| This dissertation draws quasi-experiments from the history of American education to estimate the returns to public investments in the formal schooling of young children.; The first chapter analyzes the returns to preschool provision. During the late 1960s and early 1970s, most states in the American South began funding public school kindergartens for the first time, contributing to sizable and rapid increases in kindergarten attendance in the region. Using variation in the timing of state funding initiatives, I find large reductions in grade retention rates among southern blacks aged five near the time of the reform. Similar analyses show a smaller effect on grade progression for southern whites, and little, if any, effect on early high school dropout rates for children of either race. Effects on black grade progression are similar to those observed for targeted, high quality early interventions, though not large enough to justify universal access to the program.; The second chapter, co-authored with Ethan Lewis, uses school entry legislation to estimate the effect of a year of schooling on standardized test performance during high school. Because entry laws specify an exact date, they generate sharp differences in average age at school entry, and therefore in average completed schooling among enrolled students of nearly the same chronological age. Constructing comparisons using this regression discontinuity and the National Longitudinal Survey of Youth, we find that a year of schooling during late adolescence yields small gains to performance on the Armed Forces Qualifying Test. Effects are sufficiently small to argue that age-adjusted scores yield an unbiased measure of skill upon labor market entry and that a year of schooling has a larger effect on attainment when a child is young.; The third and final chapter estimates the reliability of the standard proxy for grade repetition, whether a child is enrolled below grade given his age. I find that roughly 80 percent of students are correctly classified by the proxy. School entry legislation plays a key role in misclassification, which will impart severe attenuation bias on regression coefficients in applications where below grade is used as an outcome or explanatory variable. |
