POLICY BRIEF FEBRUARY 2021 IMPACTS OF COVID-19 ON LABOR MARKETS AND HOUSEHOLD WELL-BEING IN PAKISTAN: EVIDENCE FROM AN ONLINE JOB PLATFORM Emcet O. Taş1, Tanima Ahmed1, Norihiko Matsuda2, and Shinsaku Nomura1 SUMMARY This brief uses the administrative database of Pakistan’s largest online job platform and an online COVID-19 survey to examine the gender impacts of the COVID-19 pandemic on labor markets and other well-being indicators. The analysis shows that the pandemic led to an unprecedented level of economic insecurity, resulting in widespread job loss, business closures, slowdown in business activity, and reduced working hours. The sectors where women are more likely to be employed, such as education and health, were more severely affected, yet the post-pandemic recovery was faster for males. The pandemic has also led to a disproportionate increase in women’s unpaid care work, as well as increasing their reported rates of stress, anxiety and exposure to violence. These findings suggest that impacts resulting from COVID-19 might lead to further declines in women’s participation in the economy in Pakistan, where women’s labor force participation is already among the world’s lowest. BACKGROUND The COVID-19 pandemic has caused unprecedented laid off in downswings and hired in upswings (Elson 2010). disruptions to labor markets around the world. As lockdowns Since women are overrepresented in insecure, low-paying were put in place to mitigate spread of the disease, economic jobs, they are typically the first to be fired during crises (Cho activity plummeted, resulting in large drops in employment and Newhouse 2011), but female labor force participation in virtually every country. International organizations have can also grow if households that need the additional income predicted that 158 to 242 million full-time jobs will be lost ask female family members to look for work (Rubery 2010; globally, and 71 million people will fall below the extreme Starr 2014). poverty line in low-income countries (Asian Development Bank 2020; World Bank 2020). Economic crises can also worsen existing gender inequalities and reinforce gender roles within Economic shocks of this magnitude can have differing households. Loss of income and employment often impacts on men and women, even if the net effects of the requires households to adjust expenditures to cope with shock conceal these differences. Workers concentrated the crisis. Since women usually have responsibility for the in sectors that are insulated from an economic shock survival of household members, they may cut back on their may be protected during the crisis (Rubery 2010; Rubery consumption or increase unpaid work as an alternative to and Rafferty 2013) and since most labor markets exhibit purchasing household goods (Elson 2010). In countries occupational segregation by gender, effects on men and where social norms perceive women as caregivers and men women can be different. Social norms can also affect the as breadwinners, women might face an added burden of employment of men and women during crises. If norms unpaid work at home, including childcare and eldercare dictate preservation of men’s “breadwinner” status, when (Karamessini and Rubery 2013). During COVID-19, lockdowns jobs get scarce, women are likely to bear the brunt, getting that required adults to work from home and children to attend school remotely have added disproportionate 1 World Bank, 2 Florida International University amounts of unpaid work on women (United Nations 2020). This note examines the gender impacts of the COVID-19 market. Analysis of these data can inform decision making pandemic on labor markets and other well-being indicators while traditional labor market data are being collected. in Pakistan, where women’s labor force participation is This is important during an evolving pandemic, when quick among the world’s lowest. Although women account for decisions on response and recovery are needed. The main 49.2 percent of Pakistan’s population (Pakistan Bureau limitation of this dataset is that it represents a top segment of Statistics, Labor Force Survey), their labor market of the labor market, primarily well-educated young people participation is among the lowest in the world, especially in who live in cities and have access to the Internet.2 urban areas. The female labor force participation (FLFP) rate in Pakistan increased steadily from 13 percent in 1992 to 25 The second data source is an online COVID-19 survey of percent in 2015, before declining back to 23 percent in 2018. registered users of the Rozee.pk platform during July–August It remains considerably lower than the FLP for men, which 2020. This specialized survey includes detailed information has fluctuated around 82 percent for nearly three decades about the socioeconomic status and coping strategies of (Amir et al. 2018; Pakistan Bureau of Statistics 2019). Most about 9,000 jobseekers and 500 employers.3 The survey of the increase in FLFP from 1992 to 2015 was fueled by uses Rozee’s registered users as its sampling frame, so it may unpaid work in agriculture in rural areas, while FLFP in urban not be representative of the entire labor force in Pakistan. areas remained almost unchanged, only rising from about 7 Moreover, the number of observations is not uniform percent to 11 percent. Many explanations have been offered across questions because different parts of the survey were for the low FLFP in Pakistan’s urban areas, including limited administered to three subsamples of users to manage survey human capital, concerns about women’s safety, mobility length, and participants also had the choice of skipping the constraints, workplace discrimination, marriage, and questions. disproportionate domestic responsibilities (Amir et al. 2018). The rest of this note discusses descriptive findings using Impacts resulting from COVID-19 might lead to further both datasets, as well as multivariate results from linear and declines in women’s participation in the economy, as well as multinomial probability models included in the Appendix. putting the limited gains of the last few decades at risk. IMPACTS ON LABOR DEMAND DATA DECLINE IN JOB ADVERTISEMENTS This brief uses two data sources to examine labor market The pandemic resulted in a sudden drop in labor demand, as trends pre- and post-COVID-19, and to identify the measured by the number of job advertisements, increasing multidimensional impacts of the pandemic on businesses the competition for jobs. When countrywide lockdown and households. COVID-19 was first confirmed in Pakistan at measures were introduced in mid-March 2020, the number the end of February 2020, and the cumulative number of of job advertisements plunged by 76 percent compared to confirmed cases hit half a million by December 2020.1 As a the same period in 2019 (Figure 1a). Given that the number containment measure, federal and provincial governments of job postings in February 2020 was at 91 percent of the closed schools in mid-March and started a lockdown on previous year, the drop is clearly associated with the COVID-19 March 21, 2020. The first lockdown lasted until May 9, 2020; lockdown. With fewer jobs available, the number of job thereafter, sporadic temporary lockdowns continued until applications fell sharply to 33 percent of the previous year the writing of this brief in early 2021. The analysis in this note in the last week of March 2020, while the average number covers the period until August 2020. of applications per job increased, especially after March 18 The first data source used in this brief is the administrative (Figure 1b). In short, the labor market became increasingly database of Pakistan’s largest online job platform. Rozee.pk competitive for jobseekers, a pattern that persisted even contains information on about 75,000 job advertisements when the number of job advertisements recovered slightly and 7.5 million job applicants for the periods January– to 36 percent of the previous year by the end of August 2020 August 2019 and January–August 2020. The administrative (Figure 1a). data provide detailed information on job advertisements placed by employers and job applications submitted by jobseekers, therefore capturing real transactions in the labor 2 See Matsuda, Ahmed, and Nomura (2019) for a more detailed discussion about the demographic comparison between the Pakistan labor force survey and Rozee.pk’s online platform users. 