South Africa Child Support Grant
South African Social Security Agency / World Bank · South Africa · 2009
Summary
The South Africa Child Support Grant is one of the most thoroughly studied small-value cash transfer programs in the world. The regression discontinuity designs exploiting age cutoff changes allow credible causal identification despite the lack of a formal randomized design. The finding that early receipt — within the first year of life — produces significantly larger long-term effects than late receipt suggests a developmental window: cash transfers are not interchangeable across time. The young adult earnings effects (7.3% higher for early vs. late recipients) indicate that even a small unconditional grant can compound over two decades. The program costs approximately 0.8% of South Africa's GDP and reaches more than a quarter of the country's children.
Research question
"Does early receipt of a small unconditional child support grant improve nutrition, schooling, and health for young children?"
Methodology
Intervention
Small monthly cash transfer (initially R100/month, ~$14) to caregivers of children under 7; eligibility was means-tested; age cutoff expansions created natural experiments comparing early vs. late grant receipt
Assignment
Regression discontinuity (eligibility age cutoff expansions in 2003 and 2008)
Sample size
Multiple cohorts across South Africa; largest study covered ~3 million children
Primary outcome
Nutritional status (height-for-age, weight-for-age); school enrollment; grade attainment; labor market outcomes at ages 18–25
Effect estimate
Child stunting: −3.5% for children enrolled in first year of life; school enrollment at age 7: +4.5 pp; young adult earnings at age 20–25: +7.3% for early recipients
Decision
Program maintained and expanded; grant value and age eligibility increased multiple times; serves approximately 13 million children as of 2023; findings replicated in studies of Kenya and Ethiopia grants
Result
Positive
Child stunting: −3.5% for children enrolled in first year of life; school enrollment at age 7: +4.5 pp; young adult earnings at age 20–25: +7.3% for early recipients
Evidence strength
Moderate
Quasi-experimental design with replication support.
Replication status
Replicated
Institution
South African Social Security Agency / World Bank
Location
South Africa
Year
2009
Policy area
International Development
Mechanism
Cash transfer