Pandemic in EthiopiaCOVID-19 Monitoring Gendered impacts of the COVID-19 REPORT NO. 5 REPORT NO. 5 9 / October / 2020 Impacts on Households 12 / October / 2020 in Ethiopia Gendered Impacts of the COVID-19 Pandemic in Ethiopia: Results from a High-Frequency Phone Survey of Households Authors: Menaal Ebrahim, Alemayehu A. Ambel, Niklas Buehren, Tom Bundervoet, Adiam Hagos Hailemicheal, Girum Abebe Tefera, and Christina Wieser. The COVID-19 pandemic and its economic and social effects on households have created an urgent need for timely data to help monitor and mitigate the social and economic impacts of the crisis and protect the welfare of the least well-off Ethiopians. To track how the pandemic is affecting Ethiopia’s economy and its population and to inform interventions and policy responses, the World Bank is conducting a customized high-frequency phone survey of households (HFPS-HH). The HFPS-HH builds on the longitudinal Ethiopia Socioeconomic Survey (ESS) that the Central Statistical Agency (CSA) carried out in 2019 in collaboration with the World Bank. The HFPS-HH is representative of households with a working phone. The same households are tracked for six months, with selected respondents, typically household heads, completing phone-based interviews every three to four weeks.1 To support new response efforts to the pandemic as they become necessary, this high-frequency survey allows for a better understanding of its effects on households and how they adapt to the crisis. A series of briefs on the impact of the pandemic has thus far been published using the HFPS-HH. This special topic brief explores the gender impacts of the pandemic on households in rural and urban areas. The brief reveals the disproportionate adverse impacts of COVID-19 on women and women-headed households in Ethiopia.2 The analysis is based on a sample of 3,058 households in both urban and rural areas in all regions of Ethiopia.3 The 15-minute interview covers a diverse set of topics such as access to basic services, child educational activities during school closures, employment dynamics, household income and livelihood, income loss and coping strategies, food security and assistance received. In this brief, we focus on topics where gendered differences were striking. HIGHLIGHTS – GENDER BRIEF Four months into the pandemic, respondents, especially women, seem to adhere to preventative measures to reduce the spread of COVID-19 and remain conscious of its impacts. While access to food and medical treatment does not seem to be a particular concern for respondents, female-headed households had lower access to medical treatment in rural areas compared to their male counterparts. Given school closures, children in both male and female-headed households have gradually started engaging in learning activities at similar levels, but children in female-headed households are more often watching educational TV programs and are less likely to engage with tutors in person. Women have consistently had lower employment levels during the pandemic compared to men and face additional barriers such as health issues and caretaking responsibilities that prevent them from transitioning back to work. Female-headed households in rural areas have disproportionately experienced income losses and have more often relied on reducing food consumption as a coping strategy compared to male-headed households. 1 The data collection was undertaken by Laterite (Ethiopia) Ltd. 2 Some variables are asked at the household level while some are at the respondent level. These differences are taken into account in the analysis. 3 The original sample consisted of 3,249 households, of which 2,878 responded to Round 4 (R4) calls. Round 4 was administered in the first three weeks of August 2020. Gendered impacts of the COVID-19 REPORT NO. 5 Pandemic in Ethiopia 9 / October / 2020 Most respondents adhere to COVID-19 preventative measures and are conscious of its impacts, yet more female respondents employ these behaviors and report stronger health concerns. 4 About four months into the pandemic, over 71 percent of male and 74 percent of female respondents reported washing their hands very frequently, and 40 percent of male and 39 percent of female respondents reported wearing masks very frequently (Figure 1). Overall, people report to observe protective practices, with female respondents coming across slightly more vigilant. Respondents are also concerned about the health and financial consequences from the pandemic. 