Distributional effects of Oportunidades on early child development

The Mexican Oportunidades program is designed to increase human capital through investments in education, health, and nutrition for children in extreme poverty. Although the program is not expressly designed to promote a child's cognitive and non-cognitive development, the set of actions carried out by the program could eventually facilitate improvements in these domains. Previous studies on the Oportunidades program have found little impact on children's cognition but have found positive effects on their non-cognitive development. However, the majority of these studies use the average outcome to measure the impact of the program and thus overlook other "non-average" effects. This paper uses stochastic dominance methods to investigate results beyond the mean by comparing cumulative distributions for both children who are and children who are not aided by the program. Four indicators of cognitive development and one indicator of non-cognitive development are analyzed using a sample of 2595 children aged two to six years. The sample was collected in rural communities in Mexico in 2003 as part of the program evaluation. Similar to previous studies, the program is found to positively influence children's non-cognitive abilities: children enrolled in the program manifest fewer behavioral problems compared with children who are not enrolled. In addition, different program effects are found for girls and boys and for indigenous and non-indigenous children. The ranges where the effect is measured cover a large part of the outcome's distribution, and these ranges include a large proportion of the children in the sample. With regard to cognitive development, only one indicator (short-term memory) shows positive effects. Nevertheless, the results for this indicator demonstrate that children with low values of cognitive development benefit from the program, whereas children with high values do not. Overall, the program has positive effects on child development, especially for the most vulnerable, who are at the bottom of the distribution.