Poverty and hunger have been recognized as part of South Africa’s development challenges. These
issues are also foremostly ranked in the Sustainable Development Goals (SDGs). This study
constructed indicators of children’s multidimensional poverty index (MPI) using the Alkire-Foster
and fuzzy set approaches and examined its pro-poorness. The data were the General Household
Survey (GHS) for 2017, 2018 and 2019, and the 2016 Demographic and Health Survey (DHS). The
MPI was decomposed and analyzed with Tobit regression model, while children’s health outcomes
were analyzed with logistic regression model. The pro-poorness of multidimensional wealth index
(MWI) was analysed with Growth Incidence Curve (GIC), Pro-Poor Growth Index (PPGI) and
Poverty Equivalent Growth Rate (PEGR).
The results showed that children from Eastern Cape had the highest Alkire-Foster (AF) MPIs of 0.34.
0.33, and 0.29 in 2017, 2018, and 2019, respectively, while KwaZulu Natal had 0.31, 0.30, and 0.29.
In the combined dataset, AF MPIs were 0.26, 0.25, and 0.28 in 2017, 2018 and 2019, respectively.
Similarly, Eastern Cape had the highest average fuzzy sets child MPI (0.12), followed by KwaZulu-
Natal (0.10) and Mpumalanga (0.10). The Tobit regression results revealed that children from Eastern
Cape province had significantly higher (p<0.05) AF and fuzzy MPIs in 2017, 2018 and in the
combined dataset. Residence in traditional areas significantly increased (p<0.05) fuzzy MPIs across
the periods, while farm residence significantly increased AF and fuzzy MPIs. Being a Coloured,
Indian/Asian and White child and biological children of the households’ heads significantly reduced
MPIs (p<0.05).
The logistic regression results showed that the chance of stunting among the 0-5 significantly
increased (p<0.05) with health insurance, sharing toilets, number of living children and residence in
the Northern Cape, Free State and Gauteng provinces, while being discharged the same time with the
mothers and being a boy child reduced it. Among the 6-23, stunting significantly increased (p<0.05)
among male children, but reduced with residence in Eastern Cape, Free State, breastfeeding, milk and
dairy products’ consumption, birthweight and maternal years of education. Moreover, the change of
stunting among the 24-59 months old children reduced (p<0.05) with residence in urban areas and
birthweight but increased with residence in the Free State and KwaZulu Natal. The chance of wasting
also reduced among children 0-5 and 24-59 with birthweight, and dairy and milk products’
consumption reduced it among the 0-5. Among the 0-5 and 6-23, the number of living chidren,
increased the chances of wasting. The chance of being underweight among 0-5 children reduced with
consumption of dairy and milk products, while birthweight and mother being employed reduced it
among the 24-59 children. Among the 6-23, underweight increased with number of living children, and residence in KwaZuluNatal, North West, Gauteng, and Limpopo provinces, but decreased with
consumption of snacks and fruits and vegetables.
On the examination of pro-poor growth, the fuzzy MWIs were pro-poor between 2017 and 2018 with
PPGI of 1.33, 1.84 and 2.56 for poverty incidence, depth and severity, respectively, but only pro-poor
for poverty incidence in 2018/2019 with PPGI of 1.714. The fuzzy MWIs were largely pro-poor
among Black/African and White children and those who resided in traditional areas, Eastern Cape,
North-West, and Mpumalanga across different poverty measures between 2017 and 2019.
It was concluded that although interventions to reduce poverty in South Africa had shown some level
of pro-poorness, multidimensional poverty is still high among children. Also, MPI computation using
the AF and fuzzy set can have different and the same policy implications. It was among others
recommended that government’s efforts for poverty alleviation and enhancement of health outcomes
among South African children should focus on nutrition and marginal reforms that target African and
coloured children and those residing in Eastern Cape and KwaZulu Natal provinces. An upward
review of social grants will also benefit the poorest among the children.