AdobeStock_338992888.jpeg
VoxEU Column COVID-19 Europe's nations and regions Labour Markets

The Covid-19 shock: Employment in middle-income economies

While flexible labour markets normally facilitate economic adjustment during crises, recent Google search data suggest that the widespread Covid-19 lockdowns may impede this adjustment process. This column explores how labour market structures may determine how employment levels across middle-income countries are affected by the shock. The impending job and income losses are likely to be most severe where fewer people have permanent contracts, where many are self-employed, and where more people work for small firms and in retail. In the long term, these asymmetric impacts may further increase the demand for public-sector jobs.

Several countries have shut down entire industries in an attempt to contain the spread of the Covid-19 virus. As unemployment benefit claims in the US spiked at a record-breaking 6.6 million in the week ending 28 March 2020 (approximately ten times the peak level seen in the aftermath of the 2008-09 Global Crisis), the employment effects of the Covid-19 crisis are increasingly coming to the forefront of the policy debate. How labour markets respond to the Covid-19 shock depends on the interaction between policies and employment structures. The latter vary greatly across middle-income economies. This column examines employment structures (and distils the implications for the Covid-19 crisis response) in 38 economies across Central, Eastern, and South-Eastern Europe, as well as in Central Asia and the Southern and Eastern Mediterranean regions (see EBRD 2020).

One way to gauge how labour markets are responding to the current shock is by analysing near real-time Google search query data. These data show a sharp increase in the number of searches for unemployment and welfare (for example, ‘coronavirus and unemployment benefits’ or ‘phone number to claim unemployment’). During the two weeks starting 22 March 2020, internet searches concerning these themes increased (on average) by 11 percentage points compared with the two weeks starting 09 February 2020. They also increased by eight percentage points relative to the same period last year. Globally, such searches went even up by 58 percentage points. Importantly, such increases were higher in countries where retail services (severely affected by lockdowns) account for larger shares of the economy (see Chart 1).

Figure 1 Correlation between searches related to unemployment and welfare and retail services as a share of GDP

Sources: Eurostat, Google trends, OECD and authors’ calculations.

Notes: Google trends change is for 2 weeks starting 22 March 2020 relative to forecast based on prior trends.

In normal times, increases in searches for unemployment and welfare correlate strongly with increases in job searches (a correlation of 0.46, based on monthly observations during the period 2010-2019 across 38 economies). In the aftermath of the 2008-09 Global Crisis, this correlation even increased to 0.72 (based on country-month data for October 2008-March 2009). However, in the two weeks starting on 22 March 2020, the correlation between changes in searches related to unemployment and welfare and job searches dropped to 0.01 (see also Chart 2 which shows cross-country correlation at any point in time). This drop is approximately the same in countries that were able to pledge wage subsidies to affected firms versus countries with more limited fiscal policy responses to date (where incentives to look for jobs may, in theory, be stronger).1

Figure 2 Correlation between searches related to unemployment and welfare and job searches

Sources: Google trends and authors’ calculations.

As a result, some workers in countries with more flexible labour markets may be hit particularly hard in the near term. This may be more so in cases where governments lack the administrative capacity (or the fiscal space) to support the incomes of affected individuals. Job losses and income reductions are likely to be most severe in cases where fewer employees have permanent contracts or many are self-employed (in the formal or informal sector), or in areas where many people work for small firms. These groups of workers are likely to be more difficult to successfully target via indirect fiscal measures, such as subsidising firm payrolls (see Chart 3).

Figure 3 Share of those employed on permanent contracts, self-employed and those working for small firms, 2016 (per cent)

Sources: Life in Transition Survey 2016 and authors’ calculations.

Notes: Small firms are defined as those with less than 5 people.

In contrast, a fairly high proportion of adults in middle-income economies in Europe, Central Asia, and the Southern and Eastern Mediterranean derive their income from the state (see Chart 4). This includes those directly employed by the state in the public sector (health, education, public administration), or those working for state-owned enterprises, as well as the retired. Together these categories account for around half of respondents in a representative household survey run by the EBRD and the World Bank in 2016 (Life in Transition Survey).

