Understanding America's unfinished transformation

The United States has long ranked as the world's most educated nation, but this status has slipped over the past thirty years as bachelor's degree (BA) attainment has stagnated and those who do earn a BA are taking longer to do so. This dissertation contributes to a growing body of empirical economic research seeking to understand this stagnation in U.S. collegiate attainment and evaluate public policies that address it. In the first essay, I use Current Population Survey data to document that labor supply among "traditional" undergraduates nearly doubled between 1970 and 2003, rising from an average of 5 hours to almost 10 hours per week. Using multiple data sources to evaluate alternative hypotheses for explaining the change over time, I conclude that credit constraints may be driving the trend. In the second essay, Susan Dynarski and I examine the costs and benefits of complexity in federal financial aid for undergraduates. We use financial aid application data to show that complexity contributes little to the targeting of aid, despite evidence that its administrative and psychological costs are substantial. In the final essay, I examine the impact of the PROMISE scholarship in West Virginia, which provides financial incentives for college student performance. The program could work either by relaxing financial constraints or by inducing additional student effort. Using administrative data, I exploit discontinuities in both the eligibility formula and the timing of implementation to identify program effects. I find significant impacts on key outcomes including graduation. The concentration of impacts at the precise thresholds for annual scholarship renewal suggests that the program works by establishing clear academic goals and incentives to meet them, rather than by simply reducing the cost of college.