Exploring a universal HOUSING voucher

INTERACTIVE REPORT
Vincent Reina, Claudia Aiken, and Jenna Epstein
September 21, 2021
What is a "universal voucher"? 
​
The Housing Choice Voucher (HCV) is a portable subsidy that households can use to rent a housing unit in the private market. The voucher covers the difference between 30% of the household's income and the full cost of rent (up to a maximum rent level set by the U.S. Department of Housing and Urban Development). This means that as household income fluctuates, so does the amount of the subsidy, ensuring that the household does not become housing cost burdened.

The HCV program is currently the U.S. Department of Housing and Urban Development’s (HUD’s) largest subsidy program, serving about 2.3 million households across the country. Households generally must not have incomes above 50% of the area median income in order to qualify for the benefit. But even at this scale, the number of vouchers available is far lower than the number of households eligible to receive one. In fact, we estimate that in 2019, only about one in five households eligible for a voucher actually received one and was able to use it. And meanwhile, a global pandemic has worsened the U.S.’ already severe shortage of affordable housing options.
​​These realities have added urgency to an ongoing national conversation about the need to expand the voucher program, so that every qualifying household can access assistance.
​Why a universal voucher for housing?
 
  • ​Research has shown that vouchers reduce housing instability and homelessness, and for many households, improve household outcomes like children's educational attainment and long-run earnings.
  • Vouchers significantly reduce nonpayment of rent, which is the leading cause of eviction. 
  • The analysis below suggests that a universal voucher program would serve large numbers of extremely low-income households with children, and - in combination with the Child Tax Credit - could lift up to 2.2 million of these households out of poverty. 
  • A universal voucher would also serve millions of households with seniors, helping them achieve housing stability despite fixed incomes, and households of color, who continue face discrimination and inequity in the housing market.
  • By expanding access to stable housing, vouchers would reduce the demand for, and cost of, other welfare services, such as Medicaid, food assistance, and homelessness services. 
  • Making the voucher universal would also likely reduce the program's negative effects on employment and earnings, since households would not need to fear permanently losing access to the subsidy if their income increased; they could re-access the benefit whenever it was needed.
  • Supply-side responses are important, too. Some housing markets in the U.S. have such a constrained supply of rental housing that an effective universal voucher program will require adding new units and rehabilitating existing units. The good news is that demand-side supports like vouchers can actually make supply-side responses more financially viable. For instance, tying some vouchers to specific units ("project-basing") can give property owners a guaranteed source of rental income that they can leverage to finance new construction and rehab.
  • In addition to pairing vouchers with new and existing financing mechanisms, policymakers must work to reduce the zoning and regulatory barriers that make it difficult to build.
  • Ultimately, a universal voucher moves the U.S. closer to recognizing a fundamental right to housing. The fact that we currently lack a national safety net for households facing housing insecurity represents a moral failure that transcends cost-benefit analyses.
The Voucher Gap​

