Economics Seminar | Heterogenous Teacher Effects, Comparative Advantage, and Match Quality: Evidence from Chicago Public Schools, April 30

The Economics Seminar Series presents Dr. William Delgado discussing "Heterogenous Teacher Effects, Comparative Advantage, and Match Quality: Evidence from Chicago Public Schools" on Friday, April 30, from 3 to 4:15 p.m.

Delgado is a Postdoctoral Researcher at the Harris School of Public Policy at the University of Chicago.

For more information about the Economics Seminars, contact Dr. Roberto Duncan .

Abstract : This paper introduces a novel feature to value-added (VA) models of teacher quality and estimates the potential benefits of using this feature as a policy tool. By capturing heterogeneity in teachers’ causal impacts on student outcomes across student types (e.g., female and male), this feature identifies teachers’ comparative advantage: the added effect on a student type relative to another. In turn, this feature helps estimate match quality: the causal effects of teachers when matched to different classrooms. Specifically, I ask (i) to what extent teachers’ VA estimates vary across student types, (ii) whether these student-type specific VAs provide unbiased forecasts of teachers’ causal impacts across student subgroups, and (iii) what efficiency gains (i.e., increase in student achievement) could be realized by incorporating student-type specific VAs into policy decisions. First, I develop a flexible VA model with student-type specific teacher effects. By employing more than 1.7 million test scores from Chicago Public Schools, I estimate gender- and race-specific teacher VAs and construct comparative advantage measures (e.g., difference between female- and male-specific VAs) to capture cross-type heterogeneities
in teacher VAs. I find a large unexplained variation in teacher comparative advantage (30–
50 percent for gender and 40–80 percent for race) after controlling for traditional measures of teacher quality. Second, by exploiting changes in teaching staff as quasi-experiment, I find that comparative advantage measures accurately predict teachers’ differential impacts across student subgroups. Third, counterfactual teacher retention and reallocation policies that incorporate comparative advantage measures indicate efficiency gains of up to 8 percent. Findings suggest that heterogeneity-based performance measures, in particular comparative advantage measures of teacher quality, can be a policy tool to improve efficiency.