The Effects of Teacher Evaluation Policy on Student Achievement and Teacher Turnover: Leveraging Teacher Accountability and Teacher Development

Abstract

Like many policymakers across the globe, Tennessee policymakers have adopted theoretically substantial teacher evaluation reforms since the early 2010s. Although research suggests that the introduction of redesigned systems has not affected teacher or student outcomes, on average, specific accountability- and developmental-oriented policy levers within these systems may. Using 3 years of administrative data from low-stakes settings resembling those in many evaluation systems, the authors apply regression discontinuity and local regression designs to estimate the net policy effects of an accountability-oriented policy and a different developmentally oriented policy on teacher and student outcomes. Intent-to-treat associations suggest that assigning teachers more frequent formal observations, a developmentally oriented policy, does not affect teacher turnover or student achievement and may lower the achievement scores of the least effective teachers. Intent-to-treat associations also suggest that issuing teachers relatively lower composite effectiveness scores does not affect either outcome for most teacher groups but may improve the student achievement scores of the least effective teachers. The authors argue that the findings are relevant to a broad swath of education policymakers and imply that policymakers might use the assignment of teacher effectiveness scores for student benefit. Additionally, the authors conclude that while classroom observations might theoretically be capable of improving teaching, they do not do so in practice in Tennessee, one of the world’s most mature reformed teacher evaluation systems.