urban institute nonprofit social and economic policy research

Federal Housing Assistance and Welfare Reform

Uncharted Territory

Read complete document: PDF


PrintPrint this page
Share:
Share on Facebook Share on Twitter Share on LinkedIn Share on Digg Share on Reddit
| Email this pageE-mail
Document date: December 01, 1997
Released online: December 01, 1997

Number A-19 in Series, "New Federalism: Issues and Options for States"

The nonpartisan Urban Institute publishes studies, reports, and books on timely topics worthy of public consideration. The views expressed are those of the authors and should not be attributed to the Urban Institute, its trustees, or its funders.


Federal housing assistance was seldom mentioned in the mid-1990s’ debate over devolution of America’s social safety net. Yet in FY 1995, federal outlays for housing assistance to the poor ($19 billion) exceeded those for Aid to Families with Dependent Children (AFDC) by $7 billion.

A sizable share (about one-fifth) of households that receive AFDC also benefit from federal housing subsidies administered by the U.S. Department of Housing and Urban Development (HUD). Moreover, those receiving HUD assistance account for a much larger share of long-term welfare recipients—those likely to have the most difficulty finding and retaining employment—than welfare families that don’t receive federal housing assistance. Among AFDC beneficiaries in 1994, for example, the median cumulative period of welfare recipiency for those who also received HUD assistance was 57 months; for those not receiving HUD assistance, the comparable period was 37 months. 1

Whether or not welfare recipients also receive housing assistance will greatly influence the immediate circumstances and, possibly, the longer-term opportunities of those directly affected by welfare reforms and cutbacks in related social programs. Welfare reform may also have a marked impact on the financial condition of HUD’s housing programs. Tremendous variations in HUD assistance across states and localities (explained below), together with the new discretion states have been given to run their own welfare programs, mean that housing assistance and welfare interactions at the local level will significantly affect state responses to devolution.

What are the possible outcomes of interactions between housing assistance and welfare reform? We preface our speculation with a discussion about how federal housing assistance differs from many other safety net programs, the forms it takes, who it serves, and the relevancy of housing reforms to welfare reform.

What’s Different about Federal Housing Assistance?

Federal housing assistance is provided to reduce the rents that low-income households have to pay for decent housing. At least three features distinguish it from many other forms of federal assistance to the poor.

First, unlike the AFDC program, housing assistance has never been an entitlement. Eligible households apply for it and are put on waiting lists, where they often remain for years. The number of families selected for assistance depends mainly on how much funding Congress has made available. Only about one-quarter of all eligibles actually receive assistance. In the recent past, federal criteria have generally focused on the poorest of the poor.

Second, unlike welfare benefits, only a small fraction of federal housing assistance is allocated by the states. For the major housing assistance programs, funds are allocated to localities directly by HUD (based on a mix of need and administrative readiness). The states are not involved. A large share of the money goes to local Public Housing Authorities (PHAs), which are agents of local governments, but some goes to local governments directly and some to private for-profit and nonprofit developers for specific projects. HUD’s major programs that provide ongoing assistance to poor households have no state or local matching requirements. Furthermore, few significant independent state and local housing assistance programs exist.

Third, housing assistance has not suffered from the wide variations in state-determined benefit levels that have plagued welfare (although the quality of assisted housing and its management varies widely). Although federal housing assistance is a highly inequitable form of social support (some needy households get substantial benefits while others with the same income levels get nothing), among those who do receive it, the assistance is relatively equitable. The payment formula is uniform nationwide and takes differences in local housing costs into account. In most cases, tenants pay 30 percent of their income toward rent, and HUD covers the remainder.

Forms and Beneficiaries of Housing Support

Federal housing assistance is provided to households in one of three distinct programmatic forms, each of which has had a very different history.

Public housing projects are owned and operated by local PHAs for occupancy by low- and moderate-income households. The federal government covers the costs of development and the sizable share of operating and modernizing costs not covered by tenant rent payments. The Public Housing Program was authorized in 1937 but went out of political favor in the late 1970s for a number of reasons (including high costs and concentration of social problems). Since then, few new units have been built, and with the gradual removal of older buildings, the total stock has been declining. In 1993, a total of 1.1 million households lived in public housing.

