In just a matter of weeks, the coronavirus disease-2019 (COVID-19) pandemic has led to huge societal public health and economic challenges worldwide. The clinical effects of COVID-19 on young children are uncertain when compared with older age groups, with lower morbidity and mortality rates and no conclusive evidence supporting transmission during pregnancy; however, there is emerging evidence of increasing rates of child hyperinflammatory shock.1-3 Research on the effects of prior pandemics and disasters clearly indicates that there will be both immediate and long-term adverse consequences for many children, with particular risks faced during early childhood, when brain architecture is still rapidly developing and highly sensitive to environmental adversity.

The South African government offers three main services to improve the developmental potential and overall well-being of the country's children: cash grants, health care, and education. There is research to suggest that receiving supports to the household in multiple domains simultaneously might improve children's development over and above improvements to development that are typically experienced when receiving a single developmental support, however, no research to date has empirically examined the patterns by which these services may be received, the associated quality of these services, the multilevel characteristics of children and their households that might predict receiving particular patterns of services, or the relationship between service receipt and children's academic and cognitive development in a natural (non-experimental) environment in South Africa. This paper uses three-step latent class analysis to explore patterns of receipt of the three primary government-provided services, the child-, caregiver-, household-, and community-level factors that predict receipt of these patterns, and the associations between service receipt patterns and the academic and cognitive outcomes of low-income 7–10year old children living in peri-urban and rural regions of KwaZulu-Natal, South Africa, while accounting for potential sources of selection bias. Results revealed three service receipt patterns: primarily education, primarily cash grants and education, and all three services, showed that household level factors were most predictive of service receipt such that the most economically disadvantaged households were likely to receive the most services, and found that children receiving more services had poorer math outcomes as compared to children receiving fewer services.