Skip to main content


PhD Thesis 

Candidate: Judith Streak
Supervisor: Professor Servaas van der Berg
Institution: Stellenbosch University, Faculty of Economic and Management Sciences, Department of Economics


One of the cruelest legacies of the apartheid era in South Africa was that it created a situation in which a very large proportion of children in the African and Coloured populations live in households affected by severe poverty. The first aim of this dissertation is to make a contribution to knowledge about the South African child poverty profile and its sensitivity to the adult equivalence scale used in measurement. This contribution is based on the Income & Expenditure Survey 2005. In April 1998, following the recommendations of the Lund Committee, the Child Support Grant (CSG) cash transfer programme was introduced. This programme initially paid a benefit of R100 a month to primary caregivers who passed a means test to help them care for children age 0-6. Currently it offers R250 to primary caregivers of children age 0-15 who pass a means test. The second aim of this dissertation is to synthesize the main findings and knowledge gaps of the performance of the CSG based on an analysis of the existing research on the programme. The questions used to structure the CSG analysis are derived from an application of the Rossi et al. (2004) systematic approach for tailoring a social programme evaluation and cover the logic of the programme‘s impact theory, implementation impact and design.

Chapter one considers conceptual and methodological issues in child poverty measurement, thereby providing methodological foundations for the analysis. Chapter two reviews the existing research on child poverty in South Africa and identifies knowledge gaps that the Income and Expenditure Survey of 2005 analysis contributes towards filling. Chapter three presents the findings on the sensitivity of the child poverty profile to changes in the adult equivalence scale as well as on the dimensions of the South African child poverty profile. Chapter four describes the Rossi et al. (2004) method for tailoring a social programme evaluation and applies it, thereby laying the foundations for the CSG programme analysis, presented in chapter five. The conclusion explains how the child poverty analysis in the first half of the dissertation relates to the CSG analysis in the second and draws implications of the dissertation‘s main findings for future research.

Setting the poverty line at the 40th percentile of households calculated with different AESs, the scope and composition of child poverty are found to be relatively insensitive to the scale used. This supports the argument that it may be appropriate in South Africa to use a poverty line based on a per capita welfare measure. For the construction of the poverty profile per capita income is used as the welfare indicator, with the poverty line set at the 40th percentile of household. The profile finds that poverty remains more extensive amongst children than adults even after the massive injection of cash via the CSG into poor households with children. Large variations across provinces remain. The child poverty headcount and depth and severity measures are all found to be higher amongst children age 0-4 than children age 15-17, despite the prioritization of very young children in the roll out of the CSG programme. The finding that children age 0-4 are still most in need questions the logic behind the government‘s recent decision to expand coverage of the grant to children age 15-17 instead of allocating additional funds to support this group.

The CSG programme impact theory is found to be reasonable. A substantial amount of research on the programme‘s implementation is identified, which shows the massive expansion in its coverage since 2000 and that, in general, it has been well implemented. Delay in reaching very young eligible children, under-representation of children with non-biological caregivers and failure to reach some of the very poorest children who live in remote areas emerge as weaknesses in programme implementation. Barriers to access are identified. The difficulty of distilling the effects of the CSG programme on child poverty and its associated deprivations in the absence of a randomized community trial is explained, as is the need to use direct rather than indirect monetary indicators to isolate the impacts of the CSG on child poverty. A small yet convincing evidence base on the impacts of the programme is identified; this shows that it has been achieving its ultimate objectives of reducing child deprivation and promoting human capital development. It is argued that the existing research and current budgetary context suggests that: the income means test should not be eliminated; the CSG benefit value should not be raised substantially; school related conditions should not have been linked to receipt of the CSG benefit; and, even in the case of children age 16 and above, the grant should be paid to children via the primary caregiver.

The research priorities implied by this dissertation are organized into three separate yet interrelated areas of research. The first is further research on the dimensions of multi-dimensional child poverty in South Africa. Here the spotlight needs to be on: understanding more about which children are deprived and in what sense; similarities and differences between the composition of poverty based on indirect monetary measures and other more direct measures of deprivation; the circumstances which ensure that children age 0-4 are still most in need (at least in the resource deprivation sense). The second area is research on the implementation of the CSG programme, in which case the findings from the existing research on the weaknesses in programme implementation and concerns about targeting outcomes need to guide the research. The third area of research identified as requiring further attention is that of the CSG‘s impacts on child deprivation (wellbeing). Here the focus needs to be on establishing which children are benefiting most and least from the grant and why. Towards this end quantitative researchers need to follow the lead of other researchers and use direct indicators of child outcomes. Qualitative research on who controls the resources that flow into the household, as well as how resources are allocated inside the household, can make an important contribution to answering these questions. The CSG‘s potential to generate behavioral effects which could alter household structure and/or income earned from wages is something that is also identified as important for researchers to explore.