We argue that non-experimental impact estimators will continue to be needed for evaluations of interventions in developing countries as social experiments, for various reasons, will never be the most preferred approach. In a survey of four studies that empirically compare the performance of experimental and non-experimental impact estimates using data from development interventions, we show that the preferred non-experimental estimators are unbiased. We try to explain the reasons why the non-experimental estimators perform better in the context of development interventions than American job-market interventions. We also use the survey as a source for suggestions for implementation and assessment of non-experimental impact evaluations. Our main suggestion is to be more careful and precise in the formulation of the statistical model for the assignment into the program and also to use the assignment information for model-based systematic sampling.

Malaria is a major cause of morbidity and mortality in sub-Saharan Africa. It is also a dynamic contributor to poverty through its effects on children’s cognitive development. This paper examines the degree to which malaria in early childhood impacts on educational achievement in later childhood. The substantial decline in malaria in the region over recent years allows an assessment of its impact to be made. Focusing on Tanzania, we combine data from the Malaria Atlas Project and the 2010–2014 Uwezo household surveys (N = 246,325). We relate the district-level risk of malaria in a child’s year of birth to his/her performance in tests of acquired cognitive skills (literacy and numeracy). For causal identification, we rely on differences across districts in the pace of decline in malaria prevalence occurring over the last 15 years. We control for time-invariant district level, age, birth cohort and survey year effects, as well as district-level trends and individual and household-specific factors. In addition, we use sibling variation in birth-year exposure to malaria to strengthen our identification. A ten percentage-point decrease in malaria prevalence in birth year is associated with a 0.06 standard deviation (p = 0.000) increase in English literacy achievement. This estimate is comparable in magnitude to education intervention programs with very large effects. Our results are robust to a large number of sensitivity analyses. We find no statistically significant effects of birth-year malaria exposure on attainments in numeracy and Kiswahili, and we argue that this is probably attributable to strong ceiling effects in these test scores. We conclude that in Tanzania malaria is an important factor in geographical variation in English literacy. This indicates that malaria is a significant public health challenge to educational achievement in this country, and probably in other regions with malaria.

This study has been prepared within the context of the UNU-WIDER 2014-18 work programme.