The Distributional Effects of Charter School Entry

摘要

Charter schools in urban neighborhoods, especially those that employ "No Excuses" methods, often successfully raise their students' test scores. On the other hand, charter schools in suburban neighborhoods often do not (e.g., Abdulkadiroglu et al. 2011, Angrist et al. 2013). Though location appears to be an important factor in the success of a charter school, few papers have studied the determinants of charter school location. In this dissertation, I document the neighborhood characteristics associated with charter school entry in Chicago between 1998 and 2003, the period immediately after Illinois passed the first Charter school law. I find that charter schools are more likely to enter disadvantaged neighborhoods --- those with low student test scores, low household income, a low proportion of white residents, and high public school enrollment. Charter schools also are more likely to enter neighborhoods with low Catholic school enrollment and prevalent public transportation. While charter school entry is associated with a low level of Catholic school enrollment, event studies show that any potential crowd-out effect was likely small. To measure the benefits of charter school entry, I estimate a utility model for schools and calculate the total consumer surplus provided by charter entry from 1998 to 2013. According to the model the surplus accrued disproportionately to disadvantaged areas. Neighborhoods at or below median income obtained 73% of the total benefits from charter schools, and neighborhoods at or below median in math test scores obtained 69% of the benefits from charter schools. In addition to these benefits an analysis of the costs of Chicago Public Schools (CPS) shows that the growth of the charter sector creates significant cost savings for CPS.