
40 many advanced economies (Banerji, Lin, and Saksonovs 2015).
8Returns to education are proxied by the ratio of the average labor
30
20 income of prime-age men with higher education relative to the aver-
age labor income of prime-age men with only primary education and
10
are computed from the Luxembourg Income Study Database during
0
2000 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 1987–2013. More recent evidence suggests that skill premiums have
stagnated or marginally declined during the past decade across most
advanced economies (see Box 2.1 of the October 2017 WEO).
Sources: Eurostat, European Union Labour Force Survey; Luxembourg Income 9For a discussion of earlier trends in retirement, see Blöndal and
Study Database; Organisation for Economic Co-operation and Development; and
Scarpetta (1999), Gruber and Wise (1999), and OECD (2001).
IMF staff calculations. 10For men, the observed increase in workforce attachment at older
Note: Markers in panel 2 refer to annualized changes between year pairs. Intervals
can be of different lengths due to limited data availability. Reported statistics in ages reflects reduced retirement rates (higher participation among
panel 3 are estimated from the European Union Labour Force Survey at the the 55 and older age group) amid stable or slightly declining labor
country level over the period 2000–16. The panel reports the youth force participation at younger ages (those below 55). For women, the
population-weighted average across countries. In panel 3, “idle” youth includes observed increase is the result of a growing pool of working women
those who are neither employed, unemployed, nor enrolled in school. reaching those ages, as well as changes in retirement behavior.
11Gains in life expectancy have been generally accompanied by
increases in healthy life expectancy as documented by Salomon and

others (2012).

men versus women, as discussed. The small decline in Figure 2.5. Labor Force Participation Rates of Prime-Age Men
participation rates of prime‑age men, which remains and Women by Demographic Characteristics, 2000 and 2016
(Percent)
very high and varies little across countries, has been
more than offset by the dramatic entry of prime-age Women’s participation has increased almost across the board in advanced
women into the labor force, leading to overall gains in economies, while men’s participation has stagnated or declined, especially for the
less educated.
the participation rates of prime-age workers in most
advanced economies. While it is possible that higher 100 1. Men by Educational 2. Men by Marital Status 100
female participation has allowed some men to drop out Attainment
95 2000
of the labor force, there is little evidence to that effect 2016 95
at the country level. Correlations between changes in 90
90
prime-age female and male participation rates are, if 85
anything, positive (see Annex Figure 2.2.3), and partic- 85
80
ipation of married men has declined less than partici-
pation of single men (Figure 2.5, panel 2).12 75 80
Primary Secondary Tertiary Single Married
Because labor force participation patterns could
100 3. Men by Family 4. Men by Immigration 100
reflect significant shifts in the characteristics of Composition Status
prime-age populations—such as education, fertility, 98
95
marriage, and immigration status—Figure 2.5 provides 96
a more granular picture of the changes in the partic- 90
ipation of subgroups since 2000 for most advanced 94
economies (panels 1 and 5) and advanced European 85

92
economies (panels 2–4, 6–8).13 With the notable
90 80
exception of relatively less-educated women, the rise No Younger Older Native Immigrant
in female labor force participation has been remark- children children children
ably widespread. Across Europe, single and married 100 5. Women by Educational 6. Women by Marital Status 100
Attainment 2000
women, those with young children (below the age 90 2016 90
of 6) or older children (below the age of 15), natives
80 80
and immigrants, were significantly more likely to be
employed or looking for work in 2016 than in 2000. 70 70
For prime-age men, the decline in participation has 60 60
been the deepest for those with the lowest educational
50 50
attainment. Across all remaining groups, there has been Primary Secondary Tertiary Single Married
a small decline or stagnation in the median advanced 100 7. Women by Family 8. Women by Immigration 100
economy. This suggests that changes in population Composition Status
90 90
characteristics toward groups with lower participation,
such as the falling share of married prime-age men, 80 80
70 70
12The Council of Economic Advisers (2016) similarly finds
60 60
limited evidence that reliance on spousal income has contributed sig-
nificantly to the decline in prime-age male labor force participation 50 50
in the United States. Rising participation among prime-age women No Younger Older Native Immigrant
may be driven by falling household income; although this is difficult children children children
to examine in country-level analysis due to endogeneity concerns,
this issue is examined in greater detail when looking at people’s Sources: Eurostat, European Union Labour Force Survey; national authorities; and
decision to participate. IMF staff calculations.
13Data availability constraints allow analysis on participation by Note: Bars show median and lines show interquartile range. Panels 1 and 5 are
based on data from most advanced economies, while panels 2–4 and 6–8 are
various demographic characteristics only for a significantly shorter based on data from advanced European economies. Panels 3 and 7 report statistics
time span and a smaller sample of countries. The analysis relies on for married individuals. In panels 4 and 8, dark bars show data for 2004 instead of
individual-level data from the European Union Labour Force Survey 2000. Prime age is defined as 25–54. Young children are those below the age of
to construct country-level participation rates for subgroups of workers 6; older children are those ages 6–14. Level of educational attainment is defined
by marital status, number of children, and immigration status, and on according to the International Standard Classification of Education (ISCED).
Eurostat data, complemented with data from national authorities, to Primary education contains ISCED 2011 levels 0–2; secondary education contains
build a picture of participation by educational attainment. ISCED 2011 levels 3–4; and tertiary education contains ISCED 2011 levels 5–8.

Figure 2.6. Subgroups of the Inactive, 2000 and 2016 main income earners for their families, even a small
(Percent) decline in their labor supply could have sizable mac-
roeconomic consequences.14 Finally, detachment from
Nonparticipants consist of very different subgroups, including students, retired,
those taking care of children (“voluntarily inactive”), as well as those inactive for the labor force during a person’s peak productive years
economic reasons (“involuntarily inactive”). is associated with lower happiness and life satisfaction
for men (Winkelmann and Winkelmann 1995; Lucas
Temporary contract ended Illness or disability and others 2004; Knabe and Ratzel 2011; Krueger
Dismissed Retired
Other Students
2017), poorer health and higher mortality (Gerdtham
Family or childcare Have never worked and Johannesson 2003; Eliason and Storrie 2009; Sul-
livan and von Wachter 2009), and depressed employ-
1. Men, Ages 25–54 2. Women, Ages 25–54 ment prospects (Arulampalam, Booth, and Taylor
100 100
2000; Arulampalam, Gregg, and Gregory 2001).
80 80
60 60 The Nonparticipants
Interesting insights can be gleaned from the reasons
40 40
workers give for being out of the labor force. Fig-
20 20 ure 2.6 uses data from millions of workers surveyed
across 24 countries in Europe to break down non-
0 0 participants into those who are students, retired, not
2000 2016 2000 2016
retired but have never worked, and were previously but
3. All, Ages 15–24 4. All, Ages 55 and Older are no longer employed. It further breaks down the last
100 100
group of nonparticipants according to the reason they
80 80 reported for their detachment from the labor force.
Comparing the responses of prime-age men and
60 60 prime-age women points to important gender differ-
ences in reasons for inactivity: for instance, women are
40 40
still more likely to drop out of the labor force to look
20 20 after children, while a higher fraction of men report
illness and disability as reasons for not being employed.
0 0 The responses also suggest that a nontrivial share
2000 2016 2000 2016
of those out of the labor force may be “involuntarily
Sources: Eurostat, European Union Labour Force Survey; and IMF staff inactive”: they used to work but stopped as a result of
calculations. economic (demand-side) factors, rather than because
Note: Reported statistics are estimated from a random sample of 10,000
respondents per country per year from the European Union Labour Force Survey.
of a personal decision. Those reporting being dismissed
Categories in blue and red are subgroups of those who have worked before and from their previous job as a reason for inactivity can be
are not retired. Retired includes early retirement. seen as a lower bound for this group.15
have been sizable. The United States stands out, with 14In 2015, the composition of the labor force of the average
particularly deep declines in participation for both advanced economy was as follows: 37 percent of workers were
prime-age men, 31 percent were prime-age women, 11 percent were
women and men in the prime-age category across all ages 15–24, and 21 percent were older than 55. The composition
levels of educational attainment. of the population of the average advanced economy was as follows:
Although the decline in labor force participation of 20 percent were prime-age men, 20 percent were prime-age women,
12 percent were ages 15–24, and 31 percent were older than 55.
prime-age men appears small for the median advanced 15In line with the stylized facts already discussed, comparing the
economy, it is worrisome for several reasons. First, years 2000 and 2016 suggests that, over time, the share of students
the decline is broad-based, occurring in almost all increased, among both the young and those of prime working age,
advanced economies. Second, given that prime-age while the share of those in (early) retirement among prime-agers
fell, as did the share of those who never worked among prime-age
men are still the largest segment of the labor force in women and those 55 and older. Illness and disability became rela-
advanced economies and have traditionally been the tively more important over time as a reason for nonparticipation.

Involuntary nonparticipants drop out disproportion- Figure 2.7. The Role of Exposure to Routinization
ately from certain sectors of the economy (Figure 2.7,
panel 1). Wholesale and retail trade, manufacturing, The involuntarily inactive drop out disproportionately from highly routinizable
sectors and occupations. Vulnerability to routinization is especially pronounced for
mining and quarrying, and utilities together account men in the middle of the income distribution.
for more than half of the involuntarily inactive, even
though fewer than one-third of active workers (includ- 35 1. Activity and Inactivity 2. Involuntary Inactivity 1.0
by Sector and Routine Exposure
Exposure to routinization (index)

ing the employed and unemployed) are attached to 30 (Percent)
these sectors. Excess involuntary inactivity—measured CON MNG MNF ELC 0.5
25 Active FIN INF ACC
as the difference between the inactive individuals Involuntarily inactive PUB
20 WAT TRD
attached to a sector as a share of all nonparticipants and 0.0
15 REA TRA
the active workers attached to the same sector as a share HEA OTH
of the labor force—tends to be concentrated precisely 10 –0.5
EDU
in sectors that have a greater share of routine jobs that 5
AGR
are vulnerable to automation (Figure 2.7, panel 2). 0 –1.0
These stylized facts provide suggestive evidence –4 –2 0 2 4 6 8
AGR
MNG
MNF
CON
TRD
ACC
INF
ADM
TRA
“Excess” involuntary inactive
of the potential harm of technological progress to (percentage points)
participation rates of certain types of workers, an
issue this chapter examines in greater detail. They also 40 3. Activity and Inactivity 4. Average Routine 1.0
highlight potentially important income distributional by Occupation Exposure of Occupation
(Percent) by Gender and Income
consequences of involuntary inactivity. Displacement 30 Active (Index)
Involuntarily 0.5
of workers tends to occur disproportionately among inactive
lower- and middle-skill occupations (Figure 2.7, panel 20
3), and vulnerability to routinization is especially pro-
0.0
nounced in the middle and lower parts of the income 10 Men
distribution (Figure 2.7, panel 4). Women
0 –0.5
1 2 3 4 5 6 7 8 9 10
MAN
PROF
TECH
CLER
CRAFT
MACH
ELEM
AGRIC
Participation after the Global Financial Crisis SERV Income decile
The extent to which trends in labor force participa- High Middle Low
skill skill skill
tion have changed since the global financial crisis varies
depending on the groups of workers considered (Fig- Sources: Das and Hilgenstock (forthcoming); Eurostat, European Union Labour
ure 2.8). For young and older workers, there is little dif- Force Survey; and IMF staff calculations.
ference in the trends in participation rates for the median respondents per country per year from the European Union Labour Force Survey
economy. However, the decline in participation acceler- over the period 2000–16. In panels 1 and 3, active includes employed and
ated for prime-age men, and the rate at which prime-age unemployed, and involuntarily inactive refers to people inactive due to dismissal.
For the inactive, sector or occupation is that of last employment. In panel 2,
women joined the labor force fell after 2008. It is diffi- “excess” involuntarily inactive refers to the difference between inactive individuals
cult, however, to isolate the effect of the crisis from the in a sector as a share of all nonparticipants and the active individuals attached to
the sector as a share of the labor force. In panel 2, routine exposure is a proxy for
steady decline in the gains in women’s participation over the share of jobs in a given sector that are at risk of being automated based on
the past three decades. These patterns are broadly similar Das and Hilgenstock (forthcoming). Panel 4 in turn shows how automatable given
occupations are. ACC = accommodation and food service activities; ADM =
in countries that experienced relatively large output losses administrative and support service activities; AGR = agriculture, forestry, and
as a result of the global financial and European debt fishing; AGRIC = skilled agricultural workers; CLER = clerical workers; CON =
construction; CRAFT = craft workers; EDU = education; ELC = electricity, gas,
crises and those that were relatively shielded from their steam, and air-conditioning supply; ELEM = elementary occupations;
adverse effects (see Annex Figure 2.2.4). FIN = financial and insurance activities; HEA = human health and social work
Employment rates increased in most advanced econ- activities; INF = information and communication; MACH = plant and machine
operators; MAN = managers; MNF = manufacturing; MNG = mining
omies before the global financial crisis, but have since and quarrying; OTH = other services; PROF = professionals; PUB = public
declined in over half of them. Figure 2.9 decomposes administration and defense; REA = real estate activities; SERV = sales and service
workers; TECH = technicians; TRA = transportation and storage; TRD = wholesale
changes in employment into changes in unemploy- and retail trade; WAT = water supply, sewerage, waste management, and
ment and participation and shows that, before the remediation activities.
crisis, employment gains were matched by unemploy-
ment declines and increases in participation in most

Figure 2.8. Average Annual Changes in Labor Force ulation and changes in the labor force attachment of
Participation Rates individuals of different ages. Labor force participation
(Percentage points)
varies considerably over a person’s life, rising rapidly in
The decline in prime-age men’s participation in advanced economies became adolescence, flattening through the working years, and
more pronounced after the global financial crisis, while gains in prime-age falling with age and retirement. Hence, shifts in the
women’s participation slowed.
age distribution are an important driver of movements
1.2 1. Men, Ages 25–54 2. Women, Ages 25–54 1.2
in the aggregate participation rate. These shifts have
1.0 ** 1.0
become particularly pronounced in the past decade
in advanced economies (Figure 2.1, panel 3) as the
0.8 0.8
exceptionally large cohort of people born in the years
0.6 0.6
following World War II began reaching retirement age.
0.4 0.4
In turn, numerous interrelated factors influence
0.2 0.2
** ** individuals’ decisions to supply labor at various points
0.0 0.0
in their life as they assess the expected return to market
–0.2 –0.2
work relative to nonparticipation. Individual character-
–0.4 –0.4
1985–2000 01–07 08–16 1985–2000 01–07 08–16 istics, such as gender, educational attainment, previous
occupation, and household structure, clearly shape
1.6 3. All, Ages 15–24 4. All, Ages 55 and Older 1.6 such decisions, because they determine potential earn-
1.2 1.2 ings in the marketplace relative to nonparticipation.
0.8 ** ** 0.8 But labor market policies, institutions, and noneco-
0.4 0.4 nomic factors that govern the prospect of finding (or
0.0 0.0 retaining) a job and the relative benefit from working
–0.4 –0.4 can also affect participation. Some of these policies, such
–0.8 –0.8
as the tax-benefit system, directly affect the incentive to
–1.2 –1.2
supply labor; others, such as wage-setting institutions,
may shape supply indirectly through reduced labor
–1.6 –1.6
1985–2000 01–07 08–16 1985–2000 01–07 08–16 demand. For example, an increase in the labor tax wedge
could reduce the incentive to work or seek employment,
Sources: Organisation for Economic Co-operation and Development; and IMF staff both by reducing net wages and suppressing firms’ labor
calculations.
Note: Bars denote median; and vertical lines show interquartile range. Asterisks demand as a result of higher labor costs. Conversely,
denote statistically significant difference from 2001–07 at the 10 percent level. active labor market programs that support jobseekers
in finding vacancies may induce individuals to join
the labor force and prevent those who temporarily lose
countries, yet in about half of the sample postcrisis employment from becoming permanently detached.
employment declines translated into both rising unem- Cultural attitudes toward people’s role in society are
ployment and falling participation. also important because they determine the disutility of
Flows into inactivity suggest that the share of dis- market work—for example, through social norms or
couraged workers (inactive now, but unemployed the personally held beliefs (Fernandez 2013).
previous year) has been increasing since the crisis and Policies tailored to addressing the challenges faced
is approaching the precrisis peak (Annex Figure 2.2.5). by specific workers can also influence their labor
supply decisions. For example, provision of childcare,
Understanding Trends in Participation Rates as well as family-friendly policies that make work more
flexible, make it easier for women to combine paid
Conceptual Framework and Research Design
employment and motherhood and may discourage exit
Assessing the appropriate policy responses to from the labor market.16 For older workers, financial
counteract downward pressure on the labor supply as
a result of aging requires a clear understanding of the 16In a simple static labor supply model, parents could choose to
drivers of the aggregate labor force participation rate stay home and take care of an infant or a young child at the cost of
their hourly wage (forgone earnings) minus the price of childcare. A
and individuals’ decisions to be in the job market.
more generous childcare subsidy would increase the parent’s wage net
Two key factors underpin changes in aggregate par- of childcare costs, thus raising the opportunity cost of staying home
ticipation rates: shifts in the age structure of the pop- and increasing labor supply on the extensive margin.

Figure 2.9. Decomposition of Labor Market Shifts

(Percentage points)
Employment declines became more pronounced after the global financial crisis and increasingly translated into lower participation alongside rising unemployment.
Change in employment rate Change in unemployment rate Change in inactivity rate
20 1. 2000–16
15
10
5
0
–5
–10
–15
–20
LTU MLT KOR ISR DEU TWN CHE AUT ISL GBR SWE NLD FRA NOR FIN DNK CYP GRC
LUX EST SGP LVA NZL SVK HKG CZE AUS CAN BEL ESP SVN JPN ITA IRL PRI USA PRT
15 2. 2000–08
10
–5
–10
–15
LVA ESP NZL SVK GRC KOR ISR CZE HKG TWN SGP NOR NLD SWE GBR ISL FRA JPN
LUX EST LTU CYP SVN AUS CAN IRL AUT FIN CHE DEU MLT BEL DNK ITA PRI PRT USA
10 3. 2008–16
–5
–10
SGP KOR LTU HKG SVK JPN GBR CZE EST BEL USA CAN ITA NOR DNK SVN IRL CYP
MLT DEU ISR TWN CHE NZL ISL LUX AUT SWE FRA AUS NLD LVA PRI FIN PRT ESP GRC

Note: Employment rate, unemployment rate, and inactivity rate are defined as total employment, total unemployment, and total inactive population as a percentage of
total population, respectively. Data labels in the figure use International Organization for Standardization (ISO) country codes.
incentives embedded in pension systems and other labor market programs, and the like, can help them
social transfer programs are important considerations overcome their many disadvantages, including lack of
in retirement decisions. Policies that enable immi- information, poor access to informal networks, lack of
grants’ swift integration into labor markets, such as transferable skills and qualifications, and low language
authorization to work, access to language and active proficiency (Aiyar and others 2016).

Long-lasting changes in the demand for workers’ of workers, the analysis uses cross-country panel
skills could also influence individuals’ workforce regressions to disentangle the influence of labor poli-
attachment. For example, the secular expansion of cies and other factors on the participation of different
the service sector in many advanced economies (see population segments. While the potential set of drivers
Chapter 3 of this report) may have created significant is large, the analysis focuses on the variables most com-
employment opportunities for women, who are seen to monly discussed in the policy debate: the tax-benefit
have a comparative advantage in services, thus raising system, activation policies, wage-setting institutions,
female participation.17 On the other hand, technologi- and the role of structural changes and exposure to
cal progress that enabled routine jobs to be automated routinization. The cross-country panel approach has
may have reduced the demand for less-skilled labor in the advantage of capturing the general equilibrium
advanced economies and made certain jobs obsolete. effects of various drivers and quantifying their role in a
While these global developments benefit the economy unified framework. However, the measurement of pol-
as a whole, and create new opportunities in other icies is often imperfect, and the identification of causal
sectors, workers may be unable to take advantage of impacts can be problematic.
these opportunities due to lack of relevant skills and Alongside the analysis of macro data,
training, preferences, hardship involved in relocating individual-level data from 24 European economies
geographically, or an inadequate return compared with allow for a deeper look at the effect of individual char-
their previous earnings. acteristics, including the extent to which (past) occupa-
Participation decisions are also shaped by even more tion can be automated, on workforce attachment, and
short-lived changes in labor demand, such as those the potential for policies to shape this relationship.
caused by cyclical fluctuations (for example, Elsby,
Hobijn, and Sahin 2015). The rise in unemployment
during recessions may lead some workers to drop The Role of Aging and Cyclical Conditions
out of the labor force permanently. Diminished job To quantify the effect of aging, this section performs
prospects during recessions may also induce students a standard shift-share analysis of aggregate participation
to remain in school longer or lead parents (women of men and women. It decomposes observed changes
especially) with young children to stay at home instead in aggregate male and female participation since 2008
of seeking jobs.18 into changes in participation rates within each age
The chapter uses several complementary approaches, group while holding population shares fixed (“within
each one tailored to measure a distinct set of potential changes”), a shift in the relative sizes of age groups
drivers. It starts by quantifying the contribution of while holding participation rates fixed (“between
shifts in the age structure to aggregate participation changes”), and an interaction term. The role of aging
changes in the past decade, using a standard shift-share can be approximated by the “between changes”; in
decomposition. other words, the imputed change in participation if
Given that both the shift-share analysis and the participation rates for each age group had remained at
stylized facts presented previously point to sizable their 2008 levels.19
changes in the workforce attachment of specific groups Because the demographic inflection point coin-
cided with the global financial crisis, the analysis also
17See,
for example, Ngai and Petrongolo (2017) for a model of quantifies the role of the unusually severe recessions in
structural transformation in which relative gains in women’s labor many advanced economies. The cyclical component of
market outcomes are driven by changes toward the service-producing participation changes is estimated from the historical
sector, as well as Olivetti and Petrongolo (2016) for empirical
evidence on the role of the industrial structure in accounting for relationship between detrended aggregate participation
cross-country differences in gender outcomes. For a discussion of rates and output (or unemployment) gaps, allowing for
gender-based comparative advantage, see Feingold (1994); Galor
and Weil (1996); Baron-Cohen, Knickmeyer, and Belmonte (2005);
Christiansen and others (2016a); Rendall (2017); and Cortes, Jai-
movich, and Siu (2018), among others.
18Increasing evidence suggests that adverse initial labor market
conditions can have substantial long-term effects on the earn- 19See Box 1.1 of the October 2017 WEO for a shift-share analysis
ings of college graduates. See, for example, Genda, Kondo, and of labor force participation for selected advanced economies and Aar-
Ohta (2010); Kahn (2010); and Oreopoulos, von Wachter, and onson and others (2006) and Council of Economic Advisers (2014)
Heisz (2012). for the United States.

a differential response of labor force participation to Figure 2.10. Changes in Participation Rates, 2008–16
(Percentage points)
severe recessions.20,21
On average, the observed changes in participation of
Aging can explain the bulk of the decline in men’s participation since 2008. In most
men are broadly consistent with shifts in the popu- regions, women’s participation increased, despite the forces of aging.
lation age profile since 2008 and the drag from the
global financial crisis (Figure 2.10). Women, however, Aging, men Cyclical factors, men Residual
have become significantly more likely to work or seek Aging, women Cyclical factors, women Actual
employment, despite aging, in the average advanced
5 1. Men, All Advanced 2. Women, All Advanced 5
economy (although not in the United States), suggest- 4 Economies Economies 4
ing that policies and other factors are also at play. For 3 3
both men and women, there are notable differences 2 2
1 1
across geographical regions. In the United States, par- 0 0
ticipation has declined significantly more than aging –1 –1
–2 –2
alone would have predicted. In the average European –3 –3
and other advanced economy, on the other hand, gains –4 –4
–5 –5
in participation within each demographic group have 2009 10 11 12 13 14 15 16 2009 10 11 12 13 14 15 16
partially offset, and in some cases exceeded, the drag
from aging. 5 3. Men, Europe 4. Women, Europe 5
The role of cyclical developments is also evident. 4 4
3 3
High unemployment and poor job prospects after the 2 2
crisis depressed participation, especially in Europe and 1 1
0 0
the United States. But as the recovery has taken hold, –1 –1
the drag from cyclical developments has diminished. –2 –2
–3 –3
–4 –4
–5 –5
Drivers of Participation Rates of Specific 2009 10 11 12 13 14 15 16 2009 10 11 12 13 14 15 16
Groups of Workers
5 5. Men, United States 6. Women, United States 5
The finding that changes in aggregate participa- 4 4
tion cannot be fully explained by demographic shifts 3 3
2 2
or cyclical effects in some countries and the wide 1 1
cross-country heterogeneity in participation rates point 0 0
–1 –1
to a potentially important role for policies and other –2 –2
factors influencing the decision to keep working or –3 –3
–4 –4
seek employment. This section examines the historical –5 –5
relationship between the participation of individual 2009 10 11 12 13 14 15 16 2009 10 11 12 13 14 15 16
groups of workers and potential drivers since 1980
across 23 advanced economies. It then uses the esti- 5 7. Men, Other Advanced 8. Women, Other Advanced 5
4 Economies Economies 4
mated associations to provide an illustrative quantifi- 3 3
2 2
1 1
0 0
–1 –1
20The estimates of the cyclical effect for the United States are in
–2 –2
line with those of other studies (Erceg and Levin 2014; Aaronson –3 –3
and others 2014; Council of Economic Advisers 2014; Hall 2015; –4 –4
Balakrishnan and others 2015), despite differences in specifica- –5 –5
2009 10 11 12 13 14 15 16 2009 10 11 12 13 14 15 16
tions and revisions to estimates of potential output (Grigoli and
others 2015).
21Duval, Eris, and Furceri (2011) document that severe recessions Source: IMF staff calculations.
have significant and persistent impacts on participation, while mod- Note: Panels 1–4 and 7–8 show simple averages across countries. Other advanced
erate downturns do not. The econometric analysis relates detrended economies comprise Australia, Canada, Japan, Korea, and New Zealand.
aggregate participation rates to measures of the cyclical position in a
distributed lag specification, allowing for the sensitivity of participa-
tion rates to differ in crisis episodes. See Annex 2.3 for details.

cation of these drivers’ contributions to the observed ied by union density and the level of coordination in
changes in labor supply.22 wage bargaining.
The chapter estimates a reduced-form model of Changes in the demand for different types of work-
labor force participation, looking separately at the ers due to structural transformation and globalization
young, prime-age men, prime-age women, and older are captured in the ratio of services to manufacturing
workers. The model links their participation rates to employment, the degree of urbanization, and trade
factors that may affect the decision to supply labor, openness. Following Chapter 3 of the April 2017
controlling for all differences across countries that are WEO and Das and Hilgenstock (forthcoming),
constant over time and all shocks that affect countries the potential for technology to displace workers is
equally.23 The choice of the predictors is guided by the proxied by the “routinizability” of a country’s initial
conceptual framework outlined previously and data occupation mix interacted with the relative price
availability constraints. of investment goods in advanced economies—that
The analysis examines the tax-benefit system, as is, the automation of routine tasks. The empiri-
captured in the labor tax wedge and generosity of cal specification controls for the output gap, while
unemployment benefits, and looks at policies spe- education, measured as the share of population in the
cifically geared toward improving the job-matching age-gender group with secondary and tertiary edu-
process: spending on active labor market programs (for cation, is included as a proxy for workers’ potential
example, training programs, job-search assistance, and returns to work.26
so forth) and major policy changes that help migrants The analysis indicates that education, cyclical and
integrate in a host country. When studying women’s long-lasting shifts in labor demand, and labor market
participation decisions, the analysis expands the set of policies are strongly associated with participation rates
policies to include public spending on early childhood (Table 2.1). However, there are significant differences
education and care, length of job-protected maternity in the responsiveness of workforce attachment to these
leave, and opportunities for part-time employment.24 factors across groups of workers.
For older workers, the analysis considers the statutory In line with economic theory, education is a
retirement age and the generosity of pension plans.25 powerful predictor of labor force participation. An
Wage-setting institutions and frameworks are prox- increase in the share of workers with secondary and
especially tertiary education is associated with signifi-
cantly higher participation, particularly for prime-age
22The baseline results are based on the set of countries classified as
women and older workers. Higher education is also
advanced in the WEO for most of the time period, thus excluding positively associated with participation of prime-age
the eight countries that became advanced after 2006. The chapter’s
findings are robust to using the full set of countries currently classi- men, but to a smaller degree, in line with the much
fied as advanced.
23The empirical specification is
i,t  β  Di,t i,t − 1  + β  Z i,t 

g    = β  X,g X  g  + 
FP i,t
L D,g     + β  GAP,g GAP  Z,g
+ π i   + τ  t   + ε  i,t , 

g g g
in which LFP denotes the participation rates of worker group g in 26The empirical approach in the chapter is widely used in the
country iat time t, GAPis the cyclical position of the economy, X cross-country literature. Blanchard and Wolfers (2000); Genre,
represents the set of policies and institutions (some of these are spe- Gómez-Salvador, and Lamo (2005); Bertola, Blau, and Kahn
cific to group g), Dare factors that may shift the demand for worker (2007); Bassanini and Duval (2006, 2009); de Serres, Murtin, and
group g, Zcomprises other determinants of labor supply (education), Maisonneuve (2012); Murtin, de Serres, and Hijzen (2014); and
and π i and τ  tare country and time fixed effects. See Annex 2.4 for Gal and Theising (2015) examine determinants of employment and
further details on the empirical estimation and robustness tests, and unemployment, among others. See, for example, Jaumotte (2003);
a full description of the variables used and their sources. Genre, Gómez-Salvador, and Lamo (2010); Blau and Kahn (2013);
24Data availability on taxes on the secondary earner in the house- Cipollone, Patacchini, and Vallanti (2013); Thévenon (2013);
hold is limited, thus the variable is not included in the empirical Dao and others (2014); and Christiansen and others (2016b)
specification. for cross-country analysis of female labor force participation and
25In the baseline specification, the generosity of pension plans is employment and Blöndal and Scarpetta (1999) and Duval (2004)
measured as old-age and incapacity spending as a percent of GDP, for cross-country analysis of retirement decisions. Relative to the
purged of fluctuations resulting from cyclical and demographic fac- literature, the chapter expands the temporal coverage of the analysis,
tors. Conceptually more appropriate measures of incentives for early capturing the last decade during which significant changes in partic-
retirement, such as the change in net pension wealth from an addi- ipation occurred. The chapter’s focus on the effects of long-lasting
tional year in the labor force, or pension replacement rates, would shocks to labor demand, such as those stemming from technological
severely restrict the sample, but are examined in robustness tests. advances, and on migrant integration policies is also new.

Table 2.1. Drivers of Labor Force Participation Rates

(1) (2) (3) (4) (5)
All, Men, Women, All, All,
Ages 15–24 Ages 25–54 Ages 25–54 Ages 55+ Ages 15+
Lag of Output Gap 0.360*** 0.072*** 0.170* –0.006 0.183***
(0.112) (0.020) (0.092) (0.068) (0.044)
Routinization × Relative Price of Investment 0.303 0.302*** 1.793*** 0.505* 0.536***
(0.299) (0.048) (0.206) (0.288) (0.175)
Lag of Trade Openness 0.059*** –0.005 0.010 –0.059*** 0.012*
Other Factors
(0.022) (0.005) (0.014) (0.009) (0.007)

Relative Service Employment –0.002 –0.002 0.015*** 0.009 0.010**
(0.010) (0.002) (0.005) (0.006) (0.004)
Urbanization 0.668*** 0.101*** 0.355*** 0.194 0.249***
(0.142) (0.019) (0.071) (0.115) (0.047)
Education (percent secondary) –0.050 0.019*** 0.211*** 0.038* 0.063***
(0.042) (0.007) (0.017) (0.021) (0.017)
Education (percent tertiary) –0.275*** 0.019 0.332*** 0.389*** 0.135***
(0.057) (0.015) (0.030) (0.050) (0.031)
Tax Wedge –0.103 –0.002 –0.129*** –0.263*** –0.240***

(0.064) (0.015) (0.029) (0.037) (0.026)
Unemployment Replacement Ratio –0.002 –0.041*** –0.035 –0.081 –0.078***
(0.068) (0.007) (0.033) (0.050) (0.025)
Public Spending on ALMP 0.041*** 0.005 0.039*** –0.025** 0.031***
(0.014) (0.005) (0.006) (0.009) (0.007)
Restrictiveness of Migrant Integration Policies 0.491*** –0.047** –0.462*** 0.056 –0.207***
(0.098) (0.020) (0.049) (0.088) (0.049)
Union Density –0.009 –0.001 0.153*** –0.115*** –0.015
(0.068) (0.011) (0.044) (0.032) (0.025)
Coordination of Wage Setting 1.104*** 0.131** 0.701*** 0.040 0.256**
(0.245) (0.063) (0.219) (0.222) (0.120)
Policies
Public Spending on Early Childhood

Education and Care 3.708***
(1.210)
Share of Part-Time Employment 0.946***
(0.118)
Job-Protected Maternity Leave 0.025***
(0.006)
Statutory Retirement Age 0.661***
(0.174)
Public Spending on Old-Age Pensions –0.750***
(0.154)
Public Spending on Incapacity –0.421
(0.562)
Number of Observations 571 571 489 568 570

Countries 23 23 23 23 23
R 2 0.515 0.606 0.887 0.686 0.578
Note: The table presents results from the estimation of equation (2.3) with separate regressions for the participation rate for each group of workers on a sam-
ple of 23 advanced economies during 1980–2011 using annual data. See Annex 2.4 for the construction of the explanatory variables and Annex Table 2.1.2
for the countries in the sample. All specifications include country and year fixed effects. Driscoll-Kraay standard errors are in parentheses. ALMP = active
labor market programs.
*p < .10; **p < .05; ***p < .01.

Figure 2.11. Drivers of Participation Rates: Policies smaller variability in their participation rates seen in
(Percentage points) Figure 2.5.27
For most groups of workers, participation rates
Higher tax wedges and more generous unemployment benefits depress
participation, while spending on active labor market programs and higher levels of depend on the state of the business cycle. As expected,
wage-setting coordination are associated with higher participation. Policies that the association is significantly higher for those more
encourage the integration of migrants are associated with higher participation of
prime-age workers.
marginally attached to the workforce, such as the
young and women.
All, ages 15–24 Women, ages 25–54 All, ages 15 and older
The analysis also confirms that structural transfor-
Men, ages 25–54 All, ages 55 and older mation that may shift the demand for certain types
of workers affects their labor market involvement.
0.1 1. Tax Wedge 2. Unemployment Benefits 0.2 A relative increase in service sector employment is
typically followed by the entry of prime-age women
0.0
0.1 into the labor force, while urbanization brings gains in
–0.1 the participation of all groups, potentially by exposing
0.0 them to a larger set of job opportunities.
–0.2 Conversely, although technological change can
–0.1 benefit the economy as a whole and create new oppor-
–0.3
tunities in other sectors, it may not be fully benign
–0.4 –0.2 from the point of view of some workers. A decline
in the relative price of investment is associated with
0.10 3. Spending on Active 4. Migration Policies 0.8 lower participation rates in countries where the initial
Labor Market Programs
0.6 occupation mix is tilted toward routine-task occupa-
0.4 tions, highlighting the difficulties of workers displaced
0.05
0.2
by automation in finding alternative employment (see
Box 2.2 and Box 2.3 for subnational evidence from the
0.0
0.00
United States and Europe, respectively).28
–0.2 Participation rates are also responsive to labor
–0.4 market policies and institutions (Table 2.1; Figures
–0.05 –0.6 2.11–12). In particular:
•• The tax-benefit system has a robust relationship with
0.3 5. Union Density 6. Coordination of 2.0 participation rates. Higher labor tax wedges and
Wage Setting
1.6 more generous unemployment benefits are associated
0.2
with lower labor force attachment for most groups
1.2
0.1 of workers, in line with findings in the cross-country
0.8 literature on their effect on employment (see, for
0.0 example, Gal and Theising 2015 and its references).29
0.4
–0.1 0.0 27The negative association between labor force participation and
the share of population ages 15–24 with partial or completed tertiary

–0.2 –0.4
education likely reflects that they are still in school.
28This finding is consistent with the role of technological progress,
Source: IMF staff calculations. along with varying exposure to routine occupations, in the decline
Note: The bars denote the estimated change in participation from a one-unit
in the labor share in advanced economies documented in Chapter 3
increase in the policy variable, while the vertical lines show the 90 percent
confidence interval. See Annex 2.4 for variable definitions and specification of the April 2017 WEO and Dao and others (2017). Acemoglu and
details. Tax wedge is measured in percent of labor costs. The unemployment Restrepo (2017) provide evidence of significant employment losses
benefits gross replacement rate is measured in percent of work income. Public in local US labor markets with greater exposure to robots; Autor and
spending on active labor market policies is measured per unemployed person and Dorn (2013) examine the impact of the falling cost of automating
as percent of per capita GDP. Union density is measured as net union membership routine jobs on polarization and jobs of different skill levels.
as a proportion of wage earners in employment. Migration policy is an index 29In theory, the net effect of higher taxes on labor supply is
constructed by cumulating major changes in policies and regulations guiding the ambiguous. If higher labor taxes lower net wages, individuals may
postentry rights and other aspects of migrants’ integration, with a higher value respond by working more to maintain their income. On the other
denoting more restrictive policies. Coordination of wage setting is an index,
hand, by lowering the relative return to market work, higher taxes
ranging from 1 (decentralized) to 5 (centralized).
may lead to lower participation. The negative relationship between

•• On the other hand, policies specifically geared toward Figure 2.12. Drivers of Participation Rates: Additional Policies
(Percentage points)
improving the job-matching process are generally
associated with stronger participation rates.30 Higher
Family-friendly policies are associated with higher participation among women,
public spending on active labor market programs while retirement incentives significantly affect the participation decisions of older
tends to raise the share of young and prime-age workers.
women working or seeking employment. The
analysis also indicates that policies that encourage 0.7 1. Policies Primarily Affecting Women
the integration of migrants can help boost prime-age 0.6
workers’ participation, with more pronounced effects 0.5

on women. The positive association likely reflects the 0.4
success of these policies in narrowing the sizable par- 0.3
ticipation gaps between native and immigrant work- 0.2
ers, which are especially wide for women. However, 0.1
other channels are possible. A more migrant-friendly 0.0
policy stance may bring in more immigrants. –0.1
Child education/ Share of part-time Job-protected
Although migrants have a lower propensity to work care spending* employees* maternity leave**
than natives when they arrive, they are more likely
to be prime age than the native population and 1.2 2. Policies Primarily Affecting Older Workers
may boost aggregate participation rates through 0.8
compositional shifts (see Box 2.4). Several recent
0.4
studies have also emphasized the complementarity
of migrants’ skills to those of the native population, 0.0
which has helped boost natives’ labor market out- –0.4

comes, especially women’s.31 The negative association –0.8
between more friendly migration policies and youth –1.2
labor force participation is not surprising, given that
–1.6
integration measures include giving migrants access Implicit tax Pension Retirement Pension Incapacity
on continued replacement age spending spending
to education and training, which could lead to more work** ratio**
foreign students and increase school enrollment of
nonnative young people. Source: IMF staff calculations.
•• Women’s willingness to work or seek employment Note: The bars denote the estimated change in participation from an increase in
the policy variable, while the vertical lines show 90 percent confidence intervals.
is significantly influenced by policies that help See Annex 2.4 for variable definitions and specification details. * indicates an
them reconcile work inside and outside the house- increase in the variable by 0.1 unit. ** indicates an increase in the variable by 10
units. Public spending on childcare and education is measured as percent of GDP.
Job-protected maternity leave is measured in weeks. Statutory retirement age is
measured in years. Implicit tax on continued work is the change in the present
participation rates and the generosity of unemployment benefits, value of the stream of future pension payments net of contributions to the system
measured as the gross benefit replacement rate, is consistent with (1) from working five more years, while the pension replacement ratio is the ratio of
the positive correlation found in cross-country data between gener- mean disposable income of those ages 65–74 to the mean disposable income of
osity of unemployment benefits and unemployment levels, which those ages 50–59. Spending on old-age pensions and incapacity are measured as
could depress participation through a discouragement effect; and (2) percent of GDP and are purged of fluctuations due to cyclical and demographic
the fact that in many countries the unemployment insurance system factors. Dotted vertical lines in panel 2 denote results from different regressions.
provides a path to early retirement for older workers.
30Activation policies are proxied by spending on active labor mar-
ket programs per unemployed person as a share of GDP per capita.

To measure migrant integration policies, the chapter constructs an
index based on major policy changes in rules governing the integra-
tion of migrants, such as their postentry access to language, housing,
and cultural integration programs; social benefits; health, education,
and unemployment benefits; and the like from the DEMIG POL-
ICY database (de Haas, Natter, and Vezzoli 2014).
31See, for example, Carrasco, Jimeno, and Ortega (2008);
D’Amuri and Peri (2014); Cattaneo, Fiorio, and Peri (2015); Foged
and Peri (2015); Aiyar and others (2016); and Chapter 4 of the
October 2016 WEO.

hold (Figure 2.12). Consistent with the findings two empirical models: one that deliberately excludes
of a large body of literature, the chapter’s analysis policies and institutions as determinants of participa-
suggests that better access to childcare, longer tion and one that includes them. A comparison of how
maternity leave, and greater flexibility in work well these models account for the observed changes in
arrangements are associated with higher female participation across countries indicates that changes
labor force participation.32 in labor market policies and institutions can explain
•• For older workers, incentives for retirement have a a quantitatively meaningful fraction of the observed
powerful effect on labor force attachment.33 Raising changes in labor force participation across countries
the statutory retirement age is associated with (Figure 2.13). The correlation between actual and pre-
delayed exit from the labor market, whereas greater dicted participation is substantially higher for a model
pension plan generosity seems to encourage early that includes labor market policies compared with a
retirement. The latter finding is robust to using con- model that does not. However, there are sizable dif-
ceptually more appropriate, but less widely available, ferences in how well the empirical model can explain
measures of incentives for early retirement, such cross-country variation in participation trends across
as the implicit tax on continued work or pension population groups. Notably, a very large fraction of
replacement rates (Figure 2.12). the observed change in labor force participation of the
•• Finally, the evidence on the role of wage-setting young remains unexplained by the factors considered
institutions—unionization and the degree of wage in the analysis.
bargaining coordination—is mixed (Figure 2.11). Combining policies, education, structural shifts, and
Higher coordination of wage setting is associated technology, Figure 2.14 examines the contributions of
with greater labor force participation for most these factors to changes in participation rates between
groups of workers, consistent with the idea that 1995 and 2011. Supportive policies and educational
more coordinated bargaining systems may lead to gains have been key factors behind the dramatic
faster wage moderation during downturns as unions increase in the participation of prime-age women and
internalize the potentially detrimental effects that older workers, with structural transformation contrib-
excessive wage pressure may have on overall employ- uting positively as well. On the other hand, techno-
ment (Soskice 1990; Bassanini and Duval 2006).34 logical advances have weighed on participation for all
However, the correlation between unionization and groups of workers except the young.
participation is less robust to changes in the sample For the young, and to a certain extent prime‑age
or the inclusion of other policies. male workers, a significant share of the decline in
participation is attributed to a common component
Overall, these results suggest that policies can across advanced economies, captured by the time
influence labor force participation decisions. But can effects in the regressions. This common factor could
they help explain the sizable cross-country differences reflect the common influence of global forces, such
in observed changes in participation rates? To answer as technological progress or globalization, concurrent
this question, the chapter examines the change in the changes in policies, structural transformations, or
workforce attachment of different groups of workers other factors that may affect labor supply decisions
between 1995 and 2011—for which data are available across the advanced world, such as changing returns
for almost every policy and every country—against to education, rising life expectancy, or common scars
the changes in labor force participation predicted by from the global financial crisis. For older workers, the
latter may have delayed retirement, as captured in the
32See
Olivetti and Petrongolo (2017) and its references for a recent
positive common component, as a result of sup-
review of evidence on the economic consequences of family policies pressed returns on retirement savings as global interest
as well as, for example, Jaumotte (2003); Genre, Gómez-Salvador, rates fell, losses in financial wealth, and potentially
and Lamo (2010); Blau and Kahn (2013); Cipollone, Patacchini, and
higher debt.
Vallanti (2013); Thévenon (2013); Dao and others (2014); Chapter 3
of the April 2016 WEO; and Christiansen and others (2016b). Comparing how the various factors relate to
33See Blundell, French, and Tetlow (2016) and its references for a
participation changes across geographic regions can
review of the literature on retirement incentives and labor supply. shed light on the reasons behind their (sometimes)
34Janssen (2018) similarly finds that the costs of worker displace-
ment are higher in a more decentralized wage bargaining system, by divergent trends. For example, the analysis reveals that
studying a major reform of the wage bargaining system in Denmark. the striking difference in the participation trend for

Figure 2.13. Changes in Participation Rates, Actual versus Figure 2.14. Average Contributions to Changes in
Predicted, 1995–2011 Participation Rates, 1995–2011
(Percentage points) (Percentage points)
Policies help explain observed differences in changes in labor force participation Technological change weighs on labor force participation. Gains in education and
across advanced economies. policies have, however, more than offset this effect for prime-age women and
older workers.
Baseline Baseline plus policies

Policies Structural factors Residual
Technology and trade Output gap Actual
10 1. All, Ages 15–24 Education Year effects
5 8 1. All, Ages 15–24

0 4
Predicted
–5 0
–4
–10 Correlation baseline: 0.21
Correlation baseline plus policies: 0.28 –8
–15
–15 –10 –5 0 5 10 –12
Actual –16
All AEs Europe United States Other AEs
3 2. Men, Ages 25–54
2 Correlation baseline: 0.29 1.0 2. Men, Ages 25–54
0.5
1 Correlation baseline plus policies: 0.54 0.0
Predicted
0 –0.5
–1 –1.0
–1.5
–2
–2.0
–3 –2.5
–4 –3.0
–4 –3 –2 –1 0 1 2 3 –3.5
Actual All AEs Europe United States Other AEs
30 3. Women, Ages 25–54 16 3. Women, Ages 25–54

25 12
20
8
15
Predicted
10 4
5 0
0
Correlation baseline: 0.66 –4
–5 Correlation baseline plus policies: 0.85
–10 –8
–10 –5 0 5 10 15 20 25 30 All AEs Europe United States Other AEs
Actual
12 4. All, Ages 55 and Older
25 4. All, Ages 55 and Older
9
20
15 6
Predicted
10 3
5
0
0
Correlation baseline: 0.34
–5 Correlation baseline plus policies: 0.73 –3
All AEs Europe United States Other AEs
–10
–10 –5 0 5 10 15 20 25
Actual Source: IMF staff calculations.
Note: See Annex 2.4 for variable definitions and specification details.
Source: IMF staff calculations. AEs = advanced economies. Other AEs comprise Australia, Canada, Japan, Korea,
Note: See Annex 2.4 for variable definitions and specification details. and New Zealand.

US women relative to the average European trend can endogeneity bias arising from omitted variables and
be attributed to the more supportive policy changes reverse causality in regressions relying on aggregate
in Europe and larger gains in educational attainment data. The analysis also zooms in on the impact of
among prime-age European women. The factors technology and the extent to which policies can help
behind the rise in participation among older workers offset its effect on individuals’ decisions to drop out of
are very similar across all regions: gains in education, the labor force.
structural transformation, and the introduction of The empirical analysis models the decision of an
policies that discourage early retirement.35 However, individual to participate in the labor market as a func-
the reason that US prime-age men and youth became tion of personal characteristics (education, immigration
so much more disconnected from the labor market status, location), family composition (single versus
than their European counterparts remains puzzling, as living as part of a couple, with and without children),
evidenced by the sizable residual in the decomposition and exposure to routinization. To measure vulnerability
of the change. Many hypotheses regarding this decline to automation, the analysis uses information on the
are specific to the United States and, consequently, occupation of currently employed individuals, as well
cannot be evaluated in a cross-country setting—for as on the most recent occupation of those unemployed
example, the role of rising disability, opioid use, higher or inactive, and assigns each a routinizability score
incarceration, and improved leisure technology.36 based on their (most recent) occupation, following
Evidence from subnational data presented in Boxes 2.2 Chapter 3 of the April 2017 WEO and Das and
and 2.3 also suggests longer-lasting harm to participa- Hilgenstock (forthcoming).37
tion from technological progress in the United States In line with the aggregate findings, the analysis
than in Europe. points to large and significant effects of higher educa-
tion (Figure 2.15). Tertiary education roughly doubles
the odds of being active over attainment of up to lower
Drivers of Individual Participation Decisions
secondary education, with somewhat larger effects for
The final step of the analysis complements the women. Living in an urban area also raises partici-
cross-country findings by examining evidence from pation, likely on account of access to a more diverse
millions of individuals in Europe. The use of micro labor market with more opportunities. Natives are also
data offers important advantages relative to the more likely to participate than immigrants.
cross-country results discussed so far. It allows for a Family composition has a considerable influence on
deeper exploration of individual and household-level the decision of an individual to work or seek employ-
determinants of participation, thus mitigating the ment, although there are large gender differences.
Relative to the baseline category of being the only
35See,
among others, Blau and Goodstein (2008) and Hurt adult in a household without children, being part
and Rohwedder (2011) for evidence from the United States and
Börsch-Supan and Ferrari (2017) for evidence from Germany. of a couple and having children is associated with
36See Eberstadt (2016), Council of Economic Advisers (2016), higher participation of men but lower participation of
Krause and Sawhill (2017), and Abraham and Kearney (2018) women. Similarly, more children are associated with
for a review of the literature. Krueger (2017) discusses the poor
lower participation of women but higher participa-
health status of men not in the labor force and the rising use of
pain medication. Case and Deaton (2017) document an increase tion of men, consistent with the historical allocation
in mortality rates as a result of addiction, depression, and suicide of work across genders within a household. Interest-
(“deaths of despair”) among white prime-age adults and hypothe- ingly, the presence of other employed adults in the
size that it may be rooted in the steady deterioration of their job
opportunities. Holzer, Offner, and Sorensen (2005); Pager, West- household is associated with a higher likelihood of
ern, and Sugie (2009); and Schmitt and Warner (2010) present being active, likely pointing to common labor market
evidence of a dramatic increase in incarceration and the ex-prisoner
population in the United States, which faces significant barriers to
employment. Aguiar and others (2017) argue that the decline in 37The model is estimated on a subsample of 18 countries relative
the labor supply of young men may be linked to improvements in to the sample used in the stylized facts with detailed information
video gaming and other recreational computer activities. It should on family composition. Logit regressions relate a binary outcome
be noted, however, that the extent and direction of causality of variable capturing whether a person is in or out of the labor force
these hypotheses are difficult to establish empirically. Abraham to the above-mentioned participation determinants, controlling
and Kearney (2018) offer a rough quantification of the role of for the aggregate output gap and country and year fixed effects.
various factors in US employment rate trends since 1999 based on Annex 2.5 provides a detailed description of the empirical
existing studies. methodology.

effects. These findings should, however, be treated Figure 2.15. Change in the Odds of Being Active
as associations rather than causal effects as labor (Percent)
supply decisions and family composition are likely
Higher education is associated with higher odds of being active, while being
jointly decided.38 married and having children is associated with lower labor force participation of
Finally, in line with the country-level results, the prime-age women. Those in more routinizable occupations are more likely to
become detached from the labor force.
micro analysis points to significant negative effects of
exposure to routine tasks. Working or having worked
in an occupation that is more vulnerable to routiniza- Upper secondary education
tion is associated with lower odds of participation. This
Tertiary education
effect is larger for men and is especially pronounced for
Born in country Men, ages
workers 55 and older. The effects are both statistically
25–54
and economically significant: a unit change in routin- Urban
Women, ages
ization scores roughly corresponds to the difference in Number of children in household 25–54
All, ages 55
the routinization score of technicians and the routini- One adult with children and older
zation score of managers. Whereas about 87 percent of
Couple without children
prime-age male managers are active, about 84 percent
of prime-age male technicians are in the labor force— Couple with children
the difference in their routinization scores alone can Other household structure
explain about one-third of this 3 percentage point Other employed adult(s) in household
difference in participation rates.39
Routinizability of (last) occupation
Can policies help those vulnerable to losing their
jobs to technology remain active in the labor market? –50 0 50 100 150 200
To answer this question, the analysis examines whether
Sources: Das and Hilgenstock (forthcoming); Eurostat, European Union Labour
various country-level labor market policies, such as Force Survey; and IMF staff calculations.
spending on active labor market programs or employ- Note: Logit regressions are based on a random sample of 10,000 respondents per
ment protection, can offset some of the negative effect country per year from the European Union Labour Force Survey over the period
2000–16 and for 18 countries. Only effects significant at the 10 percent level are
of routinization on participation. It augments the shown. The base category for education is “up to lower secondary education.” For
logit model described earlier in this chapter with an family composition, the base category is “one adult without children.” Changes in
odds ratios are shown. See Annex 2.5 for specification details.
interaction between the routinization score and the
relevant policy measure. Figure 2.16 plots the effect
of a unit change in the routinization score, estimated In particular, higher spending on active labor market
at the 75th and 25th percentiles of the distribution of programs seems to attenuate the link between partici-
policies (in other words, in countries with relatively pation and routinizability of occupation. The negative
high versus relatively low spending on active labor association between routinizability and participation is
market programs, and the like). about one-third as large in countries at the 75th per-
Policies can offset at least some of the negative centile of active labor market spending as in countries
association between routinization and participation. at the 25th percentile. Disaggregated data on different
active labor market programs suggest that the finding
38While baseline specifications do not control for household is driven by spending on training, which mitigates
income due to data limitations, once a predicted income decile is some of the negative effect for prime-age women.40
included, the effect of being part of a couple and having children
on the participation of women turns positive, the effect of other
employed adults in the household turns negative, and income itself 40It should, however, be added that active labor market programs
has a negative effect. This suggests that individuals in upper deciles can be expensive; their success hinges crucially on specific design
may be able to afford to drop out of the labor force, or, alternatively, features, and evidence on their effectiveness more broadly is mixed
that some of the rise in women’s participation could be explained by (see IMF/WB/WTO 2017 for a recent literature review). Surveying
declining household income (see Annex 2.5). the evidence from North American and European studies, Heckman,
39While the baseline specification relies on a cross-country Lalonde, and Smith (1999) conclude that public employment and
panel, country-by-country estimates confirm these findings: the training programs had at best a modest positive impact on earnings
effects of vulnerability to routinization are significant and negative by raising employment probabilities. Card, Kluve, and Weber
in most countries and are typically more pronounced for men (2010) find substantial variation in estimated program effectiveness
than for women. across studies.

Figure 2.16. Policies and the Effect of Routine Exposure on for displaced workers to find other employment. This
Labor Force Participation finding underscores the importance of easing geo-
(Percent)
graphical mobility to help workers adjust to local labor
demand shocks.41
Policies, such as spending on active labor market programs and education, can
help mitigate some of the negative effects of exposure to routinization on labor Finally, while the negative effects of routinization
force participation, especially for women. The negative effects of automation are are larger for older workers, policies also provide less
also smaller in urban areas.
of an offset.
0.5
Men, ages 25–54 Women, ages 25–54
Prospects for Labor Force Participation
0.0
To conclude its analysis, this chapter examines
–0.5 the long-term prospects for labor force participation.
Using a cohort-based model, this section estimates
–1.0 trend labor force participation for finely disaggregated
age groups of men and women across 17 advanced
–1.5 economies, accounting for all age-gender-specific and
birth-year-gender-specific determinants of labor supply.
–2.0 These estimates are combined with projections on the
demographic distribution over the next 30 years to
–2.5 forecast the aggregate trend labor force participation
rate. Finally, the analysis presents three illustrative
–3.0 simulations of how these trends would evolve under
25th 75th 25th 75th 25th 75th 25th 75th 25th 75th RUR URB the assumption of significantly higher labor market
ALMP ALMP training Education Wage-setting Employment Location
spending spending spending coordination protection
participation of women and older workers and of the
implementation of policies to boost participation.
Sources: Das and Hilgenstock (forthcoming); Eurostat, European Union Labour
Force Survey; and IMF staff calculations.
Note: Bars show the effect of a one-unit increase in routine exposure on the A Cohort-Based Analysis
probability of being active for policies at given percentiles, based on logit
regressions on a random sample of 10,000 respondents per country per year from A cohort-based analysis of labor force participation
the European Union Labour Force Survey over the period 2000–16 for 24
countries. Lines show 95 percent confidence interval. Lighter colors denote that
is a widely used tool to model trend participation rates
the effects are not statistically significantly different from each other at the 10 and forecast labor supply.42 This approach exploits
percent level. See Annex 2.5 for specification details. ALMP = active labor market variation in participation across age and gender groups
programs; RUR = rural; URB = urban.
and over time for each country to uncover the underly-
ing age participation profile (age effects) and the shifts
from these profiles as a result of new cohorts entering
For both men and women, stricter employment the labor force (cohort effects).43 These cohort effects
protection (making hiring and firing more difficult)
41Encouraging people to move where there are more employment
also offsets some of the adverse individual participation
opportunities could, however, further worsen the situation for those
effect of being in a routinizable occupation, though staying behind and increase geographic polarization.
possibly at the cost of reduced labor market flexibility 42See, for example, Fitzenberger and Wunderlich (2004) for Ger-
at the country level and fewer job market prospects for many; Aaronson and others (2006, 2014), Fallick and Pingle (2007),
and Balakrishnan and others (2015) for the United States; Chapter 3
some other groups, such as youth (see, for example, of the April 2015 WEO; Euwals, Knoef, and van Vuuren (2011)
OECD 2004, 2010; Betcherman 2012). For prime-age for the Netherlands; Balleer, Gómez-Salvador, and Turunen (2014)
men, a higher level of wage-setting coordination is for selected European countries; and Blagrave and Santoro (2017)
for Chile. Annex 2.6 provides further details on the estimation
associated with a smaller negative effect of routin- methodology.
ization, as more coordinated wage bargaining may 43More precisely, the cohort-based model consists of estimating
internalize some of the negative shocks to employment. a country- and gender-specific system of equations in which the
participation rate of each five-year group between ages 15–64, and
The negative effect of routinization is smaller in
of those ages 65 and older is regressed on a constant, dummies for
urban than in rural areas, as cities may offer more different birth cohorts, and a proxy for the cyclical position of the
diverse labor markets and hence more opportunities economy. Given that a key goal of the analysis is the estimation

include all factors associated with a particular year of Figure 2.17. Age and Cohort Effects of Labor Force
birth, such as the impact of choices made early in life Participation
(Percent)
(for example, investment in education, and decisions
regarding marriage and children) that have persistent
Gains in female participation across cohorts have plateaued, and even edged
effects on labor supply as well as slowly changing social down recently, especially in the United States. The age participation profile of
norms, institutions, and preferences toward work. women remains below that of men, significantly so for the prime-age population.
Future aggregate participation is forecast by combining
100 1. Age Effects1 2. Cohort Effects1 3
the estimated age effects with projections of the distri-
2
bution of population across age groups. 80 1
Before turning to the forecast, it is useful to
0
examine the estimated age and cohort effects. Labor 60
–1
force attachment of both men and women exhibits –2
40
a well-known hump shape over the life cycle, with Men Men –3
important gender differences (Figure 2.17, panels 1, 3, 20
Women Women –4
and 4). Across all ages, men are more likely to be part –5
of the labor force than women, but the gender gap is 0 –6
15–19 30–34 45–49 60–64 1925–29 45–49 65–69 85–89
particularly pronounced during the prime-age years.
How these age profiles have shifted from cohort Europe United States Other advanced economies
to cohort is also vastly different for men and women.
Trend male participation rates have not changed 120 3. Age Effects, Men 4. Age Effects, Women 120
significantly across cohorts, except for the slight dip in
100 100
the participation of recent cohorts, which is notably
deeper in the United States. For women, there has 80 80
been a large increase in participation across cohorts, 60 60
in line with the stylized facts discussed earlier.44 For
example, women born in the 1970s are 4 percent- 40 40
age points more likely to work or seek employment 20 20

than women born in the early 1930s. Moreover, the
0 0
dispersion of cohort effects for women is significantly 15–19 30–34 45–49 60–64 15–19 30–34 45–49 60–64
smaller for later cohorts, underscoring the convergence
of women’s labor force participation across countries. 2 5. Cohort Effects, Men 6. Cohort Effects, Women 2
However, cohort effects have plateaued recently and 1 1
even edged down, especially in the United States. 0 0
This finding has important implications: the historical
–1 –1
gains in female labor force participation owing to the
–2 –2
entry of new birth cohorts and the exit of older ones
may no longer be an option for raising participation –3 –3
in many advanced economies without significant –4 –4
policy efforts. –5 –5
1925–29 45–49 65–69 85–89 1925–29 45–49 65–69 85–89

Note: Other advanced economies comprise Australia, Canada, and Japan. Age
effects describe the age-participation profile, and cohort effects describe the shifts in
of cohort effects, which requires sufficiently long data series, other the age-participation profile across cohorts. See Annex 2.6 for specification details.
1
determinants of labor supply, such as educational attainment and Lines denote median, dotted lines show population-weighted average, and shaded
areas show interquartile range.
policies, are not included because of limited temporal coverage.
44To explain the presence of cohort effects in women’s labor force
participation, Fernandez (2013) proposes a theoretical model in

which women learn from the participation behavior of earlier gen-
erations. Goldin (2006), on the other hand, attributes the positive
cohort effects to the increase in returns to education, changes in
preferences, and higher human capital accumulation.

Figure 2.18. Projected Changes in Participation Rates under Projection Scenarios

Alternative Scenarios
(Percentage points) A baseline projection scenario for trend labor force
participation up to 2050 is constructed by combining
Increasing the participation of prime-age women and older workers by the estimated age-gender-group trend rates with pro-
implementing policies aimed at boosting incentives to participate could partially
offset some of the negative effects of aging. jections of how demographic distributions will evolve
based on the United Nations World Population Pros-
Baseline, median Baseline, interquartile range pects.45 The simulation suggests that, absent policies
Baseline, weighted average to boost participation, the median trend participation
rate will fall by 5½ percentage points over the next 30
2 1. Closing Gender Gaps years (Figure 2.18). All else being constant, a decline in
0 aggregate participation of this magnitude would trans-
late into a 3 percentage point reduction in potential
–2
output by 2050 for the typical advanced economy.46
–4 The decline in participation is projected to be broad
–6 Scenario, median based, with rates hovering around 50 percent or lower
Scenario, weighted average in Belgium, France, Italy, Portugal, and Spain.
–8 Scenario, interquartile range
To give a sense of the scope for boosting labor
–10 supply, an illustrative simulation makes the stark
2015 20 25 30 35 40 45 50
assumption that prime-age women’s participation rates
gradually converge to those of prime‑age men over the
2 2. Extending Working Lives
next 20 years (Figure 2.18, panel 1).47 In this scenario,
0 the median aggregate participation rate would decline
–2 more gradually, and by the end of the projection hori-
zon, it would be 2½ percentage points higher than in
–4
the baseline scenario.
–6 Scenario, median An alternative simulation assumes that older workers
Scenario, weighted average
–8
remain in the labor force longer. Specifically, the par-
Scenario, interquartile range
ticipation rate of the 55–59 age group converges to the
–10 participation rate of the 50–54 age group over the next
2015 20 25 30 35 40 45 50
20 years, and the participation rate of the 60–64 age
2 3. Implementation of Policies group converges to that of the 50–54 age group over
the next 40 years, keeping gender gaps in participation
0
across age groups unchanged (Figure 2.18, panel 2).
–2 Raising the participation of older workers would also
–4
Scenario, median 45It is assumed that new cohorts entering the labor force do not
–6
Scenario, weighted average shift the age-participation profile and that output is equal to poten-
–8 Scenario, interquartile range tial during the projection horizon.
46For the purpose of this exercise, the labor share of income is
–10 assumed to be 56 percent, which corresponds to the average labor
2015 20 25 30 35 40 45 50
share of income in 2017 for a subset of advanced economies (Aus-
tralia, Canada, Germany, Italy, Japan, Spain, United States). The fall
Source: IMF staff calculations. in potential output is thus obtained by multiplying the average labor
Note: The “Closing Gender Gaps” scenario assumes that the participation rate of share of income by the projected fall in labor force participation
women ages 25–54 converges to the participation rate of men ages 25–54 over during 2017–50. If this were to occur at the same rate every year, it
20 years; the “Extending Working Lives” scenario assumes that the participation
would correspond to a loss in potential output of 0.09 percentage
rate of the 55–59 age group converges to the participation rate of the 50–54 age
group over 20 years and that the participation rate of the 60–64 age group point a year over 33 years.
47This scenario assumes unchanged birth rates, as higher female
converges to the participation rate of the 50–54 age group over 40 years; the
“Implementation of Policies” scenario assumes that policies converge to the 10th labor force participation need not go hand in hand with lower
or 90th percentile of the level observed among advanced economies. fertility. Sweden, for example, enjoys both one of the highest female
labor force participation rates and one of the highest fertility ratios
among advanced economies due to policies designed to support
both objectives.

make the decline in the median trend participation mask strikingly different shifts in the workforce
rate more gradual. In 2050, the median aggregate attachment of men and women. In most countries,
participation rate is projected to be 2¾ percent- the aggregate participation rates of men have declined
age points higher than in the baseline scenario. Of since the crisis, broadly in line with changes in the
course, sufficiently large increases in participation rates age structure of populations and the drag from
among older workers, especially among those older the global financial crisis. Women’s participation,
than 65, could entirely offset or even reverse the drag however, increased in most countries, despite aging
from aging.48 and adverse cyclical developments, underscoring the
Finally, the analysis attempts to quantify the extent importance of policies and other factors in shaping
to which policies can offset the projected decline in participation rates.
aggregate participation. In an illustrative scenario, Disparate developments in labor market involve-
policy settings are assumed to converge gradually over ment across different age groups of workers are evident
the next 20 years to their “best possible” levels, defined over the long term. Participation of young men and
as the 90th (or 10th) percentile of the level observed women and prime‑age men has been declining for
among advanced economies (Figure 2.18, panel the past 35 years. Participation of prime-age women
3). The coefficients estimated in the cross-country has increased dramatically since the mid-1980s, and
empirical model are used to forecast the impact of for older workers it has picked up considerably since
these policy changes on trend participation rates by the mid-1990s.
age-gender group, which are then aggregated using The chapter’s analysis suggests that changes in
projected demographic weights. This simple simulation labor market policies and institutions, together with
suggests that bringing policies to what can be viewed structural changes and gains in educational attainment,
as best practice (from the point of view of labor force account for the bulk of the increase in the labor force
participation) can offset some, but not much, of the attachment of prime-age women and older workers
drag from aging. Aggregate participation rates would in the past three decades. Conversely, technological
be about 1¼ percentage points higher than in the advances, namely automation—while beneficial for the
baseline by 2050. economy as a whole—have weighed on the labor sup-
ply of most groups of workers and can partially explain
declining prime-age male participation. Individual-level
Conclusions and Policy Implications evidence confirms the significant impact of vulner-
The increase in longevity is one of the most remark- ability to routinization. Detachment from the labor
able successes in human history (Bloom and others force is significantly more likely among individuals
2015). Yet it could have serious macroeconomic conse- whose current or past occupations are more vulnerable
quences when coupled with the decline in population to automation. But encouragingly, higher spending
growth. Because older workers participate in the labor on education and active labor market programs, and
force at much lower rates, population aging raises access to more diverse labor markets, tend to attenuate
concerns about the supply of labor in advanced econo- this negative effect.
mies, which has implications for potential growth and What does this mean for labor force participation
the sustainability of social insurance systems. prospects in advanced economies? In the absence of
This chapter documents that—despite the accelera- policy efforts, expected demographic developments
tion in population aging over the past decade—many could lead to large declines in aggregate participation
advanced economies have been able to counteract its rates. The chapter’s simulations imply that by 2050,
downward pressure on labor force participation. In overall participation rates could fall by 5½ percentage
about half of advanced economies, the aggregate labor points in the median advanced economy.
force participation rate increased after the global There is, however, scope for policies to counteract
financial crisis. Yet these aggregate developments the forces of aging by making sure those who are
willing to work can do so. In particular, reforming
48Data constraints on participation by age groups of workers older the tax-benefit system, for example, by reducing the
than 65 prevent the simulation of alternative scenarios such as rais- labor tax wedge, along with strengthening policies
ing effective retirement ages to maintain the proportion of life spent
in retirement or indexation of effective retirement ages to healthy that improve the job-matching process, can encour-
life expectancy. age individuals to keep working or seek employment.

There is also strong evidence of the effectiveness of economies over the past three decades—any efforts to
family‑friendly policies that help people combine curb international migration would thus further exac-
market work with the demands of parenthood—public erbate demographic pressure.50
spending on early childhood education and care, Finally, technological advances that transform pro-
flexible work arrangements, and parental leave—in duction processes and reduce the need for labor could
attracting women to the labor force. For older workers, help alleviate the challenges to aggregate growth from
reducing the incentives to retire early, by raising statu- aging. But policymakers should be mindful of the dif-
tory retirement ages or making pension systems more ficult adjustment such transformations may entail for
actuarially fair, could lengthen working lives, although some sectors, occupations, and geographic areas and
care should be taken that reforms do not jeopardize deal with the concerns of workers displaced by tech-
other goals, such as a basic social safety net for vulnera- nology, including through effective support for retrain-
ble individuals.49 ing, skill building, and occupational and geographic
However, the chapter’s simple illustrative simula- mobility. As the chapter’s findings suggest, increasing
tions suggest that even if countries converge to the investment in education and training can not only
best (observed) policy settings for encouraging labor make the workforce more resilient to changing labor
supply, expected demographic shifts may still depress needs, but also encourage labor force participation.
participation rates in advanced economies, taking a Investing more in the education of the young is also
toll on economic activity. Unless technological progress critical to prepare them for the jobs of the future.
delivers offsetting productivity gains, many countries
may need to reconsider immigration policies to boost
domestic labor supply, alongside policies to encourage
older workers to postpone retirement. Although receiv-
ing migrants can pose challenges for host countries, the
50As discussed in Chapter 4 of the October 2016 WEO, cultural
chapter’s analysis suggests that net migration accounts
and language differences, as well as concerns about displacement
for roughly half of the population growth in advanced of native workers, can stir social tensions and provoke a political
backlash against migration in host countries. The prompt integration
of migrants is key to alleviate such concerns. In source countries,
49It is important to recognize that some of these policies may migration can weigh on long-term growth prospects if it is associated
entail significant fiscal costs, while others may be politically challeng- with brain drain, though such effects can be mitigated by remit-
ing because of their cross-generational distributional consequences. tances or diaspora networks.

Box 2.1. Youth Labor Force Participation in Emerging Market and Developing Economies versus
Advanced Economies
Median labor force participation rates for the overall Figure 2.1.1. Labor Force Participation by
working-age population in advanced and emerging Age Group
market and developing economies have fluctuated (Percent)
around 60 percent over the past 25 years. Youth labor
force participation, however, has fallen in both groups Working-age population, median
of economies (Figure 2.1.1).1 Whether these declines Working-age population, interquartile range
Youth, median
are a cause for concern depends largely on whether Youth, interquartile range
they reflect primarily growth in school enrollment or
an increasing share of idle youth. This is particularly 80 1. Advanced Economies
important in emerging market and developing econo-
70
mies, where young people comprise about 18 percent
of the population on average, about 6 percentage 60
points higher than their share in advanced economies.2
50
Motivated by these considerations, this box looks at
how youth labor force participation has evolved in 40
recent years across advanced and emerging market and
30
developing economies.3
Low and declining youth labor force participation 20
rates are a greater cause for concern in emerging 1990 95 2000 05 10 15
market and developing economies than in advanced
economies. In both cases, there has been an uptick 80 2. Emerging Market and Developing Economies
in youth human capital investment (Figure 2.1.2). 70
For the median advanced economy, secondary school
enrollment rose more than 10 percentage points since 60
1990, to about 97 percent in 2010. The pickup in 50
schooling has been even more dramatic in emerging
market and developing economies—median secondary 40
enrollment rose almost 40 percentage points, to about 30
70 percent. However, the lower overall schooling rate
and similar youth labor force participation suggest 20
1990 95 2000 05 10 15
that a larger share of emerging market and develop-
ing economy youth is neither in the labor force nor
Sources: International Labour Organization; and IMF staff
studying. There are, moreover, significant differences in calculations.
The authors of this box are John Bluedorn and Davide Malac-
rino with research assistance from Daniela Muhaj.
1Age ranges defining the youth population sometimes differ
enrollment rates across regions—in emerging Europe
across data sets and publications. Unless indicated otherwise, the
International Labour Organization definition of 15–24 years old
enrollment rates are nearly the same as in advanced
is used. Working-age population is 15–64 years old. economies, while sub-Saharan Africa, although
2Country group median population shares in 2015 (United improved, is well behind.
Nations, Department of Economic and Social Affairs, Population The gender gap in youth labor force participation is
Division 2017). also much larger in emerging market and developing
3Ahn and others (forthcoming) investigate in greater depth the
patterns and drivers of youth labor market outcomes in emerging

economies (Figure 2.1.3). Median youth labor force
market and developing economies, including potential policy participation has trended down for both females and
implications. males in advanced economies: the initial female partic-

Box 2.1 (continued)

Figure 2.1.2. Median Secondary Enrollment Figure 2.1.3. Youth Labor Force Participation
by Geographic Region by Gender
(Percent) (Percent)
100 Male, median Male, interquartile range

Female, median Female, interquartile range
80 80 1. Advanced Economies
70
60 60
50
40 40
30
20 20
1990 95 2000 05 10 15
AEs EAP LAC SA
EMDEs ECA MENA SSA 80 2. Emerging Market and Developing Economies
0
1990 95 2000 05 10
70
Sources: Lee and Lee, Long-Run Education Dataset (2016); 60

and IMF staff calculations.
Note: Adjusted gross enrollment ratios are shown. 50
AEs = advanced economies; EAP = East Asia and Pacific;
ECA = Europe and central Asia; EMDEs = emerging market 40
and developing economies; LAC = Latin America and
Caribbean; MENA = Middle East and North Africa; 30
SA = south Asia; SSA = sub-Saharan Africa.
20
1990 95 2000 05 10 15
Sources: International Labour Organization; and IMF staff

ipation gap of about 10 percentage points has shrunk calculations.
to just a couple of percentage points in recent years.
By contrast, the gender gap remains very large in
emerging market and developing economies, at about
20 percentage points. counterfactual likelihood can also be calculated for
Individual-level census data allow for deeper inves- each young woman: the predicted probability of labor
tigation into the dynamics of the youth gender gap force participation if she were male, holding all other
across countries. From these data, for each country observable characteristics constant. The average dif-
and year, the predicted probability of participating ference between these two quantities at the individual
in the labor market can be calculated for each young level yields an alternative measure of the gender gap
woman, given her observable characteristics.4 The
force, or unoccupied) conditional on individual-level observable
4More specifically, the analysis estimates multinomial logit characteristics (such as marital status, parent or not, educational
probability models by country-year, gender, and age group attainment, and others). The models are then used to calculate
(young and not young) for individual labor market outcomes predicted probabilities at the individual level, which can be
(that is, in school, unemployed, employed, out of the labor aggregated up to get a sense of the average behavior.

Box 2.1 (continued)

in labor force participation facing young women in Figure 2.1.4. Implied 10-Year Improvement
that country and year, with the advantage that it takes in Country Gender Gaps for Youth
account of the effects of non-gender-related individual (Percentage points)
characteristics.
90
There is a wide range of youth gender gaps across
countries, from about 5 percentage points to almost Advanced economies
80 Emerging market and
70 percentage points in the latest year for which data developing economies
are available (Figure 2.1.4).5 That said, there has been 70
a broad-based improvement—most points lie below
the 45‑degree line, indicating that the gender gap 60
Latest year available

has shrunk. While this decline is encouraging, there 50
is still a long way to go to fully close the gender gap
in youth labor force participation. As discussed in 40
Elborgh‑Woytek and others (2013), Gonzales and
others (2015a), and Ahn and others (forthcoming), 30
potential policy responses include a mix of labor mar- 20
ket, social policy, and other reforms.
10
5See Minnesota Population Center (2017). IPUMS Inter- 0

national underlying data set sources: Argentina (National 0 10 20 30 40 50 60 70 80 90
10 years earlier
Institute of Statistics and Censuses), Austria (National Bureau of
Statistics), Bangladesh (Bureau of Statistics), Bolivia (National
Institute of Statistics), Botswana (Central Statistics Office), Brazil Sources: Integrated Public Use Microdata Series
(Institute of Geography and Statistics), Cambodia (National International; and IMF staff calculations.
Institute of Statistics), Colombia (National Administrative Note: A country’s gender gap is defined as the average
Department of Statistics), Costa Rica (National Institute of across the individual-level difference between a young
woman’s predicted probability of labor force participation
Statistics and Censuses), Dominican Republic (National Statistics
given her observables and her predicted probability given
Office), Ecuador (National Institute of Statistics and Censuses), the same observables if she were male (a counterfactual).
El Salvador (General Directorate of Statistics and Censuses), See text footnote 4 for a brief description of the underlying
France (National Institute of Statistics and Economic Studies), probability models. Each country shown has at least two
Ghana (Ghana Statistical Services), India (Ministry of Statistics years of census data, but the time difference varies between
and Programme Implementation), Indonesia (Statistics Indo- 5 and 20 years. For comparability across countries, the
nesia), Iran (Statistical Center), Kyrgyz Republic (National Statis- latest data are taken as given and the change in gender gap
tical Committee), Malaysia (Department of Statistics), Mexico is normalized to back out the implied gender gap 10 years
(National Institute of Statistics, Geography, and Informatics), earlier for each country. Youth are defined as 15–29 years
old.
Nicaragua (National Institute of Statistics and Censuses), Pan-
ama (Census and Statistics Directorate), Peru (National Institute
of Statistics and Informatics), Portugal (National Institute of Sta-
tistics), Romania (National Institute of Statistics), South Africa
(Statistics South Africa), Tanzania (National Bureau of Statistics),
Trinidad and Tobago (Central Statistical Office), United
States (Bureau of the Census), Uruguay (National Institute of
Statistics), Venezuela (National Institute of Statistics), Zambia
(Central Statistical Office).

Box 2.2. Permanently Displaced? Labor Force Participation in US States and Metropolitan Areas
The decline in US labor force participation over the Figure 2.2.1. Labor Force Participation and
past two decades has been widely documented and, as Change in Labor Force Participation by State
highlighted in this chapter, deviates from the evolution
of participation in many advanced European economies. 1. Change in Participation Rate, 2000–16
Many hypotheses have been put forth for this (Percentage points)
puzzling decline (alongside the effects of aging). These
include cyclical effects and the severity of the Great
Recession, structurally lower labor demand brought on
by the forces of trade and technology (especially for < = –6
[–5; –6]
those with low skills), and lower labor supply (because [–4; –5]
of incarceration, disability, and pain) as well as waning [–3; –4]
cohort effects for women’s participation and the role [–3; 0]
>=0
of policy.1
This box examines regional differences in labor force
80 2. Change in Labor Force Participation
participation in the United States to shed light on Rate, 1976–2000
Participation rate, 2000 (percent)

MN
the factors that may underpin participation declines. 75 IA SD NE
UT WI
It documents a broad-based decline in participation, ND
70 NV
especially in rural areas. Moreover, it finds that lower MO
WA HI
participation in metropolitan areas is strongly associ- 65
FL
ated with exposure to routinization and offshoring.
60
This supports hypotheses about the role of deterio-
rating job opportunities for some workers as a result 55
of technology and globalization in their increasing
50
detachment from the workforce (in line with the find- 50 55 60 65 70 75 80
ings of Acemoglu and Autor 2011, Council of Eco- Participation rate, 1976 (percent)
nomic Advisers 2016, and Krause and Sawhill 2017).
Broad-Based Decline across States Rate, 2000–16
75
The decline in participation rates is very broad based ND
across US states (Figure 2.2.1, panel 1). Between 2000 70 DC
and 2016, participation declined in almost all states,2 AK
65 DE
but declines were most pronounced in the Southeast3 NV
and parts of the Midwest and West.4 The decline was 60
OR MI GA
much smaller in the Mid-Atlantic5 and New England.6 AL
55 SC
These declines stand in marked contrast to pre-2000 MS KY
developments, when participation increased almost 50
50 55 60 65 70 75 80
across the board by an average of more than 5 per- Participation rate, 2000 (percent)
centage points between 1976 and 2000 (Figure 2.2.1,
panels 2 and 3). Sources: US Bureau of Labor Statistics; US Census Bureau;
Note: Red markers denote states with decreases (panel 2) or
The authors of this box are Benjamin Hilgenstock and particularly pronounced decreases (panel 3). Gold markers
Zsóka Kóczán. denote states with increases (panel 3) or particularly
1See, for example, Aaronson and others (2006); Fallick and pronounced increases (panel 2). Labels in the figure use
International Organization for Standardization (ISO) state
Pingle (2007); Blau and Kahn (2013); Council of Economic
codes.
Advisers (2014, 2016); Balakrishnan and others (2015); Case
and Deaton (2017); Krause and Sawhill (2017); and Krueger
(2017). See Abraham and Kearney (2018) for a recent review.
2The District of Columbia is treated as a state for the purpose More Pronounced outside Metropolitan Areas
of this box. Similar patterns can be observed at the metropolitan
3Alabama, Georgia, Kentucky, Mississippi, South Carolina.
4Alaska, Michigan, Nevada, Oregon. area level (Figure 2.2.2). Labor force participation rates
5Maryland, New Jersey, Pennsylvania. declined between 2000 and 2016 in three-quarters of
6Connecticut, Massachusetts. metropolitan areas; among the 50 most populated areas,

Box 2.2 (continued)

Figure 2.2.2. Change in Labor Force Figure 2.2.3. Decomposition of Labor Market
Participation Rate by Metropolitan Area Changes in Metropolitan Areas
(Percentage points)
Rate, 1990–2000
Atlanta 4
75 Boston
Washington Dallas Change in employment rate
70 Chicago Change in unemployment rate
3
Philadelphia Change in inactivity rate
65 Miami
New York Houston 2
60 Los Angeles
1
55
50 0
50 55 60 65 70 75 80
–1
Rate, 2000–16
–2
75
70 Los Angeles –3
New York Boston
65 –4
Atlanta 1990–2000 00–16 00–08 08–16
60
Chicago
Dallas Sources: US Bureau of Labor Statistics; US Census Bureau;
55 Philadelphia
Miami Houston Washington
50 Note: Employment rate, unemployment rate, and inactivity
50 55 60 65 70 75 80 rate are defined, respectively, as total employment, total
Participation rate, 2000 (percent) unemployment, and total inactive population as a
percentage of total population. Numbers represent simple
4 3. Changes in Metropolitan Areas versus averages across metropolitan areas.
Changes in States, 2000–16
2
0
–2 However, declines were typically larger in a state as
–4 a whole than in its metropolitan areas, exacerbating
–6
urban-rural differences (Figure 2.2.2, panel 3; in line
Change, surrounding state(s) with the findings of Weingarden 2017).
–8
Change, metropolitan area
–10
The Role of the Crisis and Changing Margins
of Adjustment
Detroit
Atlanta
Minneapolis
Los Angeles
San Francisco
Phoenix
Seattle
Dallas
Houston
Chicago
Miami
Washington
Boston
Philadelphia
New York
The decline in participation became more widespread

after the global financial crisis, when lower employ-
ment increasingly translated into lower participation
Sources: US Bureau of Labor Statistics; US Census Bureau;
and IMF staff calculations. (Figure 2.2.3). Before 2000, employment increased, on
Note: In panels 1 and 2, red markers display metropolitan average, and was matched by declines in unemployment
areas with decreases in labor force participation rates. The and increases in participation. After 2000, employment
10 largest areas by 2016 population are labeled. In panel 3, if
metropolitan areas are assigned to multiple states, blue bars declined, matched by increasing unemployment and
show population-weighted averages of surrounding states. falling participation. Although most of the decline
in employment translated into rising unemployment
before the crisis, after the crisis participation fell sharply.
only 16 (typically with already high participation) dis-
Drivers of Labor Force Participation
played increases, of which most were comparably small.7
Cross-sectional regressions at the metropolitan area
7Because several metropolitan areas cross state lines, Fig-
level examine the association between 2000–16 changes
ure 2.2.2, panel 3, assigns metropolitan areas to states based on in labor force participation rates and cyclical condi-
the definition of the US Office of Management and Budget. tions, aging, and education, as well as the impact of

Box 2.2 (continued)

Figure 2.2.4. Routine and Offshoring technology and trade, captured by the initial exposures
Exposure by State to routinization and offshoring.8 These results confirm
the significant effects of cyclical conditions, aging, and
1. Routine Exposure by State, 2000 education highlighted in the chapter (Table 2.2.1).9
(Index) Furthermore, metropolitan areas with higher initial
exposures to automation and offshoring due to their
occupational employment compositions saw larger
subsequent declines in participation rates.10 This suggests
that automation and offshoring may have permanently
displaced some workers, even if their effects on the econ-
omy as a whole were beneficial through the creation of
> = .55 job opportunities in other sectors or productivity gains.
[.50; .55] In the short and medium term, support should thus
[.45; .50]
< = .45
be provided to workers displaced as a result of automa-
tion and globalization to dampen the negative effects of
labor market shocks that may be highly concentrated in
2. Offshoring Exposure by State, 2000 some sectors, occupations, or geographic areas.
(Index)
8Exposures to routinization and offshoring act as proxies for
the share of jobs at risk of being automated or offshored (see

Chapter 3 of the April 2017 WEO; and Das and Hilgenstock,
forthcoming). Regressions include state fixed effects.
9State-level cyclical effects are also documented by Erceg and
Levin (2014), Council of Economic Advisers (2014), and Bal-

> = .3 akrishnan and others (2015). Dao, Furceri, and Loungani (2014)
[.2; .3] highlight the increasing role of participation as an absorber of
[.1; .2]
state-level labor demand shocks. Sanchez, Shen, and Peng (2004)
< = 0.1
look at the impact of mobility on employment outcomes at the
metropolitan area level.
Sources: US Bureau of Labor Statistics; US Census Bureau; 10Figures 2.2.4 and 2.2.1, panel 1, suggest that this link holds
and IMF staff calculations. at the state level as well: exposure to routinization and offshoring
was especially high in the Southeast and Midwest, which also
exhibited the largest declines in participation.
Table 2.2.1. Drivers of Labor Force Participation Rates in US Metropolitan Areas

(1) (2) (3) (4) (5)
Average Real GDP Growth 0.442*** 0.444*** 0.368***
(0.145) (0.144) (0.140)
Change in Old-Age-Dependency Ratio –0.144*** –0.130*** –0.152***
(0.040) (0.041) (0.038)
Change in Postsecondary Share 0.037 0.040* 0.053**
(0.023) (0.023) (0.022)
Initial Exposure to Routinization –2.811** –2.492**
(1.153) (1.222)
Initial Exposure to Offshoring –4.212*** –4.929***
(0.935) (0.962)
Observations 370 370 335 335 335
R 2 0.289 0.319 0.360 0.369 0.414
Note: Standard errors are in parentheses. The dependent variable is change in labor force participation rate.
*p < 0.1; **p < 0.05; ***p < 0.01.

Box 2.3. Still Attached? Labor Force Participation Trends in European Regions

In addition to the significant cross-country variation Figure 2.3.1. Change in Labor Force
in Europe documented in this chapter, there are also Participation by Region, 2000–16
large within-country differences in labor force partici- (Percentage points)
pation.1 As in the United States (Box 2.2), rural areas
saw larger drops or smaller increases in labor force
participation rates than urban areas, though declines >=5
typically started later and were less broad based than [0; 5]
[–5; 0]
in the United States. However, European regions < –5
more exposed to routinization and offshoring through
their initial occupation mix did not experience larger
declines in participation over a longer horizon, once
labor markets had time to adjust.2
Heterogeneity across Regions

Labor force participation declined in about one-third
of European regions between 2000 and 2016. Although
some countries exhibit similar patterns across regions
(for instance, participation declined in all regions in
Norway and Romania and increased in all regions in
Spain and Sweden), others (such as France, Germany,
Portugal, and the United Kingdom) show signifi-
Sources: Eurostat; and IMF staff calculations.
cant within-country differences (Figure 2.3.1).3 This
contrasts with broad-based declines observed across US
states and metropolitan areas (Box 2.2). Participation
declined in only about 27 percent of European regions However, as in the United States, there is a divide
between 2000 and 2008 and in about 45 percent of between urban and rural regions, with the latter show-
regions between 2008 and 2016.4,5 ing larger decreases or smaller increases in participa-
tion rates (Figure 2.3.2, panel 3).
The authors of this box are Benjamin Hilgenstock and
Zsóka Kóczán. The Role of the Crisis and Changing Margins
1A great deal of literature analyzes the drivers of (especially
of Adjustment
women’s) participation in Europe at the country level, focusing
predominantly on the role of policy (such as incentives for Margins of adjustment changed in Europe too,
part-time work and family-friendly measures—see, for example, though later than in the United States (Figure 2.3.3).
Genre, Gómez-Salvador, and Lamo 2010; Cipollone, Patac- While in the United States employment started to
chini, and Vallanti 2013; Thévenon 2013; and Miani and decline around 2000, employment increased, on
Hoorens 2014), and cohort effects (see, for example, Balleer,
Gómez-Salvador, and Turunen 2014 for a cross-country study
average, in European regions until the crisis, matched
and Euwals, Knoef, and van Vuuren 2011 for the Netherlands). by falling unemployment and rising participation. As
Dauth, Findeisen, and Suedekum (2014) look at the impact of employment started to decline after the crisis, this
trade on German labor markets. The key contribution of this translated into rising unemployment, with, on average,
box is its focus on variation at the regional level, in particular the still small increases in participation.
impact of technology.
2In the following the term “regions” refers to Eurostat’s Nomen-
Drivers of Labor Force Participation
clature of Territorial Units for Statistics (NUTS) 2 level regions
wherever data are available. When these data are not available, As in Box 2.2, cross-sectional regressions (here at the
NUTS 1 and NUTS 0 regions are used instead. Most of the box level of European regions) examine the link between
includes regions in advanced Europe as well as in emerging Europe; 2000–16 changes in labor force participation and cycli-
for consistency with the chapter, the regression analysis focuses on
advanced Europe only. Simple averages are used throughout.
3See also Centre for Cities (2018) on the economic divide in
the United Kingdom. and Switzerland); continued to fall in 10 percent (for example,
4Regional participation rates are not available before 2000. Romania and the United Kingdom); started to decline in 35 per-
5Furthermore, this hides a great deal of underlying disparity: cent (for example, Belgium, Denmark, the Netherlands, and
participation continued to increase after as well as before the Portugal); but started to increase in 18 percent (for example, the
crisis in 38 percent of regions (for example, Austria, Germany, Czech Republic, Poland, and the Slovak Republic).

Box 2.3 (continued)
Figure 2.3.2. Change in Labor Force Figure 2.3.3. Decomposition of Labor Market
Participation Rate by Region Changes
(Percentage points)
Rate, 2000–08
4
75 Paris London
Lisbon Change in employment rate
70 Madrid Berlin
3 Change in unemployment rate
65 Brussels Athens Change in inactivity rate
60 Budapest
2
55 Rome
Warsaw
50
1
45
40
40 45 50 55 60 65 70 75 80 0
–1
Rate, 2008–16
75 London –2
70 Berlin
Warsaw
65 Budapest –3
2000–08 08–16
60 Brussels
Madrid
55 Rome Paris Sources: Eurostat; and IMF staff calculations.
50 Lisbon Note: Employment rate, unemployment rate, and inactivity
Athens rate are defined, respectively, as total employment, total
45 unemployment, and total inactive population as a percentage
40 of total population. Numbers represent simple averages
40 45 50 55 60 65 70 75 80 across regions.
20 3. Population Density and Changes in (Box 2.2), European regions more exposed to routini-
Labor Force Participation Rates
15 zation and offshoring as a result of their 2000 occupa-
2000–16 (percentage points)
Change in participation rate,
10 tional mix experienced, if anything, larger participation

5 gains during 2000–16 (Figures 2.3.4 and 2.3.5).7
There are several possible explanations for this
0
finding, which seems contrary to the patterns observed
–5
in the chapter and across US metropolitan areas. First,
–10
R 2 = 0.0344
unlike the analysis in the main text, which examines
–15 the participation consequences of annual variation in
–20 the relative price of investment as a proxy for firms’
0 1 2 3 4 5 6 7 8 9 10
ln (average population density, 2000–16) incentives to automate routine tasks, this specification
focuses on changes in participation over a longer time
Sources: Eurostat; and IMF staff calculations. horizon, allowing labor markets to adjust to demand
Note: In panels 1 and 2, red markers display regions with shocks. The positive correlation across European
decreases in labor force participation rates. The 10 largest regions could thus be picking up an added-worker
regions by 2016 population are labeled.
effect. Secondary earners may enter the labor market
as a result of lower household income. This would be
cal conditions, aging, and education.6 Subnational evi-
7Dauth, Findeisen, and Suedekum (2014) find that the
dence confirms the significant effects of aging, cyclical
rise of China and eastern Europe in the world economy
conditions, and education highlighted in this chapter
caused substantial job losses in German regions specializing in
(Table 2.3.1). However, unlike in the United States import-competing industries, but caused employment gains in
export-oriented industries, with an overall positive effect of trade
6Regressions control for country fixed effects. integration on employment.

Box 2.3 (continued)
Table 2.3.1. Drivers of Labor Force Participation Rates in European Regions
(1) (2) (3) (4) (5)
Average Real GDP Growth 0.457 1.061*** 1.176***
(0.325) (0.383) (0.387)
Change in Old-Age-Dependency Ratio –0.282*** –0.211*** –0.218***
(0.056) (0.072) (0.072)
Change in Postsecondary Share 0.187*** 0.145** 0.117*
(0.053) (0.069) (0.070)
Initial Exposure to Routinization 4.258** 5.435***
(1.995) (1.815)
Initial Exposure to Offshoring 4.157** 5.518***
(1.968) (1.846)
Observations 148 148 223 140 139
R 2 0.645 0.644 0.646 0.730 0.729
Note: Standard errors are in parentheses. The dependent variable is change in labor force participation rate.
*p < 0.1; **p < 0.05; ***p < 0.01.
Figure 2.3.4. Initial Routine Exposure by Figure 2.3.5. Initial Offshoring Exposure by
Region, 2000 Region, 2000
(Index) (Index)
> = .2 > = .2
[.1;.2] [.1;.2]
[–.1;.1] [–.1;.1]
< –.1 < –.1
Sources: Eurostat, European Union Labour Force Survey; and Sources: Eurostat, European Union Labour Force Survey; and
IMF staff calculations. IMF staff calculations.
consistent with the sharp rise in female participation States. The latter is consistent with the significantly
observed in most European regions and the rise in larger contribution of policy to labor force participa-
two-earner households documented in the chapter. tion in Europe relative to the United States, which is
Second, institutional frameworks and policies in documented in this chapter.
Europe may have allowed those potentially affected Striking within-country differences in the evolution
by routinization and offshoring to remain attached to of labor force participation have important implica-
the workforce and/or encouraged new entrants to the tions for policy—they call for more explicit recogni-
labor market. The former is consistent with the smaller tion of the spatial dimension of economic vulnerability
changes in the occupational mix in European coun- given that short- and medium-term costs not only are
tries over this time period, which suggests that fewer concentrated in particular sectors and occupations but
jobs were automated or offshored than in the United also affect different places in different ways.