Figure 4 Share of adult population dependent on the state for their income, 2016 (per cent)

Sources: Life in Transition Survey 2016 and authors’ calculations.

Notes: Sum of primary respondents employed in the public sector or state-owned enterprises, those not currently working but actively looking for work, retired and students, as a share of interviews conducted per country.

Those paid by the state (directly or via state enterprises) are most likely to be shielded from the employment effects of the crisis. Indeed, Google search query data show that in economies with state-dominated employment, searches for unemployment and welfare increased significantly less than in economies with private-sector dominated employment (see Chart 5).

Figure 5 Google searches related to unemployment and benefits saw smaller increases where state-owned enterprises (SOEs) account for a larger share of employment

Sources: Life in Transition Survey 2016, Google trends and authors’ calculations.

Notes: Google trends change is for 2 weeks starting 22 March 2020 relative to forecast based on prior trends.

Low-income households are likely to be disproportionately hit by the Covid-19 crisis. Recent studies for the UK and US show that workers on low incomes are more likely to have lost their job because of the pandemic. They also show that the self-employed, those not paid a salary, and workers with variable hours chosen at their employers’ discretion (e.g. zero-hours contracts) are more likely to have been negatively affected by the downturn (Adams-Prassl et al 2020a, 2020b). Poorer workers are also likely to be concentrated in the occupations and sectors most affected by closures (such as retail services) and are least likely to be able to work from home. A recent study estimates that in the US only around a third of jobs can be performed from home, and that these employees tend to be paid around 55% more than others (Dingel and Neiman 2020).

In the middle-income economies of Europe, Central Asia, and the Southern and Eastern Mediterranean, poorer individuals are also least likely to have permanent contracts, are more likely to be self-employed, and are more likely to work for small firms (see Chart 6). Men are less likely to have permanent contracts than women and this difference is especially pronounced in the bottom half of the income distribution. Men are also more likely to be self-employed (this trend holds throughout the income distribution).

Figure 6 Share of those employed on permanent written contracts, self-employed and those working for small firms, by household income decile, 2016 (per cent)

Sources: Life in Transition Survey 2016 and authors’ calculations.

Notes: Small firms are defined as those with less than 5 people.

Evidence from the 2008-09 Global Crisis shows that lower-income households in ‘Emerging Europe’ and Central Asia were hit the hardest at that time. For instance, poorer households were more likely to experience job losses, reductions in hours, and closures of family businesses (EBRD 2011). This time, the polarising effect of the crisis may be compounded by the extent of the labour market freeze under lockdowns. Those on temporary contracts and the self-employed are likely to be hit hard, while those employed by the state will remain shielded. Unless governments find ways to cushion the impact on the vulnerable segments of the private sector, the economic fallout from Covid-19 may increase people’s preferences for public-sector jobs (and the associated financial security). Through this channel, the Covid-19 epidemic may further tilt middle-income economies towards state employment, potentially exacerbating the problems associated with a lack of entrepreneurship and innovation.

References

Adams-Prassl, A, T Boneva, M Golin and C Rauh (2020a), “Inequality in the Impact of the Coronavirus Shock: New Survey Evidence for the UK”, Cambridge-INET Working Paper Series No: 2020/10, Cambridge Working Papers in Economics: 2023.

Adams-Prassl, A, T Boneva, M Golin and C Rauh (2020b), “Inequality in the Impact of the Coronavirus Shock: New Survey Evidence for the US”, Cambridge-INET Working Paper Series No: 2020/09 Cambridge Working Papers in Economics: 2022.

Dingel, J and B Neiman (2020), “Who can work from home?”, University of Chicago.

European Bank for Reconstruction and Development (EBRD) (2011), “Crisis: The people’s perspective”, Transition Report 2011.

European Bank for Reconstruction and Development (EBRD) (2020), “Covid-19: From crisis to recovery”, Regional Economic Prospects, April.

Endnote

1 For coding of policy response by country, see EBRD (2020).

735 Reads