52 percent of male and 59 percent of female respondents reported being very worried for themselves or their family members falling ill from COVID-19, and Figure 1: COVID-19 related preventative measures and apprehensions, R3 percent. about 60 percent of both male and female respondents 71% 74% Male Female 61% consider the pandemic a 59% 58% 52% substantial threat to their 40% 39% finances. While women are equally apprehensive of the financial threat from COVID- 19, more of them fear its health impacts. HANDWASHING WEARING VERY WORRIED SUBSTANTIAL VERY MASKS VERY SELF OR PERCEIVED FREQUENTLY FREQUENTLY FAMILY FINANCIAL FALLING ILL THREAT FROM COVID BEHAVIORS APPREHENSIONS Female-headed households seem to have similar access to food staples but have poorer access to medical treatment in rural areas compared to their male counterparts. Access to food staples and medical treatment has generally not been a big concern for urban households, as at least 77 percent of all urban households reported being able to access basic food items and medical treatment as needed (Table 1). Female-headed households also seem to access food staples and medical treatment at similar levels as their male counterparts in urban areas. In rural areas, where access had generally been lower compared to urban areas, male and female-headed households access food staples equally. However, merely 86 percent of female-headed rural households were able to access medical treatment as needed in the previous month compared to 94 percent of male-headed households. Table 1: Ability of households to buy certain items when needed, in round four of the survey Rural Urban National Male HH Female HH Male HH Female HH Male HH Female HH Medical treatment 94% 86% 94% 94% 94% 91% Teff 52% 48% 79% 77% 63% 66% Wheat 61% 52% 85% 84% 71% 73% Maize 75% 81% 85% 78% 78% 80% Edible oil 74% 64% 85% 86% 77% 75% Percentages in bold represent statistically significant differences between male and female-headed households. Children in male and female-headed households have been engaging in learning activities at similar levels since school closures but are using different tools. On March 16, 2020, Ethiopia closed all primary and secondary schools. Not surprisingly, children’s involvement in any type of learning activities dropped drastically, especially right after restrictions were introduced and even more so in rural areas (Figure 2). However, as of the latest survey round, about half of all households reported having children that currently participate in some form of learning activity. Equal proportions of male and female-headed households have children engaged in learning activities over time. The most frequent learning engagement for both primary and secondary school children in urban areas is completing assignments (Table 2). In rural areas, children are mostly watching and listening to educational programs on TV and radio (an initiative by the Ministry of Education in Ethiopia). However, there are differences in the type of children’s learning activities between male and female-headed households. For primary school children, fewer female-headed households (8 percent) had children that met with a tutor compared to male-headed households (32 percent). Women are perhaps more cautious about COVID-19 measures and less willing for their children to meet with a tutor. For secondary school children, far more female-headed households prefer educational TV programs (72 percent vs 17 percent) and less female-headed households prefer educational radio programs (4 percent vs 47 percent). 4 Questions about COVID-19 related behaviors and beliefs were included in the first and third round of the survey. Gendered impacts of the COVID-19 REPORT NO. 5 Pandemic in Ethiopia 9 / October / 2020 Figure 2: Children engaged in learning activities, R1-R4.5 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Pre-covid R1 R2 R3 R4 Rural Male Rural Female Urban Male Urban Female National Male National Female Table 2: Educational activities children engage with during school closures, R4. Primary Secondary Rural Urban Rural Urban Male Female Male Female Male Female Male Female HH HH HH HH HH HH HH HH Completed assignments 21% 26% 45% 55% 28% 20% 48% 29% Used mobile learning apps 6% 4% 12% 11% 21% 12% 30% 22% Watched educational TV programs 11% 21% 28% 22% 17% 72% 27% 20% Listened to educational radio programs 50% 49% 12% 14% 47% 4% 8% 21% Met with tutor 32% 8% 14% 13% 17% 19% 13% 30% Percentages in bold represent statistically significant differences. Since the beginning of the outbreak, women have had lower employment rates compared to men with many reporting personal health or family responsibilities as the main reason for not working. Some of the most noticeable impacts of COVID-19 have been on employment. The State of Emergency declaration prohibits firms from laying off workers, so the data shows only a slight initial decline in employment for men and women, in both rural and urban areas (Figure 3). In round 1 of the survey, about 85 percent of men reported having had a job in the past week, compared to 69 percent of women. Of those that reported being unemployed during round 1, 40 percent (of both men and women) reported being employed right before the oubreak. So, it is likely that some respondents lost their jobs because of the pandemic. By the latest round, about 90 percent of men and 75 percent of women reported having work. Rural (mostly agriculture-based) jobs seemed to have recovered quickly for men and women (with about 85 percent of all rural respondents reporting being employed) while urban wage employment has had slower recovery (65 percent and 78 percent of women and men reporting being employed respectively). While the employment rate for both men and women have had a similar trend over time, women have consistently been at lower employment levels after the outbreak relative to men. This, however, may not imply that the COVID-19 pandemic disproportionately affected women’s jobs, as women typically have lower employment levels. When asked about the reasons for not currently working, the small proportion of respondents not working in rural areas mostly reported being temporarily absent and the jobs being seasonal work as the main reason (Figure 4). In urban areas, many report seasonal work as the main reason for not working, however 28 percent of women reported illness and taking care of relatives as main reasons for not working compared to just 6 percent of men. Therefore women, who are often caretakers, have additional challenges with maintaining stable employment. 