Publicly assisted housing refers to privately owned and operated projects for which HUD providessubsidies to supplement what eligible tenants can afford to pay for rent. This strategy became a significant alternative to public housing in the late 1960s. Several different variations within this approach (differing combinations of tax, interest, and operating subsidies) were attempted over the subsequent decade, the most recent being the Section 8 New Construction and Substantial Rehabilitation programs. In all of these, HUD awarded funds directly to prospective private developers who applied for them. The Reagan administration stopped any further development of publicly assisted housing in 1983. But a large number of poor families still live in projects developed under these programs (1.7 million households in 1993), and federal subsidy obligations for them remain substantial.

A new approach that was introduced in 1987 and fits the "publicly assisted" definition is the Low-Income Housing Tax Credit (LIHTC). In contrast to earlier programs, largely developed by profit-oriented firms, a much larger share of the LIHTC developers comprises nonprofits. The LIHTC alone does not allow developers to reduce rents enough to make them affordable to very low income households. But cities can add on funds they receive from HUD’s continuing block grants to localities—the Community Development Block Grant (CDBG) and the HOME program—to make a share of LIHTC units more affordable. At this point, however, the number of households assisted in such projects remains small in relation to the total still living in projects supported by the earlier programs in this category.

Tenant-based assistance allows eligible households to select their own units in the private housing market and receive a subsidy to cover part of their rent. Introduced in the mid-1970s, tenant-based assistance has been given in two forms—Section 8 certificates and vouchers. With recent rule changes, however, there is now very little difference in the way they operate. This approach enjoys bipartisan support and has accounted for the bulk of the new housing assistance provided since the early 1980s. The number of households assisted in this way has grown consistently since then, reaching a total of 1.2 million in 1993.

In 1993, a total of 4.05 million households benefited from the main HUD programs that provide ongoing assistance based on income—28 percent from public housing, 42 percent from the older publicly assisted projects, and 30 percent via tenant-based assistance (table 1). Each of the three forms serves a mix of types of low-income households, but the mixes vary. High shares of households in public housing (46 percent) are headed by nonelderly women and receive income from welfare or Supplemental Security Income (44 percent). Those receiving tenant-based assistance also have sizable shares in the same categories (60 percent and 45 percent, respectively). In contrast, only 40 percent of the households in publicly assisted projects are headed by nonelderly females and 28 percent receive welfare or SSI income. Households headed by African Americans account for 54 percent of the total in public housing, 35 percent in publicly assisted housing, and 34 percent in tenant-based assistance.

AFDC Quality Control data indicate that out of a total of 4.98 million AFDC households in 1993, 1.06 million, or 21 percent, also received HUD assistance. And, as noted earlier, HUD programs account for an even higher proportion of long-term welfare households—those recipients most likely to lose support in the current reform process and to have a harder time finding employment.

Housing Reforms—Proposed and in Process

Outcomes for families that receive both housing assistance and welfare will be influenced by recent reforms in HUD programs as well as the welfare reform process.

While the 104th Congress was garnering headlines for its proposals to reform most federal safety net programs, the boldest reforms for housing assistance were inventions of the Clinton administration. HUD Secretary Cisneros proposed consolidating several small categorical programs into major block grants to localities, significantly restructuring publicly assisted housing programs (to reduce federal subsidies and enhance their financial viability), and transforming the operations of public housing. Over a phase-in period, public housing residents would be given tenant-based assistance and could choose to stay in their old projects or move elsewhere. PHAs would have to compete with other providers of rental housing to retain tenants. 2

In the short term, Cisneros also proposed giving PHAs much more flexibility in spending their capital allocations (including more freedom to demolish old projects), creating measures to encourage income mixing in public housing and publicly assisted projects, and expanding social service and job training programs to help current public housing residents move toward self-reliance. These steps sought to address the feature of large public housing projects that has drawn the sharpest criticism: the spatial concentration of the poor and, in particular, of poor minorities.

Congress has stopped short of endorsing the full proposal for transforming public housing in the original Cisneros package, though several themes of draft legislation emerging from congressional committees are variants of those originally suggested by Cisneros. It is noteworthy that, in contrast to its stance on many other social programs, Congress has not proposed to fundamentally alter the federal role in housing assistance by transferring any administrative or allocative authority to the states.

So far, Congress has failed to enact permanent housing reform legislation, but it has implemented some changes for public housing (via appropriation bills) that now seem likely to be sustained. These include eliminating past federal selection preferences and establishing ceiling rents 3 (both changes are designed to facilitate income mixing) and suspending past restrictions that inhibited demolition.

