WP/20/217

COVID-19 and Inequality in Asia: Breaking the Vicious

Cycle

by Emilia Jurzyk, Medha Madhu Nair, Nathalie Pouokam,

Tahsin Saadi Sedik, Anthony Tan, Irina Yakadina

IMF Working Papers describe research in progress by the author(s) and are published
to elicit comments and to encourage debate. The views expressed in IMF Working Papers
are those of the author(s) and do not necessarily represent the views of the IMF, its
Executive Board, or IMF management.
© 2020 International Monetary Fund WP/20/217

IMF Working Paper

Asia and Pacific Department

COVID-19 and Inequality in Asia: Breaking the Vicious Cycle 1

Prepared by Emilia Jurzyk, Medha Madhu Nair, Nathalie Pouokam,Tahsin Saadi

Sedik, Anthony Tan, Irina Yakadina

Authorized for distribution by Alison Stuart

October 2020

IMF Working Papers describe research in progress by the author(s) and are published to
elicit comments and to encourage debate. The views expressed in IMF Working Papers are
those of the author(s) and do not necessarily represent the views of the IMF, its Executive Board,
or IMF management.

Abstract
The COVID-19 pandemic risks exacerbating inequality in Asia. High frequency labor
surveys show that the pandemic is having particularly adverse effects on younger workers,
women and people that are more vulnerable. Pandemics have been shown to increase
inequalities. As a result, income inequality, which was already high and rising in Asia before
the pandemic, is likely to rise further over the medium term, unless policies succeed in
breaking this historical pattern. Many Asian governments have implemented significant
fiscal policy measures to mitigate the pandemic’s effect on the most vulnerable, with the
impact depending on the initial coverage of safety nets, fiscal space, and degree of
informality and digitalization. The paper includes model-based analysis which shows that
policies targeted to where needs are greatest are effective in mitigating adverse distributional
consequences and underpinning overall economic activity and virus containment.

JEL Classification Numbers: D63; D74.

Keywords: COVID-19, Inequality, Susceptible-Infected-Recovered Macro Model, Fiscal
Policy.
Author’s E-Mail Address: [email protected]; [email protected]; [email protected];
[email protected]; [email protected]; [email protected].

1
We are grateful to Helge Berger, Davide Furceri, Jonathan D. Ostry, Changyong Rhee, Alison Stuart, Rui Xu,
Jiae Yoo, and seminar participants at the IMF for helpful comments and suggestions.
 3

Table of Contents
I. Introduction ........................................................................................................................... 4

II. Key Inequality Trends in Asia Before COVID-19............................................................... 5

III. Labor Market Surveys Indicate Rising Inequality .............................................................. 9

IV. Policies to Break the Vicious Cycle ................................................................................. 15

V. Policy Analysis: More Targeted Measures, More Lives Saved ......................................... 17

VI. Conclusions....................................................................................................................... 20

References ............................................................................................................................... 22
 4

I. INTRODUCTION

Before the COVID-19 pandemic, rising income inequality was already among key challenges
in Asia. While in the past, rapid growth in Asia had been accomplished with an equitable
distribution of the gains, since the early 1990s fast-growing Asian economies have been
unable to replicate the “growth with equity” miracle (Jain-Chandra and others 2016). Indeed,
since then, the region has witnessed rising income inequality. The change in inequality
(delta) in Asia is higher than in other regions, with the level of inequality surpassing the
world average. The rise in inequality is even more striking on a population-weighted basis on
the account of higher inequality in the most populous countries in Asia.

Absent the right policies, the COVID-19 pandemic poses a high risk of further worsening
inequality. The virus pushed the world economies into a Great Lockdown, which triggered
the worst recession since the Great Depression (IMF, 2020a; Deb et al. 2020). Evidence from
other regions also suggests that the pandemic is worsening distributional outcomes, for
example in the United States (Shibata 2020) and the United Kingdom (Haioglu, Känzig, and
Surico 2020). Against this background, the paper answers the following questions: (i) what is
the likely effect of the COVID-19 on inequality in Asia; and (ii) what policies could protect
the most vulnerable while helping the recovery.

To answer the first question, the paper documents the long-term trends in inequality and uses
the 2020 high-frequency labor survey data to identify the impact of the COVID-19 shock on
the workers and sectors most affected, and provides preliminary evidence on the varied
employment effects of the pandemic.

To answer the second question, information from the IMF Policy Tracker is used to analyze
the policy responses of Asian governments since the start of the pandemic to protect the most
vulnerable. In addition, the paper uses a novel and extended version of Susceptible-Infected-
Recovered (SIR)-macro model by Engler, Pouokam, Rodriguez, and Yakadina
(forthcoming)—with different types of agents and a fiscal policy block—to assess how
various fiscal measures could help soften the crisis impact on widening inequality.

The key findings range from empirical contributions to model simulations and policy
analysis. The COVID-19 pandemic is taking its toll on Asia’s labor markets, and the picture
is bleak. High-frequency labor market indicators have sharply deteriorated, and substantially
more than during the Global Financial Crisis (GFC). Aggregate hours worked have declined
both at the extensive (employment rate) and intensive margins (hours worked per employee).
Unemployment has surged while labor force participation has plunged—an early sign of
scarring effects. Moreover, the pandemic is having particularly adverse effects on the already
vulnerable: younger workers and women. During the pandemic Asia’s youth have suffered
greater job losses than other age categories and the gender pay gap—already the second
largest globally—has widened.