Box 2.4. Storm Clouds Ahead? Migration and Labor Force Participation Rates
As discussed in the chapter, slowing population Figure 2.4.1. Contributions of Natural
growth and rising life expectancy will put significant Population Growth and Net Migration to Total
downward pressure on labor supply. Even sizable gains Population Growth
in labor market participation of those more margin- (Percent)
ally attached to the labor force, such as women and
older workers, could be ultimately outweighed by the Total population growth
Natural population growth
pressure of aging. In this context, many argue that Net migration
international migration could bring significant bene-
fits, by boosting labor supply in recipient economies 0.8 1. All Advanced Economies
while leveraging the demographic dividend in other 0.7
parts of world. Net migration has accounted for about 0.6
half of the population growth in advanced economies 0.5
since the mid-1980s, while natural population growth 0.4
(measured as the difference between fertility and mor-
0.3
tality) has been falling (Figure 2.4.1).
0.2
This box examines the effects of migration on
future labor force participation in (receiving) advanced 0.1
economies, as well as the drivers of migrants’ decision 0.0
1985 90 95 2000 05 10 15
to participate.
It documents that migration assumptions, already 0.6 2. European Advanced Economies
embedded in population projections for advanced
0.5
economies, play a very significant role in alleviating
aging pressures. In the absence of migration, the 0.4
decline in participation would be significantly deeper.
0.3
Support for migrants’ rapid labor market integration
will yield significant further gains. 0.2
Migration: The Role of Age Composition Effects 0.1
One way migrants affect the labor supply in recip- 0.0

1985 90 95 2000 05 10 15
ient economies is through age composition. Migrants
are more likely to be of prime working age than 1.4 3. United States
natives because they typically arrive after they have
1.2
completed their education and often leave when they
retire (Figure 2.4.2, panel 1). Because participation is 1.0
highest among those of prime working age, age com- 0.8
position has significant implications for overall labor 0.6
force participation.
0.4
Figure 2.4.2, panel 2, illustrates the expected
evolution of aggregate labor force participation in 0.2
advanced European economies under Eurostat’s 0.0
alternative migration scenarios; differences stem solely 1985 90 95 2000 05 10 15
from changes in the age composition of the countries’ Sources: United Nations; and IMF staff calculations.
populations as a result of net migration.1 Under the Note: Panel 1 is based on a balanced sample of 34
advanced economies. Natural population growth refers to
the difference between fertility and mortality.
The authors of this box are Benjamin Hilgenstock and
Zsóka Kóczán.
1The Eurostat baseline scenario is broadly based on trend
extrapolation until 2050 (EC 2017). It would imply, for

instance, an increase in Germany’s migrant stock from the
current 14 percent to 29 percent. The high (low) migration

Box 2.4 (continued)

baseline scenario, the average aggregate participation Figure 2.4.2. Projected Evolution of Labor
rate would decline by 7.4 percentage points by 2050. Force Participation Rates
Allowing for an increase in net migration could offset
some of this decline: the drop would be 0.8 percentage 20 1. Age Profile of Natives and Migrants
18 (Percent of population)
point less under the assumption of high migration
16 Natives
(it would be 0.8 percentage point more under low Migrants
14
migration). More restrictive immigration policies
12
would significantly exacerbate the negative effect of 10
population aging on participation. Strikingly, if no 8
new migration is allowed, the decline in participation 6
would be 2.7 percentage points larger. These effects 4
would be especially large in high-migration countries 2
(Figure 2.4.2, panel 3). 0
0–9
10–19
20–29
30–39
40–49
50–59
60–69
70–79
80–89
90–99
Participation Effects of Migration
While migration can boost aggregate participation 62 2. Participation Rate, All European Advanced
rates through compositional shifts, it is important to Economies, 2015–50
60
(Percent)
recognize that participation rates differ significantly 58 Baseline migration
between migrants and natives, and these differences High to low migration
56
vary by gender and age.
54
Disaggregated data from 24 advanced European
economies suggest that young migrants are more likely 52
to be in the labor force than young natives (42 percent 50 No migration
versus 36 percent; young natives are, on average, more 48 United Nations baseline
likely to be in education), but participation among 46
2015 20 25 30 35 40 45 50
migrants 55 and older is slightly lower than for natives
in the same age group (5 percent versus 6 percent).2 A
3. Change in Participation Rate, 2015–50
close look at prime-age workers shows that participa- 8
(Percentage points)
tion of prime-age men is very similar for natives and 4 Baseline migration No migration
migrants. The most significant difference relates to the High to low migration UN baseline
0
participation of prime-age women, with significantly
–4
lower participation among migrant women (75 per-
cent versus 81 percent; Figure 2.4.3, panel 1). –8
However, migrant participation rates converge –12
toward those of natives over time: participation
–16
increases with years in the host country, especially
for prime-age women (Figure 2.4.3, panel 2). This –20
DEU GBR FRA ITA ESP NLD SWE BEL
Sources: Eurostat; United Nations; and IMF staff calculations.

scenarios refer to a one-third increase (decrease) in net migration Note: In panels 1 and 2, countries included are AUT, BEL,
relative to the baseline (so, for Germany, would result in migrant CYP, CZE, DEU, DNK, ESP, EST, FIN, FRA, GBR, GRC, IRL, ITA,
stocks of 25 and 33 percent, respectively, by 2050). The United LTU, LUX, LVA, MLT, NLD, NOR, PRT, SVK, SVN, and SWE.
Nations baseline scenario assumes a continuation of recent Labels in the figure and note use International Organization
migration trends for nonrefugee flows until 2050, but also for Standardization (ISO) country codes. Detailed migration
considers the country’s migration policy stance (see UN 2017 for scenarios are based on Eurostat data and are compared to
the United Nations baseline scenario.
details). While, on average, this produces estimates broadly con-
sistent with the European Union’s low-migration scenario, this
is not necessarily the case for individual countries. Figure 2.4.2,
panel 2, shows population-weighted averages across countries.
2Disaggregated data are from Eurostat’s European Labour
Force Survey. The statistics described above are from a random

sample of 10,000 respondents per country per year.

Box 2.4 (continued)

Figure 2.4.3. Labor Force Participation Rates Figure 2.4.4. Change in the Odds of Being
of Prime-Age Natives and Migrants, 2000–16 Active
(Percent) (Percent)
110 1. Participation Rates by Gender and

Immigration Status Upper secondary
100 education
90 Tertiary education
Men, 25–54,
80 Number of children in native
household Men, 25–54,
70
Native men One adult with children migrant
60 Migrant men
Native women Couple without children
50 Migrant women Women, 25–54,
40 native
Couple with children Women, 25–54,
All DEU GBR FRA ITA ESP NLD SWE BEL
Other household migrant
structure All, 55 and older,
110 2. Evolution of Participation Rates by Gender Other employed adult(s) native
and Immigration Status in household All, 55 and older,
100 migrant
Routinizability of (last)
90 occupation
80
–50 0 50 100 150 200 250 300
70
60 Sources: Eurostat, European Union Labour Force Survey; and
50 IMF staff calculations.
Native men Native women Note: Logit regressions are based on a 10,000 respondents
40 Migrant men Migrant women per country per year random sample of 18 countries. Only
30 effects significant at the 10 percent level are shown. The
1 2 3 4 5 6 7 8 9 >=10 base category for education is “up to lower secondary
education.” For family composition the base category is “one
adult without children.” The routine exposure coefficient is
Sources: Eurostat, European Union Labour Force Survey; and scaled by the difference between the 25th and 75th
IMF staff calculations. percentiles. Regressions also control for age, gender,
Note: In panel 1, “All” refers to European advanced urban/rural location, year, country and region fixed effects,
economies as listed in the note to Figure 2.4.2. Panel 2 is and the output gap. Standard errors are clustered at the
based on the eight countries listed in panel 1. x-axis in panel country-year level.
2 denotes years since migration.
Migrants’ Participation Decisions

effect holds up even when controlling for individual
What is holding back migrants’ involvement in the
and household characteristics: an additional year in
labor market? Figure 2.4.4 builds on the logit specifi-
the host country is estimated to increase the odds of
cation estimated in the chapter—looking at the effects
participation by 5–6 percent.
of individual and household characteristics on individ-
Abstracting from the age composition effect dis-
ual participation decisions—but here it is examined
cussed earlier, allowing migrants’ participation rates to
separately for migrants and natives.
increase to natives’ participation rates would result in
In many ways, migrants’ participation decisions are
an additional 1.4 percentage point increase in overall
shaped by the same factors that shape those of natives.
participation (relative to a no convergence scenario),
Those who are more educated participate more, house-
even holding the relative shares of the age groups in
the population constant.3
2000–16 average shares and examines what the overall rate
would be if migrants’ participation equaled that of natives. Given
3This example keeps population shares of the eight groups that young migrants’ participation exceeds young natives’ this
(young, prime-age men, prime-age women, and people 55 is assumed to stay constant; prime-age and 55-plus migrants’
and over for natives and migrants separately) constant at their participation are assumed to increase to natives’ levels.

Box 2.4 (continued)

hold composition matters, and the threat of automa- composition (see also Chapter 4 of the October 2016
tion is linked to a lower likelihood of being active. World Economic Outlook [WEO]). This could help
However, the results also point to significant differ- mitigate some of the future negative effects of aging
ences relative to natives. Although higher education and help make social safety nets more sustainable in
increases the odds of being active for both migrants these economies.
and natives, the effects are significantly smaller for Higher migration flows could contribute to labor
migrants, likely pointing to difficulties in the recog- supply and the host economy more broadly as well—
nition of foreign qualifications or language barriers to increasing output per capita by boosting demand and
labor market integration. investment, contributing to technological progress,
The effects of household composition are much larger and increasing labor productivity, including through
for migrants: being married and having children has skill complementarity.4
larger negative effects on the participation of migrant
women than on that of native women. Local labor mar- 4See Chapter 4 of the October 2016 WEO for a summary;
ket effects are also weaker for migrant women. see also Peri and Sparber (2009); Hunt and Gauthier-Loiselle
(2010); Farré, González, and Ortega (2011); D’Amuri and Peri
Policies for Migrant Integration (2014); Ortega and Peri (2014); Alesina, Harnoss, and Rapoport
These results suggest that policies that support (2015); Cattaneo, Fiorio, and Peri (2015); Peri, Shih, and Spar-
migrant integration, such as recognition of educational ber (2015); Aiyar and others (2016); and Jaumotte, Koloskova,
and Saxena (2016). At the same time, the impact of migration
qualifications or language training, could increase the on average wages or employment of native workers is found to
positive effect of migration on participation in (receiv- be limited (see Card 1990; Peri 2014; IMF 2015; and Aiyar and
ing) advanced economies, beyond its effects on age others 2016).

Annex 2.1. Data Sources and Country Coverage Kong Special Administrative Region, Macao Spe-
The primary data sources for this chapter are the cial Administrative Region, Malta, Puerto Rico, San
Organisation for Economic Co-operation and Devel- Marino, and Taiwan Province of China). However, due
opment, IMF World Economic Outlook (WEO) data- to data limitations, the included economies vary across
base, and United Nations World Population Prospects. the analyses, as indicated in Annex Table 2.1.2. The
The micro-level analysis is based on data from the shift‑share analysis relies on a sample of 32 advanced
2000–16 European Union Labour Force Surveys by economies during 1980–2016 for which detailed data
the European Commission, which are available from on labor force participation by age group and gender
Eurostat. All data sources used in the main analysis are available. The cross-country analysis on the role
(excluding boxes) are listed in Annex Table 2.1.1. of policies and other factors is based on annual data
The sample consists of the 39 economies classified for 23 advanced economies during 1980–2011, which
as advanced economies in Table B of the April 2018 were classified as advanced economies for the entire
WEO, excluding the smallest economies (that is, Hong sample period and for which data on policy variables
Annex Table 2.1.1. Data Sources

Indicator Source
Labor Force Participation OECD, Employment database
Labor Force Participation by Education Eurostat; National authorities
Employment Rate OECD, Employment database
Unemployment Rate IMF, WEO database
Output Gap IMF, WEO database
Crisis Indicator Gourinchas and Obstfeld (2012)
Relative Price of Investment IMF, WEO database
Routine Exposure Das and Hilgenstock (forthcoming) based on Autor and Dorn (2013), Eurostat, and
population censuses
Trade Openness IMF, WEO database
Sectoral Employment of Industry and Services World Bank, World Development Indicators database; European Union, Level Analysis
of Capital, Labour, Energy, Materials, and Service inputs (EU KLEMS)
Urban Population World Bank, World Development Indicators database
Population by Education (primary, secondary, tertiary) Barro-Lee Educational Attainment data set
Labor Tax Wedge OECD, Tax database; Bassanini and Duval (2006); Chapter 3 of the April 2016 WEO
Unemployment Benefits OECD, Benefits and Wages: Statistics
Public Spending on ALMP OECD, Social Expenditure database
Migration Policies International Migration Institute, DEMIG POLICY database
Union Density OECD, Employment database
Coordination of Wage Setting Amsterdam Institute for Advanced Labour Studies, Database on Institutional
Characteristics of Trade Unions, Wage Setting, State Intervention, and Social Pacts
Public Spending on Early Childhood Education and Care OECD, Social Expenditure database
Part-Time Employment OECD, Employment database
Job-Protected Maternity Leave OECD, Family database
Statutory Retirement Age International Social Security Association, Social Security Programs throughout the World
Old-Age-Pension Spending OECD, Social Expenditure database
Incapacity Spending OECD, Social Expenditure database
Implicit Tax on Continued Work Duval (2004); Chapter 3 of the April 2016 WEO
Pension Replacement Ratio Luxembourg Income Study database
Population Projections United Nations World Population Prospects, 2017 revision
School Enrollment OECD, Education database
Returns to Education Luxembourg Income Study database
Education Spending Eurostat
Employment Protection OECD, Employment database
Source: IMF staff compilation.
Note: ALMP = active labor market programs; OECD = Organisation for Economic Co-operation and Development; WEO = World Economic Outlook.

Annex Table 2.1.2. Country Coverage

Exercise Countries
Shift-Share Australia, Austria, Belgium, Canada, Cyprus, Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece,
Analysis Iceland, Ireland, Israel, Italy, Japan, Korea, Latvia, Lithuania, Luxembourg, Netherlands, New Zealand, Norway,
Portugal, Slovak Republic, Slovenia, Spain, Sweden, Switzerland, United Kingdom, United States
Aggregate Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Japan, Korea,
Analysis Luxembourg, Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, Switzerland, United Kingdom, United States
Micro-Level Austria, Belgium, Cyprus, Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Iceland, Ireland, Italy,
Analysis Latvia, Lithuania, Netherlands, Norway, Portugal, Slovak Republic, Slovenia, Spain, Sweden, Switzerland, United Kingdom
Cohort-Based Australia, Belgium, Canada, Denmark, France, Germany, Greece, Israel, Italy, Japan, Luxembourg, Norway, Portugal,
Analysis Spain, Sweden, United Kingdom, United States
are available. Micro-level analysis is based on annual Annex Figure 2.2.1. Changes in Labor Force Participation
data for 24 advanced European economies during Rates, 1985–2016
2000–16. Information on family composition is not
0 1. Men, Ages 25–54
available for Denmark, Finland, Iceland, Norway,
ESP GRC

–1 NLD JPN
Sweden, or Switzerland, so regressions including these

FIN PRT LUX
–2 CAN KOR
variables are estimated on a subset of 18 economies. DNK DEU SWE
–3 ISR GBR HKG
The cohort-based analysis relies on annual data for 17
AUS BEL
advanced economies from 1985 to 2016 for which –4 FRA
gender-specific labor force participation rate data are –5 IRL
USA
available for quinquennial age groups starting at age 15 –6 R 2 = 0.05
and up to 64 and covering ages 65 to 99. –7 ITA NOR
–8
89 90 91 92 93 94 95 96 97 98 99
Labor force participation rate, 1985 (percent)
Annex 2.2. Additional Stylized Facts
This section provides further stylized facts on con- 60 2. Women, Ages 25–54
vergence across economies in participation rates, the
50 ESP
R 2 = 0.87
link between the participation of men and women, LUX NLD
40
and the effects of the global financial crisis. GRC BEL
30 IRL
Annex Figure 2.2.1 examines whether there is ISR PRT
ITA FRA
evidence for convergence across economies in partici- 20 DEU CAN
HKG
KOR
pation rates. While this seems to be limited for men, 10 AUS
JPN NOR
GBR DNK
gains in female participation were indeed substantially 0
USA
larger in economies where women were historically FIN SWE
–10
less likely to be part of the workforce. As a result, as 30 40 50 60 70 80 90 100
documented in the chapter, the dispersion in women’s Labor force participation rate, 1985 (percent)
participation across advanced economies has narrowed

since 1985. calculations.
The rise in women’s labor force participation is Note: Labels in the figure use International Organization for Standardization (ISO)
country codes.
also consistent with the rising share of two-earner
households. Based on micro data from the European
Union Labour Force Survey, Annex Figure 2.2.2
shows that the share of households with one adult
working and one adult not working has fallen since some men to drop out of the labor force and finds
2000, while the share of households with both adults no evidence for this at the country level. Correla-
working has increased. tions between changes in prime-age female and male
Annex Figure 2.2.3 examines the hypothesis that participation rates are, if anything, positive, though
women’s increasing participation may have allowed relatively weak.

Annex Figure 2.2.2. Share of Households by Employment Annex Figure 2.2.4. Average Annual Changes in Labor Force
Composition, 2000–16 Participation Rates
(Percent) (Percentage points)
45 1.2 1. Men, Ages 25–54 2. Women, Ages 25–54 1.2

All adults working full time
At least one adult working and one adult not working 1.0 1.0
0.8 0.8
40
0.6 0.6
0.4 0.4
0.2 0.2
35
0.0 0.0
–0.2 –0.2
30 –0.4 –0.4
1985–2000 01–07 08–16 1985–2000 01–07 08–16
1.5 3. All, Ages 15–24 4. All, Ages 55 and Older 1.5

25
1.2 1.2
0.9 0.9
0.6 0.6
20
2000 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 0.3 0.3
0.0 0.0
Sources: Eurostat, European Union Labour Force Survey; and IMF staff –0.3 –0.3
calculations. –0.6 –0.6
–0.9 –0.9
respondents per country per year from the European Union Labour Force Survey
over the period 2000–16. –1.2 –1.2
1985–2000 01–07 08–16 1985–2000 01–07 08–16
calculations.
Note: Bars denote median; and vertical lines denote interquartile range. In all
panels, left bars for each time period show countries above the median in terms
Annex Figure 2.2.3. Changes in Labor Force Participation of real GDP loss during 2008–12 and right bars show countries below the median.
Rates of Prime-Age Men and Women, 1985–2016
(Percentage points)
60
Change in participation rate for prime-age women
50 Annex Figures 2.2.4 and 2.2.5 analyze the effects

R 2 = 0.04
of the global financial crisis and European debt crises
40 further. Annex Figure 2.2.4 compares economies with
above- and below-median GDP losses during the
30 crisis and finds that the dynamics of their labor force
participation rates were broadly similar. Annex Fig-
20
ure 2.2.5 in turn examines the evolution of transition
probabilities—flows from employment, unemploy-
10
ment, and inactivity into employment, unemployment,
and inactivity—over time. This suggests that the share
0
of discouraged workers (those who are inactive but
were unemployed the previous year) has been increas-
–10
–8 –7 –6 –5 –4 –3 –2 –1 0 ing since the crisis and is approaching the precrisis
Change in participation rate for prime-age men peak. This figure also illustrates the spike in flows from
employment into unemployment during the global
Sources: Organisation for Economic Co-operation and Development; and IMF staff financial crisis, as well as flows from unemployment
calculations.
Note: Prime age is defined as between 25 and 54. back into employment after the crisis.

Annex Figure 2.2.5. Flows into Employment, Unemployment, participation rate, LFP  a, can be rewritten as the par-
and Inactivity ticipation rates of workers of gender ain age group g,
(Percent)
weighted by their share in the male or female popula-
6 1. Flows into Employment by Previous Status 96 tion, respectively:
5 a,g   _
pop  a,g 
95   ∑ g = 1
a  =
LFP i,t n    LFP i,t i,t
pop     
, (2.1)
i,t
4
94 in which idenotes the country, tis the time index, a is
3 the gender, gis the age group (15–24, 25–54, 55–64,
93 65 and over), and popis the population. The aging
2
effect is obtained as the difference between the actual
1 92
Employment (right scale) Unemployment Inactivity participation rate and the one obtained by holding
0 91 constant the gender- and group-specific participa-
2000 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 a,g , but allowing
tion rates at their 2008 level,a,g 
LFP  i,2008
_____
pop i,t
70 2. Flows into Unemployment by Previous Status
the population shares,  pop     
,to vary as observed
i,t
in the data.
60
50
40 Estimating the Role of Cyclical Conditions
30 Economic contractions generally result in greater
20
unemployment and lower labor force participation
as some workers get discouraged and permanently
10
Employment Unemployment Inactivity separate from the workforce, and others choose to
0
2000 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 delay entry. To capture the effect of the cycle on labor
force participation, the chapter estimates the follow-
9 3. Flows into Inactivity by Previous Status 97 ing regression:
8 96
7 95
  ∑ k = 0
LFP i,t*  = 1  β 
  k UG i,t − k  + ∑ k = 0
1  δ 
  k Crisis i,t − k
6 94 + ∑ k = 0
1  γ 
  k UG i,t − k Crisis i,t − k  + π i 
5 93
4 92 + τ  t  + ε  it, (2.2)
3 91
2
Employment Unemployment Inactivity (right scale) 90 in which LFP  *is the detrended aggregate labor
1 89 force participation rate, obtained by applying the
0 88 Hodrick-Prescott (HP) filter to the labor force partic-
2000 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16
ipation rate; UGis the unemployment gap, defined
as the gap between current unemployment and the
Sources: Eurostat, European Union Labour Force Survey; and IMF staff calculations.
Note: Previous status refers to labor force status in the previous year. Reported nonaccelerating inflation rate of unemployment
statistics are estimated from a random sample of 10,000 respondents per country (NAIRU);51 Crisisis a dummy variable that takes a
per year from the European Union Labor Force Survey over the period 2000–16.
value of 1 when there is either a currency crisis, a
sudden stop, a debt crisis, or a banking crisis, based on
the Gourinchas-Obstfeld database; and π i and τ  t are
country and time fixed effects.
Annex 2.3. The Role of Aging and The regression is estimated using annual data during
Cyclical Factors 1980–2016, and the cyclical effect at time tis obtained
as the predicted value of the regression. The difference
Shift-Share Analysis
in the predicted cyclical component relative to its 2008
A standard shift-share analysis is performed to estab- value captures the role of the cycle in the change in
lish how demographic changes in advanced economies
have contributed to the trends in participation rates 51The NAIRU is constructed as in Chapter 3 of the April 2013
since 2008. The gender-specific aggregate labor force World Economic Outlook.

aggregate participation since then. While the find- country and time fixed effects.53 Some of the evidently
ings rely on a specification with a single lag (k = 1) endogenous variables are included in the specifica-
estimated in a panel setting, the results are qualitatively tion with a one-year lag. The groups comprise young
similar if a richer lag structure is used instead, or if workers (15–24), prime-age men (25–54), prime-age
the sensitivity of labor force participation to the cycle women (25–54), and older workers (55 and over); an
is allowed to vary across economies. Results are also additional equation is estimated for a group encom-
robust to employing the Corbae-Ouliaris (CO) filter passing all workers 15 and older.
instead of the HP filter to obtain the detrended aggre- Given the complex correlation structure of the error
gate labor force participation rate in equation (2.2), as term with dependence across economies, autocorrela-
well as to calculating it as deviations from a three-year tion due to the slow-moving nature of the dependent
moving average, limiting the distortions generated by variable, and heteroscedasticity, the Driscoll and Kraay
the endpoint problem of the HP filter. (1998) correction to the standard errors is used to
make statistical inferences. The findings are robust to
various alternative corrections of standard errors as
Annex 2.4. The Role of Policies and Other discussed later.
Factors: Aggregate Cross-Country Analysis The analysis then decomposes the contributions
This analysis estimates a reduced-form specification from each regressor to changes in participation of
of labor force participation that relates the participa- group g between years tand tʹ as
βˆ
  S,g  (S  gi,tʹ
tion rate of specific groups of workers to factors that
may affect the decision to supply labor. It controls for   S,g    =
C i,t,tʹ    −    g  ),
  Si,t (2.4)
all differences across economies that are constant over in which S = {X, D, GAP, Z} and C S,g
    is the contribu-
i,t,tʹ
time and shocks that affect all economies. While the tion of variable S.
potential set of drivers is large, the analysis, guided The key variables included in the analysis are
by the conceptual framework described in the main the following:
text, focuses on factors that can be measured relatively •• The cyclical position is captured using the output
consistently across economies and over time and that gap. The results are not sensitive to using alternative
are most commonly discussed in policy debates.52 measures, such as the unemployment rate.
More specifically, the aggregate analysis is based on the •• Exposure to technological progress is measured
estimation of the equation following Chapter 3 of the April 2017 World
Economic Outlook (WEO) and Das and Hilgen-
i,t  β  Di,t
g    = β  X,g X  g  + 
LFP i,t D,g     + β  GAP,g GAP 
i,t − 1 
+ β  Z,g Z i,t  + π ig   + τ  gt   + ε  gi,t ,  (2.3) stock (forthcoming). The relevant variable is the
interaction between the relative price of investment
in which LFPdenotes the participation rate of worker and the country’s exposure to routinization through
group gin country iin year t, GAPis the cyclical its initial occupational mix. The latter consists of
position of the economy, Xrepresents the set of scores that rely on occupation-level measures by
policies and institutions (some of which are specific to Autor and Dorn (2013), which order occupations
group g), Ddenotes a set of factors that may shift the by their share of routine tasks, and then use the
demand for worker group g, Zincludes other deter- employment shares of these occupations to construct
minants of labor supply (education), and πi and τt are country-level measures of routinizability. The aver-
age relative price of investment across all advanced
economies is used to minimize endogeneity con-
52The vast theoretical literature on labor supply offers a large cerns and capture changes that are due to global
number of models with different assumptions, including about technological progress (rather than, for example,
(1) the ability of consumers to transfer capital across periods and
to consider more generally a life-cycle framework; (2) the extent
country-specific capital taxation policies).
to which labor supply decisions are made by the household rather
than the individual worker; (3) the role of uncertainty about future
income, household composition, and health status; and (4) how 53Results from panel unit root tests suggest that the time series
government programs affect the incentives to work (see Blundell and of labor force participation rates for different age groups are trend
Macurdy 1999 for a review). Developing a macroeconomic theory of stationary. Because of limited data availability for some of the
labor supply encompassing all these features for different groups of explanatory variables, using a dynamic specification in the presence
workers is beyond the scope of this chapter. of country fixed effects would return biased estimates (Nickell 1981).

•• Potential shifts in the demand for different types of country, with a higher value denoting more restric-
labor due to structural transformation are measured tive policies.
as the ratio of employment in the service sector •• Union density is measured as net union membership
relative to employment in the industrial sector and as a proportion of wage earners in employment. The
the share of urban population. variable is published by the OECD.
•• Educational attainment is from the Barro-Lee •• Coordination of wage setting is an index of the
database (Lee and Lee 2016) and is measured as the centralization of bargaining, published by the
share of the population within a specific age-gender Amsterdam Institute for Advanced Labour Studies
group with the highest level of education reported as Database on Institutional Characteristics of Trade
primary, secondary, or tertiary. Unions, Wage Setting, State Intervention and Social
•• The labor tax wedge is defined as the ratio between Pacts. The index runs from 1 to 5 with values
the average tax paid by a single-earner family (one defined as (1) fragmented wage bargaining, con-
parent at 100 percent of average earnings with two fined largely to individual firms or plants; (2) mixed
children) and the corresponding total labor cost for industry and firm-level bargaining, weak govern-
the employer. The labor tax wedge is available from ment coordination through minimum wage setting
the Organisation for Economic Co-operation and or wage indexation; (3) negotiation guidelines based
Development (OECD) for 2000–16 and is extended on centralized bargaining; (4) wage norms based on
back to 1979 using Bassanini and Duval (2006) centralized bargaining by peak association with or
and Chapter 3 of the April 2016 WEO. The latter without government involvement; and (5) maxi-
series is available only in odd years; the value of the mum or minimum wage rates/increases based on
labor tax wedge in even years is obtained by linear centralized bargaining.
interpolation. •• Policies that help reconcile work inside and outside
•• The generosity of the unemployment benefits the household are proxied by public spending on
system is measured as the gross replacement rate, early childhood education and care as a percent of
which is equal to the gross unemployment benefit GDP; the proportion of employees with a part-time
levels as a percentage of previous gross earnings and contract to total employees; and job-protected
is published by the OECD. The OECD summary maternity leave, defined as the total number of
measure with the best coverage is the average of weeks of job-protected maternity, parental, and
the gross unemployment benefit replacement rates extended leave available to mothers, regardless of
for two earnings levels, three family situations, and income support. These variables are published
three durations of unemployment. Such measures by the OECD.
are available in odd years and are interpolated for •• Retirement incentives are proxied by the statutory
even years. The reported values are for the average retirement age and by the generosity of pension
worker from 2001 to 2011 and the average pro- plans. A database of statutory retirement ages is
duction worker from 1981 to 2005. The two series compiled from various publications of Social Security
are spliced. Programs throughout the World. Several alterna-
•• Public expenditure on active labor market programs, tives are used to capture the generosity of pension
published by the OECD, is calculated as active labor plans. The measure with the best country and time
market program spending per unemployed person coverage is old-age and incapacity spending as a
in percent of GDP per capita, following Gal and percent of GDP from the OECD. This measure is
Theising (2015). first purged of fluctuations resulting from cycli-
•• Restrictiveness of migration policy is obtained from cal and demographic factors (namely, share of the
the DEMIG POLICY database compiled by the population in different age groups and health status,
International Migration Institute, which codes all proxied by life expectancy) that may mechanically
changes to the existing legal framework relevant for generate a negative correlation with the labor force
migration (see also de Haas, Natter, and Vezzoli attachment of older workers. As a robustness check,
2014). The chapter focuses on major changes in the analysis considers the (conceptually more appro-
policies guiding the postentry rights or other aspects priate but less widely available) implicit tax on con-
of migrants’ integration. These changes are cumu- tinued work, calculated as the change in the present
lated starting in 1980 to construct an index for each value of the stream of future pension payments

net of contributions to the system from working tions of the standard errors. In particular, the
five more years for typical workers at different ages conclusions are broadly unchanged when employ-
(see Duval 2004 and Chapter 3 of the April 2016 ing the Beck and Katz (1995) estimator in col-
WEO). An alternative measure also considered is the umn (4), correcting the standard errors only for
aggregate replacement ratio, calculated as the ratio heteroscedasticity and autocorrelation in column
of the mean disposable income of people ages 65–74 (5) and adopting the Newey-West correction for
to the mean disposable income of those ages 50–59, the standard errors in column (6).
from the Luxembourg Income Study Database. This •• Cyclical effects: Possible distortions arising from
variable can be computed for selected years based inability to control for cyclical effects are controlled
on the availability of household survey data and is for by estimating the equation on five-year averages,
interpolated for the missing years. which could also rule out the possibility that the
results depend on some undetected local unit root.
Annex Tables 2.4.1–2.4.5 present the key results The dependent variables in the regressions are trend
from the cross-country panel regressions, along with stationary, which excludes the possibility of unde-
numerous robustness checks. Annex Table 2.4.1 con- tected cointegrating relationships with the explanatory
tains the estimated coefficients for the regression on variables. While some of the explanatory variables
the young, Annex Table 2.4.2 on the prime-age male are locally nonstationary, most of these are shares
workers, Annex Table 2.4.3 on the prime-age female bounded between 0 and 100. The results based on
workers, Annex Table 2.4.4 on older workers, and five-year averages presented in column (7) are broadly
Annex Table 2.4.5 on the aggregate participation rate. comparable to those of the baseline estimates.
Each table shows the results from the baseline specifi- •• Global financial crisis: The significance and the
cation discussed in the main text (Table 2.1, column magnitude of the coefficients are not affected by
1) and establishes its robustness to alternative mea- dropping 2008 and 2009 from the sample, as shown
sures, specification, error structure, and the like. in column (8).
•• Logistic transformation: Given that participation rates •• Other advanced economies: Broadening the sample
are bounded by 0 and 100 by construction, the to economies that were classified as advanced in the
analysis is repeated using the logistic transformation WEO database after 2006 does not generally alter
of the dependent variable in column (2). the results, as can be seen in column (9).
•• Alternative corrections to standard errors •• Alternative measures of the output gap: In column
are as follows: (10), the analysis replaces the output gap with the
oo Cross-equation correlation: There may be correla- unemployment rate. This specification returns qual-
tion across the error terms of the estimations itatively comparable results. However, in this case,
for different worker groups. Estimating a sys- older workers’ participation rates turn out to be
tem including one equation for each group in a sensitive to the cyclical conditions of the economy.
seemingly unrelated regression framework returns •• Sample selection: The analysis rules out the pos-
similar results in column (3). sibility that single economies drive the results by
oo Cross-sectional dependence: Tests by Pesaran (2004) estimating the same specification dropping one
and Frees (1995) reject the null hypothesis of economy at a time. The estimates display remarkable
cross-sectional independence, but the results stability, as is shown in column (11). This exercise
of the test by Friedman (1937) suggest that also allays concerns that the findings on the role of
cross-sectional dependence is not present. The certain variables may be an artifact of measurement
results are generally robust to alternative correc- errors in the series of some economies.

Annex Table 2.4.1. Drivers of Youth (Ages 15–24) Labor Force Participation Rates: Robustness
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11)
HAC Five- Replacing Output Dropping
Logistic Standard Newey-West Year Excluding Including Gap with One Country
Baseline Transformation SUR Beck and Katz Errors Standard Errors Averages GFC All AEs Unemployment Rate at a Time
Lag of Output Gap 0.360*** 0.015*** 0.219*** 0.226*** 0.360*** 0.360*** 0.473** 0.366*** 0.286*** –0.519*** 0.354
(0.112) (0.005) (0.073) (0.063) (0.080) (0.103) (0.183) (0.121) (0.100) (0.100) (0.292; 0.393)
Routinization x Relative Price 0.303 0.012 0.313 –0.143 0.303 0.303 0.358 0.358 0.295 0.404 0.297
of Investment (0.299) (0.012) (0.280) (0.377) (0.262) (0.344) (0.299) (0.313) (0.291) (0.293) (0.184; 0.489)
Lag of Trade Openness 0.059*** 0.003*** 0.045*** 0.020 0.059*** 0.059*** 0.059* 0.053** 0.043* 0.021 0.057
(0.022) (0.001) (0.016) (0.019) (0.018) (0.019) (0.026) (0.024) (0.022) (0.028) (0.045; 0.071)
Relative Service Employment –0.002 –0.000 –0.026*** –0.020** –0.002 –0.002 –0.004 –0.004 0.002 0.004 –0.001
(0.010) (0.000) (0.008) (0.009) (0.009) (0.011) (0.012) (0.009) (0.009) (0.009) (–0.01; 0.001)
Lag of Urbanization 0.668*** 0.030*** –0.089 0.575*** 0.668*** 0.668*** 0.560*** 0.700*** 0.685*** 0.715*** 0.669
(0.142) (0.006) (0.098) (0.135) (0.085) (0.159) (0.139) (0.170) (0.139) (0.175) (0.626; 0.702)
Education (percent secondary) –0.050 –0.002 0.007 –0.042 –0.050* –0.050 –0.049 –0.062 –0.050 –0.050 –0.049
(0.042) (0.002) (0.022) (0.032) (0.026) (0.036) (0.060) (0.049) (0.042) (0.049) (–0.076; –0.031)
Education (percent tertiary) –0.275*** –0.012*** –0.105*** –0.227*** –0.275*** –0.275*** –0.253* –0.290*** –0.286*** –0.294*** –0.273
(0.057) (0.002) (0.038) (0.047) (0.042) (0.054) (0.105) (0.064) (0.048) (0.062) (–0.303; –0.247)
Tax Wedge –0.103 –0.005* –0.021 –0.029 –0.103** –0.103* –0.086 –0.082 –0.094 0.057 –0.104
(0.064) (0.003) (0.046) (0.050) (0.048) (0.058) (0.060) (0.064) (0.059) (0.074) (–0.127; –0.058)
Unemployment Replacement –0.002 –0.000 0.111*** 0.009 –0.002 –0.002 –0.001 0.016 –0.009 0.007 0.003
Ratio (0.068) (0.003) (0.035) (0.044) (0.039) (0.059) (0.140) (0.070) (0.065) (0.066) (–0.036; 0.034)
Public Spending on ALMP 0.041*** 0.002** –0.004 0.031 0.041*** 0.041** 0.023 0.048*** 0.045*** 0.030* 0.041
(0.014) (0.001) (0.013) (0.023) (0.014) (0.019) (0.033) (0.014) (0.014) (0.017) (0.036; 0.048)
Restrictiveness of Migrant 0.491*** 0.021*** 0.421*** 0.277** 0.491*** 0.491*** 0.521*** 0.492*** 0.464*** 0.421*** 0.487
Integration Policies (0.098) (0.004) (0.079) (0.108) (0.090) (0.146) (0.091) (0.109) (0.094) (0.114) (0.414; 0.539)
Union Density –0.009 –0.000 –0.065 –0.116** –0.009 –0.009 –0.021 –0.020 –0.001 0.011 –0.01
(0.068) (0.003) (0.046) (0.057) (0.046) (0.066) (0.091) (0.069) (0.061) (0.065) (–0.026; 0.016)
Coordination of Wage Setting 1.104*** 0.045*** 0.942*** 0.081 1.104*** 1.104*** 1.848*** 1.117*** 1.088*** 0.694** 1.105
(0.245) (0.010) (0.241) (0.180) (0.251) (0.329) (0.451) (0.252) (0.233) (0.284) (0.852; 1.255)
Number of Observations 571 571 489 571 571 571 132 525 593 525
Countries 23 23 23 23 23 23 23 23 25 23
R 2 0.515 0.521 0.922 0.515 0.515 0.573 0.529 0.517 0.540
Note: The table presents results from estimation equation (2.3) with the participation rate of youth (ages 15–24) as the dependent variable on a sample of 23 advanced economies (AEs) during 1980–2011 using annual data.
See Annex 2.4 for the construction of the explanatory variables and Annex Table 2.1.2 for the countries in the sample. All specifications include country and year fixed effects. Column (1) reports the baseline estimation results;
column (2) reports the results after applying the logistic transformation to the dependent variable; column (3) reports the estimates from a seemingly unrelated regressions (SUR) estimation of a four-equation system (one for each
International Monetary Fund | April 2018

group of workers); column (4) shows the results using the Beck and Katz (1995) estimator; column (5) reports the estimates with heteroscedasticity and autocorrelation consistent (HAC) standard errors, without the correction for
cross-sectional dependence; column (6) shows the results with the Newey-West correction for the standard errors; column (7) shows the results based on a sample of five-year averages; column (8) reports the results dropping
global financial crisis (GFC) years 2008 and 2009 from the sample; column (9) reports the coefficients when Czech Republic and Slovak Republic, which recently joined AEs, are added to the sample; column (10) shows the
results when the lag of the output gap is replaced with the lag of the unemployment rate; and column (11) reports the median coefficient from a distribution of estimates obtained by dropping one country at a time from the sample.
117
Driscoll-Kraay standard errors are reported in parentheses in columns (1), (2), (7)–(10); bootstrapped standard errors are reported in parentheses in column (3); HAC standard errors assuming a panel-dependent correlation struc-
ture are reported in column (4). Column (11) reports the 10th and 90th percentile of the estimated coefficients in parentheses. ALMP = active labor market programs.
*p < .10; **p < .05; ***p < .01.
Annex Table 2.4.2. Drivers of Prime-Age Male (Ages 25–54) Labor Force Participation Rates: Robustness
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11)
118
Lag of Output Gap 0.072*** 0.012*** 0.058*** 0.020 0.072*** 0.072*** 0.117** 0.070*** 0.062*** –0.002 0.07
(0.020) (0.003) (0.017) (0.016) (0.018) (0.022) (0.037) (0.022) (0.019) (0.029) (0.06; 0.081)
Routinization × Relative Price 0.302*** 0.070*** 0.204*** 0.217*** 0.302*** 0.302*** 0.266*** 0.303*** 0.284*** 0.315*** 0.302
of Investment (0.048) (0.011) (0.067) (0.057) (0.057) (0.074) (0.072) (0.048) (0.037) (0.049) (0.285; 0.33)
Lag of Trade Openness –0.005 –0.001 –0.012*** –0.015*** –0.005 –0.005 –0.003 –0.007 –0.005 –0.010 –0.005
(0.005) (0.001) (0.004) (0.005) (0.004) (0.005) (0.009) (0.006) (0.004) (0.006) (–0.006; –0.003)
Relative Service Employment –0.002 –0.000 –0.000 –0.000 –0.002 –0.002 0.001 –0.002 –0.003 –0.002 –0.002
(0.002) (0.000) (0.002) (0.002) (0.002) (0.002) (0.002) (0.002) (0.002) (0.002) (–0.003; –0.001)
Lag of Urbanization 0.101*** 0.015*** 0.006 0.105*** 0.101*** 0.101*** 0.072** 0.105*** 0.104*** 0.110*** 0.101

(0.019) (0.004) (0.023) (0.027) (0.019) (0.020) (0.023) (0.023) (0.020) (0.026) (0.091; 0.114)
Education (percent secondary) 0.019*** 0.003** 0.037*** 0.022*** 0.019** 0.019** 0.015** 0.022*** 0.018** 0.023*** 0.019
(0.007) (0.001) (0.008) (0.009) (0.008) (0.009) (0.005) (0.007) (0.007) (0.008) (0.016; 0.026)
Education (percent tertiary) 0.019 0.001 0.023** 0.030* 0.019* 0.019 0.017 0.027 0.018 0.029 0.018
(0.015) (0.002) (0.010) (0.016) (0.010) (0.012) (0.019) (0.019) (0.014) (0.018) (0.012; 0.027)
Tax Wedge –0.002 –0.003 –0.009 0.005 –0.002 –0.002 –0.002 –0.001 0.004 0.002 –0.002
(0.015) (0.002) (0.010) (0.011) (0.010) (0.010) (0.027) (0.016) (0.014) (0.017) (–0.007; 0.001)
Unemployment Replacement –0.041*** –0.007*** –0.037*** –0.024** –0.041*** –0.041*** –0.031*** –0.041*** –0.039*** –0.044*** –0.04
Ratio (0.007) (0.001) (0.008) (0.010) (0.008) (0.011) (0.008) (0.008) (0.008) (0.008) (–0.045; –0.034)
Public Spending on ALMP 0.005 0.001 –0.000 0.005 0.005 0.005 0.005 0.007 0.006 0.010* 0.005
(0.005) (0.001) (0.003) (0.005) (0.003) (0.005) (0.009) (0.005) (0.005) (0.005) (0.004; 0.006)
Restrictiveness of Migrant –0.047** –0.007** –0.052*** –0.019 –0.047** –0.047** –0.092** –0.053** –0.048** –0.056** –0.046
Integration Policies (0.020) (0.003) (0.019) (0.024) (0.020) (0.023) (0.025) (0.020) (0.021) (0.022) (–0.062; –0.034)
Union Density –0.001 0.002 –0.016 –0.023* –0.001 –0.001 –0.011 0.000 –0.004 –0.000 –0.001
(0.011) (0.002) (0.011) (0.012) (0.010) (0.013) (0.019) (0.011) (0.011) (0.013) (–0.01; 0.007)
Coordination of Wage Setting 0.131** 0.018* 0.074 0.073* 0.131** 0.131* 0.302** 0.134* 0.139** 0.131 0.131
(0.063) (0.010) (0.057) (0.040) (0.055) (0.069) (0.090) (0.068) (0.062) (0.080) (0.117; 0.15)
Countries 23 23 23 23 23 23 23 23 25 23
R 2 0.606 0.622 0.997 0.606 0.606 0.695 0.622 0.600 0.611
Notes: The table presents results from estimation equation (2.3) with the participation rate of prime-age men (ages 25–54) as the dependent variable on a sample of 23 advanced economies (AEs) during 1980–2011 using annual
data. See Annex 2.4 for the construction of the explanatory variables and Annex Table 2.1.2 for the countries in the sample. All specifications include country and year fixed effects. Column (1) reports the baseline estimation results;
cross-sectional dependence; column (6) shows the results with the Newey-West correction for the standard errors; column (7) shows the results based on a sample of five-year averages; column (8) reports the results dropping
global financial crisis (GFC) years 2008 and 2009 from the sample; column (9) reports the coefficients when Czech Republic and Slovak Republic, which recently joined AEs, are added to the sample; column (10) shows the
results when the lag of the output gap is replaced with the lag of the unemployment rate; and column (11) reports the median coefficient from a distribution of estimates obtained by dropping one country at a time from the sample.
Driscoll-Kraay standard errors are reported in parentheses in columns (1), (2), (7)–(10); bootstrapped standard errors are reported in parentheses in column (3); HAC standard errors assuming a panel-dependent correlation
structure are reported in column (4); HAC standard errors are reported in parentheses in column (5); and Newey-West corrected standard errors are reported in column (6). Column (11) reports the 10th and 90th percentile of the
estimated coefficients in parentheses. ALMP = active labor market programs.
*p < .10; **p < .05; ***p < .01.
Annex Table 2.4.3. Drivers of Prime-Age Female (Ages 25–54) Labor Force Participation Rates: Robustness
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11)
Lag of Output Gap 0.170* 0.008* 0.180*** 0.114** 0.170*** 0.170* 0.114 0.201* 0.119 –0.407** 0.167
(0.092) (0.004) (0.052) (0.050) (0.056) (0.095) (0.223) (0.098) (0.072) (0.151) (0.139; 0.188)
Routinization × Relative Price 1.793*** 0.072*** 1.866*** 1.245*** 1.793*** 1.793*** 1.565*** 1.720*** 1.578*** 1.692*** 1.781
of Investment (0.206) (0.008) (0.206) (0.192) (0.222) (0.326) (0.229) (0.204) (0.205) (0.237) (1.672; 1.914)
Lag of Trade Openness 0.010 0.000 0.016 0.009 0.010 0.010 0.013 0.022 –0.003 0.014 0.009
(0.014) (0.001) (0.012) (0.013) (0.013) (0.016) (0.022) (0.020) (0.011) (0.018) (0.002; 0.02)
Relative Service Employment 0.015*** 0.000* 0.015*** 0.010* 0.015** 0.015** 0.017** 0.016*** 0.020*** 0.016*** 0.015
(0.005) (0.000) (0.005) (0.006) (0.006) (0.007) (0.006) (0.005) (0.005) (0.005) (0.01; 0.019)
Lag of Urbanization 0.355*** 0.021*** 0.373*** 0.313*** 0.355*** 0.355*** 0.343*** 0.372*** 0.398*** 0.341*** 0.35
(0.071) (0.004) (0.068) (0.061) (0.073) (0.075) (0.048) (0.079) (0.078) (0.077) (0.299; 0.393)
Education (percent secondary) 0.211*** 0.010*** 0.203*** 0.247*** 0.211*** 0.211*** 0.187*** 0.215*** 0.195*** 0.221*** 0.209
(0.017) (0.001) (0.023) (0.030) (0.025) (0.031) (0.016) (0.020) (0.018) (0.022) (0.19; 0.236)
Education (percent tertiary) 0.332*** 0.016*** 0.268*** 0.360*** 0.332*** 0.332*** 0.249*** 0.332*** 0.319*** 0.360*** 0.333
(0.030) (0.001) (0.038) (0.048) (0.042) (0.053) (0.049) (0.035) (0.031) (0.042) (0.285; 0.374)
Tax Wedge –0.129*** –0.002 –0.134*** –0.095*** –0.129*** –0.129*** –0.141 –0.125*** –0.104*** –0.115*** –0.13
(0.029) (0.001) (0.032) (0.026) (0.035) (0.041) (0.071) (0.028) (0.029) (0.027) (–0.158; –0.104)
Unemployment Replacement –0.035 –0.003 –0.036 –0.028 –0.035 –0.035 0.044 –0.040 –0.030 –0.048 –0.034
Ratio (0.033) (0.002) (0.025) (0.021) (0.026) (0.033) (0.095) (0.035) (0.033) (0.034) (–0.047; –0.025)
Public Spending on ALMP 0.039*** 0.002*** 0.040*** 0.038*** 0.039*** 0.039*** 0.022 0.038*** 0.046*** 0.042*** 0.039
(0.006) (0.000) (0.009) (0.008) (0.010) (0.013) (0.016) (0.007) (0.007) (0.008) (0.031; 0.042)
Restrictiveness of Migrant –0.462*** –0.019*** –0.464*** –0.330*** –0.462*** –0.462*** –0.449*** –0.470*** –0.436*** –0.496*** –0.463
Integration Policies (0.049) (0.002) (0.056) (0.063) (0.060) (0.082) (0.083) (0.052) (0.047) (0.057) (–0.491; –0.418)
Union Density 0.153*** 0.004** 0.165*** 0.084** 0.153*** 0.153*** 0.050 0.127** 0.156*** 0.116** 0.151
(0.044) (0.002) (0.033) (0.041) (0.036) (0.047) (0.094) (0.046) (0.042) (0.046) (0.114; 0.173)
Coordination of Wage Setting 0.701*** 0.026** 0.675*** 0.190 0.701*** 0.701*** 1.658** 0.640** 0.687*** 0.603** 0.707
(0.219) (0.010) (0.164) (0.126) (0.177) (0.235) (0.444) (0.259) (0.219) (0.247) (0.64; 0.771)
Public Spending on Early 3.708*** 0.250*** 3.423*** 2.151*** 3.708*** 3.708*** 5.855** 3.628*** 3.670*** 3.709*** 3.699
Childhood Education and Care (1.210) (0.071) (0.622) (0.799) (0.683) (0.951) (2.146) (1.295) (1.177) (1.276) (3.122; 4.285)
Share of Part-Time 0.946*** 0.045*** 0.932*** 0.735*** 0.946*** 0.946*** 0.982*** 0.943*** 1.021*** 0.889*** 0.956
Employment (0.118) (0.006) (0.066) (0.064) (0.073) (0.098) (0.168) (0.126) (0.109) (0.104) (0.868; 0.994)
Job-Protected Maternity 0.025*** 0.001*** 0.026*** 0.013 0.025*** 0.025*** 0.087*** 0.024*** 0.020*** 0.025*** 0.024
Leave (0.006) (0.000) (0.007) (0.009) (0.008) (0.009) (0.011) (0.006) (0.007) (0.006) (0.021; 0.028)
Countries 23 23 23 23 23 23 23 23 25 23
R 2 0.887 0.870 0.971 0.887 0.887 0.891 0.881 0.879 0.879
Note: The table presents results from estimation equation (2.3) with the participation rate of prime-age women (ages 25–54) as the dependent variable on a sample of 23 advanced economies (AEs) during 1980–2011 using annual

cross-sectional dependence; column (6) shows the results with the Newey-West correction for the standard errors; column (7) shows the results based on a sample of five-year averages; column (8) reports the results dropping global
financial crisis (GFC) years 2008 and 2009 from the sample; column (9) reports the coefficients when Czech Republic and Slovak Republic, which recently joined AEs, are added to the sample; column (10) shows the results when
119
the lag of the output gap is replaced with the lag of the unemployment rate; and column (11) reports the median coefficient from a distribution of estimates obtained by dropping one country at a time from the sample. Driscoll-Kraay
standard errors are reported in parentheses in columns (1), (2), (7)–(10); bootstrapped standard errors are reported in parentheses in column (3); HAC standard errors assuming a panel-dependent correlation structure are reported
in column (4); HAC standard errors are reported in parentheses in column (5); and Newey-West corrected standard errors are reported in column (6). Column (11) reports the 10th and 90th percentile of the estimated coefficients in
parentheses. ALMP = active labor market programs.
*p < .10; **p < .05; ***p < .01.
Annex Table 2.4.4. Drivers of Older Workers’ (Ages 55 and over) Labor Force Participation Rates: Robustness
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11)
120
Lag of Output Gap –0.006 –0.000 0.009 0.025 –0.006 –0.006 0.178 0.003 0.000 –0.268*** –0.008
(0.068) (0.003) (0.056) (0.037) (0.055) (0.083) (0.131) (0.078) (0.065) (0.079) (–0.045; 0.014)
Routinization × Relative Price 0.505* 0.009 1.038*** 0.198 0.505*** 0.505 0.473 0.468 0.742*** 0.472 0.503
of Investment (0.288) (0.015) (0.222) (0.229) (0.184) (0.372) (0.298) (0.292) (0.241) (0.289) (0.219; 0.593)
Lag of Trade Openness –0.059*** –0.002*** –0.066*** –0.012 –0.059*** –0.059*** –0.051*** –0.063*** –0.045*** –0.063*** –0.06
(0.009) (0.000) (0.013) (0.008) (0.013) (0.014) (0.013) (0.015) (0.012) (0.013) (–0.07; –0.044)
Relative Service Employment 0.009 0.001*** 0.007 0.002 0.009 0.009 0.007 0.008 0.005 0.008 0.01
(0.006) (0.000) (0.006) (0.004) (0.006) (0.006) (0.006) (0.006) (0.006) (0.006) (0.004; 0.014)
Lag of Urbanization 0.194 0.014** –0.056 0.118 0.194*** 0.194* 0.138 0.225* 0.189* 0.223* 0.194
(0.115) (0.007) (0.084) (0.092) (0.064) (0.114) (0.172) (0.127) (0.111) (0.126) (0.095; 0.245)
Education (percent secondary) 0.038* 0.001 0.019 –0.016 0.038* 0.038 0.004 0.037* 0.053** 0.037* 0.036

(0.021) (0.001) (0.021) (0.021) (0.020) (0.027) (0.043) (0.022) (0.020) (0.022) (0.016; 0.059)
Education (percent tertiary) 0.389*** 0.018*** 0.321*** 0.260*** 0.389*** 0.389*** 0.296** 0.384*** 0.397*** 0.386*** 0.387
(0.050) (0.002) (0.059) (0.057) (0.058) (0.093) (0.085) (0.053) (0.046) (0.056) (0.3; 0.44)
Tax Wedge –0.263*** –0.012*** –0.185*** –0.062** –0.263*** –0.263*** –0.332*** –0.255*** –0.245*** –0.255*** –0.268
(0.037) (0.002) (0.035) (0.029) (0.032) (0.049) (0.070) (0.039) (0.040) (0.040) (–0.288; –0.208)
Unemployment Replacement –0.081 –0.006** –0.036 –0.039 –0.081*** –0.081* –0.073 –0.051 –0.079 –0.051 –0.08
Ratio (0.050) (0.002) (0.029) (0.024) (0.029) (0.043) (0.052) (0.049) (0.052) (0.047) (–0.088; –0.042)
Public Spending on ALMP –0.025** –0.001 –0.024** –0.003 –0.025** –0.025** –0.039** –0.026** –0.027*** –0.027*** –0.025
(0.009) (0.001) (0.010) (0.008) (0.010) (0.012) (0.012) (0.010) (0.010) (0.009) (–0.029; –0.018)
Restrictiveness of Migrant 0.056 0.001 0.131* 0.066 0.056 0.056 0.126 0.063 0.108 0.063 0.055
Integration Policies (0.088) (0.005) (0.068) (0.058) (0.069) (0.092) (0.132) (0.087) (0.092) (0.084) (–0.024; 0.11)
Union Density –0.115*** –0.006*** –0.126*** –0.118*** –0.115*** –0.115*** –0.077 –0.127*** –0.125*** –0.125*** –0.114
(0.032) (0.002) (0.036) (0.031) (0.032) (0.038) (0.052) (0.038) (0.026) (0.036) (–0.146; –0.096)
Coordination of Wage Setting 0.040 0.010 –0.016 0.102 0.040 0.040 0.803* 0.109 0.106 0.088 0.029
(0.222) (0.011) (0.185) (0.078) (0.173) (0.214) (0.363) (0.239) (0.231) (0.246) (–0.066; 0.111)
Statutory Retirement Age 0.661*** 0.035*** 0.677*** 0.495** 0.661*** 0.661** 0.505 0.591*** 0.943*** 0.594*** 0.658
(0.174) (0.010) (0.196) (0.209) (0.204) (0.321) (0.308) (0.178) (0.204) (0.179) (0.456; 0.815)
Public Spending on Old-Age –0.750*** –0.038*** –0.597*** –0.306** –0.750*** –0.750*** –0.873*** –0.826*** –0.596*** –0.840*** –0.749
Pension (0.154) (0.009) (0.196) (0.126) (0.176) (0.255) (0.095) (0.152) (0.161) (0.179) (–0.839; –0.566)
Public Spending on Incapacity -0.421 –0.025 –0.689** –0.008 –0.421 –0.421 –0.659 –0.208 –0.320 –0.203 –0.426
(0.562) (0.031) (0.348) (0.300) (0.295) (0.404) (0.831) (0.586) (0.570) (0.584) (–0.634; –0.163)
Countries 23 23 23 23 23 23 23 23 25 23
R 2 0.686 0.681 0.925 0.686 0.686 0.737 0.665 0.690 0.666
Note: The table presents results from estimation equation (2.3) with the participation rate of older workers (ages 55 and older) as the dependent variable on a sample of 23 advanced economies (AEs) during 1980–2011 using annual
cross-sectional dependence; column (6) shows the results with the Newey-West correction for the standard errors; column (7) shows the results based on a sample of five-year averages; column (8) reports the results dropping global
financial crisis (GFC) years 2008 and 2009 from the sample; column (9) reports the coefficients when Czech Republic and Slovak Republic, which recently joined AEs, are added to the sample; column (10) shows the results when
the lag of the output gap is replaced with the lag of the unemployment rate; and column (11) reports the median coefficient from a distribution of estimates obtained by dropping one country at a time from the sample. Driscoll-Kraay
standard errors are reported in parentheses in columns (1), (2), (7)–(10); bootstrapped standard errors are reported in parentheses in column (3); HAC standard errors assuming a panel-dependent correlation structure are reported
in column (4); HAC standard errors are reported in parentheses in column (5); and Newey-West corrected standard errors are reported in column (6). Column (11) reports the 10th and 90th percentile of the estimated coefficients in
parentheses. ALMP = active labor market programs.
* p < .10; ** p < .05; *** p < .01.
Annex Table 2.4.5. Drivers of Aggregate Labor Force Participation Rates: Robustness
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
Replacing Dropping
Logistic HAC Standard Newey- West Five-year Including Output Gap with One Country
Baseline Transformation Beck and Katz Errors Standard Errors Averages Excluding GFC All AEs Unemployment rate at a Time
Lag of Output Gap 0.183*** 0.008*** 0.090*** 0.183*** 0.183*** 0.250** 0.193*** 0.136*** –0.364*** 0.182
(0.044) (0.002) (0.025) (0.036) (0.042) (0.096) (0.045) (0.039) (0.037) (0.143; 0.2)
Routinization x Relative Price 0.536*** 0.022*** 0.272** 0.536*** 0.536*** 0.552* 0.506*** 0.653*** 0.548*** 0.533
of Investment (0.175) (0.007) (0.120) (0.118) (0.138) (0.247) (0.167) (0.156) (0.153) (0.459; 0.61)
Lag of Trade Openness 0.012* 0.000 0.003 0.012 0.012 0.016 0.020** 0.003 0.004 0.013
(0.007) (0.000) (0.007) (0.009) (0.010) (0.011) (0.008) (0.007) (0.007) (0.003; 0.016)
Relative Service Employment 0.010** 0.000** 0.002 0.010*** 0.010** 0.016*** 0.010** 0.013*** 0.016*** 0.011
(0.004) (0.000) (0.003) (0.004) (0.005) (0.004) (0.004) (0.004) (0.004) (0.006; 0.012)
Lag of Urbanization 0.249*** 0.011*** 0.208*** 0.249*** 0.249*** 0.240*** 0.260*** 0.257*** 0.268*** 0.25
(0.047) (0.002) (0.042) (0.039) (0.071) (0.047) (0.056) (0.045) (0.063) (0.202; 0.294)
Education (percent secondary) 0.063*** 0.003*** 0.018 0.063*** 0.063*** 0.058** 0.061*** 0.062*** 0.066*** 0.064
(0.017) (0.001) (0.014) (0.015) (0.017) (0.017) (0.019) (0.016) (0.017) (0.047; 0.074)
Education (percent tertiary) 0.135*** 0.006*** 0.108*** 0.135*** 0.135*** 0.121** 0.115*** 0.136*** 0.060** 0.134
(0.031) (0.001) (0.027) (0.032) (0.035) (0.034) (0.029) (0.031) (0.026) (0.119; 0.158)
Tax Wedge –0.240*** –0.010*** –0.073*** –0.240*** –0.240*** –0.275*** –0.223*** –0.226*** –0.125*** –0.242
(0.026) (0.001) (0.020) (0.021) (0.027) (0.029) (0.024) (0.025) (0.030) (–0.253; –0.216)
Unemployment Replacement –0.078*** –0.003*** –0.032** –0.078*** –0.078*** –0.083* –0.068** –0.076*** –0.069*** –0.076
Ratio (0.025) (0.001) (0.013) (0.017) (0.028) (0.041) (0.025) (0.024) (0.023) (–0.085; –0.067)
Public Spending on ALMP 0.031*** 0.001*** 0.017*** 0.031*** 0.031*** 0.034** 0.030*** 0.033*** 0.015* 0.031
(0.007) (0.000) (0.006) (0.006) (0.009) (0.013) (0.007) (0.007) (0.008) (0.024; 0.034)
Restrictiveness of Migrant –0.207*** –0.008*** –0.084** –0.207*** –0.207*** –0.245*** –0.191*** –0.198*** –0.230*** –0.211
Integration Policies (0.049) (0.002) (0.038) (0.040) (0.070) (0.055) (0.054) (0.047) (0.052) (–0.255; –0.184)
Union Density –0.015 –0.001 –0.064*** –0.015 –0.015 0.021 –0.030 –0.007 –0.004 –0.016
(0.025) (0.001) (0.018) (0.021) (0.031) (0.024) (0.025) (0.023) (0.022) (–0.033; –0.001)
Coordination of Wage Setting 0.256** 0.011** 0.020 0.256** 0.256* 0.289 0.274** 0.238* –0.027 0.26
(0.120) (0.005) (0.065) (0.112) (0.148) (0.302) (0.108) (0.121) (0.115) (0.203; 0.312)
Number of Observations 570 570 570 570 570 132 524 592 524
Countries 23 23 23 23 23 23 23 25 23
R2 0.578 0.569 0.983 0.578 0.578 0.596 0.560 0.567 0.602
Note: The table presents results from estimation equation (2.3) with the participation rate of all workers (ages 15 and older) as the dependent variable on a sample of 23 advanced economies (AEs) during 1980–2011 using annual
column (2) reports the results after applying the logistic transformation to the dependent variable; column (3) shows the results using the Beck and Katz (1995) estimator; column (4) reports the estimates with heteroscedasticity

and autocorrelation consistent (HAC) standard errors, without the correction for cross-sectional dependence; column (5) shows the results with the Newey-West correction for the standard errors; column (6) shows the results based
on a sample of five-year averages; column (7) reports the results dropping global financial crisis (GFC) years 2008 and 2009 from the sample; column (8) reports the coefficients when Czech Republic and Slovak Republic, which
recently joined AEs, are added to the sample; column (9) shows the results when the lag of the output gap is replaced with the lag of the unemployment rate; and column (10) reports the median coefficient from a distribution of
estimates obtained by dropping one country at a time from the sample. Driscoll-Kraay standard errors are reported in parentheses in columns (1), (2), (6)–(9); HAC standard errors assuming a panel-dependent correlation structure
121
are reported in column (3); HAC standard errors are reported in parentheses in column (4); and Newey-West corrected standard errors are reported in column (5). Column (10) reports the 10th and 90th percentile of the estimated
coefficients in parentheses. ALMP = active labor market programs.
*p < .10; **p < .05; ***p < .01.
Annex 2.5. The Role of Individual and Annex 2.6. Prospects for Labor Force
Household Characteristics: Micro‑Level Analysis Participation: Cohort-Based Analysis
The micro-level analysis relies on the European The cohort-based analysis relies on Organisation
Union Labour Force Survey for 24 advanced econo- for Economic Co-operation and Development data
mies during 2000–16. It estimates logit models on a on participation rates for a balanced sample of 17
random sample of 10,000 people per country per year. advanced economies during 1985–2016. It estimates a
The dependent variable is a dummy variable indicating system of 11 seemingly unrelated regressions (one for
whether someone is in or out of the labor force.54 each age group) for each country, and separately for
Explanatory variables include age; gender (for the 55 men and women, of the following form:
and older group); and whether the person was born in the
  = α  a,g  +  _1 T
n      ∑ t − g
LFP ta,g         β  a,g C  a        + λ  a,g X    + ε  a,g
country or abroad, whether the person lives in an urban g t − g t t  , (2.5)
or rural area, and the person’s highest level of education
in which α  a,gis
a gender- and age-specific constant;
completed (lower secondary, upper secondary, or tertiary). a      is a set of birth cohort- and gender‑specific
C t − g
The regressions also control for measures of family com-
dummy variables, which take the value 1 if the birth
position: the number of children; other employed adults
cohort t - g appears in the age group g in year t; β  a,g is
in the household; and whether the individual lives in a
a gender- and birth-year-specific fixed effect (that is,
household of a single adult without children (the baseline
the cohort effect), which is divided by the number of
category), a single adult with children, or a couple with
birth cohorts in the age group n g; Xt  is the output gap;
or without children. Finally, regressions control for the
and λ  a,gis a coefficient on the output gap that varies
routinization score of an individual’s current occupation
by gender and age group.55 Within each gender group
(if currently employed) or last occupation (if currently
and country, the coefficient for each birth cohort β  a,g
unemployed or inactive). Country, region, and year
is constrained to be the same across equations. In addi-
fixed effects are included. Results are robust if interacted
tion, each birth cohort appears in at least two equa-
country-year fixed effects are included instead. Standard
tions, which implies that the sample covers cohorts
errors are clustered at the country-year level.
born between 1925 and 1994.
The baseline specification does not control for
A series of tests ensures that the results are broadly
income due to data limitations (Annex Table 2.5.1,
robust to the application of a logistic transformation
columns 1–3). However, results are broadly robust to
to the dependent variable, replacing the output gap
controlling for the income decile of employed individ-
with the unemployment rate, and dropping more birth
uals and the predicted income decile (based on age,
cohorts at the end of the sample.
gender, education, location, immigration status, and
Age-group-specific trend labor force participa-
sector and occupation of last employment) for unem-
tion rates are obtained as the predicted values of the
ployed or inactive people (for whom income infor-
cohort-based model estimates, assuming a zero output
mation is not available). Once a (predicted) income
gap. The aggregate trend labor force participation rate
decile is included, the effect on women’s participation
is calculated as the three-year moving average of the
of being part of a couple and having children turns
age group’s specific trend labor force participation rates
positive, the effect of other employed adults in the
multiplied by its population share.
household turns negative, and income itself has a
Projected scenarios for trend labor force participa-
negative effect (Annex Table 2.5.1, columns 4–6). This
tion rely on the United Nations World Population
suggests that individuals in upper deciles may be able
Prospects data, under the assumptions of medium
to afford to drop out of the labor force. The results on
fertility and migration flows and policies based on
vulnerability to routinization and education are very
historical trends. Projections assume no effects from
similar to those in the baseline.
new cohorts entering the labor force. Three illus-
trative scenarios are built on the following assump-
54Main labor force status is coded as employed (if a person has
tions. The first assumes that for people of prime
a job or profession, including unpaid work for a family business,
apprenticeship, or paid traineeship), unemployed, or out of the labor age (25–54), women’s participation rates gradually
force (including people who are students, retired, permanently dis-
abled, in compulsory military service, fulfilling domestic tasks, and 55For example, in 1985, the birth cohort dummy variable for
otherwise inactive). This coding is assigned based on respondents’ those born between 1970 and 1974 takes the value 1 for the equa-
answers about their activity during the reference week. tion of the 15–19 age group.

Annex Table 2.5.1. Determinants of Being in the Labor Force

(1) (2) (3) (4) (5) (6)
Men, Women, All, Men, Women, All,
Ages 25–54 Ages 25–54 Ages 55+ Ages 25–54 Ages 25–54 Ages 55+
Age 1.158*** 1.320*** 1.396*** 1.261*** 1.347*** 1.356***
(0.011) (0.014) (0.113) (0.018) (0.021) (0.151)
Age Squared 0.998*** 0.997*** 0.998*** 0.997*** 0.997*** 0.998***
(0.000) (0.000) (0.001) (0.000) (0.000) (0.001)
Male 1.196*** 1.539***
(0.031) (0.046)
Upper Secondary Education 1.719*** 1.709*** 1.209*** 1.737*** 1.855*** 1.102**
(0.032) (0.033) (0.036) (0.056) (0.060) (0.046)
Tertiary Education 2.759*** 2.961*** 1.594*** 2.217*** 2.763*** 1.240***
(0.082) (0.077) (0.059) (0.097) (0.115) (0.063)
Born in Country 1.489*** 1.333*** 1.091** 1.761*** 1.520*** 1.167**
(0.035) (0.024) (0.046) (0.051) (0.050) (0.075)
Urban 1.008 1.024* 1.019 0.896*** 0.864*** 0.866***
(0.019) (0.013) (0.027) (0.027) (0.022) (0.037)
Number of Children in Household 1.049*** 0.816*** 0.960* 1.094*** 0.869*** 1.039
(0.009) (0.007) (0.020) (0.012) (0.012) (0.035)
One Adult with Children 1.042 0.846*** 1.785*** 1.045 0.846*** 1.217
(0.059) (0.026) (0.394) (0.087) (0.039) (0.330)
Couple without Children 1.356*** 0.906*** 0.842*** 1.757*** 1.741*** 1.161***
(0.035) (0.034) (0.025) (0.083) (0.128) (0.051)
Couple with Children 1.726*** 0.757*** 1.446*** 2.141*** 1.248*** 2.429***
(0.052) (0.028) (0.128) (0.114) (0.088) (0.350)
Other Household Structure 0.937** 0.868*** 0.812*** 1.212*** 1.334*** 1.726***
(0.027) (0.030) (0.038) (0.063) (0.092) (0.138)
Other Employed Adult(s) in Household 1.497*** 1.152*** 1.703*** 0.992 0.601*** 0.636***
(0.035) (0.038) (0.091) (0.043) (0.046) (0.079)
Routinization Score of Occupation 0.825*** 0.900*** 0.716*** 0.467*** 0.490*** 0.488***
(0.011) (0.010) (0.013) (0.012) (0.012) (0.016)
Lagged Output Gap 1.037*** 1.023*** 1.031*** 1.042*** 1.030*** 1.037***
(0.006) (0.004) (0.007) (0.008) (0.008) (0.012)
Predicted Income Decile1 0.952*** 0.950*** 0.952***
(0.001) (0.001) (0.002)
Number of Observations 491,820 474,240 86,441 474,434 443,687 63,982
Note: Logit regressions are based on a random sample of 10,000 respondents per country per year of 19 countries, exponentiated coefficients. All specifica-
tions include country, region, and year fixed effects. The base category for education is “up to lower secondary education.” For family composition the base
category is “one adult without children.” Standard errors clustered at the country-year level.
*p < .10; **p < .05; ***p < .01.
1Predicted income decile uses the actual income decile for those currently employed and predicts the income decile for those currently unemployed/inactive
using age, gender, education, migration status, location, sector, and occupation; and country, region, and year fixed effects.
converge to those of men over the next 20 years. The References

second scenario assumes that the participation rate of Aaronson, Stephanie, Tomaz Cajner, Bruce Fallick, Felix
those ages 55–59 converges to the rate of the 50–54 Galbis-Reig, Christopher Smith, and William Wascher.
age group over the next 20 years and that the rate 2014. “Labor Force Participation: Recent Developments and
for the 60–64 age group becomes the same as for the Future Prospects.” Brookings Papers on Economic Activity 45
(2): 197–275.
50–54 age group over the next 40 years. The third
Aaronson, Stephanie, Bruce Fallick, Andrew Figure, Jonathan
scenario assumes that policies converge to the best
Pingle, and William Wascher. 2006. “The Recent Decline in
possible levels, defined as the 90th (or 10th) percen- the Labor Force Participation Rate and Its Implications for
tile of the level observed among advanced economies, Potential Labor Supply.” Brookings Papers on Economic Activity
over the next 20 years. The impact is then simulated 37 (1): 69–154.
using the coefficients estimated in the cross-country Abraham, Katharine, and Melissa Kearney. 2018. “Explaining
empirical model. the Decline in the US Employment-to-Population Ratio:

A Review of the Evidence.” NBER Working Paper 24333, Baron-Cohen, Simon, Rebecca Knickmeyer, and Matthew Bel-
National Bureau of Economic Research, Cambridge, MA. monte. 2005. “Sex Differences in the Brain: Implications for
Acemoglu, Daron, and David Autor. 2011. “Skills, Tasks and Explaining Autism.” Science 310 (4): 819–23.
Technologies: Implications for Employment and Earnings.” Bassanini, Andrea, and Romain Duval. 2006. “Employment
In Handbook of Labor Economics, edited by Orley Ashenfelter Patterns in OECD Countries: Reassessing the Role of Policies
and David Card. 4: 1043–171. Amsterdam: Elsevier. and Institutions.” OECD Economic Department Working
Acemoglu, Daron, and Pascual Restrepo. 2017. “Robots and Paper 486, Organisation for Economic Co-operation and
Jobs: Evidence from US Labor Markets.” NBER Work- Development, Paris.
ing Paper 23285, National Bureau of Economic Research, ———. 2009. “Unemployment, Institutions, and Reform Com-
Cambridge, MA. plementarities: Reassessing the Aggregate Evidence for OECD
Aguiar, Mark, Mark Bils, Kerwin Kofi Charles, and Erik Hurst. Countries.” Oxford Review of Economic Policy 25 (1): 40–59.
2017. “Leisure Luxuries and the Labor Supply of Young Beck, Nathaniel, and Jonathan Katz. 1995. “What to Do (and
Men.” NBER Working Paper 23552, National Bureau of Not to Do) with Time-Series Cross-Section Data.” American
Economic Research, Cambridge, MA. Political Science Review 89 (3): 634–47.
Ahn, Jaebin, Zidong An, John Bluedorn, Gabriele Ciminelli, Bertola, Giuseppe, Francine Blau, and Lawrence Kahn. 2007.
Zsóka Kóczán, Davide Malacrino, Daniela Muhaj, and “Labor Market Institutions and Demographic Employment
Patricia Neidlinger. Forthcoming. “Youth Labor Markets in Patterns.” Journal of Population Economics 20 (4): 833–67.
Emerging Market and Developing Economies: Drivers and Betcherman, Gordon. 2012. “Labor Market Institutions: A
Policies.” IMF Staff Discussion Note, International Monetary Review of the Literature.” World Bank Policy Research Work-
Fund, Washington, DC. ing Paper 6276, World Bank, Washington, DC.
Aiyar, Shekhar, Bergljot Barkbu, Nicoletta Batini, Helge Berger, Blagrave, Patrick, and Marika Santoro. 2017. “Labor Force Par-
Enrica Detragiache, Allan Dizioli, Christian Ebeke, Huidan ticipation in Chile: Recent Trends, Drivers, and Prospects.”
Lin, Linda Kaltani, Sebastian Sosa, Antonio Spilimbergo, IMF Working Paper 17/54, International Monetary Fund,
and Petia Topalova. 2016. “The Refugee Surge in Europe: Washington, DC.
Economic Challenges.” IMF Staff Discussion Note 16/02, Blanchard, Olivier, and Justin Wolfers. 2000. “The Role of
International Monetary Fund, Washington, DC. Shocks and Institutions in the Rise of European Unem-
Alesina, Alberto, Johann Harnoss, and Hillel Rapoport. 2015. ployment: The Aggregate Evidence.” Economic Journal 110
“Birthplace Diversity and Economic Prosperity.” NBER (462): C1–33.
Working Paper 18699, National Bureau of Economic Blau, David, and Ryan Goodstein. 2008. “Can Social Secu-
Research, Cambridge, MA. rity Explain Trends in Labor Force Participation of Older
Arulampalam, Wiji, Alison Booth, and Mark Taylor. 2000. Men in the United States?” Journal of Human Resources 45
“Unemployment Persistence.” Oxford Economic Papers (2): 328–63.
52 (1): 24–50. Blau, Francine, and Lawrence Kahn. 2013. “Female Labor
Arulampalam, Wiji, Paul Gregg, and Mary Gregory. 2001. Supply: Why Is the US Falling Behind?” American Economic
“Unemployment Scarring.” The Economic Journal 111 Review 103 (3): 251–56.
(475): 577–84. Blöndal, Sveinbjörn, and Stefano Scarpetta. 1999. “The Retire-
Autor, David, and David Dorn. 2013. “The Growth of Low-Skill ment Decision in OECD Countries.” OECD Economics
Service Jobs and the Polarization of the US Labor Market.” Department Working Paper 202, Organisation for Economic
American Economic Review 103 (5): 1553–97. Co-operation and Development, Paris.
Autor, David, David Dorn, and Gordon Hanson. 2016. “The Bloom, David, Somnath Chatterji, Paul Kowal, Peter
China Syndrome: Local Labor Market Effects of Import Lloyd-Sherlock, Martin McKee, Bernd Rechel, Larry Rosen-
Competition in the United States.” American Economic Review berg, and James Smith. 2015. “Macroeconomic Implications
103 (6): 2121–68. of Population Ageing and Selected Policy Responses.” The
Balakrishnan, Ravi, Mai Dao, Juan Solé, and Jeremy Zook. Lancet 385: 649–57.
2015. “Recent US Labor Force Dynamics: Reversible or Blundell, Richard, Eric French, and Gemma Tetlow. 2016.
Not?” IMF Working Paper 15/76, International Monetary “Retirement Incentives and Labor Supply.” In Handbook of
Fund, Washington, DC. the Economics of Population Aging, edited by John Piggott and
Balleer, Almut, Ramon Gómez-Salvador, and Jarkko Turunen. Alan Woodland. 1: 457–566. Amsterdam: Elsevier.
2014. “Labor Force Participation across Europe: A Blundell, Richard, and Thomas Macurdy. 1999. “Labor Supply:
Cohort-Based Analysis.” Empirical Economics 46: 1385–415. A Review of Alternative Approaches.” In Handbook of Labor
Banerji, Angana, Huidan Lin, and Sergejs Saksonovs. 2015. Economics, edited by Orley C. Ashenfelter and David Card.
“Youth Unemployment in Advanced Europe: Okun’s Law and 3A: 1559–695. Amsterdam: Elsevier.
Beyond.” IMF Working Paper 15/5, International Monetary Börsch-Supan, Axel, and Irene Ferrari. 2017. “Old-Age Labor
Fund, Washington, DC. Force Participation in Germany: What Explains the Trend

Reversal among Older Men? And What the Steady Increase Labor Market.” NBER Working Paper 24274, National
Among Women?” NBER Working Paper 24044, National Bureau of Economic Research, Cambridge, MA.
Bureau of Economic Research, Cambridge, MA. Council of Economic Advisers. 2014. “The Labor Force Partic-
Burniaux, Jean-Marc, Romain Duval, and Florence Jaumotte. ipation Rate since 2007: Causes and Policy Implications.”
2004. “Coping with Aging: A Dynamic Approach to Executive Office of the President of the United States, July.
Quantify the Impact of Alternative Policy Options for Future ———. 2016. “The Long-Term Decline in Prime-Age Male
Labor Supply in OECD Countries.” OECD Economics Labor Force Participation.” Executive Office of the President
Department Working Paper 371, Organisation for Economic of the United States, June.
Co-operation and Development, Paris. D’Amuri, Francesco, and Giovanni Peri. 2014. “Immigration,
Calvo, Esteban. 2006. “Does Working Longer Make People Jobs, and Employment Protection: Evidence from Europe
Healthier and Happier?” Work Opportunities for Older before and during the Great Recession.” Journal of the Euro-
Americans Series 2. Center for Retirement Research, Chest- pean Economic Association 12 (2): 432–64.
nut Hill, MA. Dao, Mai, Mitali Das, Zsóka Kóczán, and Weicheng Lian. 2017.
Canon, Maria, Marianna Kudlyak, and Yang Liu. 2015. “Youth “Why Is Labor Receiving a Smaller Share of Global Income?
Labor Force Participation Continues to Fall, but It Might Be Theory and Empirical Evidence.” IMF Working Paper
for a Good Reason.” Regional Economist (January). 17/169, International Monetary Fund, Washington, DC.
Card, David. 1990. “The Impact of the Mariel Boatlift on the Dao, Mai, Davide Furceri, Jisoo Hwang, Meeyeon Kim, and
Miami Labor Market.” Industrial and Labor Relations Review Tae-Jeong Kim. 2014. “Strategies for Reforming Korea’s
43 (2): 245–57. Labor Market to Foster Growth.” IMF Working Paper
Card, David, Jochen Kluve, and Andrea Weber. 2010. “Active 14/137, International Monetary Fund, Washington, DC.
Labour Market Policy Evaluations: A Meta-Analysis.” Eco- Dao, Mai, Davide Furceri, and Prakash Loungani. 2014.
nomic Journal 120 (548): F452–F477. “Regional Labor Market Adjustments in the United States
Carrasco, Raquel, Juan F. Jimeno, and A. Carolina Ortega. 2008. and Europe.” IMF Working Paper 14/26, International Mon-
“The Effect of Immigration on the Labor Market Perfor- etary Fund, Washington, DC.
mance of Native-Born Workers: Some Evidence for Spain.” Das, Mitali, and Benjamin Hilgenstock. Forthcoming. “Labor
Journal of Population Economics 21 (3): 627–48. Market Consequences of Routinization in Developed and
Case, Anne, and Angus Deaton. 2017. “Mortality and Morbid- Developing Economies.” IMF Working Paper, International
ity in the 21st Century.” Brookings Papers on Economic Activity Monetary Fund, Washington, DC.
(Spring): 397–476. Dauth, Wolfgang, Sebastian Findeisen, and Jens Suedekum.
Cattaneo, Cristina, Carlo V. Fiorio, and Giovanni Peri. 2015. 2014. “The Rise of the East and the Far East: German Labor
“What Happens to the Careers of European Workers when Markets and Trade Integration.” Journal of the European
Immigrants ‘Take Their Jobs’?” Journal of Human Resources 50 Economic Association 12 (6): 1643–75.
(3): 655–93. de Haas, Hein, Katharina Natter, and Simona Vezzoli. 2014.
Centre for Cities. 2018. “Cities Outlook 2018.” London. “Compiling and Coding Migration Policies. Insights from
Christiansen, Lone, Huidan Lin, Joana Pereira, Petia Topalova, the DEMIG POLICY Database.” International Migration
and Rima Turk. 2016a. “Gender Diversity in Senior Positions Institute Working Paper 87, Oxford, United Kingdom.
and Firm Performance: Evidence from Europe.” IMF Working de Serres, Alain, Fabrice Murtin, and Christine Maisonneuve.
Paper 16/50, International Monetary Fund, Washington, DC. 2012. “Policies to Facilitate the Return to Work.” Compara-
———. 2016b. “Individual Choice or Policies? Drivers of tive Economic Studies 54: 5–42.
Female Employment in Europe.” IMF Working Paper 16/49, Driscoll, John, and Aart Kraay. 1998. “Consistent Covariance
International Monetary Fund, Washington, DC. Matrix Estimation with Spatially Dependent Panel Data.”
Cipollone, Angela, Eleonora Patacchini, and Giovanna Vallanti. Review of Economics and Statistics 80 (4): 549–60.
2013. “Women Labor Market Participation in Europe: Duval, Romain. 2004. “Retirement Behaviour in OECD Countries:
Novel Evidence on Trends and Shaping Factors.” IZA Impact of Old-Age Pension Schemes and Other Social Transfer
Discussion Paper 7710, IZA Institute of Labor Economics, Programmes.” OECD Economic Studies 2003 (2): 7–50.
Bonn, Germany. Duval, Romain, Mehmet Eris, and Davide Furceri. 2011. “The
Clements, Benedict, Kamil Dybczak, Vitor Gaspar, Sanjeev Effects of Downturns on Labour Force Participation: Evi-
Gupta, and Mauricio Soto. 2015. “The Fiscal Consequences dence and Causes.” Economics Department Working Paper
of Shrinking Population.” IMF Staff Discussion Note 15/21, 875, Organisation for Economic Co-operation and Devel-
International Monetary Fund, Washington, DC. opment, Paris.
Cohen, Sheldon. 2004. “Social Relationships and Health.” Amer- Eberstadt, Nicholas. 2016. Men without Work: America’s Invisible
ican Psychologist 59 (8): 676–84. Crisis. West Conshohocken, PA: Templeton Press.
Cortes, Guido Matias, Nir Jaimovich, and Henry E. Siu. 2018. Elborgh-Woytek, Katrin, Monique Newiak, Kalpana Kochhar,
“The End of Men and Rise of Women in the High-Skilled Stefania Fabrizio, Kangni Kpodar, Philippe Wingender,

Benedict Clements, and Gerd Schwartz. 2013. “Women, Genda, Yuji, Ayako Kondo, and Souichi Ohta. 2010.
Work and the Economy: Macroeconomic Gains from Gender “Long-Term Effects of a Recession at Labor Market Entry in
Equity.” IMF Staff Discussion Note 13/10, International Japan and the United States.” Journal of Human Resources 45
Monetary Fund, Washington, DC. (1): 157–96.
Eliason, Marcus, and Donald Storrie. 2009. “Does Job Genre, Veronique, Ramon Gómez-Salvador, and Ana Lamo.
Loss Shorten Life?” The Journal of Human Resources 44 2005. “The Determinants of Labour Force Participation in
(2): 277–302. the European Union.” In Labour Supply and Incentives to
Elsby, Michael, Bart Hobijn, and Aysegül Sahin. 2015. “On the Work in Europe, edited by Ramon Gómez-Salvador, Ana
Importance of the Participation Margin for Labor Market Lamo, Barbara Petrongolo, Melanie Ward, and Etienne
Fluctuations.” Journal of Monetary Policy 72: 64–82. Wasmer. Cheltenham, United Kingdom, and Northampton,
Erceg, Christopher, and Andrew Levin. 2014. “Labor Force MA: Edward Elgar.
Participation and Monetary Policy in the Wake of the Great ———. 2010. “European Women: Why Do(n’t) They Work?”
Recession.” Journal of Money, Credit and Banking 46: 3–49. Applied Economics 42 (12): 1499–514.
Erikson, Erik, Joan Erikson, and Helen Kivnick. 1986. Vital Gerdtham, Ulf-G, and Magnus Johannesson. 2003. “A Note on
Involvement in Old Age. New York: Norton. the Effect of Unemployment on Mortality.” Journal of Health
European Commission (EC). 2017. “Technical Note: Summary Economics 22 (3): 505–18.
Methodology of the 2015-based Population Projections.” Goldin, Claudia. 2006. “The Quiet Revolution That Trans-
Annex to Eurostat Metadata. Luxembourg. formed Women’s Employment, Education, and Family.”
Euwals, Rob, Marike Knoef, and Daniel van Vuuren. 2011. “The American Economic Review 96 (2): 1–21.
Trend in Female Labour Force Participation: What Can Be Gonzales, Christian, Sonali Jain-Chandra, Kalpana Kochhar, and
Expected for the Future?” Empirical Economics 40: 729–53. Monique Newiak. 2015a. “Fair Play: More Equal Laws Boost
Fallick, Bruce, and Jonathan Pingle. 2007. “A Cohort-Based Female Labor Force Participation.” IMF Staff Discussion
Model of Labor Force Participation.” Finance and Eco- Note 15/02, International Monetary Fund, Washington, DC.
nomics Discussion Series 2007–9, Federal Reserve Board, Gonzales, Christian, Sonali Jain-Chandra, Kalpana Kochhar,
Washington, DC. Monique Newiak, and Tlek Zeinullayev. 2015b. “Catalyst
Farré, Lidia, Libertad González, and Francesc Ortega. 2011. for Change: Empowering Women and Tackling Income
“Immigration, Family Responsibility and the Labor Supply of Inequality.” IMF Staff Discussion Note 15/20, International
Skilled Native Women.” B.E. Journal of Economic Analysis & Monetary Fund, Washington, DC.
Policy 11 (1): Article 34. Goos, Maarten, Alan Manning, and Anna Salomons. 2014.
Feingold, Alan. 1994. “Gender Differences in Personality: A “Explaining Job Polarization: Routine-Biased Technological
Meta-Analysis.” Psychological Bulletin 116 (3): 429–56. Change and Offshoring.” American Economic Review 104
Fernandez, Raquel. 2013. “Cultural Change as Learning: The (8): 2509–26.
Evolution of Female Labor Force Participation over a Cen- Gourinchas, Pierre-Olivier, and Maurice Obstfeld. 2012. “Stories
tury.” American Economic Review 103 (1): 472–500. of the Twentieth Century for the Twenty-First.” American
Fitzenberger, Bernd, and Gaby Wunderlich. 2004. “The Chang- Economic Journal: Macroeconomics 4 (1): 226–65.
ing Life Cycle Pattern in Female Employment: A Comparison Grigoli, Francesco, Alexander Herman, Andrew Swiston, and
of Germany and the UK.” Scottish Journal of Political Economy Gabriel Di Bella. 2015. “Output Gap Uncertainty and
51 (3): 302–28. Real-Time Monetary Policy.” Russian Journal of Economics 1
Foged, Mette, and Giovanni Peri. 2015. “Immigrants’ Effect on (4): 329–58.
Native Workers: New Analysis on Longitudinal Data.” IZA Grigoli, Francesco, and Adrian Robles. 2017. “Inequality Over-
Discussion Paper 8961, IZA Institute of Labor Economics, hang.” IMF Working Paper 17/76, International Monetary
Bonn, Germany. Fund, Washington, DC.
Frees, Edward. 1995. “Assessing Cross-Sectional Correlations in Gruber, Jonathan, and David Wise. 1999. Social Security
Panel Data.” Journal of Econometrics 69 (2): 393–414. and Retirement around the World. Chicago: University of
Friedman, Milton. 1937. “The Use of Ranks to Avoid the Assump- Chicago Press.
tion of Normality Implicit in the Analysis of Variance.” Journal of Hall, Robert. 2015. “Quantifying the Lasting Harm to the US
the American Statistical Association 32 (200): 675–701. Economy from the Financial Crisis.” In NBER Macroeco-
Gal, Peter, and Adam Theising. 2015. “The Macroeconomic nomics Annual 2014, edited by Jonathan Parker and Michael
Impact of Policies on Labour Market Outcomes in OECD Woodford. 29: 71–128. Chicago: University of Chicago Press.
Countries: A Reassessment.” OECD Economics Depart- Heckman, James, Robert Lalonde, and Jeffrey Smith. 1999.
ment Working Paper 1271, Organisation for Economic “The Economic and Econometrics of Active Labor Mar-
Co-operation and Development, Paris. ket Programs.” In Handbook of Labor Economics, edited by
Galor, Oded, and David N. Weil. 1996. “The Gender Gap, Fertil- Orley Ashenfelter and David Card. 3: 1865–2097. Amster-
ity, and Growth.” American Economic Review 86 (3): 374–87. dam: Elsevier.

Holzer, Harry, Paul Offner, and Elaine Sorensen. 2005. “Declin- Minnesota Population Center. 2017. Integrated Public Use
ing Employment among Young Black Less-Educated Men: Microdata Series, International: Version 6.5, various data
The Role of Incarceration and Child Support.” Journal of sets. Minneapolis: University of Minnesota. http://doi.org/10
Policy Analysis and Management 24 (2): 329–50. .18128/D020.V6.5.
Hunt, Jennifer, and Marjolaine Gauthier-Loiselle. 2010. “How Murtin, Fabrice, Alain de Serres, and Alexander Hijzen. 2014.
Much Does Immigration Boost Innovation?” American Eco- “Unemployment and the Coverage Extension of Collective
nomic Journal: Macroeconomics 2 (2): 31–56. Wage Agreements.” European Economic Review 71: 52–66.
Hurt, Muchale, and Susann Rohwedder. 2011. “Trends in Labor Ngai, L. Rachel, and Barbara Petrongolo. 2017. “Gender Gaps
Force Participation: How Much Is Due to Changes in Pen- and the Rise of the Service Economy.” American Economic
sions?” Journal of Population Ageing 4 (1–2): 81–96. Journal: Macroeconomics 9 (4): 1–44.
International Monetary Fund (IMF). 2015. “International Nickell, Stephen. 1981. “Biases in Dynamic Models with Fixed
Migration: Recent Trends, Economic Impacts, and Policy Effects.” Econometrica 49 (6): 1417–426.
Implications.” Staff Background Paper for G20 Surveillance Olivetti, Claudia, and Barbara Petrongolo. 2016. “The Evolution
Note. Washington, DC. of Gender Gaps in Industrialized Countries.” Annual Review
IMF, World Bank, and World Trade Organization (IMF/WB/ of Economics 8: 405–34.
WTO). 2017. “Making Trade an Engine of Growth for All: ———. 2017. “The Economic Consequences of Family Policies:
The Case for Trade and for Policies to Facilitate Adjustment.” Lessons from a Century of Legislation in High-Income
Policy Papers for Discussion at the Meeting of G20 Sherpas, Countries.” Journal of Economic Perspectives 31 (1): 205–30.
March 23–24, Frankfurt. Oreopoulos, Philip, Till von Wachter, and Andrew
Janssen, Simon. 2018. “The Decentralization of Wage Bargain- Heisz. 2012. “The Short- and Long-Term Career Effects
ing and Income Losses after Worker Displacement.” Journal of of Graduating in a Recession.” American Economic Journal:
the European Economic Association 16 (1): 77–122. Applied Economics 4 (1): 1–29.
Jaumotte, Florence. 2003. “Female Labour Force Participation: Organisation for Economic Co-operation and Development
Past Trends and Main Determinants in OECD Countries.” (OECD). 2001. Aging and Income. Financial Resources and
Economics Department Working Paper 376, Organisation for Retirement in 9 OECD Countries. Paris.
Economic Co-operation and Development, Paris. ————. 2004 “Employment Protection Regulation and
Jaumotte, Florence, Ksenia Koloskova, and Sweta Saxena. 2016. Labour Market Performance.” OECD Employment Outlook
“Impact of Migration on Income Levels in Advanced Economies.” 2004, Chapter 2: 61–126. Paris.
Spillover Note 8, International Monetary Fund, Washington, DC. ————. 2010 “Institutional and Policy Determinants of
Kahn, Lisa. 2010. “The Long-Term Labor Market Consequences Labour Market Flows.” OECD Employment Outlook 2010,
of Graduating College in a Bad Economy.” Labour Economics Chapter 3: 167–210. Paris.
17 (2): 303–16. Ortega, Francesc, and Giovanni Peri. 2014. “Openness and
Kazandjian, Romina, Lisa Kolovich, Kalpana Kochhar, and Income: The Role of Trade and Migration.” Journal of Inter-
Monique Newiak. 2016. “Gender Equality and Economic national Economics 92: 231–51.
Diversification.” IMF Working Paper 16/140, International Pager, Devah, Bruce Western, and Naomi Sugie. 2009. “Sequencing
Monetary Fund, Washington, DC. Disadvantage: Barriers to Employment Facing Young Black and
Knabe, Andrea, and Steffen Ratzel. 2011. “Scarring or Scaring? White Men with Criminal Records.” Annals of the American
The Psychological Impact of Past Unemployment and Future Academy of Political and Social Sciences 623 (May): 195–213.
Unemployment Risk.” Economica 78 (310): 283–93. Peri, Giovanni. 2014. “Do Immigrant Workers Depress the
Krause, Eleanor, and Isabel Sawhill. 2017. “What We Know and Wages of Native Workers?” IZA World of Labor 2014: 42.
Don’t Know about Declining Labor Force Participation: A Peri, Giovanni, Kevin Shih, and Chad Sparber. 2015. “STEM
Review.” Brookings Institution, Washington, DC. Workers, H-1B Visas, and Productivity in US Cities.” Journal
Krueger, Alan. 2017. “Where Have All the Workers Gone? An of Labor Economics 33 (3): S225–55.
Inquiry into the Decline in the US Labor Force Participation Peri, Giovanni, and Chad Sparber. 2009. “Task Specialization,
Rate.” Brooking Papers on Economic Activity (September). Immigration, and Wages.” American Economic Journal: Applied
Lee, Jong-Wha, and Hanol Lee. 2016. “Human Capital in the Economics 1 (3): 135–69.
Long Run.” Journal of Development Economics 122: 147–169. Pesaran, M. H. 2004. “General Diagnostic Tests for Cross Sec-
Lucas, Richard, Andrew Clark, Yannis Georgellis, and Ed tion Dependence in Panels.” Cambridge Working Papers in
Diener. 2004. “Unemployment Alters the Set Point for Life Economics 0435, University of Cambridge.
Satisfaction.” Psychological Science 15 (1): 8–13. Rendall, Michelle. 2017. “Brain versus Brawn: The Realization
Miani, Celine, and Stijn Hoorens. 2014. “Parents at Work: of Women’s Comparative Advantage.” IEW Working Papers
Men and Women Participating in the Labour Force.” Rand 491, University of Zurich.
Europe, Short Statistical Report No. 2. https://www.rand.org/ Sahay, Ratna, Martin Čihák, Papa N’Diaye, Adolfo Barajas,
pubs/research_reports/RR348.html Annette Kyobe, Srobona Mitra, Yen Nian Mooi, and Seyed

Reza Yousefi. 2017. “Banking on Women Leaders: A Case for Sullivan, Daniel, and Till von Wachter. 2009. “Job Displace-
More?” IMF Working Paper 17/199, International Monetary ment and Mortality: An Analysis Using Administrative
Fund, Washington, DC. Data.” Quarterly Journal of Economics 124 (3): 1265–306.
Salomon, Joshua, Haidong Wang, Michael Freeman, Theo Thévenon, Olivier. 2013. “Drivers of Female Labour Force Par-
Vos, Abraham Flaxman, Alan Lopez, and Christopher ticipation in the OECD.” OECD Social, Employment and
Murray. 2012. “Healthy Life Expectancy for 187 Countries, Migration Working Paper 145, Organisation for Economic
1990–2010: A Systematic Analysis for the Global Burden Co-operation and Development, Paris.
Disease Study 2010.” The Lancet 380: 2144–62. United Nations (UN), Department of Economic and Social Affairs,
Sanchez, Thomas, Qing Shen, and Zhong-Ren Peng. 2004. Population Division. 2017. World Population Prospects: The 2017
“Transit Mobility, Jobs Access and Low-Income Labour Revision, Methodology of the United Nations Population Estimates
Participation in US Metropolitan Areas.” Urban Studies 41 and Projections. Working Paper ESA/P/WP250, New York.
(7): 1313–31. Weingarden, Alison. 2017. “Labor Market Outcomes in Met-
Schmitt, John, and Kris Warner. 2010. “Ex-Offenders and the ropolitan and Non-Metropolitan Areas: Signs of Growing
Labor Market.” CEPR Issue Brief, Center for Economic and Disparities.” FEDS Notes, September 25, Federal Reserve
Policy Research, Washington, DC. Board, Washington, DC.
Soskice, David. 1990. “Wage Determination: The Changing Winkelmann, Liliana, and Rainer Winkelmann. 1995. “Hap-
Role of Institutions in Advanced Industrialized Countries.” piness and Unemployment: A Panel Data Analysis for Ger-
Oxford Review of Economic Policy 6 (4): 36–61. many.” Applied Economics Quarterly 41 (4): 293–307.