5 In R3 and R4, we ask questions on primary and secondary education engagements, so the estimates in the figure for R3 and R4 are just weighted averages. Gendered impacts of the COVID-19 REPORT NO. 5 Pandemic in Ethiopia 9 / October / 2020 Figure 3: Currently employed, R1-R4. Figure 4: Reasons for no longer working, R4. 100% 90% Need to care for ill relative 80% Ill 70% 60% Retired 50% Temporarily absent 40% 30% Lost job due to COVID 20% Contract ended 10% 0% Seasonal/casual worker R1 R2 R3 R4 Other (specify) Rural Male Rural Female 0% 20% 40% 60% 80% Urban Male Urban Female Urban Female Urban Male National Male National Female Rural Female Rural Male While female-headed households in urban areas have experienced income losses at similar levels to male-headed households, female-headed households in rural areas appear more affected than their male counterparts. The pandemic has led to lower income through loss of jobs but also through other channels. In rural areas, the main income source for most households is agriculture, with wage employment and non-farm businesses being distant second and third (Table 3). Female-headed households are slightly less dependent on agricultural income in comparison to male-headed households. But agricultural income losses did hit female-headed households harder for the first few months of the pandemic (Figure 5).6 In the first survey round, about 40 percent of both male and female- headed households reported diminished or a complete loss in agricultural income, but by the next survey round, this number decreased to 25 percent of male-headed households and increased to 44 percent of female-headed households, and then converging to about 25 percent by the last round. In urban areas, the most common income source is wage employment. However, female-headed households have more varied sources of income such as remittances and government assistance. Wage losses have been experienced similarly by male and female-headed households. Initially, about 33 percent of both male and female-headed households experienced diminished or completely lost wages, which reduced to 14 percent and 17 percent for both types of households by the latest round respectively. When looking at total household income, the group hit the hardest during the pandemic are rural female-headed households, which mirrors the trend in loss of agricultural income (Figure 6). Table 3: Income sources, R4.7 Figure 5: Percentage reporting reduced or completely diminished Rural Rural Urban Urban farming and wage income, R1-R4. Male Female Male Female Farming, fishing and 50% livestock 63% 48% 9% 6% 40% Non-farm business 10% 17% 28% 25% 30% Wage employment 16% 13% 59% 50% 20% Remittances from within 3% 8% 5% 11% 10% Remittances from 0% abroad 1% 2% 2% 4% R1 R2 R3 R4 Income from property and investments 4% 5% 13% 14% Farming income: Rural Male Pension 0% 2% 6% 10% Farming income: Rural Female Government assistance 6% 18% 2% 5% Wages: Urban Male Wages: Urban Female NGO assistance 2% 4% 0% 1% Percentages in bold represent statistically significant differences. 6 While we do not directly ask why agricultural income decreased, questions about decreased activities on household farms reveal self and hired labor mobility restrictions as the main reasons for the decrease in activity which could partly explain the decrease in agricultural income as well. 7 The estimates presented reflect the percentage of respondents that said ‘Yes’ when asked if they received any income from a source. So, the estimates are not to be added up to a 100. The table also excludes some sources that respondents didn’t indicate any income from. Gendered impacts of the COVID-19 REPORT NO. 5 Pandemic in Ethiopia 9 / October / 2020 While both male and female-headed households use common coping strategies, female-headed households were less reliant on savings in urban areas and more reliant on reducing food consumption in rural areas. In urban areas, the most widely used coping strategy for the loss of income was to draw on savings, but a smaller proportion of female-headed households relied on savings, whereas a higher proportion relied on assistance from family and friends (Figure 7). In rural areas, the most common strategy was to reduce food consumption, though about 60 percent of the households in rural areas reported not using any coping strategy at all. A higher proportion of female- headed households reduced food and non-food consumption whereas a lower proportion of female-headed households did nothing compared to male-headed households. Figure 6: Percentage reporting reduced or completely diminished total household income, R1-R4. Figure 7: Coping strategies for loss in income, R4. 