Despite a growing emphasis on leadership skill in the labor market and higher education, few economists have studied the development of and return to leadership skill. This paper uses data from the National Education Longitudinal Study of 1988 to estimate the impact of leadership experience in high school on one of its most likely consequences - subsequent educational attainment.  Using three estimation methods to address the non-random selection of students into leadership positions, I find a large, positive impact of high school leadership on subsequent educational attainment that is significant in both a statistical and an economic sense.  The most conservative estimates suggest that students who are leaders in high school complete 0.35 more years of education than their non-leader peers.  High school leadership is also predicted to increase the probability of attending a post-secondary institution by a minimum of 5 percent and to increase the probability of holding a college degree by at least 9.5 percent.  Interestingly, similar to many empirical studies on the return to schooling, the instrumental variables estimates, which rely on variation in school leadership opportunities, birth order and twin indicators for identification, are two to three times the magnitude of the corresponding ordinary least squares and propensity score estimates.  This result suggests that failure to control for unobserved heterogeneity leads to estimates that understate the true impact of leadership.  An alternative interpretation put forth by Card (2001) is that the IV estimates reflect a relatively high return to leadership by the small group of students who are affected by the instruments.  I also find evidence of a differential impact of leadership for students from low versus high income households.  In terms of total years of education and post-secondary attendance, high school leadership appears to disproportionately benefit students from lower income households, while with respect to college graduation, leaders from high income households seem to derive at least as great or greater benefit from their leadership experience than their low-income peers.

"Revisiting Gruber (2004): Does growing up in a unilateral divorce regime really lead to negative later-life outcomes?" Working Paper, Department of Economics, UNC-CH

Abstract:
While divorce rates peaked in the 1980s, an estimated 33% of first marriages now end in divorce or separation within the first 10 years of marriage and debates over the effects of divorce laws are still very much alive. While there is a plethora of research documenting the effect of divorce on children, there has been relatively little empirical work seeking to assess the effect of the divorce legislation on children and later life outcomes.  This paper seeks to narrow this gap in the literature by revisiting earlier results and by providing updated evidence on the impact of unilateral divorce laws on child outcomes. It builds on Jonathon Gruber’s 2004 paper published in the Journal of Labor Economics, which, in addition to addressing the effect of unilateral divorce on divorce rates, looks at the effect of unilateral divorce exposure both as a child and as an adult on a variety of outcomes including education, income, marriage and suicide. In order to test the robustness of Gruber’s estimates and to assess whether these trends persist with the inclusion of more recent data, I first replicate his results as precisely as possible. Then, I extend Gruber’s analysis by including 2000 census data which was not available at the time of his publication.  While my replication of Gruber is not perfect, for the most part, I am reasonably successful in replicating his major findings using data from the 1960-1990 time period. In my extension, I find many of the updated results found using the 1960-2000 timeframe to be smaller in magnitude and, in many instances, much less precise than those estimated with the original sample. Updated results still provide evidence of a positive, albeit smaller, effect of current unilateral divorce exposure on divorce probabilities.  However, contrary to Gruber’s findings, I find little evidence of strong negative effects of youth exposure to unilateral divorce on later life outcomes once the 2000 data are included.

Abstract:
This paper uses a sequential dynamic empirical model to study the longer term impacts of a young child’s height on subsequent illnesses, entry into school, and height and weight in Cebu, Philippines. Moving a child from the 10th percentile of the age two height distribution to the 90th percentile appears to have moderate impacts on the child’s later physiological and intellectual progress. Taller children at age two have a ten percentage point lower propensity to experience a serious illness between the ages of two and seven (37% versus 47%), and the failure to recognize the endogeneity of the height at age two underestimates this impact by about a factor of four. While taller children do appear to be more likely to receive subsequent immunizations, there appears to be little impact of these immunizations on reducing the incidence of preventable diseases. The failure to control for the endogeneity of the child’s height at age two also overstates by a factor of two its impact on height at age seven, and it also severely overestimates the impact of age two height on a child’s weight at age seven. Taller children at age two do appear to have higher intelligence test scores at age seven, and estimates of this impact do not appear to be influenced by whether or not one controls for the endogeneity of height at age two. Higher stature at age two has a moderate indirect impact on whether a child starts school on time, with much of this impact operating through the lower incidence of disease between the ages of two and seven. The estimated indirect impact would only be half as large if one did not control for the endogeneity of the child’s height at age two. Taken as a group, these results suggest that better nutrition and health behaviors before age two that result in taller children could have important impacts on a child’s subsequent early intellectual development by reducing the child’s susceptibility to diseases after age two and before entering school.