Who Pays Child Maintenance in South Africa? Race, Age and Income Hold the Answer

Across South Africa, roughly 2.13 million children are growing up without their fathers in the household, and only a fraction of those absent fathers contribute financially to their children’s upbringing. Although the country’s Maintenance Act of 1998 gives courts sweeping powers to compel parents to support their children, enforcement has been inconsistent, and the burden of raising children after separation falls disproportionately on mothers and maternal relatives. A new longitudinal study published in SN Social Sciences offers one of the most detailed statistical portraits yet of the men who do pay child maintenance, and the findings reveal a landscape shaped as much by history and social expectation as by money itself.

Researchers M. Muchemwa and C. Odimegwu, both based in the Demography and Population Studies Programme at the University of the Witwatersrand, drew on the National Income Dynamics Study, or NIDS, a nationally representative panel survey that has tracked thousands of South Africans across five waves between 2008 and 2017. From a pooled dataset of more than 28,000 male person-wave observations, the authors isolated 718 observations from 636 men who reported making cash contributions to children living outside their households. Because the same men were surveyed repeatedly over nearly a decade, the researchers turned to a statistical technique designed for exactly this kind of repeated-measures data: generalized estimating equations, or GEE.

The methodological choices in the study are worth unpacking, because they shape what the numbers actually mean. Child maintenance payments are continuous monetary amounts that are strongly right-skewed, meaning most men pay modest sums while a few pay very large ones. To model such data, the authors specified a Gamma distribution with a log link function, a combination that ensures predicted payments remain positive and expresses coefficients as multiplicative, or percentage, effects rather than absolute changes in rands. Because repeated observations from the same individual violate the independence assumption underlying ordinary regression, the GEE framework explicitly models the within-person correlation across survey waves. Three candidate correlation structures were compared using the quasi-likelihood under the independence model criterion, and the unstructured structure, which allows the correlation to differ between every pair of waves, produced the lowest QIC value of 14,483.032 and was retained for the final analysis.

The headline result is stark: race, age, occupation and income were all significantly associated with the amount of child maintenance paid, even after controlling for education, employment status, marital status, debt, socioeconomic status, health, religion, province and household size. Indian and White fathers paid significantly more than Black fathers, with a regression coefficient of 1.0542 that was significant at the one percent level and consistent across all three correlation specifications. Age mattered too, with men aged 53 to 57 paying more than younger men aged 18 to 28. And income emerged as a powerful driver: the Gamma model with a log link indicates that every additional 1,000 rand in household net income was associated with roughly a 3.4 percent increase in the amount paid, a coefficient of 0.0000345 that held at the one percent significance level.

One counterintuitive finding deserves attention. Men in unskilled occupations paid an estimated 44.8 percent more than men with no occupation at all, after adjusting for other characteristics. This suggests that having any foothold in the labour market, even at the lowest rung, enables financial contribution in a way that complete joblessness does not. It also hints at the psychological dimension of the provider role: men in work may feel a stronger sense of obligation and capability, while unemployed fathers, unable to meet deeply entrenched social expectations of provision, may withdraw altogether. Previous South African research has documented how feelings of failure and emasculation push men away from their children, and how mothers or maternal families sometimes deny fathers access when they cannot provide material support.

The racial pattern in the data cannot be divorced from South Africa’s history. The authors interpret their findings through the father involvement framework first proposed by Michael Lamb and colleagues in 1985, which decomposes fatherhood into three dimensions: interaction, availability and responsibility. Child maintenance payments serve here as an observable indicator of the responsibility dimension, the father’s role in ensuring that the child’s needs are met. Within this framework, the racial gap in payments reflects structural inequalities that constrain some fathers’ capacity to enact financial responsibility. Black fathers in South Africa remain disproportionately affected by unemployment, labour migration and economic marginalisation, legacies that limit both their financial capacity and their ability to maintain consistent involvement in their children’s lives. Yet the authors are careful to note that money is not the whole story: even wealthy Black fathers have faced legal action for non-payment, indicating that willingness, relational dynamics and enforcement all shape behaviour independently of capacity.

The age gradient tells a complementary story about life stage. Older men are more likely to have accumulated resources, achieved employment stability and internalised the social expectations attached to responsible fatherhood, enabling them to fulfil the responsibility dimension of involvement. Younger men, still transitioning into stable economic and social roles, may lack both the means and the readiness to assume this responsibility. Comparable findings from Finland, where a register-based study showed older parents paying higher maintenance amounts even after adjusting for income, suggest the pattern is not unique to South Africa but reflects a broader developmental logic of paternal responsibility.

The study’s technical rigour is matched by an unusually candid discussion of its limitations. The NIDS Contributions Given module does not distinguish between court-ordered formal maintenance and informal, non-regulated support, so the analysis captures all reported financial transfers to children rather than legal compliance specifically. More importantly, respondents who reported no contribution to anyone outside their household were never asked about non-resident children, making it impossible to identify fathers who should pay but do not. The analysis therefore describes the payment amounts among observed contributors, not the probability of paying. The sample was also unbalanced, with 559 men contributing a single observation, 72 contributing two and only five contributing three, and the racial comparison rests on just 23 Indian or White respondents against 620 African respondents, limiting the precision of that comparison. A sensitivity analysis using an inverse hyperbolic sine transformation, which allowed five zero-payment observations to be retained, confirmed the robustness of the positive associations for race, income, age and survey year.

What emerges from this study is a picture of child maintenance in South Africa as a dynamic, evolving behaviour rather than a static obligation. Payments were significantly higher in 2012, 2015 and 2017 than in the 2008 baseline wave, suggesting that financial contributions have grown over time even as the underlying inequalities persist. The policy implications are concrete: the authors call for programmes that improve employment opportunities, awareness campaigns that emphasise the importance of fathers’ financial contributions, and community-based fatherhood initiatives, including counselling and parenting workshops, that engage absent fathers and address the cultural and economic barriers that keep them from supporting their children. With evidence that only 40.5 percent of children with living fathers receive financial support from them, the stakes of translating these statistical insights into action could hardly be higher, both for the millions of children whose wellbeing depends on paternal support and for a society still reckoning with the structural forces that shape who gets to be a provider.

Subject of Research: Demographic and socioeconomic determinants of child maintenance payments among men in South Africa

Article Title: Demographic and socioeconomic characteristics of men paying child maintenance in South Africa: a generalized estimating equation approach

Article References: Muchemwa, M., & Odimegwu, C. (2026). Demographic and socioeconomic characteristics of men paying child maintenance in South Africa: a generalized estimating equation approach. SN Social Sciences, 6(10), Article 480. https://doi.org/10.1007/s43545-026-01739-4

Image Credits: AI Generated

DOI: 10.1007/s43545-026-01739-4

Keywords: child maintenance, child support, fatherhood, South Africa, National Income Dynamics Study, generalized estimating equations, longitudinal analysis, father absence, income inequality, race and ethnicity, social policy, demography