3
CHAPTER
MANUFACTURING JOBS: IMPLICATIONS FOR PRODUCTIVITY
AND INEQUALITY
The declining share of manufacturing jobs in overall among more recent developers it seems to be peak-
employment has been a concern for policymakers and the ing at relatively low shares of total employment and
broader public alike in both advanced economies and at levels of national income below those in market
some developing economies. This concern stems from the economies that emerged earlier.1 The share of jobs in
widely held belief that manufacturing plays a unique the service sector has risen almost everywhere, replac-
role as a catalyst for productivity growth and income ing jobs in either manufacturing (mostly in advanced
convergence and a source of well-paid jobs for less-skilled economies) or agriculture (in developing economies;
workers. Against that backdrop, this chapter aims to Figure 3.2). From a long-term economic perspective,
provide new evidence on the role of manufacturing in the shift of capital and labor into different forms of
the dynamics of output per worker and in the level and economic activity is accepted as “structural transforma-
distribution of labor earnings. The two main takeaways tion”—the natural consequence of changes in demand,
from the analysis are that (1) a shift in employment from technology, and tradability.
manufacturing to services need not hinder economy-wide The implications of the reduced share of manufac-
productivity growth and the prospects for developing turing in employment has been much debated, with
economies to gain ground toward advanced economy researchers and policymakers focusing on two ques-
income levels, and (2) while the displacement of workers tions: (1) Does it hinder overall growth? (2) Does it
from manufacturing to services in advanced economies raise inequality?
has coincided with a rise in labor income inequality, Regarding the first question, the growth of productiv-
this increase was mainly driven by larger disparities in ity and of income has historically appeared to slow once
earnings across all sectors. These findings imply that the factors of production begin to shift from manufacturing
goal of supporting equitable growth would be better served to services (Baumol 1967; Imbs 2016). This phenom-
by policy efforts to raise productivity across all sectors and enon could be especially worrisome for developing
make the gains from higher productivity more inclusive. economies where employment shares are shifting from
Facilitating the reallocation of labor to productively agriculture to services, bypassing manufacturing, given
dynamic sectors, including by removing barriers to that skipping a traditional industrialization phase could
entry and trade in the service sector and supporting the hinder their ability to narrow income gaps vis-à-vis
reskilling of workers affected by structural change, is advanced economies (Rodrik 2016). However, whether
crucial to raise productivity and combat inequality. an expanding service sector necessarily weighs on
economy-wide productivity growth is an open question.
The service sector comprises subsectors with potentially
Introduction varying productivity levels and growth rates; recent
In many countries, manufacturing appears to have advances in technology and in the tradability of services
faded as a source of jobs (Figure 3.1). Its share in may have accelerated the productivity gains in some of
employment in advanced economies has been declin- them. The impact of the shifts in employment shares on
ing for nearly five decades. In developing economies, aggregate productivity would therefore depend on the
manufacturing employment has been more stable, but exact mix of subsectors that are gaining or losing share.
The authors of this chapter are Wenjie Chen, Bertrand Gruss 1In this chapter, emerging market and developing economies, or
(lead), Nan Li, Weicheng Lian, Natalija Novta, and Yu Shi, with developing economies for short, are an augmented group consisting
support from Felicia Belostecinic, Hao Jiang, Evgenia Pugacheva, of all emerging market and developing economies currently classified
and Jilun Xing and contributions from Jorge Alvarez and Ke Wang. as such by the World Economic Outlook (WEO) plus those that
We are grateful to Jesus Felipe for sharing his data on manufacturing have been reclassified as advanced economies since 1996 (the latter
employment. Joseph P. Kaboski was the external consultant. The including, for instance, Hong Kong Special Administrative Region,
chapter benefited from comments by Mary Hallward-Driemeier, Israel, Korea, and Singapore). See Annex 3.1 for data sources and
Andrei Levchenko, and Dani Rodrik. sample coverage.

Figure 3.1. Share of Manufacturing in Aggregate Employment Figure 3.2. Sectoral Employment Shares
(Percent) (Percent)
Manufacturing employment has been in relative decline for nearly five decades in The share of service sector jobs in overall employment has risen almost
advanced economies, and it seems to be peaking at low shares of total everywhere, reflecting a shift away from manufacturing employment in advanced
employment among more recent developers. economies and mostly a shift from agriculture in developing economies.
30 Agriculture Manufacturing Services Others
1. Advanced Economies
100
20 80
60
40
10 20
Emerging market and developing economies 0

Advanced economies 1970–79 80–89 90–99 2000–15
China
2. Emerging Market and Developing Economies
0 100
1970–79 80–89 90–99 2000–09 10–15
80
Note: The solid lines and shaded areas denote the simple average and the 60
interquartile range across economies, respectively. The sample comprises
21 advanced economies and 44 emerging market and developing economies with
sectoral employment data since 1970. See Annex 3.1 for data sources and country 40
coverage.
20
0
1970–79 80–89 90–99 2000–15
The second question arises because low- and
middle-skilled workers have traditionally earned higher Sources: Groningen Growth and Development Centre database; World KLEMS
wages in manufacturing than in services (Helper, database; and IMF staff calculations.
Note: The employment sectoral shares in each panel are computed as the
Krueger, and Wial 2012; Langdon and Lehrman weighted average across all economies in the group with weights given by total
2012; Lawrence 2017); a reduced employment share employment of each country. “Others” includes mining, construction, and utilities.
for manufacturing would thus tend to worsen income
inequality. Countries where inequality in labor earn-
Changes in the share of manufacturing jobs in
ings has risen since 1980 have typically experienced
employment have been accompanied by even more
a decline in the share of manufacturing employment
diverse changes in the output share of manufacturing
(Figure 3.3). But analysis of the mechanisms underly-
across countries.2 Moreover, a few developing econo-
ing that correlation has been sparse. Countries where
mies have experienced sizable increases in the share of
the share of manufacturing employment has declined
manufacturing in both employment and output since
more may also have been more exposed to other
the early 1970s, most notably China. This heteroge-
inequality-enhancing trends (such as technological
nous picture could reflect reallocation of production
change and the automation of routine tasks), with
across countries or country variations in the demand
a consequent rise in labor income inequality within
for manufactures, or a mix of both.
all sectors. The significance of the latter explanation
Against that backdrop, this chapter has two related
warrants review because it could mean that, to combat
goals: (1) to contribute to a better understanding of the
inequality, policy should focus on ensuring more inclu-
sive gains from structural transformation rather than 2In this chapter, output is measured in constant prices (or equiva-
on supporting manufacturing employment. lently, as value added in real terms).

CHAPTER 3 Manufacturing Jobs: Implications for Productivit y and Inequalit y
Figure 3.3. Change in Manufacturing Employment Share and pose a major challenge for developing economies
Inequality, 1980–2010 in narrowing per capita income gaps vis-à-vis
(Percentage points on x-axis; points on y-axis)
advanced economies?
•• Income inequality: Does manufacturing uniquely offer
Inequality in labor earnings has tended to increase more in economies that have
registered a steeper decline in the share of manufacturing employment. higher incomes or a more uniform distribution of
earnings across employees? How much of the increase
20 in inequality observed in many countries over the
CHN past few decades is associated with changes in the
AUT
15 relative size of the manufacturing sector?
DEU SWE
10 The main findings of the chapter are as follows:
Change in Gini coefficient
USA DNK
GBR MEX IND •• The heterogenous evolution of manufacturing
5 HKG output and employment shares across countries
KOR reflects a mix of forces: diverse trends in domestic
0 incomes and the associated variation in the demand
FRA
ARG NGA for manufactures, varying productivity trends in
–5 BRA
CHL manufacturing and other sectors, and specialization
and reallocation of production based on comparative
–10 Emerging market and developing economies advantages, facilitated by international trade and
Advanced economies
financial integration. Even though output has out-
–15 paced employment in the manufacturing sector in
–35 –30 –25 –20 –15 –10 –5 0 5 10
Change in manufacturing employment share most countries since the early 1970s, reflecting com-
paratively fast productivity growth in the sector, the
Sources: Standardized World Income Inequality Database (Solt 2016); and IMF same pattern has not held at the global level. The
staff calculations.
broadly parallel movement of global manufacturing
Note: The changes are calculated between the averages during 1980–89 and
2010–16. The Gini coefficient is based on income before taxes and transfers and output and employment shares reflects a change in
ranges from 0 to 100. Data labels use International Organization for the country composition of global manufacturing
Standardization (ISO) country codes.
employment in favor of developing economies,
where output per worker tends to be lower.
•• The rise of services and the decline or leveling-off of
ongoing transformation of manufacturing activity within manufacturing as a source of employment need not
countries and at the global level, and (2) to examine hinder economy-wide productivity growth. Some
whether manufacturing is indeed special in terms of out- service industries have higher productivity levels and
put per worker and the level and distribution of labor growth rates than manufacturing overall. Since the
earnings, so as to provide insight into how policies can early 2000s, the rise in the service share of employ-
help ensure strong and inclusive growth under structural ment has contributed positively to economy-wide
transformation. In pursuit of these goals, the chapter productivity growth in most developing economies.
seeks answers to the following questions: Moreover, productivity levels in services tend to con-
•• Trends and drivers: How have manufacturing employ- verge to the global frontier (that is, to the productiv-
ment and output shares evolved within countries and ity level in the most productive countries), just as in
at the global level since the 1970s? What were the manufacturing. The rise in the employment share of
mechanisms behind these changes? Which service those service sectors therefore can boost the growth
subsectors have expanded during the past five decades? of aggregate productivity and aid the convergence of
•• Per capita income growth: How diverse are trends in income per worker across countries.
output per worker (the main driver of income per •• While labor earnings in manufacturing are indeed
capita) across the various service subsectors and how somewhat higher and more uniformly distributed
do they compare with those in manufacturing? Have than in services, the main driver of the rise in labor
shifts in employment shares between sectors weighed income inequality in advanced economies since the
on economy-wide labor productivity growth? Does 1980s has been an increase in inequality within
the relative expansion of service sector employment all sectors.

Figure 3.4. Share of Manufacturing in Aggregate A key question for policy is whether the service-led
Employment and Output growth patterns observed in many developing countries
(Percent)
since the early 2000s will continue to hold or whether
they were a byproduct of a temporary boom in global
At the global level, the share of manufacturing in employment and output (real
value added) has changed little since 1970. However, that remarkable global demand. Higher commodity earnings and easy bor-
stability masks pronounced changes in shares at the country level. rowing conditions, for instance, may have temporarily
boosted the demand for nontraded services produced
Employment Nominal value added Real value added with less-skilled labor and facilitated the shift of labor
out of agriculture—where productivity tends to be
30 1. World
relatively low (Diao, McMillan, and Rodrik 2017).
25 In many developing countries, less buoyant growth in
domestic demand in the period ahead may restrain the
20
expansion of nontraded services, while skill shortages
15 may hold back the expansion of the traded, produc-
10 tively dynamic ones. The uncertainty surrounding
future productivity trends and sizable gaps in output
5
per worker among developing countries calls for strong
0 policy efforts to boost productivity in all sectors and
1970 75 80 85 90 95 2000 05 10 15
help channel labor to the most dynamic and productive
activities through skill development and the removal
of barriers to entry and trade in service sectors. In
25 countries where manufacturing jobs are disappearing
20 outright, policymakers ought to facilitate the reskilling
of former manufacturing workers and reduce the costs
15
of their reallocation, while strengthening safety nets to
10 alleviate the adverse consequences of joblessness and job
5
transitions for the workers and their communities.
The rest of the chapter is structured as follows. The
0 next section provides an overview of manufacturing
1970 75 80 85 90 95 2000 05 10 15
trends at the country and global levels and discusses
30 3. Emerging Market and Developing Economies the mechanisms underlying changes in the relative
share of manufacturing in economic activity. It also
25
provides some statistics on the rise in service jobs.
20 The subsequent two sections focus on the differences
15 between manufacturing and services in terms of pro-
ductivity trends and on the level and distribution of
10
labor earnings. The concluding section discusses how
5 policy can ensure strong and inclusive growth under
0
ongoing structural transformation.
1970 75 80 85 90 95 2000 05 10 15
Source: IMF staff calculations. Structural Transformation: Key

Note: The employment (value added) manufacturing share in each panel is
computed as the weighted average share across all economies in the group, with Trends and Drivers
weights given by total employment (GDP in US dollars at market exchange rates)
of each country. Dashed lines in panel 3 denote emerging market and developing The share of manufacturing jobs in global employ-
economies excluding China. ment has been remarkably stable over nearly five
decades (Figure 3.4, panel 1). The sector employs
about the same share of the world workforce now—
about one in seven workers—as it did in the 1970s.
Its share in global output (value added measured at
constant prices) remained broadly stable between the

Figure 3.5. Estimated Trends in Manufacturing Employment and Output Shares, 1960–2015
(Percentage points per year)
While the share of manufacturing jobs in aggregate employment has declined in all advanced economies and many developing economies since 1970, changes in
the output share have been more diverse, and a few economies registered sizable increases in both their manufacturing employment and output shares.
Employment Output
1. Advanced Economies 2. Emerging Market and Developing Economies
CHE HKG
GBR POL
AUT ARG
BEL SGP
AUS CHL
DEU VEN
NLD PHL
NOR NGA
SWE COL
FRA EGY
USA BRA
DNK PAK
FIN ZAF
CAN MEX
PRT IND
IRL KOR
GRC IDN
JPN CHN
ESP THA
ITA MYS
–1.2 –1.0 –0.8 –0.6 –0.4 –0.2 0.0 0.2 0.4 0.6 0.8 –1.2 –1.0 –0.8 –0.6 –0.4 –0.2 0.0 0.2 0.4 0.6 0.8

Note: The figure shows the average annual growth rate in manufacturing employment and real-value-added shares during 1960–2015 (depending on data
availability) for the 20 largest advanced and emerging market and developing economies ranked by 2015 GDP in US dollars at market exchange rates. Data labels
use International Organization for Standardization (ISO) country codes.
1970s and the early 2000s and has been on a slight Among developing economies, the median change
upward trend ever since.3 in manufacturing employment and output shares since
The global stability of manufacturing employment 1970 has been close to zero (Figure 3.6, panels 2 and
and output shares masks pronounced changes at the 4). If China is excluded, the group at the aggregate
country level (Figure 3.5). The share of manufactur- has seen little change in the output and employment
ing in total advanced economy output has remained share of manufacturing (Figure 3.4, panel 3). China,
unchanged since the 1970s, but with diverse (and Indonesia, Korea, Malaysia, and Thailand have seen
offsetting) changes at the individual country level sizable gains in shares since 1970 (Figure 3.5, panel 2),
(Figure 3.4, panel 2; Figure 3.6, panel 3). At the same although in some of these economies the manufactur-
time, almost all advanced economies individually, ing sector still employs a relatively small fraction of the
and the advanced economy group at the aggregate, workforce (for instance, in Indonesia the manufactur-
experienced steady declines in the share of manufacturing employment share has remained about 13 percent
ing jobs in total employment over almost five decades since the mid-1990s; in Thailand it was below 15 per-
(Figure 3.4, panel 2; Figure 3.6, panel 1), underscoring cent in 2010; in China, by contrast, the share was
that labor productivity in manufacturing increased about one-fifth in 2013).
faster than in all the other sectors taken together. For most developing economies, manufacturing
shares peaked around the middle of the sample period:
3The share of manufacturing in global output measured at current output and employment shares increased over the
prices shows a substantial decline over the past five decades as faster 1970s and 1980s in most countries but have declined
productivity gains in manufacturing have lowered the prices of
manufactures relative to those of other products, such as services and in about two-thirds since the 1990s (Figure 3.6, panels
agricultural goods. 2 and 4). Very few countries have experienced rising

Figure 3.6. Cross-Country Distribution of Estimated Trends in Figure 3.7. Peak of Manufacturing Employment Share
Manufacturing Shares, 1970–2015 (Percent)
(Percentage points per year)
Compared with economies that developed earlier, the manufacturing employment
shares of many developing economies have peaked at lower levels and lower
The share of manufacturing jobs in total employment has declined steadily in most income levels.
advanced economies since 1970, while in half of those economies the
manufacturing output share increased until the 1990s. For most developing
economies, manufacturing employment and output shares peaked around the 50
middle of the sample period. Emerging market and developing
economies
Advanced economies
Manufacturing employment share at the peak

Employment 40
0.5 1. Advanced Economies 2. Emerging Market and 1.5

Developing Economies 30
1.0
0.0
0.5
20
0.0
–0.5
–0.5 10
–1.0 –1.0
1970 80 90 2000 1970 80 90 2000
0
6.5 7.0 7.5 8.0 8.5 9.0 9.5 10.0 10.5
Output Log of real GDP per capita (2011 international dollars) at the peak
1.0 3. Advanced Economies 4. Emerging Market and 1.0 Source: IMF staff calculations.
Developing Economies Note: The sample is restricted to economies that show declining manufacturing
employment shares since 1990 or earlier. The x-axis shows the income level
when manufacturing employment shares peaked.
0.5 0.5
levels (Figure 3.7).4 Only a few developing economies

0.0 0.0 in which the manufacturing sector was already relatively
large by 1980—Hong Kong Special Administrative
Region, Korea, Mauritius, Poland, Romania, Singapore,
–0.5 –0.5 and Taiwan Province of China—experienced a peak in
1970 80 90 2000 1970 80 90 2000
the manufacturing employment share higher than in
Source: IMF staff calculations. the average advanced economy. Accordingly, the services
Note: The figure shows the cross-country distribution of the average annual share of employment has started to rise at a lower level
percent change in the employment and real-value-added manufacturing shares.
The horizontal line inside each box represents the median; the upper and lower of per capita income in today’s developing economies
edges of each box show the top and bottom quartiles; and the red markers denote than it has in today’s advanced economies.
the top and bottom deciles. The red diamond denotes the weighted average for
the group. X-axis labels indicate the start of each decade. Boxes for 2000
represent data for 2000–15, whenever available. Drivers of Manufacturing Output and
Employment Shares
A spectrum of explanations can help reconcile the
manufacturing employment and output shares in the
stable manufacturing output and employment shares at
2000s (for instance, Bangladesh, China, Ethiopia, and
the global level with diverse changes across countries.
Malawi). Moreover, many of the developing economies
At one extreme, shifts in manufacturing output and
with declining manufacturing shares never experienced
employment shares could reflect zero-sum reallocations
strong expansion of the shares to begin with, unlike
in supply, with manufacturing production moving
most of the economies that developed earlier. As a result,
from locations where production costs are higher to
compared with those of earlier developers, the manufac-
turing employment shares of many developing econo- 4Dasgupta and Singh (2006) and Rodrik (2016) call this phenom-
mies have typically peaked at lower shares and income enon “premature deindustrialization.”

lower-cost economies (mostly developing economies) Figure 3.8. Share of Manufacturing in Final Consumption
that have become increasingly integrated into the global versus Income per Capita, 1980–2011
(Percent)
trading system. At the other extreme, changes in output
and employment shares could reflect trends in incomes
As real incomes rise from low levels, the share of manufactured goods in
and demand. Demand for manufactures increases faster consumption increases. As incomes rise further, however, the proportion spent on
than demand for food and services in the earlier stages services grows at the expense of manufactures.
of a country’s development. In the later stages, the
50
demand for services expands the fastest, but the decline
in the relative price of manufactures could dampen the
relative shift away from their consumption as income
40
grows. Under a demand-based explanation, the global
share of manufacturing output would initially be stable
or even increase (as has been the case since 2000) as 30
global incomes converge, with fast-growing developing
economies consuming relatively more manufactured
goods while the slower-growing advanced economies 20
consume less. The global share of manufacturing output
would be expected to decline in the long term as all
economies increasingly need more services. 10
In reality, the explanation for a stable global

picture amid country variations is probably some-
0
where between these two interpretations. The global 4 5 6 7 8 9 10 11 12 13 14
performance likely reflects both some reallocation of Log of real GDP per capita (2011 international dollars)
manufacturing production toward countries with lower
Sources: World Bank International Comparison Program (ICP) database; and IMF
production costs and country variations in the demand staff calculations.
for manufacturing. Note: The black line shows the estimated relationship between the share of
manufacturing in final consumption and income per capita based on a quadratic
Each of the potential drivers of changes in manu- estimation using country fixed effects. Final consumption expenditure shares are
facturing output and employment shares—variations based on ICP data (1980, 1985, 1996, 2005, and 2011 vintages) and include
in demand for manufactures as incomes rise and their consumption by households and the government. Countries with a population less
than 1 million in 2014 are excluded.
relative price falls, and cross-border integration—has
been studied widely in the literature. Studies dating
back to the 19th century (Engel 1895) as well as ciency gains in production imply that a given increase
recent work (Kongsamut, Rebelo, and Xie 2001; Buera in output requires smaller increases in labor and other
and Kaboski 2009, 2012; Herrendorf, Rogerson, and inputs over time; a relatively faster rise of output per
Valentinyi 2013; Boppart 2014) emphasize changing worker in manufacturing has thus come with slower
consumption patterns as real income per capita grows. employment growth in manufacturing than in other sec-
The final consumption share of manufactured goods tors. The faster rise of productivity has also led unit pro-
exhibits a hump-shaped relationship to real income per duction costs in the manufacturing sector to fall more
capita (Figure 3.8). As individuals’ real income rises rapidly than in other sectors, lowering the relative price
from low levels, the share they spend on food declines of manufactures in the vast majority of countries over
(Engel’s law), and the share they spend on manufac- the past five decades (Figure 3.9). The greater affordabil-
tured products rises. As incomes grow further, however, ity of manufactured goods has tempered the decline in
the proportion spent on services rises at the expense of the relative demand for manufactures driven by higher
manufactures. incomes and shifts in preferences, but not to an extent
A second factor is linked to the faster rise in produc- that prevented the productivity-driven decline in the
tive efficiency in manufacturing than in other sectors share of manufacturing in employment (Figure 3.10).5
(Kuznets 1966; Baumol 1967; Ngai and Pissarides
5The decline in the relative price of manufactures affects consumer
2007), which has lowered the amount of labor needed
behavior in two ways. First, it raises disposable incomes, allowing
to produce a given amount of manufacturing output consumers to spend more on both goods and services (an income
and has made manufactures more affordable. Faster effi- effect). Second, consumers may spend relatively more on the

Figure 3.9. Relative Price of Manufacturing, 1970–2015 Figure 3.10. Estimated Change in Manufacturing Shares and
Relative Prices, 1960–2015
The relative price of manufactures declined in most economies over the past five (Percentage points per year)
decades, reflecting faster productivity gains in manufacturing than in other sectors.
The decline in the relative price of manufactures boosted the relative demand for
0.8 1. Total Factor Productivity in Manufacturing goods, but not to an extent that would prevent the shift of labor from
and Relative Price manufacturing to services.
Log relative price (demeaned)
0.6
0.4 Emerging market and developing economies Advanced economies
0.2 1.0 1. Manufacturing Employment Share

0.0
0.5
Employment share
–0.2
0.0
–0.4
–0.4 –0.3 –0.2 –0.1 0.0 0.1 0.2 0.3 0.4
Log total factor productivity (demeaned) –0.5
–1.0
4 2. Change in Log Relative Price
3 (Percentage points per year)
–1.5
2 –4 –3 –2 –1 0 1 2 3 4
Log relative price
1
0
1.0 2. Manufacturing Real-Value-Added Share
–1
–2
Real-value-added share
–3
0.5
–4
–5
1970 80 90 2000 1970 80 90 2000
Advanced economies Emerging market and 0.0
developing economies
Sources: Groningen Growth and Development Centre database; United Nations

database; World KLEMS database; and IMF staff calculations. –0.5
–4 –3 –2 –1 0 1 2 3 4
Note: Panel 1 shows the logarithm of the relative price of manufacturing value
Log relative price
added (relative to that of the aggregate economy) in each country and year against
the logarithm of the total factor productivity in manufacturing. Both variables are
expressed as deviations from the country average across the sample period. Panel Sources: Groningen Growth and Development Centre database; United Nations
2 shows the cross-country distribution of the average annual change in the database; and IMF staff calculations.
logarithm of the relative price of manufacturing. The horizontal line inside each Note: The relative price denotes the price of manufacturing value added relative to
box represents the median; the upper and lower edges of each box show the top that of the whole economy.
and bottom quartiles; and the red markers denote the top and bottom deciles.
X-axis labels indicate the start of each decade. Boxes for 2000 represent data for
2000–15, whenever available.
Swiecki 2017; Wood 2017). Declining trade costs
affect the patterns of specialization across countries,
increasing the share of manufacturing in output and
A third factor—trade and financial integration—can employment in countries that have comparative advan-
give rise to new influences on manufacturing shares. tage in that sector and lowering them in countries that
International trade allows the sectoral composition do not.6 Increased access to foreign finance that lowers
of domestic demand to differ from that of domestic
supply as goods and some services can be traded across 6Trade also allows specialization within manufacturing: the
borders (Matsuyama 2009; Uy, Yi, and Zhang 2013; manufacture of products requiring primarily low-skilled labor
would shift to countries with an abundance of such workers, while
the production of other types of manufacturing would shift to
now-more-affordable manufactured goods (a substitution effect). countries with an abundance of highly skilled workers and lower
However, the second channel is not strong: consumers do not sub- user costs of capital. The share of manufacturing in output may
stitute services with manufactured goods to a great extent when the remain unchanged in both groups, while the manufacturing share of
latter get cheaper (Buera and Kaboski 2009; Herrendorf, Rogerson, employment would rise where low-skilled labor was most abundant
and Valentinyi 2013; Lawrence 2017). and decline elsewhere.

the cost of capital can accentuate the specialization ing shares (by about 5 percent of global spending on
patterns in capital-scarce economies, especially where manufactures) as production has shifted to developing
financial frictions and credit rationing are more prev- economies. In developing economies, the increase in
alent. The reallocation of manufacturing to countries manufacturing gross output has exceeded the rise in
with comparative advantage also lowers the relative final expenditures on manufactures.
price of manufactures globally, raising the demand for The difference between changes in manufacturing
manufactures. gross output and expenditure shares (that is, the extent
Trade and financial integration also speed up the of reallocation) in the 1995–2011 period has been
adoption of technological advancements and their large for some countries (Figure 3.11, panels 2 and
diffusion across borders (Chapter 4). Faster diffusion 3).10 Among large advanced economies, gross out-
of innovations allows countries to converge to the put declined more than final spending in France (by
productivity frontier more quickly and shortens the 4 percent of GDP), the United States (3 percent), and
period in which an increasing share of labor needs to Japan (1.5 percent). The difference between the change
be employed in the manufacturing sector (the so-called in gross output and final spending is also negative in
industrialization phase of development).7 Faster diffu- several developing economies in the sample, includ-
sion also raises global competition among producers ing India, Mexico, Russia, and Turkey. In contrast, in
and puts downward pressure on manufacturing prices China, Germany, Ireland, and Korea, the rise in the
everywhere, which also tends to raise the final demand manufacturing output share is larger than the rise in
for manufactures.8 the expenditure share. The difference in the case of
Global sectoral expenditure and production data can China (about 10 percent of GDP) stands out, as it
give a sense of the extent of production reallocation represents about 2½ percent of global spending on
over the past two decades. Figure 3.11 compares the manufactures. Not all of the reallocations of gross
change in spending on manufactures with changes in output have been met by equal shifts in domestic man-
the domestic gross output of manufactures.9 Devel- ufacturing value added, however. Some of the realloca-
oping economies’ shares in both global gross output tion has fallen on the service-value-added component
and final expenditures of manufactures rose between of manufacturing output as well (Box 3.1).
1995 and 2011, while those of advanced economies The broadly parallel movements of global manu-
fell (Figure 3.11, panel 1). But the changes in gross facturing output and employment shares might seem
output shares have not matched the changes in expen- puzzling given the relatively fast pace of productivity
diture shares one-for-one. In advanced economies, growth in the sector, which would be expected to
gross output shares have declined more than spend- drive a growing wedge between the global output and
employment shares of manufacturing over time (as has
7Huneeus and Rogerson (2016) argue that productivity growth happened in virtually all advanced economies and most
in manufacturing (relative to other sectors) may be faster for current developing economies). The explanation is a gradual
developing economies than for earlier developers due to catch-up
effects, helping to explain why manufacturing employment shares are
shift in the composition of global manufacturing
peaking at lower levels in developing economies. employment toward developing economies, where pro-
8Rodrik (2016) argues that developing economies “imported”
ductivity tends to be lower but the demand for manu-
deindustrialization as they opened to trade (including those that may
factures higher and the unit production costs lower.11
not have experienced much technological progress), by becoming
exposed to the downward pressure on the relative price of manufac- The bilateral relationships between manufacturing
tures originating from productivity gains in advanced economies. output and employment shares and their possible
9A finished manufactured product embeds value added by both
drivers are helpful in gaining a sense of the mechanisms
the domestic and foreign manufacturing and nonmanufacturing
sectors. A vehicle purchased by a consumer, for instance, embeds underlying structural transformation. However, empiri-
domestically and foreign-produced manufactured parts as well as cally estimating the relative importance of each of these
domestically and foreign-produced engineering and marketing
services. The domestic gross output of the manufacturing sector is
the sum of all the domestically produced content of its final output. 10The difference between changes in manufacturing gross output
Spending on manufactures in a given country equals the sum of the and spending on manufactured goods over time for individual coun-
gross output of the domestic manufacturing sector, net imports of tries can reflect a faster expansion of spending and not necessarily a
finished manufactured goods, and net imports of intermediate inputs decline in gross manufacturing output.
by the manufacturing sector. Gross output and spending data used 11Felipe and Mehta (2016) also document the impact of changes
for this exercise are from the World Input-Output Database, which in the country composition of manufacturing activity on the shares
covers 1995–2011. See Annex 3.2 for details. of manufacturing in output and employment at the global level.

Figure 3.11. Manufacturing Gross Output and Final mechanisms for a broad set of countries is very chal-
Expenditure on Manufacturing Goods, 1995–2011 lenging. The complexity of the underlying mechanisms
aside, only the ex post outcomes of the causal drivers—
Developing economies exhibited a larger expansion in the share of global gross production costs and relative prices, trade intensity, and
manufacturing output than in the share of global spending on manufactures,
which suggests some production reallocation toward those economies as a group. income levels—are observed, not the exogenous forces
But there is considerable variation across economies. driving structural change. The recent literature has
therefore largely sought to explain structural transforma-
90 1. Share of Global Output and Final Expenditure
(Percent) tion patterns using global general equilibrium models,
80
70
Gross output typically focusing on one mechanism at a time.12
Final expenditure
60
50
The Rise of Services
40
30 A striking feature of structural transformation is the
20 expansion of the service sector. The share of services in
10 global employment has increased by about 16 percent-
0 age points since the 1970s. While the increase in the
1995 2011 1995 2011
AEs EMDEs share of service jobs in overall employment is largely
the flip side of declining manufacturing employment
25 2. Change in Manufacturing Output versus Change in in advanced economies (Figure 3.2, panel 1), in devel-
Manufacturing Expenditure: AEs
20 oping economies it mostly reflects a shift of labor from
(Percent of GDP)
15 agriculture (Figure 3.2, panel 2).
10 Employment in nonmarket services (government,
5 education, health) expanded rapidly in the group of
0 advanced economies, contributing about one-third
–5 of the overall expansion in service employment since
–10 1970 (Figure 3.12).13 Within market services—which
–15 contributed the remaining two-thirds of the expansion
IRL
LUX
NLD
DEU
AUT
SWE
BEL
ESP
JPN
AUS
GBR
FIN
CAN
PRT
GRC
ITA
USA
FRA
DNK
in the share of services—financial intermediation, real

estate, and business activity services were the subsec-
25 3. Change in Manufacturing Output versus Change in tors with the fastest growth. In developing economies,
Manufacturing Expenditure: EMDEs employment in market services contributed the lion’s
20
(Percent of GDP)
15 share of the overall increase in services employment,
10 with particularly large expansions in wholesale and
5 retail trade, and hotels and restaurants.
0 Manufacturing output increasingly embeds inputs
–5 from services—the so-called servicification of manu-
–10 facturing (National Board of Trade of Sweden 2010;
–15 Baldwin 2016; Hallward‑Driemeier and Nayyar 2017).
Using recently available data on global input-output
CZE
CHN
KOR
HUN
TWN
EST
IDN
SVN
POL
MEX
RUS
IND
MLT
ROU
CYP
LTU
BGR
TUR
BRA
LVA
SVK
Sources: World Input-Output Database; and IMF staff calculations. 12Herrendorf, Rogerson, and Valentinyi (2014) survey recent
Note: Panel 1 shows the share of each group of economies in global manufacturing
gross output (bars) and in global final expenditure on manufactures (squares). contributions to the literature on structural transformation.
13The classification of service industries into market and non-
Panels 2 and 3 show the difference between the change in domestic gross
manufacturing output and in final expenditure on manufactures between 1995 and market services follows the guidelines of the System of National
2011, expressed as a share of average GDP during the period. Domestic gross Accounts. Market services consists of wholesale and retail trade
manufacturing output includes value added from all domestic sectors embedded and repair of goods; hotels and accommodation; transport, storage,
in manufacturing production (foreign value added is excluded). See Annex 3.2 for and communications; financial intermediation; real estate, renting,
details. Data labels use International Organization for Standardization (ISO) country and business activities; other community and personal activities;
codes. AEs = advanced economies; EMDEs = emerging market and developing and activities of private households. Nonmarket services consist of
economies.
government (public administration, defense, and social security);
education; and health. See Annex 3.1 for a list of sectors, individual
industries, and abbreviations.

linkages, analysis in Box 3.1 documents that the share Figure 3.12. Change in Services Employment Share,
of service inputs in manufacturing production has 1970–2015
(Cumulative change, percentage points)
risen in most countries over the past two decades.
However, the share of manufactures in final expendi-
Market services account for about two-thirds of the overall expansion in service
ture has been declining in most countries, and service employment since 1970 in advanced economies, and more than 80 percent in
inputs into manufactures still account for a small developing economies.
fraction of overall value added in the service sector, so
25 1. Market and Nonmarket Service Industries
the servicification of manufacturing has not contrib-
uted meaningfully to the increased share of services in 20
overall output over the past two decades. 15
The changing service content of manufacturing out-
put poses challenges to the measurement of structural 10
change. Available statistics measure only imperfectly 5

changes in the weight of different tasks and activi-
0
ties in the economy. The increasing fragmentation of
manufacturing production implies that some activities –5
Market
services
Nonmarket
services
Market
services
Nonmarket
services
formerly carried out within manufacturing firms (such
as marketing, legal services, logistics) are unbundled
and outsourced. The reclassification of these activities as Advanced economies Emerging market and
services in official statistics could overstate the extent of developing economies
structural transformation.14 At the same time, firms in
the manufacturing sector are increasingly producing and 14 2. Market Service Industries
12
selling auxiliary services that are bundled with finished
10
goods; including such service activities in manufacturing 8
production may understate the true extent of structural 6
transformation.15 Available data do not permit reliable 4
quantification of the relative magnitude of these two 2
0
opposing effects, and partial evidence from existing
–2
studies suggests that their net effect on measures of sec- –4
toral employment and output shares is ambiguous. –6
Trade &
accommodation
Transport &
communications
Financial &
business
Community &
households
Trade &
accommodation
Transport &
communications
Financial &
business
Community &
households
Growth and Development beyond
Manufacturing
Advanced economies Emerging market and
Manufacturing has historically been considered developing economies
more technologically progressive than the service sec-
tor, so the reallocation of production from the former Source: IMF staff calculations.
to the latter has generally raised concern regarding the Note: The figure shows the cross-country distribution of the cumulative change in
the employment share of individual service industries between the average in the
growth of aggregate productivity—the most important 1970s and the average during 2000–15. The horizontal line inside each box
determinant of a country’s standard of living (Baumol represents the median; the upper and lower edges of each box show the top and
bottom quartiles; and the red markers denote the top and bottom deciles.
1967; Kaldor 1967).16 The countries that achieved Nonmarket services consists of government, education, and health. All other
service industries are market services. See Annex 3.1 for a list of sectors,
14Berlingieri individual industries, and abbreviations.
(2014); and Bernard, Smeets, and Warzynski (2017).
15Pilat and Wölfl (2005); National Board of Trade of Sweden
(2010); and Crozet and Milet (2017).
16Many of the key attributes of the manufacturing sector—relatively
high levels of innovation, foreign direct investment (facilitating

technological diffusion), economies of scale, high degrees of tradability,
and strong interlinkages with other sectors—have traditionally
been considered critical to long-term growth and development.
Hallward‑Driemeier and Nayyar (2017) note that these characteristics
vary considerably across manufacturing subsectors and over time.

substantial income convergence toward more devel- In seeking to shed light on whether nonmanufac-
oped economies since the 1960s typically experienced turing sectors can increasingly drive growth and help
strong increases in manufacturing employment and narrow income gaps across countries, the analysis fol-
exports (Jones and Olken 2005; Johnson, Ostry, and lows McMillan and Rodrik (2011) and Rodrik (2013)
Subramanian 2007). The observation that the indus- and focuses on the growth of labor productivity as a
trialization phase among developing economies is not normative benchmark. Labor productivity is defined
as vigorous as it was in countries that developed earlier as output at constant prices divided by the number
has thus led some to doubt their ability to narrow of workers in the economy or a given sector. When
income gaps with advanced economies. Rodrik (2013, cross-country comparisons of sectoral productivity levels
2016) provides compelling evidence in favor of these are involved, output is expressed in international dollars
concerns, documenting that labor productivity in man- using sector-specific purchasing power parity (PPP),
ufacturing in a sample of 130 economies has tended to which helps ensure that the comparisons are not affected
converge to the frontier, regardless of policies, institu- by price differences across countries. The analysis also
tions, and other country characteristics (unconditional provides some evidence of differences in total factor
convergence), whereas labor productivity for the overall productivity (TFP) growth rates by disaggregated sector,
economy (and hence the nonmanufacturing sector) has with TFP defined as the output for a given combina-
not.17 This unique attribute implies a pivotal role for tion of labor and capital inputs, a measure of overall
manufacturing in the development process; a stag- efficiency gains that (unlike labor productivity) does
nant manufacturing sector could present a daunting not vary with the amount of capital per worker but is
obstacle for developing economies in catching up with available for a relatively limited set of countries.19
advanced economy per capita income levels. Consistent As a final word of caution regarding this analysis,
with this observation, McMillan and Rodrik (2011) productivity data by disaggregated sector are avail-
document that structural transformation between 1990 able only for a subset of the Rodrik (2013) database.
and 2005 tended to be growth-reducing in developing Wherever possible, the analysis uses a variety of data
countries that did not experience increases in the share sets to ascertain robustness. At the same time, the data
of manufacturing employment. sets used in the chapter include sector-specific PPPs
Nonmanufacturing activities form a very diverse that facilitate the comparison of sectoral productivity
group, however.18 Productivity dynamics vary substan- across countries, which was not possible in the Rodrik
tially within services, and shifts of employment shares (2013) study.
within the nonmanufacturing sector have been sizable, The road map for the rest of the subsection is as
especially in developing economies (where activity has follows. The discussion next turns to evidence on
shifted from agriculture to services). These observations productivity levels and growth rates across disag-
highlight the value of assessing the productivity effects gregated service and manufacturing subsectors. The
of structural transformation using data at a more subsequent section examines whether shifts in employ-
disaggregated sectoral level than for manufacturing ment shares between sectors have tended to benefit or
and the rest of the economy. If productivity converges harm aggregate productivity. The final section looks at
toward the international frontier for some types of whether productivity convergence is unique to man-
services, and employment shares shift toward these ufacturing or whether it is a feature of some service
subsectors, then structural transformation that bypasses sectors as well.
manufacturing need not hinder economy-wide produc-
tivity growth.
Productivity in Services: Lagging Behind?
17Convergence requires productivity to grow faster in countries Many studies have stressed that productivity growth
where its initial level is relatively low. among the diverse set of market and nonmarket indus-
18Productivity dynamics vary substantially within manufacturing
tries is likewise diverse, ranging from the slowest to the
activities, as well as across firms within narrowly defined manufac-
turing and nonmanufacturing industries. Hsieh and Klenow (2009)
document that that the “misallocation” of capital and labor across
manufacturing firms in China and India hinder economy-wide 19In addition to being available for a small set of countries on a
total factor productivity. Dias, Marques, and Richmond (2016) find sectoral basis, TFP measures (unlike labor productivity measures)
that the extent of resource misallocation in Portugal is larger in the do not lend themselves to straightforward decompositions of
service sector than in manufacturing. within-sector and structural transformation effects.

fastest in the economy.20 Some service industries at the Figure 3.13. Difference in Labor Productivity Growth
upper end of productivity growth are among the most between Manufacturing and Services before and after 2000
(Percentage points)
intensive users of information and communication
technologies (Stiroh 2002). Recent advances in those
The difference between productivity growth in manufacturing and services has
technologies are likely to have played an important shrunk since 2000 in most economies. The average productivity growth in the
role in boosting the productivity of the sectors that use services sector has recently exceeded that of manufacturing in many developing
economies.
them (Bosworth and Triplett 2003, 2007; Jorgenson
and Timmer 2011).21 10 Emerging market and
A first look at labor productivity by aggregated sector developing economies
reveals that the manufacturing sector as a whole typi- 8 Advanced economies
cally sees faster productivity gains than the service sector
6
(most observations of the productivity growth differ- SGP
PER
CRI KOR
ential between manufacturing and services are positive 4 CHL
MUS
EGY BOL
in Figure 3.13, both before and after 2000). However, THA
After 2000
2 MYS BRA
the differential has shrunk since 2000 in most countries ZAF MWI MEX
HKG
GHA PHL COL
(that is, most observations lie below the 45-degree line 0
MAR TWN ARG BWA
in the same figure). Moreover, average productivity
–2 IND KENTZA CHN
growth in services in many developing economies, VEN IDN
including China, India, and some in sub-Saharan Africa, –4 SEN
has recently exceeded that of manufacturing. ETH
–6
Disaggregated labor productivity data show that ZMB NGA
some service industries register as fast growth in –8
output per worker as the top-performing manufac- –8 –6 –4 –2 0 2 4 6 8 10
Before 2000
turing industries (Figure 3.14). The distribution of
labor productivity growth in manufacturing industries Sources: Groningen Growth and Development Centre database; World KLEMS
over the past five decades is somewhat to the right database; and IMF staff calculations.
Note: The figure shows the average annual growth differential between labor
of that of service industries. However, in a sample of productivity growth in manufacturing and services during 2000–15 on the y-axis
19 advanced and 43 developing economies during and during 1965–99 on the x-axis. Observations below the diagonal line denote a
1965–2010, labor productivity growth in some broad decline in the productivity growth differential. Labor productivity is defined as
value added per worker at constant national prices. Data labels use International
service industries is comparable to productivity growth Organization for Standardization (ISO) country codes.
in manufacturing as a whole (Figure 3.14, panel 1). A
similar picture appears from data for 13 manufactur-
ing industries and 13 service industries available for a panel 3).22 The main takeaway is that there is a sizable
smaller number of economies (Figure 3.14, panel 2). overlap between productivity growth among the service
The data for the United States, which is available at and the manufacturing subsectors.
a finer disaggregation level (20 manufacturing indus- Finally, the levels of labor productivity for a sample
tries and 39 service industries), shows an even larger of 19 advanced and 43 developing economies in
degree of overlap between labor productivity growth 2005 suggest that, within each country, workers in
in manufacturing and service subsectors (Figure 3.14, goods-producing sectors are not necessarily more
20See, for instance, Baumol, Blackman, and Wolff (1985); Jor- productive than service sector workers (Figure 3.15).
genson and Timmer (2011); Verma (2012); Young (2014); Duarte More precisely, labor productivity in two out of four
and Restuccia (2017); and Duernecker, Herrendorf, and Valentinyi market service industries (transport and communica-
(2017). Productivity in service industries is particularly difficult to tions; financial intermediation and business activities)
measure (Triplett and Bosworth 2000), but previous work suggests
that correction for mismeasurement of output in services would is comparable to, or higher than, in manufacturing.
likely lead to higher productivity growth in services than recorded in The finding of strong productivity growth among
official data (Gordon 1996). services is good news for developing economies where
21Communication and digital technologies may help increase
productivity growth in some service industries by facilitating

the share of manufacturing in overall activity has
international trade in services (Heuser and Mattoo 2017; Loungani
and others 2017; Box 3.2), which heightens competition, facilitates 22TFP data also reveal substantial overlap between productivity
cross-border knowledge spillovers, and enhances economy of scale. growth in manufacturing and service subsectors (Annex Figure 3.3.1).

Figure 3.14. Distribution of Labor Productivity Growth of Figure 3.15. Sectoral Labor Productivity, 2005
Individual Industries (Difference with respect to economy-wide labor productivity
(Kernel density) percentage points)
There is a sizable overlap between labor productivity growth among the service Within each economy, labor productivity is not necessarily lower in service
and the manufacturing subsectors, with some service industries exhibiting industries than in goods-producing sectors. Labor productivity in two out of four
productivity growth rates as high as the top-performing manufacturing industries. market service industries is higher than the economy-wide average, and
comparable or higher than in manufacturing.
Manufacturing Services 200
0.25 1. Aggregate Manufacturing and Broad Service Industries 150

(62 economies, 1965–2010)
0.20
100
0.15
50
0.10
0
0.05
–50
0.00
–10 –8 –6 –4 –2 0 2 4 6 8 10 12 14 16 –100
0.25 2. Disaggregated Manufacturing and Service Industries –150

(31 economies, 1970–2010)
0.20 –200
Agriculture
Manufacturing
Trade &
accommodation
Transport &
communications
Financial &
business
Community &
households
Government
0.15
0.10
0.05
Sources: Groningen Growth and Development Centre database; World KLEMS
0.00 database; and IMF staff calculations.
–10 –8 –6 –4 –2 0 2 4 6 8 10 12 14 16 Note: The figure shows the cross-country distribution across 62 economies of the
percentage difference between labor productivity in each sector (value added per
worker) and aggregate labor productivity in 2005. The horizontal line inside each
0.25 3. Disaggregated Manufacturing and Service Industries box represents the median; the upper and lower edges of each box show the top
(United States, 1947–2010) and bottom quartiles; and the red markers denote the top and bottom deciles.
0.20
0.15
leveled off. However, these productively dynamic
0.10 service industries may not necessarily account for a
large share of employment and thus may play a limited
0.05 role in driving aggregate productivity. Moreover, their
expansion in the future may be constrained by the
0.00
–10 –8 –6 –4 –2 0 2 4 6 8 10 12 14 16 availability of skilled workers or the pace of expansion
in domestic demand. Ancillary evidence, however,
Sources: Groningen Growth and Development Centre database; World KLEMS suggests that these factors may not necessarily act
database; and IMF staff calculations.
Note: The figure shows the distribution of average labor productivity growth per as binding impediments to service-led productivity
decade in individual manufacturing and service industries (expressed as deviations growth in the short term.
from the average labor productivity growth across sectors in each country and
decade). Panel 1 is based on data for aggregate manufacturing and 5 service Service industries with favorable productivity dynam-
industries in 19 advanced economies and 43 emerging market and developing ics account for a meaningful share of employment and
economies. Panel 2 is based on data for 13 manufacturing and 13 service
industries in 19 advanced economies and 12 emerging market and developing
can play a key role in driving aggregate productivity
economies. Panel 3 is based on data for 20 manufacturing and 39 service growth. For instance, the service industries that rank
industries in the United States (Jorgenson, Ho, and Samuels 2012). See Annex 3.3 in the top third of the labor-productivity growth
for details.
distribution (Figure 3.14, panel 2) during 2000–10

accounted, on average, for almost half of total employ- labor productivity growth can be achieved in two
ment in market services, about 30 percent of total ways.25 First, productivity can increase within sec-
service employment, and close to 20 percent of overall tors through an increase in capital per worker, higher
employment.23 Some service industries simultaneously total-factor productivity, or a reallocation of labor
registered above-average labor productivity growth and and capital toward the more productive firms within
rising employment shares during the 2000s, thanks the sectors. The so-called “within” component of the
to strong demand (for example, financial intermedia- decomposition captures the contribution of productiv-
tion in Hungary, Russia, and Slovenia; postal services ity growth within sectors to economy-wide productiv-
and telecommunications in Korea and Lithuania; and ity growth. Second, economy-wide labor productivity
wholesale trade in the Czech Republic and Latvia). And can increase if workers shift from sectors where their
although employment in some of the tradable service productivity is low to sectors where it is high. This
industries—such as financial intermediation—are skill second part—the so-called “between” or “structural
intensive, the skill intensity of other service industries change” component—captures the effect of labor real-
with relatively high labor-productivity growth, includ- locations across sectors with varying productivity levels.
ing telecommunications, is comparable to that of When employment shares increase in high-productiv-
manufacturing (Annex Figure 3.3.3). ity sectors, structural change will be beneficial for
The growth of nontraded service sectors could economy-wide labor productivity growth.
indeed be constrained by the pace of expansion in The results of the decomposition using data span-
domestic demand. Notwithstanding the increased ning 10 sectors in a sample of 62 economies covering
tradability of services in the recent past, especially 2000–10 confirm that productivity gains within sectors
among highly productive services, such as telecom- can account for the bulk of aggregate labor productivity
munications, financial intermediation, and business growth in both advanced and developing economies
activities, international trade in services is still rather (Figure 3.16, panel 1). Importantly, the results also
limited (Box 3.2). That said, recent studies suggest that show that structural change has not exerted a drag on
the domestic demand for services exhibiting strong aggregate productivity. In advanced economies, where
productivity growth may increase in relative terms employment shares have steadily shifted from manufac-
over time as they become more affordable.24 And, turing primarily to nonmarket service industries (Fig-
given that barriers to international trade are higher ure 3.16, panel 2) and intersectoral productivity gaps
for services than for goods (Miroudot, Sauvage, and are relatively small, the contribution of structural change
Shepherd 2013), there is potential for service exports has been negative but quantitatively negligible. This
to gather speed if appropriate policy actions are taken. finding is to be expected: as documented in McMillan
and Rodrik (2011), gaps between productivity levels in
sectors narrow over time as countries develop.
Has Structural Transformation Weighed on Aggregate The contribution of structural change to aggregate
Productivity Growth? productivity growth in developing economies has been
To gauge the impact of shifts in employment shares positive in all regions since 2000—a period when labor
across disaggregated sectors, this section follows a has shifted from low-productivity agriculture to man-
decomposition analysis put forth by McMillan and ufacturing in some cases, and to market services more
Rodrik (2011) and Diao, McMillan, and Rodrik prominently (Figure 3.16, panel 2). Consistent with
(2017). The approach recognizes that economy-wide the findings in McMillan, Rodrik, and Verduzco-Gallo
(2014) and Diao, McMillan, and Rodrik (2017),
23The service industries that rank in the top third of the
the analysis shows that the positive contribution of
labor-productivity growth distribution during 2000–10 are postal structural change since 2000 has been particularly large
services and telecommunications, financial intermediation, and in sub-Saharan Africa. This is explained by the strong
wholesale and retail trade (Annex Figure 3.3.2).
24Duernecker, Herrendorf, and Valentinyi (2017) find that the labor shifts out of agriculture in the region during this
elasticity of substitution between services with high and low pro- period, combined with still-large productivity shortfalls
ductivity growth in the United States is larger than 1. This degree in agriculture relative to manufacturing and market
of substitutability implies that the demand for services with high
productivity growth and declining relative prices can substitute for
services with lower productivity growth, leading to an expansion of
their employment share despite fast productivity gains. 25See Annex 3.3 for details.

Figure 3.16. Structural Transformation and Aggregate Labor appears to have been driven by the particularly strong
Productivity Growth, 2000–10 growth of aggregate demand in the region (supported
by external transfers or higher commodity-based
The contribution of structural change in developing economies has been positive
since 2000, when labor predominantly shifted from agriculture to market services.
revenues), suggesting that overall productivity growth
may slow down as demand loses momentum, unless
8 1. Contribution to Aggregate Labor Productivity Growth productivity growth picks up within sectors.
7 (Percent) That said, the growth of productivity within sec-
Within-sector growth
6 Structural change tors differs widely and accounts for the bulk of the
5 variation in overall productivity growth across regions
4 (Figure 3.16, panel 1). The contribution to aggregate
3 productivity growth of both manufacturing and market
2
services in 2000–10 was much larger in Asia (1.9 per-
1
cent and 2.1 percent a year, respectively) than in
0
sub-Saharan Africa (almost nil in manufacturing and
–1
AEs All EMDEs Emerging Emerging Emerging Emerging 0.8 percent in market services) and in Latin America
Asia EUR LAC SSA
(about 0.2 percent in each). Therefore, the challenge
for many developing economies is not only to facilitate
Agriculture Manufacturing Market services the reallocation of labor to high-productivity sectors,
Nonmarket services Other
but also to raise productivity growth in all sectors.
1.4 2. Change in Sectoral Employment Shares
1.2 (Percentage points per year)
1.0 Implications for Income Gaps across Countries
0.8
0.6 Labor shifts into sectors with relatively high and
0.4 fast-growing productivity (by the standards of the
0.2
0.0 country) may not be enough to narrow the gap
–0.2 vis-à-vis the frontier if productivity grows even faster at
–0.4
–0.6 the frontier. Rodrik’s (2013) finding of unconditional
–0.8 convergence in manufacturing suggests that the lack
–1.0
AEs All EMDEs Emerging Emerging Emerging Emerging of income convergence at the country level might be
Asia EUR LAC SSA a result of the relatively small share of manufacturing
employment in developing economies and that con-
vergence would hasten if the share of manufacturing
Note: Panel 1 shows the contribution of within-sector labor productivity growth employment could be raised.
and changes in sectoral employment shares to the (weighted average) annual Even if the productivity of the nonmanufacturing
aggregate labor productivity growth in each group of economies (economies are
weighted by total employment) based on data for 10 broad sectors. See Annex 3.3 sector as a whole does not converge to the world econ-
for details. Panel 2 shows the change in sectoral employment shares for five omy’s highest levels, some of its subsectors might. This
groups of sectors. Nonmarket services consists of government; education; and
health. All other service industries are market services. See Annex 3.1 for a list of section tests this proposition.27
sectors, individual industries, and abbreviations. AEs = advanced economies; The empirical approach, following Bernard and Jones
EMDEs = emerging market and developing economies; EUR = Europe;
LAC = Latin America and the Caribbean; SSA = sub-Saharan Africa. (1996) and Sorensen (2001), tests whether productivity
growth in a sector is faster when the initial gap between
its productivity level and productivity at the technolog-
services (Annex Figure 3.3.4).26 One concern, stressed ical frontier is larger. This would imply that the greater
by Diao, McMillan, and Rodrik (2017), however, is the shortfall, the faster the convergence to the frontier
that the recent growth-enhancing structural change
26In earlier work, McMillan and Rodrik (2011) find that struc- 27Testing unconditional productivity convergence for individual
tural change contributed negatively to economy-wide productivity sectors is challenging because it requires data on output per worker
growth in sub-Saharan Africa over 1990–2005, a period when the at comparable international prices across countries. This section uses
share of agriculture in employment declined by only 0.1 percentage new data on sector-specific PPP from the Groningen Growth and
point. In contrast, the share of agriculture within overall employ- Development Centre database. Nonmarket service industries are
ment declined by 5 percentage points over 2000–10 and the bulk of excluded from the analysis because of lack of reliable sectoral PPP
this change occurred during the second half of the period. data. See Annex 3.3 for details.

Figure 3.17. Estimation Results, Beta-Convergence

(Coefficient)
There is strong evidence of unconditional productivity convergence to the global frontier (that is, to the productivity level in the most productive countries) for
manufacturing as well as for several service industries.
1. Extended Sample (nine sectors) 2. Reduced Sample (26 sectors)

Utilities
Utilities Mining
Agriculture
Construction
Construction Basic metal
Other mineral
Electrical equipment
Mining Machinery
Wood
Chemicals
Agriculture Manufacturing Recycling
Textile
Manufacturing Petroleum
Paper
Transport equipment
Transport & communications Food
Rubber
Finance
Trade & accommodation Business
Community
Hotels & restaurants
Financial & business Services Households
Post & telecommunications
Transport
Community & households Real estate
Trade
–2.5 –2.0 –1.5 –1.0 –0.5 0.0 0.5 –5 –4 –3 –2 –1 0 1
Sources: Groningen Growth and Development Centre database; World KLEMS database; and IMF staff calculations.
Note: This figure shows the estimated coefficients from the unconditional convergence regression for each sector based on labor productivity. Solid bars denote that
the coefficient is statistically significant at the 95 percent level. A negative and significant coefficient denotes evidence of productivity convergence across countries.
Nonmarket service industries are excluded due to a lack of reliable sectoral purchasing-power-parity data. Panel 1 corresponds to an extended sample of 19
advanced economies and 20 emerging market and developing economies with data for nine market sectors from 1965 to 2015. Panel 2 corresponds to a reduced
sample of 19 advanced economies and 11 emerging market and developing economies with data for 26 market sectors during 1970–2010. See Annex 3.3 for details.
level of productivity—a concept of convergence known still employs a large share of the workforce (see, for
as beta-convergence (Barro and Sala-i-Martin 1992). instance, the discussion in Chapter 2 of the April 2017
Starting with a sample of 19 advanced economies WEO and Box 1.3 of the October 2017 WEO).28
and 20 developing economies, the analysis provides Another indicator of convergence describes
strong evidence of unconditional convergence of whether the dispersion of sectoral productivity
productivity to the frontier for manufacturing, in line across countries has narrowed over time, a measure
with Rodrik (2013), as well as for several nonmanu- called sigma-convergence. Indeed, the dispersion
facturing sectors (Figure 3.17, panel 1). Importantly, of productivity across countries declined over time
the results suggest significant convergence in three of in all sectors that exhibited significant evidence of
the four market service sectors under study: trade and beta-convergence (Figure 3.18; Annex Figure 3.3.5).29
accommodation, transport and communications, and In the case of the service sectors, the extent of
financial and business services. convergence seems to have accelerated since the
In addition, this sample exhibits no unconditional mid-1990s or early 2000s—a time when the trad-
convergence for agriculture, which employed about
two-thirds of the workforce in developing economies 28Sectoral convergence, however, does not necessarily imply
in the 1970s and almost half as recently as the first aggregate convergence. Even if sectoral productivity has converged to
the frontier level of productivity in all sectors, there will still be dif-
decade of the 21st century. The lack of unconditional
ferences in aggregate productivity levels if the relative size of sectors
convergence in agriculture is an important find- varies across countries (Bernard and Jones 1996).
ing because it may explain the difficulty in finding 29Given that examination of sigma-convergence requires a
evidence of unconditional convergence in aggregate balanced sample, Figure 3.18 is based on a smaller sample than
the beta-convergence, comprising 28 countries for the period
income per worker in broader samples of countries, 1971–2010 (excluding eastern European countries for which sectoral
including lower-income countries where agriculture data are available only since 1995).

Figure 3.18. Sigma-Convergence Figure 3.19. Productivity Gap in 2005

(Standard deviation of log labor productivity, PPP adjusted) (Difference in productivity level with respect to the United States,
percentage points)
The dispersion of productivity levels across countries declined over time in several
service industries, providing further evidence of convergence. The extent of The productivity gap vis-à-vis the United States in 2005 was larger for
productivity convergence in service industries has accelerated since the goods-producing sectors than for the service sector. Resource shifts from goods-
mid-1990s or early 2000s. producing sectors to the service sector need not harm convergence prospects.
0.90 1. Manufacturing 2. Trade & 0.90 400

Accommodation
350 Goods
0.85 Services
0.85 300
0.80 250
0.80 200
0.75
150
0.75 0.70 100

1971 80 90 2000 10 1971 80 90 2000 10
50
0.75 3. Transport & 4. Financial & Business 1.30
Communications 0
1.20 –50
0.70
1.10 –100
–300 –250 –200 –150 –100 –50 0 50
0.65
Difference in real GDP per hour (international dollars)
1.00
0.60 Sources: Inklaar and Timmer (2014); and IMF staff calculations.
0.90
Note: Productivity gap is calculated as the log difference between sector-specific
purchasing-power-parity-adjusted value added per worker in the United States
0.55 0.80 and that in each country. The goods-producing sector includes agriculture, mining,
1971 80 90 2000 10 1971 80 90 2000 10 manufacturing, utilities, and construction. The services sector includes the rest.

Note: See Annex 3.3 for details. PPP = purchasing power parity. The evidence of convergence in services productivity
notwithstanding, the level of productivity in services
may be further away from the technological frontier
ability of services increased considerably (Heuser and than in agriculture or manufacturing. In that case, the
Mattoo 2017; Box 3.2). prospects for narrowing the gaps in income per worker
Further analysis using a reduced sample of 19 as labor shifts from goods-producing sectors to services
advanced economies and 11 developing economies would be jeopardized, at least temporarily. However,
with more granular sectoral detail reveals that almost in most countries, the productivity gap vis-à-vis the
half of the manufacturing industries (including chem- United States in 2005 was larger for goods-producing
icals, food, paper, and rubber) show no evidence of sectors than for the service sector (Figure 3.19), espe-
convergence (Figure 3.17, panel 2). Among services, cially among lower-income countries.
eight out of nine market industries show evidence of The main message that emerges from the vari-
unconditional convergence (including financial inter- ous parts of analysis in this section is that skipping a
mediation, postal services and telecommunications, traditional industrialization phase need not be a drag
and business services).30 on economy-wide productivity growth for developing
economies. Some service industries have the potential to
30There could be some concern that labor productivity convergence boost the growth of aggregate productivity and aid the
comes primarily from capital deepening. A robustness exercise on a convergence of income per worker across countries. But
reduced sample provides evidence of unconditional TFP convergence sustaining the recent improvements in living standards
in some market service sectors (for example, financial intermedi-
ation, business services, and wholesale and retail trade; see Annex in many developing countries will require policy actions
Table 3.3.2). See also the discussion in McMillan and Rodrik (2011). to strengthen productivity growth within all sectors.

Implications for Income Inequality the shift between manufacturing and nonmanufactur-
Historically, manufacturing industries are widely ing employment.
perceived to have been a major source of high-quality The micro-level data used for the analysis are from
jobs. The decline in the share of manufacturing jobs in the Luxembourg Income Study database. Because of
employment, especially among advanced economies, data limitations, the manufacturing sector is represented
has thus fueled concern that the disappearance of what by the broader industrial sector.34 The data used here
are thought to be relatively well-paying manufacturing cover labor income from household surveys in an unbal-
jobs would hurt the living standards of affected work- anced panel of 20 advanced economies since the 1980s.
ers and contribute to a variety of social ills.31 Under
this mechanism, the shift of workers from well-paying Are earnings higher and more equal in industry than
manufacturing to lower-paid jobs in the service sector
in services?
contributes to the “hollowing out” of the income
distribution by moving workers from the middle to The data show that labor compensation in industry is
the lower end of the income scale, leading to higher indeed somewhat higher than in services for comparable
earnings inequality. A large body of research has inves- skill levels (Figure 3.20).35 For medium-skilled workers
tigated the causes of growing income inequality and in the two sectors, earnings are practically indistin-
polarization, focusing primarily on the roles of trade guishable. The median difference in labor earnings
and automation.32 Few studies, however, have sought between industry and services for high- and low-skilled
to isolate the effects of structural transformation on the workers is about 6 percentage points and 9 percentage
distribution of labor income.33 points, respectively. Nonetheless, the skill premium is
Against this backdrop, this section uses micro-level more important in explaining the variation in earnings
data for a set of advanced economies to examine if pay across workers than their sector of employment: the gap
is systematically higher and more evenly distributed in between earnings for middle- versus low-skilled workers
the manufacturing sector, as is often assumed. It then within a sector is about twice as large as the gap between
gauges the extent to which changes in income inequal- low-skilled workers in industry and services.
ity can be attributed to shifts in employment shares Similarly, there is somewhat less labor income inequal-
across sectors, exploiting the initial disparity of earn- ity in the industrial sector than in the service sector (as
ings within and across types of employment. The main indicated by the two leftmost boxes in Figure 3.21, panel
takeaway of the analysis is that only a limited portion 1). But the data also show that countries with a relatively
(less than one-fourth under an extreme assumption) of high degree of earnings inequality within the service
the rise in income inequality could have resulted from sector tend to have high inequality within the industrial
sector as well (Figure 3.21, panel 2).36
31For example, Helper, Krueger, and Wial (2012) document that 34The broad sectors considered for this analysis are agriculture,
average earnings in manufacturing jobs are about 8 percent higher industry (which consists of manufacturing, construction, mining,
than in nonmanufacturing jobs when differences in worker and and utilities), services, and a residual category. The Luxembourg
job characteristics are controlled for. Lawrence (2017) stresses that Income Study database offers an alternative sectoral classification
manufacturing has historically provided the opportunity for relatively that distinguishes the manufacturing sector. However, using this
unskilled workers to earn relatively high wages; he notes that in classification would significantly reduce the sample size. Moreover,
the United States, the manufacturing sector employed more than manufacturing accounts, on average, for about two-thirds of employ-
one-third of men without a college degree in 1970 and 17 percent in ment in the broad industrial sector, and distributional statistics on
2015. Autor, Dorn, and Hanson (2017) highlight the effects of the labor income for manufacturing and overall industry are comparable
loss in manufacturing jobs on family formation dynamics. in countries where data are available for both sectors.
32The literature on job polarization and labor income inequality has 35Average labor earnings in services are higher than in manu-
focused mostly on occupations rather than industries. It indicates that, facturing, but this is because the service sector as a whole employs
since the 1980s, employment and wages in several advanced econo- more high-skilled workers than does manufacturing. Skill levels are
mies tended to grow faster for high- and low-skill occupations than determined according to the following classification of occupations
for middle-skill occupations (Autor, Katz, and Kearney 2006; Goos, in the International Standard Classification of Occupations (ISCO):
Manning, and Salomons 2014). Autor, Dorn, and Hanson (2013, managers and professionals (ISCO 1 and 2) are shown as high skill;
2016) and Acemoglu and Restrepo (2017) argue that trade and tech- laborers/elementary (ISCO 9) as low skill; and other skilled workers
nology are changing the manufacturing sector in the United States by (ISCO 3–8, 10) as medium skill.
lowering the demand for labor, especially for the middle-skill group. 36While this section focuses on advanced economies, potentially
33An exception is Bárány and Siegel (2018), who argue that lower earnings in expanding service sector jobs is also a concern
employment shifts across industries in the United States have for developing countries (Hallward-Driemeier and Nayyar 2017).
enhanced the polarization of the job market. Box 3.3 looks at the experience of individual workers in Brazil

Figure 3.20. Average Gross Wages in Industry and Services Figure 3.21. Labor Income Inequality in the 2000s
in the 2000s (Points)
(Difference with respect to average economy-wide gross wages,
percentage points) Labor income inequality is somewhat lower in industry than in services, but
country characteristics dominate in explaining within-sector inequality in both
Labor earnings in industry are somewhat higher than in services for high- and industry and services.
low-skilled workers and broadly comparable for medium-skilled workers.
0.30 1. Selected Components of Overall Labor Income Inequality
60
0.25
0.20
40
0.15
20 0.10
0.05
0
0.00
Industry Services Cross-sector Total
Within-sector inequality average income
–20 differences
0.30 2. Labor Income Inequality, Industry versus Services

–40
0.25
GRC
–60 0.20 ESP
Industry Services Industry Services Industry Services Industry Services ITA
AUS USA
Industry
High skill Medium skill Low skill Total 0.15 ISL GBR
FRA LUX CHE JPN
0.10
Sources: Luxembourg Income Study database; and IMF staff calculations. FIN AUT NLDCAN
Note: The figure shows the cross-country distribution of the difference between IRL BEL DEU
average (among individuals) gross wages by sector of employment and occupation 0.05 DNK SWE
and average economy-wide gross wages for full-time working household
members for whom occupation data are available. The horizontal line inside each 0.00
0.00 0.05 0.10 0.15 0.20 0.25 0.30
box represents the median; the upper and lower edges of each box show the top
Services
and bottom quartiles; and the red markers denote the top and bottom deciles. The
sample includes 12 countries; data correspond to the latest year available during
2000–09. Sources: Luxembourg Income Study database; and IMF staff calculations.
Note: For each country, the analysis is based on data from the latest year available
during 2000–09. The measure of inequality used is generalized entropy based on
disposable income (see Annex 3.4 for details). Panel 1 shows the cross-country
distribution of labor income inequality in industry and services; the contribution of
How did the shift in workers between industry and differences in average earnings across sectors to aggregate inequality; and
services affect the distribution of labor income? aggregate inequality. The horizontal line inside each box represents the median;
the upper and lower edges of each box show the top and bottom quartiles; and the
To isolate the effects of shifts in sectoral employ- red markers denote the top and bottom deciles. Data labels in panel 2 use
ment shares on earnings inequality, the analysis offers International Organization for Standardization (ISO) country codes.
a thought experiment. If the average pay differen-

tials between sectors and the levels of inequality
within them had stayed at their initial levels, how distribution if the level and distribution of earnings
much would the shifts in sectoral employment in each sector had remained at their initial levels.
shares have changed the inequality in earnings? A Shifts in employment shares between industry and
decomposition along these lines suggests that the services contributed only about 15 percent of the
shift in manufacturing workers to services would not rise in economy-wide income inequality (keeping the
have significantly worsened economy-wide income dispersion and relative level of earnings constant at
their initial values).37 Instead, between the 1980s and
between 1996 and 2013—a period during which the service sector
was expanding while manufacturing employment remained broadly 37A definitive test of whether the shift of middle-skilled workers
stable. Though the findings cannot be generalized, the analysis does from manufacturing to services implies erosion of their income
not find significant wage gains for workers who move to manufac- would require data over time at the individual level, which are not
turing jobs from other sectors. available for a broad set of countries.

2000s, most of the rise in earnings inequality within Figure 3.22. Contribution to Change in Overall Labor Income
countries came from the rise in pay inequality within Inequality between the 1980s and 2000s
(Points)
services and industry (Figure 3.22).38
The increase in earnings dispersion within sectors
Most of the increase in overall labor income inequality between the 1980s and
could result, however, in part from the movement 2000s is explained by rising inequality within sectors rather than by shifts in the
of workers across sectors for two reasons. First, the relative size of employment between industry and services.
dislocation of manufacturing workers to low-skill
0.06
(and low-wage) jobs in services would “mechanically”
increase the share of workers at the lower portion
of the income distribution and raise income polar- 0.04
ization and inequality. With the average income of
middle-skilled workers in the industry sector almost 0.02
30 percent higher than that of low-skilled service
sector workers (Figure 3.20), the disappearance of
0.00
middle-skill manufacturing jobs could imply a large
pay cut for workers in that group who move to
low-skill jobs in the service sector. Second, a spurt in –0.02
the number of workers competing for lower-skill jobs
can put downward pressure on wages at the lower end –0.04
of the earnings distribution, widening the gap between
incomes at the high and low ends of the spectrum.
–0.06
To assess the quantitative relevance of the first Within-sector Change in Cross-sector Total change
inequality sector size average income
channel, a stylized exercise assumes that, in the eight differences
economies with available data since the 1980s and
where manufacturing employment fell in absolute Sources: Luxembourg Income Study database; and IMF staff calculations.
terms, all manufacturing jobs lost between the 1980s Note: The figure shows the cross-country distribution of the change in aggregate
labor income inequality between 1980–89 and 2000–09 and the contribution from
and 2000s were those of middle-skilled workers who changes in inequality within sectors, changes in the relative size of sectors, and
moved to low-skill and low-wage jobs in services (set changes in the difference of average income levels across sectors. The horizontal
line inside each box represents the median; the upper and lower edges of each box
to the 25th percentile of wages in low-skill service show the top and bottom quartiles; and the red markers denote the top and bottom
jobs). In this scenario, overall labor income inequality deciles. The measure of inequality used is generalized entropy based on disposable
would have increased, on average, by about 9 per- income. The sample consists of 13 countries (see Annex 3.4 for details).
cent of the actual increase in inequality between the

1980s and 2000s and up to one-fourth in any of the
countries considered. In summary, the findings in this section suggest
Testing whether the dislocation of manufacturing that changes in aggregate labor income inequality
workers to low-skill jobs exerts downward pressure on are predominantly explained by rising labor income
wages for all workers at the lower end of the earn- inequality within sectors. As analyzed widely in the
ings distribution is beyond the scope of this chapter literature, the key drivers behind greater pay inequality
but could be a fruitful area for future research. Autor over time seem to be the dislocation of middle-skilled
(2015) argues that the slow wage growth in low-skill workers through technology and trade—and the resul-
jobs during 1999–2007 in the United States may have tant downward pressure on wages for medium- and
been related to middle-skilled workers—including low-skill jobs—rather than shifts in the relative size of
those displaced from highly routinized jobs—taking employment between industry and services.
low-skill jobs. A word of caution regarding these findings is
nonetheless warranted. First, displaced middle-skilled
38The analysis is based on a decomposition of the overall change in manufacturing workers may end up experiencing
labor income inequality between the 1980s and 2000s for a sample prolonged unemployment spells or dropping out of
of 13 economies into the contribution of within-sector changes in the labor force rather than taking low-wage jobs in
inequality, changes in the relative size of each sector, and changes in
average incomes across sectors. The year used for each country varies
services, leading to an increase in overall inequality
depending on survey data availability. See Annex 3.4 for details. that would not be captured in the analysis based on

workers’ labor earnings. Indeed, the analysis in Chap- Shifts of employment shares toward services during
ter 2 shows that workers in routinizable occupations the past two decades may have been enabled in part by
were more likely to involuntarily drop out of the strong domestic demand, which has lost momentum in
labor force. Second, some valuable nonwage attributes many developing economies, especially among commod-
of manufacturing jobs appear less widespread in other ity exporters. To help maintain productivity-benefiting
sectors. Manufacturing jobs tend to be characterized structural change, policymakers need to ensure that the
by formal employment arrangements with associated growth of domestic demand and available workforce
benefits for workers, such as access to retirement skills do not impede the expansion of highly productive
plans, paid holidays and sick leave, and health and service activities. Reducing barriers to international trade
life insurance. They also tend to provide relatively and investment in services, which tend to be particularly
stable arrangements, relying less on part-time or high in developing economies (Miroudot, Sauvage, and
temporary contracts than other sectors (Chapter 2 Shepherd 2013; Koske and others 2015), would expand
of the October 2017 WEO), and may offer collec- the service sector’s opportunities for tradability, scale,
tive bargaining via unions (Jaumotte and Osorio and productivity growth. Adapting the rules in multi-
Buitron 2015). Finally, even if shifts in employment lateral trade agreements to cover areas such as digital
shares between sectors contributed little to aggregate trade and e-commerce (as discussed in Chapter 1 and
inequality, the negative consequences of declining Box 2.2 of the October 2016 WEO) would also help
manufacturing jobs can be sizable for some groups. in that regard. To facilitate the reallocation of workers
Transitional costs associated with sectoral reallocation to sectors where their efficiency is higher, policy should
can be substantial for individual workers, due to both also ensure that workforce skills are aligned with those
prolonged unemployment spells and lower earnings needed in highly productive and expanding sectors of
in subsequent jobs (Walker 2013). These individual the economy.
costs can have nonnegligible aggregate incidence in The analysis in the chapter also indicates that
regions that had developed as manufacturing hubs. within-sector productivity growth remains anemic
in developing economies outside East Asia. In many
countries, raising productivity in agriculture—which
Conclusions and Policy Implications remains the primary employer and still exhibits very
This chapter finds that the decline in the share of low levels of productivity—is key to facilitate the
manufacturing jobs in overall employment need not transition of workers to dynamic industries in manu-
hurt growth or raise inequality. Some service sectors facturing and services. More generally, a comprehensive
can match the productivity levels and growth rates approach is needed to strengthen productivity across
of manufacturing, so the relative expansion of those all sectors, including by bolstering human capital and
services could help national income approach advanced physical infrastructure and improving the business and
economy levels in economies that appear to be bypass- investment climate. Reforms aimed at removing obsta-
ing a traditional industrialization phase. Some service cles to the efficient movement of factors of production
sectors exhibit signs of productivity convergence between firms and promoting competition are also
to the frontier, and the shift of employment shares key, especially in services where barriers to entry tend
from agriculture toward services since the 2000s has to be higher and the extent of competition lower than
benefited economy-wide productivity in many develop- in goods-producing sectors (Koske and others 2015;
ing countries. Chapter 3 of the April 2016 WEO). For example, the
However, these findings do not necessarily mean that extent of government involvement in network sectors
income convergence is assured—whether manufacturing (such as electricity, gas, rail transportation, air trans-
is expanding or not—or that recent favorable trends in portation, postal services, and telecommunications)
output per worker can be extrapolated into the future. and barriers to entry in network sectors and other
Strong policy efforts are needed to facilitate the reallo- services (professional, freight transport, and retail
cation of activity toward higher-productivity sectors and distribution services) are still pervasive and partic-
bolster productivity growth across all sectors.39
be counterproductive as they might preserve low-productivity firms
39Policies that do not respond to a specific market failure but and postpone an efficient reallocation of resources (Fournier and
focus solely on the relative size of the manufacturing sector could Johansson 2016).

ularly large among developing economies. Services the types of skills required in routinizable occupations.
deregulation would facilitate the entry of new firms To ensure inclusive gains from technological prog-
into the sector and promote competition (Koske and ress, policy should help workers cope with its adverse
others 2015; Adler and others 2017). Moreover, the side effects. A range of factors—including financial
consequent productivity gains in services can gener- constraints, strong ties to their local area, and lack of
ate positive spillovers for downstream and upstream needed skills—may have prevented workers displaced
industries, including in manufacturing (Fernandes and from manufacturing jobs from taking adequate employ-
Paunov 2012; Bourlès and others 2013; Lanau and ment in other sectors. Expanding access to training
Topalova 2016). and education programs aligned with the needs of the
The chapter also finds that changes in the inequality evolving economy (including job-search assistance and
of labor income in advanced economies are predomi- training) as well as safety nets and redistribution pol-
nantly explained by rising earnings inequality within all icies targeted at displaced workers can help soften the
sectors. Higher pay inequality has nonetheless coincided blow imposed by structural transformation on workers
with lower shares of employment in manufacturing and and their communities. Regions with a heavy reliance
higher shares of service sector jobs, reflecting trends, on declining manufacturing jobs may require specifi-
such as automation, that have affected the demand for cally targeted policy measures to facilitate the transition.
151
Box 3.1. The Changing Service Content of Manufactures

Services account for an increasing share of the total Figure 3.1.1. Nonmanufacturing Value-Added
value of manufactured goods—an increase sometimes Content in Gross Manufacturing Output,
called the “servicification” of manufacturing.1 The 1995–2011
change in input-output linkages between the service (Percent)
and manufacturing sectors from 1995 to 2011 implies
50 25
that the share of service inputs in the total production
value of manufactures increased by about 6 percentage
points, on average, across countries. This increase can
reflect, for instance, rising consumer demand for goods 48 23
that are more intensive in service inputs (for instance,
design and software), or the fact that combining
production inputs that are increasingly diffused geo- 46 21
graphically requires more service inputs (for instance,
logistics and communications). However, service
inputs in manufactured goods account for a small frac- 44 19
tion (about 12 percent) of overall value added in the
service sector, and the share of manufactures in total
final expenditure has been steadily declining during 42 17
this period (Figure 3.1.1). The lion’s share of the
expansion of services in aggregate value added—6 per-
centage points out of 7 percentage points, on aver- 40 15
1995 97 99 2001 03 05 07 09 11
age, between 1995 and 2011—corresponds to an
increase in final expenditure on services—rather than Sources: World Input-Output Database; and IMF staff
to an increase in the share of service inputs used by calculations.
other sectors. Note: The solid line (right scale) shows global spending on
Services can augment the value of finished man- manufactures as a share of global total spending. The
shaded area (left scale) depicts the share of
ufactured goods in two ways: (1) as inputs in the nonmanufacturing value-added content in gross
manufacturing process, or (2) as auxiliary activities manufacturing output.
bundled with finished goods when sold to consumers.
Examples of service inputs include design, research
and development, and information technology; exam- box uses worldwide input-output data to quantify the
ples of auxiliary service activities include financing, service content of manufacturing gross output.3,4
logistics, and installation. At the global level, the contribution to gross manu-
Input-output tables contain information on facturing output by nonmanufacturing activities—such
supply-use relations between industries within and as agriculture, mining, and services—increased from
across countries. To date, studies measuring the extent 42 percent of total gross manufacturing output in 1995
of servicification have been limited to individual to 47 percent in 2011 (Figure 3.1.1). About two-thirds
countries or exports rather than overall output.2 This of the nonmanufacturing contribution to gross
Trade in Value-Added database to document the evolution of

The author of this box is Wenjie Chen. services trade in global value chains.
1This term is used in Baldwin, Forslid, and Ito (2015); 3The data used for this analysis come from the 2013 release of the
Miroudot and Cadestin (2017); and Hallward-Driemeier and World Input-Output Database, which covers the world economy
Nayyar (2017), among others. over 1995–2011 (including data for 40 individual economies,
2For instance, the National Board of Trade of Sweden (2010) accounting for more than 85 percent of world GDP) and from the
and Lodefalk (2013) show that, in Sweden, the services share of corresponding socioeconomic accounts (Timmer and others 2015).
total inputs used in manufacturing doubled over 1975–2005, The computations used for this box are described in Annex 3.2.
with most of the gains reached by 1995. Baldwin, Forslid, and 4The gross output of the manufacturing sector is the sum
Ito (2015) use input-output data for a group of Asian econo- of the value added of the sector and the intermediate inputs it
mies and document a surge in the value-added share of services uses, whether produced domestically or abroad. Domestic gross
in manufactured exports. Heuser and Mattoo (2017) use the output can be constructed by extracting the foreign value-added
Organisation for Economic Co-operation and Development content of intermediate inputs.

Box 3.1 (continued)

Figure 3.1.2. Services Value-Added Content in Figure 3.1.3. Change in Services Value-Added
Gross Manufacturing Output, 1995 and 2011 Content in Manufacturing Gross Output,
(Percent) 1995–2011
(Percentage points)
1995 2011
30
70
20
60
50
40 10
30
20
10 0
0
LUX
BEL
NLD
AUS
ESP
PRT
DEU
AUT
SWE
GBR
CAN
JPN
GRC
FIN
IRL
FRA
USA
ITA
DNK
–10
80 2. Emerging Market and Developing Economies

70 –20 All manufacturing
Basic metals and fabricated metal

Chemicals and chemical products
Refined petroleum and nuclear fuel
Electrical and optical equipment
Food, beverages, and tobacco
Leather and footwear
Machinery, nec.
Manufacturing, nec; recycling
Other nonmetallic mineral
Paper; printing and publishing
Rubber and plastics
Textiles and textile products
Transport equipment
Wood and products of wood and cork
60
50
40
30
20
10
0
TWN
RUS
CYP
IND
HUN
EST
CZE
MEX
SVN
POL
LTU
MLT
ROW
KOR
TUR
BGR
ROU
CHN
IDN
BRA
LVA
SVK
Sources: World Input-Output Database; and IMF staff

calculations. Sources: World Input-Output Database; and IMF staff
Note: Data labels use International Organization for calculations.
Standardization (ISO) country codes. ROW = rest of the Note: The figure shows the cross-country distribution of changes
world. in the service value-added content in gross manufacturing
output between 1995 and 2011 for each of the 14
manufacturing industries. The horizontal line inside each box
represents the median; the upper and lower edges of each box
show the top and bottom quartiles; and the red markers denote
manufacturing output come specifically from service the top and bottom deciles. nec = not elsewhere classified.
industries. For the median economy in the sample, the
contribution of services to gross manufacturing output
was about one-third of manufacturing gross output in
2011 (Figure 3.1.2), albeit with considerable variation, As documented in the main text, the prices of
ranging from about 15 percent in Indonesia to 50 per- manufactures relative to services have been declin-
cent in France and 70 percent in Luxembourg. Across ing in most economies. The increase in the share of
all economies in the sample, the services value-added services in the total production value of manufactures
share in gross manufacturing output increased by an could thus reflect that the price of services value
average of about 6 percentage points, or about 0.4 per- added has outpaced that of manufacturing. Indeed,
centage point a year between 1995 and 2011. The ser- when calculated at constant (real) prices, the rise in
vices contribution increased across the whole spectrum the services share of gross manufacturing output is
of manufacturing industries (Figure 3.1.3). smaller than it is at current prices, and even declined

Box 3.1 (continued)

Figure 3.1.4. Change in Services Nominal in many economies over the 1995–2009 period
and Real-Value-Added Content in (Figure 3.1.4).5 Nonetheless, the share measured in
Manufacturing Gross Output, 1995–2009 real prices increased in about two-thirds of the sample
(Percentage points) economies.
Finally, despite the higher service content of
Nominal Real manufactures documented above, the increase in the
share of service inputs in the total production value
30 1. Advanced Economies of manufactures during 1995–2009 did not play an
important role in the overall expansion of services in
20 the economy. The expansion of services value added
as a share of total value added (by almost 7 percent-
age points, on average, between 1995 and 2011) was
10
mostly due to an increase in final demand for services
(about 6 percentage points, on average), rather than
0 due to an increase in the use of services as intermedi-
ate inputs by other sectors.
–10
AUS
CAN
PRT
NLD
AUT
JPN
IRL
DEU
ESP
GRC
BEL
FIN
SWE
GBR
USA
ITA
FRA
DNK
30 2. Emerging Market and Developing Economies
20
10
–10
IDN
MEX
KOR
CZE
ROU
CHN
RUS
HUN
IND
SVN
TWN
BGR
POL
TUR
LTU
EST
MLT
CYP
BRA
LVA
SVK
Sources: World Input-Output Database; and IMF staff

calculations.
Note: Data labels use International Organization for 5Data limitations restrict the comparison to the 1995–2009
Standardization (ISO) country codes. period. The results for 1995–2007 are similar.

Box 3.2. The Rise of Services Trade

Cross-border trade in services has been growing to 10 percent in 2014 (Figure 3.2.1). The increase in
steadily over the past four decades, and now accounts the tradability of services is widespread across countries
for about one-fifth of global exports (Figure 3.2.1). (Figure 3.2.2, panel 2).
The service share of exports has expanded in most In terms of industries, the increase in service
advanced and developing economies (Figure 3.2.2, exports has been particularly large in “modern”
panel 1), with the expansion being particularly pro- services that can be delivered at a distance, such as
nounced in the latter group. In one-fourth of these telecommunications, computer and information ser-
economies, the service share of exports increased by vices, intellectual property, financial intermediation,
more than 20 percentage points since the early 1980s. and other business activities, including research and
Much of the rise in the share of service exports comes development and professional services (Figure 3.2.3).3
from the decline in trading costs, in turn resulting from The share of modern services exports in total services
advances in information and communication technol-
ogies.1 The rise of global value chains (GVCs) has also 3Following Loungani and others (2017), modern services
been intricately linked to the rise of services trade. As in typically refer to those that do not require the physical proximity
the case of goods, the emergence of GVCs has allowed of buyer and seller. All other services are classified as traditional,
although the boundaries between traditional and modern are
for international specialization in service tasks, and
becoming increasingly blurred as technology evolves.
services have been increasingly traded as components
within GVCs.2 Indeed, many services have become as
tradable as manufactured goods (see Gervais and Jensen Figure 3.2.2. Increase in Service Trade,
2014). As a result, cross-border trade as a share of global 1980–2014
services output has risen from about 3 percent in 1970 (Percentage points)
50 1. Change in Share of Services in

The author of this box is Ke Wang.
1Copeland and Mattoo (2007) and Francois and Hoekman Total Exports, 1980–2014
40
(2010) review the growing literature on trade in services.
2Heuser and Mattoo (2017) provide a comprehensive overview 30
of the role of services trade within global value chains. 20
10
Figure 3.2.1. Exports of Services 0

(Percent) –10
25 14 –20
AEs EMDEs
12 50 2. Change in Share of Services Exports in

20 Services Output, 1980–2014
40
10
30
15
8
20
6 10
10
0
4
–10
5 AEs EMDEs
Share in total world exports 2
Share in services output (right scale) Sources: Loungani and others (2017); and IMF staff
0 0 calculations.
1970 75 80 85 90 95 2000 05 10 14 Note: The horizontal line inside each box represents the
median; the upper and lower edges of each box show the
top and bottom quartiles; and the red markers denote the
Sources: Loungani and others (2017); and IMF staff top and bottom deciles. AEs = advanced economies;
calculations. EMDEs = emerging market and developing economies.

Box 3.2 (continued)

Figure 3.2.3. Services Exports by Industry, exports increased from about 32 percent in 1990 to
1990–2014 50 percent in 2014. The fastest-growing segment was
(Percent) telecommunications and computer and information
services industries, whose exports in 2014 reached
100
10 percent of total services exports, up from 1 percent
90 in 1990. The travel industry accounts for a sizable
fraction of the services exports of developing econo-
80 Business services mies, although its relative importance has diminished
Financial services
70 Intellectual property over time.
Insurance and Barriers to international trade are larger for service
60 pension exports than for goods, and particularly large in devel-
Telecommunications,
50 computer, and info. oping countries (Miroudot, Sauvage, and Shepherd
Travel 2013; Koske and others 2015; Heuser and Mattoo
40 Transport 2017). Moreover, service sectors facing lower trade
Other traditional
30 services costs tend to be more productive and exhibit higher
productivity growth (Miroudot, Sauvage, and Shep-
20
herd 2013). Policy action to reduce barriers to trade
10 in the service sector would enhance its tradability and
help boost productivity growth in services.
0
1990 2014
Sources: Loungani and others (2017); and IMF staff

calculations.
Note: Blue (red) shades represent traditional (modern)
services.

Box 3.3. Are Manufacturing Jobs Better Paid? Worker-Level Evidence from Brazil
Jobs in the service sector, rather than in man- ment allows for an examination of the relationship
ufacturing, are increasingly replacing agricultural between the rising role of service employment and
employment in developing economies. This box uses a wage inequality.1
rich micro-level data set from Brazil to answer the fol- Wages in manufacturing are not much higher than
lowing questions: Are wages higher in manufacturing in services (Figure 3.3.2). After controlling for age,
than in services for workers with comparable skills? Do education, and labor market regions of workers, the
workers who switch to manufacturing jobs from jobs wage gap across sectors at the outset of the period was
in agriculture or services obtain initial wage gains and only about 6 percentage points, and by 2013 it was
faster wage growth? Are labor earnings more uniformly close to zero.
distributed in manufacturing than in services? In line with a moderate differential between wages
About 10 percent of Brazil’s workforce moved from in manufacturing and elsewhere, workers who switched
agriculture to service activities between 1996 and from agriculture or services to the manufacturing sector
2013, while the share of manufacturing jobs remained during the sample period did not obtain much of an
broadly stable (Figure 3.3.1). Wage inequality fell initial wage gain. The average boost was no larger than
during that period. A panel data set that tracks the
wages of Brazilian workers and their sector of employ- 1The data are from household surveys and a large matched
set of employer-employee records on workers’ income, hours,
The author of this box is Jorge Alvarez. education, and other demographic characteristics.
Figure 3.3.1. Sectoral Employment Shares Figure 3.3.2. Wage Gap between
and Wage Inequality Manufacturing and Services
(Percent, unless noted otherwise) (Percentage points)
Agriculture Manufacturing 12
Services Inequality (right scale)
Total difference
10
Residual difference
80 1.0
8
70
60 0.9 6
50 4
40 0.8 2
30
0
20 0.7
–2
1996 98 2000 02 04 06 08 10 12 13
10
Sources: Alvarez (2017); and IMF staff calculations.