  • As of 2020, there are approximately 2.6 million vouchers available through the HCV program, 88% of which are in use. Although selection varies by housing authority, federal rules require that at least 75% of households awarded a voucher must be extremely low-income, that is, have an annual income at or below 30% of area median income (AMI) or up to the poverty line, whichever is higher.
  • If a voucher were given to every renter household meeting HUD income limits for extremely low-income households (at or below 30% AMI) and not already living in public housing, there would be 10.4 million eligible households. This total would rise to 17.7 million eligible households if the voucher were universal for households with very low incomes (at or below 50% AMI). ​
  • The gap between the actual number of vouchers in use today, and the number that would be available under a universal voucher program, varies substantially across states and metropolitan areas. The gap is largest in places like California, Texas, New York, and Florida, which have hundreds of thousands of housing vouchers today, but which also have thousands more eligible households who have not received a voucher.
  • A more useful metric may be a weighted voucher gap ratio. The ratio of actual to potential vouchers is divided by the renter population. Thus, the states and metros with the highest values have the largest gaps in proportion to their renter populations.
  • For example, Wyoming currently has about 2,600 voucher households, but over 28,000 households would be eligible for a universal voucher at 50% AMI - a ratio of 1:11. The state also has a relatively small overall population of renters, only about 68,000 households. Thus, the voucher gap ratio is weighted more heavily. Wyoming has the highest weighted voucher gap ratio in the nation.
THE VOUCHER GAP RATIO, WEIGHTED BY RENTER POPULATION, BY STATE
This map shows the ratio of vouchers available in 2020 to the number of vouchers available under a universal voucher system, normalized by total renter population. Hover over any state to see a visual comparison between the current and estimated number of vouchers. Toggle between tabs to explore scenarios for a universal voucher at 30% and 50% of area median income. Zoom out to view Alaska and Hawaii.
THE VOUCHER GAP RATIO, WEIGHTED BY RENTER POPULATION, BY METROPOLITAN AREA
This map shows the ratio of vouchers available in 2020 to the number of vouchers available under a universal voucher system, normalized by total renter population. Hover over any metropolitan area to see a visual comparison between the current and estimated number of vouchers. Toggle between tabs to explore scenarios for a universal voucher at 30% and 50% of area median income. Zoom out to view metros in Alaska and Hawaii.
THE VOUCHER GAP RATIO, WEIGHTED BY RENTER POPULATION, IN THE 25 LARGEST U.S. METROS
Among the top 25 most populous metropolitan areas, Orlando, Charlotte, and Phoenix have the largest gaps in existing versus potential vouchers under a universal voucher program at 30% AMI, with respect to their overall renter populations. Chicago, Los Angeles, and New York have far larger absolute differences in existing versus potential vouchers, but only because they have very large overall numbers of renters.
Who would use a universal voucher? 
​
  • Not all households eligible for a voucher will use one. The most recent national study of voucher success rates estimated that 69% of households successfully lease a unit with the voucher they are offered; the same study also provided voucher success rate estimates for 50 public housing authorities across the country, which varied between 37% in Alameda and 100% in Detroit and Newark. More recent studies suggest that voucher success rates may be even lower, however, especially for seniors and households of color.
  • Research has shown that a variety of policy tools can increase voucher success rates and improve households' outcomes. One such tool is source of income (SOI) legislation, which prohibits landlords from rejecting prospective tenants who want to use a voucher towards their rent. Another is the Small Area Fair Market Rent (SAFMR) program, which adjusts voucher rent limits on a zip code basis, rather than setting a limit for the entire metropolitan area, thus enabling voucher holders to access neighborhoods with higher rents. There is some evidence that SAFMRs improve access to high-opportunity neighborhoods. The maps below show jurisdictions that have SOI protections and/or are implementing SAFMR.
  • Another factor affecting voucher use is length of stay. The median household that exited the HCV program in 2015 had used the voucher for 4.8 years. This duration varies by householder age, presence of children, and other variables. We can assume that in any given year, some households will be entering and others will be exiting the universal voucher program.
  • Finally, if a universal voucher program were implemented, new households might form to take advantage of the subsidy. Qualifying households currently composed of multiple families and living in overcrowded quarters would separate into multiple households, each one with its own voucher, since current voucher guidelines require this.
  • Based on these adjustments, we estimate that a universal voucher at 50% AMI would serve 5.6 million renter households in the average year (or 3.1 million renter households at 30% AMI), compared to the approximately 2.3 million households the HCV program serves today. We would expect this number to go up if more localities adopted SOI protections, or if more housing authorities implemented SAFMR and/or launched landlord engagement and mobility counseling programs.
  • Households using the voucher in the average year would make up about 13% of all renter households in the U.S (or 7%, if the voucher were universal at 30% AMI). This would vary somewhat by state, with the lowest share (9%) in Delaware—where renters tend to have relatively high incomes—and the highest (18%) in Michigan, where the average renter household’s annual income was only $32,701 in 2019.
UNIVERSAL VOUCHER ELIGIBILITY AND EXPECTED USE, BY STATE
Hover over any U.S. state in the map below to view state-level estimates of the current number of vouchers available, plus how many renter households would be eligible for and/or would use a universal voucher. Toggle between tabs to explore scenarios for a universal voucher at 30% and 50% of area median income. Zoom out to view Alaska and Hawaii.

Overlay data show the states and counties with source of income (SOI) legislation that helps protect voucher households from landlord discrimination, and in which voucher use rates are likely to be highest (these data are derived from the Poverty & Race Research Action Council, which provides an in-depth analysis of SOI protections updated April 2021). Another overlay shows Small Area Fair Market Rent (SAFMR) sites, where voucher rent limits are varied by zip code, which has the potential to increase voucher holders' access to high-opportunity neighborhoods.
Children, seniors, and households of color ​

