Do poor children benefit academically from economic integration in schools and neighborhoods? Evidence from an affluent suburb's affordable housing lotteries.

摘要

In the wake of the 2007 Supreme Court decision striking down voluntary school racial desegregation plans, the "next" form of integration might be the placement of poor students in non-poor schools to help close the large income achievement gap in the U.S. To test whether affluent schools or neighborhoods improve low-income students' academic achievement, this project draws on an original dataset about approximately 850 public housing and 3,200 housing voucher elementary-age children from 2000--2007 who lived in a large, prosperous county that voluntarily adopted a zoning mechanism to disperse affordable housing into a variety of poor and non-poor neighborhoods. Two features of the study enable me to contribute new information to the economic integration literature. First, one-third of public housing students and one-fourth of housing voucher students examined obtained a "strong dose" of economic integration since they attended low-poverty schools and lived in low-poverty neighborhoods throughout their elementary career. Second, the county's public housing authority randomly assigned public housing families to neighborhoods and, by extension, to schools.;The study finds that large academic benefits accrue from economic integration in schools (but not neighborhoods) for public housing children by the end of their elementary school years. As of the sixth grade, public housing children in the low-poverty schools performed approximately 13 percentile points better on the standardized math tests than fellow public housing children in mid-poverty schools (0.5 of a standard deviation) and six percentile points better than public housing children in high-poverty schools (0.2 of a standard deviation). Those in low-poverty schools also did eight percentile points better (0.3 of a standard deviation) than public housing peers in moderate- and moderately high poverty schools. These results hold despite public housing children's enrollment in non-accelerated math classes where other less-advantaged students within low-poverty schools clustered. Further, the benefit of low-poverty schools did not proxy the effect of racial or even academic integration. This suggests that low-poverty schools positively impact poor students over the long run via an amalgam of inputs, such as teacher composition, parental and student composition, and social environment.