REPORTS & PUBLICATIONS

The coronavirus pandemic is the largest social and economic shock in our lifetime. The rapid spread of this virus around the world and the economic devastation it has left in its wake is unlike anything we have seen before, at least not in our lifetimes. The local and international landscape is constantly morphing and changing in unpredictable ways making policy formulation and implementation as hard as it can possibly be. Policies that are helpful and sensible today may be harmful and illogical tomorrow. Every month seems to yield new information and consequences that were unforeseen even six weeks before. Indeed, “There are decades where nothing happens, and there are weeks where decades happen” (Lenin).

It is within this context that we have convened a national consortium of 30 social science researchers from five South African universities to conduct the National Income Dynamics Study (NIDS) Coronavirus Rapid Mobile Survey (CRAM) over the course of May 2020 – July 2021. Because decision-making is only as good as the data on which it is based, the NIDS-CRAM project exists to collect, analyze and disseminate data on a broadly representative sample of South African individuals, and to report on their employment and welfare in light of the COVID-19 pandemic.

As South Africa started its mass vaccination campaign on 17 May, many experts raised concerns about whether the country’s ambitious target of administering 60m shots in less than 10 months would be achievable. This paper examines whether vaccine acceptance is as an impediment to vaccine take-up and roll out based on the most recent wave 5 NIDS-CRAM data. Comparing these results to the wave 4 results, we assess the stability of vaccine beliefs over a period of two months. Almost half (47%) of those who ‘disagreed strongly or somewhat or did not know’ in February or March 2021 subsequently changed their minds over the following two months and either had been vaccinated or agreed to be vaccinated when asked again in April or May.

Within the context of increasing mental distress noted since the beginning of the pandemic, this study uses the 5Cs framework to analyse the role that depressive symptoms play in driving vaccine behaviour in South Africa. The study further explores pathways of this association through vaccine distrust, risk perception and efficacy. Multivariate regression analysis taking into account endogeneity concerns reveal that vaccine distrust is the most important predictor of vaccine hesitancy. This study has established that the indirect effects are highly relevant and need to considered closely while analysing the relationship between mental distress and vaccine behaviour. Lastly, the study also found significant feedback effect of mental distress with vaccine distrust as well as risk perception. Therefore, improved vaccine trust can lead to not just increased vaccine acceptance and reduced risk perception; but also, better mental health.

Since the onset of the COVID-19 pandemic, children have been put at greater risk of dropping out of school, lagging behind and losing learning, as well as food insecurity and emotional health deterioration. All these issues, as they arise in the South African context, are considered in this policy paper. Specifically, we make use of all five waves of the National Income Dynamics Study – Coronavirus Rapid Mobile Survey (NIDS-CRAM) data to estimate learner dropout/non-return to school, learning losses, changes in and the determinants of parent and caregiver concern and prevalence of depressed mood, child hunger, and access to school meals. These analyses build on previous work conducted on earlier waves of NIDS-CRAM that showed school feeding to lag behind pre-pandemic levels, the majority of parents and caregivers to be very worried about learners returning to school during the COVID-19 pandemic, and 50-75% of a year of lost learning amongst foundation and intermediary phase learners in resource constrained schools (Mohohlwane et al, 2020; Mohohlwane et al, 2021; Shepherd et al, 2021).

Globally, women have been disproportionately affected by the COVID-19 crisis. In line with this, research using earlier waves of the NIDS-CRAM survey found that women in South Africa were particularly hard hit. Relative to men, they were much more likely to lose their jobs during the initial strict lockdown phase, and their recovery was slower as the economy started to reopen (Casale and Posel 2020; Casale and Shepherd 2020; 2021a). Despite these uneven effects in the labour market, women were less likely than men to benefit from the COVID-specific government income support measures put in place to help cushion the blow to unemployed and furloughed workers. In addition to these gendered outcomes, there were also inequalities in the home. The time that women spend on childcare was found to be relatively more responsive to school closures and re-openings than the time men spend on childcare, with far more women than men citing childcare responsibilities as a constraint to their labour market activities (Casale and Shepherd 2020). In this final policy paper update, we track these uneven effects using the full five waves of the NIDS-CRAM survey, providing a comprehensive overview of how the first year of the pandemic has affected gender inequality in SA. We find that in March 2021, men’s employment and working hours were back to pre-COVID levels, while women’s employment and working hours remained below the February 2020 baseline figures. In addition to uneven effects in the labour market, inequalities in the time spent on childcare and in the income support for unemployed or furloughed workers persist. The evidence from the NIDS-CRAM survey over the last year therefore suggests that the pandemic has resulted in a rise in gender inequality in South Africa, undoing some of the gains of the previous two decades.

In this paper we use data from waves 1 to 5 of NIDS-CRAM to investigate employment and labour force participation outcomes in South Africa for four different age groups: youth (aged 18 to 24), prime-age adults (aged 25-39), middle-age adults (aged 40-54) and older adults (aged 55-64). We contrast outcomes just before and just after the advent of the COVID-19 pandemic and lockdown (February and April 2020) with outcomes one year later (March 2021), and study transitions between the periods. We find that although the employment-to-population ratio in March 2021 was near identical to the ratio in February 2020 (56.6% versus 56.4%), there had been a lot of churn between the two periods, with 23% of the February employed no longer employed a year later, and 30% of those without employment finding employment by March 2021. Youth experienced the largest employment-to-population ratio increase between February 2020 and March 2021 (32.5% to 35%), whereas older adults experienced the largest decrease from 45% to 41%. Between April 2020 (the peak of lockdown restrictions) and March 2021 there were decreases in the share of discouraged work seekers across age groups, and simultaneous increases in labour force participation for all groups except older adults. Tentative findings based on industry data revealed that the industries with the most notable increases in employment totals were the community, social and personal services and wholesale and retail trade industries. These industries also displayed high or increasing youth intensity of employment, potentially explaining part of the positive youth outcomes observed in NIDS-CRAM.

This paper conducts an analysis of labour market dynamics in South Africa using all five waves of NIDS-CRAM, as well as five waves of Statistics South Africa’s Quarterly Labour Force Survey (QLFS) 2020-Q1 to 2021-Q1. We find that much of the differences in estimates of labour force states including employment, unemployment and not economically active, are due to different reference periods between the two surveys. NIDS-CRAM asks about labour force activity during a reference month, whereas the QLFS utilises a reference week at the time of interview. This leads to higher estimates of employment in NIDS-CRAM compared to the QLFS for both a pre-pandemic baseline and over the entire period investigated (February 2020 to March 2021). The reference month allows NIDS-CRAM to very accurately identify labour force states during particular Lockdown levels. The QLFS, on the other hand, is not able to identify labour force states in Lockdown levels because no Lockdown level has been in effect for an entire quarter of the year. Instead, the QLFS is better suited to estimating the cross-sectional labour force states over each quarter and annually. This implies that the instruments are complementary, not contradictory, improving our understanding of the impact of Covid-19 on the labour force when taken together. Results using NIDS-CRAM show that the labour market in South Africa is very responsive to Lockdown regulations, fluctuating dramatically when bans or restrictions on trading hours and curfews are put in place, while recovering relatively quickly when they are relaxed. The QLFS shows that employment rates in South Africa have still not recovered to their pre-pandemic levels, remaining at approximately 10 percent below pre-pandemic levels for the population as a whole. This points to the fact that lockdown restrictions can have tremendously negative economic impacts that are difficult to recover from over the course of a year, and that wherever possible, all efforts should be made to geographically differentiate Lockdown levels to minimize their national impact, while maximizing the potential for localized economic recoveries.

Job retention policy – which aims to preserve jobs at firms experiencing a temporary reduction in activity by alleviating labour costs and supporting the incomes of workers – has served as one of the main tools used by governments across the world to mitigate job losses in response to the COVID-19 pandemic, particularly in the form of wage subsidies. In South Africa, the government’s Temporary Employer-Employee Relief Scheme (TERS), introduced in April 2020, served as a core component of the country’s economic policy response benefiting millions of vulnerable workers. In this paper, we provide a detailed, quantitative, descriptive analysis of TERS receipt over time by making use of representative, longitudinal survey data collected over the course of 2020 and 2021: all fives waves of the National Income Dynamics Study: Coronavirus Rapid Mobile Survey (NIDS-CRAM).

COVID-19 has exposed the fault lines that distinguish the lives and livelihoods of different social and racial groups in South Africa. Analysis of the pandemic’s stark spatial divides has been neglected in comparison. Yet the crisis has unfolded in different ways across the country, and the inequalities in income and well-being between social groups have tended to get accentuated by the differences between places. This paper explores the spatial impact of the pandemic on people and places over the last 12 months. The socio-economic consequences are explored both in terms of different parts of the country (metros, smaller cities/towns and rural areas) and between neighbourhood types within urban areas (suburbs, townships, shack dwellers and peri-urban areas). A positive finding is that employment levels and food insecurity have improved for most communities over the past year, following the initial shock caused by the strict lockdown. The COVID-19 SRD grant was also well targeted towards poorer communities, although it has since been terminated. A sobering message is that many people and places throughout the country remain vulnerable to hardship and misery, and that the poorest communities have suffered most during the crisis. Levels of hunger are still very high, even 12 months since the start of the pandemic. Urban shack dwellers stand out as being at particular risk of harm and have been most severely impacted all-round.