1 Government of Pakistan. “Pakistan Confirmed Cases.” https://covid.gov.pk/stats/pakistan 3 The active users of Rozee.pk were sent reminders to participate in the survey until the (accessed January 4, 2021). desired number of observations was reached. 2 | POLICY BRIEF Figure 1. Job Postings and Applications (year-over-year comparison), January 1–August 30, 2020 a. Ratio of job postings in 2020 to 2019 (year-over-year b. Number of applications and applications per job (year- comparison) over-year comparison) 1.79 1.78 1.50 1.50 1.43 1.43 1.10 0.97 1.00 0.91 1.00 0.98 0.93 0.88 0.78 0.88 0.72 0.50 0.50 0.33 0.33 0.00 0.24 0.00 Jan 1 Mar 1 May 1 Jul 1 Sep 1 Jan 1 Mar 1 May 1 Jul 1 Sep 1 Day of Week (2020) Day of Week (2020) Applications Applications per Job c. Number of job postings in selected sectors (year-over- d. The number of job postings (year-over-year year comparison) comparison) that prefer male versus female applicants 3.00 2.44 2.50 2.35 2.14 2.02 2.00 2.00 1.91 1.58 1.50 1.46 1.50 1.10 1.08 1.06 1.00 1.01 1.03 1.00 0.95 0.85 0.84 0.82 0.50 0.50 0.28 0.00 0.14 0.14 0.00 Jan 1 Mar 1 May 1 Jul 1 Sep 1 Jan 1 Mar 1 May 1 Jul 1 Sep 1 Day of Week (2020) Day of Week (2020) Education/Health ICT Professional Manufacturing Male Female No Preference Source: Rozee administrative database, 2019–2020. N=74,663. The fall in job postings affected all population groups, but in employment after the COVID-19 lockdowns. Specifically, the industries in which female employment is concentrated in late March and April 2020, job postings in manufacturing, were disproportionately affected.4 Although Pakistan’s education, health, and professional services all dropped to 24 labor market is male-dominated, females employed percent of the previous year’s level. In contrast, traditionally outside of the agriculture sector are largely concentrated male-dominated industries were more resilient—for in manufacturing as craft and trade-related workers, in example, the drop in the information and communications education as professionals, and in human health and social technology (ICT) sector stopped at 38 percent of the 2019 work activities as technicians and associate professionals level (Figure 1c). (Pakistan Bureau of Statistics, Labor Force Survey, Table-20). On the Rozee platform, these sectors saw the sharpest drops There was no noticeable difference between the number of jobs that sought male versus female applicants, but the post-pandemic recovery was faster for jobs that sought male applicants. A unique feature of the Rozee job portal is that 4 Lahore, Karachi, Islamabad, and Rawalpindi, which account for approximately 85 percent it allows employers to specify whether they are looking for of the job postings on the portal, all experienced large plunges in job postings after March 18, 2020. Jobs requiring secondary or higher secondary certificates reduced vacancies to a male or a female candidate to fill their vacancies. Many 10 percent and 16 percent relative to the previous year, while job postings requiring higher employers do not indicate a preference, but those that do education (college diploma, bachelors, and postgraduate degrees) remained relatively more resilient at 28 percent relative to the previous year. are twice as likely to seek male workers rather than females FEBRUARY 2021 | 3 (20 percent of jobs advertised seek males, compared to 9 by female workers (such as community health workers, percent that seek females). Job postings that sought men vaccinators, and midwives) did not operate fully during the and women declined in the week of March 25, 2020, to pandemic (World Health Organization 2020). 14 percent and 17 percent of the previous year’s levels, respectively. In August 2020, however, jobs preferring male The education industry is another sector where females’ applicants showed a more robust rebound than those formal employment is heavily concentrated, and where preferring female applicants. In August, male-preferred the impacts of the pandemic have been severe.6 The online jobs rebounded to 149 percent of the previous year, while COVID-19 survey of employers indicates that 69 percent of female-preferred jobs rebounded to 114 percent (Figure 1d). the companies in the education sector were temporarily The sectoral composition suggests that the first wave of job closed—the highest among all industries—and another recovery mainly came from professionals and ICT, both male- 12 percent were permanently closed (Figure 2a).7 Those dominated sectors, while the second wave of job recovery that remained open during the pandemic reported a fall in largely came from manufacturing. their revenues (67 percent), slowdown in their activities (50 percent), and decline in the demand for their services (50 percent) (Figure 2b). Further, they reported difficulties in REDUCED BUSINESS ACTIVITY operation due to workers’ absenteeism (67 percent) (Figure The online COVID-19 survey reveals more about the 2c).8 As a result, the education industry implemented some pandemic’s impacts on businesses across Pakistan, including of the toughest measures against their employees, together closures, reduced working hours, and slowdown in with the retail and wholesale and health industries. Two- business activity. The survey of employers shows that 32 thirds of the education firms cut working hours, one-half percent of the businesses that responded were temporarily delayed wage payments to their employees, one-third cut closed due to the pandemic, while another 8 percent were wages, and another one-third implemented permanent permanently closed (Figure 2a). Of the firms that remained layoffs (see Figure 2d). open, 51 percent reduced their working hours (Figure 2d), 49 percent of the firms reported a reduction in the demand IMPACTS ON LABOR SUPPLY for their products, and 66 percent showed a decline in their revenues (Figure 2b). With the decrease in demand INCREASE IN UNEMPLOYMENT AND JOB SEARCH and revenue, 25 percent of the open businesses reported The onset of the pandemic was accompanied by a noticeable slowing down their activities either very much or immensely decline in employment and an increase in the number of (multivariate models in Appendix Tables 1–3 show the people looking for jobs. The jobseekers sampled for the determinants of business activity). online COVID-19 survey confirmed that the labor market These impacts were disproportionately felt in the health became tighter, as previously indicated by administrative industry, where women’s formal wage employment is heavily data from Rozee.pk. The share of individuals working post- concentrated. In the online COVID-19 survey, 53 percent of COVID-19 fell from 65 to 51 percent of respondents, while the firms in the health industry reported being temporarily those looking for jobs increased from 68 to 88 percent of closed—among the highest rates for firms reporting respondents (Figure 3). The magnitude of these changes is temporary closures (Figure 2a). In addition, 89 percent of validated by multivariate analysis (Appendix Table 4) of the firms in the health industry reported worker absenteeism as change in the probability of working and job search before a problem during the pandemic, which is the highest among and after COVID-19. firms reporting on that metric (Figure 2c). Multivariate These labor market patterns were triggered by substantial job findings confirm these patterns (see Appendix Tables 1 losses across urban Pakistan, with disproportional impacts and 2).5 Taken together, these findings are consistent with on women and other vulnerable groups. Fifty-two percent of reports documenting that healthcare services unrelated to COVID-19 slowed down, many primary health care services were suspended, and the low-paid health care jobs held 6 According to the World Development Indicators (WDI), women’s share in primary school teaching was 55.6 percent in 2019. The WDI indicator includes both public and private schools, while most education institutions on the Rozee platform were private. 