DOCUMENT RESUME

ED 316 600 UD 027 224

TITLE Housing and Homelessness: A Report.INSTITUTION National Alliance to End Homelessness, Washington,

DC.

PV DATE Jun 88NOTE 72p.; Table, pages 64-67, contains photoreduced type

of poor legibility.PUB TYPE Reports Research/Technical (143)

EDRS PRICE MFO1 /PCO3 Plus Postage.DESCRIPTORS At Risk Persons; Demography; *Homeless People;

*Housing; Housing Industry; *Housing Needs; *HousingOpportunities; Human Services; *IndividualCharacteristics; Low Rent Housing

IDENTIFIERS Single Room Occupancy

ABSTRACTThis report focuses on options for rehousing the

individuals and families who are currently homeless in America, andon strategies for preventing homelessness of additional people. Asmany as 736,000 persons are estimated to be homeless on any givennight, and between 1.3 million and 2 million different individualsmay experience homelessness at some point during the year. Thefollowing characteristics of the homeless are outlined: (1) familiescomprise about 25 to 33 percent of the homeless population, themajority of which are female-headed; (2) mentally-ill individuals andalcohol or substance abusers comprise 50 percent of the homeless; and(3) the remaining 25 percent of the homeless population includesemployable individuals, those for whom some form of training isrequired before employment, and small segments comprised of theelderly, the physically disabled, and veterans. The followingstrategies are suggested: (1) expand the supply of low-incomhousing; (2) preserve the existing subsidized housing stock; (3reassess welfare allotments; (4) improve single-room occupancy (SRO)housing; (5) reduce housing costs through innovative design andregulation; (6) create alliances among public, private, and nonprofithousing providers; and (7) develop counseling and assistance programsfor those at risk of homelessness. Statistical data are included onfour tables. Estimates of the homeless population by city, theresults of local studies of the characteristics of the homeless, anda list of 38 references are appended. (FMW)

Reproductions supplied by EDRS are the best that can be madefrom the original document.

*

//,/?r,

HOUSING AND HOMELESSNESS

eport of

The National Alliance to End

U S DEPARTMENT OF EDUCATION/a° p Lti f. di+ .001,61 i-N5kri r oriel irTqw.,,e,,,p,,,

DOCATIC)NAI FiF I()Olqt:157 IN* t/f4MAT IONCF:NTI FI rE FCC,

4-*This Etta L.Amt-at hag tree, reprtsta. eLl ns/ta. raved In on mu. itemr,r, r), ofqanIzatal,,

1/1.1.,Ir chanura r,avt tart*fr Math, to 01,p,t,fai,Rtoy

. . . .

ttoltS (1/ v.e* Ur op.a.014 SPIVC1 ,r 111,S thm1,1,4,11 CIO At e!Aiiity rppre!wrIt ofI IF t3, tw+1,

omelessness

"P[J 10 REPRODUCE THISIVIATL.RIAT HAS Eit...1::N GRANTED HY

r-)

tins 1,-, /

r**)

TO THE EDUCATIONAL RESOURCES;NfrOtiMAT ;ON CL NT ER ;F:MC)

HOUSING AND HOMELESSNESS

was prepared by theNational Alliance to End Homelessness

Housing Council

JUNE, 1988

Copyright@ 1988 by The National Alliance to End Homelessness

No part of this book may be represented by any mechanical,photographic, or electronic process, or in the form of aphonographic recording, nor may it be stored in a retrievalsystem, transmitted, or otherwise copied for public orprivate use, without written permission from the publisher.

Printed in the United States of America

Additional copies of this report are available for $5 eachfrom the National Alliance to End Homelessness,1518 K Street, N.W., Suite 206, Washington, DC 20005(202) 638-1526

Art77_

WE NATIONAL ALLIANCE ID END HOMELESSNESS, INC.

THE NATIONAL ALLIANCE TO END HOMELESSNESS

BOARD OF DIRECTORS

CLIFFORD L. ALEXANDER, JR., CHAIRMANWashington, D.C.

SUSAN G. BAKER, VICE CHAIRMANWashington, D.C.

EILEEN M. EVANS, SECRETARYWashington, D.C.

SUSAN CULLMAN KUDLOW, TREASURERWashington, D.C.

ANN L. ALLENWashington, D.C.

ROBERT M. BEGGANAlexandria, Virginia

PATRICIA CARLILEWashington, D.C.

JOHN D. DRIGGSPhoenix, Arizona

MARGARET M. GRAHAMWashington, D.C.

THOMAS L. KENYONWashington, D.C.

DAVID C. McCOURTWashington, D.C.

COL. ERNEST A. MILLERWashington, D.C.

ROBERT M. MOSBACHERHouston, Texas

BETTY ROBERTSChevy Chase, Maryland

RABBI MARTIN SIEGELColumbia, Maryland

STAFF

THOMAS L. KENYON, Executive DirectorNAN P. ROMAN, Director of Policy

SHERRY C. TYLER, Administrative Assistant

1518 K STREET NW SUITE 206 WASHINGTON. DC 20005 TELEPHONE 1202) 638 -152b

HOUSING COUNCIL MEMBERS

J. RODERICK HELLER III, CO-CHAIRMANPresident and Chief Executive Officer

National Corporation for Housing PartnershipsWashington, D.C.

