Siyanda Baduza, Kelle Howson, Thato Setambule and Thobani Khumalo (Institute for Economic Justice)
More than half of the target population has not been able to access a lifesaving South African social assistance transfer aimed at alleviating extreme poverty amongst working-age adults. While the overall scale of exclusion from the “COVID-19 Social Relief of Distress Grant” (SRD grant) has long been known, research has not yet systematically evaluated exactly how people are excluded, nor have we understood who is most at risk of exclusion.
The Institute for Economic Justice (IEJ) has recently published the results of a large study exploring the extent, drivers, and impacts of systemic exclusion from the SRD grant. Our research uncovers how automated poverty targeting methods deployed by state agencies have exacerbated existing social inequalities, barring millions from accessing their rights.
Background to the SRD grant: mass exclusion predetermined by budget allocation
Before 2020, able-bodied persons aged from 18 to 59 had no access to tax-financed social assistance in South Africa. With the onset of the COVID-19 pandemic, the Government introduced the Social Relief of Distress (SRD) grant in May 2020 as a monthly emergency cash transfer for able-bodied adults living in poverty. In light of the country’s ongoing crisis of adult poverty, the grant has remained in place.
The grant is targeted at people with monthly means (“income or financial support”) below R624 (US$33). This means-test threshold is well below the country’s extreme (or “food”) poverty line, currently R796 (US$42) per month. Despite this highly restrictive means-test, approximately 17 to 18 million working-age adults are estimated to be eligible.1 As of September 2024, applications had reached 17.4 million. However, the budget allocation for the grant is sufficient to cover only around 8.5 million monthly recipients. Monthly approvals have remained below this budget cap, averaging around 8 million since April 2022.
Fig. 1: Estimates of Adults Below Full Poverty Line, SRD Grant Eligibility, SRD Grant Applications, Approvals and Payments – Compared to National Treasury Budget Cap on Recipient Numbers (2022/2023)
While the regulations governing the grant contain the proviso that payments will stop being disbursed once the budget cap has been reached, this has apparently not been invoked. Instead, all rejections of applications have been made on the grounds of ineligibility. (The reason for approvals remaining constant while rejections have increased is unclear.)
Objectives and methodology
The scale of—and underlying reasons for—mass exclusion from the grant are plainly evident. The objectives of our study were to: determine empirically how erroneous exclusions occur and on what grounds; identify those most at risk of erroneous exclusion; and explore the perceptions and experiences of those affected.
We used a mixed-methods approach, combining survey data from 900 participants living in poverty who experienced exclusion in at least one month during our reference period (December 2023 to February 2024) with 58 semi-structured qualitative interviews drawn from this group.
Findings: The majority of exclusions are erroneous
Using our survey data on self-reported income and financial support, we conducted an independent eligibility assessment against the official SRD grant criteria.2 We found that an average of 75.9 per cent of respondents were eligible each month. However, only 10.3 per cent of this group had actually received payment in the month they applied, indicating an average monthly erroneous exclusion rate (of eligible respondents) of 89.7 per cent. While this cannot be taken as an indication of overall exclusion, since we selected individuals who had experienced exclusion and were living in poverty, it nonetheless reveals a staggering level of erroneous exclusion of people in need. Erroneous exclusions far outweighed erroneous inclusions (at 16.8 per cent), a troubling finding given that exclusion violates human rights and undermines the grant’s core purpose of reducing hunger and poverty.
The high erroneous exclusion rate alone does not explain how people were excluded. Therefore, we examined the points at which exclusion occurred across all stages of the grant process, including application, eligibility verification, payment and appeals.
Barriers to application
Whereas applications for other welfare programmes in South Africa can be made in person, applications for the SRD grant can only be made online. Additional requirements and security features mean that applicants need continuous access to a camera-enabled smartphone and a bank account in their name. This is despite the fact that 21.4 per cent of people in South Africa had no internet access (fixed or mobile) in 2023, and many don’t have their own bank account. These are likely to encompass the poorest households, in effect, those in the SRD grant target group.
Among those in our sample who were eligible for the grant but had never received it, 25 per cent reported not having applied. Unsurprisingly, people without smartphones and bank accounts were less likely to have applied despite being eligible. We also found that those with lower education levels found it harder to navigate the application process and were thus less likely to apply. This is in part related to the fact that—though the vast majority of SRD grant applicants do not speak English at home—the application process is only available in English.
