UC Riverside

Chapter 1 examines the effects of providing teachers with performance measures on student achievement and how this policy may differentially impact students. In response to many policy initiatives, many states adopted new teacher evaluation systems comprised of multiple measures of teacher performance, including metrics based on students' performance on standardized tests: Student Growth Measures (SGMs). I construct an original information data set detailing each state's implementation policy and link it to the nationwide data. Using the difference-in-differences and event studies, I find that releasing the SGMs to teachers negatively impacts students' math scores, and the impact becomes more prominent with time. By looking at the change in the distribution of scores, I find that this unexpected adverse effect of the policy is driven by the deterioration among previously high-performing districts and schools.

Chapter 2 examines the policy of providing Value-Added (VA) measures to teachers on student performance in Ohio and North Carolina. Using the within-state variation of the policy implementation, I find that the distribution of students' performance shifted downward in schools with VA policy, suggesting that VA is detrimental to high-performing students. These results show that SGMs have unintended effects on student achievement, undermining students' performance at the top end of the distribution.

Chapter 3 discusses how students' exposure to drinking culture a year before reaching the legal drinking age affects their educational outcomes. I exploit a discrepancy in school cohort cutoffs in South Korea, which leads some students to be exposed to peers with the legal right to drink at an earlier age. I find that the students exposed to a peer of drinking age consume more alcohol, but this does not translate into a higher college dropout rate.