The Heritage Foundation

Backgrounder #1081 on Smart Growth

May 6, 1996

May 6, 1996 | Backgrounder on Smart Growth

Time for Bipartisam Reform of Public Housing


Housing and Urban Development (HUD) Secretary Henry Cisneros's bold plan to introduce the benefits of market competition into the troubled public housing industry has been stymied in Congress by an alliance of moderate Republicans and liberal Democrats. Yet, despite the opposition, these ideas -- which first took hold during the Reagan and Bush Administrations (especially during Jack Kemp's tenure at HUD) -- have shown remarkable resiliency. With the support of House reformers and leaders, and with bold intervention from the federal courts, there is a good chance that these needed reforms in America's New Deal housing policies will survive lobbying efforts of a public housing industry determined to preserve its costly taxpayer-supported monopoly over housing for the poor.

Ensuring this will be no easy task, however, as both the House and Senate committees with jurisdiction over housing policy already have produced bills that would reverse 25 years of market-based reform and, in the process, serve fewer needy families for the taxpayer dollars spent. The Senate's version (S.1260) has passed the floor. The companion version in the House (H.R. 2406) is scheduled for a vote in early May, but confronts a competing reform bill (H.R. 2198) developed by freshman Representatives Sue Myrick (R-NC) and Sam Brownback (R-KS). Significantly, the freshman bill is co-sponsored by the House leadership.

This clash over housing reform is far more than a dispute over spending levels. It is a dispute over how government should help the poor and a chance for the new congressional leadership to end the public housing industry's stranglehold -- a stranglehold that has hurt tenants as well as taxpayers and turned many of America's public housing projects into costly disasters. The leadership's attention, support, and intervention will be essential, both in defending reform against the public housing industry's lobbying juggernaut and in focusing lawmakers on the issues at stake.

What is at stake today is what has been at stake in every year since the early 1970s, when HUD first attempted to extract itself from the costly housing policies of the New Deal: the choice between tenant-based assistance and project-based assistance. Rent vouchers and certificates, the primary means of tenant-based assistance, provide eligible families with the cash equivalent to rent the apartment of their choice, subject to a cap based on local "fair market rents." They empower tenants and introduce market discipline into the assisted-rental market, forcing landlords and developers to seek to provide the best value for money.

In contrast, with project-based assistance (including public housing), the eligible family is offered an apartment in a building constructed and operated exclusively to house the poor. With its origins in the Depression, project-based public housing was developed as much for its direct job-creating potential as it was to meet the housing needs of the indigent; it is this dual purpose, among other factors, that explains why project-based assistance, on average, costs up to twice as much as vouchers for each family assisted. These units typically are owned and operated by government or nonprofit organizations, although some for-profit developers have been given significant tax and other financial incentives to provide such buildings. This approach empowers landlords and developers, whose incentive is to lobby Washington and produce housing as expensively as possible, not to provide value to tenants or the taxpayer.


The significant cost difference is one of the key reasons for the shift, under both Democrat and Republican Administrations, toward vouchers and away from public housing. But the different approaches also have led to dramatic differences in the quality of life for assisted families, and this has been another compelling reason for HUD's repeated attempts to extract itself from this Depression-era program.

Because public housing serves only the poor, typically in large complexes of a hundred or more units, families assisted by this program are rigidly segregated from the rest of society, always by income class and also by education and workforce participation. In the major urban areas, public housing residents are segregated almost exclusively by race. Indeed, one would be hard-pressed to find another American institution that induces measures of racial segregation as effectively as does public housing.

Apparently the courts agree. In April of this year, in settling a desegregation suit brought against HUD by the American Civil Liberties Union of Maryland on behalf of several Baltimore public housing tenants, HUD agreed to desegregate its Baltimore public housing projects, to tear down the decrepit buildings, and to provide the majority of the displaced tenants with vouchers that can be used to rent privately owned apartments within the greater Baltimore metropolitan area (see box).

