Petar Stankov
Economic Freedom and Welfare Before
and After the Crisis
“Stankov provides a timely and perceptive analysis of the complex interaction
between economic freedom and reforms of the widely discussed “Washington
Consensus” and growth in incomes, inequality and multiple measures of
individual and societal welfare. This should be required reading for anyone
trying to understand the rise of populist political movements in recent years.”
—Randall K. Filer, Professor of Economics, Hunter College and the CUNY
Graduate Center, President, The CERGE-EI Foundaion
“Anyone interested in the political economy of which economic policies achieve
the best results will find a most comprehensive analysis covering the globe
applying thorough quantitative analysis. Stankov concludes some but not all
liberalising policies do improve welfare but frequently lead to greater inequality.
This then leads into a novel exploration of how such circumstances generate the
populism one sees so widespread today. Nothing could be more timely.”
—Oleh Havrylyshyn, CASE Senior Fellow
“With this volume Stankov offers both a comprehensive catalogue and review
of the literature on economic freedom and a collection of new results concerning
policy and welfare convergence that is timely and has international appeal.
Economists and others who are researching and teaching in fields related to the
area of economic freedom will find this book indispensable.”
—Franklin G. Mixon, Jr., Columbus State University, USA
Petar Stankov
Economic Freedom
and Welfare Before
and After the Crisis
Petar Stankov
University of National and World Economy
Sofia, Bulgaria
Reviewed by
Franklin Mixon, Columbus State University
Joshua Hall, West Virginia University
Robert Lawson, Southern Methodist University
ISBN 978-3-319-62496-9
ISBN 978-3-319-62497-6
DOI 10.1007/978-3-319-62497-6
(eBook)
Library of Congress Control Number: 2017948308
© The Editor(s) (if applicable) and The Author(s) 2017
This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether
the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of
illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and
transmission or information storage and retrieval, electronic adaptation, computer software, or by similar
or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, express or implied, with respect to the material contained herein or
for any errors or omissions that may have been made. The publisher remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations.
Cover illustration: Getty Images/Caracter design
Printed on acid-free paper
This Palgrave Macmillan imprint is published by Springer Nature
The registered company is Springer International Publishing AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
To my family who taught me freedom
and the perils of using it unwisely.
Acknowledgements
I thank Palgrave Macmillan for their exceptional professionalism in
dealing with the book proposal, the first draft and the revised versions
of the book. I would like to express my sincere gratitude to the three
referees whose critical comments contributed to improvement of the first
draft.
I also thank the Economics Department of the University of National
and World Economy (UNWE) in Sofia, Bulgaria, and the Economics
Department of the American University in Bulgaria (AUBG) for
providing excellent teaching and research environments. Specifically,
I would like to thank Ivaylo Beev, Shteryo Nozharov, Kristina Stefanova,
Dimitar Damyanov, and Entsislav Harmandzhiev (all from the UNWE)
for their input during a research seminar at the Department, and
Aleksandar Vasilev (AUBG) for his customarily sharp comments.
A big thanks goes to Martin Rode (University of Navarra) for sharing
The Wild Bunch! data and to Andreas Heinö (Timbro Institute) for
sharing the Timbro Authoritarian Populism data. I was very lucky to have
rapid responses from both of them at a crucial moment of redrafting.
Deborah Novakova (CERGE-EI) provided a native English reading
vii
viii
Acknowledgements
of the manuscript. Further, CERGE-EI secured additional financial
support through its invaluable Career Integration Fellowship.
Finally, thanks to Geri Stankova for putting up with the rest—you
know you rock, girl.
Thank you all.
Sofia, Bulgaria
May 2017
Petar Stankov
Contents
1 Introduction
1
2 Contemporary Views on Welfare and Reforms
9
3 Policies and Reforms
43
4 Policy Convergence Vs. Welfare Convergence
69
5 Welfare and Reforms: Evidence
99
6 Crises, Welfare, and Populism
135
7 Conclusion
165
Index
169
ix
List of Figures
Fig.
Fig.
Fig.
Fig.
Fig.
Fig.
Fig.
Fig.
Fig.
Fig.
Fig.
Fig.
Fig.
Fig.
Fig.
Fig.
Fig.
Fig.
Fig.
Fig.
Fig.
3.1
3.2
3.3
3.4
3.5
3.6
3.7
3.8
3.9
3.10
3.11
4.1
4.2
4.3
4.4
4.5
4.6
4.7
4.8
4.9
4.10
Government intervention since 1970
Legal system and security of property rights since 1970
Monetary policies since 1970
Free trade policies since 1970
Regulatory policies since 1970
Size of government reforms since 1970
Property rights reforms since 1970
Monetary reforms since 1970
Trade reforms since 1970
Overall regulatory reforms since 1970
Financial, labor, and business reforms: a 10-year angle
Convergence in government intervention: 1970–2014
Convergence in property rights protection: 1970–2014
Convergence in monetary policies: 1970–2014
Convergence in trade policies: 1970–2014
Convergence in regulatory policies: 1970–2014
Sigma convergence in policies: 1970–2014
Income per capita convergence: 1970–2014
Consumption per capita convergence: 1970–2014
Life expectancy convergence: 1970–2014
Income inequality convergence: 1970–2014
44
47
49
52
55
59
60
61
62
63
64
76
77
78
78
79
84
85
86
87
88
xi
xii
Fig.
Fig.
Fig.
Fig.
Fig.
Fig.
List of Figures
5.1
5.2
5.3
5.4
5.5
6.1
Fig. 6.2
Fig. 6.3
Fig. 6.4
Government intervention and welfare: 1970–2014
Property rights and welfare: 1970–2014
Monetary reforms and welfare: 1970–2014
Trade reforms and welfare: 1970–2014
Deregulation and welfare: 1970–2014
The crisis, economic freedom, and populism:
Ireland vs. Greece
The crisis, economic freedom, and populism:
Chile vs. Venezuela
The crisis and economic freedom in land-locked countries
The crisis and economic freedom in large open economies
100
102
104
106
109
152
156
159
161
List of Tables
Table 4.1
Table 4.2
Table 5.1
The speed of policy convergence: 1970–2014
The speed of welfare convergence: 1970–2014
Size of government, income, and consumption:
1970–2014
Table 5.2 Size of government, life expectancy, and inequality:
1970–2014
Table 5.3 Property rights, income, and consumption: 1970–2014
Table 5.4 Property rights, life expectancy, and inequality:
1970–2014
Table 5.5 Monetary stability, income, and consumption: 1970–2014
Table 5.6 Monetary stability, life expectancy, and inequality:
1970–2014
Table 5.7 Free trade, income, and consumption: 1970–2014
Table 5.8 Free trade, life expectancy, and inequality: 1970–2014
Table 5.9 Deregulation, income, and consumption: 1970–2014
Table 5.10 Deregulation, life expectancy, and inequality: 1970–2014
Table 6.1 Political economy of populism before and after the crisis
Table 6.2 Populism as a rhetorical style before and after the crisis
Table 6.3 Authoritarian populism and crises
81
92
115
116
117
118
119
120
121
122
123
124
144
147
149
xiii
1
Introduction
The world has witnessed an unprecedented wave of economic freedom
reforms over the last 45 years. This book is about finding out to what
degree they made sense. They would make sense if the widespread policy
convergence toward market-oriented reforms has made nations better-off.
It has long been established that some market-oriented reforms increase
living standards and accelerate economic growth. However, being betteroff means much more than that.
Suppose an economy grows over a certain period of time, and this
growth is a result of conscious efforts by policy makers to make the
business environment more growth-friendly. However, there is a risk that
economists and policy makers could be blinded by this seemingly good
fortune. If most of the additional wealth created while the economy was
growing goes to a tiny proportion of the population, then political tensions
within the country will be growing.
Those political tensions are likely to lead to a soaring number of voters
discontent with the market-oriented reforms. As a result, sooner rather
than later, they would elect a government favoring large-scale redistribution policies in favor of the many at the expense of the few, as Meltzer
and Richard (1981) suggest, among others. The recent populist wave in
© The Author(s) 2017
P. Stankov, Economic Freedom and Welfare Before and After the Crisis,
DOI 10.1007/978-3-319-62497-6_1
1
2
P. Stankov
both Europe, Latin America, and even the USA suggests that the postCrisis growth is indeed producing large numbers of discontent voters.
Recent evidence by Rode and Revuelta (2015) and by Heinö (2016) not
only documents this populist resurgence across the globe but also portrays
the tendency among many elected populist politicians to overshoot with
redistribution policies and thus to stifle economic freedom. In turn, this
could lead to stifled growth prospects for the economy exactly when it
needs growth most.
If this is the case, then a good-for-growth policy will not be sufficient to
gain political support, especially in the aftermath of the Great Recession.
An effective pro-growth policy opens up opportunities for businesses to
grow but should also find ways to extend political support for further
market-oriented reforms. Finding such ways is certainly not easy but it is
not impossible. We need to start thinking about welfare in a broader sense
than just income per capita growth. Luckily, recent literature suggests ways
to expand the welfare concept.
In the spirit of Jones and Klenow (2016), among others, in this book
welfare is understood as an increase in living standards and consumption
over time, gains in life expectancy to enjoy the possible increase in living
standards, and reductions in income inequality. It is these four components of welfare that this book is focused on.
Therefore, this work is about the changes in welfare across countries
and over time, in which welfare is defined as a collection of the above
four elements. At the same time, the core goal is to analyze the impact
of market-oriented reforms on changes in welfare across the globe since
1970.
My thesis is that, despite the large-scale market reforms which brought
certain gains in income per capita, those gains were not significant enough
to boost welfare in other politically important ways. As a result, political
support for more market-oriented reforms has become limited, and voter
discontent is dominating the policy agenda against further market reforms,
especially since the Great Recession. In turn, this almost certainly produces
populist agendas, with a great degree of inevitability, on both the demand
and the supply side of the political market.
To produce evidence in favor of this thesis, I bring forward a number of
testable hypotheses. First, I scrutinize whether there was a general back-
1 Introduction
3
lash against market-oriented reforms after the Crisis. Second, I study if the
world has become a more uniform place in terms of policies, reforms, and
welfare over time. Third, I test if those reforms have brought significant
increases in welfare over the last 45 years. Fourth, I test whether macroeconomic shocks can explain the dynamics of populism. Finally, I present
a number of case studies from Europe, Latin America, Africa, and Asia to
illustrate how economic freedom reforms correlate with welfare, and with
political support for populist movements over time.
The methods used to produce the evidence in the book are diverse. In a
broad sense, they are a collection of qualitative and quantitative methods.
As most of the analyses are based on data, emphasis is given to quantitative
methods. These include, but are not limited to, ordinary least squares
(OLS), fixed effects panel data, and instrumental variable regressions. A
large part of the evidence produced is also graphical. There are two types
of graphs used: distributional plots and linear fits. The distributional plots
are based on kernel density estimations, while the linear fits are based on
pure linear bivariate estimations.
The main source of data on market reforms is the Economic Freedom
of the World (EFW) 2016 data produced by Gwartney et al. (2016). Its
time span is from 1970 to 2014. The time span in the reforms data also
limits the analysis in this work to a period from 1970 to 2014 for both
reforms and welfare.
The policies and reforms data are presented in the annual Economic
Freedom of the World report. The motivation behind constructing the
historical indices in the report and their use for empirical analyses of welfare
is presented by Gwartney and Lawson (2003) and Gwartney (2009). At
present, the index of economic freedom includes policies and reforms in
five broad domains:
1. Size of Government, measuring broadly the government intervention in
the economy;
2. Legal System and Security of Property Rights, measuring broadly the
capacity of the government to protect property rights;
3. Sound Money, measuring various elements of monetary policies;
4
P. Stankov
4. Freedom to Trade Internationally, measuring the government stance on
free trade; and
5. Government Regulation, measuring policies with respect to the credit
market, labor market, and doing business broadly.
Within each of these policy domains, the report monitors the status quo
and the development of more specific policies. Within both the broad
indexes and the subindexes, the current situation is assigned a number
ranging from 0 to 10. This number is aimed to measure how close the
respective policy is to an economy free from unproductive government
involvement. An index value of 0 is assigned to a policy status quo in
which there exists extensive government involvement. An index value of
10 is awarded to policies which are most market-oriented.
Reforms are measured by the change in a given index from a current
period to the next. If an economy scores a positive change in the index,
then it has made its policies more market-friendly. In other words, there
was more economic freedom in that policy domain during that particular
period. Alternatively, a negative change in the index means that policies
within the country over the given period moved toward more unnecessary
government intervention and have become more market-unfriendly.
There is data on the freedom indices dating back to 1970. The indices
are recorded at 5-year intervals from 1970 to 2000, and annually since
then. Despite the valid criticism of the indices (Caudill et al. 2000;
De Haan et al. 2006; Ram 2014), they allow for various types of analyses.
One approach is to focus on a rather short-term picture, e.g., a policy
stance in a given year in a given domain in a given country, or a snapshot
of the differences in policies across countries at a given point in time.
Another approach to the data is to look at a reform process within a country and within a certain policy domain. As the reform process is measured
by the change in an index over time, the changes in the index can be seen
at 5-, 10-, 20-, and even longer-term intervals across countries. Also, as
some countries reform a bit and then fall into reform fatigue, the reform
dynamics can also be explored both across countries and over time. Therefore, data as rich as the EFW allows for both a cross-country comparison
within a certain policy domain at a given point in time, and a longerterm, dynamic overview of the direction of policy changes in a number of
countries.
1 Introduction
5
Chapter 2 reviews the literature on how the changes in economic freedom affect welfare measures in a number of studies on both developing
and developed countries. As it turns out, no single economic freedom
reform has had a linearly positive and significant effect on welfare across
countries and over time, a non-linearity suggested more than 20 years ago
for economic growth by Barro (1997).
Chapter 3 illustrates how economic freedom policies and reforms have
developed within each policy area. Policy snapshots are taken at 6 different
moments in time: 1970, 1980, 1990, 2000, 2008, and 2014. Economic
freedom policies are illustrated by distributional plots. Those plots measure the approximate share of countries with a certain value of the index.
Thus, one can monitor how the worldwide distribution of a certain policy
changes over time for each of the 5 broad policy areas.
The policy snapshots, however, do not give a complete picture of policy
developments over time. Those developments can be monitored not by
plotting the distributions of the index values but rather by plotting the
changes in the indices within a certain period. The two plots complement
each other but they also address different questions. While plotting the
index values at a point in time will produce an idea of a policy stance,
plotting the change in the same index will deliver a better understanding
of the underlying reform patterns over the same period. Those reform
patterns are presented in Chap. 3.
We can combine a policy status quo at a certain point in time with
the reform processes before or after setting this policy. This angle on the
reform process is particularly informative of a phenomenon called policy
convergence : countries gradually becoming more similar in their policies
within each policy domain over a certain time period. Policy convergence
both before and after the Crisis is studied in Chap. 4.
Chapter 4 also analyzes welfare convergence : countries gradually becoming more similar in their welfare over a certain time period. The welfare
data is taken from three sources: The Penn World Table 9.0 (PWT9.0),
the World Development Indicators (WDI), and from Milanovic (2014).
The Penn World Table (PWT), version 9.0, is produced by Feenstra
et al. (2015). Along with the WDI, it is one of the most comprehensive
sources of country-level GDP per capita and growth data. It also features data on consumption per capita over time which enables anyone to
6
P. Stankov
analyze consumption growth across countries over time. The PWT9.0 is
also a database featuring the income, output, inputs, and productivity of
182 countries between 1950 and 2014. It was released on June 9, 2016.
The updated version used in this book was released on August 18, 2016.
The PWT9.0 is also used to derive the geometrically averaged compound
growth rates of income per capita and consumption per capita for each of
the periods under consideration.
The data on life expectancy are taken from the WDI database produced
by The World Bank (2016). It contains information on life expectancy
from 154 countries and territories since 1960, all of which can be matched
with the reforms data. It also contains data on income inequality, and more
specifically, on Gini coefficients. However, there is a more comprehensive
data set on income inequality and that is Milanovic (2014), which I use for
the income inequality component of welfare. Milanovic (2014) produces a
standardized Gini coefficient for 166 countries since 1950, which includes
2218 observations. Of those, only a small number are matchable with
the reforms data. However, as it contains more comprehensive income
inequality data, the other sources have an even lower matchable potential.
Equipped with the above data, Chap. 4 discovers graphical and regression evidence of both policy and welfare convergence across countries
over time. However, the fact that policies have converged and the world
has become more similar in terms of welfare does not mean that policy
convergence has lead to welfare convergence. Therefore, we need more
information on the existence of any positive and statistically significant
correlation between welfare and reforms.
Chapter 5 produces this information in two ways. First, graphical
evidence is explored, which plots reforms data and changes in welfare.
However, as graphical evidence observation can be misleading, a more
rigorous approach is employed to study the relationship between welfare
and reforms. This approach is to study the relationship by using panel
OLS models, fixed effects panel models, and instrumental variable estimations. Chapter 5 presents the results from those estimations along with
the graphical evidence.
In fact, despite the existence of some graphical evidence in favor of a
causal relationship between economic freedom reforms and welfare, the
econometric evidence to this end is far weaker. There is conclusive evidence
1 Introduction
7
that economic freedom reforms raise income per capita but do not have
a robust effect on the other measures of welfare. This is at odds with the
majority of results reviewed earlier by Hall and Lawson (2014). As the
next chapter suggests, there are multiple reasons for these differences.
Chapter 6 discusses some of the political consequences of macroeconomic shocks. Specifically, it reviews the impact of recessions, inflation,
unemployment, austerity, and income inequality on the rise of populism
across the globe. Recent efforts by Rode and Revuelta (2015) and by Heinö
(2016) produced much-needed longitudinal data sets on populism. I link
these with the available macrodata to produce an empirical investigation of
the political economy of populism. Fixed effects panel methods show that
recessions are the most consistent predictor of populist resurgences after
the Great Recession. Unemployment also plays a role in spurring left-wing
populist support. Surprisingly, austerity and income inequality rarely play
a statistically significant role in shaping populist popularity. Case studies
from around the world bring additional support to the empirical evidence.
The evidence suggests that more economic freedom raises income per
capita, and income per capita growth insures against the rise of populism.
As the price of populism is often decades of stagnation, this book argues
that freedom reforms do make sense, however small their impact on welfare
is beyond GDP.
References
Barro, R. 1997. Determinants of economic growth: A cross-country empirical study.
MIT Press.
Caudill, S.B., F.C. Zanella, and F.G. Mixon. 2000. Is economic freedom one
dimensional? A factor analysis of some common measures of economic freedom.
Journal of Economic Development 25 (1): 17–40.
De Haan, J., S. Lundstrom, and J. Sturm. 2006. Market-oriented institutions
and policies and economic growth: A critical survey. Journal of Economic Surveys
20 (2): 157–191.
Feenstra, R.C., R. Inklaar, and M.P. Timmer. 2015. The next generation of the
Penn World Table. American Economic Review 105 (10): 3150–3182 (Updated:
August 18, 2016).
8
P. Stankov
Gwartney, J. 2009. Institutions, economic freedom, and cross-country differences
in performance. Southern Economic Journal 75 (4): 937–956.
Gwartney, J., J. Hall, and R. Lawson. 2016. 2016 economic freedom dataset. Fraser
Institute.
Gwartney, J., and R. Lawson. 2003. The concept and measurement of economic
freedom. European Journal of Political Economy 19 (3): 405–430. Economic
Freedom.
Hall, J.C., and R.A. Lawson. 2014. Economic freedom of the world: An accounting of the literature. Contemporary Economic Policy 32 (1): 1–19.
Heinö, A.J. 2016. Timbro Authoritarian Populism Index 2016. Sweden: Timbro
Institute, Stockholm.
Jones, C.I., and P.J. Klenow. 2016. Beyond GDP? Welfare across countries and
time. American Economic Review 106 (9): 2426–2457.
Meltzer, A.H., and S.F. Richard. 1981. A rational theory of the size of government.
Journal of Political Economy 89 (5): 914–927.
Milanovic, B.L. 2014. All the Ginis, 1950–2012 (Updated in Autumn 2014).
Ram, R. 2014. Measuring economic freedom: A comparison of two major
sources. Applied Economics Letters 21 (12): 852–856.
Rode, M., and J. Revuelta. 2015. The wild bunch! An empirical note on populism
and economic institutions. Economics of Governance 16 (1): 73–96.
The World Bank. 2016. World development indicators, 1960–2016 (Updated
Nov. 2016).
2
Contemporary Views on Welfare
and Reforms
2.1
The Concept of Welfare
in the Twenty-First Century
The traditional neoclassical approach to studying welfare is to focus on
Pareto optimality as a criterion for welfare maximization. The debate on
what welfare is, how it can be measured, and how it can be used for
applied economic analysis has been ongoing at least as far back as Marshall’s Principles (Marshall 1890) and his successor at Cambridge, Pigou’s
The Economics of Welfare (Pigou 1920). During the 1930s, the cardinal
approach evolved into using ordinal utility functions, perhaps due to the
contributions of Robbins in his critique of the Cambridge school (Robbins
1932).
The utilitarian approach is admittedly too narrow to capture the significant aspects of welfare other than consumption per capita driven
by income per capita and relative prices. That is why the more recent
neoclassical treatments, e.g. Atkinson (2011), and some heterodox
approaches (Gowdy 2004; Munda 2016; Ng 2003; Schubert 2012)
expand traditional utilitarian welfare economics in important ways. For
example, Ng (2003) proposes the introduction of happiness as a direct
© The Author(s) 2017
P. Stankov, Economic Freedom and Welfare Before and After the Crisis,
DOI 10.1007/978-3-319-62497-6_2
9
10
P. Stankov
measure of welfare, and Gul and Pesendorfer (2007) advocate for measuring “true utility” as a gauge of happiness in a subjective sense as opposed
to “choice utility” which, according to the authors, is plagued by internal inconsistencies. In addition, Gowdy (2004) engages in a discussion of
whether altruism has any place in welfare conceptualization, and Schubert (2012) acknowledges the inherent dynamics of preferences and the
importance of learning at the individual level to adequately measure welfare over time. A more recent discussion by Munda (2016) proposes the
use of different metrics of welfare for different theoretical and empirical
purposes, rather than an all-encompassing single measure.
As a result, the debate on the essence and limitations of the concept
of welfare, which has been active at least since the 1930s and 1940s
(Samuelson 1943; Stigler 1943; Wolfe 1931), has moved far beyond the
traditional orthodoxy. Holcombe (2009, p. 209) reviews the debate and
concludes that “no economist would argue that people are materially better off today than a century ago because the economy is closer to Pareto
optimality.” To effectively conceptualize welfare, contemporary authors
suggest a focus on factors that improve well-being over time (Sen and
Nussbaum 1993; Fleurbaey 2009).
The factors leading to improved well-being are not themselves viewed in
unanimous ways. In a perhaps reductionist fashion and for purely empirical purposes, the contemporary literature represented most recently by
Jones and Klenow (2016) has narrowed the numbers of these factors to
four: (1) an increase in consumption per capita and (2) leisure over time,
(3) gains in life expectancy (reducing mortality, respectively) and (4) a
reduction in income and consumption inequality. The motivation to focus
on those four elements of “consumption-equivalent” welfare is twofold.
First, the authors assert that “standard economic analysis is arguably wellequipped to deal with” these welfare measures (Jones and Klenow 2016,
p. 2426). Second, these measures are included in a larger set of recommendations to improve welfare measurement, as suggested by Stiglitz et al.
(2009).
Jones and Klenow argue that, across their sample of both developed
and developing countries, the correlation between the traditional GDP/c.
measure of welfare and their novel measure is 0.98 in levels (Jones and
Klenow 2016, p. 2427) and 0.97 in growth rates (Jones and Klenow
2 Contemporary Views on Welfare and Reforms
11
2016, p. 2444). In a narrow-minded statistical sense, then, it appears that
the GDP/c. and the Jones–Klenow measure are virtually indistinguishable. However, there are important economic and behavioral differences
between the two indicators which the pure correlations fail to spot. For
example, according to the authors, the average GDP/c. in Western Europe
is about 67% of the one in the USA, but when the additional leisure time,
the longer life expectancy and the lower income inequality in Europe are
taken into account, welfare in Western Europe appears much closer to
that of the USA (p. 2427).
The opposite is true for the developing countries, where GDP/c. appears
closer to the one in the developed world than their actual welfare. The
Jones–Klenow welfare measure in developing countries is considerably
lower than GDP/c. suggests because of the much lower life expectancy and
the significantly higher income inequality in those countries. Therefore,
we can safely accept that GDP/c. is different from the contemporary
understanding of welfare in important ways.
Nevertheless, ignoring living standards measured by per capita income
in a study of welfare would be unwise for at least three reasons. First,
the traditional welfare measurement across countries and over time has
focused on GDP/c. as perhaps the single most important factor behind
increases in welfare, however imperfect a measure of welfare it admittedly
is. Second, using GDP/c. is convenient from an empirical standpoint for
international comparisons. This is because GDP/c. is available for virtually
all internationally recognized countries and territories. In some cases, the
data availability goes as far back as the 1950s, and in most cases, the data
begins in the 1960s or 1970s. Using a longer historical comparison across
countries is important because data on economic freedom reforms goes
back to the 1970s as well. Therefore, boosting the time span for the welfare
data also improves the credibility of any study relating welfare to market
reforms, including this one.
Third, GDP/c. provides a useful reference point for the additional
measures of welfare outlined above. By studying how economic freedom
reforms affect living standards and growth rates across countries and over
time, we set up a benchmark against which we can compare the effects economic freedom has on other welfare measures. This kind of comparison
12
P. Stankov
across welfare measures would not be possible in the absence of GDP/c.,
although consumption per capita provides a good substitute.
Consumption per capita, however, is more appropriate as a complement
to GDP/c. rather than a substitute. The reason is that some countries may
experience a take-off period due to high investment rates. As a result,
their welfare would increase if measured by GDP/c. but will be stagnant
if measured by consumption per capita. As these two measures potentially
capture different welfare dynamics over time, it would be interesting to
see if market reforms affect them differently, and if yes, how.
If we agree to include per capita consumption as a welfare gauge, we
also agree with including the other two measures proposed by Jones and
Klenow: life expectancy and income inequality. Despite the fact that average incomes within some countries grow, the way this growth is distributed
across income groups may vary significantly from one country to the next.
This will not only lead to rising within-country income inequality, but
will also deepen global income disparities. In turn, as we will see in the
last chapter, this may produce undesired political consequences in the long
term.
Influential studies have documented the significant differences in both
life expectancy (Becker et al. 2005; Peltzman 2009) and income inequality (Piketty 2014; Piketty and Saez 2014), among others, across countries
and over time. Therefore, both of these measures are well suited to complement GDP/c. and consumption per capita as measures of welfare. The
measures discussed by Jones and Klenow which I leave out of this study
for data availability reasons are leisure and environmental quality. These
two indicators could perhaps be incorporated in future empirical studies
of how welfare depends on market reforms. The literature on this dependence is reviewed next.
2.2
Theories and Evidence on How Reforms
Affect Welfare
Economists around the world have long been working to model the
relationship between economic freedom reforms and changes in welfare.
A recent broad review of the literature is produced by Hall et al. (2015).
2 Contemporary Views on Welfare and Reforms
13
Most studies focus on income and growth, and their dependence on
various institutional determinants, including the elements of economic
freedom. For example, Açemoglu et al. (2005) review a set of historical
examples and develop a theory of dynamic institutional change in which
political power and economic resources are key in further development of
market-friendly property rights and other institutions. They put forward
the argument that “economic institutions encouraging economic growth
emerge when political institutions allocate power to groups with interests
in broad-based property rights enforcement, when they create effective
constraints on power-holders, and when there are relatively few rents to
be captured by power-holders” (p. 385). That is why, they assert, efficient
institutions stand at the foundation of modern economic growth.
Alfonso-Gil et al. (2014) provide a very long-term presentation of how
liberties in general correlate with economic growth for a sample of 149
countries between 1850 and 2010. They present dynamic panel data evidence that, in the long term, civil liberties are positively associated with
economic growth. As much as the long-term picture is informative, it does
not allow inclusion of other potentially important institutional factors for
growth. By shortening the time span, other authors do exactly that. For
example, Fabro and Aixalá (2012) study a sample of 79 countries
between 1976 and 2005. This study provides evidence that economic
freedom, civil liberties and political rights “are important for economic
growth either through a better allocation of resources or, indirectly,
through the stimulation of investment in physical and human capital”
(p. 1059). A methodologically improved treatment of the relationship is
offered by Faria and Montesinos (2009). Rather than running simple OLS
regressions, they provide instrumental variable estimations in which more
economic freedom has a causal impact on growth and development.
This is in line with many previous findings in the empirical literature,
e.g. Gwartney et al. (2004), Nyström (2008), Mijiyawa (2008), among
others. Their results imply that, based on the empirically established positive link between economic freedom, capital accumulation, entrepreneurship, and growth, policy makers need to pursue a policy agenda of raising
economic freedom, including improving property rights.
Based on the empirical studies above, it is expected that the institutions of economic freedom would improve resource allocation and would
14
P. Stankov
therefore help capital accumulation. As a result, they would also raise living standards and may also accelerate growth, as the earlier evidence by
Assane and Grammy (2003), de Haan and Sturm (2000), Doucouliagos
and Ulubasoglu (2006) and Justesen (2008) suggests. However, better
resource allocation and capital accumulation alone are not sufficient to
spur growth, according to Hall et al. (2010). By developing a growth
theory in which capital productivity and allocation depend on local institutions, they conclude that “increases in physical and human capital lead
to output growth only in countries with good institutions. In countries
with bad institutions, increases in capital lead to negative growth rates
because additions to the capital stock tend to be employed in rent-seeking
and other socially unproductive activities” (p. 385).
The above study is one of the many accounts where the intuitively
expected positive effect of institutions and of economic freedom on welfare is jeopardized. For example, Xu and Li (2008) provide additional
evidence on the effect based on data from 104 countries between 1972
and 2003. They conclude that the expected positive effect of economic
and political freedom on growth is “realized and detectable at later stages of
social and economic development” (p. 183). Babecký and Campos (2011)
also document a “remarkable variation” in the effects of overall reforms
on growth by conducting one of the largest meta-studies in the reformgrowth literature. Campos and Horváth (2012) explain the variations in
the reform estimates by how the reform indices are measured in the first
place.
Irrespective of how the freedom indices are measured, it will soon
become clear that there is no single economic freedom that affects welfare
in a linear way. This means economic freedom may provide the necessary
conditions for increasing welfare but, more often than freedom advocates
would like to admit, is hardly sufficient to affect growth, consumption,
life expectancy, and income inequality in positive ways in the long run.
This is because various nations adopt different institutions of economic
freedom at different stages of development, and even identical institutions
may lead to very different welfare implications. Merlevede (2003), among
others, finds that an economy closer to a market economy will benefit more
from introducing a market-oriented mechanism. What stands behind the
difference in the effects of those mechanisms is how reformers enforce
2 Contemporary Views on Welfare and Reforms
15
newly adopted rules and norms over time. It is relatively easy to transplant
institutions, but then adherence to them makes the welfare difference,
according to Crafts and Kaiser (2004).
Further studies narrow down the empirical focus on specific economic
freedom measures. For example, Rode and Coll (2012) identify areas of
economic freedom which matter more for growth than others. They also
identify reforms which could potentially have a long-lasting effect on
growth, and others which exert only a short-lived impact. They conclude that improving the legal structure and the security of property
rights has a long-lasting positive effect on growth. At the same time,
according to the authors, the size of government and labor market regulations have an inverse relationship with growth, at least in the short term.
Williamson and Mathers (2011) also test for the significance of the economic freedom variables, but add another possibly important dimension
to the growth regressions—the impact of culture. They conclude that culture is important for growth, but once economic freedom is taken into
account, the impact of culture is gradually diminished. This suggests a
plausible supremacy of economic freedom over culture in igniting economic growth.
Economic growth has been shown to be positively related to economic
freedom in general on a panel of countries by Wu and Davis (1999). This
early evidence has spurred a considerable attention to the overall relationship between freedom and growth. For example, Karabegovic et al. (2003)
study the within-country evidence of how economic freedom affects the
level and growth of economic activity based on 10 Canadian provinces
and 50 US states. They conclude that economic freedom is positively
associated with both at the state level. Their results are confirmed later
by Murphy (2016) and Barnatchez and Lester (2017). Paldam (2003)
presents the cases of the five Southeast Asian countries that have managed
to raise themselves out of poverty since the 1950s: Japan, Hong Kong, Singapore, South Korea, and Taiwan. He finds that virtually all five countries
have adopted economic freedom reforms on their way to becoming rich.
Bengoa and Sanchez-Robles (2003) review the Latin American evidence, and Fidrmuc (2003), Kenisarin and Andrews-Speed (2008) and
Peev and Mueller (2012) do the same for Central and Eastern Europe
(CEE). All three studies support the previous findings of a positive
16
P. Stankov
relationship between freedom and income levels and growth. Bengoa and
Sanchez-Robles (2003, p. 529) add that the “host country requires, however, adequate human capital, economic stability and liberalized markets
to benefit from” increased levels of overall economic freedom.
The dependence of other welfare measures on economic freedom has
also been extensively studied. Carter (2007) examines evidence of the role
of economic freedom in income inequality dynamics. Based on a sample of
39 countries totaling 104 observations, he finds support for the hypothesis
that economic freedom reduces income inequality. However, the effect is
found to be different across different levels of economic freedom, which
means the effect may be nonlinear.
This is confirmed by Apergis (2015) and Apergis and Cooray (2017),
who provide more recent evidence on the effect of economic freedom on
income inequality. For low levels of economic freedom, raising freedom
increases inequality, while for high levels of freedom, introducing further reforms makes economies more equal. An early attempt to generalize
the argument of a non-monotonic impact of property rights and other
institutions on welfare was carried out by Morris and Adelman (1989).
They were among the first to conclude that institutions are indeed very
important at early stages of development, but the way institutions and the
economic dynamics interact is very different across various development
stages, a result which was later confirmed by Xu and Li (2008).
For example, for some regions of the world, there is conclusive evidence that market reforms raise income inequality. The evidence for
Africa is provided by Enowbi Batuo and Asongu (2015). This is, perhaps because most African countries have low levels of economic freedom
in the first place. The evidence is consistent with that of Apergis (2015).
Bennett and Vedder (2013) examine US state data between 1979 and
2004. Their data demonstrates the non-monotonic relationship between
economic freedom and income inequality. They add evidence that even
within a single country the relationship can have an inverted U-shape.
Consistent with previous evidence, they also find that states with a higher
initial level of economic freedom decrease income inequality more than
states with lower initial levels of freedom. In addition, they estimate that
furthering market-oriented reforms can produce higher income inequality for the US states with lower initial levels of economic freedom.
2 Contemporary Views on Welfare and Reforms
17
As will be demonstrated in this book, the evidence based on a longer
time span and international data is also mixed, as has been previously
shown by McCleery and Paolis (2008).
The literature above has demonstrated that an overall nonlinear association between economic freedom and welfare exists. This is confirmed
for each of the five measures of economic freedom as well. In theory,
government intervention has an ambiguous effect on growth. Barro
(1990) derives an augmented endogenous growth model with government
services. As predicted by the crowding out effect, his paper concludes that
government consumption expenditures reduce growth and saving, while
productive government expenditures generally increase them, at least in
the short run. Bajo-Rubio (2000) generalizes Barro’s argument and concludes that, indeed, the link between per capita growth and the size of
government is non-monotonic.
A plausible reason is outlined by Anshasy and Katsaiti (2013). They find
that the size of government rarely matters for growth, but the degree of
procyclicality does. They also take the degree of procyclicality as a measure
of the quality of fiscal policy management. In other words, they conclude
that it is not the size but the quality of government that matters for welfare. A further related explanation for this non-monotonicity is offered by
Cooray (2011). The author finds that the quality of government is positively correlated with financial sector development, which in turn matters
for growth. At the same time, larger governments reduce the efficiency of
the financial sector.
Larger governments are also associated with more corruption, especially in developing economies. This is found, for example, by Kotera
et al. (2012), who study this relationship for both developing and developed economies. Their sample consists of 82 countries and runs from
1995 to 2008. They find that “government size can lead to a decrease
in corruption if the democracy level is sufficiently high and, in contrast,
can lead to an increase in corruption if it is too low” (p. 2340). Therefore,
another plausible explanation for the nonlinear effect of the size of government on welfare is that, perhaps, voters in older democracies can tolerate
larger governments because their governments provide sufficient quality of
services for both citizen and businesses. As a result, despite the larger government, growth is supported in well-developed democracies. However,
18
P. Stankov
in underdeveloped countries and in new democracies, larger governments
are used to, among other things, allocate resources from private businesses to political insiders and vice versa. At the same time, significantly
improving the quality of public services is not high in the priorities list of
the governments in underdeveloped countries and in new democracies.
As this leads to a significant crowd-out effect, in those countries larger
governments do not lead to higher growth.
This logic is supported by additional evidence from Guseh (1997),
Wu et al. (2010) and Yamamura (2011). Guseh (1997) differentiates the
effect of government size on growth across economic and political systems.
He finds that “growth in government size has negative effects on economic
growth, but the negative effects are three times as great in nondemocratic
socialist systems as in democratic market systems” (p. 175). The evidence
by Wu et al. (2010) is also mixed. They observe that larger governments
increase growth, but not at lower levels of development. In support of this
evidence, Yamamura (2011) concludes that larger government size leads
to lower capital accumulation in non-OECD countries, but does not
lead to significantly lower capital accumulation in the OECD countries
themselves.
Contrary to that evidence, Fölster and Henrekson (2001) and Dar and
Amirkhalkhali (2002), among others, detect a universal crowd-out effect.
They conclude that the size of government has a negative correlation
with growth not only for developing but also for developed countries,
including the OECD. However, Agell et al. (2006) respond with criticism
to Fölster and Henrekson (2001). Agell et al. (2006) believe that in a crosscountry setting it is very difficult to find any robust effect of government
intervention on growth. This conclusion is supported in this book, which
produces additional evidence of a non-robust effect of government size on
growth and other welfare dynamics.
Larger governments may also reduce output volatility, which can also
affect other welfare dynamics. This is suggested by Fatás and Mihov (2001)
based on a sample of 22 OECD countries and 50 US states, and by
Jetter (2014) based on a larger panel of 90 countries. Fatás and Mihov
(2001) find that “a one percentage point increase in government spending
relative to GDP reduces output volatility by eight basis points” (p. 3).
Jetter (2014) adds to that evidence and concludes that governments play
2 Contemporary Views on Welfare and Reforms
19
a different role for stabilizing the economy depending on their political
regimes. In democracies, output volatility is predictive of lower subsequent
growth, while in autocratic regimes governments manage to carry forward a growth-enhancing political agenda after episodes of output volatility. Carmignani et al. (2011) go one step further and outline areas of
government intervention which may be beneficial for mitigating output
volatility. Those, according to the authors, “include domestic political
institutions, de facto central bank independence and a stable nominal
exchange rate regime” (p. 781).
Overall, there is no single recipe for how much government is optimal for both output growth and longer-term stability. In democracies, it
seems the optimal size of government is different from that in autocracies.
The literature also suggests that in developed economies more government may lead to higher growth, while in less-developed economies this
is not the case. At the same time, there is evidence that in well-established
democracies, more government means poorer responses to output volatility, while stronger governments can potentially mitigate output volatility
in non-democratic societies. Ultimately, as suggested by Facchini and
Melki (2013), the optimal size of government is not universal and would
be country-specific.
Similar conclusions can be reached for the second element of economic
freedom: property rights (PRs). Some studies identify the origins of
improved property rights, whereas others focus on the link between better
property rights and welfare. Lagerlöf (2013, p. 312) offers one explanation
for the origin of better property rights: “faster technological progress can
lead to a decline in violence and improved property rights protection,
similar to the path followed by Europe” over the course of economic
history. Sonin (2003) studies those mechanisms for Russia to explain why
a country which becomes a market-oriented economy may quickly turn
its policy agenda to a bad equilibrium: The elite chooses poorly protected
PRs and substitutes them with privately protected PRs, a story advanced
also by Açemoglu et al. (2005).
A paper by Sunde et al. (2008) offers an explanation for the reasons
democratic institutions produce various qualities of rule of law and PRs.
They claim that democracy leads to better rule of law only when income
inequality is low. As this book shows, income inequality rose differently
20
P. Stankov
across Central and Eastern European nations during their transitions
since 1989. In turn, the difference in inequality expansion might be able
to explain why almost identical institutional reforms at the onset of the
transition have led to dramatically different institutional qualities some
25 years later.
As Ogilvie and Carus (2014, p. 403) point out, “economic history has
been used to support both the centrality and the irrelevance of secure
property rights to growth, but the reason for this is conceptual vagueness”, an issue also discussed by Haggard and Tiede (2011). Both teams of
researchers call for a much more detailed understanding of the structure
of property rights before the effects of property rights on welfare can be
disentangled. Further, Haggard and Tiede (2011) claim that the effects
of PRs protection are ultimately uncertain, though the property rights
literature does sheds light on those effects.
One example of a theoretical work to study the effects of property rights
on welfare is that of Gradstein (2004). He asserts that higher levels of
economic development lead to the establishment of better property rights
and also that stronger property rights reinforce economic development and
welfare. Therefore, we can safely assume that the level of PR protection is
endogenous to growth and welfare in general.
To understand the impact of PRs in a more detailed way, Kapeliushnikov et al. (2013) take on some of the PR measurement issues and find
that PRs are important for generating positive growth in a transition economy, provided other institutional factors are already in place. Voigt and
Gutmann (2013, p. 66) bring a bit more detail into those factors and
advance the argument that “the mere promise of secure property rights is
unlikely to have any effects unless accompanied by some commitment to
enforce these rights.” According to the authors, a credible commitment
device is, for example, an independent judiciary that has the constitutional
rights to enforce protection of PRs. In two related papers, they extend the
argument by distinguishing between de jure and de facto independent
judiciaries, and then testing for their effects on growth. Feld and Voigt
(2003) do the first part of the analysis, while in a later work they find
that de jure judicial independence (JI) “is not systematically related to
economic growth, whereas de facto JI is highly significantly and robustly
correlated with growth” (Voigt et al. 2015, p. 197).
2 Contemporary Views on Welfare and Reforms
21
A significant part of the more recent literature deals with the
growth effects of intellectual property rights (IPRs) protection. Mondal
and Gupta (2008) present a general equilibrium model in which strengthening IPRs has a mitigating effect on unemployment only under certain
conditions and would normally have a negative effect on innovation. At
about the same time, Furukawa (2007) extends the endogenous growth
theory literature with IPRs. His conclusion is that strengthening IPRs
does not necessarily generate a positive effect on growth, especially in a
rapidly integrating world. Gancia and Bonfiglioli (2008) build on this line
of argumentation to find that, indeed, if a weak-IPR country is integrating
with a strong-IPR country, then the innovative activity in the strong-IPR
country declines.
Another factor which may contribute to the differences in the PR effects
across countries is trade. Early evidence that more open economies benefit
more from improving property rights has been published by Gould and
Gruben (1996). Dinopoulos and Segerstrom (2010) build on this evidence
with a model of North–South trade, in which improving IPRs in the South
leads to a permanent increase in wages, employment, and innovation
activity in the South. At the same time, the North does not benefit much
from improving IPRs in the South. However, it would be interesting to
see how this models fares against evidence of winners and losers from the
Great Recession. This is because, if we look at the European experience
per se, it seems that growth in the technologically less-developed South,
not the advanced North, has been lower in the aftermath of the Crisis.
To this end, Manca (2010) presents evidence that the strengthening PRs
has the potential to slow down the income convergence process, especially
for countries far from the technology frontier, because much of the innovation in those countries is accomplished through imitation. However,
stronger PRs raise the costs of imitation. Then, if a country lacks the
capacity for substantial product or process innovation, stronger PRs will
slow down their convergence.This logic is consistent with Chu et al. (2014,
p. 239) who develop an intuitive explanation for the reasons IPRs affect
different economies differently. They bring forward the argument that
“optimal intellectual property rights (IPR) protection is stage-dependent.
At an early stage of development, the country implements weak IPR
protection to facilitate imitation. At a later stage of development, the
22
P. Stankov
country implements strong IPR protection to encourage domestic innovation. Therefore, the growth-maximizing and welfare-maximizing levels
of patent strength increase as the country evolves towards the world technology frontier.” Jordan (2001) goes one step further and is among the first
to advocate total removal of IPRs. He argues that “protections often taken
for granted—patents, copyrights, and other intellectual property rights—
are largely unknown or are ineffective in many places in the world today.
Without such protections, incentives for creative talents to design and
develop new products and services are substantially weakened” (p. 20).
Apart from output growth and income per capita growth, other elements of welfare are also found to depend on property rights. For example, Chu and Peng (2011) set up a growth model with R&D and income
inequality. The model predicts that improving IPRs will lead not only to
higher growth, but also to greater inequality. Jayadev and Bowles (2006)
support this conclusion with their own empirical evidence of strengthening property rights and ensuing increases in inequality.
Inequality aside, Kwan and Lai (2003) develop a theory of endogenous growth with IPR and, similarly to others, determine an optimal
level of IPRs. They conclude that stronger IPRs can lead to increases in
consumption. The empirical effects of strengthening IPRs on innovation
are studied by Krammer (2009) and Ang (2011). The authors find positive and significant effects of improving IPRs on innovation in transition
economies.
The theories of property rights may also help to explain why some
countries experience resource curses, while other resource-rich countries
turn their natural resource abundance into a welfare blessing. López and
Schiff (2013) develop a theory in which PRs have a special role to play in
resource-rich economies with weakly defined property rights. They reach
the conclusion that, with weakly defined PRs, the economy will quickly
reach an overuse of the resource, resulting in a resource curse. Improving property rights, however, also improves the chances of the country
to benefit from the natural resource endowment. Farhadi et al. (2015)
find empirical evidence for this theory. On a sample of 99 countries,
they demonstrate that the resource curse can be turned into a blessing by
introducing more economic freedom. In a more detailed argument, Boschini et al. (2013) reveal which elements of economic freedom have the
2 Contemporary Views on Welfare and Reforms
23
potential to turn the curse into a blessing so that resource-abundant countries benefit from their natural endowments. They reveal that improving
property rights, as measured in the International Country Risk Guide, has
the potential to reverse the resource curse and improve welfare.
By setting up a theoretical framework, Chu et al. (2012) demonstrate
that property rights can not only lead to improved growth but also mitigate
growth volatility. They also compare the model predictions against US
data and find that about 10% of growth volatility can be explained by
improving (intellectual) property rights. Perhaps the entire set of PRs has
a more potent impact on reducing growth volatility. Indeed, weaker PR
protection is found to have an overall negative effect on output stability
by Barbier (2004). He concludes that weaker PRs contribute to a more
frequent incidence of “boom-and-bust” cycles in Latin America.
Therefore, we can conjecture that, similar to other areas of economic
freedom, PRs have a nonlinear relationship with welfare. Trade and monetary stability also affect welfare in a nonlinear way.