0 0.6
1996 98 2000 02 04 06 08 10 12 13 Note: The figure shows the average difference in mean log
wages between workers in manufacturing and those in
services based on household surveys from the Pesquisa
Sources: Alvarez (2017); and IMF staff calculations. Nacional por Amostra de Domicílios. The residual difference
Note: The measure of inequality is the variance of log wages is the average difference after controlling for the age,
based on the the Pesquisa Nacional por Amostra de education, gender, race, and region of individual workers.
Domicílios household survey data. Survey data are not Survey data are not available for 2000, 2003, and 2010.
available for 2000, 2003, and 2010.

Box 3.3 (continued)

Figure 3.3.3. Wages of Workers Switching to Figure 3.3.4. Inequality in Manufacturing
Manufacturing Jobs and Services
(Wage relative to level at time of switching sectors, (Variance of log labor earnings)
percentage points)
Within firms Between firms
15
1.0 1. Manufacturing
12
9 0.8
6 0.6
Wage gain
3 0.4
0 0.2
–3
0.0
1988 90 92 94 96 98 2000 02 04 06 08 10 12
–6 From services
From agriculture
–9 1.0 2. Services
–12 0.8
–2 –1 0 1 2 3 4 5
0.6
Source: Calculations from Alvarez (2017) using Relação
Anual de Informações Sociais panel data on formal workers. 0.4
Note: The figure shows the average relative wage of
individual workers who shifted to manufacturing jobs
(relative to their wage level at the time of the transition), 0.2
controlling for time and worker fixed effects. X-axis labels
indicate the number of years before and after switching 0.0
sectors (1 = first year in new sector). 1988 90 92 94 96 98 2000 02 04 06 08 10 12
Sources: Alvarez and others (2018); and IMF staff

calculations.
the expected gain from an additional year of experience Note: The measure of inequality used is the variance of log
in the original sector (Figure 3.3.3).2 labor earnings and is computed using Relação Anual de
Informações Sociais panel data on formal workers.
Moreover, wage gains for workers who switched to
a manufacturing job were no faster in the subsequent
years than they were before the switch, once the explaining the overall level of inequality in each sector
common trend in wages across all workers and worker and the decline since the mid-1990s. Less important
characteristics are accounted for. is inequality within service sector firms versus that
At least in the formal sector, wage inequality is not within manufacturing firms.
higher in services than in manufacturing, and the In sum, differences between the services and manu-
decline of inequality in the two sectors over the past facturing sectors in terms of the level and dispersion of
few decades is very similar (Figure 3.3.4). The analysis wages have remained small in Brazil over two decades
also shows that the dispersion of wages across firms that saw an expansion of the services share of employ-
within the two sectors plays an important role in ment and a decline in overall inequality. Changes that
affect all sectors, such as the increase in the minimum
2The analysis shows gains from transitioning after controlling
wage (Engbom and Moser 2018), and other firm-level
for time effects. As discussed in Alvarez (2017), these expected factors (Alvarez and others 2018), appear to have played
gains are equivalent to the sectoral premiums after controlling for
differences in both observable and unobservable characteristics
a more prominent role in driving overall labor income
of workers in the two sectors. Similar trends are seen when using inequality than changes in the relative size of manufac-
wages or earnings. turing versus service jobs in overall employment.

Annex 3.1. Data Sources and Country Coverage Annex Table 3.1.3 provides a summary of the main
All data sources used in the chapter are listed in sectoral compositions used throughout the chapter and
Annex Table 3.1.1. The country coverage for the differ- the correspondence with United Nations International
ent sections is presented in Annex Table 3.1.2. In this Standard Industrial Classification (Revision 3.1) sectors.
chapter, advanced economies are those that are classified Data from multiple sources are used to enhance
as such by the World Economic Outlook in 1996. All the coverage of sectoral employment and output series
other economies are considered emerging market and (Annex Table 3.3.1): Groningen Growth and Devel-
developing economies (developing economies for short). opment Centre (GGDC), Organisation for Economic
Annex Table 3.1.1. Data Sources

Indicator Source
Final Expenditure on Manufacturing Goods IMF staff calculations based on World Input-Output Database
Generalized Entropy IMF staff calculations based on Luxembourg Income Study database
Gross Hourly Wage IMF staff calculations based on Luxembourg Income Study database
Manufacturing Consumption Share IMF staff calculations based on World Bank, International Comparison Program database
Manufacturing Gross Output IMF staff calculations based on World Input-Output Database
Purchasing Power Parity Penn World Table 9.0
Real GDP per Capita IMF, World Economic Outlook database
Relative Price of Manufactured Goods IMF staff calculations based on GGDC; UN National Accounts Official Country Data
database
Sectoral Employment Felipe and Mehta (2016); GGDC; ILO; national sources; OECD; UNIDO; World KLEMS
database
Sectoral Labor Productivity GGDC; World KLEMS database
Sectoral Purchasing Power Parity GGDC; Inklaar and Timmer (2009); World KLEMS database
Sectoral TFP Jorgenson, Ho, and Samuels (2013); World KLEMS database
Sectoral Value Added (at current and constant prices) GGDC; UN National Accounts Official Country Data database; World KLEMS database
Note: GGDC = Groningen Growth and Development Centre; ILO = International Labour Organization; TFP = total factor productivity; UN = United Nations;
WTO = World Trade Organization.
Annex Table 3.1.2. Sample of Economies Included in the Analytical Exercises

Exercise3
Group1 Economies2 I II III IV V VI
Argentina,* Australia, Austria, Belgium, Brazil,* Canada, Chile,* China,* Denmark, Finland,
France, Germany, Greece, India,* Indonesia,* Ireland, Italy, Japan, Korea,* Luxembourg,
A X X X
Mexico,* Netherlands, Poland,* Portugal, South Africa,* Spain, Sweden, United Kingdom,
United States
Bolivia,* Botswana,* Colombia,* Costa Rica,* Egypt,* Ethiopia,* Ghana,* Hong Kong SAR,*
B Kenya,* Malawi,* Malaysia,* Mauritius,* Morocco,* Nigeria,* Peru,* Philippines,* Senegal,* X X
Singapore,* Taiwan Province of China,* Tanzania,* Thailand,* Venezuela,* Zambia*
Bangladesh,* El Salvador,* Guatemala,* Honduras,* Norway, Pakistan,* Panama,* Puerto
C X
Rico,* Romania,* Suriname,* Switzerland, Syria,* Trinidad and Tobago*
Cyprus,* Czech Republic,* Estonia,* Hungary,* Latvia,* Lithuania,* Malta,* Russia,* Slovak
D X X
Republic,* Slovenia*
E Austria, Finland, Germany, Ireland, Netherlands, United States X X X
F Australia, Canada, Denmark, France, Italy, Sweden, United Kingdom X X
G Belgium, Greece, Iceland, Luxembourg, Spain, Switzerland X X
H Japan X
1 Group of economies according to their use in different analytical exercises.
2 Asterisk (*) denotes emerging market and developing economies as classified by the IMF, World Economic Outlook, plus economies used in the exercises
that have been reclassified as advanced economies since 1996 (Cyprus, Czech Republic, Estonia, Hong Kong SAR, Korea, Latvia, Lithuania, Malta, Puerto
Rico, Singapore, Slovak Republic, Slovenia, Taiwan Province of China).
3 Analytical exercises performed in the chapter: I = stylized facts (Figures 3.1, 3.4–3.7, 3.9, 3.10); II = sectoral employment (Figure 3.2) and productivity
(Figures 3.12–3.16); III = beta convergence (Figure 3.17); IV = inequality decomposition, 2000s (Figure 3.21); V = inequality decomposition over time
(Figure 3.22); VI = wages (Figure 3.20).

Annex Table 3.1.3. Sectors, Individual Industries, and Abbreviations Used in the Chapter
Sector Group 10-Sector Name Sectors Included (ISIC Revision 3.1) Examples of Industries Included
Agriculture Agriculture Agriculture; fishing
Manufacturing Manufacturing Manufacturing Food, tobacco, textiles, apparel, leather,
wood, paper, coke, chemicals, rubber,
other nonmetallic products, basic metals,
electrical equipment, machinery, transport
equipment, recycling, petroleum
Market Services Trade & accommodation Wholesale and retail trade; repair Wholesale and retail trade; sale, maintenance,
of goods; hotels and restaurants and repair of motor vehicles
Transport & communications Transport, storage and Land, water, and air transport; post and
communications telecommunications
Financial & business Financial intermediation; real Financial intermediation, insurance and
estate and business activities pensions, real estate, renting of machinery
and equipment, computer (including
hardware consulting, production of
software, and data processing), research
and development, other business activities
(including professional services)
Community & households Community and personal services; Sewage and sanitation; recreational and
activities of private households other service activities; activities of private
household as employers of domestic staff
Nonmarket Services Government Public administration and defense; Public administration and defense; education;
education; health health
Other Utilities Electricity, gas, and water supply
Construction Construction
Mining Mining and quarrying
Note: ISIC = International Standard Industrial Classification.
Co-operation and Development (OECD), World input-output linkages across countries and industries for
KLEMS, International Labour Organization (ILO), each year between 1995 and 2011. It covers 40 econo-
United Nations Industrial Development Organization mies (19 advanced and 21 developing economies, repre-
database (UNIDO), and Felipe and Mehta (2016). senting more than 85 percent of world GDP), along with
The main source for sectoral employment data is the a residual for the noncovered part of the world economy,
GGDC 10-sector database. The country and time and 35 industries. The data also contain final expenditure
coverage are extended using, in order of preference, and value added by industry for each country.41
World KLEMS, OECD, UNIDO, and ILO, as well The analysis follows the consumption value-added
as national sources for individual countries.40 Data procedure described in Herrendorf, Rogerson, and
from Felipe and Mehta (2016) provide manufacturing Valentinyi (2013) to decompose a given value of final
employment data for additional countries. Sectoral expenditure into its underlying value-added compo-
value-added data are from the GGDC 10-sector data- nents. Using this approach on global input-output data
base, World KLEMS, UN National Accounts Official allows to decompose the value of global final spending
Country Data database, and national authorities. on finished manufactured products into the value added
The analysis on inequality relies on the Luxembourg from each country and sector (that is, both manufac-
Income Study database and the Standardized World turing and nonmanufacturing) that is embedded in
Income Inequality database. those manufactures. Summing the resulting decomposed
value added across sectors for a given country gives
Annex 3.2. Value-Added Decomposition the measure of domestic gross output of manufactures
underlying the calculations in Figure 3.11. Summing the
Data from the World Input-Output Database resulting value added across nonmanufacturing sectors in
(WIOD) is used for the analysis underlying Figure 3.11 all countries gives the measure reported in Figure 3.1.1
and Box 3.1. The WIOD provides data on global (Box 3.1). The results reported in Figure 3.1.2 are
40National sources are used for Australia, Canada, China, Brazil,
Costa Rica, El Salvador, Guatemala, India, Korea, Malaysia, Philip- 41Timmer and others (2015) provide more details about the con-
pines, Thailand, and Trinidad and Tobago. struction of the database and discuss additional features.

obtained by summing the decomposed value added Annex Figure 3.3.1. Distribution of Total Factor Productivity
across service industries in each country while those in Growth of Individual Industries
(Kernel density)
Figure 3.1.3 correspond to the decomposed value added
in each service industry. The calculation of value added 0.7 1. Disaggregated Manufacturing and Service Industries
in constant (real) prices in Figure 3.1.4 requires the use (United States, 1947–2010)
0.6
of the sector specific value-added price indices in the Manufacturing
WIOD Socio Economic Accounts data. 0.5 Services
0.4
Annex 3.3. Sectoral Productivity, Aggregate 0.3

Growth, and Convergence 0.2
This annex provides additional details on the analysis 0.1

shown in the section “Growth and Development beyond 0.0
–10 –8 –6 –4 –2 0 2 4 6 8 10 12 14 16
Manufacturing.” Annex Table 3.1.2 presents the sample
of economies included in the analyses in this section.
0.25 2. Disaggregated Manufacturing and Service Industries
(20 economies, 1970–2010)
Sectoral Productivity Analysis 0.20
Data
0.15
Sectoral labor productivity is constructed as value
added at constant prices in a given sector divided by 0.10
the number of workers in that sector (Figure 3.14,
panel 1), or divided by total hours worked in the sec- 0.05
tor (Figure 3.14, panels 2 and 3), using data from the
0.00
Groningen Growth and Development Centre (GGDC) –10 –8 –6 –4 –2 0 2 4 6 8 10 12 14 16
10-sector database; World KLEMS; Jorgenson, Ho,
and Samuels (2013); and national sources (see Annex Sources: Groningen Growth and Development Centre database; World KLEMS
Table 3.1.1). The data reported in Figure 3.14, panel Note: The figure shows the distribution of average total factor productivity growth
1, are available for a sample of 62 economies (19 per decade in individual manufacturing and service industries (expressed as
advanced and 43 developing economies) spanning deviations from the average total factor productivity growth across sectors in each
country and decade). Panel 1 is based on total factor productivity growth data for
10 broad sectors during 1965–2015. Data at a more 20 manufacturing and 39 service industries in the United States (Jorgenson, Ho,
disaggregated sectoral level are available for fewer and Samuels 2013). Panel 2 is based on total factor productivity growth data for
13 manufacturing and 13 service industries in 16 advanced economies (including
countries: a sample of 31 economies (19 advanced and the United States) and 4 emerging market and developing economies.
12 developing economies) for 1970–2010 spanning 13
manufacturing and 13 service industries (Figure 3.14,
panel 2), and data for the United States for 1947–2010 Annex Figure 3.3.2 is based on the same sample of
spanning 20 manufacturing and 39 service industries countries and industries as Figure 3.14, panel 2, but
(Figure 3.14, panel 3). Total factor productivity (TFP) shows the cross-country distribution of the average
data based on sectoral value added (reported in Annex relative sectoral labor productivity growth (relative to
Figure 3.3.1) are available for a reduced set of 20 econ- economy-wide labor productivity growth) over 2000–10
omies (16 advanced and 4 developing economies).42,43 for 13 individual service industries and manufactur-
ing as a whole. The average sectoral labor productivity
growth across countries during that period is then
42The source of sectoral TFP data is World KLEMS Growth
used to split service industries into two groups: a
and Productivity Accounts (ISIC Rev. 3, 2011 release, and Rev. 4, high-productivity-growth group, corresponding to the
2017 release). four industries that have on average the highest labor
43Under the assumptions of perfect competition, full capacity
productivity growth (postal services and telecommu-
utilization, and constant return to scale, TFP growth for each sector
is calculated based on the standard growth accounting methodology: nications, financial intermediation, wholesale trade,
and retail trade); and a low-productivity-growth group,
t  ∆ lnKt      − v¯   t  ∆ lnLt     − (1  − v¯   t  − v¯   t  )
¯   k,i
∆ lnTFP ti  = ∆ lnYt  i  −  v  i L,i i k,i L,i ∆ lnM  i  ,
t
¯   k,i
in which i denotes country, v  L,i
t  and v¯   t  denote the two-period aver- output. Yt  i  denotes the (constant-price-based) gross output, K ti  mea-
age (t and t – 1) share of capital and labor input in nominal gross sures capital service, L ti  is labor input, and M ti  is intermediate input.

Annex Figure 3.3.2. Sectoral Labor Productivity Growth, Annex Figure 3.3.3. Skill Composition of Workers by Sector,
2000–10 2000–07
(Difference with respect to economy-wide labor productivity growth, (Share of workers by skill level, percent)
percentage points)
60
Manufacturing
8
Services—high
productivity
50 Services—low
6
productivity
4 40
2 30
0
20
–2
10
–4
0
–6 Low skill Medium skill High skill
Manufacturing
Sales of vehicles
Wholesale trade
Retail trade
Transport
Post & telecommunications
Real estate
Business
Hotels & restaurants
Finance
Community & households
Government
Education
Health

Note: High (low) productivity services correspond to the service industries that
rank in the top third (bottom two-thirds) of the labor-productivity growth
distribution during 2000–10 for the sample used in Figure 3.14, panel 2. The
share of low-, medium-, and high skilled-workers corresponds to the average in
each group of industries over 2000–07 and across the 40 economies included in
the World Input-Output Database. Skill categories are based on the level of
education (see Timmer and others 2015).
Market Nonmarket
services services
yt  = ∑  i yt  i  θ  it  , (3.1)
Note: The figure shows the cross-country distribution across 19 advanced in which y tdenotes the aggregate value added per
economies and 12 emerging market and developing economies of the percentage worker (at constant prices); y ti  is value added per
difference between average labor productivity growth in each sector and
aggregate labor productivity growth over 2000–10. The horizontal line inside each worker in sector i; and θ  it   is the employment share of
box represents the median; the upper and lower edges of each box show the top sector i. The economy-wide growth rate of labor pro-
and bottom quartiles; and the red markers denote the top and bottom deciles.
Nonmarket services consist of government, education, and health. All other ductivity can then be decomposed in two components:
service industries are market services.
y   i 
 
 θ  t − k
g t − k,t = ∑  i     ____
   i       t − k i   
y t − k 
  g 
 t − k,t
    

{
               
 within-sector growth
which includes the remaining service industries. Annex y   i   
  _   (1  + g  ) (θ  it    −  θ  ti − k)
+ ∑  i    t − k
  i 
Figure 3.3.3 shows the skill intensity of employment in y   
t − k
  
 t − k,t   , (3.2)

these two groups of service industries, as well as in man-  
{
  structural change
ufacturing as a whole, based on industry-level data on yt 
where g t − k,t =  ____
y t − k      − 1is aggregate productivity growth
educational attainment from the World Input-Output i   =  ____yt  i 
from period t − kto t; g t − k,t      −  1 is
Database for 40 economies (19 advanced and 21 devel- i 
y t − k   y  i 
t − k 
oping economies) over 2000–07. productivity growth in sector i; and   ____ y t − k 
  is the
relative productivity level in sector i.
Decomposition of Aggregate Labor Productivity The first component measures the “within” effect—
This section describes the decomposition exer- the growth contribution of sector iholding the shares
cise conducted to gauge the contribution of struc- and the level of productivity as in the initial period. The
tural change reported in Figure 3.16. Aggregate second component measures the aggregate contribution
value added per worker can be expressed as the of the movement of workers across sectors with differ-
(employment-share-weighted) average of value added ent levels and growth rates of productivity. Following
per worker in each individual industry: Timmer and others (2015); and Diao, McMillan,

Annex Figure 3.3.4. Sectoral Labor Productivity, 2010

(Ratio with respect to economy-wide labor productivity)
30 1. Emerging Asia 6 20 2. Emerging Latin America and the Caribbean 2.5

18
25 5 16 2.0
20 4 14
12 1.5
15 3 10
8 1.0
10 2 6
5 1 4 0.5
2
0 0 0 0.0
Mining
Utilities
Agriculture
Manufacturing
Construction
Trade &
accommodation
Transport &
communications
Financial &
business
Government
Community &
households
Mining
Utilities
Agriculture
Manufacturing
Construction
Trade &
accommodation
Transport &
communications
Financial &
business
Government
Community &
households
5 3. Emerging Europe 3 30 4. Emerging Sub-Saharan Africa 16
14
4 25
12
2 20
3 10
15 8
2 6
1 10
4
1 5
2
0 0 0 0
Mining
Utilities
Agriculture
Manufacturing
Construction
Trade &
accommodation
Transport &
communications
Financial &
business
Government
Community &
households
Mining
Utilities
Agriculture
Manufacturing
Construction
Trade &
accommodation
Transport &
communications
Financial &
business
Government
Community &
households
Note: The figure shows the distribution across economies in each region of the ratio of labor productivity (value added per worker) in each sector with respect to
economy-wide labor productivity in 2010 (at 2005 prices). The horizontal line inside each box represents the median; the upper and lower edges of each box show
the top and bottom quartiles; and the red markers denote the top and bottom deciles. Values for mining and utilities are reported on the left scale.
and Rodrik (2017), a year-by-year growth rate is first sion rates are needed to convert sectoral value added
calculated (that is, kis set to 1), and then the average expressed in national currencies to units that are com-
annual growth rates for the within and between terms in parable across countries. Using PPP rates for aggregate
a given period T (that is, 2000–10) for each sector are output would be problematic as they fail to capture the
reported by taking the simple average: systematic differences in sectoral relative prices across
y   i 
  countries (Sorensen 2001; Rogerson 2008). Follow-
1    ∑    ∑   θ  i    ____
  =  __
g Twithin t − 1 i
   gt    , (3.3)
t∈T T i t − 1 y  t − 1 ing Rodrik (2013), the baseline analysis is based on
1    ∑   ∑     _ y  i    sectoral labor productivity data. TFP data based on
g Tstructural
  =  __ t∈T i y 
t − 1
   (1  + g ti  )(θ  it    −  θ  t − 1)
i   . (3.4)
T t − 1 sectoral value added are used in robustness exercises.
Nonmarket service industries, such as government,
Cross-Country Productivity Convergence Analysis health, and education, are excluded from the conver-
Data gence exercise because public funding and provision
Testing for productivity convergence across countries make output prices hard to measure (Heston 2013).
requires a cross-comparison of their sectoral produc- The analysis is conducted using two samples that
tivity. Sectoral purchasing-power-parity (PPP) conver- offer different country and sectoral coverage:

Annex Table 3.3.1. Estimation Results, Beta-Convergence—Extended Sample (Nine Sectors)

A. Baseline Specification B1. Baseline Specification B2. Cross-Section Regression
Full Sample Balanced Panel Balanced Panel
Sector Beta S.E. t R 2 Beta S.E. t R 2 Beta S.E. t R 2
Agriculture 0.13 0.18 0.73 0.34 0.27 0.19 1.41 0.44 0.01 0.23 0.04 0.00
Mining –1.31 0.30 –4.39 0.37 –1.00 0.33 –3.03 0.35 –0.67 0.31 –2.16 0.15
Manufacturing –0.87 0.26 –3.37 0.27 –0.58 0.28 –2.07 0.28 –1.16 0.49 –2.39 0.23
Utilities –2.35 0.50 –4.67 0.29 –1.77 0.42 –4.21 0.38 –1.39 0.27 –5.18 0.48
Construction –1.58 0.40 –3.94 0.46 –1.49 0.44 –3.36 0.49 –0.20 0.63 –0.32 0.01
Trade and Accommodation –1.10 0.33 –3.35 0.40 –0.94 0.33 –2.87 0.45 –0.78 0.26 –3.06 0.18
Transport and Communications –1.31 0.45 –2.94 0.19 –1.19 0.47 –2.55 0.18 –0.95 0.40 –2.39 0.24
Financial and Business –1.04 0.35 –2.95 0.13 –0.99 0.36 –2.79 0.13 –1.62 0.42 –3.87 0.50
Community and Households –0.50 0.37 –1.33 0.16 –0.64 0.33 –1.91 0.21 –0.46 0.22 –2.12 0.16
Note: Robust standard errors are reported. Panel A shows the estimation results based on 10-year nonoverlapping windows with decade dummies and an
unbalanced panel of 39 countries. Panel B1 shows the results for a balanced panel of 28 countries. Panel B2 is based on a cross-section regression over the
same sample as panel B1. S.E. = standard errors.
•• Extended sample, based on GGDC and World ˆ   i,t = α + βln Pi,t

P   + Dt  +  ε  i,t, (3.5)
KLEMS data—an unbalanced panel of value added
in which Pˆ   i,tdenotes the trend growth rate of produc-
per worker for nine market sectors for 39 coun-
tivity (labor productivity or TFP) for a given sector in
tries during 1965–2015. Data on sectoral PPP in
country irelative to the United States over the time
2005 are obtained from the GGDC productivity
period t; Pi,t
  is the sector-specific PPP-adjusted produc-
level database.44
tivity level in country irelative to the United States in
•• Reduced sample, based on World KLEMS—an
the initial year of the period; Dt is the period dummy
unbalanced panel covering 26 market sectors during
that controls for common factors; and ε  i,t indicates
1970–2007 with data on value added per hour for
an average relative productivity shock between the
30 countries and data on TFP for 20 countries.
two countries.46 Each period corresponds to a 10-year
Internationally comparable data on sectoral TFP
nonoverlapping window.
and labor productivity levels are from the 1997
A negative estimate of βfor a given sector indicates
benchmark World KLEMS database (for details, see
evidence of productivity convergence across countries:
Inklaar and Timmer 2009).45
the larger the initial gap in productivity between two
countries, the higher the rate of productivity growth in
Following Sorensen (2001), the industry-specific
the country with lower sectoral productivity (relative
productivity growth series for each country are used to
to growth in the higher-productivity country). This
extrapolate the productivity level of 2005 or 1997 over
concept of convergence is known as beta-convergence.
the whole sample period.
The estimation results using the extended sample
Methodology and Baseline Results covering nine market sectors are shown in Annex
Table 3.3.1. The results in panel A are based on the
The baseline specification for testing uncondi-
baseline regression on labor productivity based on
tional productivity convergence in each sector follows
10-year nonoverlapping periods and a broad sam-
the specification in Bernard and Jones (1996) and
ple of 39 countries during 1965–2015 as shown in
Sorensen (2001):
Figure 3.17 (panel 1).47 The estimation results using
a reduced sample covering 26 sectors are reported in
44The internationally comparable level of industry productivity Annex Table 3.3.2, in which panel A corresponds to
is computed as the nominal value added in 2005 per worker in
a given industry-country deflated by the output PPP in the same ˆ  is constructed as the trend coefficient from a regression
46P
i,t
industry-country. See Inklaar and Timmer (2014) for details. of the log level on a constant and a linear trend to minimize the
45The measures of industry productivity from World KLEMS are problems associated with measurement error, business cycles and
adjusted not only by PPPs for gross output but also by PPPs for end-sample issues, as in Bernard and Jones (1996).
intermediate input (a so-called double deflation method). Double 47Excluding observations from 2008 onward, which could reflect
deflation is preferable but is not possible in the GGDC 10-sector developments associated with the global financial crisis, does not
sample due to data limitations. change the results.

Annex Table 3.3.2. Estimation Results, Beta-Convergence—Reduced Sample (26 Sectors)
A. Baseline Specification B1. Baseline Specification B2. Cross-Section Regression C1. Baseline Specification C2. Baseline Specification
Labor Productivity, Labor Productivity, Labor Productivity,
Labor Productivity Balanced Panel Balanced Panel Sample as C2 Total Factor Productivity
Sector Beta S.E. t R 2 Beta S.E. t R 2 Beta S.E. t R 2 Beta S.E. t R 2 Beta S.E. t R 2
Agriculture –0.78 0.38 –2.04 0.37 –0.49 0.35 –1.41 0.47 –0.36 0.32 –1.10 0.08 –0.51 0.46 –1.12 0.57 –0.24 0.49 –0.48 0.58
Mining –1.26 0.52 –2.40 0.52 –0.92 0.63 –1.46 0.53 –0.94 0.62 –1.52 0.16 –1.31 0.82 –1.60 0.49 –2.40 0.95 –2.54 0.47
Utilities –1.60 0.50 –3.21 0.32 –2.09 0.48 –4.38 0.48 –1.78 0.28 –6.32 0.69 –2.10 0.87 –2.42 0.56 –2.99 0.57 –5.23 0.46
Construction –0.44 0.37 –1.18 0.54 –0.49 0.42 –1.18 0.50 –0.94 0.58 –1.61 0.11 0.15 0.56 0.26 0.60 0.10 0.67 0.15 0.57
Manufacturing
Food –0.34 0.42 –0.81 0.15 –0.36 0.36 –1.01 0.21 –1.03 0.59 –1.74 0.15 0.33 0.68 0.48 0.26 –0.11 0.80 –0.14 0.28
Textiles –1.31 0.60 –2.18 0.21 –1.71 0.74 –2.30 0.25 –0.98 0.57 –1.72 0.11 –0.55 0.76 –0.72 0.26 –0.84 0.73 –1.16 0.15
Wood –1.97 0.63 –3.12 0.37 –1.51 0.75 –2.02 0.43 –1.38 0.71 –1.93 0.27 –0.40 0.50 –0.79 0.48 –0.99 0.58 –1.73 0.48
Paper –0.82 0.49 –1.68 0.31 –0.87 0.63 –1.37 0.39 –1.59 0.41 –3.92 0.30 0.50 0.78 0.64 0.36 0.17 0.72 0.24 0.15
Petroleum –0.90 1.40 –0.64 0.63 –2.99 1.04 –2.88 0.82 –3.22 0.52 –6.25 0.64 –1.12 1.29 –0.87 0.72 –0.82 1.11 –0.74 0.72
Chemicals –1.52 0.83 –1.84 0.19 –0.81 0.86 –0.94 0.33 –1.08 0.70 –1.54 0.09 –1.08 0.68 –1.59 0.42 –1.47 1.05 –1.40 0.55
Rubber –0.28 0.65 –0.43 0.25 –0.45 0.46 –0.98 0.39 –0.61 0.53 –1.15 0.07 –0.30 0.67 –0.45 0.38 0.66 0.55 1.21 0.41
Other Mineral –2.57 0.52 –4.93 0.35 –1.67 0.59 –2.80 0.28 –1.23 0.37 –3.31 0.25 –1.24 0.61 –2.04 0.38 –0.99 0.59 –1.67 0.42
Basic Metal –4.08 0.61 –6.69 0.64 –3.22 0.58 –5.58 0.49 –2.32 0.32 –7.16 0.73 –2.97 0.81 –3.65 0.57 –3.40 0.83 –4.08 0.55
Machinery –2.22 0.61 –3.65 0.31 –1.95 0.45 –4.30 0.38 –2.06 0.47 –4.34 0.48 –1.61 0.91 –1.77 0.25 –1.03 1.12 –0.92 0.37
Electrical Equipment –2.52 0.78 –3.22 0.55 –2.96 0.79 –3.75 0.59 –3.82 0.54 –7.02 0.71 –0.04 1.40 –0.03 0.62 0.55 1.72 0.32 0.60
Transport Equipment –0.75 0.57 –1.32 0.17 –0.77 0.41 –1.87 0.30 –0.88 0.23 –3.85 0.25 –0.78 0.46 –1.70 0.45 –0.86 0.53 –1.62 0.40
Recycling –1.46 0.39 –3.79 0.41 –1.03 0.40 –2.60 0.53 –0.86 0.32 –2.66 0.26 –1.02 0.40 –2.57 0.67 –1.19 0.41 –2.89 0.63
Market Services
Trade –0.72 0.31 –2.36 0.49 –0.19 0.29 –0.67 0.54 –0.25 0.32 –0.77 0.02 –1.10 0.46 –2.39 0.59 –1.23 0.45 –2.73 0.59
Hotels and Restaurants –1.97 0.57 –3.49 0.30 –1.43 0.51 –2.77 0.38 –1.05 0.22 –4.82 0.58 –0.41 0.51 –0.81 0.28 –0.46 0.56 –0.82 0.23
Transport –1.15 0.34 –3.41 0.30 –0.86 0.34 –2.50 0.41 –0.92 0.38 –2.41 0.27 –0.50 0.32 –1.57 0.35 –0.32 0.33 –0.95 0.37
Post and Telecomm. –1.23 0.37 –3.36 0.59 –1.41 0.46 –3.05 0.63 –1.22 0.54 –2.25 0.42 –1.13 0.69 –1.63 0.61 –1.45 0.87 –1.67 0.57
Finance –3.90 0.81 –4.81 0.48 –4.64 0.61 –7.60 0.58 –2.46 0.24 –10.15 0.79 –4.59 0.67 –6.90 0.62 –3.95 1.23 –3.22 0.49
Real Estate –0.94 0.47 –2.00 0.33 –0.67 0.48 –1.41 0.36 –0.84 0.26 –3.24 0.18 –0.75 0.46 –1.63 0.56 –1.07 0.66 –1.61 0.45
Business –2.29 0.62 –3.71 0.22 –2.83 0.59 –4.83 0.31 –1.94 0.40 –4.86 0.51 –2.57 0.57 –4.50 0.37 –3.25 0.71 –4.56 0.39
Community –2.19 0.73 –3.02 0.22 –1.69 0.58 –2.92 0.34 –1.31 0.25 –5.19 0.60 –2.03 0.93 –2.18 0.35 –3.81 0.90 –4.23 0.42
Households –1.67 1.58 –1.06 0.11 –3.51 0.65 –5.38 0.33 –2.34 0.31 –7.44 0.81 –4.56 0.75 –6.08 0.38 –5.92 1.19 –4.98 0.61
Sources: Inklaar and Timmer (2009); World KLEMS database; and IMF staff calculations.
Note: Robust standard errors are reported. Panel A is shows the estimation results from the baseline specification (using 10-year nonoverlapping windows with decade dummies) over an unbalanced panel of 30 countries. Panel
B1 shows the results from the baseline specifcication using a balanced panel of 20 countries over 1973–2007, and panel B2 the results based on a cross-section regression over the same sample. Panel C1 shows the results

from the baseline specification using a unbalanced panel of 20 countries over 1970–2007 and panel C2 shows the results from the baseline specification over the same sample but using total factor productivity instead of labor
productivity. See Annex Table 3.1.2 for countries used in each regression. S.E. = standard errors.
165
Annex Figure 3.3.5. Sigma-Convergence

(Log standard deviation of PPP-adjusted labor productivity)
1.24 1. Agriculture 1.45 2. Mining 3. Manufacturing 0.90

1.22
1.40 0.88
1.20
1.18 1.35 0.85
1.16 1.30 0.83
1.14
1.12 1.25 0.81
1.10 1.20 0.79
1.08
1.15 0.76
1.06
1.04 1.10 0.74
1971 80 90 2000 10 1971 80 90 2000 10 1971 80 90 2000 10
1.00 4. Utilities 0.70 5. Construction 6. Trade & Accommodation 0.86

0.95 0.68
0.66
0.90 0.82
0.64
0.85 0.62
0.80 0.60 0.78
0.75 0.58
0.56
0.70 0.74
0.54
0.65 0.52
0.60 0.50 0.70
1971 80 90 2000 10 1971 80 90 2000 10 1971 80 90 2000 10
0.76 7. Transport & Communications 1.25 8. Financial & Business 9. Community & Households 1.00
0.74 1.20
0.95
0.71 1.15
0.69 1.10 0.90
0.66 1.05
0.85
0.64 1.00
0.61 0.95 0.80
0.59 0.90
0.75
0.56 0.85
0.54 0.80 0.70
1971 80 90 2000 10 1971 80 90 2000 10 1971 80 90 2000 10
Sources: World KLEMS database; and IMF staff calculations.

Note: Each graph shows the cross-country log standard deviation of purchasing-power-parity-adjusted labor productivity (sigma-convergence) for each of the
nine market sectors available in the Groningen Growth and Development Centre 10-sector database. PPP = purchasing power parity.
the results based on labor productivity on 10-year shocks to the growth process are relatively large,
nonoverlapping periods during 1970–2010, as shown sigma-convergence may not be achieved (Young,
in Figure 3.17 (panel 2).48 Higgins, and Levy 2008). Annex Figure 3.3.5 shows
Beta-convergence across sectors is a necessary but the standard deviation of log labor productivity across
not sufficient condition for convergence in pro- countries for each of the nine market sectors in the
ductivity levels. Even if beta-convergence holds, if GGDC 10-sector database.
Robustness Exercises
48Annex Tables 3.3.1 and 3.3.2 report robust standard errors.
Clustering standard errors at the country level does not change the
Mean Reversion. One concern with using 10-year
results of the analysis. nonoverlapping windows is that the evidence on con-

vergence may reflect mean reversion over the business that surveys are conducted at irregular time intervals
cycle. Panel C of Annex Table 3.3.1 shows the results that differ across countries, the cross-country statistics
from a robustness exercise using a cross-section anal- reported in the chapter correspond to the latest survey
ysis over a balanced panel. For each sector, the trend year available for each country within a specific range,
growth rate of labor productivity over the period as noted in the analysis. Because of data limitations,
1965–2010 is regressed on its level in 1965 and a three broad sectors are considered: agriculture, industry
constant. The results reported in panel B2 confirm (which consists of manufacturing, mining, electricity
that the baseline results are not due to mean rever- and construction), and services.50
sion (panel B1 shows for comparability purposes the The analysis on labor earnings across sectors and
results from the baseline specification but using the skills reported in Figure 3.20 uses LIS data on gross
same balanced sample). hourly wage of full-time working household mem-
The robustness exercise for the reduced (26 sectors) bers for whom skill data are available.51 The skill
sample is shown in Annex Table 3.3.2, panel B2 (panel levels—high, medium, and low—are based on the
B1 shows for comparability purposes the results from International Standard Classification of Occupations
the baseline specification but using the same balanced (ISCO) classification of occupations into manag-
sample). For each sector, the trend growth rate of labor ers and professionals (ISCO 1 and 2), other skilled
productivity over the period 1973–2007 is regressed workers (ISCO 3–8, 10), and laborers/elementary
on its level in 1973 and a constant.49 The results are (ISCO 9), respectively. Average gross hourly wages
broadly unchanged. for each sector-skill are expressed relative to average
Total Factor Productivity. Given that changes economy-wide wages.
in labor productivity may reflect capital deepen- The measure of inequality used is the Generalized
ing rather than actual productivity, an additional Entropy (GE[0]) index, or mean log deviation, which
robustness exercise uses the standard TFP instead of has the advantage of being decomposable, unlike the
labor productivity. The results are reported in Annex Gini coefficient (Shorrocks 1980; Mookherjee and
Table 3.3.2, panel C2 (panel C1 shows for compara- Shorrocks 1982). The mean log deviation, or GE(0), is
bility purposes the results using labor productivity over given by:52
the same sample). The results are similar to those based
GE(0) = −  __n1   ∑ i ln(  _ y ¯i   ),
y 
on labor productivity in a comparable sample: (1) (3.6)
several manufacturing industries show no evidence of in which nis the number of households, y i is income
beta-convergence, and (2) several service sectors show   ¯ is the mean of y i.
of household i, and y 
significant evidence of beta-convergence. The economy-wide GE(0) index can be decomposed
as a weighted sum of the extent of inequality in each
Annex 3.4. Manufacturing and Inequality sector (within-sector inequality) and the contribution
arising from differences between average incomes
This annex provides additional details on the analy-
across sectors (between-sector inequality):
sis carried out in the section “Implications for Income
Inequality.”
The analysis in the section on income inequal-
ity uses micro data from the Luxembourg Income
Study (LIS) to compute labor earnings inequality at 50The information on the sector of employment might not be avail-
the sectoral and the aggregate level. The LIS offers able if the household head is unemployed or out of the labor force, or
if the data are missing. In those cases, the household is assigned to a
survey-based data at the household and personal level separate “missing data” sector to ensure that the aggregate inequality
on labor income, employment sector, and occupa- measure is calculated for the entire population and the sum of the
tion for a broad set of countries, including a set of components equals the economy-wide level of inequality.
51The hourly wages are top and bottom coded to address extreme
advanced economies since the early 1980s. Given
values. At the bottom, negative or zero wages are set to “missing.”
At the top, wages greater than 10 times the median for a given
491973 is chosen as the initial year to maximize the country country-year are set to 10 times the median.
52The general formula for Generalized Entropy is
coverage, as data for Japan starts from only 1973. Given that data
y  α
nα(α  −  1) i [(  y 
for countries in Central and Eastern Europe are not available before GE(α) =  _______1      ¯i   )   − 1],
  ∑   _
1995, this cross-sectional analysis includes fewer country observa-
tions (20 countries). when α ≠ 0,1.When α = 0, GE is defined as in equation 3.6.

GE(0) = ∑ k  v k GE
  
(0   + ∑ k v k ln(   ,
)  k  _    ) 
1   (3.7) Autor, David H. 2015. “Why Are There Still So Many Jobs? The
λ  k
History and Future of Workplace Automation.” Journal of
{
{
within between Economic Perspectives 29 (3): 3–30.
 k
n Autor, David H., David Dorn, and Gordon H. Hanson. 2013.
in which v k =  __ n   is the population share of sector
 ¯ k
 __
y  “The China Syndrome: Local Labor Market Effects of Import
k, and λ  k =    y ¯    is the relative mean income of sector
Competition in the United States.” American Economic Review
k. The sector of employment of the household head
103 (6): 2121–68.
is used to calculate inequality at the sector level. ––––––. 2016. “The China Shock: Learning from Labor-Market
The cross-sector average income differences reported Adjustment to Large Changes in Trade.” Annual Review of
in Figure 3.21 correspond to the between-sector Economics 8: 205–40.
inequality term. ––––––. 2017. “When Work Disappears: Manufacturing
Changes in inequality over time can be analyzed Decline and the Falling Marriage-Market Value of Young
by applying the difference operator to both sides of Men.” NBER Working Paper 23173, National Bureau of
equation 3.7: Economic Research, Cambridge, MA.
Autor, David H., Lawrence F. Katz, and Melissa S. Kearney.
GE(0)  t + 1  − GE(0)  t = ∑ k  v k,t ∆ GE(0)  k  2006. “The Polarization of the US Labor Market.” American
Economic Review 96 (2): 189–94.
+ ∑ k  GE(0)  k,t + 1 ∆ v k  Baldwin, Richard. 2016. “Factory-Free Europe?” In The
Factory-Free Economy: Outsourcing, Servitization, and The
− ∑ k  ln(λ  k,t + 1) ∆ v k 
Future of Industry, edited by Lionel Fontagné and Ann Harri-
− ∑ k v k,t ∆ ln( λ  k). (3.8) son. Oxford: Oxford University Press.
Baldwin, Richard, Rikard Forslid, and Tadashi Ito. 2015. “Unveil-
Equation 3.8 is an exact decomposition of the ing the Evolving Sources of Value Added in Exports.” Papers
change in generalized entropy over time into four and Reports: Joint Research Program Series 161, Institute of
terms that can be interpreted as (1) the effect of inter- Developing Economies–Japan External Trade Organization.
temporal changes in within-sector inequality, (2) the Bárány, Zsófia L., and Christian Siegel. 2018. “Job Polarization
effect of changes in sectoral employment shares on the and Structural Change.” American Economic Journal: Macro-
“within” component, (3) the effect of changes in sec- economics 10 (1): 57–89.
Barro, Robert J., and Xavier Sala-i-Martin. 1992. “Conver-
toral employment shares on the “between” component,
gence.” Journal of Political Economy 100 (2): 223–51.
and (4) the effect of changes in the relative average sec-
Baumol, William J. 1967. “Macroeconomics of Unbalanced
toral income levels (Mookherjee and Shorrocks 1982).
Growth: The Anatomy of Urban Crisis.” American Economic
In the analysis reported in Figure 3.22, the second and Review 57 (3): 415–26.
third terms are added and referred to as “changes in Baumol, William J., Sue Anne Batey Blackman, and Edward
sector size.” N. Wolff. 1985. “Unbalanced Growth Revisited: Asymptotic
Stagnancy and New Evidence.” American Economic Review 75
(4): 806–17.
References Berlingieri, Giuseppe. 2014. “Outsourcing and the Rise in
Acemoglu, Daron, and Pascual Restrepo. 2017. “Robots and Services.” CEP Discussion Paper 1199, Centre for Economic
Jobs: Evidence from US Labor Markets.” NBER Work- Performance, London School of Economics.
ing Paper 23285, National Bureau of Economic Research, Bernard, Andrew, and Charles Jones. 1996. “Comparing Apples
Cambridge, MA. to Oranges: Productivity Convergence and Measurement
Adler, Gustavo, Romain Duval, Davide Furceri, Sinem Kiliç across Industries and Countries.” The American Economic
Çelik, Ksenia Koloskova, and Marcos Poplawski-Ribeiro. Review 86 (5): 1216–238.
2017. “Gone with the Headwinds: Global Productivity.” Bernard, Andrew, Valerie Smeets, and Frederic Warzynski.
Staff Discussion Note 17/04, International Monetary Fund, 2017. “Rethinking Deindustrialization.” Economic Policy
Washington, DC. 32 (89): 5–38.
Alvarez, Jorge. 2017. “Structural Transformation and the Agri- Boppart, Timo. 2014. “Structural Change and the Kaldor
cultural Wage Gap.” IMF Working Paper 17/289, Interna- Facts in a Growth Model with Relative Price Effects and
tional Monetary Fund, Washington, DC. Non-Gorman Preferences.” Econometrica 82 (6): 2167–196.
Alvarez, Jorge, Felipe Benguria, Christian Moser, and Niklas Bosworth, Barry P., and Jack E. Triplett. 2003. “Services Produc-
Engbom. 2018. “Firms and the Decline in Earnings Inequal- tivity in the United States: Griliches’ Services Volume Revis-
ity in Brazil.” American Economic Journal: Macroeconomics ited.” Paper prepared for CRIW Conference in Memory of Zvi
10 (1): 49–89. Griliches, Brookings Institution, Washington, DC, September.

––––––. 2007. “Services Productivity in the United States: Grili- Francois, Joseph, and Bernard Hoekman. 2010. “Services Trade
ches.” In Hard-to-Measure Goods and Services: Essays in Honor and Policy.” Journal of Economic Literature 48 (3): 642–92.
of Zvi Griliches, edited by Ernst R. Berndt and C. R. Hulten. Gervais, Antoine, and J. Bradford Jensen. 2014. “The Trad-
Chicago: University of Chicago Press. ability of Services: Geographic Concentration and Trade
Bourlès, Renaud, Gilbert Cette, Jimmy Lopez, Jacques Mairesse, Costs.” Discussion Paper CES 14–03, US Census Bureau,
and Giuseppe Nicoletti. 2013. “Do Product Market Regula- Washington, DC.
tions in Upstream Sectors Curb Productivity Growth? Panel Goos, Maarten, Alan Manning, and Anna Salomons. 2014.
Data Evidence for OECD Countries.” Review of Economics “Explaining Job Polarization: Routine-Biased Technological
and Statistics 95 (5): 1750–768. Change and Offshoring.” The American Economic Review 104
Buera, Francisco J., and Joseph P. Kaboski. 2009. “Can Tradi- (8): 2509–526.
tional Theories of Structural Change Fit the Data?” Journal of Gordon, Robert J. 1996. “Problems in the Measurement and
the European Economic Association 7 (2–3): 469–77. Performance of Service-Sector Productivity in the United
––––––. 2012. “The Rise of the Service Economy.” American States.” NBER Working Paper 5519, National Bureau of
Economic Review 102 (6): 2540–569. Economic Research, Cambridge, MA.
Copeland, Brian, and Aaditya Mattoo. 2007. “The Basic Eco- Hallward-Driemeier, Mary, and Gaurav Nayyar. 2017. “Trouble
nomics of Services Trade.” In A Handbook of International in the Making? The Future of Manufacturing-Led Develop-
Trade in Services. Oxford: Oxford University Press. ment.” World Bank, Washington, DC.
Crozet, Matthieu D., and Emmanuel Milet. 2017. “The Servi- Helper, Susan, Timothy Krueger, and Howard Wial. 2012.
tization of French Manufacturing Firms.” In The Factory-Free “Why Does Manufacturing Matter? Which Manufacturing
Economy: Outsourcing, Servitization, and the Future of Indus- Matters?” Brookings: Metropolitan Policy Program.
try, edited by Lionel Fontagné and Ann Harrison. Oxford: Herrendorf, Berthold, Richard Rogerson, and Ákos Valentinyi.
Oxford University Press. 2013. “Two Perspectives on Preferences and Structural Trans-
Dasgupta, Sukti, and Ajit Singh. 2006. “Manufacturing, formation.” The American Economic Review 103 (7): 2752–789.
Services, and Premature Deindustrialization in Developing –––––. 2014. “Growth and Structural Transformation.” In
Countries: A Kaldorian Analysis.” UNU-WIDER, United Handbook of Economic Growth, edited by Philippe Aghion
Nations University Research Paper 2006/49. and Steven N. Durlauf. Amsterdam: Elsevier.
Diao, Xinshen, Margaret McMillan, and Dani Rodrik. 2017. Heston, Alan. 2013. “Government Services: Productivity Adjust-
“The Recent Growth Boom in Developing Economies: A ments.” In Measuring the Real Size of the World Economy:
Structural Change Perspective.” NBER Working Paper 23132, The Framework, Methodology, and Results of the International
National Bureau of Economic Research, Cambridge, MA. Comparison Program—ICP. Washington, DC: World Bank.
Dias, Daniel A., Carlos Robalo Marques, and Christine Rich- Heuser, Cecilia, and Aaditya Mattoo. 2017. “Services Trade and
mond. 2016. “Misallocation and Productivity in the Lead Up Global Value Chains.” Policy Research Working Paper Series
to the Eurozone Crisis.” Journal of Macroeconomics 49: 46–70. 8126, World Bank, Washington, DC.
Duarte, Margarida, and Diego Restuccia. 2017. “Relative Prices Hsieh, Chang-Tai, and Peter J. Klenow. 2009. “Misallocation
and Sectoral Productivity.” NBER Working Paper 23979, and Manufacturing TFP in China and India.” Quarterly
National Bureau of Economic Research, Cambridge, MA. Journal of Economics 124 (4): 1403–448.
Duernecker, Georg, Bethold Herrendorf, and Ákos Valentinyi. Huneeus, Federico, and Richard Rogerson. 2016. “Industrializa-
2017. “Structural Change within the Service Sector and the tion among Early and Late Developers.” Manuscript.
Future of Baumol’s Disease.” Manuscript. Imbs, Jean. 2016. “Structural Change in the OECD: Some
Engbom, Niklas, and Christian Moser. 2018. “Earnings Inequal- Facts.” In The Factory-Free Economy: Outsourcing, Servitiza-
ity and the Minimum Wage: Evidence from Brazil.” CESifo tion, and the Future of Industry, edited by Lionel Fontagné
Working Paper Series 6393, CESifo Group, Munich. and Ann Harrison. Oxford: Oxford University Press.
Engel, Ernst. 1895. “Die Lebenskosten Belgischer Inklaar, Robert, and Marcel P. Timmer. 2009. “Productivity
Arbeiter-Familien Früher und Jetzt.” C. Heinrich. Convergence across Industries and Countries: The Impor-
Felipe, Jesus, and Aashish Mehta. 2016. “Deindustrialization? A tance of Theory-Based Measurement.” Macroeconomic Dynam-
Global Perspective.” Economics Letters 149: 148–51. ics 12 (S2): 218–40.
Fernandes, Ana M., and Caroline Paunov. 2012. “Foreign Direct –––––. 2014. “The Relative Price of Services.” Review of Income
Investment in Services and Manufacturing Productivity: and Wealth 60 (4): 727–46.
Evidence for Chile.” Journal of Development Economics 97 Jaumotte, Florence, and Carolina Osorio Buitron. 2015.
(2): 305–21. “Inequality and Labor Market Institutions.” IMF Staff
Fournier, Jean-Marc, and Åsa Johansson. 2016. “The Effect of Discussion Note 15/14, International Monetary Fund,
the Size and the Mix of Public Spending on Growth and Washington, DC.
Inequality.” OECD Economics Department Working Paper Johnson, Simon, Jonathan D. Ostry, and Arvind Subrama-
1344, OECD Publishing, Paris. nian. 2007. “The Prospects for Sustained Growth in Africa:

Benchmarking the Constraints.” IMF Working Paper 07/52, McMillan, Margaret, Dani Rodrik, and Íñigo Verduzco-Gallo.
International Monetary Fund, Washington, DC. 2014. “Globalization, Structural Change, and Productivity
Jones, Benjamin F., and Benjamin A. Olken. 2005. “The Anat- Growth, with an Update on Africa.” World Development
omy of Start-Stop Growth.” NBER Working Paper 11528, 63: 11–32.
National Bureau of Economic Research, Cambridge, MA. Miroudot, Sébastien, Jehan Sauvage, and Ben Shepherd. 2013.
Jorgenson, Dale W., Mun S. Ho, and Jon Samuels. 2013. “A “Measuring the Cost of International Trade in Services.”
Prototype Industry-Level Production Account for the United World Trade Review 12 (4): 719–35.
States, 1947–2010.” In Second World KLEMS Conference. Miroudot, Sébastien, and Charles Cadestin. 2017. “Services
Cambridge: Harvard University. in Global Value Chains: From Inputs to Value-Creating
Jorgenson, Dale. W., and Marcel Timmer. 2011. “Structural Activities.” OECD Trade Policy Papers 197, OECD Pub-
Change in Advanced Nations: A New Set of Stylized Facts.” lishing, Paris.
Scandinavian Journal of Economics 113 (1): 1–29. Mookherjee, Dilip, and Anthony Shorrocks. 1982. “A Decom-
Kaldor, Nicholas. 1967. “Strategic Factors in Economic Develop- position Analysis of the Trend in UK Income Inequality.” The
ment.” Ithaca: New York State School of Industrial and Labor Economic Journal 92 (368): 886–902.
Relations, Cornell University. National Board of Trade of Sweden. 2010. “Servicification of
Kongsamut, Piyabha, Sergio Rebelo, and Danyang Xie. 2001. Swedish Manufacturing.” National Board of Trade, Sweden.
“Beyond Balanced Growth.” Review of Economic Studies Ngai, L. Rachel, and Christopher A. Pissarides. 2007. “Struc-
68 (4): 869–82. tural Change in a Multisector Model of Growth.” American
Koske, Isabell, Isabelle Wanner, Rosamaria Bitetti, and Omar Economic Review 97 (1): 429–43.
Barbiero. 2015. “The 2013 Update of the OECD’s Database Pilat, Dirk, and Anita Wölfl. 2005. “Measuring the Interaction
on Product Market Regulation: Policy Insights for OECD between Manufacturing and Services.” OECD Science,
and Non-OECD Countries.” OECD Economics Department Technology, and Industry Working Paper 2005/05, OECD
Working Paper 1200, OECD Publishing, Paris. Publishing, Paris.
Kuznets, Simon. 1966. Economic Growth and Structure. Ports- Rodrik, Dani. 2013. “Unconditional Convergence in Manufac-
mouth, NH: Heinemann Educational Books. turing.” Quarterly Journal of Economics 128 (1): 165–204.
Lanau, Sergi, and Petia Topalova. 2016. “The Impact of –––––. 2016. “Premature Deindustrialization.” Journal of Eco-
Product Market Reforms on Firm Productivity in Italy.” nomic Growth 21: 1–33.
IMF Working Paper 16/119, International Monetary Fund, Rogerson, Richard. 2008. “Structural Transformation and the
Washington, DC. Deterioration of European Labor Market Outcomes.” Journal
Langdon, David, and Rebecca Lehrman. 2012. “The Benefits of of Political Economy 116 (2): 235–59.
Manufacturing Jobs.” United States Department of Com- Shorrocks, Anthony. 1980. “The Class of Additively Decompos-
merce, Economics and Statistics Administration Working able Inequality Measures.” Econometrica 48 (3): 613–25.
Paper 01–12, Washington, DC. Solt, Frederick. 2016. “The Standardized World Income Inequal-
Lawrence, Robert Z. 2017. “Recent Manufacturing Employment ity Database.” Social Science Quarterly 97 (5): 1267–81.
Growth: The Exception That Proves the Rule.” NBER Work- Sorensen, Anders. 2001. “Comparing Apples to Oranges: Productiv-
ing Paper 24151, National Bureau of Economic Research, ity Convergence and Measurement across Industries and Coun-
Cambridge, MA. tries: Comment.” American Economic Review 91 (4): 1160–67.
Lodefalk, Magnus. 2013. “Servicification of Manufacturing— Stiroh, Kevin J. 2002. “Information Technology and the US Pro-
Evidence from Sweden.” International Journal of Economics ductivity Revival: What Do the Industry Data Say?” American
and Business Research 6 (1): 87–113. Economic Review 92 (5): 1559–576.
Loungani, Prakash, Saurabh Mishra, Chris Papageorgiou, and Ke Swiecki, Tomasz. 2017. “Determinants of Structural Change.”
Wang. 2017. “World Trade in Services: Evidence from A New Review of Economic Dynamics 24: 95–131.
Dataset.” IMF Working Paper 17/77, International Monetary Timmer, Marcel. P., Erik Dietzenbacher, Bart Los, Robert Stehrer,
Fund, Washington, DC. and Gaaitzen J. de Vries. 2015. “An Illustrated User Guide to the
Matsuyama, Kiminori. 2009. “Structural Change in an World Input–Output Database: The Case of Global Automotive
Interdependent World: A Global View of Manufactur- Production.” Review of International Economics 23 (3): 575–605.
ing Decline.” Journal of the European Economic Associa- Triplett, Jack E., and Barry P. Bosworth. 2000. “Productivity in
tion 7 (2–3): 478–86. the Services Sector.” In Presentation at the American Economic
McMillan, Margaret, and Dani Rodrik. 2011. “Globalization, Association Meetings.
Structural Change, and Productivity Growth.” NBER Work- Uy, Timothy, Kei-Mu Yi, and Jing Zhang. 2013. “Structural
ing Paper 17143, National Bureau of Economic Research, Change in an Open Economy.” Journal of Monetary Econom-
Cambridge, MA. ics 60 (6): 667–82.

Verma, Rubina. 2012. “Can Total Factor Productivity Explain Young, Alwyn. 2014. “Structural Transformation, the Mismea-
Value Added Growth in Services?” Journal of Development surement of Productivity Growth, and the Cost Disease of
Economics 99 (1): 163–77. Services.” American Economic Review 104 (11): 3635–667.
Walker, W. Reed. 2013. “The Transitional Costs of Sectoral Young, Andrew T., Matthew J. Higgins, and Daniel Levy. 2008.
Reallocation: Evidence from the Clean Air Act and the Work- “Sigma Convergence versus Beta Convergence: Evidence
force.” Quarterly Journal of Economics 128 (4): 1787–835. from US County-Level Data.” Journal of Money, Credit and
Wood, Adrian. 2017. “Variation in Structural Change around Banking 40 (5): 1083–903.
the World, 1985–2015: Patterns, Causes, and Implications.”
WIDER Working Paper 2017/34.

4
CHAPTER
IS PRODUCTIVITY GROWTH SHARED IN A GLOBALIZED ECONOMY?
How easily do knowledge and technology flow across Against this background, this chapter takes a closer
countries? Has this relationship changed over the past look at the process of international technology dif-
decades, a period when the world has become economically fusion. It examines whether globalization means that
more integrated and the international competition land- knowledge from technology leaders is spreading faster
scape transformed? And did this help productivity growth, than it used to, and how this impacts the capacity of
both at the country and the global level? These questions other economies to innovate and be more productive.
are important because technology tends to advance at The methodology also lends itself to discussing the
different speeds across countries. As a consequence, making influence of another aspect of globalization—increased
new technologies more widely available creates opportu- international competition. Better understanding of
nities for raising productivity and incomes. Against this how productivity growth is shared across the global
backdrop, the chapter offers new empirical evidence on economy can help explain cross-country differences in
the evolution of international technology diffusion and its income per capita and technology and shed light on
impact on productivity. Using a rich data set on patents, the policies that can influence them.
cross-patent citations, research and development spending, Specifically, the chapter will ask:
and productivity, it finds that globalization has indeed •• How has the technological innovation land-
intensified the global diffusion of knowledge and technology scape evolved?
and helped spread growth potential across countries. The •• How strong is the diffusion of knowledge across
positive impact has been particularly strong for emerging countries? Has knowledge become more globalized?
market economies, fostering cross-country income conver- •• Do foreign knowledge flows increase domestic inno-
gence, thanks to their increased use of the available foreign vation and productivity, both in advanced econo-
knowledge. But technology leaders can also benefit from the mies and emerging market economies?
innovation of others. The right set of policies maximizes •• What impact does greater international competition
benefits for all involved, including policies to enhance inter- have on innovation and technology diffusion?
connectedness and build absorptive capacity. An appropriate •• Which policies help increase inward technol-
degree of protection for intellectual property rights is key ogy diffusion?
to preserve the ability of innovators to recover costs while
ensuring that new knowledge supports growth globally. To answer these questions, the chapter exploits a
high-quality micro patenting data set, the Worldwide
Patent Statistical Database (PATSTAT). The database,
Introduction which is maintained by the European Patent Office,
Technology is a key driver of improvements in can be used to construct measures of technological
income and standard of living. Historically, techno- innovation (patenting) and diffusion (cross-patent cita-
logical developments have been concentrated in a few tions) across countries and across different sectors.1
large industrialized economies (Figure 4.1). There- Use of patent and research and development (R&D)
fore, the way technology diffuses across countries is data allows precise identification of knowledge gener-
central to how global growth is generated and shared ation and diffusion. At the same time, these data have
across countries. Globalization has likely changed limits in that not all innovations are patented. Innova-
the diffusion process, with a large body of literature tions in services, for example, are less patentable and
highlighting the importance of trade and foreign direct typically are protected through forms of intellectual
investment (Keller 2004, 2010). property that tend to be more difficult to document
across countries and over time. Therefore, the patent
The authors of this chapter are Aqib Aslam, Federica Coelli,
Johannes Eugster, Giang Ho, Florence Jaumotte (team leader), 1For previous work using patents or citations data, see Branstetter
Carolina Osorio Buitron, and Roberto Piazza, with support from (2001); Peri (2005); MacGarvie (2006); Madsen (2007); and
Pankhuri Dutt, Chanpheng Fizzarotti, and Menexenia Tsaroucha. Aghion, Howitt, and Prantl (2015).

Figure 4.1. International Patent Families by Publication Year industrialized economies. Innovation sharing has taken
(Average 1995–2014)
place through many channels, including the interna-
tional use of patents and trade. Another mechanism
through which globalization appears to have boosted
Japan the diffusion of knowledge and technology is by increas-
Other ing international competition, which in turn has raised
incentives to innovate and adopt foreign technologies.
By making increasing use of available foreign knowl-
edge and technology, emerging market economies have
boosted their own innovation activity and lifted pro-
ductivity. Indeed, increased diffusion of knowledge to
emerging market economies has partly offset the effects
of the recent slowdown in innovation at the technology
frontier. More intense diffusion of leading technologies
EU G–3 United States to emerging market economies helps explain why their
productivity growth has generally been stronger than
in advanced economies, helping to drive cross-country
income convergence for many countries in recent years.
Sources: European Patent Office, PATSTAT database; and IMF staff calculations. The effects have been substantial: over 2004–14, knowl-
Note: EU G–3 = France, Germany, and the United Kingdom.
edge and technology flows from the global frontier
explain about 40 percent of average sectoral productivity
analysis in this chapter is complemented by an examina- growth in emerging market economies.
tion of productivity measures to establish whether the Finally, knowledge and technology do not flow only
identified patterns of international technology diffusion in one direction—indeed, the chapter finds evidence
are accurate indicators of productivity developments. that technology leaders themselves benefit from each
The first part of the chapter lays out a conceptual other’s innovation. This underlines the production
model for the production and diffusion of innovation. and diffusion of knowledge and technology as a key
It also documents trends in R&D, patenting, and pro- mechanism through which globalization delivers global
ductivity, both at the technology frontier and in other benefits. And even though until recently much of the
advanced and emerging market economies. The strength production of knowledge and technology was con-
of international technology diffusion and its effects on centrated in a small number of advanced economies,
productivity are then examined, with estimates of the China and Korea have now emerged as significant con-
impact of technology leaders’ knowledge flows on inno- tributors to the global technology frontier. Therefore,
vation and productivity in economies that are recipients there may be scope in the future for spillovers from
of that knowledge. Because global value chains (GVCs) these new innovators to the traditional innovators.
are a potentially important channel of knowledge spill- This chapter is a contribution to the ongoing debate
overs, the analysis is complemented by a detailed look at on the benefits and drawbacks of globalization. While
their effect on technology diffusion in emerging market the negative side effects of globalization have received
economies. The final part of the chapter discusses the much attention in public debates, the chapter high-
complex relationship between international competition, lights that there are upsides too: globalization helps the
market concentration, and innovation. It provides some diffusion of knowledge and technology across borders,
evidence of the impact of such structural changes on spreading their benefits more globally. From a policy
innovation and technology diffusion. perspective, greater global interconnectedness is thus key
The findings of the chapter show that globaliza- to maximizing inward technology diffusion and boost-
tion has intensified the diffusion of knowledge and ing economies’ growth potential. But as economists
technology across borders, helping to spread potential have long emphasized, assimilating and productively
growth among countries and boost it at the global level. using foreign knowledge often requires investments in
This productivity spillover is important because, until domestic R&D and in human capital, which enhance
recently, the production of knowledge and technology absorptive capacity (for example, Cohen and Levinthal
has been concentrated mostly in a handful of large 1989; Griffith, Redding, and Van Reenen 2004).

CHAPTER 4 Is Productivity Growth Shared in a Globalized Economy?
Figure 4.2. Technology Diffusion
Domestic
Research and Development Innovation
Foreign
Research and Development Use Productivity
Knowledge Flows
Source: IMF staff illustration.
The chapter provides some evidence suggesting that •• Available foreign knowledge: A common measure
strong institutions that uphold the rule of law benefit is the cumulated stock of past R&D spending,
innovation, but it does not examine specifically the corrected for the loss of some of the knowledge’s rel-
optimal extent of intellectual property rights protection, evance over time (see Annex 4.1). This is the main
which includes patents. This is a complex issue and could measure of foreign knowledge used in the analysis.
not be dealt with conclusively at this chapter’s broad •• Extent of use of the stock of foreign knowledge: Foreign
level of analysis. Protection for innovators’ ideas provides knowledge is transmitted internationally through
appropriate incentives and the ability to recover costs. But various channels. The strength of this transmission
the policy design should maintain sufficient competition determines to what extent foreign knowledge is
and allow for follow-on innovations by competitors, domestically usable. However, measuring transmis-
as well as prevent the abuse of power to the detriment sion is difficult. The main channels mentioned in
of consumers. Finally, concerns that globalization may the literature are foreign direct investment (FDI),
exacerbate inequality within countries also apply to the international trade, and migration (see Keller 2004
growth benefit from inward technology diffusion. It is and 2010 for an extensive discussion of the empir-
therefore important for policymakers to ensure that these ical evidence).3 Within these channels, knowledge
growth benefits are shared broadly across the population. flows can entail market transactions—for instance,
trade or the licensed use of foreign patents—or occur
Conceptual Framework through demonstration effects and outright copying of
patented or nonpatented foreign innovations that have
Domestic innovation draws on knowledge generated
become domestically available. In this case, knowledge
by domestic and foreign research efforts (Figure 4.2).2
flows incorporate a significant externality component.
While domestic R&D can affect domestic innovation
•• Impact of foreign knowledge flows on the production of
directly, it is useful to distinguish the steps through
domestic innovation and on the economy’s productiv-
which foreign knowledge influences domestic innova-
ity: Foreign knowledge flows—as measured by the
tion: the availability of foreign knowledge, the extent of
product of the available foreign knowledge and the
its use domestically, and the impact of knowledge flows
on domestic innovation and productivity more generally. 3Most empirical studies test only one channel at a time. In prac-
tice, all the channels are correlated, making it difficult to disentangle

individual contributions. Testing for the role of trade or FDI is also
2See Grossman and Helpman (1991) for models of endogenous subject to endogeneity concerns, as trade and FDI linkages with
growth, based on the idea that knowledge gained from past research technology leaders will likely be influenced by the innovativeness or
efforts increases the productivity of current research efforts. productivity of the country examined.

extent to which that stock of knowledge is used—do An international patent family features one patent
impact domestic innovation. They can also con- application in at least two distinct patent offices. The
tribute to raising domestic productivity, not only idea is to exclude many patents with lower economic
by boosting domestic innovation, but also directly value, as the low expected payoff would not warrant
through the adoption of foreign technologies in the the extra cost of application, examination, and mainte-
production process (for example, through the licens- nance in a foreign country. The approach also reduces
ing of foreign technology or technology embodied the impact of possible idiosyncrasies in patenting
in imports or FDI). activity across patent offices.
The top three patent families include an application
to at least one of the top three patent offices (Euro-
Measuring Innovation
pean Patent Office, Japan Patent Office, United States
Measuring innovation is no simple task. This Patent and Trademark Office). Relative to the previous
section discusses the advantages and limitations of the measure, this implies more consistency as it involves a
approach taken in the chapter. The analysis is centered very limited number of patent offices. The drawback is
on two variables widely used in the literature: R&D that count measures tend to favor inventors and appli-
spending and patent data. These measures have two cants from Europe, Japan, and the United States.
advantageous attributes: In recognition that there is no perfect measure, the
•• Direct quantification of innovation activity: R&D empirical analyses—which use sector- or firm-level
spending captures firms’ research input. Patent data data for each country—include country-year fixed
are a measure of the outcome of research activity. To effects to absorb the fiscal, institutional, cultural, and
be patentable, an idea needs to be novel, inventive legal factors that affect incentives to patent or cite
(“non-obvious to persons skilled in the art”), and other patents across countries and time.4
capable of industrial application (OECD 2009). A second drawback of using patent data is that not
Both variables are available internationally and at all innovations are patentable. Certain sectors, such as
disaggregated levels and can be used to study the manufacturing, display more patentability than others—
strength of the innovation link between industries such as services, which rely more on forms of intellectual
and across countries. property protection that are less systematically recorded.5
•• A proxy for the domestic use of foreign knowledge: Patent This and related data issues make it hard to investigate
citations provide a direct way to quantify the strength technology diffusion in nonmanufacturing activities,
of international knowledge flows—the extent to and suggest a focus on manufacturing sectors. Therefore,
which recipient countries actually make use of the the degree to which this chapter’s results extend to other
available stock of global knowledge. Citation data are sectors depends on how well patenting correlates with
readily available, thanks to the need for precise and overall innovation activities, including those that do not
comprehensive citations for patent registration. lead to patenting. While impossible to test precisely, some
support is found for this assumption.6 Nevertheless, mac-
Nevertheless, patent and R&D measures have their roeconomic interpretation requires some care.
limitations. First, patenting can be a noisy measure
of innovation capacity. There are multiple reasons 4This would also address the case where local firms have a lower
propensity to patent either domestically, because the actual protec-
why the incentive to patent an innovation can differ
tion of patents in the domestic economy is weak, or internationally,
between countries and across time, including differ- because the domestic market is large enough that they do not need
ences in the procedures and requirements of patent to patent abroad. Similar points can also be made for R&D spend-
offices. As a result, the number or economic value of ing, since incentives to precisely measure and classify innovation
efforts are subject to significant heterogeneity across sectors and
ideas per patent can vary significantly, which makes countries, including their tax treatment; differing public support
international comparison of simple patent counts systems; and other legal, institutional, and cultural differences.
5For example, copyrights, which are used to protect the intellectual
harder. To improve comparability, this chapter follows
ownership of texts, software, and other expressions of creative work,
the practice developed in the literature to construct
do not generally require registration, which complicates record keeping
quality-adjusted patent measures (Box 4.1 discusses even if the information is public. By definition, this also holds for
the concepts and measurement issues related to patent trade secrets. Open-source software is another example of technology
indicators). The preferred measures focus on interna- diffusion that does not involve patents or patent citations.
6For instance, recent country rankings based on broader measures
tional or top three patent “families,” which group indi- of innovativeness by Bloomberg Finance L.P. correlate strongly with
vidual applications for the same underlying technology. those based on the patent measures used in this analysis.

Despite some limitations, patents are an attractive Figure 4.3. Patenting and Research and Development at the
measure to capture innovation, which is also reflected Frontier
in their frequent use in the economic literature. Patents 80,000 1. Patenting
are related to new ideas with the objective of, or at least (International patent families by publication year)
70,000
potential for, economic exploitation. The key advantage United States Korea
60,000
is, however, the precision with which the idea can be Japan EU G–3
50,000 China
attributed to its creator at a particular moment and to
40,000
other ideas through the link of citations.
30,000
Technology diffusion can stimulate innovation, but
20,000
may also affect productivity directly through simple
10,000
adoption of existing technology. To test for this more
direct channel, various productivity indicators are 0
1990 92 94 96 98 2000 02 04 06 08 10 12 14
examined. This provides a broader, albeit less precise,
measure of technological progress and complements the 500,000 2. Gross Domestic Expenditure on R&D
patent-data analysis. The disadvantages of these measures 450,000 (Millions of constant US dollars, PPP)
compared with patent counts are that their quantifica- 400,000 United States Korea
350,000 Japan EU G–3
tion is subject to significant measurement uncertainty China
300,000
(especially for total factor productivity) and that they 250,000
include components extraneous to innovation (for 200,000
example, labor productivity increases with investments 150,000
in physical and human capital). Their main advantage is 100,000
50,000
that all innovations, regardless of their specific channel
0
of diffusion, are expected to translate into changes in 1981 85 89 93 97 2001 05 09 13 15
productivity eventually. Use of productivity measures
also helps to disentangle the effect of foreign R&D on Sources: European Patent Office, PATSTAT database; Organisation for Economic
Co-operation and Development; and IMF staff calculations.
domestic innovation (patents) from its contribution to Note: EU G–3 = France, Germany, and the United Kingdom; PPP = purchasing
the efficiency of domestic production (productivity). power parity.
A final issue is whether patent citations are a good
proxy for the extent to which foreign knowledge
logical advances have been concentrated in a few large
becomes available for domestic use through the various
industrialized countries.
transmission channels. For instance, a popular alter-
The United States, Japan, Germany, France, and
native proxy is the intensity of international trade.
the United Kingdom (henceforth the G5) accounted
This approach has its own drawbacks, however, as a
for about three-fourths of international patent fami-
significant fraction of trade in goods is not associated
lies during 1995–2014 (see Figure 4.1). They are also
with any technology diffusion. Indeed, a key advantage
responsible for the bulk of R&D spending over those
of using the propensity to cite foreign patents is that it
years (Figure 4.3). For this reason, the aggregated activ-
provides a direct measure of knowledge use and, at the
ity of the G5 is used as a proxy for the global tech-
same time, correlates well with other indirect mea-
nology frontier and as the main source of technology
sures, such as the propensity to import.7 On balance,
diffusion worldwide in the chapter’s analysis.
patent citations are the more attractive indicator of the
However, this is not to imply that other emerging
extent of use of foreign knowledge, but the chapter
market or advanced economies have not contributed
also offers estimates based on the intensity of trade as a
to the evolution of global knowledge. For example,
robustness check.
in recent years Korea and China have joined the top
five leaders in a number of sectors, either based on the
The Innovation Landscape stock of R&D and/or the stock of international pat-
The evolution of innovation can be tracked by ents (Figure 4.4). Their rise is particularly pronounced
examining data across different measures, countries, in the electrical and optical equipment sector and, for
and time periods, which confirms that global techno- Korea especially, in machinery equipment.
The dynamics in innovation between economies
7See for example MacGarvie (2006).
at the technology frontier and others are diverg-

Figure 4.4. Countries at the Technology Frontier Figure 4.5. Slowing Patenting and Productivity
United States Japan Germany United Kingdom 9 1. G5: R&D and Patent Stock Growth
France Korea China 8 (Percent)
7
6
1. Country Rank Based on Aggregate Patent Stock1
0 5
4
5 3
2 Top three patent families
10
1 International patent families
15 0 Gross expenditure on R&D
–1
20 1990 92 94 96 98 2000 02 04 06 08 10 12 14
25
14 2. Patenting Growth 25
30 (Annual percent growth, average across country sectors)
1995 97 99 2001 03 05 07 09 11 13 14 12
1995–2003 20
10 2004–14
8 15
0 2. Country Rank Based on Aggregate R&D Stock2
6 10
2
4
4 5
2
6
0 0
G5 Non-G5 AEs EMs excluding China
8 China (right scale)
10
3.5 3. Labor Productivity Growth 14
12 (Annual percent growth, average across country sectors)
1995 97 99 2001 03 05 07 09 11 13 14 3.0 12
1995–2003
2.5 2004–14 10
Sources: European Patent Office, PATSTAT database; Organisation for Economic
Co-operation and Development; and IMF staff calculations. 2.0 8
Note: R&D = research and development.
1
Based on international patent families. 1.5 6
2
Cumulated gross domestic expenditure on R&D (in millions of constant US dollars 1.0 4
at purchasing power parity).
0.5 2
0.0 0
G5 Non-G5 AEs EMs excluding China
ing (Figure 4.5). Since the early 2000s, the G5 has China (right scale)
experienced a pronounced slowdown in growth of
patenting—and to a lesser extent R&D—mirroring Sources: European Patent Office, PATSTAT database; KLEMS database;
Organisation for Economic Co-operation and Development; United Nations
the well-documented slowdown in labor productivity Industrial Development Organisation; and IMF staff calculations.
and total factor productivity.8 The slowdown was much Note: AEs = advanced economies; EMs = emerging market economies;
G5 = France, Germany, Japan, the United Kingdom, and the United States;
milder in advanced economies outside the G5 and in R&D = research and development.
emerging market economies. Growth in innovation
and productivity held up much better, especially in
emerging market economies. Diverging dynamics
of innovation related to advances in information and
could reflect issues particular to the frontier and/or
communication technology (ICT) is fading, while
changes in the way innovation is diffused from the
ongoing progress in the digital domain, artificial
frontier to other regions. To elaborate:
intelligence, automation, and machine learning will be
•• Issues specific to the frontier: There are two main
felt some years after their introduction (Brynjolfsson,
hypotheses behind the slowdown at the frontier. One
Rock, and Syverson 2017) because the benefits take
proposes that the impact of the most recent large wave
time to materialize as new general-purpose technol-
8Patenting in the United States has picked up in recent ogies. More pessimistic views (for example, Gordon
years, however. 2012; Bloom and others 2017) contend that really

Figure 4.6. The Evolution of Cross-Patent Citations within and across Regions
1. 1995
United States
Other Asia
EU (28)
Japan
2. 2014
EU (28) Other Asia

Japan
United States
Sources: European Patent Office, PATSTAT database; and IMF staff calculations.
Note: Figure shows the evolution in citation flows between (blue) and within (red) key countries and regions. For a given year, the thickness of the arrows is
proportional to the respective numbers of citations. For visibility, the increase in citations over time could not be reflected proportionally (approximate scaling factor
2014 versus 1995 is 1.5 in the figure; actual is 2.5). EU (28) = AUT, BEL, BGR, CYP, CZE, DEU, DNK, ESP, EST, FIN, FRA, GBR, GRC, HRV, HUN, IRL, ITA, LTU, LUX,
LVA, MLT, NLD, POL, PRT, ROU, SVK, SVN, SWE; Other Asia = China and Korea. Data labels use International Organization for Standardization (ISO) country codes.
good ideas become harder to come by over time, •• Changes in technology diffusion: While knowledge
leading to a secular decline in productivity growth. creation at the frontier seems to have slowed for now,
Keeping productivity growth constant would require past ICT progress and increases in globalization have
increasingly larger R&D investment in this scenario.9 opened the potential for knowledge to travel faster
and farther. Figure 4.6 shows a map of knowledge
flows in which the red arrows represent cross-patent
9Autor and others (2016) have pointed to the increased trade compe- citations within a country or region, and the blue
tition from China as a possible explanation for the decline in US firms’ arrows point to citations across countries or regions.
innovation, since it reduced profits and overall operations, including
R&D spending, of trade-exposed firms. This conclusion, however, is at Similar to other measures, the map illustrates a
odds with that of Bloom, Draca, and Van Reenen (2016), who find a changing international constellation. While in 1995
positive effect of the China shock on European innovation activity, and the United States and—to a lesser extent—Europe
seems less consistent with aggregate data, which show no protracted
slowdown in R&D spending in the United States.
and Japan were dominating global patent citations,

Figure 4.7. Knowledge Diffusion across Barriers over Time measured in a more formal way than in the previous
section. Many economists believe knowledge flows are
1.4 1. Reduction of Knowledge Flow with Additional Barriers
(Number of citations to G5 countries relative to within country-
localized, because barriers, such as geography, language,
1.2 sector citations) or technological differences, weaken their diffusion.
Individual effect Cumulative effect
1.0 These barriers can attenuate knowledge diffusion directly
0.8 or indirectly, because they reduce economic transactions
0.6
such as trade, FDI, and migration, which are important
channels for the transfer of knowledge. This section uses
0.4
a gravity model to estimate the impact of these barriers
0.2 on the intensity of knowledge flows and then examines
0.0 whether their effect has become less important over time
home
diff_country
diff_border
diff_lang
tech_spec 50th
tech_dev 50th
1,000 km
5,000 km
10,000 km
(see Annex 4.2 and Peri 2005).
The focus is on international knowledge diffusion
from the frontier, proxied by the G5 countries and
within broadly defined industrial sectors.10 Focusing
on the G5 countries misses the changing role of some
0.35 2. Diffusion of Knowledge from G5
(Predicted share of knowledge that diffuses, average across economies, particularly China and Korea, but cap-
0.30 recipient country-sectors, five-year averages) tures the bulk of the contribution to global patenting
0.25 and R&D stocks for most of the sample. Korea and
0.20 China are thus treated as recipients, even though, in
the future, they are likely to become more important
0.15
sources of global knowledge flows.11,12
0.10
The analysis uses country-sector rather than
0.05 Advanced economies Emerging market economies
economy-wide data, which makes it possible to control
0.00 for factors specific to each citing and cited country sector
1995 97 99 2001 03 05 07 09 11 13 14
in each period. Such factors include the quantity of
patenting and institutional or cultural characteristics that
Note: G5 = France, Germany, Japan, the United Kingdom, and the United States. influence the propensity to patent or to cite other patents.
Panels are derived from coefficients of same-sector regression on citations to G5 The sectoral approach is also appropriate for studying
countries. Tech_spec 50th denotes the 50th percentile of the variable tech_spec;
and tech_dev 50th denotes the 50th percentile of the variable tech_dev. knowledge diffusion because the potential for technologi-
km = kilometers. cal progress varies across industries, and the sectoral com-
position of a country’s economic activity influences the
extent of knowledge and technology diffusion. A draw-
China and Korea (depicted together as “other Asia”) back of using sector-level data is that it limits the extent
have become increasingly more integrated into global to which conclusions can be drawn about the aggregate
citations. The map in Figure 4.6 also shows a general economy. Nevertheless, the average sector-level effects
intensification of patent citations over time, captured provide a sense of the broader effects on the economy.
by the increase in the size of the arrows. However, A key summary of the analysis is the predicted
this alone does not mean that the stock of global relative frequencies of citations for each country sector
knowledge was diffusing faster. As discussed earlier, (henceforth denoted ϕ  ˆ and used in the subsequent sec-
citations are a function of the amount of innovation tion). These can be interpreted as the share of knowl-
as well as the propensity to patent and cite other edge that diffuses from the cited to the citing relative
patents, which is influenced by institutional and legal to what diffuses within the cited country sector (see
differences across countries and over time. The next Annex 4.2). Figure 4.7 (top panel) shows the share of
section derives a measure of knowledge flows that
deals with these issues. 10Intrasectoral spillovers are significantly stronger than spillovers
across sectors, reflecting in part the broad definition of the sectors
used in the analysis. Annex 4.2 provides evidence substantiating this.
Determinants of Knowledge Flows 11Annex 4.2 shows that the empirical results are robust to
excluding China.
The strength of knowledge flows from the technology 12In the case of China, an additional consideration is the absence
frontier, and how those flows have changed, can be of sufficiently long historical sectoral R&D data.

knowledge diffusing from the G5 across cumulative trends that could affect domestic innovation but be
barriers between same-sector pairs over 1995–2014. mistakenly attributed to the trend in foreign knowl-
While naturally at 1 in the home country sector, this edge flows. The sector-level effects are later aggregated
share declines by roughly ½ when information crosses to provide evidence suggestive of the impact on the
a national border (diff_country). While the effect of broader economy.13
contiguity (diff_border) is more moderate, a differ- Knowledge flows are measured by weighting the
ent language (diff_lang) again significantly decreases G5 knowledge stock—measured by their R&D
this share. Differences in technological specializa- stock—with the time-varying bilateral shares of
tion (tech_spec) and in technological development knowledge flows ϕˆ   estimated in the previous section
(tech_dev) also lead to a reduction in knowledge (see Figure 4.2).14 As discussed, the weighting method
flows. Adding technological, linguistic, and geographic used here implicitly captures various channels of knowl-
distances results in average shares of knowledge dif- edge transmission, including trade, FDI, and migration.
fusion of 15–20 percent. Thus, knowledge flows are An alternative and simpler weighting method based on
relatively localized. time-varying trade linkages at the sectoral level is also
Next, the analysis investigates how knowledge used in a robustness exercise, capturing more directly
diffusion from the G5 has changed over time, based possible knowledge transmission through trade exposure
on different regressions for each five-year period. with technology leaders (Annex 4.3).
Figure 4.7 (panel 2) shows the evolution of the The analysis then estimates how innovation (patent
average degree of knowledge diffusion for advanced flow) or productivity in the recipient country sector (P)
and emerging market economies. While emerging depends on its own R&D stock (R  c) and the weighted
market economies have notably increased their access total R&D stock of the five technology leaders (R  l).
to information available at the frontier over time, Building on the work of Peri (2005), Coe, Helpman,
this does not hold for advanced economies, which— and Hoffmaister (2009), and Acharya and Keller
particularly since the global financial crisis of 2008— (2009), the approach can be summarized as
have experienced less diffusion of knowledge, possibly
lnP  i,c,t = D c,t  +  γlnR  i,c,t  +  μln∑ l≠c ϕ  i,c,l,t R  i,l,t 
related to the postcrisis slowdown in trade. The deep-
+ ε  i,c,t, (4.1)
ening integration of emerging market economies in
knowledge flows is mostly driven by a change in the in which i denotes the industrial sector, c the country
effect of the distance in technological development receiving spillovers, l the technology leaders (that is, the
(tech_dev). In earlier periods, knowledge flows weak- G5 countries), and t the time period. The coefficient on
ened with distance from the technological frontier, the weighted foreign R&D stock (μ) captures the aver-
but this source of divergence has faded and has been age efficiency of use of foreign knowledge. The equation
replaced by a convergence trend in more recent years. is estimated using sector-level data for a broad sample
These patterns remain the same even when excluding of advanced and emerging market economies from
China, suggesting a broader pattern across emerging 1995 to 2014. The regression includes country-year
market economies. fixed effects to control for time-varying factors that may
drive innovation or productivity trends.
Impact on Innovation and Productivity

The previous section focused on knowledge flows Impact on Innovation
between the technology frontier and other countries. The estimates suggest that knowledge flows from the
It has shown that national and linguistic borders are G5 are important in stimulating the flow of domes-
important, but that the combined effect of gravity has tic innovation, as proxied by patenting, indicating
decreased for emerging market economies, increasing significant learning from the technological frontier
their access to knowledge available at the frontier.
This section examines the impact of these knowledge 13In general, the sectoral approach clearly establishes the causality
flows on innovation activity and productivity in recip- of the effect, but it does not capture aggregate general equilib-
rium effects.
ient countries. Again, the analysis uses country-sector 14Using the predicted values rather than actual values helps avoid
data instead of aggregate data. This better identifies a potential endogeneity problem because they are based on highly
the effects of interest, as it controls for aggregate exogenous variables and exclude the fixed effects.

