Part I: Charter schools are publicly funded schools that, in exchange for expanded accountability, receive more autonomy and experience fewer regulations than traditional public schools. Previous work has found mixed evidence on the impacts of charter schools on both charter and non-charter students. However, these studies focus almost exclusively on test scores and may not fully account for endogenous movements of students and location of schools. Using data from an anonymous large urban school district, I investigate how charter schools affect both charter and non-charter students. In the first chapter I look at the effects of charter schools on charter students. I find that charter schools generate improvements in student behavior and attendance but the effects on test scores differ by subject. These results change little after correcting for selection based on changes in outcomes, endogenous attrition, or persistence. In the second chapter I investigate whether charters affect students who remain in non-charter schools. I find little evidence of charter school impacts on non-charter students. However I also find evidence that regressions using school fixed-effects may be biased. Changes in peer characteristics do not appear to play a large role in the charter impacts.

Part II: Strains on the Federal budget have created worries that Federal funding of aid for higher education will fall in the future. If this happens, state governments will need to try to re-allocate their higher education spending more efficiently. One possible way to do this would be to shift funding away from public provision towards demand-side subsidies so that more students could attend private colleges. However, this will only work if private colleges provide benefits to students over public. In order to answer this question, I use highly detailed and rich data sets to assess whether there are benefits to attending private colleges over public ones. For males the wage return is small and statistically insignificant during their twenties but statistically significant at around 11 percentage points by their mid-thirties. For females the wage returns are negative and statistically insignificant. Both males and females exhibit increases in the likelihood of finishing a bachelor's degree.