After Friedman and Schwartz (1963) gained mainstream academic and
policy attention, sound money has become widely accepted as a prerequisite for growth and output stability, and through growth, as a condition
for raising welfare over time. Monetary stability then penetrated policy
agendas across the globe. This includes maintaining price stability as the
primary role of central banks in contemporary economies, including the
Eurozone, the UK, Australia, New Zealand, and more recently, to a major
extent, the USA. Among others, Gwartney et al. (2001, p. 183) argue
that monetary stability in the early 1980s and later has been at the core of
achieving “strong and steady economic growth” in the USA, which provides a natural platform for establishing a policy agenda for the rest of the
world. Bordo (2000) also reviews the role of sound money in the economy
by supporting the views of Friedman and Schwartz. He finds that strong
price stability has a positive impact on the resilience of an economy to
deal with financial shocks, which contributes positively to an economy’s
welfare.
Contemporary research into the role of sound money has also focused
on its impact on other aspects of welfare. Bjørnskov and Foss (2008)
provide empirical support for the hypothesis that inflation stability raises
entrepreneurship levels, while Feldmann (2007) examines evidence from
24
P. Stankov
87 countries between 1980 and 2003 on its role in reducing unemployment. Both studies conjecture that inflation stability increases welfare.
An additional line of research examines the impact of political and
economic freedom on sound money. For example, Aisen and Veiga (2008)
cover a sample of 160 countries between 1960 and 1999 to examine the
relationship between political instability and central bank independence
on price stability. They find that the more politically unstable a country
is, and the less independent the central bank is, the more volatile the
inflation rates are. As we will later see, sound money is one of the most
robust factors in welfare improvement.
Ho and Jorgenson (1994) review the literature on trade liberalization
and its effect on the USA. They build a theory to explain the positive association, and then test the significance of the effects of trade reforms in the
USA. They find a significantly higher positive effect of trade reforms than
previously expected due to previously ignored dynamic effects of trade.
Baldwin (1992) also builds a dynamic growth model with trade. He finds
that in the medium-term large dynamic welfare gains from trade liberalization due to capital accumulation exist. Willenbockel (1998) extends
the conclusions from this model and argues that the medium-term welfare
gains are actually preceded by significant losses due to a drop in aggregate
investment and income after trade liberalization.
Numerous other empirical studies have scrutinized trade reform propositions. Berggren and Jordahl (2005) establish a positive correlation
between trade openness and growth by questioning the previous evidence of surprisingly negative effects provided by Carlsson and Lundstrom
(2002). Berggren and Jordahl (2005) find the negative effects to be due
to one of the sub-components of the freedom to trade indices. They also
add that Carlsson and Lundstrom’s negative effect is not robust to adding
newer economic freedom data.
Trade is also found to have a positive impact in a number of studies on
developing countries, e.g. Rutherford and Tarr (2002) and Jinjarak et al.
(2013). Rutherford and Tarr (2002) develop a growth theory with trade
liberalization. They decidedly support the conclusion that trade liberalization positively affects welfare. Jinjarak et al. (2013) identify the exogenous
component of trade reforms by the timing of the trade adjustment agreements between recipient countries and the World Bank. Jinjarak et al.
2 Contemporary Views on Welfare and Reforms
25
(2013, p. 415) claim that “[i]n comparison to a pre-reform period and
to the non-recipient group, the recipient countries registered 0.2 percent
higher growth of real GDP per capita, 5.0 percent higher import growth,
and 2.5 percent higher export growth over a period of three to five years
after trade reform.”
Early evidence of those positive effects in a developing country is produced by Krishna and Mitra (1998). They study the 1991 wave of trade
liberalization in India and conclude that trade reforms did modestly contribute to an increase in welfare in India. They also document increases in
competition in the liberalized industries, as well as increase in productivity growth, which is key to raising income levels over the long term. The
evidence by Alessandrini et al. (2011) sides with this argument. They find
that the Indian trade liberalization reforms have spurred industry specialization and have also contributed to the growth of India’s medium- and
high-tech industries.
Trade liberalization has also contributed to income convergence of postWar Europe. This conclusion is reached on European data by Ben-David
(2001) and is preceded by theoretical work by Walz (1998). A positive
impact of trade reforms on welfare is also revealed by Naito (2012), who
builds a growth theory with trade and endogenously determined trade
status. The paper concludes that a reduction in trade costs, even in one
trading partner, raises welfare in both trading countries. The author also
supports this conclusion with a number of empirical tests.
Further studies on the effects from trade reforms qualify the above
theoretical and empirical conclusions. Christiansen et al. (2013, p. 347)
contend that “[d]omestic financial reforms and trade reforms are robustly
associated with economic growth, but only in middle-income countries. In
contrast, there is no evidence of a systematic positive relationship between
capital account liberalization and economic growth. [...] Sufficiently developed property rights are a precondition for reaping the benefits of financial
and trade reforms”. Ahmed (2013) also agrees that in order to work for
growth, economic freedom reforms, including trade and financial liberalization, need to be set up in an environment of well-protected property
rights and complemented by high levels of human capital. Human capital
is also found by Gibson (2005) to be a crucial lever to place a country on
a growth trajectory after trade liberalization.
26
P. Stankov
Apparently, most African countries lack those conditions, because the
more recent findings by Menyah et al. (2014) also confirm that financial and trade liberalization reforms did not exert a significant impact on
growth in 21 African countries. Yet, in an earlier study of trade reform
effects in 12 sub-Saharan economies, Onafowora and Owoye (1998) document a significant positive effect of trade liberalization on growth in most
reforming countries.
The mixed evidence on the effects of financial and trade reforms on
welfare goes at least as far back as studies by Greenaway et al. (1997)
and Diao et al. (1999). Greenaway et al. (1997) study the effect of trade
reforms on economic growth in a number of developing countries and
conclude that trade reforms after 1985 had a negative impact on growth
for that particular set of countries around the wave of trade liberalization
in the 1980s. Diao et al. (1999) also argue that the reform may have a
negative welfare implications in the long run, whereas the effects in the
short run are mostly positive. Even the short-run positive effects are not
guaranteed, according to Dijkstra (2000).
Additional, more recent, empirical support for the nonlinear impact of
trade liberalization in developing countries is published by Caselli (2013).
Their conclusion is also supported by a number of case studies on developing and emerging economies, including Argentina (Bas 2012), Bolivia
(Jenkins 1997), Korea in its rapid development stage between 1966 and
1988 (Kim 2000; Pyo 1990), Malawi (Mulaga and Weiss 1996), Sri Lanka
(Rahapakse and Arunatilake 1997), Tunisia (Belloumi 2014) and Zimbabwe (Mehlum 2002). In principle, the authors argue, trade reforms
should be able to raise firm-level productivity and also capital accumulation. However, the actual effects of the trade reforms would be uncertain. The literature finds three possible explanations. It is either: (i) the
imprecise way productivity or other outcome variables are measured or
(ii) because the reform is not credible enough in the long term to induce
sufficiently high expansion of capital accumulation or (iii) because “liberalisation raises or lowers growth depending upon the initial level of the
barrier” to trade (Baldwin and Forslid 1999, p. 797).
Current levels of economic freedom may indeed hold the key to generating positive welfare gains from trade reforms. According to Freund and
Bolaky (2008), when a country implements trade reforms, how supportive
2 Contemporary Views on Welfare and Reforms
27
the local business environment is for starting a new business matters more
for growth than financial conditions. The reason is that when trade opens,
there is often a large cross-industry reallocation of resources. However, in
different countries, this resource reallocation will ultimately depend on
how easy it is to start and close businesses. Therefore, the business regulations, as well as other forms of government intervention (Dinopoulos and
Unel 2011), and excessive competition on the input markets (Goo and
Park 2007) might play a key role in maximizing the welfare gains of trade
reforms.
Trade reforms have a significant impact on increasing inequality as
well. The intuition is well developed by Carneiro and Arbache (2003).
They build a general equilibrium model of the impact of trade reforms,
and find that trade reforms may benefit skilled workers more, especially
in export-oriented sectors. Within-country evidence also supports that
view. By studying Mexico’s regional disparities before and after entering
NAFTA, both Chiquiar (2005) and Nicita (2009) find that NAFTA did
not contribute to narrowing the gap in regional disparities. Similar to previous research on the country-level (Cragg and Epelbaum 1996; Harrison
and Hanson 1999), regions within Mexico which benefited most from the
trade reform were those initially endowed with sufficient levels of human
and physical capital, including adequate infrastructure. Iacovone (2012)
supports this view with firm-level data. He concludes that “more advanced
firms benefited disproportionately more from the liberalization” (p. 474).
Other studies on the effect of trade reforms on income inequality review
the experience of Chile (Bussolo et al. 2002) and Brazil (Castilho et al.
2012). Bussolo et al. (2002) reveal that one of the channels through
which trade reform affects inequality is the degree of local labor market regulations, while Castilho et al. (2012) confirm earlier studies for
Mexico which document increasing regional income disparities after the
trade reform. Gelan (2002) expands this view with a calibration exercise for Ethiopia. The author also notices that trade effects on growth
will ultimately depend on the local product and labor market regulations.
With more flexible underlying regulations, the country will experience a
positive impact of trade liberalization. However, with rigid labor market
conditions, a “trade reform adversely affects overall economic growth”
(p. 707). Acharya (2011) also studies the effects of trade reform on
28
P. Stankov
inequality in Nepal, and Naranpanawa and Arora (2014) do the same
for India. Both studies find that trade reforms benefit the rich more than
the poor, thereby exacerbating income inequality in developing countries which undertake trade reforms. Helpman (2016) provides recent
evidence on the relationship and broadly confirms that trade has contributed to rising inequality across countries but perhaps not so strongly
within countries, as the above case studies suggest.
The effect of trade on income inequality may be positive but also only
short-lived, according to Harris and Robertson (2013). They build a theory of open economy growth with trade reform. They do acknowledge
the negative effect of the reform on income inequality, but also call for a
dynamic viewpoint when assessing the effects. In the long run, the authors
argue, significant capital and skill accumulation would prevail over the
short-lived negative effects on inequality. To support this dynamic viewpoint, they calibrate the model for China and India. Evidence from Brazil
and Mexico also supports the view that the effects of the reforms may actually appear negative due to mismeasurement of the dependent variables
(de Carvalho Filho and Chamon 2012).
Other trade models are in disagreement with the conclusions of
Harris and Robertson (2013) and de Carvalho Filho and Chamon (2012).
A recent work by Auer (2015) builds a model of heterogeneous agents who
invest in certain types of skills after trade reforms. Their results demonstrate that “while the static gains from trade may lead to convergence, the
dynamic gains from trade occur to initially rich countries, thus leading to
cross-country divergence of income and welfare” (p. 107). Later in this
book, additional evidence is produced which sides with the hypotheses
that trade liberalization increases income per capita but at the same time
raises income inequality.
We can safely conclude that no single economic freedom so far has
exerted a uniform effect on welfare. This is valid not only for growth
and inequality but also for other aspects of welfare, e.g. subjective wellbeing and the human development index. Gehring (2013) studies the
effect of economic freedom in general on subjective well-being in a panel
of 86 countries between 1990 and 2005. The author finds a positive
effect on subjective well-being, especially from strengthening property
rights, improving the index of sound money, and deregulation. However,
2 Contemporary Views on Welfare and Reforms
29
country-fixed effects moderate the effects, which means that, other than
reforms, unobserved country characteristics may be even more important
in explaining not only objective welfare but also subjective well-being.
Indeed, the author elaborates that “societies that are more tolerant and
have a positive attitude toward the market economy profit the most”
from deepening market-oriented reforms (p. 74). Graafland and Compen
(2015) extend this evidence on a sample of 120 countries. They find that
various aspects of economic freedom affect life satisfaction differently.
Specifically, they conclude that “life satisfaction is positively related to the
quality of the legal system and negatively related to small government size”
(p. 789).
Davies (2009) studies how the size of government can affect another
measure of welfare: the Human Development Index (HDI). It turns
out the size of government does not play a linear role for the HDI either.
The author also discusses the optimal size of government with respect to
the HDI and argues that it may be country-specific. Designing countryspecific and time-specific policies could also be key to a growth-enhancing
policy agenda in virtually all reform areas, according to Huynh and JachoChávez (2009). Using nonparametric estimation methods, they also find
that the relationships between economic freedom reforms and growth are
highly nonlinear. This is valid also for economic regulation.
On the one hand, deregulation reduces the rents that regulation creates for workers, incumbent producers, and service providers. This view
has gained widespread popularity among academics and policy makers
alike since the seminal works by Stigler (1971), Posner (1974) and Peltzman (1976) contributed to the understanding of the political economy of
regulation. On the other hand, deregulation allows newly created competition on the product, labor, and capital markets to determine the winner of those rent transfers. Thus, by spurring productivity and efficiency
gains (Winston 1993), economic deregulation ultimately contributes to
an overall increase in economic growth. Additional growth is achieved
primarily through increased employment and real wages (Blanchard and
Giavazzi 2003), which affect both production and consumption, and
through increased investment (Alesina et al. 2005).
However, a more recent take on the efficiency gains from deregulation
in the developing world provides a word of caution. The key contention
30
P. Stankov
in this newer line of literature is that deregulation influences different
economies differently, depending on their position on the technology ladder and on the quality of their institutions. For example, Açemoglu et al.
(2006) claim that certain restrictions on competition may benefit technologically less-developed countries, while Estache and Wren-Lewis (2009)
find that the optimal regulatory policies in developed and developing
countries are different because of differences in the overall institutional
quality of those countries.
In addition, Aghion et al. (2007) use industry-level data to demonstrate
that within each economy, industries closer to the technology frontier
will be affected more by deregulation. They will innovate more than the
backward industries in order to prevent entry by new firms. As a result,
countries closer to the technology frontier benefit more from deregulation.
The alleged benefits of economic deregulation in many industries have
prompted more focused debates on the growth effects of specific types
of reforms, such as capital, labor, and product-market deregulation. All
of these debates are, and perhaps will always be, inconclusive about the
ultimate effects of deregulation on welfare. The results in this book confirm that effects of deregulation on welfare are not always significant, and
although deregulation did raise income per capita, it also raised income
inequality.
Perhaps the best summary of how policy makers design reforms and how
reforms affect growth is given by Rodrik (2005, p. 967): “...[P]rotection of
property rights, market-based competition, appropriate incentives, sound
money, and so on—do not map into unique policy packages. Reformers
have substantial room for creatively packaging these principles into institutional designs that are sensitive to local opportunities and constraints.
Successful countries are those that have used this room wisely.”
In what follows, I review the patterns of large-scale economic freedom
reforms since 1970, with an emphasis on how they differ before and after
the onset of the Great Recession. Then, I provide evidence on the welfare
implications of those reforms. The existing literature sets the stage for
those results very well: They will still be far from conclusive. Trade reforms
and deregulation will raise income per capita, but will also swell income
inequality. Protection of property rights and monetary policy stability will
also produce more income per capita but, unlike trade and deregulation
2 Contemporary Views on Welfare and Reforms
31
reforms, will shrink income inequality. The least eventful relationship is
between the size of government and welfare. In most of the estimations,
it will be statistically insignificant.
References
Açemoglu, D., P. Aghion, and F. Zilibotti. 2006. Distance to frontier,
selection, and economic growth. Journal of the European Economic Association
4 (1): 37–74.
Açemoglu, D., S. Johnson, and J.A. Robinson. 2005. Chapter 6, Institutions as
a fundamental cause of long-run growth. In Handbook of economic growth, ed.
P. Aghion and S.N. Durlauf, vol. 1, Part A of Handbook of Economic Growth,
pp. 385–472. Elsevier.
Acharya, S. 2011. Making unilateral trade liberalisation beneficial to the poor.
Socio-Economic Planning Sciences 45 (2): 60–71.
Agell, J., H. Ohlsson, and P.S. Thoursie. 2006. Growth effects of government
expenditure and taxation in rich countries: A comment. European Economic
Review 50 (1): 211–218.
Aghion, P., A. Alesina, and F. Trebbi. 2007. Democracy, technology, and growth.
Working Paper 13180, National Bureau of Economic Research.
Ahmed, A.D. 2013. Effects of financial liberalization on financial market development and economic performance of the SSA region: An empirical assessment.
Economic Modelling 30: 261–273.
Aisen, A., and F.J. Veiga. 2008. Political instability and inflation volatility. Public
Choice 135 (3–4): 207–223.
Alesina, A., S. Ardagna, G. Nicoletti, and F. Schiantarelli. 2005. Regulation and
investment. Journal of the European Economic Association 3 (4): 791–825.
Alessandrini, M., B. Fattouh, B. Ferrarini, and P. Scaramozzino. 2011. Tariff liberalization and trade specialization: Lessons from India. Journal of Comparative
Economics 39 (4): 499–513.
Alfonso-Gil, J., M. Lacalle-Calderón, and R. Sãnchez Mangas. 2014. Civil liberty
and economic growth in the world: A long-run perspective, 1850–2010. Journal
of Institutional Economics 10 (3): 427–449.
Ang, J.B. 2011. Financial development, liberalization and technological deepening. European Economic Review 55 (5): 688–701.
Anshasy, A.A.E., and M.-S. Katsaiti. 2013. Natural resources and fiscal performance: Does good governance matter? Journal of Macroeconomics 37: 285–298.
32
P. Stankov
Apergis, N. 2015. Economic freedom and income inequality: Further
evidence from 58 countries in the long-run. Financial Theory and Practice 39
(4): 349–370.
Apergis, N., and A. Cooray. 2017. Economic freedom and income inequality:
Evidence from a panel of global economies-a linear and a non-linear long-run
analysis. Manchester School 85 (1): 88–105.
Assane, D., and A. Grammy. 2003. Institutional framework and economic development: International evidence. Applied Economics 35 (17): 1811–1817.
Atkinson, A.B. 2011. The restoration of welfare economics. American Economic
Review 101 (3): 157–61.
Auer, R.A. 2015. Human capital and the dynamic effects of trade. Journal of
Development Economics 117: 107–118.
Babecký, J., and N.F. Campos. 2011. Does reform work? An econometric
survey of the reform-growth puzzle. Journal of Comparative Economics 39 (2):
140–158.
Bajo-Rubio, O. 2000. A further generalization of the Solow growth model: The
role of the public sector. Economics Letters 68 (1): 79–84.
Baldwin, R.E. 1992. Measurable dynamic gains from trade. Journal of Political
Economy 100 (1): 162–174.
Baldwin, R.E., and R. Forslid. 1999. Incremental trade policy and endogenous
growth: A q-theory approach. Journal of Economic Dynamics and Control 23
(5–6): 797–822.
Barbier, E.B. 2004. Agricultural expansion, resource booms and growth in Latin
America: Implications for long-run economic development. World Development
32 (1): 137–157.
Barnatchez, K., and R. Lester. 2017. The relationship between economic freedom
and economic dynamism. Contemporary Economic Policy 35 (2): 358–372.
Barro, R.J. 1990. Government spending in a simple model of endogeneous
growth. Journal of Political Economy 98 (5): S103–S125.
Bas, M. 2012. Input-trade liberalization and firm export decisions: Evidence
from Argentina. Journal of Development Economics 97 (2): 481–493.
Becker, G.S., T.J. Philipson, and R.R. Soares. 2005. The quantity and quality of
life and the evolution of world inequality. American Economic Review 95 (1):
277–291.
Belloumi, M. 2014. The relationship between trade, FDI and economic growth
in Tunisia: An application of the autoregressive distributed lag model. Economic
Systems 38 (2), 269–287. Symposium: Performance of Financial Markets.
2 Contemporary Views on Welfare and Reforms
33
Ben-David, D. 2001. Trade liberalization and income convergence: A comment.
Journal of International Economics 55 (1): 229–234. Intranational & International Economics.
Bengoa, M. and B. Sanchez-Robles. 2003. Foreign direct investment, economic
freedom and growth: New evidence from Latin America. European Journal of
Political Economy 19 (3): 529–545. Economic Freedom.
Bennett, D.L., and R.K. Vedder. 2013. A dynamic analysis of economic freedom
and income inequality in the 50 U.S. states: Empirical evidence of a parabolic
relationship. Journal of Regional Analysis & Policy 43 (1): 42–55.
Berggren, N., and H. Jordahl. 2005. Does free trade really reduce growth? Further
testing using the economic freedom index. Public Choice 122 (1–2): 99–114.
Bjørnskov, C., and N.J. Foss. 2008. Economic freedom and entrepreneurial activity: Some cross-country evidence. Public Choice 134 (3–4): 307–328.
Blanchard, O., and F. Giavazzi. 2003. Macroeconomic effects of regulation and
deregulation in goods and labor markets. The Quarterly Journal of Economics
118 (3): 879–907.
Bordo, M.D. 2000. Sound money and sound financial policy. Journal of Financial
Services Research 18 (2): 129–155.
Boschini, A., J. Pettersson, and J. Roine. 2013. The resource curse and its potential
reversal. World Development 43: 19–41.
Bussolo, M., A. Mizala, and P. Romaguera. 2002. Beyond Heckscher-Ohlin:
Trade and labour market interactions in a case study for Chile. Journal of Policy
Modeling 24 (7–8): 639–666.
Campos, N.F., and R. Horváth. 2012. Reform redux: Measurement, determinants and growth implications. European Journal of Political Economy 28 (2):
227–237.
Carlsson, F., and S. Lundstrom. 2002. Economic freedom and growth: Decomposing the effects. Public Choice 112 (3–4): 335–344.
Carmignani, F., E. Colombo, and P. Tirelli. 2011. Macroeconomic risk and the
(de)stabilising role of government size. European Journal of Political Economy
27 (4): 781–790.
Carneiro, F., and J.S. Arbache. 2003. The impacts of trade on the Brazilian labor
market: A CGE model approach. World Development 31 (9): 1581–1595.
Carter, J.R. 2007. An empirical note on economic freedom and income inequality. Public Choice 130 (1–2): 163–177.
Caselli, M. 2013. Credit constraints, inequality and the growth gains from trade.
Economics Letters 121 (1): 43–47.
34
P. Stankov
Castilho, M., M. Menéndez, and A. Sztulman. 2012. Trade liberalization,
inequality, and poverty in Brazilian states. World Development 40 (4):
821–835.
Chiquiar, D. 2005. Why Mexico’s regional income convergence broke down.
Journal of Development Economics 77 (1): 257–275.
Christiansen, L., M. Schindler, and T. Tressel. 2013. Growth and structural reforms: A new assessment. Journal of International Economics 89 (2):
347–356.
Chu, A.C., G. Cozzi, and S. Galli. 2014. Stage-dependent intellectual property
rights. Journal of Development Economics 106: 239–249.
Chu, A.C., C.K. Leung, and E. Tang. 2012. Intellectual property rights,
technical progress and the volatility of economic growth. Journal of Macroeconomics 34 (3): 749–756.
Chu, A.C., and S.-K. Peng. 2011. International intellectual property rights:
Effects on growth, welfare and income inequality. Journal of Macroeconomics
33 (2): 276–287.
Cooray, A. 2011. The role of the government in financial sector development.
Economic Modelling 28 (3): 928–938.
Crafts, N., and K. Kaiser. 2004. Long-term growth prospects in transition
economies: A reappraisal. Structural Change and Economic Dynamics 15 (1):
101–118. The Macroeconomics of Accession: Growth, Convergence and Structural Adjustment.
Cragg, M. I., and M. Epelbaum. 1996. Why has wage dispersion grown in Mexico? Is it the incidence of reforms or the growing demand for skills? Journal of
Development Economics 51 (1), 99–116. Interamerican Seminar on Macroeconomics.
Dar, A.A., and S. Amirkhalkhali. 2002. Government size, factor accumulation,
and economic growth: Evidence from OECD countries. Journal of Policy Modeling 24 (7–8): 679–692.
Davies, A. 2009. Human development and the optimal size of government. The
Journal of Socio-Economics 38 (2): 326–330.
de Carvalho Filho, I., and M. Chamon. 2012. The myth of post-reform income
stagnation: Evidence from Brazil and Mexico. Journal of Development Economics
97 (2): 368–386.
de Haan, J., and J.-E. Sturm. 2000. On the relationship between economic
freedom and economic growth. European Journal of Political Economy 16 (2):
215–241.
2 Contemporary Views on Welfare and Reforms
35
Diao, X., T. Roe, and E. Yeldan. 1999. Strategic policies and growth: An applied
model of R&D-driven endogenous growth. Journal of Development Economics
60 (2): 343–380.
Dijkstra, A.G. 2000. Trade liberalization and industrial development in Latin
America. World Development 28 (9): 1567–1582.
Dinopoulos, E., and P. Segerstrom. 2010. Intellectual property rights, multinational firms and economic growth. Journal of Development Economics 92 (1):
13–27.
Dinopoulos, E., and B. Unel. 2011. Quality heterogeneity and global economic
growth. European Economic Review 55 (5): 595–612.
Doucouliagos, C., and M.A. Ulubasoglu. 2006. Economic freedom and economic growth: Does specification make a difference? European Journal of Political Economy 22 (1): 60–81.
Enowbi Batuo, M., and S.A. Asongu. 2015. The impact of liberalisation policies
on income inequality in African countries. Journal of Economic Studies 42 (1):
68–.
Estache, A., and L. Wren-Lewis. 2009. Toward a theory of regulation for developing countries: Following Jean-Jacques Laffont’s lead. Journal of Economic
Literature 47 (3): 729–770.
Fabro, G., and J. Aixalá. 2012. Direct and indirect effects of economic and
political freedom on economic growth. Journal of Economic Issues 46 (4):
1059–1080.
Facchini, F., and M. Melki. 2013. Efficient government size: France in the 20th
century. European Journal of Political Economy 31: 1–14.
Farhadi, M., M.R. Islam, and S. Moslehi. 2015. Economic freedom and productivity growth in resource-rich economies. World Development 72: 109–126.
Faria, H.J., and H.M. Montesinos. 2009. Does economic freedom cause prosperity? An IV approach. Public Choice 141 (1–2): 103–127.
Fatás, A., and I. Mihov. 2001. Government size and automatic stabilizers: International and intranational evidence. Journal of International Economics 55 (1):
3–28. Intranational and International Economics.
Feld, L.P., and S. Voigt. 2003. Economic growth and judicial independence:
Cross-country evidence using a new set of indicators. European Journal of Political Economy 19 (3): 497–527. Economic Freedom.
Feldmann, H. 2007. Economic freedom and unemployment around the world.
Southern Economic Journal 74 (1): 158–176.
Fidrmuc, J. 2003. Economic reform, democracy and growth during postcommunist transition. European Journal of Political Economy 19 (3): 583–604.
Economic Freedom.
36
P. Stankov
Fleurbaey, M. 2009. Beyond GDP: The quest for a measure of social welfare.
Journal of Economic Literature 47 (4): 1029–1075.
Fölster, S., and M. Henrekson. 2001. Growth effects of government expenditure
and taxation in rich countries. European Economic Review 45 (8): 1501–1520.
Freund, C., and B. Bolaky. 2008. Trade, regulations, and income. Journal of
Development Economics 87 (2): 309–321.
Friedman, M., and A.J. Schwartz. 1963. A monetary history of the United States,
1867–1960. Princeton University Press.
Furukawa, Y. 2007. The protection of intellectual property rights and endogenous
growth: Is stronger always better? Journal of Economic Dynamics and Control
31 (11): 3644–3670.
Gancia, G., and A. Bonfiglioli. 2008. North-South trade and directed technical
change. Journal of International Economics 76 (2): 276–295.
Gehring, K. 2013. Who benefits from economic freedom? Unraveling the effect
of economic freedom on subjective well-being. World Development 50: 74–90.
Gelan, A. 2002. Trade liberalisation and urban-rural linkages: A CGE analysis
for Ethiopia. Journal of Policy Modeling 24 (7–8): 707–738.
Gibson, B. 2005. The transition to a globalized economy: Poverty, human capital
and the informal sector in a structuralist CGE model. Journal of Development
Economics 78 (1): 60–94.
Goo, Y., and H. Park. 2007. Economic growth and convergence with international differences in technology. Journal of Macroeconomics 29 (1): 145–168.
Gould, D.M., and W.C. Gruben. 1996. The role of intellectual property rights
in economic growth. Journal of Development Economics 48 (2): 323–350.
Gowdy, J.M. 2004. Altruism, evolution, and welfare economics. Journal of Economic Behavior & Organization 53 (1): 69–73.
Graafland, J., and B. Compen. 2015. Economic freedom and life satisfaction:
Mediation by income per capita and generalized trust. Journal of Happiness
Studies 16 (3): 789–810.
Gradstein, M. 2004. Governance and growth. Journal of Development Economics
73 (2): 505–518.
Greenaway, D., W. Morgan, and P. Wright. 1997. Trade liberalization and growth
in developing countries: Some new evidence. World Development 25 (11):
1885–1892.
Gul, F., and W. Pesendorfer. 2007. Welfare without happiness. The American
Economic Review 97 (2): 471–476.
Guseh, J.S. 1997. Government size and economic growth in developing countries: A political-economy framework. Journal of Macroeconomics 19 (1):
175–192.
2 Contemporary Views on Welfare and Reforms
37
Gwartney, J., K. Schuler, and R. Stein. 2001. Achieving monetary stability at
home and abroad. Cato Journal 21 (2): 183–203.
Gwartney, J.D., R.G. Holcombe, and R.A. Lawson. 2004. Economic freedom,
institutional quality, and cross-country differences in income and growth. Cato
Journal 24 (3): 205–233.
Haggard, S., and L. Tiede. 2011. The rule of law and economic growth: Where
are we? World Development 39 (5): 673–685.
Hall, J.C., R.S. Sobel, and G.R. Crowley. 2010. Institutions, capital, and growth.
Southern Economic Journal 77 (2): 385–405.
Hall, J.C., D. Stansel, and D. Tarabar. 2015. Economic freedom studies: A survey.
In Economic behavior, economic freedom, and entrepreneurship, ed. R. Cebula,
J. Hall, F. Mixon, and J. Payne, pp. 38–50. Edward Elgar Publishing.
Harris, R.G., and P.E. Robertson. 2013. Trade, wages and skill accumulation in
the emerging giants. Journal of International Economics 89 (2): 407–421.
Harrison, A., and G. Hanson. 1999. Who gains from trade reform? Some remaining puzzles. Journal of Development Economics 59 (1): 125–154.
Helpman, E. 2016, Dec. Globalization and wage inequality. Working Paper
22944, National Bureau of Economic Research.
Ho, M.S., and D.W. Jorgenson. 1994. Trade policy and U.S. economic growth.
Journal of Policy Modeling 16 (2): 119–146.
Holcombe, R.G. 2009. A reformulation of the foundations of welfare economics.
Review of Austrian Economics 22 (3): 209–224.
Huynh, K.P., and D.T. Jacho-Chávez. 2009. Growth and governance:
A nonparametric analysis. Journal of Comparative Economics 37 (1), 121–143.
Symposium: Labor Regulation In Developing Countries, Symposium: Labor
Regulation In Developing Countries.
Iacovone, L. 2012. The better you are the stronger it makes you: Evidence on the
asymmetric impact of liberalization. Journal of Development Economics 99 (2):
474–485.
Jayadev, A., and S. Bowles. 2006. Guard labor. Journal of Development Economics 79 (2), 328–348. Special Issue in honor of Pranab Bardhan.
Jenkins, R. 1997.Trade liberalisation in Latin America:The Bolivian case. Bulletin
of Latin American Research 16 (3): 307–325.
Jetter, M. 2014. Volatility and growth: Governments are key. European Journal
of Political Economy 36: 71–88.
Jinjarak, Y., G. Salinas, and Y.M. Tsikata. 2013. The effect of World Bank trade
adjustment assistance on trade and growth, 1987–2004: Is the glass half full or
half empty? Economic Systems 37 (3): 415–430.
38
P. Stankov
Jones, C.I., and P.J. Klenow. 2016. Beyond GDP? Welfare across countries and
time. American Economic Review 106 (9): 2426–57.
Jordan, J.L. 2001. Hayekian economic infrastructure as a foundation for sustained
prosperity. Contemporary Economic Policy 19 (1): 20–26.
Justesen, M.K. 2008. The effect of economic freedom on growth revisited: New
evidence on causality from a panel of countries 1970–1999. European Journal
of Political Economy 24 (3): 642–660.
Kapeliushnikov, R., A. Kuznetsov, N. Demina, and O. Kuznetsova. 2013. Threats
to security of property rights in a transition economy: An empirical perspective.
Journal of Comparative Economics 41 (1): 245–264.
Karabegovic, A., D. Samida, C.M. Schlegel, and F. McMahon. 2003. North
American economic freedom: An index of 10 Canadian provinces and 50 US
states. European Journal of Political Economy 19 (3): 431–452. Economic Freedom.
Kenisarin, M.M., and P. Andrews-Speed. 2008. Foreign direct investment in
countries of the former Soviet Union: Relationship to governance, economic
freedom and corruption perception. Communist and Post-Communist Studies
41 (3): 301–316.
Kim, E. 2000. Trade liberalization and productivity growth in Korean manufacturing industries: Price protection, market power, and scale efficiency. Journal
of Development Economics 62 (1): 55–83.
Kotera, G., K. Okada, and S. Samreth. 2012. Government size, democracy,
and corruption: An empirical investigation. Economic Modelling 29 (6):
2340–2348.
Krammer, S.M. 2009. Drivers of national innovation in transition: Evidence
from a panel of Eastern European countries. Research Policy 38 (5): 845–860.
Krishna, P., and D. Mitra. 1998. Trade liberalization, market discipline and productivity growth: New evidence from India. Journal of Development Economics
56 (2): 447–462.
Kwan, Y.K., and E.L.-C. Lai. 2003. Intellectual property rights protection and
endogenous economic growth. Journal of Economic Dynamics and Control 27
(5): 853–873.
Lagerlöf, N.-P. 2013. Violence and property rights. Journal of Economic Dynamics
and Control 37 (1): 312–328.
López, R., and M. Schiff. 2013. Interactive dynamics between natural and manmade assets: The impact of external shocks. Journal of Development Economics
104: 1–15.
Manca, F. 2010. Technology catch-up and the role of institutions. Journal of
Macroeconomics 32 (4): 1041–1053.
2 Contemporary Views on Welfare and Reforms
39
Marshall, A. 1890. Principles of economics. London: Macmillan.
McCleery, R.K., and F.D. Paolis. 2008. The Washington Consensus: A postmortem. Journal of Asian Economics 19 (5–6): 438–446.
Mehlum, H. 2002. Zimbabwe: Investments, credibility and the dynamics following trade liberalization: On the investment response during trade reform.
Economic Modelling 19 (4): 565–584.
Menyah, K., S. Nazlioglu, and Y. Wolde-Rufael. 2014. Financial development,
trade openness and economic growth in African countries: New insights from
a panel causality approach. Economic Modelling 37: 386–394.
Merlevede, B. 2003. Reform reversals and output growth in transition economies.
Economics of Transition 11 (4): 649–669.
Mijiyawa, A.G. 2008. Sustained economic growth: Do institutions matter, and
which one prevails? Cato Journal 28 (3): 385–420.
Mondal, D., and M.R. Gupta. 2008. Intellectual property rights protection and
unemployment in a North South model: A theoretical analysis. Economic Modelling 25 (3): 463–484.
Morris, C.T., and I. Adelman. 1989. Nineteenth-century development experience and lessons for today. World Development 17 (9): 1417–1432.
Mulaga, G., and J. Weiss. 1996. Trade reform and manufacturing performance
in Malawi 1970–1991. World Development 24 (7): 1267–1278.
Munda, G. 2016. Beyond welfare economics: Some methodological issues. Journal of Economic Methodology 23 (2): 185–202.
Murphy, R.H. 2016. Economic freedom of North America at state borders.
Journal of Institutional Economics 12 (4): 885–893.
Naito, T. 2012. A Ricardian model of trade and growth with endogenous trade
status. Journal of International Economics 87 (1): 80–88. Symposium on the
Global Dimensions of the Financial Crisis.
Naranpanawa, A., and R. Arora. 2014. Does trade liberalization promote regional
disparities? Evidence from a multiregional CGE model of India. World Development 64: 339–349.
Ng, Y. 2003. From preference to happiness: Towards a more complete welfare
economics. Social Choice and Welfare 20 (2): 307–350.
Nicita, A. 2009. The price effect of tariff liberalization: Measuring the impact on
household welfare. Journal of Development Economics 89 (1): 19–27.
Nyström, K. 2008. The institutions of economic freedom and entrepreneurship:
Evidence from panel data. Public Choice 136 (3–4): 269–282.
Ogilvie, S., and A. Carus. 2014. Chapter 8, Institutions and economic growth
in historical perspective. In Handbook of economic growth, ed. P. Aghion and
S.N. Durlauf, vol. 2 of Handbook of Economic Growth, 403–513. Elsevier.
40
P. Stankov
Onafowora, O.A., and O. Owoye. 1998. Can trade liberalization stimulate economic growth in Africa? World Development 26 (3): 497–506.
Paldam, M. 2003. Economic freedom and the success of the Asian tigers: An
essay on controversy. European Journal of Political Economy 19 (3), 453–477.
Economic Freedom.
Peev, E., and D.C. Mueller. 2012. Democracy, economic freedom and growth in
transition economies. Kyklos 65 (3): 371–407.
Peltzman, S. 1976. Toward a more general theory of regulation. Journal of Law
and Economics 19 (2): 211–40.
Peltzman, S. 2009. Mortality inequality. Journal of Economic Perspectives 23 (4):
175–90.
Pigou, A. 1920. The economics of welfare. London: Macmillan.
Piketty, T. 2014. Capital in the twenty-first century. Cambridge, MA: Harvard
University Press.
Piketty, T., and E. Saez. 2014. Inequality in the long run. Science 344 (6186):
838–843.
Posner, R.A. 1974. Theories of economic regulation. Bell Journal of Economics
5 (2): 335–358.
Pyo, H.K. 1990. Export-led growth, domestic distortions, and trade liberalization: The Korean experience during the 1980s. Journal of Asian Economics
1 (2): 225–247.
Rahapakse, P., and N. Arunatilake. 1997. Would a reduction in trade barriers promote intra-SAARC trade? A Sri Lankan perspective. Journal of Asian Economics
8 (1): 95–115.
Robbins, L. 1932. An essay on the nature and significance of economic science.
London: Macmillan.
Rode, M., and S. Coll. 2012. Economic freedom and growth. Which policies
matter the most? Constitutional Political Economy 23 (2): 95–133.
Rodrik, D. 2005. Chapter 14, Growth strategies. In Handbook of economic growth,
ed. P. Aghion and S.N. Durlauf, vol. 1, Part A of Handbook of Economic Growth,
pp. 967–1014. Elsevier.
Rutherford, T.F., and D.G. Tarr. 2002. Trade liberalization, product variety and
growth in a small open economy: A quantitative assessment. Journal of International Economics 56 (2): 247–272.
Samuelson, P.A. 1943. Further commentary on welfare economics. The American
Economic Review 33 (3): 604–607.
Schubert, C. 2012. Is novelty always a good thing? Towards an evolutionary
welfare economics. Journal of Evolutionary Economics 22 (3): 585–619.
Sen, A., and M. Nussbaum. 1993. The quality of life. Oxford University Press.
2 Contemporary Views on Welfare and Reforms
41
Sonin, K. 2003. Why the rich may favor poor protection of property rights.
Journal of Comparative Economics 31 (4): 715–731.
Stigler, G.J. 1943. The new welfare economics. The American Economic Review
33 (2): 355–359.
Stigler, G.J. 1971. The theory of economic regulation. Bell Journal of Economics
2 (1): 3–21.
Stiglitz, J.E., A. Sen, and J.-P. Fitoussi. 2009. Report by the Commission on
the Measurement of Economic Performance and Social Progress. Stiglitz-SenFitoussi Commission.
Sunde, U., M. Cervellati, and P. Fortunato. 2008. Are all democracies equally
good? The role of interactions between political environment and inequality
for rule of law. Economics Letters 99 (3): 552–556.
Voigt, S., and J. Gutmann. 2013. Turning cheap talk into economic growth: On
the relationship between property rights and judicial independence. Journal
of Comparative Economics 41 (1): 66–73. Symposium in Honor of Thrainn
Eggertson.
Voigt, S., J. Gutmann, and L.P. Feld. 2015. Economic growth and judicial independence, a dozen years on: Cross-country evidence using an updated set of
indicators. European Journal of Political Economy 38: 197–211.
Walz, U. 1998. Does an enlargement of a common market stimulate growth and
convergence? Journal of International Economics 45 (2): 297–321.
Willenbockel, D. 1998. Growth effects of anticipated trade liberalization and the
Baldwin multiplier. Economics Letters 59 (2): 231–235.
Williamson, C.R., and R.L. Mathers. 2011. Economic freedom, culture, and
growth. Public Choice 148 (3–4): 313–335.
Winston, C. 1993. Economic deregulation: Days of reckoning for microeconomists. Journal of Economic Literature 31 (3): 1263–89.
Wolfe, A.B. 1931. On the content of welfare. The American Economic Review 21
(2): 207–221.
Wu, S.-Y., J.-H. Tang, and E.S. Lin. 2010. The impact of government expenditure
on economic growth: How sensitive to the level of development? Journal of
Policy Modeling 32 (6): 804–817.
Wu, W., and O.A. Davis. 1999. The two freedoms, economic growth and development: An empirical study. Public Choice 100 (1–2): 39–64.
Xu, Z., and H. Li. 2008. Political freedom, economic freedom, and income
convergence: Do stages of economic development matter? Public Choice 135
(3–4): 183–205.
Yamamura, E. 2011. Decomposition of the effect of government size on growth.
Economics Letters 112 (3): 230–232.
3
Policies and Reforms
3.1
Economic Policies Since 1970
This section presents the world distribution of economic freedom policies
in the five broad policy domains at six moments in time: in 1970, 1980,
1990, 2000, 2008, and 2014. The first four moments are chosen so that
policies are monitored at time intervals which would allow for meaningful policy change to happen both within and across countries. The last
two moments are specifically chosen to observe significant policy changes
before and after the Great Recession. 2014 is the last year on which the
EFW index has been produced to date. As a result, analysis beyond 2014
is not possible.
3.1.1 Government Intervention
The Size of Government index measures the following:
• Government consumption: the share of government spending in total
consumption;
© The Author(s) 2017
P. Stankov, Economic Freedom and Welfare Before and After the Crisis,
DOI 10.1007/978-3-319-62497-6_3
43
44
P. Stankov
• Transfers and subsidies: general government transfers and subsidies as
a share of GDP;
• Government enterprises and investment: government investment as a
share of total investment;
• Top marginal tax rate: top marginal income and payroll tax rates.
Figure 3.1 reviews government intervention policies since 1970. Panel
3.1a presents the world distribution of the overall Size of Government
index. Overall government intervention in the economy becomes more
prevalent between 1970 and 1980. It is evident that the share of countries
with an index located between 6 and 10 is being reduced between 1970
and 1980. At the same time, the share of economies which are less free
from government intervention is going up, most notably for values of the
index between 4 and 6. The 1980s see some reversal in that trend, and by
(a)
(b)
Overall Size of Government Policies, 1970-2014
0
0
.05
.1
Density
Density
.1
.15
.2
.2
.3
.25
Government Consumption Policies, 1970-2014
0
1
2
3
4
5
6
7
8
Distribution of the overall Size of Government index
1970
1990
2008
9
10
0
1
2
3
4
5
6
7
8
Distribution of the index of Government Consumption
1980
2000
2014
1970
1990
2008
(c)
9
10
1980
2000
2014
(d)
Policies for Transfers and Subsidies, 1970-2014
0
0
.1
.05
Density
.2
Density
.1
.3
.4
.15
Govt. Enterprises and Investment Policies, 1970-2014
0
1
2
3
4
5
6
7
8
9
Distribution of the Government Enterprises and Investment index
1970
1990
2008
1980
2000
2014
10
0
1
2
3
4
5
6
7
8
9
10
Distribution of the Transfers and Subsidies index
1970
1990
2008
1980
2000
2014
Fig. 3.1 Government intervention since 1970. Source Own calculations based on
Economic Freedom of the World Data, http://freetheworld.com
3 Policies and Reforms
45
1990, more countries enjoy less government intervention; a policy trend
also observed through the 1990s and until 2008. The years after the Crisis
see some mild reversal to increased government intervention. Specifically,
fewer countries occupy the index territory between 6 and 8, and more
countries fall between 3 and 5.
Changes in government consumption since 1970 are less discernible.
Panel 3.1b reviews those developments. Due to the fact that the government in most countries finances the military, police, education, and
health care, it is not surprising that changes in the worldwide distribution of government consumption expenditures are only modest. The most
significant changes appeared to be between 1970 and 1980, after which
the notable changes appear after the 2008 crisis. This was perhaps due to
some governments stepping in to prop up their ailing private sectors.
Panel 3.1c presents one of the more interesting policy developments in
this domain since the 1970s. Similar to the behavior of the overall Size
of Government index, the distribution of the subindex of Government
Enterprises and Investment (GEI) moved slightly to the left from 1970
to 1980. This movement was suggestive of more government investment
in their economies in 1980 than in 1970. This was to be expected due
to two relatively deep recessions in the mid-1970s and end-1970s. Those
recessions presented governments around the world with the need to invest
more to both support their own economies for immediate purposes and
to improve their long-term technological potential. This prompted the
beginning of the so-called supply-side policies, which became the fashion
in economic reforms in the beginning of the 1980s and were later imitated
by new democracies in the 1990s. The two decades of the 1980s and 1990s
witnessed governments stepping back somewhat from their active role in
investment policies. This is indicated by a noticeable decrease in the share
of countries with an index value between 0 and 4, and a gradual increase
in the share of countries with a value above 5. The process saw its peak
around 2008, when most countries were situated between 7 and 9. After
the crisis, the overall index notched down a bit, perhaps for similar reasons
to those witnessed between 1975 and 1980.
46
P. Stankov
3.1.2 Property Rights
The Property Rights index measures the following:
• Judicial independence: the ability of the judiciary to appear independent from political influence of members of government, citizens, or
firms;
• Impartial courts: the ability of the legal framework to provide the necessary conditions for settlements of private disputes and to challenge
the legality of government actions;
• Protection of property rights: a subjective evaluation of how effective
property rights protection is in the country;
• Military interference in rule of law and politics: the military’s involvement in politics based on the International Country Risk Guide Political
Risk Component G: Military in Politics.
• Integrity of the legal system: the strength of the legal system and the
popular observance of the law, based on the International Country Risk
Guide Political Risk Component I for Law and Order;
• Legal enforcement of contracts: the time and money required to collect
a debt;
• Regulatory costs of the sale of real property: the time measured in days
and monetary costs required to transfer ownership of property;
• Reliability of police: a subjective evaluation of how reliable police forces
are in enforcing law and order in the country;
• Business costs of crime: a subjective evaluation of how expensive crime
and violence are for businesspeople in the country.
Figure 3.2 presents the development of the Legal Systems and Security
of Property Rights index since 1970. The overall development of policies
in this area are given in Panel 3.2a. Depending on data availability, the
rest of the figure presents more specific developments in areas such as:
property rights protection, legal enforcement of contracts, and regulatory
restrictions on the sale of real property.
