Paper # Author Title
Most investigations of the importance of and the determinants of adult cognitive skills assume that (a) they are produced primarily by schooling and (b) schooling is statistically predetermined. But these assumptions may lead to misleading inferences about impacts of schooling and of pre-schooling and post-schooling experiences on adult cognitive skills. This study uses an unusually rich longitudinal data set collected over 35 years in Guatemala to investigate production functions for adult (i) reading-comprehension and (ii) nonverbal cognitive skills as dependent on behaviorally-determined pre-schooling, schooling and postschooling experiences. Major results are: (1) Schooling has significant and substantial impact on adult reading comprehension (but not on adult nonverbal cognitive skills) but estimates of this impact are biased upwards substantially if there are no controls for behavioral determinants of schooling in the presence of persistent unobserved factors such as genetic endowments and/or if family background factors that appear to be correlated with genetic endowments are included among the first-stage instruments. (2) Both preschooling and post-schooling experiences have substantial significant impacts on one or both of the adult cognitive skill measures that tend to be underestimated if these pre- and post-schooling experiences are treated as statistically predetermined in contrast to the upward bias for schooling, which suggests that the underlying physical and job-related components of genetic endowments are negatively correlated with those for cognitive skills. (3) The failure in most studies to incorporate pre- and post-schooling experiences in the analysis of adult cognitive skills or outcomes affected by adult cognitive skills is likely to lead to misleading over-emphasis on schooling relative to these pre-and post-schooling experiences. (4) Gender differences in the coefficients of the adult cognitive skills production functions are not significant, suggesting that most of the fairly substantial differences in adult cognitive skills favoring males on average originate from gender differences in schooling attainment and in experience in skilled jobs favoring males. These four sets of findings are of substantial interest in themselves. But they also have important implications for broader literatures, reinforcing the importance of early life investments in disadvantaged children in determining adult skills and options, pointing to limitations in the cross-country growth literature of using schooling of adults to represent human capital, supporting hypotheses about the importance of childhood nutrition and work complexity in explaining the "Flynn effect" of substantial increases in measured cognitive skills over time, and questioning the interpretation of studies that report productivity impacts of cognitive skills without controlling for the endogeneity of such skills. Download Paper
Early childhood nutrition is thought to have important effects on education, broadly defined to include various forms of learning. We advance beyond previous literature on the effect of early childhood nutrition on education in developing countries by using unique longitudinal data begun during a nutritional experiment during early childhood with educational outcomes measured in adulthood. Estimating an intent-to-treat model capturing the effect of exposure to the intervention from birth to 36 months, our results indicate significantly positive, and fairly substantial, effects of the randomized nutrition intervention a quarter century after it ended: increased grade attainment by women (1.2 grades) via increased likelihood of completing primary school and some secondary school; speedier grade progression by women; a one-quarter SD increase in a test of reading comprehension with positive effects found for both women and men; and a one-quarter SD increase on nonverbal cognitive tests scores. There is little evidence of heterogeneous impacts with the exception being that exposure to the intervention had a larger effect on grade attainment and reading comprehension scores for females in wealthier households. The findings are robust to an array of alternative estimators of the standard errors and controls for sample attrition. Download Paper
The assessment of the impact of social programs is the subject of lively, sometimes heated debate over whether program evaluation is best conducted either by comparing mean outcomes from a randomized intervention or by using econometric techniques with nonrandom samples. This paper contributes to this debate through an examination of PROGRESA, a Mexican anti-poverty and human resource program, on child nutritional status. PROGRESA was randomly assigned at the locality level. However, a shortage in the availability of one component- a nutritional supplement provided to preschool children -led local administrators to exercise discretion in its delivery, systematically favoring those children with poorer nutritional status. While comparisons of mean outcomes suggest that PROGRESA had no or a negative effect on nutritional status, estimates that control for this heterogeneity using child specific fixed effects find that PROGRESA had significant and substantial positive impacts in increasing stature. The long-term consequences of these improvements are non-trivial; its impact working through adult height alone may result in a 2.9% increase in lifetime earnings. Download Paper