7 The multinomial logit estimates suggest that educational institutes were 45 percentage 5 The multinomial logit estimation controlling for firm size, industry, and location shows points less likely than the reference ICT industry to be open, as well as 44 percentage points that, compared to ICT businesses (reference category), firms in the health industry were more likely to be temporarily closed (Appendix, Table 1). 30 percentage points more likely to be temporarily closed in the period between April and 8 The linear probability model shows that, compared to ICT businesses (reference category), August 2020 (Appendix, Table 1). In addition, in a linear probability model, firms in the health firms in the education industry were 55 percentage points more likely to report slowdown industry were 30 percentage points more likely than the reference ICT sector firms to report in activities and 46 percentage points more likely to report worker absenteeism as an issue worker absenteeism as a problem during the pandemic (Appendix, Table 2). (Appendix, Table 2). 4 | POLICY BRIEF Figure 2. The condition of firms that remained in operation during COVID-19 b. Share of open firms that experienced low demand, a. Share of firms by operational status reduced revenue and activity slowdown All 60 32 8 All 49 66 25 ICT 68 25 7 ICT 41 58 20 All 60 32 8 All 49 66 25 Manufacturing 64 27 9 Manufacturing 59 71 26 ICT 68 25 7 ICT 41 58 20 Retail and wholesale 78 22 0 Retail and wholesale 71 86 43 Manufacturing 64 27 9 Manufacturing 59 71 26 Hotel, restaurant, food services 59 35 6 Hotel, restaurant, food services 60 60 25 Retail and wholesale 78 22 0 Retail and wholesale 71 86 43 Transportation 73 18 9 Transportation 71 65 13 Hotel, restaurant, food services 59 35 6 Hotel, restaurant, food services 60 60 25 Finance 50 38 12 Finance 39 36 33 Transportation 73 18 9 Transportation 71 65 13 Education 19 69 12 Education 50 67 50 Finance 50 38 12 Finance 39 36 33 Health 47 53 0 Health 33 75 17 Education 19 69 12 Education 50 67 50 Other sectors 59 35 6 Other sectors 50 80 34 Health 47 53 0 Health 33 75 17 Other sectors Open Temporarily 59 closed Permanently 35 closed 6 Demand lowered Revenue lowered Activity: slowed a lot Other sectors 50 80 34 Open Temporarily closed Permanently closed d. Share of openDemand lowered Revenue lowered Activity: slowed a lot firms that laid off workers, cut salary/ c. Share of open firms that experienced input disruptions All 59 81 working hoursAll and/or 28 22delayed 28 wages 51 30 ICT 52 81 ICT 25 22 18 26 47 All 59 81 All 28 22 28 30 51 Manufacturing 64 77 Manufacturing 34 18 31 44 47 ICT 52 81 ICT 25 22 18 26 47 Retail and wholesale 57 100 Retail and wholesale 33 33 17 50 67 Manufacturing 64 77 Manufacturing 34 18 31 44 47 Hotel, restaurant, food services 50 88 Hotel, restaurant, food services 20 0 50 20 40 Retail and wholesale 57 100 Retail and wholesale 33 33 17 50 67 Transportation 59 100 Transportation 35 29 24 24 47 Hotel, restaurant, food services 50 88 Hotel, restaurant, food services 20 0 50 20 40 Finance 69 75 Finance 27 33 31 25 69 Transportation 59 100 Transportation 35 29 24 24 47 Education 67 0 Education 29 33 50 33 67 Finance 69 75 Finance 27 33 31 25 69 Health 89 100 Health 11 11 22 44 63 Education 67 0 Education 29 33 50 33 67 Other sectors 64 79 Other sectors 28 26 43 24 55 Health 89 100 Health 11 11 22 44 63 Worker absenteeism Rise in price of raw materials (if used) Temporary layo Permanent layo Delayed wage Wage cut Hour cut Other sectors 64 79 Other sectors 28 26 43 24 55 Worker absenteeism Rise in price of raw materials (if used) Temporary layo Permanent layo Delayed wage Wage cut Hour cut Source: Rozee online COVID-19 survey of employers. N=433 (panel a), N=216 (panel b), N=126 (panel c), and N=263 (panel d). the individuals who reported working before COVID-19 were Figure 3. Share of respondents working and looking for no longer working when the online survey was conducted jobs, before and after COVID-19 in July–August 2020 (Figure 5). Of this group, 46 percent 100 88 91 87 experienced permanent layoffs, 13 percent experienced 80 68 75 69 65 66 temporary layoffs, and 26 percent experienced cuts in their 60 51 54 55 salary or working hours (Figure 4). Women were more likely 39 40 to lose their jobs than men—a greater share of previously 20 employed female respondents (59 percent) reported that they lost their jobs after the pandemic, compared with 0 Before After Before After Before After previously employed male respondents (50 percent). Job lockdown lockdown lockdown lockdown lockdown lockdown All Women Men losses also hit other vulnerable groups, including those Working Looking for a job living in secondary cities, as well as low-skilled or entry-level Source: Rozee online COVID-19 survey of jobseekers. N=2,388 (before lockdown) workers who had the highest rate of job loss (managers had and N=6,182 (after lockdown). the lowest).9,10 Job losses were substantial in the education sector, which losing their jobs. Meanwhile, 79 percent of those working employs a high share of women, but male-dominated sectors in hotels, restaurants, and food services and 76 percent of such as hotels, restaurants, food service, and transportation those working in transportation lost their jobs (Figure 5). also were hit hard. In the survey of jobseekers, 64 percent The linear and multinomial probability models indicate that of individuals who worked in the education sector reported people who worked in the female-dominated education sector were 23 percentage points more likely to lose their jobs compared to those in the ICT sector. Those working in 9 The effect of the pandemic on job losses was lower in cities like Islamabad, Karachi, Lahore, hotels, restaurants, and food services were 47 percentage and Rawalpindi than in other urban centers (Appendix, Table 5). 