One of the immediate observations on the overall development of property rights policies is that countries gradually become more similar. This is
evident from the rising mass of countries around the values of the overall
47
3 Policies and Reforms
(a)
(b)
Property Rights Protection Policies, 1970-2014
.2
.25
Density
.15
0
0
.05
.1
.1
Density
.2
.3
.25
Overall Property Rights Policies, 1970-2014
0
1
2
3
4
5
6
7
8
9
10
0
1
Distribution of the overall index of Property Rights
1970
1990
2008
3
4
1980
2000
2014
5
6
1970
1990
2008
(c)
7
8
9
10
1980
2000
2014
(d)
Restrictions on Property Sales, 1970-2014
.15
Density
.1
0
0
.05
.05
.1
.15
.2
.2
.25
0.25
Policies to Enforce Contracts, 1970-2014
Density
2
Distribution of the index of Property Rights Protection
0
1
2
3
4
5
6
7
8
Distribution of Legal Enforcement of Contracts index
1970
1990
2008
1980
2000
2014
9
10
0
1
2
3
4
5
6
7
8
9
10
Distribution of Restrictions on the Sale of Real Property index
1970
1990
2008
1980
2000
2014
Fig. 3.2 Legal system and security of property rights since 1970. Source Own calculations based on Economic Freedom of the World Data, http://freetheworld.com
index between 4 and 6 in Panel 3.2a. There is also a slight movement of
the distribution of the index to the left after 2008. This means policies
around the globe backtracked on the protection of property rights after
the Crisis.
This is also evident from the distribution of the specific index of Property Rights Protection displayed in Panel 3.2b. Most countries before and
after the Crisis are similar for values of the index below 4. However, there
is a marked shift in countries with better protection of property rights
toward less reliable property rights protection policies after the Crisis.
This is evident from two trends: (1) the diminishing number of countries
with an index value above 6 and (2) the increase of the mass of countries
around the index values between 4 and 6.
48
P. Stankov
The weakening of property rights protection policies is also complemented by a less pronounced yet evident development in the index of
Contract Enforcement. Similarly to the index of PRP, the mass of countries with an index value of 6 goes down after the Crisis. At the same time,
there is a marked increase in the share of countries situated around an
index value of between 3 and 5. The trends before and after the Great
Recession are depicted in Panel 3.2c.
At the same time, after the Crisis, governments around the world
generally adopt a slightly more business-friendly approach to transferring property. If transferring property is relatively easy, then productive
resources will be shifted to their most efficient use more cheaply, so that
economic agents incur lower production costs. This will stimulate businesses to produce more, at any given market price. In fact, governments
do make property transfers cheaper. This is indicated by the increase of
the mass of countries around the index value between 7 and 9. At the
same time, the number of countries with an index value below 7 declines.
The trends in the ease of transferring property before and after the Great
Recession are depicted in Panel 3.2d.
3.1.3 Monetary Policies
The Sound Money index effectively measures monetary policies by
including:
• Money growth: average annual growth of the money supply (M1 aggregate) relative to real GDP growth;
• Standard deviation of inflation over the last five years, whereby inflation
is gauged by the GDP deflator or the CPI index if the deflator is
unavailable;
• Inflation in the most recent year measured by the CPI or the GDP
deflator;
• Freedom to own foreign currency bank accounts: the ability to open
bank accounts in foreign currency both domestically and abroad.
49
3 Policies and Reforms
(a)
(b)
.6
Density
.4
.2
0
0
.1
Density
.2
.3
.8
Money Growth Policies, 1970-2014
.4
Overall Monetary Policies, 1970-2014
0
1
2
3
4
5
6
7
8
Distribution of the overall index of Sound Money
1970
1990
2008
9
10
0
1
2
3
4
5
6
7
Distribution of the Money Growth index
1980
2000
2014
1970
1990
2008
(c)
8
9
10
9
10
1980
2000
2014
(d)
Density
.4
0
0
.2
.2
Density
.4
.6
.8
Most Recent Inflation, 1970-2014
.6
Price Stability Policies, 1970-2014
0
1
2
3
4
5
6
7
Distribution of the Price Stability index
1970
1990
2008
1980
2000
2014
8
9
10
0
1
2
3
4
5
6
7
8
Distribution of the index of Most Recent Inflation
1970
1990
2008
1980
2000
2014
Fig. 3.3 Monetary policies since 1970. Source Own calculations based on Economic
Freedom of the World Data, http://freetheworld.com
Figure 3.3 presents both the overall Sound Money index and selected
specific subindices indicating how monetary policies developed over time.
The Sound Money index is not a measure of monetary policies per se.
However, it does include indicative measures of the effects of central bank
policies. For example, when the central bank adopts a more expansive
monetary policy, then money growth will increase, suggestive of higher
expected inflation. If that is the case, then the Money Growth index will go
down. In addition, when the central bank adopts a more expansive monetary policy, price stability may be undermined. This would also lead to a
higher standard deviation of inflation over time, which is measured by the
Price Stability subindex (the Standard Deviation of Inflation). An excessively expansionary monetary policy then will also lower the Price Stability
index, as higher price stability means higher index values. Finally, the Most
Recent Inflation index measures the inflation rate in the year preceding
the index preparation. In it, higher recent inflation would lower the value
50
P. Stankov
of the index, and vice versa. Keeping all this in mind, it is easy to argue
that although the Sound Money index is not a direct measure of monetary
policy actions, it is a direct measure of their consequences. Therefore, the
Sound Money index can be thought of as a gauge of monetary policies
over time.
The overall developments in the Sound Money index are presented
in Panel 3.3a. Because of the higher inflationary periods and the higher
inflation instability in the late 1970s, the entire distribution of overall
monetary policies has moved to the left from 1970 to 1980. The period
between 1980 and 1990 saw a moderate reversal of the inflationary policy
trend, especially considering the restrictive policies aimed to stem inflation
in the beginning of the 1980s. The next decade was characterized by a
continuation of the trends established in the 1990s, with an increasing
number of countries shifting toward increased stability and predictability
of monetary policy. This trend continued until the Great Recession, with
the distribution of countries with an index value of 7 markedly increasing
at the expense of countries with an index value below 5. Perhaps, due to
noticeable deflationary trends after the Crisis, there was an increase in the
number of countries scoring higher on this index, with a marked increase
in the mass of countries around an index value of 9.
One of the ways central banks conduct their policies is through affecting
money growth. The index of Money Growth is depicted in Panel 3.3b.
Similarly to the overall index of monetary policies, the index moves back
from 1970 to 1980, suggesting more pro-inflationary policies between
1970 and 1980. The trend is reversed after 1980, with a gradual increase
in the mass of countries by 1990, and then a further increase by 2000.
Due to the economic boom in the run-up to the Crisis, money growth
has risen, and the mass of countries have moved back a little. However,
after the Crisis, most countries pursue more stable money growth policies,
with the mass of countries settling around a very healthy index value of 9.
The longer-term shifts in monetary policies are moderate for the entire
period. This can be seen not only from the dynamics of the previous two
indices, but also from the Price Stability index (the standard deviation
of inflation). The PS index is indicative of the price volatility in the last
5 years before the index was recorded. As seen in Panel 3.3c, there are
almost negligible shifts in the price volatility index before and after the
3 Policies and Reforms
51
Crisis. The most notable negative trend in longer-term price stability was
recorded before 1980. After that, most countries focus on pursuing policies
to ensure price stability, a trend hallmarked by the introduction of inflation
targeting in more than a few countries since 1991, and most notably the
introduction of the Euro in 1999. There are only negligible differences in
the index values of Price Stability before and after the Crisis, suggestive
of a learning process about the possible detrimental effects of significant
longer-term price volatility.
While the Price Stability index is indicative of the longer-term price
dynamics, the last subindex presented in Panel 3.3d captures only the
inflation in the last year before the index was recorded. Despite the shorterterm angle, the index dynamics are very similar to the one in the previous
panels. The index suggests that inflation was a problem at the end of the
1970s, and was gradually becoming a problem in the year or two before
the Crisis. After the Great Recession, it seems that most countries enjoy
stable prices, as about three-quarters of the countries scatter around an
index value of close to 10.
3.1.4 Free Trade
Policies to support trade across borders encourage economies to
grow both domestically and internationally. The Freedom to Trade
Internationally index measures exactly those policies. It includes:
• Tariffs: the amount of taxes on international trade as a share of exports
and imports, mean tariff rates, and the standard deviation of tariff rates;
• Regulatory trade barriers: non-tariff trade barriers, and compliance cost
of importing and exporting;
• Controls on the movement of capital and people: based on existing
foreign property ownership and/or investment restrictions, capital controls, and freedom of foreigners to visit for tourist and short-term business purposes;
• Black-market exchange rates: the percentage difference between the
official and the parallel (black-market) exchange rate to measure
exchange rate controls.
52
P. Stankov
(a)
(b)
.3
Density
.2
.1
0
0
.1
Density
.2
.3
.4
Tariff Policies, 1970-2014
.4
Overall Trade Policies, 1970-2014
0
1
2
3
4
5
6
7
8
Distribution of the overall index of Freedom to Trade
1970
1990
2008
9
10
0
1
2
3
4
5
6
7
8
Distribution of the index of Tariff Policies
1980
2000
2014
1970
1990
2008
(c)
9
10
1980
2000
2014
(d)
.3
Density
.2
.1
0
0
.1
Density
.2
.3
.4
Capital Mobility and Labor Mobility Controls, 1970-2014
.4
Regulatory Trade Barriers, 1970-2014
0
1
2
3
4
5
6
7
8
Distribution of the index of Regulatory Trade Barriers
1970
1990
2008
1980
2000
2014
9
10
0
1
2
3
4
5
6
7
8
9
10
Distribution of the index of Capital and Labor Mobility
1970
1990
2008
1980
2000
2014
Fig. 3.4 Free trade policies since 1970. Source Own calculations based on Economic
Freedom of the World Data, http://freetheworld.com
If tariff rates go up, then governments are putting a brake on trade, which
may slow growth and impede economic freedom. That is why, when tariffs
rise, the index measuring those tariffs goes down. Similarly, regulatory
trade barriers hamper trade and impose restrictions on the movement of
capital and people. Therefore, the Free Trade index goes up when nontariff barriers and capital and labor restrictions are removed. Figure 3.4
presents the developments in those elements of free trade policies since
1970.
Panel 3.4a displays the dynamics of the overall index of free trade policies since the 1970s. It is evident that those changes are gradual for the
20 years before 1990. Only after 1990, have trade policies around the
world become significantly more liberalized. This is seen from the marked
shift in the number of countries with low index values to the right. Specifically, the mass of countries with index values below 5 has been signifi-
3 Policies and Reforms
53
cantly reduced, while the mass of countries above that value has notably
increased. The process continued in the run-up to the Great Recession.
The index does not demonstrate the world has moved to more restrictive
trade policies in the aftermath of the Crisis.
Panel 3.4b produces further evidence that trade policies have become
more liberalized since 2008. This is seen from the slight shift of the worldwide distribution of the index of Tariff Policies to the right, which means
that tariff rates are predominantly going down. This movement is a continuation of the trade policy trends of the three decades before the Crisis and
is in stark contrast with the reversals observed after the Great Depression
(Eichengreen and Irwin 2010; Gawande et al. 2015; Irwin 2012). This
is not surprising, as in those three decades, many more countries have
become members of the World Trade Organization and have generally
moved toward freer trade.
However, unlike more liberalized tariff policies, the non-tariff barriers
(NTBs) come marching in after the Crisis.This is seen from Panel 3.4c, and
has also been documented by Goldberg and Pavcnik (2016). The distribution of the index of NTBs, which seemed normal in 2008, has formed two
distinct hump-shapes in the aftermath of the Great Recession. At the same
time, one hump has moved to the right of the earlier distribution, while
the other has moved left. This means that some countries have become
more liberalized than before as a response to the Crisis, while others have
rapidly moved to protect their own industries by introducing various nontariff barriers to trade. Perhaps, this is a natural response given the expected
retaliation and trade destruction that would follow if governments adopt
harder protectionist measures, including outright increases in tariffs. It
would be interesting to see if more leftist or right-wing governments have
pursued these soft protectionist measures after 2008.
These measures are seen not only in NTBs, but also in some other
regulations related to trade—the free movement of capital and labor across
borders. The dynamics of the index measuring those controls are presented
in Panel 3.4d. It is interesting to note that, after the Crisis, government do
adopt more restrictive measures on capital and labor movements. However,
it is perhaps even more interesting to note that this policy reversal was
54
P. Stankov
happening in the years before the Great Recession, at least since 2000.
This is seen from the leftward shifts of the index distribution both before
and after 2008.
3.1.5 Government Regulation
As we have seen in the literature review, government regulation has long
spurred debates among economists and policy makers. On the one hand,
regulations exist which protect consumers and firms alike. Not applying
them would make almost everyone worse-off. On the other hand, there
are redundant regulations which increase the costs of firms but do not
add value to the product they produce. Therefore, a balance is required
between too little and too much regulation. The index of Government
Regulation attempts to capture the trends in how governments around the
world approach that balancing act. It is based on measuring policies in the
following areas:
• Credit market regulations: the percentage of bank deposits held in privately owned banks, the share of private credit to total credit extended
in the banking sector, and market determination of deposit and lending
interest rates;
• Labor market regulations: the prevalence and cumulative duration
of fixed-term contracts, minimum wage legislation, hiring and firing restrictions, the existence of a centralized wage bargaining process,
working hours regulations and mandated costs of worker dismissal,
and military conscription;
• Business regulations: compliance costs related to permits, regulations and reporting requirements, stringency of product standards,
the amount of time and money it takes to start a new business, the
prevalence of undocumented payments and government favoritism to
well-connected firms and individuals when deciding upon policies and
contracts, the time in days and monetary costs required to obtain a
license to construct a standard warehouse, and the time required per
year for a business to prepare, file, and pay taxes on corporate income,
value added and sales taxes, and taxes on labor.1
55
3 Policies and Reforms
(a)
(b)
Financial Policies, 1970-2014
0
0
.1
.1
Density
.2
Density
.2
.3
.3
.4
.4
Overall Regulatory Policies, 1970-2014
0
1
2
3
4
5
6
7
8
9
10
0
1
Distribution of the overall index of Regulation
1970
1990
2008
(c)
2
3
4
5
6
7
8
9
10
9
10
Distribution of the Credit Market Regulations index
1980
2000
2014
1970
1990
2008
(d)
.2
Density
.4
.6
.8
Business Regulations, 1970-2014
0
0
.1
Density
.2
.3
.4
Labor Market Policies, 1970-2014
1980
2000
2014
0
1
2
3
4
5
6
7
8
Distribution of the Labor Market Regulations index
1970
1990
2008
1980
2000
2014
9
10
0
1
2
3
4
5
6
7
8
Distribution of the Business Regulations index
1970
1990
2008
1980
2000
2014
Fig. 3.5 Regulatory policies since 1970. Source Own calculations based on Economic Freedom of the World Data, http://freetheworld.com
Figure 3.5 displays the developments of the overall index of Government Regulation since 1970 and its subindices related to credit, labor,
and product market regulations. The developments in the overall index
are presented in Panel 3.5a. The figure presents evidence that from 1970 to
1990, the overall dynamics of regulation were relatively subdued, with not
much change in either the mean or the median of the distribution. Still,
it is noticeable that countries become less similar in the way their governments approach regulation before 1990. This is seen from the widening
dispersion of the distribution of the overall index between 1980 and 1990.
After 1990, however, the tides turn in favor of more market-oriented
regulations. This is seen from the marked shift of the distribution to the
right between 1990 and 2000, and the increase in the share of countries with higher values of the overall index. At the same time, there was
another change between 1990 and 2000: The policy distribution became
less diverse. We are again witnessing a policy convergence process. The
market-oriented reforms in the overall index continued until immediately
56
P. Stankov
before the Great Recession. The surprising fact is that the overall index did
not decline after 2008. It seems governments around the world avoided
taking measures that would punish business in the aftermath of the Crisis.
Creating a friendlier business environment after 2008 seems sensible, considering the recent evidence by Feldmann (2017) and the international
competition for investment and the new normal subdued growth rates.
One of the dimensions of regulatory policies is the Credit Market Regulations (CMR). It does not monitor financial regulations per se, but
rather focuses on the results of financial regulations in terms of financial
deepening, i.e., more access to finance for more people and firms. Even
though the indicator is not a direct measure of financial regulations, it is an
adequate measure of how burdensome local policies for financial institutions are. Panel 3.5b reviews those policies across the globe over time. The
figure presents evidence that financial globalization was gradually becoming a more widespread phenomenon from 1980. This is evident from the
initially modest process of a rightward shift in the distribution of CMR
policies, which gradually gained ground, especially since 1990. Between
1990 and 2000, there was a marked decrease in the number of countries
with an index value below 6, and a simultaneous, and rapid, increase in
the number of countries with a value above 7.
This process took on an even more rapid pace after 2000. In the preCrisis years, many more countries joined the bandwagon of liberalizing
financial regulations, which is indicated by the rapid growth of the distribution of the index above index values of 7. The process did not reverse
after the Great Recession. It seems that there is still not enough political
consensus for a financial regulatory backlash within the countries where
the financial sector already enjoys a significant degree of freedom. Again,
this seems a sensible regulatory response to the Crisis because it has long
been established that financial development stimulates economic growth
(Demirgüç-Kunt and Levine 2008; Levine 1998, 2005). Perhaps, governments are a bit hesitant to put a brake on one of the factors for growth
exactly when their economies need growth most. Judging by the behavior
of this index, it seems that governments around the world are learning to
promote growth by maintaining financial liberalization, especially in view
3 Policies and Reforms
57
of recent evidence by Blau (2017), who claims that economic freedom
reforms are less likely to instill financial crises.
Labor market policies support the trends outlined above. The distribution of labor regulations since the 1970s is given in Panel 3.5c. It is
obvious that labor market policies gradually become more liberalized, with
the most stringent labor market conditions recorded at the very beginning
of the period—in 1970. Since then, the worldwide distribution of labor
market regulations is moving to the right. Again, the trend gained speed
after 1990, when it took a massive step toward labor market liberalization
by 2000. The trend continued immediately before 2008, and even after
the Crisis. The latter is seen from the decline in the share of countries
below index values of 6 and a marked increase in the number of countries
with index values above 7 by 2014.
Deregulation trends are also observed in the area of business regulation.
Panel 3.5d monitors those trends since 2000, as data is missing prior
to that year. Contrary to labor and credit market regulations, business
regulations were growing a bit more strict between 2000 and 2008. This
is seen from the decrease in the share of countries with index values of
above 7 before the Crisis and a leftward shift in the mass of those countries
by 2008. However, the trend was reversed after 2008, and by 2014, more
governments are introducing business-friendly regulations. This is seen
from the diminishing share of countries with index values below 6 and a
rapid increase in the share of countries above that value. At the same time,
business environments in this area have become more similar across the
globe after the Crisis. This policy convergence process is evident from the
fact that a great mass of countries arrive at index values between 6 and 7.
3.2
Reform Patterns
This section presents each of the five policy areas in a different way. Rather
than looking at the levels of the indices, I consider the changes in those
indices over certain periods of time. The index changes give an approximate idea of how fast the given policy area was reformed.
Policies can become either more market-oriented or more governmentoriented at different periods of time. Going pro-market means the value
58
P. Stankov
of a certain index goes up, whereas going pro-government means the value
of the index goes down. Theoretically, if a country had an index value of
zero at the beginning of the period and 10 at the end, then its reform
measure is +10. Alternatively, if the country had the most free policies
in the beginning of the period and was scoring 10 but then reversed its
policies completely, then its reform measure is −10. Thus, the theoretical
interval over which the reforms are measured is between −10 and +10.
The graphs depicting those reforms are generated by using the so-called
kernel density estimation. This is a technique used to deliver a smooth
distribution of values for a given interval, e.g., between −10 and +10. This
makes the graphs more easily interpretable than the usual histograms. It is
also important to note that the density need not represent the actual share
of countries going through certain reform values. However, the change in
those distributions over time can indeed provide insight into the direction
of the reform process.
Within each of the 5 policy areas, the reforms processes are sliced into
two groups of periods. The first subfigure displays the reform distributions
over two periods. The first distribution is the one before the Crisis. The
reform in this subfigure is the policy change between 1970 and 2008. The
second distribution goes beyond the Crisis and includes the period from
2008 to 2014. However, it trims the starting point at 1975 rather than
1970. The reason for the split is that it is easier to see if the reform process
was more intensive before or after the Great Recession.
The second subfigure narrows the interval to shorter periods. Initially,
the reforms are plotted at 10-year periods: 1970–1980, 1980–1990, and
1990–2000. Then, as the Great Recession nears, the period is shortened
to 8 years: 2000–2008, and finally to 6 years: 2008–2014. Shortening
the time span gives a graphical way to monitor the reform processes
immediately before and after the Crisis.
Those two subfigures are built within each of the five overall reform
areas. When it comes to regulation, however, separate figures with shorter
reform dynamics are produced for each of the subindices of credit, labor,
and product market regulations as well. The reason is that deregulation,
especially financial deregulation, has often been blamed for setting the
stage for the Crisis. It is interesting to see if there has been any policy
3 Policies and Reforms
59
backlash in the aftermath of the Great Recession, especially in terms of
government regulations in various fields, and particularly on financial
markets.
3.2.1 Reforms in the Size of Government
Figure 3.6a presents a long-term view of reforms in government intervention. It shows that the distribution of government intervention reforms
running up to 2008 is smoother and with a larger share of countries
strengthening the role of government. The distribution which starts in
1975 and runs through 2014 has a smaller left tail and a larger mass
around the mean. This means that after 1975, fewer countries enacted
more prominent government roles in the economy, and more countries
made at least minor changes to free the economy from unnecessary government intervention.
The evidence in Fig. 3.6b shows that, whereas the 1970s produced some
dramatic increases in the size of governments, the three decades after 1980
witnessed prominent reductions in government intervention. This was the
case right up to the Crisis. After that, the reform tides turned, similarly to
the responses observed after the Great Depression (Garrett et al. 2010).
Immediately after the Crisis, governments took a stronger position in
the economy, with more than half of countries stepping up its role in
the economy. This is perhaps only normal given the depth of the Great
(b)
(a)
Size of Government Reforms: Shorter Periods
.4
Density
.2
.15
0
.05
.1
Density
.2
.25
.6
Size of Government Reforms: Before and After the Crisis
0
-7 -6 -5 -4 -3 -2 -1
0
1
2
3
4
5
6
7
Distributions of the change in the Size of Government index between 1970-2014
-7
-6
-5
-4
-3
-2
-1
0
1
2
3
4
5
6
7
Distributions of the change in the Size of Government index between 1970-2014
Change b/w 1970-2008
Change b/w 1975-2014
Change b/w 1970-1980
Change b/w 1990-2000
Change b/w 2008-2014
Change b/w 1980-1990
Change b/w 2000-2008
Fig. 3.6 Size of government reforms since 1970. Source Own calculations based
on Economic Freedom of the World Data, http://freetheworld.com
60
P. Stankov
Recession. Still, unlike the 1970s when the government took massive steps
to intervene in the economy, the 6 years after the Crisis produced a far
less dramatic government comeback to the economy. When a government
increased its role in the economy after the Crisis, it was done quite carefully.
3.2.2 Property Rights Reforms
Figure 3.7a demonstrates that property rights reforms were mostly
gradual between 1970 and 2014. It is evident that most reforms took place
after 1975, just as with the government intervention reforms. Almost no
country deteriorated its property rights index between 1975 and 2014
with a value of more than one, and the vast majority actually improved
their property rights protection, with reforms becoming bolder and more
popular across countries.
Additional details on property rights reforms is presented in Fig. 3.7b.
By observing the changes between 2008 and 2014, we can conclude that
changes were only gradual and timid, unlike those in any other decade
preceding the Crisis. Other decades saw the majority of countries taking
either a noticeable step forward (like the 1980s and 1990s) or backward
(like 1970s and early 2000s). Similarly to the Size of Government reforms,
the reforms in Property Rights were most common in the 1980s and 1990s,
while reform reversals were undertaken in the 1970s and in the few years
before the Crisis.
Property Rights Reforms: Before and After the Crisis
(b)
Property Rights Reforms: Shorter Periods
.6
.4
Density
0
.1
.2
.2
Density
.3
.8
1
.4
(a)
0
-6
-5
-4
-3
-2
-1
0
1
2
3
4
5
6
Distributions of the change in the Property Rights index between 1970-2014
-6
-5
-4
-3
-2
-1
0
1
2
3
4
5
6
Distributions of the change in the Property Rights index between 1970-2014
Change b/w 1970-2008
Change b/w 1975-2014
Change b/w 1970-1980
Change b/w 1990-2000
Change b/w 2008-2014
Change b/w 1980-1990
Change b/w 2000-2008
Fig. 3.7 Property rights reforms since 1970. Source Own calculations based on
Economic Freedom of the World Data, http://freetheworld.com
3 Policies and Reforms
Monetary Policy Reforms: Before and After the Crisis
(b)
.6
0
.05
.2
.4
Density
.15
.1
Density
Monetary Policy Reforms: Shorter Periods
.8
.2
(a)
61
0
-8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8
Distributions of the change in the Sound Money index between 1970-2014
-8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8
Distributions of the change in the Sound Money index between 1970-2014
Change b/w 1970-2008
Change b/w 1975-2014
Change b/w 1970-1980
Change b/w 1990-2000
Change b/w 2008-2014
Change b/w 1980-1990
Change b/w 2000-2008
Fig. 3.8 Monetary reforms since 1970. Source Own calculations based on Economic Freedom of the World Data, http://freetheworld.com
3.2.3 Monetary Reforms
Figure 3.8 reviews the monetary policy reform dynamics since 1970. The
first thing to note from the figure is that the monetary reforms distribution
looks similar, overall, to those observed for government intervention and
property rights. The tails of the distribution of the reforms contain much
more extreme values than the size of government and the property rights
reforms, especially on the positive side of reforms. This is evident from
Fig. 3.8a. The initial period running up to the Crisis entailed much more
hesitant reforms, allowing financial deepening and stemming inflation.
However, the second period was more stable, with many more countries
enjoying monetary stability and financial development.
Just as with property rights, the most active period for monetary reforms
were the 1980s and 1990s. This is indicated in Fig. 3.8b. Interestingly, the
few years before the Crisis were not the most active in terms of financial
reforms. This is indicated by the approximately equal shares of countries
reforming and countries reversing reforms between 2000 and 2008. There
is another surprising trend in monetary policies after the Crisis. It seems
that the Crisis instilled reforms more often than not. Although modest,
the changes in the overall index of Sound Money after 2008 are overwhelmingly positive, unlike in the years before 2008.
Positive reform values after 2008 may be due to a variety of factors,
including how the index of Sound Money is constructed in the first place.
62
P. Stankov
Mingling together inflation stability and financial development may lower
the overall index exactly when the banking sector provides the economy
with the highest volumes of financing, which comes at a price: somewhat
higher inflation. This is exactly what happened in the few years before the
Crisis, and is perhaps what lowered the index of Sound Money. As a result,
the world may have been experiencing higher levels of financial integration
than ever, which the index may not have captured. At the same time, in the
years after 2008, as price levels have been hovering right above deflation
zones across the globe, inflation seems stable and predictable, without
much deviation over time. Perhaps, this is exactly what overwhelmingly
drove the index into positive territories after the Crisis.
3.2.4 Free Trade Reforms
Figure 3.9 documents changes in the Freedom to Trade Internationally
(FTI) index since 1970. It conveys one definite message: The world has
become a more trade-friendly place. The trade reforms are overwhelmingly
directed at making trade across borders easier and cheaper, suggested by
the large positive parts of the overall FTI distributions before and after the
Great Recession in Fig. 3.9a. Similarly to other reforms analyzed earlier,
trade reforms went ahead very strongly after the 1970s.
We further disaggregate the reforms across time in Fig. 3.9b. By doing
this, we notice that the most active decade for trade reforms was between
(a)
(b)
Trade Reforms: Before and After the Crisis
0
.05
.2
Density
.4
.6
Density
.15
.1
.8
.2
1
Trade Reforms: Shorter Periods
0
-8
-8
-7
-6 -5 -4 -3 -2 -1 0
1
2
3
4
5
6
7
Distributions of the change in the FTI index between 1970-2014
Change b/w 1970-2008
Change b/w 1975-2014
8
-7
-6 -5 -4 -3 -2 -1 0
1
2
3
4
5
6
7
Distributions of the change in the FTI index between 1970-2014
Change b/w 1970-1980
Change b/w 1990-2000
Change b/w 2008-2014
8
Change b/w 1980-1990
Change b/w 2000-2008
Fig. 3.9 Trade reforms since 1970. Source Own calculations based on Economic
Freedom of the World Data, http://freetheworld.com
3 Policies and Reforms
63
1990 and 2000, perhaps driven by the efforts of Central and Eastern
Europe and Latin America to adopt a Western-style institutional framework embedded in the Washington Consensus platform.
Outside of this exceptional transition decade between 1990 and 2000,
trade reforms have not been fundamentally different from one decade to
another. Most countries have not done much to either free or reverse freedom to trade. When they did, those reforms or reform reversals were not
very bold in either direction. Naturally, there were some exceptions indicated by the long but shallow negative tail of the 1970s distribution, and
the positive tail of the 1980s distribution. However, the Great Recession
did stall trade reforms. Most reforms were very timid after 2008, if there
were any overall reforms at all. Yet, the patterns of reversals after the Great
Recession are nowhere near the destructive trade policies of the 1930s
observed by Crucini and Kahn (1996) and Irwin (2012), among others.
3.2.5 Regulatory Reforms
In Fig. 3.10, overall deregulation reforms look very similar to trade reforms
since 1970. The vast majority of countries took positive steps to reform
their credit market, labor market, and business regulations, so that they
became more business-friendly. This is seen from Fig. 3.10a, which also
demonstrates that deregulation has gained momentum, and the process
has been stronger since 1975.
Regulatory Reforms: Before and After the Crisis
(b)
Regulatory Reforms: Shorter Periods
0
.1
.2
Density
.4
Density
.2
.3
.6
.8
.4
(a)
0
-5
-4
-3
-2
-1
0
1
2
3
4
5
Distributions of the change in the overall Regulation index between 1970-2014
-5
-4
-3
-2
-1
0
1
2
3
4
5
Distributions of the change in the overall Regulation index between 1970-2014
Change b/w 1970-2008
Change b/w 1975-2014
Change b/w 1970-1980
Change b/w 1990-2000
Change b/w 2008-2014
Change b/w 1980-1990
Change b/w 2000-2008
Fig. 3.10 Overall regulatory reforms since 1970. Source Own calculations based
on Economic Freedom of the World Data, http://freetheworld.com
64
P. Stankov
Figure 3.10b provides the most detailed picture of when deregulation
occurred with the most vigor and when it slowed down. The two decades
marked with the boldest deregulation steps were 1990–2000 and 2000–
2008. In those two decades, the highest share of economies underwent
deregulation reforms. As with most reforms, the Great Recession stalled
further deregulation reforms for a large number of countries. This is seen
from the fact that most countries did not take any further steps to liberalize
their credit, labor, and business regulations after 2008.
To see which avenue of deregulation was most active before the Crisis,
we need to look at the various types of deregulation reforms. As with trade
and overall deregulation, 1980s and 1990s saw an unprecedented wave
of financial liberalization across the globe, with the 1990s being slightly
stronger than the 1980s. Figure 3.11a depicts the changes in Credit Market Regulations (CMR) index in shorter periods. Gradually, an increasing
number of countries backtracked on the speed of their financial liberaliza(b)
Labor Reforms: Shorter Periods
1.5
Financial Reforms: Shorter Periods
0
0
.1
.5
Density
Density
.2
.3
1
.4
.5
(a)
-10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10
Distributions of the change in the CMR index between 1970-2014
Change b/w 1970-1980
Change b/w 1990-2000
Change b/w 2008-2014
Change b/w 1980-1990
Change b/w 2000-2008
(c)
-5
-4
-3
-2
-1
0
1
2
3
4
Distributions of the change in the LMR index between 1970-2014
Change b/w 1970-1980
Change b/w 1990-2000
Change b/w 2008-2014
5
Change b/w 1980-1990
Change b/w 2000-2008
0
.2
Density
.4
.6
.8
Business Regulation Reforms: Shorter Periods
-5
-4
-3
-2
-1
0
1
2
3
4
Distributions of the change in the BR index between 1970-2014
Change b/w 1970-1980
Change b/w 1990-2000
Change b/w 2008-2014
5
Change b/w 1980-1990
Change b/w 2000-2008
Fig. 3.11 Financial, labor, and business reforms: a 10-year angle. Source Own calculations based on Economic Freedom of the World Data, http://freetheworld.com
3 Policies and Reforms
65
tion before the Great Recession, which is indicated by the movement of
the reform distribution to the left between 2000 and 2008. This process of
waning financial liberalization was strengthened by the Crisis, which is a
typical reaction to many systemic crises (Abiad and Mody 2005; Stankov
2012). The process was similar to the policy responses to the Great Depression observed by Mitchener and Wandschneider (2015). After 2008, the
share of financial reformers and non-reformers was roughly similar.
Unlike financial reforms, labor reforms took a dramatic step forward in
the few years preceding the Crisis. This is indicated in Fig. 3.11b which
shows the change in the Labor Market Regulations (LMR) index. In that
particular period, we witness the smallest share of countries backtracking
on their labor market liberalization. At the same time, we see the largest
share of countries undertaking bolder steps to free their labor markets. If
we look at the timing of the most active financial and labor market reforms,
we quickly notice that labor market reforms follow financial reforms, and
normally follow with a 5–10 year lag. As one of the more politically sensitive subjects, it is intuitive that labor market reforms become politically
viable only after other reforms bear fruit.
If we notice the behavior of product market regulation reforms at the
same time, we observe that the most active period for instigating those
reforms was actually after the Great Recession. It is not very often that
we observe a reform area being invigorated by the Crisis. This is depicted
in Fig. 3.11c. It is interesting as most countries reversed their business
freedom immediately before the Crisis. It remains to be seen whether
product market deregulation was due to governments overshooting with
too many regulations in the previous period, or because they thought
deregulation in those areas would be an adequate response to the local
business needs and subdued productivity growth in the aftermath of the
Crisis.
If the above reform dynamics are typical, we can conclude that business
regulation reforms follow labor regulation reforms which in turn follow
financial deregulation. It is also interesting to address whether that is the
typical reform sequence, at least since 1990. Perhaps, many reforming
governments wanted to undertake financial reforms first, because outside
factors convinced them that financial liberalization was the way to go
immediately after 1990, and only then, should they proceed with more
66
P. Stankov
politically sensitive reforms. That is the story of some new democracies
after 1990, including most CEE countries. The process by which their
policies have become more similar to the Western world is termed policy
convergence. It has been studied earlier by Gassebner et al. (2011), among
others. The global policy convergence is reviewed in detail next.
Note
1. See Gwartney et al. (2016, pp. 273–285) for a more detailed presentation
of the methodology and data sources used to design each of the subindices.
References
Abiad, A., and A. Mody. 2005. Financial reform: What shakes it? What shapes
it? American Economic Review 95 (1): 66–88.
Blau, B.M. 2017. Economic freedom and crashes in financial markets. Journal of
International Financial Markets Institutions & Money 47: 33–46.
Crucini, M., and J. Kahn. 1996. Tariffs and aggregate economic activity: Lessons
from the Great Depression. Journal of Monetary Economics 38 (3): 427–467.
Demirgüç-Kunt, A., and R. Levine. 2008. Finance, financial sector policies, and
long-run growth. Policy Research Working Paper Series 4469, The World Bank.
Eichengreen, B., and D.A. Irwin. 2010. The slide to protectionism in the Great
Depression: Who succumbed and why? Journal of Economic History 70 (4):
871–897.
Feldmann, H. 2017. Economic freedom and human capital investment. Journal
of Institutional Economics 13 (2): 421–445.
Garrett, T.A., A.F. Kozak, and R.M. Rhine. 2010. Institutions and government
growth: A comparison of the 1890s and the 1930s. Federal Reserve Bank of
St. Louis Review 92 (2): 109–119.
Gassebner, M., N. Gaston, and M.J. Lamla. 2011. The inverse domino effect: Are
economic reforms contagious? International Economic Review 52 (1): 183–200.
Gawande, K., B. Hoekman, and Y. Cui. 2015. Global supply chains and trade
policy responses to the 2008 crisis. World Bank Economic Review 29 (1):
102–128.
Goldberg, P.K., and N. Pavcnik. 2016. The effects of trade policy. Working Paper
21957, National Bureau of Economic Research.
3 Policies and Reforms
67
Gwartney, J., J. Hall, and R. Lawson. 2016. 2016 economic freedom dataset. Fraser
Institute.
Irwin, D.A. (2012). Trade policy disaster: Lessons from the 1930s, Chapter The
Great Depression and the Rise of Protectionism, 1–48. MIT Press.
Levine, R. 1998. The legal environment, banks, and long-run economic growth.
Journal of Money, Credit and Banking 30 (3): 596–613.
Levine, R. 2005. Finance and growth: Theory and evidence. In Handbook of
Economic Growth, ed. P. Aghion and S.N. Durlauf, vol. 1, Part A of Handbook
of Economic Growth, Chapter 12, 865–934. Elsevier.
Mitchener, K.J., and K. Wandschneider. 2015. Capital controls and recovery
from the financial crisis of the 1930s. Journal of International Economics 95 (2):
188–201.
Stankov, P. 2012. Banking crises and reversals in financial reforms. CERGEEI Working Papers 474, The Center for Economic Research and Graduate
Education—Economics Institute, Prague.
4
Policy Convergence Vs. Welfare
Convergence
4.1
Policy Convergence
4.1.1 Policy Convergence: Definitions and
Importance
The reform graphs in the last chapter imply the existence of a certain
policy trend which begs a deeper look: Economic freedom policies in
most countries around the globe have become more similar since the
1970s. This process is known as policy convergence.
In political science, policy convergence studies are an integral part of
comparative public policy. In those studies, policy convergence is defined
as “any increase in the similarity between one or more characteristics of
a certain policy (...) across a given set of political jurisdictions (...) over
a given period of time” (Knill 2005, p. 768). Economic growth theory
has helped political science distill two types of policy convergence: β- and
σ -convergence. Just as economists use β- and σ -convergence to study if
economies become more similar in certain ways over time, political scientists apply these concepts to the study of policy dynamics. Knill defines
β-convergence in policies as a process in which “laggard countries catch
© The Author(s) 2017
P. Stankov, Economic Freedom and Welfare Before and After the Crisis,
DOI 10.1007/978-3-319-62497-6_4
69
70
P. Stankov
up with leader countries over time, implying, for instance, that the former strengthen their regulatory standards more quickly and fundamentally than the latter” and σ -convergence as a process in which “there is a
decrease in variation of policies among the countries under consideration”
over time (p. 769). Bennett (1991) offers an excellent early review of the
political science literature on the issue.
In economics, policy convergence is understood as a sequence of events
in which “the macroeconomic policies pursued by countries move toward
being identical” (Jackman and Moore 2008, p. 1108). Chang et al. (2013)
require two or more countries to be governed by either left-wing or rightwing parties at the same time for this move to effectively happen.
There are also efforts to study policy convergence between parties or
candidates within countries (Alesina 1988; Köppl-Turyna 2014; Krasa and
Polborn 2012; Laussel and Riezman 2005). In those studies, policy convergence is treated as the political platforms of various parties or candidates
growing closer to the median voter preferences to maximize the chances of
an election win. Alesina (1988) demonstrates that within-country policy
convergence is time-inconsistent due to different party preferences before
and after elections, while Laussel and Riezman (2005) find that voters
prefer non-converging policy, especially in the context of free trade.
Different voter preferences for certain policies across countries and
potentially different effects of those policies mean that we can define and
analyze policy convergence in each policy area. For example, for the case
of monetary policies, “narrowing and finally closing the gaps in macroeconomic stability” (Kasman et al. 2008, p. 341) is also considered to be
policy convergence. Jackman and Moore (2008) acknowledge that the
most thriving field of policy convergence research deals with issues in the
European Monetary Union (EMU). An early form of monetary policy
convergence is the increasing number of countries aspiring to join the
EMU (Andrews 1994). More recently, monetary policy convergence in
the EMU is thought of as the ability of member states to adhere to EMU
membership criteria, especially to the interest rate spreads and price-level
dynamics within the EMU, both studied by Kutan and Yigit (2004, 2005).
Convergence tests for the EMU have been developed by Phillips and Sul
(2007, 2009), and evidence on convergence within the European Union
(EU) has been provided by Monfort et al. (2013). Those studies find that
4 Policy Convergence Vs. Welfare Convergence
71
there are enormous policy shifts toward higher mobility of labor, capital,
and goods within the EU—that policy convergence exists. However, they
also document that these policies may lead to different long-term levels
of income per capita. In other words, as much as they try to imitate the
policies and reforms of the early reformers, some countries will possibly
never catch up in terms of living standards.
Policy convergence has occurred in other policy areas as well: privatization, trade liberalization, and agreements with international financial
institutions, as noted earlier by Brune et al. (2004), by Simmons and Elkins
(2004), and by Belloc et al. (2014). Boockmann and Dreher (2003) study
the specific role the International Monetary Fund (IMF) and the World
Bank played in the process of policy convergence, using a panel of 85 countries between 1970 and 1997. Contrary to the above studies, they do not
find strong evidence of coercion. Their evidence suggests that the World
Bank indeed had a positive impact on the adoption of more economic
freedom reforms, while the existence and the number of IMF programs
did not exert a significant impact on the country-level economic freedom
index. In addition, Belloc et al. (2014) find that left-wing and right-wing
parties prioritize freedom reforms differently even if they agree to unite
under the same policy agenda. They argue that left-wing parties will pursue price and entry liberalization more often than privatization, while the
right-wing governments will pursue privatization first and then liberalize,
as was the case in the network industries in 30 OECD countries up until
2007.
Fiscal policy convergence has also been studied by, among others,
Cassette et al. (2013) and Slemrod (2004). Slemrod (2004) notes that
policy convergence in tax regimes occurred at least partly due to international competition for large businesses. Cassette et al. (2013) find that fiscal
policy decisions are indeed mutually interdependent among 18 OECD
countries running up to 2008, at least in the discretionary component of
fiscal parameters, that is, the “parts of public spending and tax receipts
over which governments retain full discretion” (Cassette et al. 2013,
p. 79).
Fiscal convergence is even more evident in the EU, where the main
source of tax revenue for most governments—the Value-Added Tax
(VAT)—is gradually being harmonized across countries. It is worth men-
72
P. Stankov
tioning, however, that “fiscal policy convergence is harder to achieve than
monetary convergence because of the extra political constraints associated
with fiscal policy reforms.” (Haug et al. 2000, p. 429) This means that
there are areas of policy reform where convergence is harder to achieve,
due to the inner workings of the domestic political process. For example, fiscal systems are far more disintegrated than monetary systems, even
in the EMU, let alone in other supranational integration communities.
Therefore, because of different tolerance of taxation and different preferences for public services provision, there is perhaps less need for fiscal
convergence than for monetary conversion, despite the recent evidence of
some limited fiscal convergence.
Policy convergence was also apparent in other areas of market-oriented
reforms, such as the size of government and government regulations.
Schuster et al. (2013) document that OECD countries converge in deregulation of utilities and telecoms, as well as in the provision of industry
subsidies. Their overall conclusion is that between 1980 and 2007 there
was a “clear trend towards diminishing state influence” (Schuster et al.
2013, p. 95) in state ownership, regulating network-based services and
subsidies to various industries. In addition, capital account liberalization
convergence is studied by Steiner (2010) and by Bicaba and Coricelli
(2015) who re-emphasize the importance of learning from best practices
when designing appropriate domestic policies. Bertola (2016) builds a
model and empirically studies the link between capital mobility and labor
market policies. Increased capital mobility is found to act as a driver of
labor policy convergence. As we saw in the previous chapter, that is exactly
what happens to a typical country.
Policy convergence helps explain business cycle dynamics. Chang et al.
(2013) examine the business cycle synchronization across 14 developed
countries between 1980 and 2010. They find that policy convergence—
in an ideological sense—has a positive impact over the business cycle
correlations. Degiannakis et al. (2016), among others, find that fiscal
policy contributes to business cycle synchronization over time in the EMU.
Yet, the impact of fiscal policy is found to be country specific and varying
over time. Similar conclusions are found for monetary policy convergence
by Bearce (2009).
4 Policy Convergence Vs. Welfare Convergence
73
Jackman and Moore (2008) outline two broad factors driving crosscountry policy convergence: (1) demand for policy convergence from voters for economic and ideological reasons, and (2) external influences. They
study the similarities in monetary, fiscal, trade, and financial policies of
26 Latin-American Countries and find an increasing correlation in policies over time. They contend that Latin-American policy convergence is
influenced by four factors: the size of the country measured by GDP and
by population, concurrence of economic shocks, and geographical and
ideological distance.
Additional factors contributing to the process of policy convergence are
explored by Meseguer (2006). According to the author, policy convergence
across countries is driven by the ability of governments to learn what is
good for their countries based on what has worked well in other countries
before. This ability to learn is aided by the existence of large-scale policy
experiments in Latin America and Eastern Europe in the 1980s and the
1990s, perhaps one of the largest global waves of market-oriented reforms
of all time. The differences in those reforms across countries and over
time have provided a platform to analyze what worked and what didn’t,
hence, for updating both government and voters’ beliefs of what would
work in the future for their own countries. This type of policy-learning
process has been recommended as beneficial for growth also by
Barro (1997, p. 11), who asserts that “...advanced countries would contribute more to the welfare of poor nations by exporting their economic
systems, notably property rights and free markets.”
However, Mukand and Rodrik (2005) demonstrate that learning from
the successful policies of a certain leader country may promote growth only
for a limited set of countries. The countries in this limited set are those
which are a priori similar to the leader country. They test this proposition
for Central and Eastern Europe and find sufficient support for its validity:
Countries farther from Brussels grow more slowly than closer countries,
despite the similarities in their policy-learning process.
Meseguer (2006) adopts two alternatives to learning which can also
potentially drive policy convergence: coercion and imitation. On the one
hand, coercion to convergence happens when international financial institutions impose policy choices on the domestic government in exchange
for loans or debt restructuring. On the other hand, imitation leads to
74
P. Stankov
policy convergence when the domestic governments choose policies not
for their alleged consequences for the local economy but because those
policies tend to be internationally fashionable at the moment. Similar
factors are intuitively explored in the comparative public policy literature
by Holzinger and Knill (2005). Indeed, Meseguer (2006) finds sufficient
evidence that all three mechanisms for policy convergence—learning, imitation, and coercion—worked in a large set of developing and developed
countries between 1951 and 1990.
Hall (2016) also notes an additional mechanism at work behind convergence. If voters are mobile and can “vote with their feet,” governments
would likely adopt measures to make the political and economic environment more like the countries to which voters aspire. Thus, Hall argues,
the capacity of voters to emigrate is a key factor driving reform imitation
across countries over time.
In sum, there is no unified understanding of policy convergence, either
within countries or across countries over time. In this book, policy convergence is understood as policies across countries becoming more similar
over time. There are two ways to explore the dynamics of those similarities. First, the policies of a given country converge to those of another
country within a certain policy domain if the policies grow more similar to
the status quo in another country. The status quo within each policy area
can be gauged by using numerous available policy-specific indexes. Policy
convergence happens if the gap between the indices between two countries
becomes smaller. This is essentially how Knill understands β-convergence
in policies. Second, I also adopt Knill’s definition of σ -convergence in
policies. That is, a set of countries converges if the standard deviation of
their policy indices decreases over time. The time variation of policies can
also be measured with the help of available policy indices.
Studying policy convergence is important for two primary reasons.
First, in order to understand the drivers behind the large-scale global push
toward similar policy agendas over the last 45 years. Second, in order
to understand the effects of policy convergence on a country’s welfare.
