Income poverty & grantsIncome poverty & grants

Children living in poverty

Author/s: Katharine Hall
Date: March 2022

Definition

This indicator shows the number and share of children living in households that are income-poor.

Four different poverty lines are used:

  • An upper-bound poverty line that allows just enough money for basic nutrition and other essentials such as clothing (2021 value = R1335);
  • A lower-bound poverty line that allows enough for essentials such as clothing but only if some nutritional costs are sacrificed (2021 value = R890);
  • A food poverty line that only allows enough for basic nutrition, and no other essentials (2021 value = R624);
  • An ultra-low international poverty line, linked to the SDGs (US$1.90 or 2021 PPP value = R412).
The values of the South African poverty lines are increased in line with inflation each year, so that the real values remain constant.

What do the numbers tell us?

This indicator shows the number and share of children living in households that are income-poor. As money is needed to access a range of services, income poverty is often closely related to poor health, reduced access to education and physical environments that compromise personal safety.

International law and the Constitution recognise the link between income and the realisation of basic human rights and acknowledge that children have the right to social assistance (social grants) when families cannot meet children’s basic needs. Income poverty measures are therefore important for determining how many people need social assistance, and for evaluating the state’s progress in realising the right to social assistance and in reducing poverty.

No poverty line is perfect. Using a single income measure tells us nothing about how resources are distributed between family members, or how money is spent. But this measure does give some indication of how many children are living in households with severely constrained resources.

The measure used is the Statistics South Africa “upper-bound” poverty line, set at R779 per person per month in 2011 prices. Poverty lines increase with inflation and in 2019 the real value of the upper-bound line was R1227.1 Per capita income is calculated by adding all reported income for household members older than 15 years, and social grants received by anyone in the household, and dividing the total household income by the number of household members.

Statistics South Africa proposed two other poverty lines:

  • A “lower-bound” poverty line is calculated by adding to the food poverty line the average expenditure on essential non-food items by households whose food expenditure is below but close to the food poverty line. The value of the lower-bound poverty line in 2011 prices was R501 per person per month (R810 in 2019 prices). Those living below this line would not be able to pay for the minimum non-food expenses or would be sacrificing their basic nutrition in order to pay for non-food expenses.

  • A “food” poverty line is based on the cost of the minimum nutritional requirement of 2,100 kilocalories per person per day, without any allowance for non-food basic necessities. The value of the food poverty line in 2011 prices was R335 per person per month (R561 in 2019). Anyone living below this line will be malnourished and their health and survival will be at risk.

We use the upper-bound poverty line as our main indicator for tracking child poverty as this is linked to the minimum requirement for basic nutrition as well as other basic needs such as clothing and shelter. In other words, this is the only poverty line that meets the minimum requirement for children’s basic needs.

South Africa has very high rates of child poverty. In 2019, 56% of children lived below the upper-bound poverty line. Income poverty rates have fallen substantially since 2003, when 78% (14.1 million) children were defined as “poor” at this income threshold. The reduction in the child poverty headcount is partly the result of a massive expansion in the reach of the Child Support Grant over the same period. Although there have been reductions in the child poverty rate, large numbers of children still live in poverty: in 2019, 11.2 million children lived below the upper-bound poverty line.

There are substantial differences in poverty rates across the provinces. Using the upper-bound poverty line, nearly three-quarters of children in the Eastern Cape and Limpopo are poor, and child poverty rates are also high in the Free State, Mpumalanga, North West and KwaZulu-Natal (all over 60%). Gauteng and the Western Cape have the lowest child poverty rates – at 35% and 27% respectively. Child poverty remains most prominent in the rural areas of the former homelands, where 76% of children are below the poverty line. The urban child poverty rate, by contrast, is 41%.

There are glaring racial disparities in income poverty: while 61% of African children lived in poor households in 2019, and 33% of Coloured children were defined as poor, only 5% of White children lived below this poverty line. There are no significant differences in child poverty levels across gender or between different age groups in the child population.

Using Statistics South Africa’s lower-bound poverty line (which does not provide enough for basic essentials), 44% of children (8.9 million) were poor in 2019, and 33% (6.6 million children) were below the food poverty line, meaning that they were not getting enough nutrition.

The international ultra-poverty line used to track progress towards the Sustainable Development Goals (SDGs) is $1.90 per person per day. This translated to R385 per person per month in 2018, using the IMF purchasing power parity conversion. This poverty line is extremely low – below survival level – and is not appropriate for South Africa. No child should be below it. In 2003, 52% of children (9.3 million) lived below the equivalent of the SDG poverty line. By 2019, this decreased to 22% but was still high in real numbers (4.3 million).

The Sustainable Development Goals replaced the Millennium Development Goals in 2015 and set the global agenda for development by 2030. Target 1.1 is to eradicate extreme poverty using the same international poverty line of $1.90 per person per day. Target 1.2 is that by 2030 countries should reduce by at least half the proportion of men, women and children of all ages living in poverty in all its dimensions, according to national definitions. In terms of income poverty, this would mean reducing the number of children below the upper-bound poverty line by at least 6 million, from a baseline of 12.2 million in 2015.


1 Statistics South Africa (2021) National Poverty Lines. Statistical Release no. P0310.1. Pretoria: Statistics South Africa.


Technical notes

The General Household Survey asks a set of questions to establish whether household members over 15 years are economically active. For those who are economically active and report their earnings, these amounts are standardised to monthly values.

For those who report earnings in income bands rather than discrete amounts, each income bracket is split into deciles for those who indicated an income in that bracket, and a uniform distribution of income is assigned within each income bracket decile, for those who indicated an income in that bracket.

