Website CV
Fields
Education, Political Economy, Econometrics, Theory
Job market paper
Paying for Knowledge: Evidence From a Voluntary Certification Program
How does a substantial merit-based salary increase impact employee performance in the public sector? This study leverages the nationwide rollout of Armenia's teacher certification reform, which permanently doubles base pay for teachers who pass a high-stakes subject exam. Administrative records for approximately 36,000 teachers and 400,000 students from 2020-2025 are used to measure teacher effectiveness via teacher value added from standardized course grades. The analysis first evaluates whether subject knowledge, captured by first-attempt certification scores, predicts pre-treatment effectiveness. In mathematics, a one-standard-deviation higher score predicts 0.18 SD higher pre-treatment value added in algebra and 0.15 SD in geometry, with weaker associations in language arts. The causal effect of the raise on subsequent effectiveness is then estimated using a cohort-stacked sharp regression discontinuity at the 60 percent passing threshold, mapped into event time, and a staggered event-study difference-in-differences. Crossing the threshold does not increase productivity for marginal passers in the first post-exam year, and by year three dynamic RD estimates are sizable and negative (about -0.43 to -0.45 SD). Away from the margin, staggered DiD yields small and imprecise average effects overall (about 0.05-0.07 SD), with any gains concentrated among the highest scorers. Subject-knowledge tests therefore screen effectively in mathematics, but conditioning a large permanent raise on a single exam does not improve classroom performance at the passing margin, suggesting renewal or targeted professional development rather than one-off threshold-based pay.