May 6, 1985 | Backgrounder on Welfare and Welfare Spending

A Strategy for Helping America's Homeless

(Archived document, may contain errors)

i I 43 1 May 6, 1985 A STRATEGY FOR HELPING AMERICA'S HOMELESS INTRODUCT I ON c To judge from recent media reports, America faces a growing crisis of homelessness million homeless Americans has been repeated so of ten that it has now acquired the status of conventional wisdom. Yet this figure vastly exaggerates and distorts the number of homeless: even the meaningless. If In fact, America s homeless probably number no The staggering figure of two to three author of the study first containing the figure describes it as I I I more than 350,000 The vast majority of today's homeless, moreover, are not otherwise typical Americans who have suffered massive economic catastrophe; rather, they are either dependent on drugs o r alcohol or they are mentally ill and on the streets because of the movement more than a decade ago to empty a large share of the nation's mental institutions. A smaller group among the homeless are the very poor, often welfare families, unable to find af fordable housing due ta changes in the nation's rental housing market over two decades..

The homeless are not neglected and ignored In fact, efforts to feed and house them have been growing in recent years cally, the greater visibility of the homeless stems in part from these attempts to help. The federal government has several programs to ai d the homeless only treats the problem it does not cure it. An attack must be made on the underlying causes of homelessness and this requires that the states accept the obligations they have sought to avoid.

New federal programs will simply produce another coalition of bureaucrats and activists with a vested interest in prolonging the homeless I'crisis Ironi But emergency food and shelter 2 There are several steps to alleviate some of the causes of homelessness that the federal government could take immedi ately.

Federal law should be amended to require states to provide ade quate community mental health care before releasing patients from state institutions and to admit them if community care is inade quate. The federal government should also ensure that fe deral redevelopment grants are not used to lessen the low-income urban housing stock subsidies from costly construction to rehabilitation,and direct aid to the poor to make most efficient use of available funds Federal housing policy should continue to re d irect HOW HOMELESS ARE THERE The CCNV Estimate During congressional hearings in 1980, Mitch Snyder of the Community for Creative Non-Violence (CCNV), a homeless advocacy group, charged that official data on the homeless were woefully inadequate. To l1corr e ctf1 this, CCNV undertook a survey of its own. It was a completely nonscientific study which presented only the sketchiest data from a very limited number of locales but nonetheless has been treated as hard fact.l There is no indication in the 1980 CCNV r e port that the sampling of 14 Itkey cities which included one of just 55,000 residents) was random, or that any demographic factors were taken into account in selection. No systematic explanation of the study's methodology has ever been produced by CCNV, d espite congressional requests.

It appears that an unspecified number of unstructured telephone interviews were conducted with private and public service providers and agencies to elicit their opinions.

CCNV chairman Snyder claims that information was rece ived from Ifmore than 100 agencies and organizations in 25 cities and states averaging four calls per locale, although the report provides homeless estimates only for 14 cities.2 n I I There is little reason to believe respondents for this survey were cho sen by other than arbitrary and subjective criteria.

The CCNV data yielded widely differing estimates of homelessness for each locale, and no attempt was made to explain, reconcile or verify these estimates. The 1980 report includes, for example one inform ant's bizarre estimate of 250,000 homeless in Chicago. I Some estimates appear to be for metropolitan areas, others for Community for Creative Non-Violence, A Forced March to Nowhere, Washington D.C September 1, 1980.

Written testimony by Mitch Snyder, in Joint Hearing on "D Report on Homelessness, House Banking and Government Operations Committees, May 24 1984, pp. 33-

34. For a new edition of the study, the information was supplemented by calling "another couple of hundred people" in unspecified locales . cities annual family only; some appear to be point-in-time counts, others counts In one case, numbers of families, rather than members, are enumerated.

In short, the CCNV-data are useless as a means of estimating the number of America's homeless. Indeed , Snyder never claimed that this study was scientific, nor did his initial 1980 report include a U.S. homeless total. Nevertheless, according to a 1982 version of the report At that time [1980] we concluded that approximately 1 percent of the population, o r 2.2 million people lacked shelter...we are convinced the number of homeless people in the United States could reach 3 million or more during 1983.'13 The 1980 study reported some sketchy data and no homeless total the 1982 version omitted some of the or iginal data, expanded the narrative, and claimed to have produced a total of 2.2 million in 1980.

