STUDY LABOUR AND SOCIAL JUSTICE POVERTY AND IN­EQUALITY DURING THE PANDEMIC The diverging experiences of Kenya and Senegal Rudolf Traub-Merz September 2022 • The governments of Kenya and Senegal both reacted quickly to contain the Covid-19 pandemic. But they applied very different lockdown policies and social relief measures to assist vulnerable households. • This paper looks at the impact of these policies on income and income distribution in the informal economy and provides a case study that illustrates how policies can be applied to reduce poverty and inequality even during a pandemic. • The report is based on representative household surveys in the pre-Covid and post-peak Covid period and draws on data from the FES-IDOS-ILO research project on Informal Employment, Social Security and Political Trust in SubSaharan Africa.

LABOUR AND SOCIAL JUSTICE POVERTY ANDIN­EQUALITY DURING THE PANDEMIC The diverging experiences of Kenya and Senegal

Contents 1 INTRODUCTION 2 2 CONTAINMENT POLICIES AND IMPACT ON BUSINESS AND WORK 3 3 SOCIAL POLICY INTERVENTIONS TO MITIGATE POVERTY DURING THE PANDEMIC 6 4 INCOME DECLINE, POVERTY AND INEQUALITY 8 4.1 Preliminary remarks  8 4.2 Income distribution and poverty  8 4.3 Income distribution and inequality 11 5 SOLIDARITY IN TIMES OF CRISIS – SOCIAL ASSISTANCE TO COPE WITH THE PANDEMIC 13 5.1 Who benefits from government aid 13 5.1.1 Beneficiaries of government aid from ongoing programmes 13  of government aid from Covid-related social programmes 15 5.1.3 Types of Covid aid 15 5.1.4 Amount of Covid aid 16 5.1.5 Beneficiaries and non-beneficiaries of Covid aid 16 5.2 Social assistance from non-state sources 18 6 LOCKDOWN POLICIES, GOVERNMENT AID, POVERTY AND INEQUALITY – SOME CONSIDERATIONS 20 7 SUMMARY AND CONCLUSIONS 22 References 23 Websites 23 List of Figures 24 List of Tables 24 1

FRIEDRICH-EBERT-STIFTUNG – Poverty and in­equality during the pandemic 1 INTRODUCTION Early in 2020, many countries were confronted with the outbreak of the Covid-19 pandemic and, in different intensities, implemented lockdown policies and restricted economic activities. Governments set up support programmes for business and public service staff and enacted social relief measures to vulnerable households to deal with the misery arising from loss of income. This report looks at two countries, Senegal and Kenya, and compares the impact of virus containment and social mitigation policies on poverty and inequality within their informal economies. It shows, with the help of country-wide representative surveys, that lockdown policies and social relief measures deeply impacted poverty and inequality. Insofar as the strength of the implemented measures differed substantially, outcome indicators of poverty and inequality can be related to different policies. Our findings demonstrate that well-designed social assistance programmes were instrumental not only in preventing the worsening of social indicators when a global health crisis struck, but even in reducing poverty and inequality and providing a basis for a socially inclusive economic recovery thereafter. The report is structured in seven sections. Section 2 illustrates differences in government responses to containing the spread of the virus by using the Oxford Covid-19 stringency index. The discrepancies in restrictive policies are substantiated by testimonies from our survey on business closures and work stoppages during the pandemic. Section 3 looks at social programmes announced at the onset of the pandemic and provides a first understanding of the relative weights of the poverty mitigation strategies. Section 4 analyses the trends in income and inequality, and measures the extent to which key social indicators in the two countries worsened or improved. Section 5 looks at beneficiaries of government aid programmes, and establishes an understanding of the extent to which proposed aid was actually delivered to vulnerable households. It further looks at other sources of social help and establishes an understanding of the relative importance of social security provided via state aid and derived from community-based solidarity. Section 6 presents a short discussion of the extent to which changing trends in poverty and inequality can be explained by disparate government policies. Section 7 closes the report with a short summary. The data are taken from FES-IDOS-ILO 1 longitudinal studies implemented in two rounds in Kenya(October 2018 and December 2020) and in Senegal(June 2019 and April 2022). The surveys were conducted as face-to-face interviews in the vernacular languages. The dates of the inquiries allow a comparison of pre-Covid data with post-peak data, when some of the economic lockdown policies had been lifted again. At the time of the second round of data collection, neither country had returned to the more liberal environment of the pre-Covid time and some restrictions on the movement of people remained in place. 1 Main findings are published in: FES/IDOS/ILO(2022). Details on the survey – including technical notes on sample design and sampling process and an operational definition of informal employment – are presented in the appendix. 2

Containment policies and impact on business and work 2 CONTAINMENT POLICIES AND IMPACT ON BUSINESS AND WORK When the pandemic hit and policies to curtail the spread of the virus began to be applied from March 2020 onwards both countries slipped into recession, ending protracted periods of high growth. Kenya saw its GDP growth rate fall from five per cent(2019) to 0.3 per cent(2020), while Senegal slipped from 4.6 per cent(2019) to 1.3 per cent (2020). GDP per capita in Kenya was down by –2.5 per cent and in Senegal by –1.2 per cent. Both countries returned to economic growth in 2021, indicating that the downward pressures emanating from the Covid-19 policies had been overcome(Figure 1). To measure the intensity of the containment policies applied to economic activities we use two approaches. The Oxford Covid-19 Government Response Tracker provides a stringency index that measures the government response based on nine indicators, including school closures, workplace closures, restrictions on public gatherings, internal travel bans and international travel controls.² Our own survey provides data on business closures and work stoppages that either confirm or contradict the results of the Covid stringency index. The two approaches together are used to determine the differences between the two countries’ containment regimes. Judging from the stringency index both governments reacted to the health crisis quickly and with far-reaching interventions. From mid-March 2020 onwards, they implemented serious lockdown policies that peaked in Senegal at 78(out of 100) and in Kenya at 89. Later in the year, some measures were relaxed but applied again when a new wave of the virus hit. Restrictive and liberalisation policies have alternated ever since, although with a tendency to lower the restrictions. By early 2022, socio-economic life had not fully returned to pre-Covid levels and some precautionary measures remained in place(see Figure 2). The curves indicate that government measures in Senegal were nearly always less restrictive than in Kenya. If the two countries’ daily scores are related to each other the restriction intensity in Senegal throughout 2020 on average was only 69 per cent of what it was in Kenya, while in 2021 it was only 54 per cent. Senegal not only established a less restrictive regime to contain the spread of the virus during the first year of the pandemic, but it returned more quickly to a more liberal economic environment in the second year. The stringency index is a composite measurement built on several indicators, any of which may not have the same clamp-down effect on economic activities as others. A difference in scores is indicative but may not fully translate into similar slowdown pressure for the economy. Furthermore, if policies vary at subnational level, the index exposes the Figure 1 GDP growth rates 2017–2022, Kenya and Senegal 8% 7% 6% 5% 4% 3% 2% 1% 0% –1% 7.4% 3.8% 2017 6.2% 4.8% 2018 5.0% 4.6% 2019 1.3% –0.3% 2020 Senegal Kenya Note:* Projected. Source: IMF at https://www.imf.org/en/Countries/SEN; https://www.imf.org/en/Countries/KEN 3 7.2% 6.1% 2021 5.7% 5.0% 2022*

FRIEDRICH-EBERT-STIFTUNG – Poverty and in­equality during the pandemic Figure 2 Stringency Index Government Response Tracker, Kenya and Senegal 100 90 80 70 60 50 40 30 20 10 0 2020 ar 13 M 2020 Jun 01 2020 Aug 30 202 2 0 021 Dec Feb 10 18 2021 Jun 01 2021 Oct 12 2022 ay 16 M Kenya Senegal Source: Data of the Oxford Covid-19 Government Response Tracker are available at https://view officeapps.live.com/op/view.aspx?src=https%3A%2F%2Fraw.githubusercontent.com%2FOxCGRT%2Fcovid-policy-tracker%2Fmaster%2Fdata%2Ftimeseries%2FOxCGRT_timeseries_all.xlsx&wdOrigin=BROWSELINK strictest response level and not the average of all regions. Differences between two countries may therefore be larger or smaller, depending on the extent to which a country applied a regionally stratified approach to the spread of the virus. Our survey allows us to measure the intensity of economic restrictions from a different angle. Instead of evaluating government policies, it looks at the effect of policy restrictions at the grassroots level. For a first reference, we asked whether respondents had had to close down their business or workplace because of the pandemic. The question was directed primarily at own-account workers and micro-enterprise owners. Judging from our survey, the informal economy was at no point brought to a full standstill. While some business activities were seriously affected, others managed to continue largely unabated. There are, however, substantive differences between the two countries(Figure 3). In Kenya, nearly half (46.4 per cent) of respondents were faced with a temporary closing down of their economic activities, while in Senegal, this share was only 20 per cent. The data clearly confirm that the government of Senegal applied a less restrictive economic policy to contain the spread of the virus than did the government of Kenya. 2 A second question on work stoppages was directed to all economic status groups and reached out equally to employees and family support workers. The question included a specification of the length of work stoppages. There are again substantive differences between the two countries(see Figure 4). In Senegal, some 24 per cent of all employment groups reported a work stoppage, while the share for Kenya was 39 per cent. Some of the lower pressure in Senegal appears to be depleted by the fact that average stoppage time there was 18.33 weeks compared with 13.69 weeks for Kenya. But this may have been caused partly by differences in the length of the period considered. Stoppage time for Senegal relates to two years whereas for Kenya it is only nine months. Taking this into consideration, average stoppage time in Senegal during the first year of the pandemic may even be lower than in Kenya. The differences between the restrictive policies applied in Senegal and Kenya and measured by the Covid stringency index are confirmed by the experience of the informally employed at grassroots level. Kenya enacted a containment regime that led to far more curbs on economic opportunities in the informal economy than did Senegal. Consequently, the downward pressure on income and poverty during the first year of the pandemic was much stronger in Kenya than it was in Senegal. 2 Insofar as the question for Senegal covered two pandemic years, whereas in Kenya it concerned only about nine months, we can assume that differences in workplace and business closures in the first year of the pandemic were even more disparate than suggested by our data. 4