  • As of 2020, about 41% of voucher users are households with children, 29% are households headed by seniors (aged 62 or older), and 70% are households of color.
  • Because of the demographic makeup of households with extremely low incomes, if the voucher program were to become universal to all households at or below 50% AMI, it would continue to benefit large numbers of children, elderly persons, and persons of color. Nationally, 6.1 million households with children would be eligible for a universal voucher (making up 34% of all households eligible), 4.8 million households with seniors (27%), and 9.5 million households of color (54%).
  • The shares are similar if the voucher program were to become universal to all households at or below 30% AMI (34%, 27%, and 55%, respectively).
  • If we only consider the households that are expected to use a universal voucher in any given year, the share that are households with children falls to 27%, because these households typically have the shortest lengths of stay in the voucher program. The share that are households with seniors rises to 42%, and the share that are households of color is 49%.
  • Unsurprisingly, states with large populations - Texas, California, New York and Florida - also tend to see the largest eligible numbers of households with children and/or seniors and of households of color. But D.C. and Louisiana actually have the highest numbers of eligible households with children in proportion to their overall renter households with children (over 30% each). D.C. and Massachusetts have the highest shares of renter households with seniors who would become eligible for a voucher (40% or higher). And Connecticut and Louisiana are among the states with the most eligible households of color compared to the overall number of renter households of color (34% and 33%, respectively).
CHILDREN, SENIORS, AND HOUSEHOLDS OF COLOR
​Use the toggle at the bottom left of the map to view the number of households with children, households with seniors (aged 62+), and households of color that would be expected to use a voucher in any given year under a universal voucher program. Hover over any state to see a comparison between the estimated number of households that would be eligible for, and the number that would be expected to use, a voucher. Zoom out to view Alaska and Hawaii.
Households lifted out of poverty ​

  • Today's Housing Choice Voucher program serves about 1.8 million households that have extremely low incomes, i.e. are at or below 30% of the area median income for a household of their size. The number of voucher holders living in poverty is likely somewhat lower than this number.
  • If the voucher program were to become universal at 50% AMI, a voucher would be available to nearly 10 million households currently living in poverty, and the value of the voucher alone could lift up to 4.9 million of these households out of poverty.
  • At 30% AMI, a universal voucher would be available to about 8.9 million impoverished households , and the value of the voucher could lift up to 3.9 million of them out of poverty.
  • Many prospective universal voucher households with children could also claim the 2021 Child Tax Credit. Combined, the two subsidies could lift up to 6.1 million households out of poverty at 50% AMI, or 5 million at 30% AMI.
  • Large states like Texas and New York would see hundreds of thousands of households lifted out of poverty, but California, New Jersey, and Maryland are the top three states for households lifted out of poverty in proportion to their overall renter households in poverty (over 49% in each). 
  • Over half (55%) of households lifted out of poverty by a universal voucher at 50% AMI would be households of color; the share is even higher (59%) at 30% of AMI. Nearly a third (32%) would be households with at least one child, and nearly a quarter (23%) would be households with seniors.
HOUSEHOLDS LIFTED OUT OF POVERTY
​Use the toggle at the bottom left of the map to view the number of households with children, households with seniors, and households of color expected to be lifted above the poverty line by the value of a universal housing voucher at 30% AMI.
Where in the U.S. would voucher households live? 
​
  • The majority (about 14 million, or 79%) of renter households eligible for a universal voucher at 50% AMI live in metropolitan areas. The remaining 3.7 million live in non-metropolitan, or rural, areas. This is not surprising, given that most renters living in metro areas.
  • Similarly, among renter households eligible for a universal voucher at 30% AMI, about 81% (8.4 million) live in metropolitan areas while the remaining 2.1 million live in non-metropolitan, or rural, areas. 
  • Among households expected to use the universal voucher in any given year, the Los Angeles metro area is home to by far the largest number, at over 334,000 households (or over 200,000 households in a 30% AMI universal voucher scenario), followed by New York, Chicago, and Houston.
  • Vouchers would serve an especially high share of households with children in Houston (38%), and an especially high share of households of color in both Los Angeles and Houston (over 70% each).
  • Rural voucher holders would be especially numerous in Georgia, where nearly 246,000 rural renters would be eligible for a universal voucher at 50% AMI (or 143,000 at 30% AMI) and nearly 78,000 (or 42,000 at 30% AMI) would be expected to use the voucher in a given year. Georgia as a whole does not have source of income protections, and very few cities and counties within the state have adopted such protections, leaving voucher holders vulnerable to discrimination in the private rental market.
  • Many of the metropolitan areas with the largest concentrations of prospective voucher households already have source of income protections in place, but there are notable exceptions, including Greater Houston, Phoenix, and Detroit.
  • Similarly, many of the housing authorities in metropolitan areas with the largest concentrations of prospective voucher households have already implemented Small Area Fair Market Rents, but exceptions include the Housing Authority of the City of Los Angeles, the Seattle Housing Authority, and the City of Phoenix Housing Authority.
UNIVERSAL VOUCHER ELIGIBILITY AND EXPECTED USE, BY METROPOLITAN AREA
Hover over any metropolitan area in the map below to view metro-level estimates of the current number of vouchers available, plus how many renter households would be eligible for and/or would use a universal voucher. Toggle between tabs to explore scenarios for a universal voucher at 30% and 50% of area median income.
UNIVERSAL VOUCHER ELIGIBILITY AND EXPECTED USE, BY RURAL AREA
Click on any state in the map below to view estimates of the current number of vouchers available in rural areas (that is, any area outside of a metropolitan area), plus how many renter households would be eligible for and/or would use a universal voucher. Toggle between tabs to explore scenarios for a universal voucher at 30% and 50% of area median income.
How much would a universal voucher cost? 
​
  • This report is the first installment in a two-pronged exploration. A detailed cost-benefit analysis of a universal voucher scenario is forthcoming.
Additional resources 
​
  • In her 2020 book, The Voucher Promise, Dr. Eva Rosen delves into the benefits of, and obstacles to, using a housing voucher in the U.S.
  • The Center on Budget and Policy Priorities has published extensively on the value of housing vouchers and the need to expand the voucher program.
  • The Urban Institute has estimated the cost of expanding the voucher program to everyone who qualifies. On June 9, 2021, Urban Institute researcher Mary Cunningham delivered testimony in support of a universal voucher to the House Financial Services Committee.
  • More information on the challenge that landlord participation presents to a universal voucher program is available from the Pew Charitable Trusts. 
  • Shelterforce explores the concept of a universal voucher in the context of current legislative proposals in Congress.
 Acknowledgments 
​
The Annie E. Casey Foundation and the Stoneleigh Foundation generously supported this research. We thank them for their support but acknowledge that the findings and conclusions presented in this report are those of the authors alone, and do not necessarily reflect the opinions of these funders.
METHODS


