Both the Administration and Congress are considering proposals that would dramatically alter how the Federal Government provides housing assistance to low-income households. Beginning with public housing in 1937 and continuing with new programs in the 1960s and 1970s, the Federal Government subsidized the construction and rehabilitation of rental housing reserved exclusively for use at reduced rents by low-income persons and families. In 1974 the Federal Government began to allow low-income households to find rental units in the private market and to pay some or all of their rent and utility payments. The two approaches have come to be called "project-based" and "tenant-based" to distinguish whether the subsidy is tied to the housing unit or the household. In project-based programs, if a tenant family moves, they lose their subsidy. In tenant-based programs, if a tenant family moves, they take their subsidy with them. The proposals under consideration by the executive and legislative branches would convert some or most project-based programs to tenant-based programs. Therefore, it is important to understand how well tenant-based programs work.

HUD currently has two variants of the tenant-based approach, the Section 8 certificate program and the Section 8 voucher program. Drawing on recent studies by HUD's Office of Policy Development and Research and a forthcoming report, a great deal of information is available on how well these programs work. Because the differences between the Section 8 certificate and voucher programs are minor relative to the differences with project-based programs, this paper discusses them jointly.

Where do they work well?

Critics of tenant-based programs have argued that they don't work well in certain areas--rural markets, metropolitan markets with low vacancy rates, or suburban markets--or for certain groups-- the elderly, large families, or the handicapped. Critics also contend that many families won't use tenant-based programs because of the difficulty in finding units that meet both the rent and quality requirements and that have landlords who are willing to participate in the programs.

Data from the 1991 American Housing Survey (AHS) show that Section 8 certificate and voucher holders are found in all regions and in urban, suburban, and nonmetropolitan areas.¹ The distribution of program participants among the four Census Regions is very close to the distribution of the population of renters whose incomes would qualify them for these programs. Only in the Northeast is there any substantial difference.² Fourteen percent of Section 8 certificate and voucher holders live in the Northeast, compared to 19 percent of income-eligible renters.

The Section 8 certificate and voucher programs are actually more suburban and nonmetropolitan programs than central city programs. The proportion of program participants who live in suburbs--31 percent--equals the proportion of income-eligible suburban renters. The proportion of program participants in nonmetropolitan areas is greater than the proportion of income-eligible nonmetropolitan renters, 29 percent vs. 18 percent.

The suburban share may surprise some critics. In metropolitan areas HUD sets a rent level, called the Fair Market Rent (FMR), that determines which units are available under the Section 8 certificate and voucher programs.³ Because the FMR is based on rents throughout the metropolitan area, the FMR tends to be somewhat high relative to rents in the central city and somewhat low relative to rents in the suburbs. However, analysis of data from the 1990 census shows that moderately priced units, affordable under the FMR, can be found throughout metropolitan areas. In Table 1 HUD selected 12 diverse metropolitan areas and looked at rents, census tract by census tract. Eliminating a relatively small number of tracts with fewer than 10 two-bedroom rental units, HUD found that at least 70 percent of the census tracts in each metropolitan area had an adequate supply of modest-price units; that is, at least 30 percent of the two-bedroom rental units in a tract had rents lower than the FMR.

Although an early concern with the Section 8 certificate and voucher programs was the number of households that received certificates or vouchers but never used them, the success rate among program participants has improved. In 1993, 87 percent of enrollees in the Section 8 certificate and voucher programs, outside of New York City, successfully found a unit and received benefits. The most recent comparison point was the 1985-87 period, when a study of large urban public housing authorities found that 73 percent of program enrollees, outside of New York City, were successful. New York City is treated separately because rent control reduces moves by renters and because the New York City sample was disproportionately elderly and handicapped.⁴ In New York City the success rate improved between the two studies from 33 percent to 62 percent. Among elderly households 86 percent of participants, outside of New York City, found units and received payments.⁵

Over the past 20 years, most of the operational feasibility questions about tenant-based assistance have been answered positively. The Section 8 certificate and voucher programs provide assistance to almost 1.5 million households throughout the country.

Whom do they serve?

