Housing mobility protects against alcohol use for children with socioemotional health vulnerabilities: An experimental design.

Abstract

Neighborhood context may influence alcohol use, but effects may be heterogeneous, and prior evidence is threatened by confounding. We leveraged a housing voucher experiment to test whether housing vouchers' effects on alcohol use differed for families of children with and without socioemotional health or socioeconomic vulnerabilities. In the Moving to Opportunity (MTO) study, low-income families in public housing in five US cities were randomized in 1994 to 1998 to receive one of three treatments: (1) a housing voucher redeemable in a low-poverty neighborhood plus housing counseling, (2) a housing voucher without locational restriction, or (3) no voucher (control). Alcohol use was assessed 10 to 15 years later (2008 to 2010) in youth ages 13 to 20, N=4600, and their mothers, N=3200. Using intention-to-treat covariate-adjusted regression models, we interacted MTO treatment with baseline socioemotional health vulnerabilities, testing modifiers of treatment on alcohol use. We found treatment effect modification by socioemotional factors. For youth, MTO voucher treatment, compared with controls, reduced the odds of ever drinking alcohol if youth had behavior problems (OR=0.26, 95% CI [0.09, 0.72]) or problems at school (OR=0.46, [0.26, 0.82]). MTO low-poverty treatment (vs. controls) also reduced the number of drinks if their health required special medicine/equipment (OR=0.50 [0.32, 0.80]). Yet treatment effects were nonsignificant among youth without socioemotional vulnerabilities. Among mothers of children with learning problems, MTO voucher treatment (vs. controls) reduced past-month drinking (OR=0.69 [0.47, 0.99]), but was harmful otherwise (OR=1.22 [0.99, 1.45]). For low-income adolescents with special needs/socioemotional problems, housing vouchers protect against alcohol use.