Between August 2005 and September 2006 Statistics South Africa undertook the IES 2005 to gather data on the income sources and expenditure patterns of a nationally representative sample of 21 144 households. This type of survey is usually conducted every five years.
Table 1: Number of children, child poverty headcount and number of poor children in South Africa and by age, racial, gender and geographical location |  | Source: Streak, Yu & Van der Berg calculations using IES 2005 data |
|
Released for analysis early in 2008, IES 2005 is the most upto- date national level data source for measuring poverty in South Africa. Whilst IES 2005 had been used before to provide an updated profile of poverty in South Africa it had not been used to analyse child poverty. How the analysis was donePoverty was conceptualised in narrow resource deprivation terms and measured in money terms using the IES income data. The per capita income, which is calculated by dividing household income by the number of individuals in the household, was used as the Adult Equivalence Scale (AES) to convert household income into individual income and an indicator of welfare for ranking individuals from poorest to richest. The poverty line was set at the poorest 40% of household per capita income in IES 2005, which amounted to a R4 560 per capita per annum in 2000 Rand values. We estimated child poverty headcount (i.e. the percentage of children below the poverty line) for South Africa as whole and explored variations in the child poverty headcount by age, racial classification, gender and racial classification. Child poverty profile
Table 1 shows the total number of children in South Africa, the child poverty headcount rate (percentage of children below the poverty line) and total number of poor children in South as a whole and by age, racial group, gender, urban/rural location and province. It also shows the poverty headcount for adults older than 18.
The profile suggests that child poverty (at 65.5%) remains more extensive than poverty amongst adults (45.2%), confirming that children are more often to be found in poorer households. It indicates that despite the massive injection of transfers into households with poor children through the introduction and expansion of child support grants, poverty amongst children is still substantial. The poverty rate was found to be highest among children in the youngest age group of 0-4, followed by children aged 5-14 and then by those aged 15-17. This is surprising in view of the fact that the child support grants did not, at the time of the survey, extend to the oldest group, so one would have expected households containing only older children to perhaps experience more poverty. The analysis confirmed that the poverty rate remains far higher among African and coloured than Asian and white children. These were little difference in the poverty rate for girls versus boys. The child poverty headcount emerged as far higher in rural than urban areas, and large variation was found across provinces. In Limpopo, the province with the highest child poverty headcount, the rate was 78%. In the Western Cape, the province with the lowest, it was 37.9%. The child poverty rates in Eastern Cape and KwaZulu-Natal were similar to Limpopo. It is important to address child poverty for targeting resources, to understand variations in the shares of poor children across different ages, racial groups, gender and geographical location. In this regard it was found that: - African children constitute 72.5% of the total number of poor children, and coloured children 24.2%.
- Nearly two thirds of children identified as poor live in rural as opposed to urban areas.
- Whilst the headcount poverty rate emerged as highest in Limpopo, the poverty share was found to be much higher in more populous provinces, particularly KwaZulu-Natal and the Eastern Cape, which together were found to contain 46% of poor children.
- Western Cape has the smallest share of poor children (5%).
| Figure 1: Provincial shares of poor children in South Africa |  |
ImplicationsNo firm conclusion on trends in child poverty over time can be drawn from this analysis. This is firstly, because rigorous comparison of findings from earlier data sets is undermined by differences in survey data collection methods, and secondly because of different welfare measures used across studies. The analysis confirms the need for government to target spending on poor children. Amongst provinces, KwaZulu-Natal, Limpopo and Eastern Cape are still most in need of resources to address child poverty. It suggests that rural areas, the African and Coloured populations, should continue to receive the bulk of attention in child poverty alleviation efforts. The analysis also finds support for government's current policy stance of prioritizing children in their earliest years. Judith Streak, is a senior research manager in the Child, Youth, Family and Social Development programme, and Derek Yu and Servaas van der Berg are both at the Department of Economics, University of Stellenbosch. For a more detailed report, see Streak, S. Yu, D, & Van der Berg, S, How invariant is South African Child Poverty to the choice of equivalence scale or poverty measure? Stellenbosch Working Paper Series No.WP13/2008.
|
