JMP: Draft, October 2023
Abstract: This paper provides evidence that student behavior issues contribute to teacher turnover among U.S. middle school teachers. Using detailed administrative data on student behavior, discipline, and teacher movement in North Carolina middle schools, I show that when teachers experience an increase in reported disciplinary offenses at their school–or among the students in the grade they teach–they are more likely to leave the school or the profession. Among first-year teachers, these effects are largest for more effective teachers. I measure student behavior using only offenses that require mandatory reporting to the state, suggesting that differential reporting by teachers or schools is not driving the results. Further, I compare teachers to others at their school using school and school-by-year fixed effects models, suggesting that school-level changes in student composition are not driving the results. I also show that a more punitive disciplinary response to student offenses does not lead to higher teacher turnover for most teachers. These findings suggest that schools and teacher preparation programs should focus on strategies to help teachers manage student behavior issues.
Education Economics (2023)
Abstract: This paper provides evidence on the effect of local prescription opioid use on academic achievement of 3rd–8th graders between 2009 and 2018. Using county fixed effects models, I find that when counties have higher levels of prescription opioid use, students score lower on standardized assessments two to three years later, with variation by student subgroups and magnitudes comparable to effective interventions. I find the largest magnitudes in counties with higher poverty rates and states with below-median state education spending. As test score effects predict adult outcomes, these findings point to economic and public health challenges when affected children become adults.
Education Economics (2018), with Erin Velez and Melissa Cominole
This paper measures the effects of undergraduate student loan debt on graduates' post-college outcomes: employment, additional enrollment, family formation, home ownership, and net worth. The analysis uses data from a nationally representative sample of 2007--08 bachelor's degree recipients. Because a graduate's debt burden is not randomly assigned, we use an instrumental variable--enrollment-weighted average in-state tuition over four years--to estimate the effect of debt on post-baccalaureate outcomes while minimizing selection bias. We find that four years after graduating, undergraduate debt is related to borrowers' earnings, job choice, decisions to marry and have children, and net worth.
Provides descriptive statistics from a novel data collection of the characteristics of technology bootcamps to classify programs based on their modality (online vs. offline), intensity, total tuition fees, and length of program.