Policy convergence may contribute to countries becoming more similar
over time. However, there is no conclusive evidence to confirm that it
has led to welfare convergence across countries over longer periods of
time. Moreover, policy convergence may lead to different distributional
4 Policy Convergence Vs. Welfare Convergence
75
consequences within the converging nations. In turn, this may turn the
political tides against further convergence, especially if the winning interest
groups do not adequately compensate losing groups for rents lost in the
process of adjustment. The latter point was suggested, among others, by
Stahl and Turunen-Red (1995) in the context of free trade. It can naturally
be extended to any policy convergence process, especially in developing
countries.
The distributional consequences of any reform matter more politically
in the developing world, especially if large groups of people are expected
to lose in the short-run due to the reform. Therefore, two issues related to
the consequences of policy convergence are particularly noteworthy. First,
do politically converging countries also converge in welfare? If yes, then
a second issue also becomes relevant: Do politically converging countries
create a political environment that favors further policy convergence?
The next section offers a glance at policy convergence since 1970.
It derives policy trends in five areas: government intervention, property
rights, monetary policies, free trade, and regulations. Moreover, the section
presents evidence of both β- and σ -convergence in all five areas. I use the
Economic Freedom of the World (EFW) data by Gwartney et al. (2016).
I present both direct graphical evidence of policy convergence and regression models to derive β-convergence in policies. The book thereby extends
the recent evidence by Heckelman (2015) and Hall (2016), who review
policy convergence specifically based on the EFW data for the periods
between 1995–2014, and 1980–2010, respectively.
4.1.2 Graphical Evidence of Policy Convergence
In each figure, the level of a certain policy index in the beginning of
a certain period is plotted on the X-axis. Then, the change in the same
index over the subsequent period is plotted on the Y-axis. We are interested
in convergence in two particular periods: the 38-year period between 1970
and 2008, i.e., the period before the Great Recession (Panel A), and the
subsequent 6-year period after 2008 (Panel B). I skip the presentation of
policy convergence for the entire 44-year period between 1970 and 2014
76
P. Stankov
because the graphical evidence of convergence is almost identical to what
happened between 1970 and 2008.
Within each figure, we can observe leader countries and laggard countries at the beginning of each period. Laggard countries are those with lower
levels of economic freedom, whereas leader countries have higher economic freedom. If the line slopes downwards, this means laggard countries
changed their policies more than the leader countries—leaders reformed
little, and laggards instituted more reforms. If that is the case, then laggards are catching up with the leaders, at least in terms of policies and
reforms. This catch-up process means that countries become similar in
terms of policies, providing affirmative evidence for policy convergence.
Moreover, this evidence exists not only within each policy domain, but
also within each of the time periods under consideration.
Consistent with earlier evidence by Nieswiadomy and Strazicich (2004),
the graphical evidence in favor of policy convergence is overwhelming. If
we look at the size of government reforms, we will notice that those countries with large governments in the beginning of the period are trying to
reduce them throughout the entire period between 1970 and 2008. At the
same time, countries with smaller governments in the beginning of the
period have increased their government involvement, which is indicated
by the negative change in the Size of Government index for the many
countries with an index level above 6 in 1970 (Fig. 4.1). If we slice the
data into two periods—before and after the Crisis—we also note a sig(a)
(b)
TGO
JAM
CAF
EGY
SWE
0
1
2
ZMB
COD TTO
IND KEN
FJI
CHL
OMN GAB
ISRBEN ZWE MDG
TUN
RWA
GHA
GBR
MAR
TWN NIC MUS
TZA
AUS
USA
HND
HUN DEU ECU
PAK
IRN TURNER
CIV
PRTNOR
CAN
CMR
SYR
BRA
SGP
NZL
IDN
URY
ARG LKA
BOL
SEN
CRI HTI
PER
BDI
MLTMYS
DOM PHL
HKG
ITA
COL
THA
SLV
MEX
FRA ZAF
PRY
ISL
NLD AUT
IRL
KOR
FIN
BEL
LUX
MWI
GTM
BWA
GRC
CYP
ESP
BRB
DZA
COG
VEN
JPN
MLINGA
3
4
5
6
7
Size of Government Level, 1970
Country Code
Linear Fit
8
9
10
Size of Government: Levels Vs. Reforms, 2008-2014
Size of Government Reforms, 2008-2014
-4
-2
0
2
4
Size of Government Reforms, 1970-2008
0
-4
-2
2
6
4
Size of Government: Levels Vs. Reforms, 1970-2008
GNB
CMR
JOR GTM
BWA
TZA GEO
BDIMOZ
IND
SEN MNG
LKAPRY
KWT
GBRMYS
MDG
NGA MEX
NIC
BFA
SRB
DOM
AGOCHN
CZE
BRB
MKD
ROU
PHL
GHA
AZE
IRLTCD
NZL
COL
HUN
PRT
CAN
LTU VNM
IDN
ZAF
MNE
BHR
LVA
MLT
NAM
CHE
CHL HTI
BGR
KEN
PAN
AUT
POL
ISR
BGD
DZA
CYP
KAZ
BOLAUSBRA
HND HKG
TWN
SYR
MMR
SLV
BHS
SWENLD
BEL SVN
THA
ESP
MWI
LUXHRVDEU
KGZ
BLZ
NPL
EST
UGA
PAK
SGP
UKR
URY
ETH
PER
DNK VEN
LSO
USA
SVK
CAF
KOR
NOR GAB
TUR
COG
CIVBEN
RUS
SLE CRI
FRA
MAR
ARE
ALB
FIN ITA
MLI ECU
MUS
MDA
TUN
FJI
ISL
ARM
RWA
EGY
OMN
BIH
JAM
PNG
GRC
JPNARG
IRN
NER
CODZWE
ZMB TGO
GUY
TTO
MRT
0
1
2
3
4
5
6
7
Size of Government Level, 2008
Country Code
8
9
10
Linear Fit
Fig. 4.1 Convergence in government intervention: 1970–2014. Source Own calculations based on Economic Freedom of the World Data, http://freetheworld.com
77
4 Policy Convergence Vs. Welfare Convergence
(a)
(b)
Property Rights: Levels Vs. Reforms, 1970-2008
Property Rights: Levels Vs. Reforms, 2008-2014
Property Rights Reforms, 2008-2014
-1
0
1
2
Property Rights Reforms, 1970-2008
2
0
6
4
CHL
EGY
PER
TUN
MAR
IND
COL
IRN
ECU
NGA DZA TUR
ARG IDN
MEX
PHL
KEN
SWE
NOR
SGP
NZL
HKGDNK
CHE
AUS
LUX
FRA GBR
CAN
DEU
PRT
NLD
ITAISR ESP
IRL
JPN
TWN USA
BRA
BEL
-2
VEN
0
1
2
3
FIN
ISL
KOR
GRC
MYS
THA
ZAF
4
5
6
7
Property Rights Level, 1970
Country Code
Linear Fit
8
RWA
LSO
GEO
BIH
SLE NPL
MYS
GNB
BOL
SYR
BEL
POL
KEN
KAZ
SEN MKD
TGO BDI
JPN
MNG
MMR
PRYCMR
ZAF
ECU
UGA
GUY
PNG
DZA
LKAPAN FJIMUS
ARE
PHL
KGZ
COG
IRL
BEN
ZWEGTM
MAR
FIN
BGR
LVAPRTEST GBR
PAK
JAM
MOZ
ITA
NLDCHE
CRI
HND
IDN
ISR
CZE
DOM
GAB
SVN
NIC
SRB
TWN
ARM
GHA
ROU
HRV
NER
NZL
URY
HTI AGO
LTU
LUX
COD
GRC
NOR
ZMB
ESPCHL
AUS
SLVTTOUKR
COL
BGD
FRA
CAF
HKG
SGP
CAN
HUN
MDG ARG
ETHRUS
ISL
KOR
BHR USA
AUT
NGA MLIBRA
PER
SWE
AZE BHS OMNDEU
TURTZA
IRN
KWT
VNM
DNK
BWA
MEX ALB
CHN
MWI SVK
CYP MLT
VEN
MRT
MNE NAM
BFA
IND
JOR
THA
BRB
EGY
TUN
MDA
BLZ
TCD
-2
PAK
CIV
9
10
0
1
2
3
4
5
6
7
Property Rights Level, 2008
Country Code
8
9
10
Linear Fit
Fig. 4.2 Convergence in property rights protection: 1970–2014. Source Own
calculations based on Economic Freedom of the World Data, http://freetheworld.com
nificant policy convergence effect in government intervention, with the
convergence process being a bit slower after 2008.
We notice similar trends for how property rights policies convergence
(Fig. 4.2). In general, countries which had the weakest property rights at
the beginning of the period then improve faster than others. At the same
time, countries which had the strongest property rights at the beginning
of a period weakened them more quickly. Naturally, there are notable
exceptions. Venezuela, for example, had just about average property rights
protection in 1970, but worsened them exceptionally quickly. After the
Crisis, Venezuela moved further back, and today, it is among the countries
with the weakest property rights protection policies across the globe.
Similarly to the areas of government intervention and property rights, in
Fig. 4.3, we notice a very strong policy convergence in monetary policies.
The linear fit in the figure for the longer period is so good that there is not
much left to explain the change in the Sound Money index, except the
initial levels of monetary policies. Again, Venezuela stands out as one of
the most stable countries in terms of monetary policies in the beginning of
the period and yet managed to deteriorate its monetary policies to belowaverage standards by the end of the period. The bulk of this deterioration
happened after the Great Recession. Zimbabwe also had one of the worst
instability ratings in monetary policies before the Crisis. However, it seems
that Zimbabwe managed to pull itself out of continuing trouble after 2008,
and has dramatically changed its policy stance ever since, boasting one of
the most remarkable increases in the index of Sound Money in the short
period after the Crisis. Other notable cases of a relatively rapid monetary
78
P. Stankov
(a)
(b)
VEN
0
1
2
3
4
5
6
7
Sound Money Index Level, 1970
Country Code
8
9
4
2
DOM
ETH
0
TUR CYP KOR
JOR
NZL
GBR
SWE
ZMB
IDN
HUN
ITA
NOR
URYPER
JPN
CHL
IRL
ISR
ESP
DNK
FRA
PRT
CHN
SLV
ISL
GTM
BRA
KEN
SGP FIN
COLGHA
GRC
RWA
TZACOD
AUT
LUX
PHLEGY
CMR DZA
HKG
ZAF
CAF
SYR
THA
MLT
TWN
NPL
USA
BHR
MLI
NLD
DEU
IND
TUN
SEN
CAN
BEL
CHE
AUS
MAR
NER
OMN
ECU
TGO
CIV
MYS
BLZTTOPRY
IRN
HND
MUS
BEN
MWI
BWA
PAN
ARG
LKA
GAB
TCD
HTI
DOM
MEX
JAM
MDG
PAK
NIC
BOLKWT
GUY
MMR
COG
CRI
SLE
BGD
FJIBHS
PNG
ZWE
BDI
BRB
Monetary Policy Reforms, 2008-2014
5
0
NGA
ZWE
-2
10
Monetary Policies: Levels Vs. Reforms, 2008-2014
UGA
-5
Monetary Policy Reforms, 1970-2008
Monetary Policies: Levels Vs. Reforms, 1970-2008
10
0
1
2
RWA
CRI
BDI
NGA
AGO
IDN
MMR
UKR
KENHRV
ROU
COD
JOR
AZEPNG KGZ
BRB
LKA
MNG
COG
VNM BEN
OMN
NIC LTU
MNE
JAM
ECU
MKD
BOL
PRY
TGO
GTM
MLI
BGD
MARMDA
BIH
ZMB
CIV
NAM
URY
GNB
NER
MUS
LSO
EGY
PAK
TCD
ZAF
RUS
TZA
BGR
PAN
KAZ
POL
ARE
ARM
TUR
ISR
SLV
CZE
CMR
SWE
SRB
LVA
MDG
PER
BRA
CHE
SGP
GBR
DNK
NOR
GUY
HND
SVK
BFA
SEN
ITA
ESP
COL BWA
NLD
MYS
PRT
BEL
HUN
ALB
IND
TUN
FRA
MLT
MWI NPL
THA
IRL
CHL
GRC
DEU
KOR
MRT
FIN
BHS
AUT
CHN
CAN
EST
LUX
HTITTO
BHR
GEO
TWN
SLE MEX
UGA
GAB
HKG
NZL
AUS
KWT
JPN
USA
FJI
DZA
CAF
ISL
IRN
SVN
PHL
MOZ
BLZ
ARG
GHA
SYR
CYP
VEN
3
4
5
6
7
Sound Money Index Level, 2008
Linear Fit
Country Code
8
9
10
Linear Fit
Fig. 4.3 Convergence in monetary policies: 1970–2014. Source Own calculations
based on Economic Freedom of the World Data, http://freetheworld.com
policy deterioration after the Crisis are: Cyprus, Syria, Ghana, Belize, and
Argentina.
Trade reforms have also exhibited significant policy convergence since
1970, as Fig. 4.4 shows. Most policy makers across the globe realized the
potential benefits of freer trade by 2008. That is perhaps why very few
countries plunge into reform reversals until 2008, and why most free
their trade further. Interestingly, those countries which reversed free trade
reforms before 2008 are exactly those which are the flagmen of free trade
policies in the 1970s: USA, Canada, Germany, Belgium, and Luxembourg.
And then, unsurprisingly, there is Venezuela, which managed to reverse
most of its competitive policy advantage. Between 1970 and 2008, it
repositioned itself from one of the leaders of free trade into the group of
countries with the most unfavorable international trade policies.
(a)
(b)
Free Trade: Levels Vs. Reforms, 2008-2014
4
Free Trade Reforms, 1970-2008
10
0
5
Free Trade: Levels Vs. Reforms, 1970-2008
VEN
0
1
2
3
4
5
6
7
Free Trade Index Level, 1970
Country Code
Linear Fit
8
9
Free Trade Reforms, 2008-2014
-2
0
2
LUX
BEL
MMR
NER
BDI
ZWE
VEN
ALB
MNG HTI
ARM
RWA
MLI
AZE
KGZ
KAZLSO
SVK
PHL
MEX
JPN
NAM
MYS
PAN
MRT
ZMB
UGA
BWA
TUN
CZE
BFA
POL
BIH
TCD RUS
MDG
HUN
NPL
BGR
HRV
MKD
BRB
GRC
ITA
MLT
ROU
VNM
MUS
AUS
CHN
ESP
PRT
IDN
MOZ
BEL
CAF AGO GAB
SLE
BEN
LVA
MDA
FRA
MNE
NLD
SWE
AUT
DOM
KOR
SEN
LUX
ISL
DNK
SRB
COL
IRL
SVN
TGO
BLZ
FIN
THA
TWN
TTO
CAN
JOR
CHE
OMN
EST
LTU
TZAMAR
SLV
DEU
ARE
BHS
NZL SGP
FJI
TURBHR
ISR
COG
HKG
NIC
ZAF
PNG
KEN
ETH BGD
GTM
GEO
CRI
MWI
GNB
PAK
CHL
BRA
NOR
PRY
NGA
CIV
USA
CYPGBR
HND
COD CMR
UKR
BOL
KWT
JAM
URY
EGY GHA
IND
PER
GUY
ECU
DZA
LKA
SYR
IRN
ARG
-4
CHL
MUS
PAK PHL
DOM ISR
PER
NGABOL
ECU
TUN
SLV GBR
HND
KEN
MDG
GTM
IDN
TZA SEN
BRA KOR CYP
ISL
BHS
MLT
PRYESP
COD SYR COL GUY
SWE IRL
MAR
FJI
TWN
NZL
NLD
GRCPRT
MEXNIC
SGP
GAB MWI
THA
AUT
HKG
FIN
NOR
FRA
MMR
AUS
CHE
MYS
JPN CRI DNK
ZAF
COG
URY
PAN
ITA
IRN
USA
CANDEU
-5
TUR
GHA
LKA
NPL
10
0
1
2
3
4
5
6
7
Free Trade Index Level, 2008
Country Code
8
9
10
Linear Fit
Fig. 4.4 Convergence in trade policies: 1970–2014. Source Own calculations based
on Economic Freedom of the World Data, http://freetheworld.com
79
4 Policy Convergence Vs. Welfare Convergence
(a)
(b)
CAN
2
MMR
COD
TGO
GNB
NER
TZA
LTU
AGO
NPL
DEU
BEN
PHLCYP
MLT
DZA CHN
JAM
RUS
TWN ROU
ALB
ISR
BGD
LSO
CIV
ISL
TUR
POL
UKR
MDG
PRT
CMR
ITALUX
MYS
ARM
ETH
MDA
BIH
SWE
BRA
PRY
VNM
KOR
BWA
KAZ
PER
NGA
LKA
CZE
HRV
TTO
MAR
GTM
OMN SGP
BEL
ARG
AZE
ZMB
MEX
URY
NOR
IDN
COL
BOL
NLD
CAF
LVAUGA
EST
FIN
INDDOMKGZ
ECU
MOZ
BGR
BFA
CHE
ARE
MKD HKG
KEN
COG
GRC
IRL
FRA
MRT
SEN
NIC
NAM
MLI
HUN
ESP
DNKNZLFJI
PAK
BDI
HTI
MNG
GHA
PAN
HND
THA
CRI
TUN
MUS
CHL
CAN
JPN
SVK
KWT
ZAF
PNG
MNE
USA
AUS
GEO
SRB
JOR
BHR
EGY
GAB
SVN
AUT
ZWE
RWA BHS
MWI
BRB
GUY
SLE
IRN
SLV GBR
TCD
1
SYR
-1
0
2
SGP
AUS
IDNTWN
ISR
JOR
CHN
IRL
GRC
TZA
PRT TUN
AUT
NZL
THA BEL
NLD DNK
HKG
USA
MDG
SYR
ITAESPZAF
SWEMYS
GTM FRA
GBR
NOR CHE
PAK
KEN
COD
DEU
IND
MMR
ARG
FIN
LUX JPN
VEN
BRA
0
Regulatory Reforms, 2008-2014
4
ROU
-2
Regulatory Reforms, 1970-2008
3
Overall Regulation: Levels Vs. Reforms, 2008-2014
6
Overall Regulation: Levels Vs. Reforms, 1970-2008
0
1
2
3
4
5
6
7
Overall Regulation Level, 1970
Country Code
Linear Fit
8
9
10
VEN
BLZ
0
1
2
3
4
5
6
7
Overall Regulation Level, 2008
Country Code
8
9
10
Linear Fit
Fig. 4.5 Convergence in regulatory policies: 1970–2014. Source Own calculations
based on Economic Freedom of the World Data, http://freetheworld.com
Unlike the period before the Crisis, after 2008, policy trends turned
not only in Venezuela but also in many other countries, which reversed
free trade reforms. Most did so timidly, except Argentina, which reversed
its free trade with strong political bravado. However, even though many
countries did reverse on free trade, the negative slope of the fitted line
after 2008 suggests that evidence of convergence in free trade policies still
exists.
Figure 4.5 presents graphical evidence of policy convergence in yet
another area: government regulation. Countries that started with burdensome and expensive regulations deregulated most over the 38-year period
before the Crisis. As with free trade, there are notable differences in how
the process panned out before and after the Great Recession. Specifically,
before the Crisis, there were very few countries which made their government regulation more burdensome than it was in 1970. Many leader
countries either did not deregulate further or deregulated a little. At the
same time, laggard countries in the beginning of the period deregulated
most. After 2008, we can observe many more countries re-regulating their
credit, labor, and product markets. Still, as with most other policies, evidence of policy convergence persists. The overall evidence of convergence
in regulation exists even if we go one step down to the subindex levels.
This is also the case for the period after the Great Recession. The next
section illustrates how fast the process of β-convergence in all five areas
of economic freedom actually was.
80
P. Stankov
4.1.3 The Speed of Policy Convergence
The speed of policy convergence is how fast various economies converge to
a certain equilibrium level of a given policy, if we assume such equilibrium
exists. We examine the speed of policy convergence in various time periods
by running a simple regression model. In this model, the change in the
index for a given period is regressed on the initial value of the index. The
estimated model is:
Ii = α + β Ii + εi ,
(4.1)
where Ii is the difference between the initial and the end values of
index I . The Ii in levels is the value of the respective index in the beginning of the period. The index I measures each of the five reform areas:
Government intervention (G), Property Rights (PR), Monetary Policy
(MP), Free Trade (FT), and Government Regulation (R), respectively, as
defined in the text. Government regulation is further disaggregated into
its main subindices: CMR, LMR, and BR, as defined in the text, where
data was available for the respective time period. If indeed leader countries
reformed less, then the coefficient estimates should be negative. This is
exactly in line with the results obtained by running the above model. The
results are presented in Table 4.1.
Table 4.1 is organized as follows. Each column of the table corresponds
to a given period of time. Within each period, the table presents the estimates of β for each of the 5 broad policy areas, as well as for credit, labor,
and business regulations. We are able to study convergence in business
regulations only after 2000 because of data limitations. The way the table
is organized enables us to read the results in two ways. First, within each
time period, we can study if a certain policy area was converging faster
than others. Second, for each policy, we can observe the speed of convergence across time, thereby seeing if countries converge faster within
certain decades.
Column (1) presents estimates for the entire period between 1970 and
2014. We can see three groups of policies. The first group consists of those
policies converging very quickly. Those are monetary policies and credit
market regulations, for which β estimates are above −0.9 and are highly
Table 4.1
G
Obs
Adj. R 2
MP
Obs
Adj. R 2
FT
Obs
Adj. R 2
R
Obs
Adj. R 2
CMR
Obs
Ad. R 2
(1)
1970–2014
(2)
1970–1990
(3)
1990–2008
(4)
1970–1980
(5)
1980–1990
(6)
1990–2000
(7)
2000–2008
(8)
2008–2014
−0.773***
(0.077)
90
0.526
−0.399***
(0.070)
50
0.391
−0.933***
(0.080)
107
0.562
−0.786***
(0.055)
75
0.734
−0.674***
(0.096)
46
0.516
−0.938***
(0.041)
96
0.847
−0.664***
(0.079)
90
0.436
−0.281***
(0.085)
50
0.169
−0.556***
(0.135)
107
0.131
−0.288***
(0.074)
74
0.162
0.117
(0.114)
46
0.001
−0.452***
(0.063)
96
0.348
−0.589***
(0.067)
114
0.400
−0.378***
(0.065)
111
0.232
−0.817***
(0.050)
120
0.691
−0.707***
(0.040)
111
0.739
−0.579***
(0.052)
115
0.521
−0.765***
(0.044)
121
0.718
−0.516***
(0.077)
90
0.331
−0.353***
(0.075)
50
0.305
−0.473***
(0.099)
107
0.172
−0.214**
(0.089)
75
0.061
−0.056
(0.077)
46
−0.011
−0.324***
(0.051)
96
0.291
−0.427***
(0.072)
108
0.242
−0.125**
(0.059)
90
0.039
−0.225**
(0.102)
112
0.034
−0.204***
(0.049)
102
0.140
0.062
(0.063)
103
−0.000
−0.175***
(0.050)
112
0.094
−0.359***
(0.069)
114
0.186
−0.277***
(0.075)
111
0.103
−0.676***
(0.061)
120
0.502
−0.481***
(0.050)
110
0.457
−0.527***
(0.047)
115
0.526
−0.595***
(0.050)
119
0.540
−0.470***
(0.057)
123
0.354
−0.266***
(0.038)
123
0.287
−0.506***
(0.053)
123
0.421
−0.442***
(0.040)
122
0.501
−0.285***
(0.064)
123
0.132
−0.514***
(0.054)
123
0.422
−0.183***
(0.055)
141
0.067
−0.105***
(0.025)
141
0.107
−0.232***
(0.045)
141
0.155
−0.108**
(0.044)
141
0.034
−0.276***
(0.039)
141
0.261
−0.267***
(0.055)
141
0.137
81
(continued)
4 Policy Convergence Vs. Welfare Convergence
Obs
Adj. R 2
PR
The speed of policy convergence: 1970–2014
82
LMR
Obs
Adj. R 2
BR
Obs
Adj. R 2
P. Stankov
Table 4.1
(continued)
(1)
1970–2014
(2)
1970–1990
(3)
1990–2008
(4)
1970–1980
(5)
1980–1990
(6)
1990–2000
(7)
2000–2008
(8)
2008–2014
−0.487**
(0.175)
20
0.263
−0.146
(0.144)
20
0.001
−0.342***
(0.099)
45
0.199
−0.099
(0.138)
20
−0.026
−0.056
(0.058)
21
−0.004
−0.440***
(0.066)
44
0.499
−0.103
(0.104)
74
−0.000
−0.474***
(0.074)
74
0.356
−0.155***
(0.033)
140
0.135
−0.515***
(0.057)
140
0.369
Notes The table reports β estimates from an OLS equation for each of the above time intervals. Within each of the intervals,
the estimated equation is: Ii = α + βIi + εi , where (I )i is the difference between the initial and the end values of index
I. The Ii in levels is the value of the respective index in the beginning of the period. The index I measures any of the five
reform areas: Government intervention (G), Property Rights (PR), Monetary Policy (MP), Free Trade (FT), and Government
Regulation (R), respectively, as defined in the text. Government regulation is further disaggregated into its main subindices:
CMR, LMR and BR, as defined in the text, where data was available for the respective time period. Data source: EFW 2016
index. Symbols: * p < 0.10, ** p < 0.05, *** p < 0.01
4 Policy Convergence Vs. Welfare Convergence
83
statistically significant. The second group of policies includes government
intervention, overall government regulation, and free trade, for which
policies do converge quickly but not with the vigor we observed previously.
The β estimates for these policies stand between −0.6 and −0.8. The final
group of policies is the one in which we can still observe a highly significant
convergence process, and yet, it was only half as fast as the one for monetary
policies and financial regulations. Property rights reforms and labor market
regulations are notably the slowest policy areas to converge across countries
between 1970 and 2014.
Columns (2) and (3) address the issue of policy convergence within 18to 20-year periods. Thus, we can observe the differences between the speed
of convergence across countries in each of those two periods. It is easy to
see that policy convergence was faster in almost all areas between 1990
and 2008 than in the previous period back to 1970. Monetary policies and
financial regulations still lead the way with the fastest convergence speeds,
followed closely by trade policies, government intervention, and overall
regulation. Property rights and labor market regulations still converge in
the slowest manner between 1990 and 2008.
Columns (4) through (8) slice time into decades, and even smaller
periods of 6 to 8 years. Importantly, we are able now to see the differences
between the speed of convergence before and after the Crisis. If we adjust
the β estimates to represent annual changes in the indices, we notice that
the fastest policy convergence period was actually immediately before the
Crisis, while it was slowest immediately after 2008. The one exception was
business regulations, which converged faster after the Crisis. The dramatic
differences in how countries converge in policies before and after the Crisis
would be a worthy topic of a separate study.
It is also interesting to mention that, in addition to the convincing
evidence of β-convergence, we can observe graphical evidence of σ convergence. This means cross-country differences in policies gradually
disappear over time. The cross-country difference in policies is measured
with the standard deviation (SD) of a given policy at a certain point in
time. Then, the behavior of those standard deviations is monitored over
time. If the SDs are growing smaller, there is evidence of σ -convergence.
That evidence exists and is presented in Fig. 4.6.
84
P. Stankov
Standard Deviation of a Policy Index
1.5
2.5
3
1
2
Sigma Convergence in Policies: 1970-2014
1970
1975
1980
1985
1990
1995
Time
Government Intervention
Monetary Policies
Regulation
2000
2005
2010
2015
Property Rights
Free Trade
Fig. 4.6 Sigma convergence in policies: 1970–2014. Source Own calculations based
on Economic Freedom of the World Data, http://freetheworld.com
To conclude, with occasional exceptions, no matter how you slice time,
and no matter at which policy index you narrow your focus, policy convergence is there to be seen. It is powerful and is highly statistically significant.
It is as if a gravity force is moving the world closer to an equilibrium policy model. However, no matter how similar countries become in terms of
policies, it still remains to be seen if that policy convergence translates to
welfare convergence. That is what the next section and the next chapter
are about.
4.2
Welfare Convergence: Graphical
Evidence
There is now an abundance of literature on income convergence across
countries over time. The basic idea is that poorer countries should, in theory, grow faster than richer countries. This is mainly because the returns to
investment in capital are higher at lower levels of capital, and poorer countries are naturally less endowed with capital than richer ones. However,
85
4 Policy Convergence Vs. Welfare Convergence
those theories of convergence have long been under scrutiny by empirical economists who find that evidence of convergence exists only among
similar countries, for example, the OECD (Ben-David 1998; Bentzen
2005; Strazicich et al. 2004) or the least-developed countries (Zind 1991)
or even within countries over time (Carlino and Mills 1993). Once the
entire number of countries is taken into consideration, evidence of absolute income convergence becomes weak (e.g., Barro and Sala-i-Martin
(1992, 1997) and, more recently, King and Ramlogan-Dobson (2015),
among others).
This section examines convergence not only in income per capita but
also in the other measures of welfare. Figure 4.7 presents graphical evidence of some income convergence, while Fig. 4.8 does so for consumption, Fig. 4.9 presents evidence of life expectancy convergence, and finally,
Fig. 4.10 plots initial income inequality against the change in income
inequality within a country over a subsequent period. If we observe a negative relationship between the initial values of some of the welfare measures
(b)
0
QAT
BRN
ARE
6
7
8
9
10
11
12
30
20
GNQ
BIH
VGB
ARG
EGY IRN
IRL
SGP
QAT
KWT
LBN
KOR
MLTTWN
CHN
MDV
BTN
POL
DOM
NOR LUX
BWA
MYS
GRD
YEM VNM
BHR
IND CPV
PRT
NLD
ISR
TUN
OMN
ESP
KNA
MUS
TCA
PAN
BRA
AUT
THA
HKG
PSE
LKA
PER
CHL
GBR
GRC
MMRLSO
HTI
LAO BOL
DNK
UGA
MLI
BEL
BHS
CRI
SWE
BRB
ITA
DEU
GTM
AUS
BMU
FIN
SWZ
NZL
SYC
VCT
CYP
FRA
USA
NPL
SYR
JAM
CAN
URY
ISL
MOZ BFA
PHL
KHM
HUN
JPN
ALB
IDN
GHA
BEN
COG
SVN
ATG
SDN
MEX
TTO
MAC
MKD
ABW CYM
NIC
TUR
DZA
HND
CHE
DMA
GAB
ROU
BLZ
MAR
JOR
PAK
EST SAU
AIA
NAM
SUR
GNB
LCA
ZAF
GMB
MRT
ECU
COL
FJI
TZA
BGD
SRB
MDG
LBR NGA
PRYIRQ
HRV CZE
BRN
CMR
DJI
STP
KEN
CIV
SEN
MWI
LTU
ZMB
CAF
SVK
TCD
TGO
NER
MNG UZB VEN
TKM
BDIRWA COM
MNE
GIN AGOZWE
BGR
LVA MSR
SLE
ETH
BLR
KAZ
RUS
ARM
UKR
COD
AZE
MDA
KGZ GEO
TJK
10
5
MAC
KOR
SGP
IDN
CHN
EGY
MNG
TWN
MDV
SLV
MLT
KNA
ROU
MYS
LAO
MMR
HKG
VNM
GRD
THA TCA ABW
ARG
POL
ATG
HUN IRL
BTNPRY
CPV
BRA
MUS
IRN
TUN
IND
DOM
PANSYC
CYP
NOR LUX
BGR
PRT OMN
IRQ
ECU
BOL
SWZ
LKA
AUT
LCA
SUR
MAR
JOR
ALB
VCT
COG
ESP
PER
DEU
AIA
CHL
TTO
PSE
TUR
LSO SDN
COL
BMU
ITA
BHR
JPN
BEL
FIN
GBR
NLD
PAKPHL DMA
ISL
GAB
MLI NPL
GRC
VGB
ISR
ETH
CRI
FRA
MSR
DNK
LBNURY
CAN
SWE
AUS
FJI
USA
NZL
BLZ DZA
CHE
BFA
KWT
NAMMEX BHS
CYM
HNDGTM
CMR
BGD
MOZ
KHM
UGA
AGO
VEN
STP
RWA KEN
HTI SEN
SYR NGAJAMZAF
GHA
TZAMRT
CIV
SAU
BDI
SLE
BEN
BRB
TCD ZMB
TGO
GNB
MWI
MDG
GMB
GIN
COM
ZWE
NIC
NER
LBR COD
CAF
DJI
SLV
0
Compound GDP/c. Growth, 1990-2000
BWA
Initial GDP/c. and Growth, 1990-2000
-10
10
Initial GDP/c. and Growth, 1970-2014
GNQ
-5
Compound GDP/c. Growth, 1970-2014
(a)
13
6
7
8
Initial Ln(GDP/c.), 1970
Linear Fit
Country Code
(d)
20
AZE
YEMSYR
0
10
OMN
QAT
TTO
KAZ
ARM
MNG
AGO
MAC
GEO UKR VEN
IRN
MMR
SLV
ZMB
SAU
KWT
ROU
JOR
CHN LBN
LVA
RUS
EST
KHMMDA
BLR
LTU
BHR
MDV
SUR
IND
PER
VNM
ECU
BGR
ALB
SVK
GAB
MNE
CPV
LAO
PRY
BRN
ETH
BIHTHA
MOZ
HRV
TZATJK
SGP
EGY
IDN
RWA
TCAAIA BMU
CHL
ATG
SRB
BOL
NAM
MYS
PAN
GRD
COG
BTN
LKA
NOR
MWI
UZB
BGDSDN
DZA
TCD
TKM
COL
HKG
HUN
MLI
IRQ
POL
CYP
GRC
CZE
SVN
KOR
VCT
PAK
UGA
BRA
TUR
KNA ESP
KGZ HNDMAR
MKD
ZAF
IRL
MRT
LCA
JAMDOM
STP
BWA
GTM
LSO
BEN
FIN
CHE LUX
MEX
TUN
ARG
URY
AUS
PRT
CRI
SLE
NPL
LBR
DEU
BFA
NLD
AUT
SWE
ISL
DNK
GHA
FRA
TWN
CAN
BLZDMA
BEL
GBR
CIV
NZL
KEN
MSR MLT
USA
ITA
NER
SEN
DJI NICPHL FJI
JPN
CMR
ABW
SWZ MUSSYCBHS
TGO
PSE
COM
CYM
GMB
BDI CAF GNB
BRB ISR
HTI
VGB
ARE
GIN
COD
MDG
ZWE
7
8
9
10
11
Initial Ln(GDP/c.), 2000
Country Code
12
13
Linear Fit
12
13
Linear Fit
30
COD
20
NGA
6
11
Initial GDP/c. and Growth, 2008-2014
MDG
MMR
ZWE
MAC
TKM
LAO
IRQ
MNG
LKA
IDN
ETH
CHN PAN
URY
BGD
IND
MDA
ZMB
VNMSLV
AGO
BDI
KAZ
UZB
NPL
GEO
STP
PRY
CIV
BLR
GHA
COGPHL
POL
EGY
NAM
LBR RWATZA MRT
SDN
LTU
AZE
SUR
CHL
ROU
COL
BRA
ALB
BOL
MAR
SYC
PER
KEN
PAK
BWA
JOR
TUR
GNQ
BFA
ECU
MKD
KHM
UGA
DOM
KGZ
TJK
EST
NGA
TGO
MUS
SLE
THA
CRI
GINTCD
TWN
RUS
FJI
HUN
ARM
BGR
SGP
MYS
DJIPSE
ARG
KOR
NIC
BIHSRB
SVK
SAU
GTM
LSO
LVA
MLT
NZL
KNA
MOZ
BTN
SEN
CZEDEU
CMR
NER
AUT
MNE
HNDCPV
CHE LUX QAT
BLZ
BEL
ISR
HKG
BEN
MEX
FRA
SWZ
DNK
MSR
IRL
PRT
UKR
ZAF
MLI
SWE
TUN
MDV
NOR
LCA
USA
JAM
GBR
HTI
JPN
AUS
HRV
DMA
GRD
VCT
CAN
NLD
SVN
COM
DZA
LBN
FIN
ITA
GNB
ISL
ESP
IRNATG
VEN
BHR
GMB
YEM
TTO
SXM BMU
CUW
GAB BHS
VGB
ABW
CYM
ARE
BRN
GRC
SYR
CYPOMN
MWI
KWT
TCA
BRB AIA
CAF
10
Compound GDP/c. Growth, 2008-2014
GNQ
0
30
Initial GDP/c. and Growth, 2000-2008
-10
Compound GDP/c. Growth, 2000-2008
(c)
10
Initial Ln(GDP/c.), 1990
-10
Country Code
9
ARE
5
6
7
8
9
10
11
12
13
Initial Ln(GDP/c.), 2008
Country Code
Linear Fit
Fig. 4.7 Income per capita convergence: 1970–2014. Source Own calculations
based on Economic Freedom of the World Data, http://freetheworld.com
86
P. Stankov
(b)
QAT
BRB
6
7
8
9
10
11
12
15
GNQ
10
SWZ
PSE
SGP
MDV
CHL
LAO CPV
CYM
MLT
TWN
DOM
BLZ
MYS KOR IRL
ISR
ARE
GIN
LKA NAM
CYP
IND BTNIDN
POL
BFA
SLV PAN
HTI
ARG
VNM
KHM
BEN
MUS
BWA
THA
URY
JOR KNA HKGABW
NIC
YEM
SDN
MLI
TUN
GTM
NPL GHA
NGA
PAK
CRI
OMN
LBR
EGY
TUR
MAC
MOZ MMR
ATG
TCA
SYC
LSO HND
PHL
KEN
BOL
BRA
TZA
ETH
BHR
GRD
AUS
BHS
PER
PRY
NOR
PRT
USA
ISLLUX
IRQ
MEX
NLD
TTO
IRN
LBN
COL
GBR
MAR
JAM
GMB
SYR
SVN
ESP
FJI
MRT
BGD
BRN
MWITCD
NZL
GRC
SEN
DMA
AUT
NER
TGO
SAU
BRB
VCT
DNK
ZAF
MKD
DEU
CAN
CIV
MDG
BEL
FRA
CMR
DJI
ECU
JPN
GNB
KWT
STP
COM
SWE
ITA
QAT
VEN
FIN
CHE
ARM DZALCA
CAF
EST CZE
BMU
GAB
ROU
HUN
ALB
COG
MSR
ZMB
SVK
BLR SUR
UZB
RUS
ZWE
HRV
RWA
LTU
BDI
AGO
AZE
MNETKM
MNG
COD
BGR
SLE
SRB
LVA
UKR KAZ
MDAGEO
TJK
KGZ
13
5
6
7
Initial Ln(Cons./c.), 1970
Country Code
(d)
TCA
QAT
AZE
IRQ
UKR
NGA
KAZ
MDV
BLR
MNE
CHN ALB
ARM ROU
RUS
KGZ
MNG
GEO
YEM
MDA
AIA
VEN
ARE
VCTMYS LTU
KWT
ZMB VNM
OMN
TZA
COG
LVA
BGR
RWA
EST
GRD
ATG
SYC
IND
JOR
SRB
SAU
BHR
UGA STP BTN BWA
MWI
DOM
TKM
BDI
IRN
CHLTTO MAC
SYR
THA
BFA
HND
CPV
LKA
NAM
GHA
BIHPAN
ECU
SDN
MUS SVK CYP
MRT
TUN
DZA
TCD
ZAF
SEN
PER
DMA
BGD
IDN
MAR
BEN
BRN
PAK
COD
IRL
PHL
COM
TURKOR
GMB
EGY
SGP
MSR
NPL
GTM
CMR
NER MLI
LBN
HRV
COL
POL
LCA
MKD
FJI
NZL
GRC
AUS
CRI
UZB
KEN NIC
DJI
ESP
GAB
PRY
MEX
SLVBRA
NOR
ARG
CAN
PSE
CZE
ISR
ISL
HKG
LAO BOL
KNA
HUN
SVN
MDG LSO
JAM
FIN
GNB
GBR
BMU
LUX
TGO
BLZ
USA
SWZ
CYM
ABW
GIN CIV
TWN
MLT
DNK
FRA
SWE
NLD
ZWE
URY
AUT
PRT
BEL
CHE
CAF
BHS
JPN
HTI
DEU
SUR
BRB ITA
10
ETH
5
MOZ SLE
KHM
VGB
5
6
7
8
9
Initial Ln(Cons./c.), 2000
Country Code
10
11
12
10
Linear Fit
11
12
Linear Fit
20
ZWE
15
AGO
COD
BDI
10
TJK
MMR
9
Initial Cons./c. and Growth, 2008-2014
QAT
UZB
MNG
TJK BTN
TCD MMR
CHN
AZE
GIN
AGO
NAM
OMN
BHR
IND
RWA
LAO
MOZ
MAC
IRQ KAZ
LKA
SAU
ETH
KEN NGA PER
COG
UGA
TZA
BWA MYS
ZMB
KHM
NPL
COM
GHA
VNM
PSE
URY
GNQ
CHL
PRY
DJI
IDN
CIV
NER TGO
ARG
STP
CMR
DZA
PHLGAB
LBN
MRT BOL
BFA BGD
EGY
BLR
COL
MARBLZ
ECU
HKG
DOM
LBR
THA
TUN
LSOPAK
TTO
MLI
BRN
CRI
SYC
PAN
TUR
GTM
BRA
SGP
ISR
BEN
GMB
KGZ
SENSDNNIC
MSR
HND
MDASUR
MUS
RUS
LUX
KOR
AUS
ZAF
TWN
NOR
NZL
ARM
MEX
POL
CAN
HTI
KWT
CPV ALB
MLT
CUW
GNB
MDV
CHE
SWE
MWI
UKR
IRN
JPN
USA
JOR SVK
GEO
DEU
SLE
BEL
SWZ
VCT
GBR
AUT
FIN
CZEFRA
MDG
MKD
DNK
EST
ROU
VEN
BGR
FJISLV
NLD
GRD
BIH
ISL
ABW
SRB
BHS
LVA
SVN
LTU
HUN
VGB
IRL
ITA
KNA
MNE
CYM
PRT
ESP
CYPBMU
ATG
DMATCA
HRV
YEM
JAM
LCA
SXM
ARE
BRB
CAF
TKM
GRC
AIA
SYR
5
GNQ
0
LBR
Compound Cons./c. Growth, 2008-2014
15
Initial Cons./c. and Growth, 2000-2008
0
Compound Cons./c. Growth, 2000-2008
(c)
8
Initial Ln(Cons./c.), 1990
Linear Fit
-5
Country Code
VGB
AIA
BIH
CHN
UGA
5
MYS
Initial Cons./c. and Growth, 1990-2000
0
6
4
ARE
-5
TCA
OMN
MAC
IDN
AIA
TWN SGP
CYM
ETH
MNG KORABWHKG
BTN
THA
LKA
PSEIRQ
VNM
IND
LAO
TUN
EGY
MMR
TZA
MLT
DOM
BHR
SYC
ZWE
MUS
LSO
PAK
KWT
PRY
IRN
NGA
CYP
COG CPV
KEN
NAM
DZA
MAR
GRD
MOZ
CHL
TURJOR ISR
BFA
PHLECU
UGA
BDI
NPL
BLZ
SDN
SAU
BRA
PAN
COL
MLI
STP
VGB
GINTCDMRT HND
ROUKNA
MWI
BRN
ATG
CMR
GAB
FJI
RWACOM
CRI
MEX ISLAUS
ALB GTM
SYR
VEN
GHA
IRL LUX
GMB
VCT
BEN
AGO ZAF
TTO
BOL
NOR
SEN
CIV
DMA
POL
LCA
ARG
PER
KHM
BHS
CAN
TGO
LBR
GNB
USA
JPN
PRT
ESP
HTI
BGD
FIN
DJIURY
GBR
SLV
BGR
NZL
FRA
NER
ZMB
AUT
GRC
NLD
BEL
NIC
DEU
LBN
COD
ITA
JAM
SWE
BMU
CHE
HUN DNK
SLE
SUR
MDG
CAF
MSR
2
Compound Cons./c. Growth, 1990-2000
8
BWA
MDV
SWZ
GNQ
-10
Initial Cons./c. and Growth, 1970-2014
CHN
0
Compound Cons./c. Growth, 1970-2014
(a)
5
6
7
8
9
10
Initial Ln(Cons./c.), 2008
Country Code
11
12
13
Linear Fit
Fig. 4.8 Consumption per capita convergence: 1970–2014. Source Own calculations based on Economic Freedom of the World Data, http://freetheworld.com
and their subsequent growth, then we can tentatively conclude that there
is graphical evidence of welfare convergence across countries over time.
Each of the figures presents 4 sub-figures corresponding to various
periods of time. The first sub-figure reviews longer-term development in
welfare convergence, typically a 44-year period. Then, welfare convergence
is studied after 1990 in 6- to 10-year periods, similarly to the analysis of
policy convergence.
Indeed, as the empirical evidence of absolute income convergence previously suggested, there is almost no income convergence over time across
countries. This is seen in Fig. 4.7. We would expect a very poor country to
grow much faster than a richer one. However, the average very poor country grew only 1–2 percentage points faster than a very rich one between
1970 and 2014, as Fig. 4.7a suggests. This corresponds to a halving of the
income differences in a period of about 50 years, which is a dire prospect
for a poor country.