BETTY ROBERTS, CO-CHAIRMANMember, Executive Committee

National Alliance to End HomelessnessWashington, D.C.

NANCY ANDREWSProgram Investment Officer

Ford FoundationNew York, New York

BROTHER JOSEPH BERGCatholic Charities, U.S.A.

Washington, D.C.

MICHAEL CARLINERStaff Vice-President

Economics and Housing PolicyNational Association of Home Builders

Washington, D.C.

BLAKE CHAMBLISSArchitect

Anderson, Mason DaleDenver, Colorado

JAMES H. COUSINSSenior Vice-PresidentLegislative Affairs

U.S. League of Savings InstitutionsWashington, D.C.

LYNN EDWARDSAssistant Director

Resource DevelopmentThe Enterprise Foundation

Columbia, Maryland

DAVID GLOTH, M.D.Real Estate DeveloperMaiden, Massachusetts

6

KIM HOPPERNational Coalition for the Homeless

Yonkers, New York

HARRIET IVEYVice-President

Community Re'!ationsFannie Mae

Washington, D.C.

HERSHEL LIPOWDirector

Division of Housing and Community DevelopmentMemphis, Tennessee

DAVID MAXWELLChairman

Fannie MaeWashington, D.C.

COL. ERNEST MILLLRDirector

Nation,.- Public AffairsThe Salvation ArmyWashington, D.C.

JOSHUA MUSSPresident

JMA AssociatesWashington, D.C.

FRANK PARENTLEconomist

Economic Research DepartmentAFL/CIO

Washington, D.C.

EDWARD L. QUINNPresident

The Enterprise FoundationColumbia, Maryland

ANDY RAUBESONDirector

Sinejle Room Occupancy CorporationLos Angeles, California

JOHN B. SLIDELLSenior Vice-President

Oxford Development CorporationGreenbelt, Maryland

JAMES W. STIMPSON (OBSERVER)Deputy Assistant Secretary

Office of Policy Development and Research, HUDWashington, D.C.

LAUkA TAVORMINAProgram Officer

Local Initiatives Support CorporationNew York

JUDITH WALKERCommissioner

Department of Human ServicesChicago, Illinois

BARRY ZIGASPresident

National Low Income Housing Coalition&

Executive SecretaryLow Income Housing Information Service

Washington, D.C.

Consultant

DR. ANN B. SCHNAREVicE-PresidentICF Incorporated

Council Staff

SARAH BECKERDirector

Council ProgramsThe National Alliance to End Homelessness

The Alliance wishes to thank Dr. Ann B. Schnare andKathleen G. Heintz of ICF Incorporated for their importantcontribution in the development of a methodology for thenational estimate and their work in drafting the report.

CONTENTS

EXECUTIVE SUMMARY 1

INTRODUCTION 8

I. ESTIMATES OF THE SIZE OF THE HOMELESS POPULATION . . II

Evaluating the Accuracy of the Local Estimates, 15Constructing National Estimates, 18

Extrapolation Using Street-to-Shelter Ratios:How Reliable are They?, 18

Extrapolations From Local EstimatesChoosing the Population Base, 19

Revising the Estimates, 22Growth of the Homeless Population, 23Homelessness Over the Course of a Year, 24Characteristics, 25

Families, 27Mentally Ill, 29Veterans, 29Youth, 30Migrants, 30Disabled, 30Other Factors, 30

Sex, 30Age, 31Race, 31Income, 31Duration, 32Rural, 32Substance Abuse, 33

Summary, 33

II. HOUSING IHE HOMELESS 35

Housing: Federal Issues, 37Public Housing, 38Preservation of Existing, Subsidized Stock, 38New Construction, 40Reassessing Welfare Allotments, 40

Housing: State and Local Issues, 41SROs, 41Zaning, 43Design and Regulatory Charges, 43

Housing Providers, 44Role of States and Financing Agencies, 44Role of Nonprofits, 45Role of the Private Sector, 45

Service for Tenants, 46Eviction Assistance, 47Housing Counseling, 47

Special Needs of the Homeless, 48Benefits, 49

APPENDIXES 51

A. Estimates of the Homeless Population by City, 51B. Characteristics of the Homeless: Results of Local, 55

StudiesC. Reference:, 59

LIST OF TABLES

1 National Estimates of the Homel-ss Population, 142 Comparison of Estimates: Local Experts, HUD, and

the Results of Counts, 173 Variation in Street-to-Shelter Ratios, 204 Length of Time Homeless, 26

1

EXECUTIVE SUMMARY

Homelessness in America is a severe, persistent, andgrowing problem. While the size of the homeless poWationis difficult to estimate, as many as 736,000 persons couldbe hoseless on a given night in 1988, and considerably morecould experience homelessness over the course of the year.The roots of homelessness are complex and only partiallyunderstood. However, they appear to reflect the confluenceof a number of different 'orces: the deinstitutionalization(and noninstitutionalization) of the chronically mentallyill; continued high unemployment in some parts of thecountry and in certain industries; the rise in the number ofpoor, female-headed households; cutbacks in publicassistance benefits; and declines in the supply ofaffordable housing, as well as federal funding for thecreation of such housing. While the high cost of housing isonly one of the factors behind the phenomenon ofhomelessness, it has produced a situation in which thosewith the least ability to compete are allowed to drop out ofthe nousing market entirely.

