The Economywide Impact of The COVID-19 in Ethiopia

Ethiopian Economics Association
(EEA)
The economywide impact of the COVID-19

in Ethiopia: Policy and Recovery options
Lulit Mitik Beyene, Tadele Ferede and Getachew Diriba
Policy Working Paper 03/2020
July 2020
The economywide impact of the COVID-19 in
Ethiopia: Policy and Recovery options1
Lulit Mitik Beyene2, Tadele Ferede3 and Getachew Diriba4
Policy Working Paper 03/2020
July 2020
1
This paper is released urgently to make it available for reference; hence it has not
undergone a peer-review process.
2
Researcher, Partnership for Economic Policy (lulit.mitik@gmail.com)
3
Associate Professor, Department of Economics, Addis Ababa University
(tadele.ferede@aau.edu.et)
4
Chief Executive Officer Ethiopian Economics Association (gdiriba@yahoo.com).
The economywide impact of the COVID-19 in Ethiopia: … Policy Working Paper 03/2020
Copyright © Ethiopian Economics Association (EEA)

All rights reserved.
ISBN: 978 99944 54 76-1
The views expressed herein are those of the authors. They do not
necessarily reflect the views of the Ethiopian Economics Association, its
Executive Committee, or its donors.
The works of Ethiopian Economics Association including this publication

is made possible by the generous grants of the Bill & Melinda Gates
Foundation (BMGF) and Open Society Initiative for Eastern Africa
(OSIEA).
ii
Table of Contents
1. Introduction ............................................................................................ 1
2. Methodology ........................................................................................... 3
2.1. Model descriptions ......................................................................... 3
2.2. Transmission channels and scenarios........................................... 5
2.2.1. TRANSMISSION CHANNELS ........................................................... 5
2.2.2. IDENTIFYING SHOCKS ................................................................... 7
2.2.3. SCENARIOS ................................................................................... 9
3. Simulation results: Main findings ...................................................... 15
3.1. The short-term effects with government interventions ............ 15
3.1.1. GROWTH EFFECT ........................................................................ 15
3.1.2. FISCAL EFFECT ........................................................................... 19
3.1.3. SECTORAL EFFECT ...................................................................... 21
3.1.4. EMPLOYMENT EFFECT ................................................................ 24
3.1.5. WELFARE EFFECT ....................................................................... 27
3.2. Prospects in the medium-to-long term ....................................... 30
4. Conclusions and recommendations .................................................... 40
References .................................................................................................... 42
Annex 1 - The short-term effects without government interventions . 43
iii
Executive Summary
This study assesses the potential economywide impacts of the COVID-19 in

Ethiopia. It uses a dynamic computable general equilibrium (CGE) model
calibrated to a social accounting matrix for FY 2010/11 and covers the period
from FY 2010/11 to FY 2029/30. The analysis accounts for the main channels
through which the COVID-19 affects the economy. The domestic transmission
channels include reduced labor market participation, lower productivity, and
rising domestic trade costs. External channels include higher international trade
costs, a drop in export demand, lower import supply, a reduction in foreign direct
investment (FDI), reduction in remittances, and lower import price of oil. The
impact of the COVID-19 crisis is analysed using three scenarios, namely business
as usual (or the baseline), and the COVID-19 scenario considered under mild and
severe assumptions. Economic impacts resulting from the COVID-19 crisis are
expected to have differentiated impacts on a wide range of economic and social
indicators. The results capture the impact on national accounts, the fiscal
framework, the external accounts, the labor market, and household welfare. The
distributional impact is captured through the effect on key economic sectors and
on household groups distinguished by income quintiles and rural urban areas.
The COVID-19 pandemic has spread to the Ethiopian economy through multiple
international and domestic channels. The infection rate is increasing, and the
death toll is rising. All regions are affected with higher spread in the capital city,
Addis Ababa. It is likely that the duration of the pandemic will be at a minimum
six months. The scenarios designed are based on a six-month duration with mild
and severe scenarios.
The COVID-19 pandemic is likely to have significant growth effects even under
an optimistic scenario of mild shock and quick recovery. Our estimates suggest
that GDP would be lower than in the no-COVID-19 scenario by 127 billion ETB
in FY 2019/20 and between 159 and 310 billion ETB in FY 2020/21. On this
basis, the economy would grow by 2.6 percent in FY 2019/20. Under an amplified
(or severe) pandemic scenario, GDP growth would only reach 0.6 percent in FY
2020/21.
iv
Figure A: GDP level and loss (in billion ETB)
5,500 0
5,000 -200
4,500
-400
4,000
-600
3,500
-800
3,000
-1,000
2,500
2,000 -1,200
1,500 -1,400
2019 2020 2021 2022 2024 2026 2028 2030
Loss Mild case Loss Severe case GDP Severe case

GDP No-COVID-19 GDP Mild case
Source: Simulation results
The COVID-19 pandemic is likely to have adverse effects on key sector of the
Ethiopian economy. Not all sectors are equally affected by the crisis with higher
contraction of manufacturing activities followed by agriculture. Export intensive
industries such as textile and leather manufacturing, export-oriented agriculture,
transportation services, accommodation and food services are likely to be hit the
hardest. The effect on agriculture is much larger when we introduce a disruption
in import supply of fertilizer and other chemicals and a reduction in export
demand for Ethiopian coffee. The construction sector suffers from the investment
reduction.
Employment is likely to be hit hard. The employment level is between 8.6 percent
and 16.5 percent lower than the baseline. Job losses would be severe in all the
export-oriented sectors. Rural employment is slightly more affected than the
urban employment and the unskilled workers are negatively affected more than
the skilled and semi-skilled.
v
Figure B: GDP annual growth

10.0
9.0
8.0
7.0
6.0
5.0
4.0
3.0
2.0
1.0
0.0
2019 2020 2021 2022 2024 2026 2028 2030
No-COVID-19 Mild case Severe case
The COVID-19 crisis is likely to have medium-to-long-term negative effects. All

the indicators show the likelihood of medium to long term effects even in an
optimistic scenario. While GDP growth rate is expected to converge to the no-
COVID-19 baseline, the GDP losses are not likely to be recovered. The size of
the Ethiopian economy would remain well below the no-COVID-19 baseline
level. The recovery would have a V shape (i.e. quick recovery) in the mild case
and a U shape in the severe case.
The negative impact of the crisis on household welfare would be severe. Real
consumption expenditure could be between 4.6 and 12 percent lower than in the
reference scenario in the short term. In the absence of appropriate measures, the
most vulnerable populations are likely to be more severely impacted. The welfare
loss of the bottom 20 percent is between 1.6 and 2.5 times higher than the top 20
percent without any in-kind and/or cash transfers.
vi
Figure C: Household consumption (percent deviation relative to no-

COVID-19 scenario)
0.0
-2.0
-4.0
-6.0
-8.0
-10.0
-12.0
-14.0
-16.0
-18.0
-20.0
r-f q1 r-f q2 r-f q3 r-f q4 r-f q5 r-nf q1 r-nf q2 r-nf q3 r-nf q4 r-nf q5 u q1 u q2 u q3 u q4 u q5
2020 with relief measures 2020 no relief measures

2021 with relief measures mild 2021 no relief measures mild
2021 with relief measures severe 2021 no relief measures severe
Notes: r-f= rural farm; r-nf=rural nonfarm; u=urban; and q 1-q5=quintiles.

Welfare losses would persist over time. By the end of the time horizon of the
model, household welfare at the national level would be 1.9 percent and 10.7
percent lower than the no-COVID-19 scenario in the mild and severe cases,
respectively. Even if the transmission channels of the COVID-19 to the Ethiopian
economy partly regain their pre-crisis level, households in the lowest income
categories are likely to have higher welfare reduction than households in the
highest quintile.
The COVID-19 pandemic is likely to have a substantial effect on public finance.

Fiscal deficit is likely to widen in absolute terms and in percentage of GDP.
Government revenue would decline. At the same time, expenditure would
increase to deliver emergency health care services and food assistance, and
increase containment efforts, all of which will widen fiscal deficit.
vii
Government response plan is of paramount importance for the medium- and long-
term perspectives, especially if the impact of the pandemic is to be more severe.
The adverse impact of the pandemic on investment would be even larger without
government intervention. Household welfare would fall more sharply without
relief and recovery measures. Furthermore, without the assistance of development
partners, deficit financing would result in the deterioration of the fiscal
framework with a risk of jeopardizing macroeconomic stability and debt
sustainability.
The anticipated 3.4 billion USD government response plan meant for emergency
responses, to support businesses and protect jobs would certainly contribute to
protecting the livelihoods of workers and businesses. However, it is not enough
to put the economy on a higher growth path that would reduce that gap with pre-
COVID-19 situations.
There is uncertainty on the duration of the pandemic in Ethiopia and worldwide.

