The Effect of Teach for America on the Distribution of Student Achievement in Primary School: Evidence from a Randomized Experiment
Claremont Colleges - Robert Day School of Economics and Finance; Institute for the Study of Labor (IZA)
Southern Methodist University (SMU) - Department of Economics
Claremont McKenna College; Claremont McKenna College
IZA Discussion Paper No. 7296
Using data from a randomized experiment and fixed effect quantile regression (FEQR), we look at the effects of having a TFA teacher on test scores across the entire achievement distribution of primary school students in disadvantaged neighborhoods. While we find that TFA teachers neither help nor hurt students in terms of reading test scores, we find positive and statistically significant effects of TFA over the entire math achievement distribution for the full sample and the effects are fairly uniform. We find a similar effect of TFA across the math test score distribution irrespective of student gender, although the FEQR estimates for female students are two to three times larger than for male students.In addition, we find that there is significant heterogeneity in the effects of TFA for Hispanic and black students and for students taught by novice teachers. Finally, we find that the effect of TFA is homogeneous across the math achievement distribution irrespective of certification type. Taken together, these patterns suggest that allowing highly qualified recent college graduates and mid-career professionals, who in the absence of TFA would not have taught in these disadvantaged neighborhoods, should have a positive influence not just on students at the top of the math achievement distribution but across the entire math achievement distribution.
Number of Pages in PDF File: 32
Keywords: student achievement, random assignment, fixed effect quantile regression
JEL Classification: C21, I21, I28, J24
Date posted: March 30, 2013
© 2015 Social Science Electronic Publishing, Inc. All Rights Reserved.
This page was processed by apollo7 in 0.266 seconds