Not available / text

The Moving to Opportunity (MTO) randomized housing experiment was designed to evaluate the causal effects of random voucher assignment on the economic outcomes of adults and the educational outcomes of their children. MTO recruited predominantly non-Hispanic Black and Hispanic families living in public housing in neighborhoods with over 40% poverty and, in most cases, racial/ethnic minority concentration of 80% or more. Post hoc, MTO also examined the effect of voucher receipt on the long-term mental health outcomes among adolescents and, generally, found harmful effects among boys, but protective effects among girls. On average, voucher receipt led to residing in more affluent, and slightly less racially and ethnically segregated neighborhoods. These neighborhood differences could have led to greater exposure to perceived racial/ethnic and class discrimination among non-Hispanic Black and Hispanic adolescents from high-poverty neighborhoods, which, in turn, may have affected their mental health, with a stronger harmful effect among boys than girls. A prior qualitative study found that compared to no voucher, boys whose families moved with a voucher reported more instances of racial and class discrimination in the new neighborhoods. However, girls whose families moved with a voucher did not report more instances of discrimination than girls whose families did not receive a voucher. In addition, Black and Hispanic boys have been shown to be more sensitive to differences in affluence between themselves and their neighbors than girls. For example, low-income boys, but not girls, living alongside more affluent neighbors exhibited more antisocial behavior compared to their peers who lived alongside low-income neighbors. Examining whether aspects of neighborhood poverty and racial/ethnic composition, which may be associated with class and racial/ethnic discrimination, are mechanisms behind the unintended consequences of Section 8 housing voucher receipt may inform more nuanced housing policy decisions. Identifying the pathways for the unintended harmful effects on the mental health of boys may help policymakers distribute housing subsidies in a way that is beneficial to all family members. To examine the potential pathway for the surprising findings in the MTO, I completed three specific aims in this dissertation: a systematic literature review and two empirical aims. First, I conducted a systematic literature review to examine the evidence that sex moderates the established association between perceived racial/ethnic discrimination and mental health outcomes in Black and Hispanic preadolescents to young adults. Second, I estimated the effect of Section 8 voucher receipt on perceived individual-level racial/ethnic discrimination, stratifying analyses by the selected effect modifiers (e.g., voucher type, sex, and/or city). Third, I estimated the indirect effects of voucher receipt on the risk of long-term mental health outcomes among adolescents through neighborhood affluence and racial/ethnic composition, stratifying by sex. The results of the systematic literature review suggest that, among pre-adolescents to young adults, there was no strong evidence to support the moderating effect of sex on the association between perceived racial/ethnic discrimination and depressive symptoms or behavioral problems. However, the review found support for the moderating role of sex on the association between perceived discrimination and the outcomes of anxiety and PTSD. In the first empirical study, I found that Low Poverty Voucher (LPV) receipt reduced the risk of perceived racial/ethnic discrimination by police among boys only. In Los Angeles, receipt of LPV and, separately, Traditional Voucher (TRV) receipt reduced the risk of perceived school and neighborhood discrimination, respectively. In contrast, in Chicago, TRV receipt increased the risk of perceived discrimination at a store or at a restaurant. Voucher receipt did not have an estimated effect on the risk of perceived racial/ethnic discrimination in Boston or New York. In the second empirical study, among boys, part of the estimated harmful effect of voucher receipt (LPV and TRV combined) on the risk of externalizing disorders was mitigated by a bundle of mediators, which included residing in neighborhoods with less than 20% poverty, more college graduates, and slightly less racial/ethnic segregation over the duration of follow-up. However, the same bundle of mediators explained part of the harmful effect of voucher receipt on the risk of mood disorders among boys. Among girls, part of the protective effect of voucher receipt on externalizing disorders and mood disorders operated through this bundle of mediators. I utilized causal inference methods (i.e., calculated average treatment effect) to estimate the effect of Section 8 housing vouchers on perceived racial/ethnic discrimination in Chapter 3, as well as their indirect effect, through neighborhood affluence and racial/ethnic composition, on long-term adolescent mental health outcomes in Chapter 4. Overall, the findings of this dissertation suggest that there is heterogeneity in the effect of perceived racial/ethnic discrimination and Section 8 voucher receipt on adolescent mental health outcomes. Furthermore, such effects vary depending on the mental health outcome. For example, among boys, the harmful effect of voucher receipt partially operated through the bundle of mediators on mood disorders, but the same mediators had a protective effect on externalizing disorders. Additional research is needed into the causes of these heterogeneous indirect effects on mental health outcomes among boys. Future housing interventions should consider utilizing this information to avoid harming the subgroups they intend to aid.