Work In Progress
"Paying for Lack of Performance? Effects of Principal Incentive Pay on Students and Teachers," [JMP] (Draft coming soon!)
Abstract: With national test scores at historic lows, incentives that motivate effective school leadership are increasingly important, yet principal pay typically follows fixed schedules rather than student outcomes. This paper studies the effects of a principal incentive pay program implemented in North Carolina on principal effort, teacher effectiveness and student test scores. I employ two models to estimate the causal effects of principal incentive pay on student test scores: (i) a difference-in-differences matching model and (ii) a synthetic difference-in-differences model, both using school-grade panel data and comparing North Carolina (NC) to Georgia (GA). I find the policy induced large declines in student outcomes, reducing reading and math scores by 0.09-0.11 and 0.13-0.14 SD, respectively. To decompose policy channels, I combine NC administrative data with the Teacher Working Conditions Survey to construct principal-effort indices along four dimensions and estimate teacher value added pre/post reform. I then relate teacher value added to principal effort and characteristics to understand how principals influence teachers. I find that teacher effectiveness fell by 0.05 SD in math and 0.01 SD in reading, and principal effort devoted to school climate declined by 0.03 SD. Teachers are most responsive to principal effort in school climate, professional development, and reduced administrative load. My results suggest that incentive pay induced principals to exert effort towards margins that teachers are less responsive towards.
"The Impact of Workplace Harassment on the Occupational Outcomes of Women: A Structural Approach to an Equilibrium Problem," with Ornella Darova
Abstract: This paper investigates the impact of workplace harassment on various occupational outcomes, utilizing a novel empirical approach that combines reduced-form strategies with a structural equilibrium framework. Leveraging the restricted-use Longitudinal Employer-Household Dynamics (LEHD) database, we exploit sharp thresholds in Title VII coverage and the diffusion of the #MeToo movement to study the effect of workplace harassment on labor supply. We employ regression discontinuity designs and event studies to estimate the causal effects of vicarious liability on firm demographics, wage gaps, and employee turnover. We then exploit the findings in a structural model that estimates a job market equilibrium.
Working Papers
"Peer Suspension Effects on Student Misbehavior," with Margaux Luflade and Maria Zhu
Abstract: School suspensions are a widely debated disciplinary tool, with policymakers and researchers divided on their effectiveness. While much of the controversy focuses on the suspended students themselves, suspensions may also produce network effects, potentially deterring or encouraging peer misbehavior. This paper leverages administrative data from North Carolina public schools to estimate the impact of misbehavior and suspensions on subsequent peer infractions and suspensions. Our findings indicate that school discipline does not deter peer misconduct; instead, an additional average weighted peer suspension increases the propensity of being suspended by up to 6%. We demonstrate that this effect is driven by peer suspension, rather than peer misbehavior. The results are robust to an alternative peer group definition and withstand falsification tests. Overall, these findings suggest that suspensions negatively influence peers of disciplined students by increasing their propensity for future misbehavior.