The coronavirus pandemic has changed the nature of work, with physical distancing regulations aimed at preventing infections necessitating work-from-home arrangements. According to previous studies, many individuals have expressed a preference for working from home due to fear of contracting the virus at work. However, not all work can be performed from home. Moreover, the literature indicates that jobs that are amenable to be performed from home generally pay more, while the ability to work from home will likely increase income inequality. Therefore, we ascertained the magnitude of the socioeconomic inequalities in ability to work from home among South African employees during the pandemic. We used data from the last four waves of the National Income Dynamics Study-Coronavirus Rapid Mobile Survey, a nationally representative longitudinal telephonic survey of South African adults conducted between May 2020 and May 2021. Using years of education as the ranking variable and calculating concentration indices, we found that the ability to work from home was pro-rich (i.e. concentrated more on the better off) in all the study periods. The results were robust to the use of different ranking variables like per capita household income and monthly wage as well as varying the age cut-off. There was no gender difference in the inequalities. Casual employment, urban residence, being married/cohabiting, age, and household size dampened the degree to which ability to work from home favoured those in higher socioeconomic classes. Conversely, being non-African, living in a house/flat, and having more education enhanced the pro-richness of ability to work from home. This study highlights the significant inequalities associated with ability to work from home, a likely important determinant of labour market attachment and economic outcomes in the pandemic and post-pandemic periods. Interventions targeted at inequality-enhancing factors like race, housing and education may be important in lowering these inequalities.

It is expected that the Coronavirus pandemic will exacerbate inequality in wellbeing compared to the pre-pandemic situation. However, there are theories (e.g., the Conservation of Resource (COR) theory) that acknowledge situation-specific lower wellbeing for individuals who typically have more resources. The argument is that perception of loss might occur differently across the socioeconomic spectrum such that individuals with higher socioeconomic status experience more loss. Therefore, given the pandemic situation, it is possible that indicators of poor wellbeing (e.g., depression) becoming less concentrated among the poor, contrary to expectation. Given the above, we examine income-related inequality in self-assessed health and depressive symptoms in South Africa. This is done using both pre-pandemic data (i.e. National Income Dynamic Study (NIDS)) and data collected during the pandemic (National Income Dynamic Study-Coronavirus Rapid Mobile (NIDS-CRAM)). Consistent with expectation, we find that poor self-assessed health has not only remain concentrated amongst the poor, but this concentration has increased compared to the pre-pandemic situation. However, contrary to expectation, depressive symptoms have become less concentrated amongst the poor compared to the pre-pandemic period. We note that while there may be an alternative explanation for this change in trend, it may also be due to situation-specific lower wellbeing for individuals who typically have more resources. We argue that this has implication for tracking population health in a crisis.

The COVID-19 pandemic has affected people’s mental health. There are the immediate effects that include fear, anxiety, loneliness, and uncertainty about the future. But there are also secondary impacts flowing from national responses to the pandemic such as lockdowns, including school closures, halting of school feeding, as well as more distal economic impacts such as global trade slowdowns and massive increases in unemployment. We have evidence, from previous waves of NIDS-CRAM data, that rates of depressive symptoms have been consistently higher than before the pandemic. The risk of screening positive for depressive symptoms had increased from 24% to 29% between waves 2 and 3 of NIDS-CRAM. In 2017, before the pandemic, this risk was 21%. The risk of screening positive for “severe” depressive symptoms increased from 5.2% to 7.1% between waves 2 and 3 of NIDS-CRAM. Our findings from wave 5, collected between 6 April and 11 May 2021, indicate that the risk of screening positive for depressive symptoms has remained stable between wave 3 and 5 at around 29%. The risk of screening positive for “severe” depressive symptoms, at wave 5, was 4.9%. While depressive symptoms have been and continue to be prevalent in the context of the pandemic, it is not clear what is accounting for this. Mental health is impacted by an array of factors. Many of these are internal to the individual – for instance genetics, disposition, and developmental history. But a significant amount of the variation is explained by environmental factors.

Globally, large international organisations have been tracking school closures and learning losses due to COVID-19 induced lockdowns. Yet few studies have been conducted to assess how childcare markets or access to early childhood development (ECD) opportunities have been impacted by COVID-19, especially in low-to-middle-income countries. Using a longitudinal telephonic survey of adults, known as the National Income Dynamics Study – Coronavirus Rapid Mobile Survey (NIDS-CRAM), this paper provides empirical evidence from South Africa on ECD attendance trends since the onset of the pandemic in early 2020 to May 2021. Pre-pandemic, about 39% of respondents living with children aged 0-6 indicated that at least one child had attended an ECD programme in February 2020. After a series of troughs in reported ECD attendance to as low as 7% in July/August 2020 and early February 2021, interspersed with temporary partial recoveries, a very significant recovery is seen by May 2021. In the fifth NIDS-CRAM data collection period between 6 April and 11 May 2021, ECD attendance had edged much closer towards pre-pandemic attendance levels. Of respondents living with children aged 0-6 at the time of the April/May 2021 interview, 36% reported that at least one child aged 0-6 had attended an ECD programme in the past 7 days. Analysis of the NIDS-CRAM data series also reveals that parents’ or caregivers’ ability to afford ECD fees and pay these fees will continue to be the key determining factor for sustaining and improving access to ECD programmes in the absence of wider and deeper public financing for the sector.

The results of the fifth and final wave of the National Income Dynamics Study – Coronavirus Rapid Mobile (NIDS-CRAM) survey reported in this paper provide an update and analysis of the trend in household food insecurity and hunger during the past year in South Africa. In a previous paper in May 2021, we noted that “indicators of hunger and a lack of money to buy food, or what we will refer to as ‘food insecurity’ for brevity, have remained stubbornly high, and do not appear to have changed substantially since June 2020.” Unfortunately, this statement remains true in wave 5.

1. Nic Spaull (SU)  – Overview
2.  Tim Köhler (UCT) – Vaccines
3. Charles Wiysonge (SAMRC) – Respondent
4. David Harrison (DGMT) – Respondent
5. Nompumelelo Mohohlwane (DBE) – Schooling
6. Servaas van der Berg (SU) – Hunger
7. Reza Daniels (UCT) – Employment
8. Open Q&A – Authors + Audience

The coronavirus pandemic is the largest social and economic shock in our lifetime. The rapid spread of this virus around the world and the economic devastation it has left in its wake is unlike anything we have seen before, at least not in our lifetimes. The local and international landscape is constantly morphing and changing in unpredictable ways making policy formulation and implementation as hard as it can possibly be. Policies that are helpful and sensible today may be harmful and illogical tomorrow. Every month seems to yield new information and consequences that were unforeseen even six weeks before. Indeed, “There are decades where nothing happens, and there are weeks where decades happen” (Lenin).

It is within this context that we have convened a national consortium of 30 social science researchers from five South African universities to conduct the National Income Dynamics Study (NIDS) Coronavirus Rapid Mobile Survey (CRAM) over the course of May 2020 – June 2021. Because decision-making is only as good as the data on which it is based, the NIDS-CRAM project exists to collect, analyze and disseminate data on a broadly representative sample of South African individuals, and to report on their employment and welfare in light of the COVID-19 pandemic.

The arrival of vaccines signalled a new era in South Africa’s fight against the COVID-19 pandemic, with the focus shifting from containment to eliminating the pandemic. However, given uncertainty about vaccine delivery dates, logistical concerns, and vaccine hesitancy, it looks increasingly unlikely that the virus will be under control by the end of 2021. In this study, we focus on vaccine hesitancy, analysing the recently released NIDS-CRAM wave 4 data that asked respondents about their willingness to be vaccinated, if vaccinations were available. We track reasons for hesitancy as well as correlates with hesitancy. We find that 71% of South African adults say they would get vaccinated if a COVID-19 vaccine was available. Among the 29% of respondents who were vaccine hesitant, the three leading reasons for their hesitancy were that they were worried about the side effects (31%), did not believe it was effective (21%) or did not trust vaccines in general (18%). Encouragingly, those with chronic conditions and the elderly – who are at a higher risk of becoming severely ill or dying if infected with the virus – were more willing than the general population to accept a vaccine. Respondents who said that they trusted social media as a source of information and those under 25 had considerably and significantly higher levels of vaccine hesitancy.

Widespread vaccine acceptance is key to achieving herd immunity through vaccination against COVID19, especially because the available vaccines do not have 100% efficacy. The study finds that vaccine acceptance is lower than that of non-pharmaceutical intervention like wearing facemask in public. Based on the vaccine intentions of the respondents of National Income Dynamics Study- Coronavirus Rapid Mobile (NIDS-CRAM) survey, only 55% are fully accepting of the vaccine, while a further 16% are moderately accepting of vaccines. Together, vaccine acceptance is estimated at 70.8% (with 95% confidence interval of 70%-72%) and vaccine hesitancy against COVID19 is estimated at 29.2% (with 95% confidence interval of 28%-30%) amongst the adult South African population. The reasons cited for vaccine hesitancy are primarily due to concerns regarding the side effects and efficacy of vaccine.

