Improved Neighborhoods Don't Raise Academic Achievement

"The results of this very large scale experiment indicate no evidence of improvement in reading scores, math scores, behavior problems, or school engagement overall for any age group... Interventions focused exclusively on neighborhoods, rather than on factors directly related to the child, family, and school, are unable to solve the myriad problems of children growing up in poverty."

It is assumed that if poor families move to better neighborhoods their children will perform better in school, but until now the data to support this proposition have been difficult to isolate and even more difficult to interpret with anything like a consensus. In Neighborhoods and Academic Achievement: Results from the Moving to Opportunity Experiment (NBER Working Paper No. 11909), Lisa Sanbonmatsu, Jeffrey Kling, Greg Duncan, and Jeanne Brooks-Gunn analyze a rich mine of information regarding such families but find no academic improvement for any of the children.

The data, collected in 2002, arise out of the U.S. Department of Housing and Urban Development's experimental Moving to Opportunity (MTO) for Fair Housing program in the late 1990s. In this program, three groups of low-income families in New York, Los Angeles, Boston, Baltimore, and Chicago were offered housing assistance via lotteries. An "experimental" group could use housing vouchers in low-poverty neighborhoods. The second group, called "Section 8," could use traditional housing vouchers with no restrictions. A control group received no vouchers but was eligible for public housing. The data from these three groupings encompassed more than 5,000 children ranging from pre-schoolers to high schoolers.

Assessments of these children, based on test-score analysis and extensive interviewing, allowed for measuring the impact of the change in residential neighborhood on academic achievement, free of family and individual attributes. The authors here also theorized that neighborhoods might influence the educational norms, social values, and community resources outside of school. Their belief was that a "better" neighborhood provides a more positive impact on children, especially on the youngest learners, who are considered the most adaptable of children to new social environments, and on those just beginning school. The researchers accordingly tested that hypothesis, as well as the overall magnitude of impacts on educational outcomes, including both test scores and behavioral gains. Moreover, they investigated the possibility of differential effects arising from ethnicity, race, gender, and educational risk factors.

Their results indicate no evidence of improvement in reading scores, math scores, behavior problems, or school engagement overall for any age group. At best, the data confirm an earlier study of MTO results for Baltimore that suggested a positive impact on children aged 5-11, but this team's longer-run analysis (four to seven years) showed that the pupils did not sustain their gains. These results, the researchers theorize, may reflect something in particular about Baltimore schools, or may just be happenstance. Overall, the data for the five cities show no appreciable educational or social improvement.

The authors consider a number of explanations for this lack of improvement. One possibility is that members of the experimental group may have moved from poor to middle-class neighborhoods, as their plan allowed, but for a variety of reasons may have had to move again to less affluent neighborhoods. At the same time, the control group, whose families were permitted to use their housing vouchers anywhere, may not have moved to areas appreciably better than those they had been living in. In any event, neither group moved to truly affluent neighborhoods.

Another explanation for the stagnant academic improvement, the researchers say, is that while the voucher recipients generally moved to better neighborhoods, three-fifths of them did not move to racially or ethnically integrated neighborhoods. The authors surmise that "discrimination may limit the availability of high-quality schools or other public services in minority neighborhoods," although they admit that they had no way of confirming this.

Another factor is that, even after moving to a better neighborhood, some parents continued to send their children to school in their former neighborhoods because they felt the children would be more comfortable in familiar surroundings. And, another possible explanation for the lack of academic improvement is that better neighborhoods do not automatically mean better schools. Indeed, the researchers note that the MTO participants rarely if ever found their children attending the top-rated schools. The new school environment almost always showed, for example, no improvement in pupil-teacher ratios or in how the students rated the school climate.

The researchers suggest further that the potential gains associated with neighborhood improvement may have been offset by the disruption of relocation and integration into a new school district. Studies do show that the move to better neighborhoods is associated with positive health outcomes for female adults and youth, fewer criminal arrests among female youth, and an overall reduction in exposure to drugs and violence. Yet despite these pluses, the authors conclude: "It appears that interventions focused exclusively on neighborhoods, rather than on factors directly related to the child, family, and school, are unable to solve the myriad problems of children growing up in poverty. From a policy perspective, residential mobility programs such as Section 8 (now called Housing Choice Vouchers) and even more dramatic MTO experimental treatment do not appear to have large impacts on the academic problems of children who live in public housing in high-poverty neighborhoods.

-- Matt Nesvisky

The Digest is not copyrighted and may be reproduced freely with appropriate attribution of source.
NBER Videos

National Bureau of Economic Research, 1050 Massachusetts Ave., Cambridge, MA 02138; 617-868-3900; email:

Contact Us