Containment policies and impact on business and work Figure 3 Closure of businesses or workplaces because of the Covid-19 outbreak, Kenya and Senegal 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 80.1% 53.6% Status unchanged 27.3% 11.0% temporarily closed/inactive because of lockdowns 15.4% 8.7% temporarily closed/inactive because of economic downturn due to Covid-19 0.3% 3.7% permanently closed Senegal Kenya N(Kenya)= 2035; N(Senegal)= 712. Question Kenya:»Has the status of your business or workplace changed over the past 12 months because of the COVID-19 outbreak?« Question Senegal:»Le statut de votre entreprise ou de votre lieu de travail a-t-il changé au cours des 2 dernières années en raison de l’épidémie de COVID-19?« The questions for Kenya and Senegal differ in regard to the time reference to reflect the different dates of the interviews. In both cases, they refer to the full period of the pandemic. Figure 4 Work stoppages as a result of the Covid-19 outbreak, Kenya and Senegal 80% 75.9% 70% 60% 60.6% 50% 40% 30% 20% 10% 12.8% 4.3% 6.2% 6.9% 10.3% 7.7% 9.4% 5.9% 0% no discontinuation discontinue:<5 weeks discontinue: 5–10 weeks discontinue: 11–20 weeks discontinue:>20 weeks Senegal Kenya N(Kenya)= 2379; N(Senegal)= 1171. Mean work stoppage: Kenya= 13.69 weeks; Senegal= 18.33 weeks. Question: Kenya:»Did you have to discontinue your work in the last 12 months due to the lockdowns or COVID-19 outbreak? Q80.3.1 If YES, for how many weeks did you have to discontinue your work?« Question Senegal:»Avez-vous dû interrompre votre travail au cours des 2 dernières années en raison des confinements ou de l’épidémie de COVID-19? Q80.3.1 Si OUI, pendant combien de semaines avez-vous dû interrompre votre travail?« Note: The questions differ slightly to reflect the different dates of the interviews. In both cases, they refer to the full period of the pandemic. 5

FRIEDRICH-EBERT-STIFTUNG – Poverty and in­equality during the pandemic 3 SOCIAL POLICY INTERVENTIONS TO MITIGATE POVERTY DURING THE PANDEMIC Over the past decade both countries have implemented social protection programmes to assist vulnerable households and individuals. When the pandemic emerged, both countries continued these social protection measures while establishing new aid project related to the outbreak of Covid-19. We shall now look at ongoing and newly set-up programmes and try to obtain an understanding of the strength and coverage of social aid related to Covid-19. KENYA The main pillars of Kenya’s social protection policies are the National Safety Net Programme(NSNP, called Inua Jamii) and the Hunger Safety Net Programme(HSNP). The NSNP is a cash transfer programme for three vulnerable groups: older persons; orphans and vulnerable children; and people with severe disabilities. The HSNP targets four northern counties in the arid zone and suffering from drought. It is difficult to obtain accurate records of coverage and payments. For the flagship NSNP, the HSNP website mentions cash payments to more than 500,000 households on a regular basis and to some 375,000 households in Northern Kenya, depending on weather conditions. 3 Beneficiaries appear to receive 2,000 Kenyan shillings(KES), or 19 US dollars, per month. The HSNP targets some 100,000 households with beneficiaries being entitled to 5,400 KES every two months. 4 According to a research report, in aggregate the cash transfer programmes collectively reach 1.23 million households(Doyle and Ikutwa 2021: 2), but it remains unclear whether this reflects targets or actual beneficiaries. In response to the pandemic, the government set up three new short-term social assistance projects for vulnerable households not enrolled in the NSNP or HSNP. It announced (a) a multi-agency Covid-19 cash transfer covering some 699,000 households;(b) the National Council for Persons with Disabilities(NCPWD) cash transfer to some 33,333 households; and(b) the Kazi Mtaani urban public works programme to some 296,000 young people. In all, the government expected to provide special Covid-19 support to more than 700,000 households and almost 300,000 young people(Doyle and Ikutwa 2022: 31, Table 4). The budget details of the programmes are as follows(Doyle and Ikutwa 2021: 16): – 10 billion KES for the multi-agency Covid-19 cash transfer to which the State Department for Social Protection (SDSP) committed itself; – 500 million KES allocated to the NCPWD for payments in arrears and the new cash transfer; – 342 million KES for Kazi Mtaani phase 1(April–May 2020) from existing allocations under the State Department for Housing and Urban Development’s budget; – 10 billion KES for Kazi Mtaani phase 2 July 2020–March 2021. 5 Coverage and financing of the protection schemes remain more or less in the dark. The 10 billion KES for the Covid-19 cash transfer was announced as a new programme to be implemented with new funding, whereas other measures may(partly) represent re-designed projects for which already earmarked money was transferred. Equally, the figures quoted are targets and no reports are available on actual disbursements. SENEGAL Senegal’s main protection pillar since 2013 is the bourse familial, a conditional cash transfer to some 300,000 households. Annually, trimester payments of 25,000 CFA francs (FCFA) are paid to eligible households. The costs of the programme are set at some 22.5 billion FCFA(see PBPO 2021; IPSOS n.d.) In response to the pandemic, the government announced the Economic and Social Resilience Programme(ESRP) for vulnerable households and individuals not enrolled in ongoing programmes. Under the»Force Covid-19 Fund«, financed by an extra-budgetary fund of 1,000 billion West African francs 3»So far, more than 500,000 households are receiving cash transfers on a regular basis and an additional 374,806 households in Northern Kenya receive cash assistance in the case of extreme weather events.« See: www.hsnp.or.ke 4 The government disburses»KES 537 million under the Hunger Safety Net Programme(HSNP) to 99,494 poor and vulnerable households in Marsabit, Wajir, Mandera and Turkana counties«. Beneficiaries are entitled to 5,400 KES every two months(www.hsnp.or.ke). 5»The programme was unveiled in July last year(2020 – RT) as a form of social protection initiative to cushion the youth and vulnerable citizens in informal settlements from the effects of the coronavirus pandemic. Beneficiaries … each working for 11 days a month. They earn Sh 455 per day.« Nairobi News, 21 October 2021. See also: https://www.kenyans.co.ke/news/61618-governmentextends-kazi-mtaani-program 6

Social policy interventions to mitigate poverty during the pandemic (XOF)(1,610,000,000 US dollars) it included three types of social policies: XOF 15.5 billion for the payment of electricity bills of households for a two-month period(directed to approximately 975,522 households; XOF 3 billion to cover the water bills of 670,000 households for a two-month period; and XOF 69 billion, for the purchase of food for the benefit of one million eligible households. COVERAGE To get an impression of the significance of the social shields erected in response to the pandemic we relate the targeted number of beneficiaries to the two countries’ populations. In 2020, 53.77 million Kenyans were living in 11.79 million households(Table 1). If we take the»optimistic« figure of 1.23 million households being covered by existing social protection programmes and 700,000 households being temporarily added for Covid-19 protection, we arrive at 1.93 million households benefiting either from ongoing or from newly set up relief measures. If we further assume that the youth programme includes an additional 300,000 urban households, we arrive at 2.23 million households. Taking all programmes together, around 18.9 per cent of all households benefit from one social assistance programme or another. Looking only at Covid support programmes, this proportion falls to 5.9 per cent for the main cash and food support and 2.5 per cent for the urban youth works programme. As substantive as these figures may sound, they are certainly not sufficient to reach a majority of the households deemed to be poor. The targets set in Senegal look more impressive. In 2020, some 16.74 million Senegalese were living in 2.32 million households. Setting one million households as a target for food deliveries would reach 43 per cent of all households, whereas aiming at 1.6 million beneficiaries for the electricity and water bill subsidy would benefit 69 per cent. Many households are not connected to the electricity and water supply, however, and so cannot benefit from this subsidy for social-technical reasons. But many others would benefit from the water/electricity bill waiver, as well as from food delivery. While we cannot determine the exact number of targeted households, we can assume that the social mitigation programme was targeting at least 50–60 per cent of all households, if not more. A simple way of comparing the volumes of the Covid support packages is by looking at subsidies per targeted household. The targeted 0.7 million households in Kenya should receive on average 327.6 international dollars(Purchasing Power Parity, PPP) from the main support programme, whereas the one million households in Senegal should receive on average 274.4 dollars(PPP). While the average level of support for vulnerable households can be seen as fairly similar in the two countries, coverage shows marked discrepancies. The Covid aid programme in Senegal was designed to cover more than half of all households, whereas in Kenya it was only supposed to reach some 6 to 8 per cent. Depending on the criteria selected, social relief in Senegal was seven to eight times higher than social relief in Kenya. Table 1 Social Covid aid programmes: scope and coverage, Kenya and Senegal Population(2020) Average size of household(HH; persons)* Number of households(nationwide) Targeted households for Covid aid(key programme) Kenya 53.77 million 4.56 11.79 million 0.7 million cash/ food Targeted households for Covid aid(key programme;% of all households) 5.9% Targeted households for Covid aid(additional programme) Targeted households for Covid aid(additional programme;% of all households) Social Covid budget(main programme; LCU) Social Covid budget(additional programme; LCU) Social Covid budget(total(LCU) Social Covid budget(total;$ international)** Social Covid budget(main programme;$ international) Subsidy per household(nationwide; average;$ international) Subsidy per person(nationwide, average;$ international) Subsidy per targeted households(main programme;$ international) 0.3 million urban young people 2.5% 10.5 billion 10.342 billion 20.842 billion. 455.2 million 229.3 million 38.61 8.46 327.6 * Based on the findings of the FES surveys: Kenya December 2020; Senegal June 2019. ** Conversion rates to international dollars in 2020: Kenya(KEN)= 45.7909842; Senegal(FCFA)= 236,874359. Senegal 16.74 million 7.23 2.32 million 1.0 million food deliveries 43.1% 1.6 million water/ electricity 69.0% 65 billion 22.5 billion 87.5 billion 369.4 million 274.4 million 159.22 22.07 274.4 7

FRIEDRICH-EBERT-STIFTUNG – Poverty and in­equality during the pandemic 4 INCOME DECLINE, POVERTY AND INEQUALITY 4.1 PRELIMINARY REMARKS Some explanations concerning comparison of income levels and income differentials gained from the two inquiry rounds and the two countries are necessary before presenting the findings. Our data allow us to determine levels of income, poverty and inequality for both countries at two points in time, the pre-Covid period and the post-peak of restriction policies. By adjusting for inflation and converting local currency into international dollars, income levels can be directly related. Some limitations arise from the fact that the time lapses between the two inquiry rounds are not congruent and the second inquiry in Senegal(April 2022) took place in the midst of an economic recovery, whereas in Kenya the second inquiry(December 2020) was close to the deep-cutting recession measures enacted at the beginning of the pandemic. Without additional data between the two collection points, it is not possible to know whether changes in income reflect a linear trend or first moved in one direction before turning back. Again, income in Kenya at the end of 2020 may have been more strongly affected by pandemic-related crisis policies than in Senegal in April 2022. A further problem emanated from disparate inquiries into income. The pre-Covid inquiry had asked respondents to self-classify their incomes into pre-fixed ranges of income classes, whereas the second round in the post-COVID peak surveys collected data on income in absolute numbers. To compare income data arranged in income classes with income declared in absolute figures we applied two approaches. The absolute figures from the second-round inquiry could easily be transposed to the income group ranges used in the first round. Income ranges had to be adjusted to take account of inflation. In the case of Kenya, we raised the income ranges of the second round by 13 per cent to take account of the inflation between October 2018 and December 2020, whereas for Senegal, we applied an inflation rate of five per cent for the period June 2019 and April 2022. While these arrangements of income data allow us to compare real income changes within a country, the same method does not work for a comparison of income levels between countries. The surveys in Senegal and Kenya were part of a multi-country research project for which income clusters were designed as multiples of the statutory minimum wage(MW) (<0.5 MW; 0.5 MW–1 MW;<1 MW–2 MW;>2 MW–8 MW; >8 MW). 6 Converted into international dollars, the income ranges in various countries are not identical and MW units in one country may not correspond to the same MW units in other countries with regard to purchasing power. Our data allow us to determine the income development trend within a country but not trends between them. We can partly overcome this data limitation for the post-peak survey, where we are able to use absolute income figures converted into international dollars. This allows us to compare poverty levels directly between two countries. 4.2 INCOME DISTRIBUTION AND POVERTY Income data can be presented in various ways and differ in particular in regard to sources, units and whether it is gross or net income. 7 In our analysis we define income as what is left» after paying tax and after any contributions were deducted« and do not distinguish between labour and non-labour income. Income means disposable income and we focus either on household income, for which the earnings of all members of a household are counted together, or on individual income, which reflects only the income of a single person. INCOME DISTRIBUTION Distribution of individual incomes clustered in multiples of the minimum wage is summarised in Figure 5(Kenya) and Figure 6(Senegal). In 2018 in Kenya, 82.8 per cent of individuals in informal employment earned an income equal to or below the reference minimum wage. 8 This share increased slightly to 83.8 per cent in 2020. The number of individuals 6 For Kenya, which does not designate a single statutory minimum wage but a multitude of rates, differing by location and skills, the approach had to be adjusted. For the first-round inquiry we selected 13,000 KES as a kind of middle or average monthly minimum wage, which for the 2020 survey was raised by 13 per cent to 14,690 KES. 7 In economies with large subsistence production, cash and non-cash income are further criteria for differentiation. As informal employment does not include subsistence work, we ignore in this analysis the otherwise important aspect of subsistence production. 8 See footnote 7 for the meaning of the minimum wage in the case of Kenya. 8

Income decline, poverty and inequality on higher than the minimum wage fell accordingly from 17.8 per cent to 16.3 per cent. The difference appears to be small at first glance, but still implies that 9.2 per cent or nearly a tenth of individuals who in 2018 belonged to the better-income earners(above 1 minimum wage) had lost this status in 2020. The downward trend in income is more pronounced when the low-income ranges are considered. The segment 0.5–1 minimum wage, in 2018 comprising 28.63 per cent, fell to 20.3 per cent in 2020, thus losing 29 per cent of members. The lowest income segment(below 0.5 minimum wage) expanded accordingly by 17.2 per cent. Due to the different stretching of income ranges, it is not possible to conclude that the income depression at the lower end of the income hierarchy was more pronounced than at the upper side, or the other way round. It is however safe to conclude that individual incomes in 2020 were on average lower than in 2018. The data for Senegal show a completely different picture. Here, the lowest income segment(<0.5 minimum wage) shrank by an astonishing 58 per cent, implying that more than every second person managed to migrate upward into a better earning income class. The climbing effect becomes manifest in the higher income clusters. The non-poor category>1 MW–2 MW jumped from 20 per cent to nearly 30 per cent and membership of the well-off categories (> 2 MW-8 MW and>8 MW) increased from 7 to 15.7 per cent. Quite clearly, the informally employed in Senegal managed to substantially improve their livelihoods between June 2019 and April 2022, whereas the situation in Kenya, reflecting the period October 2018 to December 2020 was the opposite. Shifting from individual incomes to the distribution of household incomes, we see these trends somehow modified. For Figure 5 Monthly individual income in multiples of the minimum wage, Kenya 70% 60% 63.5% 54.2% 50% 40% 30% 20% 10% 28.6% 20.3% 13.3% 12.4% 0% >0.5 MW 0.5 MW–1 MW>1 MW–2 MW 3.8% 3.8% >2 MW–8 MW 0.1% 0.1% >8 MW Oct 2018 Dec 2020 N= 1133(2018); N=2147(2020). Question:»Can you give me an estimate of your average monthly income(in KSH)?«[Note: after paying tax and after any contributions were deducted][Read out options]. Income ranges: 2018: 0;>0-<6,500; 6,500-13,000;>13,000-25,000; >25,000-100,000;>100,000. Income ranges for 2020 were increased by 13 per cent. See main text for linking income clusters into multiples of the minimum wage(MW). Figure 6 Monthly individual income in multiples of the minimum wage, Senegal 50% 40% 30% 20% 45.1% 18.9% 35.4% 27.9% 29.9% 20.0% 10% 0% > 0.5 MW 0.5–1 MW>1 MW–2 MW 14.4% 6.7% >2 MW–8 MW 0.3% 1.3% > 8 MW June 2019 April 2022 N=1164(2019); N=1143(2022). Question see Figure 5. Income ranges(FCFA) 2019:<27,500; 27,500-55,000;>55,000-110,000;>110,000-450,000;>450,000. Income ranges 2022: all ranges increased by five per cent. 9

