Policy Research Working Paper 9556 The Evolving Socioeconomic Impacts of COVID-19 in Four African Countries Ann Furbush Anna Josephson Talip Kilic Jeffrey D. Michler Development Economics Development Data Group March 2021 Policy Research Working Paper 9556 Abstract The paper provides evidence on the evolving socioeconomic and food security, the gains have been relatively modest. impacts of the COVID-19 pandemic among households in Additionally, households have received very little outside Ethiopia, Malawi, Nigeria, and Uganda. The data allow esti- assistance and their ability to cope with shocks remains mating the immediate economic impacts of the pandemic, limited. School closures have created a vacuum in education beginning in April 2020, and tracking how the situation delivery and school-aged children have struggled to receive evolved through September 2020. Although households education services remotely. have started to see recovery in income, business revenues, This paper is a product of the Development Data Group, Development Economics. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/prwp. The authors may be contacted at tkilic@worldbank.org. The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. Produced by the Research Support Team The Evolving Socioeconomic Impacts of COVID-19 in Four African Countries Ann Furbush1, Anna Josephson1, Talip Kilic2, and Jeffrey D. Michler1 1 University of Arizona 2 World Bank JEL Codes: C83, I15, I31, O12. Keywords: SARS-CoV-2, COVID-19, behavioral change, income loss, household enterprises, access to basic needs, food insecurity, concerns, access to basic needs, access to education, Ethiopia, Malawi, Nigeria, Uganda, Sub-Saharan Africa. Introduction The socioeconomic impacts of the COVID-19 pandemic are not yet fully understood as the disease continues to affect individuals and households around the world. Governments have worked to attenuate these socioeconomic impacts by limiting the spread of the virus and mitigating the negative health outcomes of the disease through various policy measures. These policies include limiting travel, imposing quarantines and lockdowns, and closing businesses and schools. The effects of the pandemic have been felt worldwide, though little evidence yet exists on the ongoing impacts for individuals and households in Africa (one exception is the chapters in Djankov and Panizza 2020). We rely on direct measurements of socioeconomic indicators to present evidence on the evolving effects of the pandemic on households, adults, and children living in four African countries (Ethiopia, Malawi, Nigeria, and Uganda), as well as the actions that households are taking to mitigate these impacts. Data and Methods To examine the evolving effects of the pandemic, we use longitudinal data from high-frequency national phone surveys in Ethiopia, Malawi, Nigeria, and Uganda. In each country, these phone surveys aim to conduct monthly phone interviews, for a period of 12 months. Supported by the World Bank Living Standards Measurement Study (LSMS), the implementing agency for the phone survey in Ethiopia, Malawi, Nigeria, and Uganda, is, respectively, Laterite Ethiopia, Malawi National Statistical Office, Nigeria Bureau of Statistics, and Uganda Bureau of Statistics. The anonymized, unit-record phone survey data, as well as basic information documents, interviewer manuals, and questionnaires, associated with each monthly survey round are made publicly available through the World Bank Microdata Library, under the High- Frequency Phone Survey collection (World Bank 2020a,b,c,d). The sample for these surveys is drawn from households that had been interviewed during the latest round of the national longitudinal household survey implemented by the respective national statistical office, with assistance from the World Bank LSMS-Integrated Surveys on Agriculture (LSMS-ISA) initiative. The pre-COVID-19 LSMS-ISA-supported surveys were designed to be representative at the national, regional, and urban/rural levels. These surveys include the Ethiopia Socio-economic Survey (ESS) 2018/19, Malawi Integrated Household Panel Survey (IHPS) 2019, Nigeria General Household Survey (GHS) - Panel 2018/19, and Uganda National Panel Survey (UNPS) 2019/20. Specific details on each survey are available through the World Bank Microdata Library. 2 We examine the effects of the pandemic on and coping strategies employed by 10,865 households across the four countries, between April and September 2020. Details on the timing of the rounds in each country, as well as the response rates and number of households interviewed in each round are presented in Table 1. We use survey weights that are disseminated in the public use data sets. These weights are calibrated to correct for selection bias associated with not interviewing households that do not own mobile phones or that cannot be reached despite repeated call attempts. The correction for selection bias allows us to provide estimates of the total number of households, adults, and children associated with any of the reported outcomes. 