Association for Public Policy Analysis & Management

As the federal government started to grapple with the role of housing policy in addressing racial segregation, President Nixon noted that freedom requires both the right and the ability to choose: “The right to move out of a mid-city slum… means little without the means of doing so" (Polikoff (2006), p 92). This thinking motivates one of the central goals of federal housing policy, which is to provide the means for people to choose housing in high-quality neighborhoods (Quadel (2001)).

Despite this policy goal, participants in the largest federal housing program, the Housing Choice Voucher (HCV) program, typically live in neighborhoods with limited opportunities. Since HCV vouchers are eligible for use in any neighborhood, the low-opportunity locations of voucher holders are typically interpreted as being driven by tenant preferences. Landlord choices could also drive the low-opportunity locations of voucher holders by limiting the neighborhoods in a voucher holder's choice set. Landlords, particularly in high rent neighborhoods, might avoid voucher tenants due to higher maintenance costs, compliance costs, or their own or other tenants' (discriminatory) preferences. The difficulty of observing landlord behavior has limited the study of this important driver of location outcomes in the economics literature.

The HCV program traditionally limits vouchers to the one “Fair Market Rent" (FMR) for all apartments (with the same number of bedrooms) within a metro area. Could paying more in high rent neighborhoods lead more landlords to accept vouchers in such neighborhoods? Some cities have received exemptions from HUD to implement neighborhood-level rent limits, and HUD has moved to expand small area rent limits to several additional metro areas. Such a policy might be expected to convince more landlords in high rent neighborhoods to participate in the HCV program and move voucher recipients out of high poverty neighborhoods.

This paper studies an expansion of neighborhood-level rent limits in Washington, DC in two waves of an audit study, sending emails to landlords from fictional potential tenants. Our two waves of experimental data from 2015 and 2017 straddle a policy change that increased rent limits in high rent DC neighborhoods from 130% to 175% of HUD's metro-wide FMR level. As in Phillips (2017), we find that landlords assess a large penalty to voucher tenants. Inquiries that expressed a desire to pay with a voucher receive 27 percentage points fewer responses. The penalty for voucher tenants is larger in high rent neighborhoods. The voucher penalty is 10 percentage points larger in a neighborhood at the 90th percentile of the rent distribution than at the 10th percentile.

We find no evidence that indexing rents to small areas affects landlord acceptance of voucher tenants or ultimate lease-up locations in an economically significant way. Similarly, we find only limited evidence that voucher lease-up rates increase in neighborhoods with increased limits relative to those with limits that do not change. We do find some evidence that a small number of tenants are able to move to neighborhoods that previously had no voucher tenants.