FRIEDRICH-EBERT-STIFTUNG – Poverty and in­equality during the pandemic Kenya, income depression at household level is more pronounced between 2018 and 2020 than at individual income level. In October 2018, 28.2 per cent of households had incomes above the minimum wage reference line; in 2020, this group had shrunk to only 20 per cent(Figure 7). Nearly 30 per cent of households had fallen from a higher income bracket, while at individual income level, it was only nine per cent. The same effect is manifest in Senegal, but in the opposite direction(Figure 8). The share of households in the lowest income segment(<0.5 MW) fell by 79 per cent(6.6 per cent/31 per cent). In April 2022, only 27.2 per cent compared with the pre-Covid 53.6 per cent had to cope with a household income below the minimum wage. The share of non-poor households and better-off households grew accordingly from 22.6 per cent to 38.9 per cent and from 23.8 per cent to 33.8 per cent, respectively. earning jobs, they are counted in both balances. But households may move up the income ladder, not just because the main breadwinner earns more but because a second person starts earning as well. POVERTY Income levels derived from individual incomes are not a good measurement of poverty as a household’s consumption level depends strongly on total income and household size. Poverty is usually related to household income per capita. Per capita income can be converted into international currency for comparative purposes. This approach is applied to our data from the post-Covid peak surveys but it is not possible to calculate per capita income for income clusters, the method used for the preCovid inquiry. It is thus not possible to compare per capita trends between the two inquiry rounds. The household income data confirm a downward trend for income in Kenya for all income ranges. No income class managed to remain unaffected and large numbers of households were gripped by losses in income and thus ended up in lower income classes. All groups in informal employment were affected. Senegal witnessed the opposite effects. All household groups were pulled upwards. A broad recovery benefitted most households across all income levels. The stark differences between individual income and household income are caused by changes in the number of household members earning incomes. If a person earned an income during the pre-Covid period but lost their job or business during the pandemic, the fall in income did not pull down individual incomes, as the person in question was no longer considered an income earner, but it did depress household incomes. The same effect is manifest in the opposite direction. If more household members return to cash We use the well-known definition of 1.90 international dollars per capita per day(2011 PPP) as the threshold for»extreme poverty«, but equally apply other income ranges, as suggested by the World Bank, namely 1.90–3.20 inter­national dollars for»the moderate poor«,>3.20–5.50 international dollars for the»lower middle-income class« and>5.50 international dollars for the»upper income middle-class«. With these income demarcations, we get a differentiated income stratification for the whole of informal labour. Details of calculations and the findings are assembled in Table 2. In Kenya(December 2020) 70 per cent of all informal households are classified as extremely poor. Only about 19 per cent belong to the lower or upper middle-income class. Poverty in Senegal is at a lower level. Some 46 per cent are extremely poor whereas some 29 per cent are members of the lower or upper middle-income class. Figure 7 Monthly household income in multiples of the minimum wage, Kenya 70% 60% 59.6% 50% 40% 30% 20% 10% 0% 44.2% <0.5 MW 27.5% 20.4% 0.5 MW–1 MW 18.0% 14.2% >1 MW–2 MW 9.6% 5.7% >2 MW–8 MW 0.6% 0.1% >8 MW 2018 2020 Kenya: N= 1097(2018); N=2169(2020). Question:»Can you give me an estimate of your average household income in the last four weeks (read out each option)«. Income ranges(KES): 2018:<6,500; 6,500–13,000;>13,000–25,000;>25,000–100,000;>100,000. Income ranges: 2020: all ranges increased by 13 per cent. See main text for linking income clusters to multiples of the minimum wage(MW). 10

Income decline, poverty and inequality Figure 8 Monthly household income in multiples of the minimum wage, Senegal 45% 40% 35% 30% 25% 20% 15% 10% 5% 0% 31.0% 6.6% > 0.5 MW 22.6% 20.6% 38.9% 22.6% 0.5 MW–1 MW> 1 MW–2 MW 32.4% 21.5% > 2 MW–8 MW 2.3% 1.4% > 8 MW June 2019 April 2022 Senegal: N= 1147(2019); N= 1107(2022). Question: see Figure 7. Income ranges(XOF) 2019:<27,500; 27,500–55,000;>55,000–110,000; >110,000–450,000;>450,000. Income ranges 2022: all ranges increased by 5 per cent. Table 2 Poverty by international poverty lines(household income per capita), Kenya and Senegal Poverty lines Income range $ (2011 PPP) Extreme poverty Moderate poverty Lower middle-income class Upper middle-income class 0.00–1.90 1.91–3.20 3.21–5.50 >5.50 Income range KES* (2020, month) 0–2,654 2,655–4,469 4,470–7,682 >7,682 Income range XOF* per capita (April 2022) 0–13,727 13,728–23,119 23,120–39,735 >39,735 Kenya household income per capita (December 2020) 70.0% 11.3% 8.1% 10.6% Note:* PPP conversion rate for Kenya(KES) 2020= 45.7909842. PPP conversion rate for Senegal(XOF) 2020= 236.87. No conversion rate yet available for 2022. Transposed from daily to monthly rates with 30.5 days. Senegal household income per capita (April 2022) 46.3% 24.6% 16.8% 12.3% Even though the share of extremely poor households in Senegal is lower than in Kenya, it is much more than what could be expected from the fact that less than seven per cent fall into the lowest class if grouped by household income(see Figure 8). The reason is to be found in the size of households and in the number of non-earning household members. ­Senegal features at the top of size rankings and the size of the household pull the figures down when per capita income is considered. Even though we are not able to calculate per capita poverty indices for the pre-Covid period the findings on income changes can be used to indicate poverty trends. Income in Kenya in pre-Covid time was slightly higher, so we can assume that poverty in Kenya had been lower than what we found in December 2020. The opposite is the case for Senegal. Here, income increased at all levels and poverty indices declined between June 2019 and April 2022. While we cannot determine the exact courses of these trends, it is certain that poverty levels in the two countries had been closer before the pandemic struck and moved apart thereafter. 4.3 INCOME DISTRIBUTION AND INEQUALITY An increase in poverty is not synonymous with an increase in inequality. Whereas poverty is measured as income below a certain level, inequality measures income in relation to other income. The Lorenz curve is used to indicate the share of percentiles in total income, whereas the Gini index shows the extent to which income distribution deviates from the ideal of perfect distribution. The Gini coefficient varies from 0 to 1, with 0 representing perfect equality(everybody has the same income) and 1 representing perfect inequality(one person earns all income, all others earn nothing). An increase in the coefficient thus suggests that income is becoming more unevenly distributed. Our pre-Covid income data do not allow us to directly calculate the Gini index. We can, however, transform the income ranges into a numerical scale for arithmetic processing by using the middle value of an income range as reference. 9 This 9 If income falls within the range 1–<6,500, it is now considered to be 3,250 KES. If income falls within the range of 6,500–13,000, it is now considered to be 9,750 KES. If Income falls within the range of>13,000–25,000, it is now considered to be 19,000 KES. The lowest income remains at zero, whereas the highest income range (>100,000) becomes 200,000 KES. The same logic applies to the ­Senegal clusters. 11

FRIEDRICH-EBERT-STIFTUNG – Poverty and in­equality during the pandemic Table 3 Gini coefficients, various income categories, 2018 and 2020, Kenya and Senegal Type of income distribution Individual income Household income Household income per capita Kenya October 2018 December 2020 0.478 0.491 0.565 0.506 n.a. 0.586 June 2019 0.534 0.624 n.a. Senegal April 2022 0.445 0.404 0.438 N – Kenya: household 2018= 1097; individual 2018= 1133; household 2020= 2169; individual 2020= 2147; household per capita 2020= 2169; N – Senegal: Household 2019= 1147; individual 2019= 1164; household 2022= 1107; Individual 2022= 1143; household per capita 2022= 1107. approximation blurs the data to some extent as it ignores income inequality within income ranges and assumes that all members within a range earn exactly the mean in that range. The»real« level of inequality may thus be higher or lower, depending on how actual incomes spread within a range. Even then, the calculation with approximated income levels still provides a good insight into the level of inequality within the informal labour segment in 2018. The data for 2020 do not need this approximation and are calculated directly. We did not apply the approximation method for income per capita, as the size of a household would further blur the data. The findings on inequality as measured by the Gini coefficient are summarised in Table 3: – The Gini rates are high for both countries. The informal economy does not feature low levels of inequality and cannot be seen as falling behind inequality in other segments of the economy. Due to the size of the informal labour market, inequality in the informal economy is a strong determinant of inequality in society overall. – Inequality in Kenya between the two inquiry rounds has remained unchanged on individual income accounts and declined in regard to household income. Unchanged inequality at individual income level implies that pressure from Covid-19-policies was felt equally across all income classes; a reduction of inequality at household level implies that better-off households featured several income earners during the pre-Covid period, and when restrictive policy measures were applied, many second earners lost access to income and the difference between individual income and household income disappeared. – Senegal experienced a reduction of inequality on all accounts. Between the two inquiry rounds, the Gini coefficient fell for both individual and household incomes. Inequality during the pre-Covid period had been higher in Senegal than in Kenya. This was reversed. In the postpeak period, inequality in Senegal fell below inequality in Kenya. – The higher level of inequality per capita again reflects household size. Households tend to be larger at the lower end of the income scale, which necessarily increases per capita inequality. 12