87
4 Policy Convergence Vs. Welfare Convergence
30
35
40
45
50
55
60
65
70
75
10
RWA
BTN
NPL
MDG
BGD
NER
IRN
TUR
BOL IND
MOZ ETH
NIC
GTM
LAO
LBR
OMN LBN
PER
KHM
SLV
BRA
KOR
CPV
CHL NZLHKG
AGO
EGY
ECU
MAR
HND
CZE
ARM
MEX
DZA
BDI
BIH
MMR
GMBHTI YEM
SAU
MLT
LUX
POL
ITA
DEU
VNM
ARE
CHN
COL
PAK
SGP
FIN
MNG IDNAZE
SYR
AUT
VEN
FRA
CHE
MLI GNB BEN
PRT
ISR
ALB
TUN
MUS
BRN
AUS
JPN
ARG
URY
ESP
SVKSVN
MYS
SWE
NOR
MKD
BHR
PRY
PAN
BRB
FJI DOM
HUN
GBR
CAN
JOR
DNK
CRI
BEL
MRT
BHS
ISL
GEO
CYP
LKA
ROU
USA
QAT
UGA
GIN
NLD
LVA
KWTIRL
BFA
GRC
EST
LTU
TJK PHLSUR
TCDCOD
HRV
SEN
TTO
TZA
THA
MWINGA
JAM
KGZ
GHA
BGR
MDA
ZMB
MNE
CMRTGO
GAB
UKR
KAZ
RUSBLZ
CAF
COG
CIV
NAM
ZAF
KEN
SWZ
LSO
BWA
SLE
0
IRN
BGD DZA
TUN
TUR
BOL
YEM
GMBMWI
GIN
SAU
LBR ETH
MAR
NIC
PER
HND
MDG
IND
KOR
LAO
RWA
GTM
BFA
CPV
EGY
CHL
TZA
SLV ECU
GAB
BEN
CHN
SLE
VNMARE
AGOMOZ
BRA
THAMEX
DOM
MMR IDN
COD HTI
SGP
MNG
JOR
PRT
GNBBDI TGOMRT PAK
COL BHRLBN
CRI
HKG
LUX
NAM
PAN
GHA
SVN
BRN
JPN
NGA
MLT
AUS
ITA
MUS BIH
ESP
ZMB
FIN
ISR
SYR
ALB AUT
FRA
GRC
MYS
TCD
FJI
BRB
KGZ LKA
IRL
NZL
QATDEU
BWA
CHE
UGA KEN
BEL
TJK
ARG
BHS
KAZ VEN
CAN
HRV
GBR
MKD
COG
CZE
CAF CMR
KWT
URY
USA
ISL
CIV
NLD
NOR
CYP
GEO
PHL SURPRY
JAM
POL
SWE
EST
DNK
ROU
HUN
MDA
SVK
MNE
AZE
TTO
ARM
BLZ
LVA
ZAF
BGR
LTU
ZWE
RUS
SWZ
UKR
LSO
-10
Change in Life Expectancy, 1990-2000
OMN
NER
Initial Life Expectancy and its Change, 1990-2000
ZWE
-20
30
10
20
MLI
NPL
SENKHM
20
(b)
Initial Life Expectancy and its Change, 1970-2014
BTN
0
Change in Life Expectancy, 1970-2014
(a)
80
30
35
40
Life Expectancy at Birth, 1970
Linear Fit
KHM
BTN
NAMSEN
NPL
AGO
LAOBOL
ZWE
TJKBGD
MDG
KOR
NGAMOZGIN
MNG
IRL
LBN
TUR
EST
SVN
AZEMAR
GMB
NIC
LKA
HRV
OMN
IND
GHA
DZA
BEN
IRN
CHN
SGP
CHL
PER
ARM
LUX
BRA
CPV
CIV
THA
HUN
BDITGO HTI
ALB
SLV
CHE
AUT
SRB
MMR RUSGTM
NLD
PRT
ESP
AUS
FRA
LVA
FIN
GRC
CZE
GEO
ISR
BHS
JAM
BEL
NOR
ISL
COL
SURPRY
GBR
ARE
DNK
DEU
POL
YEMPAK KAZ
MDA
ECU
NZL
DOM
ITA
TUN
CAF TCD GNB
BRN
SVK
GAB
CAN
HND
IDN
JPN
VNM
HKG
SWE
MLT
MEX
FJI
URY
ARG
ROU
USA
MRT
PAN
EGY
JOR
BGR
BRB
BIH
BHR
MKD
MNE
VEN
SYR
MYS
CYP
CRI
CMR
SAU
PHLBLZ
TTO MUS
KWT QAT
UKR
KGZ LTU
SWZ
LSO
50
55
60
65
70
Life Expectancy at Birth, 2000
Country Code
Linear Fit
75
80
75
80
85
Linear Fit
10
ZWE
MWI
ZMB
COGTZA
NER
KEN RWA
KAZ
ETH
ZAF
NAM
UGA
GIN
SEN
TGO
BWA
EST
GAB
BOL
MLI
COD
UKR
MNG
MOZ
MDA ROU
BDI
LAOKHM
CMR
TCD
BTN
NPL
RUS
BGR
AGO
NGA GNB BFA LBR HTI MDG
BGD MAR
KOR
IND
IRN
CIV
PRT
HUN
LTU
MLT
DNK
SVK
KGZ
ESP
NIC
CHL
SGP
BRN
MNE
LVA
SVN
POL
BRA
SLV
LUX
TUR
SRB
OMN
MUS
YEMMMR
HKG
MRT
SWZ
DZA
PAK TJK
FIN
GBR
LBN
HRV
GTM
GMB
SUR
JAM
GRC
BEN
PER
CZE
GHA
THA
ECU
ITA
ISR
CHN
ALB
NOR
FRA
EGY
DOM
QAT
CAN
IDN
DEU
CYP
PAN
HND
COL
FJI
ARE
IRL
NLD
NZL
GEO
TTO
JPN
ARG
MEX
BHS
BRB
VNM
CRI
JOR
BIH
BEL
AUT
USA
MKD
SAU
URY
SWE
AUS
CHE
CPV
BHR
MYS
VEN
ARM
AZEPRY
PHL
KWT
LKA
BLZ
ISL
SYC
TUN
5
TZA
ETH
COG
MLI
KEN
COD
LBR
NER
BFA
45
70
Initial Life Expectancy and its Change, 2008-2014
SLE
CAF
LSO
0
Change in Life Expectancy, 2008-2014
15
10
5
0
ZMB
MWI
UGA
ZAF
40
65
15
(d)
BWA
RWA
35
60
Country Code
Initial Life Expectancy and its Change, 2000-2008
SLE
55
Life Expectancy at Birth, 1990
-5
Change in Life Expectancy, 2000-2008
(c)
50
SYR
-5
Country Code
45
40
45
50
55
60
65
70
75
80
85
Life Expectancy at Birth, 2008
Country Code
Linear Fit
Fig. 4.9 Life expectancy convergence: 1970–2014. Source Own calculations based
on Economic Freedom of the World Data, http://freetheworld.com
The prospects were even gloomier if we zoom in on certain time periods.
For example, the relationship between the initial values of income per
capita and its subsequent growth rate for the 10-year periods between
1990 and 2000, and between 2000 and 2008, was virtually flat.
The only time in which the income convergence theory goes hand in
hand with the empirical findings is the period after the Great Recession.
Between 2008 and 2014, there is a definite tendency for richer countries
to grow more slowly than poorer countries. There are two reasons. First,
the Crisis originated in the richer countries. Therefore, they were naturally
hit harder by the Recession. Second, poorer countries have been recently
documented to be “decoupling” from the richer world (Kose et al. 2012).
They are trading far less intensively with developed countries, and far
more among themselves. What this means is that it takes one or two large
developing economies to pull many others from poverty (Altenburg et al.
2008). After 2008, these two countries were China and India.
88
P. Stankov
(b)
BGR
GBR
USA
IDN
CZE
KOR
COL
PAK
IRN
JPN
BRA
ITA
DEU
FRA
20
(c)
25
30
35
40
45
50
55
60
Change in Income Inequality, 1990-2000
0
10
20
30
-20
-10
ZAF
MKD
CRI
USA
GBR
CAN
ARM
PRT
SWE NLD
ESP
CYP
AZE
JPN
DNK
IRN
HND
VEN
THA
BRA
NOR
AUS
15
20
25
30
35
40
45
50
Country Code
Linear Fit
(d)
BOL
UKR EST
LVA
BGRPOLMDA
SVKCZE KGZ
ROU
LTU
TUN
CHN
FRA ESP
SWE IND
GBR
MKD
CAN
USA
CRI
IRN
VEN
THA
CHL
ZAF
DNK NOR
JAM
25
BOL
IDN
SVK
EST
CZE FIN
MDA
UKR
FRA
KAZ
Country Code
RUS
GEO
20
PRY
LTU
POL
Initial Income Inequality (Gini), 1990
LKA
15
RUS
KGZ
ROU
BGR GEO
LVA
Initial Income Inequality (Gini), 1970
Initial Income Inequality and its Change, 1990-2000
FIN
Initial Income Inequality and its Change, 1990-2008
Change in Income Inequality, 1990-2008
5
-5
0
10
15
-10
Change in Income Inequality, 1970-2008
-5
10
15
-10
0
5
Initial Income Inequality and its Change, 1970-2008
30
35
40
45
50
55
60
65
60
65
Linear Fit
ZAF
MKD
ROU
KGZ
MNG
BGR
CRI
TUR
URY
FRA CAN
MRT
USA
DEU LTU LVAGBR
RUS
ITA
NOR
AUT
LUX
HRV
CHE
GRC
SVNSVK
SWE
IRN
EGYPOL
GEO
SRB
THA
IRL
CZE
MNE
ESP
VEN
HUN
BEL
DNK
FIN
UKR
20
25
30
35
COL
PER
DOM
ARG MEX
SLV
ECU
EST
MDA
40
BOL
45
50
Initial Income Inequality (Gini), 1990
Initial Income Inequality (Gini), 2000
Country Code
Country Code
Linear Fit
55
Initial Income Inequality and its Change, 2000-2008
Change in Income Inequality, 2000-2008
-10
0
15
-5
5
10
(a)
55
60
65
Linear Fit
Fig. 4.10 Income inequality convergence: 1970–2014. Source Own calculations
based on Economic Freedom of the World Data, http://freetheworld.com
The graphical evidence of consumption convergence in Fig. 4.8 is definitely stronger. Throughout the observable time, the relationship between
the initial values of consumption and its subsequent growth was negative,
as suggested previously by Pretty (2013). Initially, this convergence was
slow, especially if we monitor the relationship at shorter periods of time.
It gained momentum after 1990 and accelerated immediately before the
Crisis. The after-Crisis period was the one in which consumption was
converging across countries at its fastest rate in decades.
After the Crisis, we also observe a large number of countries actually
reducing their consumption per capita. Most of them were rich which,
given the severity of the Crisis, explains why consumption in those countries would be hit harder than in poorer countries. There were also exceptions. For example, Syria entered into a long civil war, Yemen was under
a heavy presence of terrorist groups, and some countries were being torn
apart by lingering ethnic and religious conflicts. The one fact we can
4 Policy Convergence Vs. Welfare Convergence
89
definitely see is that the world is becoming more similar in terms of consumption opportunities after the Crisis.
Is the case the same with life expectancy? Fig. 4.9 offers a clue. Life
expectancy convergence has been documented in some empirical studies already (Becker et al. 2005; Moser et al. 2007). For the entire 44year period between 1970 and 2014, we can observe that, typically, life
expectancy grows faster in countries in which people are expected to live
shorter lives. That is the long-term picture we can observe in Fig. 4.9a.
In somewhat shorter intervals, the situation is similar, with the exception of the decade between 1990 and 2000. Life expectancy convergence
was strong in the 8 years before the Crisis, and it continued after. We can
clearly see that in Figs. 4.9c,d.
There is one more interesting fact about life expectancy convergence.
In each of the shorter-term sub-figures, we can observe a slower growth of
life expectancy at the right tails of the life expectancy distribution. That
means lives are expected to grow much more slowly once they reach a
certain threshold, especially within certain shorter time spans. We can see
the life expectancy growth flattening out when life expectancy grows closer
to 70 years, a fact observed also earlier by Peltzman (2009, p. 180).
The last element of welfare which we examine for convergence is income
inequality. The hypothesis of income inequality convergence gains some
support from the graphical evidence presented in Fig. 4.10. The figure
compares the initial Gini coefficient at a given point in time with its
change within a following period.
Similarly to other elements of welfare, there is evidence of income
inequality convergence since 1970. Despite the relative scarcity of data,
the relationship in Fig. 4.10a is clear. Typically, if a country starts with
high levels of income inequality in 1970, it will gradually bring them
down, whereas low levels of income inequality in 1970 turn out to be
good predictors of an increase in the Gini in the subsequent four decades
before the Crisis.
The evidence in Fig. 4.10b suggests that if the Gini coefficient rises
above 40–45, then we can expect the given country to lower it in the
subsequent period, with some notable exceptions. At the same time, below
that threshold, it seems countries raise their income inequality.
90
P. Stankov
The 1990 s also witnessed a dramatic increase in income inequality for
the region which underwent the most reforms in the period: Central and
Eastern Europe (CEE). CEE countries started with the lowest levels of
income inequality in the world before 1990, only to see them rise rapidly
by 2008. This is seen in Fig. 4.10b.
In addition, Fig. 4.10b shows that income inequality in the USA
behaves very similarly to inequality in Latin American countries. In 1970,
USA started with just about the same income inequality as Germany and
Italy, and slightly lower than the income inequality in France. Gradually,
however, the Gini in the USA increased by a relatively large amount for a
developed economy, while most other developed countries have managed
to decrease it. Meanwhile, the UK experienced even greater increases in
income inequality than the USA. It is noteworthy that, in 1970, the UK
started with a Gini not much higher than the one in CEE.
Looking at the graphical evidence, it seems that the world has become
a more similar place in terms of welfare. Unequal countries have become
relatively more equal, while equal countries have become relatively more
unequal. Countries in which people lived longer have increased their average lifespans by less than countries in which people lived shorter lives.
Countries in which people consumed more raised their consumption per
capita by less than poorer countries. Welfare convergence was happening
slowly only in terms of income per capita. The next section examines
exactly how fast the process was.
4.3
Is Welfare Convergence Significant?
After studying the graphical evidence, it appears that there is a process of
convergence in welfare across countries over time. However, the graphical evidence suggests that convergence is strong only for some welfare
measures (consumption per capita, life expectancy, and income inequality) and virtually nonexistent for others (income per capita). This section
shows whether the graphical evidence bears any statistical significance.
4 Policy Convergence Vs. Welfare Convergence
91
To put the welfare convergence hypothesis under a more rigorous empirical testing, I estimate the following OLS equation:
Wit = α + βWit + εit ,
(4.2)
where Wit is one of the four welfare proxies: the growth rate of GDP/c.,
the growth rate of consumption per capita, the change in life expectancy,
and the change in income inequality measured by the Gini coefficient
over a certain time period. As before, the eight periods studied are: 1970–
2014, 1970–1990, 1990–2008, 1970–1980, 1980–1990, 1990–2000,
2000–2008, and 2008–2014.The Wit variable is the value of the respective
welfare measure at the beginning of the respective period. As there is still
no Gini index data for 2014, the estimates for income inequality run till
2008 rather than 2014. The results for each of the welfare measures are
presented in Table 4.2. Those results can also be interpreted as the speed
of welfare convergence.
Table 4.2 is organized in a similar way to the one presenting the results
from testing for policy convergence. For each of the welfare measures, β
estimates are presented in each of the time periods. The changes in the
welfare measures are given in rows, while the time periods are given in
columns.
The table adds another argument to our expectation that countries
across the globe are not converging significantly in income over time, at
least not over long periods of time. This is evident from the first three
columns of the table, where the β-s are derived in 20- and 44-year periods. When we go one step further to estimate convergence over shorter
periods, we see that convergence is observed in two of the five periods only:
1980–1990, and 2008–2014. At the same time, there is significant evidence of income per capita divergence between 1970 and 1980. The two
opposing signs in two consecutive decades is perhaps what drives convergence in the entire 20-year period between 1970 and 1990 insignificant.
The next 20 years were also noted for a lack of income per capita convergence.
Despite the above evidence, the world was becoming more similar in
terms of consumption opportunities. For most of the time periods under
consideration, consumption per capita was growing faster in countries
92
Table 4.2
The speed of welfare convergence: 1970–2014
GDP/c. growth
−0.21
(0.15)
No. obs
133
0.012
Adj. R 2
Cons./c. growth −0.47***
(0.14)
No. obs
133
0.121
Adj. R 2
Life Exp
−0.35***
(0.04)
No. obs
152
Adj. R 2
0.410
Gini
−0.37**
(0.15)
No. obs
14
Adj. R 2
0.260
−0.21
0.02
0.46*
(0.23)
(0.21)
(0.24)
133
154
133
−0.000
−0.006
0.018
−0.21
−0.54***
0.14
(0.21)
(0.15)
(0.30)
133
154
133
0.003
0.094
−0.005
−0.18*** −0.14*** −0.10***
(0.03)
(0.04)
(0.02)
152
152
152
0.203
0.096
0.174
−0.10
−0.28*** −0.24*
(0.17)
(0.10)
(0.13)
12
41
13
−0.058
0.174
0.187
−0.61*
(0.34)
133
0.030
−0.43**
(0.21)
133
0.031
−0.09*
(0.05)
152
0.064
−0.06
(0.14)
19
−0.050
0.26
−0.32
−1.35***
(0.22)
(0.32)
(0.30)
154
154
154
−0.001
0.002
0.218
−0.38* −0.87*** −1.62***
(0.20)
(0.20)
(0.27)
154
154
154
0.010
0.141
0.264
−0.01
−0.11*** −0.11***
(0.03)
(0.02)
(0.02)
152
153
154
−0.006
0.285
0.396
−0.46***
−0.06
–
(0.11)
(0.08)
–
35
54
–
0.333
0.003
–
Notes The table reports β estimates from estimating the following OLS equation: Wit = α +βWit +εit , where
Wit is one of the following: GDP/c. growth rate, consumption/c. growth rate, the change in life expectancy or
the change in the Gini coefficient over the studied 8 periods, and Wit is the value of the respective dependent
variable at the beginning of the period. Robust standard errors are presented in parentheses. Data source:
PWT9.0, WDI data. Note that the Gini index estimates run till 2008 rather than 2014. Symbols: * p < 0.10, **
p < 0.05, *** p < 0.01
P. Stankov
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
1970–2014* 1970–1990 1990–2008 1970–1980 1980–1990 1990–2000 2000–2008 2008–2014
4 Policy Convergence Vs. Welfare Convergence
93
which started with low levels of consumption per capita. The process not
only seems significant, but it also accelerates over time since 1990. We
can see that the β estimates are gradually increasing from −0.38 for the
period between 1990 and 2000 to −1.62 for the after-Crisis period. Convergence in consumption per capita also existed for the entire 44-year
period between 1970 and 2014. The 20-year period, in which consumption per capita convergence was not evident, was between 1970 and 1990.
Table 4.2 presents additional evidence of life expectancy convergence.
This is perhaps the strongest type of welfare convergence we observe over
time. The statistical significance of the estimates is very strong over most
of the time intervals under study, and the observed signs are expectedly
negative. In only one period do we find no statistically significant convergence of life expectancy, although the sign is performing in line with
expectations. This is the period between 1990 and 2000.
Further, let us relate the annual changes in life expectancy to their initial
levels. We notice that life expectancy convergence is gradually becoming
faster over time. A typical country in which people live 10 years less than
people from another country will add approximately 3.5 years more to
its expected lifetime for the entire 44-year period until 2014. This means
that, each year, people in the first country gain about a month more than
lifetime added in the second country. After the Crisis, this rate is doubling.
Time will tell if this is only a contemporaneous after-Crisis phenomenon
or is a new trend in life expectancy convergence which economists and
policy makers need to start looking more closely into.
Income inequality convergence is another characteristic of the data
which has significant statistical support, though this was not always the
case. This is seen from the negative signs throughout the table. The income
inequality convergence is statistically significant for the entire 38-year
period between 1970 and 2008 as a whole. Within this longer period, however, it is significant only half of the time—between 1990 and 2008—and
insignificant between 1970 and 1990. In fact, the period between 1970
and 1980 saw a statistically significant income inequality convergence,
which was immediately offset in the following decade. This is the reason
we do not observe the expected convergence within the entire 20-year
period between 1970 and 1990. At the same time, the magnitude and
the statistical significance of the convergence process between 1990 and
94
P. Stankov
2000 were sufficient to render the income inequality convergence for the
entire 18-year period between 1990 and 2008 statistically significant. This
is despite the lack of significant results in the period between 2000 and
2008. Overall, the results are conclusive regarding income inequality convergence over longer periods of time but are very much open to discussion
over the shorter time spans under study.
This chapter has shown that both economic freedom policies and welfare converge. But are economic freedom policies the driver of welfare
convergence? The answer to this question follows in the next chapter.
References
Alesina, A. 1988. Credibility and policy convergence in a two-party system with
rational voters. The American Economic Review 78 (4): 796–805.
Altenburg, T., H. Schmitz, and A. Stamm. 2008. Breakthrough? China’s and
India’s transition from production to innovation. World Development 36 (2):
325–344.
Andrews, D.M. 1994. Capital mobility and monetary adjustment in Western
Europe, 1973–1991. Policy Sciences 27 (4): 425–445.
Barro, R. 1997. Getting it right: Markets and choices in a free society. MIT Press.
Barro, R.J., and X. Sala-i-Martin. 1992. Convergence. Journal of Political Economy
100 (2): 223–251.
Barro, R.J., and X. Sala-i-Martin. 1997. Technological diffusion, convergence,
and growth. Journal of Economic Growth 2 (1): 1–26.
Bearce, D.H. 2009. EMU: The last stand for the policy convergence hypothesis?
Journal of European Public Policy 16 (4): 582–600.
Becker, G.S., T.J. Philipson, and R.R. Soares. 2005. The quantity and quality of
life and the evolution of world inequality. American Economic Review 95 (1):
277–291.
Belloc, F., A. Nicita, and S.M. Sepe. 2014. Disentangling liberalization and privatization policies: Is there a political trade-off ? Journal of Comparative Economics
42 (4): 1033–1051.
Ben-David, D. 1998. Convergence clubs and subsistence economies. Journal of
Development Economics 55 (1): 155–171.
Bennett, C.J. 1991. What is policy convergence and what causes it? British Journal
of Political Science 21 (2): 215–233.
4 Policy Convergence Vs. Welfare Convergence
95
Bentzen, J. 2005. Testing for catching-up periods in time-series convergence.
Economics Letters 88 (3): 323–328.
Bertola, G. 2016. Labor policies and capital mobility in theory and in EMU.
European Economic Review 87: 62–77.
Bicaba, Z. T., and F. Coricelli. 2015. Learning to open up: Capital account
liberalizations in the post-Bretton Woods era. CEPR Discussion Papers 10389,
C.E.P.R. Discussion Papers.
Boockmann, B., and A. Dreher. 2003. The contribution of the IMF and the
World Bank to economic freedom. European Journal of Political Economy 19
(3): 633–649. Economic Freedom.
Brune, N., G. Garrett, and B. Kogut. 2004. The International Monetary Fund
and the global spread of privatization. IMF Staff Papers 51 (2): 195–219.
Carlino, G.A., and L.O. Mills. 1993. Are U.S. regional incomes converging?: A
time series analysis. Journal of Monetary Economics 32 (2): 335–346.
Cassette, A., J. Creel, E. Farvaque, and S. Paty. 2013. Governments under influence: Country interactions in discretionary fiscal policy. Economic Modelling
30: 79–89.
Chang, K., Y. Kim, M. Tomljanovich, and Y.-H. Ying. 2013. Do political parties foster business cycles? An examination of developed economies. Journal
of Comparative Economics 41 (1): 212–226. Symposium in Honor of Thrainn
Eggertson.
Degiannakis, S., D. Duffy, G. Filis, and A. Livada. 2016. Business cycle synchronisation in EMU: Can fiscal policy bring member-countries closer? Economic
Modelling 52, Part B: 551–563.
Gwartney, J., J. Hall, and R. Lawson. 2016. 2016 Economic freedom dataset.
Fraser Institute.
Hall, J.C. 2016. Institutional convergence: Exit or voice? Journal of Economics
and Finance 40 (4): 829–840.
Haug, A.A., J.G. MacKinnon, and L. Michelis. 2000. European monetary union:
A cointegration analysis. Journal of International Money and Finance 19 (3):
419–432.
Heckelman, J.C. 2015. Economic freedom convergence clubs. In Economic
behavior, economic freedom, and entrepreneurship, ed. R.J. Cebula, J.C. Hall,
F.G. Mixon, and J.E. Payne, 102–114. Edward Elgar Publishing.
Holzinger, K., and C. Knill. 2005. Causes and conditions of cross-national policy
convergence. Journal of European Public Policy 12 (5): 775–796.
Jackman, M., and W. Moore. 2008. Economic policy co-movement in Latin
America and the Caribbean. Journal of Policy Modeling 30 (6): 1107–1114.
96
P. Stankov
Kasman, A., S. Kirbas-Kasman, and E. Turgutlu. 2008. Monetary policy convergence of potential EMU accession countries: A cointegration analysis with
shifting regimes. Economic Modelling 25 (2): 340–350.
King, A., and C. Ramlogan-Dobson. 2015. Is Africa actually developing? World
Development 66: 598–613.
Knill, C. 2005. Introduction: Cross-national policy convergence: Concepts,
approaches and explanatory factors. Journal of European Public Policy 12 (5):
764–774.
Köppl-Turyna, M. 2014. Campaign finance regulations and policy convergence:
The role of interest groups and valence. European Journal of Political Economy
33: 1–19.
Kose, A., C. Otrok, and E. Prasad. 2012. Global business cycles: Convergence
or decoupling. International Economic Review 53 (2): 511–538.
Krasa, S., and M.K. Polborn. 2012. Political competition between differentiated
candidates. Games and Economic Behavior 76 (1): 249–271.
Kutan, A.M., and T.M. Yigit. 2004. Nominal and real stochastic convergence of
transition economies. Journal of Comparative Economics 32 (1): 23–36.
Kutan, A.M., and T.M. Yigit. 2005. Real and nominal stochastic convergence:
Are the new EU members ready to join the Euro zone? Journal of Comparative
Economics 33 (2): 387–400. Returns to Schooling in Translation: The Chinese,
European, and Russian Experiences.
Laussel, D., and R. Riezman. 2005. The sources of protectionist drift in representative democracies. European Economic Review 49 (7): 1855–1876.
Meseguer, C. 2006. Learning and economic policy choices. European Journal of
Political Economy 22 (1): 156–178.
Monfort, M., J.C. Cuestas, and J. Ordóñez. 2013. Real convergence in Europe:
A cluster analysis. Economic Modelling 33: 689–694.
Moser, K., V.M. Shkolnikov, and D.A. Leon. 2007. World mortality
1950–2000: Divergence replaces convergence from the late 1980s. In HIV,
Resurgent infections and population change in Africa, ed. M. Caraël, and
J.R. Glynn, 11–25. Dordrecht: Springer Netherlands.
Mukand, S., and D. Rodrik. 2005. In search of the holy grail: Policy convergence,
experimentation, and economic performance. The American Economic Review
95 (1): 374–383.
Nieswiadomy, M., and M. Strazicich. 2004. Are political freedoms converging?
Economic Inquiry 42 (2): 323–340.
Peltzman, S. 2009. Mortality inequality. Journal of Economic Perspectives 23 (4):
175–90.
4 Policy Convergence Vs. Welfare Convergence
97
Phillips, P.C.B., and D. Sul. 2007. Transition modeling and econometric convergence tests. Econometrica 75 (6): 1771–1855.
Phillips, P.C.B., and D. Sul. 2009. Economic transition and growth. Journal of
Applied Econometrics 24 (7): 1153–1185.
Pretty, J. 2013. The consumption of a finite planet: Well-being, convergence,
divergence and the nascent green economy. Environmental and Resource.
Economics 55 (4): 475–499.
Schuster, P., C. Schmitt, and S. Traub. 2013. The retreat of the state
from entrepreneurial activities: A convergence analysis for OECD countries,
1980–2007. European Journal of Political Economy 32: 95–112.
Simmons, B.A., and Z. Elkins. 2004. The globalization of liberalization: Policy diffusion in the international political economy. American Political Science
Review 98 (1): 171–189.
Slemrod, J. 2004. Are corporate tax rates, or countries, converging? Journal of
Public Economics 88 (6): 1169–1186.
Stahl, D.O., and A.H. Turunen-Red. 1995. Tariff games: Cooperation with random variation in political regimes. European Journal of Political Economy 11
(2): 215–238.
Steiner, A. 2010. Contagious policies: An analysis of spatial interactions among
countries’ capital account policies. Working Papers 80, Institute of Empirical
Economic Research.
Strazicich, M.C., J. Lee, and E. Day. 2004. Are incomes converging among
OECD countries? Time series evidence with two structural breaks. Journal of
Macroeconomics 26 (1): 131–145.
Zind, R.G. 1991. Income convergence and divergence within and between LDC
groups. World Development 19 (6): 719–727.
5
Welfare and Reforms: Evidence
In this chapter, I initially study bivariate correlations between economic
freedom reforms and welfare. These correlations are first explored graphically and then by using more technical panel data regression methods.
Welfare is understood as a collection of four variables that change over
time: GDP per capita, to gain an idea of whether people within a certain economy are growing richer at the time of a certain reform or soon
afterward; consumption per capita to see if the increases or decreases in
income are accompanied by a change in consumption while the reform
was taking place; life expectancy; and income inequality.
For each of the reforms, graphical correlations are studied for the entire
44-year period between 1970 and 2014. The longer period is then disaggregated into shorter periods of either 18–20 years or 6–10 years. Thus,
we are able to monitor how reforms are associated with changes in welfare
across time. Because of data limitations, we are able to see the correlations
between reforms and income inequality only until 2008. We are able to
monitor the dynamics of the rest of the welfare variables through 2014,
the year on which we also have the latest data on reforms.
© The Author(s) 2017
P. Stankov, Economic Freedom and Welfare Before and After the Crisis,
DOI 10.1007/978-3-319-62497-6_5
99
100
P. Stankov
5.1
Welfare and Reforms: Graphical
Evidence
5.1.1 Government Intervention and Welfare
Figure 5.1a explores the correlations between the changes in the Size of
Government (SoG) index and corresponding average GDP/c. growth. We
can see that for the 44 years between 1970 and 2014 the correlation is
mildly negative, which signals that, over long periods of time, less government intervention may be also associated with slower growth of income
per capita. The statistical significance of this effect, however, is refuted by
the more rigorous regression evidence, which will be presented later in
this chapter.
(b)
-5
-4
-3
-2
-1
0
1
2
3
4
5
8
6
BWA
MYS OMN
IDN
TWN
SGP
LKA
IND
TUN
EGY
TZA
MLT
DOMZWE
PAK
PRY MUS
IRN
NGA
COG DZA CYP
KEN
ISR CHL
ECU
TUR
PHL
BDI
BRAMAR
COL
HND
MWI
GTM
CMR
GAB
FJIGHA
RWAAUS
VEN
ISL
ZAFCRIMEX SYR
LUX
IRL
BEN
TTO
NOR
SEN
CIV
ARG PER BOL
CAN
TGO
USA
PRT
ESP FIN
HTI
SLVNERURY
NZL
FRA AUT
ZMBGBR
GRC
BELNLD
DEU CODNIC
ITA
JAM
SWE
HUN
MDG
CAF
4
KOR
MLI
JPN
2
5
0
JPN
MLI
SGP
IDN TWN
EGY
SLV
MLTMYS
THAHKG
IRL
ARG
HUN
BRA
MUS
IRN
PRY
TUN
IND
DOM
CYP
NOR
LUX
PRT
BOL
AUTPER ECU
LKA
OMN
MAR
COG
ESP
DEU
CHL
TTO
TUR PAK
COL
PHL
ITA
BEL
GBR
NLD
URY
GAB
GRC DZA FINISL
CRI
FRA
CAN
SWE
AUS
FJIISR
USA
GTM
MEX NZL
HND
CMR
VEN
RWA
ZAF
KEN
HTI
GHA
TZA JAM
CIVSYR SEN ZMB
NGA
BDI
BEN
BRB
TGO
MWI
MDG
ZWE
NIC
NER
COD
CAF
SoG Reforms and Consumption/c. Growth, 1970-2014
BRB
-5
-4
-3
Size of Government Reforms, 1970-2014
Country Code
-2
-1
0
1
2
3
4
5
Size of Government Reforms, 1970-2014
Linear Fit
Country Code
(c)
Linear Fit
(d)
SoG Reforms and Income Inequality, 1970-2008
SLV
10
JPN
GRC
VEN
COG
ECU
GAB
BEN
IND
HTI
MEX
BRA
THA
DOM
SGP
IDN PRTCOD
PAK BDI
TGO
COL
CRI
HKG
LUX
GHA
NGA
MLTMUS
AUS ZMB
ITA AUT
ESPFIN
ISR
SYR
FRA
LKA
MYS
DEU
FJI
BRBBEL
IRL BWA
NZL
ARG
CMR CAF
CAN
GBRKEN
URY
NORUSASWE
CYP ISL NLD
PRY CIV
PHL
JAM
HUN
TTO
ZAF
ZWE
-5
-4
-3
-2
-1
0
1
2
Size of Government Reforms, 1970-2014
Country Code
Linear Fit
3
4
10
IDN
5
GBR
USA
5
NIC
MDG
EGY
CHL
TZA
KOR
0
20
MWI
KOR
COL
PAK
IRN
BRA
JPN
-5
SEN
OMN
NER
IRN
TUR TUN
BOL
MAR
PER
HND
RWA
GTM
DZA
ITA
DEU
-10
MLI
Change in Gini, 1970-2008
30
15
SoG Reforms and Life Expectancy, 1970-2014
0
Change in Life Expectancy (no. yrs.), 1970-2014
HKG
THA
0
10
BWA
KOR
-5
Compound GDP/c. Growth, 1970-2014
Size of Government Reforms and Growth, 1970-2014
Consumption per Capita Growth, 1970-2014
(a)
FRA
-5
-4
-3
-2
-1
0
1
2
3
4
5
Size of Government Reforms, 1970-2008
Country Code
Linear Fit
Fig. 5.1 Government intervention and welfare: 1970–2014. Source a and b Own
calculations based on PWT9.0 and EFW data. c and d Own calculations based on
WDI and EFW data
5 Welfare and Reforms: Evidence
101
Figure 5.1b reveals that the association between government intervention and consumption per capita is also zero. It seems that despite efforts
to decrease government over time, those efforts did not produce more
entrepreneurial activity to spur both consumption and income per capita.
If we slice time into shorter intervals, we will see that the correlations also
remained unchanged before or after the Great Recession. This implies that
perhaps government intervention is not a crucial factor affecting income
and consumption per capita growth.
Life expectancy also seems unaffected by the Size of Government.
Over the longest period on which we have data on both changes in life
expectancy and changes in government intervention, the association is
virtually flat. This is seen in Fig. 5.1c.
The one area in which we observe a more definite relationship between
government intervention and welfare is income inequality. This relationship is presented in Fig. 5.1d. For the entire period between 1970 and
2008, smaller governments were correlated with higher increases in income
inequality. The link was obtained by using too few observations, so it is
worth checking if it persists when more observations are added over time
in the regression results. Indeed, for some estimations, it does.
5.1.2 Property Rights and Welfare
Unlike government intervention, which was basically unrelated to most
welfare components through time, improving property rights (PRs) is different. Most improvements in property rights are correlated with improvements in living standards, consumption, life expectancy, and income
inequality for the average citizen. This is shown in Fig. 5.2.
Figure 5.2a displays the link between property rights and GDP/c.
growth for the 44 years between 1970 and 2014. Chile is a leader in
property rights improvement for the entire period, but it only enjoys an
average growth rate of income per capita. At the same time, countries
and territories like South Korea, Singapore, Indonesia, Taiwan, and Hong
Kong barely improved PRs. However, their income per capita has grown
enormously for the entire period, virtually almost doubling their living
standards every 12 years since 1970. This is indicative of the fact that
102
P. Stankov
(b)
Compound GDP/c. Growth, 1970-2014
-5
0
5
10
Property Rights Reforms and Growth, 1970-2014
KOR
SGP
IDN
HKG MYS
IRL ARG
BRA
IRN
TUN
IND
NOR
LUX
PRT
ECU
MAR
ESP
DEU
TUR ISL
COLPAK
PHL
ITA
JGBR
PN
BEL
FIN
NLD
GRC DZA
FRA
DNK
CAN
SWE
AUS
USA ISR
CHE NZL
MEX
VEN ZAF
KEN
NGA
TWN
THA
-5
-4
-3
-2
-1
0
1
2
3
EGY
PER
CHL
4
5
Consumption per Capita Growth, 1970-2014
0
2
4
6
8
(a)
6
PR Reforms and Consumption/c. Growth, 1970-2014
MYS
IDN
SGP
HKG
KOR
IND TUN
IRN PAK
NGA
KEN
ISR PHL
DZA
ECU MAR
TUR
BRA
COL
MEX
VEN ZAF
ISL
AUS
LUX
IRL
ARGNOR
CAN
USA ESP
JPN
PRT
FIN
GBR
NZL
FRA
GRC
NLD
BEL
DEU
ITADNK
SWE
CHE
TWN
THA
-5
-4
-3
Property Rights Reforms, 1970-2014
Country Code
-1
0
1
2
3
CHL
PER
4
5
6
7
Property Rights Reforms, 1970-2014
Linear Fit
Country Code
(c)
Linear Fit
(d)
PR Reforms and Income Inequality, 1970-2008
IRN
DZA
TUR
TUN
MAR
KORIND
ECU
VEN
PER
EGY
Change in Gini, 1970-2008
0
5
10
-5
15
PR Reforms and Life Expectancy, 1970-2014
CHL
MEX
BRA THA
SGP
PRTIDN
PAK
HKG COL
LUX
JPN
AUS NGA
ITA
ESP
FIN
ISR
FRA
GRC
DEU
IRL
NZLMYS
CHE
BEL
ARG
KEN
CAN
GBR
USA NLD
ISL
PHL NOR
SWE
DNK
USA
-4
-3
-2
-1
0
1
2
3
4
Property Rights Reforms, 1970-2014
Country Code
Linear Fit
5
6
7
IDN
KOR
COL
PAK
IRN
ITA
ZAF
-5
GBR
BRAJPN
DEU
-10
Change in Life Expectancy (no. yrs.), 1970-2014
0
10
20
30
-2
EGY
FRA
-5
-4
-3
-2
-1
0
1
2
3
4
Property Rights Reforms, 1970-2008
Country Code
5
6
7
Linear Fit
Fig. 5.2 Property rights and welfare: 1970–2014. Source a and b Own calculations
based on PWT9.0 and EFW data. c and d Own calculations based on WDI and EFW
data
improving PRs is not a definite recipe for improving living standards,
especially in developing economies. At the same time, we can also presume that economies which start with better protection of PRs in the first
place may grow faster than others, but may also not experience the need
to reform much. That could be one explanation for high growth rates in
countries that did not significantly improve PRs.
If we slice the data into shorter periods, we see that the 1990s and
early 2000s are a time of unquestionably positive association between
improvement in property rights and improvement in living standards.
That is why so many studies published in this period inevitably conclude
that improving PRs is a road to richness. After the Great Recession, the
PRs-growth association is still positive. This makes it a relatively robust
relationship one can establish on the link between reforms and income
per capita growth.
5 Welfare and Reforms: Evidence
103
Income per capita growth, however, is not the only welfare gauge we
are interested in. Income per capita is only as good as its ability to raise
consumption. Figure 5.2b addresses this ability. By comparing the two
figures, we can see that they are actually very similar. Improvements in
PRs were associated with an increase in consumption for the entire time
period between 1970 and 2014. However, the regression evidence we will
see later is far less convincing on the link between PRs and consumption.
Life expectancy also changed across the period, and for the better. The
graphical evidence for the entire 44 years suggests a possible factor: property rights improvement. The link between PRs and life expectancy is
presented in Fig. 5.2c. The link is very clear and strong. People began to
live longer in the countries which improved PRs more.
Property rights also had a changing relationship with income inequality
over the years. If we look at the entire 38-year period in the run-up to
the Crisis, we observe no correlation between improving PRs and making
societies more equal. This is seen from Fig. 5.2d. This may be due to
the fact that there are too few observations to meaningfully conclude the
correlation was there, or because the underlying relationship is inherently
nonexistent. However, if we group the data into clusters of countries, we
see that within those clusters, countries improving their PRs systems faster
also shrink their income inequality, contrary to the effects observed in a
historical context by Dow and Reed (2013).
5.1.3 Monetary Reforms and Welfare
Monetary policy reforms affect welfare in a number of ways. As seen from
Fig. 5.3a, they positively affect income per capita. For the entire period
between 1970 and 2014, countries enjoying stable inflation also grew faster
than countries which did not achieve long-term price stability. If we focus
on smaller time intervals, the link persists, which means price stability and
lower inflation volatility have the potential to positively affect growth rates
of income per capita. This is not surprising, as low and predictable inflation
increases the planning horizon of businesses. In turn, this produces higher
investment in the local economy.
104
P. Stankov
(b)
0
VEN
UGA
-10
-8
-6
-4
-2
0
2
4
6
8
8
CHN
BWA
MYS OMN
6
IDN
TWN
SGP
HKG
KOR
THA
LKA
IND
TUN
EGY
MMR
TZA
MLT ZWE DOM
BHR
MUS
KWT
PRY
IRNPAK
NGA
CYP
COGPHL
KEN
JOR
ISR
DZA
MAR
ECU
TUR
NPL
BLZ BDI
BRACHL
PAN
MLI COL
TCD
HND
MWI
CMR GTMRWA
GAB GHA
FJI
CRI ZAF
MEX
SYR
ISL
AUS
LUX
IRL
BEN
TTO
BOL
NOR
SEN
CIV
ARG
PER
BHS
CAN
TGO
PRT
ESPURY
HTI USA
BGD
FIN JPN
GBR
SLV
NZL
NER GRC FRA
ZMB
NLD AUT
NICBEL
DEU
CODDNK
ITA
JAM
SWE
CHE
HUN
SLE
MDG
CAF
4
5
KOR
SGP
IDN
CHN
TWN
MLT EGY SLV
MYS
MMR
HKG
THA
IRL
ARG
HUN
BRA
MUS
IRN PAN
PRY
TUN
IND
DOM
NOR JOR
LUX
PRT
ECU
BOL
LKA
AUT CYP
OMN
MAR
COG
ESP
PER
DEU
CHL
TTO
TUR
COL
PHL
NPL
ITA
BHR
BEL
FIN JPN
GBR
NLD
URY
PAK
ISL
GAB
MLI
GRC
ISR
CRI
FRA
DZA
DNK
CAN
SWE
AUS
FJI BLZ
USA
NZL
CHE
KWT
MEX
HND
CMR GTM
BGD
BHS
RWA
ZAF
KEN
HTIJAM
SYR
SEN
GHA TZA
CIV
ZMB NGA
SLE
BEN
BRB BDI
TCD
TGO
MWI
MDG
NIC NER ZWE
COD
CAF
Monetary Reforms and Consumption/c. Growth, 1970-2014
VEN
2
Consumption per Capita Growth, 1970-2014
10
BWA
-5
Compound GDP/c. Growth, 1970-2014
Monetary Reforms and Growth, 1970-2014
BRB
10
-7
-6
-5
-4
-3
-2
Monetary Reforms, 1970-2014
Country Code
-1
0
1
2
3
4
5
6
7
8
9
Monetary Reforms, 1970-2014
Linear Fit
Country Code
(c)
Linear Fit
(d)
Monetary Reforms and Income Inequality, 1970-2008
UGA
-7
-6
-5
-4
-3
-2
-1
0
1
2
3
4
5
Monetary Reforms, 1970-2014
Country Code
Linear Fit
6
7
8
9
10
5
IDN
0
KOR
COL
PAK
IRN
BRA
JPN
-5
VEN
GBR
USA
ITA
DEU
-10
10
20
NPL
SEN
OMN
MLI
NER
IRN
DZA
BGD
TUN
TUR
BOL
MWI MAR
NIC
PER
HND
MDG
IND
RWAKOR
EGY GTM
CHL
ECU TZA SLV
GAB BEN
CHN
SLE HTI
MEX MMR
BRA
THA
DOM
COD
SGP
IDN
JOR
PAK BHR
TGO
COLPRT
CRI
BDI
HKG
LUX
PANGHA
NGA
MLT AUT
AUS
ITAZMB
MUS
ESP
FIN JPN
ISR
SYR TCD
FRA
LKA
GRC
MYS
DEU
FJIARG
BRB
IRL
NZL
BWA
CHE CMR
BEL
BHS
KENGBR
CAN
COG
CAF
KWT USA
URY
CIV
NLD ISL
NOR
CYP
PRY
PHL
JAM
SWE
DNK
HUN
TTO
BLZ
ZAF
ZWE
Change in Gini, 1970-2008
30
15
Monetary Reforms and Life Expectancy, 1970-2014
0
Change in Life Expectancy (no. yrs.), 1970-2014
UGA
0
(a)
FRA
-4
-3
-2
-1
0
1
2
3
4
5
6
7
8
Monetary Reforms, 1970-2008
Country Code
Linear Fit
Fig. 5.3 Monetary reforms and welfare: 1970–2014. Source a and b Own calculations based on PWT9.0 and EFW data. c and d Own calculations based on WDI and
EFW data
The graphical evidence is weaker for the impact of monetary reforms
on the growth of consumption per capita, as shown in Fig. 5.3b. Nevertheless, the expected positive association between taming inflation and
consumption growth persists.
Unlike income and consumption per capita, life expectancy seems relatively unaffected by monetary policy reforms, as shown in Fig. 5.3c. The
relationship is flat for most of the period, even if we split time into shorter
intervals. The regression estimates, however, deliver a stronger message:
Countries which improve the stability of money also gain a longer expected
lifetime.
It is somewhat hard to believe that monetary policy stabilization had
a direct impact on life expectancy. However, we may also note that stable money has a direct impact on income and consumption per capita
growth. At the same time, an increase in income and consumption has the
ability to make people live longer (Becker et al. 2005; Mackenbach 2013;
5 Welfare and Reforms: Evidence
105
Oeppen 2006). Therefore, indirectly, monetary policy stabilization could
affect life expectancy.
Figure 5.3d demonstrates the alleged impact monetary reforms have on
income inequality. The graphical evidence suggests that countries introducing more stable monetary policies are also prone to higher income
inequality. This is irrespective of how one slices time.
There is a rather intuitive explanation for this empirical observation,
and it complements the credible story for the link between property rights
and income inequality. On the one hand, monetary stability is one of
the preconditions for businesses to grow and for entrepreneurial talent to
flourish. However, if they do, then some people’s wealth is going to grow
faster than the wealth of others, especially immediately after the stabilizing
reforms. Put simply, when monetary policy is stable, the “pie” grows. Due
to their different entrepreneurial talents, some people will create more of
this growth than others. Then, it is natural to expect that they would be
appropriately rewarded with bigger pieces of the “pie” over time. This is
why in most cases one would expect to see rising income inequality in
times of rising monetary policy predictability. In turn, this would result
in a positive relationship between the index of monetary reforms and the
change in the Gini coefficient.
However, there is a downside to this logic. In the above intuition, we
assumed the economy grew, and some people gain more rewards from this
growth. But what if the economy does not grow and suffers prolonged
and deep recessions accompanied by outbursts of inflations? In the above
example, we also tend to forget that rising inflation and price volatility
can be powerful redistribution mechanisms, resulting in swelling income
inequality (Albanesi 2007; Ghossoub and Reed 2017). During periods of
high inflation, borrowers gain and savers lose wealth. This is exactly what
happened in Central and Eastern Europe for most of the 1990s.
Price liberalization in the beginning of the 1990s created galloping
inflation in some countries, such as Bulgaria, Romania, Russia, and Slovakia. At the same time, those inflationary periods redistributed wealth
between savers (typically, the majority of working class people) and borrowers (typically, the new private entrepreneurs). This increased income
inequality in those countries more than in others. If we partition time into
shorter intervals, more data would become observable and it would feed
well into the intuition above.
106
P. Stankov
5.1.4 Trade Reforms and Welfare
Of all the large-scale market-oriented reforms, trade reforms perhaps exert
the most robust impact on welfare, at least when it comes to the graphical evidence in Fig. 5.4. Figure 5.4a produces a surprisingly uneventful
relationship. It seems that the long-term correlation between income per
capita growth and trade liberalization is mildly positive but not as large as
the literature has so far suggested. This is so perhaps because the literature
uses shorter time intervals to estimate the relationship. At those shorter
time intervals, trade liberalization correlates positively with income per
capita, as the regression estimates suggest.