Table 4.1. Impact of Foreign Knowledge on Domestic Innovation and Productivity

Dependent Variable Patent Flow Labor Productivity Total Factor Productivity
(1) (2) (3) (4) (5) (6)
Changing Changing Changing
Sample Period (1995–2014) Baseline Diffusion Baseline Diffusion Baseline Diffusion
Foreign R&D Stock, weighted1 0.350*** 0.199*** 0.057*** 0.040* 0.053** 0.018
[0.055] [0.057] [0.020] [0.022] [0.021] [0.037]
Foreign R&D Stock*2000–04 0.137*** 0.039*** 0.026*
[0.031] [0.012] [0.014]
Foreign R&D Stock*2005–09 0.191*** 0.043** 0.052**
[0.039] [0.018] [0.024]
Foreign R&D Stock*2010–14 0.259*** –0.009 0.072**
[0.048] [0.026] [0.030]
Own R&D Stock 0.448*** 0.441*** 0.118*** 0.118*** 0.060** 0.058*
[0.061] [0.060] [0.022] [0.022] [0.023] [0.030]
Observations 3,487 3,487 3,721 3,721 1,192 959

R 2 0.779 0.784 0.758 0.759 0.958 0.955
Country-Year Fixed Effects Yes Yes Yes Yes Yes Yes
Note: R&D = research and development. Robust standard errors (clustered at country-sector level) in brackets.
***p < 0.01, **p < 0.05, *p < 0.1.
1Regression equations for labor productivity and total factor productivity use the lag value of the weighted foreign R&D stock variable.
(Table 4.1, column [1]). For example, on average, a that foreign knowledge boosts the innovation capacity
1 percent increase in the knowledge-flow-weighted of firms, and highlights the role played by technology
foreign R&D stock is associated with about a ⅓ of sourcing—the research carried out in the main tech-
1 percent increase in the count of patent families by nological leaders—to circumvent the local character
the recipient country sector. Moreover, cross-border of knowledge and access the knowledge of technolog-
technology diffusion seems to have intensified, as ical leaders.
indicated by the steady and significant increase in
the coefficient on the weighted foreign R&D stock
between 1995 and 2014 (Table 4.1, column (2)). And Impact on Productivity
while the acceleration in technology diffusion over Foreign knowledge also plays a role in boosting
time is visible for recipients in advanced economies, domestic productivity (Table 4.1, columns [3] and
it is more pronounced for emerging market recipients [5]). This is true for both emerging market economies
(see Annex 4.3 for details). and advanced economies, though the effect is larger for
An alternative specification using simple trade emerging market economies. Separate estimations for
weights instead of citation weights to proxy for the recipients indicate that industries in emerging market
use of the foreign R&D stock produces broadly economies benefit significantly more than those in
consistent estimates, demonstrating the robustness of advanced economies from the role of foreign knowl-
the results (Annex Table 4.3.1). These results are also edge flows in channeling technological transfer into
robust to sensitivity checks, including the use of other higher labor productivity (Annex Table 4.3.2).
quality-adjusted patent measures, or the alternative Interestingly, while the impact of foreign knowl-
estimation method provided by dynamic ordinary least edge flows on innovation has remained strong (and
squares (OLS).15 Measuring the stock of G5 knowl- even strengthened) over time, the picture is mixed
edge by their weighted patent stock—instead of their for the spillover to productivity (Table 4.1, columns
weighted stock of R&D—to capture foreign knowl- [4] and [6]). Indications are that the impact on total
edge flows confirms that G5 patents make a significant factor productivity has strengthened over the past two
contribution to innovation in other countries. Using a decades,16 but the effect on labor productivity seems
similar framework, Box 4.2 presents firm-level evidence
15Dynamic OLS can address possible nonstationarity and cointe- 16The estimation sample for total factor productivity is smaller
gration of the patent and R&D series in a panel setting. and consists mainly of advanced economies.

to have weakened in the postcrisis years of 2010–14.17 Figure 4.8. Contribution of Foreign Knowledge to Labor
This could be consistent with arguments discussed Productivity Growth
(Annual percent growth, average across country sectors)
earlier—that innovations make increasingly less impact
(Bloom and others 2017). Another—more benign— 3.0 1. By Country Groups, 1995–2014
explanation could be that the protracted period of
2.5 Domestic R&D
subdued investment following the global financial crisis Foreign R&D
reduced technology diffusion, as investment goods are 2.0 Annual labor productivity growth
an important conduit for embodied new technolo-
1.5
gies to integrate into production processes (Adler and
others 2017). 1.0
Although based on sector data alone, the effects 0.5

of foreign knowledge flows on labor productivity are
0.0
economically meaningful. For illustrative purposes, All Advanced Emerging market
using the estimates in Table 4.1, one can calculate economies economies
the effect of observed changes in the weighted foreign
2.0 2. By Subperiod: Advanced Economies
R&D stock and domestic R&D stock on the growth in
domestic labor productivity in each country-sector— Domestic R&D
1.5 Foreign R&D
assuming everything else remains the same (see Annual labor productivity growth
Annex 4.3).18 These contributions can then be averaged
over countries and sectors included in the analysis 1.0
to give a sense of the magnitude of the effects. The
estimates suggest that during 1995–2014, developments 0.5
in domestic and foreign R&D combined would have
generated about 1 percentage point average sectoral 0.0
1995–2003 2004–14
labor productivity growth a year, which is about 60 per-
cent of the observed sectoral labor productivity growth,
5 3. By Subperiod: Emerging Market Economies1
consistent with there being other sources of productiv-
Domestic R&D and human capital
ity improvements. The impact of knowledge flows from 4 Foreign R&D
the G5 alone amounted to about 20 percent of the Annual labor productivity growth
explained average growth in sectoral labor productivity 3
in the sample and one-eighth of the observed average

2
growth in sectoral productivity (Figure 4.8).
The effects vary for advanced economies and emerg- 1
ing market economies in the following ways:
0
•• Technology diffusion boosted productivity growth 1995–2003 2004–14
in emerging markets more strongly, providing a
counteracting force to the slowing innovation trends Source: IMF staff estimates.
at the frontier. From 2004 to 2014, foreign knowl- Note: R&D = research and development.
1
The decomposition by subperiods for emerging market economies is based on a
edge accounted for about 0.7 percentage point of slightly different regression specification with a less demanding data requirement,
labor productivity growth a year, or 40 percent which allows for having a significantly broader sample of emerging market
economies (Annex 4.3).
of observed sectoral productivity growth, com-
pared with 0.4 percentage point annual growth
17This is consistent with OECD (2015), which, looking at a

during 1995–2003 (see Figure 4.8). Greater use
sample of firms in advanced economies, finds evidence of a rising of existing foreign knowledge by emerging market
gap in productivity growth between global frontier firms and
other firms. See also Andrews, Criscuolo, and Gal (2016). economies—combined with the stronger impact
18To assess the impact of aggregate (country-level) variability of these knowledge flows on industries in emerg-
on the coefficients estimated in equation (4.1), the regression was ing market economies than on those in advanced
also run without country-time fixed effects. The estimated impact
of the weighted foreign R&D stock on labor productivity was
economies—has been a significant factor in main-
broadly unchanged. taining the better labor productivity performance

Figure 4.9. The Dynamics of Technology Diffusion Estimating Short-Term Dynamics

(Percent)
As a complementary approach to the long-term
0.25 1. Response of Recipient TFP to a TFP Shock in Leaders framework and robustness check, this section investi-
gates the short-term dynamics of technology diffusion
0.20 using the local projection method (see Jordà 2005).
Extending the analysis in Duval and others (forthcom-
0.15
ing), this approach focuses on the short-term impact of
0.10 a productivity or innovation shock in the technology
leaders on productivity or innovation in the recipient
0.05 country sector (see Annex 4.4 for details and for defi-
nition of the shocks). Shocks to innovation are taken
0.00
1 2 3 4 5 to be changes in the total patent stock of the technol-
ogy leaders. Again, shocks in the leaders are weighted
0.20 2. Response of Recipient Labor Productivity to a Technology by the bilateral shares of knowledge that flow from the
Shock in Leaders
G5. The empirical specification includes country-time
0.15 fixed effects to capture factors that drive the short-term
dynamics of a country’s productivity and innovation at
0.10 the country level, such as business cycles.
The impact of technology shocks is significantly
0.05 stronger in the case of innovation measures. On aver-
age, a 1 percent patent shock in the leaders would raise
0.00 the patent stock in the recipient by at least 1 percent
1 2 3 4 5 after five years (Figure 4.9). This suggests that an
acceleration of innovation in technology leaders has a
2.5 3. Response of Recipient Patenting to a Patenting Shock in Leaders
particularly strong effect on innovation in other coun-
2.0 tries.19 But the effects are also significant for broad
productivity measures: in response to a 1 percent total
1.5 factor productivity (or labor productivity) shock in
the technology leaders, total factor productivity (labor
1.0
productivity) in the average recipient country sector
0.5 is estimated to increase by about 0.15 (0.07) percent
after five years. The results indicate that technology
0.0 spillovers tend to happen relatively quickly—within
1 2 3 4 5
a few years of the initial shock—and the size is not
Source: IMF staff estimates. negligible.
Note: TFP = total factor productivity. Blue shade denotes 90 percent confidence
band. Impulse responses to a 1 percent TFP/labor productivity/patent shock
estimated using local projections. X-axes denote years; t = 1 is the year of the
shock.
Flows within the Technology Frontier
What about the G5 themselves? So far, the empiri-
cal approach has focused on the predominant pattern
of these economies compared with that of advanced of knowledge and technology flows in the sample
economies. Results are robust to excluding China, period analyzed—that is, from the frontier to other
which suggests that emerging market economies, countries. However, this does not mean that flows
more broadly, have benefited. have been going in one direction only. One way to
•• In advanced economies, the contribution of foreign shed light on this question is to apply the empirical
knowledge to labor productivity growth was much approach developed above (see equation [4.1]) to
smaller, given the slowdown at the frontier and the estimating knowledge and technology diffusion among
absence of further improvements in use of foreign
knowledge (this use even declined after the global 19This suggests that follow-on innovations respond more than
financial crisis). proportionally to the initial innovation.

the G5. The exercise is subject to additional econo- (UNCTAD 2005). Griffith and Miller (2011)
metric concerns, as it is more difficult to ensure the look at examples of how multinationals in west-
absence of endogeneity and simultaneity bias than ern Europe create new knowledge using inventors
in the earlier exercises. With this caveat in mind, the located in emerging market economies.
results suggest that G5 countries themselves benefited •• On the other hand, recent analysis suggests that
from knowledge flows from other technology leaders, GVC participation often implies that innovation is
boosting their domestic innovation. Indeed, a 1 per- relocated within multinational firms to where it can
cent increase in the knowledge-flow-weighted R&D be most efficiently undertaken (Stiebale 2016). A
stock of “other” G5 countries is associated with about considerable increase in the postacquisition inno-
a ½ percent increase in the count of patent families in vation of a merged entity is driven by inventors
the G5 country considered—slightly larger than the in the acquirer’s country, while innovation in the
⅓ of 1 percent increase obtained in the baseline for country of the acquired entity tends to decline. In
non-G5 recipient countries (Table 4.1, column [1]). the case of emerging market economies, in particu-
Using firm-level data to examine knowledge spillovers lar, the relocation of multinational firms’ innovative
through technology sourcing, Box 4.2 also provides activities could reflect efforts to overcome ineffi-
evidence that knowledge spillovers between technology ciencies associated with weak institutions, including
leaders are strong—possibly even stronger than for weak intellectual property regimes (see Zhao 2006).
nonleader recipients. Western firms respond by holding the intellectual
property that results from emerging markets’ innova-
tion in the location of the parent.20
The Impact of Global Value Chains on
Patenting: A Firm-Level Analysis What role do GVCs play in this context? At first
While the preceding sections aimed to assess the glance, trends in GVC participation and patent-
strength of international technological spillovers and ing suggest that the two appear to be related across
their effects on productivity, this section explores one emerging market economies (Figure 4.10, panel 1),
specific channel through which such transmission which would suggest a positive impact. To determine
occurs: firms’ participation in global value chains whether these countries have indeed been able to
(GVCs). Firms are increasingly part of complex produc- capitalize on their participation in GVCs by increas-
tion networks—often centered around multinational ing innovation, the analysis follows the firm-level
enterprises—that process diverse goods and services framework used by Bloom, Draca, and Van Reenen
inputs from other domestic and foreign firms. Potential (2016) (see Annex 4.5).21 Working at the firm level
gains to firms in emerging market economies could be makes it possible to distinguish two types of tech-
economically significant, because multinational enter- nological diffusion as a result of GVC participation:
prises are typically at the global productivity frontier (1) a buildup of innovation capacity in the average
(OECD 2015). Engagement with multinational enter- firm—so-called within-firm effects, and (2) differentia-
prises through GVCs provides opportunities for knowl- tion of this effect between firms with different rates of
edge spillovers to local firms along the value chains, patenting—“between-firm” effects.22 This between-firm
by pooling knowledge with domestic suppliers and
encouraging new practices, specialization in productive 20Strokova (2010) documents that intellectual property regimes
tasks, and the use of new varieties and higher-quality in emerging market economies, while improving, remain rel-
foreign goods, services, and intangible inputs. atively weak.
21Firms can also benefit from participation in GVCs through
In this way, the emerging pattern of decentralized technology adoption without necessarily innovating themselves
global production represents a key channel for firms (see, for instance, Lopez-Garcia and Taglioni 2018, for evidence on
in emerging markets to build innovative capacity, with Europe). Testing for these effects would require firm-level produc-
tivity measures, which are not broadly available for emerging market
potentially positive effects for the rest of the economy. economies in this chapter’s sample. The test in this section is more
However, opposing forces may be at work: demanding, since it examines whether participation in GVCs has
•• On the one hand, innovative activity by Western boosted emerging market firms’ innovation capacity and not just
their adoption of foreign technology.
firms in emerging market economies has increased 22Due to lack of data on absorption capacity in firms or sectors,
dramatically, albeit from relatively small levels, the analysis follows a direct approach by controlling for firms’ initial
driven by a handful of large multinational firms level of innovation (as in Bloom, Draca, and Van Reenen 2016) and

Figure 4.10. Patenting and Global Value Chain Participation of the effect of GVC participation, by controlling for
firm-level characteristics that may also determine inno-
0.18 1. Trends in Average Firm Patenting and GVC 0.32
Participation
vation capacity.23
0.16
0.31 To ensure that the impact of GVC participation on
0.14
0.12 0.30
innovation is correctly identified, the empirical strategy
0.10 attempts to tackle potential reverse causality from
0.29 patenting to GVC participation. While technology
0.08
0.06 Patent flow (log) 0.28 improvements may occur because of GVC participa-
0.04 Global value chain participation tion, firms may be pulled into GVCs because of their
(share of gross exports, right scale) 0.27
0.02 high productivity, their capacity to innovate, or even
0.00 0.26 through self-selection that comes from being set up
2000 01 02 03 04 05 06 07 08 09 10 11 12
with attributes that lend themselves to GVC partici-
22 2. GVC Participation versus Inward FDI Stocks
pation. The analysis exploits the relationship between
GVC participation and FDI to establish causality:
GVC participation (natural log)
y = 0.4709x + 13.533 it is well known that GVC participation is strongly

20
R 2 = 0.5912 correlated with FDI, given how both relate to the
international allocation of production (see Figure 4.10,
18
panel 2). Changes in GVC participation are therefore
identified using policy instruments that affect FDI and
16
trade—namely, an industry-level policy indicator of
restrictions to FDI and changes in tariffs. These have
14
0 2 4 6 8 10 12 14 fallen as GVC participation has increased (see Fig-
Inward FDI stock (natural log) ure 4.10, panel 3), and they are found to be negatively
associated with changes in GVC participation in the
25 3. FDI Restrictions and Tariffs in EMDEs 0.25 econometric analysis (see Annex 4.5). These instru-
ments help correct for the potential endogeneity of
20 0.20
GVC measures to patenting.24
15 0.15 The results show that an increase in GVC partic-
ipation leads to a reallocation of innovation activity
10 0.10 but, overall, has a positive effect on firm patenting.
Average tariffs in EMDEs (percent) The effect of a change in GVC participation on firm
5 0.05
FDI index (open=0, closed=1; right scale) patenting flows is significantly positive (a “within
0 0.00 effect”), but declines with the initial level of patent-
1997 98 99 2000 01 02 03 04 05 06 07 08 09 10 11 12 13 ing activity of the firms (“between effect”) (Table 4.2,
column [1]).
Sources: EORA Multi-Region Input-Output database; External Wealth of Nations;
European Patent Office, PATSTAT; Foreign Direct Investment statistics; IMF, Once the potential endogeneity between GVC par-
October 2016 World Economic Outlook; Orbis; United Nations Conference on Trade ticipation and patenting is controlled for, the impact
and Development; and IMF staff calculations.
Note: EMDEs = emerging market and developing economies; FDI = foreign direct of GVC participation on patenting is even stronger
investment; GVC = global value chain. (Table 4.2, column [2]). This happens both within
and between firms. The estimated effects imply that
firms that were already patenting before the increase
in GVC participation tend to see some reduction in
analysis is also used to examine how GVC participa-
tion alters sectoral composition of employment across 23The primary patent data are drawn from PATSTAT, and global
firms according to their technological intensity (mea- input-output tables are used to construct industry-level GVC par-
sured by past patenting activity). Another advantage ticipation measures (see Annex 4.5). GVC participation is measured
to working at the firm level is improved identification by the sum of (1) the domestic content in exports reused in trading
partners’ exports (forward linkages), and (2) the foreign value added
embedded in exports (backward linkages) expressed as a share of
an indirect approach correlating the country-time fixed effects of the gross exports.
main regression with country-level measures of absorption capacity, 24Standard tests confirm that the instruments satisfy the exclusion
such as education, quality of infrastructure, and the rule of law. restriction.

Table 4.2. Impact of Global Value Chain Participation on Average Firm Patenting and Employment
Employment
Dependent Variable Patent Flow (Log, five-year difference) (Log, five-year difference)
(1) (2) (3)
OLS (Matched ORBIS -
Sample Period (2002–12) OLS (PATSTAT Firms) IV (PATSTAT Firms)1 PATSTAT Firms)
Initial Patent Stock (2000) –0.07*** –0.09*** –0.02*
[–5.703] [–30.002] [–1.873]
Within-Firm Effects
GVC Participation (Five-year change) 0.28*** 0.98*** 1.82***
[3.133] [7.420] [8.002]
Between-Firm Effects
Initial Patent Stock (2000) × –1.31*** –1.67*** 0.91*
GVC Participation (Five-year change) [–4.160] [–4.963] [1.943]
Observations 4,044,066 2,928,882 87,929

R 2 0.026 0.030 0.182
Country × Year Fixed Effects Yes Yes Yes
Sector Fixed Effects Yes Yes Yes
Note: IV = instrumental variable estimation; GVC = global value chain; OLS = ordinary least squares. Robust t-statistics in brackets.
***p < 0.01, **p < 0.05, *p < 0.1.
1 Instruments include foreign personnel restrictions (percent-year difference and level), screening and approval procedures (level), and tariffs (five-year
difference). (See Annex 4.5 for details.)
their patenting flow, possibly reflecting reallocation of To gauge the role of policies in building innova-
some innovation activity to other parts of the GVC.25 tion capacity in emerging market firms, Figure 4.11
But more extensive GVC participation significantly (bottom panel) shows the correlation between the
increases the average patenting of firms that did not country-year fixed effects from the estimated patenting
previously patent. These firms represent 75 percent of relationships and a number of policy factors. Policies
the sample—90 percent excluding China. The overall aimed at improving the quality of education and
effect on patenting of the average firm is positive, with connectivity to the world through better infrastruc-
the observed 1 percent increase in GVC participation ture are key, contributing jointly to increase growth
every five years explaining one-tenth of the increase in patenting by 2 percent over five years. Box 4.3
in patenting in the average firm over the same period discusses how foreign aid can play a role in technology
(Figure 4.11, top panel). diffusion to low-income countries by helping build key
Turning to the broader impact on the economy, infrastructure technologies and investing in education.
increased GVC participation leads to higher employ- Finally, the evidence presented in Figure 4.11 (panel 2)
ment growth for the average firm and faster employ- also suggests that greater adherence to the rule of law
ment growth for patenting firms than experienced by boosts firm patenting, possibly mitigating the need for
nonpatenting firms (Table 4.2, column [3]).26,27 The multinational companies to rely on internal mecha-
larger share of workers flowing from firms that do not nisms, such as relocation of innovation activities from
innovate to high-tech firms is another way GVC par- affiliates to the parent, to overcome market failures
ticipation boosts economies’ technological intensity. caused by poor institutions.
25The latter effect is substantially weaker once patenting activity in

The Role of Greater International Competition
China is excluded (see Annex 4.5). An alternative explanation could International technology diffusion is a key channel
be the relocation of some innovation activity by an emerging market
through which globalization impacts innovation, but
firm to source technology from an advanced economy (see Box 4.2).
26Orbis and PATSTAT data are matched to produce a data set of it may not be the only one. For example, globaliza-
both patenting and nonpatenting firms. tion could also make a difference by affecting global
27Data limitations prevent testing the effects on firm-level pro-
competition. Indeed, the evolution of global competi-
ductivity. Performance measures, such as return on assets and return
on equity, also have limitations, given that they are affected by the tion and global market concentration, and their impact
division of value added between labor and capital. on innovation is a much-debated issue (see Box 4.4).

Figure 4.11. The Effects of Global Value Chain Participation concentration has declined—notwithstanding increases
and Policy Variables in domestic concentration reported in some countries
(Gutierrez and Philippon 2017; Grullon, Larkin, and
9 1. Change in Average Firm Patenting and GVC Participation
(Log difference; simulation based on full sample) Michaely 2017). Trade with China has risen over the
8
Average five-year change in patent flows past two decades, not only in the textile industry, but
7
6
GVC effect also in innovation-intensive industries such as electrical
Other
5 and optical equipment and transport equipment (Fig-
4 ure 4.12). And the rise of firms from emerging market
3 economies has transformed the international competi-
2 tion landscape more generally (Freund and Sidhu 2017),
1 contributing to a reduction in global market concentra-
0 tion in most industries. Market concentration is usually
All firms All firms with zero initial patent stock
defined at the industry level and proxied by either a
2. Predicted Relation between Selected Policy Variables and
concentration ratio (for example, the share of total
3.0 Change in Average Firm Patenting industry sales that go to the industry’s top four firms)
2.5 (Log difference, five-year change, contribution to country year or the Herfindahl-Hirschman Index. Data on the global
fixed effects) Quality of port infrastructure
2.0 Quality of math and science education concentration of patenting show a more mixed picture,
Protection of property rights though they may underestimate the extent or rise of
1.5 Integrity of the legal system
Licensing restrictions (PMR)
concentration because the PATSTAT database does not
1.0
include information on firms’ ownership structure.
0.5 If global competition indeed has increased, has it
0.0 led to more or less innovation? An extension of the
–0.5 sectoral framework of analysis (see equation 4.1) can
WEF variables Fraser Institute variables
be used to investigate this question (see also Coe,
Helpman, and Hoffmaister 2009). In this extension,
Sources: EORA Multi-Region Input-Output database; European Patent Office,
PATSTAT database; Fraser Institute, Economic Freedom of the World; World the knowledge-weighted foreign R&D stock is inter-
Economic Forum Global Competitiveness Report; and IMF staff calculations. acted with relevant structural factors (S), including
Note: Panel 1 shows result of a simulation based on the full sample. Panel 2 shows
the five-year change of contribution to country year fixed effects. GVC = global increased trade with China and measures of global
value chain; PMR = product market regulation; WEF = World Economic Forum market concentration:
Global Competitiveness Report.
lnP  i,c,t = D c,t  +  γlnR  i,c,t +  μln∑ l ≠ c  ϕ  i,c,l,t R  i,l,t 
δln∑ l ≠ c  ϕ  i,c,l,t R  i,l,t * Si,c,t
+     
While this section does not claim to provide definitive
answers, the framework used in the chapter does lend +  θ Si,c,t
    + ε  i,c,t . (4.2)
itself to exploring this issue and provides some tenta- In this specification, the coefficient on the “main
tive evidence of the effect of competition on innova- effect” (θ) captures the direct impact of the structural
tion and the diffusion of technology. factor on innovation. The total impact of the weighted
At least two opposing forces are at work in the foreign knowledge stock on innovation is now given by
relationship between competition and innovation μ + δS, and thus the coefficient on the interaction term
(Box 4.4). More competition and lower market con- (δ) reflects the marginal boost to knowledge diffusion
centration can depress incentives for firms to innovate coming from the structural factor (see Annex 4.3
because reduced market power means fewer rents from for details).
any innovation. However, at the same time, more com- The results suggest that the observed increase
petition and lower concentration can enhance incen- in trade competition and decline in global market
tives to innovate to escape competition and secure concentration may have helped strengthen technology
rents in the first place. And while international trade diffusion across countries (Figure 4.13).28
increases the size of the market over which rents can
be captured by winners, it also enhances the “escape 28While innovation and technology diffusion could affect com-
competition” effect (Akcigit and others 2017). petition and concentration, raising a risk of reverse causality, it is
unlikely for measures of competition used in the present analysis.
By some measures, the evidence suggests that The China trade shock largely reflected exogenous policy changes,
international competition has increased and global including China’s entry into the World Trade Organization. If

Figure 4.12. International Competition and Global Concentration Figure 4.13. The Effect of Competition on Innovation and
Technology Diffusion
60 1. Final Goods Imports from China by Sector (Percent)
(Percent; share of sector output)
50 Textiles, wearing apparel, leather, and related products 10 1. Direct Impact on Innovation
Transport equipment
40 Electrical and optical equipment
Machinery and equipment not elsewhere classified 8
30 Coke and refined petroleum products
Other manufacturing sectors 6
20
10 4
0 2 Point estimate
2000 02 04 06 08 10 12 14 90 percent confidence interval
0
0.5 2. Top Four Firms’ Share in Global Sales Increase in trade with China Decrease in global concentration
Postal (0.77, 0.72)
0.4 6 2. Impact through Knowledge Diffusion

Information technology and other info. services
Share in 2014
0.3 5
Coke, refined petroleum More
Electrical and optical equipment Lower
4 trade
0.2 Transport equipment with concentration
3 China
0.1
Chemicals 2
0.0 Average
0.0 0.1 0.2 0.3 0.4 0.5 1 10th/90th percentile
Share in 2006 0
Trade with China Global concentration
0.4 3. Top Four Firms’ Share in Global Patent Families
Coke, refined petroleum Note: Panel 1 shows the estimated change in the recipient’s patenting activity in
0.3 response to the average change in the structural factors over the sample period.
Food, beverage, tobacco
Share in 2014
Lower and upper bounds denote the 90 percent confidence interval. Panel 2 shows
the estimated response of patenting activity in the recipient to a 10 percent
0.2 increase in the weighted foreign research and development stock, for a range of
Transport equipment values of structural factors.
Chemicals
0.1 Information technology
and other info. services
Electrical and optical equipment
0.0 •• Similarly, lower global concentration—as measured
0.0 0.1 0.2 0.3 0.4 by the sales share of the top four firms—stimulates
Share in 2006
both innovation and diffusion. Its impact on diffu-
Sources: European Patent Office, PATSTAT database; Freund and Sidhu (2017);
sion is nontrivial: for example, using the estimates, a
World Input-Output Database; and IMF staff calculations. 10 percent increase in the foreign R&D stock would
boost domestic patenting by about 5.6 percent in a
low-concentration sector, whereas the boost to inno-
vation would be less than half of that (2.7 percent)
•• Increased trade with China boosts domestic innova-
in a high-concentration sector.
tion and technology diffusion, the latter by increas-
ing the efficiency with which foreign knowledge
The evidence presented within the framework of
is used (both the main effect and the interaction
analysis of this chapter, however tentative, points to
effects are positive).
a positive relationship between international compe-
anything, more innovation in a country sector would reduce import tition and innovation and technology diffusion. This
penetration from China in that sector, leading to a downward bias is broadly in line with findings reported by Bloom,
in the coefficient estimate. As for the measure of global market con- Draca, and Van Reenen (2016) and Coelli, Moxnes, and
centration, it is not likely to be influenced by individual countries’
innovation, given that the G5 countries (which are treated as the Ulltveit-Moe (2016), who estimate that increased trade
technology frontier) are excluded from the sample. has a positive effect on innovation. However, the results

seem to differ from those presented by Autor and others have benefited from each other’s research efforts and
(2016), who estimate that trade with China had a nega- knowledge. With the growing contribution of China
tive impact on innovation among US firms. Clearly, the and Korea to the expansion of the technology frontier,
discussion is ongoing, and further analysis is needed to one can expect positive spillovers from these countries
achieve a deeper understanding of the opposing forces at to the traditional technology leaders. Alongside more
work. For example, the relationship among competition, traditional channels of gains from trade, the diffusion
concentration, and innovation or technology diffusion of knowledge and technology provides a powerful
could differ over time, countries, and industries. source of mutual benefits from globalization.
From a policy angle, a main conclusion of the
chapter is that global interconnectedness fosters
Conclusions and Policy Implications foreign knowledge flows. Policies to enhance these
Globalization has a positive impact on the interna- connections—whether through GVCs, FDI, or trade—
tional diffusion of knowledge and technology. While the are well known. They include relaxing excessively strin-
negative side effects of globalization have been much gent regulations on FDI, lowering trade barriers, and
discussed in public debates, the chapter highlights a key building necessary infrastructure. Interconnectedness
benefit—the contribution of globalization to the sharing per se is not enough, though. Economists have long
of growth potential across countries. Globalization facili- argued that assimilating knowledge requires absorptive
tates the diffusion of knowledge and technology through capacity (for example, Cohen and Levinthal 1989).
the international use of patents and trade. In addition— Knowledge has an important tacit component, which
while the impact of competition on innovation is a can be comprehended only through the acquisition
complex issue that necessitates further investigation— of scientific and engineering know-how. Investments
there is evidence suggesting that, by enhancing interna- in R&D and human capital are essential not only to
tional competition, globalization has increased incentives build innovation capacity but also to maximize the
to innovate and adopt foreign technologies. absorption of existing innovations (Griffith and others
The chapter has also found that emerging market 2004; Coe and others 2009).
economies have made increasing use of existing foreign Last but not least, while the chapter has highlighted
knowledge and technology over time. This has helped the positive growth effects from globalization, poli-
soften the impact of the slowdown in innovation at the cymakers must make certain that these benefits are
frontier on emerging market economies and contrib- shared broadly across the population. This includes
uted to cross-country income convergence. Participa- ensuring that innovating firms do not exploit the
tion in GVCs has been one important factor behind newly acquired technology to gain excessive control of
this development, although not all firms have bene- a market to the detriment of consumers, supporting
fited, as multinational companies sometimes relocate policies to facilitate adjustment (for example, by invest-
innovation activities to the parent company. ment in education and reskilling), and adjusting the
Finally, the evidence suggests that knowledge does tax-benefit system to reallocate income gains in line
not flow only in one direction. Technology leaders with countries’ social preferences.

Box 4.1. Patent Data and Concepts

This chapter largely relies on patent data to capture applicability with respective weights (PATSTAT; Van
innovation and information flows; this box explains Looy and Vereyen 2015). The patent is attributed to
key concepts of the data and offers a quick glance at the aggregate sector with the largest weight.
how the data are aggregated. Coordination on patent procedures is an early
The database used is the Worldwide Patent Statisti- example of international collaboration. Progress in
cal Database (PATSTAT), which includes information harmonizing procedures has continued since the late
on about 70 million patent applications from 80 19th century. Nevertheless, international comparability
countries and the relations between them. remains impaired by cultural and legal differences.
•• A patent application is the filing to a specific patent Two examples serve as illustration:
office that seeks intellectual property protection •• Japan and the number of claims: Until 1988, each
in the given jurisdiction. Patent applications are claim (or idea) needed its own patent (Dernis
territorial, which implies that a separate patent and Khan 2004), a rule that inflated the number
needs to be filed for each country where protec- of patent applications at the Japan Patent Office.
tion is desired. Although the number of claims per patent has
•• A patent family groups applications that relate increased significantly since the 1990s, the culture
to the same technology. Each patent application and fee structure have left it significantly below
belongs to one family, but an individual application United States Patent and Trademark Office or Euro-
can be a family by itself.1 pean Patent Office numbers for most of the sample
•• Patent citations relate patents that build upon each period (Katznelson 2008).
other. Applicants must cite prior knowledge to •• China and the incentives to patent: Part of the reason
delimit the novelty and legal boundary of the appli- for the recent explosion in patenting in China is
cation. Citations can themselves be an indicator of a set of Patent Promotion Policies. Fiscal and other
information flow. incentives reduce the cost of patenting or increase
While some parts of the chapter rely on patents at the payoffs not directly related to the protection
the micro level, for others, the data are aggregated to of the intellectual property. Some ideas are thus
country and industry level so they can be matched patented that in other countries would not be.
with other variables. For this aggregation, patents are The impact of such cultural and legal differences
attributed to: can be very significant. By the simple application
•• The country of residence of the first inventor: The count, China now patents about as much as the
inventor may be different from the applicant, who rest of the world combined. Using quality-adjusted
owns the patent. Because the former is the creator measures, which weigh the patent count by proxies for
of the new knowledge, the residence of the inventor their technical or economic value, often dramatically
seems more important to identify the location of reduces this share.
innovation. The ordering of inventors in a patent Various options for quality adjustment exist.2
application generally reflects their degree of impor- As preferred measures, the chapter uses the patent
tance. Focusing only on the first one (instead of a family count and focuses only on international or top
fractional attribution to all) simplifies the process three families:
without significantly altering the picture. •• An international patent family needs to have one
•• One of 13 industrial sectors of applicability: The application in at least two distinct patent offices.
technical applicability of a patent is defined by the The idea is that this filter will capture many of the
patent office, which maps the patent into sectors of lower-value patents, as the reduced expected payoff
would not warrant the extra cost of application,
The author of this box is Johannes Eugster.
examination, and maintenance in a foreign country.
1Different applications are connected to a “priority filing,” In addition, the cultural influence of certain patent
which is the first patent filing for a technology. Under the Paris offices would be reduced.
Convention of 1883, applicants have 12 months to file patents
in other member countries and claim retroactive protection
starting on the priority date (date of initial filing). The family 2See Squicciarini, Dernis, and Criscuolo (2013) for discussion
definition used in this chapter (the DOCDB family) generally of the various measures to capture the economic and technologi-
groups patents with the exact same priorities. cal values of patented inventions.

Box 4.1 (continued)

•• The top three patent families would require an tors and applicants from Europe, Japan, and the
application to at least one of the top three patent United States.
offices (European Patent Office, Japan Patent Different measures have different strengths and weak-
Office, United States Patent and Trademark nesses; none is fully satisfactory. It is therefore crucial to
Office). Relative to the previous measure, this include appropriate fixed effects in the empirical analysis
implies more consistency, as a very limited number to capture the time-varying differences in patenting and
of patent offices are involved. The drawback is citation culture. This chapter does that wherever possi-
that count measures would tend to favor inven- ble by including country-time fixed effects.

Box 4.2. International Technology Sourcing and Knowledge Spillovers

Despite the global reach of information technology, Figure 4.2.1. Innovation Intensity
many economists believe that knowledge diffusion is (Number of firms or patents)
largely localized (Audretsch and Feldman 1996; Jaffe,
160,000 9,000
Trajtenberg, and Henderson 1993; Keller 2002). In Foreign patents
their view, being geographically close to other inven- Firms with foreign patents
140,000 (right scale) 8,000
tors is important to learn from their knowledge. By
performing innovation activities abroad—especially in 7,000
120,000
technologically advanced economies—firms can tap
6,000
into foreign knowledge more effectively and improve 100,000
productivity. Data on publicly listed firms in Organi- 5,000
sation for Economic Co-operation and Development 80,000
(OECD) countries is used in this box to provide 4,000
evidence on the evolution of international technology 60,000
sourcing and test its role as a channel for knowledge 3,000
spillovers. Data used in the analysis are from the 40,000
2,000
Worldwide Patent Statistical Database (PATSTAT)
maintained by the European Patent Office and the 20,000 1,000
Orbis database by Bureau van Dijk.
0 0
1978 83 88 93 98 2003 08 13
Evolution of Global Innovation Networks
The innovation linkages are constructed using Sources: Chen and Dauchy 2018; and European Patent
information on the source and destination countries Office, PATSTAT database.
of patents granted to publicly listed firms in OECD
countries. The source is the country of residence of
patent inventors, and the destination is the headquar-
ter country of the firm that owns the patent. Three technology sourcing as a means of gaining access to
important patterns have emerged as international foreign knowledge.
innovation linkages have steadily strengthened over the Testing for the Role of Technology Sourcing as a
past four decades. First, an increasing number of firms’ Channel of Knowledge Spillover
innovations are carried out abroad (Figure 4.2.1). Sec-
ond, the network has become increasingly multilateral: Relative to a more aggregate approach, the firm-level
on average, the number of countries in which firms approach presents a number of advantages. First, it
have an innovation presence has increased. Third, can control for home country and industry trends in
dominant hubs—countries where a dominant share innovation using fixed effects. Second, it can flexibly
of patents are invented—are apparent in the network. control for other factors that affect productivity and
In the sample, 28 percent of all patents invented in are correlated with the foreign innovations of firms.
2013 are sourced from the United States, followed by For example, firms with more foreign innovations may
Germany (14 percent), the United Kingdom (13 per- have higher productivity simply because they also have
cent), and Japan (7 percent), as shown in Figure 4.2.2. more knowledge. These firms may also be better at
Perhaps not coincidentally, these countries also have using foreign knowledge in general, because they have
the largest aggregate knowledge among OECD coun- a higher “absorptive capacity.”
tries measured by research and development (R&D) The empirical model uses a firm-level production
stock. The United States, Japan, and Germany are function augmented with firm-specific knowledge,
the top three, and the United Kingdom ranks sixth. and industry-specific domestic and foreign knowledge
The observation that the majority of foreign patents as well as a number of control variables (Griffith,
are invented in knowledge hubs is consistent with Harrison, and Van Reenen 2006; Chen and Dauchy
2018). Knowledge is measured by R&D stocks. Tech-
nology sourcing is measured by the share of a firm’s
The author of this box is Sophia Chen, with support from total worldwide patents whose inventor was residing
Hala Moussawi. See Chen and Dauchy (2018) for more details. in a foreign country in the pre-sample period between

Box 4.2 (continued)

The approach distinguishes between two groups of
Figure 4.2.2. Foreign Patents by Source
OECD countries, based on their aggregate knowledge.
Country, 2013
(Number of patents) This allows for details to be gathered about the overall
direction and effect of international technology sourc-
35,000 ing from more advanced and less advanced economies.
The underlying assumption is that countries with more
30,000
aggregate knowledge are closer to the technological
frontier. The group of technology frontier countries
comprises Japan, Germany, and the United States; the
25,000
other group includes all other OECD countries. The
results are consistent with the technology sourcing
20,000 hypothesis: firms with a stronger innovation presence
in technology frontier countries benefit dispropor-
15,000 tionately more from their aggregate R&D than firms
that lack such presence. Besides the overall positive
10,000 effect, the results show some interesting patterns in
direction and size. The interaction terms between
5,000 technology sourcing and aggregate R&D stocks in less
advanced economies are not significant, suggesting
0 that the spillovers from less advanced economies are
weak. Moreover, spillovers from technology leader
DEU
GBR
JPN
CAN
ISR
BEL
AUT
NLD
CHE
AUS
SWE
ESP
KOR
Other OECD
USA
FRA
ITA
DNK
countries’ aggregate R&D is strongest when the

recipient countries are also technology leaders. These
results are robust to alternative explanations for foreign
Sources: Chen and Dauchy 2018; and European Patent innovation—such as profit shifting—and alternative
Office, PATSTAT database. models controlling for the absorptive capacity of firms.
Note: OECD = Organisation for Economic Co-operation and The results support the idea that technology
Development. Data labels use International Organization for
Standardization (ISO) country codes.
sourcing can be an effective channel of international
knowledge spillovers. Optimal policy design to stimu-
late innovation should take into account the interna-
tionalization of innovations. For example, policies to
1997 and 2006. It is interacted with the foreign R&D incentivize the repatriation of foreign-based innova-
stock to test for its role as a channel of knowledge tions may end up compromising domestic productivity
spillovers. The regression is estimated over a panel of growth by stifling domestic innovation. Furthermore,
about 12,000 publicly listed firms in OECD countries when evaluating the effectiveness of R&D tax policy,
in 20 manufacturing and services industries between one should take into account the social returns from
2009 and 2012. global knowledge spillovers.

Box 4.3. The Role of Foreign Aid in Improving Productivity in Low-Income Developing Countries
International technology transfers through such Research has shown that, at the macro level, foreign
channels as trade, foreign direct investment (FDI), and aid can help technology transfers and boost productivity
technology licensing are an effective way to acquire in low-income countries. For instance, Walley and Cush-
technology and improve productivity (Hoekman, ing (2013) find that as well as trade, foreign aid in the
Maskus, and Saggi 2005). But low-income countries are form of technical cooperation and overseas development
less likely to be recipients of international technology assistance grants are important channels through which
transfers through these channels. This is because they research and development investment in G7 coun-
tend to be less integrated into the world economy, they tries had a spillover effect on 11 sub-Saharan African
have weaker absorptive capacities, and their technol- countries from 1980 to 2004. Using a similar approach,
ogy needs may differ from the technologies used in Tiruneh, Wamboye, and Sergi (2017) find evidence that
advanced economies (World Bank 2008). While there is foreign aid is a conduit for R&D spillover effects from
a lot of heterogeneity across low-income countries, with nine OECD member countries on labor productivity in
countries in east and south Asia benefiting from their 28 sub-Saharan African countries from 1992 to 2011.
integration into global value chains around China, other While broad growth regression-based studies have
regions still lack integration into world trade (Allard questioned the effectiveness of aid to emerging market
and others 2016). The evidence discussed in this box economies (for example, Rajan and Subramanian
suggests that, where traditional channels of technol- 2008), the new aid allocation strategies of donors are
ogy transfer—such as FDI and integration into world showing positive results in some cases. Foreign aid
trade—are weak, foreign aid can play an important and can boost technology transfers and productivity in
complementary role in bridging the gap (Figure 4.3.1). low-income countries through various channels:
•• Aid for basic infrastructure technologies: Over the
The author of this box is Pankhuri Dutt. years, official development assistance flows to
Figure 4.3.1. Sub-Saharan Africa: Net Foreign Figure 4.3.2. Official Development
Direct Investment and Aid Inflows1 Assistance Commitment by Sector
(Percent of GDP) (Percent)
7 Aid Foreign direct investment Social infrastructure and services

Economic infrastructure and services
Production sectors
6 Others
5 100
90
4
80
3 70
60
2
50
1
40
0 30
1960 65 70 75 80 85 90 95 2000 05 10 15
20
Sources: World Bank, World Development Indicators; and 10
IMF staff calculations.
1
Foreign direct investment refers to net inflows; aid refers to 0
net official development assistance and official aid received. 2005 06 07 08 09 10 11 12 13 14 15 16
Sources: Organisation for Economic Co-operation and

Development, International Development Statistics; and IMF
staff calculations.

Box 4.3 (continued)
Figure 4.3.3. Aid Commitment to Energy infrastructure endowments (see, for example, Vigil
Generation and Wagner 2012; and Donabauer, Meyer, and
(Percent) Nunnenkamp 2016).
•• Targeted aid for sustainable development:
Renewable sources Nonrenewable sources Low-income countries can benefit from tech-
nological advancements that reduce the cost of
100 technology in advanced economies. For instance,
90 the climate change initiatives and commitments
to the United Nations Sustainable Development
80 Goals (SDGs) raised the share of aid to renew-
70 able energy projects (Figure 4.3.3), introducing
new and more efficient technologies that helped
60 reduce the energy intensity (energy use per GDP)
50 in recipient countries (Kretschmer, Hübler, and
Nunnenkamp 2013). Moreover, the evidence
40 suggests that foreign aid combined with technical
30 cooperation has had a substantial and significant
long-term effect on the renewable energy capacity
20 of recipients, whereas foreign aid without technical
10 cooperation brought immediate but short-term
effects (Kim 2014).
0 •• Building absorptive capacity: Aid can also have a
2006 2011 2016
positive impact on the absorptive capacity of the
Sources: Organisation for Economic Co-operation and recipient country when it is channeled to the health
Development, Creditor Reporting System; and IMF staff and education sectors. Donabauer, Herzer, and
calculations. Nunnenkamp (2014) find that aid for education
has a statistically significant and a positive effect on
FDI flows in Latin American countries with lower
economic infrastructure sectors have increased as education outcomes and labor force skills. Similarly,
donors recognized the importance of improving Selaya and Sunesen (2012) find that aid raises the
trade-related infrastructure and productive capacities marginal productivity of private capital when it is
of recipients, including as part of the World Trade allocated to improving the supply of complemen-
Organization’s Aid for Trade initiative beginning tary inputs, such as education, health, energy, and
in 2005 (Figure 4.3.2). Many low-income coun- transport and communication.
tries need significant investments in basic infra- •• Aid as a complement to FDI: Foreign aid can be a
structure such as roads and electricity. Aid, along complementary tool to attracting FDI, both by
with domestic and foreign private investment, is improving conditions for investment, but also
an important source of financing for the develop- as a signaling device. For instance, Garriga and
ment of this sector in these countries. Within the Phillips (2014) find that foreign aid that is not
economic infrastructure sector, the transport and geo-strategically motivated has a statistically sig-
communication, energy, and banking sectors cover nificant and positive association with FDI inflows
almost 94 percent of aid. Aid targeted at infra- in postconflict recipient developing economies.
structure improvements also makes the recipient They suggest that aid allocation in a postconflict
country more attractive for foreign investment by country acts as a reliable and public information
reducing the cost of selling to recipient-country source that improves the credibility of the recipient
consumers and improving their participation in government, as aid comes with a set of financial and
global production links. Recent empirical evidence structural covenants. Empirical evidence suggests
suggests that aid in the infrastructure sector is that aid is most effective in recipients with stable
effective in improving recipient countries’ economic governments and good institutions (Burnside and

Box 4.3 (continued)
Dollar 2000; Collier and Dollar 2002; Dutta, agenda, which refers to the large gap in funding for
Mukherjee, and Roy 2015). the SDGs. China is already using all three channels
Foreign aid is not a substitute for other channels of of aid, trade, and FDI to invest in Africa and has
technology transfer, rather—when used effectively— become the continent’s largest trading partner over
it can help lay conditions that attract foreign direct the past 15 years (Busse, Erdogan, and Mühlen
investment and foster integration into global trade 2016). Africa’s demographic potential makes it key
and value chains. The new trend in aid allocation and to invest in the region and deepen its integration in
utilization is blended finance. This is where develop- the global production networks, both for the devel-
ment finance is used to attract private investments to opment of the region and for the world economy
fund the SDGs as a part of the “Billions to Trillions” more broadly.

Box 4.4. Relationship between Competition, Concentration, and Innovation

The theoretical link between competition and inno- competition appears to interact in important ways with
vation is complex. The early literature on endogenous the degree of intellectual property rights protection—
growth emphasized a Schumpeterian “rent effect,” another determinant of innovators’ rents. For instance,
according to which less product market competition some evidence suggests that stronger product market
increases post-innovation rents for the new incumbent, competition is associated with more innovation only
thus increasing the incentives to innovate. Subsequent when intellectual property rights protection is strong
literature has highlighted the importance of an addi- (Aghion, Howitt, and Prantl 2015). However, while
tional force, the “escape competition” effect: if compet- strong protection motivates multinational companies to
itive pressure is too low and profits are already large, transfer technology across countries, it reduces innova-
a firm’s incentive to exert effort on innovation to get tion in other contexts (Williams 2013; Bilir 2014).
ahead of competitors will be low. In the international A related discussion investigates the relationship
context, the rent and escape competition effects have a between market competition and concentration. Most
wider interpretation. For instance, lower international of the literature focuses on product market concen-
barriers to trade allow innovators to extract larger tration at the industry level, usually proxied by either
rents, as the market size over which they operate is the Herfindahl-Hirschman Index or the concentration
bigger. At the same time, pressure from the pool of ratio (the share of an industry’s sales that goes to the
potential competitors increases, as it is also exerted by top four firms in the industry). Theoretically, higher
foreign firms (Akcigit and others 2017). concentration could be consistent with higher competi-
The empirical literature reflects some of these tive pressure—and possibly also greater innovation—for
conflicting forces. For instance, policies that increase example, if innovative “superstar” firms were more
product market competition have been found to spur likely to appear in more competitive markets (Autor
innovation, but only up to a certain point, after which and others 2017). However, there is empirical evidence
innovation decreases (Aghion and others 2005). Several that suggests that increased concentration in the United
recent papers have examined how innovation rates in States is at least in part linked to reduced competition
advanced economies have been affected by the increased (Grullon, Larkin, and Michaely 2017; Gutierrez and
competitive pressure stemming from globalization Philippon 2017). A final crucial observation is that
and the entry of China into world trade. The effect trends in concentration are sensitive to the definition of
on innovation is found to be positive in Europe and the relevant market. For instance, while concentration
negative in the United States (Autor and others 2016; within some large countries is rising, global concentra-
Bloom, Draca, and Van Reenen 2016). Product market tion appears to be falling, thanks to the increased role
in international markets of firms from emerging market
The author of this box is Roberto Piazza. economies (Freund and Sidhu 2017).

Annex 4.1: Data, Sample, and Variable Definition
Annex Table 4.1.1. List of Variables, Variable Definitions, and Sources1

Variable Definition Source
Patent Flows (international) Patent families with an application in at least two distinct Constructed from PATSTAT
patent offices
Patent Flows (top three) Patent families with an application in at least one of top Constructed from PATSTAT
three patent offices (EPO, JPO, USPTO)
Patent Stock Cumulated patent flows constructed using perpetual Constructed from PATSTAT
inventory method (with discount rate = 10 percent)
R&D Expenditure Spending on research and development, in constant price OECD ANBERD database
PPP US dollar
R&D Stock Cumulated R&D expenditure constructed using perpetual Constructed from OECD ANBERD data
inventory method (with discount rate = 10 percent)
Labor Productivity Real value added per worker, in US dollars Constructed from KLEMS and UNIDO
data
Total Factor Productivity (TFP) TFP adjusted for varying input utilization (see Annex 2 for Constructed from KLEMS data
details)
Trade with China Imports of final goods from China as a share of sector WIOT
gross output
Global Concentration Revenue share of top four firms globally Freund and Sidhu (2017)
Aggregate R&D Stock Cumulated gross domestic expenditure on R&D (in Constructed from OECD data
constant price PPP US dollar), constructed using
perpetual inventory method (with discount rate =
10 percent)
Aggregate Human Capital Average years of schooling Barro-Lee dataset
Product Market Regulation Indicator of regulation in product markets OECD
Sector R&D Intensity R&D spending per worker Constructed from OECD and KLEMS data
Sector Skill Intensity Computed as 1–share of production worker Bureau of Labor Statistics, Occupational
Employment Statistics
Sector Turnover Business churn rate OECD
Technological Specialization Uncentered bilateral correlation between two Constructed based on PATSTAT
country-sectors’ vectors of patent applications in the 23
IPC subsection
Technological Distance Absolute ln-difference between two country-sectors in the Constructed from OECD and KLEMS data
ratio of R&D (in constant PPP terms) per number of
person engaged
Different Country Dummy for an international country pair Mayer and Zignago (2011)
Different Border Dummy for a country pair sharing no common border Mayer and Zignago (2011)
Different Language Dummy for a country pair sharing no common official Mayer and Zignago (2011)
language
International Distance Distance between the capital cities of two countries, zero Mayer and Zignago (2011)
for the same country pair
Bilateral Citations Sum of citations between two country-industry pairs Constructed based on PATSTAT
Global Value Chain (GVC) Eora multi-region input-output database and
World input-output database (2000–12)
Firm Employment Growth Five-year difference of the logarithm of employee count per Bureau van Dijk Orbis (2000–12)
firm
FDI Regulatory Restrictiveness Index Index summarizing regulation restrictions on FDI; range OECD FDI database (2000–12)
from 0 (open) to 1 (closed).
Tariffs UNCTAD TRAINS (2000–12)
IPR, Education, Infrastructure Quality Index, ranging from 1 (lowest) to 7 (best) World Economic Forum (2000–12)
PMR, Institutions Index, ranging from 1 (lowest) to 10 (best) Fraser (2000–12)
1“Notes on CEPII’s distances measures: The GeoDist Database,” CEPII Working Paper 2011–25.
Note: EPO = European Patent Office; IPC = International Patent Classification; JPO = Japan Patent Office; OECD = Organisation for Economic Co-operation and
Development; PMR = product market regulation; PPP = purchasing power parity; R&D = research and development; UNIDO = United Nations Industrial Develop-
ment Organisation; USPTO = United States Patent and Trademark Office; WIOT = World Input-Output Tables.

Annex Table 4.1.2. List of Sectors in Estimation Samples1

ISIC4 Code Sector Description
10–12 Food products, beverages, and tobacco
13–15 Textiles, wearing apparel, leather, and related products
16–18 Wood and paper products, printing, and reproduction of recorded media
19 Coke and refined petroleum products
20–21 Chemicals and chemical products
22–23 Rubber and plastics products, and other non-metallic mineral products
24–25 Basic metals and fabricated metal products, except machinery and equipment
26–27 Electrical and optical equipment
28 Machinery and equipment, not elsewhere classified
29–30 Transport equipment
31–33 Other manufacturing, repair and installation of machinery and equipment
F Construction
62–63 Information technology and other information services
1The construction and information technology services sectors are only included in the first-stage sample.
Annex Table 4.1.3. List of Countries in Estimation Samples1

Regression Advanced Economies Emerging Market Economies
Gravity model of knowledge diffusion sample Australia, Austria, Belgium, Canada, Denmark, China, Czech Republic, Estonia, Hungary,
(with technological distance based on Finland, France, Germany, Ireland, Israel, Mexico, Poland, Slovenia, Slovakia, Turkey
research and development) Italy, Japan, Korea, Netherlands, New
Zealand, Norway, Portugal, Singapore,
Spain, Sweden, Switzerland, United
Kingdom, United States
Alternative gravity model of knowledge Australia, Austria, Belgium, Canada, Denmark, Argentina Brazil, Bulgaria, Chile, China,
diffusion sample (with technological Finland, France, Germany, Ireland, Israel, Colombia, Czech Republic, Estonia,
distance based on value added) Italy, Japan, Korea, Netherlands, New Hungary, India, Indonesia, Malaysia,
Zealand, Norway, Portugal, Singapore, Mexico, Poland, Russia, Slovakia, Slovenia,
Spain, Sweden, Switzerland, United South Africa, Thailand, Turkey, Uruguay,
Kingdom, United States Vietnam
Patent and labor productivity sample Australia, Austria, Belgium, Canada, China, Czech Republic, Estonia, Hungary,
Denmark, Finland, Ireland, Israel, Italy, Mexico, Poland, Slovakia, Slovenia, Turkey
Korea, Netherlands, New Zealand, Norway,
Portugal, Singapore, Spain, Sweden,
Switzerland
Patent and labor productivity sample, Argentina, Bulgaria, Brazil, Chile, China,
expanded emerging market economy Colombia, Czech Republic, Estonia,
sample Hungary, India, Malaysia, Mexico, Poland,
Russia, Slovakia, Slovenia, South Africa,
Turkey, Uruguay
Total factor productivity sample Austria, Denmark, Finland, Italy, Netherlands, Czech Republic, Slovakia
Spain, Sweden
Patent and global value chain sample, Brazil, China, India, Indonesia, Mexico,
emerging market firm level Philippines, Poland, Russia, South Africa,
Thailand, Turkey
1The classification of countries into advanced economies and emerging economies is as of the beginning of the sample period, that is, around 1995. Israel,
Korea, and Singapore all became advanced economies around 1997 and thus are classified as advanced economies in the sample.

Annex 4.2. Determinants of Knowledge Flows: the Pseudo-Poisson-Maximum Likelihood estimator,

Additional Results a natural choice for a gravity-type model with signifi-
This annex provides details and robustness tests of cant data heteroscedasticity, many zero entries, and a
the baseline results presented in the chapter’s “Deter- large number of dummies (Santos Silva and Tenreyro
minants of Knowledge Flows” section. 2006, 2011).
A key summary of the analysis is the predicted
relative frequencies of citations for each country sector
Baseline Results (denoted ϕˆ ).
  The predicted values exclude the fixed
As discussed in the chapter, a gravity model helps effects. Given the exponential function and that all vari-
investigate the determinants of knowledge flows. It ables are zero for the same country sector, the predicted
follows Peri (2005) and models the citations made value will be equal to 1 within, and generally a fraction
in the patents of a given country sector to patents thereof across, different country sectors (see Peri 2005
from the technology frontier as a function of a set for more details). These can be interpreted as the share
of geographic, linguistic, and technological variables. of knowledge that diffuses from the cited to the citing
Dummy variables indicate whether citations involve relative to what diffuses within the cited country sector.
two distinct sectors (diff_sector) or countries (diff_ The baseline estimation focuses on same-sector
country) and whether the countries share a common pairs and restricts the cited countries to members of
border (diff_border) or an official language (diff_lang). the G5 (France, Germany, Japan, United Kingdom,
The regression also includes a measure of the distance United States). The model is estimated for two samples
between countries’ capital cities (dist_int) and differ- (see Annex 4.1):
ences in technological specialization (tech_spec) and •• The first sample uses the log difference of R&D
technological development (tech_dev). While techno- per worker to measure the distance in technologi-
logical development captures the difference in techno- cal development between citing and cited country
logical intensity (measured as the log difference either sector. In this case, the sample of citing countries
in research and development [R&D] or value added includes 23 advanced economies and 9 emerging
per worker), technological specialization captures market economies, reflecting in part the limited
compositional differences in the types of technology availability of sectoral R&D data for emerging mar-
used.29 Defining ϕas the citations, the model can be ket economies.
written as follows: •• To expand coverage of emerging market economies,
the chapter follows Peri (2005) in considering an
ϕ  i,n;j,m = exp[a + ρ  i,n  + ϑ  j,m  + b 1(diff _ sector)  alternative measure of distance in technological
+ b2  (diff _ country) + b3  (diff _ border)  development: the log difference of real value added
per worker between citing and cited country sector.
+ b4  (diff _ lang) + b5  (dist _ int) + b6  (tech _ spec)  This expands the sample to 22 emerging mar-
+ b7  (tech _ dev) + ε  i,n;j,m], (4.3) ket economies.
in which i and n denote the citing country and sector, Annex Table 4.2.1 shows the baseline results
and j and m the cited country and sector. It includes presented in the chapter, based on the R&D measure
country-sector fixed effects for both the citing and of distance in technological development. Column
cited country sector to control for differences in the (1) shows the results for the model estimated as a
amount of innovation, and institutional or cultural cross section during 1995–2014; columns (2) to (5)
factors that might influence the propensity to patent show the results for the model estimated over each
and cite other patents. The model is estimated using five-year subperiod.
In an alternative specification, the difference in
29The technological development is defined based on value
difference in technological specialization is based on com-
positional differences in patent application. Similar to Peri (2005), added per worker instead of R&D spending. While the
for each country sector, a vector is produced for which the cells are effects of geographic variables are generally comparable
the proportions of all patent applications that relate to each of the to those obtained using R&D spending, somewhat
23 International Patent Classification subsections. The variable is
then defined as 1 minus the uncentered correlation between the two more positive (or at least less negative) effects of
country industries’ proportion vectors. differences in technological specialization and develop-

Annex Table 4.2.1. Gravity Model of Knowledge Diffusion: Baseline Results for Different Time Periods
(1) (2) (3) (4) (5)
1995–2014 1995–99 2000–04 2005–09 2010–14
diff_country –0.457*** –0.595*** –0.407*** –0.370*** –0.726***
[–3.69] [–7.45] [–5.18] [–3.78] [–4.52]
diff_border –0.124 –0.333*** 0.0117 0.117 –0.435*
[–0.93] [–4.89] [0.12] [1.09] [–2.53]
diff_lang –0.810*** –0.539*** –0.708*** –0.940*** –0.815***
[–11.96] [–10.42] [–11.70] [–12.61] [–7.66]
dist_int –0.02493 0.017* –0.036** –0.050*** 0.004
[–1.51] [1.96] [–3.02] [–4.51] [0.20]
tech_spec –2.214*** –3.779*** –2.971*** –2.411*** –2.786***
[–3.30] [–8.32] [–5.96] [–4.52] [–4.03]
tech_dev_R&D –0.0655 –0.143*** –0.169*** –0.169*** 0.185
[–0.68] [–3.89] [–3.63] [–3.32] [1.48]
Citing-Country-Industry Fixed Effect Yes Yes Yes Yes Yes
Cited-Country-Industry Fixed Effect Yes Yes Yes Yes Yes
Observations 1,759 1,139 1,263 1,710 1,654
Note: Result from same-sector regression with cited countries limited to the G5 (France, Germany, Japan, United Kingdom, United States) for each sector.
Robust t-statistics (clustered at citing country-industry level) are in brackets.
***p < 0.001, **p < 0.01, *p < 0.05.
Annex Figure 4.2.1. Diffusion of Knowledge from G5 with ment are found in emerging market economies. The
Expanded Emerging Market Economy Sample size and evolution of the predicted use of information
(Predicted share of knowledge that diffuses, average across recipient is, however, very similar to the baseline used (Annex
country-sectors)
Figure 4.2.1).
0.30
Robustness
0.25 This section shows that the baseline results are
robust to different choices of the estimation sample
and other regression specifications. Three main alterna-
0.20
tives are considered:
•• Inclusion of cross-sectoral citations: The sample is
0.15 expanded to include cross-sectoral patent citations
by including, in the gravity equation, a dummy
0.10 diff_sector for the case in which the citing and cited
sectors differ. Annex Figure 4.2.2 presents the regres-
Advanced economies sion result for the share of knowledge that flows
0.05
Emerging market economies from a given country sector (ϕ  ˆ ). As can be expected,
crossing a sectoral barrier entails a significant reduc-
0.00 tion in knowledge diffusion. Accordingly, the aver-
1995 97 99 2001 03 05 07 09 11 13 14
age ϕˆ   now converges to levels just below 10 percent,
Source: IMF staff calculations. roughly half compared with the same-sector setup.
Note: The figure shows the average share of knowledge from G5 that diffuses, The detailed regression results are shown in Annex
based on a same-sector regression with the difference in technological
development based on value added per worker and using interactions to estimate Table 4.2.2.
separate coefficients for emerging markets and advanced economies. G5 = France, •• Inclusion of all countries as source: In this specifica-
Germany, Japan, United Kingdom, United States.
tion all countries in the sample, and not just the

G5, are included as potential sources of knowledge Annex Figure 4.2.2. Reduction of Knowledge Flow with
(for example, all countries are on both the cit- Additional Barriers: Including Cross-Sectoral Citations
(Share of information that diffuses across cumulative and individual
ing and the cited side). The differences with the barriers)
baseline estimation are small (as shown in Annex
Figure 4.2.3), though the effects of most barriers are Cumulative effect Individual effect Baseline
slightly larger than in the baseline, consistent with
the finding that information from nonleaders tends 1.0
to diffuse less (see Peri 2005). 0.9
•• Excluding China from the baseline regression: This
specification is the same as in the baseline, but 0.8
China is excluded from the estimation sample. As 0.7

shown in Annex Figure 4.2.4, the importance of the
0.6
national border is reduced, but this is partly com-
pensated for by the increased importance of technol- 0.5
ogy barriers. Moreover, a shift is observed between 0.4
sharing a border (getting weaker) and international
0.3
distance (getting stronger). Overall, point estimates
ˆ ’s are comparable, suggesting that
and the average ϕ  0.2
the inclusion of China, though important, is not a
0.1
key driver of the results.
0.0
home
diff_sector
diff_country
diff_border
diff_lang
tech_spec 50th
tech_dev 50th
1,000 km
5,000 km
10,000 km
Note: Square reflects baseline from Figure 4.7 for comparison. km = kilometers.
Annex Table 4.2.2. Gravity Model of Knowledge Diffusion: Including Cross-Sectoral Pairs

(1) (2) (3) (4) (5)
1995–2014 1995–99 2000–04 2005–09 2010–14
diff_sector –0.866*** –0.908*** –0.875*** –0.818*** –0.972***
[–5.15] [–4.14] [–3.87] [–5.50] [–5.82]
diff_country –0.490*** –0.672*** –0.496*** –0.466*** –0.560***
[–5.50] [–8.59] [–6.21] [–5.21] [–6.21]
diff_border –0.0735 –0.309*** –0.00757 0.114 –0.292*
[–0.67] [–4.16] [–0.09] [1.16] [–1.97]
diff_lang –0.810*** –0.542*** –0.687*** –0.899*** –0.956***
[–12.90] [–12.20] [–12.19] [–12.50] [–12.20]
dist_int –31.84* 12.03 –35.65*** –54.48*** –7.275
[–2.25] [1.38] [–3.41] [–5.34] [–0.38]
tech_spec –1.926*** –2.086*** –1.887*** –1.906*** –1.886***
[–9.70] [–7.97] [–6.62] [–9.87] [–9.62]
tech_dev_R&D –0.0610 –0.0997*** –0.0866*** –0.0660* –0.0291
[–1.70] [–5.75] [–3.38] [–2.30] [–0.65]
Citing-Country-Industry Fixed Effect Yes Yes Yes Yes Yes
Cited-Country-Industry Fixed Effect Yes Yes Yes Yes Yes
Observations 22,726 14,337 15,930 22,162 21,502
Note: Result from same-sector regression as well as cross-sectoral pairs and cited countries limited to the G5 (France, Germany, Japan, United Kingdom, United
States) for each sector. Robust t-statistics (clustered at citing-country-industry level) are in brackets.
***p < 0.001, **p < 0.01, *p < 0.05.

Annex Figure 4.2.3. Reduction of Knowledge Flow with Annex Figure 4.2.4. Reduction of Knowledge Flow with
Additional Barriers: Unrestricted Cited Sample Additional Barriers: Excluding China from Baseline
(Share of information that diffuses across cumulative and individual (Share of information that diffuses across cumulative and individual
barriers) barriers)
Cumulative effect Individual effect Baseline Cumulative effect Individual effect Baseline
1.0 1.0
0.9 0.9
0.8 0.8
0.7 0.7
0.6 0.6
0.5 0.5
0.4 0.4
0.3 0.3
0.2 0.2
0.1 0.1
0.0 0.0
home
diff_country
diff_border
diff_lang
tech_spec 50th
tech_dev 50th
1,000 km
5,000 km
10,000 km
home
diff_country
diff_border
diff_lang
tech_spec 50th
tech_dev 50th
1,000 km
5,000 km
10,000 km
Source: IMF staff calculations. Source: IMF staff calculations.
Note: Square reflects baseline from Figure 4.7 for comparison. km = kilometers. Note: Square reflects baseline from Figure 4.7 for comparison. km = kilometers.
Annex 4.3. Impact of Foreign Knowledge economies and emerging market recipients shows
on Domestic Innovation and Productivity: that foreign knowledge matters for both groups of
Additional Results for Panel Estimation of countries in boosting innovation—measured by
Long-Term Relationships patenting—and productivity (Annex Tables 4.3.1 and
This annex presents additional discussion and robust- 4.3.2, columns [1] and [2]). Foreign R&D seems
ness tests of the panel estimation results presented in the to play a comparatively more important role for
sections “Impact on Innovation and Productivity” and innovation in emerging market economies, while for
“The Role of Greater International Competition.” advanced economies domestic R&D efforts matter
more. Compared with advanced economies, emerging
Impact on Innovation and Productivity: market recipients also enjoy a stronger productiv-
Additional Robustness ity boost for a given change in the foreign stock of
The chapter estimated the long-term relationship knowledge. Focusing on the dynamics of knowledge
between the stock of foreign research and development diffusion, the impact of foreign knowledge flows on
(R&D) and domestic innovation (measured by patent domestic innovation appears to have increased more
flow) or productivity using a panel data set at the strongly over time in emerging market economies
country-sector-year level. Various robustness exercises (Annex Tables 4.3.1 and 4.3.2, columns [3] and [4]).
were conducted for both the impact on innovation •• Dynamics of knowledge diffusion: The increase
(Annex Table 4.3.1) and on productivity (Annex over time in the coefficient on foreign R&D in
Table 4.3.2). The results are summarized below. the innovation equation is robust to restricting
•• Advanced economies versus emerging market econo- the sample to be roughly balanced (that is, keep-
mies: Splitting the estimation sample into advanced ing only country sectors with a long period) to

Annex Table 4.3.1. Impact of Foreign Knowledge on Domestic Innovation: Robustness

(1) (2) (3) (4) (5) (6) (7) (8) (9)
Changing Changing EM Top Three Sector-
AE EM Diffusion- Diffusion- Recipients- Dynamic Patent Trade Year
Variables Recipients Recipients AE EM Broad OLS Families Weight Fixed Effect
Foreign R&D Stock, Weighted 0.353*** 0.342*** 0.232*** 0.115 0.240*** 0.298*** 0.359*** 0.240*** 0.508***
[0.070] [0.088] [0.078] [0.085] [0.078] [0.070] [0.057] [0.033] [0.113]
Foreign R&D Stock*2000–04 0.125*** 0.239***
[0.034] [0.064]
Foreign R&D Stock*2005–09 0.184*** 0.280***
[0.044] [0.076]
Foreign R&D Stock*2010–14 0.249*** 0.353***
[0.056] [0.083]
Own R&D Stock 0.477*** 0.361*** 0.440*** 0.346*** 0.410*** 0.464*** 0.468*** 0.724***
[0.077] [0.089] [0.091] [0.107] [0.042] [0.064] [0.066] [0.039]
Aggregate R&D Stock* 0.130***
Sector R&D Intensity [0.042]
Human Capital* 0.139*
Sector Skill Intensity [0.073]
Observations 2,345 1,142 2,132 940 2,115 1,605 3,468 3,021 3,487
R 2 0.750 0.707 0.747 0.723 0.646 0.323 0.790 0.794 0.758
Country-Year Fixed Effect Yes Yes Yes Yes Yes Yes Yes Yes No
Sector-Year Fixed Effect No No No No No No No No Yes
Note: AE = advanced economy; EM = emerging market; OLS = ordinary least squares; R&D = research & development.
Robust standard errors (clustered at country-sector level) in brackets.
***p < 0.01, **p < 0.05, *p < 0.1.
Annex Table 4.3.2. Impact of Foreign Knowledge on Domestic Labor Productivity: Robustness

(1) (2) (3) (4) (5) (6)
Changing Changing EM
Variables AE Recipients EM Recipients Diffusion-AE Diffusion-EM Recipients-Broad Dynamic OLS
Foreign R&D Stock, 0.039** 0.080** 0.021 0.074 0.073** 0.065**
Weighted (Lagged) [0.017] [0.040] [0.020] [0.046] [0.031] [0.032]
Foreign R&D Stock*2000–04 0.027** 0.060***
[0.011] [0.021]
Foreign R&D Stock*2005–09 0.050*** 0.062**
[0.018] [0.029]
Foreign R&D Stock*2010–14 –0.006 –0.034
[0.033] [0.055]
Own R&D Stock (Lagged) 0.133*** 0.103*** 0.123*** 0.108*** 0.133***

[0.022] [0.037] [0.025] [0.038] [0.023]
Aggregate R&D Stock* 0.039*
Sector R&D Intensity [0.022]
Human Capital* 0.035
Sector Skill Intensity [0.064]
Observations 1,968 1,753 1,751 1,511 2,248 1,785
R 2 0.619 0.693 0.633 0.725 0.992 0.067
Country-Year Fixed Effect Yes Yes Yes Yes Yes Yes
Note: AE = advanced economy; EM = emerging market; OLS = ordinary least squares; R&D = research & development.
***p < 0.01, **p < 0.05, *p < 0.1.

avoid sample composition effects. In addition, weights are constructed as imports of goods from
period-by-period estimation, which allows all coef- the originating country sector as a share of gross
ficients to vary over time, yields similar results. The output (Annex Table 4.3.1, column [8]).
subperiod coefficients on the foreign R&D stock •• Fixed effects: While the baseline specifications use
are all statistically significant. country-year fixed effects, in line with Peri (2005),
•• Expanded emerging market sample: Given that the the results are robust to using sector-year fixed
availability of sector-level R&D data limits the sample effects instead, which can capture sector-specific
to a small number of emerging market economies, developments that are common across countries.32
an alternative specification is estimated for a larger The coefficients on both foreign and domestic R&D
number of emerging market economies, in which the become significantly larger under the specification
domestic-sector-level R&D stock is replaced by the with sector-year fixed effects (Annex Table 4.3.1,
domestic aggregate R&D stock interacted with a sec- column [9]).
tor’s R&D intensity.30 The specification also controls •• Calculation of contributions: To calculate the con-
for a measure of human capital (that is, aggregate years tribution of foreign knowledge to productivity, the
of schooling interacted with a sector’s skill intensity).31 estimated coefficient on foreign R&D is applied
The results regarding the economic significance of the to the average annual change in the variable over
foreign R&D stock also hold for this larger sample the relevant period. The contributions by country
(Annex Tables 4.3.1 and 4.3.2, column [5]). groups are obtained from separate regression esti-
•• Dynamic ordinary least squares (OLS): Given that the mates for advanced economies and emerging market
R&D stock and patent/labor productivity series are recipients, and those by subperiods are obtained
possibly nonstationary and cointegrated, the baseline from the regression specification in which the
specification is reestimated using dynamic OLS (see coefficient on foreign R&D stock is allowed to vary
Kao and Chiang 2001). The procedure essentially over time. Only “long panels” (country sectors with
involves adding several lags and leads of the change ample coverage over time) are included in the calcu-
in the regressors and requires a strongly balanced lation of contributions to make sure that changes in
sample. The number of lags chosen is two, and the sample composition do not affect the results.
number of leads is one. The baseline results hold for
both the innovation and labor productivity specifica-
tions, with a slightly larger coefficient on the foreign The Role of Greater International Competition: Results
R&D stock (Annex Tables 4.3.1 and 4.3.2, column and Robustness
[6]). For the total factor productivity specification, Within the same framework used to estimate the
the balanced-sample requirement significantly reduces impact of foreign knowledge on domestic innovation,
the degrees of freedom, and thus the dynamic OLS the impact of competition and market concentration
estimation was not performed. on domestic innovation and the strength of technology
•• Alternative patent measure: While the baseline uses diffusion are also estimated. Annex Table 4.3.3 presents
international patent families, the results are very these estimates for measures that affect the extent of
similar using patent families with at least one appli- competition: trade with China, global market concen-
cation at one of the top three patent offices, which tration, and product market regulation.
is another measure of quality-adjusted patent counts •• Trade with China is measured as imports of final
(Annex Table 4.3.1, column [7]). goods from China as a share of the receiving coun-
•• Alternative weighting method: The baseline results are try sector’s gross output, calculated from the World
robust to using the (time-varying) bilateral trade links Input-Output Tables. This variable increases domestic
between country sectors in place of the predicted innovation directly, but also indirectly, by increasing
share of knowledge flow (ϕ  ˆ ) based on cross-patent technology diffusion (Annex Table 4.3.3, column [1]).
citations. For each receiving country sector, the trade Alternative measures using final goods trade from
the Organisation for Economic Co-operation and
30The correlation between sector-level R&D stock and this
interacted variable is about 0.49 (calculated over country sectors for
which both are available). The sector’s R&D intensity used in the 32The inclusion of both country-year and sector fixed effects
interaction term to create sectoral variation is based on US data. removes most of the variation in the data, and thus the results are
31The sector’s skill intensity is based on US data. not discussed here.

Annex Table 4.3.3. Impact of Competition on Innovation

Variables (1) (2) (3) (4)
Foreign R&D Stock 0.337*** 0.413*** 0.335*** 0.405***
[0.054] [0.046] [0.045] [0.075]
Own R&D Stock 0.494*** 0.435*** 0.447*** 0.478***
[0.063] [0.055] [0.061] [0.059]
China Trade 2.465*** 2.086***
[0.777] [0.758]
Foreign R&D Stock*China Trade 1.474*** 1.236***
[0.442] [0.394]
Global Concentration -4.021*** -4.059***
[0.923] [0.879]
Foreign R&D Stock*Global Concentration -2.121*** -2.27***
[0.559] [0.565]
PMR*Firm Turnover -0.021*** 0.02
[0.007] [0.019]
Foreign R&D Stock*(PMR*Firm Turnover) -0.01*** 0.004
[0.003] [0.008]
Observations 2,281 1,559 2,533 1,175
R 2 0.801 0.819 0.789 0.832
Country-Year Fixed Effect Yes Yes Yes Yes
Note: PMR = product market regulation; R&D = research and development.
***p < 0.01, **p < 0.05, *p < 0.1.
Development (OECD) Structural Analysis Database that changes in global concentration (at the sector
or total goods trade from the COMTRADE Data- level) are not just picking up this global trend.
base yield similar estimates. Interestingly, measures of •• Domestic competition is proxied by the OECD
imports of inputs from China do not seem to matter indicator of product market regulation (interpolated
for innovation, suggesting that the effect comes from between available years). As the indicator is only avail-
the competition channel, which is better captured by able at the country level, a difference-in-difference
trade in final goods. approach is used, in which product market regulation
•• Global market concentration is measured for each sec- is interacted with the sectoral turnover rate for the
tor as the global market share of the four largest firms United States (proxied by the average business churn
based on sales. It is calculated from the firm-level rate collected from the OECD). The assumption
Orbis data set made available by Freund and Sidhu underlying this strategy is that sectors with higher
(2017), following their methodology, which uses the turnover are more likely to be affected by regulation
largest 650 firms globally by revenue in each sector. that restricts firm entry and exit. The coefficients on
Only data for 2006 and 2014 are available, and val- both the main and interaction terms are statistically
ues for the years in between are interpolated for use significant in themselves, but become insignificant
in the regression. Global concentration has a negative when all competition variables enter the regression
impact on domestic innovation, directly and through simultaneously (Annex Table 4.3.3, columns [3] and
lower technology diffusion (Annex Table 4.3.3, col- [4]). Alternative measures of domestic concentration
umn [2]). Alternative measures such as the Herfind- based on patent data produce similar results, although
ahl Index or patent-based concentration measures their reverse causality risk may be higher.
calculated from PATSTAT data bring similar results.33 •• Additional variables: In addition to the baseline
Results are also robust to including an interaction regressors presented in Annex Table 4.3.3, education
term between foreign R&D and time dummies, and intellectual property rights protection were also
which would control for the possible presence of a considered as alternative independent variables. These
global trend in technology diffusion. This ensures measures seemed to matter for innovation and tech-
nology diffusion when included individually, but their
33However, the patent-based measures may underestimate the
extent of concentration because the PATSTAT database does not

significance was not robust to controlling for other pol-
have information on firms’ ownership structure. icies and structural factors. The results are thus omitted.

Annex 4.4. Methodology for Local Projection The specification considered corresponds to the
Method Estimation differencing of both productivity and hours. More pre-
This annex presents the estimation framework cisely, a vector autoregression (VAR) of the following
for the local projection method analysis used in the form is first estimated,
section “Impact on Innovation and Productivity” and yt  = A 1 y t − 1  +  . . . + Ap  yt − p
    + u t, (4.5)
in which y t = [    ], with ∆ xt  the change in log labor
explains the identification of productivity shocks used ∆ xt 
in this framework. ∆ nt 
productivity (measured as gross value added per hour)
Estimation Framework and ∆ nt  the change in log hours. The lag order p is
To examine the short-term dynamics of technology selected according to an Akaike information criterion,
diffusion, the impulse response of productivity and which, for annual data, in virtually all cases returns a
innovation to a technology shock in leader countries is value p = 1.
estimated using the following equation, one for each The identification of structural innovations is
time horizon h (h = 1, . . . ,5): achieved by setting restrictions on the impact matrix B
defined implicitly by
dlnY  i,c,t + h = α  h ω  i,c,l,t dln Yi,l,t
    + β   h    Xi,c,t − 1
   
ut  = B ϵ  t, (4.6)
in which ϵ  t = [  m ]is the vector of structural inno-
+ θ  ct  + ε  i,c,t, (4.4) ϵ  zt 
ϵ  t 
in which i denotes the sector, c the country receiving vations with covariance equal to the identity matrix.
the spillovers, l the technological leader, and t the The restrictions on Bare placed so that a nontechno-
time period. dlnY  i,c,t + h = lnY  i,c,t + h  −  lnY  i,c,t − 1 is the logical innovation, represented by a shock ϵ  m t  , has no
change in Y in the recipient between period t – 1 and long-term effect on x t. By premultiplying the estimated
t + h and dln Yi,l,t
  = lnY  i,l,t  −  lnY  i,l,t − 1is the shock in vector of reduced form shocks u  ˆ   t for B  −1, the above
the leader, in which the variable under investigation, Y, equation can be used to calculate the vector of esti-
could be either total factor productivity, labor produc- mated structural innovations ϵ  ˆ   t.
tivity, or the patent stock of a country sector. Similarly Finally, the series of technological shocks e  ˆ   zt  is
to the long-term approach, the shock is weighted using retrieved as the sequence of technological impacts on
bilateral country-sector weights (ω  i,c,l,t) reflecting the labor productivity:
strength of linkages between the receiving and the orig-
eˆ    zt  = B(1,1) ϵˆ   zt  . (4.7)
inating country sectors. X i,c,t − 1is a vector of controls,
including two lags of the shock in the leaders and two The data for the estimation are obtained by merg-
lags of the growth rate of domestic total factor pro- ing the ISIC 3 and ISIC 4 versions of the KLEMS
ductivity.34 Finally, θ  ctdenotes the country-year fixed data set for the G5 countries (France, Germany,
effects, capturing time-varying factors driving pro- Japan, United Kingdom, United States). Due to data
ductivity and innovation trends at the country level, availability, only the manufacturing and construction
such as the business cycle. The impulse response to a sectors are considered. For the various country-sector
technology shock in the leader countries is constructed pairs, the available data are annual and span about
from a sequence of parameter estimates{α  h}   h = 1 5       and 1970–2015 (only shocks for 1995–2015 are used in
the associated standard errors (see Jordà 2005). the local projection estimation).
Identification of Labor Productivity Shocks Identification of Total Factor Productivity Shocks

Shocks to labor productivity are identified using a The measure of total factor productivity (TFP) that
structural vector autoregression with long-term restric- enters the local projection estimation (both as shocks
tions as in Galí (1999). The estimation is performed in the leaders and as TFP in the recipients) are changes
using the vars package in R. in utilization-adjusted TFP, which is TFP adjusted
for varying input utilization, nonconstant returns
and imperfect competition following Basu, Fernald,
34Including the leads of the shock, as in Teulings and Zubanov
and Kimball (2006) to obtain a measure of “purified”
(2014), to correct for possible misspecification does not change the technology shocks. The adjustment involves estimating
results materially. a production function at the sector level. In particular,

for sector i, which belongs to a group k (k = durable A Firm-Level Analysis.” It also discusses robustness
manufacturing, nondurable manufacturing, or of the results, the instrumentation strategy, and the
nonmanufacturing): examination of the effect of institutional variables on
firm-level innovation.
dy  i = γ  k dx  i  + β  k d hi    +  dz  i, (4.8)
in which dy  iis the growth rate of real gross output; Estimation Framework
dx  i = sk i d ki    + sl i d li    + sm i d m iis the growth rate of the The country-sector-firm-level analysis in the section
composite input (consisting of capital, labor, and mate- on global value chains and patenting follows the
rials), with sk, sl, and sm denoting the share of each framework developed by Autor and others (2016)
input in gross output; d hi  is the growth rate of hours and Bloom, Draca, and Van Reenen (2016). To assess
worked (measured as the first difference in detrended whether changes in global value chain (GVC) partici-
log hours)—a proxy for unobserved input utilization; pation at the sectoral level are related to firms’ techno-
and dz  iis the residual/adjusted TFP or a measure of logical change—measured by the change in the patent
industry technology shocks. flow—and growth prospects, measured by the change
The parameters γand βare assumed to be the same in employment, the following equation is estimated:
for all sectors within a group.35 Given the potential
correlation between input growth (dx  i and d hi  ) and ∆  5 Xijkt
  = δ  X Pijk,2000
  s   +  α  X ∆  5 GVC jkt 
technology shocks in the residual, input growth is + γ  X(Pijk,2000
  s   * ∆  5 GVC jkt) 
instrumented using oil prices, growth in real govern-
ment defense spending (for the United States), or + fkt  X    + s jX   + ε  Xijkt  , (4.9)
changes in the cyclically adjusted fiscal balance (for in which the subscript idenotes firms, jdenotes sec-
other advanced economies in the sample) and a mea- tors, kdenotes countries, and t periods.
sure of monetary shocks.36 X = {P  f, N}, in which Nis the logarithm of
The exercise is conducted for 24 manufacturing employment, and P  f and P  sdenote the logarithm
and services sectors in 17 advanced economies37 over of a transformed count of patent flows and stocks,
1995–2015 (the sample period for the United States respectively.38 Pijk,2000  s  
, a firm’s patent stock at the
goes back to 1970). Sector-level data on gross output, beginning of the sample, is a measure of the firm’s
labor, capital, and intermediate input are taken from initial technological intensity. GVC jkt is the standard
the KLEMS database. measure of participation in global value chains in a
given country sector and year, computed as the sum of
(1) the domestic content in exports reused in trading
Annex 4.5. Impact of Global Value Chains partners’ exports (forward linkages) and (2) the foreign
on Firm-Level Patenting: Methodology value added embedded in exports (backward linkages)
and Robustness expressed as a share of gross exports. f  ktX   is a full set
This annex presents the estimation framework of country dummies interacted with year dummies
for the firm-level analysis presented in the section (country-year fixed effects), which are used to capture
“The Impact of Global Value Chains on Patenting: country-specific factors that support the capacity to
innovate, such as education levels and infrastructure
35This is a more restrictive assumption than in Basu, Fernald, and and macroeconomic shocks. s jX  are sector fixed effects,
Kimball (2006), which allows the returns-to-scale parameter (γ) to which control for systematic differences in patenting
differ across all sectors. This assumption allows for better perfor-
and employment trends across industries. ∆ 5 denotes
mance of the instruments.
36For the United States, monetary shocks—identified in a vector
autoregression as in Burnside (1996)—are obtained from Basu, 38To account for the zeros in patent counts when taking log-
Fernald, and Kimball (2006). For other advanced economies in the arithms, the estimation follows Bloom, Draca, and Van Reenen
sample, monetary shocks are estimated as the forecast error of the (2016) and uses the following transformation: P d = ln(1 + pat d),
policy rates, defined as the difference between the actual policy rates in which d = {f, s} and patis the untransformed patent count.
and the rate expected by analysts as of October of the same year Furthermore, data limitations prevent the construction of firm-level
using forecasts from Consensus Economics. This approach follows total factor productivity and labor productivity measures. Other firm
Furceri, Loungani, and Zdzienicka 2016. performance measures, such as return on assets and return on equity,
37Australia, Austria, Belgium, Canada, Denmark, Finland, France, were considered, but concerns about how these measures are affected
Germany, Ireland, Italy, Japan, Korea, Netherlands, Spain, Sweden, by the division of value added between labor and capital ultimately
United Kingdom, United States. excluded them from the analysis.

five-year differences, and the errors (ε  Xijkt  ) are assumed measures—backward linkages, forward linkages,
to be heteroscedastic. lagged measures, and participation only with regard
The data cover 2000–12 for eight manufacturing to advanced economies; (3) using alternative methods
sectors across 11 emerging market and developing of adjusting patent counts for their quality—such as
economies: Brazil, China, India, Indonesia, Mexico, forward citation or family-size weights or focusing only
Philippines, Poland, Russia, South Africa, Thailand, and on granted patents; (4) estimating over a different time
Turkey.39,40 The primary data are drawn from PATSTAT, period—the years after the global financial crisis were
which provides comprehensive coverage of all patenting excluded to ensure the results were not driven by the
firms. Global input-output tables are used to construct shock of the crisis; and (5) excluding from the sample
industry-level GVC participation measures. To examine either China or the electrical and machinery equipment
the employment effect, the PATSTAT data set is merged sector—each accounting for a large share of the sample.
with Orbis to produce a data set of both patenting and
nonpatenting firms. This allows employment data to be Instrumentation
obtained and the reallocation of employment between In the patenting equation, changes in GVC par-
nonpatenting and patenting firms to be examined.41 ticipation are likely to be correlated with the unob-
This framework allows for analysis of two served shocks ( ε  XP
ijkt  ), due to the possibility of reverse
types of effects: causality (innovative firms may be more likely to be
•• A “within-firm” (intensive margin) effect, captured pulled into GVCs because of their high productivity
by coefficient α  X: It measures how changes in GVC and capacity to innovate) or self-selection (firms may
participation relate to firms’ average performance be geared toward GVC participation). Therefore, the
in terms of technology upgrading or employment use of instrumental variables—the restrictiveness of
growth. As discussed in the text, the results indicate foreign direct investment (FDI) regulations, as well as
that α  X > 0,suggesting that increasing GVC partic- changes in FDI restrictions and tariffs—are consid-
ipation increases firm performance. ered to address potential endogeneity.42 The first-stage
•• A “between-firm” (extensive margin) effect, captured regression of the model can be written as
by coefficient γ  X: The latter captures whether, after
2000, the buildup of innovation or job creation   = θ Zkt  n  + 
∆  5 Gjkt   fkt  G   + s jG   + ε  Gijkt ,   (4.10)
associated with increased GVC participation is disin which ∆  5 Gjkt   = {∆  5 GVC jkt, Pijk,2000
  s   * ∆  5 GVC jkt}
proportionately larger for lower-tech firms (γ  X < 0) and Zkt  n   is the vector of instruments. As expected, all
or higher-tech firms (γ  X > 0). The results indicate the coefficients in θhave a negative sign, suggesting
that technological advances have been relatively that with stricter restrictions on FDI or higher tariffs,
larger in initially lower-tech firms (γ  P < 0), whereas integration into GVCs is expected to be lower in the
job growth has been relatively higher in higher-tech subsequent five years. Standard tests indicate that the
firms (γ  N > 0). set of instruments satisfies the exclusion restriction
that the error term be uncorrelated with sectoral-level
The results are robust to a number of tests (Annex changes in tariffs and FDI restrictions, and the degree
Table 4.5.1), including (1) clustering errors at the of restrictiveness of the latter.43
country-industry level; (2) using alternative GVC
42The Durbin-Wu-Hausmann endogeneity test indicates that
39The Czech Republic and the Slovak Republic were originally changes in GVC participation variables—the variable itself and the
included in the sample, but they have been dropped because they interaction term—are indeed endogenous. The components of the
do not have any patenting activity in PATSTAT. Although Poland is FDI restrictions used in the estimation correspond to screening and
currently considered an advanced economy, it is still included in the approval procedures and restrictions on foreign personnel. The cho-
sample because it was not considered a high-income country at the sen instruments can only be matched with five of the eight sectors
start of the sample period. in the primary data set, but rerunning the ordinary least squares
40Food and beverages, textiles and wearing apparel, wood and regression on the subsample for which the instrumental variables
paper, petroleum-chemicals and nonmetallic mineral products, metal estimation is carried out leaves the results broadly unaffected.
products, electrical and machinery, transport equipment and other 43In general, tariffs and FDI restrictions could be correlated with
manufacturing. innovation through channels other than GVCs, such as knowledge

41Initially, the relationship between GVC participation and flows more broadly or changes in the degree of competition. How-
innovative capacity is tested only for patenting firms in the sample. ever, the tests confirm the strength and validity of the instruments,
While the sample of patenting firms is much smaller, the results likely reflecting the difference in aggregation levels between GVC
qualitatively confirm those obtained using the full sample of patent- measures and instruments (sectoral) and patenting (firm level),
ing firms from the original exercise using the PATSTAT data set. making the former more exogenous.

Annex Table 4.5.1. Impact of Global Value Chain Participation on Firm-Level Innovation: Robustness
Dependent Variable Patent Flow (log, five-year difference)
Excluding
Baseline Family-Size Electrical
(Robust Clustered GVC Weighted Granted Excluding and
Errors) Errors GVC Forward Backward Patents Patents China Machinery
Sample Period (2002–12) (1) (2) (3) (4) (5) (6) (7) (8)
Initial Patent Stock (2000) –0.07*** –0.07*** –0.08*** –0.07*** –0.08*** –0.06*** –0.05*** –0.05***
[–91.317] [–5.703] [–111.620] [–90.896] [–90.624] [–82.359] [–48.643] [–50.686]
Within-Firm Effects
GVC Participation (five-year
change) 0.28*** 0.28*** 0.19*** 0.44*** 0.28*** 0.11*** 0.14*** 0.55***
[16.494] [3.133] [9.273] [13.756] [14.356] [7.269] [4.656] [28.131]
Between-Firm Effects
Initial Patent Stock (2000) × –1.31*** –1.31*** –1.03*** –1.42*** –1.36*** –0.94*** –0.08* –1.49***
GVC Participation (five-year
change) [–44.878] [–4.160] [–21.249] [–41.980] [–42.087] [–36.306] [–1.889] [–37.928]
Observations 4,044,066 4,044,066 4,044,066 4,044,066 4,044,066 4,044,066 792,584 1,684,033
R 2 0.026 0.026 0.025 0.026 0.025 0.022 0.025 0.024
Country × Year Fixed Effect Yes Yes Yes Yes Yes Yes Yes Yes
Sector Fixed Effect Yes Yes Yes Yes Yes Yes Yes No
Note: Robust t-statistics in brackets. GVC = global value chain.
***p < 0.01, **p < 0.05, *p < 0.1.
P
Correlation between Country-Year Fixed Effects and fˆ    kt  =
  ω  0  + ω  m Ikt  m   + μ kt , (4.11)
Policy Variables in which I ktm  is a vector containing m institutional
Finally, the extent to which country-specific variables, including a firm’s perceptions of the quality of
factors—estimated using country-year fixed effects infrastructure and education, the strength of the prop-
in equation (4.9) for the patenting variable (fˆ    kt )
P
  — erty rights system, and competition and the rule of law.
capture absorption capacity factors at the country level Annex Table 4.5.2 shows the correlation between
is tested by estimating these institutional variables and the country-year fixed
Annex Table 4.5.2. Relationship between Country-Year Fixed Effects and Selected Policy Variables
(1) (2) (3) (4) (5) (6) (7) (8)
Interconnectedness
Quality of Port Infrastructure 0.01080* 0.01189*
Education [1.754] [1.932]
Quality of Primary Education 0.01308**
[2.590]
Quality of Math and Science Education 0.00668* 0.00733**
Rule of Law [1.875] [2.328]
Protection of Property Rights 0.00407* 0.00553**
[1.955] [2.200]
Integrity of the Legal System 0.00301* 0.00320**
Product Market Regulation [1.906] [2.044]
Licensing Restrictions –0.00346** –0.00329**
[–2.391] [–2.118]
Constant 0.01610 0.01068 0.02919** –0.01562 0.02631** 0.03068*** 0.07200*** 0.02333
[0.701] [0.698] [2.428] [–0.752] [2.256] [3.248] [7.609] [1.122]
Observations 70 60 70 70 110 110 90 90
R 2 0.042 0.089 0.044 0.095 0.031 0.023 0.060 0.128
Note: Robust t-statistics in brackets.
***p < 0.01, **p < 0.05, *p < 0.1.

effects from the estimated patenting relationships. As Bilir, L. Kamran B. 2014. “Patent Laws, Product Life-Cycle
illustrated in the chapter, the results suggest that the Lengths, and Multination Activity.” The American Economic
country-year fixed effects in patenting are positively Review 104: 1979–2013.
correlated with firms’ perceptions of the quality of Bloom, Nicholas, Mirko Draca, and John Van Reenen. 2016.
“Trade Induced Technical Change? The Impact of Chinese
infrastructure and education, the strength of the prop-
Imports on Innovation, IT and Productivity.” Review of Eco-
erty rights system, competition, and rule of law.
nomic Studies 83(1): 87–117.
Bloom, Nicholas, Charles I. Jones, John Van Reenen, and
Michael Webb. 2017. “Are Ideas Getting Harder to Find?”
References NBER Working Paper 23782, National Bureau of Economic
Acharya, Ram C., and Wolfgang Keller. 2009. “Technology Research, Cambridge, MA.
Transfer through Imports.” Canadian Journal of Economics/ Branstetter, Lee. 2001. “Are Knowledge Spillovers International
Revue canadienne d’économique 42 (4): 1411–48. or Intranational in Scope? Microeconomic Evidence from the
Adler, Gustavo, Romain Duval, Davide Furceri, Sinem Kiliç US and Japan.” Journal of International Economics 53: 53–79.
Çelik, Ksenia Koloskova, and Marcos Poplawski-Ribeiro. Brynjolfsson, Erik, Daniel Rock, and Chad Syverson. 2017.
2017. “Gone with the Headwinds: Global Productivity.” IMF “Artificial Intelligence and the Modern Productivity Paradox:
Staff Discussion Note 17/04, International Monetary Fund, A Clash of Expectations and Statistics.” In Economics of Artifi-
Washington, DC. cial Intelligence. Chicago: University of Chicago Press.
Aghion, Philippe, Nick Bloom, Richard Blundell, Rachel Grif- Burnside, Craig. 1996. “Production Function Regressions,
fith, and Peter Howitt. 2005. “Competition and Innovation: Returns to Scale, and Externalities.” Journal of Monetary
An Inverted-U Relationship.” Quarterly Journal of Economics Economics 37 (2): 177–201.
120 (2): 701–28. ———, and David Dollar. 2000. “Aid, Policies, and Growth.”
Aghion, Philippe, Peter Howitt, and Susanne Prantl. 2015. The American Economic Review 90 (4): 847–68.
“Patent Rights, Product Market Reforms, and Innovation.” Busse, Erdogan C., and H. Mühlen. 2016. “China’s Impact on
Journal of Economic Growth 20 (3): 223–62. Africa—The Role of Trade, FDI and Aid.” KYKLOS 69 (2):
Akcigit, Ufuk, Sina Ates, and Giammario Impullitti. 2017. 228–62.
“Innovation and Trade Policy in a Globalized World.” Chen, Sophia, and Estelle Dauchy. 2018. “International Tech-
Unpublished. nology Sourcing and Knowledge Spillovers: Evidence from
Allard, Céline, Jorge I. Canales Kriljenko, Jesus R. Gonza- OECD Countries.” IMF Working Paper 18/51, International
lez-Garcia, Emmanouil Kitsios, Juan P. Trevino, and Wenjie Monetary Fund, Washington, DC.
Chen. 2016. “Trade Integration and Global Value Chains in Coe, David T., Elhanan Helpman, and Alexander W. Hoffmais-
Sub-Saharan Africa: In Pursuit of the Missing Link.” IMF ter. 2009. “International R&D Spillovers and Institutions.”
African Departmental Paper 16/02, International Monetary European Economic Review 53 (7): 723–41.
Fund, Washington, DC. Coelli, Federica, Andreas Moxnes, and Karen Helene
Andrews, Dan, Chiara Criscuolo, and Peter Gal. 2016. “The Ulltveit-Moe. 2016. “Better, Faster, Stronger: Global Innova-
Global Productivity Slowdown, Technology Divergence, and tion and Trade Liberalization.” NBER Working Paper 22647,
Public Policy: A Firm-Level Perspective.” Background Paper National Bureau of Economic Research, Cambridge, MA.
for OECD Global Forum on Productivity. Cohen, Wesley M., and Daniel A. Levinthal. 1989. “Innovation
Audretsch, David B., and Maryann P. Feldman. 1996. “R&D and Learning: The Two Faces of R&D.” Economic Journal 99
Spillovers and the Geography of Innovation and Production.” (397): 569–96.
The American Economic Review 86 (3): 630–40. Collier, Paul, and David Dollar. 2002. “Aid Allocation and Pov-
Autor, David, David Dorn, Gordon H. Hanson, Gary Pisano, erty Reduction.” European Economic Review 46: 1475–500.
and Pian Shu. 2016. “Foreign Competition and Domestic Dernis, Hélène, and Mosahid Khan. 2004. “Triadic Patent Fam-
Innovation: Evidence from US Patents.” NBER Working ilies Methodology.” OECD Science, Technology and Industry
Paper 22879, National Bureau of Economic Research, Cam- Working Papers 2004/02, Organisation for Economic Co-op-
bridge, MA. eration and Development, Paris.
Autor David, David Dorn, Lawrence F. Katz, Christina Pat- Donabauer, Julian, Birigt Meyer, and Peter Nunnenkamp. 2016.
terson, and John Van Reenen. 2017. “The Fall of the Labor “Aid, Infrastructure and FDI: Assessing the Transmission
Share and the Rise of Superstar Firms.” NBER Working Channel with a New Index of Infrastructure.” World Develop-
Paper 23396, National Bureau of Economic Research, Cam- ment 78: 230–45.
bridge, MA. Donaubauer, Julian, Dierk Herzer, and Peter Nunnenkamp.
Basu, Susanto, John Fernald, and Miles Kimball. 2006. “Are 2014. “Does Aid for Education Attract Foreign Investors? An
Technology Improvements Contractionary?” The American Empirical Analysis.” European Journal of Development Research
Economic Review 96 (5): 1418–48. 26: 597–613.