Even the upper range of the estimates Snyder presented in 1980 would yield homeless rates ranging from only several hun dredths of a percentage point to half a percentage point in half the cities he enumerates each in Louisville, Chicago, and Washington, D.C.--lead to a homeless rate over 1 percent, even using the city, rather than metropolitan area, population in all cases So where did the widely publicized s t atistic come from that two'to three million Americans, one percent of the population are homeless? Why does it still appear in CCNV documents? With disarming honesty, Snyder told a congressional panel last year Only three individual estimates-one these nu mbers are in fact meaningless. We have tried to satisfy your gnawing curiosity for a number because we are Americans with western little minds that have to quantify everything in sight, whether we can or not.

Counting'the Homeless=-The HUD.Report The Depar tment of Housing and Urban Development in 1983 I I decided to compile official data on homelessness. Last May the results of HUDIs six-month study were released.5 The report's conclusions were based on a review of nearly 100 local and national studies, ov e r 500 interviews with local observers in a nationally representative sampling of 60 metropolitan areas, site visits in ten localities, a national survey of shelter operators, a 50-state I I Mary Ellen Hombs and Mitch Snyder, Homelessness in America, A For ced March to Nowhere, Community for Creative Non-Violence, Washington, D.C., 1982 p. mi.

Joint Hearing op. cit., May 1984, p. 32 IUD Office for Policy Development and Research, A Report to the Sec'retary on the Homeless and Emergency Shelters, May 1, 1984. 4 survey of state activity, and discussions with representatives of national homeless advocacy groups.

Based on four different approaches, each with a different rationale and methodology meticulously explained in the report HUD concluded that on an average night in December 1983 to January 1984, the homeless numbered from 192,000 to 586,0

00. HUD con sidered the limost reliable range" to be from 250,000 to 350,000.

The HUD figure was a llpoint-in-timett* estimate Since. many' of the homeless are only temporarily without a residence, the number of people who experience homelessness during any year, of course would be far greater than this range. But the point-in-time estimate is a far mo r e meaningful indication of the extent of the problem It is the method, for instance, by which unemployment is measured. HUD clearly stated what its figure signifies. In contrast, the CCNV study does not explain whether its numbers are annual or point-in-t ime counts.

One of HUD's methods made use of the highest published local estimates in 37 localities. Ten of the localities were among the fourteen surveyed by the CCNV report HUD of making conscious efforts to minimize the problem, the CCNV and HUD estimat es should be compared for these ten cities Since CCNV later accused HOD (1981-83) CCNV (1980 Atlanta Ba 1 t imo re Boston Chicago Detroit Los Angeles New York Pittsburgh Seattle Washington 9 3,500 15,000 8,000 25,000 8,000 30,000 50,000 1,500 5,000 10,000 1,000 8,000 5,000 250,000 500 8,500 75,000 135 Not Clear 15,000 If Chicago, clearly an anomaly in the CCNV data, is discounted the total HUD estimates are actually 30 percent higher than those of CCNV. So why are HUD's national totals so different from CC N V's? HUD's method for extrapolating a national homeless rate from their data is clear and methodical--CCNV's method is unex plained and inexplicable 5 Reaction-to the HUD Report The HUD study used routine and rarely questioned survey and statistical techn i ques, yet it immediately encountered intense criticism-presumably because it did not find that two million homeless were camping on America's sidewalks and sleeping on grates. The press emphasized that the HUD results were disputed by homeless advocacy gr o ups, and HUD was accused of ''playing games.Il6 Critics claimed that-the HUD study,was not objective although it was conducted by professional civil servants whose tenure at HUD predated the Reagan Administation), and that the tlowlt numbers were simply c ooked to justify llcallousll Reagan Administration budget cuts. It was not noted that HUDIs numbers were ltlowll only by comparison to CCNV' s unsubstantiated figures.

Instead, the HUD report itself was compared to Nazi propaganda by the chairman of the co ngressional hearings on the HUD study.7 The criticism of HUDts methodology simply does not bear scrutiny. Many of the allegations by critics are misinformed or false. The compl.aint that people living on temporary vouchers in hotels and motels were not co u nted by HUD is si.mply erroneous, as is the accusation that HUD had used a ridiculously low estimate of 12,000 for New York City's homeless to establish their reliable range. Nor did HUD obtain llartificiallyll lower rates by counting only inner-city home l ess while dividing by the metropolitan area population to obtain the rate. HUD obtained its homeless rates by 'employing credible and consistent techniques, In the several hundred pages of testimony presented at hearings.on the study there was not a singl e valid methodological criticism that the HUD report itself did not raise or that HUD spokesmen did not subsequently answer.8 Curiously, despite the careful attention critics paid to HUD's methodology, they almost universally over looked the totally unscie ntific basis of the CCNV estimate that there are two to three million American homeless.

