Working Paper: NBER ID: w12286
Authors: Julie Berry Cullen; Randall Reback
Abstract: We explore the extent to which schools manipulate the composition of students in the test-taking pool in order to maximize ratings under Texas' accountability system in the 1990s. We first derive predictions from a static model of administrators' incentives given the structure of the ratings criteria, and then test these predictions by comparing differential changes in exemption rates across student subgroups within campuses and across campuses and regimes. Our analyses uncover evidence of a moderate degree of strategic behavior, so that there is some tension between designing systems that account for heterogeneity in student populations and that are manipulation-free.
Keywords: accountability; performance measurement; education; school ratings; gaming
JEL Codes: D82; H39; I28
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| accountability system (H83) | strategic behavior of schools (L21) |
| exemption rates (H26) | likelihood of achieving higher ratings (C52) |
| strong short-run incentives to exempt additional students (H52) | increase in the fraction of students classified as exempt (I24) |
| expected pass rates of subgroups (C46) | marginal benefit of exempting students (H52) |
| manipulation of exemption rates (H26) | inflation of exemption rates for Hispanic and Black students (I24) |
| accountability system (H83) | exemptions correlate with expected pass rates (K34) |