Income inequality is likely to keep rising over the medium term, damaging economic growth
and social cohesion (October 2020 Regional Economic Outlook: Asia and Pacific; see also
the companion paper; Saadi Sedik and Xu, forthcoming). This is similar to the findings of
Furceri, Loungani, Ostry and Pizzuto (2020) who provide evidence that major epidemics
 5

over the past two decades, even though smaller in scale than COVID-19, have led to
persistent increases in the Gini coefficient, raised income shares to higher-income deciles,
and lowered the employment to population ratio for those with basic education compared
with those with higher education. One channel through which pandemics may increase
inequality is the acceleration in automation and robotization (October 2020 Regional
Economic Outlook: Asia and Pacific; see also Saadi Sedik and Yoo, forthcoming). Indeed,
while automation may raise productivity, it also increases inequality by displacing lower paid
workers in routine manual occupations.
Lessons from past pandemics also suggest that the resulting higher levels of inequality could
undermine social cohesion and jeopardize future growth, increasing the risks of a vicious
cycle. This is especially salient for countries with already high inequality going into this
crisis ((October 2020 Regional Economic Outlook: Asia and Pacific; Saadi Sedik and Xu,
forthcoming a and b).
According to the IMF Policy Tracker, many Asian governments have implemented
significant fiscal packages to mitigate the pandemic’s effect on the most vulnerable, however
the impact varies depending on the initial coverage of social safety nets and fiscal space but
also the degree of informality and digitalization.
Finally, model-based analysis shows that policies targeted to where needs are greatest are
effective in mitigating adverse distributional consequences and underpinning overall
economic activity and virus containment. In particular, fiscal support measures when
governments have access to external financing not only help diminish the economic cost of
the pandemic but can significantly reduce the number of infections and, thus, save lives. The
favorable effects are larger for targeted than for untargeted measures. Allowing governments
to borrow externally helps support the economy throughout the pandemic recession but may
require more progressive fiscal measures to avoid excessive pandemic debt accumulation and
preserve medium-term debt sustainability.
The rest of the paper is organized as follows. Section II presents regional inequality trends
before COVID-19. Section III shows how the pandemic and its economic toll exacerbated
these trends. Section IV analyzes the policy measures by Asian governments in response to
the crisis. Section V presents model-based policy recommendations. Section IV concludes.

II. KEY INEQUALITY TRENDS IN ASIA BEFORE COVID-19

Asia has been a growth leader and has achieved remarkably high growth for sustained
periods, lifted millions out of poverty and virtually eliminated incidence of extreme poverty
(Figure 1). However, in a number of Asian economies this impressive economic performance
has been accompanied by rising inequality, particularly since the early 1990s—a break from
the region’s remarkable past (Jain-Chandra and others 2016). This means that Asia’s last 25-
plus years of growth have been less inclusive and less pro-poor (Balakrishnan et al., 2013;
and Figure 2). The change in inequality (delta), as measured by the net Gini index, in Asia is
currently higher than in other regions (Figure 3), with the level of inequality surpassing the
world average. The increase in inequality is even more striking on a population-weighted
basis, reflecting the sharp rise in inequality in the most populous countries in Asia, notably
 6

China, India and Indonesia.2 This reflects larger income gains for the top 10 percent of the
population (Figure 4).

Figure 1. Extreme Poverty (per capita household

consumption below $1.90/day)
(Share of the population living in extreme poverty)

Source: World Bank, IMF staff calculations

Note: Poverty gap is a ratio showing the average shortfall of the total population from
the poverty line.

Figure 2: Selected Asia: Income Inequality

(Net Gini Index; in Gini points; change during the period indicated in parentheses)
(a) Pre-1990 (1960s to mid-1980s) (b) Post-1990 (1990 to 2018, or latest)

Source: SWIID v8.2, IMF staff calculations.

Note: Average for Asia is based on population-weighted average.

2
Alternative measures of income equality, such as the Palma ratio, also showed consistent trend.
 7

Figure 3. Change in Income Inequality: Figure 4. Asia: Growth Incidence Curve

Regional Comparison (Annual compounded mean income/consumption
(Net Gini index, in Gini points; average growth (USD), by decile, in percent)
across region)

Source: SWIID v8.2, IMF staff calculations. Source: World Bank PovCal database, IMF staff calculations.
Note: Regional aggregation is based on population-weighted Note: Data refers to the median income/consumption growth over
average. corresponding period.

Gender. From the gender perspective, men continue to make up an overwhelming majority
of top earners globally. According to ILO data, gender income disparity in Asia remains the
second largest, after North America and Western Europe, where Asia’s men take home
around 11 percent more as compared to their female counterparts (Figure 5). Gender income
inequality remains persistent in Asia because it is partly rooted in culture and tradition in
some countries that prevent women from equal access to education and employment
opportunities (Jayachandran, 2015; UN, 2020). Asian countries that ranked among the top 20
in the United Nation’s human development index (HNI)–which measures social and
economic achievements, fared poorly in the gender development index (UNDP, 2019).