10 The multinomial logit estimates show that individuals who worked in shops are 24 points more likely to lose their jobs than those in the ICT percentage points more likely to have experienced permanent layoffs compared to individuals who worked in the ICT sector (reference category). In addition, managers had the lowest sector (Appendix Table 5). likelihood of job loss and the lowest likelihood of permanent layoffs (see Appendix, Table 5). FEBRUARY 2021 | 5 Not only were women not insulated from job losses, they were Figure 4. Share of respondents who had jobs before also less likely to look for jobs during COVID-19. Before the COVID-19, but experienced layoffs or reduction in salary pandemic, women, compared to men, were 12 percentage or working hours points less likely to be employed and 8 percentage points 15 more likely to be looking for jobs, controlling for individual, 27 14 household, and regional characteristics (Appendix Table 24 15 26 4). These preexisting disparities did not insulate women No e ect from job losses during COVID-19—in fact, men and women Permanent layo Men experienced similar rates of decline in employment (14 and Women Temporary layo 15 percentage points; Figure 3). Yet, the increase in the 17 13 All Salary/hours cut share of women looking for jobs after COVID-19 was smaller 12 46 45 than it was for men—the share of women looking for jobs 46 increased from 75 to 91 percent (a 16 percentage point increase), while for men it increased from 66 to 87 percent Source: Rozee online COVID-19 survey of jobseekers. N=1,887 (a 21 percentage point increase) (Figure 3). That women’s All Women Men propensity to look for jobs was lower than men’s, despite No e ect 15 14 15 both groups experiencing similar rates of job loss, suggests of respondents Figure 5. SharePermanent layo 46 were who 45 employed 46 before that the norm of men being the breadwinner and women experienced COVID-19, but Temporary layo job loss 13 17 12 Salary/hours All cut 52 26 59 24 50 27 the homemaker in urban Pakistan, or the belief that men ICT 41 41 52 39 have more rights to jobs during times of scarcity, may have Manufacturing 47 47 44 47 become more prevalent during the pandemic.11 Retail and Wholesale 58 58 67 57 Hotel, restaurant, food services 7979 91 77 Transportation 7676 50 77 IMPACTS ON HOUSEHOLD WELL-BEING Finance 32 32 32 37 30 Education 64 64 70 58 LOSS OF INCOME AND CONSUMPTION Health 44 44 55 39 Agriculture 51 51 60 50 The labor market effects of COVID-19 were accompanied by Other sectors 53 53 51 53 All Women Men coping strategies to mitigate the impact of income loss on household well-being. Due to work loss during COVID-19, Source: Rozee online COVID-19 survey of jobseekers. N=2,201. 68 percent of females and 61 percent of males in the online COVID-19 survey reported a reduction in household income likely to take loans to balance household expenditures and (Figure 6a). Women who were employed during COVID-19 5 percentage points less likely to receive external assistance were 11 percentage points more likely to report a decline (Figure 6b). in household incomes than men (Appendix Table 6). In addition, women who were employed were more likely than INCREASE IN UNPAID WORK men to report redistribution of household expenditures to cope with the loss of income, possibly to support food Respondents, especially women, complemented their consumption by other household members. For example, 93 coping strategies with unpaid work. It is well-established percent of employed women reported reducing spending on that women perform more unpaid work than men globally, meals, durables, clothes, education, or tobacco, as opposed but this disparity is worse in Pakistan than any other country to 89 percent of employed men (Figure 6b), a difference in the world (UN Women 2020). According to UN Women, confirmed by a linear probability model (Appendix Table Pakistani women spend additional 11 hours on unpaid 6). Despite their responsibility for household maintenance household chores and caregiving for every hour spent by on shrunk budgets, women were 10 percentage points less men on the same activities. In the online COVID-19 survey, which covered urban areas, 83 percent of women and 79 percent of men reported that the need for unpaid household 11 In Appendix Table 4, the interaction term between the female and post-lockdown dummies work increased after COVID-19 (Figure 7). A larger share is not significant in the model that includes working status as the dependent variable. This suggests that being a female during the pandemic did not have a statistically significant of women than men report an increase in unpaid work impact on women’s likelihood of employment, which is largely determined by structural gaps regardless of employment status, but this gender gap is the in the urban labor market in Pakistan. However, in the model that includes job search status as the dependent variable, the interaction term between the female and post-lockdown largest between working men and working women. The dummy is negative and statistically significant, suggesting that women were less likely to look linear probability estimates suggest that individuals with jobs for jobs during the pandemic. 6 | POLICY BRIEF Figure 6. Share of respondents who coped with reduced for consumption (by 4 percentage points) (Figure 8). In income and consumption during COVID-19 addition, 89 percent of respondents reported that children a. Household income and consumption were attending school online (Figure 7). The care needs of these household were met through increased unpaid work. 92 90 94 94 94 93 94 94 91 94 94 Households with children and elders reported a greater 94 89 92 59 90 67 59 58 93 60 91 60 89 70 increase in household chores (81 percent) compared to those 52 67 60 51 70 59 63 52 52 75 59 68 58 60 73 60 61 51 50 76 without children or elders (75 percent) (Figure 8). The linear 63 All 52 Currently 75 Currently 68 All 60 73 Currently Currently 61 All 50 Currently 76 Currently probability model (Appendix Table 6) suggests that living in All working Currently not Currently All working Currently not Currently working All Currently not Currently working working not working working not working working not a household with elderly and children is associated with a working working working All All Women Women Men Men 4 percentage point increase in the likelihood of doing more Household income lowered Income not su cient to pay for consumption Household income in Reduced spending lowered Income meals, education, not suclothes durables, cient to andpay for consumption tobacco household chores during COVID-19 than before. However, Reduced spending in meals, education, durables, clothes and tobacco the increase in unpaid work is higher among women (by 7 b. External Assistance 8 percentage points) than it is among men (Appendix Table 6). 7 9 8 8 9 8 7 4 9 8 8 4 4 9 4 4 47 44 50 4 50 46 55 INCREASE IN STRESS AND HOUSEHOLD TENSIONS 40 38 41 55 47 50 50 46 44 40 38 41 Women experienced more anxiety and stress due to All All Currently Currently working Currently Currently not All All Currently Currently working Currently Currently not All All Currently working Currently Currently not Currently COVID-19 than men. Since the COVID-19 outbreak, 76 working All working not working working Women working not working working Men working not working percent of individuals reported having experienced at least Take loan All Women Received assistance from government or NGO Men one of the following: getting angry quickly, getting into Take loan Received assistance from government or NGO Source: Rozee online COVID-19 survey of jobseekers. N=7,599 (panel a) frequent arguments, frequently praying, inability to sleep, and N=7,700 (panel b). inability to concentrate, and feeling anxious or depressed (Figure 9). These symptoms were more prevalent among were 9 percentage points less likely to report an increase in women than men—82 percent of women, compared to 74 household work and this impact was smaller for women who percent of men reported at least one of these symptoms. were working during the pandemic, at about 5 percentage However, these stress symptoms were lower among points. This finding points to the double burden of work individuals who were working (72 percent) than those who faced by working women. were not working (80 percent). As women are responsible for household management, a fall in household income and Women’s care work responsibilities have increased a rise in unpaid work might have led to more stress for them. disproportionately in households with