Using the Health Belief model and the Extended Parallel Process model, the study has identified
perceived risk of infection with the mediating role of efficacy as a key predictor of vaccine intention.
Apart from vulnerability to infection, those perceiving higher severity of risk (elderly and those with chronic illness) have lower hesitancy compared to their younger and healthier counterparts. Higher awareness of COVID19 related information and higher household income are correlated with lower
vaccine hesitancy.

There are other significant differences across socio-economic and demographic variables in vaccine hesitancy. The non-black African population group has significantly high vaccine hesitancy compared to black Africans. Males on average have lower hesitancy compared to females, and unmarried/ unpartnered individuals are found to have higher hesitancy.

From a communication perspective, it is imperative to continue risk messaging, hand in hand with clearer information on the efficacy of the vaccines. The risk messaging needs to be targeted at younger age and low-income groups who perceive the lowest risk. Those who are less informed on COVID19 are also found to be those who perceive low-risk, therefore this reinforces the need to have in place a communication plan to educate and create awareness on the pandemic. There is an all around need to enhance vaccine efficacy related messaging, particularly targeting non-blacks and religious individuals. Concerns regarding the side-effects of vaccines need to be addressed with more transparent scientific information together with the associated risk-benefit analysis.

This policy paper makes use of four waves of the National Income Dynamics Study – Coronavirus Rapid Mobile Survey (NIDS-CRAM) panel data to estimate changes in and the determinants of adult concern, child hunger and access to school meals amongst respondents residing in households with learners in the compulsory schooling system. The transition between Waves 2 and 3 of NIDS-CRAM from lockdown Level 3 to a less restrictive Level 1, as well as the transition from Waves 3 and 4 from Level 1 to a more restrictive adjusted Level 3, are expected to be linked to significant changes in food security and child hunger, and, potentially, adult concern about learner return given significant shifts in the number of infections and deaths.

Previous education policy papers produced using Waves 1 through 3 of the NIDS-CRAM data provided discussions on learner attendance as a result of phased school reopening, and how this might be used as a proxy for the effects of the pandemic on schooling and learning. In this policy paper, we build on this using panel data from the Early Grade Reading Study (EGRS) II to estimate learning losses experienced by grade 4 learners in 2020, as well as contextual factors gathered through telephonic surveys with teachers that may help explain how teachers and learners responded.

Finally, this paper presents a new analysis of deaths amongst teachers using government administrative data comparing the trends of 2020 (and the known peaks of COVID-19 infections) to those of the previous year. This new analysis presents credible information to be considered in the ongoing debate about the risks faced by teachers by going to school.

At the height of the COVID-19 pandemic, many governments across the world issued a directive for workers to stay at home and work from there where functionally possible. While this directive has become less stringent since its first issuance, it is still in force as governments strive to reduce disease transmission and save lives. Nonetheless, that not all jobs can be performed from home has prompted research into who cannot work from home as they face a higher risk of exposure to the virus relative to those who can work from home. This paper explores the correlates (demographic and job characteristics) of workers who are unable to work from home. The analysis utilises logit models and the NIDS-CRAM wave 4 dataset. Results show that men are more unable to work from home than women, while non-white race groups have lower probabilities of working from home relative to whites. Individuals who dwell in shacks/informal housing are less likely to work from home than those whodwell in a house/flat. In addition, workers in urban areas have lower probabilities of working from home than those in rural areas. Employees in formal jobs are less likely to work from home than their counterparts in informal jobs. For occupations, workers in elementary occupations and plant and machine operations are more likely to be unable to work from home than managers and professionals. Further, workers in the mining sector, private households, the manufacturing sector and community, social and personal services are more likely to be unable to work from home compared to workers in financial intermediation. These results call for interventions that minimise the risk of the vulnerable workers’ exposure to COVID-19 in the workplace.

As much global research now shows, women have been disproportionately affected by the COVID-19 crisis. In South Africa, research using the first three waves of NIDS-CRAM found that women’s employment was particularly hard hit by the initial lockdown (Casale and Posel 2020), and women also experienced a slower recovery than men as the economy started reopening (Casale and Shepherd 2021). In this policy paper, we exploit the panel nature of the NIDS-CRAM data to analyse gendered labour market dynamics over the course of the pandemic in more detail. In particular, we compare the labour market transitions between the various waves of the panel for women and men. Further, in a multivariate context, we analyse the individual- and household-level characteristics that help predict which women and men lost or gained work over the course of the pandemic, and whether these factors can help explain the ‘female penalty’.

This document updates research by Espi et al. (2021) in the light of new data about employment in January 2021 from NIDS-CRAM wave 4. First, January employment outcomes are compared to prior outcomes for three age groups: youth (18-24), prime-age adults (25-40) and middle-age adults (41-55). Then the focus turns to employment dynamics between October 2020 and January 2021.

The employment to population ratio declined similarly by a few percentage points for all age groups (and age and education subgroups) in the period, with the percentage of all working age (18-64) adults employed dropping from 55% to 52%. However, in the period there was substantial churning underlying this relatively small percentage point drop in employment, far in excess of estimates of churning from pre-COVID years. Of working age adults who were non-employed in October, 18% had found work by January, while 19% of the October employed were without work in January. Rates of job finding among the non-employed were similar across age groups, while job loss was strongly and negatively correlated with age. These results indicate a worsening in employment and transition outcomes in response to the second wave of COVID-19 infections and the associated lockdown, but that the declines in employment were small relative to the effects of the initial COVID-19 lockdown.

In the light of documented discrepancies between NIDS-CRAM and other surveys’ estimates, the authors remain cognisant that these findings may be descriptive of a particular sample or subset of the South African population.

Given that South Africa experienced significant food insecurity even before the COVID-19 pandemic, it is not surprising that the pandemic and associated control measures would result in even greater food insecurity in the country. This paper seeks to provide additional evidence on the well-known relationship between food insecurity and health. Data came from the National Income Dynamics Study-Coronavirus Rapid Mobile Survey, a four-wave longitudinal survey of adult South Africans. Health was captured using a self-reported indicator of general health status, while food insecurity was measured using information on household hunger, the frequency of hunger, and households running out of money to buy food. We performed descriptive and econometric analyses. Food insecurity has remained stubbornly high even in the face of greater re-opening of the economy, with 37-47% of the population indicating that their households ran out of money to buy food and 16-22% indicating that their households experienced hunger over the May/June 2020 to February/March 2021 period. While hunger generally declined somewhat over this period, the poorest quintile suffered more hunger than before. Moreover, among those whose households experienced hunger, between a quarter and a third struggled with hunger almost everyday or daily. Being in a household that experienced hunger was associated with a 5.5-7.3 percentage point higher probability of worse health compared to not experiencing hunger in May/June 2020-February/March 2021. Compared to being unaffected by hunger, being hungry everyday was associated with a 17-percentage point higher probability of worse health in May/June 2020. These results show the enormity of the hunger problem in South Africa and its adverse effects on welfare. In the face of economic uncertainty and the removal of COVID-19 palliatives like the grant top-ups, policymakers must think of ways to protect the vulnerable from food insecurity.

In response to the COVID-19 pandemic, numerous governments across the world recommended some preventive behaviour for their citizens, e.g. social distancing and wearing masks. The effectiveness of these recommendations depends on voluntary compliance by the public. Existing evidence on past epidemics shows that social trust, political trust and trust in sources of information underlie compliance with regulations. In this paper, we explore the role of social trust and trust in sources of information on compliance to COVID-19 preventative measures in South Africa. The analysis utilises logit models, data from waves of NIDS-CRAM and the 5th wave of NIDS survey. We find that very high levels of social trust are required to foster a positive association between social trust and the probability of compliance with COVID-19 regulations. Further, trust in local and international news predicts the probability of adherence to a higher range of preventative measures; wearing a mask, staying at home and washing/sanitising hands regularly. This is followed by trust in government information which is associated with a higher probability of wearing a mask and staying at home. Trust in social media and health workers are associated with a higher probability of wearing a mask and staying at home, respectively. Based on these findings, we suggest a need for measures to increase levels of social trust and trust in information sources in South Africa which assist the government to avert pandemics and natural disasters.

Risk perception is an important variable in the context of the COVID-19 pandemic, this is because the risk communication literature and several theories related to it argues that this variable encourages the adoption of preventative measures. However, heightened risk perception can also have negative consequences for general wellbeing (for example in terms of psychological distress). In this paper, we consider the relationship between risk perception and two health measures (self-assessed health and depressive symptoms) in the context of the COVID-19 pandemic. Furthermore, adopting the Extended Parallel Process Model, we consider the relationship between risk perception and health when risk perception interacts with self-efficacy.