Solidarity in times of crisis – social assistance to cope with the pandemic 5 SOLIDARITY IN TIMES OF CRISIS – SOCIAL ASSISTANCE TO COPE WITH THE PANDEMIC When individuals or households experience declining income or are unable to conduct business because of lockdown, they may look for alternative jobs, reduce consumption, exhaust savings or go into debt. Beyond these individual coping strategies two main avenues are left: government aid or assistance from relatives and community-based sources. Individuals or households may benefit from government programmes such as cash transfers or food assistance; or rely on support from neighbours, relatives, kinship groups or other groups. In this section, we look at the various sources of solidarity and try to establish the extent to which they make a difference. In the course of this we attempt to clarify which type of solidarity was more important than others during the pandemic. 5.1 WHO BENEFITS FROM GOVERNMENT AID POLICIES? Senegal and Kenya had established social assistance programmes even before the pandemic, but launched new assistance programmes for vulnerable people and households to help them to cope with Covid-related challenges. To ascertain whether households in informal employment actually benefitted from government aid we requested information about social assistance programmes from two perspectives: heads of households were first asked whether their households had received support from long-established programmes. Comparing the results from the two inquiry rounds, the extent to which long established programmes continued unabated or came to be»amended« could be specified. Secondly, we refocussed the questions onto Covid-­ related programmes. As both support lines are not supposed to overlap, beneficiaries could be summed to determine overall coverage of government aid during the pandemic. As households may not be fully aware of the details of particular aid programmes, we checked people’s identities to avoid double-counting. 5.1.1 Beneficiaries of government aid from ongoing programmes The number of beneficiaries of long established aid programmes are summarized in Figures 9(Kenya) and 10(Sene­ gal). The figures have to be interpreted with some care as the list contains not only support programmes provided by Figure 9 Beneficiaries(households) by type of assistance programme, Kenya 12% 10% 8% 6% 4% 2% 6.0% 3.8% 6.1% 3.8% 7.7% 5.5% 5.9% 1.7% 10.4% 5.7% 3.7% 2.3% 2.4% 1.8% 0% National Safety Net Program (NSNP) Hunger Safety Net Program (HSNP) Other social assistance/ regular cash transfer programme Public works Food for work In-school feeding Other regular payment in kind Oct 2018 Dec 2020 N(2018)=1188; N(2020)=2392. Question: In the last 12 months have you or any member of your household received regular benefits in cash or in kind?(Multiple answers possible.) Note: The figure does not show the answer option»free medical services« as this led to misunderstandings among people who were members of a health insurance scheme. They may treat the payment of a medical bill by the insurance provider as free medical service. But this issue does not affect the fact that coverage of assistance programmes declined. 13

FRIEDRICH-EBERT-STIFTUNG – Poverty and in­equality during the pandemic the government but aid projects from NGOs, international donors and churches as well. Furthermore, there may be substantive overlapping of projects. We therefore focus on programmes clearly aligned with government aid and known as main support lines, namely the»National Safety Net Program«(NSNP) and the»Hunger Safety Net Program« (HSNP) for Kenya and the» bourse des security familiale« for Senegal. Coverage rates are surprising. In Kenya, both the NSNP and the HSNP declined between the two inquiry rounds from some six per cent to below four per cent, reflecting some 35 to 40 per cent fewer beneficiaries. We do not know whether 2018 had been an outstanding year with a high number of recipients, returning to»normal« in 2020. But the pandemic increased the need for support and it might have been expected that the coverage rate would have remained at the 2018 level. This was not the case, however. While the reason remains unclear, it may be that the government cut back of the NSNP and HSNP in order to have more funds available for newly established Covid-19 projects. The case is different for Senegal. The main support line – the bourse familial – was fully maintained during the pandemic. It reached out to relatively more households than the NSNP and HSNP in Kenya and continued unabated when the health crisis hit. Figure 10 Beneficiaries(households) by type of assistance programme, Senegal 25% 23.0% 21.9% 20% 15% 10% 5% 5.6% 4.1% 0% Programme national des bourses de sécurité familiale Food(incl. food for work; school feeding) N(2019)= 1193; N(2022)= 1200. Question: see Figure 9. 0.7% 0.5% Other(cash) transfers June 2019 April 2022 1.0% 0.0% Public works 1.1% 0.2% Other Figure 11 Share of households benefitting from government aid, Kenya and Senegal 80% 70% 69.1% 60% 50% 47.2% 40% 30% 20% 10% 0% 21.9% 7.6% Regular support* 4.2% Covid aid** 11.8% All government aid*** Senegal Kenya N(Kenya)= 2391; N(Senegal)= 1200. * Recipients of regular national programmes. Senegal: bourse familiale; Kenya NSNP+ HSNP. Taken from Figures 9 and 10. ** Recipients of Covid-19 aid who did not receive support from regular national programmes; *** Local and central government are added. Question:»Since the outbreak of Covid-19 have you or any member of your household mainly received any food, cash, subsidy or other ­support from the government or anyone else?«[Read out options]»No«;»Yes, from central government«;»Yes, from local government«; »Yes, from traditional leaders or from my church«;»Yes, from my relatives«;»Yes, from neighbours or members of my community«; »Yes, from NGOs, donors, international agencies or development projects«;»Refuse[to answer]«;»Don’t know«. Note: Figures shows only beneficiaries of government support programmes. 14

Solidarity in times of crisis – social assistance to cope with the pandemic Figure 12 Type of government aid to households, Kenya Food aid 53.7% Cash transfer or unemployment benefit 46.9% Protective equipment(Face mask, sanitation kit) 4.8% Free health insurance/Subsidy of health insurance … 2.3% Business loans 0.3% Other 1.4% 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50% 55% 60% N(cases)= 385. Question:»Which type of support did you receive from the government?«[Read out options][Multiple answers are possible]. »Food aid«;»Cash transfer or unemployment benefit«;»Protective equipment(face mask, sanitation kit)«;»Business loans«;»Wage subsidy«; »Free health insurance/subsidy for health insurance premiums«;»Subsidy for electricity/water costs«;»Reduction, deferral or suspension of the payment of certain taxes or fees«;»Other(specify)«. Note: Figures do not sum to 100 per cent due to multiple answers. Figure 13 Type of government aid to households, Senegal Food aid 96.2% Cash transfer or unemployment benefit 0.5% Protective equipment(Face mask, sanitation kit) 4.3% Wage subsidy 0.6% Subsidy of electricity/water costs 13.0% 0% 10% 20% N(cases)= 748. Note: Figures do not sum to 100 per cent due to multiple answers. Question: see Figure 12 30% 40% 50% 60% 70% 80% 90% 100% 5.1.2 Beneficiaries of government aid from Covid-related social programmes How many households benefited from programmes newly established to cope with pandemic-related income losses? We asked respondents to answer the question:»Since the outbreak of Covid-19 have you or any members of your household received any food, cash, subsidy or other support from the government or anyone else?« The answer options mentioned the various sources, allowing us to collate support with the benefactors. Recipients of regular national support programmes and Covid-aid are treated separately. Counted together, they may be seen as the comprehensive social security shield in place during the pandemic. A huge coverage gap can be documented between the two countries. The government of Senegal implemented a Covid support programme that reached 47 per cent of all households, whereas in Kenya Covid support reached only 4.2 per cent. The discrepancies already noted when the governments announced the social relief packages(see Section 3) are confirmed at the ground. Adding regular ongoing support programmes, the gap increases further. During the pandemic, Senegal provided government aid to 69 per cent of all households in the informal economy, whereas Kenya covered only 12 per cent. 5.1.3 Types of Covid aid Our household survey confirms the composition of Covid aid as announced by the governments. In Kenya, about half of beneficiaries received food aid, whereas the other half benefitted from cash transfers. Other support measures were of minor importance(Figure 12). In Senegal(Figure 13), nearly all(96.2 per cent) received food deliveries. The share of households benefitting from electricity and water relief was fairly low(13 per cent). This can be seen as an indication that the subsidy for electricity and water reached households mainly in the formal economy and largely bypassed households in the informal economy, many of whom do not even have access to the national grid. 10 10 According to our survey, 71.6 per cent of households in the informal economy are connected to the grid. The share of households receiving the subsidy should be in the range of 50 per cent to 60 per cent. This implies that most of the electricity and water subsidy was not targeting households whose members engage in informal labour but those whose members are engaged in formal labour. 15