This report focuses on the housing needs of thehomeless, specifically options for rehousing thoseindividuals and families who are currently homeless andstrategies for preventing additional homelessness. Asbackgrind to this discussion, the National Alliance to EndHomelessness explores two of the major questions that mustbe answered before effective planning can begin, namely:How many homeless are there? What are their characteristicsand needs?

The limitations of such an effort must be emphasizedfrom the start. Given the available data, definitiveanswers simply do not exist. Current studies can becriticized on a number of methodological grounds, and themajority are local, not national, in scope. Particularlywhere numbers are concerned, definitional problems arise.For example, most estimates focus on those who sleep inshelters, in places such as park, or abandoned buildings, orin rented hotel and motel rooms paid for by public orprivate agencies. However, other groups might also beconsidered homeless: those who are doubled-up with friendsor relatives, those who are living in short-term rentals, orthose who are in institutional care and have no address. Inaddition, one might identify a large number of persons who

t

2

are "at risk" of homelessness, including those who sufferfrom the types of chronic problems observed among thecurrent homeless, as well as those who are living in housingunits that are so costly (relative to their incomes) thatthe situation must be described as tenuous at best.

Assessing the characteristics of the homeless is noeasier than pinpointing their numbers. Although many localstudies provide extensive data on homeless persons, thedifferent populations they cover (for example, street vs.shelter) and the varying definitions used make patternsdifficult to ascertain. Furthermore, local studies showtremendous variation by city. This is not surprising sincea host of local factors could affect the size andcomposition of the population, including the availabilityand quality of mental health care or the nature of thehousing market itself. This variability does, however,argue for caution in attempting to arrive at a "nationalportrait" of the homeless; it also argues for permittingmaximum local flexibility in developing programs designed tomeet the needs of this diverse population.

Despite these problems, Housing and Homelessnessattempts to cull insights from the existing literature inorder to present a snapshot of the homeless that can be usedto help guide planning activities. Section 1 begins with adiscussion of the size of the population, followed by anidentification of major subgroups among the homeless.Section II then turns to a set of housing strategies thatcan be used to provide housing for those currently homelessand to reduce the flow of new households into the homelesspopulation.

COUNTING THE HOMELESS

No one really knows how many people are homeless today.Estimates differ widely at both the local and nationallevel, and no single set of estimates is clearly superior tothe rest. After reviewing the existing evidence, theAlliance estimates that as many as 736,000 persons may hehomeless on a given night, and that between 1.3 million and2 million different individuals way experience homelessnessat some point during the year,

This estimate of the number of homeless on a givennight is based on data originally assembled by the U.S.Department of Housing and Urban Development (HUD), butinterpreted in a different light. While the HUD studyattempted to collect data for metropolitan areas--includingcentral cities and their surrounding suburbs--most citieshave charged that the HUD figures represent city, notmetropolitan estimates. The Alliance has accepted thiscriticism and has adjusted the data accordingly.

3

However, unlike other studies, the Alliance does notassume that the rate of homelessness in the cities is equalto that in the suburbs. Instead, data is used from theWashington, D.C. area (the only data available) to estimatea city-suburban rate differential, which is applied to anational total. The Alliance estimates that the rate ofhomelessness in the suburbs (9/10,000) is one-third of thecity rate (28110,000). This approach produces a total of355,000 homeless individuals as of 1934.

Over the intervening four years, homelessness hascertainly increased. However, no reliable data exist on theactual rate of growth. For example, if the homelesspopulation has been growing at a rate close to that of theoverall population (that is about 1 percent per year) thiswould yield a 1988 population of under 400,000 persons. Onthe other hand, local officials have estimated that thedemand for shelter has been increasing at a rate of rot'ghly20 percent per year. Applying this growth rate to the 1984estimate would yield a current population of 736,000. Whflethe truth is probably somewhere in-between, the 736,000figure is a reasonable upperbound estimate of the number ofhomeless on a given night as of the present time.

Nightly counts are appropriate for planning sheltc.beds. However, in planning permanent solutions, it is moreappropriate to consider the number who become homeless atsome time during the year. if people are homeless for arelatively short period of time, nightly and annual countscan differ significantly. The Alliance used therelationship between "nightly" and "annual" counts that wasderived in a study of Chicago's homeless to arrive at anational estimate of the number of individuals likely to behomeless at some time during the year. Using a 1988 base of736,000, this procedure produces an upperbound estimate ofbetween 1.3 million and 2 million persons who could behomeless at some time during the year.

CHARA,7ERISTICS OF THE HOMELESS

From the broadest perspective, the homeless populationcan be divided into two groups: those who are members offamilies, and those who are homeless alone (individuals).The problem of family homelessness has received a great dealof attention in recent years, and families are the fastestgrowing segment of the homeless population. P review ofexisting studies suggests that about 25 to 33 percent of thehomeless population is made up of persons in families.