3 Table 1: Round-specific response rates for World Bank LSMS-supported high-frequency phone surveys on COVID-19 Round 1 Round 2 Round 3 Round 4 Round 5 Ethiopia 22 Apr - 14 May - 3 Jun - 27 Jul - 24 Aug - Dates 13 May 3 Jun 26 Jun 14 Aug 17 Sep Response Rate 60% 96% 94% 89% 85% # of Attempted Interviews 5,374 3,249 3,241 3,249 3,249 # of Completed Interviews 3,249 3,107 3,058 2,878 2,770 Total # of Households in Pre-COVID-19 Survey 6,770 Malawi 26 May - 2 Jul - 12 Aug 14 Sep - Dates 14 Jun 16 Jul 27 Aug 29 Sep Response Rate 74% 95% 94% 95% # of Attempted Interviews 2,337 1,729 1,722 1,711 # of Completed Interviews 1,729 1,646 1,616 1,618 Total # of Households in Pre-COVID-19 Survey 3,181 Nigeria 20 Apr - 2 Jun - 2 Jul - 9 Aug - 7 Sep - Dates 11 May 16 Jun 16 Jul 24 Aug 21 Sep Response Rate 65% 93% 93% 95% 96% # of Attempted Interviews 3,000 1,950 1,925 1,881 1,856 # of Completed Interviews 1,950 1,820 1,790 1,789 1,773 Total # of Households in Pre-COVID-19 Survey 4,976 Uganda 3 Jun - 31 Jul - 14 Sep - Dates 20 Jun 21 Aug 7 Oct Response Rate 93% 93% 91% # of Attempted Interviews 2,421 2,410 2,352 # of Completed Interviews 2,257 2,230 2,147 Total # of Households in Pre-COVID-19 Survey 3,098 Note: Rows report the response rate, number of attempted interviews, and number of completed interviews for each country in each round. Dates during which interviews were conducted for each round are also reported. The total number of households in the pre-COVID-19 surveys is reported in the bottom row, which does not vary by round and also includes households that do not have any phone contact information and that are outside the scope of the phone survey. The response rate is calculated as (# of completed interviews)/(# of attempted interviews). The number of attempted interviews is declining over time since the surveys do not attempt to recontact households that refuse to be interviewed in a given round. 4 We examine a series of outcomes of the pandemic, including behavior and perceptions related to COVID-19; losses in income and business revenues; changes in food security; and variation in access to resources including education, food, and medical services. Any reported statistically significant inter- temporal differences in our outcomes of interest come from statistical tests with significance set at the 95% level. The tests are conducted through linear regressions in accordance with the method outlined by Josephson et al. (2020). This analysis allows us to estimate the impacts of the pandemic, focusing on heterogeneity in effects across time. The code to reproduce all of the analysis in this paper is available in Furbush (2021). Tracking how people’s lives are differentially affected by the COVID-19 pandemic over time can enable governments and policy makers to better understand the circumstances faced by their citizenry and to make data-driven, informed policy decisions. The longitudinal data collected through the high-frequency phone surveys and our examination of these data over time cultivate this understanding by documenting the changing consequences of the pandemic. Behavior and Perceptions during COVID-19 As the SARS-CoV-2 virus spread in 2020, countries in Sub-Saharan Africa followed worldwide trends by closing schools and issuing stay-at-home orders. Ethiopia closed schools and suspended public gatherings on March 16. This was followed by the declaration of a state of emergency on April 8, which included closing non-essential businesses and limiting international and domestic travel. Nigeria's response primarily occurred at the state-level. Most states closed schools and suspended large gatherings by March 24. By early April, most states had closed all non-essential businesses and suspended inter-state travel. Uganda closed schools, limited large gatherings, and closed the international border on March 18. By March 30, Uganda had closed all non-essential businesses and suspended public and private transport. The President of Malawi declared a state of disaster on March 20, which included closing schools and limiting the size of public gatherings. Following other countries, Malawi issued a stay-at-home order on April 14. However, the order faced legal challenges, with the High Court barring the government from implementing the lockdown. As a result, Malawi, unlike the other countries in our study, instituted no stay-at-home order or lockdown. The cross-country differences in