FRIEDRICH-EBERT-STIFTUNG – Poverty and in­equality during the pandemic Figure 14 Value of Covid aid from government by beneficiary($ international), Kenya and Senegal 225 197.94 200 175 150 132.92 125 100 75 50 25 0 $ international 101.4 85.3 % median income Senegal Kenya Notes: Includes only aid given to heads of household. See footnote 15 for explanation. Support from central government and local government aggregated to»government aid«. Local currency converted into$ international at 2020 rates. Question:»Since the COVID-19 outbreak, have you received any food, cash, subsidy or other support from the government/employer?«[Read out options] Question:»Please provide an estimate of the overall food aid or cash received(in KES) from government and/or employers since the outbreak of Covid-19«. 5.1.4 Amount of Covid aid Broad aid coverage may not be meaningful if the level of support is low. We asked respondents to assess the value of the Covid-related support they had received from government since the outbreak of the pandemic. 11 For aid in kind, respondents were asked to put a monetary value on it and provide an aggregate figure. In Senegal, Covid aid provided by government amounted on average to 49,055 XOF(207.09 dollars – PPP), while in Kenya, it came to 5,550 KES(121.19 dollars – PPP)(see Figure 14). Measured in terms of purchasing power parity, average support for individual households in Senegal was 71 per cent higher than in Kenya. 12 is higher because income levels there are lower. In both countries, government-provided Covid support was sufficient to compensate a median household for one month’s income. The level of Covid aid for a single household can be considered fairly balanced between the two countries or, depending on what criteria are used for comparison, Senegal does better in terms of its aid programme. What remains decisive, however, are the huge disparities in coverage. Taking all aspects together, the aid programme in Senegal reached about ten times more households than the one in Kenya. From this standpoint, Covid aid in Senegal can be called substantial, while in Kenya it was not much more than a drop in the bucket. If we relate the value of Covid aid to the median income, we arrive at similar rates for both countries. In Senegal, Covid aid amounts to 98 per cent of the median monthly income, while in Kenya, it comes to 111 per cent. The rate for Kenya 11 The wording of the question did not distinguish between»received by yourself« and»received by the household«. Some households may have received aid through different members but usually household members inform each other about donations received and, in most cases, respondents provided an aggregate assessment. Government aid to households may be only slightly higher – if at all – than what we were told by our respondents. 12 In a simple comparison we can relate these findings to the Covid aid announcements by the governments. In Senegal, Covid aid was promised in the amount of 274.4 million dollars(PPP) or 274.4 dollars (PPP) for each of the one million targeted households(see Table 1). This is higher than the 207.09 dollars(PPP) of our survey, but does not deviate too far, if the administrative and handling costs of food deliveries are taken into account. Furthermore, we found coverage of 47 per cent, whereas the national target share was set at 43 per cent. It could be that more households were reached than originally planned and average food deliveries were lowered to meet a higher target. In the case of Kenya, main Covid aid amounted to 229.3 million dollars or 327.6 dollars(PPP) for each of the 700,000 targeted households. This is far above the 121.19 dollars in our survey, implying serious implementation problems for the aid programme. 5.1.5 Beneficiaries and nonbeneficiaries of Covid aid The distributive impact of Covid aid can be assessed by comparing recipients and non-recipients. Covid benefits target vulnerable households and individuals and income is a criterion for eligibility. But are officials in charge of allocating aid aware of household incomes and able to respond to specific needs? If poverty is used to decide on beneficiaries, a majority of households in both countries qualify. But is assistance graded? Do people defined as extremely poor get more than those rated only moderately poor? Is there an inverse allocation of benefits which serves to compensate to some extent income inequality? Do those who receive more assistance belong to the very poor, whereas those receiving less assistance have higher incomes? If the very poor benefit more, government aid would function as an inequality-reducing policy. If the allocation of assistance follows the income inequality profile, it would be neutral. If the relatively better-off receive more assistance than the very poor, the government assistance programme would even deepen inequality. 16

Solidarity in times of crisis – social assistance to cope with the pandemic Figure 15 Beneficiaries and non-beneficiaries of government Covid aid by monthly income(KES), Kenya 38% 36% 34% 32% 30% 28% 26% 24% 22% 20% 18% 16% 14% 12% 10% 8% 6% 4% 2% 0% 13.7% 11.3% 0–1,000 35.9% 28.6% 29.0% 15.4% 10.2% 18.8% 11.1% 14.5% 5.9% 5.1% 1,001–2,000 2,001–5,000 5,001–10,000 10,001–20,000 20,001–50,000 N= 2033(non-beneficiaries); N= 117(beneficiaries). non-beneficiaries beneficiaries 0.5% 0.0% >50,000 Figure 16 Beneficiaries and non-beneficiaries of government Covid aid by monthly income(XOF), Senegal 28% 26% 24% 22% 20% 18% 16% 14% 12% 10% 8% 6% 4% 2% 0% <10,000 10,000– 20,000 20,001– 40,000 40,001– 60,000 60,001– 80,000 80,001– 100,000 100,001– 150,000 N(non-beneficiaries)= 785; N(beneficiaries)= 351. non-beneficiaries beneficiaries 150,001– 200,000 <200,000 Figures 15 and 16 reproduce the profiles of recipients and non-recipients of government Covid aid. In Kenya, lower-income segments exhibit an overrepresentation of aid beneficiaries, while higher-income segments are underrepresented. Covid aid, accordingly, was targeted towards lower earning groups and produced a redistributive effect. A similar compensatory effect cannot be detected for Senegal, where Covid aid distribution largely follows income distribution. If we look at beneficiary status by per capita income, however, a redistributive effect becomes visible for Senegal(Figure 17). Government aid covers more households with lower income per capita. But the effect is rather small. level is possible, two findings on the distributive impact of Covid aid on informal labour are important. Neither Kenya nor Senegal focused on the extremely poor. Even though a redistributive effect of Covid aid allocation on income can be noted, it was rather small. In both countries, officials did not take much effort to take income into account when allocating aid to households. Rather, they included households at all income levels, some even at the high-end of the income scale who may not have been eligible to receive aid at all. Due to the high level of poverty prevailing in both countries, the misallocation of aid for the poor to the non-poor was limited. A large majority of those who benefitted deserved to do so. Our data analysis is limited to the allocation of Covid aid to those in informal employment, as we have no information on the distribution of aid to other groups in society. While no statement as to the overall effect of Covid aid at national The key discrepancy in the application of Covid aid is the scope of the programme. The financial support for Covid aid in Senegal outperformed by many times Kenya’s financial package. Government aid in Senegal reached a majority of 17

FRIEDRICH-EBERT-STIFTUNG – Poverty and in­equality during the pandemic Figure 17 Beneficiaries and non-beneficiaries of government Covid aid by monthly per capita income(XOF), Senegal 43% 40% 38% 35% 33% 30% 28% 25% 23% 20% 18% 15% 13% 10% 8% 5% 3% 0% <2,000 2,000– 5,000 5,000– 10,000 10,000– 20,000 20,000– 30,000 30,000– 40,000 40,000– 50,000 50,000– 60,000 non-receivers receivers N(non-receivers)= 557; N(receivers)= 256. Note: Includes only head of households as receivers and non-receivers of Government aid. 60,000 a,m, households, whereas in Kenya government aid was no more than a drop of water on a hot stone. Government aid in Senegal was substantial, whereas it was merely symbolic in Kenya. 5.2 SOCIAL ASSISTANCE FROM NONSTATE SOURCES Social assistance may come from a variety of sources. In addition to state aid, relatives, neighbours and friends, traditional and religious leaders or ethnic communities may render help to households and individuals. To understand the varieties of help and their relative significance, we compared the sources and quantified values. This allows to determine the relevance of social security provided by the state and support from community solidarity. The term»community solidarity« is used as an aggregate term for aid from relatives, neighbours, other members of the community and traditional leaders. Kenya and Senegal produce huge disparities when we compare state-provided aid with non-state aid made available during the pandemic(Figure 18). In Senegal, assistance from community-based sources was hardly forthcoming and a large majority of households(94.6 per cent) received Covid aid from the government, not from relatives, neighbours or other community-based sources. This does not imply that community solidarity in Senegal has died out or that a readiness to provide help to vulnerable households and individuals has vanished at community level. Rather, people at community level are aware of the government’s extensive aid programme and social pressure to help each other may not be as keen as in cases where the government does not step in. During the pandemic Senegal may provide a case of state-provided security crowding out community solidarity. 13 13 See Strupat/Klohn(2018) on this effect for Ghana. In Kenya, community-based solidarity played a much stronger role during the pandemic. It reached even more households than did state-provided aid. The incidence of community-based aid, however, was by far not sufficient to compensate for large vacuum left by the government. Taken together, state and non-state solidarity reached only 13.9 per cent of households, far behind the aid coverage in Senegal. Communities took more care of vulnerable households in Kenya, but their capacity was limited. NGOs and international donors also provided support and played an important role for selected communities. At the national level, they could not play a meaningful role and compensate for shortcomings from other sources, however. Considering all sources, aid coverage in Senegal was 3.7 times higher in Senegal than in Kenya. Here again, the level of aid has to be considered in order to understand the impact of support from various sources during the pandemic. Our data do not allow us to sum all financial help at household level, as we have information only on the amount of aid provided to income earners. We partly circumvent this problem by considering support to income receivers only when they are heads of households. 14 This still excludes aid to households that, at the time of the inquiry, had no income earner at all. However, the data are still indicative in assessing the relative weights of aid from different sources and complement our analysis of disparities between solidarity provided at state- and non-state-level. 14 We assume that(a) the head of household is the best-informed person of a households and is aware, if other members of the household have received Covid-support;(b) that the head of household includes all sources of support to the household, even if he or she have been asked about support provided to him or her. These assumptions make sense as many questions around this one considered household issues as well and from the interview dynamic, the impact on the household is clearly in the mind of the respondent. 18

