Where Children Live Matters: Housing Policy Effects on Education Outcomes

Social scientists have long suspected that the demographic characteristics and attributes of neighborhoods affect the educational outcomes of the children who live in them. However, this hypothesis is challenging to test because people tend to self-select the neighborhoods in which they live. As a result, exogenous variation — a randomized reason for residence in a given neighborhood instead of self-selection — is necessary in order to estimate the causal effect of neighborhoods on the educational outcomes of young people. Recent research following a natural experiment in Denver, Colorado establishes a link between location and education, suggesting that housing policy can improve educational achievement among Latino and African American youth.

Galster, et al. find strong evidence that neighborhoods affect education outcomes. For example, higher neighborhood occupational prestige is associated with a decrease in the risk of dropping out of school. In addition, on the negative side, social vulnerability — measured by the percentage of low-income residents, unemployed persons, renters and female-headed households — is statistically significant in predicting the rate of students repeating a grade level.

The authors leveraged the Denver Housing Authority’s (DHA) waiting list for housing assistance as a natural experiment because it mimics a random process, which means they could examine the relationship between neighborhood and educational outcome without the confounding factor of self-selection. Through the DHA program, eligible families — defined as households earning less than 80-percent of median income for the area — are offered a housing option from among a wide range of neighborhoods throughout Denver.

Each family has the right to refuse the first offer from DHA, but the authors estimate that three-fourths of all tenants choose to reside in a unit in the first neighborhood offered. Combining this natural experiment with an instrumental variable econometric approach, Galster, et al., measure the independent effect of neighborhood of residence on educational outcomes.

The study includes 764 Latino and African American youth from families with an average household income of $13,213. The researchers focus on three educational outcomes: repeating a grade in school (retention), leaving school before a diploma was earned (dropout), and final grade point average (GPA).

In handling neighborhood data, the authors compute a neighborhood social vulnerability index by summing the percentage of poor residents, unemployed workers, renters and female-headed households. A neighborhood occupational prestige score and administrative data on crime statistics are also utilized to characterize each neighborhood. Through these methods, the authors estimate the causal effect of neighborhoods on children’s education.

The results highlight several aspects of neighborhoods that are important predictors of educational performance for Latino and African American students. The researchers postulate that neighborhoods with many professionals and role models could expose students to potential career paths and incentivize educational success. In communities with elevated property crimes, on the other hand, economic disincentives for academic achievement may be present if illegal activity offers more lucrative earning potential than legal labor market participation. Among Latino youth in particular, neighborhoods containing more property crimes were associated with a greater risk of dropout. The authors argue that these results can be generalized to low-income African American and Latino families who remain on the waiting list long enough to obtain public housing.

One difficulty in the study is that determining whether the results measure the effect of neighborhoods distinctly from that of schools is impeded by the lack of information about local schools. Yet, if school attributes are closely correlated with neighborhood characteristics, then local schools are one mechanism through which neighborhoods affect educational outcomes. The model presented in this research is unable to separate the two effects, but this shortcoming only calls into question the specific ways in which neighborhoods impact educational outcomes, not whether the effect occurs. The Denver Housing Authority takes advantage of this effect by offering families the opportunity to move to a new neighborhood with a better school, lower crime, more professionals and greater social stability.

Moving forward, housing policy offers a potential opportunity to improve educational outcomes for minority youth. The authors’ findings differ from the oft-cited Moving to Opportunity (MTO) study, yet both uncover the potential of housing policy as a means to affect educational outcomes. One significant difference between this research and the MTO study is the duration of time children lived in disadvantaged neighborhoods, with the focus on younger children in the MTO study suggesting that early childhood environments may have long-lasting impacts. More research is needed to address the ways in which neighborhoods affect children’s development, but it remains evident that a child’s neighborhood affects his or her educational outcomes.

Article Source: Galster, George, et al. “Neighborhood Effects on Secondary School Performance of Latino and African American Youth: Evidence From a Natural Experiment in Denver.” Journal of Urban Economics 93 (2016): 30–48.

Featured photo: cc/(SeanXu, photo ID: 502919920, from iStock by Getty Images)