Association for Public Policy Analysis & Management

At a cost of 19.3 billion dollars, the federal Housing Choice Voucher (HCV) program provides housing subsidies to more than 2.5 million children under the age of eighteen.  Despite its large size, evidence on its educational impacts to date is limited. In this paper we draw on rich administrative data from the nation’s largest school district to shed light on whether – and to what extent ­­ this large national program improves educational outcomes for children whose families receive housing vouchers.

There are several reasons to believe that housing vouchers will improve educational outcomes. First, because vouchers are portable housing subsidies, recipients may live in safer neighborhoods and/or attend better schools. Second, recipients may have higher quality and/or less crowded homes. Third, subsidized households may pay less in rent and enjoy an effective increase in income.  Finally, housing subsidies might increase housing stability over the longer run, helping families and children avoid potential disruptive effects of moving to new schools and communities.

Using data on the largest school district in the country, New York City, and a panel of subsidized housing tenants from the department of Housing and Urban Development (HUD) we match over 88,000 voucher families to public school records and follow their schooling and residential experiences.

To identify the effects of vouchers, we use a difference­in­differences strategy, exploiting the random timing of voucher receipt to compare changes in outcomes of students whose families receive a housing voucher to changes in outcomes of students whose families will receive one in the future. These estimates can be interpreted as causal if timing of voucher receipt is conditionally random and there is good reason to believe this is so.  As of March 17, 2014 there were 121,999 families on the waiting list for vouchers, and it fair to assume that a family is unable to anticipate voucher receipt in any given year and that whether a family receives a voucher in year t rather than t+1 is unrelated to prior performance or family characteristics.

In a second analysis, we compare changes in outcomes among children in voucher families to changes in outcomes among children whose families receive other forms of housing assistance, specifically public housing. We find that students in voucher households perform slightly better in both ELA and math in the years that they are in the program. The largest impacts of the program occur within the first two years of voucher receipt. Compared to students in public housing, students with vouchers perform consistently better, suggesting a portable subsidy delivers bigger impacts than a place-based subsidy. In preliminary work examining potential mechanisms for these effects, we find that students are more likely to make a residential move in the initial year in the program, suggesting that families may be using vouchers to move into higher quality neighborhoods and schools. In addition, we explore whether parent and student perception of schools changes after the receipt of a HCV and exploring alternative outcomes including daily attendance.