THE EVOLVING LABOR MARKET IMPACTS OF COVID-19 IN DEVELOPING COUNTRIES The early labor market impacts of the COVID-19 pandemic resulted in widespread disruption to livelihoods. Previous analysis showed that between April and July 2020, across a sample of 39 countries, an average of 34 percent of workers stopped work, 20 percent of employees experienced partial or no payments for work performed, and 9 percent changed jobs during the early part of the pandemic.1 This brief discusses how labor markets have evolved since the initial phase of the crisis in the spring and early summer of 2020. It uses harmonized data from high-frequency phone surveys (HFPS) conducted in 33 developing countries and provides information on the changing labor market impacts of the crisis in these countries from the initial phase of the pandemic in April 2020 through December 2020.2 The main findings are: Employment recovered substantially between the second and third quarters of 2020. The average 1. employment rate increased by 10 percentage points, consistent with increases in time spent at the office according to Google Mobility data. There is little evidence of a continued large recovery in employment between the third and final quarters 2. of 2020, with only a 2-percentage point increase during the last two quarters of the year. The average share of households reporting income loss declined by 25 percentage points since the 3. early phase of the crisis but remained high as an average of 40 percent of households reported losses in the third quarter. Sectoral reallocation coincided with the re-opening of many economies and showed a net movement 4. towards commerce and mining, manufacturing, public utilities, and construction. The share of employment in the commerce sector increased by 0.8 percentage points while the sector comprising mining, manufacturing, public utilities, and construction increased by 1.2 percentage points. 1 Khamis, Melanie; Prinz, Daniel; Newhouse, David; Palacios-Lopez, Amparo; Pape, Utz Johann; Weber, Michael. 2021. The Early Labor Market Impacts of COVID-19 in Developing Countries: Evidence from High-Frequency Phone Surveys (English). Policy Research Working Paper No. WPS 9510; COVID-19 (Coronavirus) Washington, D.C.: World Bank Group. 2 The data for this brief looks at countries that collected labor market information in at least two quarters. This is the case for 33 countries overall. For the main labor market indicator, “current work” with 27 countries covering quarter 2 and quarter 3 and 14 countries cover quarter 3 and quarter 4 (Figure 1). For 24 of these countries Google Mobility data and current work were available as well (Figure 2a) and for 9 countries for the last two quarters (Figure 2b). For 18 of these countries complete total income information for comparing pre-pandemic and the last survey were available (Figure 3). For 21 of these countries the current sector of work was available (Figure 4). 1 THE LABOR MARKET IMPACT: EMPLOYMENT RECOVERS As a measure of labor market activity, this section looks occurred in quarter 3, brought employment closer to at respondents’ answers to the question “Last week, that pre-pandemic levels but labor market activity has not is from Monday [ DATE  ] up to Sunday [ DATE  ], did you do yet fully returned to pre-pandemic levels. any work for pay, do any kind of business, farming or other activity to generate income, even if only for one For the 14 countries with responses for quarter 3 and hour?” and aggregates the responses across countries quarter 4 only, the changes in current work increased and quarters, with appropriate survey weights. only slightly: 70.5 percent in quarter 3 to 72.5 percent in quarter 4, constituting a much smaller increase Across the entire set of countries, the share of respon- compared to the earlier quarter of only 2 percentage dents that are “currently working” increased substantially points (Figure 1).3 from the second to the third quarter (Figure 1). For the 27 countries that have information on “current work” We cross-check the reported current work measure for quarter 2 and quarter 3, 54 percent reported currently against the workplace mobility measure from the Google working in quarter 2 while in quarter 3, 64 percent Community Mobility Reports for the different quarters.4 reported currently working, which amounts to a signifi- Figure 2 panels a and b reinforce the observed patterns: cant increase (10 percentage points). For these countries, improvements in workplace mobility and improved 68 percent of respondents reported in quarter 2 to have labor market outcome, measured by current work, worked before the pandemic which constitutes a 14 occurred over the quarters. For quarter 2 to quarter percentage point difference to working in the second 3 and for quarter 3 to quarter 4, the figures show a quarter of 2020. An increase in employment, which positive correlation between workplace mobility and current work. FIGURE 1 In the Latin America and the Caribbean (LAC) region, Impact of COVID-19 on Current Work with data available for countries in quarter 2 and quarter  3, quarter 3 exhibits higher proportions of 100 respondents reporting “currently working” than for 90 quarter 2. For the Sub-Saharan Africa (SSA) region, a 80 similar pattern shows sustained increases in reports of “current work” over quarter 2, quarter 3, and in some 70 countries even for quarter 4. The averages of the LAC 60 and SSA regions, where the data was available, exhibit a clear upward and evolving trend to increases in current Percent 50 work that respondents report. 40 30 In short, current work increased significantly from quarter 2 to quarter 3, but increases flattened 20 out over the last two quarters of 2020. For most 10 countries where data are available, the current 0 work measure is positively related with Google Average 27 countries Average 14 countries Mobility patterns of workplace mobility. This suggests a link between the easing of pandemic Pre-pandemic Q2 Q3 Q4 mobility restrictions and the observed recovery of the labor market. Source: World Bank High-Frequency Phone Surveys 2021. Note: Quarters 2, 3, and 4 for the year 2020 are represented as Q2, Q3, and Q4. 3 As a robustness check, the means were also estimated with clustered standard errors at the geographical administrative 1 unit for each country and the message is largely sustained, even with these larger standard errors. 4 Data for mobility trends for places of work as compared to the baseline, Google COVID-19 Community Mobility Report. 2 FIGURE 2 Google Mobility and Current Work, Q2 vs Q3 change and Q3 vs Q4 change A. Q2 vs Q3 change B. Q3 vs Q4 change 50 MMR 50 40 40 30 NGA 30 PER PRY SLV BOL PHL 20 UGA 20 Current Work Current Work GTM CRI HND ECU MUS BGR 10 DOM 10 IRQ BFA MEX COL BFA MLI MOZ ZWE 0 0 MNG ARG MOZ UGA KHM CHL KHM –10 –10 MNG –20 POL –20 MMR –30 –30 –30 –20 –10 0 10 20 30 40 50 –30 –20 –10 0 10 20 30 40 50 Google Mobility Google Mobility Source: World Bank High-Frequency Phone Surveys 2021; Google COVID-19 Community Mobility Reports 2021. Notes: (1) Panel A—24 countries. Panel B—9 countries. (2) Quarters 2, 3, and 4 for the year 2020 are represented as Q2, Q3, and Q4. INCOMES SHOW SOME SIGNS OF RECOVERY, FIGURE 3 BUT LOSSES REMAIN PREVALENT Total Income Change, Q2 pre-pandemic vs Q3 last survey Turning to income losses, we examine total income changes across two time dimensions and comparison 40 points: (1) How total household income change in 20 quarter 2 compared to before the crisis, and (2) for 0 subsequent quarters, how total household income Percent has changed since the previous round of the survey –20 (Figure 3). –40 Income losses, reported in the later quarters, became –60 less prevalent. The phone survey responses indicate –80 that household income declines were common, with Q2 Q3 Pre-pandemic Last survey two-thirds of households reporting income loss from pre-pandemic levels in quarter 2 and only about 5 Loss Increase percent reporting an income increase. The remaining Source: World Bank High-Frequency Phone Surveys 2021. 28 percent reported no change. In the third quarter, Notes: (1) 18 countries, Q3 countries. Negative percent represents there seemed to be a slight recovery for households the percent of respondents that report income loss, positive on average, with only 41 percent reporting income percent represents the percent of respondents that report losses compared to the previous round of the survey income increase. (2) Quarters 2, 3, and 4 for the year 2020 are represented as Q2, Q3, and Q4. and 8 percent reporting income increases.5 5 Based on 18 countries that report both pre-pandemic income measures and income since the last survey. 