Our results show that across the waves of the NIDS-CRAM data, risk perception is correlated with poor health under both measures of health. In our analysis we are unable to ascertain whether this is as a result of reverse causality between risk perception and health. Furthermore, we found limited evidence that when risk perception is not balanced with self-efficacy the correlation between poor health and risk perception becomes stronger. These results underscore the importance of giving careful consideration to the crafting of risk communication.

This policy brief updates trends on ECD attendance in South Africa with wave 4 of the National Income Dynamics – Coronavirus Rapid Mobile Survey (NIDS-CRAM).3 In addition to updating trends on ECD attendance, we explore the reasons for non-attendance of children at ECD programmes and provide self-reported responses on who is looking after children that are currently not attending ECD programmes.

Wave 4 of the National Income Dynamics Study – Coronavirus Rapid Mobile (NIDS-CRAM) survey provides an update to the widely reported and sobering findings from the first three waves of the study. Indicators of hunger and a lack of money to buy food, or what we will refer to as ‘food insecurity’ for brevity, have remained stubbornly high, and do not appear to have changed substantially since June 2020. We conclude that given the protracted nature of the pandemic, slow recovery, and its continuing social and economic impact, the government needs to place hunger and food security at the center of its Covid-19 response.

• Introduction and Overview – Nic Spaull (SU)
• Employment findings – Reza Daniels (UCT)
• Respondent: Ismail Momoniat (NT)
• Hunger findings – Servaas van der Berg (SU)
• Respondent:Ruth Hall (UWC)
• Education – Nompumelelo Mohohlwane (DBE)
• ECD – Gabrielle Wills (SU)
• Respondent: (Ed+ECD): Linda Zuze (SU)
• Health – Brendan Maughan Brown (UCT)
• Overall respondent: Rudi Dicks (Presidency)

The coronavirus pandemic is the largest social and economic shock in our lifetime. The rapid spread of this virus around the world and the economic devastation it has left in its wake is unlike anything we have seen before, at least not in our lifetimes. The local and international landscape is constantly morphing and changing in unpredictable ways making policy formulation and implementation as hard as it can possibly be. Policies that are helpful and sensible today may be harmful and illogical tomorrow. Every month seems to yield new information and consequences that were unforeseen even six weeks before. Indeed, “There are decades where nothing happens, and there are weeks where decades happen” (Lenin).

It is within this context that we have convened a national consortium of 30 social science researchers from five South African universities to conduct the National Income Dynamics Study (NIDS) Coronavirus Rapid Mobile Survey (CRAM) over the course of May 2020 – June 2021. Because decision-making is only as good as the data on which it is based, the NIDS-CRAM project exists to collect, analyze and disseminate data on a broadly representative sample of South African individuals, and to report on their employment and welfare in light of the COVID-19 pandemic.

We use Wave 3 of NIDS-CRAM data to provide an update to previous estimates (Jain, Bassier, et al., 2020; Jain, Budlender, et al., 2020) of COVID-19-related employment impacts in South Africa. We find evidence of a substantial job market recovery in October. We previously found a 40% drop in April compared to February active employment, partially recovering to a 20% drop by June. We now find that by October the proportion of actively employed adults appears close to its February pre-pandemic level. The difference between February and October active employment (a 0.15% decline) is not statistically significant, but due to precision limitations the data still allow for a non-negligible drop in employment. However, we can confidently rule out a February to October decline in active employment greater than 4.7%. We perform a number of robustness checks which affirm the plausibility of this surprisingly large recovery. We additionally present results on earnings changes and social protection coverage.

Globally, women appear to have been disproportionately affected by the Covid-19 crisis. In South Africa, research using Waves 1 and 2 of the National Income Dynamics Study-Coronavirus Rapid Mobile Survey (NIDS-CRAM) found that women were particularly hard hit by the initial lockdown phases and school closures, both in terms of labour market outcomes and childcare responsibilities (Casale and Posel 2020; Casale and Shepherd 2020). In this policy paper update, we continue to track gendered outcomes in the labour market and in unpaid care work in the home using the recently-released third wave of the NIDS-CRAM data. With the move from Level 3 lockdown in June to the much less restrictive Level 1 lockdown in October, we find a substantial recovery in employment of roughly 2 million net jobs among adults aged 18 and over, shared almost equally between men and women. However, given how much larger the fall in women’s employment was as a result of the initial shock to the labour market, it appears women still remained behind men in terms of reaching their pre-Covid employment levels in October. Taking the data at face value, compared to February, women’s employment was still down by just under 700 000 jobs (or 8%) in October, while men’s employment was down by 200 000 jobs (or 2%) in October. Similarly, employed women were working two fewer hours per week on average in October compared to February, while men’s working hours were back up to pre-Covid levels. However, given the small sample sizes (and therefore large margins of error), it would be prudent not to place too much emphasis on the exact numbers. At most one could say that the recovery between June and October was substantial for both men and women, although it appears to have been slower for women. As expected given the reopening of schools, both women and men reported spending significantly fewer hours on childcare in October compared to June. Women were able to cut back their hours by a greater amount than men however, highlighting how women have borne the brunt of school closures thus far. Nonetheless, among those living with children, women continue to spend more hours on childcare than men, with negative implications for their labour market prospects.

Past research suggests that spells of unemployment can cause long-term labour market scarring, and that economic shocks can lead to severe, more persistent unemployment effects for youth in particular. We investigate heterogeneity in employment outcomes between February and October 2020 (as captured by NIDS-CRAM waves 1 to 3) for three age groups: youth (18-24), prime-age adults (25-40) and older adults (41-55). Thereafter, we group adults (within a balanced panel) on the basis of their employment history between 2012 and 2017 (NIDS waves 3 to 5) to study the correlation between 2020 employment outcomes and this employment history. We find evidence for a substantial recovery in employment rates between June and October, leading to October employment rates that were similar to February levels for youth and older adults. Employment recovery was positively correlated with educational attainment among prime-age adults and youth. On the other hand, April job loss was more likely to be persistent for youth relative to older groups. Employment history correlated strongly with 2020 employment outcomes: individuals with a more extensive history of employment were more likely to remain stably employed, or, among the non-employed, to find work. Importantly, the recovery of employment between April and October was driven to a substantial degree by job finding among people who were non-employed in February, and not just a ‘bounce back’ of April job losers. Together, these findings provide valuable insight into the adjustments, along dimensions of age and employment history, that have occurred in the South African labour market in response to the COVID-19 pandemic and lockdown.

The first wave of the National Income Dynamics Study – Coronavirus Rapid Mobile Survey (NIDS-CRAM) provided strong evidence of drastic increases in household and child hunger during the initial period of the Coronavirus pandemic, as well as an increase in household food insecurity. Much of the improvement since 2000 in household and child hunger that was due to the expansion of the Child Support Grant had been almost entirely reversed by the hard lockdown and the coronavirus pandemic. The second wave of NIDS-CRAM showed improvement, although hunger and food insecurity remained disturbingly high. Wave 3, surveyed in November and December, provides an update to the widely reported and sobering findings from the first two waves. Indicators of hunger and food insecurity have again worsened after the improvement in Wave 2. One possible cause is the phasing out of the top-ups to some of the social grants since November. The magnitude of the problem of food insecurity problem is still such that the need for social relief efforts remains undiminished and that the reduction and phasing out of in some of the social grants will have severe consequences in terms of hunger.

Using data from the second and third waves of the National Income Dynamics Study – Coronavirus Rapid Mobile Survey (NIDS-CRAM), we examine changes in learner attendance, the worry levels of parents and the provision of meals to learners since March 2020 until December 2020. The paper also presents an analysis of the National Institute for Communicable Diseases data on hospitalisations and deaths over time by age group to assess the health risks to children associated with school attendance. The recommendations focus on a reconsideration of timetable rules and regulations to allow more children to attend daily, better communication to improve parent concern and the availability of more alternatives to ensure increased delivery of daily meals to learners.

This updated policy paper identifies Early Childhood Development (ECD) attendance trends up to the last quarter of 2020 using wave 3 of the National Income Dynamics Study- Coronavirus Rapid Mobile Survey (NIDS-CRAM). Compared with the start of the third quarter of 2020, a partial recovery in ECD attendance is identified in the last quarter of 2020. Among NIDS-CRAM adult respondents interviewed in November/December 2020 and living with children aged 0-6, 28% reported at least one child attending an ECD programme in the past 7 days. This is up from just 7% in July/August. Extrapolating these NIDS-CRAM adult responses to the level of the child suggests that about 31% of children aged 0-6 were enrolled in ECD programmes by November/December, compared to at most 15% in July/August. However, of wave 3 respondents (living with children aged 0-6) 39% reported at least one child attending an ECD programme in February 2020. Despite the notable recovery from quarter three to quarter four of 2020, ECD attendance had not recovered to pre-lockdown levels and remained lower compared to prior years. Continued support for the sector remains imperative.

