
Download cover  174Kb ~ 1 min (1 pages)

Executive summary
This paper estimates the cost of the means test for the Child Support Grant (CSG) to Government and to applicants. The costing draws on fieldwork conducted in three sites each in the Western and Eastern Cape provinces. These six sites were chosen to represent a range of settlements вЂ“ from very rural to metropolitan. At each site, Department of Social Development (DSD) officials dealing with each of the steps in the application and processing of Child Support Grants were asked about the time spent on a typical case, and how long they estimated it would take if there was no means test. The difference between the actual time and estimate without the means test gave us the time taken up by the application of the means test.
CSG applicants generally need to interface with the South African Police Service (SAPS) to obtain some type of confirmation of their documentation. Police officers were therefore also questioned on how long it took them to process documentation related to the means test. This was included in the cost to Government.
The time taken by each official, both at the DSD and SAPS, was then multiplied by the cost of employment (COE) for the lowest level of official who could be employed on the task to give the cost related to that step. The costs for the individual officials were added together to give the cost of one application, which amounted to R18.77.
The researchers interviewed applicants who had submitted completed applications. These women were asked to list all activities they had done thus far in the application process. From this list we determined which activities had something to do with proof for the means test, and asked about the time spent on this activity as well as any costs incurred. This provided a less exact measure of the cost of the means test than the calculations in respect of officials. This is because some of the costs would have been incurred even if the means test were abolished. The interviews and calculations yielded a mean monetary cost of just under R25 and close on six hours per applicant.
The interviews and calculations gave us the cost of processing a single application in respect of the means test. This then had to be multiplied by the total number of applicants. For this we needed eligibility rates for the Child Support Grant. A range of different estimates were derived in this respect to cope with two complications. The first relates to the fact that the means test thresholds have not changed since the CSG was introduced in 1998. We therefore developed estimates both in respect of the actual and unchanged thresholds, and in respect of what the thresholds would have been were they adjusted for inflation. The second complication relates to the fact that the age group eligible for the CSG has changed over the last few years. We therefore derived estimates for 08, 010, and 013 вЂ“ the age ranges eligible for 2003/04, 2004/05 and 2005/06 respectively. We developed further estimates for the age group 017, i.e. all children.
Eligibility figures were calculated by analysing the raw data from the General Household Survey (GHS) of 2003. The calculations required assumptions about the identity of the primary caregiver (PCG) of each child вЂ“ the person eligible to apply for the grant on behalf of the child. This was based primarily on who the children were living with. The calculations also required assumptions about the income of the selected PCG and their spouse, if married. The paper details all the assumptions and manipulations that were made to arrive at the eligibility estimates. The final estimate for children aged 013 years вЂ“ using standard threshold cutoffs вЂ“ was 8,791,705 children, using the weights supplied by Statistics South Africa (Stats SA). This is 65.3% of the relevant age group вЂ“ ranging from 77.1% in rural areas to 45.8% among children living in formal dwellings in
urban areas. After adjustment to correct for seeming errors in Stats SA weights, the estimate increases slightly to 9,008,851 children. This is 64.4% of eligible children as the adjustment to the weights also affects the estimate for total children. DSD records in respect of CSG payouts in December 2003 reflect 4,245,298 children, about 78% of our estimate of the number of children
who would have been eligible for the CSG at that date. An analysis of the characteristics of those deemed to be eligible and those actually receiving grants, suggests that it is children who do not have their biological mother as their PCG who are most likely not receiving the grant, even when eligible.
Finally, we estimated the cost of the means test if PCGs applied in respect of all eligible children by multiplying the cost derived for a single case to the estimate of the number eligible. (The estimate ignores the cost of applicants who are not successful and is thus on the conservative side.) The cost ranges from R113.2 million for standard cutoffs and age group 08 to R223.8 million for all children using inflationadjusted cutoffs if one uses the adjusted weights.
The cost of the means test should be incurred only once in respect of each child if the caregiver remains the same over the period that the grant is received. The total cost would thus not be incurred in a single year. Ideally, the costs should be incurred only in respect of children born in a particular year. In real life, however, caregivers change, especially in a situation where HIV/AIDS is rampant and where PCGs have had to reapply for the children in their care as a result of the
staggered rollout of the CSG extension.