The Section 8 certificate and voucher programs serve a broad cross section of low-income households. The racial and ethnic composition of Section 8 certificate and voucher holders matches almost perfectly the racial and ethnic composition of income-eligible renter households. Sixty-four percent of Section 8 certificate and voucher holders are white, compared to 66 percent of eligible renters; 31 percent are black, compared to 30 percent of eligible renters; and 11 percent are Hispanic, compared to 13 percent of eligible renters.

Program participants have very low incomes. The median household income for Section 8 certificate and voucher holders in 1991 was $7,906, compared to $8,180 for eligible renters and $18,918 for all renters. In 1991, 48 percent of Section 8 certificate and voucher holders reported having wage or salary income; 45 percent received welfare or SSI payments. Among participants in the project-based programs, 36 percent reported having wage or salary income and 34 percent received welfare or SSI payments.

The Section 8 certificate and voucher programs serve families with children very well. Families with children represent 54 percent of all Section 8 certificate and voucher holders, compared to 43 percent of income-eligible renters. Single-parent households with children formed 32 percent of Section 8 certificate and voucher households, compared to 22 percent of income-eligible renters. Among project-based households 38 percent were families with children, and 22 percent were single-parent families.

The most striking difference between how participants and administrators have used project-based and tenant-based programs is the family/elderly split. As noted the Section 8 certificate and voucher programs serve families with children to a much greater degree than project-based programs. The opposite is true with respect to elderly households. Among Section 8 certificate and voucher holders, 21 percent have members 65 years old or older, compared to 26 percent among income-eligible renters and 41 percent for the project-based programs. Greater difficulty in using certificates and vouchers cannot explain the lower elderly participation because elderly households have a success rate almost equal to that of the typical household. An important explanation is the proportion of small units in the project-based buildings. Fifty-one percent of the units are either efficiencies or one-bedroom units; this compares to 28 percent of the units in the Section 8 certificate and voucher programs.

What housing options do they offer?

Section 8 certificates and vouchers differ fundamentally from their project-based alternatives in that they provide households with the option of receiving assistance in the unit in which they live. A recent HUD study estimated that 5.3 million income-eligible renters had severe housing problems in 1991, of which 3.9 million lived in adequate, uncrowded housing but were burdened by having to pay more than 50 percent of their income for rent.⁶ Solving the housing problems of these 3.9 million households does not require their moving, although many might choose to move if they had the opportunity. Actual Section 8 certificate and voucher program experience shows, outside of New York City, that 30 percent of participants rent in place. (In New York City, 61 percent rent in place.)

Outside New York City, 70 percent of Section 8 certificate and voucher holders move when they enter these programs.⁷ As explained earlier units affordable under the Section 8 certificate and voucher programs are available throughout a metropolitan area. Unfortunately, little is known about where these households choose to move. Another recent HUD study used data from a General Accounting Office analysis of the location of Section 8 households in four metropolitan areas: Oklahoma City, OK; Seattle, WA; Washington, DC; and Wilmington, DE.⁸ The analysis included both movers and households who elected to rent in place. In these four sites, 45 percent of households receiving certificates and vouchers obtain housing in census tracts where the poverty rate is less than 10 percent, and 91 percent obtain housing in tracts where the poverty rate is less than 30 percent. Fifty-nine percent of the households choose units in tracts where the black population is less than 20 percent. The comparable rates for participating black households are 36 percent in tracts where the poverty rate is less than 10 percent, 87 percent in tracts where the poverty rate is less than 30 percent, and 36 percent in tracts where the black population is less than 20 percent. Recognizing the limits of having only four sites, the study observed that the neighborhoods chosen by Section 8 certificate and voucher holders are generally less poor and less segregated than the neighborhoods surrounding conventional public housing projects.

The Section 8 certificate and voucher programs provide participants with a wider range of housing types from which to choose. In particular Section 8 certificate and voucher holders tend to make greater use of single-family structures than their counterparts in project-based programs. Thirty percent of Section 8 certificate and voucher holders choose to live in single-family detached structures, compared to 25 percent of all renters and 3 percent of all project-based participants. The high utilization of single-family structures reflects both the options available and the higher percentage of Section 8 certificate and voucher holders who have children. Conversely, only 7 percent of Section 8 certificate and voucher holders live in buildings with 50 or more units, compared to 9 percent of all renters and 34 percent of all project-based participants.