The stringent lockdown of the economy in response to the Covid-19 global pandemic triggered a larger negative shock on labour market outcomes of informal when compared to formal workers in developing countries. This potentially generated a disproportionate burden of stress and anxiety which cause depressive symptoms, among informal workers. These unintended consequences of COVID-19 reinforced the vulnerability and stressful working conditions that characterised informal work pre-COVID 19. Therefore, this study examines differences in depressive symptoms between informal and formal workers in South Africa during the COVID-19 lockdown period. Our analysis focuses on the June to October 2020 period, which was characterised by the easing of lockdown regulations from level 3 to level 1. Using the NIDS-CRAM wave 2 and 3 survey data, we estimate a standard logit and a fixed effects (FE) logit model to account for worker’s unobserved heterogeneity. Results show that the likelihood of experiencing depressive symptoms increased across all workers between June and October. We find no statistically significant differences between informal and formal workers’ mental health over this period. Additional results show that workers living in urban areas and households suffering from hunger had a higher risk of experiencing depressive symptoms. Based on these results, we recommend for government strategies that curb the rise in depressive symptoms among workers.

Widespread adoption of preventive behaviours is critical for controlling the COVID-19 pandemic. Key preventive behaviours promoted in South Africa are mask-wearing, handwashing, physical distancing, staying at home, and avoiding large crowds. After the first wave of COVID-19 infections peaked in South Africa in July 2020 there were widespread concerns about ‘COVID-fatigue’ – a reduction in perceived COVID risk and less attention devoted to preventive behaviours – during the lull in the epidemic between the first and second waves. Our paper examines COVID-19 risk perceptions, preventive behaviour and COVID-fatigue during this epidemic lull. The analysis is based on data from the first three surveys conducted by the National Income Dynamics Study: Coronavirus Rapid Mobile Survey (NIDS-CRAM) – first survey: 7 May to 27 June 2020, second survey: 13 July to 13 August 2020, third survey: 2 November to 13 December 2020. Key findings and policy implications are summarized here.

The coronavirus disease 2019 (COVID-19) pandemic has devastated livelihoods and health systems globally. The pandemic resulted in the closure of businesses and schools and the remote provision of many services in order to mitigate the spread of the virus. In some instances, families stopped patronizing paid caregivers due to the fear of contracting the virus and financial concerns due to massive job losses. Even with the re-opening of schools, many families are reluctant to send their children back to school on an in-person basis. One key consequence of these disruptions is a significant increase in the need for childcare services, including assistance with schoolwork. Such a substantial increase in time requirements for childcare domestically has potential mental health consequences. We therefore ascertained the relationship between childcare and mental health in South Africa. Data came from the second and third waves of the National Income Dynamics Study-Coronavirus Rapid Mobile survey. The outcome was a depression index obtained from the two-item Patient Health Questionnaire while the main covariate was the average number of hours spent in taking care of children per weekday. Employing the ordered logit model, we found a positive relationship between spending more hours on childcare and worse mental health for caregivers, with the relationship generally stronger for men. We suspect that our results are lower bound estimates of the impact of time spent on childcare on mental health in South Africa. Furthermore, the childcare-mental health relationship was significantly mediated by childcare responsibilities preventing/mitigating the ability of caregivers to work as well as preventing caregivers from searching for jobs. These findings highlight the wider health implications of COVID-19 and the need to carefully consider potential policy responses aimed at containing the virus. We advocate a multi-stakeholder approach to mitigating the mental health impact of COVID-19 by encouraging conversations and input from government, school authorities, employers and parents/guardians.

The adverse effect of the COVID-19 pandemic (and associated lockdown) on mental health, caused by increased financial and health-related anxiety, is now widely acknowledged. There is however less clarity regarding the implications of the easing of lockdown restrictions on mental health. The easing of lockdown restrictions is expected to reduce the financial stressors, provided that employment and income respond positively to the easing of lockdown. On the other hand, the increase in economic activities may elevate the subjective risk of contracting the virus thereby exacerbating health-related concerns. Using waves 2 and 3 of the NIDS-CRAM data, this paper examines the changes in the distribution and risk of screening positive for depressive symptoms in South Africa as the COVID-19 lockdown restrictions are relaxed from level 3 to 1. We also decompose the change in the proportion of respondents that screen positive for depressive symptoms between the two periods using Oaxaca-Blinder decomposition. This method decomposes the difference in proportion into components that are attributable to differences in characteristics (composition effect) and the difference in coefficients (return effect). Our analysis shows that, despite the easing of lockdown restrictions, both the distribution and the risk of screening positive for depression has increased in wave 3, relative to wave 2. The decomposition analysis shows that an overwhelming part of this change is explained through returns to covariates (or behavioural response). Furthermore, detailed decomposition shows that the returns to and changes in the composition of variables that are positively correlated with low social-economic status explains the increase in depressive symptoms, despite the easing of lockdown restrictions.

Tobacco sales were prohibited between 25 March and 17 August 2020 as part of the government’s response to the COVID-19 pandemic. A number of studies were conducted during the sales ban to investigate smokers’ responses to the sales ban. A drawback of these studies was that they were not nationally representative. The current study, as part of the NIDS-CRAM suite of surveys during the lockdown, aims to be broadly nationally representative, and as such provides an opportunity for a broadly representative ex-post investigation into smokers’ behaviour related to the national sales ban.

Latest Panel User Manual 

The coronavirus pandemic is the largest social and economic shock in our lifetime. The rapid spread of this virus around the world and the economic devastation it has left in its wake is unlike anything we have seen before, at least not in our lifetimes. The local and international landscape is constantly morphing and changing in unpredictable ways making policy formulation and implementation as hard as it can possibly be. Policies that are helpful and sensible today may be harmful and illogical tomorrow. Every month seems to yield new information and consequences that were unforeseen even six weeks before. Indeed, “There are decades where nothing happens, and there are weeks where decades happen” (Lenin).

It is within this context that we have convened a national consortium of 30 social science researchers from five South African universities to conduct the National Income Dynamics Study (NIDS) Coronavirus Rapid Mobile Survey (CRAM) over the course of May – December 2020. Because decision-making is only as good as the data on which it is based, the NIDS-CRAM project exists to collect, analyze and disseminate data on a broadly representative sample of South African individuals, and to report on their employment and welfare in light of the COVID-19 pandemic.

Many governments around the world have responded to the COVID-19 pandemic through a suite of social distancing measures (lockdown). However, this has had unintended consequences on labour markets, with informal workers disproportionately affected. South Africa’s lockdowns commenced with stringent directives (level 5) some of which were chronologically relaxed in subsequent lockdowns (level 4 to level 2). This paper explores changes in labour market outcomes (wages and hours worked) of formal/informal workers in South Africa, between April (lockdown level 5) and June (lockdown level 3). We employ a DID estimator and data from the first 2 waves of NIDS-CRAM data. Our results show a wage decrease of the same order for both formal and informal workers. Across subgroups, wages decreased more for men and urban workers who are informally employed, relative to their formally employed counterparts. We also find higher decreases in hours worked among men and women in informal work relative to those in formal work. This is also the case for informal workers in urban areas. Based on these results, we recommend for government interventions that set all workers on a recovery path. These can be targeted at enabling recovery of time spent at work which in turn boost productivity and wages. The government also needs to consider extending coverage of social relief to informal workers, as well as stabilise the broader economy for informal sector enterprises to thrive.

The data from Wave 1 of NIDS-CRAM showed that women were disproportionately affected by the COVID-19 crisis and the first month of the lockdown period in South Africa. Not only were they much more likely than men to lose their jobs between February and April or to work fewer hours compared to the pre-crisis period, they also took on a greater share of the additional childcare as a result of school closures and the suspension of all childcare services. In this policy paper, we use Wave 2 of NIDS-CRAM to explore how women and men have fared as the economy started to reopen and lockdown restrictions were relaxed. The data suggest that with the move from Level 5 lockdown in April to Level 3 lockdown in June, there was hardly any change in employment levels overall. However, women may have gained slightly relative to men. Nonetheless, given the very large job losses recorded among women between February and April (women lost 2 million jobs and men 1 million jobs), women still remained well behind men in reaching their pre-COVID employment levels in June. In contrast, men benefited more from the reopening of some school grades and childcare services in June. Compared to April, the hours men reported spending on childcare in June fell by more than the hours women reported spending on childcare. The data also show that much higher numbers of women than men found childcare to affect their ability to work, to work the same hours as before lockdown, and to search for work.

Future employment outcomes are believed to be strongly determined by an individual’s past employment experience (or lack thereof). We investigate the relationship between individuals’ labour market experience over the decade covered by NIDS (2008-2017) and their contemporary employment outcomes under the COVID-19 pandemic and lockdown as captured by NIDS-CRAM waves 1 and 2. Restricting to a balanced panel of prime-age (25 to 50) individuals who have employment status information across all periods of NIDS and NIDS-CRAM, we assign individuals to three groups on the basis of their employment history: the stable employed, the transient employed, and the persistent non-employed. We find that pre-lockdown employment status in February 2020 correlated strongly with this employment history. Nonetheless, a substantial proportion (45%) of the historically persistent non-employed were employed going into lockdown. This group was found to be young and thus more likely to have previously been engaged in education rather than labour activities. This highlights the need for care in interpreting employment history for individuals with different ages. Under lockdown, those with transient employed or persistent non-employed histories were more likely to have lost work and to be excluded from employment opportunities relative to the historically stable employed, with job loss consistently the highest for the persistent non-employed in both April and June 2020. Job gain also followed clearly differentiated lines, with a third of the historically stable employed who were without work in February finding employment by April, relative to only 8% of the persistent non-employed (although this discrepancy was less pronounced between April and June). Those with the least recorded employment were mostly younger, rural African women, while among the stable employed (who were all much more likely to be male), those who lost work in the lockdown were more likely to be African and have a rural background relative to those who retained work. These findings suggest that people’s long-term employment histories have been influential in determining their employment outcomes during COVID-19, and that understanding these histories can tell us more about the capacities and needs for intervention of those who have been affected by the lockdown.