HUD’s budget has not suffered dramatic cuts, but fiscal constraints are clearly being felt. The agency’s budget authority dropped from $20.1 billion in FY 1995 to $19.4 billion in FY 1997, and is scheduled to go up again to $25.1 billion in FY 1998. However, this last increase is meant to cover Section 8 contract renewals, not to expand service. 4 In fact, the most important change reflected in congressional action on HUD’s budget over the past few years has been the explicit elimination of funding for any net growth in the number of households receiving HUD assistance. This move terminated a bipartisan commitment to growth increments that had been in force for over two decades.

For the time being there are no indications of major reductions in support for current beneficiaries, but there is a considerable amount of belt tightening related to that support. For example, Congress has not fully funded HUD’s estimates of needs to cover public housing operating expenses. Also, Congress has required a time gap in certifying new households to replace those that have left a HUD subsidy program, and now allows PHAs to charge tenants a minimum rent of up to $50 per month, even when that amount exceeds 30 percent of income. These changes aim to slow down expenditure rates and reduce overall subsidy needs.

How Housing Assistance Might Affect Welfare Reform Outcomes

It is impossible to predict exactly how housing assistance will influence outcomes for families affected by welfare reform. Below we speculate about the possibilities and, in so doing, identify issues that state officials will need to watch carefully as the welfare reform process unfolds.

Increased risks and reduced opportunities for those not receiving housing assistance.

For current welfare recipients not now receiving housing assistance, the elimination of the growth increment in the housing budget means there is a substantially reduced chance that they will be able to receive such assistance in the future, even if they are eligible for it. New slots will open up only as current beneficiaries move off the housing assistance rolls. Moreover, as PHAs move to achieve greater income mixing, the most needy are likely to get a smaller share of the replacement slots than they did in the past. Clearly, the costs of shelter will play a critical role in the pressures put on those who lose welfare income and do not benefit from HUD assistance.

A recent HUD study estimates that 5.3 million renter households nationwide do not receive housing assistance but have "worst case housing needs," meaning they live in seriously substandard housing or pay an extremely high share (more than 50 percent) of their monthly income for rent. 5 Only a few families are in this group because the quality of their housing is deficient—for 89 percent of them, excessive rent burden is their only housing problem. Families that have to pay so much for rent are particularly vulnerable when their income is reduced for any reason. Many whose welfare and other benefits are cut back in the reform process will not only need to find a job, but one that keeps their income minimally at its previous level to let them continue making their rent payments and avoid eviction.

An important cushion, and special service opportunities, for those who do receive housing assistance.

Welfare terminees already receiving HUD subsidies, however, are in a safer position. If they are unable to find employment, by law their housing subsidy will increase when they lose their welfare income. 6 Some may be moved out of projects and be given tenant-based assistance instead. Either way, the bulk of the costs of adequate housing will be covered by the government regardless of how well these welfare terminees do in the labor market. As noted above, housing costs loom very large in the budgets of most poor families, so HUD subsidies represent a very important cushion for welfare terminees.

Another possible benefit for those who live in public housing or publicly assisted projects is that having a sizable number of welfare recipients living together in the same complex creates opportunities and potential efficiencies for service delivery that would not exist if they were spread out in different units in the private housing market. HUD is now endorsing program initiatives to provide job training, job search assistance, and other social services within its housing projects. Such programs have not yet been funded at a significant scale, but some are under way, notably through the HOPE VI program. Further, HUD is now funding a major demonstration program—Jobs Plus—that will examine the effects of saturation-level welfare-to-work services in a number of other public housing developments not supported under HOPE VI.

Through these initiatives, for the first time HUD is explicitly recognizing a responsibility to help prepare its tenants for upward mobility. Some observers, in fact, are talking about making public housing more of a "way station" that facilitates economic mobility for the poor more generally. 7 How well this may work has yet to be tested. It is not clear that HUD and the PHAs know how to accomplish this in practice.

Fiscal uncertainty for the housing programs and the chance of greater inequities over the long term.