children and elderly The linear probability estimates controlling for individual, members. Individuals living in households with elders and household, and regional characteristics support this point children reported an increased likelihood of income loss (Appendix Table 7). (by 2 percentage points) and insufficient income to pay Men and women both reported an increase in domestic violence during COVID-19, but working women, in particular, Figure 7. Share of respondents who increased time spent experienced more violence than non-working women.12 on household chores and caregiving after COVID-19 Thirty percent of individuals reported having experienced Currently not working 84 77 88 at least one type of violence (humiliation, threat, insult, felt unsafe at home, physical abuse, or physical sexual Men Currently working 75 74 89 All 79 75 89 abuse) during COVID-19 (Figure 10). Working women reported experiencing more violence (26 percent) than non- Currently not working 85 77 90 Women Currently working 81 78 91 All 83 77 91 12 The online survey included safeguards to ensure that respondents were presented the Currently not working 84 77 88 violence questions only after confirming that they were (a) alone or able to maintain privacy of their responses, and (b) willing and able to respond to questions about intra-household Currently working 76 75 90 All relations. Those respondents who reported any type of violence were provided the contact All 80 76 89 information of counseling organizations and shelters in their province. The questions included the following: When people feel stressed and economically insecure during crises, they often Do more household If lived with elders/children, knowingly or unknowingly hurt those around them. Since the beginning of March 2020, chores than before COVID did more caregiving when coronavirus spread across Pakistan, have you experienced any of the following? (a) your If lived with elders/children, children are doing home-based schooling partner saying or doing something to humiliate you in front of others; (b) threaten to hurt or harm you or someone you care about; (c) insult you or make you feel bad about yourself; Source: Rozee online COVID-19 survey of jobseekers. N=7,223 (more chores), (d) hit, slap, kick or do anything else to hurt you physically; (e) physically force you to have N=5,961 (more caregiving) and N=5,229 (home-based schooling). intercourse or force you to perform any other sexual acts against your will. FEBRUARY 2021 | 7 Figure 8. Share of respondents living in households with children and elderly, who coped with reduced income and consumption during COVID-19 81 85 80 75 78 73 8 4 9 7 5 8 50 41 52 37 35 38 92 94 92 91 91 90 60 60 61 56 58 55 63 61 69 64 62 60 Living with Not living Living with Not living Living with Not living elders/ with elders/ elders/ with elders/ elders/ with elders/ children children children children children children All Women Men Household income lowered Income not su cient to pay for consumption Reduced spending in meals, education, durables, clothes or tobacco Taken loan Received assistance from government or NGO Do more household chores than before COVID Source: Rozee online COVID-19 survey of jobseekers. N=5,753 (more chores), N=6,083 (received assistance), N=6,075 (taken loan), N=6,023 (reduced spending), N=3,572 (income insufficient) and N=5,863 (income lowered). Figure 9. Share of respondents experienced stress symptoms during COVID-19 Quick anger Currently not working 46 26 49 51 31 57 78 Frequent argument Men Currently working 39 23 42 43 26 50 70 Frequent prayer Unable to sleep All 42 25 45 47 28 53 74 Unable to concentrate Currently not working 52 33 54 56 30 68 84 Feel anxious or depressed Women 48 29 50 51 28 63 81 Experienced at least one Currently working of the stress symptoms All 50 32 52 54 29 66 82 Currently not working 48 29 50 53 31 60 80 Currently working 41 24 44 45 26 53 72 All All 44 27 47 49 28 56 76 Source: Rozee online COVID-19 survey of jobseekers. N=7,606. working women (24 percent). In contrast, a greater share points to increased rates of humiliation and household of unemployed men, 37 percent, reported experiencing tensions associated with the unemployment of men. It is at least one type of violence since COVID-19, as opposed consistent with a recent study using data from 31 developing to 27 percent for men with employment. Multivariate countries, which showed that a 1 percent increase in the analysis verifies these findings (Appendix Table 8) and male unemployment rate is associated with an increase in 8 | POLICY BRIEF Figure 10. Share of respondents experienced violence after COVID-19 Currently not working 28 13 30 30 4 3 37 Men Currently working 19 11 20 20 33 27 All 23 12 24 25 4 3 32 Currently not working 17 8 19 19 4 2 24 Women Currently working 20 7 21 22 42 26 All 18 8 20 20 4 2 25 Currently not working 24 11 26 26 4 2 33 Currently working 19 10 20 21 32 27 All All 22 11 23 23 4 2 30 Humiliation Threat Input Unsafe at home Physical abuse Physical sexual abuse Experienced at least one of the violence Source: Rozee online COVID-19 survey of jobseekers. N=5,753 (more chores), N=6,083 (received assistance), N=6,075 (taken loan), N=6,023 (reduced spending), N=3,572 (income insufficient) and N=5,863 (income lowered). the incidence of physical violence against women by 0.5 labor market altogether. Similarly, the longer the economic percentage points or 2.8 percent (Bhalotra, Kambhampati, recovery takes, the fewer women may remain in or return to Rawlings, and Siddique 2020). the labor force. COVID-19 recovery efforts that do not address the disproportional impacts on women’s employment will likely result in larger gender gaps after the pandemic. CONCLUSION For sustained recovery, there must be greater recognition Labor force participation among Pakistani women is among of women’s economic role and unpaid work, as well as the lowest in the world, especially in urban areas. With targeted social safety net and support services that address the employment prospects so unfavorable for women, any the care burden and other stresses women face at home. disproportionate gender effects of the COVID-19 pandemic These services must be targeted toward women previously will worsen these gaps. As businesses closed, either employed in sectors where the decline in employment was temporarily or permanently, jobs have disappeared for both most severe, such as healthcare and education. For example, men and women. However, the sectors where women are firms can be incentivized to provide unemployment insurance more likely to be employed, such as education and health, and other temporary benefits to their workers, while those were more severely affected. Likewise, the COVID-19 who have lost their jobs in these sectors can serve as frontline pandemic has led to a disproportionate increase in women’s workers to roll out public COVID-19 response programs for unpaid care work. The fall in household income and the rise contact tracing, testing, vaccination and remote learning. in unpaid work are likely factors in higher stress and anxiety In the short term, there is also a dire need to address the among women, as well as leading to an increase in domestic pandemic’s toll on women’s physical, emotional and mental violence. health, such as by investing in multi-purpose helplines These findings present some alarming trends that need and nongovernmental organizations that provide women to be considered in COVID-19 recovery efforts. If women with financial assistance, psychosocial support and shelter. endure the added burden of unpaid household and care Without immediate and long-term interventions to establish work for a prolonged period, they are more likely to quit the work-family policies, it is possible that the labor market participation of women will decline even further. FEBRUARY 2021 | 9 REFERENCES Pakistan Bureau of Statistics, Labor Force Survey. “Table-20 Percentage distribution of employed persons 10 years of age Amir, Saman, Aphichoke Kotikula, Rohini P. Pande, Laurent and over by major industry divisions, occupation groups and Loic Yves Bossavie, and Upasana Khadka. 