Stellenbosch University
Welcome to Stellenbosch University
New survey provides critical data in fight against COVID-19
Author: Daniel Bugan
Published: 16/07/2020

​​​​The coronavirus pandemic has devastated South Africa's social and economic landscape, making policy formulation and implementation near impossible. Within this context a new study sheds light on who is being affected, how they are being affected and how government can respond to the crisis.

The National Income Dynamics Study - Coronavirus Rapid Mobile Survey (NIDS−CRAM) sampled the responses of 7 000 South Africans over the period 7 May to 27 June 2020. Respondents were asked about their income and employment, their household welfare, receipt of grants and about their knowledge and behaviour relating to COVID-19. The result of this study was launched on Wednesday, 15 July 2020.

Over 30 social science researchers from a number of universities and research organisations led by a team from Stellenbosch University, the University of Cape Town and the University of the Witwatersrand, collaborated on the study.

Dr Nic Spaull of SU's Economics Department and the principal investigator of the NIDS−CRAM study said: “The study can be seen as a type of barometer, assessing how firms and families are being affected by the lockdown, the pandemic and the recession. But it will also report on the reach and efficacy of government's social and economic relief efforts in the coming months."

The study presents key findings from 11 papers which were summarized into three main sections: employment, hunger and health.

Respondents were asked questions relating to the period February to April 2020. April was an especially hard time in South Africa as the country was experiencing its most severe form of lockdown. Many South Africans' income and employment was severely affected.

There was an 18% decline in employment between February and April 2020: 17 million people were employed in February, but only 14 million people in April.

One in three income earners in February did not earn an income in April. Ranchhod & Daniels (2020) found that the proportion of adults who earned an income in February declined by 33% which is made up of a roughly equal share of those who lost their job and those who were furloughed (temporarily laid off). Jain et al (2020) reported a 40% decline in “active employment' which was also split equally between those who were laid off and those who were either furloughed or on paid leave.

All authors found that job losses were disproportionately concentrated among the already disadvantaged groups – such as women, African/Blacks, youth, and less educated groups – in the labour market.

Of the approximately three million net job losses between February and April, women accounted for two million, even though in February they only accounted for less than half of the workforce (47%) (Casale & Posel, 2020). Of those women who were employed in both February and April, almost half of them reported working fewer or no hours in April (compared to 42% of men).

Many of those who experienced job losses are in grant-receiving households, for example a child support grant or an old age pension. Jain et al (2020) found that approximately 30% of those who were retrenched between February and April reported no household-level grant protection at all.


A total of 47% of the respondents had run out of money to buy food in a single month, April (i.e. during lockdown).

Furthermore, 21% reported that someone in their household went hungry in the last seven days, while 15% reported that a child had gone hungry in the last seven days.

Thirteen percent reported that someone in their household had gone hungry for three or more days in the last seven days. In addition, 7% reported that someone in their household went hungry (almost every day or every day) in the last seven days.

In households with children, 8% reported frequent child hunger (3+ days in the last seven days) in their household and 4% reported perpetual hunger (almost every day or every day).

Where adults experienced hunger for four days or less in the last seven days, 47% indicated that a child did not go hungry in the household (i.e. half of children were shielded). However, when there was perpetual adult hunger (almost every day or every day in the last seven days) then adults appear less able to shield children, since only 33% of respondents indicated that a child did not go hungry.

Among those respondents who reported a child going hungry in the household in the last seven days, 84% said that there is at least one Child Support Grant (CSG) in the household. Furthermore, in households where the respondent indicated that someone had gone hungry (an adult or a child), 72% have either a CSG or Old Age Pension in the household, and 63% have at least a CSG in the household.

Pre-lockdown, three quarters of grant receiving households relied on income sources other than grants. Pandemic-induced job losses presented a major threat to the livelihoods of a large proportion of grant receiving households. This can partly explain why hunger has been exacerbated despite grant income not changing, or even increasing.

NIDS−CRAM data was supplemented with reporting from the Maternal and Child Health survey that was sampled from 15 000 new and prospective mothers on the Department of Health's MomConnect platform. All respondents were surveyed from 7 May to 27 June.

Of the 7073 participants, 23% reported they were unable to access medication, condoms or contraception in the past four weeks. Of the 1 524 respondents with a chronic condition such as HIV, TB, a lung condition, a heart condition or diabetes, 39% reported that they could not access medicine, condoms or contraception.

A total of 96% who needed to see a healthcare worker for a chronic condition in the last four weeks, said they were able to do so (NIDS-CRAM). This is encouraging since the period referred to was during Level 5 or Level 4 lockdown.

Of those that needed care for an acute condition, only 78% visited a healthcare facility (NIDS-CRAM). Nearly 9% reported symptoms in the last four weeks that indicated they might need care (for example fever, sore throat, coughing, and shortness of breath or injury).

Affluent individuals in the top quintile are almost twice as likely (53%) to believe that they will contract COVID-19 than those in the poorest quintile (25%).

While 91% of respondents reported changing their behaviour in some way to try and prevent contracting or spreading the virus, much of this effort is expended on low-impact strategies. Of those that reported changing behaviour, only 35% reported enacting first-best preventative behaviours.

Although 64% of respondents listed coughing as a symptom, and 63% listed fever, only 11% listed tiredness as a symptom. This implies that many South Africans would not be in a good position to make decisions about when it would be vital to quarantine and/or seek care for COVID-19 symptoms.

Of the 3 047 new or prospective mothers that answered the question about antiretroviral therapy (ART), 46% said they did not require ART. Of the 1 610 that required ART, 11% said they ran out of ART.

Of the 3140 women surveyed at the end of June, 60% reported attending a healthcare facility in June and a further 24% in May. Of concern is the fact that 16% reported that their last clinic visit was in April or earlier. This equates to a two-month gap in care, which represents a risk to pregnant mothers and to new lives in this important developmental phase. Of those who did not go to the clinic, 37% cited Coronavirus fears as the reason.

Inequality worsens

“Reviewing the findings presented in the 11 papers reveals that there is a high degree of agreement between researchers on what the key findings are; that employment has declined substantially and that the effects of this are largest for the most disadvantaged," said Spaull. “Inequalities along traditional lines of race, gender, occupation, earnings, location and education, have all grown significantly. An already unequal national situation has been made much worse as a result of the COVID-19 pandemic and the local and international responses to it."

Policy options
The study provides policymakers with the following options:

  • Deploy community health workers more effectively as service linkers;
  • Strengthen local coordination structures to ensure local ownership and champions that are required for changing social norms;
  • Distribute free masks to the poorest communities: Not all citizens who need them can afford appropriate face masks, or enough of them. These should be distributed with instructions as well as information flyers;
  • Provide specific and actionable recommendations on key preventative behaviours, with a focus on mask-wearing and physical distancing;
  • Provide clear, concise and consistent communication to improve knowledge of symptoms especially weakness/tiredness; and
  • Anchor messages in hope and a positive vision for the future.

Spaull added that there will be at least four further waves of data collection following the same individuals during the course of 2020.​

  • For more information about the National Income Dynamics Study/Coronavirus Rapid Mobile Survey (including reports), go to: