How do lockdowns affect economic activity in developing countries?

The severity of lockdown measures has been similar across the world. But differences between developing and advanced economies in their sectoral composition and the size of the informal economy mitigate the economic damage in poorer countries.

Since the start of the Covid-19 pandemic, the lockdown measures enacted in developing countries have been at least as stringent as in advanced economies. But while developing countries typically have low employment shares in sectors where people can work from home, they also have high employment shares in sectors that are exempt from lockdowns (such as agriculture) and in the informal economy where lockdown measures are more easily circumvented. These factors mitigate the effect of lockdowns on employment and GDP in developing countries, particularly in the poorest economies.

What is the economic background of the pandemic in developing countries?

To combat the pandemic, governments across the world have imposed restrictions on mobility – lockdowns – that invariably reduce economic activity. In advanced countries, the stringency of lockdowns is often framed as a trade-off between ‘lives’ and ‘the economy’. In developing countries, the choice may be more aptly described as between ‘lives’ and ‘lives’ (Ray and Subramanian, 2020).

This is because governments in developing countries have far fewer means to insure the livelihoods of workers resulting from the economic fallout of lockdown policies:

Governments’ low fiscal capacity makes it especially hard to mobilise resources to insure workers’ income loss, prevent firm bankruptcies, provide social assistance and support demand.
Governments’ limited administrative capacity makes it difficult to target the segments of the population most in need (Gerard et al, 2020).

In addition, economies in developing countries face a number of external shocks due to the severe global economic downturn:

A drop in exports: many developing countries depend disproportionately on exports of commodities, the prices of which have plummeted, and on international tourism, which has come to a halt. Remittances, a major source of foreign currency, are also predicted to decline.
A tightening of financial conditions: the ‘flight-to-safety’ in international financial markets has worsened the borrowing conditions of governments, banks and private firms in the developing world. Several governments have gone into default (Argentina, Ecuador, Lebanon) and an increasing number are seeking credit support from the International Monetary Fund (Djankov and Panizza, 2020).

Related question: What happens to trade in a global downturn?

While their healthcare infrastructures are less well equipped (for example, in the form of intensive care units), developing countries have one key advantage:

Their younger age structure means that Covid-19 is likely to be substantially less deadly than in advanced economies (Alon et al, 2020).

How stringent are lockdown measures by countries’ income levels?

Figure 1 reports two variables that measure the strength of lockdown policies collected by the Oxford Blavatnik School’s Coronavirus Government Response Tracker (Hale et al, 2020). The left pane portrays workplace closures (scaled from 0 to 3) averaged over quintiles of countries ordered by GDP per capita (from Q1 the poorest quintile to Q5 the richest):

Comparing the poorest and richest quintiles of countries (Q1 versus Q5) shows that the most advanced countries have reacted more strongly in closing workplaces. Meanwhile, the most radical policies measures have been in middle-income countries (Q2-Q4).

The right pane of Figure 1 depicts an alternative and broader stringency index of government responses (scaled from 0 to 100). In addition to workplace closures, it includes school closures, stay-at-home orders, restrictions on public events and gatherings, and restrictions on public transport and travel:

Overall, the index confirms that it is middle-income countries that have enacted the most severe actions to date. In addition, measures implemented in May and June have been more stringent in low-income countries than in high-income countries.

Figure 1: Average lockdown measures by quintile of countries and date

Graph showing stringency of lockdown measures

Notes: Data on country-specific workplace closures (scale: 0 to 3) and index of overall stringency of government measures (scale: 0 to 100) are sourced from Hale et al (2020). Averages by quintiles of countries are constructed based on real GDP per capita and cover 144 countries (all countries in the data, conditional on population larger than 1 million). The months correspond to the mid-point of the month. Average GDP per capita, normalised to the US, equals 0.03 in Q1, 0.10 in Q2, 0.22 in Q3, 0.46 in Q4, and 0.91 in Q5.

In summary, the severity of lockdowns across the world has been rather similar by level of development. If anything, the policies enacted in the developing world, particularly in middle-income countries, have been stronger than in advanced economies.