2018. Female sex: Pakistan & Provinces, Rural & Urban,” http://www.pbs. Labor Force Participation in Pakistan: What Do We Know? gov.pk/content/labour-force-statistics (accessed November The World Bank. 1, 2020). Asian Development Bank. 2020. “An Updated Assessment of Rubery, Jill, ed. 2010. Women and Recession (Routledge the Economic Impact of COVID-19.” ADB Briefs 133, https:// Revivals). Routledge. www.adb.org/sites/default/files/publication/604206/adb- brief-133-updated-economic-impact-covid-19.pdf (accessed Rubery, Jill, and Anthony Rafferty. 2013. “Women and November 1, 2020). Recession Revisited.” Work, Employment and Society 27(3): 414–432. Bhalotra, Sonia, Uma Kambhampati, Samantha Rawlings, and Zahra Siddique. 2020. Intimate Partner Violence: The Starr, Martha A. 2014. “Gender, Added-Worker Effects, and Influence of Job Opportunities for Men and Women. Policy the 2007–2009 Recession: Looking Within the Household.” Research Working Paper 9118. The World Bank. Review of Economics of the Household 12(2): 209–235. Cho, Yoonyoung, and David Newhouse. 2011. How Did the United Nations. 2020. “Policy Brief: The Impact of Great Recession Affect Different Types of Workers? Evidence COVID-19 on Women,” https://www.unwomen.org/-/media/ From 17 Middle-Income Countries. The World Bank. headquarters/attachments/sections/library/publications/ 2020/policy-brief-the-impact-of-covid-19-on-women-en. Elson, Diane. 2010. “Gender and the Global Economic Crisis pdf?la=en&vs=1406 (accessed November 1, 2020). in Developing Countries: A Framework for Analysis.” Gender & Development 18(2): 201–212. UN Women. 2019. Progress of the World’s Women 2019– 2020: Families in a Changing World. United Nations. Karamessini, Maria, and Jill Rubery, eds. 2013. Women and Austerity: The Economic Crisis and the Future for Gender World Bank. 2020. “Projected Poverty Impacts of COVID-19 Equality. Routledge. (Coronavirus),” http://pubdocs.worldbank.org/en/4616015 91649316722/Projected-poverty-impacts-of-COVID-19.pdf Matsuda, Norihiko, Tutan Ahmed, and Shinsaku Nomura. (accessed November 1, 2020). 2019. Labor Market Analysis Using Big Data: The Case of a Pakistani Online Job Portal. Policy Research Working Paper World Health Organization. 2020. “Pakistan’s Drive to 9063. The World Bank. Restore Essential Health Services During COVID-19,” https://docs.google.com/viewerng/viewer?url=https://www. Pakistan Bureau of Statistics. 2019. Employment Trends uhcpartnership.net/wp-content/uploads/2020/10/Stories- 2018: Pakistan. The Government of Pakistan. from-the-field_issue4_Pakistan.pdf (accessed November 1, 2020). Pakistan Bureau of Statistics, Labor Force Survey. “Table-1 Percentage distribution of population by age, sex and area: Pakistan & Provinces,” http://www.pbs.gov.pk/content/ labour-force-statistics (accessed November 1, 2020). STAY CONNECTED We gratefully acknowledge the support of the Umbrella Facility for Gender Equality (UFGE). The UFGE is a multi-donor trust fund administered by the World Bank to advance gender equality and women’s empowerment through experimentation and knowledge creation aimed at helping The research highlighted in this brief is forthcoming as a journal article: “Effects of a Multi-Faceted Education Program on Enrollment, Learning and Gender Equity: Evidence from India”, Delavallade, Clara; Alan Griffith; and Rebecca Thornton. World Bank Economic Review, forthcoming. governments and the This brief was produced private sector in collaboration focus policies with researchers andBank’s at the World programs on scalable Africa Gender Innovationsolutions with Lab (AFRGIL), SARGENDERLAB@WORLDBANK.ORG which conducts impact evaluations of development interventions and leads policy sustainable research onoutcomes. gaps inhas The UFGE how to close gender received earnings, generous productivity, assets, contributions from and agency. For more Australia, information, Canada, visit: http://www.worldbank.org/africa/gil Denmark, Germany, Iceland, the Netherlands, Norway, the Republic of Latvia, Spain, Sweden, WORLDBANK.ORG/SARGENDERLAB Switzerland, the United Kingdom, the United States, and the Bill and Melinda Gates Foundation. 10 | POLICY BRIEF APPENDIX Table 4. Linear probability models: the probability of working and looking for jobs, before and after COVID-19 Table 1. The probability of firms in operation after COVID-19 (1) (2) (1) (2) (3) Working=1, Not=0 Looking for a Mul�nomial logit—current opera�onal Job=1, Not=0 status A�er lockdown=1 -0.15*** 0.21*** Business Business Business (0.01) (0.01) open temporarily permanently Female=1 -0.12*** 0.08*** closed closed (0.02) (0.02) Ref: Size: Micro 1–10 A�er lockdown X -0.03 -0.06** Small (11–50) 0.06 -0.06 -0.01 female (0.02) (0.02) (0.06) (0.06) (0.03) Age 0.05*** -0.02*** Medium (51–300) 0.03 -0.05 0.02 (0.004) (0.004) (0.07) (0.06) (0.04) Age-squared -0.001*** 0.0002*** Large (301+) 0.11 -0.14** 0.02 (0.00001) (0.00001) (0.07) (0.07) (0.05) Married=1 0.07*** -0.03*** Ref: Industry: ICT (0.01) (0.01) Manufacturing - 0.003 0.01 -0.01 Education Ref: SS or Less (0.07) (0.07) (0.04) HSS -0.08*** -0.03 Retail and 0.14 -0.06 -0.08*** (0.03) (0.03) wholesale (0.15) (0.15) (0.02) Diploma/bachelor -0.04 -0.03 Hotel, restaurants, -0.10 0.11 -0.01 (0.03) (0.03) food service (0.12) (0.11) (0.07) Graduate 0.01 -0.03 Transporta�on 0.05 -0.06 0.01 (0.03) (0.03) (0.10) (0.09) (0.07) Household size 0.001 0.001 Finance -0.12 0.12 0.003 (0.001) (0.001) (0.12) (0.12) (0.06) Live with elderly or a 0.01 0.02 Educa�on -0.45*** 0.44*** 0.01 child=1 (0.01) (0.01) (0.09) (0.10) (0.06) Has savings for at 0.01 -0.06*** Health -0.22* 0.30** -0.08*** least a month=1 (0.01) (0.01) (0.12) (0.12) (0.02) Cities lived in Feb 2020: Ref: Islamabad Other sectors -0.08 0.09 -0.01 Karachi 0.04** -0.01 (0.06) (0.06) (0.04) (0.02) (0.02) Ref: Location: Islamabad Lahore 0.03* 0.0002 Karachi -0.01 -0.03 0.04** (0.02) (0.02) (0.07) (0.07) (0.02) Rawalpindi -0.04 0.02 Lahore -0.12* 0.03 0.10*** (0.02) (0.03) (0.07) (0.07) (0.02) Other ci�es -0.08*** 0.06*** Rawalpindi -0.10 -0.002 0.10 (0.02) (0.02) (0.12) (0.11) (0.07) Constant -0.21*** 1.01*** Other ci�es -0.32*** 0.18** 0.14*** (0.07) (0.07) (0.08) (0.08) (0.04) Observations 11,230 6,777 Observations 421 421 421 ; *** 1%, ** 5%, Source: Rozee online survey of jobseekers Note: Rozee online survey of employers; *** 1%, ** 5%, and * 10% level of significance. and * 10% level of significance. FEBRUARY 2021 | 11 Table 2. Linear probability models: the effect of COVID-19 on firms’ operation: demand, revenue, activities, and access to inputs (1) (2) (3) (4) (5) Demand Revenue Ac�vity slowed Worker If use raw materials, reduced=1 reduced=1 down: very or absenteeism is an price rise of raw extreme=1 issue=1 materials is a problem=1 Ref: Size: Micro 1–10 Small (11–50) -0.10 0.05 -0.22*** -0.17** -0.03 (0.08) (0.08) (0.08) (0.08) (0.10) Medium (51–300) -0.20** -0.13 -0.25*** -0.17* -0.05 (0.10) (0.10) (0.09) (0.10) (0.11) Large (301+) -0.17 0.04 -0.35*** -0.09 -0.06 (0.11) (0.11) (0.09) (0.10) (0.12) Ref: Industry: ICT Manufacturing 0.24** 0.15 0.16 0.19* -0.02 (0.11) (0.11) (0.09) (0.10) (0.13) Retail and wholesale 0.09 0.17 0.16 0.03 0.28** (0.25) (0.18) (0.25) (0.27) (0.12) Hotel, restaurants, food service 0.15 -0.02 -0.04 -0.13 0.21** (0.16) (0.17) (0.12) (0.17) (0.08) Transporta�on 0.45*** 0.07 0.03 0.09 0.24** (0.10) (0.13) (0.10) (0.14) (0.10) Finance 0.05 -0.09 0.16 0.15 -0.07 (0.15) (0.17) (0.15) (0.15) (0.24) Educa�on 0.16 0.23 0.55*** 0.46*** -0.72*** (0.25) (0.22) (0.19) (0.10) (0.16) Health -0.14 0.16 -0.21 0.30** 0.09 (0.18) (0.20) (0.15) (0.15) (0.08) Other sectors 0.14 0.25*** 0.19** 0.09 0.02 (0.09) (0.09) (0.09) (0.09) (0.12) Percentage of female workers 0.09 -0.02 0.26*** 0.01 0.16* 40%+ before COVID (0.09) (0.10) (0.09) (0.09) (0.09) Percentage of entry workers 40%+ 0.01 0.04 0.09 0.10 -0.10 before COVID (0.08) (0.08) (0.08) (0.08) (0.10) Ref: Location: Islamabad Karachi 0.02 0.10 0.18** -0.07 -0.03 (0.10) (0.11) (0.08) (0.10) (0.11) Lahore 0.19* 0.14 0.17** -0.08 -0.12 (0.10) (0.11) (0.08) (0.10) (0.12) Rawalpindi 0.21 0.16 0.08 -0.10 -0.19 (0.17) (0.17) (0.12) (0.17) (0.33) Other ci�es -0.002 -0.10 0.13 -0.16 0.04 (0.15) (0.15) (0.12) (0.13) (0.15) Constant 0.37*** 0.47*** 0.13 0.68*** 0.87*** (0.10) (0.11) (0.09) (0.10) (0.13) Observations 233 215 195 234 111 Note: Rozee online survey of employers; *** 1%, ** 5%, and * 10% level of significance. 