Urban versus rural. In the spatial dimension, the inequality of income (or consumption)
between urban and rural households in Asia remains large, notably in its largest economies.
In China, the gap has widened significantly since the 1990 (Figure 6). According to (Kanbur
and Zhuang, 2013), rural–urban income divide accounts for 45 percent of the economy-wide
inequality in China (India and Indonesia: 25 percent and 20 percent, respectively).
 8

Figure 5. Gender Income Inequality, Figure 6. Selected Asia: Urban-Rural Inequality

Regional Comparison of Income (or Consumption)
(Gender wage gap, in percent of average (USD, average)
male wages, 2018 or latest)

Source: ILO, IMF staff calculations. Source: World Bank

Asia refers to Australia, China, Indonesia, Korea, Mongolia,
Nepal, Philippines, Sri Lanka, Thailand and Vietnam.

Inequality of Opportunities. Apart from inequality of outcomes, Asia is also confronted with
considerable inequality of opportunities, such as access to education, health, financial
services. The widening inequality of opportunities not only preserves inequality of outcomes,
but also perpetuates social divisions (UN, 2020; Aiyar and Ebeke, 2019).

Education. Figure 7 shows that there is a large gap between educational attainments of the
wealthiest quintile and the poorest quintile in Asia, defined as the share of population aged
20-24 with less than four years of schooling. As access to basic education by the poor is
limited; they are more likely to be trapped in the cycle of poverty, further hampering
economic mobility. As well as facing a large education gap, the Asian emerging and
developing economies (EMDEs) also have the largest share of youth not in education,
employment or training (Figure 8).

Figure 7. Selected Asia: Access to Figure 8. Share of Youth Not in Education,

Education by Wealth Quintile Employment or Training (NEET), Regional
(Attained less than 4 years of education, Comparison
percent of total population aged 20-24) (Percent)

Source: World Bank, IMF staff calculations. Source: ILO, IMF staff calculations
Note: Selected Asian countries are based on available data. Note: Data are based on population-weighted averages.
 9

Healthcare. There is also a persistent and sizeable gap in access to healthcare between high-
and low- income households in Asian EMDEs. Figure 9 shows that there is a large gap in the
reproductive, maternal, newborn and child health coverage, notably in South Asia. Such
disparities could also reflect insufficient medical services in rural areas, partly due to poor
infrastructure (such as rail and road connectivity), as well as absence of universal healthcare
coverage.

Financial services. The lack of access to financial services also constrain low-income
households from investing in their future, such as for education and training. Figure 10 shows
that in some Asian EMDEs, there remains a notable gap in the share of adults with bank
accounts between the top 60 percent and the bottom 40 percent of the income distribution
(such as in Lao PDR and the Philippines).

Figure 9. Selected Asia: Access to Health by Figure 10. Selected Asia: Access to Financial
Wealth Quintile Services by Income Share
(Percent, coverage of reproductive, maternal, (Accounts at a financial institution; in percent
newborn and child health) of total population aged 15 and above)

Source: WHO, IMF staff calculations. Source: World Bank, Global Findex Database, IMF staff
Note: Selected Asian countries are based on available data. calculations.
Note: Selected Asian countries are based on available data.

III. LABOR MARKET SURVEYS INDICATE RISING INEQUALITY

In this section, we use high-frequency labor survey data to identify which types of workers-
sectors are more impacted by COVID-19 shocks and compare the impact of the pandemic to
that of the Global Financial Crisis (GFC). We focus on the effects on employment across
different industry classifications (whether they are high contact sectors, requiring physical
human interactions) and job flexibility (whether remote work is possible). We then study
worker demographics (such as gender, age, and skill/education levels) to document how the
pandemic has worsened distributional outcomes exacerbating inequality trends in Asia
discussed earlier.
 10

Pandemic Effects on Employment and Earnings

We find that Asia’s labor market conditions have deteriorated markedly, more than during
the Global Financial Crisis (GFC). The aggregate effects of the pandemic on Asia’s
employment was negative across the board, as aggregate hours worked declined both at the
extensive (employment rate) and intensive margins (hours worked per employee), surpassing
the GFC (Figures 11 and 12). As a result, unemployment rate surged and labor force
participation plunged—a first sign of scarring effects (Figures 13 and 14).

Figure 11. Asia: Change in Employment Figure 12. Asia: Change in Weekly Hours
Rate Worked
(Percentage points) (Q1 2007 = 100), (percentage points)

Source: Haver Analytics, IMF staff calculations Source: Haver Analytics, IMF staff calculations
Notes: Asia refers to Australia, Hong Kong, Indonesia, Notes: Asia refers to Australia, Hong Kong, Japan, Korea,
Japan, Korea, Malaysia, New Zealand, Philippines, Philippines and Singapore. Data are seasonally-adjusted, based
Singapore, Taiwan Province of China, Thailand and on population-weighted average.
Vietnam. Data are seasonally-adjusted, based on population-
weighted average

Figure 13. Asia: Change in Unemployment Figure 14. Asia: Change in Labor Force
Rate Participation Rate
(Percentage points) (Percentage points)

Source: Haver Analytics, IMF staff calculations Source: Haver Analytics, IMF staff calculations
Notes: Asia refers to Australia, China, Hong Kong, India, Notes: Asia refers to Australia, Hong Kong, Indonesia,
Indonesia, Japan, Korea, Malaysia, New Zealand, Philippines, Japan, Korea, Malaysia, New Zealand, Philippines,
Singapore, Taiwan Province of China, Thailand and Vietnam. Singapore, Taiwan Province of China, Thailand and
Data are seasonally-adjusted, based on population-weighted Vietnam. Data are seasonally-adjusted, based on
average population-weighted average
 11

The effects on Asia’s employment varies, with significant heterogeneity across industries, as
the lockdowns and closure of non-essential businesses had differential impact on the type of
jobs that had been lost. Essential industries (such as utilities, healthcare and groceries)
continue to remain in business, while “social jobs” ─ those that require face-to-face
interactions were prohibited from operating during the beginning of the pandemic recession.
Job losses during the pandemic can also be differentiated in terms of job flexibility,
depending on whether workers are able to work from home (remote work), 3 which in turn
depends on workers having certain skillsets/education levels. The following are the key
stylized facts. 4

• First, high contact social industries 5 that require physical interactions for the consumption
of goods (such as hotels/restaurants and retail) were the most affected, recording the
largest declines in employment (over 6 percentage points in June as compared to pre-
COVID in December 2019). These industries were closely associated with the travel and
tourism sector, which have notably been hit hard. In contrast, job creation in these sectors
remained positive during the GFC (Figures 15 and 16).

• Second, non-teleworkable industries were the second most impacted. These are mostly in
the mining, manufacturing and construction industries, where remote work is not
possible, and consistent with the findings in Brussevich and others (2020).

• Third, workers in both social and non-teleworkable industries are characterized by a

larger share of workers with lower average earnings. For example, the average monthly
wage in the social sector is less than one-third that of essential and teleworkable
industries. This shows that the pandemic has led to widening the already large income
disparities in the region, leaving lower income workers further disadvantaged and
exacerbating income inequality (Figure 17 and 18).

3
However, this would also depend on industries having the necessary pre-requisites, such as access to technology and IT
infrastructure for effective teleworking. Workers in emerging market economies are likely to face significant challenges during
strict lockdowns given limited access to technology (Brussevich and others, 2020).
4
Similar trend can be observed, for example in the U.S. (Shibata, 2020) and Euro area (Botelho, et al., 2020).
5
Social industries refer to those industries where consumption of goods require physical interactions such as wholesale, retail,
leisure/tourism and entertainment. Essential industries refer to agriculture, utilities, transport, information/communication,
health and public administration; social industries refer to wholesale/retail, hotels/restaurants, arts/entertainment;
teleworkable industries refer to finance, business/professional services and education; and non-teleworkable industries refer
to mining, manufacturing and construction. Even though health is a social industry by definition, it is regarded in this study as
an essential industry due to the health risk posed by the pandemic. See Shibata, 2020, Dingel and Newman, 2020.
 12

Figure 15. Asia: Change in Employment Figure 16. Asia: Change in Employment (All
by Industry Classification during Crises Industries), GFC versus Pandemic Recession
(Percentage points) (Percentage points)

Source: Haver Analytics, IMF staff calculations Source: Haver Analytics, IMF staff calculations
Notes: COVID-19 = coronavirus disease; GFC = global Asia refers to Australia, Hong Kong SAR, Indonesia, Japan, Korea,
financial crisis. Asia refers to Australia, Hong Kong SAR, Malaysia, New Zealand, Singapore, Taiwan Province of China,
Indonesia, Japan, Korea, Malaysia, New Zealand, Singapore, Thailand, The Philippines, and Vietnam. Data are seasonally adjusted,
Taiwan Province of China, Thailand, The Philippines, and based on June 2020 data.
Vietnam. Data are seasonally adjusted, based on June 2020
data. Essential industries refer to agriculture, utilities,
transport, information and communication, and health and
public administration; social industries refer to wholesale
and retail, hotels and restaurants, and arts and entertainment;
teleworkable industries refer to finance, business and
professional services, and education; and non-teleworkable
industries refer to mining, manufacturing, and construction.

Figure 17. Asia: Average Monthly Wages Figure 18. Asia: Change in Average Monthly
(Apr 2020) Wages from Pre-COVID levels (Dec 2019 to
(in USD) latest)
(Percentage points)

Source: Haver Analytics, IMF staff calculations Source: Haver Analytics, IMF staff calculations
Notes: Asia refers to Japan, Korea, Taiwan Province of China Notes: Asia refers to Japan, Korea, Taiwan Province of China
and Thailand. and Thailand.
 13

Pandemic Effects on Employment by Worker Demographics

Gender

The pandemic is having a major impact on the well-being of many vulnerable groups,
including in Asia. There is a growing literature documenting that women are amongst those
most heavily affected, amplifying existing socioeconomic inequalities (UN, 2020; UNDP,
2020; Boinol et al., 2019). Women represent a larger share of essential workers on the
frontline, such as first responders in the healthcare industry, and cashiers at
groceries/pharmaceutical stores (OECD, 2020). Not only that women are more exposed to
greater risk of infection, but there remains a significant gender income gap, underscoring the
notion that women continue to be underpaid as compared to their male counterparts.
Industries that were hit hard by the pandemic also predominantly employed women (Figure
19). Asia’s gender inequality is further exacerbated by the region’s high share of informality
(including the gig economy), where women account for the bulk of the share. They are more
likely to be in lower paying jobs with little security or protection.