We used the following methodology to estimate voucher eligibility and use:

  1. We constructed a household-level dataset drawn from the household-level 2019 five-year American Community Survey estimates captured in the Integrated Public Use Microdata Series (IPUMS). This dataset includes the following variables for over 1.87 million renter households: state, county, and metropolitan area identifiers; household income; gross rent; the number of people per household; other basic demographics for householders and other household members; and a household weight that estimates how many other households in the country each individual observation represents.
  2. We subdivided overcrowded multi-family households in the dataset so that households consisting of more than one family and occupying a unit with more than one person per room would form separate households, each eligible for a voucher.
  3. For households in a metropolitan area, we matched each observation with the appropriate 2019 HUD income limit for “extremely low-income households.” These limits are set at 30% of HUD median family income or the poverty line (whichever is higher) and vary by the number of persons in the household.
  4. For households outside of a metropolitan area, we applied county-level HUD income limits where the county identifier was available, and state-level HUD income limits in all other cases.
  5. If a household’s income was below the limit based on their household size, they would be offered a voucher.
  6. We expanded the dataset based on IPUMS household weights so that an observation with a weight of 5 (indicating that it represented five households with similar characteristics) would be replicated the corresponding number times to estimate the total number of eligible voucher households by state, metropolitan area, and rural (non-metropolitan) area.
  7. Rural area estimates were derived by subtracting aggregated metro-level estimates from state-level estimates. Since some metropolitan areas cross state lines, we applied the ratio of renters living in each state-specific portion of the metro area to our voucher estimates in order to perform this calculation. 
  8. We then removed households currently living in public housing from the estimates by randomly selecting a sample of households that aligned with the characteristics of public housing households in the HUD Picture of Subsidized Households along age, race, and income lines. 
  9. We assumed that not all households would use the voucher, even if offered one. The most recent national study of voucher success rates found that households offered a voucher successfully use it 69% of the time. The same study also estimated success rates for 50 public housing authorities across the countries; we applied these rates to the relevant households, and the national rate when a local rate was not available.
  10. We also took into account differing lengths of stay. Kirk McClure’s 2018 study found that the median length of stay in the Housing Choice Voucher program for households that exited in 2015 varied between 4.1 years and 8.0 years based on householder age and presence of children, with older childless households using the voucher for the longest periods. We averaged these lengths of stay over 10 years to estimate how many households would be using a voucher at any one time.
Whether you're a policymaker, researcher, journalist, student, funder, or just someone who cares about housing - we would love to hear from you. Email us at housinginitiative@design.upenn.edu.