Enrolment ratios and related puzzles in developing countries: Approaches for interrogating the data drawing from the case of South Africa

• Population-enrolment discrepancies common in developing countries.
• In South Africa, young population seems over-estimated by as much as 14%.
• Not adjusting the data results in significant monitoring and financing distortions.
• Creative analysis of available school and household data can clarify the situation.
• Demographers and education planners should collaborate more closely.

Available developing country enrolment ratios are frequently inaccurate. This results in poorly informed policy discourses. Underlying data problems can undermine funding systems. The magnitude of the problem is described, partly through reference to UNESCO and UNICEF ratios. Steps for remedying the situation within one country are proposed, on the basis of an existing South African study, and some new data analysis of South African, Brazilian, and international data. Remedies must be sensitive to local contexts, but proposed techniques for detecting fraud in the enrolment microdata, and comparing age structures across different datasets, seem relevant for different contexts.