Social grants: the research behind the controversy
We look at how different grants were evaluated in the recent SALDRU study
A UCT report for the National Treasury has caused consternation in civil society after a government strategy document used its findings to propose ending the R350 Covid Social Relief of Distress grant and replacing it with a much smaller pilot programme.
Two of the authors of the report have come out against the National Treasury using their work to justify removing the R350 SRD grant.
But what does their paper actually say?
On Monday 8 November, the 56-page working paper, called Simulation of options to replace the special COVID-19 Social Relief of Distress grant and close the poverty gap at the food poverty line, was released to the public. The paper was the result of a collaboration between the Southern Africa Labour and Development Research Unit (SALDRU), based at the University of Cape Town, and the National Treasury.
Since the Covid crisis hit, the R350-a-month SRD grant has been used as an emergency measure to reach a large number of people in desperate need. But many people who should not qualify for the grant do, and many who need help are not receiving it.
The SALDRU paper evaluates five different alternatives to the R350 SRD grant:
- Increase the Child Support Grant (CSG);
- Keep and extend the Covid SRD grant and extend it to Child Support Grant caregivers;
- Create a new Universal Basic Income Grant to working-age adults;
- Create a new “Family Poverty Grant” targeted at very poor households; and
- Extend the public works programme.
The paper is based on the assumption that current grants continue.
The question that the National Treasury wanted answered is: what is the cheapest way to make sure that everyone has enough money to eat each month?
This year in South Africa, according to the paper, the Food Poverty Line (FPL) is R624 per month. Other measures of poverty are the Lower-bound Poverty Line (LBPL), R890 per month, and the Upper-bound Poverty Line (UBPL), which is R1,335. About a quarter of the population receives an income below the Food Poverty Line.
The Food Poverty Line “gap” is the shortfall below the poverty line across the population, taken as an average. This should be thought of in the following way: imagine a two person economy where the food poverty line is R100, and one person has R50 in income and the other R120. The gap for the first person is R50 and for the second person R0 (as they earn more than the poverty line). Add these two gaps together and divide by two, and the poverty “gap” is 25%.
A key measure of comparison between programmes is ‘effectiveness’: For every R100 spent by the National Treasury, how many rands go directly to alleviating poverty? The more money spent on alleviating poverty, rather than on people who already have incomes above the poverty line, the more effective the programme.
The paper estimates that the total amount of money needed to bring everyone up to the Food Poverty Line is R45 billion in 2021 (this is if we could perfectly observe everyone’s income). So, a policy that costs R45 billion, with all the money going where it is needed, would be a perfectly effective programme, by the standards of the inquiry.
Based on this standard, in theory, how do the various proposals compare?
Increase the Child Support Grant
Two options were explored. In the first, the value of the grant would be increased to R624, the Food Poverty Line. In the second scenario, the value would be increased and so would the means-test threshold that people would become eligible for the grant, to incomes under R6,240 per month.
These grant options would cost R27.7 billion and R32.4 billion.
This option would be easy to implement, as the grant exists already. However, not many more people would receive the grant. At present, the CSG is a very efficient grant - 44% of the value of the grant goes towards reducing the poverty gap and further increases in value do not do too much more to help.
Maintain the Covid Social Relief of Distress grant
The paper examines six different options for amending the R350-a-month to better target more people and to alleviate poverty. As things stand, the grant is given to all those applicants without formal employment, with a means test used for those that appeal once they’ve failed the employment test.
Though the grant goes to 9.5 million people, the paper found that many others who are eligible do not receive it (and some who should not, do)
The most efficient way to reduce the poverty gap using this grant would be to give R350 to any adult with income less than the Food Poverty Line level of R624 per month. This would cost R71 billion per year. Here, 35% of the money spent on this grant goes towards reducing food poverty, and this would reduce the poverty gap by 55%.
Another option is to pay R624 to anyone with income less than the national minimum wage. This programme would cost R149 billion, but it would reduce the percentage of people with income below the Food Poverty Line from just over 25% to under 9%. This option would reach 19.9 million people directly, and 46.7 million indirectly. However, the efficiency of this option is low; under a quarter of the money spent would go directly to alleviating food poverty.
The Universal Basic Income Grant (UBIG) option
In this programme, a monthly universal basic income grant would be given to everyone between 18 and 59 years old.
The paper offers three varieties: give to all in that age (R624 per month, reaching 34.21 million people directly each month), give to all who earn below the national minimum wage of R3,731 per month (25 million people), and give to everyone, but use a ‘tax clawback’ on wealthier people - they would pay tax on the extra money.
Under these scenarios, the percentage of people below the Food Poverty Line would be reduced from about 25% to less than 8% of the population.
But these proposals would cost between R187 billion and R256 billion per year. This would double South Africa’s spending on all grants. The efficiency of these programmes would be between 15%-20%. In other words, most of the money would go to purposes other than reducing food poverty. This is not necessarily a bad thing - many people would still be well below the upper bound of poverty.
Furthermore, since there is no administrative barrier for people to jump to access this grant, the National Treasury would be assured that it would effectively wipe out food poverty - even if it is not the most efficient option for this task.
A Public Works programme
Another option considered is expanding the public works programme further to pay cash to a large number of people - either one million or two million - in exchange for eight days work a month. This would amount to either R795 for eight days’ labour at the current minimum wage in the Expanded Public Works Programme, or R1,446 at the national minimum wage, for either one or two million people, at a total annual cost of between R9.5 billion and R34.7 billion.
