RENTAL HOUSING ASSISTANCE¥ THE CRISIS CONTINUES The 1997 Report to Congress on Worst Case Housing Needs April 1998 Office of Policy Development and Research U. S. Department of Housing and Urban Development Table of Contents Letter to Congress Executive Summary Chapter 1: Introduction Terms and Sources Building the Report Chapter 2: Findings Major Findings Finding 1: Despite robust economic growth between 1993 and 1995, the number of very-low-income renters with worst case housing needs remained at an all-time high -- 5.3 million Finding 2: The stock of rental housing affordable to the lowest income families is shrinking and Congress has eliminated funding for new rental assistance since 1995 Finding 3: The fastest growth in worst case needs in the 1990s was among working families Finding 4: One of every three households with worst case needs now lives in the suburbs Supplementary Findings Finding 5: The most serious housing needs are concentrated among households with the lowest incomes Finding 6: Of the 12.5 million persons in households with worst case needs, almost 1.5 million are elderly and 4.5 million are children. The number of adults with disabilities in households with worst case needs is estimated between 1.1 and 1.4 million Finding 7: Worst case needs continue to shift to the West Chapter 3: Policy Implications Appendix A: Data on Housing Problems Appendix B: Glossary Appendix C: Procedures Used To Estimate Housing Needs From American Housing Survey Data Exhibits Exhibit 1: Income Categories Used in Housing Programs Exhibit 2: Extremely Low, Very Low, and Low Income: Examples of Income Limits for Four-Person Households Exhibit 3: Housing Assistance and Affordable Housing Programs Exhibit 4: Household-Level Data From HUD Programs Exhibit 5: Changes to Assisted Housing Policy: Recent and Proposed Exhibit 6: Worst Case Needs Dropped in the Mid-1980s, but Not in the 1993-1995 Recovery Exhibit 7: Annual Increase in Number of Households With Public and Assisted Housing, 1978-1996 Exhibit 8: Loss of Private Market Units Affordable for Very-Low-Income Renters Accelerated Between 1993 and 1995 Exhibit 9: Mismatches Between Extremely-Low-Income Renters and Rental Units They Can Afford Continue to Worsen Exhibit 10: Between 1991 and 1995, Growth in Worst Case Needs Was Highest Among Working Families and Individuals Exhibit 11: Substantial Numbers of Very-Low-Income Families With Children Are Working Exhibit 12: Over 1.6 Million Unassisted Working Renters With Worst Case Needs Have Extremely Low Incomes Exhibit 13: Many Assisted Families Have Earnings as Their Primary Source of Income Exhibit 14: In the Early 1990s, Worst Case Needs Grew Quickly in the Suburbs Exhibit 15: Mismatches Between Extremely-Low-Income Renters and Available Rental Units They Can Afford are Worst in the Suburbs Exhibit 16: Over Two-Thirds of Renters With Priority Problems Have Income Below 30 Percent of Median Exhibit 17: Renters With Income Below 30 Percent of Median Are the Only Groups Likely To Have Severe Housing Problems Exhibit 18: Housing Assistance Is Well Targeted to the Income Groups With Priority Problems Exhibit 19: History of the Federal Preference System Exhibit 20: Some Public Housing Projects Exclude the Poor Exhibit 21: The HOME Program Serves Extremely-Low-Income Households at Affordable Rents, Especially When Combined With Tenant-Based Rental Assistance Exhibit 22: Designated Housing Exhibit 23: Worst Case Needs by Household Type, 1995 Exhibit 24: Priority Problems Concentrate in the Poorest Households of Each Family Type Exhibit 25: Western Renters Are Underserved Relative to Needs Exhibit 26: Mismatches Between Extremely-Low-Income Renters and Available Rental Units They Can Afford are Worst in the West ---------- Letter To Congress U.S. DEPARTMENT OF HOUSING AND URBAN DEVELOPMENT THE SECRETARY WASHINGTON, D.C. 20410-0001 April 1998 TO THE CONGRESS OF THE UNITED STATES: I have the privilege of transmitting Rental Housing Assistance -- The Crisis Continues, the worst case housing needs report for 1997 as requested by the Senate Appropriations Committee in 1990. The report presents clear and compelling evidence of deep and persistent housing problems for Americans with the lowest incomes. The report makes it clear that the Federal government must significantly enhance its efforts to create more affordable housing opportunities for these households. This report has four major findings: * Despite robust economic growth between 1993 and 1995, the number of very-low-income American households with "worst case" housing needs remained at an all-time high -- 5.3 million. Households with worst case needs are defined as unassisted renters with incomes below 50 percent of the local median who pay more than half of their income for rent or live in severely substandard housing. * The stock of rental housing affordable to the lowest income families is shrinking and the Congress has eliminated funding for new rental assistance since 1995. Between 1993 and 1995 there was a loss of 900,000 rental units affordable to very-low-income families (those with incomes below 50 percent of area median income), a reduction of 9 percent. There was an even greater reduction -- 16 percent -- in the number of units affordable for extremely-low-income renters, those with incomes below 30 percent of area median. Since 1995, Congress has denied the Administration's requests for new rental assistance to mitigate worst case needs. This is a historic reversal of Federal housing policy. From the Great Depression until 1995 -- under both Democratic and Republican Administrations and in periods of economic boom and recession -- Congress always expanded the availability of rental assistance. As the report makes clear, it should do so now. * From 1991 to 1995, worst case needs increased the fastest among the working poor. Full-time work should provide a family with an income sufficient to afford a decent place to live. This study documents a disturbing trend in the opposite direction. In the economic recovery between 1991 and 1995, worst case needs increased fastest among working households. Between 1991 and 1995, the number of working poor households with worst case needs increased by 265,000, or 24 percent. In 1995, almost 1.4 million of the 5.3 million households with worst case needs had earnings equivalent to a full-time worker at the minimum wage. * One of every three households with worst case needs now lives in the suburbs. While the greatest numbers of worst case needs are in central cities, a large and fast-growing number of households in need of assistance live in the suburbs. More than 1.8 million households with worst case needs -- or one of every three worst case households -- lived in the suburbs in 1995, an increase of 9 percent from 1991. The data in this report confirm what housing practitioners already know about the overwhelming unmet need for housing assistance. More than one million families are on waiting lists nationwide forcing families to wait years before getting assistance. When the waiting list for Chicago's Section 8 certificate and voucher program was opened for two weeks in the summer of 1997, 100,000 applications poured in. Administrators were forced to tell three-quarters of these applicants that they would have to wait more than five years for assistance. There are 17,700 households on waiting lists in San Diego, 12,000 in suburban St. Louis County, and 5,000 in Spokane, Washington. The report's findings make a clear and compelling case for greater Federal attention to housing needs. Economic growth alone will not ameliorate the record-level housing needs among families with limited incomes. Not even families working full-time at minimum wage can afford decent quality housing in the private rental market. The report also makes clear that this is not just a big city problem, but affects America's growing suburbs as well. To address these problems, the Administration has asked Congress to resume the expansion of rental assistance for low-income Americans. President Clinton's fiscal year (FY) 1999 budget requests 103,000 Section 8 units for families and individuals who cannot afford to rent decent housing. Fifty thousand of these housing vouchers would be used by families making the transition from welfare to work. The Administration has also proposed expansion of programs to help build and rehabilitate more affordable housing. The FY 1999 budget expands funding for the HOME program and proposes a new HOME Bank, a new loan guarantee feature allowing communities to leverage up to five times their annual grant amounts for larger scale projects. The Administration also proposes a substantial expansion of the Low Income Housing Tax Credit program that would produce 180,000 affordable rental units over the next five years. Additional rental assistance is an important complement to these programs, to ensure that all HOME and tax credit housing is affordable to even the lowest income households. Congress is also considering legislation that will determine the income levels of households who will be admitted to public housing and be provided Section 8 rental assistance. The Clinton administration and HUD strongly support the transformation of public and assisted housing developments into healthier, mixed-income communities. But policymakers must be careful not to exclude poor families altogether from these housing developments, nor to increase by Federal policy the numbers of families with worst case needs. This report shows that mixed-income communities can be achieved while still continuing to serve families who are working but who have low incomes and serious housing needs. I look forward to working with the Congress to try to reverse the disturbing trends documented in this report. Sincerely, Andrew Cuomo ---------- Executive Summary This Congressionally mandated report documents four major findings: * The persistence of a housing affordability crisis for very-low-income renters despite the robust economic growth of the 1990s. * A reduction in the stock of affordable rental housing and the elimination of new Federal rental assistance, beginning in 1995. * A sharp increase in needs for rental housing assistance among the working poor. * The increased suburbanization of housing needs. These findings have significant implications for Federal housing policy. To begin to ameliorate this severe housing crisis, Congress should resume the expansion of Federal tenant-based rental assistance targeted to those with the most severe needs. In addition, Congress should expand programs like HOME and the Low Income Housing Tax Credit which subsidize the construction and rehabilitation of affordable rental housing. Finally, Congress should to the maximum extent possible continue to focus scarce Federal public and assisted housing opportunities on those households with the most severe housing needs, while still fostering a greater income mix in public and assisted housing developments. Major Findings Finding 1: Despite robust economic growth between 1993 and 1995, the number of very-low-income renters with worst case housing needs remained at an all-time high -- 5.3 million. * In 1995, 5.3 million very-low-income renters without housing assistance paid over half their income for housing or lived in severely substandard housing. Households with worst case needs are defined as unassisted renters with incomes below 50 percent of the local median who pay more than half of their income for rent or live in severely substandard housing. Renters with incomes below 50 percent of area median income are called "very-low-income" renters. * Households with the lowest incomes are most likely to have worst case needs. Almost 70 percent of unassisted renters with extremely low incomes had worst case housing needs in 1995. In this report, renters with incomes below 30 percent of area median income are called "extremely-low-income" renters. Nationally, 30 percent of median income approximates the poverty level, but the definition of extremely low income is adjusted for geographical differences. As a national average in 1998, "extremely low income" means less than $13,590 for a family of four or less than $10,872 for a two-person family. Finding 2: The stock of rental housing affordable to the lowest income families is shrinking, and Congress has eliminated funding for new rental assistance since 1995. * Between 1993 and 1995, there was a loss of 900,000 rental units affordable to very-low-income families, a reduction of 9 percent. There was an even greater reduction -- 16 percent -- in the number of units affordable for "extremely-low-income" renters. * Federal housing policy has done little to ameliorate these problems. Since 1995, Congress has denied Administration requests for new rental assistance and ceased funding for new incremental rental assistance to serve families with worst case needs. This is a historic reversal of Federal housing policy, which had continuously expanded Federal rental assistance in every year prior to 1995. From 1978 through 1982, an average of 224,000 additional households were provided Federal rental assistance each year. The average number of new households getting assistance dropped to approximately 146,000 during the 1980s and early 1990s. Finding 3: The fastest growth in worst case needs in the 1990s was among working families. * Full-time work should provide a family with an income sufficient to afford a decent place to live. In fact, having a low-paid job is increasingly unlikely to lift a family out of poverty or resolve worst case housing needs. Between 1991 and 1995, worst case needs rose by 265,000, or 24 percent, for renters with annual earnings of at least one full-time worker at the minimum wage. By 1995, there were 1.4 million such households with worst case needs. Finding 4: One of every three households with worst case needs now lives in the suburbs. * While the greatest numbers of worst case needs are in central cities, a large and fast-growing number live in the suburbs. The first half of the 1990s saw a suburbanization of worst case needs as more than 1.8 million of the 5.3 million households with worst case needs -- or one of every three -- lived in the suburbs in 1995. Suburban worst case needs grew by 9 percent from 1991 to 1995. Supplementary Findings Finding 5: The most serious housing needs are concentrated among households with the lowest incomes. * Almost 4 million of the 5.3 million households with worst case needs have extremely low incomes -- below 30 percent of median. Almost 7 of every 10 such households pay more than one-half their income for rent or live in severely inadequate housing when they are not assisted. * The frequency of worst case needs declines sharply as income rises. Only 26 percent of unassisted renters with incomes between 31 and 50 percent of median have worst case needs and fewer than 5 percent of renters with incomes between 51 and 80 percent of median experience such problems. Finding 6: Of the 12.5 million persons in households with worst case needs, almost 1.5 million are elderly and 4.5 million are children. The number of adults with disabilities in households with worst case needs is estimated between 1.1 and 1.4 million. * From 1993 to 1995, there was a surge in the number of very-low-income single individuals who are not elderly, even as the overall number of very-low-income households declined slightly. Importantly, the likelihood of having worst case needs also grew among non-elderly singles. Although the American Housing Survey does not measure disabilities directly, many of these individuals with worst case needs have disabilities. * It is clear that both single individuals with disabilities and other households with a disabled member face substantial and growing housing problems. The number of adults with disabilities living in households with worst case needs is estimated at between 1.1 and 1.4 million in 1995. * More than 2.1 million families with children had worst case problems in 1995. Among the 2.1 million families with children who had worst case problems, 930,000 had income from either Aid to Families with Dependent Children (AFDC) or Supplemental Security Income (SSI). * More than 1 million elderly households had worst case problems in 1995. Finding 7: Worst case needs continue to shift to the West. * The number of very-low-income renters in the West continued to increase between 1993 and 1995, while dropping in other regions. The West had the highest percentage of very-low-income renters with acute housing needs, 42 percent, compared with 32 percent in the South, 33 percent in the Midwest, and 39 percent in the Northeast. The result is that the number of Western households with worst case needs reached a record 1.56 million in 1995. * The mismatch between available extremely-low-rent units and extremely-low-income renters is large and getting larger in all four census regions. Shortages are worst in the West, however, where in 1995 for every 100 extremely-low-income renters there were only 31 units that were either already occupied by extremely-low-income renters or vacant and for rent, compared with the nationwide figure of 44 units per 100 renters. Policy Implications The findings of this report suggest that economic growth alone will not ameliorate the record-level housing needs among families with limited incomes. Not even families working full-time at the minimum wage can afford decent quality housing in the private rental market. The report also makes it clear that housing needs are not just found in big cities but increasingly in the suburbs as well. The report suggests a clear and compelling need for the Congress to provide greater support for Federal housing assistance -- by expanding both tenant-based rental assistance and programs that create and rehabilitate more affordable housing units. And Congress should act carefully in reforming the income targeting rules for public and assisted housing programs to balance the goals of achieving a greater income mix in public and assisted housing developments and providing assistance to families with the most severe housing needs. New Housing Assistance * 103,000 New Vouchers: The Administration has asked Congress to fund 103,000 new housing assistance vouchers, including 50,000 welfare-to-work vouchers to help welfare recipients get and keep jobs. This report documents the need for those vouchers to reduce the overall number of families and individuals with worst case needs and to provide the portable housing assistance critical for a successful transition to work. * Ending the Delay on Reissuing Vouchers: Congress should end immediately its cost-saving requirement placed on local housing authorities to hold for three months rental subsidies returned by families leaving the program. This practice reduces by 40,000 the number of subsidies in circulation and thus the number of families receiving housing assistance. * Expanding Production of Affordable Housing Through HOME and the Low Income Housing Tax Credit: The Administration is also seeking to expand tools to build and rehabilitate affordable housing. HUD's FY 1999 budget includes increased funding for the HOME program, along with a new HOME Bank, a loan guarantee feature that would allow communities to leverage up to five times their Federal grants for larger scale housing investments. In addition, the Administration is proposing a substantial expansion of the Low Income Housing Tax Credit that would create 180,000 new affordable rental units over the next five years. Careful Income Targeting of Federal Housing Assistance Congress is considering legislation that will determine the income levels of households who will be admitted to public housing and to receive Section 8 rental assistance. The Clinton administration and U.S. Department of Housing and Urban Development (HUD) strongly support the transformation of public and assisted housing developments into healthier, mixed-income communities. But policymakers must be careful not to exclude poor families altogether from these housing developments, nor to reduce unnecessarily the numbers of families with worst case needs who can be served by Federal housing programs. This report shows that this goal can be achieved while still continuing to serve families who are working but who have low incomes and serious housing needs. The body of the report is presented in three parts. The introduction (chapter 1) explains the background and approach of this report. Chapter 2 summarizes statistical data from HUD analyses of worst case needs, documenting major findings. Chapter 3 explores the implications of the findings for current policy decisions. Appendices provide detailed definitions and statistical tables. ---------- Chapter 1 Introduction Since the early 1990s, the U.S. Department of Housing and Urban Development (HUD) has submitted regular formal reports to Congress on worst case needs for rental housing assistance. These reports have drawn on data collected by the Bureau of the Census, U.S. Department of Commerce, in the American Housing Survey (AHS) and in the decennial censuses. The 1996 report, Rental Housing Assistance at a Crossroads, for the first time also included information from HUD administrative data about how well current housing assistance programs are serving families and individuals that otherwise would have worst case needs.1 Terms and Sources HUD has used a consistent definition of worst case needs, making it possible to track changes over time. This definition is based on (1) income limits that determine eligibility for Federal housing assistance and (2) "priority housing problems" (see appendix B, Glossary). This report uses the terms "acute housing needs" and "worst case needs" interchangeably to refer to households that: * Are renters. * Do not receive Federal housing assistance. * Have incomes below 50 percent of median family income in their area, as established by HUD. * Pay more than one-half of their income for rent and utilities or live in severely substandard housing -- until recently, households on waiting lists with either of these characteristics received preference for rental assistance programs.2 The basic source of information for analyzing the U.S. housing stock and the housing needs of U.S. households is the American Housing Survey. AHS is conducted for HUD by the Bureau of the Census, which completes about 45,000 interviews of occupied households in a biennial national sample of housing units. Smaller samples of units in 47 large metropolitan areas are surveyed on 4- to 6-year cycles. HUD's first formal report to Congress on worst case needs in 1991 was based on 1989's AHS. The second report in 1992 continued to use 1989 AHS data for the Nation, augmented with information on worst case needs from the metropolitan surveys. In 1994, HUD based its report on data from the 1991 AHS and the 1990 Decennial Census. The 1996 report was based on data from the 1993 AHS and, for the first time, included administrative data on the characteristics of households participating in the public housing and Section 8 programs. The 1996 report also included analyses of data from the Social Security Administration (SSA) to better understand the housing needs of persons with disabilities and their participation in HUD programs. Finally, that report reanalyzed and refined earlier AHS data to more reliably track growth in the number of households with worst case housing needs between 1978 and 1993.3 * In 1990, the Senate Appropriations Committee directed HUD to "resume the annual compilation of a worst case housing needs survey of the United States." HUD had reported worst case housing needs to Congress during the 1980s on an informal basis, following a request from the Chair of the HUD Subcommittee of the Senate Appropriations Committee. * Although "substandard" housing should include homelessness, the homeless are omitted from this report's counts of worst case needs because the AHS surveys and counts only persons in housing units. For a history of the Federal preferences, see exhibit 19, "History of the Federal Preference System." * HUD's previous reports to Congress are Priority Problems and "Worst Case" Needs in 1989 (June 1991, HUD-1314-PDR), The Location of Worst Case Needs in the Late 1980s (December 1992, HUD-1387-PDR), Worst Case Needs for Housing Assistance in the United States in 1990 and 1991 (June 1994, HUD-1481-PDR), and Rental Housing Assistance at a Crossroads: A Report to Congress on Worst Case Housing Needs (March 1996). Both the June 1994 and March 1996 reports are available online at http://www.huduser.org under the Publications heading. ---------- Building the Report This year's report uses data from the 1995 AHS and HUD administrative data as of January 1997. It builds on the 1996 report in the following ways: * Detailed income categories are used to examine the housing needs of households. Like the 1996 report, this report distinguishes among extremely-low-income households (those with incomes below 30 percent of median income), other very-low-income households (those with incomes between 31 and 50 percent of median income), and other low-income households (those with incomes between 51 and 80 percent of median income). The extremely-low-income category has become particularly important. Legislation under consideration since 1995 would replace the waiting list preference for worst case needs households by targeting assistance to those below 30 percent of area median income. As shown by this and earlier worst case needs reports, households with incomes below 30 percent of median are those most likely to have worst case needs. Exhibit 1 Income Categories Used in Housing Programs For many HUD programs and housing programs administered by other Federal agencies, eligibility is restricted to households whose incomes do not exceed a specific percentage of the median family income for the area in which the household lives. HUD defines median income for each metropolitan area and non-metropolitan county, and the HUD-adjusted area median family income (HAMFI) varies by location and household size.* In contrast, poverty status is determined by comparing income with national poverty thresholds that vary only by household size but not location. Because HUD's income limits vary with location and use smaller adjustments for household size, they cannot be compared directly with the Federal poverty line. Averaged across the United States, however, poverty thresholds correspond approximately with 30 percent of area median income. The number of households below a specified percentage of HUD's area median income is not related to any break on the total income distribution, such as quintiles or deciles. For example, almost one-half (45 percent) of all U.S. households and 64 percent of all renters have incomes below 80 percent of their area median income. More than 26 percent of all U.S. households have incomes less than 50 percent of area median income. The upper limits of income categories used in housing programs and in this report are as follows: 80 percent of area median income. Defined as lower income by the U.S. Housing Act and used for many rental and homeownership programs. 60 percent of area median income. Used in Low Income Housing Tax Credit and HOME programs. 50 percent of area median income. Defined as very low income by the U.S. Housing Act and used for many rental programs. 30 percent of area median income. Defined as extremely low income in pending housing authorization bills. Used as a proxy for households that, until 1995, would have received a Federal preference for rental housing assistance because they have worst case housing needs. The table below shows how many U.S. renter households fell into the different income groups relevant for housing programs in 1995. To suggest the overlap between the HUD income groups and poverty, it also shows the share of each income group whose cash income fell below the poverty line or below 150 percent of the poverty line, which is the approximate eligibility cutoff for the U.S. Department of Agriculture Food Stamp program. As in this exhibit, this report frequently refers to specific income groups as ranges of percentages of area median income because official terms are so complex. For example, incomes 51-80 percent of area median are officially "low but not very low" incomes. Exhibit 2 gives examples of HUD income cutoffs for nine large metropolitan areas. Percent Share of Households in Group With Income Income as % of Share of Below the Below HUD-Adjusted Area U.S. Poverty Level 150% of Median Family Renters the Income (HAMFI) 1995 (%) Poverty Level 0-30 25 86 99 31-50 17 15 64 51-60 8 0 19 61-80 13 0 4 *Appendix B discusses other adjustments. Source: HUD-PD&R tabulations of the 1995 American Housing Survey Exhibit 2 Extremely Low, Very Low, and Low Income: Examples of 1997 Income Limits for Four-Person Households In 1997, the average poverty threshold for a family of four was $16,400. * The most recent information from HUD administrative records is used to show how rental assistance programs serve different income and demographic groups. The 1996 report used data from the Multifamily Tenant Characteristics System (MTCS) and the Tenant Rental Assistance Certification System (TRACS) as of December 1995. This report uses MTCS and TRACS data from February 1997. Exhibit 3 Housing Assistance and Affordable Housing Programs Federal rental assistance programs operate in three basic ways: * Public housing. These units are owned by local public agencies. From 1937 to the mid-1980s, public housing was used extensively to produce additional assisted housing units. Today, there are 1.2 million occupied units of public housing. * Project-based assisted housing. These programs supported the construction and rehabilitation of 1.4 million rental units for low-income households. Deep rent subsidies are attached to projects owned by for-profit and nonprofit sponsors that must rent units to eligible households. These programs added large numbers of assisted units from 1974 to the early 1980s. * Tenant-based assisted housing. These programs provide direct rental assistance to 1.4 million renter households to enable them to find their own housing on the open market. The maximum subsidy is the difference between the tenant contribution and the local fair market rent (FMR), an average rent for standard quality housing in the area. Begun in 1974, this type of assistance has accounted for virtually all the incremental units, or additions to assisted housing, since the mid-1980s. In all three types of programs, assisted households pay rents that are a percentage of their adjusted income -- usually 30 percent. This formula allows even the poorest households to live in assisted housing. Other Federal programs produce affordable housing. There are a number of other Federal housing programs in which renters are charged fixed or flat rents, with the maximum determined by program rules. Households pay the established rent rather than a percentage of their income. Without an additional subsidy, the poorest households often cannot afford this housing. These programs include: * The Low Income Housing Tax Credit program. This tax credit program subsidizes the capital costs of units that must bear rents affordable to households with incomes at or below 60 percent of area median income. HUD estimates that this program has produced more than 600,000 units since its enactment in 1986.4 * The HOME Investment Partnership (HOME) program. This is a formula grant to States and local governments that can be used to assist existing homeowners, first-time homebuyers, or renters. Between 1992 and September 1997, HOME produced 126,000 affordable rental units. Qualifying rents must be affordable to households with incomes at or below 65 percent of area median income, or below local FMRs. * Older rental subsidy programs. The Section 221(d)(3) below market interest rate (BMIR) program and the Section 236 program were active from the early 1960s through the early 1970s. They were designed to produce housing affordable by families with incomes above the public housing income limits. Over time many projects or portions of projects in these programs became "project-based assisted housing" rather than "rental subsidy" as deep rental subsidies were attached to the units. There remain 300,000 units subsidized by these older programs that do not have deep rental subsidies. Exhibit 4 Household-Level Data From HUD Programs The Multifamily Tenant Characteristics System (MTCS) is an automated data base of households assisted by public housing and the tenant-based Section 8 certificate and voucher programs and other programs administered by HUD's Office of Public and Indian Housing. The system contains information about the demographic characteristics of each household, the level and sources of the household's income, and the address of the housing unit. The information is based on the form used by public housing authorities (PHAs) to calculate each household's rent and subsidy levels. As of February 1997, the system contained nearly 2.4 million household records, or about 85 percent of the possible total. The Tenant Rental Assistance Certification System (TRACS) is a similar system for households assisted in project-based Section 8 programs and other assisted projects administered by HUD's Office of Housing. Information in TRACS is based on forms completed by the private owner or manager of the project and submitted to HUD. As of June 1997, TRACS contained 1.6 million household records. Tables providing additional information from MTCS and TRACS data on the income levels of assisted households can be found in Recent Research Results, published by HUD's Office of Policy Development and Research (PD&R) in March 1998 or at under "publications." For data from MTCS and TRACS summarized for each housing project and each census tract, see A Picture of Subsidized Households, published by PD&R in 1996 and 1997. These reports provide income and demographic information at the census tract and project level, along with national and regional reports, based on HUD's administrative records. Another source of information on households receiving Federal housing assistance is: Characteristics of HUD-Assisted Renters and Their Units in 1993, published by PD&R in May 1997. This report is based on case-by-case matching of administrative data on the addresses of assisted housing units to AHS data. * Data from SSA are used to better describe the housing needs of persons with disabilities and their participation in HUD programs. Because AHS does not specifically ask about disabilities, worst case needs reports published before 1996 identified people with disabilities by assuming that anyone under the age of 62 who reported receiving Supplemental Security Income (SSI) had a disability.5 The 1996 report used data from an audit of the SSI program to develop much more complete estimates of the number of worst case needs households with members who have disabilities. This report does not update the analysis of SSA data, but uses proportional relationships from the 1996 report to estimate the number of households with disabilities and worst case needs in 1995. * 1995 AHS national data are used to update the evidence of mismatches between very-low-income renters and the availability of units affordable to them. In addition, a separate report, to be published shortly by PD&R, provides the detailed results and policy implications of a major study of gains and losses in affordable housing units in 41 metropolitan areas. * The policy outlook is still uncertain. The 1996 report was titled Rental Housing Assistance at a Crossroads because Congress had begun to make major changes in housing assistance policy that could result in public and assisted housing serving many fewer families and individuals who otherwise would be categorized with worst case needs. This report is titled Rental Housing Assistance -- The Crisis Continues, because it supplies evidence that worst case needs do not decline in an economic expansion, but also because the turmoil in housing assistance policy is still under way (see exhibit 5). A new dimension has been added to the uncertainty in housing assistance policy by the major overhaul of the Nation's welfare system. A year and a half after the enactment of welfare reform, uncertainty exists about its effects on the incomes of poor families with children and other categories of households affected. Exhibit 5 Changes to Assisted Housing Policy: Recent and Proposed Appropriations Actions HUD's Appropriations Acts since 1996 have included the following measures: * No funds for incremental rental housing assistance. Funds are provided only for the replacement of public and assisted housing units that leave the assisted housing stock, for other circumstances in which families and individuals who already have housing assistance need to be protected, or for very limited special purposes such as litigation settlements. * Year-to-year suspension of the Federal preferences for public and assisted housing that target assistance to households with worst case housing needs, without changes substituting income targeting requirements. * A required delay in the reissuance of tenant-based assistance by housing authorities when house-holds leave the program that, in effect, shrinks the number of households assisted at any one time. Authorization Actions Congress is considering new enabling legislation for public housing and tenant-based housing assistance. Provisions of this legislation would permanently end Federal preferences for households with worst case housing needs. * Senate Bill S. 462, passed by the Senate in September 1997, would eliminate Federal preferences permanently and require a minimum of 40 percent of households newly admitted to public housing and 65 percent of households admitted to tenant-based rental assistance programs to have incomes below 30 percent of the area median. Seventy percent of households newly admitted to public housing and 90 percent of households admitted to tenant-based rental assistance would be required to have incomes below 60 percent of median. This would mean that 30 percent of households newly admitted to public housing and 10 percent of households newly admitted to tenant-based rental assistance programs could have incomes as high as 80 percent of area median income. * House Bill H.R. 2, passed by the House of Representatives on May 14, 1997, would eliminate Federal preferences and require a minimum of 35 percent of households newly admitted to public housing to have incomes below 30 percent of area median income. Forty percent of households newly admitted to the tenant-based Section 8 program would have to have incomes below 30 percent of median. This would mean that 65 percent of households admitted to public housing and 60 percent of households newly receiving tenant-based Section 8 assistance could have incomes of up to 80 percent of the area median. If a public housing authority (PHA) exceeded the extremely-low-income targeting requirement for Section 8 tenant-based assistance, the public housing targeting requirements could be reduced by the amount of excess in targeting Section 8 tenant-based assistance. Thus, it would be possible for a PHA not to admit any extremely-low-income households to its public housing. * The Clinton administration's position, reflected in H.R. 1447 and S. 784, would eliminate Federal preferences and require a minimum of 40 percent of families newly admitted to public housing and 75 percent of families newly admitted to tenant-based assistance programs to have incomes below 30 percent of median. For public housing, at least 90 percent of newly admitted households would be required to have incomes below 60 percent of median, while the remaining 10 percent could have incomes up to 80 percent of median. For Section 8 tenant-based assistance, the remaining 25 percent of newly admitted families would have to have incomes below 50 percent of the area median. The Administration-supported bills also would require that at least 40 percent of the units in each public housing project be occupied by families with incomes not exceeding 30 percent of median, ensuring that some extremely-low-income individuals and households would be served in all projects, including the most desirable projects. * Estimate assumes 100,000 units placed into service in 1996 and 1997. * AHS supplemental questions on disabilities were included in the 1978 and 1995 national surveys, but these questions cover only physical disabilities. ---------- Chapter 2 Major Findings Acute needs for housing assistance, at an all-time high in 1993, did not drop between 1993 and 1995 despite the strong, sustained economic growth experienced during that period. In 1995, as in 1993, more than 5.3 million very-low-income renter households -- almost 12.5 million individuals -- paid more than one-half of their income for housing or lived in poor-quality housing. Among very-low-income renters, the share with worst case needs rose very slightly, to 37 percent.1 These 5.3 million households represent 4.9 percent of the Nation's population and over one-seventh of all U.S. renters. Without Federal housing assistance, they lack the income to afford adequate, market-rate housing. One missed paycheck, an unexpected medical bill, or some other emergency is all that separates many of these families from homelessness. Finding 1: Despite robust economic growth between 1993 and 1995, the number of very-low-income renters with worst case housing needs remained at an all-time high -- 5.3 million. * In the economic expansion of the mid-1980s, the number of households with worst case needs declined. No comparable drop occurred between 1993 and 1995, when housing assistance was no longer increasing each year. During the economic expansion of the mid-1980s, worst case needs dropped significantly between 1985 and 1987 (see exhibit 6). Between 1993 and 1995, economic growth brought greater declines in both poverty and unemployment rates than had occurred between 1985 and 1987. Unemployment dropped from 6.8 percent in 1993 to 5.6 percent in 1995, whereas the 1985-1987 decline in unemployment was from 7.2 to 6.2 percent. The poverty rate dropped from 15.1 percent in 1993 to 13.8 percent in 1995, whereas the 1985-1987 decline in poverty was from 14.0 percent to 13.4 percent.2 In light of the strong 1993-1995 economic growth, the absence of a decline in needs between 1993 and 1995 is both surprising and discouraging. Part of the explanation for the different pattern of worst case needs between 1985 and 1987 and 1993 and 1995 may be that the mid-1980s was a period of continued rapid growth in housing assistance (see exhibit 7), whereas the mid-1990s was not. Housing developments funded by production programs in the late 1970s and affordable by the poorest households were completed and occupied in the mid-1980s.3 In addition, Congress then was adding about 100,000 incremental units of tenant-based rental assistance each year. Exhibit 6 Worst Case Needs Dropped in the Mid-1980s, but Not in the 1993-1995 Recovery Exhibit 7 Annual increase in Number of Households With Public and Assisted Housing, 1978-1996 Source: Congressional Budget Office; based on data provided by the U.S. Department of Housing and Urban Development and the Farmers Home Association The failure of worst case needs to ameliorate during the economic expansion of the 1990s demonstrates that acute housing needs are a persistent problem in American society that cannot be solved through overall management of the economy. An explicit Federal housing policy response is needed. For as long as worst case needs have been reported, affordability rather than housing quality has been the predominant problem facing unassisted renters. In 1995, more than 95 percent of those households with worst case problems paid more than one-half of their reported income for housing. The proportion of worst case households living in housing with severe physical problems continued to drop, falling below 8 percent in 1995. Reversing earlier declines, the incidence of overcrowding rose slightly between 1993 and 1995, both among all renters (from 4.4 to 4.9 percent) and among very-low-income renters (from 6.8 to 7.9 percent). However, overcrowding is not included in estimates of worst case needs in this or previous reports to Congress because it did not automatically give a household a Federal preference for admission to assisted housing programs. Despite the slight increase in crowding, severe rent burdens were the only housing problem for almost four of five worst case households (79 percent). * A slight decline from 5.34 million worst case households in 1993 to 5.32 million in 1995 is statistically insignificant. Despite a modest decline of about 200,000 in the number of households whose incomes fell below 50 percent of the area median, among very-low-income renters, the share with worst case needs increased slightly and insignificantly, from 36.3 to 36.53 percent. * Bureau of the Census. Current Population Reports, Series P-60, Nos. 184 and 188. * Public housing and Section 8 new construction units are affordable by the poorest households because rent is set at 30 percent of the actual income of the households. ---------- Finding 2: The stock of rental housing affordable to the lowest income families is shrinking and Congress has eliminated funding for new rental assistance since 1995. HUD's 1996 report to Congress on worst case housing needs demonstrated a serious mismatch between numbers of extremely-low-income renters and numbers of units affordable to them as of 1993. This mismatch worsened markedly in just 2 years. * There were sharp losses in rental units for very-low-income renters. Between 1993 and 1995, losses of private market rental units affordable to very-low-income renters -- that is, units on which a family with income at 50 percent of median would spend 30 percent of income or less -- accelerated greatly. Between 1993 and 1995, the number of units affordable to very-low-income renters dropped by almost 900,000, from just under 10.4 million to 9.5 million units (see exhibit 8). This drop in very-low-rent units was almost four times greater than the decline in very-low-income renters. The number of very-low-income renters dropped slightly, by 229,000, between 1993 and 1995. * The gap between the lowest income renters and units affordable to them is large and growing. The largest losses of private rental stock, in percentage terms, were for units with rents affordable to extremely-low-income renters, those on which a family with income at 30 percent of median would spend 30 percent of income or less. The number of extremely-low-rent private market units dropped from 1.8 million in 1993 to 1.5 million in 1995, a 16-percent loss in just 2 years. Exhibit 8 Loss of Private Market Units Affordable for Very-Low-Income Renters Accelerated Between 1993 and 1995 The gap between the number of extremely-low-income renters and the number of extremely-low-rent housing units continued to get worse. By 1995, for every 100 extremely-low-income renters there were only 77 units with rents that would have been affordable had those units been available to them. By comparison, there were 111 affordable units for every 100 renters with incomes below 50 percent of area median income. Not all these extremely-low-rent units were rented by the lowest income renters. Many of these units were occupied by households with incomes greater than 30 percent of median. When only those units either vacant and for rent or already occupied by extremely-low-income renters are considered available, there were only 44 units per 100 households in 1995. This ratio has dropped steadily during the 1990s (see exhibit 9). Another indication that the private housing market does not supply rental units affordable to the lowest income families is that more than three quarters of the extremely-low-rent units occupied by households at any income level were "non-market" units. These include public and assisted housing, other housing with publicly subsidized rents, and units for which the occupant did not pay a cash rent. Exhibit 9 Mismatches Between Extremely-Low-Income Renters and Rental Units They Can Afford Continue to Worsen * In a historic reversal of Federal housing policy, Congress has stopped providing funding to expand rental assistance for households with worst case needs. The Federal Government began its commitment to providing low-income renters with decent quality, affordable housing in response to the Great Depression in the 1930s. Since World War II, there has been an uninterrupted growth in the number of households who have been provided with Federal rental assistance through a variety of Federal construction, rehabilitation, and rental assistance programs. The growth in the numbers of households receiving rental assistance peaked in the late 1970s, when more than 200,000 additional households were being added to the housing assistance rolls each year. From 1978 through 1982, an average of 224,000 additional households were provided Federal rental assistance each year. The average number of new households getting assistance dropped to approximately 146,000 during the 1980s and early 1990s. This trend ended abruptly in 1996, when for the first time, the number of renters receiving Federal housing assistance actually dropped (see exhibit 7). Since 1995, Congress has denied Administration requests and provided no funding for new rental assistance to serve families with worst case needs. ---------- Finding 3: The fastest growth in worst case needs in the 1990s was among working families. The failure of worst case needs to drop despite a robust economy becomes more understandable if one looks at the nature of the growth in worst case needs during a somewhat longer period of time: 1991-1995. Underlying the growth in worst case needs in the early 1990s were structural problems not alleviated -- and even exacerbated -- by the mid-1990s economic boom. These problems included lags in wages for the lowest paid workers. Working poor families and individuals faced a diminishing supply of housing at rents they could afford, as people with stronger income growth put upward pressures on rents. * Between 1991 and 1995, worst case needs among working families grew by 24 percent. The share of all households experiencing worst case needs that had earnings at least the equivalent of one full-time worker at minimum wage rose from 22 to 26 percent. As very-low-income families enter the workforce, increases in income often are not enough to make rents affordable and alleviate worst case housing needs. In fact, housing problems can be even more vexing for very-low-income workers than for those who receive their income from public assistance. Housing affordable to a family with low wages may be far from work. Finding and renting in the right location to get and keep a job may mean spending so much for housing that other needs go unmet. It may be impossible to keep up with the rent because of unexpected expenses such as car repairs or specialized clothing for work. Missing rent payments and possible eviction is a serious concern just at the time when the family is coping with child care and other new pressures associated with keeping a job. Between 1991 and 1995, worst case needs overall rose from 4.9 million renter households to 5.3 million, a growth rate of 7 percent. For unassisted very-low-income renters with children in the family and someone working the equivalent of at least full-time at minimum wage, worst case needs rose at double that rate, or 13 percent (see exhibit 10). Renters without children experienced the same pressures: Among those working at least full-time, worst case needs rose by 37 percent. Together, these two groups of working households were 1.1 million, or 22 percent, of renters with worst case needs in 1991 and grew to 1.4 million, or 26 percent, by 1995. * A family that moves to work as a result of welfare reform will likely still have worst case needs for housing assistance. The very-low-income threshold of 50 percent of the area median is at a level well above both the poverty threshold and welfare benefit levels.4 Even extremely-low-income families -- those with incomes below 30 percent of median or roughly the poverty level -- are as likely to be working as they are to be receiving income from public assistance (see exhibit 11). In the income group between 21 and 30 percent of median, almost two-thirds of families with children have earnings as their main income source and half are working the equivalent of full-time at minimum wage. It is not surprising, then, that more than one-half of families with children and worst case housing needs rely primarily on work to support themselves. Exhibit 10 Between 1991 and 1995, Growth in Worst Case Needs Was Highest Among Working Families and Individuals *Earnings exceed full-time work at minimum wage ** Non-elderly and no disability (not reporting SSI income). Source: HUD-PD&R tabulations from the 1995 American Housing Survey Exhibit 11 Substantial Numbers of Very-Low-Income Families With Children Are Working The Clinton administration strongly believes that those capable of working should do so and that housing assistance should play a key role in rewarding work and family responsibility. Helping families with worst case needs and rewarding work are not incompatible priorities. In 1995 more than 1.6 million renters with worst case needs had incomes below 30 percent of the area median and earnings as their primary income source (see exhibit 12). Proposed changes that would raise income eligibility for Federal housing assistance could -- in the name of helping low-income workers -- divert resources away from the working poor who most need assistance. A very large number of the families that will go to work as a result of welfare reform will continue to have incomes below 30 percent of the area median. A recent study found that the median annual income of families leaving welfare for work was just $5,000 in the first year and $9,000 after 5 years.5 These incomes are well below the national extremely-low-income cutoff of $12,900. If a family is working yet has an extremely low income, that family will probably have worst case housing needs. Two of every three unassisted extremely-low-income working families with children have worst case needs. Exhibit 12 Over 1.6 Million Unassisted Working Renters With Worst Case Needs Have Extremely Low Incomes * Current housing assistance programs effectively reach the working poor. Housing assistance can be crucial to working poor families, helping them to stabilize their lives and continue their climb out of poverty. Of the 4.3 million households living in public housing or receiving assistance from one of the Section 8 programs, more than 1.1 million report earnings as their primary source of income.6 Current housing assistance programs are well targeted, not just to workers but to the working poor. Nearly one-half (48 percent) of assisted families with children who report incomes between 21 and 30 percent of median have earnings as their primary source of income. Even at the very lowest income range (below 20 percent of median), almost one-fifth (18 percent) of all families with children living in assisted housing have earnings as their primary income source. Work is even more common among extremely-low-income households without children when that household is headed by an individual who is not elderly and does not have disabilities (see exhibit 13). Exhibit 13 Many Assisted Families Have Earnings as Their Primary Source of Income Housing assistance programs could be even more explicitly targeted to reward work. The Clinton administration supports providing housing assistance to working families and to families preparing themselves for work through job training and education. This change is most critical in public and assisted housing projects, where the growing children urgently need adult role models who work and adhere to mainstream values of the society. To achieve this objective, however, it is not necessary to serve families with incomes above 50 percent of median income. These families are unlikely to have severe needs for housing assistance. Instead, housing authorities and managers of privately owned assisted housing can design admission priorities that direct assistance to those who work and have incomes within the very-low- and extremely-low-income ranges. * In 1995, a family receiving the maximum Aid to Families with Dependant Children (AFDC) grant for a family of three would have been at 16 percent of the area median income in New York City, 10 percent in Chicago, 5.5 percent in Houston, and 16 percent in Los Angeles. It is not surprising that many families with incomes below 30 percent of the area median receive no welfare. * Daniel R. Meyer and Maria Cancia, Life After Welfare: The Economic Well-Being of Women and Children Following Exit from AFDC, Institute for Research on Poverty, University of Wisconsin, Discussion Paper No. 1101-96. August 1996. * The difference between these 4.3 million units and the 4.5 million total HUD-assisted rental units occurs because Indian housing and units in the Section 8 moderate rehabilitation program have not been included in the analysis for this report. ---------- Finding 4: One of every three households with worst case needs now lives in the suburbs. Because of higher poverty concentrations and lower homeownership rates, central cities have the most renters with worst case needs. But surprising numbers of renters with worst case needs live in the suburban portions of metropolitan America, and, except in the Northeast, that is where worst case needs are growing the fastest. Moving from welfare to work may require access to the suburbs -- where approximately two-thirds of new jobs are being created -- but for very-low-income renters, paying suburban rents often means paying more than half of the family's income for housing. Nationwide, 1.8 million suburban renters had worst case needs in 1995, an increase of 146,000 since 1991. In the Western part of the United States, the change was more pronounced, with worst case needs of renters in the suburbs increasing from 570,000 to 678,000, or almost one-fifth (see exhibit 14). Even in the South, where overall there was a modest decline in worst case housing needs between 1991 and 1995, the number of suburban households with worst case needs grew. In the Midwest, worst case needs declined slightly in central cities but increased by over 8 percent in the suburbs. Exhibit 14 In the Early 1990s, Worst Case Needs Grew Quickly in the Suburbs Source: HUD-PD&R tabulations from the American Housing Survey Shortages of units that are both affordable and available to extremely-low-income renters are greatest in the suburbs, the same areas in which worst case needs are growing most rapidly. Across the Nation in 1995, there were only 39 extremely-low-rent units available for every 100 extremely-low-income renters in the suburbs, compared with 43 units in central cities and 56 units in non-metro areas. (see exhibit 15). Exhibit 15 Mismatches Between Extremely-Low-Income Renters and Available Rental Units They Can Afford are Worst in the Suburbs * Even in relatively low-rent rural areas, worst case needs failed to drop. Outside of metropolitan areas, renters are less likely to have worst case housing needs, and those renters with worst case housing problems are somewhat more likely to live in housing units with severe physical problems. Nevertheless, affordability is the overwhelming housing problem in non-metropolitan America just as it is within metropolitan areas. Over three-fourths (76 percent) of non-metropolitan households with worst case needs have severe rent burdens as their only housing problem. Although housing costs are lower in non-metro areas, here as elsewhere in America the economic boom of the 1990s failed to reduce the number of households with worst case housing needs. The level of worst case needs in non-metro areas rose slightly between 1991 and 1993 and then did not drop at all between 1993 and 1995, remaining at 727,000 families and individuals (see exhibit 14). ---------- Supplementary Findings Finding 5: The most serious housing needs are concentrated among households with the lowest incomes. * The vast majority of renter households with worst case needs have extremely low incomes. The incidence of worst case needs falls sharply as income increases. It is the poor who cannot cope with America's robust, market-driven system of providing housing -- unless they receive help from publicly funded housing assistance programs. Worst case housing needs are defined as severe housing problems among unassisted very-low-income renters, renters with incomes below 50 percent of area median. The Federal preferences for housing assistance that have been suspended in recent years -- and will almost certainly be repealed by pending housing legislation -- have the effect of targeting assistance to a yet lower income group, those with incomes below 30 percent of area median. In place of the Federal preferences, new housing legislation will reserve some units for this extremely-low-income group. It will also allocate assistance to other households with incomes of up to 80 percent of the area median. The total number of renter households with severe or "priority housing problems" in 1995 includes the 5.3 million very-low-income renters with worst case needs. Another 565,000 renters also had priority housing problems, but they had incomes above 50 percent of the area median. As exhibit 16 shows, over two-thirds of renter households with severe housing problems have incomes below 30 percent of area median, which is approximately the official poverty threshold. A startling 42 percent (2.4 million) of renters with severe housing problems have incomes below 20 percent of the area median. Almost 63 percent have household incomes that fall below Federal poverty cutoffs that, it must be remembered, are not adjusted for geographical differences in incomes.7 Exhibit 16 Over Two-Thirds of Renters With Priority Problems Have Income Below 30 Percent of Median Almost 70 percent of unassisted, extremely-low-income renters have worst case housing needs. As soon as a household's income rises above 30 percent of area median income -- to between 31 and 40 percent of median -- the likelihood that the household has severe housing problems is only one in three. In the group just above the very-low-income cutoff but eligible for units produced by HOME and the Low Income Housing Tax Credit (51 to 60 percent of area median income), less than 7 percent have priority problems. Less than 5 percent of those in the group between 61 and 80 percent of median have severe housing needs (see exhibit 17). Yet legislation under consideration by Congress would take from this group up to 65 percent of households newly admitted to public housing and 60 percent of those newly receiving tenant-based Section 8 assistance (see exhibit 5). Not only are priority needs concentrated at the bottom of the income ladder, but most very-low-income renters have some housing problem, even if it is not a priority problem. In marked contrast, as income increases above the very-low-income cutoff, the probability of encountering other housing problems, which most often include paying between 31 and 50 percent of income for housing, drops rapidly. * The public housing and Section 8 programs have historically been well targeted to the income groups most likely otherwise to have acute housing needs. Public housing and the tenant-based Section 8 program currently direct about three-fourths of assistance to extremely-low-income renters, those with incomes below 30 percent of the area median (see exhibit 18). The project-based Section 8 programs also are remarkably well targeted to those who otherwise would be most likely to have acute housing needs, with 70 percent of units occupied by extremely-low-income renters. Since 1974, the basic income ceiling for Federal housing assistance has been 80 percent of the area median. Since 1981, however, mandatory quotas have directed assistance in each program to households with incomes below 50 percent of median: 75-85 percent of the assistance for public housing and project-based Section 8 (depending on the age of the project) and essentially 100 percent of the assistance for tenant-based Section 8. The deeper actual targeting of these programs, in which 70-75 percent of residents are extremely low income, is a result of several factors. One is the practice of "Federal preferences" enacted by Congress in 1978 and 1981 but not implemented until 1988. Federal preferences put households with worst case needs ahead of others on waiting lists for assisted housing. Exhibit 17 Renters With Income Below 30 Percent of Median Are the Only Groups Likely to Have Severe Housing Problems Exhibit 18 Housing Assistance is Well Targeted to the Income Groups With Priority Problems Exhibit 19 History of the Federal Preference System Before 1979 PHAs and owners used local preferences -- consistent with statutory income targeting requirements -- to determine admission to public and assisted housing. 1979 The Housing and Community Development Amendments of 1979 created the first Federal preferences for selection of occupants of public and assisted housing. Preference was to be given to those who were either involuntarily displaced from their homes or living in substandard housing. 1983 The Housing and Urban-Rural Recovery Act of 1983 added a preference for families paying more than one-half of their income for rent. 1988 Regulations implementing the three Federal preferences (see 1979 and 1983 above) were published in January 1988 and became effective in July 1988. At that time, 90 percent of households newly admitted to the public housing and project-based Section 8 programs and 100 percent of households newly admitted to the tenant-based Section 8 programs had to be Federal preference holders. 1994 Regulations were published implementing legislative changes enacted in 1990 and 1992. Federal preferences now applied to 50 percent of newly admitted households to public housing, 70 percent of newly admitted households to the project-based Section 8 programs, and 90 percent of newly admitted households to the tenant-based Section 8 programs. 1996 The continuing resolution enacted in January 1996 included a one-year suspension of the Federal preferences. This suspension was implemented by notice in February and March 1996. 1997 The Appropriations Act for fiscal year 1998 continued the suspension of Federal preferences. Federal preferences were controversial from the start because of a belief that they would lead to an excessive concentration of poor and nonworking families in particular housing projects. Exceptions intended to encourage greater economic diversity or income mixing within projects were enacted in 1992 and 1994. Then, in 1996, Congress suspended mandatory Federal preferences for one year. That suspension has been extended by subsequent appropriations laws, and the Federal preferences will almost certainly be repealed by new housing legislation. The program information shown in exhibit 18 dates from February 1997 for public housing and tenant-based assistance and June 1997 for project-based assistance. It probably reflects very little change resulting from the suspension of Federal preferences starting in 1996, since it takes time for program administrators -- and, particularly, public housing authorities -- to adopt new policies. In addition, the turnover rate for public and assisted housing is less than 15 percent per year, so for some time aggregate program data will reflect old policies. Waiting list preferences are not the only reason for the deep actual targeting of rental assistance programs. A particularly powerful factor is the calculation of the household's share of rent at 30 percent of household income. Relatively higher income households may find more attractive housing at 30 percent of their income than they could find in the assisted housing projects for which they might apply, particularly when those projects are located in relatively undesirable neighborhoods. Percent-of-income rents probably have been more important than the Federal preferences in the heavy use of housing assistance by those with incomes below 30 percent of the area median. Indeed, even before the Federal preferences were implemented in 1988, public and assisted housing programs were used primarily by households with extremely low incomes. However, some public and assisted housing projects are found in highly desirable locations and can attract relatively higher income households, even when they must pay 30 percent of their income as rent. Already there is a group of public housing projects that largely exclude the poor. Those projects identified as "high end" in exhibit 20 have fewer than 20 percent of their occupants with incomes below $10,000 per year. As the table shows, such projects are particularly likely to be located in low-poverty neighborhoods. Almost two-thirds (63.5 percent) are in census tracts with fewer than 20 percent of persons living in poverty, while only 35.4 percent of more typical projects are in these low-poverty locations. Exhibit 20 Some Public Housing Projects Exclude the Poor *A high-end project is any project in which at least 20 percent of the residents have incomes of more than $20,000 per year or at least 70 percent have wages as their primary source of income but in which fewer than 20 percent of the occupants have incomes of less than $10,000 per year. Source: HUD PD&R tabulations of census data documented in A Picture of Subsidized Households, U.S. Department of Housing and Urban Development, Office of Policy Development and Research, December 1996 Furthermore, recent legislation encourages more widespread use of "ceiling rents" for public housing. Ceiling rents permit public housing authorities to charge a maximum rent that is less than 30 percent of income to make public housing projects more competitive with private market alternatives and to help projects retain working tenants as residents when their incomes rise. Where opportunities to serve relatively higher income families exist, public housing authorities (PHAs) and owners may take advantage of them. Many PHAs would prefer to serve relatively higher income households as Federal budget reductions increase the prospect that PHAs will have less money for public housing operating costs. A change in program rules enacted in 1996 permits PHAs to keep some of the additional rental income that can be charged to such residents of public housing.8 Furthermore, serving relatively higher income households may reduce management and administrative problems and thereby reduce costs. When housing assistance is tenant based, the program's maximum subsidy, or Fair Market Rent (FMR), can have a powerful influence on whether the program is used mainly by the poor. FMRs are calculated and published by HUD to make available 40 percent of the rental housing in an area that passes basic health and safety standards and has turned over recently. In many locations, this means that families with incomes close to (or above) the very-low-income limit of 50 percent of local median income would receive a small enough subsidy that they do not have a strong incentive to join the program. (The FMR minus 30 percent of their income is a small amount.) However, the House version of the pending public housing authorization bill (see exhibit 5), would not only permit families with incomes up to 80 percent of the area median to receive tenant-based assistance, it would also permit the subsidy standard to be set at 120 percent of the FMR. This would on average make 75 percent of U.S. rental housing affordable under the program and make the program attractive to relatively higher income families. Given all these factors -- the attractive locations of some public and assisted projects, ceiling rents, incentives for PHAs to increase public housing rental income, and the possibility of higher FMRs -- the actual targeting of housing assistance to the neediest households could erode substantially over the next decade. To prevent this from occurring, the laws governing the public housing and Section 8 programs should include explicit income targeting rules that direct a substantial portion of housing assistance, overall and in each housing project, to the lowest income groups. Congress has refused for the past four years to appropriate funds for additional units of the type of housing assistance most easily focused on extremely-low-income households -- tenant-based rental assistance. Because of the subsidy formula (FMR minus 30 percent of income), tenant-based assistance can make housing affordable to the poorest families. Because the subsidies are used throughout the rental market, income mixing is not needed within the group of families that receive tenant-based assistance. * The current programs that produce affordable rental housing -- HOME and the Low Income Housing Tax Credit -- can serve extremely-low-income households at affordable rents, especially when combined with tenant-based rental assistance. Current housing policy includes two programs -- HOME and the Low Income Housing Tax Credit -- that produce affordable housing. Affordable housing is distinguished from assisted housing by the use of flat or fixed rents instead of rents that vary with a household's income (see exhibit 4). Generally, rental projects developed under HOME and the tax credit must set the flat rents at a level no higher than 18 percent of the area median income for the size household that would be expected to occupy units of different sizes (one, two, or more bedroom units). At the maximum allowable rent, the units are affordable at 30 percent of income to households with 60 percent of area median income. However, residents pay the flat rent regardless of their actual income. Rental projects developed under HOME and the Low Income Housing Tax Credit are occupied by families with a range of incomes, including extremely low incomes. But most households with extremely low incomes served by HOME and tax credit rental projects fall into one of two categories: They also have tenant-based assistance, or they have high-rent burdens, as shown for HOME in exhibit 21. Exhibit 21 The HOME Program Serves Extremely-Low-Income Households at Affordable Rents, Especially When Combined With Tenant-Based Rental Assistance An evaluation of the Low Income Housing Tax Credit by the General Accounting Office, completed in early 1997, suggests that 39 percent of households occupying tax credit units also receive a rental subsidy to make the housing truly affordable to that family. The average income of these households is at 25 percent of the area median, compared with 45 percent of median for renters of tax credit units that do not also receive a rental subsidy.9 * See exhibit 1 for the relationship between percent-of-median income levels and poverty thresholds. * Changes in governing legislation and program rules in 1996 permit PHAs to take actions to attract and retain working families by removing what were perceived to be work disincentives created by the percentage of income rent structure. By law, PHAs were permitted to adjust income used for rent calculations so that new earned income was not counted in the rent calculation for 18 months, and then phased in over a 3-year period. By regulation, HUD has given PHAs much greater freedom, permitting them to exclude all or part of the earned income of a family, or of an individual in the family, when determining overall income of the family for eligibility or rent determination purposes. A PHA takes some risk in adopting either the statutorily authorized adjustment to income or the regulatory exclusions from income, since any net loss of rental income would not be offset by increased operating subsidy payments. * United States General Accounting Office, Tax Credits: Opportunities to Improve Oversight of the Low-Income Housing Program, Report 97-55, Washington, DC, March 1997, pp. 136, 141, and 146. ---------- Finding 6: Of the 12.5 million persons in households with worst case needs, almost 1.5 million are elderly and 4.5 million are children. The number of adults with disabilities in households with worst case needs is estimated to be between 1.1 and 1.4 million. * Increases in the numbers of single adults with very low incomes and in the incidence of worst case needs among them suggest that worst case needs rose markedly among persons with disabilities. The 1996 worst case needs report, Rental Housing Assistance at a Crossroads, analyzed several sources to estimate the number of non-elderly adults with disabilities who receive housing assistance and the number who have worst case housing needs. That report concluded that in 1993 at least 17 percent of worst case households had adults with disabilities present. While there was a slight drop in the overall number of very-low-income households between 1993 and 1995, the number of non-elderly single persons living alone or with other singles in households with incomes less than 50 percent of the area median increased by almost 300,000. Moreover, worst case needs rose significantly among these households, from 44 to 48 percent. Therefore, the number of disabled adults living in households with worst case needs may well have grown to between 1.1 and 1.4 million by 1995.10 Housing assistance is an important part of this country's commitment to secure a dignified life for citizens with disabilities. Unless the number of households receiving housing assistance continues to expand, it may become an empty commitment for the hundreds of thousands of persons with disabilities who will have to pay more than one-half of their incomes for housing and/or live in physically deficient housing. In fact, the amount of public and assisted housing currently available to those with disabilities could shrink as a result of legislation that permits housing authorities and owners to bypass households with disabilities on waiting lists to create elderly-only "designated" housing projects (see exhibit 22). Exhibit 22 Designated Housing Recent legislation permits housing authorities and owners of Section 8 projects to "designate" projects for the exclusive use of the elderly or people with disabilities. If housing is designated for use by one group only, current occupants may not be displaced so long as they comply with the terms of their lease. But as units become vacant, they can be reserved for the designated type of household, even if another type of household has applied for the project and was placed on the waiting list at an earlier date. However, the legislation requires Section 8 owners to continue to serve some minimum number of persons with disabilities. Also, a PHA that designates housing exclusively for the elderly must have a plan to provide within the locality for the housing needs of very-low-income households with persons with disabilities who would have been served had the designation not been made, and the designation must be approved by HUD. As of late 1997, HUD had approved the designation of almost 19,000 units of public housing for exclusive use by the elderly. PHAs have requested fewer than the 6,300 available units of tenant-based Section 8 available for persons with disabilities affected by the designation because they have been able to use existing local resources to replace most of the designated units. * The number of elderly households with worst case needs problems remained above one million in 1995, while more than two million families with children had worst case problems. Among households with very low incomes, both families with children and households with an elderly head but no children had almost a one-in-three chance of having worst case needs. This situation occurs despite the fact that housing assistance has been heavily directed toward these two groups -- 37 percent of very-low-income elderly and 29 percent of very-low-income families with children receive housing assistance. Exhibit 23 Worst Case Needs by Household Type, 1995 *For example, non-elderly singles and childless couples. Between 1.1 and 1.4 million households are estimated to include adults with disabilities. Source: HUD-PD&R tabulations from the 1995 American Housing Survey Families with children represent the largest group of households with worst case needs -- more than 2.1 million households of a total 5.3 million worst case households. Just over 1 million elderly individuals or heads of households without children have worst case housing needs. The total number of elderly with very low incomes dropped between 1993 and 1995 by about 300,000. This may reflect a growing portion of the elderly population protected from economic distress by Social Security and private pensions. Nonetheless, many elderly remain both poor and in need of housing assistance. These elderly poor are among those Americans for whom overall economic expansion will not alleviate worst case housing needs. Continuing or returning to work or gaining additional income through marriage are often unlikely. As for all types of households, an overwhelming proportion (69 percent) of the elderly with priority housing problems are extremely low income, with incomes below 30 percent of area median (see exhibit 24). Similarly, unassisted, elderly-headed households whose incomes are below 30 percent of area median are far more likely to have acute needs (63 percent) than those with incomes between 31 and 50 percent (33 percent). Priority problems are rare among the elderly with incomes between 51 and 80 percent of median where only 72,000 (9 percent) have severe needs. Inevitably, proposals designed to attract families with a broader range of incomes into public housing and project-based assisted housing would have the consequence of reducing the number of worst case households that have access to housing assistance. These proposals should be carefully crafted to apply only to those situations in which there is a compelling need to make this tradeoff. For example, the Administration proposes to permit ceiling rents for public housing that are as low as 75 percent of the operating costs of a public housing development but would not permit such ceiling rents for developments occupied predominantly by the elderly. There is no compelling need for income mixing in housing for the elderly -- no children for whom relatively better-off families can provide role models and no need to build work incentives into the rent structure. Exhibit 24 Priority Problems Concentrate in the Poorest Households of Each Family Type * The results of welfare reform are not yet clear but are not likely to produce a substantial reduction in worst case needs. In 1995, when the current AHS was conducted, welfare reform had not yet been enacted. In the 1995 AHS, one fourth of those with worst case needs reported receiving welfare payments from either Aid to Families with Dependent Children (AFDC) or Supplemental Security Income (SSI). Among the 2.1 million families with children who had worst case problems, 930,000 (44 percent) reported AFDC or SSI payments. When worst case needs are measured on the basis of the 1997 AHS, most effects of welfare reform will still be unobservable because the data will describe conditions just one year after the enactment of the Personal Responsibility and Work Opportunity Reconciliation Act of 1996. The effect of welfare reform on the incomes of extremely-low-income families with children is still uncertain. States have great flexibility to design their own income support systems under Temporary Assistance for Needy Families (TANF), the successor to Aid to Families with Dependent Children, and some families will receive increased support for a limited time period because of generous disregard of earned income for determining benefit levels. In all States, however, some families will lose benefits because they do not cooperate with the new rules requiring work or training, while others may reach the time limits for TANF without finding jobs. Even families that do find jobs will lose the income associated with their former benefits. In the year following passage of the Act, welfare caseloads fell throughout the Nation, with the number of recipient families in the typical State falling by more than one-sixth. Favorable job markets, the specific policies implemented by the States, and the general message of the reform all played some part in this change. It is not possible at this stage to determine whether increases in earned income will exceed losses of welfare benefits for the affected families, but rigorous evaluations of State reforms started before the new Federal law provide some clues about the size of the effect that can be expected. * In Delaware, the evaluators found that after 12 months of implementation, the new welfare program raised employment by 24 percent over the old welfare program, but the average annualized gain in earnings for the caseload as a whole was just $668 per family per year.11 * After 21 months of experience with the new program in Minnesota, researchers found that the rate of employment among families who had ever received welfare rose from 38 percent under the older program to 52 percent under the new one. Earnings had risen at an annualized rate of $694 per family per year.12 * After 15 months of experience in Florida, analysts reported an increase in employment from 34 percent under the old program to 37 percent under the new one, with earnings over the past year higher by $157 per family per year under the new program.13 Based on this earlier experience, the incomes of families under national welfare reform may be higher, but not radically higher than they were under the old regime. Many affected families will have low wages and will likely have worst case needs despite working. Moreover, policies that effectively compel them to obtain access to a job may also lead some to rent more expensive housing than they did when on welfare, because housing that is convenient to jobs will be more costly than equivalent units that have less easy access. On balance, there is no reason to believe that welfare reform will lead to a significant decline in worst case needs over the next few years. * In 1995 the (AHS) found 656,000 renter households with no children or elderly persons but at least one person receiving SSI. Of these, 559,000 had very low incomes; 227,000 (41 percent) of these very-low-income households had worst case housing needs. The equivalent figures from the 1996 report, using the 1993 AHS, were 422,000 very-low-income non-elderly renters receiving SSI income, of whom 144,000 (34 percent) had worst case needs. However, HUD has always known that this AHS proxy for persons with disabilities is incomplete. The 1996 report compared the 1993 AHS with data from the Social Security Administration for Supplemental Security Income (SSI) recipients in 1994. An extrapolation from SSI figures estimated that 2,148,000 renters were non-elderly adults with disabilities, of whom 881,000 (41 percent) had worst case needs (see table 2 of the 1996 report). An equivalent updated comparison file from the Social Security Administration is not available at this writing. The range cited in the text (1.1 to 1.4 million) is an extrapolation that assumes constant relationships between the number of persons reporting SSI in the AHS and the number likely to be found in SSI data when they become available. * David Fein and Jennifer Karweit, The ABC Evaluation: The Early Economic Impacts of Delaware's A Better Chance Welfare Reform Program, Abt Associates, December 1997. * Cynthia Miller, Virginia Knox, Patricia Auspos, Jo Ann Hunter-Manns, and Alan Orenstein, Making Welfare Work and Work Pay: Implementation and 18-Month Impacts of the Minnesota Family Investment Program. Manpower Demonstration Research Corporation, October 1997. The research finding is a difference in earnings of $1,041 over 18 months, which we converted to an annualized rate in the text. * James Kemple and Joshua Haimson, Florida's Project Independence: Program Implementation, Participation Patterns, and First-Year Impacts. Manpower Demonstration Research Corporation, January 1994. ---------- Finding 7: Worst case needs continue to shift to the West. The number of very-low-income renters in the West continued to increase between 1993 and 1995, while dropping in other regions. Once again, the West had the highest percentage of very-low-income renters with acute housing needs, 42 percent, compared with 32 percent in the South, 33 percent in the Midwest, and 39 percent in the Northeast (see exhibit 25). The result is that the number of Western households with worst case needs reached a record 1.56 million in 1995. At the same time, renters living in the West who have very low incomes are considerably less likely to receive Federal housing assistance than households in other parts of the country. Only 18 percent of very-low-income renter households in the West receive housing assistance, compared with an average of 29 percent for very-low-income renters in the other three regions of the country. Virtually all public housing units and much of the project-based Section 8 stock were developed before the large population shift to the West. The recent growth in the number of tenant-based certificates and vouchers has not been large enough to balance these disparities in Federal housing assistance across the Nation.14 Exhibit 25 Western Renters Are Underserved Relative to Needs * The largest increases in worst case needs occurred in the Northeast and the West, where the mismatch between the wages of entry-level workers and the rents of even the most affordable housing continues to widen. Because affordability is the overwhelming cause of worst case housing needs, those areas of the country with high housing costs are the areas where worst case needs are growing the most rapidly. Increasing income inequality and the loss of rental units that the lowest income families can afford are particularly pronounced in these areas, as overall economic growth brings upward pressures on rent.15 Particularly in the Northeast and the West, working full-time at minimum wage or living on a modest pension does not bring in enough income to pay the rent. Between 1991 and 1995, the number of unassisted very-low-income renters with worst case housing needs increased by 160,000 in the Northeast and 226,000 in the West (see exhibit 14). Exhibit 26 Mismatches Between Extremely-Low-Income Renters and Available Rental Units They Can Afford are Worst in the West * Mismatches between extremely-low-income renters and rental units affordable to them are most severe in the West. The mismatch between available extremely-low-rent units and extremely-low-income renters is large and getting larger in all four census regions. Shortages are worst in the West, however, where in 1995 for every 100 extremely-low-income renters there were only 31 units that were either already occupied by extremely-low-income renters or vacant and for rent, compared with the nationwide figure of 44 units per 100 renters (see exhibit 26). * Since 1974, all housing assistance -- newly produced units of public housing and project-based Section 8 as well as funding for new units of tenant-based assistance -- has been allocated to different parts of the country on the basis of a formula. Although the formula has changed somewhat over time, it has failed to reflect worst case needs in any direct way and has never taken into account the effect of historical patterns of housing assistance. * Affordable Rental Housing: When to Build, When to Preserve, When to Subsidize? A Study of Housing Market Dynamics in 41 Metropolitan Areas, HUD, Office of Policy Development and Research, forthcoming May 1998. ---------- Chapter 3 Policy Implications The findings of this report suggest that economic growth alone will not ameliorate the record-level housing needs among families with limited incomes. Not even families working full-time at the minimum wage can afford decent quality housing in the private rental market. The report also makes it clear that housing needs are found not just in big cities but increasingly in the suburbs as well. The fundamental reform of the Nation's welfare system, enacted in 1996, is likely to have far-reaching effects on low-income families. These reforms occurred after the data in this report were collected, and it is too early to measure welfare reform impacts on housing needs. Clearly, though, many welfare recipients will find work. Some recipients will experience declining incomes, as they lose income support from either program sanctions or time limits. In both cases, housing assistance will continue to be needed to prevent and alleviate the distress and instability people experience when they pay so much for rent that they cannot afford other necessities. Worst case needs for housing assistance constitute a persistent structural problem of American society, requiring a significant public policy response. The report suggests a clear and compelling need for the Congress to provide greater support for Federal housing assistance -- by expanding both tenant-based rental assistance and programs that create and rehabilitate more affordable housing units. And Congress should act carefully in reforming the income -- targeting rules for public and assisted housing programs to balance the goals of achieving a greater income mix in public and assisted housing developments and of providing assistance to families with the most severe housing needs. A short-run policy response should include: * 103,000 New Housing Vouchers. The Administration has asked Congress to fund 103,000 new housing assistance vouchers, including 50,000 welfare-to-work vouchers to help welfare recipients find and keep jobs. Tenant-based assistance is a cost-effective way to reduce severe rent burdens and provide access to rental housing throughout the private market. For example, these vouchers can help extremely-low-income families live in housing supplied by HOME and the Low Income Housing Tax Credit at affordable rents. * Ending the Delay on Reissuing Vouchers. Congress should end immediately its cost-saving requirement placed on local housing authorities to hold for three months rental subsidies returned by families leaving the program. This practice reduces by 40,000 the number of subsidies in circulation and thus the number of families receiving housing assistance. * Expanding Production of Affordable Housing Through HOME and the Low Income Housing Tax Credit. The Administration is also seeking to expand tools to build and rehabilitate affordable housing. HUD's FY 1999 budget includes increased funding for the HOME program, along with a new HOME Bank, a loan guarantee feature that would allow communities to leverage up to five times their Federal grants for larger scale housing investments. In addition, the Administration is proposing a substantial expansion of the Low Income Housing Tax Credit that would create 180,000 new affordable rental units over the next five years. * Careful Income Targeting of Federal Housing Assistance. Congress is considering legislation which will determine the income levels of households who will be admitted to public housing and to receive Section 8 rental assistance. The Clinton administration and HUD strongly support the transformation of public and assisted housing developments into healthier, mixed-income communities. But policymakers must be careful not to exclude poor families altogether from these housing developments, nor to reduce unnecessarily the numbers of families with worst case needs who can be served by Federal housing programs. This report shows that this goal can be achieved while still continuing to serve families who are working but who have low incomes and serious housing needs. Three-fourths of tenant-based assistance should be reserved for families with incomes below 30 percent of the area median, which is the current practice. Forty percent of units that become vacant in public housing and in Section 8 projects should be reserved for extremely-low-income households. Each project should be required to serve a minimum number of extremely-low-income families to avoid gentrification of the most desirable and best located housing. ---------- Appendix A Data on Housing Problems Table A-1. Housing Conditions of U.S. Renters and Owners, 1995, by Relative Income Table A-2. Housing Conditions of All Renters and Owners, 1978, 1983, 1989, 1991, 1993, and 1995 Table A-3. Income Distribution and Ownership Rates of All Households With and Without Children, 1978, 1993, and 1995 Table A-4. Housing Problems of Very-Low-Income Renters by Household Type, 1978, 1983, 1989, 1991, 1993, and 1995 Table A-5. Housing Problems and Characteristics of Very-Low-Income Renters by Household Type, 1995 Table A-6. Housing Problems and Characteristics of Worst Case Renters by Household Type, 1995 Table A-7. Detailed Housing Problems of Worst Case Renters by Household Type, 1995 Table A-8. Housing Problems Among Very-Low-Income Renters by Race and Ethnicity, 1978, 1983, 1989, 1991, 1993, and 1995 Table A-9. Housing Conditions Among Very-Low-Income Renters