Over 70 per cent of unsuccessful applications were rejected on the grounds of a “failed” means test, but an average of 76 per cent of these rejections were erroneous each month (of persons we judged to be eligible). This stems directly from flaws in the state’s poverty-targeting methodologies. Means testing is carried out via an automated bank surveillance system that considers any and all monies appearing in someone’s account to be “means”. As Table 1 indicates, this includes money received on behalf of someone else (like a child or relative), loans, and ad hoc donations from friends and family intended to help the recipient survive in a given month. None of these kinds of account inflows can be considered regular income or financial support which benefits the applicant. Individual bank account surveillance can also “double-count” money transferred between household members and used for the basic needs of the household. In addition, we found that the poverty targeting system penalised and discouraged income-generating activities (such as enterprise formation or attending job interviews), thus reinforcing a poverty trap.
Table 1. Source and purpose of bank account inflows which led to grant rejection (selected responses)
| Source | Purpose |
| A friend of my mom | My mom sews clothes for church members, and she sometimes uses my bank account to receive payments |
| A loan shark. I was supposed to give it to my mom | My mom’s bank card was misplaced, so we had to use mine |
| A relative | Had to give it to somebody that did not have a bank account |
| From my boyfriend | To buy clothes for our son |
| Money lender | So I can buy food for Christmas |
| My relative | For me to start a small business |
| My sister | Buying my mom medication |
| My sister | To go for a job interview |
The risk of erroneous rejection on the basis of bank account surveillance was higher for certain population groups. Women were more at risk than men, possibly due to a ‘caregiver penalty,’ as maintenance payments for children (very common in South Africa) are picked up by means testing.
Another risk factor was ID type. Holders of “green ID books” (an older, offline yet legitimate form of identity document) were more likely to be rejected, probably due to the rollout of a biometric identity verification system that only supports the new “smart ID” cards. This again disadvantages the poorest, as the cost of switching to a smart card ID is R140 (US$7), 40 per cent of the grant value (R350) at the time of the survey.
Barriers to receiving payment: digital and financial exclusion
Nearly half of all approved grants were not paid during our three-month reference period. Most participants who had not been paid did not know why, though some pointed to an inability to complete biometric identity verification (which requires ownership of a smart ID card and a camera-enabled smartphone). A new law that came into force during our data collection gave the state the power to cancel historic pending payments. Some respondents indicated their unpaid grants had been cancelled.
People who opted to be paid into a personal bank account at the time of application were more likely to receive their grants compared with those who opted to collect via the South African Post Office. This again exacerbates the financial exclusion of the poorest.
Barriers to successfully appealing an unfair rejection
Rejected applicants are allowed to appeal, but they cannot provide new information or documentation to support their appeal, for example, to explain why a certain transaction appearing in their account was not personal income. The appeal process, therefore, simply reruns the same flawed automated bank verification check.
Most participants had appealed at least once, but only 5.3 per cent had ever had a successful outcome. Rejections resulting from automated bank account checks were much less likely to be overturned.
Because the appeals process is also exclusively online and in English, those without internet access and with lower levels of education were also less likely to appeal.
Conclusion
Our study shows exactly how, in the case of the SRD grant, poverty targeting has led to the disproportionate exclusion of the most vulnerable. People who are already disadvantaged by the digital divide, by unpaid caregiving responsibilities, and by lack of access to the financial system and to identification documents are at greater risk of falling through the cracks in the social safety net.
Erroneous exclusion has been driven in large part by a rapid digitalisation and automation of the grant infrastructure with insufficient oversight and accountability. However, these barriers have simply served to enable disbursement to remain within the arbitrary budget cap, which limits the grant’s availability to a fraction of the eligible population.
In our working paper, we propose concrete policy recommendations to address unfair exclusion from the SRD system, but note that meaningful reform depends on adequately resourcing the programme.
Read the full working paper here.
The IEJ would like to acknowledge and thank Agence Française de Développement (AFD) for making this research possible.
1 Orkin, K., Goldman, M., Kreft, B., Hlela, N., Nicklin, J., Woolard., I & Leibbrandt., M. (2023). Proposal for the extension, redesign, and repurposing of the special COVID-19 Social Relief of Distress grant for the twin goals of poverty and unemployment reduction. SA-TIED Working Paper 218.
2 Which also include age, qualifying immigration/citizenship status, and residence in South Africa.