How Do Changes In Housing Voucher Design Affect Rent and Neighborhood Quality?
Based on BFI Working Paper No. 2018-1, “How Do Changes In Housing Voucher Design Affect Rent and Neighborhood Quality?” by Peter Ganong, assistant professor in economics at UChicago’s Harris School of Public Policy, and Robert Collinson, doctoral student, New York University
- Better neighborhoods are an important factor in determining residents’ future success
- Housing vouchers are meant to encourage movement to better neighborhoods, but often fail to do so
- To induce voucher recipients to move to better neighborhoods, subsidies should be “tilted” higher for better neighborhoods and decreased for lower-quality neighborhoods
- Evidence from Dallas, home to a rent-tilting program since 2011, shows improvement in neighborhood quality for voucher recipients
Policy reforms to address this phenomenon have focused on increasing voucher size, which is the highest dollar rent that the government is willing to pay. The idea is that if you give a larger vouchers to households to use wherever they want, they will apply that voucher to lease units in neighborhoods with higher rents. Surprisingly, though, the evidence suggests otherwise. In their paper, “How Do Changes In Housing Voucher Design Affect Rent and Neighborhood Quality?”
Notes: This figure shows the two changes in voucher generosity that we study in this paper. The first increases the maximum per-unit government subsidy — which we refer to as the “rent ceiling” — uniformly in all neighborhoods in a metro area. The second increases the ceiling in higher-quality ZIP codes and lowers it in lower-quality ZIP codes.
Peter Ganong, assistant professor in economics at UChicago’s Harris School of Public Policy, and Robert Collinson, doctoral student at New York University, address this question by evaluating the effectiveness of two policy options: increasing the per-unit subsidy for all neighborhoods, or increasing the subsidy in higher-quality ZIP codes while decreasing the subsidy in lower-quality ZIP codes.
Their conclusion: when policymakers uniformly increase the subsidy for homes across all neighborhoods, some landlords may benefit, but there is little impact on observed neighborhood quality. People mostly stay where they are. The better alternative is to establish ZIP code-specific ceilings, that is, offer vouchers of varying size dependent on where they are applied: higher vouchers for better neighborhoods, lower vouchers in worse. The idea is to set vouchers high enough to induce voucher holders to move to higher-quality neighborhoods, and such that there is a subsequent rise in neighborhood quality.
Uniform vouchers, which pay the same subsidy across neighborhoods, penalize those who hope to move to better neighborhoods because they are not able to afford the higher rents. Tilting the rent ceiling, or raising vouchers to meet rents in high-quality neighborhoods and lowering them in low-quality neighborhoods, induces voucher recipients to consider better neighborhoods, and it does so without raising the cost of the program.
Some Background on Housing Vouchers
Housing Choice Vouchers (formerly known as Section 8 Vouchers) are a public program that uses the private market to provide rental units. The program served 2.3 million low-income households in 2016. The U.S. Department of Housing and Urban Development (HUD) works through local housing authorities, who administer the program and make payments on behalf of tenants to landlords. Tenants, though, are responsible for searching for units to lease on the private market.
Tenants with vouchers pay at least 30 percent of their income in rent, and the local housing authority pays the difference, up to a rent ceiling, which is from 90 percent to 110 percent of a federally-set “fair market rent” (FMR). FMR areas are defined using county boundaries, but in urban areas there is often a single FMR for all counties in a metro area.
Voucher holders renting units below the rent ceiling pay nothing when rents rise but remain below the ceiling, with the housing authority paying each extra dollar of an increase. This is important because when the rent ceiling increases, landlords can increase rents without worrying that this will cause the voucher holder to move.
A Natural Experiment in Dallas
The authors empirically estimate the impact of uniformly raising rent ceilings vs. rent tilting. Regarding uniform increases, estimate that a $1 increase in the rent ceiling raises rents by 46 cents over the following six years, while unit and neighborhood quality rise by only 5 cents over the same period. Importantly, the authors estimate no impact—a precise zero—on neighborhood quality as measured by Census tract median rent and tract poverty rate. This suggests that the benefit of this policy to landlords is eight times as large as the benefit in terms of observed quality to tenants.
Uniform rent ceilings mostly benefit landlords and do not improve quality of neighborhoods chosen by voucher recipients.
So how does rent tilting fare by comparison? The authors have a natural experiment to test its effectiveness: the Dallas-Fort Worth metro area, where housing authorities in Dallas switched to a tilted rent ceiling (also known as “Small Area Fair Market Rents” or SAFMRs) in 2011. Much as with the uniform rent ceiling increase, the authors find that landlords adjust rents accordingly—raising them in expensive ZIP codes and lowering them in low-cost ZIP codes. Because this policy makes vouchers more generous when they are used in high-quality neighborhoods, one might expect that it would improve neighborhood quality.
And so it did. The authors use neighboring Fort Worth, Texas, as a comparison group to reveal that new leases signed after the policy was implemented were in tracts where neighborhood quality is 0.23 standard deviations higher than leases signed prior to policy implementation.
(To measure neighborhood quality, the authors constructed an index using the violent crime rate, test scores, the poverty rate, the unemployment rate, and the share of children living with single mothers.)
Relative to other housing voucher policies, 0.23 standard deviations is a substantial improvement in neighborhood quality. It is about half the magnitude of the improvements in neighborhood quality for people currently living in public housing who are allocated vouchers. (See the accompanying map, which shows how tenants moved in response to tilted vouchers.)
In addition to improvements in neighborhood quality, the Dallas tilting policy is budget-neutral within the time period studied by the authors (2011-13). If no tenant had moved in response to the policy, it would have been cost-saving for the government, because voucher holders tend to live in inexpensive neighborhoods; therefore, rent increases in expensive ZIP codes were offset by larger decreases in low-cost ZIP codes. When the authors incorporate tenants’ improved neighborhood choices, the Dallas intervention had zero net cost to the government.
Tilting the rent ceiling induces voucher recipients to move to better neighborhoods, and it does so without raising the cost of the program.
The bottom line is that Ganong and Collinson’s analysis reveals that a simple budget-neutral reform to housing voucher design—moving from uniform rent ceilings to a rent-tilting scheme known as Small Area Fair Market Rents (SAFMR)—has the potential to substantially improve voucher holder neighborhood quality at no additional cost to the government.
The assertion that SAFMR improves voucher holder neighborhood quality will be testable beyond Dallas, as the rent-tilting program is currently in place in these markets:
• Chattanooga Housing Authority (TN)
• Housing Authority of Cook County (IL) (excludes Chicago)
• Housing Authority of the City of Laredo (TX)
• Town of Mamaroneck Housing Authority (NY)
• Housing Authority of Plano (TX)
• Housing Authority of City of Long Beach (CA)
In addition, two dozen other housing markets are set to implement SAFMR, including Atlanta, GA, and surrounding areas; Colorado Springs, CO; Pittsburgh, PA; San Diego-Carlsbad, CA; and Honolulu, HI.1