Section 8 certificate and voucher units compare favorably in quality with private market rental units. The 1991 AHS contains detailed data on the physical and neighborhood characteristics of rental units, both HUD-assisted units and private market units. Across the various questions, there appears to be no meaningful differences between Section 8 certificate and voucher units and all renter units in the frequency of problems with the units or their neighborhoods. In most cases certificate and voucher units have the same or a lower frequency of occurrence of specific problems. In other cases the frequency of occurrence is only one percentage point higher for certificate and voucher units. Out of 38 unit and neighborhood condition comparisons, the only exceptions are "problems with neighbors," with residents of certificate and voucher units reporting this problem 3 percentage points more than all renters (18 percent vs. 15 percent); and "trash, litter, and junk on streets and nearby properties," with residents of certificate and voucher units reporting this problem 5 percentage points more than all renters (22 percent vs. 17 percent). Only the problem with trash difference is statistically significant.

An important AHS finding is that Section 8 certificate and voucher holders who move seem to be happy with the move. Fifty-four percent say that their new unit is better than their old unit, and 47 percent say their new neighborhood is better. Only about 15 percent rate their new home or neighborhood as worse.

Conclusions

Based on the experience of the Section 8 certificate and voucher programs, tenant-based assistance passes all the major tests. Participant success rates are high; the programs are used effectively in many different settings, both regional and urban versus rural; and rental options are available throughout metropolitan areas. Tenant-based assistance appears to be especially effective both in serving families with children and in helping households living in adequate housing but paying an excessive portion of their income on housing. There are some unanswered questions. Little is known about how well the Section 8 certificate and voucher programs work for households with a disabled member. However, a recent analysis of HUD program data found that Section 8 certificate and voucher programs had a higher percentage of nonelderly households with disabilities (15 percent) than public housing (12 percent). More needs to be learned about how effectively participants use the option to move to improve their living environment and how to make this a more effective option. But the little that is known about mobility suggests that Section 8 certificate and voucher holders live in less poor and less segregated neighborhoods than their public housing counterparts.

Notes

1. At HUD's request the Bureau of the Census matched renter households from the 1991 AHS with lists of Section 8 certificate and voucher holders from Public Housing Authorities. The match identified 652 AHS households as participants in the Section 8 certificate and voucher programs. The results discussed here are contained in a forthcoming HUD report Characteristics of HUD-Assisted Renters and Their Units in 1991 by Connie H. Casey.

2. The Northeast Census region includes Maine, New Hampshire, Vermont, Massachusetts, Rhode Island, Connecticut, New York, New Jersey, and Pennsylvania.

3. Certificate holders cannot choose units with rents higher than the FMR. Voucher holders can choose units with rents higher than the FMR but must pay the full difference between the unit's rent and the FMR.

4. During the study period, the New York City Public Housing Authority used 80 percent of its Section 8 certificates and vouchers for homeless individuals and families. The study sampled from the remaining recipients were predominately elderly and handicapped.

5. The two studies referenced here are: Stephen D. Kennedy and Meryl Finkel, Section 8 Rental Voucher and Rental Certificate Utilization Study, Final Report, prepared by Abt Associates for the U.S. Department of Housing and Urban Development, Washington, DC, October 1994. Mireille L. Leger and Stephen D. Kennedy, Final Comprehensive Report of the Freestanding Housing Voucher Demonstration, prepared by Abt Associates for the U.S. Department of Housing and Urban Development, Washington, DC, May 1990.

6. Worst Case Need for Housing Assistance in the United States in 1990 and 1991: A Report to Congress, U.S. Department of Housing and Urban Development, Washington, DC, June 1994.

7. The Final Comprehensive Report of the Freestanding Housing Voucher Demonstration found that in 1987, 12 percent of the households already receiving assistance chose to move to a new unit during the first year of the study (see Volume I, page 171).

8. John Goering, Helene Stebbins, and Michael Siewert, Promoting Housing Choice in HUD's Rental Assistance Programs: A Report to Congress, U.S. Department of Housing and Urban Development, Washington, DC, April 1995.