government responses are reflected in how individuals changed their behavior at the beginning of the pandemic to help prevent the spread of the virus (Figure 1). Using the 5 Figure 1: Change in behaviour to reduce exposure to COVID-19 100 80 Percent of individuals 60 40 20 0 Ethiopia Malawi Nigeria Uganda Avoided crowds Increased hand washing Avoided physical contact Note: Figure reflects first month of available data in each country, which was April for Ethiopia, May for Nigeria, and June for Malawi and Uganda. first round of phone survey data in each country, we see that the increased adoption of handwashing was near universal, as was avoiding physical contact, in Ethiopia, Nigeria and Uganda. In Malawi, nearly 90% of individuals increased handwashing and avoided physical contact more frequently than before. However, since Malawi implemented no stay-at-home order, individuals were significantly less likely to avoid crowds compared to the three countries where gatherings and travel were limited. In Uganda and Malawi, the behavioral change questions were asked in follow-up phone survey rounds, allowing us to estimate the number of people continuing to wash hands and avoid crowds over time (Figure 2). In both countries, we see a significant decline in the number of individuals who are avoiding crowds and washing their hands more frequently. Avoiding physical contact remained high in Malawi and Uganda, though declined significantly in Uganda relative to prior rounds. Starting in July in Malawi and August in Uganda, the phone survey also included questions about mask wearing. Wearing a mask in public increased in Malawi between July and September from below 40% to almost 90%. The already high rates of mask wearing in Uganda increased slightly between July and August. The jump in mask wearing in 6 Malawi coincides with a national mask mandate which went into effect in September. Uganda had instituted a mask mandate in July, before the question was added to the survey. The low level of behavioral change during the early months of the pandemic in Malawi is reflected in a greater prevalence of beliefs in common misconceptions and false claims about the virus and the disease (Figure 3). As with behavioral change over time, only the phone surveys in Malawi and Uganda asked these questions (but limited them to the first phone survey round). Relative to Ugandans, Malawians are significantly more likely to believe that Africans are immune to the virus, that children are not affected by the virus, and that the disease is no different than the common flu. Despite Malawians' increased likelihood to hold several of these false beliefs, the sheer number of people in Uganda that lack accurate information about the virus and disease is greater than in Malawi. The most common false belief in Malawi is that the virus is just the common flu, believed by an estimated 3.1 million individuals. The most common false belief in Uganda is that the virus cannot survive warm weather, with an estimated 6.1 million individuals believing this myth. An estimated one million to three million individuals in each country subscribe to the other false beliefs, representing a substantial share of the population. The prevalence of false beliefs about the disease have been noted in nearly every country in the world and demonstrate the continued need for clear and accurate messaging to avoid the far-reaching spread of information of questionable quality (Gallotti et al., 2020). 7 Figure 2: Inter-temporal change in behaviour to reduce exposure to COVID-19 Malawi Uganda 100 100 80 80 Percent of individuals 60 60 40 40 20 20 0 0 June July Aug Sept June Aug Sept Avoided crowds Increased hand washing Avoided physical contact Wore mask in public Note: Figure presents data from each wave for Malawi and Uganda. Behavioural questions were only asked in the first round for Ethiopia and Nigeria. Questions regarding mask wearing were introduced into the Malawi survey in July and into the Uganda survey in August. 8 Figure 3: False beliefs regarding COVID-19 Malawi Africans are immune to coronavirus Uganda Malawi Coronavirus does not affect children Uganda Malawi Coronavirus cannot survive warm weather Uganda Malawi Coronavirus is just common flu Uganda 0 20 40 60 80 100 Percent False True Note: Questions about false beliefs were only asked in Malawi and Uganda and only in the first round. Economic and Social Challenges of COVID-19 As countries moved to curb the spread of the virus, millions of individuals in Ethiopia, Malawi, Nigeria, and Uganda found themselves out of work, both in the formal and informal labor markets (Nonvide, 2020). In the months immediately following the imposition of COVID-19 related restrictions, Josephson et al. (2020) found that 256 million individuals in these four countries - 77% of the population - reported a loss in income. Since that time, fewer households in all four countries reported a loss of income. In each round, the reference period for questions concerning loss of income is the prior interview, typically about a month before. So, for households asked about income in September, the question inquires about how income changed since the August interview. As of August/September, Ethiopia and Uganda saw the most substantial decline in income loss relative to the previous month, particularly with regards to business income. In the early months, over 85% of households in Ethiopia and 90% in Uganda reported loss of income from this source. By September, the share of households reporting decreased business income 9 relative to the previous interview period fell to 50% in Ethiopia and 44% in Uganda - while still representing a substantial share of the population. Though these findings suggest that income recovered relative to previous months, it is unclear how household income in each round compares to pre-pandemic times. Further, in Malawi and Nigeria the monthly recovery of income was less substantial. In these countries, a majority of households still reported a decline in income in the last available survey round. These findings suggest that economic recovery may be slow without additional assistance by governments and the international development community. Figure 4: Households reporting decreases in income Ethiopia Malawi 100 100 Percent of households 80 80 60 60 40 40 20 20 0 0 Apr May June July Aug Sept June July Aug Sept Nigeria Uganda 100 100 Percent of households 80 80 60 60 40 40 20 20 0 0 May June July Aug June July Aug Sept Farm income Business income Wage income Remittances All else Note: To have lost income, the household must have previously received income from that source in the previous 12 months. The reference period for the first round was the pre-Covid-19 period. For all subsequent rounds, the reference period was the period since the last phone call. As with income, many households saw a recovery in non-farm enterprise (NFE) revenues. The findings are encouraging, as they reveal consistent recovery in NFE revenues, particularly in Malawi and Uganda. Each month, households in these two countries reported NFE revenue improved relative to the month previous. There is also evidence of recovery in Ethiopia and Nigeria, though the recovery appears to be bumpier in these countries. In Ethiopia, after an improvement in NFE revenue in May and June, there 10 was a reversion in August relative to June. This may be related to the Hachalu Hundessa riots that occurred in late June and early July across much of the Oromia Region. The riots resulted in the imposition of business and travel restrictions by the federal government. A similar decline in NFE revenues occurred in Nigeria in July, though the reasons are not readily obvious. Some good news is that NFE revenues appear to be back on a trajectory towards recovery in both Ethiopia and Nigeria after July. Figure 5: Households reporting change in business revenue (%) April May Ethiopia June Aug Sept 0 20 40 60 80 100 June July Malawi Aug Sept 0 20 40 60 80 100 May June Nigeria July Aug Sept 0 20 40 60 80 100 June Uganda Aug Sept 0 20 40 60 80 100 Less than before Same as before Higher than before Note: Business revenue refers to revenue from non-farm enterprises (NFEs). To have lost revenue, the household must have previously received revenue from their NFE in the previous 12 months. The reference period for the first round was the pre-COVID-19 period. For all subsequent rounds, the reference period was the period since the last phone call. To further understand the impacts of the pandemic on households, we estimate the prevalence of moderate or severe food insecurity among the adult population, as measured by the Food Insecurity Experience Scale (FIES) (Figure 6). The FIES is an experience-based metric of food insecurity severity, which relies on people’s direct responses to questions about their experiences with access to adequate food. This metric makes it possible to compare prevalence rates of food insecurity across national and sub- national populations (FAO, n.d.). 11 In the high-frequency phone surveys that inform our analyses, the FIES questions had a reference period of the last 30 days. Following the FIES standard survey model, eight questions were asked, aimed to capture whether the respondent or other adult household members: (i) were worried they would not have enough to eat, (ii) were unable to eat healthy and nutritious food, (iii) ate only a few kinds of foods, (iv) had to skip a meal, (v) ate less than they thought they should, (vi) ran out of food, (vii) were hungry but did not eat, or (viii) went without eating for a whole day. For information on how the FIES is calculated, see Josephson et al. (2020). Food insecurity prevalence was highest in Malawi and Nigeria, with the prevalence of moderate or severe food insecurity greater than 60% among the adult population in each round of the survey. However, the prevalence of food insecurity in both countries declined across time. While the declines were relatively modest, they are statistically significant. In Uganda and Ethiopia, most rounds had a statistically significant change in food insecurity from the prior round. In both countries, the prevalence of moderate or severe food insecurity remained lower than the levels in Malawi and Nigeria, with the prevalence being just over 40% among the adult population, at its highest level. Over the period of June to August for which food insecurity estimates are available for all countries, the cross-country prevalence of moderate or severe food insecurity among the adult population declined from 61%, representing approximately 100.3 million adults, to 58%, representing 97.6 million adults. 12 Figure 6: Prevalence of moderate or severe food insecurity 100 80 60 Prevalence 40 20 0 Ethiopia Malawi Nigeria Uganda May June July Aug Sept Note: FIES is calculated in reference to the adult population in each country following Josephson et al. (2020). To understand how households are coping with the pandemic, respondents were asked about their concerns related to COVID-19, including if they are worried about falling ill with COVID-19 and if they are concerned about the financial threat that the virus poses. In the first phone survey rounds with available data, the concerns about contracting the virus and managing the financial ramifications were high across the four countries, ranging from about 70% of households in Ethiopia to over 95% in Malawi (Figure 7). The high level of concern in Malawi may be related to the lack of COVD-19 restrictions within the country relative to other countries. As discussed above, the Malawian President’s efforts to install a lockdown were blocked by the country’s High Court, making Malawi the only one of the four countries not to institute a stay-at-home order or lockdown. These differences in policy may have affected the level of concerns among Malawians. 13 Figure 7: Concerned about falling ill and financial threat of COVID-19 in the first phone survey round Concerned that family or self will fall ill with COVID-19 (%) Ethiopia Malawi Nigeria Uganda 0 20 40 60 80 100 Concerned about the financial threat of COVID-19 (%) Ethiopia Malawi Nigeria Uganda 0 20 40 60 80 100 No Yes Note: The figure is based on the data from the first phone survey round in each country. The data for Ethiopia, Malawi, and Uganda are from June while the data for Nigeria are from May. The same questions regarding concerns were asked in subsequent rounds in Malawi and Uganda, allowing for comparisons across time (Figure 8). In Malawi, the share of respondents that were concerned about falling ill with COVID-19 declined significantly, from over 92% in June to 84% in September. This coincides with the implementation of a mask mandate in the country, which went into effect in September. Compared to Malawi, the share of respondents that were concerned about falling ill with COVID-19 remained stable in Uganda, though at lower levels, ranging from 75% in June to 71% in September. The level of concern about the financial threat of COVID-19 was unchanged in subsequent rounds in both Malawi and Uganda, with no statistically significant changes in Malawi across time and statistically significant, though modest, decline in Uganda. The latter finding is correlated with the significant increase in income and NFE revenues in Uganda (as reported in Figures 4 and 5). 14 Figure 8: Concerned about falling ill and financial threat of COVID-19 over time Concerned that family or self will fall ill with COVID-19 (%) June July Malawi Aug Sept 0 20 40 60 80 100 June Uganda Aug Sept 0 20 40 60 80 100 Concerned about the financial threat of COVID-19 (%) June July Malawi Aug Sept 0 20 40 60 80 100 June Uganda Aug Sept 0 20 40 60 80 100 No Yes Adaptation and Resilience to COVID-19 The combination of the spread of COVID-19 itself as well as the attempts to limit its spread led to households suffering a variety of shocks to their economic well-being. These shocks were not limited to COVID-19-related illness or death of an income earner, but also included job loss, business closure, disruption of farming activity, rising input prices, falling output prices, or increasing food prices. Around 42% of households across the four countries reported suffering from one of these