“The Tamil Nadu Integrated Nutrition Project…is a growth monitoring, food supplementation, and intensive nutrition counseling programme. … In terms of the project’s impact, TINP was labeled a success by the World Bank. This followed the 1986 midterm evaluation, which found a reduction in rates of severe undernutrition. … The success of TINP was used as proof that a ‘silver bullet’ to address undernutrition existed. … The assessment at the end of [the second] phase was less glowing. The official TINP-I terminal evaluation showed a decline in the prevalence of severe undernutrition but an increase in mild and moderate undernutrition. … Interview with key Bank official revealed that there were major shortcomings with the baseline and midterm evaluations. … Other observers raised wider concerns…. The claims of successful growth monitoring…have been based on anecdotal and impressionistic evidence. … By 1999, the Bank had loaned over US$750 million to India despite the lack of substantive evidence that the design and implementation of TINP were effective. …
“Why does all this matter? … First, it matters for the potential beneficiaries of the nutrition program in question. Second… despite the lack of rigorous evaluation, TINP has come to represent the ultimate success story of the World Bank’s lending for stand-alone nutrition projects.” [Sridhar, The Battle Against Hunger, p2-5, 2008]
This isn’t a problem unique to the World Bank (but obviously the WB isn’t exempt). Billions go into programs about which all we know is “anecdotal and impressionistic.” Good evaluation is on the rise, though…
David Evans' personal blog 