Figure 2 shows one potential measure of the effect of lockdowns. It reports the average mobility change as measured by the Google Covid-19 Community Mobility Reports. The left pane shows that visits to workplaces decreased dramatically relative to baseline:

Over the period from April to June, middle- and high-income countries (Q2-Q5) have featured similar drops in workplace activity. By comparison, the lowest-income countries (Q1) have experienced a markedly lower decline of about half that size.

Figure 2: Average mobility change (in %) relative to baseline by quintile of countries and date

Graph showing average mobility change across various countries

Notes: The data are sourced from Google's COVID-19 Community Mobility Reports. Workplace mobility and residential mobility represent the percent change in visits relative to the pre-pandemic baseline over the period 3 January - 6 February 2020. Each quintile represents averages over countries ordered by GDP per capita and contains a subset of countries composing the quintiles in Figure 1, depending on data availability. Altogether, 118 countries are included.

The right pane represents time spent at home:

Over the period from April to June, individuals in countries of all income levels spent substantially more time at home. The increase was slightly lower in the poorest (Q1) than in the richest countries (Q3), while the strongest increase occurred in middle-income countries (Q2 and Q3).

The Google mobility data have the advantage of measuring activity in real time, but they are not a direct measure of economic output. We will get a clearer picture of the effect of lockdowns as data on realised GDP become available. But it will still require disentangling the direct effect of policy measures relative to other forces, such as changes in individual behaviour.

To understand how lockdowns differ across countries over the development spectrum, it is therefore useful to analyse their structural differences. Three factors that are likely to play a key role are:

The extent to which people are able to work from home.
The sectoral composition of the economy.
The size of the informal economy, in which work is neither regulated nor registered.

How does the ability to work from home differ across countries?

As lockdowns and social distancing shutter workplaces, working from home becomes key to sustaining employment and economic activity during the pandemic. Figure 3 presents two indices that act as proxies for the inability to work from home – physical/manual intensity and face-to-face intensity – and relates them to GDP per capita (Hatayama et al, 2020). The upper (lower) pane uses data from the PIAAC (STEP) survey. Both show that:

The task structure in developing countries is less suitable for working from home than in advanced economies.

Several studies have estimated the fraction of total employment that can potentially work from home by using cross-country variation in the structure of occupations and demographics. For example, developing countries have low employment shares in occupations with high ability to work from home (such as managers and professionals) and high employment shares in occupations with low ability to work from home (such as farmers and elementary occupations):

One estimate finds that roughly 40% of total employment can be executed from home in the richest countries, but only about 10% in the poorest countries (Dingel and Neiman, 2020).
Another study projects that more than 50% of urban employment is amenable to working from home in the most advanced economies, but only 35% in the least developed countries (Gottlieb et al, 2020).

These numbers imply that a complete lockdown leads to a substantially larger decline in employment (and by extension GDP) in developing countries. At the same time, the health benefits of staying at home are likely to be more limited in developing countries because people’s home environments offer less protection against infection (Brown et al, 2020).

Figure 3: Task content by level of development

Graphs showing task content by level of development

Source: Hatayama et al (2020).
Note: The vertical axis measures the corresponding task index, in standard deviations from the mean for all PIAAC/STEP countries. GDP per capita PPP comes from the World Development Indicators and corresponds to the same year of the respective PIAAC and STEP surveys.

Related question: Who can work from home and how does it affect their productivity?

How does the sectoral composition of the economy relate to lockdowns?

While countries implement distinct types of economic restrictions to stem the pandemic, they all exempt sectors and occupations considered to be essential. One such sector is agriculture. Not only is it essential, it is also concentrated in occupations categorised as ‘low-proximity’ where virus transmission is relatively low (Mongey et al, 2020). Poorer countries have a substantially higher fraction of employment and value-added in agriculture.

Similarly, developing countries have relatively low employment shares in sectors that are seen as most risky in fostering virus transmission, such as the arts and entertainment or hospitality. This implies that a lockdown targeting the same set of sectors leads to different effects across countries.