The attack on HUD left the Reagan Administration reeling.

Take the Federal Emergency Management Agency, which is charged with administering a substantial portion of f ederal funds for the homeless. FEMA recently released a study that ignores the HUD figures. When asked why HUDts figures were not used, Dennis Kwiatkowski of FEMA replied that HUD's numbers have been seriously di~credited The fact is that HUD's figures ha ve not been discredited; they only have been attacked--unconvincingly Colman McCarthy Just What the Homeless Needed," Washington Post, May 12, 1984 Joint Hearing, op. cit., May 1984, p. 9.

Ibid., pp. 281-287 and 297.

Telephone interview with Dennis Kwiatkowski, Chief of the Division of Indiiidual Assistance, March 11, 19

85. The-FEMA study was released as The National Board Emergency Food and Shelter Program Study of Homelessness March 1, 19

85. It talks of percent increases without mentioning specific n umbers reference point makes the percentage useless. So FEMA intends to base its policies only on whether the homeless problem is getting better or worse without reference to its scale Since a percent is-a ratio between two numbers, the lack-of a 6 Many c o mmentators have suggested that an argument over numbers is somehow irrelevant and that discussion should focus only on solutions. Yet the numbers are'of critical political significance at the federal level If the U.S. is swamped with millions of homeless A mericans, then a better case can be made for treating the matter as a federal problem. But if the home lessness is on the scale that the HUD figures suggest, there is little justification for asking Washington to intervene--in yet another area. Instead cu r rent. federal efforts could be redirected rather than expanded homeless-and society does have a moral responsibility for them is with state and local governments and private organizations The main burden of responsibility for the WHO ARE THE HOMELESS The HUD study used a very broad definition of homelessness.

Analysts counted a person as homeless if his or her nighttime residence was in a public or private emergency shelter, such as a church basement or government building, or in any public or private spac e not designed for shelter, for instance, a sidewalk subway, or car. The END definition also encompassed individuals and families living in welfare hotels on temporary vouchers.1 Thus the definition included the chronically homeless and those who for a va r iety of circumstances are homeless on a temporary or occasional basis, including those in the care of a welfare system of service providers, led HUD to conclude that the homeless comprise three general categories:ll A review of numerous studies, plus inte r views with hundreds 1) Mental Illness and Substance Abuse: A survey of shelter operators by Westat, Inc. for END "suggests that about half the shelter population suffers from mental illness and/or alcoholism and drug abuse.1112 This share is lower than fo r the homeless as a whole, because the hard-core street population, which does not use shelters, tends to have a higher incidence of such problems It is also low because the shelter operators appeared reluctant to characterize their clients as chronically i ll. Many local studies show a higher proportion of the chronically disabled homeless. In Boston, for example it has been estimated that between 60 and 70 percent of the homeless are mentally ill (among women, closer to 90 percent).13 A recent study in Was h ington D.C found that 73 percent of the city's sheltered homeless were schizophrenic or alcoh01ic.l lo A Report to the Secretary, op. cit pp. 7-8 l1 Ibid p. 22. p. 24 l3 l4 12 Ibid U.S. Conference of Mayors, Homelessness in America's Cities, June 1984 p. 15.

Reported in The Washington Post, April 24, 1985. 7 Since the homeless who do not use shelters are likely to have even a higher incidence of such problems, it is clear that at least half and probably closer to two-thirds of the homeless population is ch ronically disabled because of mental illness or substance abuse. The principal reason for this high proportion of chronically disabled among the homeless population is the policy known as ltdeinstitutionalization.ll As a result of the 1963 Community Menta l Health Centers Act, large numbers of mentally ill Americans were released from state mental institutions; and fewer were admitted to them. The population of the country's mental institutions was 505,000 in 19