Another key distinguishing factor that women bore the brunt of the pandemic is that women,
in many societies, are the primary care providers. Given the travel restrictions, home
quarantines, school and day-care center closures, additional burden of care on the young and
the elderly falls disproportionately on women, even as some women (and their partners) have
the option to work from home. A key reason is a “guilt gap” between women and men, where
women often feel compelled to take on more professional sacrifices (Aoyagi, 2020). Between
December 2019 and June 2020, Asia’s female participation rate declined by 1.3 percentage
points compared to a 1.0 percentage point decline for male (Figure 20), as more women
exited the labor force. Without policies to prevent scarring and support female employment
opportunities, the pandemic could potentially accentuate the already high gender inequality
in the region.

Figure 19. Asia: Share of Employment by Figure 20. Asia: Change in Labor Force
Gender (All Industries) Participation Rate (by Gender)
(Percent) (Percentage points)

Source: ILO, IMF staff calculations Source: Haver Analytics, IMF staff calculations
Notes: Asia refers to Australia, Hong Kong, Indonesia, Japan, Notes: COVID-19 = coronavirus disease; GFC = global financial
Korea, Malaysia, New Zealand, Philippines, Singapore, crisis. Asia refers to Australia, Japan, Korea, Hong Kong, Thailand,
Thailand and Vietnam. Data refers to 2018. and The Philippines. Data are seasonally adjusted. For COVID-19,
data are up to June 2020.
 14

Age

Looking at the age dimension, Asia’s youth experienced sharper job losses compared to other
workers during the pandemic, with the average youth unemployment rate rising by 1.4
percentage points by June 2020 (Figure 21). As discussed in Section II, prior to the
pandemic, Asia had one of the highest shares of youth not in employment, education or
training, particularly in developing countries. The current crisis is aggravating this inequality
trend, as youth are typically the most vulnerable to worsening economic conditions (Ahn et
al., 2019) and they are mostly employed in sectors requiring social contact (Figure 22). There
are, however, substantial variations across Asia, with the increase in youth unemployment
most notable in Australia and Thailand.

Figure 21. Asia: Change in Unemployment Figure 22. Asia: Share of Youth Employment
Rate by Age Cohorts (By Selected Industries), 2018 or latest
(Percentage points) (Percentage points)

Source: Haver Analytics. Source: ILO, IMF staff calculations

Note: Asia refers to Australia, Japan, Korea, New Zealand, Notes: Asia refers to Bangladesh, Brunei, Cambodia, Lao PDR,
Taiwan Province of China, and Thailand. Data refers to the Mongolia, Myanmar, Nepal, Tonga, Korea and Vietnam.
change in unemployment rate from December 2019 to June
2020. Data are seasonally adjusted. The horizontal line inside
each box represents the median; the upper and lower edges of
each box show the top and bottom quartiles, respectively; and
the top and bottom markers denote the maximum and the
minimum, respectively; x is the mean.

Education

Looking further into how crises events affected workers with different education background
reveals the following observations:

• Workers with primary education and below have been the most affected by the pandemic,
with the average employment rate declining by 1.5 percentage point as compared to pre-
pandemic levels—surpassing the GFC (Figure 23). This is consistent with the observation
that workers with lower levels of educational attainment tend to be more susceptible to
job losses, as these jobs generally do not allow workers the option for telework, which
typically requires at least a college degree (Shibata, 2020).
 15

• For workers with secondary education and above, the magnitude of the decline in average
employment rates relative to pre-pandemic levels were largely similar across both
recessions.
Figure 23 Asia: Change in Employment Rates by Education Level
(Percentage points)

Source: Haver Analytics, IMF staff calculations

Notes: Asia refers to Hong Kong, Korea, Taiwan Province of China and Thailand. Data refers to the change in
unemployment rate from December 2019 to June 2020. Data are seasonally adjusted. The horizontal line inside each box
represents the median; the upper and lower edges of each box show the top and bottom quartiles, respectively; and the top
and bottom markers denote the maximum and the minimum, respectively; x is the mean.

IV. POLICIES TO BREAK THE VICIOUS CYCLE

Governments in Asia have immediately responded to COVID-19 crisis with policy

packages, some sized in double-digit percentage of GDP. We compile and analyze such
policy measures using the IMF Policy Tracker and find that effective crisis responses go
hand-in-hand with the following country characteristics: (i) benefiting from a larger fiscal
space; (ii) having broader social safety nets; (iii) exhibiting lower levels of informality; and
(iv) having a higher degree of digitalization.
Looking at the first characteristic, countries with lower outstanding debt at the end of 2019
and therefore larger fiscal space have been more able to respond effectively and protect the
vulnerable, in contrast to the ones that entered the crisis with weaker initial conditions and
thus were facing greater challenges (Figure 24, panel 1).
Second, a higher share of the advanced economies in Asia introduced targeted benefits that
aimed to support the population and preserve firms and jobs. Many advanced economies
introduced cash transfers, enhanced unemployment benefits, wage subsidies, and fiscal
support to firms (Figure 24, panel 2). These measures were utilized by fewer emerging
markets and low-income countries.
 16

Figure 24. Asia’s Policy Responses

1. Asia: Fiscal Response to COVID-19 2. Targeted Help to Households and Workers
(Percent of GDP) (Percent of countries implementing the policy)

Sources: IMF World Economic Outlook database; and IMF survey of

Note: COVID-19 = coronavirus disease pandemic. AE = advanced economy;
EM = emerging market; LIC = low-income country.

Source: IMF survey of policy responses to COVID-19.

Note: COVID-19 = coronavirus disease pandemic.
Country abbreviations are International Organization for
Standardization country codes.