This approach produces a very small reduction in the percentage of people living below the Food Poverty Line. Even if 2 million people receive R1,446 per month, the share of population under the Food Poverty Line is reduced from just over 25% to just under 23%. This scenario is highly ineffective too - just 19% of spending on this programme would go to reducing the FPL gap.
Introducing a Family Grant
The Family Poverty Grant (family grant) model is most famously used in Brazil, where it was implemented by the government of Luiz Inácio Lula da Silva in 2003. It remains in force today, despite the attempts of President Jair Bolsanaro. “More than 30 million Brazilians escaped poverty between 2003 and 2014”, thanks in large part to the programme, according to The Economist. The programme is credited with reducing extreme poverty in Brazil from nearly 10% of the population to about 4% of the population.
The family grant is given to a household, rather than to individuals, and, in Brazil, is conditional - households must make sure that children attend school, for instance.
Three variations are presented. The first option offers R460 to each working age adult in a household. The second option tops up the household income so that each person earns on average R624. The third gives R460 to each working age adult, and then tops up any households with average incomes that are still below the R624 Food Poverty Line, to R624 per person.
The grant would be given to 5 million households, and would directly benefit about 20 million people. Under the first option, the percentage of people below the Food Poverty Line would drop from just over 25% to just over 13%, and in the top-up options, it would drop to zero.
The efficiency of the family grant is high - in the second variation 62% of spending would go towards reducing the food poverty gap. This would be achieved at a cost of R73 billion. Compared to other programmes, this is highly attractive - the National Treasury could eliminate food poverty in South Africa, and not double annual grant spending doing so.
The theoretical performance of the family grant is compelling.
But bringing this grant into practice is the problem.
The document used the SALDRU research to select the Family Poverty Grant as the best option, over the other alternatives like the Universal Basic Income Grant. It concluded that a Family Poverty grant of R624 (food poverty line) performed the best both from an impact perspective as it eradicates extreme poverty, as well as an efficiency perspective … “This strategy strongly supports National Treasury’s proposal for a Family/household grant, as it will have a more meaningful impact on poverty than other grants.”
The paper proposed a ‘transition strategy’ targeting 1 million households, and increasing the number of households by 1 million each year until all eligible households were covered. This ‘transition strategy’ would cut off many of the millions who have come to rely on the R350 SRD grant to survive, and was widely attacked by civil society groups, as well as economic think-tanks like the Institute for Economic Justice (IEJ).
A week after we revealed the leaked strategy, members of the SALDRU team spoke out against the National Treasury’s interpretation of the working paper. The family grant choice was “based on an incomplete reading” said Joshua Budlender, one of the authors of the working paper. On the day of the SALDRU paper’s release, Budlender and Ihsaan Bassier, also a co-author of the working paper, published an opinion article in the Daily Maverick in which they argued against the adoption of the family grant based on the findings of their paper.
In their article, Bassier and Budlender draw from the discussions in the working paper to argue that the family grant is impractical, since “the government would have to implement a means test for every person in an applying household, and they would need to know precisely which individuals live in which households.” This would require the creation of a household registry, which South Africa does not have.
The working paper points out that “the creation of such a registry would be exceedingly difficult and contentious in a country characterised by high rates of migrant labour, extended family support networks, and high rates of household dissolution and re-formation, and large populations living in informal settlements.”
Bassier and Budlender go further and show that the family grant is particularly sensitive to errors that would come with implementation.
If everyone who should get the family grant did get it, the efficiency of the grant would be 94%. But as the share of people who should not get the grant but do receive it increases, the efficiency of the grant drops much faster than the alternatives. This is called the false inclusion rate. “With a false inclusion rate of 40%, the Family Poverty Grant has an efficiency score of 58%, the SRD has 50%, and the R624 BIG is still at 53%”, say Bassier and Budlender.
Budget policy statement
Speaking to GroundUp, Maya Goldman, lead author of the working paper, said that her team had been given a mandate to look at the theoretical efficiency of various programmes only. But once they had begun to examine implementation, they felt they could not ignore it, given that implementation issues could completely change the ranking of the various options.
Goldman said they had found that based on the numbers in the household survey, many more people should be receiving the current SRD grant. She said more research on implementation was necessary and in the meantime a refinement of the SRD grant might be best.
“The SRD is an important grant in that it resolves huge gaps in our social security system and makes an important dent in poverty - and it’s cheaper than the UBIG which makes it appealing for the Treasury. The impact it can have on poverty can be increased by raising the means test from FPL to the national minimum wage. If it is to be extended, there must be an effort to resolve the implementation gaps that we identified,” said Goldman. “The UBIG is important from the point of view of the social contract - just because it’s not efficient, it doesn’t mean that we shouldn’t take a path towards it.”
There was no mention of the family grant in the medium term budget policy statement, released on 11 November. But, there was no mention of what grants would replace the R350 SRD either.
Finance Minister Enoch Godongwana refrained from committing the National Treasury to any extension of the SRD grant beyond March 2022, saying that it would be a decision made by the government between now and then, rather than by the National Treasury. He did say that any extension of the programme would need to come from cuts elsewhere or from a tax revenue windfall from a sustained commodities boom, rather than from increased taxation.
© 2021 GroundUp. This article is licensed under a Creative Commons Attribution-NoDerivatives 4.0 International License.
You may republish this article, so long as you credit the authors and GroundUp, and do not change the text. Please include a link back to the original article.
We put an invisible pixel in the article so that we can count traffic to republishers. All analytics tools are solely on our servers. We do not give our logs to any third party. Logs are deleted after two weeks. We do not use any IP address identifying information except to count regional traffic. We are solely interested in counting hits, not tracking users. If you republish, please do not delete the invisible pixel.