Family homelessness includes both intact families andfemale-headed households; however, the latter grouppredominates. :ome of these families have experienced

4

homelessness as a result of a particular crisis (fire,evitAion, divorce, domestic violence), suggesting that theirhomeless condition may be of relatively short duration.However, there is also evidence of families who are "multi-problemed" and for whom Namelessness is a relatively chroniccondition. There is also strong evidence that one of thecontributing factors to increased family homelessness is thesize of the housing allotments provided to recipients of Aidto Families with Dependent Children (AFDC). Nationally,'iese allotments average only 49 percent of the cost of a

.Jdest unit, and in some parts of the country, they are aslow as 29 percent.

Among homeless individuals--who make up the majority(67 to 75 percent) of the population--several majorsubgroups can be identified. These include the mentally illand alcohol or substance abusers. The Alliance estimatesthat each of these two subgroups account: for roughly one-third of the single homeless population, or 25 percent ofthe homeless population as a whole. Although some familymembers also fall into these groups, their proportions arerelatively low.

The mentally ill are a diverse group, but generallyexhibit a diminished capacity for self-care, interpersonalrelationships, and work or schooling, and typically requirelong-term mental health care. The large number (25 percent)of mentally ill persons among the homeless population wasoriginally attributed to deinstitutionalization policies ofthe 1960s, but the age profile of the homeless (median age35) suggests that noninstitutionalization and inadequatefunding for community programs are also at work. Littledata link mental health status to length of homelessness,but one might assume that the mentally ill figureprominently among the longer-term homeless.

The second group accounting for substantial numbers(also 25 percent) includes alcoholics and drug abusers.This condition affects men proportionately more than women,and older men show somewhat higher rates than younger men.Substance abuse might be considered a precipitating factorin homelessness for these individuals, and is clearly acondition that must be addressed as part of any rehousingstrategy. Substantial overlap :s also likely betweensubstance abuse and other barriers to employment (age,education, work experience, criminal convictions' i-Lt willhave to be considered as well.

Finally, one must address the remaining one-fourth ofthe population for whom no single classification applies.Presumably this includes both employable individuals (the"economic" homeless) with establisaed work histories, andthose for whom some form of training, education, or otherservice program is a prerequisite to employment and self-

14

. I 4

5

sufficiency. A variety of other subgroups are among thehomeless who overlap with those described above. Theseinclude the elderly (a relatively small segment), thephysically disabled (also small), and veterans. The lattergroup may account for as much as one-half of the single malepopulation.

HOUSING STRATEGIES FOR THE HOMELESS

Housing strategies for the homeless include programs orpolicies to prevent new homelessness and efforts to rehousethose who are currently homeless. The former areappropriate elements of an overall national housing policyand focus on strategies to reactivate the role of thefederal government, to expand the supply of available low-income units, and to intervene in situations of imminenthomelessness. The recommendations for housing the homelessformulated by the National Alliance to End Homelessness arepresented in Section II. The following is a brief overview.

Expanding the Supply of Low-Income Housing. Thecurrent supply of low-income housing-falls short of demand.The Alliance estimates that about 5.5 million units would berequired to meet this cap. However, HUD housing programscurrently focus on the provision of tenant-based subsidiesthat enhance a family's purchasing power, but do little toincrease the supply of new units. Since supply problems maybe localized in nature, it is ap,ropriate that state andcity governments are becoming more important actors in theconstruction and rehabilitation of lower-income housing.Nevertheless, supply issues need to be addressed at thenational level, both through the provision of funding andthrough public and private efforts to better meet this need.

Preserving the Existing Subsidized Stock.Currently, about 4 million households live in unitssubsidized by HUD, about one-half of these (2 million) inprivately owned units developed under a variety of newconstruction and rehabilitation programs. It has beenestimated that as *lany as 1.4 million of these units may beremoved from the stock over the next 20 years, due toprepayment provisions in older Federal Home Administration(FHA) mortgages, opt-out provisions in early Section 8contracts, normal subsidy expirations, and losses due tophysical or financial collapse. Both public and privategroups are working to develop progress to preserve thisvaluable stock and prevent the high levels of displacementthat would result from its loss.

Reassessing Welfare Allotments. The large andincreasing number of homeless families who are current AFDCrecipients is, in part, due to the fact that the housing

6

allotments provided under this and other income maintenanceprograms fall far short of the amounts actually needed topurchase housing. Just under one-half of all publiclyassisted families and individuals (AFDC, SupplementalSecurity Income, and General Assistance) currently pay morethan 50 percent of their incomes for housing. While therelationship between welfare allowances and housingsubsidies is complex, efforts are needed to rationalize thetwo approaches to meeting housing costs and to provide sumsthat better reflect the actual costs of today's housingmarket.

Single-Room Occupancy. One component of the stockwhich merits particular attention is single-room occupancy(SRO) units. Once considered substandard housing, SROs haveshown themselves to be an important alternative forindividuals with extremely low incomes, and several majorcities have undertaken programs to upgrade and preserve thesupply of these units. Additional family housing (typicallylarger units) and residences for the mentally ill are alsoneeded, as demonstrated by the large proportions of thehomeless that fall into these two groups.

Design and Regulatory Issues. Continued effortsueed to be directed to reducing housing costs throughinnovative construction and design and through areassessment of local policies that may tend to increasehousing cots unnecessarily. The latter include localrehabilitation standards and zoning regulations.

