Association for Public Policy Analysis & Management

The Housing Choice Voucher program is the largest source of federal rental assistance in the country. A key aim of tenant-based vouchers is to expand the housing (and neighborhood) choices of recipients. Yet two key criticisms of the voucher program relate directly to choice, namely that a sizable share of voucher recipients fail to successfully lease housing with their voucher, and that those who do, tend to be concentrated in a limited number of neighborhoods that are highly disadvantaged.

A key hurdle is landlords. The voucher program is unusual among social programs in that take-up requires both tenant and landlord participation. A recently commissioned HUD audit study of landlords in five different markets reports alarmingly high denial rates, ranging from 78 percent in Fort Worth, Texas to 15 percent in Washington, D.C. (Cunningham et al., 2018). A few excellent qualitative studies shed some light on the landlords who choose to participate in the voucher program. Based on interviews with 127 landlords in Baltimore, Dallas and Cleveland, Garboden et al (2018) report that voucher landlords appreciate the reliable rent payments and low turnover among voucher tenants. Yet many are still wary of serving voucher holders because of their pre-existing stereotypes about how they will behave as tenants. Further, landlords complain about burdensome inspection systems and other bureaucratic impositions.

Source of Income (SOI) protection laws are often lauded as a possible solution, but the existing empirical evidence does not assess the most recent decade when many housing markets were quite tight, and also yields mixed results on some outcomes (e.g., Freeman 2012; Freeman and Li 2012). This paper extends and updates earlier work by assessing a broader set of impacts at a lower level of geography, including directly assessing whether (and where) new landlords enter the program after the adoption of SOI laws. We draw on a combination of datasets, including longitudinal HUD data on the universe of voucher holders across the country, PHA data on voucher utilization, and a rarely used HUD dataset on voucher landlords. We employ a difference-in-difference approach, comparing voucher holder outcomes in jurisdictions before and after SOI protections are adopted to voucher outcomes for nearby public housing authorities (PHAs) with similar pre-trends as key outcomes of interest, but in jurisdictions without such protections. We begin by looking at voucher usage, captured by PHA utilization rates. We also examine changes changes in a range of neighborhood characteristics (including the quality of nearby schools) of new voucher recipients, and those of existing recipients who move, separately by race and family composition. Finally, going to a finer grain of geography and exploiting the HUD data on landlords, we test whether SOI protections appear to expand the number of buildings and landlords that accept voucher households. We specifically test whether any changes in the location outcomes appear to be through the participation of new landlords, or the expansion of new buildings among landlords already accepting voucher tenants in other buildings they own.