The Impact of Varying Rent Subsidies on Housing Voucher Neighborhoods by ZIP Code Level Rather than Uniform Subsidies

Using housing choice voucher subsidies pegged to ZIP Code-level rents rather than uniform subsidies across entire metropolitan areas increases the share of new voucher holders moving to neighborhoods with higher rents and lower poverty rates, without increasing costs or affecting the ability of new voucher recipients to use their vouchers, according to a new study by NYU Furman Center’s Ingrid Gould Ellen, Katherine O’Regan, and Sarah Strochak.

The paper, “Pricing for Opportunity: The Impact of Spatially Varying Rent Subsidies on Housing Voucher Neighborhoods and Take Up,” published by the Journal of Public Economics, evaluates a new approach to setting voucher subsidies adopted by the U.S. Department of Housing and Urban Development (HUD), which vary with ZIP Code rents, called Small Area Fair Market Rents (SAFMRs).  In 2018, HUD mandated that housing agencies in selected metropolitan areas use SAFMRs. HUD adopted the change to help more of the nearly 2.3 million voucher holders in the country access lower poverty, higher opportunity areas that particularly benefit low-income children. 

Researchers have previously suggested that the low use of vouchers in low-poverty neighborhoods could be caused by the uniform nature of the rental assistance. Housing Choice Vouchers allow holders to pay 30% of their income in rent, with the remainder covered by the subsidy if the unit falls at or below a specified local payment standard within certain ranges set by HUD. Payment standards have traditionally been pegged to the 40th percentile of rents in the metropolitan area. The problem is that these payment standards are typically too low for homes outside of low-poverty neighborhoods. As a result, voucher holders tend to use their subsidies for housing units in high-poverty neighborhoods with lower rents. 

At the time the rule was proposed, several advocacy groups expressed concerns that the change would reduce landlord voucher acceptance in low-rent ZIP Codes, where the new payment standards would be below the subsidies to which landlords were accustomed. They argued that while increasing subsidies in high-rent neighborhoods should broaden choices for voucher recipients, reducing subsidies in low-rent neighborhoods, where most voucher recipients live, could increase search costs and reduce the already low voucher leasing success rates.  

Ellen, O’Regan, and Strochak examined HUD administrative data for all voucher recipients from 200 housing authorities, from 2014 through 2022, which includes key demographics, destination address, as well as origin ZIP Code to assess the impacts of the 2018 reform. They used the same criteria set by HUD to select the 23 metropolitan areas that were mandated to use ZIP Code-level rent subsidies to select a relevant comparison group, which comprised voucher recipients in metropolitan areas just below the threshold HUD used for inclusion. The novel data set allowed Ellen, O’Regan, and Strochak to measure leasing success rates for voucher recipients, a key metric that prior research has been unable to track. 

Key Findings:

• The study finds that the shift from uniform to ZIP Code subsidy pricing led new voucher holders to lease homes in higher-rent and lower-poverty neighborhoods. The chart below demonstrates this finding, plotting raw, uncontrolled average differences between a measure of rent (rent ratios) and poverty percentiles of origin and destination neighborhoods of voucher recipients who successfully leased homes with their vouchers. The charts show that voucher recipients in the comparison group and SAFMR treatment group track closely in the two years prior to implementation, and then they diverge in 2018 as voucher holders in SAFMR metropolitan areas lease homes in more expensive neighborhoods. 

Figure 1: Raw change between origin and destination ZCTAs over time uncontrolled  

Note: The change between origin and destination outcomes are calculated as the destingation value minus the origin value.

• Furthermore, the paper finds that, four years after SAFMR implementation, successful voucher households in metropolitan areas experienced a reduction in neighborhood poverty rates of nearly four percentile points compared to comparable households. This can be seen in the chart below, which shows the average difference between the destination and the origin rent ratio, as well as the difference between the origin and destination poverty percentile. 

Figure 2: Event study plots for change between origin and destination characteristics

Note: Panel above shows the change between origin and destination characteristics.

• Although advocates were concerned that SAFMRs would make it more difficult for voucher recipients to use their vouchers to lease homes, this paper finds that the program’s locational impacts did not come at the expense of overall voucher success rates. Figure 3, below, demonstrates that even after the adoption of SAFMRs, trends in both success rates and search times are generally the same in SAFMRs and in comparable metropolitan areas that did not peg voucher subsidy amounts to ZIP Codes. Furthermore, Figure 8 shows that there are no statistically significant increases in tenant rent contributions, even in the low-income neighborhoods where payment standards declined. 

Figure 3: Raw search outcomes over time, uncontrolled

Note: Success rate is the share of searches that successfully use their voucher within 1 year of issuance. Search time is the elapsed time between the issuance of the voucher and successful voucher use, for households that are successful. Standard errors are clustered by PHA.

• Finally, the study examined the locational effects on subgroups. Through their analysis, the researchers find that SAFMR’s impact on neighborhood rent and poverty rates holds for virtually every subgroup. The one exception is older adults, perhaps because they have less to gain from higher rent, lower-poverty neighborhoods and more reasons to stay in existing neighborhoods. 

The paper is available for download on the Journal of Public Economics for a limited time. You can find the SSRN Working Paper here.