If the transition from welfare to jobs works effectively for housing assistance recipients, pressure on HUD’s budget could begin to decline. If it does not, however, HUD and local PHAs face some difficult challenges. Subsidies for their existing housing beneficiaries would need to be increased. Even if Congress were to hold total funding for HUD constant, we would probably see a net reduction in the number of low-income families served over time. As normal turnover occurs each year, PHAs would be able to bring in fewer new recipients from waiting lists than the number that had departed. HUD’s current service ratio (about 25 percent of eligible households) would go down, and project operating efficiencies would also be put under strain. In any given project, as the number of tenants declines, it is normally impossible to cut the costs of adequate maintenance and operating expenses proportionately.

How the developments above occur will vary across states and among localities within states, depending on the extent and composition of HUD programs and the degree of overlap between HUD assistance and welfare recipiency.

Variations in HUD assistance run along two dimensions—the state level and the local level.

State Variations. Since HUD has allocated housing assistance on the basis of how aggressive local housing authorities and developers have been in seeking and using it, as well as on the basis of estimated need, we would not expect the distribution of assistance across states to perfectly match the distribution of low-income households. None-theless, the variations are substantial. As shown in table 2, the total number of households receiving housing assistance ranges from only 5,000 in Wyoming to 417,000 in New York. More meaningfully, however, states also vary widely in the degree to which the HUD assistance they receive matches their share (relative to the nation) of low-income households. As shown in the fourth column of table 3, in 1993 in five states—Massachusetts, New Hampshire, New York, Rhode Island, and South Dakota—shares of housing assistance recipients were at least one-third above the state’s share of the nation’s low-income households. In Massachusetts, for example, the ratio between the share of families receiving housing assistance and the share of that state’s low-income families is 1.47; in Rhode Island it is 2.03. By contrast, in five other states—Arizona, Florida, Idaho, Utah, and Vermont—shares of housing recipients were at least 25 percent below that state’s share of U.S. low-income households.

States also vary significantly in the mix of the three programmatic forms of HUD assistance they receive. For example, as seen in table 2, in states such as Alabama, Alaska, Georgia, Oklahoma, South Dakota, and Tennessee, a high proportion (over 40 percent) of assisted households were in public housing in 1993. In California and Iowa, however, only a small portion (11 percent) of assisted households were in public housing. States where publicly assisted housing is dominant (shares at or above 60 percent) include the District of Columbia, Indiana, Michigan, and New Hampshire. Alaska had only 19 percent of assisted households in that category. In California, Idaho, Maine, Oregon, and Vermont, over 45 percent of assisted households received tenant-based assistance in 1993, while in Tennessee only 7 percent did.

In addition to variation among states in the amounts and types of HUD assistance received, there is little consistency among states in the overlap between housing assistance and AFDC recipiency. As seen in the sixth column of table 3, in nine states, a third or more of the AFDC caseload also received HUD assistance in 1993 (Alabama, Arkansas, Connecticut, Delaware, the District of Columbia, Massachusetts, Montana, Nebraska, and North Dakota). In contrast, the share of AFDC recipients also receiving housing assistance in California is only 8 percent, and below 16 percent in Arizona, Florida, Michigan, and Wisconsin.

Local Variations. Differences in program conditions such as those shown above between states may be even more striking between cities within states. The local scene is crucial because many of the decisions that will affect how HUD assistance impacts welfare reform will be made by local actors, particularly local PHAs.

HUD is already expanding the discretion of the PHAs to devise and implement their own operating and development strategies. Overall, the agency has been pressuring them to demolish their worst projects—mostly large high-rise developments—where social problems have been most concentrated in the past. But responses to that and other new pressures may well differ, depending on local political and economic factors. Some PHAs are building new housing, typically low-rise, to replace some of the units demolished, and some are looking for other creative ways to redevelop existing structures (where possible, to attract higher-income residents to projects that are well located). Other PHAs may consider phasing out of the business altogether, contracting out management responsibilities or even transferring housing title to local nonprofits. In most cases, some existing project residents will be displaced, be given tenant-based assistance, and have to find suitable apartments in the private housing market. Some PHAs may emphasize this strategy, pouring sizable numbers of new tenants into the local market in a short period of time.

Some of the most interesting questions about local variations relate to spatial effects within cities. Much of the public and publicly assisted housing stock is concentrated in some of the most distressed inner city neighborhoods. The more optimistic hypothesis is that reducing that stock and giving people tenant-based assistance will lessen the concentration of the poor, enabling more of them to move out to better neighborhoods with better housing. The pessimists argue that many of those with vouchers will not be able to find housing that meets program rent and quality requirements and that the results will be substantial increases in homelessness.

