How Effective are Cash Transfer Programmes at Improving Nutritional Status?
Cash transfer programmes are a widely applied social protection scheme that has achieved successes in fighting poverty worldwide. A large literature has sprung up around these programmes, yet the relationship between cash transfer programmes and the nutritional status of recipient children is unclear. Can cash transfers be counted on to improve child health, or are mediating characteristics important? In particular, conditionalities are expensive: are conditional cash transfer programmes more successful than unconditional programmes?
To answer these questions, we reviewed the literature linking nutritional status to interventions involving cash transfers in developing countries. A systematic, but not exhaustive, search identified over 30,000 articles from which we identified 18 studies of 15 social assistance programmes that reported impacts on height for age and found that programmes diverged greatly in their effectiveness. To do a more detailed analysis of the pathways by which programmes succeeded, we gathered data on a variety of study, programme, recipient, and country characteristics. Statistical meta-analysis using this data allowed us to isolate some of the factors contributing to programme success.
Our key findings include higher marginal effects in the most disadvantaged areas and that girls benefit more than boys in terms of height for age measures. We find higher marginal effects in countries with poorer health care systems. We also find that in our sample, conditional programmes accomplish about the same as unconditional programmes: the difference is not statistically significant.
However, this hides an important dichotomy. Conditionalities with health components are statistically indistinguishable from unconditional programmes, while other types of requirements strongly inhibit child growth. Because of the ambiguity and potential for both help and harm, we conclude that programme evaluation must be integrated into such interventions, and encourage the further accumulation of evidence on this topic. We anticipated that more programs would have looked at impacts on height for age, but comparatively few programs examined this outcome. One important outcome of our search is a list of evaluated programmes which can now be used as a starting point for future research qualitatively investigating causal pathways from social protection interventions to child growth behind the broad correlations we have identified.