HOW DID THE COVID-19 CRISIS AFFECT DIFFERENT TYPES OF WORKERS IN THE DEVELOPING WORLD? The COVID-19 pandemic is the worst global macroeconomic shock since the Great Depression. This brief reports which groups of workers have been hit hardest by the jobs impact following the economic fallout of COVID-19 in developing countries.1 It complements an earlier study by Khamis et al. (2021) that shows that the onset of the pandemic had major and pernicious adverse effects on the livelihoods of workers across about 40 developing countries. This brief reveals the following: • Larger shares of female, young, less educated, and urban workers stopped working, with gender differences being particularly pronounced. Although women work in different sectors than men, gender gaps in work stoppage stemmed mainly from differences within sectors rather than differential employment patterns across sectors. • Looking at people who remained employed, changes in sector of employment and employment type were similar for all groups except for age, where young workers saw a slightly larger decline in industrial employment than older workers. • Employment increased between April and October with larger gains for the groups with larger initial job losses. For most groups, however, the employment recovery was not nearly enough to reach pre‑pandemic levels. • Phone surveys give a generally accurate picture of group disparities in employment rates following the onset of the crisis and are proving to be a valuable tool for monitoring differential impacts of the crisis on workers. This analysis fills an important knowledge gap, study examines whether this is also the case in as there is limited systematic evidence about the crisis’ developing countries, also drawing on information impact on different types of workers in developing from about 40 countries across five regions. This countries. Evidence from developed countries points to information sheds light on the extent to which the traditionally disadvantaged workers in the labor market crisis is exacerbating traditional disparities and the being disproportionately affected by the pandemic (Lee potential need for policy interventions targeted to et al., 2021; Fairlie et al., 2020). Our recently released particularly affected groups. 1 We calculate statistics for each individual country using the household weights constructed by the World Bank and national statistics offices. The cross-country averages are calculated as simple averages between the 40 country-level values. 1 Work stoppage rates in April-June 2020 ranged evidence so far from developed countries shows that, between 28% and 37% and the gender gap was in general, both women and men increased the amount larger than differences in work stoppage by age, of time allocated to these activities, but the extra time education, or location (Figure 1). Women were was larger for women. At the same time, the present 8 percentage points more likely than men to stop crisis differs from previous recessions in that sectors working in the early stage of the pandemic, young and employing larger shares of women—such as travel, low-educated workers were 4 percentage points more restaurant, and other services—have been more affected likely to stop working than adult and high-educated due to social distancing measures. workers respectively, while the rate of work stoppage for workers in urban locations was 3 percentage points Not much of the gender gap in work stoppage is higher than for rural workers. In sum, female, young, explained by two often cited factors—care responsibilities less-educated urban workers were most likely to stop and gender segregation of pre-pandemic sectors of working in the initial phase of the pandemic. employment (Figure 2). Children learning activities at home explain little of the gender gap. This could be Women were more likely to stop working than related to cross-country differences in the way children men employed in the same economic sectors. participated in remote learning activities while schools The larger penalty of the pandemic on female rather were closed. Looking at the contribution of pre-pandemic than male employment have been linked to gender sectors of employment, they only explain 7% of the differences in care and domestic responsibilities as observed gender gap in work stoppage, with some well as to gender differences in pre-pandemic sectors sectors contributing positively to the gap—sectors that of employment.2 The closing of schools and nurseries typically have a larger share of female employment such since the start of the pandemic meant an increase in as other services and commerce, while some others the time allocated to housework and childcare. The contributed negatively—sectors with larger shares of male employment, such transport and communications FIGURE 1 and construction. These results indicate that gender differences in pre-pandemic sectors of employment Rate of work stoppage by group also explain little of the gender disparity in work stoppage. Instead, the gender gap was primarily 37% 36% 35% caused by female workers being much more likely to 33% stop working than men working in the same sectors. 31% 31% 28% 28% Employment is not the only important labor market outcome, as the type of job people have is related to their productivity. Looking at people who remained employed, changes in sector of employment and on employment type were similar for all groups except for age. Employment fell slightly more for youth than adults in the industrial sector, but overall, we find no major differences. Employment increased between April and October with larger gains for the groups with larger initial job losses. Data on employment between Women Men Young Adults Low High Urban Rural April and October is only available for a much smaller Gender Age Education Location set of 17  countries. However, Figure 3 shows that female, less educated, young, and, to a lesser extent, Source: World Bank High-Frequency Phone Surveys 2021. urban workers had larger employment gains during Notes: Statistics using Wave 1 of the HFPS for 40 countries when grouping by gender and age, 30 when grouping by education, this period. Nonetheless, for most groups, the and 35 by location. The rate of work stoppage is obtained as employment recovery was not nearly enough to simple averages between country-level values. reach pre‑pandemic levels. In August-October 2020, 2 Adams-Prassl et al., 2020; Del Boca et al., 2021; Albanesi et al., 2021; Alon et al., 2020; Alon et al., 2021. 