12 | POLICY BRIEF Table 3. Linear probability models: the effect of COVID-19 on firms’ human resource management – layoffs, wage management and working hours (1) (2) (3) (4) (5) Temporary layoff Permanent layoff Wage cut=1 Delayed wage=1 Hour cut=1 without pay=1 without pay=1 Ref: Size: Micro 1–10 Small (11–50) 0.04 -0.02 -0.03 -0.13* -0.04 (0.07) (0.07) (0.08) (0.08) (0.09) Medium (51–300) 0.06 0.08 -0.09 -0.14 -0.20** (0.09) (0.08) (0.10) (0.09) (0.10) Large (301+) 0.17* 0.08 -0.03 -0.13 0.03 (0.10) (0.09) (0.10) (0.10) (0.11) Ref: Industry: ICT Manufacturing 0.06 -0.13* 0.12 0.15 0.02 (0.09) (0.07) (0.10) (0.10) (0.11) Retail and wholesale 0.02 0.06 0.25 0.13 0.09 (0.24) (0.24) (0.28) (0.25) (0.30) Hotel, restaurants, food service -0.12 -0.34*** -0.22* 0.17 -0.16 (0.18) (0.09) (0.13) (0.17) (0.14) Transporta�on 0.09 0.06 -0.04 0.08 0.06 (0.14) (0.12) (0.13) (0.11) (0.15) Finance -0.03 -0.07 -0.09 0.04 0.26* (0.12) (0.12) (0.13) (0.11) (0.15) Educa�on 0.01 0.13 0.24 0.32 0.27 (0.24) (0.27) (0.26) (0.21) (0.20) Health -0.21 -0.20 0.04 -0.09 0.24 (0.15) (0.16) (0.19) (0.19) (0.19) Other sectors -0.003 -0.004 -0.04 0.25*** 0.12 (0.08) (0.07) (0.08) (0.08) (0.09) Percentage of female workers 0.12 0.12 0.04 0.24** 0.01 40%+ before COVID (0.09) (0.08) (0.09) (0.09) (0.09) Percentage of entry workers 0.23*** 0.29*** 0.09 0.13 0.15* 40%+ before COVID (0.08) (0.08) (0.08) (0.08) (0.08) Ref: Location: Islamabad Karachi 0.07 0.01 0.07 0.19** -0.13 (0.08) (0.07) (0.10) (0.08) (0.10) Lahore 0.15** 0.07 0.02 0.14* -0.14 (0.08) (0.07) (0.09) (0.07) (0.10) Rawalpindi 0.21 0.31** 0.31* 0.30* -0.24 (0.14) (0.15) (0.17) (0.17) (0.16) Other ci�es -0.06 0.07 0.07 0.17 -0.32** (0.12) (0.13) (0.14) (0.13) (0.15) Constant 0.07 0.10 0.24** 0.07 0.61*** (0.08) (0.07) (0.10) (0.08) (0.10) Observations 233 232 229 233 233 Note: Rozee online survey of employers; *** 1%, ** 5%, and * 10% level of significance. FEBRUARY 2021 | 13 Table 5. The probability of respondents who were employed before COVID-19 experiencing job loss, layoff, and cuts in salary or working hours (1) (2) (3) (4) (5) Linear Marginal effects: Mul�nomial logit Probability Predicted ME Predicted ME Predicted ME Predicted ME Model: Job loss (no effect) (permanent (temporary layoff) (salary/ working – not working layoff) hours cut) a�er COVID Female=1 0.06 0.01 0.01 -0.003 -0.02 (0.04) (0.03) (0.04) (0.02) (0.03) Age -0.03*** 0.01 -0.02* 0.01 -0.0003 (0.01) (0.01) (0.01) (0.01) (0.01) Age-squared 0.0004*** -0.0002 0.0002 -0.0001 0.00001 (0.0001) (0.0001) (0.0001) (0.0001) (0.0001) Married=1 0.03 -0.01 0.01 -0.03 0.04 (0.03) (0.02) (0.03) (0.02) (0.03) Ref: Education (SS or Less) HSS 0.13 -0.10 0.24*** -0.10 -0.04 (0.09) (0.08) (0.08) (0.08) (0.08) Diploma/bachelor 0.04 -0.10 0.18*** -0.07 -0.01 (0.08) (0.07) (0.07) (0.07) (0.07) Graduate -0.01 -0.09 0.12* -0.05 0.03 (0.08) (0.07) (0.07) (0.07) (0.07) Feb 2020: worked as manager -0.11*** 0.05 -0.11** -0.01 0.07* (0.04) (0.03) (0.05) (0.04) (0.04) Feb 2020: worked as work in office -0.01 0.03 -0.06 -0.01 0.04 (0.04) (0.03) (0.05) (0.04) (0.04) Feb 2020: worked as produc�on 0.03 -0.05 0.08 0.06 -0.08 (0.06) (0.05) (0.07) (0.05) (0.07) Feb 2020: worked as transport -0.15 0.002 -0.17 0.11* 0.06 (0.10) (0.10) (0.11) (0.06) (0.11) Feb 2020: worked as construc�on 0.11 -0.19 -0.01 0.09* 0.11 (0.09) (0.12) (0.10) (0.06) (0.09) Feb 2020: worked as work in shop 0.03 0.02 0.24* -0.05 -0.21 (0.09) (0.10) (0.12) (0.09) (0.15) Feb 2020: worked in other posi�ons 0.02 0.02 -0.11* 0.01 0.07 (0.05) (0.04) (0.06) (0.04) (0.05) Ref: Feb 2020: worked in sector ICT Manufacturing 0.10* -0.09 0.07 0.04 -0.02 (0.06) (0.06) (0.06) (0.04) (0.06) Retail and wholesale 0.16** -0.15** 0.15** -0.01 0.01 (0.06) (0.06) (0.07) (0.04) (0.06) Hotel, restaurants, food service 0.47*** -0.28*** 0.38*** 0.02 -0.13* (0.08) (0.05) (0.08) (0.06) (0.07) Transporta�on 0.23* -0.21** 0.16 -0.01 0.06 (0.12) (0.08) (0.12) (0.07) (0.11) Finance 0.01 -0.22*** 0.06 0.04 0.13** (0.05) (0.05) (0.06) (0.04) (0.05) Educa�on 0.23*** -0.24*** 0.21*** 0.08* -0.05 (0.06) (0.05) (0.06) (0.04) (0.05) Health 0.09 -0.09 0.003 0.09* 0.004 (0.06) (0.06) (0.06) (0.05) (0.06) Agriculture 0.10 -0.23*** 0.18 -0.0004 0.05 (0.11) (0.08) (0.11) (0.07) (0.10) Other sectors 0.14*** -0.16*** 0.13** 0.02 0.01 (0.05) (0.05) (0.06) (0.04) (0.05) Ref: Feb 2020: worked in private company Work at NGO -0.01 0.07* 0.07 -0.06** -0.08** (0.05) (0.04) (0.05) (0.03) (0.04) Work at government -0.23*** 0.02 -0.13*** -0.10*** 0.20*** (0.04) (0.04) (0.05) (0.02) (0.05) Other 0.13* -0.11*** 0.39*** -0.05 -0.23*** 14 | POLICY BRIEF (0.08) (0.03) (0.06) (0.05) (0.03) Household size 0.01* -0.003* 0.002 0.002 -0.0004 (0.003) (0.002) (0.003) (0.002) (0.002) (0.05) (0.05) (0.06) (0.04) (0.05) Ref: Feb 2020: worked in private company Work at NGO -0.01 0.07* 0.07 -0.06** -0.08** (0.05) (0.04) (0.05) (0.03) (0.04) Work at government -0.23*** 0.02 -0.13*** -0.10*** 0.20*** (0.04) (0.04) (0.05) (0.02) (0.05) Other 0.13* -0.11*** 0.39*** -0.05 -0.23*** (0.08) (0.03) (0.06) (0.05) (0.03) Household size 0.01* -0.003* 0.002 0.002 -0.0004 (0.003) (0.002) (0.003) (0.002) (0.002) Live with elderly and/or child=1 0.03 -0.03 -0.001 -0.02 0.05 (0.03) (0.02) (0.03) (0.02) (0.03) Has savings for at least a month in Feb 0.08*** -0.003 0.04 0.02 -0.05** 2020 (0.03) (0.02) (0.03) (0.02) (0.02) Cities lived in Feb 2020: Ref: Islamabad Karachi -0.03 -0.08** -0.004 0.01 0.07 (0.05) (0.04) (0.05) (0.04) (0.04) Lahore 0.01 -0.06 0.05 -0.02 0.03 (0.05) (0.04) (0.05) (0.03) (0.04) Rawalpindi 0.07 -0.12** 0.12* 0.01 -0.01 (0.07) (0.05) (0.07) (0.05) (0.06) Other ci�es 0.13*** -0.15*** 0.17*** -0.01 -0.02 (0.05) (0.04) (0.05) (0.03) (0.04) Constant 0.73*** … … … … (0.23) Observations 1,426 1,416 1416 1,416 1,416 Source: Rozee online survey of jobseekers; *** 1%, ** 5%, and * 10% level of significance. FEBRUARY 2021 | 15 Table 6. Linear probability models: the effect of COVID-19 on respondents’ household income, consumption, and unpaid work, and their coping strategies (1) (2) (3) (4) (5) (6) (7) (8) HH Income not Reduced Taken Received Do more If lived with If lived with income sufficient to spending loan=1 assistance HH elderly/children, elderly/children, lowered pay for in meals, from govt chores did more children are =1 consump�on=1 educa�on, or NGO=1 than caregiving=1 doing home- durables, before based clothes, or COVID=1 schooling=1 tobacco =1 Female=1 -0.02 -0.07*** 0.001 - -0.04*** 0.03* 0.02 0.04** 0.08*** (0.02) (0.02) (0.01) (0.02) (0.01) (0.02) (0.02) (0.02) Currently working=1 -0.25*** -0.19*** -0.06*** - 0.02** -0.09*** -0.03** 0.01 