Date of Award

Spring 5-15-2022

Degree Type


Degree Name

Doctor of Philosophy (PhD)




Engelhardt, Gary V.


Marijuana Legalization, Student Achievement, Student Behavior, Teacher Pensions, Teacher Quality

Subject Categories

Economics | Education Policy | Public Affairs, Public Policy and Public Administration | Social and Behavioral Sciences


This dissertation is comprised of three essays on the economics of education. The first and third chapters examine marijuana legalization and its effects on students, while the second chapter examines the impact of pension incentives on teacher quality.The first chapter examines the extent to which there are negative spillovers of recreational marijuana legalization on underage marijuana use and educational outcomes. I use two complementary identification strategies that rely on plausibly exogenous spatial and temporal variation in access to marijuana in Oregon. In November of 2014, Oregon passed Measure 91, a referendum to legalize recreational marijuana. Unlike other legal states, Oregon allowed counties that voted against the legalization measure by at least 55% to opt out. Difference-in-differences estimates suggest that self-reported access to marijuana from the Oregon Student Wellness and Oregon Healthy Teens surveys did not change in counties above versus below the vote-share threshold after legalization, but that use increased, particularly for 11th-grade girls. Additionally, using data on high schools from the Oregon Department of Education, I find that chronic absenteeism, dropout rates, and English proficiency all get worse after legalization. The second chapter, which is co-authored with Patten Priestley Mahler, studies the impact of pension incentives on teacher quality by analyzing a return-to-work policy in North Carolina that effectively removed the "push" incentives embedded in teacher pensions by allowing them to tap into their pension while teaching. Using administrative public-school data from the North Carolina Research Data Center, we estimate the impact of teachers who returned to work after retirement on student outcomes. We develop an instrumental variable identification strategy centered on the cancellation of the policy and find small improvements in both reading and math achievement for students in the same school who had one of these teachers in their grade during the policy relative to students who did not. The results suggest that schools are losing effective teachers because of pension incentives and that return-to-work policies may be a way to retain them. The final chapter estimates the effect of recreational marijuana legalization on educational outcomes using exogenous spatial variation in access to marijuana dispensaries in Washington. In November 2012, Washington passed Initiative-502, a referendum to legalize recreational marijuana. As part of the initiative, the state capped the number of dispensaries at 334. It held a lottery to assign licenses in localities where the number of license applicants exceeded the local dispensary quota, thus generating exogenous variation in dispensary locations. Using an instrumental variable strategy and data on public high schools from Washington's Office of Superintendent of Public Instruction, I find that schools near open dispensaries have worse chronic absenteeism, dropout rates, and discipline rates relative to schools near dispensaries that did not open. This is consistent with the negative effects of legalization that I estimate for Oregon in the first chapter.


Open Access