Public, Private, and Nonprofit Providers. To theextent that supply problems are iocalizeU, it is appropriatethat states and localities initiate action on their ownbehalf. States and localities have responded by adopting amore aggressive role in the development of supply-orientedprograms, including low-income housing funds to finance ,

construction and rehabilitation. Such funding, however,must be 04mented by increased federal appropriations forlow-income housing.

The primary providers of low-income housing are publichousing agencies, nonprofit organizations (includingcommunity development corporations), and the private sector.Given existing conditions, including changes in the 1986 TaxReform Act, it is unlikely that for-profit developers willfind the present rate of return on investment sufficient toengage in large-scale activities on behalf of the homeless.At present, nonprofit organizations are the predominantproviders of housing to the very low-income homelesspopulation. However, if America is to meet its nationalhousing goals, private sector involvement is required. Aneffective alliance must be created that includes all housingproviders. Such an alliance should result in well-targeted

7

allocations to the poorest of A:lerica's poor, the homeless,on a scale commensurate with their number.

Counseling and 'Issistance to Prevent IiminentHomelessness. programs need to be developed toassist families and individuals on the brink ofholaelessness. measured against the costs of homelessness,such interventions might be well worth the expense. Theyinclude: mediation of landlord tenant disputes that mightotherwise lead to eviction; negotiation and counseling toprevent foreclosure; assistance in finding housing; andshort-term financial assistance to households that are indanger of losing their current housing or cannot afford thetransaction costs of obtaining new housing.

Taking all factors together, the phenomenon ofhomelessness argues for a renewed national commitment to thegoal of ensuring safe, decent, and affordable housing to allAmericans. While the elements of such a policy will behotly debated, the problem of homelessness will not beresolved without maximum public and private commitment tomeeting the nation's housing needs.

1"

8

INTRODUCTION

Homelessness is a problem of national scope and crisisproportion in the United States. The ranks of the homelessnot only include the traditionally recognized urban bag ladyand skid row alcoholic, but increasingly reflect all ofAmerica's poor. Homeless individuals are young and old,able and disabled, employed and unemployed, mentally ill andhealthy. They are individuals, families, veterans, runawayyouth, and teenage mothers. They are low-income familiesdisplaced by escalating housing costs and countless otherswho find themselves in need, but without family or communitysupport. They live not only in urban areas, but can befound in newly impoverished rural farming communities andwealthy suburbs.

How many homeless people are there? Estimates of thesize of this population vary greatly. The National Allianceto End Homelessness estimates that on any given night in1988, 736,000 people will be homeless. Over the course of1988, the Alliance estimates as many as 2 million peoplewill be homeless. The homeless population is increasing ata significant rate per year, with the highest rate ofincrease being among families. Some groups, including theU.S. Conference of Mayors, estimate that the rate ofincrease exceeds 20 percent per year.

What his caused this crisis, particularly in a time ofnational Pc_ .omic prosperity? The causes are many.

Fede al Cutbacks. Since 1980 the budget authorityfor federal housing programs has been cut by 75 percent,from $30 billion in fiscal year (FY) 1981 to $7 billion inFY 1987 (1). In addition, the eligibility and payments;:andards for Aid to Families with Dependent Children(AFDC), Food Stamps, disability, and other federal supportprograms have been tightened. Thus, an increasing number oflow-income and poor people are forced to compete on the openhousing market with reduced disposable incomes. Those whoare not successful become homeless.

Urban Revitalization and Displacement. The inner-city redevelopment of the 1970s and F980s brought manybenefits, but it also left a wake of displaced poor. Up to

1) Barry Zigas, "Homelessness and the Low Income HousingCrisis," National Low Income Housing Coalition, Washington,D.C., 1987, p. 1.

18

50 percent of the nations stock of single-room occupancy(SRO) housing was destroyed as a result of this process.Currently, SOOIC 2.5 million people are displaced from theirhomes each year, most entering a spiral of increasinghousing costs, overcrowding, and worsening housingconditions (2). Between 1970 and 1980 the number of housingunits affordable to very low-income households (thoseearning $5,000 per year or less) shrank from 15 million to 3million, leaving only half as many units as fam:lies (3).Today, the shortage of affordable units may be as large as5.5 million (4).

Deinstitutionalization. In the 1950s more than550,000 ,6,117FrirKETTT6BeTOTATid in state mental hospitals inany given year. Today that number totals fewer than150,000. The community-based mental health system that wasplanned to support these patients was never fully funded.As a result, thousands of low-income, mentally ill peoplehave found themselves on the streets.

Destabilization of the Family. The 1980s have seenan expansion-in the nuinber of single-parent households, withreduced economic means, seeking increasingly unavailablehousing. The tragic fallout from drug and alcohol addictionand the increase in runaway youth have also made theircontribution to the ranks of the homeless.

Economic Changes. From 1960 to 1980 the nation lost20 percent of its solidly middle-income manufacturing jobs(5). These have been replaced largely by much lower-payingservice positions of little security and usually withoutbenefits such as health insurance. The resulting reductionin disposable income has led to the disillusioningphenomenon of employed homeless people.

