The Geography of Educational Opportunity: Teacher Quality

• In Event: Teacher Value Added and Long-Run Student Outcomes

Abstract

Growing economic inequality in recent decades has coincided with a period of declining intergenerational mobility in the United States. In a series of seminal papers, economics researchers have sought to characterize the “geography of economic opportunity” by measuring intergenerational mobility at the commuting zone and county levels (Chetty and Hendren, 2018a; Chetty and Hendren, 2018b; Chetty et al., 2014). These authors have correlated summary measures of intergenerational mobility with area-level characteristics. Although this exercise identifies school quality as an important correlate of intergenerational mobility, the key question is whether these correlations reflect a causal effect of school quality or whether other family and neighborhood influences affect both education and economic outcomes and account for the correlations between the two. The evidence that measurable school characteristics including expenditures explain little of the variation in school quality, the fact that much of the most compelling evidence on the variation in teacher quality is based on within-school variation in test-score growth, and the growing recognition that school effects on cognitive skills capture only one channel through which schools affect longer-term economic outcomes complicate efforts to measure the contribution of schooling to geographic differences in economic mobility.

In this paper, we extend the approaches used to estimate teacher effectiveness to quantify the contributions of teachers to skill acquisition and social mobility as measured by longer-term schooling and employment outcomes. Evidence shows that teachers constitute an important input into the education production process. Our empirical two-stage methodology first uses teachers who switch schools to estimate school-average teacher effectiveness at raising cognitive and noncognitive skills relative to other schools in the largest connected network. Comparisons of the effectiveness of teacher movers relative to peer-teachers in their origin and destination schools produce compelling estimates of school-average teacher effectiveness. The approach accounts for all school and neighborhood influences common to all students in a school at a point in time. We then aggregate these estimates for each student to create measures of average teacher effectiveness over grades four through eight. These average effectiveness measures then become the key explanatory variables in longer-term outcome regressions.

We expect our findings to provide new evidence on variation in school-average teacher quality and the channels through which teachers affect longer-term outcomes and promote educational and economic opportunity. Comparisons across different levels of aggregation including district, commuter zone and time and across demographic groups including race and gender will illuminate the character of teacher quality effects given the existing institutional structures. The analyses included in this paper and the planned extensions represent part of a larger research agenda focused on the contributions of teachers and schools to longer-term educational attainment, labor market engagement, and involvement in the criminal justice system. By characterizing the degree to which geographic variation in teacher effectiveness and school resource effects can ultimately help to explain geographic variation in intergenerational mobility, this work will contribute to research on the determinants of economic and social opportunity and provide valuable information for the development of education and social policies.

Author

• Steven g Rivkin, University of Illinois, Chicago
  Presenting Author