shocks in the early months of the pandemic. To cope with these shocks, households adopted a number of strategies (Figure 9). The estimates for adoption of a coping strategy are conditional on the experience of a shock: that is, to employ a coping strategy, a household must have experienced a shock since mid-March. More than 40% of the household population across all countries and survey rounds adopted at least one coping strategy. These strategies include living off of savings, selling assets, reducing food or non-food consumption, receiving help from family, and receiving government assistance. In all countries, the most frequently used coping strategies were to rely on savings and to reduce food consumption. However, it is striking that many households in Ethiopia, Malawi, and Uganda did nothing to cope with an experienced shock. Nigeria stands 15 apart in this regard, as Nigerian households relied heavily on reduction of food consumption. This comports with the elevated levels of food insecurity in Nigeria (as reported in Figure 6). Figure 9: Households reporting use of coping strategy Ethiopia Malawi 100 100 Percent of households 80 80 60 60 40 40 20 20 0 0 April May June Aug Sept July Aug Nigeria Uganda 100 100 Percent of households 80 80 60 60 40 40 20 20 0 0 May July June Relied on savings Reduced food cons. Reduced non-food cons. Help from family Sale of asset Did nothing Note: To adopt a coping strategy, a household must have experienced a shock to their livelihood since mid-March. Further, the phone surveys asked respondents whether they received social assistance from governments, NGOs, and other sources, either as cash, food, or in-kind. Despite the scope of the economic shock and the relatively high incidence of relying on negative coping strategies, the receipt of any form of social assistance remained low in all four countries (Figure 10). The vast majority of households, more than 80% of the population, reported having received no assistance of any form from any source. The incidence of receiving any social assistance is lowest in Ethiopia – no more than 10% of households reported having received any social assistance from any source in a given survey round. In Nigeria, the incidence of receiving food assistance declined in July and August, and this finding aligns with our findings related to the use of reduced food consumptions as a coping strategy (Figure 9). Uganda is the only country where households reported an increase in assistance over the rounds, in particular, in-kind assistance grew rapidly, with about 23% of households receiving in-kind assistance (in the form of soap, masks, and mosquito nets) in August compared to less than 3% in June. 16 Figure 10: Households receiving social assistance Ethiopia Malawi 100 100 Percent of households 80 80 60 60 40 40 20 20 0 0 April May June Aug Sept June July Aug Nigeria Uganda 100 100 Percent of households 80 80 60 60 40 40 20 20 0 0 May June July Aug June Aug Sept Cash Food In-kind Any assistance Note: The reference period for the first round was the period since mid-March. For all subsequent rounds, the reference period was the period since the last phone call. Households were asked for both the type of assistance and its source (government, NGO, religious organizations, international organizations, volunteers, or other sources). The challenges of coping with the impacts of the pandemic were exacerbated by the inability of households to access basic necessities, including medical services (Figure 11) and staple foods (Figure 12). Conditional on having sought medical services, between 5% and 23% of households were unable to access the services in months for which data are available, with the highest rates in Nigeria in May and lowest in Ethiopia in August. Of those unable to access medical care, the most common inhibitor reported in Ethiopia, Nigeria, and Uganda was lack of money with 50%, 79%, and 83% citing the issue, respectively. By contrast, only 22% of Malawians who sought medical services reported lack of money as a hindrance, while 36% reported no medical personnel and 13% reported the facility was full. The share of households unable to access medical services declined significantly between the first and most recent round of available data in each country, except Malawi, where data are only available for June and July. Generally, the ability of most households to get the medical attention they seek is encouraging and is consistent across time, suggesting that these countries are not facing an immediate strain on their health sector. This picture aligns with household experience regarding access to medicine, where only a minority of households (34% or less in 17 each country and round) report an inability to purchase the