The impact of these changes on the welfare reform process will also vary depending on the extent of the local welfare/HUD assistance overlap. For some PHAs, welfare reform may be a side issue; for PHAs that house the dominant share of the affected welfare households in their localities, welfare reform may be central. In addition, PHAs are likely to exhibit different degrees of enthusiasm and skill in mounting on-site service programs to help tenants move toward jobs and self-sufficiency. These variables will most certainly affect outcomes for welfare terminees who receive housing assistance.

State officials have little ability to control such outcomes directly, but changes in how housing assistance and welfare interact in individual localities will have important impacts on overall state performance in devolution. States may want to monitor local performance and encourage working partnerships among local housing agencies, welfare offices, and social service providers. Coordination would enhance coherence in local responses to a very complex set of challenges.


Notes

1. See Sandra J. Newman and Joseph Harkness, "The Effects of Welfare Reform on Housing: A National Analysis." Presented at the Policy Research Roundtable on the Implications of Welfare Reform for Housing, sponsored by the Fannie Mae Foundation and the Johns Hopkins University Institute for Policy Studies, Baltimore, Md., July 22, 1997.

2. Henry Cisneros’s proposals are documented in U.S. Department of Housing and Urban Development, Reinvention Blueprint, December 19, 1994, and further detailed in U.S. Department of Housing and Urban Development, HUD: Reinvention: From Blueprint to Action, March 1995.

3. In the past, when public housing tenants found work and increased their income, they had to pay 30 percent of all such increases as rent payments to remain in the project. If the increase was substantial, this provided a strong incentive for them to move out. Under the new approach, they pay 30 percent of their income only up to the ceiling rent. They never have to pay more than the ceiling rent regardless of how much their income goes up. This policy is aimed at encouraging tenants other than the very low income to stay in public housing, thus increasing the income mix among tenants.

4. "House Panel Okays HUD Funding Bill with $9.2 Billion for Section 8 Renewals," Housing and Development Reporter Current Developments, vol. 25, no. 8, June 30, 1997.

5. U.S. Department of Housing and Urban Development, Rental Housing Assistance at a Crossroads: A Report to Congress on Worst Case Housing Needs, Washington, D.C.: U.S. Department of Housing and Urban Development, 1996. Of the 5.3 million total with worst case housing needs, almost half (2.4 million) are welfare or SSI beneficiaries.

6. Proposals have been made to prohibit increases in housing subsidies when welfare recipients lose welfare benefits due to sanctions (i.e., noncompliance with program rules). But as of September 1997, these proposals have not been enacted. There is no law preventing such housing assistance increases when welfare benefits are terminated due to time limits.

7. Many of the original proponents of public housing saw it playing this role, but over the years the idea was subverted as legal and rule changes led to concentrating the poorest of the poor in public housing projects. See, for example, Sandra J. Newman and Ann B. Schnare, Beyond Bricks and Mortar: Reexamining the Purpose and Effects of Housing Assistance, Washington, D.C.: Urban Institute Press, 1992.


Tables & Figures

Table 1

Table 2
Table 2 cont.

Table 3 Table 3 cont.


About the Authors

G. Thomas Kingsley is director of the Urban Institute’s Center for Public Finance and Housing, and manager of the National Neighborhood Indicators Project. He previously served as director of the RAND Corporation’s Housing and Urban Policy Program and as assistant administrator for the New York City Housing and Development Administration.



Topics/Tags: | Governing | Housing | Poverty, Assets and Safety Net


Usage and reprints: Most publications may be downloaded free of charge from the web site and may be used and copies made for research, academic, policy or other non-commercial purposes. Proper attribution is required. Posting UI research papers on other websites is permitted subject to prior approval from the Urban Institute—contact publicaffairs@urban.org.

If you are unable to access or print the PDF document please contact us or call the Publications Office at (202) 261-5687.

Disclaimer: The nonpartisan Urban Institute publishes studies, reports, and books on timely topics worthy of public consideration. The views expressed are those of the authors and should not be attributed to the Urban Institute, its trustees, or its funders. Copyright of the written materials contained within the Urban Institute website is owned or controlled by the Urban Institute.

Email this Page