Figure 4 presents the results of a study that incorporates country-level differences in the ability to work from home as well as differences in sectoral composition (Gottlieb et al, 2020). It simulates the level of employment and GDP by two lockdown scenarios relative to pre-trend against countries’ real income per capita. Under both scenarios, employment and GDP are U-shaped in income per capita. There are two forces at work:

On the one hand, rich countries have a substantially higher ability to work from home, which mitigates declines in employment and GDP.
On the other hand, sectoral composition favours poor countries: they concentrate employment and value-added in essential sectors that are not shut down, particularly agriculture.

Figure 4: Simulation of GDP relative to trend under different lockdowns

Graph showing simulation of GDP relative to trend under a hard and soft lockdown

Source: Gottlieb et al (2020).
Note: Real GDP per capita of each country corresponds to the 2017 PPP-adjusted series from the Penn World Table, normalised to the US. The trend line is a quadratic fit of the logarithm of real GDP per capita.

How does informality relate to lockdowns?

Labour markets in developing countries are characterised by a high degree of informality – parts of the economy in which work is neither regulated nor registered. Figure 5 shows that while in the poorest economies, almost 90% of non-agricultural employment is informal, in the richest countries in the sample, that share is only 30%.

These data do not cover the most advanced economies, where the informality rate is lower still. Because informal workers by definition operate outside the government’s purview, the implication of the high rate of informality in developing countries is twofold.

On the one hand, the task of reaching and insuring workers is more difficult. As informal workers are disproportionately low-skilled and self-employed, they have few options to work from home (Saltiel, 2020).

On the other hand, high rates of informality mean that lockdowns in developing countries are less effective because informal workers can circumvent closures and pursue their activities more easily.

One simulation finds that the degree of informality is the main reason why lockdowns generate a substantially steeper GDP decline in rich countries than in poor countries (Alon et al, 2000):

The study quantifies that the effect of, say, a 12-week lockdown lowers GDP by 7.1 percentage points in the richest quartile of countries, but only by 2.9 percentage points in the poorest countries. Much of that difference is due to the assumption that informal employment is not affected by lockdown.

Figure 5: Informal employment in low and middle-income countries

Graph showing informal employment in low and middle-income countries

Source: Alon et al (2000).

What are the policy implications?

Given weak fiscal and administrative capacities to absorb shocks, and facing unprecedented adverse consequences due to the global economic and financial crisis, policy-makers in developing countries can ill afford to implement long-lasting lockdown measures. In comparison with advanced economies, lockdowns are likely to have the following impact on employment and GDP in developing countries:

A pronounced negative effect because less work can be executed from home.
A mitigated negative effect because employment is concentrated in sectors that are essential and in informal work where the enforcement of lockdowns is weak.

In the least developed economies, the latter force is likely to be stronger, implying that lockdowns have a limited adverse effect on economic activity. The flipside, however, is that lockdowns are less effective in radically curbing the transmission of Covid-19.

Where can I find out more?

The VoxEU e-book Covid-19 in Developing Economies edited by Simeon Djankov and Ugo Panizza contains articles on the specific challenges faced by developing countries.

The policy brief Social Protection Response to the COVID-19 Crisis: Options for Developing Countries by François Gerard, Clément Imbert and Kate Orkin reflects on policy responses to mitigate the economic consequences of lockdowns in developing countries.

In Lockdown Accounting, Charles Gottlieb, Jan Grobovšek, Markus Poschke and Fernando Saltiel discuss how the effects of lockdowns differ across countries due to differences in sectoral composition and the ability to work from home.

Lockdowns in developing countries should shield the elderly: Writing at VoxEU, Titan Alon, Minki Kim, David Lagakos and Mitchell VanVuren explain the impact of lockdowns in developing countries.

India’s Lockdown: Writing at VoxEU, Debraj Ray, Sreenivasan Subramanian and Lore Vandewalle recount the specific circumstances of lockdown in India.