This implies that recovery may not come as quickly as would be anticipated
putting the Ethiopian economy closer to the severe scenario rather than the mild
case. Government support is much needed not only by increasing its spending
under the COVID-19 response plan, but by creating an enabling environment that
would allow businesses to thrive and social safety nets to share the burden.
Shielding the most vulnerable groups from the pandemic is crucial to limit the
populations that would fall (or back) into poverty. In-kind and cash transfers to
the most vulnerable are likely to narrow the gap between households in the lowest
quintiles and those in the highest quintiles. Hence it is necessary to expand the
existing social assistance programs such as rural and urban safety nets to protect
the most vulnerable segments of the population.
Given the multifaced nature of COVID-19 induced challenges facing the country,
a recovery and response plan is urgently needed to achieve dual objectives of
mitigating further economic contraction and of stimulating the economy. The
recovery and response plan shall target and safeguard sectors essential for food
security, job creation and sustainable and inclusive growth. The plan needs to take
into account the differentiated impacts of the pandemic on different activities and
households.
viii
Both fiscal and monetary policy instruments that have been introduced by the
government to fight the pandemic shall be continued, enforced and implemented,
in a coordinated way, to support the effectiveness of interventions until the
economy recovers. However, support measures need to be monitored and
evaluated to facilitate the transition towards economic recovery and boost
economic dynamism.
The recovery and response plan requires rapid and predictable financing which
shall come from two sources. It is important that the international community
provides quick and coordinated support. Given the global nature of the pandemic
and uncertainty of external finance in terms of timing and amount, it is necessary
to design a strategy for mobilizing and diversifying domestic sources of finance
to manage and support the crisis and recovery in a more dependable manner. It is
also necessary to evaluate tax related supports ongoing basis as across-the-board
tax reductions, deferrals and relief may jeopardize medium-term revenue raising
capacity of the government.
ix
1. Introduction
The 21st century economies are interconnected due to advancement in information

and communication technology and global value chains. In this chain, several
actors can be identified: workers, firms, suppliers, consumers, banks and financial
intermediaries. In this web of interconnectedness, if there are disruptions in any
one of the links due to the disease or containment measures, the outcome could
be a cascading chain of disruptions both across the world and within countries.
The COVID-19 adversely impacts economic activities which in turn, if

unchecked, will lead to social crisis. Efforts to fight the virus affects economic
activity and changes the economic trajectory. Keeping workers away from work
and consumers away from consumption tend to reduce economic activity. The
COVID-19 pandemic creates all manner of economic shocks and involves
feedbacks among different economic actors. The virus triggers three types of
economic shocks (Triggs and Kharas, 2020). First, as the pandemic spreads, it
hits output by putting workers into their sickbeds creating temporary
unemployment if the spread of the virus is contained. Second, public-health
related containment measures (e.g. factory and office closures, travel bans,
quarantines, and the like) will affect the economy through reduced supply, low
demand and financial distress. Third, consumers and firms all around the world
would be in a state of a wait-and-see mode (i.e. expectations).
The COVID-19 pandemic has spread to the Ethiopian economy through multiple
international and domestic channels. The infection rate is increasing, and the
death toll is rising. All regions are affected with higher spread in the capital city,
Addis Ababa. It is likely that the duration of the pandemic will be at a minimum
six months.
This study seeks to answer three policy questions:

• What is the economy-wide effects of COVID-19 in Ethiopia?
• How long will it take for the economy to recover to its pre-shock level?
1
• Is the current level of rescue and stimulus sufficient for the economy to
reach its pre-shock level and beyond?
The paper is organized in five sections. Section 2 presents the methodology while
Section 3 discusses the simulation results and main findings. Finally, Section 4
presents conclusions and recommendations.
2
2. Methodology
2.1. Model descriptions
Economic impacts resulting from the COVID-19 crisis are expected to have
differentiated impacts on a wide range of economic and social indicators. It is
therefore important to apply a methodology that allows to capture the impact of
the shock through its multiple dimensions including on the national accounts
(GDP, consumption, investment), the fiscal framework (government revenue,
expenditure, and deficit and debt), the external accounts (trade, FDI, and the
current account), the labor market (sectoral employment and wages), and
household welfare.
The study uses a dynamic Computable General Equilibrium (CGE) model to

assess the potential short-term and medium-to-long-term impacts of the pandemic
in Ethiopia. The CGE methodology is well-suited as CGE models take into
account the economic interdependencies between the different sectors and
economic actors in the country and internationally allowing to capture direct,
indirect and feedback effects of policies and shocks. The behavioural functions
are based on sound microeconomic theory. In this model, agents adopt a
profit/utility maximization behaviour and market prices adjust to reconcile
endogenous supply and demand decisions, thus determining levels of production,
employment, and consumption. This type of method offers a comprehensive
assessment and analysis on a wide range of indicators including production,
consumption, factor markets and prices.
It is also a tool that enables to simulate relief and recovery measures. The
distributional effect is captured through the impact on households distinguished
by income quintiles across rural-farm, rural non-farm and urban settings. This
allows to identify vulnerable groups that would be most severely affected by the
shock, recognize the contribution of relief measures towards alleviating the
adverse effects, and inform the design of stimulus or recovery measures in the
short to medium terms.
3
The single-country recursive dynamic CGE model used for the study is an
adapted version of the Mitigation, Adaptation and New Technologies Applied
General Equilibrium (MANAGE) model (Van der Mensbrugghe, 2017). In
addition to the standard specifications of the MANAGE model, additional
features have been introduced. Public and private investments are distinguished.
A simple debt module was introduced to distinguish sources of government
expenditure finance. This module allows to capture changes in cost of financing
linked to interest payments on domestic and foreign debt.
The model is calibrated to a social accounting matrix (SAM) for the year 2010/11
for Ethiopia (Ahmed et al. 2017) and covers the period up to 2029/30. The SAM
has 68 production activities, 71 commodities, and 3 types of factors of production:
labor, land, and capital. There are 8 labor categories distinguished by education
level and rural-urban divide: the uneducated, and those with primary, secondary
or tertiary education allowing the distinction between the unskilled and semi-
skilled/skilled labor. There are 15 household categories distinguished by
income/consumption decile for the urban, rural farm and rural non-farm. Other
institutions include enterprises, the government, and the rest of the world. There
are several tax/subsidy accounts, including import tariffs, VAT, and other indirect
taxes as well as direct income taxes. The investment account in the SAM was
split to distinguish public and private investment.
Production is modelled using a nested Constant Elasticity of Substitution (CES)

structure in each sector. The labor market is segmented into three groups:
agricultural, non-agricultural, and public sector workers (public administration,
education, and health). Labor supply is a function of real wages for each category
of labor, and we assume partially flexible wages and labor supply. The model is
run for 20 periods, from 2011 (FY 2010/11) to 2030 (FY 2029/30).
The macroeconomic closure imposes that the government budget balance is

endogenous while government expenditure (recurrent and capital) is exogenous
and calibrated in the baseline to reflect past performances. As such, any extra
expenditure, or a reduction in revenue increases government deficit. This is the
preferred closure considering the fiscal situation of the country and the need for
prudent fiscal policy. Tax rates are also fixed at base year rates. Investment is
savings driven with exogenous propensity to save for households and firms.
4
Regarding the external sector, we assume exogenous foreign savings in foreign

currency calibrated to match historical data and projections. The nominal
exchange rate is fixed. The real exchange rate adjusts to maintain the current
account balance.
The baseline scenario was calibrated using Balance of Payments, fiscal, and other
macro data for the fiscal year 2010/11 to present. Projections were used for the
period 2019/20 to 2029/30. The sources of data include Central Statistical
Authority (CSA), Ministry of Finance (MOF), National Bank of Ethiopia (NBE),
World Bank and International Monetary Fund (IMF).
2.2. Transmission channels and scenarios
2.2.1. Transmission Channels

The pandemic generates shocks that affect the economy in multiple fronts (Figure
1). In particular, the pandemic creates economic shocks and involves feedbacks
among different economic actors. In normal circumstances, as the circular flow
diagram indicates, the economy continues running only when the money keeps
flowing around the circuit (Baldwin and di Mauro, 2020). But COVID-19
disrupts the normal flow which causes a slowdown everywhere. International
transmission mechanisms considered include remittances, export demand, import
supply, foreign direct investment, and oil prices. Internal factors include labor
supply and labor productivity, productivity of capital, and total factor
productivity. Finally, the increase in transaction costs both on the domestic and
international markets is incorporated. The impact of the shock is measured in
terms of the national accounts, the fiscal framework, the external accounts, the
labor market, and household welfare. The distributional effect is captured through
the effect on key economic sectors and household groups distinguished by income
quintiles.
The red crosses (see Figure 1) show where the shocks can potentially disrupt the
flow of economic activities. For example, households who do not get paid may
experience financial distress or even bankruptcy, as seen in many countries that
have experienced pandemic outbreak. This reduces spending on goods and
services, and this also affects the demand for imports.
5
Labor layoffs, sick leaves, quarantines, or stay-at-home is another strike area.