In contrast with public housing, vouchers allow the assisted family to choose its place of residence from private landlords in the community, provided that the required rent stays within the established limits. This freedom allows assisted families to escape the worst urban communities. As a 1994 congressional report observed, "Studies have found that recipients of tenant-based rental assistance were less likely than public housing residents to live in concentrated poor urban communities.... "1 With vouchers, a family with children can choose a safer neighborhood with good schools; an unemployed individual can relocate to a part of the community with better job prospects; and an elderly couple can choose a place close to family, friends, and convenient shopping. Whatever their specific needs, families can "self-tailor" the assistance in ways that provide the greatest benefits for a given cost. Moreover, because vouchers are a fixed amount of assistance, tenants have a strong incentive to "shop around" for the best value, and landlords and developers have the incentive to provide good value to the tenants (which does not happen when tenants must live in designated projects).

Recognizing these advantages, President Clinton's FY 1996 budget wisely proposed to shift virtually all housing assistance from the project-based form to tenant-based vouchers. Under this plan, HUD would end the multibillion-dollar direct payments to public housing authorities (PHAs) and replace them with indirect support through vouchers that would be provided directly to the tenants now living in these projects. In turn, the tenants could use these vouchers to continue to pay rent to the PHA, rent better quarters elsewhere, or join with other tenants to hire better managers. In any case, PHAs would lose their government-sponsored monopoly status, and instead would have to compete for assisted tenants by offering better service at a competitive price. Confronted with the prospective loss of revenue, PHAs would have no choice but to improve their performance and lower their costs.


The FY 1996 HUD plan was the most innovative housing proposal in 25 years and relied heavily on both privatization and competitive market forces. But Congress rejected this proposal, and instead drafted new legislation that largely shifted the housing assistance emphasis back to public housing while merely paring back some of the programmatic impedimenta that had accumulated over the 30 years of HUD's existence. In making this decision, both committees said nice things about vouchers but then went on to cite the findings of a June 1995 GAO report2 that questioned their cost, with the Senate committee stating:

The Committee believes that a total conversion to a voucher system is a "One-size-fits-all" approach that is not appropriate or will not work in all markets or in all circumstances. For example, a June 1995 study by the General Accounting Office determined that while nationwide the cost of vouchers versus the cost of operating public housing is similar, the averages conceal wide differences in these two options in different market areas.3

The House came to a similar conclusion. It acknowledged that on average vouchers were 10 percent cheaper than public housing per household served, but added: "As GAO noted, these wide variations in cost raise a number of important issues.... "4

The GAO's acknowledgement that sample observations are distributed around their mean (in other words, a bell curve) would hardly have surprised a high school math student, but this commonplace point was sufficiently alarming for both congressional housing committees to reject the HUD reform plan and continue the six-decade reliance on public housing as the primary vehicle for providing assistance to the poor. Yet the findings presented in the GAO report do nothing to support Congress's action; instead, they support the 1995 HUD proposal to "voucher out" the public housing projects.


One of the reasons HUD recommended a wholesale shift from public housing to vouchers was the significant cost difference in favor of vouchers. As acknowledged in the GAO report, HUD determined "that it cost $440 per month on average to house a family with a housing certificate, compared to an average of $481 per month in public housing."5 If spread over the 1.3 million families now in public housing, this difference, described as "similar" by the Senate, would lead to an annual savings of $640 million per year. In turn, if the savings were applied to additional vouchers, Congress would be able to assist another 118,000 poor families now eligible for HUD assistance but unable to obtain it because of limits on HUD funding.

The prospect of a GAO-acknowledged savings of $640 million alone should have encouraged Congress to begin the prompt yet orderly shift recommended by HUD Secretary Cisneros. But the GAO report presents plenty of additional information to demonstrate that the potential long-term savings from shifting to vouchers are even greater than the acknowledged 10 percent. Indeed, the true cost difference may be as much as ten times the amount HUD and GAO presented when a full accounting of all public housing costs is compiled.

For example, the same section of the GAO report that acknowledges the 10 percent difference also says, in a sentence apparently overlooked by both the House and Senate committees, that "because HUD's analysis considered the capital costs of public housing as sunk costs, it made no attempt to add an imputed debt service charge to the cost of public housing."6 In other words, if we assume that public housing costs nothing to build, then public housing has only a 10 percent cost disadvantage when compared with vouchers. The GAO comparison is a classic case of apples and oranges because the cost of the voucher, as related to the rent it pays, embodies all costs, including operating, capital, and debt, as well as other expenses from which public housing is exempt, such as federal, state, and local income taxes, local property taxes, and local licenses and fees. No such costs are included in GAO's representation of public housing's expenses.

Fortunately, the Senate Banking Committee and the GAO provide additional information in their respective reports that can be used to establish a fuller accounting of the costs associated with public housing and allow for a more accurate comparison of these costs with the cost of vouchers. When this is done, it is apparent that, on average, the true cost of public housing is as much as twice the cost of a voucher.

The search for a more accurate assessment of public housing's costs begins with an estimate of the federal government's investment in public housing, an investment comprised largely of the cost to construct and the cost to renovate and modernize the units periodically as they deteriorate over time. An estimate of this total is included in the Senate report, which states that "the Committee is committed to safeguarding the federal taxpayers' $90 billion investment in the nation's public housing industry.... "7

Combining this $90 billion estimate with the GAO's observation that there are approximately 1.4 million public housing units in existence8 yields an average per unit historic investment/cost of $65,000. Of course, today's per unit costs are much higher than the average of costs cumulated over past years. HUD estimates current public housing development costs at an average of $88,073 per unit.9 For example, HUD has just announced the planned construction of a new project in Washington, D.C., that will have an eye-popping estimated expenditure per unit of $186,500, and a new project in Baltimore entailing expenditures of $340,000 per unit when the cost of other amenities at the project is included.10

In order to pro rate this historic $65,000 cost over the useful life of the project, as well as establish an estimate for the costs associated with the debt service charges that would be incurred for this estimated average investment, a useful proxy is the monthly payment associated with a 30-year fixed rate, level payment, fully amortized mortgage. Using an interest rate of 7.5 percent as a conservative average for the government's long-term borrowing rate over the past several decades, this approach yields an additional monthly cost of $454.49.11 When added to the HUD/GAO monthly operating cost estimate of $481, this yields a more accurate total monthly public housing unit cost of $935.49, a figure that is more than twice the cost of vouchers. Using the current cost of building a new public housing unit -- $88,073 -- the monthly debt service charge would rise to $615.82 and the total monthly cost of that new unit would be $1,096.82.

With a cost difference of this magnitude between the two programs, the rationales for the policies embodied in both the housing bills developed by both the Senate and House committees (S. 1260 and H.R. 2406) fall apart. Specifically:

1. Vouchers are by far the better form of assistance.

With full program costs exposed as not "similar," public housing no longer merits the benefit of the doubt, and vouchers no longer deserve the skepticism that characterizes their treatment by Congress. As a result, the policy emphasis in both these bills must be reversed and the burden of proof placed on public housing, not on vouchers. Specifically, vouchers should be the chief vehicle for housing assistance, except when a particular public housing project is not used as a vehicle for racial segregation and can be demonstrated to be less expensive as a result of an independent audit covering all costs.

2. There are no regional cost differences of consequence once total costs of public housing are included.

If public housing overall is twice as expensive as vouchers, the assertion that there are widespread regional differences in costs is also invalid. When the excluded capital costs presented by the GAO are added, each of the examples which purport to demonstrate the lower cost of public housing actually demonstrates just the opposite; not one of these examples survives the re-estimation.12

3. An independent study commission should be established to settle the cost issue once and for all.

To date, all cost comparison studies have been conducted by institutions, or contractors to institutions, with a vested interest in the outcome of the debate. As a result, and regardless of the quality of the studies, the analysis is perceived to reflect inherent biases, or at least to be tainted with the view of its benefactor. As a congressional agency, for example, the GAO cannot be seen as wholly neutral. The reason: It carries out research according to specific requests from Members of Congress and sometimes is constrained by the requesting lawmaker in the approach it can use to explore issues and evaluate costs. Nor can studies conducted by HUD be seen as free of bias in regard to prevailing HUD policy preferences.


While the cost differences cited above should leave no doubt that all future and incremental housing assistance should be delivered through vouchers for reasons of both cost and quality, there remains the question of how to make the best use of the existing public housing stock, estimated by GAO at 1.4 million units. Although the taxpayers have invested an estimated $90 billion, this sunk cost does not justify any further investment in the system, since spending any additional dollars on public housing reduces by more than half the number of families that can be assisted, when compared with those that could be assisted through vouchers

Still, investing no more money in public housing does not mean that the existing stock of public housing units has no role to play in an orderly transition from New Deal paternalism to more cost-effective tenant empowerment. In its present state, but under proper management,13 much of the inherited public housing stock can provide adequate and cost-effective shelter for those in need, and the assisted tenants should be allowed to live there if that is their choice.

Providing tenants with the choice, and providing them with the funding rather than the landlord, puts tenants in a position to get the maximum benefit from government assistance. Providers of housing, including the public housing authorities, must compete with others in offering quality services to secure their operating revenue, which previously came directly from HUD with no effective requirement that minimal quality standards be met or cost efficiencies established.

If all public housing operators were funded only by the revenues available from federally assisted rent-paying tenants, financial survival would compel these operators to provide better services. Just as important, tenant-based assistance would force public housing authorities to decide which units to continue in operation and which to close down. Units in extreme disrepair would be rejected by voucher-bearing tenants with the right to choose. Ultimately, they would be shut down, as the federal courts forced HUD to do in Baltimore, and the money that otherwise would be wasted on uneconomic modernization could be used to assist a larger number of beneficiaries directly.


It is important for Members of Congress to keep in mind that federal housing assistance is not an entitlement, and that the level of housing assistance serves only a fraction of the eligible population. As the Congressional Budget Office has pointed out:

In 1989, the most recent year for which the detailed data used in this study were available, about 4.1 million households received assistance from the federal government in meeting their housing needs. Under the program rules in effect in 1994, almost 14.5 million additional households would have been eligible for aid; that is, their incomes were sufficiently low to qualify them. Not all of the households that were eligible, however, applied for aid, whereas many of those that did apply were placed on waiting lists because sufficient aid was not available.14

In other words, for every one household served by HUD, more than three eligible ones were not. Thus, within the confines of current budget limits, the money saved through program efficiencies could be reallocated to assist the most needy of the 14.5 million poor yet unserved families now on waiting lists. The per-unit expenditures of $186,500 and $340,000 for new buildings in Washington and Baltimore, noted earlier, represent an egregious waste of money and perpetuate the long waiting lists which characterize HUD programs.

Example: If the $115 million that HUD intends to spend to construct just 338 new units in Baltimore was applied to vouchers, HUD could have provided a year's worth of safe and decent privately owned apartments to 21,000 of the 25,000 poor Baltimore families on the housing assistance waiting list.

Given such trade-offs, Congress's continuation of new public housing funding defies reason -- and merely rewards the construction and public housing industry lobbies.

Example: Consider the $2.5 billion that the Senate Appropriations Subcommittee on VA, HUD, and Independent Agencies proposes to modernize portions of the existing public housing stock. Had this been allocated directly to certificates or vouchers, this $2.5 billion could have provided needed help to more than 475,000 poor households currently on waiting lists for government housing assistance.

While some in Congress may defend this multibillion-dollar modernization expenditure as necessary to provide additional housing units for the poor, the government's own statistics demonstrate that it is a waste of money. According to the government's American Housing Survey (see box), there are approximately 2.6 million vacant rental units in the United States, and about 39 percent of them, or just over one million, rent for $400 or less per month -- well within the $440 range of the average HUD voucher. About 49 percent, or 1.3 million units, rent for $500 or less, also within HUD's range given that most assisted tenants are expected to supplement the voucher with their own financial resources.15 In central cities, the vacancy rate is much higher because of continuing depopulation.


Although public housing authorities are financially dependent upon the federal government for virtually all aspects of their operations, Congress has not had much success in using the leverage this dependency should provide to induce improved performance. Despite numerous financial concessions, its not-for-profit status, exemptions from all taxes, and exhortations to better management, public housing still costs more than vouchers on a per unit basis.

The difficulties that both HUD and Congress have had with public housing are underscored by the 15 percent of the units under the control of authorities formally described year after year as "distressed." Although the Cisneros voucher plan would have transferred the hapless tenants in these units to private sector housing and put the poorly managed PHAs out of business, Congress chose instead to rely on modified versions of Washington-managed solutions. Yet, what little management and oversight of public housing there is comes from HUD. And it was Secretary Cisneros's understanding that trying to manage from Washington is not the answer that contributed to his voucher plan to replace failed central planning and management with the discipline of the competitive marketplace.

Aspects of both congressional committee bills could make this situation worse by diminishing executive branch oversight of the housing industry and key decisions in the choice of assistance. For example:

  • HUD's oversight of the PHAs would be diminished by the House committee bill's creation of an independent Housing Foundation and Accreditation Board, comprised largely of the recipients and beneficiaries of federal spending on housing. This board would establish benchmarks for local housing and management authorities, establish procedures for accrediting authorities, and classify authorities. It is through board accreditation that a PHA would become eligible to receive the proposed block grants under this bill. This proposal creates a legislative precedent by turning over executive branch spending authority to an independent board comprised of individuals whose chief qualification for eligibility is the possession of a conflict of interest.
  • Both bills move the responsibility for determining the future course of federal housing policy from Congress and the executive branch to the PHAs by granting to PHAs the responsibility for conducting cost comparison studies on vouchers and project-based assistance. The bills then give PHAs the discretion to act on the outcome of their own studies. With so much turning on a single cost comparison, it is hard to understand how Congress can place responsibility for cost comparison studies on the bodies that stand to benefit or lose financially from the outcome of such studies. As recommended earlier, these comparisons should be conducted by a blue ribbon commission of independent experts.


While the taxpayer bears an indirect burden for having his money wasted on poorly conceived programs, the brunt of the failure is born by society's most vulnerable households, including millions of children who are condemned to live in these public housing units or, worse, are denied any housing assistance at all because billions of dollars are wasted on costly and inefficient housing programs. As noted above, the Senate's proposed $2.5 billion modernization program will be at the expense of nearly 500,000 eligible households who will receive no assistance this year because of budget shortfalls in other program accounts.

Added to this is a Republican-endorsed change in public housing regulations that will shift the government's limited housing resources away from the poorest segments of the population by allowing local public housing authorities to "serve residents with a greater range of income."16 As the Senate Committee report puts it: "The Committee's bill promotes mixed-income conditions by repealing federal regulations that prescribe strict occupancy preferences and deregulating well-run public housing authorities."17

If these proposed changes are enacted into law, the government's long-standing commitment to use its limited housing assistance funds to help those in greatest need first will be rejected in favor of policies that enforce notions of economic diversity. With 14 million eligible but unserved households, of which about 8.4 million are categorized by the CBO as "very low income renter households,"18 this policy change, at least on the surface, would appear to be ethically odd. But upon closer inspection, it can be justified by the sort of "burn the village to save it" logic often compelling in the narrow factual confines in which government discourse is engaged.

Although they have not yet been enacted into law, HUD has wasted no time in putting these ideas into practice. In announcing, in early April 1996, a grant of $25 million to build 134 new units (equaling $186,500 per unit) near the U.S. Capitol, HUD also noted that households with incomes as high as $78,430 would be eligible to move into this project.19 HUD made no mention of its plans for the 15,983 poor people in the District of Columbia who are on government waiting lists for housing assistance.20

As noted earlier, one of the more serious drawbacks of public housing, and indeed of any government bricks and mortar housing program, is that it serves to segregate households by race and income. By warehousing such large concentrations of the poor, often in deteriorated inner-city neighborhoods where much of the nation's urban public housing is located, Congress has created volatile social environments characterized by high crime, drug abuse, child neglect, and other social pathologies that tend not to occur as frequently with the less costly voucher program that allows assisted households to self-integrate into the community at large. But rather than blame a fundamentally flawed program, Congress blames the victims, and proposes to solve the problem by serving fewer of the poor.

The House committee bill would allow public housing authorities, once compelled to take the poorest and most vulnerable on a priority basis, to rent to higher income families, and would require that no more than 25 percent of a project's units be dedicated to the very poor, characterized as those with incomes below 30 percent of the area's median family income. Expressed another way that better quantifies the human cost of the proposed change, Congress will allow public housing authorities to set aside no more than 350,000 units to serve a population that the CBO estimates at 11.9 million eligible households, of which about 8.4 million receive no housing assistance at all, in part because of programmatic financial limits. With public housing authorities allowed to discriminate against the poor, the worsening situation will likely deteriorate even further.21

However bad some of the housing legislation before Congress may be, there remains the growing possibility that federal courts increasingly will render efforts now underway by Congress and HUD as irrelevant and in violation of the law. As court-directed settlements in Baltimore and several other communities demonstrate, tenants and their legal counsel are becoming increasingly less tolerant of the pervasive de facto segregation that characterizes most of HUD's urban public housing projects, and many of the suburban ones as well. Increased use of vouchers is the only acceptable alternative to the plaintiffs, and to the courts which render the final decision.

But as the Baltimore case also demonstrates, decisions by the courts also can be remarkably coercive, and could ultimately be as disruptive and as counterproductive as were the school busing decisions of a past generation. But this need not happen if Congress enacts a new initiative that breaks with past housing policy and the federally financed de facto racial segregation that is an integral part of it. Legislative proposals such as those of Representatives Myrick and Brownback, or HUD's plan of last year, each of which is founded on tenant choice, would be consistent with American principles of civil liberties. In contrast, the legislation produced by the housing committees of both the House and Senate simply entrenches today's disastrous programs and further frustrates reform.


  1. Congressional Budget Office, The Challenges Facing Federal Rental Assistance Programs, December 1994, p. xix.
  2. U.S. General Accounting Office, Public Housing: Converting to Housing Certificates Raises Major Questions About Costs, GAO/RCED-95-195, June 1995, p. 3.
  3. The Public Housing Reform and Empowerment Act of 1995, Committee on Banking, Housing and Community Affairs, U.S. Senate, 104th Cong., 1st Sess., December 20, 1995, p. 22 (hereafter Senate Report).
  4. United States Housing Act of 1996, Committee on Banking and Financial Services, U.S. House of Representatives, 104th Cong., 2nd Sess. February 1, 1996, p. 90 (hereafter House Report).
  5. GAO, Public Housing, p. 3. The GAO reproduced these numbers from "Will It Cost More to Replace Public Housing with Certificates," Issue Brief #1, Office of Policy Development and Research, U.S. Department of Housing and Urban Development, March 1995.
  6. GAO, Public Housing, p. 4. In a footnote to this sentence, GAO defends its decision to ignore this $90 billion cost by noting that "The Federal government has already paid for the construction of public housing. Accordingly, there are no mortgages on public housing or associated debt payments." While this is true in a narrow technical sense, the federal government has, over time, paid approximately $90 billion for this public housing and has borrowed from the public to do so, as the government's $1 trillion in outstanding debt reveals.
  7. Senate Report, p. 22.
  8. Approximately 100,000 vacant units explain the difference between the 1.4 million public housing units and the 1.3 million households living in public housing.
  9. U.S. Department of Housing and Urban Development, Budget Summary Fiscal Year 1997, p. H-8.
  10. See Vernon Loeb, "After Three Decades, The Bulldozers Show Up," The Washington Post, April 3, 1996, p. B1, and Michael James, "Lafayette Courts: 40 Years from High Hopes to Oblivion," The Baltimore Sun, August 16, 1995, p. 1A. The per unit cost estimates were derived from information provided in these two articles. Repeated efforts to confirm with HUD were met with non-response.
  11. Thirty years was chosen because it is both a conservative estimate and, at present, the government's official estimate of the useful life of structures, as defined by the Internal Revenue Service for purposes of determining allowable depreciation charges. Given the exceptional wear and tear that the typical public housing unit is subject to, a shorter period of, say, 15 or 20 years would be more reasonable and also would lead to a higher monthly cost estimate. As for the interest rate, this also allows the analysis to err on the conservative side inasmuch as the average 30-year government bond rate was 9.2 percent for as long as there has been a 30-year bond (since 1977). Obviously, this too would yield a higher monthly cost. Alternatively, using the 10-year government bond rate, whose series goes back to 1953, a time span that more closely parallels the construction of today's public housing stock, the average 10-year interest rate is 6.8 percent, and this would yield a monthly debt service cost of $423.79, making public housing still the most expensive form of housing assistance.
  12. Obviously, the GAO's confusion over the properties of normally distributed observations is also rendered moot, and there is no need to elaborate further on this matter.
  13. Ronald D. Utt, "Time for New Management at America's Troubled Public Housing Projects," Heritage Foundation Backgrounder Update No. 247, May 17, 1995.
  14. CBO, The Challenges Facing Federal Rental Assistance Programs, p. xiii.
  15. The data presented in this paragraph are found in The Statistical Abstract of the United States, 1995, p. 738, in Table No. 1231. For an independent confirmation of estimates of this magnitude of the available inventory of suitably priced rental units, see CBO, The Challenges Facing Federal Rental Assistance Programs, p. 44, which notes: "Of the 28 million unsubsidized units in the United States, 16.4 million, or close to 60 percent, had rents that were no greater that the local FMR." FMR is the fair market rent, a measure calculated by HUD to determine the appropriate value for vouchers in each community. See also, U.S. Department of Housing and Urban Development, "Rental Housing Assistance at a Crossroads: A Report to Congress on Worst Case Housing Needs," March 1996. The HUD study notes that there are 175 vacant units available at rents below the FMR for every 100 worst-case-need households that needed to move -- that is, live in substandard housing.
  16. Senate Report, p. 4.
  17. House Report, p. 84.
  18. CBO, The Challenges Facing Federal Rental Assistance Programs, p. 45.
  19. Loeb, "After Three Decades, The Bulldozers Show Up."
  20. Barbara Vobejda, "Low Income Housing Crisis Is Escalating, HUD Reports," The Washington Post, March 15, 1996, p. A27.
  21. HUD recently reported that families with worse case housing needs increased by 1.1 million between 1978 and 1991, and another 400,000 by 1993. Vobejda, "Low Income Housing Crisis Is Escalating, HUD Reports."


Under the guidance of U.S. District Court Judge Marvin J. Garbis, in April 1996, HUD finally agreed to the terms of a settlement of a January 1995 desegregation suit (Thompson vs. HUD, et al., Civil No. MJG 95-309) brought against it by the American Civil Liberties Union of Maryland on behalf of several African-American families living at Lafayette Gardens, a decrepit HUD-supported public housing project operated by the Baltimore Public Housing Authority.

According to the Baltimore Sun, "The suit alleged that the city, the Housing Authority of Baltimore City, and the U.S. Department of Housing and Urban Development illegally segregated black public housing tenants for six decades, which the defendants deny. It said they would 'rebuild segregation for generations' and perpetuate poverty if allowed to redevelop the high rise sites as public housing."1

In April 1996, HUD agreed to a $300 million settlement under which, among other promises, it would tear down several of the dilapidated, segregated public housing projects and replace them with 814 new, low-rise public housing units. Money earmarked for 150 of these units could be redirected to vouchers if Congress concurs. HUD also would provide vouchers to allow another 1,342 families to move into better neighborhoods. In addition, 664 of the proposed 814 new public housing units or vouchers could be made available for homeownership, partly under a resuscitated version of former Secretary Kemp's HOPE program. Traditional public housing is nowhere part of the court-supervised desegregation settlement.

Although in principle the settlement is a significant endorsement of the Reagan-Kemp empowerment policies of the 1980s, in practice the implementation details are remarkably coercive and would deny the winning tenants the right to exercise their choice in using the voucher.

Specifically, the assisted tenants, who are almost exclusively African-American, can use their vouchers only if they agree to move into a "non-impacted" area. In HUD-speak, non-impacted areas are defined census tracts, mostly in the surrounding suburbs and predominantly white in racial composition. In response to suburban concerns that the federal government will be forcibly relocating poor, inner-city blacks to their communities, HUD has established annual quotas under which the relocation program will be conducted. For example, only 60 black families per year will be permitted to move into neighboring Baltimore County in each of the next six years until the 360-family quota of African-Americans for the county is exhausted in 2002.

Although dramatic in cost and scope, the Baltimore settlement is not the first of its kind, and is not likely to be the last, unless Congress abandons its six-decade obsession with public housing.2


Although the perception within Congress is that there is a dearth of affordable housing available to the poor, the facts indicate otherwise. The table below, prepared by the U.S. Bureau of the Census, demonstrates clearly that a substantial portion of the U.S. rental stock is available at rents well within the range of the resources available to the typical HUD voucher recipient.

National Distribution of Apartments (Private and Public) by Rent - 1993

Gross Rent # of Units (1,000)
No Cash Rent 2,414
Less than $100 551
$100-$199 2,079
$200-$299 3,152
$300-$399 4,812
$400-$499 5,701
$500-$599 4,817
$600-$699 3,683
$700-$799 2,382
$800-$999 2,257
$1,000 and above 1,625
Total Occupied Units 33,472

As the table indicates, there are 33,472,000 occupied rental units in the United States. In 1993, the most recent year for which such data are available, the vacancy rate calculated by the Bureau of the Census was 7.4 percent of the rental stock, roughly as it has been for most of this decade. This means approximately 2,674,000 additional units were vacant. Further, the number of affordable vacant units available from the private sector, estimated at 1.3 million units, is nearly equal to the entire stock of public housing units now in existence. With the private sector able to provide these units at rental rates below either the partial cost ($481) or the full cost ($935) of public housing, one is hard-pressed to explain Congress's long-standing bipartisan obsession with this failed approach.

 It should be noted that the vacancy rate is higher, and housing therefore more available, within urban areas, reflecting the long-term depopulation trends within those communities as more and more Americans choose not to make their homes in cities. In 1993 when the overall vacancy rate was 7.4 percent, the rate inside Metropolitan Statistical Areas was 7.6 percent, while that outside MSAs was 6.4 percent. Thus, it should come as no surprise that 89 percent of voucher recipients are able to find suitable, qualified, affordable rental housing from the private sector.

The claimed shortage of affordable housing is one of the more popular rationales used by the public housing industry and its defenders to justify pouring billions of dollars into renovation, modernization, and new construction. Indeed, so great is this obsession that for years Congress has forbidden HUD to tear down any public housing unit, no matter how dilapidated, unless it first agrees to replace it with a new unit. Although both the House and the Senate propose to repeal this absurd and long-standing restriction, apologists for public housing still go to great lengths to misconstrue the actual state of affairs, as the House Banking Subcommittee report demonstrates with its creative interpretation of a GAO finding. On the matter of affordability and availability, the GAO says that:

Generally, the lower the vacancy rate in an area, the more difficult it will be for assisted households to find alternative housing. In the New York City housing market, for example, where about 11 percent of the nation's public housing stock is located, the vacancy rate is less than half the national average, making alternative housing more difficult to find.3

Seven months later, the House Subcommittee on Housing had this to say on the same matter in a paragraph referencing the GAO's findings:

Also, for vouchers to succeed, housing must be available, and availability is very much market specific. For example, the vacancy rate in markets such as New York City, where about 11 percent of the nation's public housing stock is located, is less than one half the national rate. Providing a voucher in this area would be useless.4

Despite the difficulties that more than 50 years of rent control and rent stabilization have created for the New York City rental housing market, it is significant that HUD has a 62 percent success rate with vouchers and certificates.5


  1. James Bock, "City to Get $300 Million for Housing," The Baltimore Sun, April 8, 1996, p. 1A.
  2. Similar programs have been implemented in Chicago; Dallas; Allegheny County, Pennsylvania; Omaha; and Minneapolis.
  3. GAO, Public Housing, p. 7.
  4. House Report, p. 91.
  5. See Section 8 Rental Voucher and Rental Certificate Utilization Study: Final Report, October 1994, Executive Summary, p. ii. A study prepared by Abt Associates for the U.S. Department of Housing and Urban Development.

About the Author

Ronald D. Utt, Ph.D. Herbert and Joyce Morgan Senior Research Fellow