medicine they need. Fortunately, a majority of households can access medical care, including both medicine and services, though that should not obscure the fact that many households across all countries struggle to obtain these basic necessities. Figure 11: Share of households unable to access medical services Ethiopia Malawi Percent unable to purchase 100 100 80 80 60 60 40 40 20 20 0 0 April May June Aug Sept June July Nigeria Uganda Percent unable to purchase 100 100 80 80 60 60 40 40 20 20 0 0 May June July Aug June Aug Sept Note: A household's ability or inability to purchase medical services is conditional on that household having sought to obtain medical care. Access to staple foods, on the other hand, varied greatly from country to country. Less than 20% of households in Uganda and Ethiopia reported difficulties accessing staple foods, while more than 40% of households in Nigeria reported similar challenges. In Nigeria and Ethiopia, accessing food was increasingly difficult over time, though these changes are only significant in Ethiopia. Conversely, in Uganda difficulties decreased significantly over time. In Malawi, access remained relatively constant at just under 30% of households unable to access staple foods in June and July. In Ethiopia and Nigeria, households were asked which specific staples they struggled to obtain. In Ethiopia, households consistently struggled to access teff, with between 36% and 43% of households unable to buy the teff they need from month-to-month. Of those unable to access teff, about 60% of households were unable to access the food due to decreased income, and 34% were unable to access the food due to an increase in price. Similarly, in Nigeria, where households reported a relatively high prevalence of food insecurity, households indicated issues accessing yams. Of the 18 over 60% of households unable to buy the yams they need, 81% reported lack of money as a key inhibitor and 13% cited an increase in price as an issue in securing the food. Figure 12: Share of households unable to access staple foods Ethiopia Malawi Percent unable to purchase 100 100 80 80 60 60 40 40 20 20 0 0 April May June Aug Sept June July Nigeria Uganda Percent unable to purchase 100 100 80 80 60 60 40 40 20 20 0 0 May July June Sept Note: A household's ability or inability to purchase staple foods is conditional on that household having sought to obtain staple food not already in their possession. Prior to the onset of the pandemic, substantial worldwide progress had been made towards achieving Millennium Development Goal 2: universal primary education. As of 2015, primary school enrollment rates in low- and middle-income countries had reached 91% (MDG, 2015). In our data, an estimated 96% of households with school-aged children had their children attending school before the outbreak (Josephson et al., 2020). However, following school closures, the incidence of children who were previously attending school engaging in any learning activity fell to an estimated 46% in the first months of the pandemic. Children’s ability to engage in learning activities varied across countries and across time (Figure 13). In Ethiopia, only 20% of school-aged children were engaged in any sort of learning activity immediately after school closures. This number increased through June but then, following the Hachalu Hundessa riots, began to fall. In the other three countries, although child engagement in learning activities was fairly stable over time, it was much higher in Nigeria and Uganda (around 60%) than in Malawi (around 19 20%). The continued widespread lack of educational engagement is likely to have long-term impacts on school-age children. Figure 13: Households with children engaged in learning activities Ethiopia Malawi 100 100 Percent of households 80 80 60 60 40 40 20 20 0 0 April May June Aug Sept June July Nigeria Uganda 100 100 Percent of households 80 80 60 60 40 40 20 20 0 0 May June July Aug Sept June Aug Sept Note: Only households with school-aged children present are considered. Schools in all four countries remained closed through July, and countries began phased-in re- openings between August and October. Given the extended period without traditional classes, households with school-aged children sought out new ways to re-engage in learning activities. Households used technologies, such as radios, televisions, and mobile learning apps to try and mitigate educational losses (Figure 14). Educational radio programs were commonly used by households in Ethiopia, Malawi, and Nigeria, across all rounds. Further, there was an increasing incidence of sessions with a teacher or tutors for children in Ethiopia and Nigeria, coinciding with the relaxation of restrictions. Uganda employed two unique strategies of engagement, including distribution of reading materials from the government, which tapered off across rounds, and reviewing textbooks and notes from past classes, which increased across rounds. Understanding the changing landscape of engagement in learning activities is important for facilitating aid and support to those families with school-aged children. There is a need to minimize the loss of educational attainment due to the global pandemic if countries are going to continue to achieve the education-related targets under the Sustainable Development Goals. 