The COVID-19 pandemic has had severe and potentially long-lasting impacts on the South African economy since the onset of the national lockdown in March 2020, however, these effects have not been equally felt by all. Research has shown that the effects of the COVID-19 pandemic have been disproportionately felt by vulnerable groups, and disemployment effects have been concentrated amongst women. This paper aims to investigate whether wage inequality has deepened among job retainers, particularly along gender lines. Using the first two waves of the NIDS-CRAM data, this paper presents results for the conditional and unconditional gender wage gap at the mean, showing that women earned approximately 29% less than men per hour in February 2020, which expanded to approximately 43% less in June 2020. Monthly figures are more severe, with the gender wage gap estimated at approximately 30% in February 2020 and 51.6% in June 2020. Given the heterogeneity of the gender wage gap over the wage distribution, we proceed to use Recentred Influence Function (RIF) regressions to estimate the gender wage gap along the distribution. Although there is evidence that the gender wage gap exists at almost all points of the distribution in both periods, it has not changed uniformly across the wage distribution: in fact, evidence suggests that the monthly gender wage gap has deepened significantly for individuals below the 40th percentile of the distribution, but not for those at the top. Hourly wage gaps show no evidence of deepening across the distribution, suggesting that changes in wage inequality are driven by decreased working hours amongst women relative to men, perhaps due to being unable to work effectively from home, or due to increased childcare burdens during lockdown. This finding is robust to sample selection corrections carried out via a DiNardo, Fortin and Lemieux (DLF) reweighting.

We use Wave 2 of NIDS-CRAM data to provide an update to our original estimates (Jain et al., 2020) of COVID-19-related employment and poverty impacts in South Africa. Compared to the most stringent phase of South Africa’s lockdown in April, we find evidence of a limited recovery in the labour market, a decrease in poverty, and an important role for the new Social Relief of Distress grant by June. While temporary unemployment almost returned to February levels, we find that active employment was still 20% lower in June than February, mostly due to job terminations that persisted into June.

This paper is an update to Ranchhod and Daniels (2020). It uses Wave 2 of NIDS-CRAM to conduct an evaluation of the state of the South African labour market in June 2020, when the country was in Stage 3 of the national lockdown. We compare these outcomes to corresponding outcomes from April 2020, when the country was in Stage 5 of lockdown, as well as against a pre-lockdown baseline of February 2020. In our first paper, we reported that a major new feature of the South African labour market associated with the lockdown was the introduction of a large number of “furloughed workers”. These are individuals who retain an employment relationship with their employers, but receive either less pay or no payment at all, due to the impossibility of working in certain occupations and industries that are prohibited from trading under lockdown regulations. The findings from Wave 2 suggest that while some recovery of the South African labour market has taken place between April and June, we are still a long way off from the pre-lockdown levels of employment that were observed in February. In addition, there is also an extremely high degree of churning in the labour market between April and June. These net changes remain inequality enhancing rather than inequality-reducing. These results are somewhat understandable given that Statistics South Africa estimated in September 2020 that GDP in the second quarter of 2020 had reduced significantly. This paper therefore makes a timely contribution to our understanding of the unfolding labour market impacts of the pandemic and lockdowns as we charter a path towards improving the resilience of the economy.

The first wave of the National Income Dynamics Study – Coronavirus Rapid Mobile Survey (NIDSCRAM) provided strong evidence that there had been a drastic increase in both adult and child hunger in the survey months, May and June 2020, and that almost half of all households had on occasion run out of money for food during April 2020. Comparisons with information from earlier surveys (the annual General Household Survey from 2002 to 2018) showed that much of the improvements since 2000 in adult and child hunger and food security due to the expansion of the Child Support Grant had been almost entirely reversed by the hard lockdown and the coronavirus pandemic. The second wave of NIDS-CRAM shows improvement in all three measures, but that hunger and food insecurity remain disturbingly high.

In Wave 1 of NIDS-CRAM, 22% of households reported that at least one adult had gone hungry in the past 7 days, and 15% of respondents reported that a child in their household had gone hungry in the week before they were interviewed in May or June. For the month of April, 47% of respondents reported that their household had run of money (the first month of the lockdown, before social relief measures were instituted). In comparison, the General Household Survey of 2018 reported only 25% running out of money for food in the past year, a far less strict criterion. Loss of the main income source during April, the first full month of the lockdown, strongly increased the likelihood of household hunger and of running out of money to buy food (Van der Berg, Zuze & Bridgman, 2020).

These sobering findings from Wave 1 of NIDS-CRAM have been widely reported. Wave 2 of NIDSCRAM now makes it possible to update hunger statistics for July/August (relative to May/June) and statistics on running out of money to buy food in June (relative to April). After discussing the new findings, this policy update concludes that the magnitude of this problem is still such that the need for social relief efforts remains undiminished.

In response to the adverse effects of the COVID-19 pandemic, the South African government expanded its system of social assistance by increasing the amounts of all existing social grants and introducing a special COVID-19 grant, both for six months. In particular, the COVID-19 grant has brought millions of previously unreached individuals now into the system. This paper uses new data from Wave 2 of the NIDS-CRAM to analyse the distribution of application for and receipt of the COVID-19 grant, examine how the Child Support Grant (CSG) ‘per grant’ top-up in May compares to the ‘per caregiver’ top-up in place from June 2020 onwards, and investigate the costs and welfare effects of several alternative policy options to consider once the expansion of the grants system comes to an end after October. We find that application for and receipt of the COVID-19 grant has been relatively pro-poor, and that conditional on applying, certain individuals are more likely than others to be successful in their application. Despite the grant’s progressivity, we show that the extent of under-coverage is however regressive. We show that the ‘per child’ CSG top-up is more pro-poor than the ‘per caregiver’ top-up, but only marginally. This is important considering that we estimate the cost of the chosen policy to be substantially cheaper than a six-month ‘per child’ top-up. Considering alternative post-October policies, we find that an extension of the current grant policy package may be preferable to a Basic Income Grant or special public works programme, however more analysis is required.

After schools closed on the 18th of March 2020 in response to the initial spread of COVID-19 in South Africa, a phased approach to reopening schools was adopted by the government. On the 8th of June, grade 7 and grade 12 pupils returned to school, with grade 6 and grade 11 learners returning on the 6th of July. Schools were then closed again on the 24th of July. This paper describes the partial return to school that occurred during June and July, drawing mainly on the second wave of the NIDSCRAM survey. To what extent was there alignment between the grades that were gazetted to return in June and July and actual school attendance rates by children across the grades? How worried were parents and guardians about sending their children back to school and how did this vary across society? What was the state of readiness of schools to operate under the new COVID-19 regulations? Did school reopening have any observable impact on the spread of COVID-19? Lastly, what can be said about the negative impacts of not being able to attend school on children, whether through nutritional or learning losses? Based on an exploration of these questions several recommendations are then made for the resumption of schooling over the months to come.

This paper analyses the impact of the pandemic on different parts of South Africa, bearing in mind their contrasting vulnerability and resilience. It compares the severity of the initial COVID-19 shock (February-April 2020) and the subsequent trajectory (April-June) of the metros, smaller cities/towns and rural areas. It also considers the different impacts within cities – between suburbs, townships, shack areas and peri-urban areas. A key question is whether COVID-19 has aggravated pre-existing spatial disparities? A second question is whether government social support has helped to mitigate these gaps in income and well-being? The paper reveals that the pandemic has magnified the existing economic and social divides (i) between cities and rural areas, and (ii) between suburbs and townships/informal settlements within cities. Government grants have helped to offset the large economic disparities between places, but the incidence of hunger is still much higher in informal settlements, townships and rural areas than in suburbs. There is a strong case for more targeted efforts to boost jobs and livelihoods in lagging urban and rural areas. Pre-existing conditions were bad enough, but now there is further ground to make up, and the withdrawal of temporary relief grants could be a serious setback for poor communities and groups reliant on cash transfers.

