Working Paper: NBER ID: w11969
Authors: Victor Lavy
Abstract: In 1994 the city of Tel Aviv replaced its existing school integration program based on inter-district busing, with a new program that allowed students to choose freely between schools in and out of district. This paper explores the impact of this program on high school outcomes while distinguishing the effect of choice on individual students from general equilibrium effects on affected districts. The identification is based on a regression discontinuity design that yields comparison groups drawn from untreated tangent neighborhoods in adjacent cities and on instrumental variables. The results suggest that the choice program had significant general equilibrium effects on high school dropout rates, matriculation rates and program of study. The gains are more pronounced among disadvantaged children but not among students who took advantage of the option to attend out of district schools with higher mean outcomes. These results and other evidence related to the behavioral responses of schools and students to the program suggest that the positive impact of the program is mainly due to better matching between students and schools and to productivity effects of choice and competition among schools.
Keywords: school choice; education policy; quasi-experimental; dropout rates; matriculation rates
JEL Codes: I20; J24
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| improved matching between students and schools (I24) | positive outcomes (I14) |
| increased competition among schools (I24) | positive outcomes (I14) |
| opting out of neighborhood schools (R28) | long-term outcomes (I12) |
| Tel Aviv school choice program (I24) | reduction in dropout rates (I21) |
| Tel Aviv school choice program (I24) | improvement in matriculation rates (I24) |
| Tel Aviv school choice program (I24) | better performance in science subjects (D29) |
| Tel Aviv school choice program (I24) | better performance in advanced-level courses (D29) |