0.08*** (0.01) (0.01) (0.01) (0.02) (0.01) (0.01) (0.01) (0.01) Currently working X 0.11*** 0.14*** 0.05*** 0.05* -0.01 0.05* 0.02 -0.02 female (0.03) (0.03) (0.003) (0.03) (0.01) (0.02) (0.03) (0.02) Age -0.003 0.01* -0.004 0.01 -0.01*** -0.003 0.004 0.013*** (0.01) (0.01) (0.003) (0.01) (0.003) (0.004) (0.01) (0.004) Age-squared 0.00001 -0.0001 0.0001 -0.0001 0.0001*** 0.00001 -0.0001 -0.0002*** (0.0001) (0.0001) (0.00004) (0.0001) (0.00001) (0.0001) (0.0001) (0.0001) Married=1 -0.01 0.05 -0.01 0.09*** 0.004 -0.04 0.05*** 0.02 (0.03) (0.03) (0.02) (0.03) (0.02) (0.03) (0.02) (0.01) Ref: Education: SS or Less HSS 0.08* 0.05 -0.002 -0.03 -0.02 -0.03 0.01 0.0003 (0.05) (0.05) (0.03) (0.05) (0.03) (0.04) (0.05) (0.04) Diploma/bachelor 0.002 0.03 0.01 -0.09** -0.05* -0.02 0.01 -0.001 (0.04) (0.04) (0.02) (0.04) (0.03) (0.03) (0.04) (0.03) Graduate 0.02 0.04 -0.01 - -0.06** -0.04 0.03 0.01 0.11*** (0.04) (0.04) (0.02) (0.04) (0.03) (0.03) (0.04) (0.03) Agree: men should get 0.01 0.01 0.02*** 0.03** 0.02*** 0.02** 0.05*** -0.03*** the job in economic (0.01) (0.01) (0.01) (0.01) (0.01) (0.01) (0.01) (0.01) hardship Household size 0.003** -0.001 0.0004 -0.001 -0.0004 -0.001 -0.001 -0.001 (0.001) (0.001) (0.001) (0.001) (0.001) (0.001) (0.001) (0.001) Live with elderly 0.02 0.05** -0.001 0.13*** 0.02 0.04** and/or child=1 (0.02) (0.02) (0.01) (0.02) (0.01) (0.02) Live with elderly 0.02 -0.03 0.01 -0.02 0.01 0.07** and/or child X female (0.03) (0.03) (0.02) (0.03) (0.02) (0.03) Has savings for at least -0.06*** -0.14*** -0.03*** - -0.01* -0.004 0.01 0.02* a month in Feb 2020 0.13*** (0.01) (0.01) (0.01) (0.01) (0.01) (0.01) (0.01) (0.01) Cities lived in Feb 2020: Ref: Islamabad Karachi 0.01 -0.01 -0.01 -0.04 0.01 -0.02 -0.03 0.003 (0.03) (0.03) (0.01) (0.03) (0.01) (0.02) (0.03) (0.02) Lahore 0.07*** 0.07*** 0.01 0.07*** -0.002 -0.01 -0.01 0.03 (0.03) (0.03) (0.01) (0.03) (0.01) (0.02) (0.02) (0.02) Rawalpindi 0.05 -0.001 0.02 0.02 0.03 -0.003 0.03 0.02 (0.03) (0.03) (0.02) (0.03) (0.02) (0.03) (0.03) (0.03) Other ci�es 0.08*** 0.06** 0.01 0.08*** 0.04*** -0.01 0.003 0.001 (0.02) (0.02) (0.01) (0.02) (0.01) (0.02) (0.02) (0.02) Constant 0.73*** 0.47*** 1.0*** 0.44*** 0.27*** 0.89*** 0.67*** 0.64*** (0.10) (0.11) (0.08) (0.10) (0.06) (0.09) (0.10) (0.09) Observations 5,678 5,751 5,829 5,885 5,888 5,602 4,754 4,197 Source: Rozee online survey of jobseekers; *** 1%, ** 5%, and * 10% level of significance. 14 | POLICY BRIEF Table 7. Linear probability models: the effect of COVID-19 on experiencing stress symptoms (1) (2) (3) (4) (5) (6) (7) Quick Frequent Frequent Unable to Unable to Feel anxious or Experienced at least anger=1 argument=1 prayer=1 sleep =1 concentrate=1 depressed =1 one of the stress symptoms=1 Female=1 0.07*** 0.07*** 0.05*** 0.04** -0.02 0.08*** 0.05*** (0.02) (0.02) (0.02) (0.02) (0.02) (0.02) (0.02) Currently working=1 -0.06*** -0.04*** -0.07*** -0.09*** -0.06*** -0.09*** -0.08*** (0.02) (0.01) (0.02) (0.02) (0.01) (0.02) (0.01) Currently working X female 0.01 -0.02 0.02 0.02 0.01 0.01 0.03 (0.03) (0.03) (0.03) (0.03) (0.03) (0.03) (0.02) Age -0.002 -0.04 -0.001 0.001 -0.0027 0.01 0.001 (0.01) (0.01) (0.01) (0.01) (0.01) (0.01) (0.01) Age-squared -0.00001 0.00001 -0.00001 -0.0001 0.00001 -0.0001 -0.00001 (0.0001) (0.0001) (0.0001) (0.0001) (0.0001) (0.0001) (0.0001) Married=1 0.03 -0.01 0.04 0.03 -0.04 -0.05 0.01 (0.03) (0.03) (0.03) (0.03) (0.03) (0.03) (0.03) Ref: Education: SS or Less HSS -0.05 0.004 -0.06 -0.02 -0.04 -0.001 0.03 (0.05) (0.04) (0.05) (0.05) (0.04) (0.05) (0.04) Diploma/bachelor -0.07* 0.02 -0.06 -0.06 -0.004 0.05 0.06 (0.04) (0.04) (0.04) (0.04) (0.04) (0.04) (0.04) Graduate -0.06 0.01 -0.06 -0.05 -0.001 0.07 0.05 (0.04) (0.04) (0.04) (0.04) (0.04) (0.04) (0.04) Agree: men should get the 0.001 -0.06*** -0.01 -0.01 -0.05*** -0.07*** -0.02 job in economic hardship (0.01) (0.01) (0.01) (0.01) (0.01) (0.01) (0.01) Household size 0.003* 0.001 0.003** 0.002 0.001 -0.0004 0.001 (0.001) (0.001) (0.001) (0.001) (0.001) (0.001) (0.001) Live with elderly and/or 0.06*** 0.05** 0.07*** 0.07*** 0.04** 0.05** 0.06*** child=1 (0.02) (0.02) (0.02) (0.02) (0.02) (0.02) (0.02) Live with elderly and/or -0.01 0.01 -0.02 -0.01 0.02 0.02 -0.02 child X female (0.04) (0.03) (0.04) (0.04) (0.03) (0.04) (0.03) Has savings for at least a -0.05*** 0.01 -0.03** -0.03** -0.01 0.02 -0.01 month in Feb 2020 (0.01) (0.01) (0.01) (0.01) (0.01) (0.01) (0.01) Cities lived in Feb 2020: Ref: Islamabad Karachi -0.01 -0.06** -0.003 -0.01 -0.03 0.004 -0.01 (0.03) (0.03) (0.03) (0.03) (0.03) (0.03) (0.02) Lahore 0.03 -0.02 0.03 0.01 -0.01 0.01 -0.02 (0.03) (0.02) (0.03) (0.03) (0.02) (0.03) (0.02) Rawalpindi 0.06* -0.06* 0.06 0.03 -0.03 0.01 0.01 (0.04) (0.03) (0.04) (0.04) (0.03) (0.03) (0.03) Other ci�es -0.02 -0.10*** -0.01 -0.04* -0.06** -0.05* -0.04* (0.02) (0.02) (0.02) (0.02) (0.02) (0.02) (0.02) Constant 0.54*** 0.39*** 0.52*** 0.56*** 0.43*** 0.52*** 0.74*** (0.11) (0.09) (0.11) (0.11) (0.10) (0.11) (0.09) Observations 5,849 5,849 5,849 5,849 5,849 5,849 5,849 Source: Rozee online survey of jobseekers; *** 1%, ** 5%, and * 10% level of significance. FEBRUARY 2021 | 15 Table 8. Linear probability models: the effect of COVID-19 on experiencing violence (1) (2) (3) (4) (5) (6) (7) Humilia�on=1 Threat=1 Insult=1 Unsafe at Physical Physical sexual Experienced at least home =1 abuse=1 abuse=1 one of the violence=1 Female=1 -0.10*** -0.04*** - -0.11*** -0.001 -0.004 -0.13*** 0.11*** (0.02) (0.01) (0.02) (0.02) (0.01) (0.01) (0.02) Currently working=1 -0.08*** -0.03** - -0.09*** -0.01* 0.0003 -0.10*** 0.09*** (0.01) (0.01) (0.02) (0.02) (0.01) (0.01) (0.02) Currently working X female 0.11*** 0.02 0.13*** 0.13*** 0.01 -0.002 0.14*** (0.03) (0.02) (0.03) (0.03) (0.01) (0.01) (0.03) Age -0.01* 0.002 -0.01 -0.01* -0.001 0.0001 -0.01 (0.01) (0.003) (0.01) (0.01) (0.002) (0.002) (0.01) Age-squared 0.00001 -0.0001 0.00001 0.00001 0.00001 -0.00001 -0.00001 (0.0001) (0.00001) (0.0001) (0.0001) (0.00001) (0.00001) (0.0001) Married=1 0.001 0.04 -0.01 0.001 0.01 0.01 0.02 (0.03) (0.02) (0.03) (0.03) (0.02) (0.01) (0.03) Ref: Education: SS or Less HSS 0.04 -0.07* 0.04 0.04 -0.03 -0.04 0.01 (0.05) (0.04) (0.05) (0.05) (0.03) (0.03) (0.05) Diploma/bachelor 0.03 -0.07* 0.04 0.04 -0.03 -0.05** 0.01 (0.04) (0.04) (0.04) (0.04) (0.02) (0.02) (0.04) Graduate 0.01 -0.09** 0.02 0.02 -0.04 -0.05** -0.03 (0.04) (0.04) (0.04) (0.04) (0.02) (0.02) (0.04) Agree: men should get the job 0.05*** 0.04*** 0.05*** 0.05*** 0.002 0.01** 0.08*** in economic hardship (0.01) (0.01) (0.01) (0.01) (0.01) (0.004) (0.01) Household size 0.04* 0.003 0.03 0.05** 0.003 0.01 0.03 (0.02) (0.02) (0.02) (0.02) (0.01) (0.01) (0.02) Live with elderly and/or 0.01 -0.01 0.02 0.02 -0.01 - 0.001 0.0002 child=1 (0.03) (0.03) (0.03) (0.03) (0.02) (0.01) (0.04) Live with elderly and/or child X -0.06*** -0.04*** - -0.07*** -0.02*** -0.01*** -0.09*** female 0.07*** (0.01) (0.01) (0.01) (0.01) (0.01) (0.004) (0.01) Has savings for at least a 0.04* 0.003 0.03 0.05** 0.003 0.01 0.03 month in Feb 2020 (0.02) (0.02) (0.02) (0.02) (0.01) (0.01) (0.02) Cities lived in Feb 2020: Ref: Islamabad Karachi -0.05** 0.004 -0.05** -0.05** 0.02* 0.01 -0.05* (0.02) (0.02) (0.02) (0.02) (0.01) (0.01) (0.03) Lahore -0.01 -0.01 -0.02 -0.01 0.01 0.01** -0.02 (0.02) (0.02) (0.02) (0.02) (0.01) (0.01) (0.03) Rawalpindi 0.01 0.01 0.02 0.01 0.01 0.01 0.02 (0.03) (0.02) (0.03) (0.03) (0.01) (0.01) (0.04) Other ci�es 0.01 0.01 0.004 0.01 0.004 0.01*** 0.02 (0.02) (0.02) (0.02) (0.02) (0.01) (0.01) (0.02) Constant 0.43*** 0.22*** 0.43*** 0.44*** 0.12** 0.07* 0.57*** (0.10) (0.08) (0.10) (0.10) (0.05) (0.04) (0.10) Observations 4,704 4,704 4,704 4,704 4,704 4,704 4,704 Source: Rozee online survey of jobseekers; *** 1%, ** 5%, and * 10% level of significance.