When workers lose their jobs, they tend to cut back spending on less necessary
items, which leads to a decline in domestic demand.
Note that a decline in demand and/or direct supply shocks can lead to a disruption
in international and domestic supply chains. Both lead to further contraction in
output, more so in the manufacturing sectors. Firms and households will be forced
to postpone investment (due to wait-and-see behaviour of people and firms-
expectations). In addition, many businesses may experience reductions in
cashflow, if not supported, may lead to business bankruptcies (Benassy-Quéré,
2020). This sort of business closures creates further disruptions in the economy.
For example, the financial sector will not get paid, putting the financial sector
under stress.
Figure 1: COVID-19’s multiple strikes in the circular flow of income

diagram
Source: Baldwin (2020)
6
2.2.2. Identifying shocks

The shock has been divided into two major channels: domestic and external. The
domestic transmission channels include reduced labor market participation, lower
productivity, and rising domestic trade costs. External channels include higher
international trade costs, a drop in export demand, lower import supply, a drop in
FDI and remittances, and lower import price of oil. The remainder of this section
provides some details on the individual transmission channels and the underlying
assumptions.
Labor market participation: There is a general reduction in labor supply

associated with the virus as workers are unable to work due to movement
restrictions, stay-at-home directives, fear, business closures, and reduced capacity
operation. It is also likely that there is a permanent reduction in the workforce due
to coronavirus-triggered deaths. The scope of the shock was based on ILO (2020)
and IMF (2020) estimations ranging from 5 to 8 percent reduction of working
days. Working hours are estimated to decline by 4.9 percent in Africa in the
second quarter of 2020 according to ILO (2020) estimates. The IMF African
Economic Outlook of April 2020 estimates that between 5 and 8 percent of
working days could be lost.
Productivity: Because of idle capacity in the economy and due to supply chain
disruptions, both in the domestic market and globally, there will be a reduction in
sectoral and economy-wide productivity. Given that some sectors will be hit hard
by the pandemic, sector-specific shocks need to be introduced to better capture
the sensitivity of the economy to disruptions to key exports, including horticulture
(flower farms), tourism, transportation, labor-intensive manufacturing industries,
and other services sectors such as hotels. A recent study has estimated factor
productivities for the Ethiopian economy (Kidanemariam et al., 2020). The
results are reported for labor and capital productivity between 2000 and 2014.
The Ethiopian economy registered the lowest productivities in 2002 and 2003,
due to drought which slowed down in economic activities. The pandemic is
expected to lead to even a larger reduction in productivity than the drought
because it affects all sectors directly or indirectly.
Foreign direct investment: Not only foreign direct investment but also domestic
investment is likely to decline during the pandemic. FDI monthly data from the
7
National Bank of Ethiopia indicates a downward trend with FDI inflow declining
from US$ 2271.3 million (ETB 63.8 billion) in first eight months of FY 2018/19
to US$ 1721.2 million (ETB 51.6 billion) for similar months of 2019/20.
Although this reduction in FDI occurred before the spread of the pandemic, it is
a good indicator of the extent of the shock. A large reduction in FDI is expected
ranging from 24 to 70 percent.5
Remittances: Due to the global nature of the pandemic and the slowdown of
economic activity in the Ethiopian migrants’ host countries, remittance flows are
expected to decline. Monthly data from NBE show that remittances has shown a
declining trend between January and March 2020, reduced by 22.7 percent. This
decline occurred during the early period of the outbreak of the pandemic
indicating that the scope of the reduction could be much higher. Hence two
scenarios are considered: a lower remittance inflow with a 25 percent reduction
and large reduction of 70 percent.
World price of oil: Because of shrinking global demand, world price of oil has
been significantly affected. For net oil importing countries like Ethiopia, this
would be a positive shock. As prices are regulated, lower prices are not
necessarily reflected on the consumer price but rather generate income to the
government. Crude oil prices have drastically declined since the outbreak of the
pandemic in 2020, decreasing by more than 60 percent. The decline in oil price
led to a fall in natural gas prices, by one-third compared to the levels a year ago.6
Transaction costs: Higher trade transaction costs can be expected due to border
closures, delays due to slowdown in logistics, quarantines, movement restrictions,
and supply chain disruptions. This is modelled using the ‘iceberg’ approach. The
size of the shock is calibrated such that the anticipated cost increase (between 25
and 50 percent) is applied to the trade and transport margins on domestic,
imported and exported goods. The resulting reduction in exports, imports and
domestic supply are then used as reference to calibrate the corresponding
transaction cost.
5
UNCTAD also estimates a large reduction in FDI across the world.
6
https://www.power-technology.com/comment/covid-19-outbreak-impact-fossil-fuel/
8
Trade: The pandemic is affecting the Ethiopian economy through weak external
demand and the disruption of import supply chains. Export demand is reducing
because of the slowdown among the key trading partners of Ethiopia. The flower
sector, the textile and apparel manufacturing, and tourism have been severely
affected by the reduction in international demand.7 The supply of imported goods
is expected to decline due to the disruption of global value chains. This affects
the supply of intermediate inputs possibly creating shortages. The scope of the
implemented shock is 25 percent reduction.
Depending on the anticipated severity, the magnitude of the shock will be

different for each transmission mechanism. Table 1 presents the scenarios and the
amplitude of the shocks for each channel. The scenarios are all implemented
according to the anticipated duration of the crisis, that is six months. All the
shocks are implemented simultaneously.
2.2.3. Scenarios
Three scenarios are considered to capture the economy-wide effects of COVID-
19.
Scenario 1: Business as usual scenario which serves as a base run scenario to

which other scenarios are compared8.
Scenario 2: Rapid containment of the virus with limited community spread,

targeted restriction of movements or partial lockdown resulting in limited
disruption to major economic activities. This scenario assumes that the adverse
effects of the virus will be mitigated through rapid response mechanisms with
fewer cases, including fiscal, financial, and other mechanisms. For instance,
enterprises or firms may proactively take measures to contain the spread of the
virus in the workplace through provision of personal protective devices,
awareness creation, rescheduling work hours, etc. This will lead to a quick
resumption of economic activities across most sectors as workers will return to
their work place. In this scenario, the economy will experience mild productivity
7
Recently, the flower export has shown an increasing trend (in May).
8
We refer to this scenario as baseline, reference, pre-Covid-19 or as no-Covid-19
scenario.
9
loss, limited unemployment, etc. Hence, the economic and social repercussions
are kept in check. The international transmission channels are mildly affected
with FDI, export demand, import supply and remittances declining by about one-
fourth.
Scenario 3: Relatively widespread community transmission and more

