Childhood Housing and Adult Earnings: A Between-Siblings Analysis of Housing Vouchers and Public Housing

Abstract

To date, research on the long-term effects of childhood participation in voucher-assisted and public housing has been limited by the lack of appropriate data and suitable identification strategies. We create a new, national-level longitudinal data set on housing assistance and labor market earnings to explore how children’s housing affects their later earnings. While naïve estimates suggest there are substantial negative long-term consequences to childhood participation in voucher-assisted and public housing, these relationships appear to be driven largely by negative selection into housing assistance programs. To mitigate this source of bias, we employ household fixed-effects specifications that use only within-household (across-sibling) variation for identification. Compared to naïve specifications, household fixed-effects estimates are more positive for all demographic groups and, for some groups, positive and statistically significant. Black non-Hispanic females, in particular, benefit from time spent in both voucher-assisted and public housing. Exploiting the between sibling variation accounts for unobserved time-invariant family attributes that may influence outcomes but does not address time varying within household factors that may be at work. We use a number of strategies to address these issues and find our results are results are largely robust to these concerns.

JEL Classification: H43; I31; I38; J38; J62