New evidence suggests that over four months after the closure of early childhood development (ECD) programmes on 18 March 2020, the ECD sector is operating at a fraction of its pre-lockdown levels. Of the 38% of respondents from the new NIDS-CRAM survey reporting that children aged 0-6 in their households had attended ECD programmes before the lockdown in March, only 12% indicated that children had returned to these programmes by mid-July, well after programmes were allowed to reopen. Using these findings, we estimate that less than 5% of children aged 0-6 were attending ECD programmes by mid-July to mid-August compared to 38% in 2018. The last time that ECD attendance rates were as low as this was pre-2000. At this point it is not yet clear what proportion of these declines are only temporary, or whether there will be a lasting impact on ECD enrolment in the country. This dramatic contraction in the ECD sector relates to prohibitive costs to reopening ‘safely’ imposed by the regulatory environment, coupled with shocks to the demand side for ECD programmes (both in terms of reduced household incomes and parent fears of children contracting COVID-19). When viewed from a broader socio-economic lens, the threat of ECD programme closures across the nation will have impacts beyond ECD operators to the lives of millions of children, millions of households and millions of adults who rely on these ECD services. A swift intervention by government is necessary to save this important sector and limit the ripple effect of programme closures on multiple layers of society.

South Africa’s lockdown in March and April has saved lives by containing the spread of COVID-19 but it has done so at a tremendous social and economic cost. To avoid a second surge and another lockdown, it is vital to prioritise adherence to non-pharmaceutical interventions (NPIs) as a first line of defence against containing COVID-19. NPIs can save lives without having to risk livelihoods. But to have an impact, it requires sufficiently high levels of public adherence. This paper considers adherence to NPIs and risk perceptions against the backdrop of an increase in freedom of movement with the relaxation of alert level 4 to alert level 3 in July. At the same time, there was a steep rise in objective COVID-19 risk with the surge in cases. The study examines the relationship of NPI adherence and the perceived risk of contracting COVID-19, the perceived effectiveness of NPIs and the accuracy of information held. We find a large increase in perceived infection risk and mask-wearing over this period. There are encouraging signs of the widespread credibility of high-impact NPIs and few resort to unproven prevention measures and poor science.

An understanding of individual health response behaviour is important in managing the pandemic risk in the country. Given the un-emphasizable economic and social divide that exists in South Africa, it is critical to acknowledge and manage the health response of its residents within the different socioeconomic contexts that define the lived realities of individuals. The current policy paper therefore aims at assessing the socio-economic inequality in some of the major factors that drive individual health behaviour, viz, subjective risk perception, self-efficacy or the belief that good behaviour can yield desired health outcome, feasibility of adopting preventive measures as revealed by individual behaviour and lastly, the sources of information related to the pandemic. The study finds that there is significant income, education and age-related differences in the individual response to COVID-19. While there is significant increase in subjective risk perception between June and August 2020 across the board, the non-blacks have significantly higher subjective risk perception compared to the black African population. The optimism bias is observed to be more pronounced among the less affluent groups like black African, rural and young population groups. The use of facemasks has gained widespread popularity across socio-economic groups. It is however of concern that there is increasing complacency towards physical distancing especially among the low-income groups. The findings indicate the need for tailoring policies specific to the various socio-economic contexts in the country.

The negative impact of the COVID-19 pandemic suggests that it disproportionately affects vulnerable groups in terms of income and health. One aspect of health that is often less researched but is particularly important in the context of the pandemic is mental health. We explore how the distribution of depressive symptoms (as a proxy for the state of mental health) and the variables associated with depressive symptoms in existing literature have changed relative to the pre-COVID period (2017). We also estimate the relationship between employment transition (between February and April, 2020) and depressive symptoms. While we exercise caution in comparing the depressive symptoms between 2017 and 2020 because of the difference in instruments, under reasonable assumptions our analysis suggests that depressive symptoms has increased significantly in 2020 relative to 2017. Furthermore, our results show that the pandemic has actually narrowed existing inequalities due to some factors that are known to influence depressive symptoms (e.g. gender and income). However, this equality occurs at a higher level relative to 2017. While higher subjective risk perception is associated with increase in depressive symptoms for the affluent, depressive symptoms for the poor is related to financial concerns and social grants. We find that employment shock that occurred between February and April 2020 is associated with increased (decreased) depressive symptoms for those who lost (gained) employment. This relationship also differs by gender with the effect of job loss being stronger for males workers.

This paper considers how access to public sector healthcare in South Africa have been affected by the COVID-19 pandemic using three sources of evidence: the NIDS-CRAM (Coronavirus Rapid Mobile Survey) and MatCH (Maternal and Child Health) SMS survey together with routine health care utilisation data from the District Health Information System (DHIS). Our analysis shows that across almost all districts in the country there was a reduction in primary healthcare utilisation, especially in HIV testing and health visits by children under 5 years of age, irrespective of the actual district-level incidence risk of COVID-19 at the time. Preventive services such as immunisation and contraception were still below expected levels in 75% of districts in August 2020. These unmet needs are corroborated by survey results for the uninsured segment of the population where 23% reported not seeking acute care when needed and inability to access medication, contraceptives or condoms. HIV testing is climbing again in some areas, and early access to antenatal care was only slightly lower than expected, recovering quite quickly since April 2020. COVID-19 fears may have disproportionately affected the poorest and most vulnerable groups, and impacted access especially to preventive services, where the negative consequences may unfold over an extended period if interventions to ‘catch up’ are not urgently prioritised and closely monitored.

Latest Panel User Manual 

The coronavirus pandemic is the largest social and economic shock in our lifetime. The rapid spread of this virus around the world and the economic devastation it has left in its wake is unlike anything we have seen before, at least not in our lifetimes. The local and international landscape is constantly morphing and changing in unpredictable ways making policy formulation and implementation as hard as it can possibly be. Policies that are helpful and sensible today may be harmful and illogical tomorrow. Every month seems to yield new information and consequences that were unforeseen even six weeks before. Indeed, “There are decades where nothing happens, and there are weeks where decades happen” (Lenin).

It is within this context that we have convened a national consortium of 30 social science researchers from five South African universities to conduct the National Income Dynamics Study (NIDS) Coronavirus Rapid Mobile Survey (CRAM) over the course of May – December 2020. Because decision-making is only as good as the data on which it is based, the NIDS-CRAM project exists to collect, analyze and disseminate data on a broadly representative sample of South African individuals, and to report on their employment and welfare in light of the COVID-19 pandemic.

This policy paper analyses the early effects of the COVID-19 crisis and ‘hard lockdown’ period in South Africa on women’s and men’s work in the paid and unpaid (care) economies. Because women and men typically have different roles in both of these sectors, it is likely that they would experience the negative effects of the crisis unevenly, potentially exacerbating existing inequalities. Based on the first wave of the NIDS-CRAM survey, we find that women have been more severely affected than men in the early phase of the crisis in South Africa, namely the ‘hard’ lockdown period. Net job losses between February and April 2020 were higher for women than for men, with women accounting for two-thirds of the total net job losses. Among those who remained in employment, there was also a bigger fall in average hours worked per week for women than for men. Compounding these disproportionate employment losses were disproportionate increases in unpaid childcare as a result of the lockdown and school closures. While the majority of both men and women living with children reported doing large amounts of extra childcare in April, women were doing relatively more. Nearly 80% of women who were spending more time than usual on childcare were spending more than 4 extra hours a day on it, compared to 65% of men. This gender gap persists even among men and women who reported being employed in April. Understanding who bears the brunt of the crisis in the workplace and in the home is key to designing appropriate policy responses.

We use a newly-released South African panel dataset to present some of the first estimates of COVID-19-related employment and poverty impacts in a developing country. We observe a 40% net decline in active employment compared to pre-lockdown. Approximately half of this decline is comprised of job terminations as opposed to temporary lay-offs or paid leave, suggesting that labour market impacts are likely to be persistent. Women, manual workers and those at the bottom half of the income distribution have suffered disproportionately higher rates of job loss. In the absence of data on prelockdown income, we cannot directly observe poverty changes over the lockdown period; instead we compare lockdown incomes of job-losers and job-retainers after DiNardo-FortinLemieux re-weighting. We estimate that 15 – 30% of job losers fall into poverty, which corresponds to approximately 1 million individuals. Social protection efforts have provided inadequate coverage to substantially mitigate this poverty impact, with only one in five temporarily unemployed workers receiving the relief designed for them, while other job-losers must rely on the existing social grant system. About one in three job losers had no social protection at all.

In light of the South African government’s pandemic-induced expansion of the country’s social protection system to provide relief to vulnerable individuals and households, this paper aims to use newly available, nationally representative survey data – Wave 1 of the NIDS-CRAM – to provide a quantitative, descriptive evaluation of whether social grants are being successfully targeted at the most vulnerable in the context of the national lockdown and COVID-19 crisis. In particular, we highlight heterogeneity in labour market outcomes before and during the national lockdown to show that social grants are an important source of income relief for individuals in low-income households. We show that these individuals have been disproportionately burdened by adverse labour market effects induced by the lockdown with respect to employment loss, the likelihood of having a paid job to return to, and reductions in working hours and earnings. We show that grants substantially increase the incomes of poor households in relative terms and, through fiscal incidence analysis, we show that the pandemic-induced additional government spending on grants have been pro-poor, especially that on the Child Support Grant. Considering the observed heterogeneity in labour market outcomes before and during lockdown across the household income distribution and that grants are relatively well-targeted, we conclude with a set of policy recommendations for South Africa’s social protection system going forward.