70% Did nothing 60% Relied on savings 50% Reduced non-food consumption 40% Reduced food consumption 30% Credited purchases 20% Borrowed from friends and 10% family Received assistance from 0% friends and family R1 R2 R3 R4 Engaged in additional income generating activities Rural Male Rural Female 0% 20% 40% 60% 80% Urban Male Urban Female National Male National Female Urban Female Urban Male Rural Female Rural Male . While both male and female-headed households have similar levels of overall food insecurity, a higher proportion of female-headed households are compromising on a healthy and diverse diet. Food security is a major concern in Ethiopia and the pandemic has had serious impacts on the affordability of food. To monitor how Ethiopians were faring on food security indicators during the pandemic, the survey applied the Food Insecurity Experience Scale (FIES).8 Over 40 percent of all households have experienced moderate or severe food insecurity, without experiencing much improvement from the second to the fourth round of the survey (Figure 8). As we do not have information on food insecurity of these households before the pandemic, we cannot find a causal link between COVID-19 and food insecurity for our sample. When looking at food insecurity experiences that determine the overall food insecurity index, a higher proportion of female-headed households in rural areas were worried about not having enough food and ate less in terms of quantity and quality than male-headed households (Table 4). In urban areas, female-headed households appear more likely to eat unhealthy and fewer kinds of food. While we may not be able to conclude that female-headed households are more food insecure, we do see a higher proportion compromising food consumption, as was also apparent from their coping strategies. 8 In this survey FIES consists of eight questions (http://www.fao.org/policy-support/tools-and-publications/resources-details/en/c/1236494/). The full set of questions were included from Round 2 onwards. The questions are household referenced and the recall period is 30 days preceding the survey. Prevalence rates are calculated by population. Gendered impacts of the COVID-19 REPORT NO. 5 Pandemic in Ethiopia 9 / October / 2020 Figure 8: Food insecurity prevalence, R2-R4. 70 59 58 60 53 54 52 50 46 45 45 48 45 44 44 50 43 40 45 43 42 40 30 20 10 0 Female Male Female Male Female Male Rural Urban National R2 R3 R4 Table 4: Food insecurity experiences, R4. Rural Urban National Male Female Male Female Male Female Worried about not having enough food 49% 62% 45% 50% 48% 56% Unable to eat healthy food 60% 69% 50% 61% 57% 65% Ate few kinds 58% 68% 51% 58% 56% 63% Skipped meal 36% 37% 31% 34% 34% 35% Ate less 40% 42% 38% 40% 40% 41% Did not eat due to lack of resources 7% 15% 8% 9% 7% 12% Ran out of food 17% 26% 28% 30% 20% 28% Hungry but did not eat 8% 17% 17% 12% 11% 15% Percentages in bold represent statistically significant differences. This survey brief is the fifth in a series reporting on the findings of the HFPS-HH. It reports results from survey rounds 1 through 4 related to the differential effects of and responses to the COVID-19 pandemic for male and female-headed households. Data collection will continue by following up with the same households every four weeks. For each round, the survey brief, table of indicators, and microdata will be available at https://www.worldbank.org/en/country/ethiopia/brief/phone-survey-data-monitoring-covid-19-impact-on-firms- and-households-in-ethiopia The high-frequency phone survey monitors the economic and social impacts of the COVID-19 pandemic on households and their responses in terms of such topics as access to food staples, access of children to educational activities during school closures, employment dynamics, household incomes and livelihoods, income losses and coping strategies, and external assistance. The HFPS-HH sample consists of a subsample of those interviewed for the Ethiopia Socioeconomic Survey (ESS) in 2019 —households with access to a phone—covering urban and rural areas in all regions of Ethiopia. The HFPS-HH called the 5,374 households that in the ESS had provided a valid phone number. Phone penetration in rural Ethiopia is low; about 40 percent of rural households have access to a phone compared to over 90 percent of urban households. This not only means that the rural sample is relatively small but there is also a systematic difference between households that have access to a phone and those that have not. Phone-owning households are better off in terms of total consumption, educational attainment, access to improved water and sanitation, access to assets, and access to electricity. The sample of the HFMP-HH is therefore representative only of households who have access to phones in urban and rural Ethiopia. The respondent is typically the household head; where that person cannot be reached despite numerous call- backs, another knowledgeable household member is selected as the respondent. Table 5: Completed interviews in each round by gender of the household head Round 1 Round 2 Round 3 Round 4 (Apr 22- May 13) (May 14-June 3) (June 4- 26) (July 27- Aug14) Female Headed Households 995 944 934 877 Male Headed Households 2,254 2,163 2,124 2,001 Total 3,249 3,107 3,058 2,878