by Region, 1978, 1983, 1989, 1991, 1993, and 1995 Table A-10. Assistance and Worst Case Needs Among Very-Low-Income Renters by Region and Location, 1995 Table A-11. Income Distribution, Priority Housing Problems, and Assistance of Renters With and Without Children, 1978, 1993, and 1995 Table A-12. Housing Problems, Characteristics, and Earnings of Non-Elderly Renters by Relative Income and Household Type, 1995 Table A-13. Assisted Renters and Percent With Wages as Primary Source of Income by Household Type and Relative Income, 1995 Table A-14. Measures of Housing Mismatch: Numbers of Affordable Units Per 100 Renters With Incomes Below 30 Percent or 50 Percent of Area Median Income by Region, 1989, 1991, 1993, and 1995 ---------- Appendix B Glossary Household and Family Types Family -- The "families" eligible for HUD programs have traditionally included households with relatives, households with children, elderly single persons age 62 or older, and single persons with disabilities. The Cranston-Gonzalez National Affordable Housing Act of 1990 broadened the statutory definition of "family" in a way that makes all households eligible for rental programs. In this report, however, the term "family" refers only to non-elderly "family households" in which one or more persons in the household are related to the householder by birth, marriage, or adoption. Families with children -- Household with a child under age 18 present. Elderly -- Household in which the head of household or spouse is age 62 or older, and there are no children present. Nonfamily households -- Households with a single non-elderly person living alone or only with nonrelatives. Households having members with disabilities -- Ideally, this category should include all non-elderly households with adults with disabilities present. However, none of the available data sources count these households perfectly. The American Housing Survey (AHS) proxy used in this and previous reports is known to be an underestimate, because it counts only single persons living alone or with nonrelatives who report receiving Supplemental Security Income (SSI). HUD program data show appreciably more households (without children) having members with disabilities receiving rental assistance than does the AHS proxy. New data on SSI recipients who are blind or have other disabilities permit more complete counts of very-low-income renters receiving HUD assistance or having a severe rent burden. Even these data exclude very-low-income persons who have disabilities with incomes above SSI levels. Types of Income Income -- Income in AHS is based on the respondent's reply to questions about income during the 12 months prior to interview. It includes amounts reported for wage and salary income, net self-employment income, Social Security or railroad retirement income, public assistance or welfare payments, and all other money income, prior to deductions for taxes or any other purpose. Following HUD rules for income eligibility, early worst case reports also included imputed income from equity in an owned home as income for owners, but income from equity is not included in this report. In 1993, AHS began asking more detailed questions on nonwage income, and the share of households reporting nonwage income rose from 63 percent (in 1991) to 77 percent. Family income -- Reported income from all sources for the householder (the first household member 18 years or older who is the owner or renter of the housing unit) and other household members related to the householder. Household income -- Reported income from all sources for all household members. Housing Problems Overcrowding -- The condition of having more than one person per room per residence. Rent or cost burden -- Ratio between payments for housing (including utilities) and reported household income. This calculation is based on gross income. It does not make the adjustments to income required by housing assistance programs before percentage-of-income rents are determined. To the extent that respondents underreport total income, the AHS estimates overcount the number of households with cost burden. Moderate rent or cost burden -- Housing costs between 31 and 50 percent of reported income. Severe cost burden -- Housing costs exceeding 50 percent of reported income. Inadequate housing -- Housing with severe or moderate physical problems, as defined in the AHS since 1984. These definitions are presented in appendix A of the AHS published volumes in detail and in appendix B of this report. Briefly, a unit is defined as having severe physical problems if it has severe problems in any of five areas: plumbing, heating, electrical system, upkeep, and hallways. It has moderate problems if it has problems in plumbing, heating, upkeep, hallways, or kitchen, but no severe problems. Priority housing problems -- Problems qualifying for Federal preference in admission to assisted housing programs: paying more than one-half of income for rent (severe rent burden), living in severely substandard housing (including being homeless or in a homeless shelter), or being involuntarily displaced. Because the AHS sample tracks housing units and thus cannot count the homeless, AHS estimates of priority problems in this report include only households with cost burdens above 50 percent of income or severely inadequate housing. Income Categories HUD-adjusted area median family income (HAMFI) -- In 1974, Congress defined "low income" and "very low income" for HUD rental programs as incomes not exceeding 80 and 50 percent, respectively, of the area median family income, as adjusted by HUD. Statutory adjustments now include upper and lower caps for areas with low or high ratios of housing costs to income and, for each non-metropolitan county, a lower cap equal to its State's non-metropolitan average. Estimates of the median family income and the official income cutoffs for each metropolitan area and non-metropolitan county are based on the most recent Decennial Census results and then updated each year by HUD. Each base income cutoff is assumed to apply to a household of four, and official cutoffs are further adjusted by household size: one person, 70 percent of base; two persons, 80 percent; three persons, 90 percent; five persons, 108 percent; six persons, 116 percent; and so on. Low income -- Reported income not in excess of 80 percent of HAMFI or, if lower, the national median family income. In 1995, 45 percent of AHS households reported incomes that fell below the low-income cutoffs. Very low income -- Income not in excess of 50 percent of HAMFI. In 1995, 27 percent of AHS households reported income below the very-low-income cutoffs. Extremely low income -- Income not in excess of 30 percent of HAMFI. In 1995, 14 percent of AHS households reported income below 30 percent of HAMFI. Poor -- Household income below the official national poverty cutoffs for the United States for that household size. The poverty cutoff for a family of four approximates 33 percent of HAMFI. Forty-four percent of very-low-income households and 85 percent of extremely-low-income households are poor. Middle income -- For this report, adjusted incomes between 81 and 120 percent of HAMFI. About one-fourth of households (24 percent) were in this category in 1995. Upper income -- For this report, households with income above 120 percent of HAMFI. One-third of U.S. households fell into this category in 1995. Housing Assistance Status Receiving assistance -- From AHS data, includes those responding "yes" to the following AHS questions: Is the building owned by a public housing authority? Does the Federal Government pay some of the cost of the unit? Do the people living here have to report the household's income to someone every year so they can set the rent? Worst case or with acute needs -- Unassisted very-low-income renters with the priority housing problems that give them preference for admission to rental assistance programs. Rent Affordability Categories Extremely low rent -- Annual rent, including utilities, is at or below 30 percent of 30 percent of HAMFI. For rents, the HUD adjustments vary by number of bedrooms to reflect expected household size: 0 bedrooms-1 person; 1 bedroom-1.5 persons; 2 bedrooms-3 persons; 3 bedrooms-4.5 persons, etc. Very low rent -- Annual rent, including utilities, is at or below 30 percent of 50 percent of HAMFI. Location (Standard) Metropolitan Statistical Area -- From 1973 to 1983, the definitions of metropolitan location in Annual Housing Survey data corresponded to the 243 Standard Metropolitan Statistical Areas (SMSAs) used in the 1970 census. Since 1984, metropolitan location in AHS has referred to MSAs defined in 1983, based on the 1980 census. Region -- The four census regions are the Northeast, Midwest, South, and West. ---------- Appendix C Procedures Used To Estimate Housing Needs From American Housing Survey Data To accurately estimate worst case needs for housing assistance from American Housing Survey (AHS) data, it is essential to determine whether household incomes fall below HUD's official very-low-income limits (50 percent of HUD-adjusted area median family income [HAMFI]), whether a household already receives housing assistance, and whether an unassisted income-eligible household has one or more of the priority problems that confer tenant selection preference (rent burdens exceeding 50 percent of income, substandard housing, or being involuntarily displaced). This appendix discusses the procedures and definitions used with microdata from the 1995 AHS to estimate the number of households in different income categories that have worst case needs or other housing problems. * All estimates in this report base income category and rent burdens on household income for all households. * Because HUD's official income limits have been based on 1990 census data since 1992, limits based on 1990 census data were used for this report. * Area income limits. To categorize households in relation to "local" income limits as accurately as possible within the limitations of the AHS geography, household income was compared with area income limits for all households. Very-low- and low-income cutoffs for a household of four -- that is, 50 or 80 percent of HAMFI, respectively -- were defined for each unit of geography identified on the AHS national microdata tapes. Official income limits were used directly for each of the 141 MSAs identified on the AHS tapes. For housing units outside these MSAs, the AHS geography identifies only four regions, metropolitan status, and six climate zones. Average income limits were estimated for each of these 48 locations. * Categorizing households by income. For all households, income status is determined by comparing household income with the very-low- and low-income cutoffs, with appropriate adjustments for household size. Households reporting negative income were categorized as middle income if their monthly housing costs were above Fair Market Rent (FMR), since many households in this situation appear to be reporting temporary accounting losses. * Receiving housing assistance. In AHS data, households are counted as receiving Federal housing assistance if they answered "yes" to one of the following AHS questions: Is the building owned by a public housing authority? Does the Federal Government pay some of the cost of the unit? Do the people living here have to report the household's income to someone every year so they can set the rent? Although the number and characteristics of households responding affirmatively to these questions are generally consistent with program data, detailed examination reveals that households often do not report their assistance status correctly. (See Duane T. McGough, Characteristics of HUD-Assisted Renters and Their Units in 1993, May 1997.) * Severe or moderate physical problems. The definitions are those used since 1984 in AHS and defined in appendix A of published AHS volumes. A unit is considered severely inadequate if it has any one of the following five problems: * Plumbing. Lacking piped hot water or a flush toilet or lacking both bathtub and shower, all for the exclusive use of the unit. * Heating. Having been uncomfortably cold last winter for 24 hours or more or three times for at least six hours, each due to broken down heating equipment. * Upkeep. Having any five of the following six maintenance problems: leaks from outdoors, leaks from indoors, holes in the floor, holes or open cracks in the walls or ceilings, more than a square foot of peeling paint or plaster, or rats in the last 90 days. * Hallways. Having all of the following four problems in public areas: no working light fixtures, loose or missing steps, loose or missing railings, and no elevator. * Electrical. Having no electricity or having all of the following three electrical problems: exposed wiring, a room with no working wall outlet, and three blown fuses or tripped circuit breakers in the last 90 days. A unit is defined as moderately inadequate if it has any of the following five problems, but none of the severe problems: * Plumbing. Having all toilets break down simultaneously at least three times in the last three months for at least six hours each time. * Heating. Having unvented gas, oil, or kerosene heaters as the main source of heat (since these heaters give off unsafe fumes). * Upkeep. Having any three of the six upkeep problems mentioned under severe problems. * Hallways. Having any three of the four hallway problems mentioned under "severely inadequate." * Kitchen. Lacking a sink, range, or refrigerator for the exclusive use of the unit. * Weighting of AHS estimates, 1990 based. Because each housing unit in the AHS sample represents many other units, the sample data are adjusted so that each year's total matches independent estimates of the total housing stock. For 1995, these independent estimates were based on the 1990 Census of Housing (1990 weights) ---------- End of Document