For those who are economically active but did not provide a discrete income amount or indicate an income bracket (unspecified/refused), the median income for men and women in each population group is allocated. The medians are calculated separately for each year.

Total household income from earnings is calculated as the total earnings for all household members over 15 years. Total household income from social grants is calculated by allocating the grant amounts for that year for each type of grant reported to be received by household members. Total household income is derived by adding total income from earnings and grants. Per capita income is calculated by adding all reported income for household members older than 15 years, including social grants, and dividing the total household income by the number of household members.

The three South African poverty lines were set in 2011 Rand values. This are inflated using headline CPI reported by Statistics South Africa for each year. Per capita income is calculated by dividing total household income equally by the number of household members.

In addition to the three poverty lines (upper, lower and food poverty) is the $1.25-a-day international poverty line. This poverty line is used by the World Bank and other international groups for monitoring poverty rates in developing countries and is used as an ultra poverty line for reporting against the MDGs and SDGs.

There are many limitations to working with poverty lines and this method will almost certainly result in an over-estimation of the poverty rate because the General Household Survey is not designed to obtain detailed income and expenditure data. For example, it does not record information on income from dividends and investments, or detailed information about income from remittances and other non-labour sources.

Strengths and limitations of the data

The numbers are derived from the General Household Survey, a multi-purpose annual survey conducted by the national statistical agency, Statistics South Africa, to collect information on a range of topics from households in the country’s nine provinces.

The GHS uses a Master Sample frame which has been developed as a general-purpose household survey frame that can be used by all other Stats SA household-based surveys that have design requirements that are reasonably compatible with the GHS. The sample is drawn from Census enumeration areas using a stratified two-stage design with probability proportional to size sampling of PSUs in teh first stage, and sampling of dwelling units with systematic sampling in the second stage. The resulting sample consists of just over 20,000 households with around 70,000 individuals, and should be representative of all households in South Africa. It is also designed to be representative at provincial level and within provinces at metro/non-metro levels and three geography types (urban areas, rural areas under traditional authority, and farms).

The sample consists of households and does not cover other collective institutionalised living-quarters such as boarding schools, orphanages, students’ hostels, old-age homes, hospitals, prisons, military barracks and workers’ hostels. These exclusions probably do not have a noticeable impact on the findings in respect of children.

Changes in sample frame and stratification
Since 2014 the GHS has been based on the 2013 master sample that that is, in turn, based on information collected during the 2011 Population Census. The previous master sample for the GHS was used for the first time in 2008, and the one before that in 2004. These again differed from the master sample used in the first two years of the GHS: 2002 and 2003. Thus there have been four different sampling frames during history of the annual GHS, with the changes occurring in 2004, 2008 and 2013. In addition, there have been changes in the method of stratification over the years. These changes could compromise comparability across iterations of the survey to some extent, although it is common practice to use the GHS for longitudinal monitoring and many of the official trend analyses are drawn from this survey.

Weights
Person and household weights are provided by Stats SA and are applied in Children Count analyses to give population estimates on the indicators. The GHS weights are derived from Stats SA’s mid-year population estimates for the relevant year. The population estimates are based on a model that is revised from time to time when it is possible to calibrate the population model to Census data and larger population surveys such as the Community Survey.

In 2017, Stats SA revised its demographic model to produce a new series of mid-year population estimates and the GHS data were re-released with the revised population weights. All the Children Count indicators were re-analysed retrospectively, using the revised weights provided by Stats SA, based on the 2013 model. The estimates are therefore comparable over all years. The revised weights particularly affected estimates for the years 2002 – 2007.

The 2017 model drew on the 2011 census, along with vital registration, antenatal and other administrative data, but was a “smoothed” model that did not mimic the unusual shape of the age distribution found in the census. The results of the 2011 census were initially distrusted because it seemed to over-count children in the 0 – 4 age group and under-count children in the 4 – 14-year group. It is now thought that the fertility rates recorded in the 2011 population census may have been an accurate reflection of demopraphic trends, with an unexplained upswing in fertility around 2009 after which fertility rates declined again gradually. Similar patterns were found in the vital registration data as more births were reported retrospectively to the Department of Home Affairs, and in administrative data from schools, compiled by the Department of Basic Education. In effect, this means that there may be more children in South Africa than appear from the analyses presented in these analyses, where we have applied weights based on a model that it is now known to be inaccurate.

Stats SA has subsequently developed a new population model - the 2022 series, which provides revised mid-year population estimates back to 2002 and projected to 2032. However, the GHS series has not yet been reweighted.The population estimates in Children Count are therefore based on weights derived from outdated population model (2017). It is not yet clear when and how the population model will be revised again following the 2022 Census, as there are concerns around census under-count and plausibility of its findings.

Disaggregation
Statistics South Africa suggests caution when attempting to interpret data generated at low level disaggregation. The population estimates are benchmarked at the national level in terms of age, sex and population group while at provincial level, benchmarking is by population group only. This could mean that estimates derived from any further disaggregation of the provincial data below the population group may not be robust enough.

Reporting error
Error may be present due to the methodology used, i.e. the questionnaire is administered to only one respondent in the household who is expected to provide information about all other members of the household. Not all respondents will have accurate information about all children in the household. In instances where the respondent did not or could not provide an answer, this was recorded as “unspecified” (no response) or “don’t know” (the respondent stated that they didn’t know the answer).

For more information on the methods of the General Household Survey, see the metadata for the respective survey years, available on Nesstar or DataFirst