Moving to Opportunity?

Girls fare significantly better in multiple dimensions after they move to more affluent neighborhoods... In contrast, boys appear either to be unaffected or to be affected negatively by such moves.

Children who live in areas of concentrated poverty consistently perform worse in school, have more health problems, and get in trouble with the law more often than children who grow up with more affluent neighbors. Adults in these poorer neighborhoods have lower employment rates and worse health. To what extent are these outcomes attributable to living in the disadvantaged neighborhoods, as opposed to the direct effects of the family and the individual characteristics of the residents of such neighborhoods? The answer is critical to the design of education, health, housing, and other social policies aimed at assisting low-income families.

Jeffrey Kling, Jeffrey Liebman, and Lawrence Katz address the question in Experimental Analysis of Neighborhood Effects (NBER Working Paper No. 11577). Their study is based on evidence from the Moving to Opportunity (MTO) demonstration, a randomized housing mobility experiment in which families living in high-poverty, inner-city, public-housing projects were offered housing vouchers to help them move to private housing units in lower-poverty neighborhoods.

MTO is a demonstration conducted by the U.S. Department of Housing and Urban Development (HUD) in five cities: Baltimore, Boston, Chicago, Los Angeles, and New York. Through MTO, public housing residents with children were eligible to participate in a lottery resulting in random assignment to one of three groups. A control group received no new assistance, but continued to be eligible for public housing. A Section 8 group received a traditional housing voucher (known as a Section 8 voucher), without geographic restriction. An experimental group received a Section 8 voucher, restricted for one year to a census tract with a poverty rate of less than 10 percent, and mobility counseling. The random assignment to groups took place from 1994 to 1997.

Of these households, 85 percent were headed by an African-American or Hispanic woman. In their analysis of data from 4248 households, the authors draw upon baseline surveys, administrative earnings and welfare records, and a 2002 survey that collected data four to seven years after enrollment in MTO. In 2002, the families offered housing vouchers through MTO lived in safer neighborhoods with significantly lower poverty rates than those of the control group not offered vouchers.

The authors find that neighborhoods do seem to have independent effects on youth education, health, and behavior, but that these effects are complex. Girls fare significantly better in multiple dimensions after they move to more affluent neighborhoods. In the experimental and Section 8 groups, the girls showed improved mental health, better educational outcomes, and reductions in risky behaviors relative to the control group. In contrast, boys appear either to be unaffected or to be affected negatively by such moves. The boys in the experimental and Section 8 groups experienced higher rates of substance use and suffered a large increase in non-sports injuries relative to the control group.

Moves to lower-poverty neighborhoods also appear to generate a nuanced pattern of effects on the adults in the MTO demonstration. The authors find no significant overall effects of neighborhood moves on adult employment, earnings, or public assistance receipt. These results could reflect the weak job growth of the neighborhoods to which the experimental group moved. Although the offer of housing vouchers appears to have no detectable effects on several aspects of physical health (general health, asthma, physical limitations, and hypertension), the authors find significant reductions in obesity associated with MTO moves.

In addition, the MTO intervention generated substantial and significant mental health benefits for the adults in the experimental group. Using both voucher groups in analyses, the authors demonstrate that larger changes in neighborhood poverty rates are associated with larger improvements in mental health in an almost linear fashion. Overall, there is a consistent pattern of improvements in specific aspects of mental health for both groups offered vouchers relative to the control group. For example, the authors find large and significant reductions in psychological distress and increases in feeling "calm and peaceful" for the experimental group relative to the control group. The mental health benefits may have important spillover benefits, particularly to children who have been found to have more problems in school and more behavior problems when their mothers are experiencing mental health problems. Taken together, these findings suggest that health concerns may need to return to the more prominent place in housing policy discussions that they held 60 years ago - with a new emphasis on the importance of mental health

-- David R. Francis