20 Figure 14: How children are engaging in learning activities Ethiopia Malawi 100 100 Percent of households 80 80 60 60 40 40 20 20 0 0 April May June Aug Sept June July Nigeria Uganda 100 100 Percent of households 80 80 60 60 40 40 20 20 0 0 May June July Aug Sept June Aug Sept Educational radio programs Used mobile learning apps Watched education television Session with teacher Read material from government Reviewed textbooks and notes Note: Only households with school-aged children present are considered. Summary In order to develop policies targeted at mitigating the adverse health and economic impacts of the COVID- 19 pandemic, governments and aid agencies require reliable and timely data on the present circumstances faced by households. While households began to see recovery in income, business revenues, and food security, the gains were relatively modest. Additionally, households received very little outside assistance and continued to struggle to cope with shocks and to ensure that school-aged children were engaged in educational activities. These findings are limited as we have not compared the pandemic data with the pre-pandemic data for the same households. Further, any evidence on the evolving effects of the virus on households, individuals, and children will simply be a fleeting glimpse of a rapidly changing and dynamic disease environment. However, these findings still give a current picture of the circumstances faced by individuals in households in Ethiopia, Malawi, Nigeria, and Uganda. 21 While the recovery is encouraging, one must recall the disease context for these countries. Compared to countries in North America and Europe, the confirmed COVID-19 cases remain low in Ethiopia, Malawi, Nigeria, and Uganda. The picture that emerges from the data is that households saw a substantial economic hit due to the pandemic and a global economic slow-down. A concern exists that as governments in these four countries ease restrictions, and individuals become complacent, the virus will begin to spread at rates similar to those in other parts of the world. Though so far spared the high rates of infection and mortality, countries in Africa may still experience disease burdens equivalent to those elsewhere in the world. This is especially likely since these countries will not see vaccines at sufficient levels to reduce transmission until 2022 (Khamsi, 2020). Without additional assistance, households are likely to continue to suffer economically and be in a particularly precarious position if widespread outbreaks occur in those countries. Finally, our findings point to substantial scope for strengthening social protection systems to help cope with the widespread and evolving income losses, food insecurity, and insufficient access to education. References Djankov, S and U. Panizza (2020), Covid-19 in Developing Economies, CEPR Press. Food and Agriculture Organization of the United Nations (n.d.), “Voices of the Hungry: Brining experience-based food insecurity measurement to the global level”. Furbush, A (2021), "The Evolving Socioeconomic Impact of COVID-19 in Four African Countries: Replication Code, Version 1.0", Zenodo, https://doi.org/10.5281/zenodo.4549786. Gallotti, R, F Valle, N Castaldo, P Sacco and M De Domenico (2020), “Assessing the risks of ‘infodemics’ in response to Covid-19 epidemics”, Nature Human Behaviour 4: 1285-93. Josephson, A, T Kilic and J D Michler (2020), “Socioeconomic Impacts of Covid-19 in Four African Countries”, World Bank Policy Research Working Paper No. 9466. Khamsi, R (2020), “If a Coronavirus Vaccine Arrives, Can the World Make Enough?”, Nature 580: 578- 80. Nonvide, G (2020), “Policy for limiting the poverty impact of Covid-19 in Africa,” in S Djankov and U Panizza (eds), Covid-19 in Developing Economies, CEPR Press. United Nations (2015), The Millennium Development Goals Report 2015. World Bank (2020a), Ethiopia - High-Frequency Phone Survey on Covid-19 2020. World Bank (2020b), Malawi - High-Frequency Phone Survey on Covid-19 2020. World Bank (2020c), Nigeria - High-Frequency Phone Survey on Covid-19 2020. World Bank (2020d), Uganda - High-Frequency Phone Survey on Covid-19 2020. 22