Housing Choice Vouchers provide low-income households with additional income to spend on housing.  The assistance provided by vouchers is substantial.  As an example, the median voucher household with children has a family size of four, earns approximately $13,000 annually and lives in a unit that rents at $1,000 per month.  For this family, the voucher is equivalent to an increase in post-tax income of approximately $8,000 annually, increasing their income by 60 percent. Thus, vouchers have the potential to dramatically widen the neighborhoods -- and associated schools -- that low-income households can reach.  However, existing research on this program finds that voucher holders do not live near to better schools than other households with similar incomes (Horn, Ellen and Schwartz, 2014).  This project examines this question from a new angle, focusing on the full population of voucher holders and examining whether households with school aged children are more likely to use vouchers to move towards better schools. 

Relying on the Department of Housing and Urban Development’s internal assisted housing dataset, between 2002 and 2011, matched to data on the quality of the nearest elementary school within their district, we are able to observe the quality of the school in the neighborhoods where voucher holders live when they first enter the program and then as they spend more years in the program.  We begin by examining the relationship between the performance in schools near to voucher households and a range of voucher household characteristics.  We estimate the following model:

Sch_imt = a + b₁HH_imt + ηHHFE + γYRFE + e_imt        (1)

where i indexes the household, m the core based statistical area and t the time period.  In this model HHrepresents a vector of household level characteristics.  In order to look at how school quality changes throughout a household’s tenure in the voucher program we run these models with household level fixed effects.  We include an indicator for school aged children to examine whether school quality is higher for these households.  We next observe whether school quality improves for households throughout their tenure in the voucher program.  We interact the number of years a household has been in the program with the presence of school aged children to test for whether households with children are more likely to live near to better schools the longer they have been in the program.

We then examine whether households that move are more likely to move towards higher performing schools.  We add into model (1) an indicator for whether the household moved.  We then interact this variable with the presence of children to identify whether households with children that move are more likely to experience improvements in school quality.  We find evidence that voucher households do move towards better schools when they have school-aged children. We hope this research will provide deeper insight into the residential choices of voucher families, enabling policymakers to improve the voucher program in ways that further help low-income families reach neighborhoods with better schools.