63. It is approximately one-quarter of that f igure today.15 The intent was well-meaning, to provide patients with care in community-based centers, or what was termed "least restrictive settings.Il But the effect was very different. The policy abandoned many mentally ill Americans to the streets with inadequate care or no care at all. While some mental health specialists now support legislation to require that mental health patients be placed in "optimum therapeutic settings" (that is, readmitted to mental hospitals where community-based care is unava i lable others oppose changing the standard from the present "least restrictive setting," on the grounds that this would "reduce the pressure on local governments to develop appropriate community mental health facilities and would mean that states would go b ack to relying more on institutions.I1l6 In other words, homeless people should be hostages to secure better local policies 2) Personal Crises: Crises that lead to homelessness include divorce, release from a hospital or prison with no place to go, and do m estic violence. Many local studies indicate that a personal crisis may account for half of all instances of homeless ness. Since such homelessness is usually temporary, however personal crises account for only a small minority of the total at any given ti m e, probably' as low as 10 percent 3) Economic Conditions: Media reports and claims by home less advocacy groups suggest that two' of the major causes for the alleged rise in "new homelessnessll among middle-class families are unemployment, said to be due t o Reagan economic policies, and recent budget cuts in federal social programs. During the reces sion of 1981 and 1982, media reports focused on the supposed rise of this middle-class new homelessness. Comparisons were made with the Great Depression l5 l6 l 7 Federal Task Force on the Homeless, Regional Meetings Briefing Book March 7, 1985, p. 2 Views of Norman Rosenberg, Mental Health Law Project, as reported in the Washington Post, March 20, 1985 HHS Working Group on the Homeless, Report to the President, A ugust 15 1984 Reasons for Home1essness"--chart .a 8 Subsequent analysis, however, has demonstrated that there is no basis in fact for the contention that there has been economical ly induced new homelessness among the middle class.18 ployment and the rece ssion had been causing homelessness in 1981-1983, then the problem now would be declining dramatically.

Some seven million jobs have been created since the trough of the recession If homelessness is still on the rise, as homeless advocates insist, labor market conditions cannot be the main cause. This is not to deny of course, that some homelessness is caus e d by economic conditions, primarily among the very poor especially single-parent welfare households. But this is not the Itnew1l middle-class homelessness that has been widely alleged If unem Urban Policies Many urban policies have been driving families o nto the sidewalks. Most of these policies are a legacy of the 1970s.

Investment in low-income rental housing, for instance, has de clined, because there has been little incentive to invest. Tax advantages and inflation-induced capital gains, for instance h ave encouraged the conversion of multi-family units to condomi niums.

Rent control is -also to blame In over 200 cities, rent control limits severely the potential profitability of rental housing investment and thus predictably prevents the,private housin g market from responding to demand.. The President's Commis sion on Housing reported in 1982 that rent control leads to disinvestment in rental housing, either by prompting conversion to cooperative and condominium ownership or by encouraging deferral of n ecessary maintenance. This has led to the abandonment of many low-cost rental units and diminished low-income housing stock.lg The end result is a lower quality Urban redevelopment policies similarly have exacerbated the homelessness problem by wiping out many low-income residential areas, rooming houses, and it single-room occupancy" hotels to make room for higher-priced housing, hotels, and commercial buildings.

During the 1970s, for example, the U.S. lost about 1 million single-room occupancy units=-nea rly 50 percent of the total.20 The Itnew homeless1! are thus not new at all. The families in New Yorkls costly welfare hotels were there in the 1970s, but they were not part of the perceived I1homeless1I problem. For the l8 Gregg Easterbrook Housing: Exam i ning a Media Myth," The Atlantic October 1983 pp. 10-24 l9 The Report of the President's Commission on Housinp Washington, D.C 1982 DD. 91-94 2o U.S. House of Representatives, Committee on Government Operations, The Federal Response to the. Homeless Crisi s , Third Report, April 18, 1985 House Report 99-47), p.3. 9 most part, the new homeless are simply the old homeless made more visible because localities first found it financially expedient to put people out on the street, and then morally expedient to shi f t the blame for the resulting homelessness onto the shoulders of an allegedly heartless federal government THE HOmLESS AND PUBLIC POLICY The 1984 HUT) report states that early in January that year there were about 111,000 shelter spaces available nationwi d e for as many as 350,000 homeless.21 serious capacity shortage hotels and motels provided by welfare departments (and sometimes private agencies) was not counted as shelter capacity by HUD even though researchers believe the number of homeless housed in s u ch units is substantial.22 While many shelters, moreover, are forced to turn away people, indicating mismatches between shelter Zocations and concentrations of the homeless, the average national monthly occupancy rate in shelters in January 1984 was 70 pe rcent.