Together, these and other factors have resulted in awidening gap between the funds people have to pay forhousing and the cost of that housing. Based on data from

2) Chester Hartman, "The Housing Part of the HomelessnessProblem," in Homelessness: Critical Issues for Policy andPractice, Boston Foundation, 1587.

3) Zigas, E. cit.

4) Sandra Newman and Ann Schnare, Subs dizing Shelter: TheRelationship Between Welfare and Hoiiiik,_Assistance, ReportNo. 1, The Urban institute Press, Illash-i6-41on, P.C.,May 1988.

5) Mark Griuker, "The Causes of Modern Homelessness," FoodMonitor Winter:1988.

19

9

the 1983 Annual Housing Survey, the median percentage ofincome spent on rent is 60 percent for renters with incomeunder $5,000; it is 55 percent for those earning less than$7,000; and 39 percent for those earning less than $10,000(6). In federal housing assistance programs the percentageof income to be spent on rent is targeted at 30 percent.Once a family or individual falls out of the housing marketthrough eviction, disaster, or domestic upheaval, they willhave great difficulty in finding affordable replacementhousing. If families do find replar-21ent housing, the firstand last months' rent, plus the reqred security midutility deposits, are likely to be greater than tP2resources they have available. All too often the result ishomelessness.

Section I of this report examines the size andcharacteristi.:s of the homeless population to assess morerealisticall/ its housing needs. Section II discussesrecommendations on housing the homeless. It begins withstrategies to prevent homeles.ness, then examines the dualissues of housing supply and affordability, and finallylooks at a feu of the hcusing and housing-related needsspecific to the homeless population.

6) Hartman, op. cit.

20

10

I

11

ESTIMATES OF THE SIZE OF THEHOMELESS PURULATITN

Planning strategies to help the homeless is complicatedby a lack of information about the population and thedifficulty of sorting out factors thought to causehomelessness. InnJmerable commissions and task forces atall levels of government have grappled with essentially thesame questions:

How many homeless people are there?

Who are they?

Why are they homeless?

What needs to be done?

Although no one has arrived at satisfactory answers,each effort has contributed something. This section drawsfrom efforts to construct, as accurately as possible, anassessment of the size of the homeless population.

Limitations on this assessment must be recognized fromthe start. Given the nature of the available data,definitive answers to the many questions surrounding thehomeless issue simply do not exist. The numerous studiescompleted thus far can be (and have been) criticized on anumber of methodological grounds. The majority are local,not national, in scope. Nevertheless, an objective look atthe available evidence does suggest a number of broadconclusions about the magnitude and nature of the problem.

It is also important to recognize that the size of thehomeless population depends in large part on how one defineshomelessness: Who is homeless (7)? Obviously, those who

7) Interview with Cynthia Taeuber, Population Division,Census Bureau, Suitland, Maryland, March 2, 1988. Even theCensus Bureau recognizes the difficulties faced in trying to

21

12

sleep in shelters, or in places such as abandoned buildingsor parks, are homeless. Most studies also include familiesstaying in welfare hotels. However, three other groups canbe identified:

Those who are "doubled-up" with friends andrelatives. Willie some may five this way by choice, for a1WWTFIWiTy people, doubling-up is a last ditch effort to

avoid living on the street.

Those living in short-term rental accommodations.Many poor people live in SROs renterby the week or day.When their money runs out, they sleep on the streets or inshelters.

Those in institutional care. Many of those inhospitalrililnalsouoiries to which they canreturn when released. Indeed, police stations and hospitalsoften function as de facto shelters.

All three groups are similar to the more easilyidentified homeless population; and many individuals movebetween categories (8). While these groups are excludedfrom most estimates of the size of the homeless population,they constitute at the very least an important segment ofthe "at risk" population, and must be considered in devising

arrive at an accurate assessment of the homeless population.Because the Census Bureau considers the definition ofhomelessness to be necessarily a political one, the 1990decennial census count will be limited to certain componentsof the homeless population, primarily the street and shelterpopulations. The count will be conducted on one night only,March 20, 1990. Demographic data will be collected on allpopulations; socioeconomic data on the shelter populationonly. Because of tne limited search parameters,particularly on the street count, it is certain that thefinal count will underrepresent the numbers of the homeless.

8) In Chicago, about one-third of the homeless had spentsome time the previous week doubled-up or in a rented room.Travelers Aid data, from eight cities, indicate that 27percent of families with children who seek shelterassistance report spending the previous night with familiesand friends. One percent had spent the previous night in ahotel, motel, or boarding house. See Dr. Penelope Maza andDr. Judy Hall, Study of Homeless Families: PreliminaryFindings, Travelers Aid International, Washington, D.C.,1987, p. 2.

2°

13

long-term strategies to address the problems ofhomelessness.

While no one knows how many homeless people there arein the United States, The National Alliance to EndHomelessness is concerned that a reasonable assessment ofsize is needed for long-term planning. Offered here is amethodology that the Alliance believes conforms to theperceived size of the homeless population, but cautions thatno estimate is entirely accurate. An examination of thecausal factors involved in homelessness leads the Allianceto conclude that the number of people who are on the brinkof homelessness, homeless, or cyclically homelessconstitutes a tragedy far in excess of what any populationfigure might indicate.