geographic spread of the infections. The third scenario, severe COVID-19
situation, assumes an outbreak of the virus across the country. Accordingly, the
economic and social impacts are expected to be more severe. In this scenario,
rapid spread of the virus will hamper movement of people and goods due to travel
bans, business closures, stay-at-home, etc. The economy is assumed to experience
multiple and interlinked supply, demand and financial shocks, causing slowdown
of economic activity, unemployment, supply shortage, low productivity, idle
capacity, etc. International transmission channels are severely affected.
These scenarios are implemented under two assumptions linked to the severity of
the impact:
i. Mild impact:
a. Duration of the shock is six months from the last quarter of EFY
2019/20 lasting throughout the first quarter of EFY 2020/21; and
b. There is no recovery of the shock variables to their pre-COVID-19
level but variables resume their pre-COVID-19 growth rates starting
from 2021/22.
ii. Severe impact: replicates the same shocks as the mild scenario in the last
quarter of FY 2019/20 and more severe effects in the first quarter of EFY
2020/21.
The distinction between the mild and severe scenarios is linked to the anticipated
duration of the crisis and the magnitude of the effect on the transmission channels
of the crisis. The two scenarios are analysed according to several variants:
i. Analysis of results with and without considering any government relief
measures or recovery with various financing sources;
ii. Short-term and medium-to-long term effects; and
iii. Rapid recovery versus deeper and more prolonged pandemic.
10
The government COVID-19 response plan is around 3.4 billion USD. While 1.6
billion is planned for emergency response, 1.8 billion USD is for macroeconomic
interventions. The following interventions have been incorporated into the
scenario:
i. Health interventions of 439 million USD (15 billion ETB): this
emergency relief measure is implemented as a recurrent expenditure in
the EFY 2019/20.
ii. Support to businesses and micro, small and medium enterprises
(MSMEs) of 1000 million USD (33.4 billion ETB): the planned support
to businesses encompasses several types of measures aimed at enabling
large firms and MSMEs to survive this crisis. The financial support
includes postponement of debt reimbursement, corporate income tax
exemption, and expedited VAT reimbursement for larger firms. MSMEs
are likely to benefit from financial support to reduce their operational
costs, increase their liquidity, and maintain employees. These measures
are implemented in the form of production subsidies to all the economic
activities except public administration, education and health sectors. The
time frame implemented in the scenario is such that 5 billion ETB is spent
in EFY 2019/20 and the remining in EFY 2020/21.
iii. Protecting jobs through a wage support or subsidy of 328.8 million USD
(10.8 billion ETB). This intervention is planned to benefit manufacturing
and services sectors. The allocation of the fund is based on government
calculations whereby three months of salary would be covered.
Accordingly, we inject 7.2 billion ETB in construction, 1.5 in services,
2.1 in manufacturing sectors. As no information is provided on the
timeframe, we allocate the equivalent of one month in EFY 2019/20 and
two third in EFY 2020/21.
iv. Food and other related emergency costs of up to 1,170 million USD. This
expenditure is aimed at protecting the welfare of populations, in
particular the most vulnerable ones. The government plans to use existing
urban and rural safety net programs to provide income to the most
affected. It also includes emergency food assistance. We implement this
by introducing cash-transfers to both rural and urban households in
quintiles 1, 2 and 3. This way, the 40 percent poorest or most vulnerable
are targeted. In order to decide on the amount allocated to each group, we
use the CGE model simulation results by running the mild scenarios
without any government intervention in the mild case. This allowed us to
11
quantify the losses in EFY 2019/20 and 2020/21. For households in

quintiles 1 and 2, a cash-transfer from the government is implemented
representing half of their consumption budget losses in real terms.
Households in quintile 3 are compensated up to 25 percent of their
consumption losses.
When we consider government interventions, the sources of financing of these
responses or measures are important. The loss of government revenue will already
increase the fiscal deficit without any additional spending. Deficit financing is
likely to have significant crowding-out effect on private investment. The
alternative is a mix of deficit financing amounting 5 billion ETB and external
assistance in the form of grant for other expenditures.
The medium- and long-term potential effects are captured by comparing the
baseline scenario (no-COVID-19) with the COVID-19 scenario under four
assumptions:
i. No government intervention: there are no additional expenditures;
ii. Relief measures through deficit financing: expenditure increases
according to the above-mentioned relief measures;
iii. Relief measures through development partners’ assistance and some
deficit financing; and
iv. Relief measures with the above mix of financing and progressive
recovery of the economy towards its pre COVID-19 level within a period
of three years starting from EFY 2021/22. This is inspired by the
convergence of growth rates of Ebola-affected countries between 2014
and 2018.
After the 2008 global financial crisis, economic growth in Ethiopia slowed down
to 8.6 percent in FY 2008/09 from 10.7 percent in FY 2007/08. GDP growth
recovered to 12.4 percent in FY 2009/10 and 11.4 the following year. In variant
(ii) and (iii), the recovery of GDP growth towards the pre-COVID-19 growth rate
in our medium-to-long-term scenarios is informed by this. In the mild case, we
assume a rapid recovery. GDP growth rate converges to the no-COVID-19
baseline. The shock variables resume their pre-COVID-19 growth rates starting
from FY 2021/22. In the severe scenario, we assume that the shock variables do
not immediately recover their pre-COVID-19 growth rate but increase to the point
that GDP growth is about half of the 9 percent no-COVID-19 baseline rate.
12
The fourth variant of the medium-to-long-term scenario allows the shock

variables to fully or partially recover their no-COVID-19 level up to the point
where GDP growth remains within the range of past performances, that is, not
exceeding 12.4 percent. This implies that although economic growth could
surpass the baseline rates, it will not reach the level that would allow the GDP
loss to be recovered. In the mild case, we assume that recovery takes place from
FY 2021/22 to 2023/24. In the severe case, it is calibrated to begin in FY 2021/22
where the shock variables reach their no-COVID-19 growth rates. From FY
2022/23 to 2024/25, the shock variables fully or partially recover their no-
COVID-19 level up to a GDP growth rate not exceeding 11.3 percent. In both
cases, after the recovery period, the baseline growth rates are applied.
13
Table 1: Operationalizing Scenarios

Mild scenario Severe scenario Remarks
Business as This is the No- This is the No-COVID-19
usual COVID-19 scenario scenario
Labor market Loss of 5% of working 5% reduction of working ILO 2020 and
participation days days IMF 2020
Productivity growth
- Labor -2.79 % -6.41% Kidanemariam
- Capital -2.15% -2.61% et al. (2020)
FDI -24.1% -70%
Remittances -25% -60%
Export 25% reduction in export 50% reduction in export Informed by
demand demand for Ethiopian demand for Ethiopian review of
goods and services goods and services and recent trade
No reduction for coffee 25% reduction for coffee statistics.
Import There is no reduction 25% reduction of import
supply in import supply supply
Transaction Higher transaction Higher transaction costs Motivated by
costs costs on domestic, on domestic, exported, early
exported, and and imported goods anecdotal
imported goods (25%) (50%) reports (e.g.
Tamiru,
Engida and
Minten,
2020).
International Lower import price of Lower import price of oil
oil prices oil by 33% by 33%
Variant 1- deficit Variant 1- external
financing financing
Government Health interventions: 439 million USD (15 billion MOF 2020
interventions ETB) in EFY 2019/20
Food and other related emergency costs: 1,170
million USD
Support to businesses and MSMEs: 1,000 million
USD (33.4 billion ETB) with 5 billion ETB in EFY
2019/20 and 28.4 billion ETB in EFY 2020/21
Protecting jobs: 328.8 million USD (10.8 billion
ETB) with 1/3 in EFY 2019/20 and 2/3 in EFY
2020/21
BoP Financing: 500 million USD
14
3. Simulation results: Main findings
The impact of the pandemic is analysed in two steps. In the first part, we analyse
the short-term effects, that is, in FY 2019/20 and 2020/21. In the second part, we
analyse medium to long term effects. To distinguish the role of government
interventions in mitigating the adverse effects of the COVID-19, the reader can
refer to the tables in the Annex to compare the results with and without
government interventions. We focus our analysis on the results with government
relief and recovery measures as described in Table 1.
3.1. The short-term effects with government interventions
3.1.1. Growth effect

The drivers of growth in this model are key to understanding the growth effects
of the COVID-19. The level of GDP in our analytical framework is determined
by three factors: (i) the supply of workers, (ii) investment, and (iii) productivity.
Investment is driven by savings from domestic and foreign sources. Labor supply
is one of the transmission channels of the COVID-19 and is affected by the
containment measures. It is also affected by lower labor demand as economic
activity slows down. This will have a direct impact of GDP. Productivity, another
transmission channel, is exogenous in the model and assumed to be lower, hence
having a direct effect on GDP.
Our estimates suggest that the economic cost of the COVID-19 pandemic
could be significant in the short term. GDP contracts in the short term with
higher effects as the length of the duration of the pandemic increases from three
to six months9. GDP is 5.9 percent lower in FY 2019/20 and contracts by 6.7
percent in FY 2020/21 compared to the baseline scenario (Figure 2). This
represents 127 and 159 billion ETB loss of GDP in FY 2019/20 and 2020/21,
respectively. The size of the Ethiopian economy shrinks significantly more in the
severe scenario. With 13.1 percent deviation from the baseline, GDP is smaller
by 310 billion ETB (Figure 3).
9
The first three months represent the last quarter of FY 2019/20. A six months duration
extends the time frame to the first quarter of FY 2020/21.
15
Figure 2: Growth effect – mild case (in level and % deviation from baseline)
-5.0 0
-127 -159
-5.5 -50
-6.0 -100
-6.5 -5.9 -150