Dutta, Nabamita, Deepraj Mukherjee, and Sanjukta Roy. Jordà, Òscar. 2005. “Estimation and Inference of Impulse
2015. “Re-Examining the Relationship between Domestic Responses Local Projections.” The American Economic Review
Investment and Foreign Aid: Does Political Stability Matter?” 95 (1): 161–82.
International Review of Applied Economics 29 (3): 259–86. Kao, Chihwa, and Min-Hsien Chiang. 2001. “On the Estima-
Duval, Romain, Davide Furceri, Sinem Kiliç Çelik, and Marcos tion and Inference of a Cointegrated Regression in Panel
Poplawski-Ribeiro. Forthcoming. “Productivity Spillovers in Data.” In Nonstationary Panels, Panel Cointegration, and
Advanced Economies.” Dynamic Panels (Advances in Econometrics, Volume 15), edited
Eaton, Jonathan, and Samuel Kortum. 2002. “Technology, by Badi H. Baltagi, Thomas B. Fomby, and R. Carter Hill.
Geography and Trade”, Econometrica (70): 1741–779. Emerald Group Publishing Limited.
Freund, Caroline, and Dario Sidhu. 2017. “Global Competition Katznelson, Ron D. 2008. “Bad Science in Search of Bad Pat-
and the Rise of China.” PIIE Working Paper 17-3. ents.” Federal Circuit Bar Journal 17 (1).
Furceri, Davide, Prakash Loungani, and Aleksandra Zdzienicka. Keller, W. 2004. “International Technology Diffusion.” Journal of
2016. “The Effects of Monetary Policy Shocks on Inequality.” Economic Literature 42:752–82
IMF Working Paper 16/245, International Monetary Fund, Keller, Wolfgang. 2002. “Geographic Localization of Interna-
Washington, DC. tional Technology Diffusion.” The American Economic Review
Galí, Jordi. 1999. “Technology, Employment, and the Business 92 (1): 120–42.
Cycle: Do Technology Shocks Explain Aggregate Fluctua- ———. 2010, “International Trade, Foreign Direct Invest-
tions?” The American Economic Review 89 (1): 249–71. ment, and Technology Spillovers.” In Handbook of
Garriga, Ana Carolina, and Brian J. Phillips. 2014. “Foreign Aid the Economics of Innovation (2): 793–829. Amsterdam:
as Signal to Investors: Predicting FDI in Postconflict Coun- Elsevier.
tries.” Journal of Conflict Resolution 58 (2): 280–306. Kim, Jung Eun. 2014. “Expanding the Usage of Renewable
Gordon, Robert J. 2012. “Is US Economic Growth Over? Falter- Energy Through Innovation and Technology Transfer.” Dis-
ing Innovation Confronts the Six Headwinds.” NBER Work- sertations ALL, Paper 57.
ing Paper 18315, National Bureau of Economic Research, Kretschmer, Betinna, Michael Hübler, and Peter Nunnen-
Cambridge, MA. kamp. 2013. “Does Foreign Aid Reduce Energy and Carbon
Griffith, Rachel, Rupert Harrison, and John Van Reenen. 2006. Intensities of Developing Economies?” Journal of International
“How Special Is the Special Relationship? Using the Impact Development 25: 67–91.
of US R&D Spillovers on UK Firms as a Test of Technology Lopez-Garcia, Paloma, and Daria Taglioni. 2018. “Knowledge
Sourcing.” The American Economic Review 96 (5): 1859–75. Transfers from International Openness in Trade and Invest-
Griffith, Rachel, and Helen Miller. 2011. “Innovation in ment: The European Case.” In Critical Connections: Why
China: The rise of Chinese Inventors in the Production of ECA’s Regional and Global Network of Interconnections Matter
Knowledge.” IFS Working Paper W11/15, Institute for Fiscal for Inclusive Growth and Stability, edited by David M. Gould.
Studies, London. Washington, DC: World Bank.
Griffith, Rachel, Stephen Redding, and John Van Reenen. 2004. MacGarvie, Megan. 2006. “Do Firms Learn from International
“Mapping the Two Faces of R&D: Productivity Growth in a Trade?” Review of Economics and Statistics 88 (1): 46–60.
Panel of OECD Industries.” Review of Economics and Statistics Madsen, Jakob B. 2007. “Evidence for the OECD Countries.”
86 (4): 883–95. Journal of International Economics 72: 464–80.
Grossman, Gene M., and Elhanan Helpman. 1991. Innovation Organisation for Economic Co-operation and Development
in the World Economy. Cambridge, MA: MIT Press. (OECD). 2009. OECD Patent Statistics Manual. OECD
Grullon, Gustavo, Yelena Larkin, and Roni Michaely. 2017. “Are Publishing, Paris.
US Industries Becoming More Concentrated?” Unpublished ——. 2015. The Future of Productivity. OECD Publishing, Paris.
Working Paper. Peri, Giovanni. 2005, “Determinants of Knowledge Flows and
Gutiérrez, Germán, and Thomas Philippon. 2017. “Declining Their Effect on Innovation.” Review of Economics and Statistics
Competition and Investment in the US.” NBER Working 87 (2): 308–22.
Paper 23583, National Bureau of Economic Research, Cam- Rajan, Raghuram G., and Arvind Subramanian. 2008. “Aid
bridge, MA. and Growth: What Does the Cross-Country Evidence
Hoekman, Bernard M., Keith E. Maskus, and Kamal Saggi. Really Show?” Review of Economics and Statistics 90 (4):
2005. “Transfer of Technology to Developing Countries: Uni- 643–65.
lateral and Multilateral Policy Options.” World Development Santos Silva, João, and Silvana Tenreyro. 2006. “The Log of
33 (10): 1587–602. Gravity.” Review of Economics and Statistics 88 (4): 641–58.
Jaffe, Adam B., Manuel Trajtenberg, and Rebecca Henderson. Santos Silva, J.M.C., and Silvana Tenreyro. 2011. “Further
1993. “Geographic Localization of Knowledge Spillovers as Simulation Evidence on the Performance of the Poisson
Evidenced by Patent Citations.” Quarterly Journal of Econom- Pseudo-Maximum Likelihood Estimator.” Economics Letters
ics 108 (3): 577–98. 112 (2): 220–22.

Selaya, Pablo, and Eva Rytter Sunesen. 2012. “Does Foreign Aid Van Looy, Bart, Caro Vereyen, and Ulrich Schmoch 2015. “Pat-
Increase Foreign Direct Investment?” World Development 40 ent Statistics: Concordance IPC V8—NACE Rev. 2 (version
(11): 2155–176. 2.0).” Eurostat Working Paper, Luxembourg.
Squicciarini, M., H. Dernis, and C. Criscuolo. 2013. “Measur- Vigil, Mariana, and Laurent Wagner. 2012. “Does Aid for Trade
ing Patent Quality: Indicators of Technological and Economic Enhance Export Performance? Investigating the Infrastructure
Value.” In OECD Science, Technology and Industry Working Channel.” World Economy 35 (7): 838–68.
Papers 2013/03, OECD Publishing, Paris, http://dx.doi. Walley, Bernard, and Matthew Cushing. 2013. “Development
org/10.1787/5k4522wkw1r8-en. Aid and Foreign R&D Spillovers in Sub-Saharan Africa.”
Stiebale, Joel. 2016. “Cross-Border M&As and Innovative Activ- International Journal of Applied Economics 10 (2): 10–31.
ity of Acquiring and Target Firms.” Journal of International Williams, Heidi L. 2013. “Intellectual Property Rights and
Economics 99:1–15. Innovation: Evidence from the Human Genome.” Journal of
Strokova, Victoria. 2010. “International Property Rights Index: Political Economy 121 (1): 1–27.
2010 Report.” United Nations Conference on Trade and World Bank (WB). 2008. “Global Economic Prospects: Technol-
Development, New York. ogy Diffusion in the Developing World.” Washington, DC.
Teulings, Coen N., and Nikolay Zubanov. 2014. “Is Economic United Nations Conference on Trade and Development (UNC-
Recovery a Myth? Robust Estimation of Impulse Responses.” TAD). 2005. “World Investment Report 2005: Transnational
Journal of Applied Econometrics 29 (3): 497–514. Corporations and the Internationalization of R&D.” New
Tiruneh, Esubalew Alehegn, Evelyn Wamboye, and Bruno York.
S. Sergi. 2017. “Does Productivity in Africa Benefit from Zhao, Minyuan. 2006. “Conducting R&D in Countries with
Advanced Countries’ R&D?” Technology Analysis & Strategic Weak Intellectual Property Rights Protection.” Management
Management 29 (7): 804–16. Science 52: 1185–199.

STATISTICAL APPENDIX
T
he Statistical Appendix presents histori- underlying the projections for selected economies are
cal data as well as projections. It comprises described in Box A1.
seven sections: Assumptions, What’s New, With regard to interest rates, it is assumed that the
Data and Conventions, Country Notes, London interbank offered rate (LIBOR) on six-month
Classification of Countries, Key Data Documentation, US dollar deposits will average 2.4 percent in 2018 and
and Statistical Tables. 3.4 percent in 2019, that three-month euro deposits will
The assumptions underlying the estimates and pro- average –0.3 percent in 2018 and 0.0 percent in 2019,
jections for 2018–19 and the medium-term scenario and that six-month yen deposits will average 0.0 percent
for 2020–23 are summarized in the first section. The in 2018 and 0.1 percent in 2019.
second section presents a brief description of the As a reminder, with respect to the introduction of the
changes to the database and statistical tables since the euro, on December 31, 1998, the Council of the Euro-
October 2017 World Economic Outlook (WEO). The pean Union decided that, effective January 1, 1999, the
third section provides a general description of the data irrevocably fixed conversion rates between the euro and
and the conventions used for calculating country group currencies of the member countries adopting the euro
composites. The fourth section summarizes selected are as follows:
key information for each country. The fifth section
summarizes the classification of countries in the vari-
1 euro = 13.7603 Austrian schillings
ous groups presented in the WEO. The sixth section
= 40.3399 Belgian francs
provides information on methods and reporting stan-
= 0.585274 Cyprus pound1
dards for the member countries’ national account and
= 1.95583 Deutsche marks
government finance indicators included in the report.
= 15.6466 Estonian krooni2
The last, and main, section comprises the statistical
= 5.94573 Finnish markkaa
tables. (Statistical Appendix A is included here; Sta-
= 6.55957 French francs
tistical Appendix B is available online.) Data in these
= 340.750 Greek drachmas3
tables have been compiled on the basis of information
= 0.787564 Irish pound
available through April 2, 2018. The figures for 2018
= 1,936.27 Italian lire
and beyond are shown with the same degree of preci-
= 0.702804 Latvian lat4
sion as the historical figures solely for convenience;
= 3.45280 Lithuanian litas5
because they are projections, the same degree of accu-
= 40.3399 Luxembourg francs
racy is not to be inferred.
= 0.42930 Maltese lira1
= 2.20371 Netherlands guilders
Assumptions
= 200.482 Portuguese escudos
Real effective exchange rates for the advanced econo-
= 30.1260 Slovak koruna6
mies are assumed to remain constant at their average
= 239.640 Slovenian tolars7
levels measured during the period January 26 to Febru-
= 166.386 Spanish pesetas
ary 23, 2018. For 2018 and 2019, these assumptions 1Established on January 1, 2008.
imply average US dollar–special drawing right (SDR) 2Established on January 1, 2011.
conversion rates of 1.454 and 1.464, US dollar–euro 3Established on January 1, 2001.
4Established on January 1, 2014.
conversion rates of 1.240 and 1.254, and yen–US dollar 5Established on January 1, 2015.
conversion rates of 107.5 and 105.7, respectively. 6Established on January 1, 2009.
It is assumed that the price of oil will average $62.31 a 7Established on January 1, 2007.
barrel in 2018 and $58.24 a barrel in 2019.

Established policies of national authorities are assumed See Box 5.4 of the October 1998 WEO for details on
to be maintained. The more specific policy assumptions how the conversion rates were established.
International Monetary Fund | October 2017 215

What’s New Although every effort is made to ensure the WEO data are
relevant and internationally comparable, differences in both
No changes have been introduced for the April 2018
sectoral and instrument coverage mean that the data are
WEO database.
not universally comparable. As more information becomes
Data and Conventions available, changes in either data sources or instrument
coverage can give rise to data revisions that can sometimes
Data and projections for 193 economies form the statisti-
be substantial. For clarification on the deviations in sectoral
cal basis of the WEO database. The data are maintained
or instrument coverage, please refer to the metadata for the
jointly by the IMF’s Research Department and regional
online WEO database.
departments, with the latter regularly updating country
Composite data for country groups in the WEO are
projections based on consistent global assumptions.
either sums or weighted averages of data for individual
Although national statistical agencies are the ultimate
countries. Unless noted otherwise, multiyear averages of
providers of historical data and definitions, international
growth rates are expressed as compound annual rates of
organizations are also involved in statistical issues, with the
change.2 Arithmetically weighted averages are used for all
objective of harmonizing methodologies for the compila-
data for the emerging market and developing economies
tion of national statistics, including analytical frameworks,
group—except data on inflation and money growth, for
concepts, definitions, classifications, and valuation proce-
which geometric averages are used. The following conven-
dures used in the production of economic statistics. The
tions apply:
WEO database reflects information from both national
•• Country group composites for exchange rates, interest
source agencies and international organizations. rates, and growth rates of monetary aggregates are
Most countries’ macroeconomic data presented in the weighted by GDP converted to US dollars at market
WEO conform broadly to the 1993 version of the System exchange rates (averaged over the preceding three
of National Accounts (SNA). The IMF’s sector statistical years) as a share of group GDP.
standards—the sixth edition of the Balance of Payments •• Composites for other data relating to the domes-
and International Investment Position Manual (BPM6), the tic economy, whether growth rates or ratios, are
Monetary and Financial Statistics Manual and Compila- weighted by GDP valued at purchasing power parity
tion Guide (MFSMCG), and the Government Finance as a share of total world or group GDP.3 Annual
Statistics Manual 2014 (GFSM 2014)—have been or are inflation rates are simple percentage changes from the
being aligned with the SNA 2008. These standards reflect previous years, except in the case of emerging market
the IMF’s special interest in countries’ external positions, and developing economies, for which the rates are
financial sector stability, and public sector fiscal positions. based on logarithmic differences.
The process of adapting country data to the new standards •• Composites for real GDP per capita in purchasing
begins in earnest when the manuals are released. However, power parity terms are sums of individual country
full concordance with the manuals is ultimately dependent data after conversion to the international dollar in the
on the provision by national statistical compilers of revised years indicated.
country data; hence, the WEO estimates are only partially •• Unless noted otherwise, composites for all sectors for
adapted to these manuals. Nonetheless, for many countries, the euro area are corrected for reporting discrepan-
the impact on major balances and aggregates of conver- cies in intra-area transactions. Unadjusted annual
sion to the updated standards will be small. Many other GDP data are used for the euro area and for the
countries have partially adopted the latest standards and majority of individual countries, with the exception
will continue implementation over a period of years.1 of Cyprus, Germany, Ireland, and Portugal, which
The fiscal gross and net debt data reported in the
WEO are drawn from official data sources and IMF staff 2 Averages for real GDP and its components, employment, infla-
estimates. While attempts are made to align gross and net tion, factor productivity, GDP per capita, trade, and commodity
prices are calculated based on the compound annual rate of change,
debt data with the definitions in the GFSM, as a result of
except in the case of the unemployment rate, which is based on the
data limitations or specific country circumstances, these simple arithmetic average.
data can sometimes deviate from the formal definitions. 3 See “Revised Purchasing Power Parity Weights” in the July 2014
WEO Update for a summary of the revised purchasing-power-parity-

1 Many countries are implementing the SNA 2008 or European based weights, as well as Box A2 of the April 2004 WEO and Annex
System of National and Regional Accounts (ESA) 2010, and a few IV of the May 1993 WEO. See also Anne-Marie Gulde and Marianne
countries use versions of the SNA older than that from 1993. A Schulze-Ghattas, “Purchasing Power Parity Based Weights for the
similar adoption pattern is expected for the BPM6 and GFSM 2014. World Economic Outlook,” in Staff Studies for the World Economic
Please refer to Table G, which lists the statistical standards adhered to Outlook (Washington, DC: International Monetary Fund, December
by each country. 1993), 106–23.

report calendar adjusted data. For data prior to 1999, •• Argentina’s authorities discontinued the publication of
data aggregations apply 1995 European currency unit labor market data in December 2015 and released new
exchange rates. series starting in the second quarter of 2016.
•• Composites for fiscal data are sums of individual •• Argentina’s and Venezuela’s consumer prices are
country data after conversion to US dollars at the excluded from all WEO group aggregates.
average market exchange rates in the years indicated. •• Greece’s primary balance estimates for 2016 are based
•• Composite unemployment rates and employment on preliminary excessive deficit procedure (EDP)
growth are weighted by labor force as a share of data on an accrual basis (ESA 2010) provided by the
group labor force. National Statistical Service (ELSTAT) as of October
•• Composites relating to external sector statistics are 23, 2017. Fiscal data since 2010 are adjusted in line
sums of individual country data after conversion to with program definitions.
US dollars at the average market exchange rates in the •• India’s real GDP growth rates are calculated as per
years indicated for balance of payments data and at national accounts: for 1998 to 2011, with base year
end-of-year market exchange rates for debt denomi- 2004/05; thereafter, with base year 2011/12.
nated in currencies other than US dollars. •• Against the background of a civil war and weak capaci-
•• Composites of changes in foreign trade volumes and ties, the reliability of Libya’s data, especially medium-
prices, however, are arithmetic averages of percent term projections, is low.
changes for individual countries weighted by the US •• Data for Syria are excluded from 2011 onward because
dollar value of exports or imports as a share of total of the uncertain political situation.
world or group exports or imports (in the preceding •• Projecting the economic outlook in Venezuela, includ-
year). ing assessing past and current economic developments
•• Unless noted otherwise, group composites are com- as the basis for the projections, is complicated by the
puted if 90 percent or more of the share of group lack of discussions with the authorities (the last Article
weights is represented. IV consultation took place in 2004), long intervals
Data refer to calendar years, except in the case of a in receiving data (with information gaps), incomplete
few countries that use fiscal years; Table F lists the econ- provision of information, and difficulties in inter-
omies with exceptional reporting periods for national preting certain reported economic indicators given
accounts and government finance data for each country. economic developments. The fiscal accounts include
For some countries, the figures for 2017 and earlier the budgetary central government and Petróleos de
are based on estimates rather than actual outturns; Table Venezuela, S.A. (PDVSA), and data for 2016–23
G lists the latest actual outturns for the indicators in the are IMF staff estimates. Revenue includes the IMF
national accounts, prices, government finance, and bal- staff’s estimate of foreign exchange profits transferred
ance of payments indicators for each country. from the central bank to the government (buying
US dollars at the most appreciated rate and selling at
Country Notes more depreciated rates in a multitier exchange rate
•• The consumer price data for Argentina before Decem- system) and excludes IMF staff’s estimate of revenue
ber 2013 reflect the consumer price index (CPI) for from PDVSA’s sale of PetroCaribe assets to the central
the Greater Buenos Aires Area (CPI-GBA), while from bank. The effects of hyperinflation and the noted data
December 2013 to October 2015 the data reflect the gaps mean that IMF staff’s projected macroeconomic
national CPI (IPCNu). The government that took indicators need to be interpreted with caution. For
office in December 2015 discontinued the IPCNu, example, nominal GDP is estimated assuming the
stating that it was flawed, and released a new CPI for GDP deflator rises in line with IMF staff’s projection
the Greater Buenos Aires Area on June 15, 2016 (a new of average inflation. Public external debt in relation
national CPI has been disseminated starting in June to GDP is projected using IMF staff’s estimate of the
2017). At its November 9, 2016, meeting, the IMF average exchange rate for the year. Fiscal accounts for
Executive Board considered the new CPI series to be in 2010–23 correspond to the budgetary central govern-
line with international standards and lifted the declara- ment and PDVSA. Fiscal accounts before 2010 cor-
tion of censure issued in 2013. Given the differences in respond to the budgetary central government, public
geographical coverage, weights, sampling, and method- enterprises (including PDVSA), Instituto Venezolano
ology of these series, the average CPI inflation for 2014, de los Seguros Sociales (IVSS - social security), and
2015, and 2016, and end-of-period inflation for 2015 Fondo de Garantía de Depósitos y Protección Bancaria
and 2016 are not reported in the April 2018 WEO. (FOGADE - deposit insurance).
Classification of Countries developing Europe (sometimes also referred to as “central

and eastern Europe”); Latin America and the Caribbean
Summary of the Country Classification
(LAC); the Middle East, North Africa, Afghanistan, and
The country classification in the WEO divides the Pakistan (MENAP); and sub-Saharan Africa (SSA).
world into two major groups: advanced economies Emerging market and developing economies are also
and emerging market and developing economies.4 This classified according to analytical criteria. The analytical
classification is not based on strict criteria, economic criteria reflect the composition of export earnings and a
or otherwise, and it has evolved over time. The objec- distinction between net creditor and net debtor econo-
tive is to facilitate analysis by providing a reasonably mies. The detailed composition of emerging market
meaningful method of organizing data. Table A pro- and developing economies in the regional and analytical
vides an overview of the country classification, showing groups is shown in Tables D and E.
the number of countries in each group by region and The analytical criterion source of export earnings
summarizing some key indicators of their relative size distinguishes between the categories fuel (Standard
(GDP valued at purchasing power parity, total exports International Trade Classification [SITC] 3) and nonfuel
of goods and services, and population). and then focuses on nonfuel primary products (SITCs 0,
Some countries remain outside the country classifica- 1, 2, 4, and 68). Economies are categorized into one of
tion and therefore are not included in the analysis. Cuba these groups when their main source of export earnings
and the Democratic People’s Republic of Korea are exceeded 50 percent of total exports on average between
examples of countries that are not IMF members, and 2012 and 2016.
their economies therefore are not monitored by the IMF. The financial criteria focus on net creditor economies,
net debtor economies, heavily indebted poor countries
General Features and Composition of Groups in (HIPCs), and low-income developing countries (LIDCs).
the World Economic Outlook Classification Economies are categorized as net debtors when their lat-
Advanced Economies est net international investment position, where available,
was less than zero or their current account balance accu-
The 39 advanced economies are listed in Table B.
mulations from 1972 (or earliest available data) to 2016
The seven largest in terms of GDP based on market
were negative. Net debtor economies are further differen-
exchange rates—the United States, Japan, Germany,
tiated on the basis of experience with debt servicing.5
France, Italy, the United Kingdom, and Canada—
The HIPC group comprises the countries that are or
constitute the subgroup of major advanced economies;
have been considered by the IMF and the World Bank
often referred to as the Group of Seven (G7). The
for participation in their debt initiative known as the
members of the euro area are also distinguished as a
HIPC Initiative, which aims to reduce the external debt
subgroup. Composite data shown in the tables for the
burdens of all the eligible HIPCs to a “sustainable” level
euro area cover the current members for all years, even
in a reasonably short period of time.6 Many of these
though the membership has increased over time.
countries have already benefited from debt relief and
Table C lists the member countries of the European
have graduated from the initiative.
Union, not all of which are classified as advanced
The LIDCs are countries that have per capita income
economies in the WEO.
levels below a certain threshold (set at $2,700 in 2016 as
measured by the World Bank’s Atlas method), structural
Emerging Market and Developing Economies
features consistent with limited development and struc-
The group of emerging market and developing econo- tural transformation, and insufficiently close external
mies (154) includes all those that are not classified as financial linkages to be widely seen as emerging market
advanced economies. economies.
The regional breakdowns of emerging market and
developing economies are Commonwealth of Independent 5 During 2012–16, 25 economies incurred external payments
States (CIS); emerging and developing Asia; emerging and arrears or entered into official or commercial bank debt-rescheduling
agreements. This group is referred to as economies with arrears and/or
4 As used here, the terms “country” and “economy” do not always rescheduling during 2012–16.
refer to a territorial entity that is a state as understood by interna- 6 See David Andrews, Anthony R. Boote, Syed S. Rizavi, and Suk-
tional law and practice. Some territorial entities included here are winder Singh, Debt Relief for Low-Income Countries: The Enhanced
not states, although their statistical data are maintained on a separate HIPC Initiative, IMF Pamphlet Series 51 (Washington, DC: Interna-
and independent basis. tional Monetary Fund, November 1999).

Table A. Classification by World Economic Outlook Groups and Their Shares in Aggregate GDP, Exports of Goods
and Services, and Population, 20171
(Percent of total for group or world)
Exports of Goods
GDP and Services Population
Number of Advanced Advanced Advanced
Economies Economies World Economies World Economies World
Advanced Economies 39 100.0 41.3 100.0 63.6 100.0 14.4
United States 36.9 15.3 16.2 10.3 30.6 4.4
Euro Area 19 28.1 11.6 41.4 26.3 31.8 4.6
Germany 7.9 3.3 12.1 7.7 7.8 1.1
France 5.4 2.2 5.4 3.4 6.1 0.9
Italy 4.4 1.8 4.2 2.7 5.7 0.8
Spain 3.4 1.4 3.1 2.0 4.4 0.6
Japan 10.3 4.3 6.1 3.9 11.9 1.7
United Kingdom 5.6 2.3 5.6 3.5 6.2 0.9
Canada 3.4 1.4 3.5 2.3 3.4 0.5
Other Advanced Economies 16 15.7 6.5 27.3 17.3 16.0 2.3
Memorandum
Major Advanced Economies 7 74.0 30.6 53.1 33.8 71.7 10.4
Emerging Emerging Emerging
Market and Market and Market and
Developing Developing Developing
Economies World Economies World Economies World
Emerging Market and Developing Economies 154 100.0 58.7 100.0 36.4 100.0 85.6
Regional Groups
Commonwealth of Independent States2 12 7.6 4.5 7.5 2.7 4.5 3.9
Russia 5.4 3.2 5.0 1.8 2.3 2.0
Emerging and Developing Asia 30 55.2 32.4 49.5 18.0 56.6 48.5
China 31.1 18.2 29.4 10.7 22.0 18.8
India 12.7 7.4 6.0 2.2 20.9 17.9
Excluding China and India 28 11.5 6.7 14.1 5.2 13.8 11.8
Emerging and Developing Europe 12 6.1 3.6 9.9 3.6 2.8 2.4
Latin America and the Caribbean 32 13.1 7.7 14.1 5.1 9.8 8.4
Brazil 4.3 2.6 3.1 1.1 3.3 2.8
Mexico 3.3 1.9 5.3 1.9 2.0 1.7
Middle East, North Africa, Afghanistan, and
Pakistan 23 12.8 7.5 14.6 5.3 10.9 9.3
Middle East and North Africa 21 11.3 6.6 14.2 5.2 7.2 6.1
Sub-Saharan Africa 45 5.1 3.0 4.4 1.6 15.3 13.1
Excluding Nigeria and South Africa 43 2.6 1.5 2.6 0.9 11.4 9.8
Analytical Groups3
By Source of Export Earnings
Fuel 28 17.9 10.5 20.8 7.6 11.8 10.1
Nonfuel 125 82.1 48.2 79.2 28.9 88.2 75.5
Of Which, Primary Products 31 4.8 2.8 5.1 1.9 8.4 7.2
By External Financing Source
Net Debtor Economies 121 49.8 29.2 46.5 16.9 67.1 57.4
Net Debtor Economies by Debt-
Servicing Experience
Economies with Arrears and/or Rescheduling
during 2012–16 25 3.3 1.9 2.3 0.8 5.7 4.8
Other Groups
Heavily Indebted Poor Countries 39 2.5 1.4 1.9 0.7 11.5 9.8
Low-Income Developing Countries 59 7.1 4.2 6.7 2.4 22.7 19.4
1The GDP shares are based on the purchasing-power-parity valuation of economies’ GDP. The number of economies comprising each group reflects those
for which data are included in the group aggregates.

2Georgia, Turkmenistan, and Ukraine, which are not members of the Commonwealth of Independent States, are included in this group for reasons of
geography and similarity in economic structure.

3Syria is omitted from the source of export earnings and South Sudan and Syria are omitted from the net external position group composites because of
insufficient data.

Table B. Advanced Economies by Subgroup

Major Currency Areas
United States
Euro Area
Japan
Euro Area
Austria Greece Netherlands
Belgium Ireland Portugal
Cyprus Italy Slovak Republic
Estonia Latvia Slovenia
Finland Lithuania Spain
France Luxembourg
Germany Malta
Major Advanced Economies
Canada Italy United States
France Japan
Germany United Kingdom
Other Advanced Economies
Australia Korea Singapore
Czech Republic Macao SAR2 Sweden
Denmark New Zealand Switzerland
Hong Kong SAR1 Norway Taiwan Province of China
Iceland Puerto Rico
Israel San Marino
1On July 1, 1997, Hong Kong was returned to the People’s Republic of China and became a Special
Administrative Region of China.
2On December 20, 1999, Macao was returned to the People’s Republic of China and became a
Special Administrative Region of China.
Table C. European Union

Austria Germany Poland
Belgium Greece Portugal
Bulgaria Hungary Romania
Croatia Ireland Slovak Republic
Cyprus Italy Slovenia
Czech Republic Latvia Spain
Denmark Lithuania Sweden
Estonia Luxembourg United Kingdom
Finland Malta
France Netherlands

Table D. Emerging Market and Developing Economies by Region and Main Source of Export Earnings
Fuel Nonfuel Primary Products
Commonwealth of Independent States
Azerbaijan Uzbekistan
Kazakhstan
Russia
Turkmenistan1
Emerging and Developing Asia
Brunei Darussalam Lao P.D.R.
Timor-Leste Marshall Islands
Mongolia
Papua New Guinea
Solomon Islands
Tuvalu
Latin America and the Caribbean
Bolivia Argentina
Ecuador Chile
Trinidad and Tobago Guyana
Venezuela Paraguay
Suriname
Uruguay
Middle East, North Africa, Afghanistan, and Pakistan
Algeria Afghanistan
Bahrain Mauritania
Iran Morocco
Iraq Sudan
Kuwait
Libya
Oman
Qatar
Saudi Arabia
United Arab Emirates
Yemen
Sub-Saharan Africa
Angola Burkina Faso
Chad Burundi
Republic of Congo Central African Republic
Equatorial Guinea Democratic Republic of the Congo
Gabon Côte d’Ivoire
Nigeria Eritrea
South Sudan Guinea
Guinea-Bissau
Liberia
Malawi
Mali
Sierra Leone
South Africa
Zambia
1Turkmenistan, which is not a member of the Commonwealth of Independent States, is included in this group for reasons of geography and similarity in
economic structure.

Table E. Emerging Market and Developing Economies by Region, Net External Position, and Status as Heavily Indebted Poor Countries
and Low-Income Developing Countries
Low-Income Low-Income
Net External Heavily Indebted Developing Net External Heavily Indebted Developing
Position1 Poor Countries2 Countries Position1 Poor Countries2 Countries
Commonwealth of Independent States Emerging and Developing Europe
Armenia * Albania *
Azerbaijan • Bosnia and Herzegovina *
Belarus * Bulgaria *
Georgia3 * Croatia *
Kazakhstan * Hungary *
Kyrgyz Republic * * Kosovo *
Moldova * * FYR Macedonia *
Russia • Montenegro *
Tajikistan * * Poland *
Turkmenistan3 * Romania *
Ukraine3 * Serbia *
Uzbekistan • * Turkey *
Emerging and Developing Asia Latin America and the Caribbean
Bangladesh * * Antigua and Barbuda *
Bhutan * * Argentina •
Brunei Darussalam • The Bahamas *
Cambodia * * Barbados *
China • Belize *
Fiji * Bolivia * •
India * Brazil *
Indonesia * Chile *
Kiribati • * Colombia *
Lao P.D.R. * * Costa Rica *
Malaysia • Dominica *
Maldives * Dominican Republic *
Marshall Islands * Ecuador *
Micronesia • El Salvador *
Mongolia * Grenada *
Myanmar * * Guatemala *
Nauru * Guyana * •
Nepal • * Haiti * • *
Palau • Honduras * • *
Papua New Guinea * * Jamaica *
Philippines * Mexico *
Samoa * Nicaragua * • *
Solomon Islands * * Panama *
Sri Lanka * Paraguay *
Thailand * Peru *
Timor-Leste • * St. Kitts and Nevis *
Tonga * St. Lucia *
Tuvalu * St. Vincent and the
Vanuatu * Grenadines *
Vietnam * * Suriname *
Trinidad and Tobago •
Uruguay *
Venezuela •
Table E. Emerging Market and Developing Economies by Region, Net External Position, and Status as Heavily Indebted Poor Countries
and Low-Income Developing Countries (continued)
Low-Income Low-Income
Net External Heavily Indebted Developing Net External Heavily Indebted Developing
Position1 Poor Countries2 Countries Position1 Poor Countries2 Countries
Middle East, North Africa, Afghanistan, and Pakistan Democratic Republic of
the Congo * • *
Afghanistan • • *
Republic of Congo * • *
Algeria •
Côte d’Ivoire * • *
Bahrain •
Equatorial Guinea *
Djibouti * *
Eritrea * * *
Egypt *
Ethiopia * • *
Iran •
Gabon •
Iraq •
The Gambia * • *
Jordan *
Ghana * • *
Kuwait •
Guinea * • *
Lebanon *
Guinea-Bissau * • *
Libya •
Kenya * *
Mauritania * • *
Lesotho * *
Morocco *
Liberia * • *
Oman •
Madagascar * • *
Pakistan *
Malawi * • *
Qatar •
Mali * • *
Saudi Arabia •
Mauritius •
Somalia * * *
Mozambique * • *
Sudan * * *
Namibia *
Syria4 ...
Niger * • *
Tunisia *
Nigeria * *
United Arab Emirates •
Rwanda * • *
Yemen * *
São Tomé and Príncipe * • *
Sub-Saharan Africa
Senegal * • *
Angola •
Seychelles *
Benin * • *
Sierra Leone * • *
Botswana •
South Africa •
Burkina Faso * • *
South Sudan4 ... *
Burundi * • *
Swaziland •
Cabo Verde *
Tanzania * • *
Cameroon * • *
Togo * • *
Central African Republic * • *
Uganda * • *
Chad * • *
Zambia * • *
Comoros * • *
Zimbabwe * *
1Dot (star) indicates that the country is a net creditor (net debtor).
2Dot instead of star indicates that the country has reached the completion point, which allows it to receive the full debt relief committed to at the decision point.
3Georgia, Turkmenistan, and Ukraine, which are not members of the Commonwealth of Independent States, are included in this group for reasons of geography and similarity in
economic structure.
4South Sudan and Syria are omitted from the net external position group composite for lack of a fully developed database.

Table F. Economies with Exceptional Reporting Periods1

National Accounts Government Finance
The Bahamas Jul/Jun
Bangladesh Jul/Jun
Barbados Apr/Mar
Belize Apr/Mar
Bhutan Jul/Jun Jul/Jun
Botswana Apr/Mar
Dominica Jul/Jun
Egypt Jul/Jun Jul/Jun
Ethiopia Jul/Jun Jul/Jun
Haiti Oct/Sep Oct/Sep
Hong Kong SAR Apr/Mar
India Apr/Mar Apr/Mar
Iran Apr/Mar Apr/Mar
Jamaica Apr/Mar
Lesotho Apr/Mar Apr/Mar
Malawi Jul/Jun
Marshall Islands Oct/Sep Oct/Sep
Mauritius Jul/Jun
Micronesia Oct/Sep Oct/Sep
Myanmar Oct/Sep Oct/Sep
Nauru Jul/Jun Jul/Jun
Nepal Aug/Jul Aug/Jul
Pakistan Jul/Jun Jul/Jun
Palau Oct/Sep Oct/Sep
Puerto Rico Jul/Jun Jul/Jun
St. Lucia Apr/Mar
Samoa Jul/Jun Jul/Jun
Singapore Apr/Mar
Swaziland Apr/Mar
Thailand Oct/Sep
Trinidad and Tobago Oct/Sep
1Unless noted otherwise, all data refer to calendar years.

Table G. Key Data Documentation

National Accounts Prices (CPI)
System of Use of Chain- Latest

Historical Data Latest Actual National Weighted Historical Data Actual
Country Currency Source1 Annual Data Base Year2 Accounts Methodology3 Source1 Annual Data
Afghanistan Afghan afghani NSO 2016 2002/03 SNA 1993 NSO 2017
Albania Albanian lek IMF staff 2016 1996 SNA 1993 From 1996 NSO 2017
Algeria Algerian dinar NSO 2016 2001 SNA 1993 From 2005 NSO 2017
Angola Angolan kwanza MEP 2015 2002 ESA 1995 NSO 2015
Antigua and Eastern Caribbean CB 2016 20066 SNA 1993 NSO 2016
Barbuda dollar
Argentina Argentine peso NSO 2017 2004 SNA 2008 NSO 2017
Armenia Armenian dram NSO 2016 2005 SNA 2008 NSO 2016
Australia Australian dollar NSO 2017 2015/16 SNA 2008 From 1980 NSO 2017
Austria Euro NSO 2016 2010 ESA 2010 From 1995 NSO 2017
Azerbaijan Azerbaijan manat NSO 2016 2005 SNA 1993 From 1994 NSO 2017
The Bahamas Bahamian dollar NSO 2016 2012 SNA 1993 NSO 2017
Bahrain Bahrain dinar NSO 2016 2010 SNA 2008 NSO 2017
Bangladesh Bangladesh taka NSO 2016 2005 SNA 1993 NSO 2016
Barbados Barbados dollar NSO and CB 2014 19746 SNA 1993 NSO 2017
Belarus Belarusian ruble NSO 2016 2014 SNA 2008 From 2005 NSO 2017
Belgium Euro CB 2016 2015 ESA 2010 From 1995 CB 2016
Belize Belize dollar NSO 2016 2000 SNA 1993 NSO 2016
Benin CFA franc NSO 2014 2007 SNA 1993 NSO 2017
Bhutan Bhutanese NSO 2015/16 2000/016 SNA 1993 CB 2015/16
ngultrum
Bolivia Bolivian boliviano NSO 2016 1990 Other NSO 2017
Bosnia and Bosnia convertible NSO 2016 2010 ESA 2010 From 2000 NSO 2016
Herzegovina marka
Botswana Botswana pula NSO 2015 2006 SNA 1993 NSO 2016
Brazil Brazilian real NSO 2017 1995 SNA 2008 NSO 2017
Brunei Darussalam Brunei dollar NSO and GAD 2016 2010 SNA 1993 NSO and GAD 2017
Bulgaria Bulgarian lev NSO 2016 2010 ESA 2010 From 1996 NSO 2017
Burkina Faso CFA franc NSO and MEP 2016 1999 SNA 1993 NSO 2016
Burundi Burundi franc NSO 2015 2005 SNA 1993 NSO 2017
Cabo Verde Cabo Verdean NSO 2016 2007 SNA 2008 From 2011 NSO 2017
escudo
Cambodia Cambodian riel NSO 2016 2000 SNA 1993 NSO 2017
Cameroon CFA franc NSO 2016 2005 SNA 1993 NSO 2017
Canada Canadian dollar NSO 2016 2007 SNA 2008 From 1980 NSO 2017
Central African CFA franc NSO 2012 2005 SNA 1993 NSO 2015
Republic
Chad CFA franc CB 2017 2005 Other NSO 2017
Chile Chilean peso CB 2017 20136 SNA 2008 From 2003 NSO 2017
China Chinese yuan NSO 2017 2015 SNA 2008 NSO 2017
Colombia Colombian peso NSO 2017 2005 Other From 2000 NSO 2017
Comoros Comorian franc MEP 2015 2000 Other NSO 2015
Democratic Republic Congolese franc NSO 2016 2005 SNA 1993 CB 2016
of the Congo
Republic of Congo CFA franc NSO 2016 1990 SNA 1993 NSO 2017
Costa Rica Costa Rican colón CB 2016 2012 SNA 2008 CB 2016

Table G. Key Data Documentation (continued)

Government Finance Balance of Payments
Statistics
Statistics Manual
Historical Data Latest Actual Manual in Subsectors Accounting Historical Data Latest Actual in Use at
Country Source1 Annual Data Use at Source Coverage4 Practice5 Source1 Annual Data Source
Afghanistan MoF 2017 2001 CG C NSO, MoF, and CB 2017 BPM 5
Albania IMF staff 2016 1986 CG,LG,SS,MPC, Other CB 2016 BPM 6
NFPC
Algeria MoF 2016 1986 CG C CB 2016 BPM 6
Angola MoF 2016 2001 CG,LG Other CB 2016 BPM 6
Antigua and MoF 2016 2001 CG C CB 2016 BPM 6
Barbuda
Argentina MEP 2017 1986 CG,SG,SS C NSO 2017 BPM 6
Armenia MoF 2016 2001 CG C CB 2016 BPM 6
Australia MoF 2016 2014 CG,SG,LG,TG A NSO 2017 BPM 6
Austria NSO 2016 2001 CG,SG,LG,SS A CB 2016 BPM 6
Azerbaijan MoF 2015 Other CG C CB 2016 BPM 6
The Bahamas MoF 2016/17 2001 CG C CB 2017 BPM 5
Bahrain MoF 2016 2001 CG C CB 2017 BPM 6
Bangladesh MoF 2015/16 Other CG C CB 2016 BPM 6
Barbados MoF 2016/17 1986 CG C CB 2016 BPM 5
Belarus MoF 2016 2001 CG,LG,SS C CB 2016 BPM 6
Belgium CB 2016 ESA 2010 CG,SG,LG,SS A CB 2016 BPM 6
Belize MoF 2016/17 1986 CG,MPC Mixed CB 2016 BPM 6
Benin MoF 2016 1986 CG C CB 2016 BPM 6
Bhutan MoF 2015/16 1986 CG C CB 2014/15 BPM 6
Bolivia MoF 2016 2001 CG,LG,SS,NMPC, C CB 2016 BPM 6
NFPC
Bosnia and MoF 2015 2001 CG,SG,LG,SS Mixed CB 2016 BPM 6
Herzegovina
Botswana MoF 2015/16 1986 CG C CB 2015 BPM 5
Brazil MoF 2017 2001 CG,SG,LG,SS, C CB 2017 BPM 6
MPC,NFPC
Brunei Darussalam MoF 2016 Other CG, BCG C NSO, MEP, and GAD 2015 BPM 6
Bulgaria MoF 2016 2001 CG,LG,SS C CB 2016 BPM 6
Burkina Faso MoF 2016 2001 CG CB CB 2016 BPM 6
Burundi MoF 2015 2001 CG A CB 2016 BPM 6
Cabo Verde MoF 2016 2001 CG A NSO 2016 BPM 6
Cambodia MoF 2016 1986 CG,LG A CB 2016 BPM 5
Cameroon MoF 2016 2001 CG,NFPC C MoF 2016 BPM 5
Canada MoF 2016 2001 CG,SG,LG,SS A NSO 2016 BPM 6
Central African MoF 2016 2001 CG C CB 2015 BPM 5
Republic
Chad MoF 2017 1986 CG,NFPC C CB 2015 BPM 6
Chile MoF 2016 2001 CG,LG A CB 2017 BPM 6
China MoF 2017 Other CG,LG C GAD 2017 BPM 6
Colombia MoF 2017 2001 CG,SG,LG,SS Other CB and NSO 2017 BPM 6
Comoros MoF 2016 1986 CG Mixed CB and IMF staff 2016 BPM 5
Democratic Republic MoF 2016 2001 CG,LG A CB 2016 BPM 5
of the Congo
Republic of Congo MoF 2017 2001 CG A CB 2015 BPM 6
Costa Rica MoF and CB 2016 1986 CG C CB 2016 BPM 6



Côte d'Ivoire CFA franc NSO 2015 2009 SNA 1993 NSO 2017
Croatia Croatian kuna NSO 2016 2010 ESA 2010 NSO 2017
Cyprus Euro NSO 2016 2005 ESA 2010 From 1995 NSO 2016
Czech Republic Czech koruna NSO 2017 2010 ESA 2010 From 1995 NSO 2017
Denmark Danish krone NSO 2017 2010 ESA 2010 From 1980 NSO 2017
Djibouti Djibouti franc NSO 2013 1990 Other NSO 2017
Dominica Eastern Caribbean NSO 2015 2006 SNA 1993 NSO 2015
dollar
Dominican Republic Dominican peso CB 2016 2007 SNA 2008 From 2007 CB 2017
Ecuador US dollar CB 2016 2007 SNA 1993 NSO and CB 2017
Egypt Egyptian pound MEP 2016/17 2011/12 SNA 1993 NSO 2016/17
El Salvador US dollar CB 2016 1990 Other NSO 2017
Equatorial Guinea CFA franc MEP and CB 2016 2006 SNA 1993 MEP 2017
Eritrea Eritrean nakfa IMF staff 2006 2005 SNA 1993 NSO 2009
Estonia Euro NSO 2017 2010 ESA 2010 From 2010 NSO 2017
Ethiopia Ethiopian birr NSO 2016/17 2015/16 SNA 1993 NSO 2016
Fiji Fijian dollar NSO 2017 20116 SNA 1993 NSO 2017
Finland Euro NSO 2017 2010 ESA 2010 From 1980 NSO 2017
France Euro NSO 2017 2010 ESA 2010 From 1980 NSO 2017
Gabon CFA franc MoF 2016 2001 SNA 1993 NSO 2017
The Gambia Gambian dalasi NSO 2016 2004 SNA 1993 NSO 2016
Georgia Georgian lari NSO 2016 2000 SNA 1993 From 1996 NSO 2017
Germany Euro NSO 2017 2010 ESA 2010 From 1991 NSO 2017
Ghana Ghanaian cedi NSO 2016 2006 SNA 1993 NSO 2017
Greece Euro NSO 2017 2010 ESA 2010 From 1995 NSO 2017
Grenada Eastern Caribbean NSO 2016 2006 SNA 1993 NSO 2017
dollar
Guatemala Guatemalan CB 2016 2001 SNA 1993 From 2001 NSO 2016
quetzal
Guinea Guinean franc NSO 2016 2010 SNA 1993 NSO 2017
Guinea-Bissau CFA franc NSO 2015 2005 SNA 1993 NSO 2017
Guyana Guyanese dollar NSO 2016 20066 SNA 1993 NSO 2016
Haiti Haitian gourde NSO 2016/17 1986/87 SNA 2008 NSO 2016/17
Honduras Honduran lempira CB 2016 2000 SNA 1993 CB 2016
Hong Kong SAR Hong Kong dollar NSO 2017 2014 SNA 2008 From 1980 NSO 2017
Hungary Hungarian forint NSO 2017 2005 ESA 2010 From 2005 IEO 2017
Iceland Icelandic króna NSO 2017 2005 ESA 2010 From 1990 NSO 2017
India Indian rupee NSO 2016/17 2011/12 SNA 2008 NSO 2016/17
Indonesia Indonesian rupiah NSO 2017 2010 SNA 2008 NSO 2017
Iran Iranian rial CB 2016/17 2011/12 SNA 1993 CB 2016/17
Iraq Iraqi dinar NSO 2016 2007 SNA 1968/93 NSO 2016
Ireland Euro NSO 2017 2015 ESA 2010 From 1995 NSO 2017
Israel New Israeli shekel NSO 2017 2015 SNA 2008 From 1995 NSO 2017
Italy Euro NSO 2016 2010 ESA 2010 From 1980 NSO 2017
Jamaica Jamaican dollar NSO 2016 2007 SNA 1993 NSO 2016


Statistics
Statistics Manual
Côte d'Ivoire MoF 2016 1986 CG A CB 2015 BPM 6
Croatia MoF 2016 2001 CG,LG A CB 2016 BPM 6
Cyprus NSO 2017 ESA 2010 CG,LG,SS A CB 2016 BPM 6
Czech Republic MoF 2016 2001 CG,LG,SS A NSO 2017 BPM 6
Denmark NSO 2016 2001 CG,LG,SS A NSO 2017 BPM 6
Djibouti MoF 2016 2001 CG A CB 2016 BPM 5
Dominica MoF 2015/16 1986 CG C CB 2015 BPM 6
Dominican Republic MoF 2016 2001 CG,SG,LG,SS, NMPC Mixed CB 2016 BPM 6
Ecuador CB and MoF 2016 1986 CG,SG,LG,SS, NFPC C CB 2016 BPM 6
Egypt MoF 2016/17 2001 CG,LG,SS,MPC C CB 2016/17 BPM 5
El Salvador MoF and CB 2017 1986 CG,LG,SS C CB 2016 BPM 6
Equatorial Guinea MoF 2016 1986 CG C CB 2016 BPM 5
Eritrea MoF 2008 2001 CG C CB 2008 BPM 5
Estonia MoF 2017 1986/2001 CG,LG,SS C CB 2016 BPM 6
Ethiopia MoF 2015/16 1986 CG,SG,LG,NFPC C CB 2015/16 BPM 5
Fiji MoF 2016 1986 CG C CB 2016 BPM 6
Finland MoF 2016 2001 CG,LG,SS A NSO 2017 BPM 6
France NSO 2016 2001 CG,LG,SS A CB 2017 BPM 6
Gabon IMF staff 2017 2001 CG A CB 2016 BPM 5
The Gambia MoF 2016 1986 CG C CB and IMF staff 2016 BPM 5
Georgia MoF 2017 2001 CG,LG C NSO and CB 2016 BPM 5
Germany NSO 2017 2001 CG,SG,LG,SS A CB 2017 BPM 6
Ghana MoF 2016 2001 CG C CB 2016 BPM 5
Greece NSO 2016 2014 CG,LG,SS A CB 2017 BPM 6
Grenada MoF 2017 2001 CG CB CB 2016 BPM 6
Guatemala MoF 2016 2001 CG C CB 2016 BPM 6
Guinea MoF 2016 2001 CG Other CB and MEP 2016 BPM 6
Guinea-Bissau MoF 2014 2001 CG A CB 2015 BPM 6
Guyana MoF 2016 1986 CG,SS,NFPC C CB 2016 BPM 5
Haiti MoF 2016/17 2001 CG C CB 2016/17 BPM 5
Honduras MoF 2016 2014 CG,LG,SS,NFPC A CB 2015 BPM 5
Hong Kong SAR NSO 2016/17 2001 CG C NSO 2016 BPM 6
Hungary MEP and NSO 2016 ESA 2010 CG,LG,SS,NMPC A CB 2017 BPM 6
Iceland NSO 2016 2001 CG,SG,SS A CB 2017 BPM 6
India MoF and IMF staff 2016/17 1986 CG,SG C CB 2016/17 BPM 6
Indonesia MoF 2016 2001 CG,LG C CB 2017 BPM 6
Iran MoF 2016/17 2001 CG C CB 2016/17 BPM 5
Iraq MoF 2016 2001 CG C CB 2016 BPM 5
Ireland MoF and NSO 2016 2001 CG,LG,SS A NSO 2017 BPM 6
Israel MoF and NSO 2017 2001 CG,LG,SS Other NSO 2017 BPM 6
Italy NSO 2016 2001 CG,LG,SS A NSO 2016 BPM 6
Jamaica MoF 2016/17 1986 CG C CB 2016 BPM 5



Japan Japanese yen GAD 2017 2011 SNA 2008 From 1980 GAD 2017
Jordan Jordanian dinar NSO 2016 1994 SNA 1993 NSO 2016
Kazakhstan Kazakhstani tenge NSO 2016 2007 SNA 1993 From 1994 CB 2016
Kenya Kenya shilling NSO 2016 2009 SNA 2008 NSO 2017
Kiribati Australian dollar NSO 2016 2006 SNA 2008 NSO 2016
Korea South Korean won CB 2017 2010 SNA 2008 From 1980 NSO 2017
Kosovo Euro NSO 2016 2015 ESA 2010 NSO 2016
Kuwait Kuwaiti dinar MEP and NSO 2016 2010 SNA 1993 NSO and MEP 2016
Kyrgyz Republic Kyrgyz som NSO 2016 2005 SNA 1993 NSO 2017
Lao P.D.R. Lao kip NSO 2016 2012 SNA 1993 NSO 2016
Latvia Euro NSO 2017 2010 ESA 2010 From 1995 NSO 2017
Lebanon Lebanese pound NSO 2013 2010 SNA 2008 From 2010 NSO 2015/16
Lesotho Lesotho loti NSO 2015/16 2012/13 Other NSO 2017
Liberia US dollar CB 2016 1992 SNA 1993 CB 2016
Libya Libyan dinar MEP 2016 2003 SNA 1993 NSO 2017
Lithuania Euro NSO 2017 2010 ESA 2010 From 2005 NSO 2017
Luxembourg Euro NSO 2016 2010 ESA 2010 From 1995 NSO 2016
Macao SAR Macanese pataca NSO 2016 2015 SNA 2008 From 2001 NSO 2017
FYR Macedonia Macedonian denar NSO 2017 2005 ESA 2010 NSO 2017
Madagascar Malagasy ariary NSO 2015 2000 SNA 1968 NSO 2016
Malawi Malawian kwacha NSO 2011 2010 SNA 2008 NSO 2017
Malaysia Malaysian ringgit NSO 2017 2010 SNA 2008 NSO 2017
Maldives Maldivian rufiyaa MoF and NSO 2016 2014 SNA 1993 CB 2017
Mali CFA franc NSO 2016 1999 SNA 1993 NSO 2016
Malta Euro NSO 2017 2010 ESA 2010 From 2000 NSO 2017
Marshall Islands US dollar NSO 2016/17 2003/04 SNA 1993 NSO 2016/17
Mauritania Mauritanian NSO 2014 2004 SNA 1993 NSO 2017
ouguiya
Mauritius Mauritian rupee NSO 2016 2006 SNA 1993 From 1999 NSO 2017
Mexico Mexican peso NSO 2016 2013 SNA 2008 NSO 2017
Micronesia US dollar NSO 2014/15 2004 SNA 1993 NSO 2014/15
Moldova Moldovan leu NSO 2016 1995 SNA 1993 NSO 2017
Mongolia Mongolian tögrög NSO 2016 2010 SNA 1993 NSO 2016
Montenegro Euro NSO 2016 2006 ESA 1995 NSO 2016
Morocco Moroccan dirham NSO 2016 2007 SNA 1993 From 1998 NSO 2017
Mozambique Mozambican NSO 2016 2009 SNA 1993/ NSO 2017
metical 2008
Myanmar Myanmar kyat MEP 2016/17 2010/11 Other NSO 2016/17
Namibia Namibia dollar NSO 2016 2000 SNA 1993 NSO 2016
Nauru Australian dollar Other 2015/16 2006/07 SNA 1993 NSO 2016/17
Nepal Nepalese rupee NSO 2015/16 2000/01 SNA 1993 CB 2016/17
Netherlands Euro NSO 2017 2010 ESA 2010 From 1980 NSO 2017
New Zealand New Zealand dollar NSO 2017 2009/10 Other From 1987 NSO 2017
Nicaragua Nicaraguan CB 2016 2006 SNA 1993 From 1994 CB 2017
córdoba
Niger CFA franc NSO 2016 2000 SNA 1993 NSO 2017
Nigeria Nigerian naira NSO 2017 2010 SNA 2008 NSO 2017
Norway Norwegian krone NSO 2017 2015 ESA 2010 From 1980 NSO 2017


Statistics
Statistics Manual
Japan GAD 2016 2014 CG,LG,SS A MoF 2017 BPM 6
Jordan MoF 2016 2001 CG,NFPC C CB 2016 BPM 5
Kazakhstan NSO 2016 2001 CG,LG A CB 2016 BPM 6
Kenya MoF 2017 2001 CG A CB 2017 BPM 6
Kiribati MoF 2016 1986 CG,LG C NSO 2016 BPM 6
Korea MoF 2017 2001 CG C CB 2017 BPM 6
Kosovo MoF 2016 Other CG,LG C CB 2016 BPM 5
Kuwait MoF 2016 1986 CG Mixed CB 2016 BPM 6
Kyrgyz Republic MoF 2016 Other CG,LG,SS C CB 2016 BPM 5
Lao P.D.R. MoF 2016 2001 CG C CB 2016 BPM 5
Latvia MoF 2017 1986 CG,LG,SS C CB 2017 BPM 6
Lebanon MoF 2015 2001 CG Mixed CB and IMF staff 2015 BPM 5
Lesotho MoF 2016/17 2001 CG,LG C CB 2016/17 BPM 5
Liberia MoF 2016 2001 CG A CB 2016 BPM 5
Libya MoF 2017 1986 CG,SG,LG C CB 2017 BPM 5
Lithuania MoF 2016 2014 CG,LG,SS A CB 2016 BPM 6
Luxembourg MoF 2016 2001 CG,LG,SS A NSO 2016 BPM 6
Macao SAR MoF 2016 2014 CG,SS C NSO 2016 BPM 6
FYR Macedonia MoF 2017 1986 CG,SG,SS C CB 2017 BPM 6
Madagascar MoF 2016 1986 CG,LG C CB 2016 BPM 5
Malawi MoF 2016/17 1986 CG C NSO and GAD 2016 BPM 6
Malaysia MoF 2016 2001 CG,SG,LG C NSO 2017 BPM 6
Maldives MoF 2017 1986 CG C CB 2016 BPM 5
Mali MoF 2016 2001 CG Mixed CB 2016 BPM 6
Malta NSO 2016 2001 CG,SS A NSO 2016 BPM 6
Marshall Islands MoF 2016/17 2001 CG,LG,SS A NSO 2016/17 BPM 6
Mauritania MoF 2017 1986 CG C CB 2016 BPM 5
Mauritius MoF 2016/17 2001 CG,LG,NFPC C CB 2017 BPM 6
Mexico MoF 2017 2014 CG,SS,NMPC,NFPC C CB 2016 BPM 6
Micronesia MoF 2014/15 2001 CG,SG,LG,SS Other NSO 2014/15 Other
Moldova MoF 2016 1986 CG,LG,SS C CB 2016 BPM 5
Mongolia MoF 2016 2001 CG,SG,LG,SS C CB 2016 BPM 6
Montenegro MoF 2016 1986 CG,LG,SS C CB 2016 BPM 6
Morocco MEP 2017 2001 CG A GAD 2017 BPM 6
Mozambique MoF 2016 2001 CG,SG Mixed CB 2017 BPM 6
Myanmar MoF 2016/17 Other CG,NFPC C IMF staff 2016/17 BPM 5
Namibia MoF 2016 2001 CG C CB 2015 BPM 6
Nauru MoF 2016/17 2001 CG Mixed IMF staff 2014/15 BPM 6
Nepal MoF 2015/16 2001 CG C CB 2015/16 BPM 5
Netherlands MoF 2016 2001 CG,LG,SS A CB 2016 BPM 6
New Zealand MoF 2016/17 2001 CG A NSO 2017 BPM 6
Nicaragua MoF 2016 1986 CG,LG,SS C IMF staff 2016 BPM 6
Niger MoF 2017 1986 CG A CB 2017 BPM 6
Nigeria MoF 2017 2001 CG,SG,LG C CB 2016 BPM 6
Norway NSO and MoF 2016 2014 CG,LG,SS A NSO 2017 BPM 6



Oman Omani rial NSO 2015 2010 SNA 1993 NSO 2016
Pakistan Pakistan rupee NSO 2015/16 2005/066 SNA 1968/ NSO 2016/17
1993
Palau US dollar MoF 2015/16 2004/05 SNA 1993 MoF 2015/16
Panama US dollar NSO 2017 2007 SNA 1993 From 2007 NSO 2017
Papua New Guinea Papua New Guinea NSO and MoF 2015 2013 SNA 1993 NSO 2015
kina
Paraguay Paraguayan CB 2016 1994 SNA 1993 CB 2017
guaraní
Peru Peruvian nuevo sol CB 2017 2007 SNA 1993 CB 2017
Philippines Philippine peso NSO 2017 2000 SNA 2008 NSO 2017
Poland Polish zloty NSO 2017 2010 ESA 2010 From 1995 NSO 2017
Portugal Euro NSO 2017 2011 ESA 2010 From 1980 NSO 2017
Puerto Rico US dollar NSO 2014/15 1954 SNA 1968 MEP 2016
Qatar Qatari riyal NSO and MEP 2016 2013 SNA 1993 NSO and MEP 2017
Romania Romanian leu NSO 2016 2010 ESA 2010 From 2000 NSO 2016
Russia Russian ruble NSO 2017 2016 SNA 2008 From 1995 NSO 2017
Rwanda Rwandan franc NSO 2016 2014 SNA 2008 NSO 2017
Samoa Samoa tala NSO 2016/17 2009/10 SNA 1993 NSO 2016/17
San Marino Euro NSO 2016 2007 Other NSO 2017
São Tomé and São Tomé and NSO 2016 2000 SNA 1993 NSO 2016
Príncipe Príncipe dobra
Saudi Arabia Saudi riyal NSO and MEP 2016 2010 SNA 1993 NSO and MEP 2016
Senegal CFA franc NSO 2016 2000 SNA 1993 NSO 2017
Serbia Serbian dinar NSO 2016 2010 ESA 2010 From 2010 NSO 2016
Seychelles Seychellois rupee NSO 2015 2006 SNA 1993 NSO 2016
Sierra Leone Sierra Leonean NSO 2016 2006 SNA 1993 From 2010 NSO 2016
leone
Singapore Singapore dollar NSO 2017 2010 SNA 2008 NSO 2017
Slovak Republic Euro NSO 2017 2010 ESA 2010 From 1997 NSO 2017
Slovenia Euro NSO 2017 2010 ESA 2010 From 2000 NSO 2017
Solomon Islands Solomon Islands CB 2016 2004 SNA 1993 NSO 2016
dollar
Somalia US dollar CB 2016 2012 SNA 1993 CB 2014
South Africa South African rand NSO 2017 2010 SNA 1993 NSO 2017
South Sudan South Sudanese NSO 2016 2010 SNA 1993 NSO 2016
pound
Spain Euro NSO 2017 2010 ESA 2010 From 1995 NSO 2017
Sri Lanka Sri Lankan rupee NSO 2017 2010 SNA 1993 NSO 2017
St. Kitts and Nevis Eastern Caribbean NSO 2016 20066 SNA 1993 NSO 2016
dollar
St. Lucia Eastern Caribbean NSO 2016 2006 SNA 1993 NSO 2016
dollar
St. Vincent and the Eastern Caribbean NSO 2016 20066 SNA 1993 NSO 2016
Grenadines dollar
Sudan Sudanese pound NSO 2015 1996 Other NSO 2015
Suriname Surinamese dollar NSO 2016 2007 SNA 1993 NSO 2017


Statistics
Statistics Manual
Oman MoF 2016 2001 CG C CB 2015 BPM 5
Pakistan MoF 2016/17 1986 CG,SG,LG C CB 2016/17 BPM 5
Palau MoF 2015/16 2001 CG Other MoF 2015/16 BPM 6
Panama MoF 2016 1986 CG,SG,LG,SS, NFPC C NSO 2016 BPM 6
Papua New Guinea MoF 2015 1986 CG C CB 2015 BPM 5
Paraguay MoF 2016 2001 CG,SG,LG,SS,MPC, C CB 2016 BPM 5
NFPC
Peru MoF 2017 1986 CG,SG,LG,SS C CB 2017 BPM 5
Philippines MoF 2017 2001 CG,LG,SS C CB 2016 BPM 6
Poland MoF and NSO 2016 ESA 2010 CG,LG,SS A CB 2016 BPM 6
Portugal NSO 2016 2001 CG,LG,SS A CB 2017 BPM 6
Puerto Rico MEP 2015/16 2001 Other A … … …
Qatar MoF 2015 1986 CG C CB and IMF staff 2017 BPM 5
Romania MoF 2016 2001 CG,LG,SS C CB 2016 BPM 6
Russia MoF 2017 2001 CG,SG,SS Mixed CB 2017 BPM 6
Rwanda MoF 2016 1986 CG,LG Mixed CB 2016 BPM 6
Samoa MoF 2016/17 2001 CG A CB 2016/17 BPM 6
San Marino MoF 2016 Other CG Other … … …
São Tomé and MoF and Customs 2016 2001 CG C CB 2016 BPM 6
Príncipe
Saudi Arabia MoF 2016 2014 CG C CB 2016 BPM 6
Senegal MoF 2016 2001 CG C CB and IMF staff 2017 BPM 6
Serbia MoF 2016 1986/2001 CG,SG,LG,SS C CB 2016 BPM 6
Seychelles MoF 2016 1986 CG,SS C CB 2016 BPM 6
Sierra Leone MoF 2017 1986 CG C CB 2016 BPM 5
Singapore MoF 2016/17 2001 CG C NSO 2017 BPM 6
Slovak Republic NSO 2017 2001 CG,LG,SS A CB 2017 BPM 6
Slovenia MoF 2017 1986 CG,SG,LG,SS C NSO 2017 BPM 6
Solomon Islands MoF 2016 1986 CG C CB 2016 BPM 6
Somalia MoF 2016 2001 CG C CB 2016 BPM 5
South Africa MoF 2016 2001 CG,SG,SS C CB 2016 BPM 6
South Sudan MoF and MEP 2016 Other CG C MoF, NSO, and MEP 2016 BPM 5
Spain MoF and NSO 2016 ESA 2010 CG,SG,LG,SS A CB 2016 BPM 6
Sri Lanka MoF 2017 2001 CG C CB 2016 BPM 5
St. Kitts and Nevis MoF 2016 1986 CG, SG C CB 2016 BPM 6
St. Lucia MoF 2015/16 1986 CG C CB 2016 BPM 6
St. Vincent and the MoF 2016 1986 CG C CB 2016 BPM 6
Grenadines
Sudan MoF 2017 2001 CG Mixed CB 2016 BPM 5
Suriname MoF 2016 1986 CG Mixed CB 2016 BPM 5



Swaziland Swazi lilangeni NSO 2015 2011 SNA 1993 NSO 2017
Sweden Swedish krona NSO 2017 2016 ESA 2010 From 1993 NSO 2017
Switzerland Swiss franc NSO 2017 2010 ESA 2010 From 1980 NSO 2017
Syria Syrian pound NSO 2010 2000 SNA 1993 NSO 2011
Taiwan Province of New Taiwan dollar NSO 2016 2011 SNA 2008 NSO 2017
China
Tajikistan Tajik somoni NSO 2017 1995 SNA 1993 NSO 2017
Tanzania Tanzania shilling NSO 2016 2007 SNA 1993 NSO 2017
Thailand Thai baht MEP 2017 2002 SNA 1993 From 1993 MEP 2017
Timor-Leste US dollar MoF 2016 20156 SNA 2008 NSO 2017
Togo CFA franc NSO 2015 2007 SNA 1993 NSO 2016
Tonga Tongan pa’anga CB 2016 2010 SNA 1993 CB 2016
Trinidad and Tobago Trinidad and NSO 2016 2012 SNA 1993 NSO 2016
Tobago dollar
Tunisia Tunisian dinar NSO 2017 2004 SNA 1993 From 2009 NSO 2016
Turkey Turkish lira NSO 2016 2009 ESA 2010 From 2009 NSO 2017
Turkmenistan New Turkmen NSO 2017 2008 SNA 1993 From 2000 NSO 2017
manat
Tuvalu Australian dollar PFTAC advisors 2015 2005 SNA 1993 NSO 2016
Uganda Ugandan shilling NSO 2016 2010 SNA 1993 CB 2016/17
Ukraine Ukrainian hryvnia NSO 2017 2010 SNA 2008 From 2005 NSO 2017
United Arab U.A.E. dirham NSO 2016 2010 SNA 1993 NSO 2016
Emirates
United Kingdom Pound sterling NSO 2017 2015 ESA 2010 From 1980 NSO 2017
United States US dollar NSO 2017 2009 Other From 1980 NSO 2017
Uruguay Uruguayan peso CB 2016 2005 SNA 1993 NSO 2017
Uzbekistan Uzbek sum NSO 2016 1995 SNA 1993 NSO 2016
Vanuatu Vanuatu vatu NSO 2016 2006 SNA 1993 NSO 2016
Venezuela Venezuelan bolívar CB 2016 1997 SNA 2008 CB 2016
fuerte
Vietnam Vietnamese dong NSO 2017 2010 SNA 1993 NSO 2017
Yemen Yemeni rial IMF staff 2008 1990 SNA 1993 NSO,CB, and 2009
IMF staff
Zambia Zambian kwacha NSO 2016 2010 SNA 1993 NSO 2017
Zimbabwe US dollar NSO 2013 2009 Other NSO 2016


Statistics
Statistics Manual
Swaziland MoF 2016/17 2001 CG A CB 2017 BPM 6
Sweden MoF 2016 2001 CG,LG,SS A NSO 2017 BPM 6
Switzerland MoF 2016 2001 CG,SG,LG,SS A CB 2017 BPM 6
Syria MoF 2009 1986 CG C CB 2009 BPM 5
Taiwan Province of MoF 2016 1986 CG,LG,SS C CB 2016 BPM 6
China
Tajikistan MoF 2017 1986 CG,LG,SS C CB 2016 BPM 6
Tanzania MoF 2016 1986 CG,LG C CB 2016 BPM 5
Thailand MoF 2016/17 2001 CG,BCG,LG,SS A CB 2017 BPM 6
Timor-Leste MoF 2017 2001 CG C CB 2017 BPM 6
Togo MoF 2016 2001 CG C CB 2016 BPM 6
Tonga MoF 2016 2014 CG C CB and NSO 2016 BPM 6
Trinidad and Tobago MoF 2016/17 1986 CG C CB and NSO 2016 BPM 6
Tunisia MoF 2016 1986 CG C CB 2016 BPM 5
Turkey MoF 2016 2001 CG,LG,SS A CB 2017 BPM 6
Turkmenistan MoF 2015 1986 CG,LG C NSO and IMF staff 2015 BPM 6
Tuvalu MoF 2016 Other CG Mixed IMF staff 2012 BPM 6
Uganda MoF 2016 2001 CG C CB 2016 BPM 6
Ukraine MoF 2016 2001 CG,SG,LG,SS C CB 2016 BPM 6
United Arab MoF 2016 2001 CG,BCG,SG,SS C CB 2016 BPM 5
Emirates
United Kingdom NSO 2017 2001 CG,LG A NSO 2017 BPM 6
United States MEP 2017 2014 CG,SG,LG A NSO 2016 BPM 6
Uruguay MoF 2017 1986 CG,LG,SS,MPC, A CB 2016 BPM 6
NFPC
Uzbekistan MoF 2016 Other CG,SG,LG,SS C MEP 2016 BPM 6
Vanuatu MoF 2016 2001 CG C CB 2016 BPM 5
Venezuela MoF 2013 2001 BCG,NFPC C CB 2016 BPM 5
Vietnam MoF 2015 2001 CG,SG,LG C CB 2016 BPM 5
Yemen MoF 2013 2001 CG,LG C IMF staff 2009 BPM 5
Zambia MoF 2016 1986 CG C CB 2016 BPM 6
Zimbabwe MoF 2014 1986 CG C CB and MoF 2016 BPM 6
Note: BPM = Balance of Payments Manual; CPI = consumer price index; ESA = European System of National Accounts; SNA = System of National Accounts.
1CB = central bank; Customs = Customs Authority; GAD = General Administration Department; IEO = international economic organization; MEP = Ministry of Economy, Planning,
Commerce, and/or Development; MoF = Ministry of Finance and/or Treasury; NSO = National Statistics Office; PFTAC = Pacific Financial Technical Assistance Centre.
2National accounts base year is the period with which other periods are compared and the period for which prices appear in the denominators of the price relationships used to
calculate the index.

3Use of chain-weighted methodology allows countries to measure GDP growth more accurately by reducing or eliminating the downward biases in volume series built on index numbers
that average volume components using weights from a year in the moderately distant past.
4For some countries, the structures of government consist of a broader coverage than specified for the general government. Coverage: BCG = budgetary central government; CG =
central government; EUA = extrabudgetary units/accounts; LG = local government; MPC = monetary public corporation, including central bank; NFPC = nonfinancial public corporation;
NMPC = nonmonetary financial public corporation; SG = state government; SS = social security fund; TG = territorial governments.
5Accounting standard: A = accrual accounting; C = cash accounting; CB = commitments basis accounting; Mixed = combination of accrual and cash accounting.
6Base year is not equal to 100 because the nominal GDP is not measured in the same way as real GDP or the data are seasonally adjusted.

Box A1. Economic Policy Assumptions Underlying the Projections for Selected Economies
Fiscal Policy Assumptions on the authorities’ fiscal plans, with adjustments for
The short-term fiscal policy assumptions used in the the IMF staff’s assumptions.
World Economic Outlook (WEO) are normally based on Brazil: Fiscal projections for the end of 2018 take
officially announced budgets, adjusted for differences into account budget performance through January
between the national authorities and the IMF staff 2018, and the deficit target approved in the budget law.
regarding macroeconomic assumptions and projected Canada: Projections use the baseline forecasts in the
fiscal outturns. When no official budget has been 2018 federal budget and the latest provincial bud-
announced, projections incorporate policy measures get updates as available. The IMF staff makes some
that are judged likely to be implemented. The medium- adjustments to these forecasts, including for differ-
term fiscal projections are similarly based on a judg- ences in macroeconomic projections. The IMF staff’s
ment about the most likely path of policies. For cases forecast also incorporates the most recent data releases
in which the IMF staff has insufficient information to from Statistics Canada’s Canadian System of National
assess the authorities’ budget intentions and prospects Economic Accounts, including federal, provincial,
for policy implementation, an unchanged structural and territorial budgetary outturns through the fourth
primary balance is assumed unless indicated otherwise. quarter of 2017.
Specific assumptions used in regard to some of the Chile: Projections are based on the authorities’
advanced economies follow. (See also Tables B5 to B9 in budget projections, adjusted to reflect the IMF staff’s
the online section of the Statistical Appendix for data on projections for GDP and copper prices.
fiscal net lending/borrowing and structural balances.)1 China: Projections assume that the pace of fiscal
Argentina: Fiscal projections are based on the avail- consolidation is likely to be more gradual, reflect-
able information regarding budget outturn and budget ing reforms to strengthen social safety nets and the
plans for the federal and provincial governments, fiscal social security system announced as part of the Third
measures announced by the authorities, and the IMF Plenum reform agenda.
staff’s macroeconomic projections. Denmark: Estimates for 2016 are aligned with the
Australia: Fiscal projections are based on Australian latest official budget numbers, adjusted where appro-
Bureau of Statistics data; the fiscal year 2017/18 bud- priate for the IMF staff’s macroeconomic assumptions.
gets of the Commonwealth and States and Territories; For 2017–18, the projections incorporate key features
the Commonwealth’s 2017 Mid-Year Economic and of the medium-term fiscal plan as embodied in the
Fiscal Outlook and Updates by States and Territories; authorities’ Convergence Programme 2017 submitted
and the IMF staff’s estimates. to the European Union.
Austria: Fiscal projections are based on data from France: Projections for 2017 reflect the budget law
Statistics Austria, the authorities’ projections, and the and cancellation of spending taken in July 2017. For
IMF staff’s estimates and projections. 2018–22, they are based on the multiyear budget and
Belgium: Projections are based on the 2017–20 the 2018 budget, adjusted for differences in assumptions
Stability Programme and other available information on macro and financial variables, and revenue projec-
tions. Historical fiscal data reflect the May and Septem-
1 The output gap is actual minus potential output, as a
ber 2017 revisions and update of the fiscal accounts,
percentage of potential output. Structural balances are expressed
debt data, and national accounts for 2014 and 2015.
as a percentage of potential output. The structural balance is the Germany: The IMF staff’s projections for 2018 and
actual net lending/borrowing minus the effects of cyclical output beyond are based on the 2018 Draft Budgetary Plan
from potential output, corrected for one-time and other factors, and data updates from the national statistical agency,
such as asset and commodity prices and output composition adjusted for the differences in the IMF staff’s mac-
effects. Changes in the structural balance consequently include
effects of temporary fiscal measures, the impact of fluctuations
roeconomic framework and assumptions concerning
in interest rates and debt-service costs, and other noncyclical revenue elasticities. The projections do not include
fluctuations in net lending/borrowing. The computations of policy measures in the new government’s coalition
structural balances are based on the IMF staff’s estimates of agreement published in February 2018. The estimate
potential GDP and revenue and expenditure elasticities. (See of gross debt includes portfolios of impaired assets and
Annex I of the October 1993 WEO.) Net debt is calculated as
gross debt minus financial assets corresponding to debt instru-
noncore business transferred to institutions that are
ments. Estimates of the output gap and of the structural balance winding up, as well as other financial sector and EU
are subject to significant margins of uncertainty. support operations.

Box A1 (continued)
Greece: Greece’s primary balance estimates for 2016 Mexico: Fiscal projections for 2018 are broadly in
are based on preliminary excessive deficit procedure line with the approved budget; projections for 2019
(EDP) data on an accrual basis (European System onward assume compliance with rules established in
of National and Regional Accounts [ESA 2010]) the Fiscal Responsibility Law.
provided by the National Statistical Service (ELSTAT) Netherlands: Fiscal projections for 2017–23 are
as of October 23, 2017. Fiscal data since 2010 are based on the authorities’ Bureau for Economic Policy
adjusted in line with program definition. Analysis budget projections, after differences in macro-
Hong Kong Special Administrative Region: Projec- economic assumptions are adjusted for. Historical data
tions are based on the authorities’ medium-term fiscal were revised following the June 2014 Central Bureau
projections on expenditures. of Statistics release of revised macro data because of
Hungary: Fiscal projections include the IMF staff’s the adoption of ESA 2010 and the revisions of data
projections of the macroeconomic framework and of sources.
the impact of recent legislative measures, as well as fis- New Zealand: Fiscal projections are based on
cal policy plans announced in the 2018 budget. the fiscal year 2017/18 budget and 2017 Half-Year
India: Historical data are based on budgetary execu- Economic and Fiscal Update, and the IMF staff’s
tion data. Projections are based on available informa- estimates.
tion on the authorities’ fiscal plans, with adjustments Portugal: The projections for the current year are
for the IMF staff’s assumptions. Subnational data are based on the authorities’ approved budget, adjusted
incorporated with a lag of up to two years; general to reflect the IMF staff’s macroeconomic forecast.
government data are thus finalized well after central Projections thereafter are based on the assumption of
government data. IMF and Indian presentations differ, unchanged policies.
particularly regarding divestment and license auction Puerto Rico: Fiscal projections are based on the
proceeds, net versus gross recording of revenues in cer- Puerto Rico Fiscal and Economic Growth Plan
tain minor categories, and some public sector lending. (FEGP), which was finalized on February 12, 2018,
Indonesia: IMF projections are based on moderate and is pending certification by the Oversight Board.
tax policy and administration reforms, fuel subsidy In line with assumptions of this plan, IMF projec-
pricing reforms introduced since January 2015, and a tions assume federal aid for rebuilding after Hurricane
gradual increase in social and capital spending over the Maria devastated the island in September 2017. The
medium term in line with fiscal space. projections also assume revenue losses from the fol-
Ireland: Fiscal projections are based on the country’s lowing: elimination of federal funding for Puerto Rico
Budget 2018. for the Affordable Care Act starting in 2018; elimina-
Israel: Historical data are based on Government tion of federal tax incentives (starting in 2018) that
Finance Statistics data prepared by the Central Bureau had neutralized the effects of Puerto Rico’s Act 154
of Statistics. The central government deficit is assumed on foreign companies; and the effects of the Tax Cuts
to remain at the current ceiling level of 2.9 percent of and Job Act, which reduces tax advantages for US
GDP throughout the projection period, rather than companies producing in Puerto Rico. Given sizable
declining in line with medium-term fiscal targets, policy uncertainty, some FEGP and IMF assumptions
consistent with long experience of revisions to those may differ, in particular those relating to the effects
targets. of the corporate tax reform, tax compliance, and tax
Italy: The IMF staff’s estimates and projections are adjustments (fees and rates); reduction of subsidies
based on the fiscal plans included in the government’s and expenses, freezing of payroll operational costs, and
2018 draft budget plan and September 2017 Update improvement of mobility; and increasing health care
to the Economic and Financial Document. efficiency. On the expenditure side, measures include
Japan: The projections include fiscal measures extension of Act 66, which freezes much government
already announced by the government, including the spending, through 2020; reduction of operating costs;
consumption tax hike in October 2019. decreases in government subsidies; and spending
Korea: The medium-term forecast incorporates the cuts in education. Although IMF policy assumptions
medium-term path for public spending announced by are similar to those in the FEGP scenario with full
the government. measures, the IMF’s projections of fiscal revenues,

Box A1 (continued)
expenditures, and balance are different from FEGP’s. measures and the IMF staff’s higher inflation forecast.
This stems from two main differences in methodolo- For the medium term, the fiscal projections assume a
gies: first and foremost, while IMF projections are on more gradual fiscal consolidation than envisaged in the
an accrual basis, FEGP’s are on a cash basis. Second, Medium Term Programme.
the IMF and FEGP make very different macroeco- United Kingdom: Fiscal projections are based on
nomic assumptions. the country’s November 2017 Budget and the March
Russia: Projections for 2018–20 are the IMF staff’s 2018 update, with expenditure projections based
estimates, based on the authorities’ budget. Projections on the budgeted nominal values and with revenue
for 2021–23 are based on the new oil price rule, with projections adjusted for differences between the IMF
adjustments by the IMF staff. staff’s forecasts of macroeconomic variables (such as
Saudi Arabia: Staff baseline projections of total GDP growth and inflation) and the forecasts of these
government revenues reflect the impact of announced variables assumed in the authorities’ fiscal projections.
policies in the 2018 Budget. Oil revenues are based on The IMF staff’s data exclude public sector banks and
WEO baseline oil prices and the assumption that Saudi the effect of transferring assets from the Royal Mail
Arabia continues to meet its commitments under the Pension Plan to the public sector in April 2012. Real
OPEC+ agreement. Expenditure projections take the government consumption and investment are part of
2018 budget as a starting point and reflect staff esti- the real GDP path, which, according to the IMF staff,
mates of the effects of the latest changes in policies and may or may not be the same as projected by the UK
economic developments. Expenditures in 2018 include Office for Budget Responsibility.
allowances and other measures announced in the Royal United States: Fiscal projections are based on the June
Decree for one year in January 2018. 2017 Congressional Budget Office baseline, adjusted
Singapore: For fiscal year 2018/19, projections are for the IMF staff’s policy and macroeconomic assump-
based on budget numbers. For the remainder of the tions. Projections incorporate the effects of tax reform
projection period, the IMF staff assumes unchanged (Tax Cuts and Jobs Act, signed into law end of 2017)
policies. as well as the Bipartisan Budget Act of 2018 passed in
South Africa: Fiscal projections are based on the February 2018. Finally, fiscal projections are adjusted to
2018 Budget. Nontax revenue excludes transactions in reflect the IMF staff’s forecasts for key macroeconomic
financial assets and liabilities, as they involve primar- and financial variables and different accounting treat-
ily revenues associated with realized exchange rate ment of financial sector support and defined-benefit
valuation gains from the holding of foreign currency pension plans, and are converted to a general govern-
deposits, sale of assets, and conceptually similar items. ment basis. Data are compiled using SNA 2008, and
Spain: For 2017, fiscal data are the IMF staff’s pro- when translated into government finance statistics, this
jections, reflecting the cash outturn through Novem- is in accordance with GFSM 2014. Because of data
ber. For 2018 and beyond, fiscal projections are based limitations, most series begin in 2001.
on the information specified in the government’s 2018
Budgetary Plan and on the IMF staff’s macroeconomic
projections. Monetary Policy Assumptions
Sweden: Fiscal projections take into account the Monetary policy assumptions are based on the
authorities’ projections based on the 2018 Spring Bud- established policy framework in each country. In most
get. The impact of cyclical developments on the fiscal cases, this implies a nonaccommodative stance over
accounts is calculated using the Organisation for Eco- the business cycle: official interest rates will increase
nomic Co-operation and Development’s 2005 elasticity when economic indicators suggest that inflation
to take into account output and employment gaps. will rise above its acceptable rate or range; they will
Switzerland: The projections assume that fiscal policy decrease when indicators suggest that inflation will
is adjusted as necessary to keep fiscal balances in line not exceed the acceptable rate or range, that out-
with the requirements of Switzerland’s fiscal rules. put growth is below its potential rate, and that the
Turkey: The fiscal projections for 2018 are based on margin of slack in the economy is significant. On this
the authorities’ Medium Term Programme 2018–20, basis, the London interbank offered rate (LIBOR) on
with adjustments for additionally announced fiscal six-month US dollar deposits is assumed to average

Box A1 (continued)
2.4 percent in 2018 and 3.4 percent in 2019 (see Japan: Monetary policy assumptions are in line with
Table 1.1). The rate on three-month euro deposits is market expectations.
assumed to average –0.3 percent in 2018 and 0.0 per- Korea: Monetary policy assumptions are in line with
cent in 2019. The interest rate on six-month Japanese market expectations.
yen deposits is assumed to average 0.0 percent in 2018 Mexico: Monetary policy assumptions are consistent
and 0.1 percent in 2019. with attaining the inflation target.
Australia: Monetary policy assumptions are in line Russia: Monetary projections assume that policy
with market expectations. rates will be falling over the next year or two as infla-
Brazil: Monetary policy assumptions are consistent tion fell below the 4 percent target of the Central
with gradual convergence of inflation toward the Bank of Russia in the context of a tight monetary
middle of the target range. stance, and the output gap is likely to be small.
Canada: Monetary policy assumptions are in line Saudi Arabia: Monetary policy projections are based
with market expectations. on the continuation of the exchange rate peg to the
China: Monetary policy is expected to tighten with US dollar.
a gradual rise in the interest rate. Singapore: Broad money is projected to grow in line
Denmark: The monetary policy is to maintain the with the projected growth in nominal GDP.
peg to the euro. South Africa: Monetary policy will remain neutral.
Euro area: Monetary policy assumptions for euro Sweden: Monetary projections are in line with
area member countries are in line with market Riksbank projections.
expectations. Switzerland: The projections assume no change in
Hong Kong Special Administrative Region: The IMF the policy rate in 2016–17.
staff assumes that the currency board system remains Turkey: The outlook for monetary and financial con-
intact. ditions assumes no changes to the current policy stance.
India: The policy (interest) rate assumption is con- United Kingdom: The short-term interest rate path is
sistent with an inflation rate within the Reserve Bank based on market interest rate expectations.
of India’s targeted band. United States: The IMF staff expects continued
Indonesia: Monetary policy assumptions are in line gradual normalization of the federal funds target rate
with the maintenance of inflation within the central over the medium term, in line with the broader mac-
bank’s targeted band. roeconomic outlook.

List of Tables
Output
A1. Summary of World Output
A2. Advanced Economies: Real GDP and Total Domestic Demand
A3. Advanced Economies: Components of Real GDP
A4. Emerging Market and Developing Economies: Real GDP
Inflation
A5. Summary of Inflation
A6. Advanced Economies: Consumer Prices
A7. Emerging Market and Developing Economies: Consumer Prices
Financial Policies
A8. Major Advanced Economies: General Government Fiscal Balances and Debt
Foreign Trade
A9. Summary of World Trade Volumes and Prices
Current Account Transactions

A10. Summary of Current Account Balances
A11. Advanced Economies: Balance on Current Account
A12. Emerging Market and Developing Economies: Balance on Current Account
Balance of Payments and External Financing

A13. Summary of Financial Account Balances
Flow of Funds
A14. Summary of Net Lending and Borrowing
Medium-Term Baseline Scenario

A15. Summary of World Medium-Term Baseline Scenario

Table A1. Summary of World Output1

(Annual percent change)
Average Projections
2000–09 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2023
World 3.9 5.4 4.3 3.5 3.5 3.6 3.5 3.2 3.8 3.9 3.9 3.7
United States 1.8 2.5 1.6 2.2 1.7 2.6 2.9 1.5 2.3 2.9 2.7 1.4
Euro Area 1.4 2.1 1.6 –0.9 –0.2 1.3 2.1 1.8 2.3 2.4 2.0 1.4
Japan 0.5 4.2 –0.1 1.5 2.0 0.4 1.4 0.9 1.7 1.2 0.9 0.5
Emerging Market and Developing Economies 6.1 7.4 6.4 5.4 5.1 4.7 4.3 4.4 4.8 4.9 5.1 5.0
Regional Groups
Commonwealth of Independent States3 5.9 4.6 5.3 3.7 2.5 1.0 –2.0 0.4 2.1 2.2 2.1 2.3
Latin America and the Caribbean 3.0 6.1 4.6 2.9 2.9 1.3 0.3 –0.6 1.3 2.0 2.8 2.8
Pakistan 5.2 4.7 4.4 5.0 2.6 2.8 2.5 4.9 2.6 3.4 3.7 3.6
Middle East and North Africa 5.2 4.9 4.4 5.1 2.5 2.6 2.4 4.9 2.2 3.2 3.6 3.4
Sub-Saharan Africa 5.7 7.0 5.1 4.4 5.3 5.1 3.4 1.4 2.8 3.4 3.7 4.0
Memorandum
European Union 1.7 2.0 1.8 –0.3 0.3 1.8 2.4 2.0 2.7 2.5 2.1 1.7
Analytical Groups
Fuel 5.8 5.1 5.2 5.0 2.7 2.2 0.3 1.8 1.2 2.0 2.4 2.4
Nonfuel 6.2 8.1 6.7 5.5 5.7 5.3 5.2 4.9 5.5 5.6 5.6 5.5
Of Which, Primary Products 3.7 6.6 4.7 2.2 4.0 2.0 3.0 1.4 3.0 3.0 3.4 3.7
Net Debtor Economies 4.9 6.9 5.3 4.3 4.7 4.4 4.2 3.8 4.7 5.0 5.2 5.4
Net Debtor Economies by
Debt-Servicing Experience
Economies with Arrears and/or
Rescheduling during 2012–16 4.8 4.3 2.4 1.8 2.9 1.6 0.4 2.4 3.2 4.2 4.7 5.0
Memorandum
Median Growth Rate
Output per Capita4
Advanced Economies 1.8 –4.0 2.5 1.1 0.7 0.9 1.6 1.7 1.1 1.9 2.0 1.1
World Growth Rate Based on Market Exchange
Rates 2.5 4.1 3.1 2.5 2.6 2.8 2.8 2.5 3.2 3.4 3.3 2.9
Value of World Output (billions of US dollars)
At Market Exchange Rates 46,643 65,960 73,165 74,535 76,596 78,663 74,429 75,485 79,865 87,505 92,734 114,353
At Purchasing Power Parities 66,645 89,346 94,925 99,726 104,749 110,342 115,252 120,367 127,044 134,981 143,283 178,018
1Real GDP.
2Excludes the United States, euro area countries, and Japan.
3Georgia, Turkmenistan, and Ukraine, which are not members of the Commonwealth of Independent States, are included in this group for reasons of geography and similarity in economic structure.
4Output per capita is in international currency at purchasing power parity.

Table A2. Advanced Economies: Real GDP and Total Domestic Demand1
Fourth Quarter2
Average Projections Projections
2000–09 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2023 2017:Q4 2018:Q4 2019:Q4
Real GDP
Advanced Economies 1.8 3.0 1.7 1.2 1.3 2.1 2.3 1.7 2.3 2.5 2.2 1.5 2.6 2.4 2.0
United States 1.8 2.5 1.6 2.2 1.7 2.6 2.9 1.5 2.3 2.9 2.7 1.4 2.6 3.0 2.3
Euro Area 1.4 2.1 1.6 –0.9 –0.2 1.3 2.1 1.8 2.3 2.4 2.0 1.4 2.7 2.2 2.0
Germany 0.8 3.9 3.7 0.7 0.6 1.9 1.5 1.9 2.5 2.5 2.0 1.2 2.9 2.5 1.9
France 1.4 2.0 2.1 0.2 0.6 0.9 1.1 1.2 1.8 2.1 2.0 1.6 2.5 1.8 2.0
Italy 0.5 1.7 0.6 –2.8 –1.7 0.1 1.0 0.9 1.5 1.5 1.1 0.8 1.6 1.3 1.1
Spain 2.7 0.0 –1.0 –2.9 –1.7 1.4 3.4 3.3 3.1 2.8 2.2 1.7 3.1 2.5 2.1
Netherlands 1.6 1.4 1.7 –1.1 –0.2 1.4 2.3 2.2 3.1 3.2 2.4 1.9 3.4 2.9 2.3
Belgium 1.7 2.7 1.8 0.2 0.2 1.4 1.4 1.5 1.7 1.9 1.7 1.5 1.9 2.2 1.3
Austria 1.7 1.8 2.9 0.7 0.0 0.8 1.1 1.5 2.9 2.6 1.9 1.5 3.6 2.1 2.0
Greece 2.7 –5.5 –9.1 –7.3 –3.2 0.7 –0.3 –0.2 1.4 2.0 1.8 1.9 1.9 2.7 1.0
Portugal 0.9 1.9 –1.8 –4.0 –1.1 0.9 1.8 1.6 2.7 2.4 1.8 1.2 2.4 2.0 2.4
Ireland 3.9 1.8 2.9 0.0 1.6 8.3 25.5 5.1 7.8 4.5 4.0 2.8 7.8 2.2 2.0
Finland 2.0 3.0 2.6 –1.4 –0.8 –0.6 0.1 2.1 3.0 2.6 2.0 1.2 2.8 2.7 1.6
Slovak Republic 4.5 5.0 2.8 1.7 1.5 2.8 3.9 3.3 3.4 4.0 4.2 3.4 3.5 3.9 4.2
Lithuania 4.6 1.6 6.0 3.8 3.5 3.5 2.0 2.3 3.8 3.2 3.0 2.5 3.7 2.7 3.3
Slovenia 2.9 1.2 0.6 –2.7 –1.1 3.0 2.3 3.1 5.0 4.0 3.2 2.1 6.2 2.4 3.6
Luxembourg 3.0 4.9 2.5 –0.4 3.7 5.8 2.9 3.1 3.5 4.3 3.7 3.0 5.2 1.4 5.1
Latvia 4.7 –3.9 6.4 4.0 2.4 1.9 3.0 2.2 4.5 4.0 3.5 3.0 4.7 4.9 2.6
Estonia 4.1 2.3 7.6 4.3 1.9 2.9 1.7 2.1 4.9 3.9 3.2 2.9 5.3 4.2 3.2
Cyprus 3.5 1.3 0.3 –3.1 –5.9 –1.4 2.0 3.0 3.9 3.6 3.0 2.4 3.9 3.0 3.0
Malta 1.6 3.5 1.3 2.7 4.7 8.1 9.9 5.5 6.6 5.7 4.6 3.2 4.5 6.5 4.1
Japan 0.5 4.2 –0.1 1.5 2.0 0.4 1.4 0.9 1.7 1.2 0.9 0.5 2.1 0.8 –0.1
United Kingdom 1.7 1.7 1.5 1.5 2.1 3.1 2.3 1.9 1.8 1.6 1.5 1.6 1.4 1.6 1.6
Korea 4.7 6.5 3.7 2.3 2.9 3.3 2.8 2.8 3.1 3.0 2.9 2.6 3.0 3.4 2.9
Canada 2.1 3.1 3.1 1.7 2.5 2.9 1.0 1.4 3.0 2.1 2.0 1.6 2.9 2.1 1.9
Australia 3.1 2.4 2.7 3.9 2.2 2.6 2.5 2.6 2.3 3.0 3.1 2.6 2.4 3.0 3.4
Taiwan Province of China 3.8 10.6 3.8 2.1 2.2 4.0 0.8 1.4 2.8 1.9 2.0 2.0 2.9 1.8 1.8
Switzerland 1.9 2.9 1.8 1.0 1.9 2.5 1.2 1.4 1.1 2.3 2.0 1.7 1.9 2.2 1.9
Sweden 2.0 6.0 2.7 –0.3 1.2 2.6 4.5 3.2 2.4 2.6 2.2 1.9 3.3 1.9 2.6
Singapore 5.2 15.2 6.4 4.1 5.1 3.9 2.2 2.4 3.6 2.9 2.7 2.6 3.6 1.6 3.6
Hong Kong SAR 4.2 6.8 4.8 1.7 3.1 2.8 2.4 2.1 3.8 3.6 3.2 3.0 3.4 3.9 2.7
Norway 1.8 0.7 1.0 2.7 1.0 2.0 2.0 1.1 1.8 2.1 2.1 1.9 1.4 2.7 1.7
Czech Republic 3.4 2.3 1.8 –0.8 –0.5 2.7 5.3 2.6 4.3 3.5 3.0 2.5 5.2 3.1 3.2
Israel 3.5 5.5 5.2 2.2 4.2 3.5 2.6 4.0 3.3 3.3 3.5 3.0 3.0 3.1 3.8
Denmark 1.0 1.9 1.3 0.2 0.9 1.6 1.6 2.0 2.1 2.0 1.9 1.7 1.2 2.7 1.9
New Zealand 2.9 2.0 1.9 2.5 2.2 3.2 4.2 4.2 3.0 2.9 2.9 2.4 3.2 2.5 3.3
Puerto Rico 1.0 –0.4 –0.4 0.0 –0.3 –1.2 –1.1 –2.6 –7.7 –3.6 –1.2 –0.8 ... ... ...
Macao SAR ... 25.3 21.7 9.2 11.2 –1.2 –21.6 –0.9 9.3 7.0 6.1 4.3 ... ... ...
Iceland 3.5 –3.6 2.0 1.3 4.3 2.2 4.3 7.5 3.6 3.2 3.0 2.6 1.2 4.0 3.9
San Marino ... –4.8 –9.3 –7.6 –3.2 –0.9 0.6 2.2 1.5 1.3 1.3 1.3 ... ... ...
Memorandum
Major Advanced Economies 1.4 2.8 1.6 1.4 1.4 2.0 2.1 1.4 2.1 2.4 2.1 1.3 2.4 2.3 1.8
Real Total Domestic Demand
United States 1.7 2.9 1.6 2.1 1.3 2.7 3.5 1.7 2.4 3.5 3.2 1.0 2.6 3.6 2.8
Euro Area 1.3 1.5 0.7 –2.4 –0.6 1.3 2.0 2.4 2.0 2.2 1.9 1.5 1.5 2.8 1.6
Germany 0.3 2.9 3.0 –0.8 1.0 1.3 1.5 2.4 2.4 2.3 2.0 1.5 2.0 2.7 1.8
France 1.7 2.1 2.0 –0.3 0.7 1.4 1.6 1.9 2.3 1.9 1.8 1.6 2.1 2.3 1.6
Italy 0.7 2.0 –0.6 –5.6 –2.6 0.2 1.5 1.1 1.4 1.6 1.1 0.7 1.2 1.7 0.7
Spain 2.9 –0.5 –3.1 –5.1 –3.2 2.0 4.0 2.6 2.9 2.6 2.1 1.5 3.3 2.4 2.0
Japan 0.2 2.4 0.7 2.3 2.4 0.4 1.0 0.4 1.2 1.1 1.0 0.5 1.8 0.8 –0.2
United Kingdom 1.8 2.5 –0.6 1.8 1.9 3.1 2.3 2.2 1.4 1.4 1.3 1.6 1.9 1.0 1.4
Canada 2.8 5.1 3.4 2.0 2.1 1.8 0.1 0.9 3.8 2.5 1.2 1.4 5.1 1.3 1.2
Other Advanced Economies3 2.9 6.1 3.1 2.0 1.5 2.7 2.5 2.2 3.4 2.9 2.8 2.6 4.3 2.9 2.7
Memorandum
1Inthis and other tables, when countries are not listed alphabetically, they are ordered on the basis of economic size.
2From the fourth quarter of the preceding year.