Some studies have produced widely varying estimates ofthe size of the homeless population, depending on the dataand assumptions used (see Table 1). Other estimates rangefrom the low figure of 200,000 cited by the U.S. Departmentof Housing and Urban Development (HUD) (9) to the highfigure of 3 million used by the National Coalition for theHomeless. Two more recent studies (10) suggest nationaltotals higher than the HUD estimate, but lower than theCoalition figure. Both of these relied on techniques- -and,for the most part, data sources--originally developed in theHUD report. However, by varying one or two key assumptions,different conclusions were reached.

The national figures presented in Table 1 are based onlocal estimates that have been extrapolated to the nationallevel. As described in more detail below, local estimatesvary significantly, depending on their underlying source andmethodology. Furthermore, key assumptions vary as differentresearchers extrapolate from local to national data. Thus,in assessing the validity of the national estimates, onemust consider both the reliability of the local figures andthe extrapolation procedures used.

9) U.S. Department of Housing and Urban Development, "ARepnrt to the (2ecretary on the Homeless and EmergencyShelters," Washington, D.C., 1984.

10) Friedman, "Permanent Homelessness in America?," WorkingPaper No. 2013, National Bureau of Economic Research,Cambridge, Massachusetts, 1986; and William Tucker, "WhereDo the Homeless Come From ?," National Review September25:1987.

2

14

T'BLE 1 National Estimates of the Homeless Population

Source Estimate Assumptions Used

Community forCreative Nonviolence190

U.S. Department ofHousing and UrbanDevelopwent, 1984

Fried,Aan and Hall,19L7

Tucker, 1936

2,200,000

192,000

254,000

536,000

27,000

700,000

Based on a small numberof high local estialates.Apparently uses citypopulations to estimatea rate of homelessness.Applies constant rate ofhomelessness to theentire country.

Applies a street-to-shelter ratio toestilaates of thesheltered population.

Bascd on estimates for60 cities. Usesmetropolitan populationas the base. Calculatesrates separate forlarge, iJediu,a, andsHall areas.

Takes nighest localestimates. Usesmetropolitan populationas the base. AssuAes aconstant rate ofho,lielessness nationwide.

4plies a street-to-shelter ratio toest ijates of thesheltered population.

Based on estii.lates for50 cities. Uses citypopulations as the base.Allows rates to var., forlarce, mediu,o, and smallcities.

SOURCES: See footnotes 9 and 10

15

EVALUATING THE ACCURACY OF THE LOCAL ESTIMATES

Most local estimates produced to date have beenobtained through interviews with shelter providers,advocates of the homeless, city officials, and otherindividuals familiar with the local situation. Yet, even inthe same city, there may be considerable disagreementregarding the size of the homeless population, with high andlow estimates sometimes differing by a factor of 10. In LosAngeles, for example, estimates have ranged from 19,500 to39,000; in Dayton, from 75 to 1,000 (11). In some cases,this variation may reflect definitional differences (forexample, some estimates may include households that aredoubled-up, or may refer to annual and not nightly figures).However, in other cases the discrepancies reflect morefundamental differences in perception of the phenomenon.

The most systematic effort to collect and evaluatelocal estimates of the homeless population was conducted byHUD in 1984, and involved interviews with 3 to 12respondents in each of 60 metropolitan areas. Toincorporate widely varying estimates and arrive at areliable range for each community, the HUD study weightedthe estimates based on the respondent's knowledge of thesituation and the source of his or her information (12). Asa result, the HUD estimates fall between the high and lowfigures that were obtained from respondents in any area.

One of the major criticisms of the HUD approach wasthat the information it collected referred to an unknowngeographic area (13). HUD attempted to obtain informationon the number of homeless people in the entire metropolitanarea, including the central city, any satellite cities, andthe surrounding suburban ring. However, many respondentslater indicated that their estimates referred to the centralcity alone. Thus, in using the HUD figures, the Alliancehas assumed they reflect city, not metro, totals.

One way of evaluating the accuracy of the HUD estimatesis to compare them to independent "counts" of the shelteredand unsheltered homeless. Unfortunately, only a handful of

11) HUD, ok. cit.

12) Appendix A presents these data for each of the 60cities. The appendix also provides data for 15 additionalcities collected in 1986 and using procedures similar toHUD's.

13) Applebaum, "Testimony on a Report to the Secretary onthe Homeless and Emergency Shelters," Housing the Homeless,Erickson and Wilhelm, eds., 1986.

16

counts have been taken, and they differ in theirmethodologies. However, their totals are usually lower thanthose provided by local observers. In Chicago (14), forexample, a count of the homeless, in which interviewerssearched alleys, abandoned buildings, open hallways, parkedcars, and so on, produced a total of 2,344 homeless peoplebetween September and October 1985 and 2,020 betweenFebruary and March 1986 (15). This compares to the HUDestimate of 19,400 and 20,300, and a maximum local estimiof 250,000 (16). In the skid row area of Los Angeles, wtmeelocal informants had suggested a total of 12,000 to 20,000homeless people, a street and shelter survey came up with afigure of roughly 3,000 (17). In other cities (Pittsburgh,Boston, and Washington, D.C.), the discrepancies are not aslarge, although the counts tend to be lower than themajority of local estimates (see Table 2).