Third, a less frequent adoption of such measures among low-income countries and emerging
markets was likely related to a higher degree of informality. With a higher share of workers
not covered by the social insurance schemes (either because they were employed outside of
the formal sector, were self-employed or held jobs in the new gig economy), reaching
workers through channels designed for formal employment was limited. (Figure 25, panel
1).
Fourth, digitalization likely helped: countries with higher levels of digital adoption were
better placed to deliver support to population as governments were able to use tools like
digital wallets, existing electronic social security and tax rosters and other tools to identify
the vulnerable and deliver assistance. Low-income and emerging market countries that
introduced targeted cash transfers had, on average, higher digitalization scores than those
that did not introduce these measures (Figure 25, panel 2).

Figure 25. Asia’s Policy Responses

1. Fiscal Measures and Informality 2. Targeted Support and Digitalization
(Percent
Sources: ofWorld
IMF countries implementing
Economic and IMF(Percent
the policy)
Outlook database; survey of of countries
policy implementing
responses the policy)
to COVID-19.

Note: COVID-19 = coronavirus disease pandemic. AE = advanced economy; EM = emerging market; LIC = low-income country.
 17

V. POLICY ANALYSIS: MORE TARGETED MEASURES, MORE LIVES SAVED

This section compares the efficiency of various fiscal measures to alleviate the impact of the
pandemic and the lockdown, focusing on targeted fiscal support. It uses a Susceptible-
Infected-Recovered (SIR) macro model (Eichenbaum, Rebelo, and Trabandt, 2020)
extended to include skilled and unskilled workers, redistributive fiscal policy as well as
external borrowing in the form of a pandemic bond (Engler et al. 2020). Box 1 describes
key model assumptions and features.
In the seminal Eichenbaum, Rebelo, and Trabandt (2020) pandemic macro model, the virus
spreads through consumption, workplaces and in the general community. In response
consumers and workers rationally cut their consumption and hours to reduce their
probability of getting infected and possibly perishing from the virus.
We solve for optimal (consumer welfare-maximizing) fiscal policy to compare the
efficiency of various fiscal measures in a lockdown 6, namely a progressive taxation with
targeted fiscal transfers favoring the needy (unskilled, low income workers) vs. a uniform
taxation with untargeted transfers.

6
Lockdown is modeled as a time-varying, Pigouvian consumption tax with all the proceeds rebated to
households.
 18

Box 1. Key Model Assumptions and Features

This paper relies on the novel SIR-macro-fiscal-inequality model by Engler, Pouokam, Rodriguez
Guzman and Yakadina (EPRY, forthcoming). The EPRY model adds fiscal policy and inequality
dimensions to the seminal model by Eichenbaum, Rebelo and Trabandt (ERT, 2020) to perform
comparisons among different sets of—optimal and ad-hoc—fiscal policy measures in the time of a
pandemic.
The original ERT model explores the behavioral response of economic agents to the pandemic and the
government-imposed lockdowns. The decentralized, competitive equilibrium in the economy consists
of the consumption and labor supply decisions by each type of agent: Susceptible to the virus, Infected
or Recovered with a life-time immunity (SIR). The model is calibrated to weekly frequency to best
match the pandemic course and focuses on the transitional dynamics, assuming that after the pandemic
is over the economy returns to the original steady state.
In addition to the epidemiological SIR block that models the virus spread during consumption, work or
through general community interactions, ERT study how the macroeconomy responds to changes in
consumers and workers behavior who are aware of the virus channels. Their behavior causes a large
drop in consumption and hours worked as infections and deaths rise to a peak at the heights of a
pandemic. While the economic agents act rationally to optimize their welfare, they fail to fully
internalize the externality of spreading the virus through their own behavior (work and consumption),
so the government has to levy a time varying Pigouvian consumption tax that acts as a containment
(lockdown) and helps tame the pandemic and save lives. The tax proceeds are transferred to the
consumers who are assumed hand-to-mouth.
The EPRY model uses the key ERT features and adds two further features: first, it distinguishes
between rich and poor which is modeled as unskilled and skilled workers. As single-good producing
firms pay competitive wages in proportion to workers’ productivity (skills), unskilled workers earn low
incomes and consume less than skilled workers. Such low-income workers constitute the majority (70
percent) of the population, close to the average in Asia. The second feature is introducing a variety of
fiscal policy instruments so that we can identify optimal fiscal policy of a benevolent government
chooses fiscal instruments to maximize consumers welfare. Fiscal instruments consist of (i)
progressive, time-varying consumption tax rates that are different for the two labor income groups
(skilled and unskilled); (ii) progressive (or targeted) transfers to each consumer type; and (iv) external
debt to help finance the targeted transfers during the pandemic.
Third, Introducing short-term external borrowing is an innovative feature that helps study the optimality
of using targeted fiscal policy while acknowledging the tradeoffs between a better control over the
pandemic in the near term and the spike in debt-to-GDP ratios caused by both deteriorating fiscal
balances and lockdown-induced deep recession. With the goal of focusing the pandemic-related
effects, debt is modeled in a form of a short-term, external pandemic bond that can be rolled over for
250 weeks but has to be fully repaid thereafter, through uniform labor income taxes over the medium
term. The interest rate on the pandemic bond is exogenous and assumed constant interest rate as long as
debt-to-GDP ratio remains sustainable, but with a fast-rising risk premium when debt ratios became too
high.
The EPRY model is used to run three sets of simulations: the baseline with no policies (without
lockdown and fiscal transfers); the untargeted policy scenario in which the government sets the same
containment and transfers for skilled and unskilled; and the targeted policy scenario in which higher
earners are taxed more and lower earners receive more in fiscal transfers.
 19

