Essays on Late Investment in Human Capital

Abstract

This dissertation consists of three essays examining public education's role in the formation of human capital from high school through adulthood. I use quasi-experimental methods assess the returns to three types of targeted educational interventions that develop the skills of adults and older children: (1) providing English language training to adult immigrants and refugees with limited English proficiency, (2) credentialing adult dropouts whose test performance signals they are academically prepared for college-level coursework, and (3) offering disadvantaged high school students the chance to enroll in academically rigorous public charter schools that ascribe to a "No-excuses" educational philosophy.

After decades of research on job training and adult skill formation, labor economists have increasingly neglected programs of research focused on public investments in human capital for low-skilled adults and adolescents in favor of interventions that occur earlier in the life cycle. The principle underlying the "Heckman Equation," that "early investments targeted toward disadvantaged children have much higher returns than later interventions" has quickly become accepted wisdom, driving human capital research, funding, and attention toward early childhood and K-5 education (Heckman, 2006, 2012a).

While Heckman’s conclusions about complementarities in educational investments and diminishing marginal returns of human capital investments over an individual’s life cycle hold true on average, it is not clear that current, low levels of public investment in human capital interventions that begin in late adolescence or adulthood are efficient. In fact, theory suggests that interventions that target adults and older children with latent human capital may have especially high returns (Borjas, 1989; Becker, 1994). This dissertation is organized around three populations that may benefit from such interventions: adult immigrants who do not speak English fluently, high-performing uncredentialed adults and high school dropouts, and underserved high school students in low-income communities.

Chapter 1 (co-authored with Kirsten Slungaard Mumma) leverages randomized lotteries for access to English for Speakers of Other Languages (ESOL) services to estimate the causal effect of being offered a chance to attend ESOL classes on measures of civic engagement and economic integration for adult immigrants in Framingham, Massachusetts. We find that attending ESOL classes more than doubles rates of voter registration and substantially increases annual employer-reported earnings. Chapter 2 describes the national population of "college-ready" GED® testers and uses quasi-experimental variation in eligibility for GED® college readiness designations to examine whether being designated as "college-ready" makes marginally qualifying students more likely to enroll and persist in college. I find no evidence that GED® college readiness benchmarks promote better college outcomes, but in the absence of more detailed test score information these designations offer a simple heuristic to predict short-run college enrollment and persistence. Chapter 3 (co-authored with Matthew Davis) draws on another set of randomized lotteries to measure the effects of attending a so-called "No Excuses" charter high school network in Chicago on students' post-secondary enrollment and success. We find that lottery winners are more likely to matriculate to college, enroll in more selective institutions, and attend for longer than their peers who entered but did not win an enrollment lottery. Assessing the impacts of these three distinct interventions advances knowledge about which types of late human capital investments have high or low returns and why, helping policymakers better identify and create opportunities to efficiently invest in adults' and adolescents' human capital.