Consumption per capita growth has also been positively affected by,
or at least correlated with, trade reforms. This can be easily seen from
Fig. 5.4b. The positive correlation means that countries liberalizing trade
(a)
(b)
KOR
SGP
IDNSLV
TWN MLT
MYS
MMR
HKG
THA
IRL
BRA
IRN LUX
PRY
TUN MUS
PAN
CYP
NOR
PRT
ECU
BOLDOM
AUT
LKA
MAR
ESP
PER
DEU COG
CHL
TUR
COL
PHL
NPL
ITA JPN
BEL
FIN
NLD
URY
PAK
ISLGBR ISR
GAB
GRC
CRI
FRA
DNK
CAN
SWE
AUS
FJI
USA
GTM
CHENZL
MEX HND
ZAFSYR BHS
KEN
SEN NGA
GHA
TZA
MWI
MDG
NIC
COD
VEN
-10
-8
-6
Consumption per Capita Growth, 1970-2014
8
0
2
4
6
Compound GDP/c. Growth, 1970-2014
-5
0
5
10
Free Trade Reforms and Growth, 1970-2014
-4
-2
0
2
4
Free Trade Reforms, 1970-2014
Country Code
6
8
10
MYS
IDN
TWN
SGP
HKG
KOR
THA
LKA
TUN
MMR
TZA
MLT
DOM
PRY
IRN
NGA PAK MUS
CYP
COG
KEN
MAR
ECU ISR PHLCHL
NPLTUR
BRA
PAN
COLHND
MWI
GTM
GAB
FJI
CRI
GHA
AUSMEX
ZAF SYR
LUX
IRL ISL
BOL
NOR
SEN PER
BHS
CAN
USA
JPN
PRT ESP GBR
URY
FIN
SLV
NZL
FRA
AUT
GRC
NLD
BEL
NIC
DEU ITA
COD
SWE
DNK
CHE
MDG
VEN
-6
-5
-4
-3
Linear Fit
-2
-1
0
1
2
3
4
Trade Reforms, 1970-2014
Country Code
(c)
5
6
7
8
9
Linear Fit
(d)
Trade Reforms and Income Inequality, 1970-2008
15
Trade Reforms and Life Expectancy, 1970-2014
NPL
Change in Gini, 1970-2008
0
5
10
-5
SEN
IRN
TUN
BOL
NICMAR HNDPER
KORMDG
GTM
TZA ECU
SLV
GAB
TUR
MWI
CHL
MEX
BRA
THA
MMR DOM
COD
SGP
IDN
PAK
CRI PRT COL
LUX
PAN HKG
GHA
AUT ESP
MLT NGA
AUS
ITA JPN
MUS
FIN
ISR LKA
SYR
FRA
GRC
MYS
DEU
FJI
IRL
NZL
BEL
BHS
KEN
CAN COG CHE
GBR
USAURY
ISL
NLD
NOR
CYP
PRY
PHL
SWE
DNK
VEN
-5
-4
-3
-2
-1
0
1
2
3
4
Trade Reforms, 1970-2014
Country Code
5
Linear Fit
6
7
8
9
IDN
KOR
COL
IRN
JPN
PAK
BRA
ITA
ZAF
-6
GBR
USA
DEU
-10
Change in Life Expectancy (no. yrs.), 1970-2014
0
10
20
30
Trade Reforms and Consumption/c. Growth, 1970-2014
FRA
-6
-5
-4
-3
-2
-1
0
1
2
3
4
Trade Reforms, 1970-2008
Country Code
5
6
7
8
9
Linear Fit
Fig. 5.4 Trade reforms and welfare: 1970–2014. Source a and b Own calculations
based on PWT9.0 and EFW data. c and d Own calculations based on WDI and EFW
data
5 Welfare and Reforms: Evidence
107
faster also enjoyed faster growth of their consumption opportunities. This
evidence is firmly against more recent calls for abandoning free trade talks
between USA and Europe, which are gradually gaining speed on both sides
of the Atlantic Ocean, especially after the 2016 US presidential elections.
Trade reforms apparently also correlate very well with life expectancy.
Those correlations are explored in Fig. 5.4. Overall, for the long period
between 1970 and 2014, the majority of countries added between 10 and
20 years to the life expectancy of their citizens. Countries which liberalized
more added more to life spans. For example, Luxembourg, Belgium, and
Germany have all added about 10 years and at the same time limited overall
trade liberalization. Within the same 44-year period, numerous countries
added far more years to life spans, but at the same time liberalized trade
more. The bottom line for the entire set of countries between 1970 and
2014 is clear: Deeper trade liberalization is associated with longer lives.
The link is hardly direct. As with monetary stability, it probably works
through trade liberalization effects on income per capita and consumption
per capita growth. It is only natural that richer nations can afford better
health care and protection from infectious diseases (Neumayer 2004),
healthier food, and safer water.
The relationship between trade liberalization and income inequality is
shown in Fig. 5.4d. The figure is very clear about this correlation: Leaders
in trade liberalization are also more likely to experience a swift increase
in income inequality. If we zoom in shorter time periods, the evidence
persists, which means the relationship is robust, at least in the bivariate
case. The regression results depict a more nuanced relationship.
There is a relatively simple intuition for growing income inequality as a
result of trade liberalization. Suppose an autarchy begins trading with the
rest of the world. Local businesses find new markets abroad, and foreign
businesses find new customers in the local economy. Trade theory predicts
that both labor and capital owners will be better-off after trade opens.
However, their income will grow differently. Because of the differences in
the growth rates of income across factor owners, after some time capital
owners will become considerably richer than labor owners. Therefore, even
if an economy starts with zero income inequality, trade liberalization will
invariably create winners and losers.
108
P. Stankov
A large part of this growing income inequality is healthy. If capital owners were not rewarded for venturing into production and trade activities,
they would not have incentives to do so. As a result, the closed economy
would stagnate relative to the rest of the world. It is very easy to point at
Cuba and North Korea to see dramatic illustrations of this logic.
However, some part of inequality growth is also unproductive. If the
income of producers and traders grows much faster than labor income,
then they will build up sufficient resources for a regulatory and a political
capture. Once that happens, the political and regulatory process will be
slanted in favor of individuals and businesses with vast economic resources.
Naturally, they would be driving the political agenda of the local economy in their own favor, as the insightful story by Açemoglu et al. (2005)
suggests. This may mean stifling market competition or influencing state
authorities at both legislative and judicial levels to turn a blind eye to corrupt practices. As a result, trade liberalization may lead to rapidly growing
income inequality in countries with sub-prime democracies, such as the
economies in the former Soviet Union, Central and Eastern Europe and
Latin America, an effect observed by Carter (2007), and confirmed even
for developed economies by Krieger and Meierrieks (2016).
5.1.5 Deregulation and Welfare
Figure 5.5a represents the association between overall deregulation reforms
and income per capita growth since 1970.The dynamics of the relationship
are different over time, but the message is clear: Deregulating the product,
labor, and credit markets pays off. If the correlations could be interpreted
in a causal sense, then Fig. 5.5a convincingly supports the hypothesis that
deregulation leads to a significant increase in living standards. That, of
course, is the story of an average regulatory reformer. Within those countries which reformed, there is significant variation in growth rates. In
fact, there were a number of reforming countries whose growth rate was
mediocre relative to some of the countries which reversed deregulation.
All this raises an important caution when interpreting the graphical
evidence: Deregulation certainly helps create some of the conditions for
faster growth. However, some countries apparently lack the complementary policy setting to allow deregulation to fulfill its growth-enhancing
109
5 Welfare and Reforms: Evidence
(b)
Compound GDP/c. Growth, 1970-2014
-5
0
5
10
Overall Regulatory Reforms and Growth, 1970-2014
SGP
IDN
CHN
TWN
MYS
HKG MMR
THA
IRL
ARG
TUN
IND
NOR
LUX
AUTJORPRT
ESP
DEU
ITA
JPNGBR
BEL
FIN
NLD
PAK
GRC
FRA
DNK
CAN
SWE
USA
GTM
NZL AUS ISR
CHE
ROU
BRA
VEN
KENZAF
SYR
TZA
MDG
COD
-5
-4
-3
-2
-1
0
1
2
3
Overall Regulatory Reforms, 1970-2014
Country Code
4
5
6
Deregulation and Consumption/c. Growth, 1970-2014
CHN
MYS
IDN
HKG
THA
TUN
IND
PAK
KEN
TWN
SGP
MMR TZA
JOR
ISR
GTM
LUX ZAF
IRL AUSSYR
ARGNOR
CAN
JPN
PRT
ESPUSA
FIN
GBR
FRA AUTNZL
GRC
NLD
BEL COD
DEU ITA
SWE
CHE DNK
BRA
VEN
ROU
MDG
-2
-1
0
Linear Fit
1
2
3
Deregulation, 1970-2014
Country Code
(c)
4
5
6
Linear Fit
(d)
Deregulation and Income Inequality, 1970-2008
Change in Gini, 1970-2008
0
5
10
-5
15
Deregulation and Life Expectancy, 1970-2014
TUN
IND
VEN
GTM
MDG
TZA
CHN
BRA
THA
CODMMR
JOR PRTIDN SGP
PAK
HKG
JPN FINLUX ESP AUT
AUS
ITA
ISR
SYR
FRA MYSNZL
GRC
IRL
BEL
ARGCHE DEU
CANGBRKEN
NOR USA
SWENLD
DNK
ROU
GBR
-1
0
1
2
3
Deregulation, 1970-2014
Country Code
4
Linear Fit
5
6
IDN
JPN
ITA
ZAF
-2
USA
PAK
BRA
DEU
-10
Change in Life Expectancy (no. yrs.), 1970-2014
0
10
20
30
Consumption per Capita Growth, 1970-2014
8
0
2
4
6
(a)
FRA
-2
-1
0
1
2
Deregulation, 1970-2008
Country Code
3
4
5
Linear Fit
Fig. 5.5 Deregulation and welfare: 1970–2014. Source a and b Own calculations
based on PWT9.0 and EFW data. c and d Own calculations based on WDI and EFW
data
potential. There was an increase in the slope of the relationship between
deregulation and growth after the Great Recession. Effectively, this means
deregulation was rewarded more after the Crisis than before it, at least
when it comes to income per capita growth.
That was the case with consumption per capita growth as well. Consumption growth after a deregulation reform mimics the growth of income
per capita. This is easily seen in Fig. 5.5b. As with other reforms, variation
in the consumption growth is smaller for those who reversed deregulation,
and far greater for the reformers. Effectively, this means triggering any
causal effects from deregulation to consumption growth needs to be conditioned on other institutional factors as well. For example, China, Hong
Kong, Indonesia, Malaysia, and Thailand enjoyed perhaps the highest
consumption growth rates over the entire 44-year period. And yet, none
of them were among the leaders in deregulation. Similar conclusions can
110
P. Stankov
be drawn from observing the consumption–deregulation link at smaller
time intervals.
Unlike consumption, life expectancy growth seems largely unaffected
by deregulation reforms. This is evident from Fig. 5.5c. The figure demonstrates a mildly positive correlation over the entire period between 1970
and 2014—a fact which will be confirmed to a certain degree by the
regression estimates as well.
In addition, Fig. 5.5d tells us that deregulation had something to do
with the rising income inequality. The relationship is strong and positive,
and is evident in almost all studied periods, confirming recent evidence
of Pérez-Moreno and Angulo-Guerrero (2016). For the entire 38-year
period before the Crisis, a 3-point increase in the index of regulation was
associated with an average increase in the Gini coefficient of approximately
7 points. This is a rather large effect if we can interpret the relationship in
a causal sense, and indeed, the regression estimates portray a less dramatic
effect of deregulation on inequality, although they do validate the positive
sign.
When it comes to countries in Central and Eastern Europe (CEE),
the causal effect was probably indeed there to be seen. CEE countries
were among the leaders in deregulation between 1990 and 2008. At the
same time, some of them were also among the leaders in growing income
inequality, including Bulgaria, Romania, and Russia. Most of the increase
happened during the notorious 1990s, when Bulgaria and Romania raised
their Gini coefficients by approximately 10 points. At the same time,
Russia raised its Gini by nearly 14 points. That is a colossal increase in
income inequality for a single decade.
Why could income inequality be related to deregulation? Two reasons seem obvious. First, and similarly to trade reforms, deregulation
creates new opportunities for entrepreneurship by lowering the costs
of doing business. Then, depending on the underlying distribution of
entrepreneurial talent in the economy, some countries will experience
higher increase in income inequality than others. The increase will
be higher within those countries with a more diverse distribution of
entrepreneurial talent. The latter fact will surface as a higher increase in
income inequality in the data.
5 Welfare and Reforms: Evidence
111
Second, income inequality will rise more in countries that deregulate
more because of different underlying political and economic institutions.
In countries with low political accountability and weak property rights
protection, the use of talent will be channeled into rent-seeking activities
with the political class and the bureaucracy (Mudambi et al. 2002; Murphy
et al. 1991). In turn, this will channel government purchases to agents close
to the politician or the bureaucrat. Then, those agents will profiteer with
taxpayers money. The effect is expected to be stronger in societies with
poorer political accountability and property rights protection mechanisms.
Virtually all leaders in income inequality growth, irrespective of whether
they are in CEE, Latin America or Africa, are also notorious for their
corruptible judicial system and political class.
The overall link between more economic freedom and higher income
inequality is not reserved for the emerging markets alone. It was observed
earlier in the USA at the state level by Ashby and Sobel (2008), Compton
et al. (2014), Apergis et al. (2014), and in a panel of countries by Krieger
and Meierrieks (2016).
5.2
Empirical Strategy and Data
The above graphical evidence is informative but could also be misleading.
In what follows, a more rigorous approach is used to determine if the
correlations studied above are indeed there to be seen and if they are
statistically significant. Finally, we need to understand how the correlations
behave in the presence of other important factors which were excluded
from the two-dimensional graphical relationships.
To test for statistical significance and to include other potentially important factors working in the background to change welfare, I estimate the
effect of each of the reforms and each of the welfare proxies studied above.
The estimated regression model is:
Yit = β1 + β2 X it + β3 Rit + εit ,
(5.1)
where Yit is any of the four welfare proxies: log(GDP/c.), log
(Consumption/c.), life expectancy at birth, or Gini coefficient for country
112
P. Stankov
i over period t; X it is a vector of explanatory variables: log(Capital per
capita)—L(K/c.)—for the GDP per capita equation, and log(GDP per
capita)—L(GDP/c.)—for the consumption per capita, life expectancy,
and income inequality equations. X it also includes a human capital index,
HC, and the log of the country’s population, L(Pop).
Both GDP/c. and Consumption/c. data are taken from the Penn World
Table, version 9.0 (PWT9.0). Life expectancy at birth data is taken from
the World Development Indicators (WDI) data base. Income inequality
data are taken from Milanovic (2014). The HC index and the population
data are taken from the PWT9.0.
The index Rit measures the development of economic freedom reforms
in one of the five reform areas: Size of Government (SoG), Property Rights
(PR), Monetary Stability (SM), Freedom to Trade (FT), and Government
Regulation (Reg). The reforms data are taken from the 2016 Economic
Freedom Dataset (Gwartney et al. 2016).
Two sets of estimations are performed for each of the four welfare
proxies: in levels and in differences. The level estimations are done to
extend the graphical evidence and to explore correlations between welfare
and reforms, whereas the estimations in differences are done to answer the
following question: Do reforms—i.e., the changes in the policy indices
over time—affect welfare in any positive way?
Three separate models are run for each set of estimations. The first
is simply a panel ordinary least squares (OLS) model. The second is
the panel OLS model, which further includes unobservable country and
time effects. Fixed-effects methods can potentially address two issues
with the panel OLS: (i) Some of the unobserved country characteristics
affect both the likelihood of reforms and welfare, and therefore, excluding
them from the panel OLS model is wrong; (ii) the time-fixed effects could
capture the effect of contemporaneous global events on country-level welfare, e.g., the oil shocks of the 1970s, the currency crises in the 1980s and
1990s, and of course, the Great Recession.
5 Welfare and Reforms: Evidence
113
The third model runs a two-stage least squares (2SLS) estimation, in
which Rit is instrumented with the potential rents from natural resources.
More specifically, the instrumental variable (IV) for the policy changes
is the changes in the total natural resource rents. These are the sum of
oil rents, natural gas rents, coal rents, mineral rents, and forest rents,
expressed as a share of GDP in the country. The rent from a particular
natural resource is typically the potential value of the resource, if it were
extracted and sold at world prices, minus the costs of its extraction. The
rents data are taken from the WDI. As we can see, total rents do not measure actual production values from natural resources for a given country.
Rather, they measure the estimated potential rents from being endowed
with those resources. Then, the total rents are arguably uncorrelated with
the contemporaneous welfare measures.
At the same time, the political economy of reforms literature finds that
the potential rents from natural resources could predict the timing of
reforms and the change in the reform indices. Studies in this direction
have been published, among others, by Beck and Laeven (2006); Levine
(2005); Mulligan and Tsui (2008); Tsui (2011). All of those studies suggest
that resource discoveries hamper market-oriented reforms. These findings
support the validity of using the dynamics of rents as a predictor of economic freedom reforms.
As rents are uncorrelated with welfare but correlated with reforms, they
can be used to identify at least a part of the exogenous component of
reforms to the changes in welfare. To check the validity of this approach,
the presented results from the 2SLS estimations include also the p-value
of the Hansen J test. This test is commonly used in the literature to check
the validity of a given IV. In this particular case, it tests for the validity
of using rents as an instrument for reforms in the first stage of the 2SLS
estimation.
Finally, all standard errors reported are clustered on the country level.
This procedure ensures that the within-country error terms are not treated
independently across each observation in the data. Further, it ensures that
the standard errors are robust to heteroskedasticity. The estimation results
from running the three models in both levels and differences are presented
below.
114
5.3
P. Stankov
Results
There are four welfare measures which are affected by five economic freedom reforms. For each reform, two tables are constructed. The first table
is the one in which GDP/c. and Consumption/c. are the dependent variables. The second table is the one in which life expectancy and income
inequality are the dependent variables. Thus, Tables 5.1 and 5.2 report
estimates in which the change in government intervention is the main
explanatory variable; Tables 5.3 and 5.4 report estimates for property
rights; Tables 5.5 and 5.6—for monetary policies; Tables 5.7 and 5.8—
for trade reforms; and Tables 5.9 and 5.10—for the overall deregulation.
Table 5.1 presents evidence that, indeed, changes in government size
are not a crucial factor for either increasing or decreasing welfare. If we
look at how government intervention affects income growth over time,
we will notice that neither the level nor the difference estimations yield
any statistically significant results. As the literature review has demonstrated, both theory and evidence in this respect are ambiguous (Davies
2009; Fölster and Henrekson 2001). The vast majority of results report
negative effects of government intervention on total factor productivity,
growth, and entrepreneurship (Afonso and Furceri 2010; Bergh and Karlsson 2010; Bjørnskov and Foss 2008; Dar and Amirkhalkhali 2002; Lee
1996; Nyström 2008). A minority of studies inform of welfare-improving
interventions (Dinopoulos and Unel 2011). The results here demonstrate
that, contrary to the majority of earlier evidence, government intervention did not play a significant role in spurring income per capita growth,
something we already suspected the data would deliver when looking at
the graphical evidence. The results here extend the recent evidence by
Kacprzyk (2016) who do not find a statistically significant effect of reducing government intervention on growth in a sample of 28 EU countries.
The estimates demonstrate that government intervention played an
equally negligible role in increasing consumption per capita over time. As
a result, we need to approach any policy recommendations advocating for
either more or less government intervention for the long-term welfare of
an economy with caution. They both could be very much based on desired
results or ideology rather than on tangible evidence—an implication similar to the ones formulated earlier by Carlsson and Lundstrom (2002).
115
5 Welfare and Reforms: Evidence
Table 5.1
Size of government, income, and consumption: 1970–2014
Dependent Variable: Income per Capita
Estimation in levels
Estimation in differences
(1)
(2)
(3)
(4)
(5)
(6)
L(K/c.)
HC
L(Pop)
SoG
N
R2
FEs
Hansen
0.657***
(0.044)
0.380***
(0.061)
−0.042**
(0.020)
0.006
(0.009)
1128
0.897
No
0.541***
(0.061)
0.587***
(0.152)
−0.302**
(0.117)
−0.001
(0.011)
1128
0.701
Yes
0.573***
(0.075)
0.365*
(0.207)
−0.442*
(0.255)
0.256
(0.227)
987
0.293
Yes
0.781
0.597***
(0.040)
0.148
(0.104)
−0.210
(0.131)
0.006
(0.007)
988
0.173
No
0.624***
(0.057)
0.054
(0.153)
0.353
(0.415)
0.003
(0.007)
988
0.137
Yes
0.635***
(0.069)
0.079
(0.187)
0.085
(0.499)
−0.000
(0.248)
858
0.129
Yes
0.869
Dependent Variable: Consumption per Capita
Estimation in levels
Estimation in differences
L(GDP/c.)
HC
L(Pop)
SoG
N
R2
FEs
Hansen
(1)
(2)
(3)
(4)
(5)
(6)
0.579***
(0.049)
0.285***
(0.065)
−0.055***
(0.017)
−0.009
(0.006)
1128
0.913
No
0.461***
(0.061)
0.235**
(0.098)
−0.253***
(0.069)
0.000
(0.006)
1128
0.785
Yes
0.439***
(0.070)
0.202*
(0.115)
−0.408***
(0.099)
0.091
(0.096)
987
0.691
Yes
0.002
0.430***
(0.049)
0.032
(0.056)
−0.179***
(0.068)
−0.001
(0.003)
988
0.436
No
0.377***
(0.050)
0.006
(0.073)
−0.066
(0.152)
0.001
(0.003)
988
0.409
Yes
0.341***
(0.095)
0.026
(0.191)
−0.344
(0.381)
0.274
(0.344)
858
–
Yes
0.156
Notes The estimated panel OLS equation in differences is Yit = β1 + β2 X it +
β3 Rit + εit , where Yit is one of the following: log(GDP/c.), log(Consumption/c.),
life expectancy at birth, or Gini coefficient. X it is a vector of explanatory variables:
log(Capital per capita)—L(K/c.)—for the income per capita equation, or log(GDP
per capita)—L(GDP/c.)—for the consumption per capita equation; human capital—
HC; log(population)—L(Pop). Rit is a reform variable: either Size of Government
(SoG), or Property Rights (PR), or Monetary Stability (SM), or Freedom to Trade
(FT), or Regulation (Reg). Models (2), (3), (5), and (6) include country- and timefixed effects. In models (3) and (6), R is instrumented with the potential natural
resource rents, as detailed in the text. For those models, the p-value of the Hansen J
test is presented as well. Clustered standard errors are in parentheses. Data source
PWT9.0, WDI, EFW 2016 index. Symbols * p < 0.10, ** p < 0.05, *** p < 0.01
116
P. Stankov
Table 5.2
Size of government, life expectancy, and inequality: 1970–2014
Dependent Variable: Life Expectancy at Birth
Estimation in levels
Estimation in differences
(1)
(2)
(3)
(4)
(5)
(6)
L(GDP/c.)
2.573***
(0.667)
HC
7.924***
(1.037)
L(Pop)
1.641***
(0.405)
SoG
0.057
(0.131)
Observations 1128
R2
0.680
FEs
No
Hansen
L(GDP/c.)
HC
L(Pop)
SoG
N
R2
FEs
Hansen
2.279***
(0.773)
−3.327*
(1.789)
4.609***
(1.170)
−0.063
(0.134)
1128
0.704
Yes
2.343***
(0.796)
−3.216
(2.082)
4.178**
(1.891)
−0.071
(1.557)
987
0.654
Yes
0.118
1.324***
(0.342)
−1.166
(0.886)
5.659***
(1.135)
−0.144***
(0.049)
988
0.075
No
1.065***
(0.325)
−0.234
(1.154)
6.969***
(2.595)
−0.071
(0.051)
988
0.124
Yes
0.850
(0.617)
−0.296
(1.819)
5.762*
(3.439)
1.875
(2.305)
858
–
Yes
0.686
Dependent Variable: Gini Coefficient
Estimation in levels
Estimation in differences
(1)
(2)
(3)
(4)
(5)
(6)
−0.464
−2.371
−2.566
−5.162*** −6.511*** −2.718
(1.006)
(1.721)
(2.105)
(1.802)
(1.861)
(2.456)
−2.847
2.113
0.621
0.427
−1.418
2.497
(1.738)
(4.944)
(6.360)
(5.156)
(8.002)
(9.553)
−0.110
−3.327
−1.447
−5.170
5.591
16.292
(0.541)
(4.164)
(6.338)
(7.986)
(16.771)
(22.168)
1.204*** 0.712** −1.645
0.553
0.580
2.053
(0.276)
(0.309)
(2.692)
(0.412)
(0.462)
(2.237)
397
397
350
232
232
198
0.318
0.088
–
0.052
0.156
–
No
Yes
Yes
No
Yes
Yes
0.237
0.593
Notes The estimated panel OLS equation in differences is Yit = β1 + β2 X it +
β3 Rit + εit , where Yit is one of the following: log(GDP/c.), log(Consumption/c.),
life expectancy at birth, or Gini coefficient. X it is a vector of explanatory variables:
log(GDP per capita)—L(GDP/c.); human capital—HC; log(population)—L(Pop). Rit
is a reform variable: either Size of Government (SoG), or Property Rights (PR), or
Monetary Stability (SM), or Freedom to Trade (FT), or Regulation (Reg). Models
(2), (3), (5), and (6) include country- and time-fixed effects. In models (3) and (6), R
is instrumented with the potential natural resource rents, as detailed in the text.
For those models, the p-value of the Hansen J test is presented as well. Clustered
standard errors are in parentheses. Data source PWT9.0, WDI, Milanovic (2014),
EFW 2016 index. Symbols * p < 0.10, ** p < 0.05, *** p < 0.01
5 Welfare and Reforms: Evidence
Table 5.3
117
Property rights, income, and consumption: 1970–2014
Dependent Variable: Income per Capita
Estimation in levels
Estimation in differences
(1)
(2)
(3)
(4)
(5)
(6)
L(K/c.)
HC
L(Pop)
PR
N
R2
FEs
Hansen
L(GDP/c.)
HC
L(Pop)
PR
N
R2
FEs
Hansen
0.640***
(0.048)
0.391***
(0.063)
−0.020
(0.017)
0.029***
(0.010)
1020
0.902
No
0.515***
(0.068)
0.493***
(0.170)
−0.281**
(0.125)
0.044***
(0.012)
1020
0.719
Yes
0.437*** 0.599***
0.672*** 0.661***
(0.152)
(0.042)
(0.059)
(0.089)
0.218
0.017
−0.061
0.054
(0.267)
(0.110)
(0.169)
(0.440)
−0.445
−0.264**
0.061
−0.217
(0.280)
(0.134)
(0.452)
(0.810)
0.392
0.011*
0.018**
−0.035
(0.402)
(0.006)
(0.008)
(0.346)
929
881
881
800
0.100
0.177
0.162
0.091
Yes
No
Yes
Yes
0.556
0.880
Dependent Variable: Consumption per Capita
Estimation in levels
Estimation in differences
(1)
(2)
(3)
(4)
(5)
(6)
0.573***
0.421***
0.459*** 0.414***
0.363*** 0.153
(0.056)
(0.065)
(0.093)
(0.053)
(0.056)
(0.379)
0.275***
0.163*
0.214*
0.037
0.024
−0.408
(0.071)
(0.094)
(0.118)
(0.055)
(0.068)
(0.792)
−0.057*** −0.250*** −0.244* −0.229*** −0.161
0.774
(0.016)
(0.061)
(0.124)
(0.068)
(0.158)
(1.683)
0.004
0.006
−0.085
0.005
0.012**
0.508
(0.008)
(0.009)
(0.186)
(0.004)
(0.005)
(0.912)
1020
1020
929
881
881
800
0.918
0.805
0.688
0.434
0.416
–
No
Yes
Yes
No
Yes
Yes
0.007
0.556
Notes The estimated panel OLS equation in differences is Yit = β1 + β2 X it +
β3 Rit + εit , where Yit is one of the following: log(GDP/c.), log(Consumption/c.),
life expectancy at birth, or Gini coefficient. X it is a vector of explanatory variables:
log(Capital per capita)—L(K/c.)—for the income equation, or log(GDP per capita)—
L(GDP/c.)—for the consumption equation; human capital—HC; log(population)—
L(Pop). Rit is a reform variable: either Size of Government (SoG), or Property Rights
(PR), or Monetary Stability (SM), or Freedom to Trade (FT), or Regulation (Reg).
Models (2), (3), (5), and (6) include country- and time-fixed effects. In models (3)
and (6), R is instrumented with the potential natural resource rents, as detailed in
the text. For those models, the p-value of the Hansen J test is presented as well.
Clustered standard errors are in parentheses. Data source PWT9.0, WDI, EFW 2016
index. Symbols * p < 0.10, ** p < 0.05, *** p < 0.01
118
P. Stankov
Table 5.4
Property rights, life expectancy, and inequality: 1970–2014
Dependent Variable: Life Expectancy at Birth
Estimation in levels
Estimation in differences
(1)
(2)
(3)
(4)
(5)
(6)
L(GDP/c.)
HC
L(Pop)
PR
N
R2
FEs
Hansen
L(GDP/c.)
HC
L(Pop)
PR
N
R2
FEs
Hansen
2.803***
(0.611)
7.208***
(0.999)
1.557***
(0.363)
0.377***
(0.129)
1020
0.678
No
2.219***
(0.666)
−1.786
(1.514)
4.697***
(1.253)
0.387**
(0.154)
1020
0.718
Yes
2.887
1.259*** 0.911*** 0.956
(1.882)
(0.364)
(0.334)
(0.792)
−1.741
−0.814
−0.719
−0.440
(1.830)
(0.957)
(1.207)
(2.106)
5.186** 5.943*** 7.451**
5.365
(2.333)
(1.667)
(3.141)
(5.004)
−0.767
0.004
0.053
−0.407
(2.615)
(0.045)
(0.060)
(1.868)
929
881
881
800
0.622
0.063
0.119
0.037
Yes
No
Yes
Yes
0.773
0.224
Dependent Variable: Gini Coefficient
Estimation in levels
Estimation in differences
(1)
(2)
(3)
(4)
(5)
(6)
−0.097
0.390
0.934
−2.028
−2.625
2.734
(0.890)
(1.457)
(2.034)
(1.520)
(2.013)
(6.408)
−1.231
3.035
6.374
−0.941
−2.661
4.972
(1.494)
(4.178)
(4.894)
(4.918)
(6.903)
(8.399)
0.137
−1.995
−3.915
−1.684
6.258
−11.397
(0.507)
(3.690)
(4.063)
(8.556)
(16.361)
(36.943)
−0.977*** −0.916** −2.097
0.020
−0.140
−3.278
(0.310)
(0.365)
(1.299)
(0.360)
(0.412)
(4.996)
389
389
343
231
231
198
0.264
0.077
0.059
0.005
0.085
−0.262
No
Yes
Yes
No
Yes
Yes
0.371
0.520
Notes The estimated panel OLS equation in differences is Yit = β1 + β2 X it +
β3 Rit + εit , where Yit is one of the following: log(GDP/c.), log(Consumption/c.),
life expectancy at birth, or Gini coefficient. X it is a vector of explanatory variables:
log(GDP per capita)—L(GDP/c.); human capital—HC; log(population)—L(Pop). Rit
is a reform variable: either Size of Government (SoG), or Property Rights (PR), or
Monetary Stability (SM), or Freedom to Trade (FT), or Regulation (Reg). Models
(2), (3), (5), and (6) include country- and time-fixed effects. In models (3) and (6), R
is instrumented with the potential natural resource rents, as detailed in the text.
For those models, the p-value of the Hansen J test is presented as well. Clustered
standard errors are in parentheses. Data source PWT9.0, WDI, Milanovic (2014),
EFW 2016 index. Symbols * p < 0.10, ** p < 0.05, *** p < 0.01
5 Welfare and Reforms: Evidence
Table 5.5
119
Monetary stability, income, and consumption: 1970–2014
Dependent Variable: Income per Capita
Estimation in levels
Estimation in differences
(1)
(2)
(3)
(4)
(5)
(6)
L(K/c.)
HC
L(Pop)
SM
N
R2
FEs
Hansen
L(GDP/c.)
HC
L(Pop)
SM
N
R2
FEs
Hansen
0.621***
(0.045)
0.349***
(0.061)
−0.048**
(0.020)
0.031***
(0.006)
1158
0.893
No
0.523***
(0.054)
0.556***
(0.145)
−0.251**
(0.107)
0.025***
(0.007)
1158
0.709
Yes
0.525*** 0.583*** 0.592*** 0.612***
(0.071)
(0.042)
(0.053)
(0.145)
0.321*
0.047
−0.038
0.066
(0.187)
(0.104)
(0.141)
(0.150)
−0.036
−0.087
0.427
0.076
(0.258)
(0.110)
(0.309)
(0.443)
0.151
0.023*** 0.023*** 0.016
(0.142)
(0.006)
(0.006)
(0.076)
993
1019
1019
864
0.488
0.185
0.171
0.147
Yes
No
Yes
Yes
0.210
0.703
Dependent Variable: Consumption per Capita
Estimation in levels
Estimation in differences
(1)
(2)
(3)
(4)
(5)
(6)
0.592***
0.473***
0.298*** 0.440*** 0.388*** 0.262***
(0.051)
(0.064)
(0.106)
(0.050)
(0.052)
(0.084)
0.262***
0.217**
0.183
0.026
0.020
−0.048
(0.062)
(0.096)
(0.165)
(0.058)
(0.073)
(0.134)
−0.059*** −0.238*** −0.164
−0.141** −0.080
−0.236
(0.017)
(0.069)
(0.163)
(0.062)
(0.134)
(0.209)
0.003
0.008*
0.167*
0.008*** 0.014*** 0.093**
(0.004)
(0.004)
(0.098)
(0.003)
(0.003)
(0.038)
1158
1158
993
1019
1019
864
0.914
0.785
0.155
0.457
0.456
–
No
Yes
Yes
No
Yes
Yes
0.196
0.023
Notes The estimated panel OLS equation in differences is Yit = β1 + β2 X it +
β3 Rit + εit , where Yit is one of the following: log(GDP/c.), log(Consumption/c.),
life expectancy at birth, or Gini coefficient. X it is a vector of explanatory variables:
log(Capital per capita)—L(K/c.)—for the income per capita equation, or log(GDP
per capita)—L(GDP/c.)—for the consumption per capita equation; human capital—
HC; log(population)—L(Pop). Rit is a reform variable: either Size of Government
(SoG), or Property Rights (PR), or Monetary Stability (SM), or Freedom to Trade
(FT), or Regulation (Reg). Models (2), (3), (5), and (6) include country- and timefixed effects. In models (3) and (6), R is instrumented with the potential natural
resource rents, as detailed in the text. For those models, the p-value of the Hansen J
test is presented as well. Clustered standard errors are in parentheses. Data source
PWT9.0, WDI, EFW 2016 index. Symbols * p < 0.10, ** p < 0.05, *** p < 0.01
120
P. Stankov
Table 5.6
Monetary stability, life expectancy, and inequality: 1970–2014
Dependent Variable: Life Expectancy at Birth
Estimation in levels
Estimation in differences
(1)
(2)
(3)
(4)
(5)
(6)
L(GDP/c.)
HC
L(Pop)
SM
N
R2
FEs
Hansen
L(GDP/c.)
HC
L(Pop)
SM
N
R2
FEs
Hansen
2.461***
(0.621)
7.771***
(1.016)
1.694***
(0.404)
0.168*
(0.092)
1158
0.679
No
2.079***
(0.696)
−3.409**
(1.706)
4.569***
(1.111)
0.223**
(0.103)
1158
0.711
Yes
2.114*
1.253***
0.870***
0.434
(1.161)
(0.297)
(0.264)
(0.455)
−3.286
−1.701**
−0.799
−0.720
(1.999)
(0.819)
(1.048)
(1.331)
4.143*** 5.862***
6.812***
6.488***
(1.584)
(1.072)
(2.310)
(2.510)
0.259
0.110**
0.170***
0.479**
(1.078)
(0.049)
(0.051)
(0.234)
993
1019
1019
864
0.658
0.085
0.152
0.063
Yes
No
Yes
Yes
0.672
0.871
Dependent Variable: Gini Coefficient
Estimation in levels
Estimation in differences
(1)
(2)
(3)
(4)
(5)
(6)
−1.764
−2.992
−1.641
−4.927*** −5.322*** −1.757
(1.186)
(1.810)
(2.638)
(1.597)
(1.965)
(2.121)
−0.126
4.771
−9.837
4.779
1.321
4.118
(1.912)
(4.716)
(26.496)
(5.247)
(8.099)
(8.362)
−0.011
−4.525
9.817
−1.921
−9.056
5.576
(0.552)
(4.332)
(28.272)
(8.282)
(17.936)
(16.568)
0.173
−0.112
2.795
−0.345
−0.415**
−0.900
(0.208)
(0.233)
(7.098)
(0.215)
(0.205)
(0.703)
405
405
352
238
238
200
0.126
0.083
–
0.065
0.171
0.030
No
Yes
Yes
No
Yes
Yes
0.303
0.852
Notes The estimated panel OLS equation in differences is Yit = β1 + β2 X it +
β3 Rit + εit , where Yit is one of the following: log(GDP/c.), log(Consumption/c.),
life expectancy at birth, or Gini coefficient. X it is a vector of explanatory variables:
log(GDP per capita)—L(GDP/c.); human capital—HC; log(population)—L(Pop). Rit
is a reform variable: either Size of Government (SoG), or Property Rights (PR), or
Monetary Stability (SM), or Freedom to Trade (FT), or Regulation (Reg). Models
(2), (3), (5), and (6) include country- and time-fixed effects. In models (3) and (6), R
is instrumented with the potential natural resource rents, as detailed in the text.
For those models, the p-value of the Hansen J test is presented as well. Clustered
standard errors are in parentheses. Data source PWT9.0, WDI, Milanovic (2014),
EFW 2016 index. Symbols * p < 0.10, ** p < 0.05, *** p < 0.01
5 Welfare and Reforms: Evidence
Table 5.7
121
Free trade, income, and consumption: 1970–2014
Dependent Variable: Income per Capita
Estimation in levels
Estimation in differences
(1)
(2)
(3)
(4)
(5)
(6)
L(K/c.)
HC
L(Pop)
FT
N
R2
FEs
Hansen
L(GDP/c.)
HC
L(Pop)
FT
N
R2
FEs
Hansen
0.625***
(0.047)
0.399***
(0.071)
−0.049**
(0.019)
0.006
(0.010)
1097
0.893
No
0.499***
(0.057)
0.568***
(0.146)
−0.377***
(0.111)
0.012
(0.010)
1097
0.702
Yes
0.518*** 0.581*** 0.683*** 0.588***
(0.077)
(0.038)
(0.055)
(0.151)
0.417**
0.109
0.019
0.204
(0.165)
(0.110)
(0.160)
(0.237)
−0.382* −0.205*
0.396
−0.083
(0.199)
(0.119)
(0.409)
(0.528)
0.105
0.018**
0.024**
−0.084
(0.137)
(0.008)
(0.009)
(0.128)
975
951
951
840
0.602
0.170
0.163
–
Yes
No
Yes
Yes
0.206
0.903
Dependent Variable: Consumption per Capita
Estimation in levels
Estimation in differences
(1)
(2)
(3)
(4)
(5)
(6)
0.591***
0.461***
0.481*** 0.437*** 0.395*** 0.456***
(0.052)
(0.064)
(0.057)
(0.053)
(0.055)
(0.078)
0.276***
0.198**
0.256*
0.061
0.054
0.168
(0.065)
(0.099)
(0.135)
(0.055)
(0.069)
(0.140)
−0.053*** −0.244*** −0.183
−0.177** −0.049
−0.267
(0.017)
(0.070)
(0.168)
(0.070)
(0.170)
(0.230)
−0.003
−0.001
−0.156
−0.007** −0.005
−0.152*
(0.007)
(0.007)
(0.120)
(0.003)
(0.003)
(0.082)
1097
1097
975
951
951
840
0.916
0.789
0.410
0.437
0.421
–
No
Yes
Yes
No
Yes
Yes
0.134
0.560
Notes The estimated panel OLS equation in differences is Yit = β1 + β2 X it +
β3 Rit + εit , where Yit is one of the following: log(GDP/c.), log(Consumption/c.),
life expectancy at birth, or Gini coefficient. X it is a vector of explanatory variables:
log(Capital per capita)—L(K/c.)—for the income per capita equation, or log(GDP
per capita)—L(GDP/c.)—for the consumption per capita equation; human capital—
HC; log(population)—L(Pop). Rit is a reform variable: either Size of Government
(SoG), or Property Rights (PR), or Monetary Stability (SM), or Freedom to Trade
(FT), or Regulation (Reg). Models (2), (3), (5), and (6) include country- and timefixed effects. In models (3) and (6), R is instrumented with the potential natural
resource rents, as detailed in the text. For those models, the p-value of the Hansen J
test is presented as well. Clustered standard errors are in parentheses. Data source
PWT9.0, WDI, EFW 2016 index. Symbols * p < 0.10, ** p < 0.05, *** p < 0.01
122
P. Stankov
Table 5.8
Free trade, life expectancy, and inequality: 1970–2014
Dependent Variable: Life Expectancy at Birth
Estimation in levels
Estimation in differences
(1)
(2)
(3)
(4)
(5)
(6)
L(GDP/c.)
HC
L(Pop)
FT
N
R2
FEs
Hansen
L(GDP/c.)
HC
L(Pop)
FT
N
R2
FEs
Hansen
2.543***
(0.601)
6.835***
(0.982)
1.611***
(0.394)
0.497***
(0.121)
1097
0.686
No
2.143***
(0.647)
−3.057*
(1.586)
4.621***
(1.086)
0.334***
(0.109)
1097
0.723
Yes
2.479*** 1.204*** 0.936*** 0.910**
(0.858)
(0.324)
(0.300)
(0.395)
−3.194*
−0.849
−0.370
−0.127
(1.801)
(0.896)
(1.098)
(1.347)
4.733*** 4.897*** 5.869*** 4.902**
(1.624)
(1.009)
(2.056)
(2.065)
−0.308
−0.025
0.047
−0.121
(0.891)
(0.042)
(0.048)
(0.347)
975
951
951
840
0.644
0.063
0.134
0.104
Yes
No
Yes
Yes
0.852
0.140
Dependent Variable: Gini Coefficient
Estimation in levels
Estimation in differences
(1)
(2)
(3)
(4)
(5)
(6)
−0.592
−0.681
−1.086
−2.690*
−3.870
−1.246
(0.972)
(1.551)
(1.818)
(1.626)
(2.338)
(2.217)
−4.258** −0.445
0.331
2.244
4.360
0.469
(1.679)
(4.812)
(7.723)
(5.255)
(6.834)
(7.991)
0.061
−2.516
−2.705
−8.091
−9.919
5.305
(0.569)
(3.783)
(6.482)
(8.035)
(19.951)
(19.030)
1.122***
0.914**
0.306
0.395
−0.019
−1.217
(0.345)
(0.367)
(2.840)
(0.449)
(0.470)
(1.568)
395
395
349
231
231
199
0.229
0.087
0.052
0.031
0.125
0.050
No
Yes
Yes
No
Yes
Yes
0.089
0.529
Notes The estimated panel OLS equation in differences is Yit = β1 + β2 X it +
β3 Rit + εit , where Yit is one of the following: log(GDP/c.), log(Consumption/c.),
life expectancy at birth, or Gini coefficient. X it is a vector of explanatory variables:
log(GDP per capita)—L(GDP/c.); human capital—HC; log(population)—L(Pop). Rit
is a reform variable: either Size of Government (SoG), or Property Rights (PR), or
Monetary Stability (SM), or Freedom to Trade (FT), or Regulation (Reg). Models
(2), (3), (5), and (6) include country- and time-fixed effects. In models (3) and (6), R
is instrumented with the potential natural resource rents, as detailed in the text.
For those models, the p-value of the Hansen J test is presented as well. Clustered
standard errors are in parentheses. Data source PWT9.0, WDI, Milanovic (2014),
EFW 2016 index. Symbols * p < 0.10, ** p < 0.05, *** p < 0.01
5 Welfare and Reforms: Evidence
Table 5.9
123
Deregulation, income, and consumption: 1970–2014
Dependent Variable: Income per Capita
Estimation in levels
Estimation in differences
(1)
(2)
(3)
(4)
(5)
(6)
L(K/c.)
HC
L(Pop)
Reg
N
R2
FEs
Hansen
L(GDP/c.)
HC
L(Pop)
Reg
N
R2
FEs
Hansen
0.631***
(0.048)
0.355***
(0.070)
−0.035*
(0.019)
0.040***
(0.013)
1052
0.896
No
0.511***
(0.058)
0.517***
(0.141)
−0.302***
(0.109)
0.031
(0.019)
1052
0.701
Yes
0.484*** 0.579***
0.643*** 0.660***
(0.088)
(0.041)
(0.060)
(0.087)
0.382**
0.072
0.016
0.050
(0.175)
(0.109)
(0.157)
(0.186)
−0.306*
−0.200
0.290
0.055
(0.161)
(0.142)
(0.468)
(0.445)
0.287
0.039***
0.039*** 0.134
(0.257)
(0.012)
(0.013)
(0.233)
954
912
912
825
0.526
0.177
0.162
0.115
Yes
No
Yes
Yes
0.217
0.806
Dependent Variable: Consumption per Capita
Estimation in levels
Estimation in differences
(1)
(2)
(3)
(4)
(5)
(6)
0.593***
0.452***
0.238*
0.421***
0.378*** 0.145
(0.053)
(0.068)
(0.128)
(0.055)
(0.059)
(0.147)
0.275***
0.207**
0.170
0.085
0.100
0.113
(0.067)
(0.095)
(0.192)
(0.055)
(0.072)
(0.155)
−0.057*** −0.258*** −0.379** −0.213*** −0.060
−0.574
(0.018)
(0.073)
(0.185)
(0.073)
(0.168)
(0.411)
0.002
0.019
0.554*** 0.005
0.006
0.408
(0.012)
(0.012)
(0.208)
(0.006)
(0.006)
(0.256)
1052
1052
954
912
912
825
0.920
0.788
–
0.429
0.405
–
No
Yes
Yes
No
Yes
Yes
0.908
0.215
Notes The estimated panel OLS equation in differences is Yit = β1 + β2 X it +
β3 Rit + εit , where Yit is one of the following: log(GDP/c.), log(Consumption/c.),
life expectancy at birth, or Gini coefficient. X it is a vector of explanatory variables:
log(Capital per capita)—L(K/c.)—for the income per capita equation, or log(GDP
per capita)—L(GDP/c.)—for the consumption per capita equation; human capital—
HC; log(population)—L(Pop). Rit is a reform variable: either Size of Government
(SoG), or Property Rights (PR), or Monetary Stability (SM), or Freedom to Trade
(FT), or Regulation (Reg). Models (2), (3), (5), and (6) include country- and timefixed effects. In models (3) and (6), R is instrumented with the potential natural
resource rents, as detailed in the text. For those models, the p-value of the Hansen J
test is presented as well. Clustered standard errors are in parentheses. Data source
PWT9.0, WDI, EFW 2016 index. Symbols * p < 0.10, ** p < 0.05, *** p < 0.01
124
Table 5.10
P. Stankov
Deregulation, life expectancy, and inequality: 1970–2014
Dependent Variable: Life Expectancy at Birth
Estimation in levels
Estimation in differences
(1)
(2)
(3)
(4)
(5)
(6)
L(GDP/c.)
HC
L(Pop)
Reg
N
R2
FEs
Hansen
L(GDP/c.)