Even for the highest occupancy months for 1983, the occupancy rate never reached 100 percent.23 that simply providing more shelters is no guarantee that people will come in off the streets to use them This may seem to indicate a But emergency housin g in commercial This certainly demonstrates Why do men and women stay on the streets even when shelter is available? Experts cite several reasons. For.one thing, the homeless sometimes are simply unaware of the existence of a shelter. For another, the men t ally ill on occasion may be incap able of making a rational decision, even when they know of a shelter's availability. Another factor is that many shelters impose conditions that the homeless find unacceptable. Example attendance at religious services may be required, or drugs and alcohol may be banned. The Wall Street Journal recently reported that, of some 250 homeless one street who were offered rides to a shelter by a New York City official, only 36 accepted.24 During last winter's severe freeze, the m ayors of Philadelphia and New York City had to order their police departments to use force to bring the homeless to shelters to prevent their freezing to death that some of the homeless are llchoosinglt to remain on the street.

This is what prompted Ronald Reagan to note correctly 21 A Report to the Secretary, op. cit p. 34 22 Ibid p. 36. 23 The FEMA report claims occupancy rose to 92 percent by December 1984 (p 15).

HUD rate and the FEMA rate are not comparable. In fact, the FEMA method But they used a di fferent method of calculating occupancy, so the does not measure capacity at all December compared to occupancy in not capacity. 24 Joseph Perkins New Institutions Journal, February 26, 1985 their percent reflects occupancy in the highest-occupancy month in 1984 for the Homeless The Wall Street I 10 To be sure, no rational person "choosesIr homelessness, but many of the homeless are far from being rational.

What is Being Done?

The private response to the problem of homelessness has been The number of shel ters in the last four years has Only 6 percent are provided by city and extraordinary jumped 41 percent. Some 94 percent of the nation's shelters are operated by churches, synagogues, nonreligious groups, and other voluntary organizations county governmen t s.25 funded 63 percent of 1983 shelter operating expenses. Shelters additionally benefited from thousands of volunteers who, on average, donated the equivalent of four hours assistance every night for each bed in the nation's shelters. Private food assist ance to the poor and homeless is also extensive In 1983, the Second Harvest network alone (comprising 79 of the more than 300 food banks in the U.S distributed 118 million pounds of food donated by the food industry and private donors of this food, worth $ 78 million, went to food centers patronized by the homeless homeless. Across the nation, 80 percent of city and county gdvernments operate shelters, give financial grants to private shelters and service providers, lease or rehabilitate buildings for priva t e shelter providers or furnish vouchers for hotel motel, and apartment accommodation for the homeless. State governments, by contrast, have done comparatively little for the homeless. They act primarily as a conduit for federal funds to local governments. Maryland, New Jersey, California, Massachusetts and New York, however, recently have appropriated substantial sums to aid the homeless Private sources totaling $136 million About 40 percent Local government is also heavily involved in aiding the Federal a s sistance exceeds the $210 million for emergency food and shelter distributed by the Federal Emergency Management Agency (FEW) over the last two years. 'Millions of additional dollars have been applied to the problems of the homeless through such programs a s the Department of Agriculture's Emergency Feeding Programs ($1 billion in surplus ,commodities since 1982, $50 million a year toward distribution costs, and 75 million to emergency feeding centers HUDrs Community Development Block Grant ($62 million in F Y 1983 for emergency food and shelter HHS's Alcohol, Drug Abuse and Mental Health Block Grant 1.3 million for research grants in FY 1984); and HHS's Program for Runaway and Homeless Youth ($23 million in FY 1984 Federal funds from General Revenue Sharing a nd block grants are also available for the homeless. In addition, some 20 to 35 percent of the homeless receive help from various federal entitlement 25 Task Force Briefing Book, op. cit p 3. The remainder of this section is based on information contained in the.briefing and the HUD report 11 programs, including Medicaid, Medicare, Food Stamps, Aid to Families with Dependent Children, Social Security Disability Insurance, Supplemental Security Income, and Veterans Cash and Medical Benefits.

RECOMMENDATIONS Homelessness is clearly a serious problem, evenuif it is much less rampant than some advocates for the homeless claim.