Solidarity in times of crisis – social assistance to cope with the pandemic Figure 18 Share of beneficiary households by source of aid provider, Kenya and Senegal All sources 16.5% 61.3% Government 6.5% 58.0% Community-based source 2.7% 7.6% NGO/ international donor 0.6% 2.4% 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50% 55% 60% 65% Senegal Kenya N(Senegal)= 1200; N(Kenya)= 2392. Note: Numbers for government aid are slightly higher than those in Figure 11 as they include some double-counting of Covid aid and other assistance. Community-based solidarity consists of support from traditional chiefs or churches; relatives; neighbours or other members of the community. Figure 19 Value of government aid by source, if income receiver is head of household($ international) 225 197.94 200 172.15 175 150 123.27 125 100 75 50 25 0 Senegal 132.92 84.09 150.86 Kenya Government Community solidarity NGOs/ international donors N(Senegal, government)= 256; N(Senegal, community solidarity)=59; N(Senegal, NGO/international donors)= 10; N(Kenya, government)= 76; N(Kenya, community solidarity)= 191; N(Kenya, NGO/international donors)= 19. Figure 19 presents aid amounts by source, provided to heads of households and converted into international dollars. In Senegal, state support to households is slightly above aid from community members(198 dollars(PPP) in contrast to 172 dollars(PPP)), while in Kenya, government aid is some 58 per cent higher than help from community members (133 dollars(PPP) in contrast to 84 dollars(PPP)). On both scores, assistance in Senegal is higher than assistance in Kenya. Help from NGOs and international donors, is more generous in Kenya than in Senegal. If assistance levels were related to median income the differences between Kenya and Senegal would largely vanish, given the lower income levels in Kenya(see Figure 14). In this regard, an analysis of differences in what a single household may receive in support from various sources does not reveal an important new feature in understanding solidarity during the pandemic. 15 Support levels for beneficiaries in the two countries are fairly similar and it is the huge discrepancy in coverage that protects vulnerable households in Senegal better than those in Kenya. The government of Senegal distributed pandemic relief measures to a majority of households, while the government of Kenya provided only an under-sized relief package and never intended to include many people. Community solidarity was not forthcoming in sufficient strength to compensate for non-appearing government aid. When most households were faced with an income crisis, their capacity to help others equally declined. When solidarity at community level is lacking and state support is not forthcoming in sufficient quantity, pandemic-related policy measures that restrict economic activities hit the most vulnerable households hardest. 15 We desist from cross-tabulation of household coverage(Figure 17) and value of support to individual income earners(Figure 18) as the samples are not sufficiently congruent to provide a more detailed profile. 19

FRIEDRICH-EBERT-STIFTUNG – Poverty and in­equality during the pandemic 6 LOCKDOWN POLICIES, GOVERNMENT AID, POVERTY AND INEQUALITY – SOME CONSIDERATIONS The social experience of informal labour during the pandemic in Kenya and Senegal followed opposing trends. The government of Senegal was restrained in clamping down on economic activities and implemented an ambitious social relief package, whereas the Kenyan government applied a more restrictive control regime that led to more work stoppages, while doing less to protect vulnerable groups. The differences in the two countries’ policy frameworks were accompanied by disparate social developments. In Kenya, poverty increased, while in Senegal poverty declined, as did income inequality. Is there a causal relationship between these policy differences and disparate social experiences? Did poverty and inequality in Senegal decline because of the government’s pandemic response? And did poverty in Kenya increase because its government followed a different approach? While there are arguments to support such a view, some shortcomings have to be considered as well. Our longitudinal study offers a comparison on a high-aggregate level of data analysis. We provide a static-comparative analysis with two rounds of inquiries and do not have a sequence of data points at hand that would allow us to assess the oscillation of data say on a monthly or bi-monthly rhythm. We equally do not provide a breakdown into economic branches and do not look at spatial trends. Our analysis is further exacerbated by the fact that the two inquiry rounds were not conducted at the same time. Even so, they are framed in terms of pre-Covid- and post-Covid-peak time management; they compromise different economic phases. Kenya moved from economic growth to recession between October 2018 and December 2020, while between June 2019 and April 2022 Senegal witnessed a growth–recession–growth sequence. Consequently, in Kenya we inquired about income, poverty and inequality at the end of an economic recession, while in Senegal we inquired after the recovery had already set in. Some of the differences we found on poverty and inequality may indeed be explained by these distinctions. Beyond methodological considerations in data collection differences in the structures of the two economies have to be considered as well. Kenya’s economy has a larger agricultural sector(35 per cent vs 16 per cent of GDP) and depends more on rainfall patterns, and overall rainfall amounts, 16 while Senegal has stronger industry(23 per cent vs 16 per cent of GDP) and its more diversified structure may be more receptive to economic stimulus from the government. Diversification increases interdependencies between the formal and informal economy and a stimulus programme, which in Senegal surpassed that of Kenya, has provided a demand push for goods and labour in the informal economy, even if it originally targeted formal sector companies and households. Food production in agriculture in Senegal in 2021/2020 performed particularly well, showing a 25 per cent growth 17 over 2020/2019. This expansion during the first pandemic year offered the government the opportunity to purchase much needed food from local markets, boosting the incomes of peasants and food traders. Social relief would not only have prevented consumption from declining but would have injected demand directly into sectors with high earning potential for the poor. Such interdependencies have to be taken into account and more research and analysis is needed before we fully understand how government policies and market environments impacted on income development in informal employment during the pandemic. But even though we cannot determine the exact extent to which national differences in lockdown policies and social relief measures caused poverty and inequality in Senegal and Kenya to follow different tracks we are confident that they were strong contributing factors. Limiting work stoppages to half what occurred in Kenya and supplying 60 per cent of the informal economy with food allowed many households in Senegal to sail through the pandemic with few financial problems. The government of Senegal provided support to informal employment on the demand and the supply sides, much more than what the government of Kenya did. 16 See, for example Government of Kenya(2022). 17 Grain harvest 2020/2021 over 2019/2020 grew from 2.7 to 3.4 million tonnes. See https://www.gtai.de/trade/senegal/branchen/vieldynamikk-in-senegals-agrarsektor-741882. 20

Lockdown policies, government aid, poverty and inequality – some considerations A World Relief(2022) report summarizes Covid-19’s impact on the world’s poor and warns that the pandemic has increased global numbers of extremely poor people and may roll back virtually every measure of progress from recent years. While not explicitly mentioning Kenya in this report, our findings support such a projection. This is not the case for Senegal, however. Two years after the pandemic hit, poverty and inequality have declined. The government of Senegal steered society through the pandemic paying close attention to social indicators. Senegal in 2022 finds itself socially better positioned than during before Covid. This is not to state that the Senegalese government could not have done more to protect vulnerable persons. And it is not to claim that there is no longer any need to continue the use of social policies to combat poverty. The high number of people still belonging to the extremely poor necessitates further upgrading of social policy measures. 21