HC
L(Pop)
Reg
N
R2
FEs
Hansen
2.783***
(0.615)
5.952***
(0.990)
1.544***
(0.394)
0.623***
(0.189)
1052
0.673
No
2.313***
(0.646)
−3.269**
(1.536)
4.544***
(1.116)
0.424**
(0.186)
1052
0.693
Yes
2.616*** 1.357***
0.887**
−0.283
(0.934)
(0.387)
(0.343)
(0.812)
−3.265*
−0.812
0.001
0.303
(1.854)
(0.883)
(1.142)
(1.450)
4.133*** 5.210***
8.253***
4.601
(1.297)
(1.179)
(2.929)
(3.206)
−0.387
0.139
0.206
2.146
(1.898)
(0.124)
(0.128)
(1.530)
954
912
912
825
0.635
0.060
0.119
–
Yes
No
Yes
Yes
0.850
0.744
Dependent Variable: Gini Coefficient
Estimation in levels
Estimation in differences
(1)
(2)
(3)
(4)
(5)
(6)
−1.597
−2.220
−2.854
−3.980** −5.266** −1.412
(0.979)
(1.970)
(3.188)
(1.947)
(2.441)
(2.338)
−2.556
1.403
−0.606
−2.412
−3.255
2.264
(1.672)
(4.910)
(7.918)
(4.678)
(6.855)
(7.422)
0.333
2.609
7.888
−0.682
1.920
26.439
(0.543)
(4.790)
(13.508)
(8.680)
(17.713)
(26.668)
1.598*** 1.243**
4.445
1.080**
0.805
2.113
(0.411)
(0.547)
(7.200)
(0.496)
(0.604)
(2.359)
393
393
347
235
235
199
0.201
0.087
–
0.059
0.136
0.027
No
Yes
Yes
No
Yes
Yes
0.739
0.570
Notes The estimated panel OLS equation in differences is Yit = β1 + β2 X it +
β3 Rit + εit , where Yit is one of the following: log(GDP/c.), log(Consumption/c.),
life expectancy at birth, or Gini coefficient. X it is a vector of explanatory variables:
log(GDP per capita)—L(GDP/c.); human capital—HC; log(population)—L(Pop). Rit
is a reform variable: either Size of Government (SoG), or Property Right (PR), or
Monetary Stability (SM), or Freedom to Trade (FT), or Regulation (Reg). Models
(2), (3), (5), and (6) include country- and time-fixed effects. In models (3) and (6), R
is instrumented with the potential natural resource rents, as detailed in the text.
For those models, the p-value of the Hansen J test is presented as well. Clustered
standard errors are in parentheses. Data source PWT9.0, WDI, Milanovic (2014),
EFW 2016 index. Symbols * p < 0.10, ** p < 0.05, *** p < 0.01
5 Welfare and Reforms: Evidence
125
Further, the size of government did not play a significant role in increasing life expectancy. This is seen in Table 5.2. It is somewhat intuitive to
observe this relationship. Less government intervention means less government spending. However, less government spending is also likely associated with lower healthcare expenditure, which directly affects the quality of
healthcare services, in turn affecting life expectancy. If this logic is correct,
then we should expect to see statistical significance more often. However,
in only one of the estimations is this effect observable—the difference estimation without fixed effects. It is a matter of further empirical research to
question the role of government intervention on life expectancy.
The graphical correlations demonstrate that government intervention
plays a positive and significant role in increasing income inequality. Some,
yet not all, of the regression estimates confirm the positive correlation. The
significant estimates can be seen in the level equations only. This means
that countries with less government intervention also have higher income
inequality, given the effects of other explanatory factors. It is therefore
tempting to infer that less government also increases income inequality.
Yet, given the insignificance of the difference estimates, we cannot recommend increasing government intervention in order to reduce income
inequality. This is an important point, as many left-wing economists produce the following recipe to deal with inequality: Let’s strengthen the
redistributive role of the government to reduce income inequality. In the
presence of rising income inequality, it has been shown that a typical voter
prefers stronger redistributive government (Meltzer and Richard 1981;
Pecoraro 2017). However, Table 5.2 produces evidence that larger governments will not reduce income inequality in the long run. The case
studies reviewed in the next chapter suggest further caution should be
applied when using the large-scale redistributive role of the government
to address income inequality problems.
The graphical evidence is firmly in favor of the hypothesis that improving property rights protection is associated with raising welfare. The regression results demonstrate that this is not always the case, as the results of
Chu et al. (2014) and Trebilcock and Veel (2008) suggest. Indeed, improving PRs had a positive and statistically significant effect on income per
capita growth. This is seen from both the level and the difference estimations in Table 5.3. These results are in agreement with early empirical
126
P. Stankov
conclusions of Torstensson (1994) and Goldsmith (1995), and later findings of Farhadi et al. (2015), Gwartney (2009), Mehlum et al. (2005), and
Mijiyawa (2008), among others. However, the 2SLS estimates produce
insignificant results, which question the robustness of earlier evidence by
Sturm and De Haan (2001) and Gwartney et al. (2006), among many others. These non-robust estimates reiterate the issue of a possible publication
bias in the freedom-growth studies, outlined well by Doucouliagos (2005).
PRs had an even weaker impact on the growth of consumption per
capita. The graphical evidence suggested that the relationship between
the two was strong. And yet, the empirical evidence in favor of a strong
relationship is confined to the fixed-effects estimation in differences. In
most cases, that could be sufficient evidence in favor of improving PRs in
order to raise consumption across countries over time, as the evidence by
Kemper et al. (2015) has already suggested for Vietnam.
Life expectancy is also positively affected by improving PRs. This is seen
in Table 5.4 and is consistent with the evidence of Stroup (2007). Again,
the evidence is there but is somewhat mixed. On the one hand, people do
live longer in countries with better PRs. On the other hand, reforming
PRs did not lead to longer lives when factors such as GDP/c. growth and
population growth were taken into account. This may be the case because,
while the economy grows, growth may enable the government to invest
more in both healthcare improvement and judicial system effectiveness.
This would surface as a positive correlation in both the graph and the level
equation but would not necessarily show up as a statistically significant
result in a difference estimation with time and country-fixed effects like
the one presented in the last three columns of Table 5.4.
Income inequality dynamics are also not explained well by improvements in PRs. The graphical evidence was mixed, and so is the regression evidence. Indeed, if we focus on the level estimates alone, we find
the expected negative correlation between PRs and Gini. That is, countries with better protected PRs also enjoy lower income inequality, and
countries with higher income inequality also have weaker property rights,
among other elements of economic freedom (Krieger and Meierrieks
2016). However, as in most previous cases, improving PRs is not a sure
recipe for income inequality reduction, as the difference estimates suggest.
5 Welfare and Reforms: Evidence
127
However, stabilizing money growth exerts a more robust effect on welfare. Table 5.5 indicates just that. Making inflation stable and more predictable played a positive and highly significant role in raising living standards. This is valid for both level and difference estimates. This is valid
also for the long-term growth of consumption and income, although the
effects are smaller for the growth of consumption than income. The fact
that both the level and the difference estimations are positive and statistically significant sends a clear policy message. Namely, not only are
countries with more stable money richer, but stabilizing money is also a
robust tool to increase living standards and consumption.
Table 5.6 presents the effects from more stable monetary policies on life
expectancy and income inequality. As with income and consumption, the
effect on life expectancy is evident in both the levels and the difference estimations. It is also positive and statistically significant. Perhaps, the effects
work its way through the positive increase in income and consumption
after monetary stabilization. Even so, monetary stabilization does increase
life expectancy together with income and consumption.
A positive effect can also be observed on income inequality. As the fixedeffects difference estimates suggest, stabilizing money has the potential to
lower income inequality. The effect is statistically significant only in one
of the three difference models, however. Ultimately, this may mean the
alleged effect from stabilizing money to reducing income inequality is not
robust to various modeling choices. Still, despite being insignificant, the
difference estimates are always negative, which is the expected sign of the
relationship.
Table 5.7 demonstrates that trade liberalization also contributed to the
growth of income per capita, as the bulk of existing evidence suggests (e.g.,
Ben-David (1996); Krishna and Mitra (1998), among others). This is seen
particularly from the difference estimates. Both the panel OLS and the
fixed-effects estimates suggest trade liberalization affected income growth
positively and significantly. That is not surprising, given the plethora of
studies delivering empirical support to that hypothesis, and also given
trade theory.
However, a surprising fact about the effects of trade reforms is that, even
when they raise income per capita, they do not deliver a complementary
increase in consumption growth, as seen from the difference estimates.Two
128
P. Stankov
explanations are possible. First, opening up trade could potentially raise
income growth, but that income growth is unevenly distributed between
capital and labor owners. If most of the gains are channeled to capital
owners, the effect would show up as a negligible increase in consumption
due to the larger mass of labor owners, who do not gain much from the
boost in trade. Second, even when trade spurs growth, it is either possible
that the bulk of this growth is channeled to savings and therefore used to
accelerate investment rather than consumption, or that different groups
of consumers benefit from trade differently, as in Janeba (2007).
More freedom to trade is also associated with longer life expectancy.
Countries with free trade regimes were also those with more prevalent
longevity. At the same time, the estimates here are derived after controlling for income, human capital, population, and country- and time-fixed
effects. As a result, they are more credible than the bivariate correlations
alone. However, trade liberalization did not lead to a significant increase
in life expectancy, as the estimates in Table 5.8 suggest.
The positive and significant correlation in levels but not in differences
may be due to purely demographic factors. More developed countries are
typically more open to trade but at the same time have a higher share of
elderly people. This would show up as a positive correlation between trade
freedom and longevity but will not necessarily produce a significant effect
from trade liberalization to increasing life expectancy.
Table 5.8 also suggests that free trade has been associated with an
increase in income inequality. This is intuitive, as trade carries significant
distributional consequences within countries, especially across workers
with varying skills (Carneiro and Arbache 2003; Zhu and Trefler 2005),
firms with different technologies (Egger and Kreickemeier 2012; Harrison
et al. 2011), or firms with varying size and export participation (Helpman
et al. 2017). However, the fact that the difference estimates are insignificant means that opening up trade is not a leading factor behind the rise
in income inequality. Recent evidence from the US states even suggests
that more economic freedom is generally associated with lower income
inequality (Wiseman 2017).
The estimates in Table 5.9 demonstrate that not only are countries
with less burdensome regulations richer, as in Djankov et al. (2002) and
subsequent studies reviewed by Djankov (2009), but also that deregulation
5 Welfare and Reforms: Evidence
129
has a positive and significant effect on the growth of income per capita.
The latter fact is seen from the difference estimations in Table 5.9. The
same table demonstrates that the effect on consumption per capita growth
was less evident. In only one of the six models studied was the level of
regulation associated with a higher consumption level. The rest of the
estimates were statistically insignificant.
The level of regulations is also associated with life expectancy and
income inequality. This is seen from Table 5.10. Both the panel OLS
and the fixed-effects models in the level equations suggest that people in
societies with less burdensome regulations live longer, but are also less
equal. It also appears that deregulation worsened income inequality significantly, as seen from the difference estimations without fixed effects,
which is consistent with recent evidence of Pérez-Moreno and AnguloGuerrero (2016). However, including fixed effects in the model rendered
the estimates insignificant. This means that the effect of deregulation on
income inequality is not robust, unlike the above recent evidence.
As we can see from the estimates above, some reforms are significant in
levels only. This means people in countries with free economies are typically richer, consume more, live longer, and are less equal than countries
which suppress economic freedom. However, we can also note that this
positive association between boosting economic freedom and increasing
welfare is not a recipe for how to raise welfare, because most economic
freedom reforms did not produce a robust positive effect on welfare. This
is expected, given the results of much earlier studies by Ayal and Karras
(1998) and Heckelman and Stroup (2000), among others.
For example, making governments smaller did not produce a statistically significant effect on either of the welfare proxies. Improving property
rights, stabilizing monetary policies, and liberalizing trade made countries
richer. Improving property rights and monetary stability also allowed average citizens to enjoy higher levels of consumption. In addition, stable
monetary policies also led to societies living longer and enjoying lower
income inequality. Deregulation also made societies richer, but at the cost
of raising income inequality, which is not consistent with earlier evidence
of Berggren (1999) and Scully (2002). Finally, the 2SLS estimates rarely
were significant. This implies that the causal effect of the economic freedom reforms on welfare is still in question, despite previous results of Faria
and Montesinos (2009) and Justesen (2008).
130
P. Stankov
References
Açemoglu, D., S. Johnson, and J.A. Robinson. 2005. Chapter 6, Institutions as a
fundamental cause of long-run growth. In Handbook of Economic Growth, ed.
P. Aghion and S. N. Durlauf, vol. 1, Part A of Handbook of Economic Growth,
385–472. Elsevier.
Afonso, A., and D. Furceri. 2010. Government size, composition, volatility and
economic growth. European Journal of Political Economy 26 (4): 517–532.
Albanesi, S. 2007. Inflation and inequality. Journal of Monetary Economics 54
(4): 1088–1114.
Apergis, N., O. Dincer, and J.E. Payne. 2014. Economic freedom and income
inequality revisited: Evidence from a panel error correction model. Contemporary Economic Policy 32 (1): 67–75.
Ashby, N.J., and R.S. Sobel. 2008. Income inequality and economic freedom in
the US states. Public Choice 134 (3–4): 329–346.
Ayal, E., and G. Karras. 1998. Components of economic freedom and growth:
An empirical study. Journal of Developing Areas 32 (3): 327–338.
Beck, T., and L. Laeven. 2006. Institution building and growth in transition
economies. Journal of Economic Growth 11: 157–186.
Becker, G.S., T.J. Philipson, and R.R. Soares. 2005. The quantity and quality of
life and the evolution of world inequality. American Economic Review 95 (1):
277–291.
Ben-David, D. 1996. Trade and convergence among countries. Journal of International Economics 40 (3–4): 279–298.
Berggren, N. 1999. Economic freedom and equality: Friends or foes? Public
Choice 100 (3–4): 203–223.
Bergh, A., and M. Karlsson. 2010. Government size and growth: Accounting for
economic freedom and globalization. Public Choice 142 (1–2): 195–213.
Bjørnskov, C., and N.J. Foss. 2008. Economic freedom and entrepreneurial activity: Some cross-country evidence. Public Choice 134 (3–4): 307–328.
Carlsson, F., and S. Lundstrom. 2002. Economic freedom and growth: Decomposing the effects. Public Choice 112 (3–4): 335–344.
Carneiro, F., and J.S. Arbache. 2003. The impacts of trade on the Brazilian labor
market: A CGE model approach. World Development 31 (9): 1581–1595.
Carter, J.R. 2007. An empirical note on economic freedom and income inequality. Public Choice 130 (1–2): 163–177.
Chu, A.C., G. Cozzi, and S. Galli. 2014. Stage-dependent intellectual property
rights. Journal of Development Economics 106: 239–249.
5 Welfare and Reforms: Evidence
131
Compton, R.A., D.C. Giedeman, and G.A. Hoover. 2014. A distributional analysis of the benefits of economic freedom. European Journal of Political Economy
33: 121–133.
Dar, A.A., and S. Amirkhalkhali. 2002. Government size, factor accumulation,
and economic growth: Evidence from OECD countries. Journal of Policy Modeling 24 (7–8): 679–692.
Davies, A. 2009. Human development and the optimal size of government. The
Journal of Socio-Economics 38 (2): 326–330.
Dinopoulos, E., and B. Unel. 2011. Quality heterogeneity and global economic
growth. European Economic Review 55 (5): 595–612.
Djankov, S. 2009. The regulation of entry: A survey.World Bank Research Observer
24 (2): 183–203.
Djankov, S., R. La Porta, F. Lopez-de-Silanes, and A. Shleifer. 2002. The regulation of entry. The Quarterly Journal of Economics 117 (1): 1–37.
Doucouliagos, C. 2005. Publication bias in the economic freedom and economic
growth literature. Journal of Economic Surveys 19 (3): 367–387.
Dow, G., and C. Reed. 2013. The origins of inequality: Insiders, outsiders, elites,
and commoners. Journal of Political Economy 121 (3): 609–641.
Egger, H., and U. Kreickemeier. 2012. Fairness, trade, and inequality. Journal of
International Economics 86 (2): 184–196.
Farhadi, M., M.R. Islam, and S. Moslehi. 2015. Economic freedom and productivity growth in resource-rich economies. World Development 72: 109–126.
Faria, H.J., and H.M. Montesinos. 2009. Does economic freedom cause prosperity? An iv approach. Public Choice 141 (1–2): 103–127.
Fölster, S., and M. Henrekson. 2001. Growth effects of government expenditure
and taxation in rich countries. European Economic Review 45 (8): 1501–1520.
Ghossoub, E., and R. Reed. 2017. Financial development, income inequality, and
the redistributive effects of monetary policy. Journal of Development Economics
126: 167–189.
Goldsmith, A.A. 1995. Democracy, property rights and economic growth. Journal of Development Studies 32 (2): 157–174.
Gwartney, J. 2009. Institutions, economic freedom, and cross-country differences
in performance. Southern Economic Journal 75 (4): 937–956.
Gwartney, J., J. Hall, and R. Lawson. 2016. 2016 Economic Freedom Dataset.
Fraser Institute.
Gwartney, J., R. Holcombe, and R. Lawson. 2006. Institutions and the impact
of investment on growth. Kyklos 59 (2): 255–273.
Harrison, A., J. McLaren, and M. McMillan. 2011. Recent perspectives on trade
and inequality. Annual Review of Economics 3: 261–289.
132
P. Stankov
Heckelman, J., and M. Stroup. 2000. Which economic freedoms contribute to
growth? Kyklos 53 (4): 527–544.
Helpman, E., O. Itskhoki, M.-A. Muendler, and S. Redding. 2017. Trade and
inequality: From theory to estimation. Review of Economic Studies 84 (1): 357–
405.
Janeba, E. 2007. International trade and consumption network externalities.
European Economic Review 51 (4): 781–803.
Justesen, M.K. 2008. The effect of economic freedom on growth revisited: New
evidence on causality from a panel of countries 1970–1999. European Journal
of Political Economy 24 (3): 642–660.
Kacprzyk, A. 2016. Economic freedom-growth nexus in European Union countries. Applied Economics Letters 23 (7): 494–497.
Kemper, N., L.V. Ha, and R. Klump. 2015. Property rights and consumption
volatility: Evidence from a land reform in vietnam. World Development 71 (SI),
107–130.
Krieger,T., and D. Meierrieks. 2016. Political capitalism:The interaction between
income in\equality, economic freedom and democracy. European Journal of
Political Economy 45: 115–132.
Krishna, P., and D. Mitra. 1998. Trade liberalization, market discipline and productivity growth: New evidence from India. Journal of Development Economics
56 (2): 447–462.
Lee, J.-W. 1996. Government interventions and productivity growth. Journal of
Economic Growth 1: 391–414.
Levine, R. 2005. Law, endowments and property rights. The Journal of Economic
Perspectives 19 (3): 61–88.
Mackenbach, J.P. 2013. Convergence and divergence of life expectancy in Europe:
A centennial view. European Journal of Epidemiology 28 (3): 229–240.
Mehlum, H., K. Moene, and R. Torvik. 2005. Crime induced poverty traps.
Journal of Development Economics 77 (2): 325–340.
Meltzer, A.H., and S.F. Richard. 1981. A rational theory of the size of government.
Journal of Political Economy 89 (5): 914–927.
Mijiyawa, A.G. 2008. Sustained economic growth: Do institutions matter, and
which one prevails? Cato Journal 28 (Fall) (3): 385–420.
Milanovic, B.L. 2014. All the Ginis, 1950–2012 (Updated in Autumn 2014).
Mudambi, R., P. Navarra, and C. Paul. 2002. Institutions and market reform
in emerging economies: A rent seeking perspective. Public Choice 112 (1–2):
185–202.
Mulligan, C.B., and K.K. Tsui. 2008. Political entry, public policies, and the
economy. Working Paper 13830, National Bureau of Economic Research.
5 Welfare and Reforms: Evidence
133
Murphy, K.M., A. Shleifer, and R.W. Vishny. 1991. The allocation of talent:
Implications for growth. The Quarterly Journal of Economics 106 (2): 503–30.
Neumayer, E. 2004. HIV/AIDS and cross-national convergence in life
expectancy. Population and Development Review 30 (4): 727–742.
Nyström, K. 2008. The institutions of economic freedom and entrepreneurship:
Evidence from panel data. Public Choice 136 (3–4): 269–282.
Oeppen, J. 2006. Life expectancy convergence among nations since 1820:
Separating the effects of technology and income. In Perspectives on mortality forecasting—III. The linear rise in life expectancy: history and prospects, ed.
T. Bengtsson, 55–82. Swedish Social Insurance Agency.
Pecoraro, B. 2017. Why don’t voters ‘put the Gini back in the bottle’? Inequality
and economic preferences for redistribution. European Economic Review 93:
152–172.
Pérez-Moreno, S., and M. Angulo-Guerrero. 2016. Does economic freedom
increase income inequality? Evidence from the EU countries. Journal of Economic Policy Reform 19 (4): 327–347.
Scully, G. 2002. Economic freedom, government policy and the trade-off between
equity and economic growth. Public Choice 113 (1–2): 77–96.
Stroup, M.D. 2007. Economic freedom, democracy, and the quality of life. World
Development 35 (1): 52–66.
Sturm, J., and J. De Haan. 2001. How robust is the relationship between economic freedom and economic growth? Applied Economics 33 (7): 839–844.
Torstensson, J. 1994. Property-rights and economic-growth–an empirical-study.
Kyklos 47 (2): 231–247.
Trebilcock, M., and P.-E. Veel. 2008. Property rights and development: The contingent case for formalization. University of Pennsylvania Journal of International
Law 30 (2): 397–481.
Tsui, K.K. 2011. More oil, less democracy: Evidence from worldwide crude oil
discoveries. The Economic Journal 121 (551): 89–115.
Wiseman, T. 2017. Economic freedom and growth in US state-level market
incomes at the top and bottom. Contemporary Economic Policy 35 (1): 93–112.
Zhu, S., and D. Trefler. 2005. Trade and inequality in developing countries: A
general equilibrium analysis. Journal of International Economics 65 (1): 21–48.
6
Crises, Welfare, and Populism
6.1
Defining Populism
Every crisis creates winners and losers. However, only large-scale crises
create losers on a mass scale. If the losers are not adequately compensated
soon after the crisis, they will vote for politicians offering them an easy
solution, a redemption for their ills. The more people demand redemption, the more politicians will be keen to offer it. To convince the voters to
buy into their offer, politicians need to embed their agendas within controversial political discourse: populism. This chapter looks into the roots of
populism, its driving factors, and its economic consequences both before
and after the Great Recession.
In political science, populism is defined as a “specific political communication style,” a discourse, which tries to be close to the people but at
the same time takes an anti-establishment stance and excludes “specific
population segments” from an image of an ideal society (Jagers and Walgrave 2007, p. 475). Typically, populist parties are on the extreme right
or left. Also, in political science, populism is described as a set of ideas
which lead to strategies, policies, and institutions appealing to the majority
© The Author(s) 2017
P. Stankov, Economic Freedom and Welfare Before and After the Crisis,
DOI 10.1007/978-3-319-62497-6_6
135
136
P. Stankov
of voters. Typically, the representatives of those voters are fighting against
the elite establishment (Hawkins 2009), emphasizing the us-against-them
rhetoric.
The economic approach to populism is more technical. Economists
define populism as a specific set of policy priorities: “Macroeconomic
populism is an approach to economics that emphasizes growth and income
distribution and deemphasizes the risks of inflation and deficit finance,
external constraints and the reaction of economic agents to aggressive nonmarket policies.” (Dornbusch and Edwards 1990, p. 247). Dornbusch and
Edwards (1990) present arguments that populism goes through similar
phases across various economies and that it ultimately leads to welfare
deterioration for most of the voters favoring populist policy priorities.
6.2
The Political Economy of Populism
The recent rise of populism after the Crisis in Europe, the UK, and the
USA is probably unprecedented, at least since World War II. Several factors may be able to explain it. First, the severity of the Great Recession
created a large mass of workers who lost their jobs who saw significant
stagnation in their incomes and consumption. Second, normally any government would counteract those unemployment and consumption trends
by pumping government expenditures to compensate for the lack of private
spending. However, the fiscal stance after the Great Recession in Europe,
the UK, and the USA has been different than in previous recessions. Unlike
before, governments now need to curb government expenditures exactly
when voters need them most because of already high existing levels of
government debt. Those fiscal restrictions are now popularly known as
austerity. Austerity has been an additional factor fueling a sense of discontent among voters. Third, the rise of income inequality over the last few
decades has undermined the sense of fairness, especially in Europe and the
USA, which feeds well into the us-against-them rhetoric. Finally, the deep
recession, the persistent unemployment, and the long-standing austerity
coincide with large migration flows which, in the public eye, worsen the
job prospects of the incumbent workers. In turn, this further undermines
the support for mainstream political parties and shifts it to populists.
6 Crises, Welfare, and Populism
137
Thus, crises in general can affect the likelihood of populist insurgence
(Moffitt 2015). Dornbusch and Edwards (1991) also note that the depth
of recessions and the ensuing income inequality have been at the root
of the populist rise in Latin America. Research on Asia has also shown
a significant effect of crises on the likelihood of populism. For example,
Tejapira (2002) and Hewison (2005) review the rise of economic nationalism in Thailand as a result of economic stagnation which followed the
1997 East-Asian crisis. As the post-crisis reforms created a vast number
of losers, economic freedom reforms were rejected in the 2001 elections.
A far more nationalistic political agenda stepped in, fitting well into the
“populist paradigm” offered by Dornbusch and Edwards (1990).
As the number of losers during or after a recession increases, inequality
surges. New data demonstrates that inequality has been on the rise irrespective of the business cycle. Alvaredo et al. (2017) monitor inequality trends
specifically in the USA, UK, France, and China. They “observe rising top
income and wealth shares in nearly all countries in recent decades” (p. 1).
This suggests country-specific factors, e.g., policies and institutions, may
explain the rise. (Dornbusch and Edwards 1991, p. 1) agree that “populist
regimes have historically tried to deal with income inequality problems
through the use of overly expansive macroeconomic policies.”
Macroeconomic mismanagement by populists, however, leads to recessions, and to banking and fiscal crises, which can result in hyperinflation.
In turn, this worsens inequality rather than remedying it. Based on Latin
American evidence of hyperinflation, Bittencourt (2010) calls for central
bank independence and a committed fiscal authority in order to raise the
costs of implementing populist agendas. Further, Greskovits (1993) argues
that, in order to make populism less likely, all reform packages need to
contain an adequate compensation mechanism for the reform losers. If
the group of losers is large enough and the compensation is inadequate
and untimely, mounting social discontent will inevitably meet with the
supply of populist agendas.
Persistent inequality, combined with stagnant growth or an outright
economic depression, stands at the heart of voter discontent, which motivates the supply of populist agendas, according to Dornbusch and Edwards
(1990) and Kaufman and Stallings (1991). To arrive at this conclusion,
both teams of researchers review an array of populist episodes in Latin
138
P. Stankov
America before 1990. Kaufman and Stallings (1991) also predict that
populism would become a more isolated political phenomenon, a conclusion which definitely requires revision almost 10 years after the Great
Recession. In fact, even before the Great Recession, it was apparent that
populism is coming back into prominence in Latin America as a result of
persistent income inequality and stagnant growth, even in the presence of
market-oriented reforms (Roberts 2007). Leon (2014) also argues that the
use of macroeconomic redistribution to alleviate income inequality may
make populist agendas more likely, and Dornbusch and Edwards (1990)
add that in fact, large-scale redistribution proposals are a persistent feature
of populism in Latin America.
Populism has plagued politics in Latin America for the better part of
the last half century. However, it has also emerged in other regions of the
world. For example, some parts of Europe are already embracing populist
agendas. Unlike in other regions, the European brand of populism has a
distinctive trait: xenophobia. This trait is not new to Europe. The devastating consequences of World War I, combined with the perils of the
Great Depression, set the stage for an infamous populist and xenophobe
in Germany: Adolf Hitler. For a number of years now, and even before
the Great Recession, various authors have studied the nascent comeback
of populism to the European political scene. According to Jones (2007,
p. 37), populists “are making headway across Europe and from all points
on the political spectrum,” and a distinctive trait of this comeback is
its “xenophobic, anti-immigrant rhetoric” (p. 37) which, according to
the author, may be very hard to restrain. The reasons for this rhetoric
are outlined very well by Cahill (2007). He asserts that the immigration
waves from “North Africa and Eastern Europe, fear of economic dislocations under European Union enlargement, and the struggles to integrate
Muslim immigrants have breathed new life into anti-immigration platforms” (p. 79). Učeň (2007) adds that populism in Europe is inherently
anti-establishment. This in turn may appeal to the large masses of people experiencing discontent from the consequences of austerity, persistent
unemployment, and stagnant growth in Europe in the aftermath of the
Great Recession.
Ten years after their publication, these studies sound even more relevant to contemporary European politics, especially because the fear of
6 Crises, Welfare, and Populism
139
immigration is being reinforced by the current refugee crisis, and Europe
still struggles to integrate Muslim immigrants, as it always has. More
importantly from a contemporary standpoint, anti-establishment and
anti-immigration rhetoric has been increasingly moving from the political periphery to the political mainstream. Brexit succeeded mainly due to
the mounting anti-immigrant rhetoric, and the political agenda Theresa
May has adopted seems to follow it up. The anti-Islam Freedom Party of
Geert Wilders in the Netherlands was an inch away from being pivotal
in the March 2017 elections, Marine Le Pen gained the largest populist
support ever in France’s May 2017 elections, far-right parties in Germany
have been gaining traction due to their own extreme rhetoric, and farright parties already occupy government seats in other smaller countries
in 2017, e.g., Bulgaria.
The above literature suggests five major factors for the rise of populism
across the globe: recessions, unemployment, inflation, austerity, and immigration. The analysis below tests if any of those has played a statistically
significant role for the likelihood of populist resurgence. However, first,
we need to find a way to measure populism.
6.3
Measuring Populism
Despite the elusiveness of the populism concept, recent efforts have generated three data sets that can be used to understand its causes and consequences. Rode and Revuelta (2015) is the first. By using and updating the
original index of Hawkins (2009), they design a nonpartisan measure of
populism and study its effect on economic freedom in 33 countries. The
advantage of their data set is that it tracks populism in both developed and
developing countries across the globe. Its within-country time variation is
small, and yet it allows study of the political economy of populism across
countries over time. The index contains 252 observations, of which 55
are after 2007. It is important to note that the Rode and Revuelta (2015)
data is generated by studying political leaders’ speeches. Therefore, this
data presents populism as a rhetorical style.
The second data set is especially designed with Europe in mind by Heinö
(2016). Heinö (2016) monitors the rise in the European authoritarian
populism since 1980 from a different angle than that of Rode and Revuelta
140
P. Stankov
(2015). He monitors actual national election outcomes in 33 European
countries: the 28 EU members and Iceland, Montenegro, Norway, Serbia,
and Switzerland. Then, he sums up the political support for both rightwing and left-wing populist parties to arrive at an annual index of populism
in Europe, including the period after the Crisis. Actually, the data allows
for studying overall populism dynamics, as well as the underlying trends in
right-wing and left-wing populist support.The report portrays a dark trend
in the European political landscape, namely that support for populism
has been rapidly gaining ground since the Crisis. (Heinö 2016, p. 17)
maintains that “the increase is driven mainly by the exceptional successes
for left-wing populist parties in Greece, Italy and Spain, but left-wing
radicals have also been successful in countries such as Denmark, Belgium,
Ireland, Romania and Croatia.” He concludes that, based on the observed
surge in political support for populism, we can safely presume that this
political style and economic strategy are going to persist in Europe for the
foreseeable future.
The Rode and Revuelta (2015) rhetoric populism data varies little over
time, and the Heinö (2016) election outcomes data covers only Europe.
As a result, we still lack a global perspective on the rise of populism from
actual election outcomes data. This calls for an additional empirical effort
directed at both increasing time coverage, especially with more observations after the Crisis, and expanding the geographic coverage beyond
Europe. This is needed because populism is now moving from the local
political fringe to global prominence, even in democracies and in welldeveloped countries. Therefore, to adequately understand the political
economy of populism, especially after the Great Recession, we need a
global data set. To this date, it is not yet readily available. However, by
following guidance on what populism is, it is possible to construct it from
the Database of Political Institutions (DPI) initially produced by Beck
et al. (2001) and recently updated by Cruz et al. (2016a). That is exactly
what I do in this part of the book.
To produce global data on populism, initially I use the notion that
populism is nationalistic in its rhetorical style. The DPI data has a way
to measure if a ruling party or an elected chief executive is nationalistic.
The DPI records a party as nationalistic if the primary component of
its platform is “the creation or defense of a national or ethnic identity,”
6 Crises, Welfare, and Populism
141
e.g., calls for persecution of minorities or is xenophobic, or if “the party is
listed as nationalistic” in primary data bases of political orientation (Cruz
et al. 2016b, p. 9). Then, I notice that the varieties of populism are located
either on the left or right end of the political spectrum, or some parties are
outright “progressive, authoritarian or xenophobic” (Cruz et al. 2016b,
p. 8), which prevents them from being classified as left or right. Thus,
a ruling party or a chief executive who is nationalistic would also be a
populist if they are left wing or right wing (but not centrist), or if they are
thought to be progressive, authoritarian, or xenophobic.
Thus, three global binary measures of populism can be produced: (1)
chief executive populism (CEP); (2) incumbent government populism
(IGP); and (3) both CEP and IGP (BOTH). Out of the total 7303 observations in the DPI2015, the observations which can be classified as CEP
are 571 (7.8%), the IGPs are 553 (7.6%), and a simultaneous occurrence
of CEP and IGP are observed in a total of 466 cases (6.4%). Those binary
populist outcomes can then be related to recessions, unemployment, inflationary episodes, austerity, and income inequality.
For the empirical estimations that follow, I use all the three data sets
presented so far: the Rode and Revuelta (2015) data, the Heinö (2016)
data, and the newly constructed binary measures of populism. I link those
three data sets with the above explanatory factors. Data on GDP dynamics,
unemployment, inflation, and the share of government expenditures in
GDP are taken from the World Development Indicators produced by
The World Bank (2017). The data on inequality dynamics is taken from
Milanovic (2014). Unfortunately, data on net migration as a share of the
labor force is too scarce to allow for its inclusion in the models. Further
efforts are still needed to increase the availability of net migration data on
a global level, so that it can be meaningfully related to the rise of populism.
As the Rode and Revuelta (2015) data is continuous and runs from
0 to 1.9, while my own data on populism is a binary 0–1 variable, it makes
sense to normalize the former so that it spans from 0 to 1. The normalized
rhetoric populism data would allow comparison of the estimation results
across the two data sets. In addition, it would allow application of similar
estimation methods on both data sets. Details on the estimation methods
follow.
142
6.4
P. Stankov
Model
To estimate the impact of the drivers of populism on the probability of
populists assuming power, I estimate the following population-averaged
(PA) probit model:
Pr (P O Pit = 1|X it ) = F(X it β),
(6.1)
where Pr (P O Pit = 1|X it ) is the probability Pr of observing populism
(POP) of a certain type in country i in year t. The probability depends
on a number of factors discussed above. In this case, they are as follows:
Log(GDP/c.), unemployment (Unemp.), inflation (Infl.), the share of
government expenditures in GDP (G/GDP), and the Gini coefficient for
country i in year t. Per capita GDP is preferred in this case over the
underlying GDP for one important reason: it is voter preferences that
drive election outcomes, and the median voter sentiment will be much
better proxied by the per capita GDP than the total GDP.
Further, we would like to see if the Great Recession has produced any
meaningful difference in the way those factors work on the probability
of observing populism. In other words, we would like to know if the
probability of observing populism is somehow affected by the Crisis. To
answer that question, I interact an after-crisis dummy variable (AC) with
each of the core explanatory factors. The AC dummy is equal to 1 if the
year is greater than 2008, and equal to 0 otherwise. The AC dummy is
also included in the equation as an independent term.
The PA model produces an idea of how likely a populist election outcome is in an average country in the sample. However, we would also like
to know how likely it is that a populist will be elected in the same country, given the observed explanatory variables for that particular country.
We need another version of the probit model to be able to answer that
question, namely the random-effects (RA) probit model. Specifically, in
the RA model, I estimate the following equation:
Pr (P O Pit = 1|X it , u it ) = G(X it β + u it ),
(6.2)
6 Crises, Welfare, and Populism
143
where the explained and the explanatory variables are the same as before.
Table 6.1 presents the results from both the PA and the RA model
estimations.
Since populism data is available not only in the form of election outcomes but also in terms of rhetorical style, we can also notice how that
rhetorical style correlates with recessions, unemployment, inflation, austerity, and income inequality. That is exactly what I do with the Rode
and Revuelta (2015) data. I replace the above binary measures of populism with the normalized populism score (PS) from their data. Then, I
run fixed-effects panel OLS estimations, in which initially the explanatory factors are included one by one and then are included simultaneously.
Table 6.2 presents the results.
Finally, I focus on the rise of European authoritarian populism after
the Crisis. The Heinö (2016) data is suitable, as it allows for long-term
averaging of the populism trends in Europe. For each country in the sample, I average the overall populist support in three periods: 1992–1999,
2000–2007, and 2008–2015. I also do that separately for right-wing and
left-wing populist support. Then, I difference those long-term averages
to arrive at the change in voter attitudes toward populism between those
three periods. Thus, two data points are produced for most of the countries
in the sample, yielding a total of 63 observations. Next, I estimate how
important the concurrent changes in average living standards, inflation,
unemployment, long-term unemployment, share of government expenditures in GDP, and income inequality are for the rise of populism in Europe.
Those estimations are conducted using two sets of models: bivariate and
multivariate. The bivariate models estimate the correlations between the
changes in support for populism and the changes in the explanatory factors
separately, one by one. The multivariate models estimate the correlations
between the change in populism support and the changes in the explanatory factors, when all the factors are included in the models. As the Heinö
(2016) data monitors both left-wing and right-wing populism support,
as well as the overall support for populism, three dependent variables are
used: TAP, indicating the overall authoritarian populism; TAP-RW—for
the right-wing populist support; and TAP-LW—for the left-wing support.
The results from both bivariate and the multivariate models are presented
in Table 6.3. The details on the results follow.
144
6.5
P. Stankov
Results
The results in Table 6.1 are presented in two sets of three models each.
The first set of models presents the PA estimates, while the second set
presents the RE estimates. Within each set, three dependent variables
Table 6.1
Political economy of populism before and after the crisis
Population-averaged probit
(1)
(2)
(3)
CEP
IGP
BOTH
L(GDP/c.)
−0.140
(0.115)
Infl
0.000***
(0.000)
Unempl
0.018
(0.019)
G/GDP
0.002
(0.024)
Gini
0.003
(0.008)
After Crisis
0.680
(1.001)
L(GDP/c.)*AC −0.194**
(0.089)
Infl*AC
−0.130***
(0.048)
Unempl*AC −0.017
(0.032)
(G/GDP)*AC 0.044
(0.027)
Gini*AC
0.014
(0.015)
N
1255
C’ry FEs
No
−0.237**
(0.104)
0.000***
(0.000)
−0.001
(0.016)
0.024
(0.023)
0.012
(0.008)
0.491
(0.960)
−0.396**
(0.189)
−0.058
(0.037)
−0.088***
(0.033)
0.044**
(0.022)
0.050***
(0.017)
1234
No
−0.129
(0.116)
0.000***
(0.000)
0.017
(0.017)
−0.005
(0.025)
0.006
(0.008)
−0.455
(0.692)
−0.116
(0.084)
−0.133**
(0.059)
−0.119***
(0.028)
0.065***
(0.023)
0.032***
(0.012)
1213
No
Random-effects probit
(4)
(5)
(6)
CEP
IGP
BOTH
−0.434*
(0.230)
0.001
(0.001)
0.018
(0.036)
0.041
(0.044)
0.012
(0.021)
9.224
(9.828)
−2.108
(2.041)
−0.201
(0.299)
0.122
(0.166)
0.128
(0.250)
0.059
(0.100)
1255
No
−0.864***
(0.321)
0.000
(0.001)
−0.066*
(0.038)
0.101**
(0.045)
0.039*
(0.021)
4.700
(28.530)
−3.424
(8.970)
−0.514
(1.396)
−0.221
(0.786)
0.259
(0.750)
0.375
(0.985)
1234
No
−0.439*
(0.240)
0.000
(0.001)
−0.005
(0.038)
0.028
(0.045)
0.015
(0.021)
3.192
(23.566)
−2.912
(7.086)
−0.453
(1.166)
−0.220
(0.489)
0.283
(0.583)
0.324
(0.739)
1213
No
Notes The estimated Population-Averaged (PA) model is Pr (P O Pit = 1|X it ) =
F(X it β), where Pr (P O Pit ) = 1 is the probability of a populist being elected as
a chief executive (CEP), or a populist party gaining majority in the incumbent government (IGP), or both (BOTH). X it is a vector of explanatory variables detailed in
the text. Robust standard errors are presented in parentheses for the PA model.
Data source DPI2015, WDI, Milanovic (2014). Symbols * p < .10, ** p < .05, ***
p < .01
6 Crises, Welfare, and Populism
145
are used: the CEP, the IGP, and BOTH representing, respectively, the
likelihood of observing a chief executive populist, a populist party forming
the incumbent government, and both together.
The PA estimates demonstrate that declines in income per capita raise
the likelihood of populists assuming power, especially when it comes to
forming a government. An average country is more likely to give the
power to populist parties forming a government during or immediately
after recessions. However, this is not the case for choosing populist chief
executive or having both a populist chief executive and an incumbent
government. The RE estimates demonstrate that this effect is observable
not only in an average country, but also within a certain country over
time. It is far more likely to observe populists assuming power in a given
country after the same country has experienced a recession. The effect is
also present for all types of populism—CEP, IGP, and BOTH.
Inflation also plays a certain role in electing populists. The effect, however, is negligibly small (not being observable even at the third digit after
the decimal point), although it is statistically significant. Further, the effect
is observable only in the PA estimates. This implies that we can expect
countries with higher inflation to be ruled by populists more often, but
inflation is hardly the core driver of populist power. The results also show
that, historically, a short-term rise in unemployment has played a negligible role in helping populists assume power.
However, model (5) also calls for caution in that interpretation. It tells
us that populist governments are also more likely to coincide with lower
unemployment and higher government expenditure. If we revisit the “populist paradigm” literature, this is exactly what happens shortly before the
economy goes bust. Initially, an incumbent populist government delivers on its promise to put everyone to work and raise social expenditures.
After some time, it gradually becomes clear that keeping that promise is
increasingly difficult without ever-increasing government debt. Austerity
is politically unsellable so the government keeps borrowing. Sooner than
later, the government is broke, and international financial institutions step
in at a very high social price. As a result, the populist government made
the very people who elected it—those who needed jobs and more social
expenditures—worse-off in the long-term. This is because populists use
policy measures which are “diametrically at variance with the incentive
146
P. Stankov
structure required to move the economy into the desired follow-up phase
of investment and export growth” (Lago 1991, pp. 264–265).
The PA estimates after the Crisis are more informative of the support
for populism. The interaction terms of the explanatory variables with the
after-crisis dummy AC are largely significant. Specifically, recessions and
rising income inequality exert a more important impact on the likelihood
of populism after the Crisis than before. This means that voters after the
Great Recession are more sensitive to income drops and rising inequality
than before it. At the same time, periods of higher inflation and unemployment, and lower government expenditures in the after-crisis period are,
on average, associated with lower likelihood of observing populism than
before the Crisis. This seems puzzling, as we would expect the opposite
sign of the relationship.
Again, I would like to raise caution in interpreting these results as
a causal relationship, mainly because all the effects noted here and in
Table 6.2 are contemporaneous. Thus, the signs and significance could
not only capture the effects of inflation, unemployment, and austerity
on populism, but also the fact that populist governments may be able
to temporarily deliver on their promises to lower inequality and unemployment through higher government expenditures. Another reason to be
cautious in interpreting the results is the RA estimates. They demonstrate
that, effectively, there is no significant difference in how macroeconomic
shocks affect the likelihood of observing populism before and after the
Crisis within the same country over a short period of time. However, they
also demonstrate that the effects could be elsewhere: e.g., within the same
country over a longer time period.
Table 6.2 presents similar estimates of the contemporaneous correlations between macroeconomic shocks, income inequality shocks, and populism as a political discourse. This is possible due to the availability of the
Rode and Revuelta (2015) data. The estimates show that populist rhetoric
behaves somewhat differently than the originally constructed populism
data. This is seen from both the bivariate models (1) through (5), and
from the last model (6) which presents the multivariate estimates.
The results show that recessions rarely have a significant immediate
impact on populist rhetoric, and that effect is not significantly different
before and after the Great Recession. This result is different than the one
6 Crises, Welfare, and Populism
Table 6.2
Populism as a rhetorical style before and after the crisis
(1)
NPS
L(GDP/c.)
L(GDP/c.)*AC
(2)
NPS
(3)
NPS
(4)
NPS
(6)
NPS
−0.007**
(0.004)
0.000
(0.001)
−0.024
(0.027)
186
Yes
−0.008
(0.155)
−0.021
(0.016)
0.000
(0.000)
−0.002
(0.002)
0.006
(0.006)
−0.006*
(0.004)
−0.010
(0.009)
0.001
(0.005)
−0.005
(0.004)
0.000
(0.002)
0.254
(0.211)
185
Yes
0.000***
(0.000)
−0.001
(0.001)
Infl*AC
Unempl
0.003*
(0.002)
−0.009
(0.007)
Unempl*AC
G/GDP
−0.014*
(0.008)
0.000
(0.002)
(G/GDP)*AC
Gini
Gini*AC
N
C’ry FEs
(5)
NPS
−0.118
(0.093)
−0.039
(0.028)
Infl
After Crisis
147
0.394
(0.276)
252
Yes
0.031
(0.032)
252
Yes
0.114
(0.091)
246
Yes
0.033
(0.052)
252
Yes
Notes The estimated equation is P Sit = X it β + f i + u it , where P Sit is the Populism
Score (PS) from Rode and Revuelta (2015), normalized to 1, NPS. X it is a vector of
explanatory variables detailed in the text. Robust standard errors are presented
in parentheses. All models include country fixed effects. Data source Rode and
Revuelta (2015), WDI, Milanovic (2014). Symbols * p < .10, ** p < 0.05, *** p < 0.01
we observed before, when recessions played the most consistent role in
populists gaining power.
Similarly to the previous results, inflation plays a statistically significant
but politically negligible role in channeling voter discontent to populist
parties. This is perhaps because deflation is now a greater concern than
high inflation, including in Latin American countries. Further, a spike in
unemployment and a negative shock on government expenditures can also
be associated with more visible populist rhetoric. This effect is seen from
data running over more than two decades, but is not visibly different before
148
P. Stankov
and after the Crisis. Interestingly, the results are exactly opposite to the ones
we observed in Table 6.1, indicating the sometimes dramatic difference
between rhetoric before elections and actual policies after. However, all
the estimates here are rather short term, similarly to the ones presented in
Table 6.1.
The multivariate regression in model (6) in Table 6.2 also suggests that
macroeconomic shocks in general, and income inequality shocks, rarely
have a short-term effect on populist rhetoric. In addition, the effects are
not significantly different before and after the Crisis. This is seen from the
insignificant interaction terms between the macroeconomic variables and
the dummy variable indicating the period after the Great Recession. The
exception is unemployment, which plays a slightly less pronounced role
in sparking populist rhetoric than before the Crisis.