3 In short, a large number of households suffered This points to a sectoral reallocation of employment. total income decreases due to the pandemic. Figure 4 shows the percentage increase in the estimated While the share of households reporting income number of workers in each sector. While all sectors declines dropped significantly in the third quar- experienced increases in employment between Q2 and ter, relatively few households reported income Q3, some more than doubled the number of workers. gains. This suggests that the observed increase Again, commerce and mining / manufacturing / public in employment may have offset income losses utilities / construction experienced large increases. from other sources of household income, but was insufficient to translate into widespread income Moreover, looking into the details that comprise the gains. This may be because of changes in the type “other sectors” category, a large relative change occurred of work performed. for transport and communications (130.1 percent), likely EMPLOYMENT REALLOCATES FIGURE 4 ACROSS SECTORS Growth rate in employment by sector, Q2 to Q3 To better understand sectoral reallocation during this 120 time, we aggregate sectors into four categories: (1) agriculture, (2) mining / manufacturing / public utili- 100 ties  construction, (3) commerce, and (4) other ser- /  vices. Sectoral differences across quarters and possibly 80 reallocation of worker flows to and from the sectors occurred over the different quarters, which workers Percent report as their current sector of work.6 Employment 60 reallocations across sectors are an indication of changes in the quality of work, and also the differential income 40 potential in these sectors. 20 In particular, we examine how the share of employ- ment across sectors changed between quarter 2 and 0 quarter  3. The share of employment decreased in Agriculture Mining/ Commerce Other the agricultural sector and the category of “other Manufacturing/ services sectors,” which compromises transportation / com- Public utilities/ Construction munications, financial and business services, and other service activities, by 0.3 and 1.7 percentage Source: World Bank High-Frequency Phone Surveys 2021. points respectively. In contrast, commerce and min- Notes: (1) 21 countries. (2) Quarters 2, 3, and 4 for the year 2020 ing / manufacturing / public utilities / construction saw are represented as Q2, Q3, and Q4. increases between quarter 2 and quarter 3 of 0.8 and 1.2 percentage points. 6 In Khamis et al (2021) documented sectoral changes between the pre-pandemic and the early impacts of the crisis and found that the labor market impact, measured by work stoppage, was more prevalent in the industrial and service sectors than in agriculture. 4 coinciding with the easing of restrictions on mobility. towards commerce and mining  manufactur- /  While this sectoral allocation may have helped to stem ing / public utilities / construction. Furthermore, the tide of income losses observed in the second quarter, there were increases in the transport and com- these new jobs did not appear to be sufficiently produc- munications sectors, which are consistent with tive to lead to widespread gains in household incomes. increased mobility occurring during this time. In short, sectoral reallocation between quarter 2 and quarter 3 coincided with the closing and re-open- ing of many economies and showed a movement This brief was prepared by Melanie Khamis, Daniel Prinz, David Newhouse, Amparo Palacios-Lopez, Utz Pape, and Michael Weber. The production and publication of this report has been made possible through financial support from the World Bank’s Jobs Umbrella Multi-donor Trust Fund (MDTF), which is supported by the UK’s Foreign, Commonwealth & Development Office/ UK AID, the Governments of Austria, Germany, Italy, Norway, the Austrian Development Agency, and the Swedish International Development Cooperation Agency. The team is also grateful to the Poverty and Equity Global Practice and the Data for Goals group for collecting, harmonizing, and sharing the phone survey data. Aggregate indicators from the high frequency phone sur- veys are available at the High Frequency Phone Survey Dashboard at: https://www.worldbank.org/en/data/interactive/2020/11/11/ covid-19-high-frequencymonitoring-dashboard. All Jobs Group’s publications are available for free and can be accessed through the World Bank or the Jobs and Development Partnership website. Please send all queries or feedback to Jobs Group. Join the conversation on Twitter: @WBG_Jobs #Jobs4Dev. 5