Essays on the economics of education of underserved populations
Farber, Matthew Scott
MetadataShow full item record
This dissertation examines how current targeted accountability and funding provisions under federal guidelines impact the academic outcomes of the country's more underserved populations.The first chapter demonstrates that accountability at the race level leads to increased reading and math achievement for students. I investigate the impact of school-level accountability on racial subgroups within a school, using a regression-discontinuity design with student-level Texas panel data on third through eighth graders from 2004 through 2011. The targeted incentives increase passing rates by 1-2 percentage points and the scores by .03 standard deviations in both math and reading. These results persist for two to three years after intervention, but fade out by the fourth year. Furthermore, students outside the targeted group are not hindered, with no effect on passing rates and scores. A deeper analysis suggests that schools are not focusing on high-leverage students but rather implementing wide-ranging interventions. I also find that the majority of gains are due to gains among Black students, though it is not clear whether this is due to racial targeting. In the second chapter, I analyze the impact of federally designed and funded interventions on student achievement, both of targeted students and non-targeted students. Under the No Child Left Behind (NCLB) act of 2001, schools with less than 40% low-income students use federal Title I funds for a Targeted Assistance Program, where schools above 40% are free to use those same funds as general school money. This paper uses a fuzzy regression discontinuity design around the 40% threshold with student-level Texas panel data on third through eighth graders from 2004 through 2011 to investigate. The evidence suggests that there is no difference in student outcomes, on the whole or among subsamples, between the methods of using the federal funding. The third chapter of my dissertation shows that the impact of Title I funding on student achievement is complex, benefiting certain subgroups of students while impacting others negatively. I use an instrumental variable research design in order to estimate impacts while keeping external validity through exploiting the large data set available, which includes student-level panel data on Texas public school students from the years 2004 through 2011. Title I funding increases math passing rates by 3 percentage points and has no impact on neither reading passing rates nor standardized scores for either subject. Elementary school students are impacted negatively by Title I funding in both math and reading, while lower-performing and low-income middle school students show large, though insignificant, effects of the funding on both math and reading exams. Unfortunately, this study cannot speak to the impacts on high school students.