-6.7
-7.0 -200
2020 2021
Reduction in Bilion ETB (right axis) GDP (deviation from baseline)
Figure 3: Growth effect – severe case (in level and % deviation from
baseline)
0.0 0
-2.0 -127 -310 -50
-4.0 -100
-6.0 -5.9 -150
-8.0 -200
-10.0 -250
-12.0 -300
-13.1
-14.0 -350
2020 2021
Reduction in Bilion ETB (right axis) GDP (deviation from baseline)
127 billion ETB translates into a 2.6 percent economic growth in FY 2019/20.
The following year would see the economy growing by 8 percent in the mild
scenario (Figure 4). With a higher degree of contraction in economic activity in
FY 2020/21, GDP would grow by 0.6 percent in the severe case while the
economy was estimated to grow by 9 percent in the No-COVID-19 scenario
(Figure 5). This poor GDP performance is reflected in the drop in exports,
investment and final consumption.
16
Figure 4: GDP annual growth rate – mild case

10.0
8.0 9.0 9.0

8.0
6.0
4.0
2.0 2.6
0.0
2020 2021
No-COVID-19 Post-COVID-19
Figure 5: GDP annual growth rate – severe case

10.0
8.0 9.0 9.0
6.0
4.0
2.0 2.6 0.6

0.0
2020 2021
No-COVID19 Post-COVID19
Private investment would be significantly affected via direct and indirect

effects. Investment is affected through three channels. FDI is lower, fiscal deficit
creates a crowding-out, and savings have declined reducing the total investment.
Public investment is maintained exogenous in real terms and only varies with the
investment price index. Investment is 5.4 percent lower in FY 2019/20 and 5.9
percent in FY 2020/21 in the mild case (Figure 6). Under the amplified pandemic
scenario, investment is 18.3 percent lower FY 2020/21 (Figure 7).
17
Figure 6: expenditure on GDP at purchaser’s price – mild case (%

deviation from baseline)
10.0 7.3
0.0 -5.4-5.9 -16.4 -15.3 -3.9 -4.8
0.0 -4.6 -5.3
-10.0
-20.0
Private Public Private Exports Imports
consumption consumption Investment
2020 2021
Figure 7: Expenditure on GDP at purchaser’s price – severe case (%

10.0 7.3
-4.6 -12.0 0.0 -5.4 -18.8 -16.4-25.6 -3.9-16.9
0.0
-10.0
-20.0
-30.0
Private Public Private Exports Imports
consumption consumption Investment
2020 2021
The impact on trade is likely to be significant. Under both scenarios, exports,

and imports decline. Exports are affected by lower international demand, higher
transaction and transport costs, and possible disruptions of supply due to
containment measures. Exports are 16.4 percent lower in FY 2019/20 and decline
by 15.3 and 25.6 percent in FY 2020/21 in the mild and severe cases, respectively
(Figures 6 & 7). Imports are linked to three factors: higher transport and
18
transaction costs, disruption in international supply, and reduction in domestic

economic activity. With exogenous current account in foreign currency, imports
would decline by 3.9 percent in FY 2019/20 and by 4.8 and 16.9 percent in FY
2020/21 in the mild and severe cases, respectively (Figures 6 & 7).
Private consumption is likely to be negatively affected by the crisis.

Consumption level is 4.6 percent lower in FY 2019/20 and reduced by 5.3 percent
in FY 2020/21 in the mild scenario compared to a 12 percent loss in an amplified
pandemic scenario (Figures 6 & 7). When we compare private consumption with
and without government relief measures, we find that without relief measures,
consumption levels would have been 1.2 percentage points lower in FY 2019/20
and 1.1 and 1.2 percentage points lower in the mild and severe cases, respectively
in FY 2020/21.
Our estimates are in line with previous studies albeit higher effects. The IMF
projections in the most recent African Economic Outlook (April 2020) suggests
that real GDP would grow by 3.2 percent in FY 2019/20 and by 4.3 percent in FY
2020/21 from a pre-COVID-19 projection of 9 percent. The assessment of the
socio-economic impact of the COVID-19 in Ethiopia by the United Nations
Country Team suggests real GDP growth to range between 7 to 4.23 percent in
an optimistic case to 5.4 to 2.23 percent in a pessimistic case in 2020.
Government expenditure to strengthen the health sector, to protect the most

vulnerable populations, to protect jobs, and to support businesses contributes to
reducing the adverse effects of the COVID-19 pandemic on the Ethiopian economy.
Although 3.4 billion USD, equivalent to 3.5 percent of the pre-COVID-19 GDP, is
allocated, it is not enough to bring about notable improvements in economic growth.
3.1.2. Fiscal effect

The COVID-19 pandemic is likely to have a substantial effect on public
finances. The pandemic is expected to increase the fiscal deficit with varying
degrees depending on the level of development partners’ assistance. The fiscal
effect is driven by revenue loss and higher level of expenditure. Without external
assistance and for the same level of expenditure as in the no-COVID-19 scenario,
government revenue declines by of 2.5 percent in FY 2019/20 and by 3.3 percent
in FY 2020/21 in the mild case and reaching 10.4 percent in the severe case. With
19
the 3.4 billion USD response plan mainly financed through external assistance,
fiscal deficit would be 0.6 percent higher than the baseline in FY 2019/20 (Figures
8 & 9). When we consider financing the COVID-19 plan through borrowing,
public deficit as a share of GDP would be 2.5 percentage points higher than in the
baseline (Figure 26). In FY 2020/21, fiscal deficit would increase by 1.3 percent
in the mild case and by 7.3 percent higher in the severe case compared to the
deficit in the baseline (Figures 8 & 9). The impact on government debt is
negligible in the short term.
Figure 8: Public finance – mild case (% deviation from baseline)

4.0 3.7
2.0 1.3
0.6
-2.2 0.0 0.1
0.0
-2.0
-4.0
Overall balance Government revenue Public debt
2020 2021
Figure 9: Public finance – severe case (% deviation from baseline)

10.0 7.3
5.0 3.7
0.6 -9.3 0.0 0.1
0.0
-5.0
-10.0
-15.0
Overall balance Government revenue Public debt
2020 2021
20
3.1.3. Sectoral effect

The COVID-19 pandemic is likely to have adverse effects on every key sector
of the Ethiopian economy. Not all sectors are equally affected by the crisis with
higher contraction of manufacturing activities followed by agriculture. The
sectoral impact transmits through supply (of domestic and imported inputs and
factors) and demand (from domestic and export markets) effects. Production and
consumption linkages drive the direct and indirect effects at sectoral level. The
production activities are linked to each other through forward and backward
linkages. Backward linkages imply that a reduction in demand for intermediate
inputs will affect the activities producing these goods and services10. Through
forward linkages, a disruption in the of supply of intermediate inputs (and/or
higher prices) will affect production of sectors that are using these inputs.
Consumption linkages operate in such a way that lower factor income (labor, land
and capital) leads to lower final consumption demand. Labor supply and labor
productivity are two other important transmission channels. While the loss of 5
percent working days affects all the sectors similarly, activities with relatively
higher labor-intensity are likely to be more affected. Finally, tradable sectors with
a high degree of openness are hit hardest.
Manufacturing output declines by 7.4 percent in FY 2019/20 with relatively small

increase the following year (Figure 10). Agricultural production declines by 7.4
percent and 7.6 percent in FY 2019/20 and 2020/21 in the mild case. With an
amplified pandemic scenario, the reduction of agricultural output doubles
reaching 13 percent (Figure 11). Services decline by 10.2 percent in FY 2020/21
in the severe case. The effect on agriculture is much larger in the severe scenario
because we assume a disruption in import supply and a reduction in export
demand for Ethiopian coffee. This is on top of an amplification of the other
transmission channels (Table 1). Fertilizers and other chemicals used in
agricultural production are mainly imported. A disruption of supply would have
consequences on production.
10
For instance, agricultural production uses agricultural products (such as seeds),
manufacturing inputs (essentially consisting of fertilizers and chemicals) and services
(mainly financial and transport services). A reduction in agricultural activities will affect
services sectors more than the manufacturing because manufacturing inputs used in
agricultural production are imported. Similarly, a reduction in services such as hotels and
restaurants will affect demand for agricultural products.
21
Figure 10: Aggregate sectoral production – mild case (% deviation

from baseline)
Services -5.4
-4.1
-7.6
Manufacturing
-7.4
-7.2
Agriculture
-6.9
-8.0 -7.0 -6.0 -5.0 -4.0 -3.0 -2.0 -1.0 0.0
2021 2020
Figure 11: Aggregate sectoral production – severe case (% deviation

from baseline)
-10.2
Services
-4.1
-8.2
Manufacturing
-7.4
-13.0
Agriculture
-6.9
-15.0 -10.0 -5.0 0.0
2021 2020
22
At a disaggregated level, key economic sectors are likely to contract at

varying degrees. The exception would be the public administration, education
and health sectors because of the expenditures linked to health interventions in
FY 2019/20. Sector-specific effects are driven by the composition of output and
exports. Export intensive industries such as the textile and leather, export-oriented
agriculture, transportation and storage, and accommodation and food services are
hit the hardest (Figures 12 & 13). The coffee sector was not heavily affected in
FY 2019/20 because we assumed, in the scenarios, that export demand would not
be affected in FY 2019/20. As soon as we introduce a reduction in export demand
for Ethiopian coffee, the sector would be adversely affected. The higher the
amplitude of the shock, the greater the output contraction across all sectors. The
construction sector suffers from reduction in investment.
Figure 12: Sectoral production – mild case (% deviation from baseline)
Public adm., education, health