But if sensible and sensitive policies are to be formulated, it is necessary to recognize that what is loosely called Ilhomeless ness" has several very different causes--the most important and distressing of which has been the thousands of the mentally ill released from hospitals. Policy makers thus should devise assis tance that best meets the needs of each segment of this vulner able p opulation. Policy makers also must determine who is respons ible for providing assistance and the steps necessary for eliminat ing the root causes of homelessness.

Among the approaches to be considered Mental Health The federal community mental health legi slation of 1963 allowed states to save money by deinstitutionalizing the chronical for them. Services for the mentally ill traditionally have been a state function, but states currently provide only 8 percent of the funding for the homeless shelters that m ust now cater to thousands of former mental patients. ly mentally ill without providing necessary community services Whatever philosophical differences there'are in the mental health community on this issue it is clearly inhumane and callous to leave the h omeless mentally ill in the streets while they are resolved. The states should make greater use of existing federal block grant monies to deal with the mental health problem, and following the example of the five states that already have done so, appropri ate more of their own funds to this end.

Recommendation: The federal government should amend the 1963 law so that states must provide adequate mental health care in the com munities before releasing patients from state institutions.

Housing Federal housing policy needs reform, but so do state and local housing policies unprofitable and it should be abolished. To the extent that eliminating rent control would cause temporary hardship to some of the poor, direct rental subsidies in the form of housing vouchers, should be made available by states Rent control has made low-income housing 12 Recommendation: The federal qovernment should not permit local governments to undertake any urban redevelopment efforts with federal funds, such as th ose available through the Urban Develop ment Action Grant proqram, that substantially diminish the low income housinq stock.

Recommendation: The federal qovernment should continue rehabili tating existinq housing and should start giving direct housinq subs idies to the poor in.the .form of vouchers rather .than fund ing expensive new housing construction Recommendation: The federal qovernment should continue rehabili tating existinq housing and should start giving direct housinq subsidies to the poor in.the .form of vouchers rather .than fund ing expensive new housing construction.

Recommendation: The federal government should encourage innova tive local responses to housing shortaqes by tenant and community organizations, such as the recently announced deci sion by HUD to foster homeownership in public housing projects, to help stabilize the housinq stock.

General Strategy Congress should resist the temptation to set up any long-term program directed specifically at llhomelessness.lf This would not solve the problem's root causes, but would simply throw money at the symptoms and create yet another new federal bureaucracy policy reform at the federal, state and local levels.

What is really needed is fundamental housing and mental health I I I Federal involvem ent should.be limited to the coordination and leadership function that currently characterizes the Federal Task Force on the Homeless. Among the reasons for limiting Washington's role to the homeless. Moreover, traditional responsibilities and the fact th a t the states are healthier financially than the federal government strongly suggest that programs assisting the homeless should be created and financed at the state level I 1) Considerable federal assistance is already being provided I I I 2) The individu a l and intensive nature of caring for the homeless and the extensive private sector involvement make it very difficult for a federal program to be flexible enough to meet homeless needs in an appropriate manner. Homelessness I invariably involves local cha r acteristics that need to be diagnosed and treated on the local and state levels 3 It is not at all certain that a federal program would not replace, rather than supplement, present local and private efforts. Government bureaucratization of social services has discouraged private initiatives in the past.26 26 Robert L. Woodson, A Summons to Life (Cambridge, Massachusetts: Ballinger Publishing Company, 1981), especially pp 82-88; and Nathan Glazer Towards a Self-service Society," The Public Interest, Winter 1983 pp 66-90.

CONCLUSION The most reliable estimate for the number of homeless in the U.S. is between 250,000 and 350,0

00. No reliable study puts the homeless population at between two and three million.

I1estimatel1 has achieved the status of conventi onal wisdom only through repetition This The root causes of increasing homelessness are~not unemploy ment and federal budget cuts, despite the allegations of city and state officials who understandably would prefer to evade their own responsibility for th e problem and let Washington foot the bill. Rather, the causes are ill-conceived mental health and housing policies on the federal, state, and local levels. The deinstitutionalization of the mentally ill and the loss of low income housing stock due to urba n development, gentrification and rent control--developments of the 1960s and 1970s--are the chief causes of the homelessness of the early 1980s.

More federal funding is not the answer. States should face up to their obligations and bear a larger burden in assisting the homeless, including making better use of discretionary federal block grant funds. Most important, fundamental changes in mental health and housing policies at all levels of government are essential if America is serious about eliminating ho melessness S. Anna Kondratas Schultz Fellow

About the Author