The model mimics some features of the pandemic and lockdowns: (i) consumption falls
more for skilled workers with higher income and greater discretionary spending; and (ii)
unskilled (low income) workers are more exposed to the pandemic through their workplace,
as seen in transport, retail, production lines, and meat processing plants (Figure 26).
Figure 26. Matching Pandemic Behavior of Different Income Brackets

As more countries used targeted fiscal measures, the model

…where the unskilled are more exposed to the pandemic
has targeted policies that differ for higher income skilled
through their workplaces.
and lower income unskilled workers…
Confirmed New Cases
50000 (per million)
40000
30000
20000
10000
0
1 11 21 31 41
Skilled Unskilled

Lower income unskilled workers lose more hours due to as …while consumption falls more for skilled workers with
their workplaces were more affected by lockdowns… higher income.
Working Hours, Consumption,
10
(percent deviations from pre-pandemic levels) (percent deviations from pre-pandemic levels)
0
0 -10
-10 -20
-20 -30
-30 -40
-40 -50
Skilled Unskilled Skilled Unskilled
1 11 21 31 41 -60
1 11 21 31 41
Source: IMF Policy Tracker and Engler and others (2020).
Note: Panel 1 compares the share of countries that adopted targeted vs. untargeted fiscal support measures. Panel 2 shows
additional weekly infections for skilled and unskilled workers; Panels 3 and 4 present the respective declines in weekly hours and
consumption of skilled and unskilled relative to their respective pre-pandemic levels.

Our analysis shows that fiscal support measures not only mitigate the economic cost of
the pandemic but can significantly reduce the number of infections—about one-third
relative to the no-intervention baseline. By helping to protect the livelihoods of
consumers and workers and increasing their disposable income, these measures make
staying home more affordable and help reinforce greater social distancing.
Comparing two sets of optimal fiscal policies, we find that favorable effects from policies
are larger for targeted than for untargeted measures. The former help reduce inequality in
disposable income and preserve a higher consumption share of GDP for the unskilled
(Figure 27). This saves more lives because unskilled workers tend to be more exposed to
the health crisis. The reduction in infections and fatalities, in turn, helps reduce the depth
of the recession and therefore flattens the surge in the debt-to-GDP ratio. The model
suggests that, compared with untargeted transfers, targeted transfers raise GDP by some 3
percent and lower the debt-to-GDP ratio by 6 percentage points (Figure 27).
 20

Figure 27. Targeted versus Untargeted Fiscal Support

(Differences, percent of GDP)
Optimal policy with targeted transfers results in a higher
…which leads to a lower pandemic debt accumulation.
GDP relative to the one with untargeted transfers…

…while the skilled experience a significant reduction in

Targeted support leads to higher consumption share of the
their consumption share because of redistributive
unskilled in GDP…
measures.

Source: Engler and others (2020).

Note: TT = targeted transfers; UT = untargeted transfers. Panel 1 shows the weekly GDP level under the targeted fiscal support scenario (with
progressive taxes and targeted transfers) in percent deviations from weekly GDP under the untargeted scenario (with uniform taxation and
general transfers). The remaining three panels plot differences between the targeted and untargeted scenarios for the ratios-to-GDP of debt,
consumption of unskilled, and consumption of skilled, respectively.

The analysis helps hone in on policy recommendations. Given the worsening of the
underlying inequality trends exacerbated by the pandemic, it is optimal to continue targeted
support measures to mitigate the pandemic and lockdown impact. Where appropriate,
enhancing progressivity of taxes and transfers could help mitigate the anticipated buildup in
debt that could otherwise pose medium-term debt sustainability risks.

VI. CONCLUSIONS

In this paper we show, based on high-frequency labor surveys, that inequality is increasing
further during the COVID-19 pandemic because job losses have been concentrated among
low-income workers, women and youth. Information from the IMF Policy Tracker shows
that many Asian governments have implemented significant fiscal policy measures to
mitigate the pandemic’s effect on the most vulnerable, with the impact depending on the
 21

initial coverage of safety nets, fiscal space, and degree of informality and digitalization.
Although there is no one-size-fits-all best policy, our model-based analysis suggests that it
is economically and socially beneficial to provide targeted support to the unskilled.
Countries that had a lower share of workers in the informal sector could extend support
through formal channels, including wage subsidies for firms to keep workers employed, and
through enhanced unemployment benefits.
As the recovery is likely to be increasingly driven by new sectors and activities, with greater
focus on digitalization and green energy. Therefore, the jobs available in the recovery phase
are likely to require different skills from those lost during the crisis (skill mismatches).
To minimize longer-term damage, policies should also address challenges from automation,
including by revamping education curriculums to achieve more flexible skill sets and
lifelong learning, as well as new training for adversely affected workers.
 22

