Targeting performance of three large-scale, nutrition-oriented social programs in Central America and Mexico.

Abstract

This study evaluated whether three nutrition-oriented programs in Central America and Mexico have been successful in targeting those households most vulnerable to undernutrition and poverty. For each country, nationally representative data sets were used to estimate cutoff points dividing the population into 10 equal-sized groups according to child anthropometric measurements (age-standardized height) and household income (per capita household expenditures). Households meeting eligibility criteria were then assessed using special baseline surveys or national data obtained before implementation of the program. Children in these households were classified according to national deciles of height-for-age, and households were classified according to expenditure deciles. In spite of markedly differing targeting strategies, each of the programs was well targeted, with 45% (Honduras and Mexico) and 52% (Nicaragua) of children in eligible households coming from the lowest two deciles of the national distributions, and virtually none from the upper two deciles. Similar results pertained to household income. These experiences demonstrate that vulnerable households can be targeted relatively straightforwardly, and that the need to do this does not in every case imply household-level income screening.