Other services
Accommodation and food services
Transportation and storage
Wholesale and retail trade
Construction
Electricity & water
Other manufacture activity
Textile and leather
Agro industry
Mining and quarrying
Livestock and livestock products
Coffee
Agriculture cash/export crops oth.
Agricultural food crops
-20.0 -15.0 -10.0 -5.0 0.0 5.0 10.0

2021 2020
23
Figure 13: Sectoral production – severe case (% deviation from baseline)

Other services
Construction
Electricity & water
Textile and leather
Agro industry
Coffee
-25.0 -20.0 -15.0 -10.0 -5.0 0.0 5.0 10.0

2021 2020
The impact on the agricultural food crops sectors is likely to have implications
for food security. If we consider availability of food as an indicator of food
security measured by nationwide food supply, lower output in food crops may
indicate a possible food deficit. Food availability also indicates a reduction of 3.1
percent in FY 2019/20. Food imports are 3.4 percent lower in FY 2020/21 in the
mild case and 17.3 percent lower in the severe case. Food availability also
indicates a reduction of 3.1 percent in FY 2019/20 due to a combination of
reduced food production and imports.
3.1.4. Employment effect

Employment is likely to take a big hit in line with the sectoral effects. The
employment level is between 8.6 percent and up to 16.5 percent lower than the
baseline. Employment losses are significant with higher reduction in
manufacturing sectors (Figures 14 & 15). With a longer duration of the pandemic,
the employment loss does not increase notably in the mild scenario. However,
with higher amplitude of the shock, the employment effect would be severe in all
the export-oriented sectors (Figure 15).
24
Figure 14: Employment by sector – mild case (% deviation from baseline)

Other services
Construction
Electricity & water
Textile and leather
Agro industry
Coffee
-30.0 -20.0 -10.0 0.0 10.0

2021 2020

Figure 15: Employment by sector – severe case (% deviation from baseline)
Other services
Construction
Electricity & water
Textile and leather
Agro industry
Coffee
-30.0 -20.0 -10.0 0.0 10.0

2021 2020
25
Employment in rural sectors is slightly more affected than urban employment

(Figures16 & 17). The impact on rural employment goes along with the
contraction in agricultural activities and the reduced activity in other rural non-
agricultural sectors. Agriculture is the economy’s largest employer. A significant
proportion (72.6 percent) of the labor force works in agriculture and related
activities. Furthermore, the government measures introduced in the form of wage
support target sectors outside of agriculture. The loss of employment in the
agricultural sectors would affect wage laborers and not family workers. Although
it is not modelled, there are signs of urban to rural migration because of the
slowdown in urban economic activity. There are already signs of urban to rural
migration (see Degye et al., 2020). This is likely to increase the volume of the
rural unemployed.
The unskilled, who constitute most of the work force, are more impacted than the
skilled and semi-skilled. Employment of the unskilled is 9.3 percent lower in FY
2019/20 reaching 10.2 and 17.3 percent in the mild and severe scenarios,
respectively, in FY 2020/21 (Figures 16 & 17). The skilled are the least affected
by the crisis.
Figure 16: Employment by skill and location – mild case (% deviation

from baseline)
Urban -9.5
-7.8
Rural -9.7
-8.9
Total -9.7
-8.6
Skilled -2.7
0.2
Semi-skilled -3.3
-0.9
Unskilled -10.2
-9.3
-12.0 -10.0 -8.0 -6.0 -4.0 -2.0 0.0 2.0
2021 2020
26
Figure 17: Employment by skill and location – severe case (% deviation

from baseline)
Urban -15.6
-7.8
Rural -16.8
-8.9
Total -16.5
-8.6
Skilled -5.0
0.2
Semi-skilled -6.0
-0.9
Unskilled -17.3
-9.3
-20.0 -15.0 -10.0 -5.0 0.0 5.0
2021 2020
3.1.5. Welfare effect

Income loss is likely to result in welfare reduction for all households.
Household welfare at aggregate level, measured by real consumption, is 4.6
percent lower than in the reference scenario in FY 2019/20 (Figure 18). In FY
2020/21, the impact is 5.3 percent and 12 percent lower in the mild and severe
cases, respectively (Figures 18 & 19). The rural households are likely to have
higher consumption losses mainly due to the heavy toll on the rural farm
households. Without government intervention simulated through cash transfers to
the most vulnerable (and other measures), welfare in FY 2020/21 is lower than in
the reference scenario by 6.4 percent and 13.2 percent in the mild and severe
cases, respectively (Figures A1 & A2 in Annex). The rural farm households are
the most affected group due to their source of livelihood, which heavily relies on
agriculture.
27
Figure 18: Aggregate household consumption – mild case (% deviation

from baseline)
0.0
-1.0
-2.0
-3.0
-3.1
-4.0
-3.9
-4.3
-5.0 -4.6 -4.7
-4.9 -5.1
-5.3 -5.3
-6.0 -5.6
Total Rural farm Rural non- Rural Urban
farm
2020 2021
Figure 19: Aggregate household consumption – severe case (%

0.0
-2.0
-4.0 -3.1
-4.6 -4.7 -4.3
-6.0 -4.9
-8.0
-8.2
-10.0
-10.8
-12.0
-12.0
-12.9 -12.5
-14.0
Total Rural farm Rural non- Rural Urban
farm
2020 2021
28
Without targeted government support, the COVID-19 crisis is likely to be

regressive, the poorest suffering the most across the rural farm and non-
farm and the urban households. The simulation results that do not incorporate
government expenditures targeting the 40 percent poorest demonstrate the
harmful effect on the most vulnerable populations. The lower the income level,
the higher the welfare loss, where the slightest welfare loss is likely to be
detrimental. We find that the welfare loss of the bottom 20 percent is 1.6, 2.5, and
2.2 times higher than the top 20 percent for the rural-farm, the rural non-farm and
the urban, respectively, in FY 2019/20 in the mild scenario (Figures A3 & A4 in
Annex). The trend is similar in the following fiscal year and in the severe case.
If the provision of in-kind and cash transfers target the 40 percent poorest,
the disproportionate welfare loss of the poorest could be prevented. Our
results suggest that by compensating up to half of their estimated reduction in
consumption in FY 2019/20 and 2020/21, the disproportionate welfare loss would
be minimised (Figures 20 & 21). Consumption would be 5.3 percent lower for
the rural farm instead of 6.9 percent in FY 2020/21 in the mild case and decline
by 12.3 percent instead of 14.7 percent in the severe case.
Figure 20: Household consumption by decile – mild case (% deviation

from baseline)
0.0
-1.0
-2.0
-3.0
-4.0
-5.0
-6.0
-7.0
-8.0
-9.0
2020 2021
29
Figure 21: Household consumption by decile – severe case (% deviation

from baseline)
0.0
-2.0
-4.0
-6.0
-8.0
-10.0
-12.0
-14.0
-16.0
2020 2021
3.2. Prospects in the medium-to-long term
There is much uncertainty on the future of economic activity in the post-COVID-

19 years. We assess the medium-to-long-term effects based on four assumptions:
i. No government intervention;
ii. Relief measures combined with lower import prices of oil financed via
higher deficit;
iii. The above relief measures are financed through assistance from
development partners and deficit; and
iv. Relief measure with the above mix of financing sources and progressive
recovery of the economy within a period of three years from EFY
2021/22.
The key messages are summarized in the following points:

- The COVID-19 is likely to have medium-to-long-term effects.
- Although adverse effects do not amplify, the major macro indicators
remain below their pre-COVID-19 level.
30
- The higher the scope of the impact, the greater the gap with the baseline
(no-COVID-19) scenario.
- The current government intervention is not enough to put the economy
on a higher growth path (no-COVID-19) if it does not go beyond FY
2020/21.
The effect of the COVID-19 crisis is likely to remain negative in the coming
years if a progressive recovery of the shock variables (transmission channels)
does not take place. In this highly pessimistic scenario, labor supply,
productivity of labor and capital, remittances, FDI, export demand and imports
supply, would resume their baseline (or no-COVID-19) growth rates from FY
2021/22 onwards. This situation is represented by the green and yellow colours
(Figures 23 and 24). The three-year progressive recovery is represented by the
red line. The orange columns represent the situation without government
intervention. The blue colour represents the baseline or no-COVID-19 scenario.
Even if GDP growth recovers its pre-crisis 9 percent growth rate, the size of
the Ethiopian economy would remain well below the no-COVID-19 baseline
level. Without any recovery of the transmission channels of the shock, the size of
the Ethiopian economy would remain well below the baseline level (Figures 22
& 23). A progressive recovery is not likely to allow the economy to reach its pre-
COVID-19 size. GDP would be lower by 106 billion ETB than the baseline in the
mild case in FY 2029/30 and by 610 billion ETB lower in the severe case. By the
end of the time horizon of our model, 2029/30, and without recovery, GDP would
be lower than that of the baseline (no-COVID-19) by 489 billion ETB in the mild
case and by 1,259 billion ETB in the severe case (Figures 22 & 23).
31
Figure 22: GDP level and loss (right axis) -Mild case Billion ETB)
0
4,000 -200
-400
2,000 -600
2020 2021 2025 2030
Loss No Gvt. Intervention Loss Gvt. Int. External financing

Loss Gvt. Int. Deficit financing Loss Gvt. Int. External financing + recovery
No-COVID-19 No Gvt. Intervention
Gvt. Int. External financing Gvt. Int. Deficit financing
Gvt. Int. External financing + recovery
Figure 23: GDP level and loss (right axis) -Severe case Billion ETB)
0
4,000 -500
-1,000
2,000 -1,500
2020 2021 2025 2030
Loss No Gvt. Intervention Loss Gvt. Int. External financing

Loss Gvt. Int. Deficit financing Loss Gvt. Int. External financing + recovery
GDP No-COVID-19 No Gvt. Intervention
32
With a three-year recovery assumption, GDP growth rate would attain the
pre-COVID-19 situation of 9 percent growth rate in the mid-term in the mild
case. In a pessimistic scenario, GDP growth would remain below the baseline
scenario in either the mild (8.7 percent) or severe (8 percent) cases in FY 2025/26
(Figures 24 & 25). In line with the severity of the shock, the recovery would have
a V shape in the mild case and a U shape in the severe case.
Figure 24: GDP annual growth rate – Mild case

12.0
10.0
8.0
6.0
4.0
2.0
0.0
2019 2020 2021 2022 2023 2024 2025 2026 2030
No Gvt. Intervention No-COVID-19

Figure 25: GDP annual growth rate – Mild case
12.0
7.0
2.0
-3.0
2019 2020 2021 2022 2023 2024 2025 2026 2030
No Gvt. Intervention No-COVID-19
Notes: Gvt.int.=Government intervention.

33
Regardless of the scenario, fiscal deficit is likely to increase in absolute terms

and in percentage of GDP. Deficit financing of the government-COVID-19
economic response plan would lead to a severe deterioration of the fiscal balance
(Figures 26 & 27).
Figure 26: Public deficit as share of GDP – Mild case (deviation from
baseline in percentage points)
4.0
2.0
0.0
2020 2021 2022 2023 2024 2025 2026 2030
No Gvt. Intervention Gvt. Int. External financing

Gvt. Int. Deficit financing Gvt. Int. External financing + recovery
Figure 27: Public deficit as share of GDP – Severe case (deviation from
baseline in percentage points)
8.0
6.0
4.0
2.0
0.0
2020 2021 2022 2023 2024 2025 2026 2030
34
Government response plan is of paramount importance for the medium- and

long-term investment. The impact of the pandemic on investment would be
larger without government intervention in the short, medium and long terms. The
role of the assistance from development partners is crucial at this time of crisis.
Both mild and severe scenarios clearly illustrate the importance of government
interventions supported by external assistance to avoid losing the gains from the
prudent fiscal management of the past years. This would also avoid medium- and
long-term investment reductions (Figures 28 & 29). Without government
intervention, private investment would be 1.9 and 1.4 percentage points lower
compared to the level with COVID-19 response plan in the mild and severe cases
respectively in FY 2025/26.
Figure 28: Investment – Mild case (% deviation from baseline)

0.0
-2.0
-4.0
-6.0
-8.0
-10.0
-12.0
2020 2021 2022 2023 2024 2025 2026 2030

35
Figure 29: Investment – Severe case (% deviation from baseline)

-5.0
-10.0
-15.0
-20.0
-25.0
-30.0
-35.0
2020 2021 2022 2023 2024 2025 2026 2030
Household welfare would fall more sharply without relief and recovery
measures. Relief measures are only effective in the short term. By the end of the
time horizon of our model, FY 2029/30, household welfare at the national level,
under the recovery scenario, would be 1.9 and 10.7 percent lower than the
reference scenario in the mild and severe cases, respectively (Figures 30 & 31).
This is likely to undermine the poverty reduction efforts with a risk of increasing
poverty incidence, especially among those households with income just above the
poverty line.
36
Figure 30: Household consumption – Mild case (% deviation from

baseline)
0.0
-2.0
-4.0
-6.0
-8.0
-10.0
2020 2021 2022 2023 2024 2025 2026 2030
Figure 31: Household consumption – Severe case (% deviation from

baseline)
0.0
-10.0
-20.0
-30.0
2020 2021 2022 2023 2024 2025 2026 2030
The poorest are likely to have higher welfare reduction in the medium to long
term. In-kind and cash transfers to the most vulnerable are likely to narrow the gap
37
between households in the lowest quintiles and those in the highest quintiles, in the
short term, in the severe case (Figures 32 and 33). However, this is not enough
because the gap is clearly visible between the poorest and the others in the medium
to long term. The rural urban divide is also likely to widen with the duration of the
crisis between FY 2019/20 and 2020/21 with larger gap if no recovery.
Even if the transmission channels of the COVID-19 to the Ethiopian economy

partly regain their pre-crisis level, households in the lowest income categories
will have the higher welfare reduction than the highest quintile. The gap between
the rural and urban would be slightly smaller than in a no recovery scenario. All
households would have consumption level that is lower than the no-COVID-19
scenario.
Figure 32: Quintiles 1 & 5 rural farm households’ consumption

(percent deviation from baseline)
0.0
-10.0
-20.0
-30.0
2020 2021 2022 2023 2024 2025 2026 2030
r f q1 no gvt. Int. r f q5 no gvt. Int.
r f q1 Int. ext. fin. r f q5 Int. ext. fin.
r f q1 gvt. Int. + recovery r f q5 gvt. Int. + recovery
38
Figure 33: Quintiles 1 & 5 urban households’ consumption (percent

0.0
-5.0
-10.0
-15.0
-20.0
-25.0
-30.0
2020 2021 2022 2023 2024 2025 2026 2030
u q1 no gvt. Int. u q5 no gvt. Int.

u q1 Int. ext. fin. u q5 Int. ext. fin.
u q1 gvt. Int. + recovery u q5 gvt. Int. + recovery
39
4. Conclusions and recommendations
The COVID-19 pandemic is likely to have significant growth and welfare effects
even under an optimistic scenario of mild shock and quick recovery. Fiscal deficit
is likely to widen in absolute terms and as the percentage of GDP. The COVID-
19 pandemic is likely to have adverse effects on every key sector of the Ethiopian
economy. Employment is likely to be hardly hit. Although there is much
uncertainty in the future, the COVID-19 crisis is likely to have medium-to-long-
term negative effects.
GDP growth rate is expected to converge to the no-COVID-19 baseline relatively

swiftly if the scope of the shock is mild. However, the GDP and welfare losses
are not likely to be fully recovered. In an amplified scenario, the economic and
welfare losses would be higher and the gap with the no-COVID-19 baseline
would be much greater.
There is much uncertainty on the duration of the pandemic in Ethiopia and

