This research considers how exit exams in place in Texas impact student behavior and human capital formation. Employing a regression discontinuity framework, I examine the impact of passing the exam the first time it is administered for students within a small window of scores around the passing threshold. Considering behavioral responses to the administration of the Texas Assessment of Knowledge Skills, I study the impact on students' courses taken in high school, attendance, and a set of disciplinary actions. I find little effect of the exit exam on any behavioral response, offering further evidence that current exit exam policies have a small direct impact on a student's probability of graduating from high school and little else. I then consider heterogeneity among school resources to discern whether the TAKS exam has differential impacts across school campuses in Texas.
We consider the effect of federal work-study, a need-based financial aid program, on educational outcomes for individuals in two-year colleges in Texas. We find that both working in school (with any job) and participating in the federal work-study program (in a work-study job) are associated with a 12-15% increase in the probability of persisting to a second year of community college. Additionally, working in a position through the federal work-study program is associated with a higher probability of transferring to a four-year college by 3-4%. We also attempt to measure the effect of student federal work-study participation using an instrument of lagged average federal work-study earnings at the student's institution, but this strategy does not yield precisely estimated outcomes due to the limited number of cohorts available.
Spreading HOPE? Measuring the Effect of Merit Scholarships in Tennessee
Using the 2000 US Census data and American Community Survey data from 2001-2010, I evaluate the effect of merit scholarships in Tennessee on current college enrollment using difference-in-difference estimation. Unlike estimates in Georgia, effects of the Tennessee Education Lottery Scholarship are mildly negative but not statistically different from zero considering the population of youth ages 16-26, focusing on traditional college enrollees ages 18-19 and older students, ages 20-22. I argue these estimates are in line with many more recent findings examining merit scholarship programs. Finally, I employ a synthetic control method to compare these estimates with more traditional estimation strategies.