Table A3. Advanced Economies: Components of Real GDP

Averages Projections
2000–09 2010–19 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019
Private Consumer Expenditure
United States 2.4 2.5 1.9 2.3 1.5 1.5 2.9 3.6 2.7 2.8 2.9 2.6
Euro Area 1.4 0.9 0.8 –0.1 –1.1 –0.6 0.8 1.8 2.0 1.6 1.8 1.6
Germany 0.7 1.4 0.3 1.3 1.3 0.8 1.0 1.6 1.9 2.1 2.0 1.7
France 2.0 1.1 1.8 0.5 –0.2 0.5 0.8 1.4 2.2 1.3 1.4 1.7
Italy 0.6 0.2 1.2 0.0 –4.0 –2.4 0.2 1.9 1.4 1.3 1.3 1.1
Spain 2.5 0.5 0.3 –2.4 –3.5 –3.1 1.5 3.0 3.0 2.4 2.3 2.0
Japan 0.8 0.8 2.4 –0.4 2.0 2.4 –0.9 0.0 0.1 1.1 0.6 0.8
United Kingdom 2.2 1.5 0.6 –0.7 1.6 1.7 2.1 2.6 2.9 1.7 1.0 1.1
Canada 3.2 2.4 3.6 2.3 1.9 2.6 2.6 2.2 2.3 3.4 2.4 1.0
Memorandum
Major Advanced Economies 1.8 1.8 1.7 1.3 1.1 1.3 1.8 2.5 2.1 2.2 2.1 1.9
Public Consumption
Advanced Economies 2.3 0.9 1.0 –0.5 0.1 –0.3 0.6 1.5 1.6 0.9 2.1 1.6
United States 2.2 0.1 0.1 –2.7 –0.9 –2.4 –0.5 1.3 1.0 0.1 3.1 1.9
Euro Area 2.0 0.8 0.7 –0.1 –0.3 0.3 0.7 1.3 1.8 1.2 1.4 1.1
Germany 1.2 1.8 1.3 0.9 1.1 1.4 1.5 2.9 3.7 1.6 2.0 1.7
France 1.7 1.2 1.3 1.0 1.6 1.5 1.3 1.1 1.3 1.6 1.0 0.5
Italy 1.3 –0.3 0.6 –1.8 –1.4 –0.3 –0.7 –0.6 0.6 0.1 0.3 0.3
Spain 5.1 0.0 1.5 –0.3 –4.7 –2.1 –0.3 2.1 0.8 1.6 1.0 0.7
Japan 1.7 1.1 1.9 1.9 1.7 1.5 0.5 1.5 1.3 0.1 –0.3 1.1
United Kingdom 2.8 0.8 0.5 0.2 1.3 0.2 2.5 0.6 0.8 0.1 1.0 0.9
Canada 2.6 1.2 2.3 1.3 0.7 –0.7 0.5 1.6 2.2 2.2 1.2 0.6
Memorandum
Major Advanced Economies 2.0 0.6 0.7 –0.9 0.1 –0.7 0.3 1.4 1.4 0.5 2.0 1.4
Gross Fixed Capital Formation
United States 0.1 3.8 1.1 3.7 6.3 3.1 4.8 3.5 0.6 3.4 5.6 6.5
Euro Area 0.9 1.6 –0.3 1.5 –3.4 –2.5 1.9 3.3 4.6 3.5 4.4 3.5
Germany –0.4 2.9 5.0 7.4 –0.1 –1.2 3.8 1.1 2.9 4.0 3.5 2.9
France 1.6 1.9 2.1 2.1 0.2 –0.8 0.1 1.0 2.9 3.7 4.0 3.7
Italy 0.8 –0.7 –0.5 –1.9 –9.3 –6.6 –2.3 2.1 3.2 3.7 4.0 2.1
Spain 2.4 0.2 –4.9 –6.9 –8.6 –3.4 4.7 6.5 3.3 5.0 4.5 3.6
Japan –2.0 2.2 –1.6 1.7 3.5 4.9 3.1 1.7 1.1 2.5 3.4 1.6
United Kingdom –0.1 3.2 4.5 2.2 2.1 3.4 7.1 2.8 1.8 4.0 2.2 2.5
Canada 3.1 2.4 11.4 4.6 4.9 1.3 2.4 –5.2 –2.9 2.8 3.2 2.2
Memorandum

Table A3. Advanced Economies: Components of Real GDP (continued)

2000–09 2010–19 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019
Final Domestic Demand
United States 1.9 2.4 1.5 1.7 1.9 1.2 2.7 3.3 2.1 2.5 3.4 3.2
Euro Area 1.4 1.0 0.5 0.3 –1.5 –0.8 1.0 2.0 2.5 1.9 2.3 1.9
Germany 0.5 1.8 1.4 2.5 1.0 0.5 1.7 1.8 2.5 2.4 2.3 1.9
France 1.8 1.3 1.8 0.9 0.3 0.4 0.7 1.2 2.2 1.9 1.9 1.8
Italy 0.8 –0.1 0.7 –0.8 –4.5 –2.8 –0.4 1.4 1.5 1.5 1.6 1.1
Spain 2.9 0.3 –0.7 –3.0 –4.8 –3.0 1.8 3.5 2.6 2.8 2.5 2.1
Japan 0.3 1.2 1.4 0.5 2.3 2.8 0.2 0.7 0.6 1.2 1.3 1.0
United Kingdom 2.0 1.6 1.2 –0.1 1.6 1.7 3.0 2.2 2.4 1.8 1.2 1.3
Canada 3.0 2.2 5.0 2.6 2.4 1.6 2.1 0.3 1.1 3.0 2.3 1.2
Memorandum
Stock Building2
Advanced Economies –0.1 0.1 1.3 0.1 –0.2 0.0 0.1 0.1 –0.3 0.0 0.0 0.0
United States –0.2 0.1 1.5 –0.1 0.1 0.2 –0.1 0.2 –0.4 –0.1 0.0 0.0
Euro Area –0.1 0.1 0.9 0.5 –0.9 0.2 0.3 0.0 –0.1 0.1 0.0 0.0
Germany –0.2 0.0 1.4 0.5 –1.6 0.5 –0.4 –0.3 –0.1 0.0 0.0 0.0
France –0.1 0.2 0.3 1.1 –0.6 0.2 0.7 0.3 –0.1 0.4 0.0 0.0
Italy –0.1 0.1 1.3 0.2 –1.1 0.2 0.6 0.1 –0.4 –0.1 0.0 0.0
Spain 0.0 0.0 0.2 –0.1 –0.2 –0.3 0.2 0.4 0.0 0.1 0.1 0.0
Japan 0.0 0.1 1.0 0.2 0.0 –0.4 0.1 0.3 –0.2 –0.1 0.0 0.0
United Kingdom –0.2 0.1 1.4 –0.6 0.2 0.1 0.2 0.1 –0.2 –0.4 0.1 0.0
Canada 0.0 0.0 0.1 0.7 –0.3 0.5 –0.4 –0.2 –0.2 –0.4 0.0 0.0
Other Advanced Economies1 –0.1 0.1 1.9 0.2 –0.3 –0.8 0.2 0.0 –0.3 0.3 –0.1 –0.1
Memorandum
Major Advanced Economies –0.1 0.1 1.2 0.1 –0.2 0.2 0.0 0.2 –0.3 –0.1 0.0 0.0
Foreign Balance2
Advanced Economies 0.1 0.0 0.1 0.3 0.4 0.3 0.0 –0.2 –0.2 0.0 –0.1 –0.2
United States 0.0 –0.3 –0.5 0.0 0.1 0.3 –0.2 –0.7 –0.2 –0.2 –0.6 –0.6
Euro Area 0.1 0.4 0.6 0.9 1.5 0.4 0.1 0.1 –0.4 0.4 0.3 0.1
Germany 0.5 0.4 1.1 0.9 1.4 –0.3 0.7 0.1 –0.4 0.2 0.4 0.2
France –0.3 –0.2 –0.1 0.0 0.5 –0.1 –0.5 –0.5 –0.8 –0.4 0.1 0.0
Italy –0.2 0.4 –0.3 1.2 2.8 0.8 –0.1 –0.5 –0.3 0.1 0.0 0.0
Spain –0.2 0.7 0.5 2.1 2.2 1.5 –0.5 –0.4 0.7 0.3 0.3 0.2
Japan 0.1 0.1 1.6 –0.9 –0.8 –0.4 0.0 0.3 0.5 0.5 0.1 –0.1
United Kingdom –0.1 –0.1 –0.7 1.4 –0.7 –0.6 –0.5 –0.1 –0.8 0.6 0.3 0.2
Canada –0.8 0.0 –2.1 –0.3 –0.4 0.3 1.1 0.9 0.7 –0.9 –0.3 0.8
Other Advanced Economies1 0.6 0.1 0.0 0.5 0.5 0.9 0.4 –0.2 –0.1 –0.6 0.0 0.1
Memorandum
Major Advanced Economies 0.0 –0.1 –0.1 0.1 0.2 0.1 0.0 –0.3 –0.2 0.0 –0.2 –0.3
1Excludesthe G7 (Canada, France, Germany, Italy, Japan, United Kingdom, United States) and euro area countries.
2Changes expressed as percent of GDP in the preceding period.

Table A4. Emerging Market and Developing Economies: Real GDP

Average Projections
2000–09 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2023
Commonwealth of Independent States1,2 5.9 4.6 5.3 3.7 2.5 1.0 –2.0 0.4 2.1 2.2 2.1 2.3
Russia 5.4 4.5 5.1 3.7 1.8 0.7 –2.5 –0.2 1.5 1.7 1.5 1.5
Excluding Russia 7.5 5.0 6.0 3.6 4.2 1.9 –0.6 1.9 3.6 3.5 3.6 3.9
Armenia 8.5 2.2 4.7 7.1 3.3 3.6 3.3 0.3 7.5 3.4 3.5 4.0
Azerbaijan 14.6 4.6 –1.6 2.1 5.9 2.7 0.6 –3.1 0.1 2.0 3.9 2.6
Belarus 7.2 7.8 5.5 1.7 1.0 1.7 –3.8 –2.5 2.4 2.8 2.4 2.0
Georgia 5.9 6.2 7.2 6.4 3.4 4.6 2.9 2.8 4.8 4.5 4.8 5.2
Kazakhstan 8.5 7.3 7.5 5.0 6.0 4.3 1.2 1.1 4.0 3.2 2.8 3.4
Kyrgyz Republic 4.6 –0.5 6.0 –0.1 10.9 4.0 3.9 3.8 4.5 3.3 4.9 3.3
Moldova 4.6 7.1 6.8 –0.7 9.4 4.8 –0.4 4.3 4.0 3.5 3.8 3.9
Tajikistan 8.2 6.5 7.4 7.5 7.4 6.7 6.0 6.9 7.1 4.0 4.0 4.0
Turkmenistan 14.2 9.2 14.7 11.1 10.2 10.3 6.5 6.2 6.5 6.2 5.6 5.7
Ukraine3 4.5 0.3 5.5 0.2 0.0 –6.6 –9.8 2.4 2.5 3.2 3.3 4.0
Uzbekistan 6.4 8.5 8.3 8.2 8.0 8.0 7.9 7.8 5.3 5.0 5.0 6.0
Bangladesh 5.8 6.0 6.5 6.3 6.0 6.3 6.8 7.2 7.1 7.0 7.0 7.0
Bhutan 8.2 9.3 9.7 6.4 3.6 4.0 6.1 6.3 6.0 7.1 7.6 11.4
Brunei Darussalam 1.4 2.7 3.7 0.9 –2.1 –2.5 –0.4 –2.5 0.5 1.0 8.0 9.1
Cambodia 8.3 6.0 7.2 7.3 7.4 7.1 7.2 7.0 6.9 6.9 6.8 6.0
China 10.3 10.6 9.5 7.9 7.8 7.3 6.9 6.7 6.9 6.6 6.4 5.5
Fiji 0.9 3.0 2.7 1.4 4.7 5.6 3.8 0.4 3.8 3.5 3.4 3.2
India4 6.9 10.3 6.6 5.5 6.4 7.4 8.2 7.1 6.7 7.4 7.8 8.2
Indonesia 5.3 6.4 6.2 6.0 5.6 5.0 4.9 5.0 5.1 5.3 5.5 5.6
Kiribati 1.4 –0.9 1.6 4.6 4.3 –0.6 10.3 1.1 3.1 2.3 2.4 1.8
Lao P.D.R. 7.0 8.0 8.0 7.8 8.0 7.6 7.3 7.0 6.8 6.8 7.0 6.8
Malaysia 4.7 7.5 5.3 5.5 4.7 6.0 5.0 4.2 5.9 5.3 5.0 4.9
Maldives 6.3 7.1 8.4 2.3 7.1 7.6 2.2 4.5 4.8 5.0 5.0 5.4
Marshall Islands 1.9 1.2 3.5 2.9 –0.8 –0.4 1.9 2.0 1.9 1.8 1.7 1.5
Micronesia 0.5 3.3 1.0 –1.7 –3.0 –2.5 3.9 2.9 2.0 1.4 0.9 0.6
Mongolia 5.6 7.3 17.3 12.3 11.6 7.9 2.4 1.2 5.1 5.0 6.3 7.0
Myanmar 11.1 5.3 5.6 7.3 8.4 8.0 7.0 5.9 6.7 6.9 7.0 7.5
Nauru ... 13.6 11.7 10.1 34.2 36.5 2.8 10.4 4.0 –3.0 0.0 2.0
Nepal 4.1 4.8 3.4 4.8 4.1 6.0 3.3 0.4 7.5 5.0 4.0 3.8
Palau ... 3.0 5.1 3.9 –2.1 5.4 11.4 0.5 –1.0 1.0 4.0 2.0
Papua New Guinea 2.8 10.1 1.1 4.6 3.8 12.5 8.0 2.4 2.5 2.9 2.6 3.3
Philippines 4.4 7.6 3.7 6.7 7.1 6.1 6.1 6.9 6.7 6.7 6.8 7.0
Samoa 3.2 –2.0 5.6 0.4 –1.9 1.2 1.6 7.1 2.4 2.5 2.8 2.2
Solomon Islands 1.2 6.8 13.2 4.6 3.0 2.3 2.5 3.5 3.2 3.0 2.9 2.9
Sri Lanka 5.1 8.0 8.4 9.1 3.4 5.0 5.0 4.5 3.1 4.0 4.5 5.0
Thailand 4.3 7.5 0.8 7.2 2.7 1.0 3.0 3.3 3.9 3.9 3.8 3.5
Timor-Leste5 ... 8.5 7.7 5.5 2.5 4.1 4.0 5.3 –0.5 2.8 5.7 5.2
Tonga 1.1 3.2 1.8 –1.1 –0.6 2.9 3.5 3.1 3.1 3.2 2.9 1.4
Tuvalu ... –3.1 7.9 –3.8 4.6 1.3 9.1 3.0 3.2 3.5 3.1 2.0
Vanuatu 3.3 1.6 1.2 1.8 2.0 2.3 0.2 3.5 4.2 3.8 3.5 3.0
Vietnam 6.9 6.4 6.2 5.2 5.4 6.0 6.7 6.2 6.8 6.6 6.5 6.5
Albania 5.9 3.7 2.5 1.4 1.0 1.8 2.2 3.4 3.9 3.7 3.8 4.0
Bosnia and Herzegovina 4.2 0.8 0.9 –0.7 2.4 1.1 3.1 3.2 2.7 3.2 3.5 4.0
Bulgaria 5.0 1.3 1.9 0.0 0.9 1.3 3.6 3.9 3.6 3.8 3.1 2.8
Croatia 3.0 –1.4 –0.3 –2.2 –0.6 –0.1 2.3 3.2 2.8 2.8 2.6 2.2
Hungary 2.4 0.7 1.7 –1.6 2.1 4.2 3.4 2.2 4.0 3.8 3.0 2.2
Kosovo ... 3.3 4.4 2.8 3.4 1.2 4.1 4.1 4.1 4.0 4.0 4.0
FYR Macedonia 3.1 3.4 2.3 –0.5 2.9 3.6 3.9 2.9 0.0 2.8 3.0 3.5
Montenegro ... 2.7 3.2 –2.7 3.5 1.8 3.4 2.9 4.2 3.1 2.4 3.0
Poland 3.9 3.6 5.0 1.6 1.4 3.3 3.8 2.9 4.6 4.1 3.5 2.8
Romania 4.8 –2.8 2.0 1.2 3.5 3.1 4.0 4.8 7.0 5.1 3.5 3.1
Serbia 5.1 0.6 1.4 –1.0 2.6 –1.8 0.8 2.8 1.8 3.5 3.5 4.0
Turkey 3.8 8.5 11.1 4.8 8.5 5.2 6.1 3.2 7.0 4.4 4.0 3.6

Table A4. Emerging Market and Developing Economies: Real GDP (continued)
Average Projections
2000–09 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2023
Latin America and the Caribbean 3.0 6.1 4.6 2.9 2.9 1.3 0.3 –0.6 1.3 2.0 2.8 2.8
Antigua and Barbuda 2.8 –7.2 –2.1 3.5 –0.1 5.1 4.1 5.3 2.8 3.5 3.0 2.0
Argentina 2.3 10.1 6.0 –1.0 2.4 –2.5 2.7 –1.8 2.9 2.0 3.2 3.3
The Bahamas 1.0 1.5 0.6 3.1 –0.6 –1.2 –3.1 0.2 1.3 2.5 2.2 1.5
Barbados 1.4 0.3 0.8 0.3 –0.6 0.1 1.0 1.6 0.9 0.5 0.8 1.0
Belize 4.9 3.3 2.1 3.7 0.7 4.0 3.8 –0.5 0.8 1.8 2.0 1.7
Bolivia 3.7 4.1 5.2 5.1 6.8 5.5 4.9 4.3 4.2 4.0 3.8 3.7
Brazil 3.4 7.5 4.0 1.9 3.0 0.5 –3.5 –3.5 1.0 2.3 2.5 2.2
Chile 4.2 5.8 6.1 5.3 4.1 1.8 2.3 1.3 1.5 3.4 3.3 3.0
Colombia 4.0 4.0 6.6 4.0 4.9 4.4 3.1 2.0 1.8 2.7 3.3 3.5
Costa Rica 4.2 5.0 4.3 4.8 2.3 3.5 3.6 4.5 3.2 3.6 3.6 3.5
Dominica 2.6 0.7 –0.2 –1.1 0.8 4.2 –3.7 2.6 –4.2 –16.3 12.2 1.5
Dominican Republic 4.2 8.3 3.1 2.8 4.7 7.6 7.0 6.6 4.6 5.5 5.0 5.0
Ecuador 3.9 3.5 7.9 5.6 4.9 3.8 0.1 –1.6 2.7 2.5 2.2 1.8
El Salvador 2.0 1.4 2.2 1.9 1.8 1.4 2.3 2.4 2.4 2.3 2.3 2.2
Grenada 2.3 –0.5 0.8 –1.2 2.4 7.3 6.4 3.7 3.5 3.6 3.6 2.7
Guatemala 3.3 2.9 4.2 3.0 3.7 4.2 4.1 3.1 2.8 3.2 3.6 3.6
Guyana 1.8 4.4 5.4 4.8 5.2 3.8 3.1 3.3 2.1 3.5 3.7 27.8
Haiti 0.8 –5.5 5.5 2.9 4.2 2.8 1.2 1.5 1.2 2.0 3.0 2.0
Honduras 4.5 3.7 3.8 4.1 2.8 3.1 3.8 3.8 4.8 3.5 3.6 3.8
Jamaica 0.9 –1.4 1.4 –0.5 0.2 0.6 0.8 1.5 1.0 1.5 1.8 2.4
Mexico 1.4 5.1 3.7 3.6 1.4 2.8 3.3 2.9 2.0 2.3 3.0 2.9
Nicaragua 2.9 4.4 6.3 6.5 4.9 4.8 4.9 4.7 4.9 4.7 4.5 4.5
Panama 5.5 5.8 11.8 9.2 6.6 6.0 5.8 5.0 5.4 5.6 5.8 5.5
Paraguay 1.9 13.1 4.3 –1.2 14.0 4.7 3.0 4.0 4.3 4.5 4.1 4.0
Peru 5.0 8.5 6.5 6.0 5.8 2.4 3.3 4.1 2.5 3.7 4.0 3.8
St. Kitts and Nevis 3.2 –2.9 –0.8 –0.8 6.6 5.1 4.9 3.1 2.6 3.5 3.2 2.7
St. Lucia 2.2 –1.6 0.6 0.2 0.3 3.6 –0.9 3.4 3.0 2.5 2.3 1.6
St. Vincent and the Grenadines 3.1 –2.3 0.2 1.3 2.5 0.3 0.9 0.8 1.0 2.1 2.5 2.8
Suriname 4.5 5.2 5.8 2.7 2.9 0.3 –2.6 –5.1 0.0 1.4 2.0 3.0
Trinidad and Tobago 6.3 3.3 –0.3 1.3 1.0 –0.3 1.5 –6.0 –2.6 0.2 0.2 1.9
Uruguay 2.2 7.8 5.2 3.5 4.6 3.2 0.4 1.5 3.1 3.4 3.1 3.0
Venezuela 3.7 –1.5 4.2 5.6 1.3 –3.9 –6.2 –16.5 –14.0 –15.0 –6.0 –1.5
and Pakistan 5.2 4.7 4.4 5.0 2.6 2.8 2.5 4.9 2.6 3.4 3.7 3.6
Afghanistan ... 8.4 6.5 14.0 5.7 2.7 1.3 2.4 2.5 2.5 3.0 5.0
Algeria 3.9 3.6 2.8 3.4 2.8 3.8 3.7 3.3 2.0 3.0 2.7 0.5
Bahrain 5.6 4.3 2.0 3.7 5.4 4.4 2.9 3.2 3.2 3.0 2.3 2.3
Djibouti 3.2 4.1 7.3 4.8 5.0 6.0 6.5 6.5 6.7 6.7 6.7 6.0
Egypt 5.0 5.1 1.8 2.2 3.3 2.9 4.4 4.3 4.2 5.2 5.5 6.0
Iran 4.8 5.7 3.1 –7.7 –0.3 3.2 –1.6 12.5 4.3 4.0 4.0 4.1
Iraq 10.9 6.4 7.5 13.9 7.6 0.7 4.8 11.0 –0.8 3.1 4.9 2.4
Jordan 6.5 2.3 2.6 2.7 2.8 3.1 2.4 2.0 2.3 2.5 2.7 3.0
Kuwait 5.3 –2.4 10.9 7.9 0.4 0.6 –1.0 2.2 –2.5 1.3 3.8 2.9
Lebanon 4.9 8.0 0.9 2.8 2.6 2.0 0.8 1.0 1.2 1.5 1.8 2.8
Libya4 4.2 3.2 –66.7 124.7 –36.8 –53.0 –13.0 –7.4 70.8 16.4 1.4 1.5
Mauritania 4.3 4.8 4.7 5.8 6.1 5.6 0.8 1.6 3.2 2.7 4.5 5.3
Morocco 4.8 3.8 5.2 3.0 4.5 2.7 4.5 1.2 4.2 3.1 4.0 4.6
Oman 3.5 4.8 –1.1 9.3 4.4 2.8 4.7 1.8 –0.3 2.1 4.2 2.3
Pakistan 4.7 2.6 3.6 3.8 3.7 4.1 4.1 4.5 5.3 5.6 4.7 5.0
Qatar 12.1 18.1 13.4 4.7 4.4 4.0 3.6 2.2 2.1 2.6 2.7 2.7
Saudi Arabia 3.4 4.8 10.3 5.4 2.7 3.7 4.1 1.7 –0.7 1.7 1.9 2.3
Somalia ... ... ... 1.2 1.9 2.4 2.5 2.4 1.8 2.5 2.8 3.5
Sudan6 5.5 5.2 –3.7 –10.6 2.2 3.2 3.0 3.5 3.2 3.7 3.5 3.0
Syria7 4.4 3.4 ... ... ... ... ... ... ... ... ... ...
Tunisia 4.6 2.6 –1.9 3.9 2.4 2.3 1.1 1.0 1.9 2.4 2.9 4.2
United Arab Emirates 4.9 1.6 6.4 5.1 5.8 3.3 3.8 3.0 0.5 2.0 3.0 3.1
Yemen 4.1 7.7 –12.7 2.4 4.8 –0.2 –37.1 –34.3 –13.8 –0.5 17.9 6.3

Table A4. Emerging Market and Developing Economies: Real GDP (continued)
Average Projections
2000–09 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2023
Angola 11.3 3.5 3.9 5.2 6.8 4.7 3.0 –0.8 0.7 2.2 2.4 4.9
Benin 4.2 2.1 3.0 4.8 7.2 6.4 2.1 4.0 5.6 6.0 6.3 6.1
Botswana 3.4 8.6 6.0 4.5 11.3 4.1 –1.7 4.3 2.2 4.6 4.5 4.2
Burkina Faso 5.3 8.4 6.6 6.5 5.8 4.3 3.9 5.9 6.4 6.0 6.0 5.3
Burundi 3.4 5.1 4.0 4.4 5.9 4.5 –4.0 –1.0 0.0 0.1 0.4 0.5
Cabo Verde 6.0 1.5 4.0 1.1 0.8 0.6 1.0 3.8 4.0 4.3 4.0 4.0
Cameroon 3.9 3.4 4.1 4.5 5.4 5.9 5.7 4.5 3.2 4.0 4.5 5.5
Central African Republic 1.0 3.0 3.3 4.1 –36.7 1.0 4.8 4.5 4.0 4.0 4.0 4.0
Chad 8.3 13.6 0.1 8.8 5.8 6.9 1.8 –6.4 –3.1 3.5 2.8 4.2
Comoros 2.0 2.1 2.2 3.0 3.5 2.0 1.0 2.2 2.5 3.0 3.0 3.0
Democratic Republic of the Congo 3.1 7.1 6.9 7.1 8.5 9.5 6.9 2.4 3.4 3.8 4.0 4.7
Republic of Congo 4.6 8.7 3.4 3.8 3.3 6.8 2.6 –2.8 –4.6 0.7 4.6 0.2
Côte d’Ivoire 0.7 2.0 –4.2 10.1 9.3 8.8 8.8 8.3 7.8 7.4 7.1 6.4
Equatorial Guinea 25.3 –8.9 6.5 8.3 –4.1 –0.7 –9.1 –9.7 –4.4 –8.5 –2.8 1.0
Eritrea –0.9 2.2 8.7 7.0 4.6 2.9 2.6 1.9 5.0 4.2 3.8 4.3
Ethiopia 8.4 10.6 11.4 8.7 9.9 10.3 10.4 8.0 10.9 8.5 8.3 8.0
Gabon 0.6 6.3 7.1 5.3 5.5 4.4 3.9 2.1 0.8 2.7 3.7 4.7
The Gambia 3.7 6.5 –4.3 5.6 4.8 0.9 4.3 2.2 3.5 5.4 5.2 4.8
Ghana 5.4 7.9 14.0 9.3 7.3 4.0 3.8 3.7 8.4 6.3 7.6 5.1
Guinea 2.9 4.2 5.6 5.9 3.9 3.7 3.5 6.6 6.7 5.8 5.9 5.0
Guinea-Bissau 2.7 4.6 8.1 –1.7 3.3 1.0 6.1 5.8 5.5 5.5 5.5 5.0
Kenya 3.4 8.4 6.1 4.6 5.9 5.4 5.7 5.8 4.8 5.5 6.0 6.0
Lesotho 3.7 6.3 6.7 4.9 2.2 3.0 2.5 3.1 3.1 1.7 2.6 1.8
Liberia ... 6.4 7.7 8.4 8.8 0.7 0.0 –1.6 2.5 3.2 4.7 5.3
Madagascar 3.0 0.3 1.5 3.0 2.3 3.3 3.1 4.2 4.1 5.1 5.6 5.1
Malawi 4.2 6.9 4.9 1.9 5.2 5.7 2.9 2.3 4.0 3.5 4.5 6.5
Mali 5.2 5.4 3.2 –0.8 2.3 7.0 6.0 5.8 5.3 5.0 4.7 4.7
Mauritius 4.4 4.1 3.9 3.2 3.2 3.6 3.5 3.9 3.9 3.9 4.0 4.0
Mozambique 7.6 6.7 7.1 7.2 7.1 7.4 6.6 3.8 2.9 3.0 2.5 9.9
Namibia 3.8 6.0 5.1 5.1 5.6 6.4 6.0 1.1 –1.2 1.2 3.3 3.5
Niger 4.3 8.4 2.2 11.8 5.3 7.5 4.0 5.0 5.2 5.1 5.4 5.6
Nigeria 8.3 11.3 4.9 4.3 5.4 6.3 2.7 –1.6 0.8 2.1 1.9 2.0
Rwanda 8.3 7.3 7.8 8.8 4.7 7.6 8.9 6.0 6.1 7.2 7.8 7.5
São Tomé and Príncipe 3.9 4.5 4.8 4.5 4.3 4.1 4.0 4.1 4.0 5.0 5.5 5.5
Senegal 4.0 4.3 1.9 4.5 3.6 4.1 6.5 6.7 7.2 7.0 7.0 6.5
Seychelles 1.9 5.9 5.4 3.7 6.0 4.5 5.0 4.5 4.2 3.3 3.3 3.3
Sierra Leone 8.7 5.3 6.3 15.2 20.7 4.6 –20.5 6.3 3.5 3.5 5.6 7.3
South Africa 3.6 3.0 3.3 2.2 2.5 1.8 1.3 0.6 1.3 1.5 1.7 1.8
South Sudan ... ... ... –52.4 29.3 2.9 –0.2 –13.8 –11.1 –3.8 –2.6 –0.7
Swaziland 3.8 3.5 2.0 3.5 4.8 3.6 1.1 0.0 0.2 –0.9 0.2 2.2
Tanzania 6.2 6.4 7.9 5.1 7.3 7.0 7.0 7.0 6.0 6.4 6.6 6.5
Togo 1.5 6.1 6.4 6.5 6.1 5.9 5.7 5.1 4.4 4.9 5.2 5.8
Uganda 7.5 7.7 6.8 2.2 4.7 4.6 5.7 2.3 4.5 5.2 5.8 7.3
Zambia 6.8 10.3 5.6 7.6 5.0 4.7 2.9 3.7 3.6 4.0 4.5 4.5
Zimbabwe8 –6.1 15.4 16.3 13.6 5.3 2.8 1.4 0.7 3.0 2.4 4.2 5.0
1Data for some countries refer to real net material product (NMP) or are estimates based on NMP. The figures should be interpreted only as indicative of broad orders of magnitude because
reliable, comparable data are not generally available. In particular, the growth of output of new private enterprises of the informal economy is not fully reflected in the recent figures.
2Georgia, Turkmenistan, and Ukraine, which are not members of the Commonwealth of Independent States, are included in this group for reasons of geography and similarity in economic
structure.
3Data are based on the 2008 System of National Accounts. The revised national accounts data are available beginning in 2000 and exclude Crimea and Sevastopol from 2010 onward.
4See country-specific notes for India and Libya in the “Country Notes” section of the Statistical Appendix.
5In this table only, the data for Timor-Leste are based on non-oil GDP.
6Data for 2011 exclude South Sudan after July 9. Data for 2012 and onward pertain to the current Sudan.
7Data for Syria are excluded for 2011 onward owing to the uncertain political situation.
8The Zimbabwe dollar ceased circulating in early 2009. Data are based on IMF staff estimates of price and exchange rate developments in US dollars. IMF staff estimates of US dollar values
may differ from authorities’ estimates. Real GDP is in constant 2009 prices.

Table A5. Summary of Inflation

(Percent)
Average Projections
2000–09 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2023
GDP Deflators
United States 2.2 1.2 2.1 1.8 1.6 1.8 1.1 1.3 1.8 2.3 2.2 1.7
Euro Area 2.0 0.7 1.0 1.3 1.2 0.9 1.4 0.8 1.1 1.3 1.6 2.0
Japan –1.1 –1.9 –1.7 –0.8 –0.3 1.7 2.1 0.3 –0.2 0.5 1.1 0.8
Consumer Prices
United States 2.6 1.6 3.1 2.1 1.5 1.6 0.1 1.3 2.1 2.5 2.4 2.1
Euro Area2 2.1 1.6 2.7 2.5 1.3 0.4 0.0 0.2 1.5 1.5 1.6 2.1
Japan –0.3 –0.7 –0.3 –0.1 0.3 2.8 0.8 –0.1 0.5 1.1 1.1 1.3
Emerging Market and Developing Economies3 6.8 5.6 7.1 5.8 5.5 4.7 4.7 4.3 4.0 4.6 4.3 4.1
Regional Groups
Commonwealth of Independent States4 13.7 7.2 9.8 6.2 6.5 8.1 15.5 8.3 5.5 4.6 4.8 4.2
Latin America and the Caribbean 6.2 4.2 5.2 4.6 4.6 4.9 5.5 5.6 4.1 3.6 3.5 3.4
Pakistan 6.7 6.6 9.3 9.8 9.2 6.7 5.6 4.7 6.3 8.2 6.8 5.7
Middle East and North Africa 6.5 6.2 8.7 9.7 9.4 6.5 5.8 4.9 6.6 8.7 7.1 5.8
Memorandum
European Union 2.5 2.0 3.1 2.6 1.5 0.5 0.0 0.2 1.7 1.9 1.8 2.2
Analytical Groups
Fuel 10.2 6.7 8.6 8.0 8.1 6.4 8.8 6.9 5.3 6.1 5.9 5.5
Nonfuel 5.9 5.3 6.7 5.3 4.9 4.2 3.8 3.7 3.7 4.3 4.0 3.8
Of Which, Primary Products5 ... ... ... ... ... ... ... ... ... ... ... ...
Net Debtor Economies 7.7 6.7 7.7 7.0 6.3 5.7 5.5 5.1 5.4 5.6 5.2 4.6
Rescheduling during 2012–16 8.8 9.7 10.0 7.8 6.7 10.3 15.4 8.7 16.8 15.7 11.5 6.7
Memorandum
Median Inflation Rate
Emerging Market and Developing Economies3 5.1 4.1 5.4 4.5 3.8 3.2 2.7 2.7 3.3 3.6 3.4 3.0
1Excludes the United States, euro area countries, and Japan.
2Based on Eurostat’s harmonized index of consumer prices.
3Excludes Argentina and Venezuela. See country-specific notes for Argentina and Venezuela in the “Country Notes” section of the Statistical Appendix.
5Data are missing because of Argentina, which accounts for more than 30 percent of the weights of the group. See country-specific notes for Argentina in the “Country Notes” section of the Statistical
Appendix.

Table A6. Advanced Economies: Consumer Prices1

End of Period2
2000–09 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2023 2017 2018 2019
United States 2.6 1.6 3.1 2.1 1.5 1.6 0.1 1.3 2.1 2.5 2.4 2.1 2.1 2.6 2.3
Euro Area3 2.1 1.6 2.7 2.5 1.3 0.4 0.0 0.2 1.5 1.5 1.6 2.1 1.4 1.6 1.8
Germany 1.6 1.2 2.5 2.1 1.6 0.8 0.1 0.4 1.7 1.6 1.7 2.7 1.6 1.6 1.9
France 1.9 1.7 2.3 2.2 1.0 0.6 0.1 0.3 1.2 1.5 1.6 1.9 1.2 1.4 2.0
Italy 2.3 1.6 2.9 3.3 1.2 0.2 0.1 –0.1 1.3 1.1 1.3 1.6 0.7 1.6 1.3
Spain 3.0 1.8 3.2 2.4 1.4 –0.1 –0.5 –0.2 2.0 1.7 1.6 1.8 1.1 1.6 1.6
Netherlands 2.3 0.9 2.5 2.8 2.6 0.3 0.2 0.1 1.3 2.0 2.2 2.4 1.7 2.1 2.2
Belgium 2.1 2.3 3.4 2.6 1.2 0.5 0.6 1.8 2.2 1.6 1.8 2.0 1.6 1.5 2.0
Austria 1.9 1.7 3.5 2.6 2.1 1.5 0.8 1.0 2.2 2.2 2.2 2.1 2.3 2.2 2.2
Greece 3.2 4.7 3.1 1.0 –0.9 –1.4 –1.1 0.0 1.1 0.7 1.1 1.7 1.0 0.9 1.2
Portugal 2.6 1.4 3.6 2.8 0.4 –0.2 0.5 0.6 1.6 1.6 1.6 2.1 1.6 1.8 2.5
Ireland 2.9 –1.6 1.2 1.9 0.6 0.3 0.0 –0.2 0.3 0.9 1.3 1.9 –0.1 1.8 1.4
Finland 1.8 1.7 3.3 3.2 2.2 1.2 –0.2 0.4 0.8 1.2 1.7 2.0 0.5 1.7 1.7
Slovak Republic 5.2 0.7 4.1 3.7 1.5 –0.1 –0.3 –0.5 1.3 1.9 1.9 2.0 2.0 1.9 2.0
Lithuania 3.0 1.2 4.1 3.2 1.2 0.2 –0.7 0.7 3.7 2.2 2.2 2.5 3.8 2.2 2.2
Slovenia 4.9 1.8 1.8 2.6 1.8 0.2 –0.5 –0.1 1.4 1.7 2.0 2.0 1.7 1.7 2.0
Luxembourg 2.7 2.8 3.7 2.9 1.7 0.7 0.1 0.0 2.1 1.4 1.8 2.0 0.9 1.4 3.1
Latvia 5.8 –1.2 4.2 2.3 0.0 0.7 0.2 0.1 2.9 3.0 2.5 2.3 2.2 3.0 2.5
Estonia 4.3 2.7 5.1 4.2 3.2 0.5 0.1 0.9 3.7 3.0 2.5 2.1 3.8 1.8 2.5
Cyprus 2.6 2.6 3.5 3.1 0.4 –0.3 –1.5 –1.2 0.7 0.4 1.6 2.0 –0.5 2.2 2.1
Malta 2.5 2.0 2.5 3.2 1.0 0.8 1.2 0.9 1.3 1.6 1.8 2.0 1.3 1.7 1.8
Japan –0.3 –0.7 –0.3 –0.1 0.3 2.8 0.8 –0.1 0.5 1.1 1.1 1.3 0.6 0.8 2.1
United Kingdom 1.8 3.3 4.5 2.8 2.6 1.5 0.0 0.7 2.7 2.7 2.2 2.0 3.0 2.7 2.0
Korea 3.1 2.9 4.0 2.2 1.3 1.3 0.7 1.0 1.9 1.7 1.9 2.0 1.5 2.1 2.0
Canada 2.1 1.8 2.9 1.5 0.9 1.9 1.1 1.4 1.6 2.2 2.2 2.0 1.8 2.2 2.1
Australia 3.2 2.9 3.3 1.7 2.5 2.5 1.5 1.3 2.0 2.2 2.4 2.5 2.0 2.3 2.4
Taiwan Province of China 1.0 1.0 1.4 1.9 0.8 1.2 –0.3 1.4 0.6 1.3 1.3 2.0 1.2 1.3 1.3
Switzerland 1.0 0.7 0.2 –0.7 –0.2 0.0 –1.1 –0.4 0.5 0.7 1.0 1.0 0.6 0.9 0.9
Sweden 1.9 1.9 1.4 0.9 0.4 0.2 0.7 1.1 1.9 1.5 1.6 2.0 1.8 1.5 1.8
Singapore 1.5 2.8 5.2 4.6 2.4 1.0 –0.5 –0.5 0.6 1.2 1.0 1.0 0.4 1.8 0.4
Hong Kong SAR –0.2 2.3 5.3 4.1 4.3 4.4 3.0 2.4 1.5 2.2 2.1 2.6 1.5 2.2 2.1
Norway 2.1 2.4 1.3 0.7 2.1 2.0 2.2 3.6 1.9 1.9 2.0 2.0 1.6 1.8 2.0
Czech Republic 2.8 1.5 1.9 3.3 1.4 0.3 0.3 0.7 2.4 2.3 2.0 2.0 2.4 2.2 2.0
Israel 2.0 2.7 3.5 1.7 1.5 0.5 –0.6 –0.5 0.2 0.7 1.3 2.0 0.4 1.0 1.5
Denmark 2.1 2.3 2.8 2.4 0.8 0.6 0.5 0.3 1.1 1.4 1.7 2.0 1.0 1.6 1.8
New Zealand 2.7 2.3 4.1 1.0 1.1 1.2 0.3 0.6 1.9 1.7 2.1 2.0 1.6 2.2 2.0
Puerto Rico 2.8 2.5 2.9 1.3 1.1 0.6 –0.8 –0.3 1.9 2.2 0.8 1.6 3.1 2.2 0.8
Macao SAR ... 2.8 5.8 6.1 5.5 6.0 4.6 2.4 1.2 2.2 2.4 2.8 2.0 2.2 2.4
Iceland 6.2 5.4 4.0 5.2 3.9 2.0 1.6 1.7 1.8 2.4 2.3 2.5 1.9 2.7 2.5
San Marino ... 2.6 2.0 2.8 1.6 1.1 0.1 0.6 0.9 1.0 1.1 1.2 0.9 1.0 1.1
Memorandum
1Movements in consumer prices are shown as annual averages.
2Monthly year-over-year changes and, for several countries, on a quarterly basis.

Table A7. Emerging Market and Developing Economies: Consumer Prices1

End of Period2
2000–09 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2023 2017 2018 2019
Commonwealth of Independent States3,4 13.7 7.2 9.8 6.2 6.5 8.1 15.5 8.3 5.5 4.6 4.8 4.2 4.7 4.9 4.7
Russia 13.9 6.9 8.4 5.1 6.8 7.8 15.5 7.1 3.7 2.8 3.7 4.0 2.5 3.5 4.0
Excluding Russia 13.3 8.1 13.3 9.2 5.7 8.8 15.5 11.3 9.9 9.2 7.2 4.5 10.1 8.3 6.4
Armenia 3.6 7.3 7.7 2.5 5.8 3.0 3.7 –1.4 0.9 3.5 4.0 4.0 2.6 4.0 4.0
Azerbaijan 7.0 5.7 7.8 1.1 2.5 1.5 4.1 12.6 13.0 7.0 6.0 4.0 10.0 7.0 5.0
Belarus 31.6 7.7 53.2 59.2 18.3 18.1 13.5 11.8 6.0 6.0 6.0 5.0 4.6 6.0 6.0
Georgia 6.3 7.1 8.5 –0.9 –0.5 3.1 4.0 2.1 6.0 3.6 3.0 3.0 6.7 3.0 3.0
Kazakhstan 9.2 7.1 8.3 5.1 5.8 6.7 6.7 14.6 7.4 6.4 5.6 2.1 7.1 6.0 5.2
Kyrgyz Republic 8.5 8.0 16.6 2.8 6.6 7.5 6.5 0.4 3.2 4.5 5.0 5.0 3.7 5.5 5.2
Moldova 11.7 7.4 7.6 4.6 4.6 5.1 9.6 6.4 6.6 4.7 5.1 5.1 7.3 4.0 6.0
Tajikistan 16.0 6.5 12.4 5.8 5.0 6.1 5.8 5.9 7.3 6.3 6.0 6.0 6.7 6.3 6.0
Turkmenistan 7.6 4.4 5.3 5.3 6.8 6.0 7.4 3.6 8.0 9.4 8.2 6.0 10.4 9.4 8.2
Ukraine5 12.9 9.4 8.0 0.6 –0.3 12.1 48.7 13.9 14.4 11.0 8.0 5.0 13.7 9.0 6.5
Uzbekistan 15.8 12.3 12.4 11.9 11.7 9.1 8.5 8.0 12.5 19.5 12.9 7.3 18.9 16.9 10.1
Emerging and Developing Asia 4.0 5.1 6.5 4.6 4.6 3.4 2.7 2.8 2.4 3.3 3.3 3.5 2.8 3.5 3.3
Bangladesh 5.6 9.4 11.5 6.2 7.5 7.0 6.2 5.7 5.7 6.0 6.0 5.5 5.9 6.0 6.0
Bhutan 4.8 5.7 7.3 9.3 11.3 9.9 6.3 3.9 3.4 4.1 4.6 4.7 3.5 4.5 4.5
Brunei Darussalam 0.6 0.2 0.1 0.1 0.4 –0.2 –0.4 –0.7 –0.1 0.1 0.3 0.2 0.0 0.3 0.2
Cambodia 4.6 4.0 5.5 2.9 3.0 3.9 1.2 3.0 2.9 3.3 3.2 3.0 2.2 3.3 3.2
China 1.8 3.3 5.4 2.6 2.6 2.0 1.4 2.0 1.6 2.5 2.6 3.0 1.8 2.8 2.6
Fiji 3.4 3.7 7.3 3.4 2.9 0.5 1.4 3.9 3.4 3.3 3.0 3.0 2.8 3.3 3.0
India 5.9 9.5 9.5 10.0 9.4 5.8 4.9 4.5 3.6 5.0 5.0 4.8 4.7 5.2 4.9
Indonesia 8.5 5.1 5.3 4.0 6.4 6.4 6.4 3.5 3.8 3.5 3.4 3.0 3.6 3.5 3.6
Kiribati 3.5 –3.9 1.5 –3.0 –1.5 2.1 0.6 1.9 2.2 2.5 2.5 2.5 2.2 2.5 2.5
Lao P.D.R. 7.8 6.0 7.6 4.3 6.4 4.1 1.3 1.6 0.8 2.3 3.1 3.1 0.1 2.6 2.9
Malaysia 2.2 1.7 3.2 1.7 2.1 3.1 2.1 2.1 3.8 3.2 2.4 2.5 3.5 3.0 2.5
Maldives 3.2 6.2 11.3 10.9 3.8 2.1 1.0 0.5 2.8 1.5 1.7 2.3 0.5 2.3 2.4
Marshall Islands ... 1.8 5.4 4.3 1.9 1.1 –2.2 –1.5 0.7 1.1 1.5 2.1 0.7 1.1 1.5
Micronesia 3.1 3.7 4.1 6.3 2.2 0.7 –0.2 0.5 0.5 2.0 2.0 2.0 0.5 2.0 2.0
Mongolia 8.8 10.2 7.7 15.0 8.6 12.9 5.9 0.6 4.6 6.4 6.8 6.3 7.2 7.8 6.9
Myanmar 18.9 8.2 2.8 2.8 5.7 5.1 10.0 6.8 5.1 5.5 5.8 5.7 5.5 5.8 6.1
Nauru ... –2.0 –3.4 0.3 –1.1 0.3 9.8 8.2 5.1 2.0 2.0 2.0 1.6 2.0 2.0
Nepal 5.5 9.6 9.6 8.3 9.9 9.0 7.2 9.9 4.5 6.0 5.8 5.5 2.7 5.7 5.7
Palau ... 1.4 4.7 3.6 3.4 4.1 0.9 –1.0 0.9 2.0 2.0 2.0 0.9 2.0 2.0
Papua New Guinea 7.5 5.1 4.4 4.5 5.0 5.2 6.0 6.7 5.2 2.9 2.5 2.4 3.9 2.6 2.5
Philippines 4.9 3.8 4.7 3.2 2.9 4.2 1.4 1.8 3.2 4.2 3.8 3.0 3.3 4.1 3.7
Samoa 5.7 –0.2 2.9 6.2 –0.2 –1.2 1.9 0.1 1.3 2.9 2.5 3.0 1.0 2.6 2.5
Solomon Islands 9.1 1.0 7.4 5.9 5.4 5.2 –0.6 0.5 –0.4 1.3 1.4 4.2 –2.2 1.9 5.0
Sri Lanka 9.7 6.2 6.7 7.5 6.9 2.8 2.2 4.0 6.5 4.8 4.8 4.9 7.1 4.7 4.8
Thailand 2.4 3.3 3.8 3.0 2.2 1.9 –0.9 0.2 0.7 1.4 0.7 2.0 0.8 0.8 1.0
Timor-Leste ... 5.2 13.2 10.9 9.5 0.7 0.6 –1.3 0.6 1.8 2.7 4.0 0.8 2.5 2.8
Tonga 8.0 3.5 6.3 1.1 2.1 1.2 –1.1 2.6 8.0 3.0 2.5 2.5 7.5 2.5 2.5
Tuvalu ... –1.9 0.5 1.4 2.0 1.1 3.2 3.5 2.4 2.7 2.5 2.6 2.9 3.4 2.8
Vanuatu 2.9 2.8 0.9 1.3 1.5 0.8 2.5 0.8 3.1 4.8 3.4 3.0 3.8 4.6 3.4
Vietnam 6.5 9.2 18.7 9.1 6.6 4.1 0.6 2.7 3.5 3.8 4.0 4.0 2.6 4.0 4.0
Emerging and Developing Europe 12.7 5.7 5.5 6.1 4.5 4.1 3.2 3.2 6.2 6.8 6.3 5.2 6.8 6.6 6.0
Albania 2.7 3.6 3.4 2.0 1.9 1.6 1.9 1.3 2.0 2.5 2.8 3.0 1.8 2.8 2.9
Bosnia and Herzegovina 2.9 2.1 3.7 2.0 –0.1 –0.9 –1.0 –1.1 1.3 1.5 1.6 2.0 1.6 1.9 1.7
Bulgaria6 6.7 3.0 3.4 2.4 0.4 –1.6 –1.1 –1.3 1.2 2.0 2.1 2.1 1.8 2.1 2.1
Croatia 3.2 1.0 2.3 3.4 2.2 –0.2 –0.5 –1.1 1.1 1.5 1.5 1.8 1.2 1.2 1.4
Hungary 6.1 4.9 3.9 5.7 1.7 –0.2 –0.1 0.4 2.4 2.7 3.3 3.0 2.1 3.2 3.2
Kosovo ... 3.5 7.3 2.5 1.8 0.4 –0.5 0.3 1.5 1.0 1.9 2.2 0.7 1.8 2.0
FYR Macedonia 2.6 1.5 3.9 3.3 2.8 –0.3 –0.3 –0.2 1.4 1.8 1.9 2.0 2.4 1.8 2.0
Montenegro 10.1 0.4 3.5 4.1 2.2 –0.7 1.5 –0.3 2.4 2.8 1.8 2.0 1.9 2.7 1.8
Poland 3.5 2.6 4.3 3.7 0.9 0.0 –0.9 –0.6 2.0 2.5 2.5 2.5 2.1 2.9 2.2
Romania 15.6 6.1 5.8 3.3 4.0 1.1 –0.6 –1.6 1.3 4.7 3.1 3.1 3.3 3.5 3.2
Serbia 20.2 6.1 11.1 7.3 7.7 2.1 1.4 1.1 3.1 2.7 3.0 3.0 3.0 3.0 3.0
Turkey 21.7 8.6 6.5 8.9 7.5 8.9 7.7 7.8 11.1 11.4 10.5 8.0 11.9 10.9 10.0

Table A7. Emerging Market and Developing Economies: Consumer Prices1 (continued)
End of Period2
2000–09 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2023 2017 2018 2019
Latin America and the Caribbean7 6.2 4.2 5.2 4.6 4.6 4.9 5.5 5.6 4.1 3.6 3.5 3.4 4.1 3.6 3.5
Antigua and Barbuda 1.8 3.4 3.5 3.4 1.1 1.1 1.0 –0.5 2.5 1.4 2.0 2.0 2.8 2.0 2.0
Argentina8 8.4 10.5 9.8 10.0 10.6 ... ... ... 25.7 22.7 15.4 8.0 24.8 19.2 13.6
The Bahamas 2.3 1.6 3.1 1.9 0.4 1.2 1.9 –0.3 1.4 2.2 2.5 2.1 2.0 2.4 2.6
Barbados 3.7 5.8 9.4 4.5 1.8 1.8 –1.1 1.5 4.4 5.4 2.9 2.7 6.6 2.2 3.5
Belize 2.5 0.9 1.7 1.2 0.5 1.2 –0.9 0.7 1.1 1.3 1.9 1.7 1.1 1.6 2.1
Bolivia 4.8 2.5 9.9 4.5 5.7 5.8 4.1 3.6 2.8 3.5 4.5 4.5 2.7 4.5 4.5
Brazil 6.9 5.0 6.6 5.4 6.2 6.3 9.0 8.7 3.4 3.5 4.2 4.1 2.9 3.9 4.3
Chile 3.5 1.4 3.3 3.0 1.9 4.4 4.3 3.8 2.2 2.4 3.0 3.0 2.3 2.6 3.0
Colombia 6.3 2.3 3.4 3.2 2.0 2.9 5.0 7.5 4.3 3.5 3.4 3.0 4.1 3.4 3.0
Costa Rica 10.9 5.7 4.9 4.5 5.2 4.5 0.8 0.0 1.6 2.8 3.0 3.0 2.6 3.0 3.0
Dominica 2.0 2.8 1.1 1.4 0.0 0.8 –0.8 0.0 0.6 1.4 1.6 1.6 1.4 1.4 1.8
Dominican Republic 12.2 6.3 8.5 3.7 4.8 3.0 0.8 1.6 3.3 4.4 3.5 4.0 4.2 3.7 3.8
Ecuador 15.3 3.6 4.5 5.1 2.7 3.6 4.0 1.7 0.4 1.0 1.4 1.8 –0.2 2.5 1.4
El Salvador 3.5 1.2 5.1 1.7 0.8 1.1 –0.7 0.6 1.0 1.7 1.9 2.0 2.0 2.1 2.0
Grenada 2.8 3.4 3.0 2.4 0.0 –1.0 –0.6 1.7 2.0 2.0 1.8 1.9 0.5 1.8 1.9
Guatemala 7.0 3.9 6.2 3.8 4.3 3.4 2.4 4.4 4.4 4.2 3.7 3.8 5.7 4.2 3.5
Guyana 6.1 4.3 4.4 2.4 1.9 0.7 –0.9 0.8 2.1 2.6 2.9 3.3 2.3 2.8 3.0
Haiti 14.8 4.1 7.4 6.8 6.8 3.9 7.5 13.4 14.7 11.1 6.0 5.0 15.4 8.0 5.0
Honduras 8.2 4.7 6.8 5.2 5.2 6.1 3.2 2.7 3.9 4.7 4.5 4.5 4.7 5.0 4.5
Jamaica 10.9 12.6 7.5 6.9 9.4 8.3 3.7 2.3 4.4 4.9 5.0 5.0 5.2 5.1 5.0
Mexico 5.2 4.2 3.4 4.1 3.8 4.0 2.7 2.8 6.0 4.4 3.1 3.0 6.8 3.6 3.0
Nicaragua 8.9 5.5 8.1 7.2 7.1 6.0 4.0 3.5 3.9 6.2 7.2 7.2 5.7 6.3 7.4
Panama 2.4 3.5 5.9 5.7 4.0 2.6 0.1 0.7 0.9 2.2 2.5 2.3 0.5 2.2 2.5
Paraguay 8.2 4.6 8.2 3.7 2.7 5.0 3.1 4.1 3.6 4.2 4.0 4.0 4.5 4.0 4.0
Peru 2.6 1.5 3.4 3.7 2.8 3.2 3.5 3.6 2.8 1.6 2.0 2.0 1.4 2.3 2.0
St. Kitts and Nevis 3.4 0.9 5.8 0.8 1.1 0.2 –2.3 –0.7 0.1 1.1 2.0 2.0 0.2 2.0 2.0
St. Lucia 2.8 3.3 2.8 4.2 1.5 3.5 –1.0 –3.1 0.1 1.5 1.6 1.5 2.2 1.4 1.5
St. Vincent and the Grenadines 2.9 0.8 3.2 2.6 0.8 0.2 –1.7 –0.2 2.0 1.5 1.5 1.5 2.2 1.5 1.5
Suriname 15.3 6.9 17.7 5.0 1.9 3.4 6.9 55.5 22.0 8.9 9.6 3.9 9.3 11.2 7.8
Trinidad and Tobago 6.3 10.5 5.1 9.3 5.2 5.7 4.7 3.1 1.9 2.7 2.1 3.2 1.3 2.7 2.1
Uruguay 8.5 6.7 8.1 8.1 8.6 8.9 8.7 9.6 6.2 7.0 6.1 6.1 6.6 6.6 6.5
Venezuela8 20.8 28.2 26.1 21.1 43.5 57.3 111.8 254.4 1,087.5 13,864.6 12,874.6 12,874.6 2,818.4 12,874.6 12,874.6
Afghanistan, and Pakistan 6.7 6.6 9.3 9.8 9.2 6.7 5.6 4.7 6.3 8.2 6.8 5.7 7.2 7.3 7.4
Afghanistan ... 2.2 11.8 6.4 7.4 4.7 –0.7 4.4 5.0 5.0 5.0 5.0 3.0 5.0 5.0
Algeria 3.2 3.9 4.5 8.9 3.3 2.9 4.8 6.4 5.6 7.4 7.6 13.9 4.9 10.4 6.9
Bahrain 1.6 2.0 –0.4 2.8 3.3 2.7 1.8 2.8 1.4 2.9 4.9 1.5 1.4 2.7 4.7
Djibouti 3.4 4.0 5.1 3.7 2.4 2.9 2.1 2.7 0.7 1.0 2.0 2.5 –1.0 1.0 2.0
Egypt 7.0 11.7 11.1 8.6 6.9 10.1 11.0 10.2 23.5 20.1 13.0 7.0 29.8 10.4 15.2
Iran 14.7 12.3 21.5 30.6 34.7 15.6 11.9 9.1 9.9 12.1 11.5 10.2 10.2 11.2 13.4
Iraq ... 2.4 5.6 6.1 1.9 2.2 1.4 0.4 0.1 2.0 2.0 2.0 0.4 2.0 2.0
Jordan 3.6 4.8 4.2 4.5 4.8 2.9 –0.9 –0.8 3.3 1.5 2.5 2.5 2.5 2.5 2.5
Kuwait 2.9 4.5 4.9 3.2 2.7 3.1 3.7 3.5 1.5 2.5 3.7 2.7 1.5 2.5 3.7
Lebanon 2.4 4.0 5.0 6.6 4.8 1.9 –3.7 –0.8 4.5 4.3 3.0 2.5 5.0 3.5 2.5
Libya8 –0.1 2.5 15.9 6.1 2.6 2.4 9.8 25.9 28.0 24.3 14.4 9.0 30.0 20.0 10.0
Mauritania 6.2 6.3 5.7 4.9 4.1 3.8 0.5 1.5 2.3 3.7 5.0 4.4 1.2 4.7 5.0
Morocco 1.9 1.0 0.9 1.3 1.9 0.4 1.5 1.6 0.8 1.4 2.0 2.0 1.9 2.0 2.0
Oman 2.5 3.3 4.0 2.9 1.2 1.0 0.1 1.1 1.6 2.5 3.5 3.0 1.6 2.5 3.5
Pakistan 7.5 10.1 13.7 11.0 7.4 8.6 4.5 2.9 4.1 5.0 5.2 5.0 3.9 5.4 5.0
Qatar 5.5 –2.4 2.0 1.8 3.2 3.4 1.8 2.7 0.4 3.9 3.5 2.2 ... ... ...
Saudi Arabia 1.6 3.8 3.8 2.9 3.5 2.2 1.3 2.0 –0.9 3.7 2.0 2.1 –0.9 3.7 2.0
Somalia ... ... ... ... ... ... ... ... ... ... ... ... 5.2 2.8 2.6
Sudan9 8.6 13.0 18.3 35.4 36.5 36.9 16.9 17.8 32.4 43.5 39.5 22.5 25.2 42.0 37.0
Syria10 4.8 4.4 ... ... ... ... ... ... ... ... ... ... ... ... ...
Tunisia 2.8 3.3 3.5 5.1 5.8 4.9 4.9 3.7 5.3 7.0 6.1 3.9 6.4 6.5 5.9
United Arab Emirates 5.5 0.9 0.9 0.7 1.1 2.3 4.1 1.6 2.0 4.2 2.5 2.0 2.0 4.2 2.5
Yemen 10.9 11.2 19.5 9.9 11.0 8.2 61.4 –20.3 4.9 23.0 20.0 5.0 16.0 30.0 10.0

Table A7. Emerging Market and Developing Economies: Consumer Prices1 (continued)
End of Period2
2000–09 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2023 2017 2018 2019
Sub-Saharan Africa 10.3 8.1 9.4 9.2 6.6 6.3 7.0 11.3 11.0 9.5 8.9 7.5 10.3 9.6 9.3
Angola 62.4 14.5 13.5 10.3 8.8 7.3 10.3 32.4 31.7 27.9 17.0 6.5 26.3 24.6 15.0
Benin 3.2 2.2 2.7 6.7 1.0 –1.1 0.3 –0.8 0.1 2.9 2.9 2.7 3.0 2.8 3.0
Botswana 8.7 6.9 8.5 7.5 5.9 4.4 3.1 2.8 3.3 3.7 3.8 4.0 2.9 3.5 3.7
Burkina Faso 2.8 –0.6 2.8 3.8 0.5 –0.3 0.9 –0.2 0.4 2.0 2.0 2.0 2.1 2.0 2.0
Burundi 10.7 6.5 9.6 18.2 7.9 4.4 5.6 5.5 16.6 12.7 22.1 14.5 10.5 18.9 24.7
Cabo Verde 2.0 2.1 4.5 2.5 1.5 –0.2 0.1 –1.4 0.8 1.0 1.5 2.0 0.3 1.0 1.5
Cameroon 2.6 1.3 2.9 2.4 2.1 1.9 2.7 0.9 0.6 1.1 1.3 2.0 0.8 1.1 1.3
Central African Republic 3.4 1.5 1.2 5.9 6.6 11.6 4.5 4.6 3.8 3.5 3.2 3.0 3.6 3.6 3.4
Chad 3.5 –2.1 1.9 7.7 0.2 1.7 6.8 –1.1 –0.9 2.1 2.6 3.0 7.2 –2.3 5.4
Comoros 4.4 3.9 2.2 5.9 1.6 1.3 2.0 1.8 1.0 2.0 2.0 2.0 2.9 3.5 2.0
Democratic Republic of the Congo 61.5 23.5 14.9 0.9 0.9 1.2 1.0 18.2 41.5 25.8 13.7 3.5 55.0 29.5 15.8
Republic of Congo 2.9 0.4 1.8 5.0 4.6 0.9 3.2 3.2 0.5 1.5 1.6 3.0 1.8 1.8 1.9
Côte d’Ivoire 3.0 1.4 4.9 1.3 2.6 0.4 1.2 0.7 0.8 1.7 2.0 2.0 1.1 2.0 2.0
Equatorial Guinea 5.6 5.3 4.8 3.4 3.2 4.3 1.7 1.4 0.7 0.6 2.8 2.8 –0.2 1.3 4.0
Eritrea 18.7 11.2 3.9 6.0 6.5 10.0 9.0 9.0 9.0 9.0 9.0 9.0 9.0 9.0 9.0
Ethiopia 10.3 8.1 33.2 24.1 8.1 7.4 10.1 7.3 9.9 11.2 8.6 8.0 13.6 10.0 8.0
Gabon 1.1 1.4 1.3 2.7 0.5 4.5 –0.1 2.1 3.0 2.8 2.5 2.5 3.0 2.8 2.5
The Gambia 6.6 5.0 4.8 4.6 5.2 6.3 6.8 7.2 8.0 5.8 5.0 4.8 6.9 5.1 4.8
Ghana 17.7 6.7 7.7 7.1 11.7 15.5 17.2 17.5 12.4 8.7 8.0 6.0 11.8 8.0 8.0
Guinea 15.1 15.5 21.4 15.2 11.9 9.7 8.2 8.2 8.9 8.2 8.0 7.8 9.5 8.0 8.0
Guinea-Bissau 3.0 1.1 5.1 2.1 0.8 –1.0 1.5 1.5 1.1 2.0 2.2 3.0 –1.3 2.0 2.3
Kenya 7.3 4.3 14.0 9.4 5.7 6.9 6.6 6.3 8.0 4.8 5.0 5.0 4.5 5.1 5.0
Lesotho 7.3 3.3 6.0 5.5 5.0 4.6 4.3 6.2 5.6 5.5 5.5 5.0 6.0 5.5 5.5
Liberia 9.8 7.3 8.5 6.8 7.6 9.9 7.7 8.8 12.4 11.7 10.5 6.2 13.9 11.0 10.0
Madagascar 10.4 9.2 9.5 5.7 5.8 6.1 7.4 6.7 8.1 7.8 6.8 5.0 8.1 7.9 6.2
Malawi 14.1 7.4 7.6 21.3 28.3 23.8 21.9 21.7 11.5 10.4 7.6 5.0 7.1 9.0 7.5
Mali 2.5 1.3 3.1 5.3 –0.6 0.9 1.4 –1.8 1.8 1.4 1.7 2.2 1.1 1.6 1.7
Mauritius 5.9 2.9 6.5 3.9 3.5 3.2 1.3 1.0 3.7 5.1 4.5 3.7 4.2 5.9 4.7
Mozambique 10.5 12.7 10.4 2.1 4.2 2.3 2.4 19.2 15.3 6.7 5.7 5.5 7.2 6.5 5.5
Namibia 7.6 4.9 5.0 6.7 5.6 5.3 3.4 6.7 6.1 5.8 5.8 5.8 5.2 5.7 5.8
Niger 3.1 –2.8 2.9 0.5 2.3 –0.9 1.0 0.2 2.4 3.9 2.0 2.0 4.8 1.9 2.1
Nigeria 12.3 13.7 10.8 12.2 8.5 8.0 9.0 15.7 16.5 14.0 14.8 14.0 15.4 14.5 16.6
Rwanda 8.1 2.3 5.7 6.3 4.2 1.8 2.5 5.7 4.8 2.8 5.0 5.0 0.7 5.0 5.0
São Tomé and Príncipe 15.9 13.3 14.3 10.6 8.1 7.0 5.3 4.6 5.5 5.4 4.7 3.0 5.8 5.0 4.5
Senegal 2.0 1.2 3.4 1.4 0.7 –1.1 0.1 0.9 1.4 1.5 1.5 1.5 0.3 1.5 1.5
Seychelles 8.6 –2.4 2.6 7.1 4.3 1.4 4.0 –1.0 2.9 3.8 3.0 3.0 3.5 3.8 3.3
Sierra Leone 7.5 17.8 18.5 13.8 9.8 8.3 9.0 11.5 18.0 13.9 11.2 7.2 13.8 13.0 9.9
South Africa 6.0 4.3 5.0 5.6 5.8 6.1 4.6 6.3 5.3 5.3 5.3 5.5 4.7 5.6 5.3
South Sudan ... ... ... 45.1 0.0 1.7 52.8 379.8 187.9 104.1 108.2 39.0 117.7 96.4 125.1
Swaziland 7.5 4.5 6.1 8.9 5.6 5.7 5.0 8.0 6.3 5.4 5.4 5.5 4.7 6.0 4.9
Tanzania 6.5 7.2 12.7 16.0 7.9 6.1 5.6 5.2 5.3 4.8 5.0 5.0 4.0 5.0 5.0
Togo 3.0 1.4 3.6 2.6 1.8 0.2 1.8 0.9 –0.7 0.4 1.2 3.0 –1.6 2.4 0.2
Uganda 6.4 3.7 15.0 12.7 4.9 3.1 5.4 5.5 5.6 3.6 4.3 5.0 3.3 4.0 4.5
Zambia 17.2 8.5 8.7 6.6 7.0 7.8 10.1 17.9 6.6 8.2 8.0 8.0 6.1 8.0 8.0
Zimbabwe11 –5.5 3.0 3.5 3.7 1.6 –0.2 –2.4 –1.6 1.3 5.2 6.3 2.3 3.5 7.9 4.9
1Movements in consumer prices are shown as annual averages.
2Monthly year-over-year changes and, for several countries, on a quarterly basis.
3For many countries, inflation for the earlier years is measured on the basis of a retail price index. Consumer price index (CPI) inflation data with broader and more up-to-date coverage are
typically used for more recent years.

4Georgia, Turkmenistan, and Ukraine, which are not members of the Commonwealth of Independent States, are included in the group for reasons of geography and similarity in economic
structure.
5Starting in 2014, data exclude Crimea and Sevastopol.
7Excludes Argentina and Venezuela.
8See country-specific notes for Argentina, Libya, and Venezuela in the “Country Notes” section of the Statistical Appendix.
11The Zimbabwe dollar ceased circulating in early 2009. Data are based on IMF staff estimates of price and exchange rate developments in US dollars. IMF staff estimates of US dollar values may
differ from authorities’ estimates.

Table A8. Major Advanced Economies: General Government Fiscal Balances and Debt1
(Percent of GDP unless noted otherwise)
Average Projections
2000–09 2012 2013 2014 2015 2016 2017 2018 2019 2023
Major Advanced Economies
Net Lending/Borrowing –4.2 –6.4 –4.3 –3.6 –3.0 –3.3 –3.4 –3.5 –3.7 –2.7
Output Gap2 0.4 –2.0 –1.9 –1.3 –0.7 –0.7 –0.1 0.6 1.1 0.7
Structural Balance2 –4.1 –5.2 –3.8 –3.2 –2.8 –3.2 –3.3 –3.6 –4.2 –3.0
United States
Net Lending/Borrowing3 –4.6 –7.9 –4.4 –4.0 –3.5 –4.2 –4.6 –5.3 –5.9 –5.0
Output Gap2 1.0 –2.2 –1.9 –1.1 0.0 –0.2 0.3 1.2 1.8 1.1
Net Debt 45.2 80.5 81.3 80.8 80.5 81.5 82.3 81.4 82.7 90.2
Gross Debt 65.3 103.5 105.4 105.1 105.3 107.2 107.8 108.0 109.4 116.9
Euro Area
Net Lending/Borrowing –2.4 –3.6 –3.0 –2.6 –2.1 –1.5 –0.9 –0.6 –0.5 0.1
Output Gap2 0.6 –2.0 –2.8 –2.5 –1.9 –1.4 –0.5 0.3 0.6 0.5
Net Debt 55.2 72.2 74.6 75.0 73.9 73.2 71.0 68.9 66.9 58.6
Gross Debt 68.7 89.4 91.3 91.8 89.9 88.9 86.6 84.2 81.7 71.7
Germany
Net Lending/Borrowing –2.2 0.0 –0.1 0.3 0.6 0.8 1.1 1.5 1.7 1.4
Output Gap2 –0.3 0.5 –0.3 0.1 0.1 0.2 0.9 1.6 1.7 0.8
Structural Balance2 –2.2 –0.2 0.1 0.5 0.6 0.8 0.9 0.6 0.7 1.0
Net Debt 52.3 58.4 57.4 54.2 51.2 48.5 45.1 41.5 38.1 27.2
Gross Debt 63.9 79.8 77.4 74.7 71.0 68.2 64.1 59.8 55.7 42.4
France
Output Gap2 0.3 –1.9 –2.4 –2.5 –2.5 –2.3 –1.8 –1.1 –0.5 0.1
Net Debt 56.5 80.0 83.1 85.6 86.5 87.5 87.7 87.0 86.9 79.7
Gross Debt 65.4 90.7 93.5 95.0 95.8 96.6 97.0 96.3 96.2 89.0
Italy
Net Lending/Borrowing –3.2 –2.9 –2.9 –3.0 –2.6 –2.5 –1.9 –1.6 –0.9 0.0
Output Gap2 0.1 –2.8 –4.1 –4.1 –3.2 –2.6 –1.6 –0.7 –0.2 0.2
Structural Balance2,4 –3.9 –1.5 –0.6 –1.0 –0.7 –1.3 –1.5 –1.3 –0.7 –0.1
Net Debt 94.9 111.6 116.7 118.8 119.5 120.2 119.9 118.5 116.5 106.5
Gross Debt 103.2 123.4 129.0 131.8 131.5 132.0 131.5 129.7 127.5 116.6
Japan
Output Gap2 –1.3 –3.7 –2.3 –2.6 –2.0 –1.8 –0.8 –0.2 0.1 –0.1
Net Debt 93.6 146.7 146.4 148.5 147.6 152.8 153.0 152.6 150.8 146.3
Gross Debt5 168.9 229.0 232.5 236.1 231.3 235.6 236.4 236.0 234.2 229.6
United Kingdom
Output Gap2 0.9 –2.2 –2.0 –0.8 –0.3 –0.2 0.0 0.0 0.0 0.0
Net Debt 36.8 76.0 77.2 79.1 79.6 79.1 78.2 77.4 77.0 73.6
Gross Debt 41.7 84.5 85.6 87.4 88.2 88.2 87.0 86.3 85.9 82.5
Canada
Net Lending/Borrowing 0.5 –2.5 –1.5 0.2 –0.1 –1.1 –1.0 –0.8 –0.8 –0.7
Output Gap2 0.5 –0.2 0.4 1.2 0.1 –0.5 0.4 0.5 0.6 0.0
Structural Balance2 0.2 –2.4 –1.7 –0.7 –0.2 –0.8 –1.2 –1.1 –1.1 –0.7
Net Debt 31.3 28.3 29.3 28.0 27.7 28.5 27.8 27.4 26.6 23.5
Gross Debt 74.6 84.8 85.8 85.0 90.5 91.1 89.7 86.6 83.8 74.3
Note: The methodology and specific assumptions for each country are discussed in Box A1. The country group composites for fiscal data are calculated as the sum of the US dollar values
for the relevant individual countries.
1Debt data refer to the end of the year and are not always comparable across countries. Gross and net debt levels reported by national statistical agencies for countries that have adopted
the System of National Accounts (SNA) 2008 (Australia, Canada, Hong Kong SAR, United States) are adjusted to exclude unfunded pension liabilities of government employees’
defined-benefit pension plans. Fiscal data for the aggregated major advanced economies and the United States start in 2001, and the average for the aggregate and the United States is
therefore for the period 2001–07.
2Percent of potential GDP.
3Figures reported by the national statistical agency are adjusted to exclude items related to the accrual-basis accounting of government employees’ defined-benefit pension plans.
4Excludes one-time measures based on the authorities’ data and, if unavailable, on receipts from the sale of assets.
5Includes equity shares; nonconsolidated basis.

Table A9. Summary of World Trade Volumes and Prices

2000–09 2010–19 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019
Trade in Goods and Services
World Trade1
Volume 5.0 4.9 12.5 7.1 3.0 3.5 3.8 2.7 2.3 4.9 5.1 4.7
Price Deflator
In US Dollars 3.4 0.5 5.5 11.2 –1.7 –0.7 –1.7 –13.2 –4.1 4.4 6.1 1.2
In SDRs 2.2 1.0 6.6 7.4 1.3 0.1 –1.7 –5.8 –3.4 4.6 1.2 0.6
Volume of Trade
Exports
Imports
Emerging Market and Developing Economies 9.0 5.9 14.3 11.5 5.3 5.2 4.2 –0.9 1.8 6.4 6.0 5.6
Terms of Trade
Advanced Economies –0.2 0.1 –0.9 –1.6 –0.6 0.9 0.3 2.0 0.9 –0.2 0.6 0.2
Emerging Market and Developing Economies 1.5 0.1 1.9 4.1 0.6 –0.6 –0.6 –4.2 –1.4 0.6 1.1 –0.5
Trade in Goods
World Trade1
Volume 4.8 5.0 14.5 7.0 2.3 3.3 3.1 2.2 2.2 5.4 5.3 4.8
Price Deflator
In US Dollars 3.4 0.4 6.4 12.6 –1.4 –1.3 –2.5 –14.3 –5.0 4.8 6.3 1.0
In SDRs 2.1 0.9 7.5 8.8 1.6 –0.5 –2.5 –7.0 –4.3 5.1 1.3 0.4
World Trade Prices in US Dollars2
Manufactures 1.7 0.3 2.3 4.2 2.8 –3.0 –0.5 –2.2 –5.2 1.4 1.9 1.3
Oil 13.1 –0.6 27.9 31.6 1.0 –0.9 –7.5 –47.2 –15.7 23.3 18.0 –6.5
Nonfuel Primary Commodities 5.3 1.6 26.7 18.1 –10.2 –1.5 –3.9 –17.6 –1.5 6.8 5.6 0.5
Food 5.5 1.4 12.3 20.5 –2.9 0.4 –4.1 –17.4 2.7 2.2 2.6 1.8
Beverages 5.2 –0.3 14.1 16.6 –18.6 –11.9 20.7 –3.1 –5.0 –9.3 –3.5 4.9
Agricultural Raw Materials 0.1 2.4 33.2 22.7 –12.7 1.6 2.0 –13.5 –5.7 2.3 3.6 –0.9
Metal 9.4 1.8 48.2 13.5 –16.8 –4.3 –10.1 –23.0 –5.4 22.2 13.0 –1.1
World Trade Prices in SDRs2
Manufactures 0.5 0.8 3.4 0.7 6.0 –2.2 –0.4 6.1 –4.5 1.7 –2.9 0.7
Oil 11.8 –0.1 29.3 27.2 4.1 –0.1 –7.5 –42.7 –15.1 23.6 12.5 –7.1
Nonfuel Primary Commodities 4.1 2.1 28.0 14.2 –7.4 –0.7 –3.9 –10.5 –0.9 7.0 0.6 –0.1
Food 4.2 1.9 13.5 16.5 0.1 1.2 –4.1 –10.3 3.4 2.5 –2.2 1.2
Beverages 4.0 0.3 15.3 12.7 –16.1 –11.2 20.8 5.2 –4.4 –9.0 –8.0 4.2
Agricultural Raw Materials –1.1 2.9 34.6 18.5 –10.0 2.4 2.0 –6.1 –5.1 2.6 –1.3 –1.6
Metal 8.1 2.4 49.8 9.7 –14.3 –3.5 –10.1 –16.4 –4.8 22.5 7.7 –1.7
World Trade Prices in Euros2
Manufactures –1.0 1.3 7.4 –0.7 11.3 –6.1 –0.5 17.1 –4.9 –0.6 –7.2 0.2
Oil 10.2 0.5 34.3 25.5 9.3 –4.1 –7.6 –36.8 –15.4 20.8 7.5 –7.6
Nonfuel Primary Commodities 2.6 2.7 32.9 12.7 –2.8 –4.6 –4.0 –1.3 –1.3 4.6 –3.8 –0.6
Food 2.7 2.4 17.9 14.9 5.1 –2.8 –4.2 –1.1 3.0 0.2 –6.6 0.6
Beverages 2.5 0.8 19.8 11.2 –11.9 –14.7 20.7 16.1 –4.8 –11.1 –12.1 3.7
Agricultural Raw Materials –2.6 3.5 39.8 17.0 –5.5 –1.7 1.9 3.6 –5.5 0.2 –5.7 –2.1
Metal 6.6 2.9 55.5 8.3 –10.0 –7.3 –10.2 –7.8 –5.1 19.8 2.9 –2.2

Table A9. Summary of World Trade Volumes and Prices (continued)

2000–09 2010–19 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019
Trade in Goods
Volume of Trade
Exports
Fuel Exporters 4.9 2.6 6.8 5.7 2.6 2.2 –0.6 2.8 1.3 0.8 1.9 3.0
Nonfuel Exporters 8.8 6.4 18.7 8.3 4.4 5.8 3.9 0.5 3.0 7.8 5.9 5.9
Imports
Emerging Market and Developing Economies 8.8 5.9 15.6 11.1 5.1 4.7 2.5 –0.5 2.3 7.0 6.2 5.8
Fuel Exporters 10.9 3.0 8.2 11.7 8.2 3.9 0.7 –7.3 –5.0 3.1 6.6 1.4
Nonfuel Exporters 8.4 6.5 17.5 10.9 4.4 4.9 2.9 1.0 3.9 7.8 6.1 6.6
Price Deflators in SDRs
Exports
Advanced Economies 1.3 0.7 4.3 6.4 0.3 0.2 –2.2 –6.1 –2.4 4.4 2.0 0.9
Fuel Exporters 9.2 0.7 21.3 25.6 4.6 –2.6 –6.7 –29.6 –13.3 16.1 8.1 –3.7
Nonfuel Exporters 3.0 1.4 9.0 8.4 2.4 –0.7 –1.6 –0.8 –5.4 4.6 –0.6 0.2
Imports
Advanced Economies 1.7 0.7 6.2 8.7 1.5 –0.6 –2.2 –7.9 –3.5 4.3 1.3 0.8
Emerging Market and Developing Economies 3.0 1.2 10.7 8.5 2.6 –0.7 –2.6 –4.9 –5.8 5.7 0.2 0.0
Fuel Exporters 3.4 1.0 8.0 6.5 3.3 0.2 –2.4 –3.2 –3.8 3.1 –1.5 0.6
Nonfuel Exporters 2.9 1.3 11.3 8.9 2.4 –0.9 –2.7 –5.2 –6.2 6.2 0.6 –0.1
Terms of Trade
Advanced Economies –0.4 0.0 –1.8 –2.1 –1.1 0.8 0.1 1.9 1.2 0.1 0.7 0.1
Regional Groups
Commonwealth of Independent States3 3.8 0.1 12.6 20.6 1.6 –6.8 –1.5 –19.9 –13.3 10.3 7.7 –3.2
Emerging and Developing Asia –0.7 0.0 –6.0 –2.3 1.3 0.9 2.2 8.6 0.3 –3.4 –1.4 0.7
Emerging and Developing Europe 0.5 –0.3 –3.6 –1.9 –1.1 1.5 1.3 2.4 1.7 –2.1 –1.1 0.3
Latin America and the Caribbean 2.3 0.2 7.1 5.1 –1.7 –1.3 –2.2 –9.1 1.3 4.6 0.1 –1.1
Pakistan 5.0 –0.7 8.5 12.6 –0.1 –0.1 –4.6 –25.7 –6.2 10.2 8.5 –3.6
Middle East and North Africa 5.3 –0.7 8.5 12.8 0.4 0.0 –4.7 –26.4 –6.7 10.7 8.6 –3.7
Sub-Saharan Africa 3.6 0.8 12.3 12.6 –1.4 –1.2 –3.4 –15.5 –1.4 8.1 3.1 –2.3
Analytical Groups
Fuel 5.6 –0.3 12.3 17.9 1.2 –2.7 –4.4 –27.2 –9.8 12.6 9.7 –4.3
Nonfuel 0.2 0.2 –2.1 –0.5 0.0 0.2 1.2 4.6 0.8 –1.5 –1.2 0.3
Memorandum
World Exports in Billions of US Dollars
Goods and Services 12,353 22,699 18,713 22,284 22,596 23,304 23,729 21,070 20,669 22,654 25,273 26,701
Goods 9,792 17,755 14,907 17,931 18,135 18,549 18,635 16,223 15,757 17,422 19,474 20,515
Average Oil Price4 13.1 –0.6 27.9 31.6 1.0 –0.9 –7.5 –47.2 –15.7 23.3 18.0 –6.5
In US Dollars a Barrel 49.17 75.54 79.03 104.01 105.01 104.07 96.25 50.79 42.84 52.81 62.31 58.24
Export Unit Value of Manufactures5 1.7 0.3 2.3 4.2 2.8 –3.0 –0.5 –2.2 –5.2 1.4 1.9 1.3
1Average of annual percent change for world exports and imports.
2As represented, respectively, by the export unit value index for manufactures of the advanced economies and accounting for 83 percent of the advanced economies’ trade (export of goods) weights;
the average of UK Brent, Dubai Fateh, and West Texas Intermediate crude oil prices; and the average of world market prices for nonfuel primary commodities weighted by their 2002–04 shares in
world commodity exports.
4Percent change of average of UK Brent, Dubai Fateh, and West Texas Intermediate crude oil prices.
5Percent change for manufactures exported by the advanced economies.

Table A10. Summary of Current Account Balances

(Billions of US dollars)
Projections
2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2023
Advanced Economies –16.1 –49.1 5.7 200.6 217.4 295.8 316.7 396.2 345.8 287.2 414.3
United States –430.7 –444.6 –426.2 –349.5 –373.8 –434.6 –451.7 –466.2 –614.7 –727.3 –739.7
Euro Area –7.7 –12.4 174.1 293.1 326.6 371.7 408.2 442.4 460.8 478.1 510.5
Germany 192.3 229.7 248.9 252.5 291.0 301.1 297.5 296.6 346.8 364.1 413.6
France –22.2 –28.3 –32.8 –24.5 –36.2 –10.7 –21.0 –36.8 –39.1 –26.6 –8.2
Italy –72.6 –68.3 –7.0 21.1 41.3 28.4 50.4 56.1 56.1 50.8 23.2
Spain –56.2 –47.4 –3.1 20.7 14.9 13.5 23.8 21.7 23.7 26.2 34.6
Japan 221.0 129.8 59.7 45.9 36.8 134.1 188.1 195.4 194.5 199.0 238.9
United Kingdom –92.3 –62.3 –113.1 –151.8 –161.4 –150.0 –153.9 –106.7 –110.1 –103.2 –101.0
Canada –58.2 –49.6 –65.7 –59.4 –43.2 –55.9 –49.3 –49.3 –57.3 –48.6 –42.1
Other Advanced Economies1 283.0 266.4 272.2 347.8 360.1 365.5 348.5 346.1 393.5 406.4 450.3
Emerging Market and Developing Economies 280.1 379.1 355.7 177.5 168.8 –59.7 –90.6 –25.2 –23.7 –76.0 –345.5
Regional Groups
Commonwealth of Independent States2 68.9 107.3 67.5 18.0 57.8 52.4 –0.3 26.6 63.5 54.2 51.7
Russia 67.5 97.3 71.3 33.4 57.5 68.8 25.5 40.2 76.8 67.3 67.6
Excluding Russia 1.4 10.0 –3.8 –15.4 0.2 –16.4 –25.9 –13.6 –13.3 –13.1 –15.8
Emerging and Developing Asia 233.3 98.2 122.1 99.3 231.1 312.0 226.6 151.0 122.9 131.4 –6.7
China 237.8 136.1 215.4 148.2 236.0 304.2 202.2 164.9 166.7 179.1 132.2
India –47.9 –78.2 –87.8 –32.3 –26.8 –22.1 –15.3 –51.2 –66.6 –67.4 –121.2
ASEAN-53 45.4 49.4 6.3 –3.6 22.4 30.7 43.5 49.3 38.5 36.2 –1.2
Emerging and Developing Europe –86.9 –119.5 –81.9 –72.0 –59.9 –35.3 –32.8 –49.6 –62.2 –63.2 –72.6
Latin America and the Caribbean –95.3 –111.3 –136.6 –162.8 –184.5 –174.0 –96.2 –85.4 –116.9 –133.6 –171.5
Brazil –75.8 –77.0 –74.2 –74.8 –104.2 –59.4 –23.5 –9.8 –33.5 –40.1 –51.1
Mexico –5.0 –12.4 –18.4 –30.9 –23.7 –29.3 –22.8 –18.8 –23.3 –28.5 –31.0
Pakistan 170.6 413.1 411.4 331.2 188.0 –123.3 –130.3 –28.0 17.8 –9.1 –64.5
Sub-Saharan Africa –10.5 –8.8 –26.9 –36.2 –63.7 –91.5 –57.6 –39.8 –48.7 –55.6 –81.9
South Africa –5.6 –9.2 –20.3 –21.6 –18.7 –14.0 –9.6 –7.9 –10.6 –11.9 –15.4
Analytical Groups
Fuel 309.6 619.9 597.1 465.4 311.3 –77.4 –86.2 63.6 145.4 113.0 60.3
Nonfuel –27.8 –240.8 –241.4 –288.0 –142.5 17.7 –4.3 –88.8 –169.1 –189.0 –405.8
Of Which, Primary Products –11.7 –31.0 –68.0 –81.8 –55.4 –54.8 –43.8 –56.0 –61.9 –69.6 –87.0
Net Debtor Economies –282.7 –381.4 –433.5 –399.9 –369.6 –309.7 –217.2 –232.3 –310.2 –336.3 –490.5
Rescheduling during 2012–16 –21.2 –31.5 –50.5 –50.1 –35.1 –47.7 –52.5 –39.1 –35.4 –38.0 –59.3
Memorandum
World 264.0 330.0 361.4 378.1 386.3 236.1 226.2 371.0 322.2 211.2 68.8
European Union –23.9 62.4 188.4 269.2 286.0 338.0 332.5 417.2 470.7 495.1 529.2
Low-Income Developing Countries –15.5 –19.6 –31.5 –39.6 –45.8 –77.2 –44.4 –30.8 –48.9 –58.8 –89.5
Middle East and North Africa 170.0 408.1 413.9 333.7 190.0 –122.0 –126.8 –15.9 33.4 5.9 –46.5

Table A10. Summary of Current Account Balances (continued)

(Percent of GDP)
Projections
2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2023
Advanced Economies 0.0 –0.1 0.0 0.4 0.5 0.7 0.7 0.8 0.7 0.5 0.7
United States –2.9 –2.9 –2.6 –2.1 –2.1 –2.4 –2.4 –2.4 –3.0 –3.4 –3.0
Euro Area –0.1 –0.1 1.4 2.2 2.4 3.2 3.4 3.5 3.2 3.2 2.9
Germany 5.6 6.1 7.0 6.7 7.5 8.9 8.5 8.0 8.2 8.2 7.8
France –0.8 –1.0 –1.2 –0.9 –1.3 –0.4 –0.9 –1.4 –1.3 –0.9 –0.2
Italy –3.4 –3.0 –0.3 1.0 1.9 1.5 2.7 2.9 2.6 2.2 0.9
Spain –3.9 –3.2 –0.2 1.5 1.1 1.1 1.9 1.7 1.6 1.7 1.8
Japan 3.9 2.1 1.0 0.9 0.8 3.1 3.8 4.0 3.8 3.7 4.0
United Kingdom –3.8 –2.4 –4.2 –5.5 –5.3 –5.2 –5.8 –4.1 –3.7 –3.4 –2.9
Canada –3.6 –2.8 –3.6 –3.2 –2.4 –3.6 –3.2 –3.0 –3.2 –2.5 –1.7
Economies 1.2 1.4 1.3 0.6 0.5 –0.2 –0.3 –0.1 –0.1 –0.2 –0.7
Regional Groups
Commonwealth of Independent States2 3.2 4.0 2.4 0.6 2.1 2.8 0.0 1.3 2.8 2.3 1.9
Russia 4.1 4.7 3.2 1.5 2.8 5.0 2.0 2.6 4.5 3.8 3.4
Emerging and Developing Asia 2.4 0.8 1.0 0.7 1.5 2.0 1.4 0.9 0.6 0.6 0.0
China 3.9 1.8 2.5 1.5 2.2 2.7 1.8 1.4 1.2 1.2 0.6
India –2.8 –4.3 –4.8 –1.7 –1.3 –1.1 –0.7 –2.0 –2.3 –2.1 –2.6
ASEAN-53 2.7 2.6 0.3 –0.2 1.1 1.5 2.0 2.1 1.5 1.3 0.0
Brazil –3.4 –2.9 –3.0 –3.0 –4.2 –3.3 –1.3 –0.5 –1.6 –1.8 –1.9
Mexico –0.5 –1.0 –1.5 –2.4 –1.8 –2.5 –2.1 –1.6 –1.9 –2.2 –2.0
and Pakistan 6.1 12.7 12.4 9.8 5.4 –4.0 –4.2 –0.9 0.5 –0.3 –1.5
South Africa –1.5 –2.2 –5.1 –5.9 –5.3 –4.4 –3.3 –2.3 –2.9 –3.1 –3.4
Analytical Groups
Fuel 6.4 10.5 9.7 7.3 5.1 –1.6 –1.9 1.3 2.7 2.0 0.9
Nonfuel –0.2 –1.2 –1.1 –1.2 –0.6 0.1 0.0 –0.3 –0.6 –0.6 –0.9
Memorandum
World 0.4 0.5 0.5 0.5 0.5 0.3 0.3 0.5 0.4 0.2 0.1
European Union –0.1 0.3 1.1 1.5 1.5 2.1 2.0 2.4 2.4 2.4 2.2

Table A10. Summary of Current Account Balances (continued)

(Percent of exports of goods and services)
Projections
2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2023
Advanced Economies –0.1 –0.3 0.0 1.4 1.5 2.2 2.4 2.8 2.2 1.7 2.0
United States –23.2 –20.9 –19.2 –15.2 –15.7 –19.2 –20.5 –20.0 –24.5 –27.4 –22.6
Euro Area –0.3 –0.4 5.4 8.6 9.2 11.5 12.7 12.6 ... ... ...
Germany 13.3 13.6 15.3 14.8 16.3 19.0 18.5 17.0 17.5 17.3 16.1
France –3.1 –3.4 –4.1 –2.9 –4.2 –1.4 –2.8 –4.7 –4.7 –3.0 –0.8
Italy –13.5 –11.1 –1.2 3.4 6.5 5.2 9.1 9.2 7.9 6.7 2.5
Spain –15.3 –11.0 –0.8 4.7 3.3 3.4 5.8 4.8 4.5 4.7 4.9
Japan 25.4 13.9 6.5 5.5 4.3 17.1 23.2 22.4 20.6 20.3 22.4
United Kingdom –13.4 –7.8 –14.3 –18.7 –18.9 –19.0 –20.5 –13.3 –11.9 –10.9 –10.0
Canada –12.4 –9.1 –11.9 –10.7 –7.6 –11.4 –10.4 –9.7 –10.2 –8.1 –5.7
Economies 4.0 4.5 3.9 2.0 2.1 –0.7 –1.2 –0.3 –0.2 –0.7 –2.9
Regional Groups
Commonwealth of Independent States2 10.3 12.1 7.4 2.0 6.8 8.9 –0.1 4.3 8.9 7.5 6.1
Russia 15.3 17.0 12.1 5.6 10.2 17.5 7.7 9.8 16.1 14.2 12.1
Emerging and Developing Asia 8.3 2.8 3.3 2.6 5.8 8.2 6.2 3.7 2.7 2.7 –0.1
China 14.8 6.8 9.9 6.3 9.6 12.9 9.2 6.8 6.3 6.4 4.0
India –12.6 –17.3 –19.4 –6.9 –5.7 –5.3 –3.5 –10.4 –12.1 –11.1 –13.9
ASEAN-53 6.1 5.5 0.7 –0.4 2.3 3.4 4.8 4.7 3.4 2.9 –0.1
Brazil –32.7 –26.3 –26.4 –26.8 –39.5 –26.5 –10.8 –3.9 –12.5 –14.4 –15.4
Mexico –1.6 –3.4 –4.8 –7.7 –5.7 –7.3 –5.7 –4.3 –5.0 –5.7 –4.9
Pakistan 13.6 27.0 24.3 20.8 13.4 –9.9 –11.7 –2.4 1.7 –0.3 –3.9
South Africa –5.2 –7.3 –17.3 –19.0 –17.0 –14.5 –10.8 –7.6 –9.5 –10.2 –11.4
Analytical Groups
Fuel 16.5 25.3 22.6 18.4 13.7 –4.1 –5.5 3.7 7.7 6.0 3.0
Nonfuel –0.6 –4.1 –4.0 –4.5 –2.2 0.3 –0.1 –1.4 –2.3 –2.5 –4.2
Memorandum
World 1.3 1.5 1.5 1.6 1.7 1.2 1.1 1.6 1.3 0.8 0.2
European Union –0.4 0.8 2.6 3.5 3.6 4.7 4.6 5.3 5.2 5.2 4.5
structure.