The results on populist rhetoric are possibly different than those from
actual election outcomes for one more reason: the inclusion of country
fixed effects. As suggested by the literature, unobservable local factors may
be the key drivers of populism rather than the intuitive macroeconomic
and inequality dynamics. They could also be different because policies differ from political rhetoric. The overarching message, however, is that when
all factors are taken together, populism seems insusceptible to explanation
by short-term phenomena, even if they are intuitive. This is also valid
irrespective of how one defines populism: as nationalists or xenophobes
occupying the left- and the right-end of the political spectrum assuming
power, or as a political discourse.
The results above demonstrate the limitations of a static approach to
studying populism. To make the conclusions more convincing, we need
global data on actual election outcomes of populist parties, no matter
if they have been elected or not. Thus, we will observe how the factors
above actually drive changes in populist support even at its nascent stage.
Although it is much needed, such data is existent only for Europe (Heinö
2016). Its extension to a global data set would underpin a dynamic understanding of populism on a global level.
Then, we might need both a longer-term view of populism and its
drivers, and a refinement in how we define populism in the first place.
In the forthcoming analysis based on the results in Table 6.3, populism
is understood as actual electoral support for nationalistic far-right and
6 Crises, Welfare, and Populism
Table 6.3
Authoritarian populism and crises
Bivariate estimates
(1)
(2)
(3)
TAP
TAP-RW TAP-LW
L(GDP/c.)
Infl.
Unempl.
LT-Unempl.
G/GDP
Gini
C’ry FEs
N
adj. R 2
149
−19.100**
(8.976)
0.001
(0.035)
0.567**
(0.210)
0.154*
(0.086)
0.398
(0.482)
0.642**
(0.291)
Yes
−9.056*
(5.327)
−0.022
(0.051)
0.224
(0.170)
0.038
(0.079)
0.582
(0.571)
0.570
(0.367)
Yes
−10.061*
(5.546)
0.023
(0.017)
0.343***
(0.116)
0.116**
(0.050)
−0.181
(0.257)
0.074
(0.153)
Yes
Multivariate estimates
(4)
(5)
(6)
TAP
TAP-RW TAP-LW
−27.142
(16.819)
−0.347
(0.223)
0.569*
(0.310)
−0.336**
(0.159)
−0.443
(0.700)
0.579
(0.475)
Yes
53
0.335
−18.458*
(1.397)
−0.481*
(0.239)
0.271
(0.283)
−0.295**
(0.111)
0.163
(0.522)
0.313
(0.450)
Yes
53
0.197
−8.724
(1.756)
0.134
(0.110)
0.298
(0.226)
−0.041
(0.120)
−0.605
(0.425)
0.268
(0.195)
Yes
53
0.400
Notes The estimated fixed-effects panel OLS equation is P Sit = X it β + u it ,
where P Sit is either the overall Timbro Authoritarian Populism index score (TAP),
or the right-wing TAP (TAP-RW), or the left-wing TAP (TAP-LW). The index has been
produced by Heinö (2016). For the bivariate models in columns (1), (2), and (3), X it
is either of the explanatory variables detailed in the text. All those variables have
been included simultaneously in the multivariate models (4), (5), and (6). Constants
are included, but their estimates are not reported. The bivariate models capture
different number of observations, typically 63. The bivariate models yield different
Adj. R-squared coefficients, ranging between −0.016 and 0.227. Robust standard
errors are presented in parentheses. All models include country fixed effects. Data
source Heinö (2016), WDI, Milanovic (2014). Symbols * p < 0.10, ** p < 0.05, ***
p < 0.01
far-left parties irrespective of whether they have been elected. This is a
crucial advantage of the Heinö (2016) data over the previous two data
sets: it allows for a study of populism even before it enters the political
mainstream. Further, it allows for longer-term estimations of how the
changes in attitudes toward populism are associated with the changes in
macroeconomic and social conditions. The results of those estimations
are presented in Table 6.3.
The approach in Table 6.3 may be different, but the lessons are similar.
The deeper the recession, the more likely it is that populists will gain
150
P. Stankov
political support, and possibly assume power in a few electoral cycles. This
is especially valid for right-wing populism, which is significantly affected
by recessions in both the bivariate and the multivariate estimations. Leftwing populism is also affected by recessions in the bivariate model.
Populism seems unaffected by the long-term inflation trends. Most
of the estimates are insignificant, with the exception of the multivariate
model for right-wing populism. In this model, higher average inflation
gives populists less steam. Again, this might be due to deflation occurring
in the after-crisis period, which in turn is typically associated with deeper
recessions like the one after 2008.
Bivariate estimates also demonstrate that left-wing populism is positively affected by the current and the long-term unemployment. This is a
key difference between left-wing and right-wing populism. While recessions play the major role in spurring right-wing populism, the left-wing
is also driven by unemployment patterns.
Surprisingly, austerity and income inequality rarely play a statistically
significant role in shaping populist voter attitudes. In only one of the
six studied models, did income inequality significantly affect the election
outcome for populist parties, although the estimates had the expected
positive sign in all models. This is surprising given how much awareness
has been built around income inequality after the Crisis.
To conclude, despite being elusive as a concept, populism can be measured in at least three different ways. In addition, its dynamics can be
related to underlying social and economic trends, as expected. Crises do
induce resurgences of populism. In turn, this may fuel further economic
and political crises, as in Dornbusch and Edwards (1990). Populism resurgence seems to be a trait of most isolated cases of local crises, notably in
Latin America in the past, and more recently in Europe, UK, and the
USA. This may be surfacing as a global phenomenon only now because
the Great Recession was the first major peacetime global recession since
the Great Depression. Therefore, we need to start thinking about the economic underpinnings of populist resurgence in a more systematic fashion.
Both Rode and Revuelta (2015) and Heinö (2016) offer a way to go in this
direction by measuring populism either as a discourse, or as actual electoral support for predefined left-wing or right-wing populists. Those two
6 Crises, Welfare, and Populism
151
ways to understand populism do have pitfalls but nevertheless seem the
most promising avenues that the literature has come up with to this date.
An additional avenue for research would be to build case studies of
the political landscape dynamics in countries which are similar in certain
characteristics, in which economic freedom has affected welfare and hence
populist support in different ways. This is exactly what the next section
does.
6.6
Freedom, Populism, and Welfare: Case
Studies
The purpose of the case studies below is to look into after-crisis policy
responses and their welfare and political implications in further detail. In
some cases, after-crisis policy responses are diverging, thereby strengthening the argument that economic freedom ultimately has a positive impact
on welfare. At the same time, the after-crisis economic dynamics gives
rise to divergent political dynamics as well. Specifically, as the empirical
estimations above have shown, the recent rise of populism is related to
deeper and longer recessions.
The case studies have been chosen based on two criteria: geographical representation, and freedom-welfare or freedom-populism dynamics.
Thus, pairs of similar countries from various world regions that experienced diverging economic freedom and welfare patterns have been chosen.
Whenever data is available or easily constructible in the spirit of Heinö
(2016), patterns of populist party popularity are added.
Ireland and Greece are discussed in the Eurozone; Chile and Venezuela
represent Latin America; and Bolivia and Paraguay along with Burundi and
Rwanda represent smaller land-locked economies in Latin America and
Africa. Finally, I compare economic freedom patterns in China and Japan,
and relate them to income per capita and income inequality dynamics in
large open Asian economies. Typically, economic and political dynamics
are reviewed as far back as 1980, whenever data is available. They exhibit
the expected positive correlation described earlier by Lawson and Clark
(2010).
152
P. Stankov
6.6.1 The Eurozone
Ireland and Greece are interesting because of their divergent welfare
patterns since the Great Recession despite both being members of the
Eurozone. Figure 6.1a shows that, at least since 1980, Ireland was the
faster-growing economy of the two and with a freer economy throughout
(b)
GDP per Capita and Economic Freedom
Unemployment and Economic Freedom
1985
1990
1995
2000
2005
2010
7.00
6.00
1980
1985
1990
1995
Year
GDP/c., Ireland
EFW-Total, Ireland
(c)
Unempl., Ireland
EFW-Total, Ireland
(d)
GDP per Capita and Populism
2005
2010
2015
Unempl., Greece
EFW-Total, Greece
Long-Term Unemployment and Populism
60
70
40
60
20
50
40
1985
1990
1995
2000
2005
2010
0
30
0
Long-Term Unemployment, %
40
20
100
TAP Index
60
150
Ireland Vs. Greece, 1980-2015
50
2015
1980
1985
1990
1995
Year
GDP/c. in IRL
TAP, Ireland
(e)
GDP/c. in GRC
TAP, Greece
LT Unempl., IRL
TAP, Ireland
(f)
Government Spending and Populism
2005
2010
2015
LT Unempl., GRC
TAP, Greece
Income Inequality and Populism
1980
1985
1990
1995
2000
2005
2010
2015
Year
Gov't Spending, IRL
TAP, Ireland
60
40
36
20
34
0
30
0
10
32
Gini Coefficient
40
TAP Index
20
15
20
38
60
40
Ireland Vs. Greece, 1980-2015
25
Ireland Vs. Greece, 1980-2015
Gov't Spending, % of GDP
2000
Year
TAP Index
GDP/c., 2000=100
2000
Year
GDP/c., Greece
EFW-Total, Greece
Ireland Vs. Greece, 1980-2015
1980
EFW Summary Index
20
15
10
2015
5.00
5.00
5
Unemployment, %
25
8.00
7.00
6.00
EFW Summary Index
100
GDP/c., 2000=100
50
1980
8.00
Ireland Vs. Greece, 1980-2015
150
Ireland Vs. Greece, 1980-2015
TAP Index
(a)
1980
1985
1990
1995
2000
2005
2010
2015
Year
Gov't Spending, GRC
TAP, Greece
Gini, IRL
TAP, Ireland
Gini, GRC
TAP, Greece
Fig. 6.1 The crisis, economic freedom, and populism: Ireland vs. Greece. Source
a EFW and WDI. b EFW and WDI. c Own calculations. Data source Heinö (2016)
and WDI. d Heinö (2016) and WDI. e Heinö (2016) and WDI. f Heinö (2016) and
Milanovic (2014)
6 Crises, Welfare, and Populism
153
the time period. It also demonstrates that both countries were hit hard
by the Great Recession, with Ireland bottoming-out sooner than Greece.
In 2015, Ireland was already fast approaching 150% of its 2000-level of
GDP per capita, while Greece was still struggling to regain its 2000-level
of per capita GDP.
Some part of the reason for this welfare divergence can be traced to
the divergent patterns of economic freedom after the Crisis. We can note
that both countries experienced a slight downward movement in their
economic freedom just before the Crisis. However, Ireland was quick to
reverse this immediately after, while Greece has continued to fall downward in the Economic Freedom rankings ever since. As more economic
freedom imposes more flexibility in how product-, labor-, and financial
markets operate, then we can safely conclude that freedom reforms brought
more resilience to the Irish economy to respond to the perils of the Great
Recession.
One of those perils is unemployment. The patterns relating to unemployment and economic freedom in both countries since the 1990s are
depicted in Fig. 6.1b. As both economies grew steadily before 2008, unemployment fell to historically low levels, in the case of Greece, or stayed
approximately constant at natural levels, in the case of Ireland. Around
2008, the tides turned for both countries. They experienced a steep rise in
unemployment, with Greece reaching a staggering 26% in 2013. At the
same time, Ireland also saw its unemployment rise to levels not seen since
the 1990s.
The difference between the two countries is their divergent patterns of
economic freedom. While Ireland responded to the Crisis with more freedom, which coincided with a declining growth of unemployment, Greece
took the opposite stance. This saw a continued rise in unemployment in
Greece, while it started declining in Ireland. Naturally, we can blame the
depth of the Crisis for the divergent unemployment patterns, but a significant part of the divergence is perhaps attributable to the differences in
economic freedom.
The depth of the Great Recession had a very different impact on how
the political environment developed in those two countries as well. Specifically, the patterns of populist support in Ireland and Greece were dramatically different. The continued decline of income per capita in Greece
154
P. Stankov
beyond 2011 coincided with a rapid increase in political support for populist parties. At the same time, populist support in Ireland barely budged,
although it did increase slightly. This could indicate that voters adjust
their political preferences only after a certain patience threshold has been
reached, e.g., several years of consecutive decline in income per capita, as
in Fig. 6.1c, or several years of increase in long-term unemployment, as in
Fig. 6.1d.
Ireland and Greece witnessed not only divergent patterns of income per
capita, unemployment, and populist support, but also how government
changed its spending in response to the Crisis. Since 1995, Greece has
traditionally had a larger government than Ireland. In both countries,
the share of government purchases in GDP, which traditionally measures
the size of government, was growing before the Crisis. Both also saw a
marked decline in government intervention after 2008, with a far steeper
decline in Ireland, especially after 2013.The sharper decline in government
expenditures in Ireland, which coincided with a smaller increase in populist
support, indicates that austerity is perhaps not the core factor behind the
rise of populism, at least not in Europe. This is clearly visible in Fig. 6.1e.
At the same time, income inequality was rising equally fast in both countries after 2008, with dramatically different political outcomes. Despite
rising income inequality, Irish voters did not change their attitudes toward
populist parties, while Greek voters have elected a populist government
to negotiate its debt restructuring with their international creditors. What
this tells us is that voters can tolerate a rise in income inequality after a deep
recession as long as the median voter becomes better-off fast enough. However, if the income pattern of the average household mimics an economic
depression, as in the Greek scenario, it is far more likely that populists
will occupy the executive power—a lesson taught with even more vigor in
Latin America.
6.6.2 Latin America
Latin America has long exhibited stark contrasts in its political environment, economic freedom policies, and economic performance. The two
countries illustrating those contrasts best are Chile and Venezuela.
6 Crises, Welfare, and Populism
155
Figure 6.2 presents some of the stark differences between Chile and
Venezuela. Between 1985 and 2015, Chile has tripled its income per
capita. Today, it is the only country in South America that is also a member of the Organization for Economic Cooperation and Development
(OECD)—the club of the world’s richest countries. Meanwhile, Venezuela
has experienced a secular stagnation around its 1980 level of income per
capita. Today, Venezuela’s per capita income is at or lower than it used to
be in 1980—a remarkable stagnation despite the country holding one of
the richest oil reserves on the globe.
A significant part of the income variation across the two countries
can be attributed to how policies have panned out since 1980. While
Chile has been a champion of market-oriented reforms in South America,
Venezuela has steadily deteriorated its economic freedom. Interestingly,
until approximately the mid-1980s, Venezuela was the freer economy, as
indicated in Fig. 6.2a. However, it rapidly reduced its economic freedom,
initiating a perilous welfare trend.
One of the very few positive achievements of macroeconomic policy
management in Venezuela is a faster decline in income inequality than that
of Chile. By 2010, Venezuela had managed to reduce its Gini coefficient to
around 40—an impressive decline given its enormous income inequality
in 1980. Meanwhile, Chile started with about the same income inequality as Venezuela in 1980: an astounding Gini coefficient of around 55.
Figure 6.2b shows that more than 30 years later, Chile has not managed
to reduce it significantly.
Politically, the two countries also paid a very different price for their
macroeconomic policy management. Since the mid-1990s, voters in
Venezuela have elected left and far-left populists whose policy priorities are
macroeconomic redistribution at the expense of economic stability and
growth. The significant populist support in Venezuela also coincides with
a marked shift toward less economic freedom, as suggested for a crosssection of countries by Rode and Revuelta (2015). Meanwhile, Chilean
voters have barely noticed the existence of left-wing parties, let alone leftwing populists. Today, not only is Chile one of the most economically free
countries in the world, but its voters also barely notice the existence of
politicians whose agendas emphasize the need for macroeconomic redistribution, a story consistent with the recent findings of Pecoraro (2017).
156
P. Stankov
1990
1995
2000
2005
2010
1980
1985
1990
1995
Gini, Chile
EFW-Total, Chile
(d)
Left-Wing Populism and Economic Freedom
1985
1990
1995
2000
2005
2010
40
60
150
20
100
GDP/c., 2000=100
0
1980
Populist Support, Venezuela
EFW-Total, Venezuela
1985
1990
1995
(f)
2005
2010
2015
2000
2005
2010
2015
GDP/c., Venezuela
Populist Support, Venezuela
Unemployment and Left-Wing Populism
60
40
12
20
10
0
8
Unemployment, %
14
Chile Vs. Venezuela, 1980-2015
6
60
40
20
0
1995
Voter Support for Left-Wing Populists
60
40
20
Inflation, CPI, Y-on-Y, %
0
1990
2000
GDP/c., Chile
Populist Support, Chile
Chile Vs. Venezuela, 1980-2015
1980
1985
1990
1995
Year
Inflation, Chile
Populist Support, Chile
Gini, Venezuela
EFW-Total, Venezuela
50
2015
Inflation and Left-Wing Populism
1985
2015
Year
Populist Support, Chile
EFW-Total, Chile
1980
2010
GDP per Capita and Left-Wing Populism
Year
(e)
2005
Chile Vs. Venezuela, 1980-2015
EFW Summary Index
3.00 4.00 5.00 6.00 7.00 8.00
60
40
20
0
Voter Support for Left-Wing Populists
GDP/c., Venezuela
EFW-Total, Venezuela
Chile Vs. Venezuela, 1980-2015
1980
2000
Year
GDP/c., Chile
EFW-Total, Chile
(c)
3.00 4.00 5.00 6.00 7.00 8.00
60
55
50
45
Gini Coefficient
2015
Year
Voter Support for Left-Wing Populists
1985
Chile Vs. Venezuela, 1980-2015
2000
2005
2010
Voter Support for Left-Wing Populists
1980
Income Inequality and Economic Freedom
40
100
50
GDP/c., 2000=100
150
3.00 4.00 5.00 6.00 7.00 8.00
Chile Vs. Venezuela, 1980-2015
EFW Summary Index
(b)
GDP per Capita and Economic Freedom
EFW Summary Index
(a)
2015
Year
Inflation, Venezuela
Populist Support, Venezuela
Unempl., Chile
Populist Support, Chile
Unempl., Venezuela
Populist Support, Venezuela
Fig. 6.2 The crisis, economic freedom, and populism: Chile vs. Venezuela.
Source a Own calculations. Data source EFW and WDI. b Own calculations. Data source EFW and Milanovic (2014). c Data source EFW,
http://www.electionguide.org.
d
Data
source
WDI,
Nohlen
(2005),
Nohlen (2005), http://www.electionguide.org. e Data source WDI, Nohlen
(2005), http://www.electionguide.org. f Data source WDI, Nohlen (2005),
http://www.electionguide.org
The detrimental effects of macroeconomic redistributive policies pursued by populists have already been discussed. It seems that more than
25 years after the influential study of Latin American populism
6 Crises, Welfare, and Populism
157
by Dornbusch and Edwards (1990), the lessons from the rest of the continent have not been learned in Venezuela. For a long time now, its politicians have produced a textbook example of an economy trapped in a
resource curse as discussed by Sachs and Warner (1999), which has been
further complicated by persistently loyal voters to populist platforms.
This may be about to change. For the 30 years running till 2010,
Venezuelan voters have trusted mostly left-wing parties pursuing macroeconomic redistribution policies. Since Hugo Chavez was elected in 1998,
economic freedom has stagnated. The negative correlation between leftwing governments and economic freedom has been recently documented
for a number of OECD and EU countries by Jaeger (2017). In Venezuela,
initially the economy picked up, fueled by rising oil prices. The rising
economy in the run-up to 2008 brought surging support for the economic policies pursued by Chavez. His party swept more than 60% of
votes in the elections before the Great Recession.
However, the decline in oil prices and the global recession hit the
Venezuelan economy hard. The policy response to the Crisis was more
populist measures which, unlike in the boom years before the Crisis, did
not help this time around. The economy continued to plunge, and voters finally noticed. The electoral results on parliamentary and presidential
elections in Venezuela since 2010 portray a certain populism fatigue—a
decline in the share of voters supporting Chavist policies, which mimics
the renewed decline in their income per capita. This is clearly indicated
in Fig. 6.2d.
At the same time, populism fatigue was exacerbated by rising inflation
and unemployment. Those are seen in Fig. 6.2e,f. This will inevitably cause
a surge in income inequality as well when more data becomes available.
All this fits very well into the populist paradigm: sooner rather than
later, macroeconomic mismanagement pursued by populists hurts most
exactly the people who were supposed to benefit from populist policies.
The tragic Spring 2017 protests across Venezuela illustrate the dire political
consequences of the long-term pursuit of populist redistribution and of
stifling economic freedom.
Meanwhile in Chile, the rapid increase in income per capita, combined
with long-term price stability and a steady decline in unemployment has
158
P. Stankov
produced one of the most stable political environments in Latin America
which, since the early 1970s, is void of any influential left-wing populists.
The parallels between Greece and Ireland and Venezuela and Chile produce an intuitive conclusion which further supports the empirical results
in earlier chapters. Specifically, long-term income stagnation gives rise to
populist movements. If populists are elected to power, economic freedom suffers. In turn, this produces a hostile business environment which
worsens the prospects for a welfare increase. The only meaningful choice
variable populists have—income inequality—turns out to be a poor policy
target. Instead, it is better left as an outcome of prudent macroeconomic
policies like those pursued by Ireland and Chile. Yet, some milder forms of
redistribution are perhaps inevitable. This is to pre-empt a further surge in
populist movements, given the recent uptick in voter support for populism
in traditional economic freedom strongholds like Chile and Ireland.
6.6.3 Land-Locked Countries: Latin America and
Africa
Land-locked countries are interesting to analyze because trading with
them is harder, which makes domestic economic policies—including economic freedom policies—all the more important for welfare. Bolivia and
Paraguay are the only two land-locked countries in Latin America. Both
have accelerated their efforts to become more free from government intervention since 2000, and both are growing strong. In fact, Paraguay’s GDP
per capita and overall economic freedom pattern have very much in common: Whenever economic freedom rose, GDP per capita followed suit,
and whenever economic freedom was at risk, the average citizen of that
country suffered a blow to their income.
Paraguay suffered its own lost decade in terms of income per capita. The
years between 1995 and 2007 saw the average citizen of that country experiencing a marked decline in per capita income, which had risen steadily
in the decade before. Bolivia experienced not one but two lost decades
in the period between 1980 and 2000. However, realizing its economy
was not doing well, Bolivian governments have pursued more economic
freedom since 1985. As soon as the government started introducing more
159
6 Crises, Welfare, and Populism
(b)
GDP per Capita and Economic Freedom
Income Inequality and Economic Freedom
1990
1995
2000
2005
2010
6.00
5.00
4.00
3.00
1980
1985
1990
1995
GDP/c., Paraguay
EFW-Total, Paraguay
Gini, Bolivia
EFW-Total, Bolivia
(d)
2000
2005
2010
2015
Income Inequality and Economic Freedom
7.00
60
6.00
50
5.00
Gini Coefficient
40
1980
1985
1990
1995
Year
GDP/c., Burundi
EFW-Total, Burundi
Gini, Paraguay
EFW-Total, Paraguay
4.00
4.00
1995
2015
30
5.00
6.00
7.00
EFW Summary Index
200
150
100
GDP/c., 2000=100
50
1990
2010
Burundi Vs. Rwanda, 1980-2015
8.00
Burundi Vs. Rwanda, 1980-2015
1985
2005
20
GDP per Capita and Economic Freedom
1980
2000
Year
GDP/c., Bolivia
EFW-Total, Bolivia
(c)
EFW Summary Index
60
55
50
2015
Year
8.00
1985
45
Gini Coefficient
3.00
40
4.00
5.00
6.00
EFW Summary Index
140
120
100
GDP/c., 2000=100
80
1980
7.00
Bolivia Vs. Paraguay, 1980-2015
7.00
Bolivia Vs. Paraguay, 1980-2015
EFW Summary Index
(a)
2000
2005
2010
2015
Year
GDP/c., Rwanda
EFW-Total, Rwanda
Gini, Burundi
EFW-Total, Burundi
Gini, Rwanda
EFW-Total, Rwanda
Fig. 6.3 The crisis and economic freedom in land-locked countries. Data Source
a EFW and WDI. b EFW and Milanovic (2014). c EFW and WDI d EFW and Milanovic
(2014)
freedom, the economy slowed its decline, and recovery ensued soon thereafter. Today, Bolivia’s average income is 40% larger than in 2000, which is
at least partly due to its growing economic freedom, as seen from Fig. 6.3a.
At the same time, income inequality is also correlated with economic
freedom in the long run. This is easily seen in Fig. 6.3b. The marked
improvement in economic freedom in both countries between 1985 and
1995 preceded a sharp increase in income inequality, measured by the Gini
coefficient. Around 1995, the governments in both countries slowed their
progress with economic freedom, which coincided with a steady decline in
income inequality at least until the Crisis. Their after-crisis policy efforts
to spur economic freedom expectedly coincided with a period of renewed
growth in income inequality. Therefore, land-locked countries in Latin
America become richer when more economic freedom is adopted, but at
the same time, experience rising income inequality.
160
P. Stankov
The lessons from two small land-locked African economies are very similar. They can be seen in Fig. 6.3c. Following an aggressive improvement
of economic freedom since 1995, Rwanda’s people today are, on average, twice as rich as they were in 2000. This remarkable improvement,
however, was not always present. In the years preceding 1995, Rwanda’s
economic freedom policies were wavering, and its income per capita was
declining. Burundi’s GDP per capita, on the contrary, increased before
1990, coinciding with a gradual long-term improvement in its economic
freedom score. However, since 1990, Burundi has experienced a dramatic
slump in its income per capita. By the year 2000, its citizens lost a third of
their income, which went hand in hand with a deterioration of economic
freedom. Ever since 1995, Burundi has been trying to gradually improve
economic freedom, with varying success. Ten years later, the decline in per
capita income has stopped, and a noticeable, yet still meager, recovery of
income per capita has ensued. This has coincided with bolder advancements in economic freedom policies.
As in Latin America, bold advances in economic freedom in African
countries also comes with a hefty increase in income inequality. This is
seen in Fig. 6.3d. Between 1985 and 2005, Rwanda dramatically improved
its economic freedom. At the same time, however, its income inequality
doubled. Only recently, after years of remarkable economic growth, has
its inequality growth began to abate. Still, a Gini coefficient of above 50
signals an extremely polarized population in terms of economic opportunities, despite the extraordinary improvement in economic freedom over
the last 20 years. The data on Burundi is too scarce to enable any meaningful conclusions on the link between inequality and freedom.
Therefore, we can safely conclude that economic freedom is indeed
conducive to income per capita growth, perhaps even more so in landlocked countries. However, more freedom is also correlated with higher
income inequality, as our graphical analysis has shown in earlier chapters.
If that is the case for small land-locked countries in Latin America and
Africa, can we observe similar trends in large open economies? The next
section elaborates.
161
6 Crises, Welfare, and Populism
(b)
1990
1995
2000
2005
2010
2015
7.00
6.00
5.00
4.00
1980
1985
1990
1995
Year
GDP/c., China
EFW-Total, China
8.00
50
45
40
35
Gini Coefficient
30
7.00
6.00
5.00
4.00
1985
EFW Summary Index
400
300
200
GDP/c., 2000=100
100
0
1980
Income Inequality and Economic Freedom
China Vs. Japan, 1980-2015
8.00
China Vs. Japan, 1980-2015
EFW Summary Index
GDP per Capita and Economic Freedom
25
(a)
2000
2005
2010
2015
Year
GDP/c., Japan
EFW-Total, Japan
Gini, China
EFW-Total, China
Gini, Japan
EFW-Total, Japan
Fig. 6.4 The crisis and economic freedom in large open economies. Source a EFW
and WDI. b EFW and Milanovic (2014)
6.6.4 Large Open Economies: China and Japan
China and Japan are two of the largest economies on the globe, experiencing different reform and income patterns over the last 35 years. They
also have remarkably different growth trajectories, observed in Fig. 6.4a.
By 2000, the average Chinese person was four times richer than they were
in 1980. By 2015, income per capita almost quadrupled yet again! This
exceptional growth path is in dramatic contrast to Japanese income patterns over the same period of time. Today, the income per capita in Japan
is hardly different from what it was in 1990.
Figure 6.4a reveals a possible reason for this astonishing difference in
China and Japan’s income growth patterns, which complements the traditional explanations of income convergence. By 1990, the Japanese economy was still improving its economic freedom, and its GDP per capita
grew. After 1990, it worsened its economic freedom, and income stagnation followed suit. Japan is still a free economy by any standards. However,
it has not done much to improve its economic freedom since 1990. In
contrast, China has done a great deal. It started as one of the less free
countries in 1980. Since 1990, Chinese policies have always been consistent with a gradual yet very noticeable improvement in its overall freedom
rankings. Therefore, similarly to small, land-locked countries, large open
economies like China and Japan have also gained additional income per
capita after becoming freer economies.
162
P. Stankov
However, as with many other countries, more economic freedom has
meant rising income inequality, as seen from 6.4b. In 1990, the year
in which China began its march toward economic freedom, its income
inequality was comparable to that of many developed European nations.
25 years later, its inequality looks similar to that of many Latin American
countries. It is only recently, after decades of remarkable growth, that
income inequality growth has stagnated. It even notched down, similarly
to other nations experiencing rapid long-term growth. The long-term
trends in Japan have been similar yet somewhat more subtle. While the
Japanese economy was growing and its economic freedom improved before
1990, its income inequality also rose. When it slumped into stagnation of
both income and economic freedom after 1990, its income inequality was
also trendless between 1993 and 2005—it was comparable to the more
equal European nations.
The broad lessons are confirmed for large open economies as well as
for small land-locked countries. In sum, economic freedom works well
for income per capita, and less so—for income inequality. This broad
conclusion is in line with the large-scale review of the literature by Hall
and Lawson (2014).
References
Alvaredo, F., L. Chancel, T. Piketty, E. Saez, and G. Zucman. 2017. Global
inequality dynamics: New findings from WID.world. Working Paper 23119,
National Bureau of Economic Research.
Beck, T., G. Clarke, A. Groff, P. Keefer, and P. Walsh. 2001. New tools in comparative political economy: The database of political institutions. The World
Bank Economic Review 15 (1): 165–176.
Bittencourt, M. 2010. Democracy, populism and hyperinflation(s): Some evidence from Latin America. Working Papers 169, Economic Research Southern
Africa.
Cahill, B. 2007. Of note: Institutions, populism, and immigration in Europe.
SAIS Review 27 (1): 79–80.
Cruz, C., P. Keefer, and C. Scartascini. 2016a. The database of political institutions 2015. Last updated Jan., 2016.
6 Crises, Welfare, and Populism
163
Cruz, C., P. Keefer, and C. Scartascini. 2016b. Database of Political Institutions:
Changes and Variable Definitions. Inter-American Development Bank. A codebook to the DPI2015.
Dornbusch, R., and S. Edwards. 1990. Macroeconomic populism. Journal of
Development Economics 32 (2): 247–277.
Dornbusch, R., and S. Edwards. 1991. Introduction to “The macroeconomics
of populism in Latin America”. In The macroeconomics of populism in Latin
America, ed. R. Dornbusch and S. Edwards, 1–4. University of Chicago Press.
Greskovits, B. 1993. The use of compensation in economic adjustment programmes. Acta Oeconomica 45 (1/2): 43–68.
Hall, J.C., and R.A. Lawson. 2014. Economic freedom of the world: An accounting of the literature. Contemporary Economic Policy 32 (1): 1–19.
Hawkins, K.A. 2009. Is Chávez populist? Measuring populist discourse in comparative perspective. Comparative Political Studies 42 (8): 1040–1067.
Heinö, A.J. 2016. Timbro authoritarian populism index, 2016. Sweden: Timbro
Institute, Stockholm.
Hewison, K. 2005. Neo-liberalism and domestic capital: The political outcomes
of the economic crisis in Thailand. The Journal of Development Studies 41 (2):
310–330.
Jaeger, K. 2017. Economic freedom in the early 21st century: Government ideology still matters. Kyklos 70 (2): 256–277.
Jagers, J., and S. Walgrave. 2007. Populism as political communication style: An
empirical study of political parties’ discourse in Belgium. European Journal of
Political Research 46 (3): 319–345.
Jones, E. 2007. Populism in Europe. SAIS Review 27 (1): 37–47.
Kaufman, R. R. and B. Stallings. 1991. The political economy of Latin American
populism. In The macroeconomics of populism in Latin America, ed. R. Dornbusch and S. Edwards, 15–43. University Of Chicago Press.
Lago, R. 1991. The illusion of pursuing redistribution through macropolicy:
Peru’s heterodox experience, 1985–1990. In The macroeconomics of populism in
Latin America, ed. R. Dornbusch and S. Edwardspp, 263–330. University of
Chicago Press.
Lawson, R.A., and J.R. Clark. 2010. Examining the Hayek-Friedman hypothesis
on economic and political freedom. Journal of Economic Behavior & Organization 74 (3): 230–239.
Leon, G. 2014. Strategic redistribution: The political economy of populism in
Latin America. European Journal of Political Economy 34: 39–51.
Milanovic, B.L. 2014. All the Ginis, 1950–2012. Updated in Autumn 2014.
164
P. Stankov
Moffitt, B. 2015. How to perform crisis: A model for understanding the key
role of crisis in contemporary populism. Government and Opposition 50 (2):
189–217.
Nohlen, D. 2005. Elections in the Americas: A Data Handbook, vol. 2. South
America: Oxford University Press.
Pecoraro, B. 2017. Why don’t voters ‘put the Gini back in the bottle’? Inequality
and economic preferences for redistribution. European Economic Review 93:
152–172.
Roberts, K.M. 2007. Latin America’s populist revival. SAIS Review 27 (1): 3–15.
Rode, M., and J. Revuelta. 2015. The wild bunch! An empirical note on populism
and economic institutions. Economics of Governance 16 (1): 73–96.
Sachs, J.D., and A.M. Warner. 1999. The big push, natural resource booms and
growth. Journal of Development Economics 59 (1): 43–76.
Tejapira, K. 2002. Post-crisis economic impasse and political recovery in Thailand: The resurgence of economic nationalism. Critical Asian Studies 34 (3):
323–356.
The World Bank. 2017. World Development Indicators, 1960–2016. Last
updated Feb. 01: 2017.
Učeň, P. 2007. Parties, populism, and anti-establishment politics in East Central
Europe. SAIS Review 27 (1): 49–62.
7
Conclusion
This book reviews the dynamics of market-oriented reforms and their
impact on welfare between 1970 and 2014. The need to analyze the longterm impact of reforms is motivated by the seismic changes the Great
Recession has brought to both welfare and the political landscapes around
the world. In this work, welfare is understood more broadly than in the
conventional sense. Its measure is not only the traditional income per
capita, but also how much average citizens consume, how long they live,
and how they fare in terms of income inequality. Throughout the work,
reforms are monitored in five areas: government intervention, protection of property rights, monetary policies, free trade, and government
regulation.
The data on welfare comes from three comprehensive sources: the Penn
World Table 9.0, the World Development Indicators, and the income
inequality data by Milanovic (2014). The data on reforms are taken from
the Economic Freedom of the World 2016 annual report, which includes
data from 1970 to 2014. Both qualitative and quantitative methods are
employed through the book, with an emphasis on quantitative analysis, uncovering the dynamics of reforms from 1970. They also reveal
the impact of those reforms on changes in welfare across countries and
over time, as well as before and after the Great Recession. The literature
© The Author(s) 2017
P. Stankov, Economic Freedom and Welfare Before and After the Crisis,
DOI 10.1007/978-3-319-62497-6_7
165
166
P. Stankov
reviewed in Chap. 2, including the most recent work by both theoretical
and empirical social scientists, does not provide a unanimously positive
answer to the question this book asks: Did more economic freedom produce
more welfare?
Chapter 3 demonstrates that most policies were moving toward less
government since 1970 until the global Crisis of 2008. Thus, reforms
were undertaken in the spirit of the neoliberal policy agenda of letting
markets do their job and limiting governments to simply set the stage for
ever stronger private growth.
After 2008, however, governments around the world overwhelmingly
reverted to protective policy mood, as was the case with many previous
episodes of crises and wars (De La Escosura 2016), and despite earlier
evidence to the contrary by Pitlik and Wirth (2003). This is especially
valid for government intervention and property rights. The evidence for
monetary policies, free trade, and government regulation after the Crisis
is mixed. In those broad areas, governments either did not reform further
or progressed one step forward, followed by two steps backward. That is
the case, for example, with government regulation and free trade. Within
the area of overall government regulations, business regulations have been
made easier, but both financial and labor regulations have been made
tightened. In trade reforms, while governments did not embark on outright trade wars, as was the case in the aftermath of the Great Depression,
they did impose harder non-tariff barriers to trade. Overall, the Crisis has
stalled the momentum of economic freedom reforms.
Chapter 4 demonstrates a significant feature of reforms in all areas: policy convergence, i.e., policies becoming more similar across countries and
over time. We can look at this process in two ways. First, countries lagging
in economic freedom were catching up with the leaders. In other words, we
observed the so-called β-convergence in policies. Second, the entire distribution of policies was decreasing in diversity over time, a process known as
σ -convergence in policies. Both types of policy convergences were highly
significant. As a result, the world has become a far more uniform place in
terms of policies than it used to be 45 years ago.
However, Chap. 4 also shows that, despite converging policies, income
differences across countries persisted. If you were born in a poor country,
your income would not grow significantly faster than the income of a
7 Conclusion
167
person in a rich country. There is convincing evidence of convergence in
consumption per capita, life expectancy, and income inequality.
Chapter 5 tests if there was any significant relationship between policy
and welfare convergence. In other words, it addressed an important political question for most developing countries: If we adopt Western policies,
are we going to gradually become as well-off as they are? As it turns out
from the results in Chap. 5, and contrary to the bulk of earlier evidence
(Ali and Crain 2002; Dawson 2003; Grubel 1998), more often than not
the answer is no.
More specifically, reducing the size of government did not bring significantly more income per capita, nor increase consumption per capita,
nor make lives longer or reduce income inequality in most cases. Improving property rights protection did have a positive effect on income and
consumption growth. It did not produce a significant long-term effect
on life expectancy or income inequality. Unlike property rights, monetary
reforms produced robust positive welfare gains across all welfare measures.
Trade reforms did have a positive effect on income growth but less on
consumption growth. They seem to not affect life expectancy or income
inequality in any systematic way, although indeed people live longer and
enjoy lower income inequality in countries with freer trade. Deregulation
reforms also had a positive impact on income growth, but raised income
inequality. In fact, the case studies presented later in the book demonstrate
that overall economic freedom reforms work well for income per capita
and at the same time raise income inequality.
If that is the case, it is natural that policy agendas would shift from
pro-market to pro-redistribution sooner rather than later. The Crisis has
definitely played a significant part in this. As the average household still
struggles with the consequences of the Great Recession, far-left and farright populists ride the wave of social discontent. Any sensible policy
maker in the middle ground would also have a finger on the pulse of their
voters. In order to preempt a populist resurgence in their own countries,
they could rely on a mix of more economic freedom with stronger income
redistribution. Elements of this redistribution agenda have already been
suggested, among others, by Piketty (2015). A version of this preemptive
policy agenda sold well in the May 2017 French presidential elections
swept by Emmanuel Macron. It remains to be seen if Macron-omics will
168
P. Stankov
evolve into a policy trend across Europe and the rest of the world or will
prove to have been a one-off event. A careful redistribution to reduce
discontent is a far smaller price to pay than the likely damage that could
be caused by a long-run populist government.
However, if politicians pursue stronger redistribution, they should
tread lightly, because large-scale macroeconomic redistribution may push
investors to more business-friendly places around the globe. Reducing
investment is the last thing a sensible politician—left or right—needs in
the aftermath of the Great Recession. As the last chapter demonstrates,
voters go to political extremes predominantly when their income goes
down.
If economic freedom is good for anything, it is per capita income. As
a result, more economic freedom does make sense, even if it does not
do much about welfare beyond raising incomes. But if it thus prevents
populism from marching in, so be it. Because the price of populism is
decades of stagnation. And no sensible voter would like that. Therefore,
enter more freedom.
References
Ali, A.M., and W.M. Crain. 2002. Institutional distortions, economic freedom,
and growth. Cato Journal 21 (3): 415–426.
Dawson, J. W. 2003. Causality in the freedom-growth relationship. European
Journal of Political Economy 19 (3): 479–495. Economic Freedom.
De La Escosura, L.P. 2016. Economic freedom in the long run: evidence from
OECD countries (1850–2007). Economic History Review 69 (2): 435–468.
Grubel, H. 1998. Economic freedom and human welfare: Some empirical findings. Cato Journal 18 (2): 287–304.
Milanovic, B. L. 2014. All the Ginis, 1950–2012. Updated in Autumn 2014.
Piketty, T. 2015. Putting distribution back at the center of economics: Reflections
on ‘Capital in the Twenty-First Century’. Journal of Economic Perspectives 29
(1): 67–88.
Pitlik, H., and S. Wirth. 2003. Do crises promote the extent of economic liberalization?: An empirical test. European Journal of Political Economy 19 (3):
565–581.
Index
A
Austerity, 7, 136, 138, 141, 145,
146, 150, 154
B
Business environment, 1, 158
C
Capital
human, 14, 16, 25, 27
physical, 14, 27
Capital accumulation, 14, 24, 26,
28
Consumption, 2, 14, 22, 85, 101,
103
Consumption convergence, 88, 89,
167
Consumption per capita, 6, 9, 10,
12, 93, 99, 101, 104, 106,
107, 109, 111, 114, 126,
127, 129, 165, 167
Corruption, 17
D
Data
consumption per capita, 5
Database of Political Institutions
(DPI), 140
economic freedom, 3, 4, 43, 112
GDP per capita, 5
income inequality, 6, 112, 141
life expectancy, 6
Penn World Table (PWT), 5,
112
populism, 7, 139, 140
rents, 113
welfare, 5, 6, 112, 165
World Development Indicators
(WDI), 5, 112, 141
Democracy, 19
Developed countries, 5, 10, 17, 30,
74, 90, 139, 140
Developing countries, 5, 10, 11,
17, 24, 26, 28–30, 74, 75,
102, 139, 167
© The Editor(s) (if applicable) and The Author(s) 2017
P. Stankov, Economic Freedom and Welfare Before and After the Crisis,
DOI 10.1007/978-3-319-62497-6
169
170
Index
E
Economic freedom, 1–7, 11–16,
22, 25, 26, 28–30, 43, 52, 71,
76, 79, 99, 112–114, 129,
137, 139, 151, 153–155,
157–162, 166–168
Economic growth, 1, 13–15, 17,
18, 20, 23, 25–27, 29, 30,
56, 69, 87, 108, 136–138,
155, 158, 160–162
F
Financial liberalization, 25, 56, 64,
65
Free trade, 4, 51, 62, 75, 83, 107,
112, 128, 165, 166
G
Gini, 6, 89, 90, 105, 110, 111,
126, 142, 155, 159, 160
Government intervention, 3, 4,
17–19, 44, 59, 75, 83, 100,
101, 114, 125, 158, 165, 166
Graphical evidence, 3, 6, 75, 76,
79, 85, 86, 90, 103, 104,
106, 108, 111, 125
The Great Depression, 166
The Great Recession, 2, 7, 43,
48, 50, 53, 54, 56, 58–60,
62–65, 75, 77, 79, 87,
101–103, 109, 135, 136,
138, 142, 146, 148, 150,
152, 153, 157, 165, 168
I
Income convergence, 21, 84–87,
91
Income inequality, 2, 7, 10–12, 14,
16, 19, 22, 27, 28, 30, 85,
89, 90, 99, 101, 103, 105,
107, 108, 110–112, 114,
125–129, 136, 137, 141,
143, 146, 150, 151, 154,
155, 157–160, 162, 165, 167
convergence in, 89, 93, 94, 167
Income per capita, 2, 6, 7, 9,
11, 12, 28, 30, 71, 87, 91,
99–104, 106–109, 111, 114,
125, 127, 129, 136, 142,
151, 153–155, 157, 158,
160–162, 165, 167, 168
Institutional change, 13
Intellectual property rights, 21, 22
L
Life expectancy, 2, 6, 10–12, 14,
85, 99, 101, 103, 104, 107,
110, 111, 114, 125–129,
165, 167
Life expectancy convergence, 89,
93, 167
Living standards, 2, 71, 101, 102,
108, 127, 143
M
Market-oriented reforms, 1–3, 16,
29, 55, 72, 73, 106, 113,
155, 165
Methods
fixed effects, 3, 6, 7, 112, 143
instrumental variable, 3, 6, 13,
113
OLS, 3, 6, 13, 112
probit model, 142
Index
171
Monetary policy, 3, 30, 48–50, 61,
75, 77, 80, 83, 103–105,
114, 127, 129, 165, 166
Monetary stability, 23, 112, 127
60, 61, 75, 77, 83, 101–103,
105, 111, 112, 114, 125,
126, 129, 165–167
Publication bias, 126
N
Non-tariff barriers, 53
R
Recession, 7, 45, 105, 136, 137,
141, 146, 149, 151
Redistribution, 1, 2, 105, 125, 136,
155, 157, 158, 167, 168
Reform sequencing, 65
Reforms, 1, 3–6, 11, 12, 16, 29,
30, 57, 58, 60, 61, 65, 99,
102, 129, 161, 165, 166
measurement of, 4, 58
Regulation, 4, 15, 29, 54, 55, 75,
79, 83, 112, 166
credit market, 4, 54–56, 58, 63,
64, 79, 80, 83, 108, 166
deregulation, 29, 30, 63–65, 72,
79, 108–111, 114, 128, 129,
167
labor market, 4, 27, 54, 55, 57,
58, 63, 65, 79, 83, 108, 166
political economy of, 29
product market, 4, 27, 54, 55,
57, 58, 63, 65, 79, 108, 166
Rents, 13, 29, 113
natural resources, 113
Resource allocation, 14
Resource curse, 22
P
Policy agenda, 2, 13, 23, 29
Policy coercion, 71, 73, 74
Policy convergence, 1, 5, 6, 55, 57,
66, 69–79, 84, 166, 167
speed of, 80, 83
Policy imitation, 73, 74
Policy learning, 72–74
Policy makers, 1, 13, 29, 30, 54,
78, 93
Political capture, 108
Political market, 2
Populism, 1–3, 7, 135, 136,
138–143, 145, 148, 150,
151, 153–155, 158, 168
authoritarian, 139, 143
as a discourse, 135, 139, 140,
143, 146–148, 150
fatigue, 157
left-wing, 140, 143, 150, 155,
167
measurement of, 139, 141, 150
political economy of, 7, 139,
140
right-wing, 140, 143, 150, 167
Populist paradigm, 137, 145, 157
Post-Crisis growth, 2
Price stability, 49–51, 103, 157
Property rights, 3, 13, 15, 16, 19,
20, 22, 23, 25, 30, 46–48,
S
Size of government, 3, 15, 17–19,
29, 31, 43, 45, 59–61, 72,
76, 100, 101, 112, 114, 125,
129, 142, 143, 154, 167
Supply-side policies, 45
172
Index
T
Tolerance of taxation, 72
Trade liberalization, 24–28, 52, 53,
106–108, 127–129
Trade reforms, 24–28, 30, 62, 63,
78, 106, 107, 114, 166
V
Voter discontent, 2, 136–138, 147,
167
W
Welfare, 2, 3, 5, 6, 9, 10, 12, 14,
18–20, 22–25, 28, 31, 74,
85, 99, 101, 106, 112, 114,
125, 127, 129, 136, 151,
153, 155, 158, 165–168
measurement of, 10–12, 111
Welfare convergence, 5, 6, 74, 84,
86, 90, 91, 93, 167
speed of, 91