2 FIGURE 2 FIGURE 3 Oaxaca-Blinder decomposition of the gender Change in employment between April difference in work stoppage and October Children learning activities 38% Women –6% Gender Other services 23% Men Public Administration –2% 34% Finance & Business Young 5% Age Transport & Communication 29% Adult –5% Commerce Low 44% Construction Education educated –12% Public utility services High 28% educated –21% Manufacturing 30% Urban 2% Location Mining –1 0 1 2 28% Rural 5% Source: World Bank High-Frequency Phone Surveys 2021. Notes: Bars show the portion of the total gender difference in work stoppage explained by different observed characteristics. Difference Aug.-Oct. vs Apr.-June Model run using Wave 1 of the HFPS. Model controls for young, Difference Aug.-Oct. vs pre-pandemic low-educated, urban indicator variables and country fixed effects. Omitted sector: Primary activities. Weights were adjusted Source: World Bank High-Frequency Phone Surveys 2021. to add up to 1 in each country. Included countries: Bulgaria, Notes: Statistics using 17 countries with available information Bolivia, Chile, Colombia, Costa Rica, Dominican Rep., Ecuador, when grouping by gender and age, 14 when grouping by Croatia, Madagascar, Peru, Philippines, Paraguay, South Sudan. education and 13 by location. The contribution of all sectors together is 0.006. Female-Male observed difference is 0.091. female and male employment rates were still 6% and mainly household heads, provide reasonably accurate 2% below their pre-pandemic levels, respectively. Adult measures of disparities in employment levels by gender, employment was 5% below, while for young workers education, and urban/rural location following the onset the employment rate was 5% above the pre-pandemic of the crisis, though they perform less well in capturing value. Because of limitations in the data, it is difficult disparities between age groups. to know if the jobs gained were of similar quality to those lost. As an illustration, Figure 4 shows the comparison of the group differences by gender, age, and location in Phone surveys give an accurate picture of group Kenya. The left bar shows the estimates for all household disparities in employment rates following the members, the middle bar shows the estimates for onset of the crisis. The HFPS data we are using greatly survey respondents only, and the right bar shows the over-represent sample heads and therefore overestimate estimates for survey respondents that were reweighted employment rates. However, in five countries the in an attempt to correct for sampling bias. The gender surveys were careful to collect labor market information and urban/rural differences are similar in all three from all adult members and not just one. In these five cases. The age difference improved when applying the countries, the data from one respondent, even if it is reweighting technique, but it is still far from the result 3 for all household members. This indicates that in Kenya, FIGURE 4 the reweighted HFPS delivers an accurate estimation of Comparison of group differences in employment differences across gender and locations but the difference levels during-COVID between different samples by age is not precisely estimated. Similar results were in Kenya found for the other four countries. In sum, the HFPS, when used with appropriate caution, is proving to A. By gender, women vs men be a most valuable tool for the timely monitoring of group differences in the employment impact All of the pandemic. The COVID-19 pandemic shock hit some groups in the labor market disproportionately harder. In particular, Survey respondents female, young, less-educated and to a lesser extent urban workers bore the brunt of the burden of the labor market impact from the COVID-19 shock. Reweighted survey There was some job recovery by these same groups in the countries where data were available but −50 −40 −30 −20 −10 0 not nearly to an extent enough to compensate Percentage for the initial collapse in employment and income. These results shed new light on the labor market conse- B. By age, young vs adults quences of the COVID-19 crisis in developing countries, and suggest that real-time phone surveys, despite All their lack of representativeness, are a valuable source of information to measure differential employment impacts across groups during a crisis such as the current pandemic. Survey respondents Reweighted survey −50 −40 −30 −20 −10 0 Percentage C. By location, urban vs rural All Survey respondents Reweighted survey −50 −40 −30 −20 −10 0 Percentage Source: World Bank High-Frequency Phone Surveys 2021. Notes: Group differences in percentages. Data from the World Bank COVID-19 Rapid Response Phone Survey collected between May and June of 2020. 4 REFERENCES Adams-Prassl, A., Boneva, T., Rauh, C., and Golin, M. (2020). “Inequality in the Impact of the Coronavirus Shock: Evidence from Real Time Surveys.” Journal of Public Economics, Volume 189, September 2020, 104245. Albanesi, J., and Kim, J. (2021). “The Gendered Impact of the COVID-19 Recession on the US Labor Market.” Working Paper 28505. National Bureau of Economic Research, Cambridge, MA. Alon, T. M., Doepke, M., Olmstead-Rumsey, J., and Tertilt, M. 2020. “This Time It’s Different: The Role of Women’s Employment in a Pandemic Recession.” Working Paper 27660. National Bureau of Economic Research, Cambridge, MA. Alon, T., Coskun, S., Doepke, M., Koll, D., and Tertilt, M. 2021. “From Mancession to Shecession: Women’s Employment in Regular and Pandemic Recessions.” Working Paper 28632. National Bureau of Economic Research, Cambridge, MA. Del Boca, D., Oggero, N., Profeta, P., and Rossi, M. 2020. “Women’s and Men’s Work, Housework and Childcare, Before and During COVID-19.” Review of Economics of the Household, 18: 1001–17. Fairlie, R. W., Couch, K., and Xu, H. 2020. “The Impacts of COVID-19 on Minority Unemployment: First Evidence from April 2020 CPS Microdata.” Working Paper 27246. National Bureau of Economic Research, Cambridge, MA. Khamis, M., Prinz, D., Newhouse, D., Palacios-Lopez, A., Pape, U., and Weber, M. 2021. “The Early Labor Market Impacts of COVID-19 in Developing Countries: Evidence from High-Frequency Phone Surveys.” Policy Research Working Paper No. 9510, World Bank, Washington, DC. Kugler, M., Viollaz, M., Duque, D., Gaddis, I., Newhouse, D., Palacios-Lopez, A., and Weber, M. 2021. “How Did the COVID-19 Crisis Affect Different Types of Workers in the Developing World?” Policy Research Working Paper No. 9703, World Bank, Washington DC. Lee, S. Y., Park, M., and Shin, Y. 2021. “Hit Harder, Recover Slower? Unequal Employment Effects of the COVID-19 Shock.” NBER Working Paper No. 28354. National Bureau of Economic Research, Cambridge, MA. This brief was prepared by Maurice Kugler, Mariana Viollaz, Daniel Duque, Isis Gaddis, David Newhouse, Amparo Palacios-Lopez, 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. 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