Table A11. Advanced Economies: Balance on Current Account

(Percent of GDP)
Projections
2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2023
Advanced Economies 0.0 –0.1 0.0 0.4 0.5 0.7 0.7 0.8 0.7 0.5 0.7
United States –2.9 –2.9 –2.6 –2.1 –2.1 –2.4 –2.4 –2.4 –3.0 –3.4 –3.0
Euro Area1 –0.1 –0.1 1.4 2.2 2.4 3.2 3.4 3.5 3.2 3.2 2.9
Germany 5.6 6.1 7.0 6.7 7.5 8.9 8.5 8.0 8.2 8.2 7.8
France –0.8 –1.0 –1.2 –0.9 –1.3 –0.4 –0.9 –1.4 –1.3 –0.9 –0.2
Italy –3.4 –3.0 –0.3 1.0 1.9 1.5 2.7 2.9 2.6 2.2 0.9
Spain –3.9 –3.2 –0.2 1.5 1.1 1.1 1.9 1.7 1.6 1.7 1.8
Netherlands 7.4 9.1 10.8 9.9 8.6 8.7 8.4 9.8 9.6 8.9 7.7
Belgium 1.8 –1.1 –0.1 –0.3 –0.9 –0.1 0.1 0.1 0.3 0.2 0.5
Austria 2.9 1.6 1.5 1.9 2.5 1.9 2.1 2.1 2.5 2.0 1.9
Greece –11.4 –10.0 –3.8 –2.0 –1.6 –0.2 –1.1 –0.8 –0.8 –0.6 0.0
Portugal –10.1 –6.0 –1.8 1.6 0.1 0.1 0.6 0.5 0.2 –0.1 –1.4
Ireland –1.2 –1.6 –2.6 2.1 1.6 10.9 3.3 12.5 9.8 8.7 6.5
Finland 1.2 –1.8 –1.9 –1.6 –1.3 –1.0 –1.4 0.7 1.4 1.9 2.2
Slovak Republic –4.7 –5.0 0.9 1.9 1.1 –1.7 –1.5 –1.5 –0.3 0.5 0.7
Lithuania –1.3 –4.5 –1.4 0.8 3.2 –2.8 –1.1 1.0 –0.1 –0.6 –2.8
Slovenia –0.1 0.2 2.1 4.4 5.8 4.4 5.2 6.5 5.7 5.2 2.8
Luxembourg 6.7 6.0 5.6 5.5 5.2 5.1 4.8 5.5 5.4 5.3 5.0
Latvia 2.0 –3.2 –3.6 –2.7 –1.7 –0.5 1.4 –0.8 –1.9 –2.2 –2.5
Estonia 1.8 1.3 –1.9 0.5 0.3 2.0 1.9 3.2 2.0 0.7 –2.5
Cyprus –11.3 –4.1 –6.0 –4.9 –4.3 –1.5 –4.9 –4.7 –4.1 –4.6 –4.5
Malta –4.7 –0.2 1.7 2.7 8.8 4.5 6.5 10.2 9.9 9.5 9.6
Japan 3.9 2.1 1.0 0.9 0.8 3.1 3.8 4.0 3.8 3.7 4.0
United Kingdom –3.8 –2.4 –4.2 –5.5 –5.3 –5.2 –5.8 –4.1 –3.7 –3.4 –2.9
Korea 2.6 1.6 4.2 6.2 6.0 7.7 7.0 5.1 5.5 5.8 5.6
Canada –3.6 –2.8 –3.6 –3.2 –2.4 –3.6 –3.2 –3.0 –3.2 –2.5 –1.7
Australia –3.7 –3.1 –4.3 –3.4 –3.1 –4.7 –3.1 –2.3 –1.9 –2.3 –2.3
Taiwan Province of China 8.3 7.8 8.9 10.0 11.7 14.3 13.6 13.8 13.6 13.5 13.5
Switzerland 14.8 7.9 10.3 11.3 8.5 10.9 9.4 9.3 9.7 9.4 9.0
Sweden 6.0 5.6 5.6 5.2 4.5 4.5 4.2 3.2 3.1 3.1 3.0
Singapore 23.4 22.1 17.0 16.5 18.7 18.6 19.0 18.8 18.9 18.7 16.0
Hong Kong SAR 7.0 5.6 1.6 1.5 1.4 3.3 4.0 3.0 3.1 3.2 3.5
Norway 10.9 12.4 12.5 10.3 10.5 7.9 3.8 5.1 6.1 6.5 7.0
Czech Republic –3.6 –2.1 –1.6 –0.5 0.2 0.2 1.1 1.1 0.3 0.4 –1.4
Israel 3.7 2.2 0.5 3.1 4.0 4.9 3.5 3.0 2.6 2.7 3.0
Denmark 6.6 6.6 6.3 7.8 8.9 8.8 7.3 7.6 7.6 7.2 6.3
New Zealand –2.3 –2.8 –3.9 –3.2 –3.2 –3.1 –2.3 –2.7 –2.6 –3.0 –2.8
Puerto Rico ... ... ... ... ... ... ... ... ... ... ...
Macao SAR 39.4 40.9 39.3 40.2 34.2 25.3 26.9 30.4 32.1 33.1 34.8
Iceland –6.6 –5.2 –3.9 5.9 4.0 5.3 7.7 3.6 3.3 2.6 3.8
San Marino ... ... ... ... ... ... ... ... ... ... ...
Memorandum
Major Advanced Economies –0.8 –0.8 –1.0 –0.8 –0.7 –0.5 –0.4 –0.3 –0.6 –0.7 –0.5
Euro Area2 0.5 0.8 2.2 2.8 3.0 3.7 3.6 3.8 3.8 3.7 3.4
1Data corrected for reporting discrepancies in intra-area transactions.
2Data calculated as the sum of the balances of individual euro area countries.

Table A12. Emerging Market and Developing Economies: Balance on Current Account
(Percent of GDP)
Projections
2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2023
Commonwealth of Independent States1 3.2 4.0 2.4 0.6 2.1 2.8 0.0 1.3 2.8 2.3 1.9
Russia 4.1 4.7 3.2 1.5 2.8 5.0 2.0 2.6 4.5 3.8 3.4
Armenia –13.6 –10.4 –10.0 –7.3 –7.6 –2.6 –2.3 –2.6 –2.8 –2.8 –4.6
Azerbaijan 28.4 26.0 21.4 16.6 13.9 –0.4 –3.6 3.5 5.6 7.0 7.2
Belarus –14.5 –8.2 –2.8 –10.0 –6.6 –3.3 –3.5 –1.8 –2.5 –2.7 –1.7
Georgia –10.3 –12.8 –11.7 –5.8 –10.7 –12.0 –12.8 –9.3 –10.5 –9.5 –7.7
Kazakhstan 0.9 5.3 0.5 0.5 2.8 –2.8 –6.7 –2.9 –1.4 –1.3 –0.3
Kyrgyz Republic –2.2 –2.9 3.7 –13.3 –16.0 –16.0 –12.1 –7.8 –13.6 –12.2 –10.1
Moldova –7.5 –11.7 –6.7 –4.2 –3.9 –5.7 –4.0 –4.7 –3.7 –4.7 –4.0
Tajikistan –9.6 –7.3 –9.2 –7.8 –2.8 –6.0 –3.8 –2.6 –5.2 –4.7 –4.0
Turkmenistan –12.9 –0.8 –0.9 –7.3 –6.1 –15.6 –19.9 –11.5 –9.0 –7.8 –7.6
Ukraine 2 –2.2 –6.3 –8.1 –9.2 –3.9 –0.3 –4.1 –3.7 –3.7 –3.5 –4.0
Uzbekistan 7.1 5.7 1.2 2.8 1.7 0.7 0.7 3.7 0.2 –1.1 –2.3
Emerging and Developing Asia 2.4 0.8 1.0 0.7 1.5 2.0 1.4 0.9 0.6 0.6 0.0
Bangladesh 0.4 –1.0 0.7 1.2 1.2 1.7 0.6 –1.2 –2.0 –2.3 –1.7
Bhutan –22.2 –29.8 –21.4 –25.4 –26.4 –28.3 –29.1 –20.5 –19.6 –15.9 –3.8
Brunei Darussalam 36.6 34.7 29.8 20.9 30.7 16.0 9.6 6.1 5.0 13.1 18.6
Cambodia –9.3 –5.9 –8.2 –13.0 –9.8 –9.3 –8.8 –8.8 –10.7 –9.5 –7.8
China 3.9 1.8 2.5 1.5 2.2 2.7 1.8 1.4 1.2 1.2 0.6
Fiji –4.5 –5.1 –1.4 –9.7 –7.6 –3.6 –5.0 –4.5 –5.2 –4.5 –3.5
India –2.8 –4.3 –4.8 –1.7 –1.3 –1.1 –0.7 –2.0 –2.3 –2.1 –2.6
Indonesia 0.7 0.2 –2.7 –3.2 –3.1 –2.0 –1.8 –1.7 –1.9 –1.9 –2.0
Kiribati –2.2 –13.1 –4.4 8.3 25.0 46.7 19.4 9.0 17.0 7.1 –15.7
Lao P.D.R. –16.5 –15.3 –26.0 –28.4 –20.0 –18.0 –12.0 –13.0 –14.9 –13.7 –9.2
Malaysia 10.1 10.9 5.2 3.5 4.4 3.0 2.4 3.0 2.4 2.2 1.9
Maldives –7.3 –14.8 –6.6 –4.3 –3.2 –7.4 –24.5 –22.1 –18.0 –15.2 –10.5
Marshall Islands –20.9 2.0 0.1 –5.3 1.9 16.5 8.5 5.5 4.5 3.8 –2.3
Micronesia –15.4 –18.8 –13.4 –10.1 1.2 4.2 3.3 3.6 3.2 3.1 3.2
Mongolia –13.0 –26.5 –27.4 –25.4 –11.3 –4.0 –6.3 –8.8 –6.4 –8.3 2.0
Myanmar –1.1 –1.8 –4.0 –4.9 –2.2 –5.1 –3.9 –5.3 –5.4 –5.6 –5.8
Nauru 46.3 26.1 38.1 18.8 –13.5 –9.5 1.7 0.7 –0.7 0.1 2.1
Nepal –2.4 –1.0 4.8 3.3 4.5 5.0 6.3 –0.4 –3.6 –3.1 –2.6
Palau –9.0 –11.8 –11.2 –11.6 –15.0 –7.7 –10.4 –13.6 –13.4 –13.8 –13.4
Papua New Guinea –20.4 –24.0 –36.1 –30.8 1.3 13.3 16.7 16.8 20.2 19.2 15.9
Philippines 3.6 2.5 2.8 4.2 3.8 2.5 –0.3 –0.4 –0.5 –0.6 –1.2
Samoa –7.0 –4.3 –6.3 –0.4 –8.1 –2.8 –4.2 –1.3 –1.8 –3.4 –4.0
Solomon Islands –32.9 –8.3 1.7 –3.4 –4.3 –3.0 –3.9 –4.4 –5.0 –6.4 –5.9
Sri Lanka –1.9 –7.1 –5.8 –3.4 –2.5 –2.4 –2.4 –2.9 –2.7 –2.5 –2.1
Thailand 3.4 2.5 –0.4 –1.2 3.7 8.0 11.7 10.8 9.3 8.6 3.5
Timor-Leste 42.0 41.4 41.0 42.3 27.0 6.4 –21.6 –13.0 –22.6 –24.7 –21.3
Tonga –19.0 –16.8 –12.3 –8.3 –10.7 –14.7 –12.7 –10.9 –12.1 –11.9 –9.7
Tuvalu –42.2 –63.6 –36.4 –17.1 –15.2 –23.8 –32.0 –19.2 –20.5 –29.7 –25.3
Vanuatu –5.4 –8.1 –6.5 –3.3 –0.3 –10.6 –4.1 –9.0 –9.2 –8.6 –7.4
Vietnam –3.8 0.2 6.0 4.5 4.9 –0.1 3.0 4.1 3.0 2.4 0.5
Albania –11.3 –13.2 –10.1 –9.3 –10.8 –8.6 –7.6 –7.2 –6.7 –6.7 –6.1
Bosnia and Herzegovina –6.0 –9.5 –8.7 –5.3 –7.4 –5.7 –5.1 –5.2 –5.9 –6.5 –4.8
Bulgaria –1.7 0.3 –0.9 1.3 0.1 0.0 2.3 4.5 3.0 2.3 0.1
Croatia –1.1 –0.7 –0.1 0.9 2.0 4.5 2.5 3.7 3.0 2.1 0.3
Hungary 0.3 0.7 1.8 3.8 1.5 3.5 6.0 3.6 2.5 2.4 2.0
Kosovo –11.6 –12.7 –5.8 –3.6 –7.0 –8.7 –8.9 –8.7 –8.9 –8.6 –7.7
FYR Macedonia –2.0 –2.5 –3.2 –1.6 –0.5 –2.0 –2.7 –1.3 –1.5 –1.8 –2.7
Montenegro –22.7 –17.6 –18.5 –14.5 –15.2 –13.2 –18.1 –18.9 –19.0 –17.8 –10.2
Poland –5.4 –5.2 –3.7 –1.3 –2.1 –0.6 –0.3 0.0 –0.9 –1.2 –2.1
Romania –5.1 –5.0 –4.8 –1.1 –0.7 –1.2 –2.1 –3.5 –3.7 –3.7 –3.5
Serbia –6.4 –8.6 –11.5 –6.1 –6.0 –4.7 –3.1 –4.6 –4.5 –4.1 –3.8
Turkey –5.8 –8.9 –5.5 –6.7 –4.7 –3.7 –3.8 –5.5 –5.4 –4.8 –3.3

Table A12. Emerging Market and Developing Economies: Balance on Current Account (continued)
(Percent of GDP)
Projections
2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2023
Antigua and Barbuda ... ... ... ... 2.0 6.8 0.2 –7.0 –12.1 –2.3 –1.0
Argentina –0.4 –1.0 –0.4 –2.1 –1.6 –2.7 –2.7 –4.8 –5.1 –5.5 –5.9
The Bahamas –7.9 –10.9 –14.3 –14.2 –20.2 –14.3 –7.7 –16.4 –13.6 –8.8 –4.7
Barbados –5.6 –11.8 –8.4 –8.4 –9.3 –6.1 –4.4 –3.7 –3.0 –2.9 –2.5
Belize –2.9 –1.1 –1.2 –4.5 –7.8 –9.8 –9.0 –7.7 –6.0 –5.8 –5.0
Bolivia 3.9 0.3 7.2 3.4 1.7 –5.8 –5.7 –5.8 –5.4 –5.2 –4.2
Brazil –3.4 –2.9 –3.0 –3.0 –4.2 –3.3 –1.3 –0.5 –1.6 –1.8 –1.9
Chile 1.4 –1.6 –3.9 –4.0 –1.7 –2.3 –1.4 –1.5 –1.8 –1.9 –2.2
Colombia –3.0 –2.9 –3.1 –3.3 –5.2 –6.4 –4.3 –3.4 –2.6 –2.6 –2.3
Costa Rica –3.2 –5.3 –5.1 –4.8 –4.8 –3.6 –2.6 –3.1 –3.1 –3.0 –3.3
Dominica ... ... ... ... –7.1 –1.9 0.8 –17.8 –37.1 –21.5 –10.4
Dominican Republic –7.5 –7.5 –6.5 –4.1 –3.3 –1.9 –1.1 –0.2 –1.0 –1.4 –2.6
Ecuador –2.3 –0.5 –0.2 –1.0 –0.5 –2.1 1.5 –0.4 –0.1 0.3 0.9
El Salvador –2.5 –4.8 –5.4 –6.5 –4.8 –3.6 –2.0 –2.2 –3.2 –3.1 –3.9
Grenada ... ... ... ... –4.4 –3.8 –3.2 –6.6 –7.1 –6.4 –7.0
Guatemala –1.4 –3.4 –2.6 –2.5 –2.1 –0.2 1.5 1.4 1.1 0.6 –1.3
Guyana –8.4 –12.2 –11.3 –13.3 –9.5 –5.1 0.3 –4.2 –5.2 –4.7 44.1
Haiti –1.5 –4.3 –5.7 –6.6 –8.5 –3.1 –1.0 –2.9 –4.1 –3.0 –3.1
Honduras –4.3 –8.0 –8.5 –9.5 –6.9 –4.7 –2.7 –1.7 –3.9 –4.0 –3.9
Jamaica –8.0 –12.2 –11.1 –9.2 –7.5 –3.2 –2.7 –2.8 –2.9 –2.9 –0.8
Mexico –0.5 –1.0 –1.5 –2.4 –1.8 –2.5 –2.1 –1.6 –1.9 –2.2 –2.0
Nicaragua –8.9 –11.9 –10.7 –10.9 –7.1 –9.0 –8.6 –6.2 –7.8 –7.7 –8.4
Panama –10.3 –12.6 –10.0 –9.4 –13.1 –7.9 –5.5 –6.1 –6.0 –4.3 –3.6
Paraguay 0.2 0.8 –1.2 2.1 –0.2 –1.1 1.5 –1.8 –2.0 –1.2 –0.5
Peru –2.4 –1.8 –2.8 –4.7 –4.4 –4.8 –2.7 –1.3 –0.7 –1.1 –2.0
St. Kitts and Nevis ... ... ... ... –4.9 –9.7 –11.4 –12.6 –13.1 –12.1 –10.2
St. Lucia ... ... ... ... 3.4 6.9 –1.9 0.3 –1.0 –0.6 –0.1
St. Vincent and the Grenadines ... ... ... ... –25.7 –14.9 –15.8 –14.4 –13.5 –13.1 –10.6
Suriname 14.9 9.8 3.3 –3.8 –7.9 –16.5 –3.1 8.9 6.3 5.2 1.8
Trinidad and Tobago 18.9 16.8 13.2 20.4 14.5 3.8 –10.7 –5.6 –3.0 –4.0 –6.5
Uruguay ... ... –4.0 –3.4 –3.0 –0.7 1.6 1.6 0.6 –0.1 –1.5
Venezuela 1.9 4.9 0.8 2.0 2.3 –6.6 –1.6 2.0 2.4 3.6 2.5
and Pakistan 6.1 12.7 12.4 9.8 5.4 –4.0 –4.2 –0.9 0.5 –0.3 –1.5
Afghanistan 29.4 26.6 10.8 0.3 5.5 7.5 7.1 1.6 0.6 –0.2 –1.9
Algeria 7.5 9.9 5.9 0.4 –4.4 –16.5 –16.6 –12.3 –9.3 –9.7 –4.0
Bahrain 3.0 8.8 8.4 7.4 4.6 –2.4 –4.6 –3.9 –3.2 –3.3 –4.3
Djibouti 2.8 –13.1 –18.8 –23.3 –25.1 –31.8 –30.4 –23.2 –19.5 –18.8 –11.5
Egypt –1.9 –2.5 –3.6 –2.2 –0.9 –3.7 –6.0 –6.5 –4.4 –3.9 –3.2
Iran 4.2 10.4 6.0 6.7 3.2 0.3 4.0 4.3 7.0 6.3 6.0
Iraq 1.6 10.9 5.1 1.1 2.6 –6.5 –8.6 0.7 0.2 –1.6 –4.4
Jordan –7.1 –10.3 –15.2 –10.4 –7.3 –9.1 –9.3 –8.7 –8.5 –7.9 –6.4
Kuwait 31.8 42.9 45.5 40.3 33.4 3.5 –4.5 2.0 5.8 3.6 –0.3
Lebanon –20.2 –15.2 –25.7 –29.5 –31.2 –19.6 –23.3 –25.0 –25.8 –25.2 –23.4
Libya 3 21.1 9.9 29.9 0.0 –78.4 –54.4 –24.7 2.2 –10.7 –10.9 –6.0
Mauritania –8.2 –5.0 –24.1 –22.0 –27.3 –19.8 –14.9 –10.0 –9.9 –8.4 1.4
Morocco –4.4 –7.6 –9.3 –7.6 –5.9 –2.1 –4.4 –3.8 –3.6 –3.5 –2.0
Oman 8.3 13.0 10.2 6.6 5.2 –15.9 –18.4 –11.5 –6.2 –6.0 –8.6
Pakistan –2.2 0.1 –2.1 –1.1 –1.3 –1.0 –1.7 –4.1 –4.8 –4.4 –3.8
Qatar 19.1 31.1 33.2 30.4 24.0 8.4 –5.5 1.3 2.5 1.8 1.6
Saudi Arabia 12.7 23.6 22.4 18.1 9.8 –8.7 –3.7 2.7 5.4 3.6 –2.2
Somalia ... ... ... –3.6 –5.3 –4.7 –6.3 –6.7 –7.2 –6.5 –5.3
Sudan4 –1.8 –0.4 –8.6 –10.1 –8.1 –10.2 –8.9 –5.5 –6.2 –6.8 –5.8
Syria5 –2.8 ... ... ... ... ... ... ... ... ... ...
Tunisia –4.8 –7.4 –8.3 –8.4 –9.1 –8.9 –8.8 –10.1 –9.2 –7.8 –5.8
United Arab Emirates 4.2 12.6 19.7 19.0 13.5 4.9 1.4 4.7 5.3 5.1 4.0
Yemen –3.4 –3.0 –1.7 –3.1 –1.7 –6.5 –5.2 –1.0 –6.5 –3.8 –2.7

Table A12. Emerging Market and Developing Economies: Balance on Current Account (continued)
(Percent of GDP)
Projections
2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2023
Angola 9.1 12.6 12.2 6.7 –3.0 –10.0 –5.1 –4.5 –2.2 –0.1 –2.9
Benin –8.2 –7.3 –7.4 –7.4 –8.6 –9.0 –9.4 –9.4 –8.5 –7.9 –5.1
Botswana –2.8 3.1 0.3 8.9 15.4 7.8 11.7 10.8 8.3 7.5 6.7
Burkina Faso –2.3 –4.0 –6.7 –11.3 –8.1 –8.5 –7.3 –8.3 –7.5 –6.5 –6.7
Burundi –12.2 –14.4 –18.6 –19.3 –18.5 –17.7 –13.1 –12.7 –13.2 –11.9 –9.9
Cabo Verde –12.4 –16.3 –12.6 –4.9 –9.1 –3.2 –2.8 –8.8 –9.5 –10.0 –9.9
Cameroon –2.5 –2.7 –3.3 –3.6 –4.0 –3.8 –3.2 –2.5 –2.5 –2.4 –2.4
Central African Republic –10.2 –7.6 –4.6 –3.0 –5.6 –9.0 –9.1 –10.2 –9.3 –9.2 –5.4
Chad –8.5 –5.8 –7.8 –9.1 –8.9 –13.6 –9.2 –5.2 –4.3 –5.5 –4.0
Comoros –0.4 –6.0 –5.5 –7.0 –6.3 0.0 –7.4 –4.9 –6.9 –8.5 –7.1
Democratic Republic of the Congo –10.5 –5.2 –4.6 –5.0 –4.6 –3.7 –3.1 –0.5 0.3 –0.9 –2.1
Republic of Congo 7.3 14.0 17.7 13.8 1.4 –54.1 –74.1 –12.7 3.0 4.8 –9.2
Côte d’Ivoire 1.9 10.4 –1.2 –1.4 1.4 –0.6 –1.1 –1.2 –1.5 –1.3 –1.5
Equatorial Guinea –20.2 –5.7 –1.1 –2.5 –4.3 –17.7 –11.8 –0.5 –0.9 –13.3 0.1
Eritrea –6.1 3.2 2.7 3.6 4.0 –1.4 –2.1 –2.4 –1.5 –2.1 –2.6
Ethiopia –1.4 –2.5 –6.9 –5.9 –6.4 –10.2 –9.0 –8.1 –6.5 –6.3 –4.1
Gabon 14.9 24.0 17.9 7.3 7.6 –5.6 –10.1 –4.8 –1.5 –1.9 7.4
The Gambia –16.3 –11.7 –7.0 –10.3 –10.8 –15.0 –8.9 –14.3 –18.4 –16.9 –13.6
Ghana –8.6 –9.0 –11.7 –11.9 –9.5 –7.7 –6.7 –4.5 –4.1 –4.0 –4.1
Guinea –6.4 –18.4 –20.0 –12.5 –13.4 –15.4 –31.9 –23.0 –19.1 –10.0 –10.6
Guinea-Bissau –8.3 –1.3 –8.4 –4.6 0.5 2.3 1.3 0.1 –3.3 –2.6 –1.7
Kenya –5.9 –9.2 –8.4 –8.8 –10.4 –6.7 –5.2 –6.4 –6.2 –5.7 –5.2
Lesotho –8.9 –13.4 –8.4 –5.5 –4.9 –4.5 –7.4 –6.9 –12.2 –12.1 –13.8
Liberia –20.7 –17.6 –17.3 –21.6 –26.3 –26.5 –18.5 –22.4 –22.5 –22.4 –19.9
Madagascar –10.2 –7.0 –7.6 –5.9 –0.3 –1.9 0.6 –3.4 –4.0 –4.8 –4.8
Malawi –8.6 –8.6 –9.2 –8.4 –8.3 –9.4 –13.6 –10.0 –8.9 –8.1 –7.6
Mali –10.7 –5.1 –2.2 –2.9 –4.7 –5.3 –7.2 –6.2 –6.9 –6.4 –6.4
Mauritius –10.3 –13.8 –7.3 –6.3 –5.7 –4.9 –4.4 –6.0 –7.4 –8.7 –2.1
Mozambique –16.1 –25.3 –44.7 –42.9 –38.2 –40.3 –39.2 –16.1 –16.9 –44.6 –117.7
Namibia –3.5 –3.0 –5.7 –4.0 –10.8 –12.6 –14.1 –1.4 –3.6 –5.1 –8.3
Niger –19.8 –25.1 –16.1 –16.8 –15.4 –20.5 –15.5 –13.2 –16.1 –16.7 –12.0
Nigeria 3.6 2.6 3.8 3.7 0.2 –3.2 0.7 2.5 0.5 0.4 0.2
Rwanda –7.2 –7.4 –11.2 –8.7 –11.8 –13.3 –14.3 –6.8 –8.4 –9.2 –5.9
São Tomé and Príncipe –22.9 –27.7 –21.9 –13.8 –21.9 –12.6 –6.0 –13.0 –11.3 –9.9 –7.0
Senegal –4.4 –8.0 –10.9 –10.5 –9.0 –7.0 –5.5 –9.4 –7.9 –7.5 –6.4
Seychelles –19.4 –23.0 –21.1 –11.9 –23.1 –18.6 –18.3 –16.0 –14.4 –13.9 –12.1
Sierra Leone –22.7 –65.0 –31.8 –17.5 –18.2 –17.4 –19.4 –21.9 –18.9 –21.6 –19.7
South Africa –1.5 –2.2 –5.1 –5.9 –5.3 –4.4 –3.3 –2.3 –2.9 –3.1 –3.4
South Sudan ... 18.2 –15.9 –3.9 –1.6 –7.2 1.8 –6.0 –6.1 –4.3 –1.8
Swaziland –8.6 1.0 12.7 19.3 21.6 26.7 16.7 14.6 15.4 15.0 15.7
Tanzania –7.7 –10.8 –11.6 –10.6 –10.1 –8.4 –4.5 –3.8 –5.4 –6.0 –4.7
Togo –5.8 –7.8 –7.6 –13.2 –10.0 –11.0 –9.6 –8.2 –7.8 –6.4 –2.5
Uganda –8.0 –9.9 –6.8 –7.2 –7.8 –6.7 –3.4 –4.5 –6.9 –9.5 –3.1
Zambia 7.5 4.7 5.4 –0.6 2.1 –3.9 –4.5 –3.3 –2.6 –1.9 –1.3
Zimbabwe 6 –14.3 –20.1 –13.1 –16.6 –14.2 –9.5 –3.4 –2.6 –2.6 –2.4 –2.5
structure.
2Starting in 2014, data exclude Crimea and Sevastopol.
3See country-specific notes for Libya in the “Country Notes” section of the Statistical Appendix.
6The Zimbabwe dollar ceased circulating in early 2009. Data are based on IMF staff estimates of price and exchange rate developments in US dollars. IMF staff estimates of US dollar
values may differ from authorities’ estimates.

Table A13. Summary of Financial Account Balances

Projections
2010 2011 2012 2013 2014 2015 2016 2017 2018 2019
Advanced Economies
Financial Account Balance –168.1 –224.1 –130.0 230.6 342.0 322.2 438.4 424.4 335.8 231.8
Direct Investment, Net 340.7 358.9 111.9 155.4 213.5 35.7 –14.5 433.1 315.0 327.3
Portfolio Investment, Net –969.0 –1,111.5 –246.3 –539.9 30.6 176.1 374.4 –126.9 –149.1 –287.5
Financial Derivatives, Net –114.1 –6.4 –98.3 73.9 –12.4 –105.3 45.3 41.3 29.2 51.2
Other Investment, Net 261.5 191.7 –155.0 399.5 –11.7 4.0 –137.2 –70.6 68.0 64.3
Change in Reserves 352.9 349.8 273.2 153.1 134.9 226.8 179.6 170.4 92.9 93.7
United States
Financial Account Balance –446.4 –525.6 –448.9 –404.0 –326.8 –333.2 –377.7 –349.2 –606.4 –763.8
Direct Investment, Net 85.8 173.1 126.9 104.7 101.2 –195.0 –167.8 75.7 –102.7 –95.5
Portfolio Investment, Net –620.8 –226.3 –498.3 –30.7 –120.8 –53.6 –196.7 –247.5 –481.7 –608.1
Financial Derivatives, Net –14.1 –35.0 7.1 2.2 –54.3 –25.2 15.8 26.4 –12.7 9.3
Other Investment, Net 100.9 –453.4 –89.0 –477.1 –249.4 –53.0 –31.0 –202.1 –9.3 –69.5
Change in Reserves 1.8 15.9 4.5 –3.1 –3.6 –6.3 2.1 –1.7 0.0 0.0
Euro Area
Financial Account Balance –16.9 –40.9 184.3 443.7 351.9 314.9 343.3 466.6 ... ...
Direct Investment, Net 82.3 124.9 59.4 23.8 95.6 268.2 246.6 –0.5 ... ...
Portfolio Investment, Net –81.4 –383.3 –175.8 –156.9 35.0 107.0 461.0 432.6 ... ...
Financial Derivatives, Net –4.4 5.5 38.9 42.1 65.5 96.8 20.5 10.5 ... ...
Other Investment, Net –27.1 197.7 242.9 528.5 150.0 –168.7 –402.1 25.5 ... ...
Change in Reserves 13.7 14.3 19.0 6.2 5.8 11.7 17.4 –1.4 ... ...
Germany
Financial Account Balance 123.7 167.7 194.3 300.0 316.3 259.6 269.1 311.4 346.8 364.1
Direct Investment, Net 60.6 10.3 33.6 26.0 96.6 59.9 23.8 47.7 52.7 47.5
Portfolio Investment, Net 154.1 –51.4 66.8 209.6 175.0 217.9 230.4 226.1 278.6 288.7
Financial Derivatives, Net 17.6 39.8 30.9 31.8 42.3 29.2 36.1 10.1 31.1 30.2
Other Investment, Net –110.7 165.1 61.1 31.4 5.6 –45.0 –23.1 28.9 –15.6 –2.3
Change in Reserves 2.1 3.9 1.7 1.2 –3.3 –2.4 1.9 –1.4 0.0 0.0
France
Direct Investment, Net 34.3 19.8 19.4 –13.9 47.2 –2.6 29.0 33.7 41.6 46.6
Portfolio Investment, Net –155.0 –335.1 –50.6 –79.3 –23.8 51.5 –4.0 –19.3 –32.5 –33.2
Financial Derivatives, Net –4.1 –19.4 –18.4 –22.3 –31.8 12.0 0.7 7.7 16.4 25.1
Other Investment, Net 115.5 263.8 –3.6 98.2 –2.9 –82.4 –60.0 –61.2 –67.2 –68.0
Change in Reserves 7.7 –7.7 5.2 –1.9 1.0 8.0 2.5 2.3 2.6 2.8
Italy
Financial Account Balance –107.1 –79.9 –4.1 29.0 68.5 40.0 73.7 53.4 58.3 53.0
Direct Investment, Net 21.3 17.2 6.8 0.9 3.1 2.7 –3.3 –9.0 0.8 1.0
Portfolio Investment, Net 62.5 25.6 –22.4 –5.4 5.5 109.1 176.6 111.4 79.9 49.3
Financial Derivatives, Net 6.6 –10.1 7.5 4.0 –4.8 2.6 –3.3 –6.6 –2.8 –0.7
Other Investment, Net –198.9 –113.9 2.1 27.5 65.9 –75.0 –95.1 –45.5 –19.5 3.3
Change in Reserves 1.4 1.3 1.9 2.0 –1.3 0.6 –1.3 3.0 0.0 0.0

Table A13. Summary of Financial Account Balances (continued)

Projections
2010 2011 2012 2013 2014 2015 2016 2017 2018 2019
Spain
Financial Account Balance –58.9 –43.4 0.5 41.6 14.8 23.1 27.9 24.9 27.3 30.0
Direct Investment, Net –1.9 12.8 –27.2 –24.6 8.6 31.0 18.5 24.1 26.7 28.1
Portfolio Investment, Net –46.6 43.1 53.7 –83.6 –12.1 10.2 55.1 12.5 –3.2 –2.1
Financial Derivatives, Net –11.4 2.9 –10.7 1.4 1.7 –1.1 –2.9 0.0 0.0 0.0
Other Investment, Net 0.0 –116.2 –18.2 147.8 11.5 –22.6 –51.8 –11.7 3.7 4.0
Japan
Financial Account Balance 247.3 158.4 53.9 –4.3 58.9 178.6 266.7 153.1 191.1 195.6
Direct Investment, Net 72.5 117.8 117.5 144.7 118.6 131.0 134.6 146.2 134.0 147.0
Portfolio Investment, Net 147.9 –162.9 28.8 –280.6 –42.2 131.5 282.2 –54.4 –48.9 –46.2
Financial Derivatives, Net –11.9 –17.1 6.7 58.1 34.0 17.7 –16.7 30.4 32.2 33.4
Other Investment, Net –5.5 43.4 –61.1 34.8 –60.1 –106.7 –127.7 7.4 63.3 50.4
Change in Reserves 44.3 177.3 –37.9 38.7 8.5 5.1 –5.7 23.6 10.5 11.0
United Kingdom
Direct Investment, Net –10.1 53.4 –34.8 –11.2 –176.1 –116.2 –219.5 84.6 67.5 57.4
Portfolio Investment, Net –201.0 –215.5 275.0 –284.3 22.6 –212.3 –193.2 –87.7 0.0 0.0
Financial Derivatives, Net –69.3 7.4 –65.8 63.4 31.2 –128.6 29.3 12.5 1.9 –9.1
Other Investment, Net 158.6 98.6 –279.1 91.8 –37.5 286.1 231.1 –105.6 –194.8 –167.9
Canada
Direct Investment, Net 6.3 12.5 12.8 –12.0 1.3 22.2 36.3 10.2 21.7 22.5
Financial Derivatives, Net ... ... ... ... ... ... ... ... ... ...
Other Investment, Net 41.4 34.3 –13.4 –22.5 –15.9 –43.8 25.8 6.2 4.1 14.9
Other Advanced Economies 1
Direct Investment, Net 93.5 –6.5 –34.8 26.3 –7.5 –109.9 –76.9 –99.0 –78.5 –72.8
Portfolio Investment, Net –57.1 46.8 148.7 138.4 180.1 334.0 273.2 276.7 321.3 341.5
Financial Derivatives, Net –15.2 31.1 –28.3 –33.5 –23.5 –14.2 0.7 –27.9 –35.0 –37.0
Other Investment, Net –16.8 88.5 –104.0 143.1 97.1 –83.4 –10.9 43.7 119.1 104.1
Economies
Financial Account Balance 137.1 237.3 113.2 38.7 15.8 –276.2 –431.5 –254.9 4.8 –56.4
Direct Investment, Net –455.9 –530.9 –491.4 –482.9 –415.4 –340.5 –279.9 –375.3 –388.5 –396.4
Other Investment, Net –22.9 166.5 408.8 89.5 409.5 461.8 386.2 135.7 343.6 325.1
Change in Reserves 837.2 745.0 432.0 590.8 128.4 –515.5 –482.4 164.6 143.6 102.6


Projections
2010 2011 2012 2013 2014 2015 2016 2017 2018 2019
Regional Groups
Commonwealth of Independent States2
Direct Investment, Net –8.5 –15.2 –27.6 –3.6 19.2 0.9 –34.6 –5.0 –0.3 –1.8
Portfolio Investment, Net –14.2 17.9 3.5 –0.2 28.8 12.0 –2.4 –15.8 –4.8 –4.2
Other Investment, Net 36.0 64.3 44.3 27.5 73.0 39.4 28.8 21.9 22.1 27.2
Change in Reserves 60.9 32.0 30.5 –21.5 –114.1 –6.8 9.8 61.7 48.4 35.0
Financial Account Balance 148.4 65.2 9.7 33.0 149.5 86.9 –40.0 –85.0 131.1 137.8
Portfolio Investment, Net –91.3 –58.0 –115.6 –64.7 –123.9 82.7 31.9 –50.5 –21.9 –23.9
Financial Derivatives, Net 0.2 –0.3 1.5 –2.0 0.4 –1.5 –10.2 –11.5 –10.0 –10.6
Other Investment, Net –97.5 –28.7 207.4 –78.7 280.8 461.5 346.2 –58.9 244.2 244.9
Emerging and Developing Europe
Financial Derivatives, Net 0.0 1.6 –2.9 –1.4 0.3 –1.8 0.1 –0.9 0.8 0.8
Other Investment, Net –52.5 –30.1 6.5 –12.9 8.6 12.7 –2.8 18.7 –3.9 –10.4
Change in Reserves 35.9 14.6 27.8 18.5 –0.2 –10.4 23.6 –12.7 6.9 11.1
Financial Derivatives, Net 0.7 5.5 2.3 1.7 4.4 1.4 –1.1 4.4 0.6 0.7
Other Investment, Net –0.2 14.1 29.5 48.2 1.8 30.0 62.8 59.2 44.6 42.8
Change in Reserves 90.9 108.0 59.0 11.7 39.0 –29.0 20.8 29.0 19.1 8.3
Afghanistan, and Pakistan
Financial Account Balance 122.5 320.3 285.8 306.7 180.3 –133.1 –208.8 –55.4 5.4 –20.2
Portfolio Investment, Net 25.2 74.2 56.8 72.1 131.9 69.5 –6.9 –18.0 11.1 18.6
Other Investment, Net 60.1 129.5 100.5 114.9 61.6 –63.3 –48.3 101.1 39.7 18.3
Change in Reserves 86.1 138.2 154.7 128.1 16.2 –139.3 –146.4 –121.9 –17.9 –28.0
Sub-Saharan Africa
Financial Derivatives, Net –0.2 –1.7 –1.7 –0.8 –1.5 –0.4 0.9 1.1 0.0 0.0
Other Investment, Net 31.2 17.5 20.5 –9.4 –16.2 –18.4 –0.5 –6.3 –3.1 2.3
Change in Reserves 0.5 20.7 21.0 2.9 –7.9 –14.1 –10.4 14.7 14.1 3.2


Projections
2010 2011 2012 2013 2014 2015 2016 2017 2018 2019
Analytical Groups
Fuel
Financial Account Balance 251.8 512.2 445.8 375.9 223.6 –85.7 –165.2 53.7 128.8 101.4
Direct Investment, Net –31.2 –23.8 –33.3 10.8 7.0 7.4 –29.5 –7.3 –12.8 –13.9
Portfolio Investment, Net 20.2 87.0 47.2 78.2 162.8 79.9 –7.2 –25.1 20.6 21.6
Other Investment, Net 145.8 253.0 196.0 186.4 155.1 9.2 33.7 147.7 90.3 84.0
Change in Reserves 115.1 194.7 234.5 100.2 –106.5 –189.5 –162.6 –61.9 30.5 9.6
Nonfuel
Financial Derivatives, Net 0.7 5.8 –1.0 –2.4 3.6 –2.2 –10.3 –7.0 –8.6 –9.2
Other Investment, Net –168.3 –86.6 212.8 –96.8 254.4 452.7 352.5 –12.0 253.3 241.1
Net Debtor Economies
Other Investment, Net –83.9 –65.8 –58.1 –21.1 –11.4 32.7 31.9 65.5 29.0 15.3
Rescheduling during 2012–16
Portfolio Investment, Net –11.2 0.9 –1.5 –10.2 –0.4 1.8 –1.4 –21.1 –16.1 –6.5
Other Investment, Net 7.3 6.4 –3.2 –11.8 0.7 –22.2 –23.1 10.2 5.7 4.6
Change in Reserves 14.2 –11.8 –24.7 3.2 –8.3 6.1 –2.5 0.0 8.7 –0.1
Memorandum
World
Financial Account Balance –31.0 13.2 –16.8 269.3 357.8 46.0 6.8 169.4 340.6 175.4
Note: The estimates in this table are based on individual countries’ national accounts and balance of payments statistics. Country group composites are calculated as the sum of the US dollar
values for the relevant individual countries. Some group aggregates for the financial derivatives are not shown because of incomplete data. Projections for the euro area are not available because
of data constraints.
structure.

Table A14. Summary of Net Lending and Borrowing

(Percent of GDP)
Projections
Averages Average
2000–09 2004–11 2012 2013 2014 2015 2016 2017 2018 2019 2020–23
Advanced Economies
Net Lending and Borrowing –0.8 –0.7 0.1 0.5 0.5 0.5 0.7 0.8 0.7 0.5 0.6
Current Account Balance –0.9 –0.7 0.0 0.4 0.5 0.7 0.7 0.8 0.7 0.5 0.6
Savings 21.9 21.4 21.3 21.5 22.1 22.4 21.9 22.0 22.2 22.3 22.6
Investment 22.6 22.0 21.0 20.9 21.2 21.2 21.0 21.2 21.5 21.8 22.1
Capital Account Balance 0.0 0.0 0.0 0.0 0.0 –0.1 0.0 0.0 0.0 0.0 0.0
United States
Net Lending and Borrowing –4.5 –4.3 –2.6 –2.1 –2.1 –2.4 –2.4 –2.3 –3.0 –3.3 –3.3
Current Account Balance –4.5 –4.3 –2.6 –2.1 –2.1 –2.4 –2.4 –2.4 –3.0 –3.4 –3.3
Savings 17.7 16.5 17.7 18.3 19.3 19.4 18.0 17.5 17.2 17.4 17.8
Investment 21.9 20.8 19.4 19.8 20.1 20.4 19.7 19.8 20.2 20.8 21.1
Capital Account Balance 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.0 0.0 0.1
Euro Area
Net Lending and Borrowing –0.1 0.0 1.5 2.4 2.6 3.1 3.4 3.3 ... ... ...
Current Account Balance –0.2 –0.1 1.4 2.2 2.4 3.2 3.4 3.5 3.2 3.2 3.0
Savings 22.8 22.7 22.3 22.4 22.9 23.8 24.0 24.4 24.6 24.8 24.9
Investment 22.5 22.2 20.1 19.6 19.9 20.0 20.4 20.9 21.1 21.3 21.6
Capital Account Balance 0.1 0.1 0.1 0.2 0.1 –0.1 0.0 –0.2 ... ... ...
Germany
Net Lending and Borrowing 3.4 5.5 7.0 6.7 7.5 8.9 8.6 8.0 8.2 8.2 7.9
Current Account Balance 3.4 5.6 7.0 6.7 7.5 8.9 8.5 8.0 8.2 8.2 7.9
Savings 23.7 25.3 26.3 26.2 27.0 28.1 27.7 27.7 28.0 28.2 28.2
Investment 20.3 19.8 19.3 19.5 19.5 19.1 19.2 19.7 19.7 20.0 20.3
France
Net Lending and Borrowing 1.0 –0.1 –1.2 –0.8 –1.2 –0.4 –0.9 –1.4 –1.3 –0.9 –0.3
Current Account Balance 1.0 –0.2 –1.2 –0.9 –1.3 –0.4 –0.9 –1.4 –1.3 –0.9 –0.3
Savings 23.3 22.6 21.4 21.4 21.5 22.3 22.1 22.0 21.9 22.2 22.6
Investment 22.4 22.7 22.6 22.3 22.7 22.8 23.0 23.4 23.2 23.1 22.9
Italy
Net Lending and Borrowing –0.8 –1.8 –0.1 0.9 2.1 1.8 2.6 2.9 2.7 2.3 1.5
Current Account Balance –0.9 –1.9 –0.3 1.0 1.9 1.5 2.7 2.9 2.6 2.2 1.4
Savings 20.2 19.2 17.5 17.9 18.9 18.9 19.8 20.4 20.2 20.1 19.8
Investment 21.1 21.1 17.9 17.0 17.0 17.3 17.1 17.5 17.7 17.9 18.4
Capital Account Balance 0.1 0.1 0.2 0.0 0.2 0.2 –0.2 0.0 0.1 0.1 0.1
Spain
Net Lending and Borrowing –5.5 –6.0 0.3 2.2 1.6 1.8 2.2 1.9 1.8 1.9 2.0
Current Account Balance –6.2 –6.5 –0.2 1.5 1.1 1.1 1.9 1.7 1.6 1.7 1.7
Savings 22.2 21.1 19.8 20.2 20.5 21.5 22.4 22.8 23.1 23.3 23.5
Investment 28.3 27.6 20.0 18.7 19.5 20.4 20.5 21.1 21.5 21.7 21.8
Japan
Savings 27.9 26.9 23.6 24.1 24.7 27.1 27.4 28.0 28.3 28.4 28.7
Investment 24.7 23.5 22.7 23.2 23.9 24.0 23.6 24.0 24.6 24.7 24.7
Capital Account Balance –0.1 –0.1 0.0 –0.1 0.0 –0.1 –0.1 –0.1 –0.1 –0.1 –0.1
United Kingdom
Savings 14.4 13.3 11.5 10.5 11.8 11.8 11.1 12.8 13.3 13.7 14.4
Investment 17.3 16.6 15.7 16.1 17.1 17.0 16.9 16.9 17.0 17.1 17.5
Capital Account Balance 0.0 0.0 0.0 –0.1 –0.1 –0.1 –0.1 –0.1 –0.1 –0.1 –0.1

Table A14. Summary of Net Lending and Borrowing (continued)

(Percent of GDP)
Projections
Averages Average
2000–09 2004–11 2012 2013 2014 2015 2016 2017 2018 2019 2020–23
Canada
Net Lending and Borrowing 1.1 –0.4 –3.6 –3.2 –2.4 –3.6 –3.2 –3.0 –3.2 –2.5 –1.8
Current Account Balance 1.1 –0.4 –3.6 –3.2 –2.4 –3.6 –3.2 –3.0 –3.2 –2.5 –1.8
Savings 23.0 22.8 21.3 21.7 22.5 20.5 20.0 20.7 20.8 21.5 22.5
Investment 21.9 23.2 24.9 24.9 24.9 24.1 23.2 23.7 23.9 24.1 24.4
Other Advanced Economies1
Savings 29.8 30.4 30.3 30.4 30.5 30.9 30.2 30.2 30.5 30.5 30.2
Investment 25.6 25.9 26.0 25.1 25.2 24.8 24.7 25.3 25.4 25.4 25.5
Capital Account Balance –0.1 0.0 0.0 0.1 –0.1 –0.4 0.1 –0.2 0.0 –0.1 –0.1
Economies
Net Lending and Borrowing 2.6 2.9 1.3 0.7 0.6 0.0 –0.2 0.0 0.1 –0.1 –0.4
Current Account Balance 2.5 2.8 1.3 0.6 0.5 –0.2 –0.3 –0.1 –0.1 –0.2 –0.5
Savings 29.5 32.2 33.5 32.8 33.0 32.6 31.8 32.2 32.8 32.9 32.5
Investment 27.3 29.6 32.4 32.4 32.6 32.9 32.1 32.3 32.9 33.0 33.0
Regional Groups
Commonwealth of Independent States2
Savings 27.4 27.1 27.2 24.3 25.1 26.1 24.7 25.9 26.8 26.9 26.4
Investment 21.1 22.0 24.8 23.6 22.9 23.0 24.3 24.4 23.8 24.4 24.3
Capital Account Balance –0.5 –0.3 –0.2 0.0 –1.5 0.0 0.0 0.0 0.0 0.0 0.0
Savings 38.4 42.4 43.7 43.0 43.6 42.4 41.0 40.9 40.9 40.6 39.5
Investment 35.2 38.8 42.6 42.3 42.0 40.4 39.6 40.0 40.2 40.0 39.2
Emerging and Developing Europe
Savings 19.7 19.9 20.5 21.5 22.1 22.9 22.5 23.2 23.1 23.3 23.6
Investment 24.2 25.7 24.9 25.0 25.0 24.7 24.1 25.5 25.9 25.9 26.0
Savings 20.1 21.2 20.0 19.2 17.9 18.1 17.2 17.5 17.3 17.5 18.2
Investment 20.4 21.3 22.5 22.4 21.7 22.0 19.4 19.1 19.4 19.8 20.6
and Pakistan
Net Lending and Borrowing 8.0 9.6 11.9 10.0 6.1 –3.6 –4.1 –0.8 0.8 0.0 –1.1
Current Account Balance 8.3 10.1 12.4 9.8 5.4 –4.0 –4.2 –0.9 0.5 –0.3 –1.3
Savings 34.7 37.4 37.9 36.1 33.0 24.9 24.1 26.4 28.6 27.8 26.4
Investment 27.0 28.0 25.9 25.9 26.8 28.4 27.6 27.0 27.4 27.2 26.8
Sub-Saharan Africa
Net Lending and Borrowing 2.0 2.4 –0.6 –1.8 –3.4 –5.6 –3.6 –2.2 –2.6 –2.7 –2.9
Current Account Balance 0.7 0.9 –1.7 –2.2 –3.8 –6.0 –4.1 –2.6 –2.9 –3.1 –3.2
Savings 20.3 21.6 19.5 18.7 18.3 16.1 16.1 17.8 17.6 18.1 18.1
Investment 19.7 20.6 21.0 21.0 21.9 21.8 19.8 20.2 20.5 21.1 21.2

Table A14. Summary of Net Lending and Borrowing (continued)

(Percent of GDP)
Projections
Averages Average
2000–09 2004–11 2012 2013 2014 2015 2016 2017 2018 2019 2020–23
Analytical Groups
Fuel
Net Lending and Borrowing 9.3 10.1 9.3 7.4 4.7 –1.5 –1.8 1.2 2.9 2.2 1.2
Current Account Balance 9.6 10.4 9.7 7.3 5.1 –1.6 –1.9 1.3 2.7 2.0 1.1
Savings 33.9 35.0 34.7 32.0 30.2 26.5 24.8 27.1 28.7 28.3 26.6
Investment 24.7 25.0 25.4 24.9 25.4 28.3 26.2 25.5 25.4 25.5 24.8
Capital Account Balance 0.0 0.0 –0.1 0.0 –0.7 –0.1 –0.1 0.0 0.0 0.0 0.0
Nonfuel
Net Lending and Borrowing 0.8 0.8 –0.9 –1.0 –0.4 0.2 0.1 –0.2 –0.4 –0.5 –0.7
Current Account Balance 0.6 0.6 –1.1 –1.2 –0.6 0.1 0.0 –0.3 –0.6 –0.6 –0.8
Savings 28.4 31.4 33.2 33.1 33.7 33.8 33.1 33.2 33.5 33.6 33.4
Investment 28.0 30.9 34.2 34.2 34.3 33.8 33.2 33.5 34.1 34.2 34.2
Net Debtor Economies
Savings 22.1 23.4 23.0 22.5 22.5 22.1 22.3 22.9 23.0 23.3 24.1
Investment 23.6 25.2 26.2 25.3 25.1 24.5 23.9 24.5 25.0 25.4 26.2
Rescheduling during 2012–16
Savings 20.9 20.7 14.7 13.4 14.3 11.8 12.1 13.8 15.9 16.5 17.4
Investment 22.2 22.7 20.6 19.2 18.3 17.8 18.7 19.1 20.5 21.1 22.7
Memorandum
World
Savings 23.9 24.6 26.0 26.0 26.4 26.4 25.8 26.1 26.4 26.6 26.9
Investment 23.8 24.4 25.3 25.3 25.6 25.8 25.2 25.6 26.0 26.4 26.7
Note: The estimates in this table are based on individual countries’ national accounts and balance of payments statistics. Country group composites are calculated as the sum of the US
dollar values for the relevant individual countries. This differs from the calculations in the April 2005 and earlier issues of the World Economic Outlook, in which the composites were
weighted by GDP valued at purchasing power parities as a share of total world GDP. The estimates of gross national savings and investment (or gross capital formation) are from individual
countries’ national accounts statistics. The estimates of the current account balance, the capital account balance, and the financial account balance (or net lending/net borrowing) are
from the balance of payments statistics. The link between domestic transactions and transactions with the rest of the world can be expressed as accounting identities. Savings (S ) minus
investment (I ) is equal to the current account balance (CAB ) (S − I = CAB ). Also, net lending/net borrowing (NLB ) is the sum of the current account balance and the capital account
balance (KAB ) (NLB = CAB + KAB ). In practice, these identities do not hold exactly; imbalances result from imperfections in source data and compilation as well as from asymmetries in
group composition due to data availability.
structure.

Table A15. Summary of World Medium-Term Baseline Scenario

Projections
Averages Averages
2000–09 2010–19 2016 2017 2018 2019 2016–19 2020–23
Annual Percent Change
World Real GDP 3.9 3.9 3.2 3.8 3.9 3.9 3.7 3.7
Memorandum
Potential Output
Major Advanced Economies 1.9 1.3 1.4 1.5 1.6 1.7 1.6 1.5
World Trade, Volume1 5.0 4.9 2.3 4.9 5.1 4.7 4.2 3.9
Imports
Exports
Terms of Trade
Advanced Economies –0.2 0.1 0.9 –0.2 0.6 0.2 0.4 0.0
Emerging Market and Developing Economies 1.5 0.1 –1.4 0.6 1.1 –0.5 0.0 0.0
World Prices in US Dollars
Manufactures 1.7 0.3 –5.2 1.4 1.9 1.3 –0.2 0.8
Oil 13.1 –0.6 –15.7 23.3 18.0 –6.5 3.5 –2.1
Nonfuel Primary Commodities 5.3 1.6 –1.5 6.8 5.6 0.5 2.8 –0.5
Consumer Prices
Interest Rates Percent
Real Six-Month LIBOR2 1.1 –0.7 –0.2 –0.3 –0.2 0.9 0.0 1.5
World Real Long-Term Interest Rate3 2.1 0.5 0.4 –0.2 0.0 0.5 0.2 0.9
Current Account Balances Percent of GDP
Advanced Economies –0.9 0.4 0.7 0.8 0.7 0.5 0.7 0.6
Emerging Market and Developing Economies 2.5 0.4 –0.3 –0.1 –0.1 –0.2 –0.2 –0.5
Total External Debt
Debt Service
1Data refer to trade in goods and services.
2London interbank offered rate on US dollar deposits minus percent change in US GDP deflator.
3GDP-weighted average of 10-year (or nearest-maturity) government bond rates for Canada, France, Germany, Italy, Japan, the United Kingdom, and the
United States.

1
CHAPTER
WORLD ECONOMIC OUTLOOK

SELECTED TOPICS
World Economic Outlook Archives

World Economic Outlook: Financial Systems and Economic Cycles September 2006
World Economic Outlook: Spillovers and Cycles in the Global Economy April 2007
World Economic Outlook: Globalization and Inequality October 2007
World Economic Outlook: Housing and the Business Cycle April 2008
World Economic Outlook: Financial Stress, Downturns, and Recoveries October 2008
World Economic Outlook: Crisis and Recovery April 2009
World Economic Outlook: Sustaining the Recovery October 2009
World Economic Outlook: Rebalancing Growth April 2010
World Economic Outlook: Recovery, Risk, and Rebalancing October 2010
World Economic Outlook: Tensions from the Two-Speed Recovery—Unemployment, Commodities,
and Capital Flows April 2011
World Economic Outlook: Slowing Growth, Rising Risks September 2011
World Economic Outlook: Growth Resuming, Dangers Remain April 2012
World Economic Outlook: Coping with High Debt and Sluggish Growth October 2012
World Economic Outlook: Hopes, Realities, Risks April 2013
World Economic Outlook: Transitions and Tensions October 2013
World Economic Outlook: Recovery Strengthens, Remains Uneven April 2014
World Economic Outlook: Legacies, Clouds, Uncertainties October 2014
World Economic Outlook: Uneven Growth—Short- and Long-Term Factors April 2015
World Economic Outlook: Adjusting to Lower Commodity Prices October 2015
World Economic Outlook: Too Slow for Too Long April 2016
World Economic Outlook: Subdued Demand—Symptoms and Remedies October 2016
World Economic Outlook: Gaining Momentum? April 2017
World Economic Outlook: Seeking Sustainable Growth: Short-term Recovery, Long-term Challenges October 2017
World Economic Outlook: Cyclical Upswing, Structural Change April 2018
I. Methodology—Aggregation, Modeling, and Forecasting

Measuring Inequality: Conceptual, Methodological, and Measurement Issues October 2007, Box 4.1
New Business Cycle Indices for Latin America: A Historical Reconstruction October 2007, Box 5.3
Implications of New PPP Estimates for Measuring Global Growth April 2008, Appendix 1.1
Measuring Output Gaps October 2008, Box 1.3
Assessing and Communicating Risks to the Global Outlook October 2008, Appendix 1.1
Fan Chart for Global Growth April 2009, Appendix 1.2
Indicators for Tracking Growth October 2010, Appendix 1.2
Inferring Potential Output from Noisy Data: The Global Projection Model View October 2010, Box 1.3
Uncoordinated Rebalancing October 2010, Box 1.4
World Economic Outlook Downside Scenarios April 2011, Box 1.2
Fiscal Balance Sheets: The Significance of Nonfinancial Assets and Their Measurement October 2014, Box 3.3

Tariff Scenarios October 2016, Scenario box

World Growth Projections over the Medium Term October 2016, Box 1.1
II. Historical Surveys

Historical Perspective on Growth and the Current Account October 2008, Box 6.3
A Historical Perspective on International Financial Crises October 2009, Box 4.1
The Good, the Bad, and the Ugly: 100 Years of Dealing with Public Debt Overhangs October 2012, Chapter 3
What Is the Effect of Recessions? October 2015, Box 1.1
III. Economic Growth—Sources and Patterns

Asia Rising: Patterns of Economic Development and Growth September 2006, Chapter 3
Japan’s Potential Output and Productivity Growth September 2006, Box 3.1
The Evolution and Impact of Corporate Governance Quality in Asia September 2006, Box 3.2
Decoupling the Train? Spillovers and Cycles in the Global Economy April 2007, Chapter 4
Spillovers and International Business Cycle Synchronization: A Broader Perspective April 2007, Box 4.3
The Discounting Debate October 2007, Box 1.7
Taxes versus Quantities under Uncertainty (Weitzman, 1974) October 2007, Box 1.8
Experience with Emissions Trading in the European Union October 2007, Box 1.9
Climate Change: Economic Impact and Policy Responses October 2007, Appendix 1.2
What Risks Do Housing Markets Pose for Global Growth? October 2007, Box 2.1
The Changing Dynamics of the Global Business Cycle October 2007, Chapter 5
Major Economies and Fluctuations in Global Growth October 2007, Box 5.1
Improved Macroeconomic Performance—Good Luck or Good Policies? October 2007, Box 5.2
House Prices: Corrections and Consequences October 2008, Box 1.2
Global Business Cycles April 2009, Box 1.1
How Similar Is the Current Crisis to the Great Depression? April 2009, Box 3.1
Is Credit a Vital Ingredient for Recovery? Evidence from Industry-Level Data April 2009, Box 3.2
From Recession to Recovery: How Soon and How Strong? April 2009, Chapter 3
What’s the Damage? Medium-Term Output Dynamics after Financial Crises October 2009, Chapter 4
Will the Recovery Be Jobless? October 2009, Box 1.3
Unemployment Dynamics during Recessions and Recoveries: Okun’s Law and Beyond April 2010, Chapter 3
Does Slow Growth in Advanced Economies Necessarily Imply Slow Growth in Emerging Economies? October 2010, Box 1.1
The Global Recovery: Where Do We Stand? April 2012, Box 1.2
How Does Uncertainty Affect Economic Performance? October 2012, Box 1.3
Resilience in Emerging Market and Developing Economies: Will It Last? October 2012, Chapter 4
Jobs and Growth: Can’t Have One without the Other? October 2012, Box 4.1
Spillovers from Policy Uncertainty in the United States and Europe April 2013, Chapter 2,
Spillover Feature
Breaking through the Frontier: Can Today’s Dynamic Low-Income Countries Make It? April 2013, Chapter 4
What Explains the Slowdown in the BRICS? October 2013, Box 1.2
Dancing Together? Spillovers, Common Shocks, and the Role of Financial and Trade Linkages October 2013, Chapter 3
Output Synchronicity in the Middle East, North Africa, Afghanistan, and Pakistan and in the
Caucasus and Central Asia October 2013, Box 3.1
Spillovers from Changes in U.S. Monetary Policy October 2013, Box 3.2
Saving and Economic Growth April 2014, Box 3.1

SELECTED TOPICS
On the Receiving End? External Conditions and Emerging Market Growth before, during,
and after the Global Financial Crisis April 2014, Chapter 4
The Impact of External Conditions on Medium-Term Growth in Emerging Market Economies April 2014, Box 4.1
The Origins of IMF Growth Forecast Revisions since 2011 October 2014, Box 1.2
Underlying Drivers of U.S. Yields Matter for Spillovers October 2014, Chapter 2,
Spillover Feature
Is It Time for an Infrastructure Push? The Macroeconomic Effects of Public Investment October 2014, Chapter 3
The Macroeconomic Effects of Scaling Up Public Investment in Developing Economies October 2014, Box 3.4
Where Are We Headed? Perspectives on Potential Output April 2015, Chapter 3
Steady As She Goes—Estimating Sustainable Output April 2015, Box 3.1
Macroeconomic Developments and Outlook in Low-Income Developing Countries—
The Role of External Factors April 2016, Box 1.2
Time for a Supply-Side Boost? Macroeconomic Effects of Labor and Product Market
Reforms in Advanced Economies April 2016, Chapter 3
Road Less Traveled: Growth in Emerging Market and Developing Economies in a Complicated
External Environment April 2017, Chapter 3
Growing with Flows: Evidence from Industry-Level Data April 2017, Box 2.2
Emerging Market and Developing Economy Growth: Heterogeneity and Income Convergence
Over the Forecast Horizon October 2017, Box 1.3
Manufacturing Jobs: Implications for Productivity and Inequality April 2018, Chapter 3
Is Productivity Growth Shared in a Globalized Economy? April 2018, Chapter 4
Recent Dynamics of Potential Growth April 2018, Box 1.3
IV. Inflation and Deflation and Commodity Markets

The Boom in Nonfuel Commodity Prices: Can It Last? September 2006, Chapter 5
International Oil Companies and National Oil Companies in a Changing Oil Sector Environment September 2006, Box 1.4
Commodity Price Shocks, Growth, and Financing in Sub-Saharan Africa September 2006, Box 2.2
Has Speculation Contributed to Higher Commodity Prices? September 2006, Box 5.1
Agricultural Trade Liberalization and Commodity Prices September 2006, Box 5.2
Recent Developments in Commodity Markets September 2006,
Appendix 2.1
Who Is Harmed by the Surge in Food Prices? October 2007, Box 1.1
Refinery Bottlenecks October 2007, Box 1.5
Making the Most of Biofuels October 2007, Box 1.6
Commodity Market Developments and Prospects April 2008, Appendix 1.2
Dollar Depreciation and Commodity Prices April 2008, Box 1.4
Why Hasn’t Oil Supply Responded to Higher Prices? April 2008, Box 1.5
Oil Price Benchmarks April 2008, Box 1.6
Globalization, Commodity Prices, and Developing Countries April 2008, Chapter 5
The Current Commodity Price Boom in Perspective April 2008, Box 5.2
Is Inflation Back? Commodity Prices and Inflation October 2008, Chapter 3
Does Financial Investment Affect Commodity Price Behavior? October 2008, Box 3.1
Fiscal Responses to Recent Commodity Price Increases: An Assessment October 2008, Box 3.2
Monetary Policy Regimes and Commodity Prices October 2008, Box 3.3
Assessing Deflation Risks in the G3 Economies April 2009, Box 1.3
Will Commodity Prices Rise Again When the Global Economy Recovers? April 2009, Box 1.5

Commodity Market Developments and Prospects October 2009, Appendix 1.1

What Do Options Markets Tell Us about Commodity Price Prospects? October 2009, Box 1.6
What Explains the Rise in Food Price Volatility? October 2009, Box 1.7
How Unusual Is the Current Commodity Price Recovery? April 2010, Box 1.2
Commodity Futures Price Curves and Cyclical Market Adjustment April 2010, Box 1.3
Commodity Market Developments and Prospects October 2010, Appendix 1.1
Dismal Prospects for the Real Estate Sector October 2010, Box 1.2
Have Metals Become More Scarce and What Does Scarcity Mean for Prices? October 2010, Box 1.5
Oil Scarcity, Growth, and Global Imbalances April 2011, Chapter 3
Life Cycle Constraints on Global Oil Production April 2011, Box 3.1
Unconventional Natural Gas: A Game Changer? April 2011, Box 3.2
Short-Term Effects of Oil Shocks on Economic Activity April 2011, Box 3.3
Low-Frequency Filtering for Extracting Business Cycle Trends April 2011, Appendix 3.1
The Energy and Oil Empirical Models April 2011, Appendix 3.2
Commodity Market Developments and Prospects September 2011, Appendix 1.1
Financial Investment, Speculation, and Commodity Prices September 2011, Box 1.4
Target What You Can Hit: Commodity Price Swings and Monetary Policy September 2011, Chapter 3
Commodity Market Review April 2012, Chapter 1,
Special Feature
Commodity Price Swings and Commodity Exporters April 2012, Chapter 4
Macroeconomic Effects of Commodity Price Shocks on Low-Income Countries April 2012, Box 4.1
Volatile Commodity Prices and the Development Challenge in Low-Income Countries April 2012, Box 4.2
Commodity Market Review October 2012, Chapter 1,
Special Feature
Unconventional Energy in the United States October 2012, Box 1.4
Food Supply Crunch: Who Is Most Vulnerable? October 2012, Box 1.5
Commodity Market Review April 2013, Chapter 1,
Special Feature
The Dog That Didn’t Bark: Has Inflation Been Muzzled or Was It Just Sleeping? April 2013, Chapter 3
Does Inflation Targeting Still Make Sense with a Flatter Phillips Curve? April 2013, Box 3.1
Commodity Market Review October 2013, Chapter 1,
Special Feature
Energy Booms and the Current Account: Cross-Country Experience October 2013, Box 1.SF.1
Oil Price Drivers and the Narrowing WTI-Brent Spread October 2013, Box 1.SF.2
Anchoring Inflation Expectations When Inflation Is Undershooting April 2014, Box 1.3
Commodity Prices and Forecasts April 2014, Chapter 1,
Special Feature
Commodity Market Developments and Forecasts, with a Focus on Natural Gas October 2014, Chapter 1,
in the World Economy Special Feature
Commodity Market Developments and Forecasts, with a Focus on Investment April 2015, Chapter 1,
in an Era of Low Oil Prices Special Feature
The Oil Price Collapse: Demand or Supply? April 2015, Box 1.1
Commodity Market Developments and Forecasts, with a Focus on Metals in the World Economy October 2015, Chapter 1,
Special Feature
The New Frontiers of Metal Extraction: The North-to-South Shift October 2015, Chapter 1,
Special Feature Box 1.SF.1
Where Are Commodity Exporters Headed? Output Growth in the Aftermath
of the Commodity Boom October 2015, Chapter 2

SELECTED TOPICS
The Not-So-Sick Patient: Commodity Booms and the Dutch Disease Phenomenon October 2015, Box 2.1
Do Commodity Exporters’ Economies Overheat during Commodity Booms? October 2015, Box 2.4
Commodity Market Developments and Forecasts, with a Focus on the April 2016, Chapter 1,
Energy Transition in an Era of Low Fossil Fuel Prices Special Feature
Global Disinflation in an Era of Constrained Monetary Policy October 2016, Chapter 3
Commodity Market Developments and Forecasts, with a Focus on Food Security and October 2016, Chapter 1,
Markets in the World Economy Special Feature
How Much Do Global Prices Matter for Food Inflation? October 2016, Box 3.3
Commodity Market Developments and Forecasts, With a Focus on the Role Technology and April 2017, Chapter 1,
Unconventional Sources in the Global Oil Market Special Feature
Commodity Market Developments and Forecasts October 2017, Chapter 1,
Special Feature
Commodity Market Developments and Forecasts April 2018, Chapter 1,
Special Feature
What Has Held Core Inflation Back in Advanced Economies? April 2018, Box 1.2
The Role of Metals in the Economics of Electric Vehicles April 2018, Box 1.SF.1
V. Fiscal Policy

Improved Emerging Market Fiscal Performance: Cyclical or Structural? September 2006, Box 2.1
When Does Fiscal Stimulus Work? April 2008, Box 2.1
Fiscal Policy as a Countercyclical Tool October 2008, Chapter 5
Differences in the Extent of Automatic Stabilizers and Their Relationship with Discretionary Fiscal Policy October 2008, Box 5.1
Why Is It So Hard to Determine the Effects of Fiscal Stimulus? October 2008, Box 5.2
Have the U.S. Tax Cuts Been “TTT” [Timely, Temporary, and Targeted]? October 2008, Box 5.3
Will It Hurt? Macroeconomic Effects of Fiscal Consolidation October 2010, Chapter 3
Separated at Birth? The Twin Budget and Trade Balances September 2011, Chapter 4
Are We Underestimating Short-Term Fiscal Multipliers? October 2012, Box 1.1
The Implications of High Public Debt in Advanced Economies October 2012, Box 1.2
The Good, the Bad, and the Ugly: 100 Years of Dealing with Public Debt Overhangs October 2012, Chapter 3
The Great Divergence of Policies April 2013, Box 1.1
Public Debt Overhang and Private Sector Performance April 2013, Box 1.2
Is It Time for an Infrastructure Push? The Macroeconomic Effects of Public Investment October 2014, Chapter 3
Improving the Efficiency of Public Investment October 2014, Box 3.2
The Macroeconomic Effects of Scaling Up Public Investment in Developing Economies October 2014, Box 3.4
Fiscal Institutions, Rules, and Public Investment October 2014, Box 3.5
Commodity Booms and Public Investment October 2015, Box 2.2
Cross-Border Impacts of Fiscal Policy: Still Relevant October 2017, Chapter 4
The Spillover Impact of U.S. Government Spending Shocks on External Positions October 2017, Box 4.1
Macroeconomic Impact of Corporate Tax Policy Changes April 2018, Box 1.5
VI. Monetary Policy, Financial Markets, and Flow of Funds

How Do Financial Systems Affect Economic Cycles? September 2006, Chapter 4
Financial Leverage and Debt Deflation September 2006, Box 4.1
Financial Linkages and Spillovers April 2007, Box 4.1
Macroeconomic Conditions in Industrial Countries and Financial Flows to Emerging Markets April 2007, Box 4.2
Macroeconomic Implications of Recent Market Turmoil: Patterns from Previous Episodes October 2007, Box 1.2

What Is Global Liquidity? October 2007, Box 1.4

The Changing Housing Cycle and the Implications for Monetary Policy April 2008, Chapter 3
Is There a Credit Crunch? April 2008, Box 1.1
Assessing Vulnerabilities to Housing Market Corrections April 2008, Box 3.1
Financial Stress and Economic Downturns October 2008, Chapter 4
The Latest Bout of Financial Distress: How Does It Change the Global Outlook? October 2008, Box 1.1
Policies to Resolve Financial System Stress and Restore Sound Financial Intermediation October 2008, Box 4.1
How Vulnerable Are Nonfinancial Firms? April 2009, Box 1.2
The Case of Vanishing Household Wealth April 2009, Box 2.1
Impact of Foreign Bank Ownership during Home-Grown Crises April 2009, Box 4.1
A Financial Stress Index for Emerging Economies April 2009, Appendix 4.1
Financial Stress in Emerging Economies: Econometric Analysis April 2009, Appendix 4.2
How Linkages Fuel the Fire April 2009, Chapter 4
Lessons for Monetary Policy from Asset Price Fluctuations October 2009, Chapter 3
Were Financial Markets in Emerging Economies More Resilient than in Past Crises? October 2009, Box 1.2
Risks from Real Estate Markets October 2009, Box 1.4
Financial Conditions Indices April 2011, Appendix 1.1
House Price Busts in Advanced Economies: Repercussions for Global Financial Markets April 2011, Box 1.1
International Spillovers and Macroeconomic Policymaking April 2011, Box 1.3
Credit Boom-Bust Cycles: Their Triggers and Policy Implications September 2011, Box 1.2
Are Equity Price Drops Harbingers of Recession? September 2011, Box 1.3
Cross-Border Spillovers from Euro Area Bank Deleveraging April 2012, Chapter 2,
Spillover Feature
The Financial Transmission of Stress in the Global Economy October 2012, Chapter 2,
Spillover Feature
The Great Divergence of Policies April 2013, Box 1.1
Taper Talks: What to Expect When the United States Is Tightening October 2013, Box 1.1
Credit Supply and Economic Growth April 2014, Box 1.1
Should Advanced Economies Worry about Growth Shocks in Emerging Market Economies? April 2014, Chapter 2,
Spillover Feature
Perspectives on Global Real Interest Rates April 2014, Chapter 3
Housing Markets across the Globe: An Update October 2014, Box 1.1
U.S. Monetary Policy and Capital Flows to Emerging Markets April 2016, Box 2.2
A Transparent Risk-Management Approach to Monetary Policy October 2016, Box 3.5
Will the Revival in Capital Flows to Emerging Markets Be Sustained? October 2017, Box 1.2
VII. Labor Markets, Poverty, and Inequality

The Globalization of Labor April 2007, Chapter 5
Emigration and Trade: How Do They Affect Developing Countries? April 2007, Box 5.1
Labor Market Reforms in the Euro Area and the Wage-Unemployment Trade-Off October 2007, Box 2.2
Globalization and Inequality October 2007, Chapter 4
The Dualism between Temporary and Permanent Contracts: Measures, Effects, and Policy Issues April 2010, Box 3.1
Short-Time Work Programs April 2010, Box 3.2
Slow Recovery to Nowhere? A Sectoral View of Labor Markets in Advanced Economies September 2011, Box 1.1
The Labor Share in Europe and the United States during and after the Great Recession April 2012, Box 1.1
Jobs and Growth: Can’t Have One without the Other? October 2012, Box 4.1

SELECTED TOPICS
Reforming Collective-Bargaining Systems to Achieve High and Stable Employment April 2016, Box 3.2
Understanding the Downward Trend in Labor Shares April 2017, Chapter 3
Labor Force Participation Rates in Advanced Economies October 2017, Box 1.1
Recent Wage Dynamics in Advanced Economies: Drivers and Implications October 2017, Chapter 2
Labor Market Dynamics by Skill Level October 2017, Box 2.1
Worker Contracts and Nominal Wage Rigidities in Europe: Firm-level Evidence October 2017, Box 2.2
Wage and Employment Adjustment After the Global Financial Crisis: Firm-level Evidence October 2017, Box 2.3
Labor Force Participation in Advanced Economies: Drivers and Prospects April 2018, Chapter 2
Youth Labor Force Participation in Emerging Market and Developing Economies versus
Advanced Economies April 2018, Box 2.1
Storm Clouds Ahead? Migration and Labor Force Participation Rates April 2018, Box 2.4
Are Manufacturing Jobs Better Paid? Worker-Level Evidence from Brazil April 2018, Box 3.3
VIII. Exchange Rate Issues

How Emerging Market Countries May Be Affected by External Shocks September 2006, Box 1.3
Exchange Rates and the Adjustment of External Imbalances April 2007, Chapter 3
Exchange Rate Pass-Through to Trade Prices and External Adjustment April 2007, Box 3.3
Depreciation of the U.S. Dollar: Causes and Consequences April 2008, Box 1.2
Lessons from the Crisis: On the Choice of Exchange Rate Regime April 2010, Box 1.1
Exchange Rate Regimes and Crisis Susceptibility in Emerging Markets April 2014, Box 1.4
Exchange Rates and Trade Flows: Disconnected? October 2015, Chapter 3
The Relationship between Exchange Rates and
Global-Value-Chain-Related Trade October 2015, Box 3.1
Measuring Real Effective Exchange Rates and Competitiveness:
The Role of Global Value Chains October 2015, Box 3.2
Labor Force Participation Rates in Advanced Economies October 2017, Box 1.1
Recent Wage Dynamics in Advanced Economies: Drivers and Implications October 2017, Chapter 2
Labor Market Dynamics by Skill Level October 2017, Box 2.1
Worker Contracts and Nominal Wage Rigidities in Europe: Firm-Level Evidence October 2017, Box 2.2
Wage and Employment Adjustment After the Global Financial Crisis: Firm-Level Evidence October 2017, Box 2.3
IX. External Payments, Trade, Capital Movements, and Foreign Debt

Capital Flows to Emerging Market Countries: A Long-Term Perspective September 2006, Box 1.1
How Will Global Imbalances Adjust? September 2006, Box 2.1
External Sustainability and Financial Integration April 2007, Box 3.1
Large and Persistent Current Account Imbalances April 2007, Box 3.2
Multilateral Consultation on Global Imbalances October 2007, Box 1.3
Managing the Macroeconomic Consequences of Large and Volatile Aid Flows October 2007, Box 2.3
Managing Large Capital Inflows October 2007, Chapter 3
Can Capital Controls Work? October 2007, Box 3.1
Multilateral Consultation on Global Imbalances: Progress Report April 2008, Box 1.3
How Does the Globalization of Trade and Finance Affect Growth? Theory and Evidence April 2008, Box 5.1
Divergence of Current Account Balances across Emerging Economies October 2008, Chapter 6
Current Account Determinants for Oil-Exporting Countries October 2008, Box 6.1
Sovereign Wealth Funds: Implications for Global Financial Markets October 2008, Box 6.2
Global Imbalances and the Financial Crisis April 2009, Box 1.4

Trade Finance and Global Trade: New Evidence from Bank Surveys October 2009, Box 1.1
From Deficit to Surplus: Recent Shifts in Global Current Accounts October 2009, Box 1.5
Getting the Balance Right: Transitioning out of Sustained Current Account Surpluses April 2010, Chapter 4
Emerging Asia: Responding to Capital Inflows October 2010, Box 2.1
Latin America-5: Riding Another Wave of Capital Inflows October 2010, Box 2.2
Do Financial Crises Have Lasting Effects on Trade? October 2010, Chapter 4
Unwinding External Imbalances in the European Union Periphery April 2011, Box 2.1
International Capital Flows: Reliable or Fickle? April 2011, Chapter 4
External Liabilities and Crisis Tipping Points September 2011, Box 1.5
The Evolution of Current Account Deficits in the Euro Area April 2013, Box 1.3
External Rebalancing in the Euro Area October 2013, Box 1.3
The Yin and Yang of Capital Flow Management: Balancing Capital Inflows with Capital Outflows October 2013, Chapter 4
Simulating Vulnerability to International Capital Market Conditions October 2013, Box 4.1
The Trade Implications of the U.S. Shale Gas Boom October 2014, Box 1.SF.1
Are Global Imbalances at a Turning Point? October 2014, Chapter 4
Switching Gears: The 1986 External Adjustment October 2014, Box 4.1
A Tale of Two Adjustments: East Asia and the Euro Area October 2014, Box 4.2
Understanding the Role of Cyclical and Structural Factors in the Global Trade Slowdown April 2015, Box 1.2
Small Economies, Large Current Account Deficits October 2015, Box 1.2
Capital Flows and Financial Deepening in Developing Economies October 2015, Box 1.3
Dissecting the Global Trade Slowdown April 2016, Box 1.1
Understanding the Slowdown in Capital Flows to Emerging Markets April 2016, Chapter 2
Capital Flows to Low-Income Developing Countries April 2016, Box 2.1
The Potential Productivity Gains from Further Trade and Foreign Direct Investment Liberalization April 2016, Box 3.3
Global Trade: What’s Behind the Slowdown October 2016, Chapter 2
The Evolution of Emerging Market and Developing Economies’ Trade Integration with
China’s Final Demand April 2017, Box 2.3
Shifts in the Global Allocation of Capital: Implications for Emerging Market and
Developing Economies April 2017, Box 2.4
Macroeconomic Adjustment in Emerging Market Commodity Exporters October 2017, Box 1.4
Remittances and Consumption Smoothing October 2017, Box 1.5
A Multidimensional Approach to Trade Policy Indicators April 2018, Box 1.6
The Rise of Services Trade April 2018, Box 3.2
Role of Foreign Aid in Improving Productivity in Low-Income Developing Countries April 2018, Box 4.3
X. Regional Issues

EMU: 10 Years On October 2008, Box 2.1
Vulnerabilities in Emerging Economies April 2009, Box 2.2
East-West Linkages and Spillovers in Europe April 2012, Box 2.1
The Evolution of Current Account Deficits in the Euro Area April 2013, Box 1.3
Still Attached? Labor Force Participation Trends in European Regions April 2018, Box 2.3
XI. Country-Specific Analyses

Why Is the U.S. International Income Account Still in the Black, and Will This Last? September 2005, Box 1.2
Is India Becoming an Engine for Global Growth? September 2005, Box 1.4
Saving and Investment in China September 2005, Box 2.1
China’s GDP Revision: What Does It Mean for China and the Global Economy? April 2006, Box 1.6

SELECTED TOPICS
What Do Country Studies of the Impact of Globalization on Inequality Tell Us?

Examples from Mexico, China, and India October 2007, Box 4.2
Japan after the Plaza Accord April 2010, Box 4.1
Taiwan Province of China in the Late 1980s April 2010, Box 4.2
Did the Plaza Accord Cause Japan’s Lost Decades? April 2011, Box 1.4
Where Is China’s External Surplus Headed? April 2012, Box 1.3
The U.S. Home Owners’ Loan Corporation April 2012, Box 3.1
Household Debt Restructuring in Iceland April 2012, Box 3.2
Abenomics: Risks after Early Success? October 2013, Box 1.4
Is China’s Spending Pattern Shifting (away from Commodities)? April 2014, Box 1.2
Public Investment in Japan during the Lost Decade October 2014, Box 3.1
Japanese Exports: What’s the Holdup? October 2015, Box 3.3
The Japanese Experience with Deflation October 2016, Box 3.2
Permanently Displaced? Labor Force Participation in US States and Metropolitan Areas April 2018, Box 2.2
XII. Special Topics

Climate Change and the Global Economy April 2008, Chapter 4
Rising Car Ownership in Emerging Economies: Implications for Climate Change April 2008, Box 4.1
South Asia: Illustrative Impact of an Abrupt Climate Shock April 2008, Box 4.2
Macroeconomic Policies for Smoother Adjustment to Abrupt Climate Shocks April 2008, Box 4.3
Catastrophe Insurance and Bonds: New Instruments to Hedge Extreme Weather Risks April 2008, Box 4.4
Recent Emission-Reduction Policy Initiatives April 2008, Box 4.5
Complexities in Designing Domestic Mitigation Policies April 2008, Box 4.6
Getting By with a Little Help from a Boom: Do Commodity
Windfalls Speed Up Human Development? October 2015, Box 2.3
Breaking the Deadlock: Identifying the Political Economy Drivers of Structural Reforms April 2016, Box 3.1
Can Reform Waves Turn the Tide? Some Case Studies Using the Synthetic Control Method April 2016, Box 3.4
A Global Rush for Land October 2016, Box 1.SF.1
Conflict, Growth, and Migration April 2017, Box 1.1
Tackling Measurement Challenges of Irish Economic Activity April 2017, Box 1.2
Within-Country Trends in Income per Capita: The Case of the Brazil, Russia, India, China,
and South Africa April 2017, Box 2.1
Technological Progress and Labor Shares: A Historical Overview April 2017, Box 3.1
The Elasticity of Substitution Between Capital and Labor: Concept and Estimation April 2017, Box 3.2
Routine Tasks, Automation, and Economic Dislocation around the World April 2017, Box 3.3
Adjustments to the Labor Share of Income April 2017, Box 3.4
The Effects of Weather Shocks on Economic Activity: How Can Low-Income Countries Cope? October 2017, Chapter 3
The Growth Impact of Tropical Cyclones October 2017, Box 3.1
The Role of Policies in Coping with Weather Shocks: A Model-Based Analysis October 2017, Box 3.2
Strategies for Coping with Weather Shocks and Climate Change: Selected Case Studies October 2017, Box 3.3
Coping with Weather Shocks: The Role of Financial Markets October 2017, Box 3.4
Historical Climate, Economic Development, and the World Income Distribution October 2017, Box 3.5
Mitigating Climate Change October 2017, Box 3.6
Smartphones and Global Trade April 2018, Box 1.1
Has Mismeasurement of the Digital Economy Affected Productivity Statistics April 2018, Box 1.4
The Changing Service Content of Manufactures April 2018, Box 3.1
Patent Data and Concepts April 2018, Box 4.1
International Technology Sourcing and Knowledge Spillovers April 2018, Box 4.2
Relationship between Competition, Concentration, and Innovation April 2018, Box 4.4

IMF EXECUTIVE BOARD DISCUSSION OF THE OUTLOOK,
APRIL 2018
The following remarks were made by the Chair at the conclusion of the Executive Board’s discussion of the
Fiscal Monitor, Global Financial Stability Report, and World Economic Outlook on April 2, 2018.
E
xecutive Directors broadly shared the key protectionism and inward-looking policies. Record-
messages of the flagship reports and found high levels of global debt, geopolitical tensions, and
the analytical chapters topical, relevant, and climate events also threaten global growth prospects.
insightful. They welcomed the broadbased Against this backdrop, Directors underscored that
recovery of the global economy, supported by a pickup the cyclical upswing provides a golden opportunity to
in investment and trade. Directors observed that global advance policies and reforms to strengthen medium-
growth is expected to rise further in the near term. term prospects and reduce vulnerabilities. Priorities are
Meanwhile, inflation remains muted in many coun- to raise potential output, ensure the gains are widely
tries. Subdued labor productivity growth and popula- shared, enhance economic and financial resilience, and
tion aging continue to hold back growth in advanced safeguard debt sustainability. Directors stressed that
economies. While the recent commodity price increase a multilateral framework that is open, resilient, and
has supported a recovery in commodity-dependent adhered to by all can support growth and benefit the
emerging market and developing economies, the ongo- global economy. Enhanced commitment to multilateral
ing adjustment processes continue to weigh on growth. cooperation is particularly needed to reduce trade bar-
Directors agreed that risks around the short-term riers and distortionary trade practices, and to promote
outlook are broadly balanced, but beyond the next a rule-based multilateral trading system that works for
several quarters, risks are tilted to the downside. On all. Directors also called for multilateral cooperation to
the upside, the cyclical pickup in advanced economy further reduce incentives for cross-border profit shift-
growth may prove stronger than expected as slack in ing and tax evasion, avoid tax competition, implement
labor markets may be larger than currently assessed. the postcrisis financial regulatory reform agenda, and
On the downside, a sharp tightening of global finan- address other shared challenges such as refugees, secu-
cial conditions could have negative repercussions for rity threats, cyber risks, and climate change. Reducing
growth, while financial vulnerabilities accumulated excess external imbalances requires policy efforts to lift
over years of low interest rates could amplify the the contribution of domestic sources of growth above
impact of asset price movements on the financial sys- overall GDP growth in surplus countries and to boost
tem, putting growth at risk in the medium term. Most potential output and saving in deficit countries.
Directors noted that the tax reform in the United Directors concurred that monetary accommodation
States is procyclical and may trigger inflation pressure should continue in advanced economies with infla-
and a faster-than-anticipated withdrawal of monetary tion below target. Where output is close to potential
accommodation, as well as widen global imbalances, and inflation is rising toward target, a gradual, data-
although the view was also expressed that the reform dependent, and well-communicated withdrawal of
would boost investment and efficiency, and thus move monetary support is warranted. Directors supported
the US economy to a higher, sustainable growth path. the call for fiscal policy to start rebuilding buffers now,
An abrupt tightening of global financial conditions, where appropriate, to create room for an eventual
especially if accompanied by capital flow reversals, downturn and prevent fiscal vulnerabilities from becom-
could be challenging for several emerging markets and ing a source of stress. Fiscal adjustment is warranted in
low-income developing countries, notwithstanding most countries, calibrated to avoid procyclicality and
improved resilience of their financial systems. Down- anchored on fiscal reforms that increase productivity and
side risks are particularly evident from escalating trade promote human and physical capital. In countries that

have ample fiscal space and are operating at or close to and developing economies enhance their resilience to
capacity, fiscal policy should be used to facilitate growth- external shocks. Directors welcomed China’s progress in
enhancing structural reforms. Directors also saw a role reducing financial vulnerabilities and encouraged fur-
for fiscal policy in promoting equality, and for labor and ther efforts to strengthen its regulatory and supervisory
immigration policies in boosting labor supply. frameworks, particularly in the shadow banking sector.
Directors agreed that digitalization presents both Directors noted that low-income developing coun-
opportunities and risks. Digitalization can reduce tax tries face multiple challenges in their effort to progress
compliance costs, improve spending efficiency, and toward the 2030 Sustainable Development Goals.
enhance social protection. At the same time, it cre- They expressed concern over the broad-based increase
ates challenges for fiscal policy and the international in public debt burdens, the increasing number of
tax system. Directors noted that mitigating risks countries at high risk of debt distress, and data gaps.
from digitalization would require a comprehensive These underscore the urgent need for fiscal prudence,
reform agenda, adequate resources, and a coordinated improved debt management capacity, and greater debt
approach toward a long-term vision of the interna- transparency on the part of both debtors and credi-
tional tax architecture. tors, as well as concerted efforts from the international
Directors welcomed the increased resilience of community. Several countries need to make room in
the banking system and stressed the importance of their budgets to accommodate higher spending on
completing and implementing the postcrisis regula- social services such as health care and education, and
tory reform agenda. They encouraged policymakers public investment, by mobilizing domestic revenues
to develop and deploy micro and macroprudential and improving spending efficiency. Commodity
tools to address financial vulnerabilities, and to closely exporters and those vulnerable to climate-related events
monitor risks related to credit allocation and increas- face additional complex challenges of diversifying their
ingly synchronized house prices across countries. The economies. While country circumstances differ, com-
global implications of Brexit-related challenges also call mon priorities for promoting economic diversification
for close cross-border cooperation. Directors concurred and employment include increasing access to credit,
that, while crypto assets do not pose an immediate expanding vocational skills training, and improving the
threat to financial stability, if widely used, they may quality of infrastructure.
raise issues about investor and consumer protection, Directors expressed concern over the stalled progress
money laundering, and tax evasion. in the catching-up process of emerging market and
Directors agreed that enhancing the quality of credit developing economies. They noted that, to facilitate
intermediation, avoiding credit booms that lead to income convergence, policies should aim to strengthen
excessive risk taking, and, where feasible, permitting governance, improve educational and health outcomes,
exchange rate flexibility can help emerging market and lower entry barriers for new firms.

Highlights from IMF Publications
Unleashing Growth and Digital Revolutions in Fiscal Policies and Gender Equality
Strengthening Resilience in Public Finance $25. ©2017. Paperback
the Caribbean $25. ©2017. Paperback ISBN 978-1-51359-036-3. 250pp.
$25. ©2017. Paperback ISBN 978-1-48431-522-4. 362pp.
ISBN 978-1-48431-519-4. 378pp.
Fiscal Politics Financial Integration in Shifting Commodity Markets

$40. ©2017. Paperback Latin America: A New Strategy in a Globalized World
ISBN 978-1-47554-790-0. 548pp. for a New Normal $25. ©2017. Paperback
$25. ©2017. Paperback ISBN 978-1-48431-032-8. 112pp.
ISBN 978-1-51352-024-7. 177pp.
To order, visit bookstore.imf.org/we418

Text Book Full Report of Word Economy PDF

Uploaded by

Document Informationclick to expand document information

Document Informationclick to expand document information

Copyright:

Available Formats

Text Book Full Report of Word Economy PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Text Book Full Report of Word Economy PDF

Uploaded by

Copyright:

Available Formats

Wor l d Eco no m i c a nd F i na nci a l S ur v e y s

World Economic Outlook

Cyclical Upswing, Structural Change

Cover and Design: Luisa Menjivar and Jorge Salazar

Joint Bank-Fund Library

Names: International Monetary Fund.

ISBN 978-1-48433-827-8 (paper)

Recommended citation: International Monetary Fund. 2018. World Economic Outlook:

Publication orders may be placed online, by fax, or through the mail:

Assumptions and Conventions x

Executive Summary xvi

Chapter 1. Global Prospects and Policies 1

Chapter 2. Labor Force Participation in Advanced Economies: Drivers and Prospects 71

International Monetary Fund | April 2018 iii

Chapter 3. Manufacturing Jobs: Implications for Productivity and Inequality 129

Chapter 4. Is Productivity Growth Shared in a Globalized Economy? 173

Statistical Appendix 215

iv International Monetary Fund | April 2018

Table B. Advanced Economies by Subgroup 220

World Economic Outlook, Selected Topics 271

IMF Executive Board Discussion of the Outlook, April 2018 281

International Monetary Fund | April 2018 v

vi International Monetary Fund | April 2018

Figure 1.7. Advanced Economies: Monetary and Financial Market Conditions 7

International Monetary Fund | April 2018 vii

Figure 2.13. Changes in Participation Rates, Actual versus Predicted, 1995–2011 89

viii International Monetary Fund | April 2018

International Monetary Fund | April 2018 ix

ASSUMPTIONS AND CONVENTIONS

x International Monetary Fund | April 2018

Corrections and Revisions

Print and Digital Editions

Copyright and Reuse

International Monetary Fund | April 2018 xi

xii International Monetary Fund | April 2018

International Monetary Fund | April 2018 xiii

xiv International Monetary Fund | April 2018

International Monetary Fund | April 2018 xv

xvi International Monetary Fund | April 2018

International Monetary Fund | April 2018 xvii

xviii International Monetary Fund | April 2018

GLOBAL PROSPECTS AND POLICIES

International Monetary Fund | April 2018 1

4 4. Advanced Economies 5. Emerging Market and 9

2 International Monetary Fund | April 2018

Rising Commodity Prices 25 4. Emerging Market and Developing Economies Excluding

January, the highest level since 2015, following

1Box 1.1 discusses the role of the so‑called tech cycle in explaining

the rebound in trade in Asian economies and elsewhere.

International Monetary Fund | April 2018 3

7 1. Advanced Economies: Export Growth 180

7 2. Advanced Economies: Import Growth 60

10 3. Emerging Market and Developing Economies: Export Growth

4 International Monetary Fund | April 2018

Headline Inflation Has Picked up, but Core Inflation 0 4

International Monetary Fund | April 2018 5

6 International Monetary Fund | April 2018