FRIEDRICH-EBERT-STIFTUNG – Poverty and in­equality during the pandemic 7 SUMMARY AND CONCLUSIONS Kenya and Senegal are two countries with diverging social experiences emanating from the pandemic and the policies enacted to control the spread of the virus. This paper looks at informal employment and establishes the extent to which the dominant components of the labour market were exposed to poverty and inequality. Two main differences in policy interventions were noted at the outset. Kenya applied a more rigid policy regime than Senegal, which contracted economic activities in Kenya more. At the same time, Senegal enacted social relief measures several times the magnitude of Kenya’s to protect vulnerable households. Data on income and inequality for both countries are taken from a pre-Covid survey and a second survey conducted in the post-Covid peak period. Due to technical and legal problems related to travel restrictions during the pandemic the collecting dates in the two countries could not be harmonized. The time deferrals in the two longitudinal studies may explain some of the differences found, but do not undermine the main findings. In Kenya, some nine months after the pandemic broke out, poverty in informal employment had substantively increased, while in Senegal, two years after the beginning of the pandemic, poverty and inequality were substantially lower than during the pre-Covid time. Poverty increases in Kenya can be considered an expected or»normal« response to economic repression; in the same way, Senegal, with more recovery time, could be expected to »return« to pre-Covid indices of poverty and inequality but »surprised« with its strongly improved indicators. Our survey found empirical proof that in both countries relief measures were implemented»according to plan« with regard to coverage. In Senegal, government aid under the Economic and Social Resilience Programme(ESRP), in particular the 69 billion FCFA package for the purchase of food for one million eligible households, reached some 58 per cent of all households while in Kenya, only some 6.5 per cent of households benefitted from social assistance provided by the state. Community-based solidarity was more forthcoming in Kenya than in Senegal, but it could not compensate for the shortfall of state aid. In Kenya, most households had to face the economic crisis emanating from the pandemic without any support, whereas in Senegal, a large majority of vulnerable households could rely on food deliveries from the government. It is not possible to conclude from our data categorically that the less contractive economic policy and the more generous social policy were the only factors that caused poverty and inequality to decline in the informal economy in Senegal. Other factors may have been at work as well to drive social indicators into this direction. We do not hesitate to conclude, however, that the magnitude of the social relief package in Senegal was instrumental in preventing many households from falling deeper into an income crisis; and the less restrictive regime allowed more households to continue their economic activities unabated than was the case in Kenya. In this regard, Senegal has to be seen not only as a positive example of a country that mastered the social threats from the pandemic better than others; it even used the challenging time to reduce poverty and inequality and finds itself socially better positioned than before the health crisis struck. 22

REFERENCES Doyle, Alexandra /Nivea, Ikutwa(2021): Towards shock-responsive ­social protection: lessons from the COVID-19 response in Kenya. Research report(March); available at: https://www.opml.co.uk/files/Publications/ A2241-maintains/maintains-covid-19-srsp-responses-kenya-case-studyfinal.pdf?noredirect=1 FES/IDOS/ILO(2022): A Majority Working in the Shadows. A six-country opinion survey on informal labour in sub-Saharan Africa. Edited by Traub-Merz, Rudolf / Ö­ hm, Manfred / Leininger, Julia / Bonnet, Florence /  Maihack, Henrik. Bonn: Friedrich-Ebert-Stiftung 2022. Government of Kenya(2022): The short rainy season assessment ­report. Kenya Food Security Steering Group(KFSSG)(February); available at: https://reliefweb.int/report/kenya/kenya-2021-short-rains-season-­ assessment-report-february-2022 IBP(2021): International Budget Partnership, Transparency and Credibility of COVID Response in Senegal, May, available at: https://international budget.org/covid/wp-content/uploads/2021/08/Senegal-Brief-ManagingCOVID-Funds.pdf IPSOS(n.d.): Responding to COVID-19: Highlights of a Survey in Senegal; available at: https://www.ipsos.com/sites/default/files/ct/publication/ documents/2020-05/senegal_report_0.pdf Strupat, Christoph/Klohn, Florian(2018): Crowding out of solidarity? Public health insurance versus informal transfer networks in Ghana, in: World Development 104(C), 212–221. World Relief(2022): Pandemic and Poverty: Covid-19 Impact on the World's Poor(March); available at: https://reliefweb.int/report/world/ pandemic-and-poverty-covid-19-impact-world-s-poor WEBSITES  https://graphics.reuters.com/world-coronavirus-tracker-and-maps/­ countries-and-territories/kenya/  https://www.socialprotection.or.ke/social-protection-components/­ social-assistance/national-safety-net-program/hunger-safety-netprogramme-hsnp  https://ourworldindata.org/covid-stringency-index  https://view.officeapps.live.com/op/view.aspx?src=https%3A%2F%2 Fraw.githubusercontent.com%2FOxCGRT%2Fcovid-policy-tracker% 2Fmaster%2Fdata%2Ftimeseries%2FOxCGRT_timeseries_all.xlsx&wd Origin=BROWSELINK Summary and conclusions 23

FRIEDRICH-EBERT-STIFTUNG – Poverty and in­equality during the pandemic LIST OF FIGURES LIST OF TABLES Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 Figure 11 Figure 12 Figure 13 Figure 14 Figure 15 Figure 16 Figure 17 Figure 18 Figure 19 GDP growth rates 2017–2022, Kenya and Senegal 3 Stringency Index Government Response Tracker, Kenya and Senegal 4 Closure of businesses or workplaces because of the Covid-19 outbreak, Kenya and Senegal 5 Work stoppages as a result of the Covid-19 outbreak, Kenya and Senegal 5 Monthly individual income in multiples of the minimum wage, Kenya 9 Monthly individual income in multiples of the minimum wage, Senegal 9 Monthly household income in multiples of the minimum wage, Kenya 10 Monthly household income in multiples of the minimum wage, Senegal 11 Beneficiaries(households) by type of assistance programme, Kenya 13 Beneficiaries(households) by type of assistance programme, Senegal 14 Share of households benefitting from government aid, Kenya and Senegal 14 Type of government aid to households, Kenya 15 Type of government aid to households, Senegal 15 Value of Covid aid from government by beneficiary ($ international), Kenya and Senegal 16 Beneficiaries and non-beneficiaries of government Covid aid by monthly income(KES), Kenya 17 Beneficiaries and non-beneficiaries of government Covid aid by monthly income(XOF), Senegal 17 Beneficiaries and non-beneficiaries of government Covid aid by monthly per capita income(XOF), Senegal 18 Share of beneficiary households by source of aid provider, Kenya and Senegal 19 Value of government aid by source, if income receiver is head of household($ international) 19 Table 1 Table 2 Table 3 Social Covid aid programmes: scope and coverage, Kenya and Senegal  7 Poverty by international poverty lines(household i­ncome per capita), Kenya and Senegal 11 Gini coefficients, various income categories, 2018 and 2020, Kenya and Senegal 12 24

IMPRINT ABOUT THE AUTHOR IMPRINT Traub-Merz, Rudolf: Consultant. Coordinator of the FES-IDOS-ILO survey project»Informal Employment, Social Security and Political Trust in Sub-Saharan Africa«. Previously, he has served as head of FES offices in various major cities, including Shanghai and Moscow. Published by: Friedrich-Ebert-Stiftung e.V. Godesberger Allee 149 53175 Bonn Germany ABOUT THIS STUDY This publication is part of the project»Informal Employment, Social Security and Political Trust in sub-Saharan Africa«. You can find further publications of this project at https://www.fes.de/referat-afrika/publikationen. Email: info@fes.de Issuing Division: International Cooperation Division Responsibility for content and editing: Henrik Maihack| Head of Africa Department Phone:+49-30-269-35-7440| Fax:+49-30-269-35-9217 www.fes.de/en/africa-department Contact/Orders: Janine.Kaliga@fes.de Typesetting: Ludger Stallmeister The views expressed in this publication are not necessarily those of the Friedrich-Ebert-Stiftung. Commercial use of media published by the Friedrich-Ebert-Stiftung(FES) is not permitted without the written consent of the FES. Publications by the Friedrich-Ebert-Stiftung may not be used for electioneering purposes. ISBN: 978-3-98628-257-8 © 2022 www.fes.de/bibliothek/fes-publikationen

POVERTY AND IN­EQUALITY DURING THE PANDEMIC The diverging experiences of Kenya and Senegal • The governments of Kenya and Senegal reacted quickly to contain the spread of the Covid-19 virus. They applied different lockdown policies, however, and different social relief measures to assist vulnerable households in withstanding the health crisis. Kenya clamped down harder on economic activities than did Senegal, whereas Senegal provided a more generous state aid package to poor households than Kenya. The report discusses the impact of these different policy approaches on poverty and inequality in the two countries’ informal economies. • Our data show significant discrepancies with regard to poverty and inequality. In Kenya, poverty in the informal labour force increased substantially between October 2018 and December 2020, whereas in Senegal, poverty and in­ equality declined between June 2019 and April 2022. Kenya’s experience may be considered a»normal response« to an economic recession but the improvement in Senegal’s social indicators makes that country stand out from the general trend of declining income during the pandemic. We cannot exclude other factors, but conclude that the sheer magnitude of the social relief package in Senegal was instrumental in improving poverty and inequality indicators. • The data are taken from country-wide representative empirical studies conducted as part of the FES-IDOS-ILO project on»Informal Employment, Social Security and Political Trust in Sub-Saharan Africa«. The project ran from 2018 to 2022 and includes opinion surveys of views on access to health services, and on political trust and reasons for joining groups, including levels of interest in trade union membership. The general report on the project was recently published under the title A Majority Working in the Shadows. More studies on income, poverty and inequality are being prepared and will soon be published. Further information on the topic can be found here: www.fes.de/en/africa-department