REFERENCES

Ahn JaeBin, An Zidong, Bluedorn John, Ciminelli Gabriel, Kóczán Zsóka, Malacrino
Davide, Muhaj Daniela, and Neidlinger Patricia. 2019. “Work In Progress; Improving Youth
Labor Market Outcomes in Emerging Market and Developing Economies,” IMF Staff
Discussion Note 19/02, International Monetary Fund, Washington, DC.
Aiyar Shekhar and Ebeke Christian. 2019. “The Missing Link between Income Inequality
and Economic Growth: Inequality of Opportunities.” VOX-EU and Centre for Economic
Policy Research, London.
Balakrishnan, Ravi, Steinberg Chad, and Syed Murtaza. 2013. “The Elusive Quest for
Inclusive Growth: Growth, Poverty and Inequality in Asia”, IMF Working Paper 13/152, ,
International Monetary Fund, Washington, DC.
Boinol Mathieu, Mcisaac Michelle, Xu Lihui, Wuliji Tana, Diallo Khassoum, and Campbell
Jim. 2019, “Gender Equity in the Health Workforce: Analysis of 104 Countries”, Working
Paper 1, World Health Organization.
Botelho Vasco, Consolo Agostino, and Dias da Silva António. 2020. “A Preliminary
Assessment of the Impact of the COVID-19 Pandemic on the Euro Area Labour Market.”
ECB Economic Bulletin 5/2020, European Central Bank.
Brussevich Mariya, Dable-Norries Era, and Khalid Salma. 2020. “Who Will Bear the Brunt
of Lockdown Policies? Evidence from Tele-Workability Measures Across Countries.” IMF
Working Paper 20/88, International Monetary Fund, Washington, DC.
Deb, Pragyan, Davide Furceri, Jonathan D. Ostry, and Nour Tawk. 2020a. “The Economic
Effects of COVID-19 Containment Measures.” IMF Working Paper 20/158, International
Monetary Fund, Washington, DC.Deb, Pragyan, Davide Furceri, Jonathan D. Ostry, and
Nour Tawk. 2020b. “The Effect of Containment Measures on the COVID-19 Pandemic.”
IMF Working Paper 20/159, International Monetary Fund, Washington, DC.
Dingel I. Jonathan and Neiman Brent. 2020. “How Many Jobs Can be Done at Home.”
NBER Working Paper 26948, National Bureau of Economic Research, Cambridge,
Massachusetts.Eichenbaum Martin S., Sergio Rebelo, and Mathias Trabandt. 2020. “The
Macroeconomics of Epidemics.” NBER Working Paper 26882, National Bureau of
Economic Research, Cambridge, MA.
Engler, P., Nathalie Pouokam, Diego Rodríguez Guzman, and Irina Yakadina. Forthcoming.
“Fiscal Redistribution and Inequality in the Time of a Pandemic.” IMF Working Paper,
International Monetary Fund, Washington DC.
Franks, Jeffrey, Bertrand Gruss, Carlos Mulas Granados, Manasa Patnam, and Sebastian
Weber. Forthcoming. “Exiting from Lockdowns: Early Evidence from Reopenings in
Europe.” IMF Working Paper, International Monetary Fund, Washington DC.
Furceri, Davide, Prakash Loungani, Jonathan D. Ostry, and Pietro Pizzuto. 2020. “Will
COVID-19 Affect Inequality? Evidence from Past Pandemics.” COVID Economics 12
(May): 138–157.
Haioglu, Sinem, Diego R. Känzig, and Paolo Surico. 2020. “The Distributional Impact of the
Pandemic.” CEPR Discussion Paper 15101, Centre for Economic Policy Research, London.
 23

International Labor Organization (ILO). 2020. “Global Employment Trends for Youth
2020.” ILO, Geneva.
Jain-Chandra Sonali, Kinda Tidiane, Kochhar Kalpana, Piao Shi, and Schauer Johanna. 2016.
“Sharing the Growth Dividend: Analysis of Inequality in Asia.” IMF Working Paper 16/48,
International Monetary Fund, Washington, DC.
Jayachandran, Seema. 2015. “The Roots of Gender Inequality in Developing Countries.”
Annual Review of Economics No. 7, 63-88.
Kanbur Ravi and Zhuang Juzhong. 2013. “Urbanization and Inequality in Asia.”, Asian
Development Review, Vol. 30(1): 131-147.
Organization for Economic Co-operation and Development (OECD). 2020. “Women at the
Core of the Fight Against COVID-19 Crisis”, OECD Policy Responses to COVID-19.
Shibata Ippei. 2020. “The Distributional Impact of Recessions: The Global Financial Crisis
and the Pandemic Recession.” IMF Working Paper No. 20/96, International Monetary Fund,
Washington, DC.
World Health Organization (WHO). 2020. “Considerations in Adjusting Public Health and
Social Measures in the Context of COVID-19: Interim Guidance.” WHO reference number
WHO/2019-nCoV/Adjusting_PH_measures/2020.1 (April 16, 2020), World Health
Organization, Geneva.
United Nations Development Program (UNDP). 2019. “Human Development Reports.”
United Nations.
United Nations Development Program (UNDP). 2020. “The Social and Economic Impact of
COVID-19 in the Asia-Pacific Region, United Nations.
United Nations (UN). 2020a. “Education as the Pathway towards Gender Equality.” UN
Chronicle, United Nations.
United Nations (UN). 2020b. “World Social Report 2020: Inequality in a Rapidly Changing
World.” United Nations.
United Nations (UN). 2020c. “The Impact of COVID-19 on Women”, Policy Brief, United
Nations.