The COVID-19 pandemic is a health crisis that will exacerbate the South African unemployment crisis. As a temporary measure, the lockdown has prevented a substantial number of people from going to work. In addition, the longer term implications of a global recession have not been adequately recognized. This paper conducts an analysis of labour market dynamics in South Africa during the initial period of lockdown, from the end of March to the end of April 2020, using the first wave of the NIDS-CRAM (2020) survey. Within our sample of over 6000 adults aged 18 to 59, we found that there was a very large decrease in employment. The fraction of the sample that was conventionally classified as employed decreased from 57% in February to 48% in April. If we further exclude temporarily absent workers, this fraction decreases further to 38%. Thus, about one of every 3 employed people in our sample either lost their job or did not work and received no wages during April. This has extremely large implications for poverty and welfare. We further analyse the labour market by comparing across demographic groups as defined by race, by gender, by age groups, by geographic areas, and by education levels. The over-arching finding from this analysis is that the job losses were not uniformly distributed amongst the different groups. In particular, groups who have always been more vulnerable – such as women, African/Blacks, youth, and less educated groups – have been disproportionately negatively affected. In addition to the poverty implications, this will also likely affect the inequality situation in South Africa.

There is widespread recognition, both internationally and in South Africa, that the measures to curb the spread of COVID-19 have impacted particularly negatively on informal workers, whose jobs are precarious, who often depend on daily earnings for survival, and who lack legal and social protections. However, it is also likely that these impacts have been experienced unevenly by different groups of workers within the informal economy. In many contexts, the current moment has been described as a ‘triple crisis’ consisting of a health, economic and care crisis that impacts on women more than men. This paper analyses the first wave of the NIDS-CRAM survey in order to identify how the effects of the COVID-19 crisis differ within the informal economy and, in particular by gender and type of employment (by self-employment, informal wage employment and casual employment). We find that just under a third (31%) of informal workers who did not lose their livelihoods completely, were ‘locked out’ of employment in April – compared with 26% of those in formal employment. Among those who were employed informally in February and April, women in the informal economy saw a decrease of 49% in the typical hours worked in April while men in informal employment saw a 25% decrease in typical hours. The decrease in hours worked within the informal economy was greatest for the self-employed where average hours decreased by a third and typical hours decreased by more than 50%. Not surprisingly, these large reductions in hours coincided with earnings losses in the informal economy. Among the informal self-employed who were working in both months, average earnings decreased by 27% and typical earnings by 60%. For women in informal self-employment, typical earnings decreased by nearly 70% between February and April. These findings suggest that, as the pandemic unfolds in South Africa, current interventions need to be significantly scaled-up and far better targeted at informal workers, in general, and women informal workers in particular.

The COVID-19 lockdown, the requirement that people “shelter-in-place”, the physical closure of learning institutions, and the suspension or loss of employment, changed the living arrangements of a sizeable share of South Africans. For some, the change was temporary, while for others, it may be more longlasting. As the effects of the pandemic and ensuing economic contraction ripple through the labour market, more people are expected to move into different households and to rely on kin and social networks in the absence of other means of economic support. In this paper, we explore the spatial connectedness of the South African population, both from the perspective of households which are “stretched” to include non-resident members, and from the perspective of adults who moved during the early stages of the lockdown. In doing so, we highlight the importance, but also the limitations, of living arrangements and kin networks in helping to absorb the costs of the COVID-19 crisis.

Child hunger in South Africa declined after introduction of the child support grant. Despite the 2007/8 recession and slow economic growth since, child hunger more than halved from being reported in 35% of households in 2002 to 16% in 2018 in the General Household Survey. However, the pandemic reversed many of the gains. In newly collected data from Wave 1 of the NIDS-CRAM study, 15% of respondents reported that a child in their household had gone hungry in the week before they were interviewed in May or June. For the month of April, 47% of respondents reported that their household had run of money (the first month of the lockdown, before social relief measures were instituted). A much lower percentage, only 25%, reported that they had run out of money for food in the past year in 2018, a far less strict criterion. Loss of main income source between February and April, largely as a result of the lockdown, is strongly associated with a higher likelihood of child hunger in households. There is some shielding of children by adults: Child hunger is much lower than for other household members, and many households running out of money for food do not report child hunger. Drawing down savings, borrowing and depending on the generosity of others may have acted as a shield for child hunger in this initial lockdown period. This cannot continue indefinitely. A short analysis of social grants and school meals points to their importance for reducing child hunger, but also highlights their limitations at a time of great economic trauma.

In response to the COVID-19 crisis, and the anticipated socio-economic impacts of a hard lockdown, the South African government has leveraged three channels of social protection to protect livelihoods: social insurance, a social assistance programme of grants and localised social relief efforts. Using a mixed methods approach, this policy report highlights that while the response from government and the social sector has been commendable, large groups of households are experiencing tremendous hardship as a direct consequence of the lockdown. New evidence from a telephonic survey suggests that 2 of every 5 adults reported that their household lost its main source of income since lockdown started in South Africa on the 27th of March 2020. This has had devastating consequences for food insecurity and household hunger. Of interviewed adults, 47% reported that their household ran out of money to buy food in April. Between May and June 2020, 21% reported that someone in the household went hungry in the last 7 days and 15% reported that a child went hungry in the last 7 days. To stave off mass, chronic hunger we simply cannot let up on the support being provided to households through all three channels of social protection. Failure to do so could deepen an emerging humanitarian crisis, hamper economic recovery, and threaten socio-political stability.

The rapid spread of COVID-19 in South Africa threatens to amplify an unequal and polarised health system, with poor and vulnerable populations bearing a greater share of the COVID-19 infection and mortality burden. Available evidence suggests that preventative measures would have a protective effect against the spread of the virus. However, the success of these measures depends on whether the public receives, internalises and acts on appropriate messaging. Given our reliance on preventative measures for containing the spread of COVID-19, compliance with preventative measures is lower than desired. Energy is frequently misdirected: with higher compliance with low-impact measures aimed at preventing atypical transmissions via surfaces and lower compliance with a first-best set of preventative measures such as avoiding people, physical distancing and mask-wearing that aim to prevent droplet transmission, which is more typical. It is also concerning that only 6% of respondents in our NIDS-CRAM survey knew the three most common COVID-19 symptoms. This lack of knowledge may hamper early identification which is key to stopping the spread of COVID-19. It is discouraging to see that high-risk groups such as the elderly and those with chronic.

Our study examined the unintended health consequences of the COVID-19 pandemic in South Africa. We considered access to acute and chronic care amongst the general population with our NIDSCRAM (Coronavirus Rapid Rapid Mobile Survey) telephone interviews, and access to vaccinations, ARTs and antenatal care by pregnant women and women with infants in the public sector with our MATCH (Maternal and Child Health) SMS survey. We find that within the range of care and treatment types that we examined, the impact has been varied, with the highest-stakes type of care seemingly least affected. Chronic care and care of infants have been least affected. We find that 22% of those who needed acute health care, did not seek care. Also, 16% of pregnant women and mothers of infants in our public sector sample have not been to the clinic for two months. We find it particularly concerning that 11% of our sample of public sector expectant mothers or mothers with infants have run out of ARTs. Overwhelmingly, fear of contracting the Coronavirus is the most frequently cited reason for not seeking care or not accessing treatment.

The coronavirus pandemic (COVID-19) has resulted in an unprecedented dislocation of society especially in South Africa. The South African government has imposed a number of measures aimed at controlling the epidemic, chief being a nationwide lockdown. This has resulted in income loss for firms and individuals, with vulnerable populations (low earners, those in informal and precarious employment, etc.) more likely to be adversely affected through job losses and the resulting income loss. Income loss will likely result in reduced ability to access healthcare and a nutritious diet, thus adversely affecting health outcomes. Given the foregoing, we hypothesize that the economic dislocation caused by the coronavirus will disproportionately affect the health of the poor. Using the fifth wave of the National Income Dynamics Study (NIDS) dataset conducted in 2017 and the first wave of the NIDS-Coronavirus Rapid Mobile Survey (NIDS-CRAM) dataset conducted in May/ June 2020, this paper estimated income-related health inequality in South Africa before and during the COVID-19 epidemic. Health was a dichotomized self-assessed health measure, with fair or poor health categorized as “poor” health, while excellent, very good and good health were categorized as “non-poor” health. Household per capita income was used as the ranking variable. Concentration curves and indices were used to depict the income-related health inequalities. Furthermore, we decomposed the COVID-19 era income-related health inequality in order to ascertain the significant predictors of such inequality. The results indicate that poor health was pro-poor in the pre-COVID-19 and COVID-19 periods with the latter six times the value of the former. Being African (relative to white), per capita household income and household experience of hunger significantly explained income-related health inequalities in the COVID-era, while being in paid employment had a nontrivial if statistically insignificant effect. Addressing racial disparities, tackling hunger, income inequality and unemployment will likely mitigate income-related health inequalities in South Africa during the COVID-19 epidemic.

Latest Panel User Manual