worldwide. This implies that recovery may not come as quickly as anticipated putting
the Ethiopian economy closer to the severe scenario rather than the mild case. The
role of the government is of paramount importance not only by increasing its
spending under the COVID-19 response plan, but by creating an enabling
environment that would support businesses to thrive and allow social safety nets
to share the burden.
The anticipated 3.4 billion USD response plan is meant for emergency responses
and to support businesses and protect jobs. Under the current allocation, this plan
would certainly contribute protecting the livelihoods of workers and businesses.
However, it is not enough to put the economy on a higher growth path that would
reduce that gap with the pre-COVID-19 projections. Furthermore, without the
assistance of development partners, deficit financing would result in the
deterioration of the fiscal framework with a risk of jeopardizing macroeconomic
stability and debt sustainability.
Shielding the most vulnerable groups is crucial to limit the populations that would
fall (or back) into poverty. Hence it is necessary to expand the existing social
40
assistance programs such as rural and urban safety nets to protect the most
vulnerable segments of the population.
Given the multifaced nature of COVID-19 induced challenges facing the country,
a recovery and response plan is urgently needed to achieve dual objectives of
mitigating further economic contraction and of stimulating the economy. The
recovery and response plan shall target and safeguard sectors essential for food
security, job creation and sustainable and inclusive growth. The plan needs to take
into account the differentiated impacts of the pandemic on different activities and
households.
Both fiscal and monetary policy instruments that have been introduced by the
government to fight the pandemic shall be continued, enforced and implemented,
in a coordinated way, to support the effectiveness of interventions until the
economy recovers. However, support measures need to be monitored and
evaluated to facilitate the transition towards economic recovery and boost
economic dynamism.
The recovery and response plan requires rapid and predictable financing which
shall come from two sources. It is important that the international community
provides quick and coordinated support. Given the global nature of the pandemic
and uncertainty of external finance in terms of timing and amount, it is necessary
to design a strategy for mobilizing and diversifying domestic sources of finance
to manage and support the crisis and recovery in a more dependable manner. It is
also necessary to evaluate tax related supports ongoing basis as across-the-board
tax reductions, deferrals and relief may jeopardize medium-term revenue raising
capacity of the government.
41
References
Ahmed, H.A., T. Tebekew and J. Thurlow (2017). 2010/11 Social Accounting

Matrix for Ethiopia: A Nexus Project SAM. International Food Policy
Research Institute, Washington DC.
Baldwin, R. (2020). “Keeping the lights on: Economic medicine for a medical
shock”, VoxEU.org, 13 March.
Degye et al. (2020). Assessment of COVID-19 Effects and Response Measures
in Ethiopia: Livelihoods and Welfare Implications, Policy Working
Paper 03/2020, Ethiopian Economics Association.
Dominique van der Mensbrugghe (2017). Mitigation, Adaptation and New
Technologies Applied General Equilibrium (MANAGE) Model Version
2.0f, Center for Global Trade Analysis, Purdue University.
FDRE (2020). Economic Impact Responses Assessment and Policy, Addis
Ababa.
International Labor Organization (2020). ILO Monitor: COVID-19 and the
world of work. Second edition Updated estimates and analysis. 7 April
2020.
Kidanemariam Berhe, Mulu Gebreeyesus, Tsegay Kidanemariam & Kenichi
Ohno. (2020). Ethiopia Productivity Report 20202. Policy Studies
Institute (PSI), Addis Ababa.
Tamru, S., Engida, E. and Minten, B. (2020). Impacts of the COVID-19 crisis
on coffee value chains in Ethiopia, ESS-IFPRI.
Triggs, A. and Kharas, H. (2020). The triple economic shock of COVID-19 and
priorities for an emergency G-20 leaders meeting. Available at:
https://www.brookings.edu/blog/future-development/2020/03/17/the-
triple-economic-shock-of-covid-19-and-priorities-for-an-emergency-g-
20-leaders-meeting/
United Nations Country Team in Ethiopia (2020). Socio-economic impact of the
COVID-19 in Ethiopia. One UN assessment. May 2020.
42
Annex 1 - The short-term effects without government interventions
Figure A1: Aggregate household consumption – mild case Figure A2: Aggregate household consumption –
(% deviation from baseline) severe case (% deviation from baseline)
0.0 0.0
-2.0 -5.0 -3.6

-5.8 -4.9
-6.4 -6.2
-4.0 -3.6 -10.0 -8.6
-4.2 -10.9
-4.9 -13.2
-6.0 -5.7 -15.0 -14.2
-5.8 -6.2 -14.7
-6.4 -6.4 -6.7
-6.9
-8.0 -20.0
Total Rural farm Rural non- Rural Urban Total Rural Rural non- Rural Urban
farm farm farm
2020 2021 2020 2021
Source: Simulation results Source: Simulation results
43
Figure A3: Household consumption by decile – mild Figure A4: Household consumption by decile – severe
case (% deviation from baseline) case (% deviation from baseline)
2020 2021 2020 2021
0.0 0.0
-2.0
-2.0 -4.0
-4.0 -6.0
-8.0
-6.0 -10.0
-12.0
-8.0 -14.0
-10.0 -16.0
-18.0
-12.0 -20.0
Rural farm q5
Urban q1
Urban q2
Urban q3
Urban q4
Urban q5
Urban q1
Urban q2
Urban q3
Urban q4
Urban q5
Rural farm q1
Rural farm q2
Rural farm q3
Rural farm q4
Rural farm q5
Rural farm q1
Rural farm q2
Rural farm q3
Rural farm q4
Rural n-farm q1
Rural n-farm q2
Rural n-farm q3
Rural n-farm q4
Rural n-farm q5
Rural n-farm q1
Rural n-farm q2
Rural n-farm q3
Rural n-farm q4
Rural n-farm q5
Source: Simulation results Source: Simulation results
44
Inquiries should be addressed to:
Ethiopian Economics Association (EEA)

Ethiopian Economic Policy Research Institute (EEPRI)
P. O. Box 34282
Tel: 011 645 32 00, 3329, 3041
Fax: 011 645 30 20
e-mail: eea@ethionet.et
Website: www.eeaecon.org

The Economywide Impact of The COVID-19 in Ethiopia

Uploaded by

Copyright:

Available Formats

The Economywide Impact of The COVID-19 in Ethiopia

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Economywide Impact of The COVID-19 in Ethiopia

Uploaded by

Copyright:

Available Formats

Ethiopian Economics Association

The economywide impact of the COVID-19

Lulit Mitik Beyene, Tadele Ferede and Getachew Diriba

Policy Working Paper 03/2020

Lulit Mitik Beyene2, Tadele Ferede3 and Getachew Diriba4

Policy Working Paper 03/2020

Copyright © Ethiopian Economics Association (EEA)

ISBN: 978 99944 54 76-1

The works of Ethiopian Economics Association including this publication

This study assesses the potential economywide impacts of the COVID-19 in

Figure A: GDP level and loss (in billion ETB)

Loss Mild case Loss Severe case GDP Severe case

Figure B: GDP annual growth

No-COVID-19 Mild case Severe case

Source: Simulation results

The COVID-19 crisis is likely to have medium-to-long-term negative effects. All

Figure C: Household consumption (percent deviation relative to no-

2020 with relief measures 2020 no relief measures

Notes: r-f= rural farm; r-nf=rural nonfarm; u=urban; and q 1-q5=quintiles.

The COVID-19 pandemic is likely to have a substantial effect on public finance.

There is uncertainty on the duration of the pandemic in Ethiopia and worldwide.

The 21st century economies are interconnected due to advancement in information

The COVID-19 adversely impacts economic activities which in turn, if

This study seeks to answer three policy questions:

2.1. Model descriptions

The study uses a dynamic Computable General Equilibrium (CGE) model to

Production is modelled using a nested Constant Elasticity of Substitution (CES)

The macroeconomic closure imposes that the government budget balance is

Regarding the external sector, we assume exogenous foreign savings in foreign

2.2. Transmission channels and scenarios

2.2.1. Transmission Channels

Labor layoffs, sick leaves, quarantines, or stay-at-home is another strike area.

Figure 1: COVID-19’s multiple strikes in the circular flow of income

Source: Baldwin (2020)

2.2.2. Identifying shocks

Labor market participation: There is a general reduction in labor supply

Depending on the anticipated severity, the magnitude of the shock will be

Scenario 1: Business as usual scenario which serves as a base run scenario to

Scenario 2: Rapid containment of the virus with limited community spread,

Scenario 3: Relatively widespread community transmission and more

quantify the losses in EFY 2019/20 and 2020/21. For households in

The fourth variant of the medium-to-long-term scenario allows the shock

Table 1: Operationalizing Scenarios

3. Simulation results: Main findings

3.1. The short-term effects with government interventions

3.1.1. Growth effect

-6.5 -5.9 -150

Source: Simulation results

Reduction in Bilion ETB (right axis) GDP (deviation from baseline)

Source: Simulation results

Figure 4: GDP annual growth rate – mild case

8.0 9.0 9.0

Source: Simulation results

Figure 5: GDP annual growth rate – severe case

8.0 9.0 9.0

2.0 2.6 0.6

Source: Simulation results

Private investment would be significantly affected via direct and indirect

Figure 6: expenditure on GDP at purchaser’s price – mild case (%