Association for Public Policy Analysis & Management

Historically, rental subsidies provided by the Housing Choice Voucher program, the United States’ largest housing assistance program, generally have been capped at the 40^th percentile of rents paid by recent movers in the respective metropolitan area, with relatively little scope for variation within a metropolitan area. Rents across a metropolitan area tend to be higher in neighborhoods with lower poverty and better access to amenities and services, including education, employment access, and environmental quality. Vouchers allow assisted households to choose any unit that meets the rent and quality parameters of the program. However, rents in higher opportunity neighborhoods are often above the 40^thpercentile cap. In practice, assisted households are frequently concentrated in high-poverty neighborhoods with limited access to the amenities and services associated with resident opportunity.

In this paper we evaluate the effectiveness of using ZIP Code-based subsidy caps, called Small Area Fair Market Rents (SAFMRs) instead of metropolitan area-based caps, called Fair Market Rents (FMRs). This change is in effect in five public housing authorities (PHAs) across the nation through HUD’s Small Area Fair Market Rent Demonstration and additionally in the PHAs in the Dallas, TX metropolitan area. With SAFMRs becoming mandatory in 24 metropolitan areas across the country and optional for all PHAs in the Fall of 2017, this paper provides evidence of the effect of the change in locations where it was already made.

We show that, as intended, SAFMRs increase the pool of units potentially available to housing choice voucher holders in high-opportunity neighborhoods and decrease the number of units potentially available in low-opportunity neighborhoods.

We compare voucher household locations before and after the introduction of SAFMRs in the demonstration metropolitan areas to changes in voucher household locations over the same time periods in a large sample of jurisdictions that continued to use FMRs. We find that SAFMRs appear to affect tenant locations. Of voucher households that move, the share locating in ZIP Codes with relatively high rents within the SAFMR metropolitan area increases by 10 percentage points after the introduction of SAFMRs, whereas no such change is observed in unaffected jurisdictions. This change results in an increase in the share of voucher households living in ZIP Codes that offer higher levels of opportunity.

We conclude by showing that this increase in the share of voucher holders moving to higher-opportunity neighborhoods was accompanied by a reduction in overall average subsidy cost per assisted household to the government. In brief, the higher subsidy costs for households that live in ZIP Codes with increased assistance caps are more than offset by lower subsidy costs for households that live in ZIP Codes with reduced rent caps.