What would account for these discrepancies? Theunsheltered homeless are, for several reasons, verydifficult to count. Many homeless peopie maintain a goodappearance, and so are overlooked by casual observation.Furthermore, their sleeping places are often hidden due tofears of harassment or victimization. The Boston surveycautioned that "dead end alleys . . . abandoned buildings,parking garages and MBTA bus and train yards" were notinvestigated (18). Therefore, the counts should be regardedas minimum estimates and revised upward to allow for the

14) nussi et al., "The Condition of the Homeless inChicago," 0767

15) An additional 540 were estimated to be temporarilyhoused in hospitals, jails, and other settings.

16) Kondratas, "A Strategy for Helping America's Homeless,"1985.

17) Hamilton, Rabinovitz, and Alschuler, "Faces of Misery:The Skid Row Homeless in Los Angeles," 1987. The 3,000estimate does not include roughly 8,000 persons residing indowntown hotels, some of whom may receive vouchers.Approximately 2,000 individuals were located in missions,and 1,000 were counted among the street population. Otherestimates for the area (from various sources) are reportedin Ropers and Robertson, "Basic Shelter Research Project,"1985.

18) Boston Emergency Shrlter Commission, "The OctoberProject," 1983.

21.)

17

TABLE 2 Comparison of Estimates: LocalResults of Counts

Experts, HUD, and the

Lowa

LocalHighLocal HUD

b

City Estimate Estimate Estimate Counts

Boston 2,300 5,000 3,200 2,115Chicago 17,000 26,000 19,850 2,700Phoenix 300 1,500 1,075 2,477Pittsburgh 50 1,500 888 875Washington,

D.C. 3,000 10,500 4,700 2,562 4

a Low and high estimates as reported to HUD interviewers.

b Midpoint of "most reliable range."

Data for Boston, Phoenix, and Pittsburgh as reported by HUD.Chicago data from Peter Rossi et al., "The Condition of theHomeless of Chicago," University T Massachusetts at Amherst,1986. D.C. data from Frederic Robinson, "Homeless People in theNation's Capital," University of the District of Coludbia, 1985,

d Adjustments to the observed figure of 2,562 resulted in a"probabl" number of 4,347 and a "possible" number of 6,454.

18

"concealed.homeless." In the Washington, D.C. study (19) itis suggested that such adjustments would raise the"probable" homeless total by 70 percent and the "possible"total by 152 percent (20). There may also be a discrepancybetween those counts that reflect a short duration of time(usually a few days) and those that encompass a period ofmonths or a year.

Taking all evidence together, the Alliance feels thatthe HUD estimates provide a workable basis for developingnational estimates, assuming that they apply to the centralcities and not to the metropolitan areas. The HUDstatistics are typically higher than the number implied by(unadjusted) local counts, suggesting they are notunrealistically low. At the same time, they are based on anumber of estimates for each community, and thus serve tobalance the widely varying perspectives of different serviceproviders and other observers of the ler,' scene. While theHUD data can be criticized on several Is, noalternative, consistent data source exis,. that covers arange of geographic areas wide enough to produce reasonableestimates at the national level.

CONSTRUCTING NATIONAL ESTIMATES

To obtain an estimate for the United States as a whole,,! must extrapolate from the local data in some way. Most

07 the disagreement between the various national estimatesreflects the different assumptions and extrapolationprocedures used. Two basic approaches are possible:(1) extrapolations based on street-to-shelter ratios, and(2) extrapolations from city estimates. Many methodologicalproblems E.re involved in extrapolations from city estimates(for example, the cnoice of the "true" local number), butthe Alliance believes the street-to-shelter approach is evenless reliable.

Extrapolations Using Street-To-Shelter Ratios:How Reliable Are They?

The first extrapolation approach relies on two types ofdata:

19) Robinson, "Homeless People in the sPation s Capital,"1985.

20) HUD has concluded that the most reliable range of localestimates was between 3,000 and 6,400 for the entiremetropolitan area, while the "probable" figure TITFT7ed fromthe city count was 4,347 and the "possible" total 6,454.

19

s local estimates of the number of people in shelterson a given night, which are then extrapolated to thenational level; and

a local counts of the number of homeless on thestreets, which are used to estimate typical street-to-shelter ratios.

National estimates of the total number of sheltered andunsheltered homeless people are derived by estimating thenational shelter population and applying a street-to-shelterratio to arrive at an overall count.

Using this approach, HUD derived a national estimate of192,000 homeless people (Table 1). This was based on anaverage nightly shelter population of 69,000 in January 1984and a street-to-shelter ratio of 1.79 to 1. The latter wasobtained by taking the average of the ratios found in threelocal counts. A second ratio-based estimate was prepared byFriedman and Hall for the Naticnal Bureau of EconomicResearch (NBER) in 1986 (21). Their figure of 279,000homeless people relies on an upwardly revised shelter count(22) and the use of a slightly higher ratio derived from a

series of interviews conducted in New York City in 1935.

The NBER report concludes that HUD was "essentiallycorrect" in its estimate of the homeless population size,however, the deficiencies of the ratio method are difficultto overlook. To illustrate the deficiencies, Table 3 showsthe street-to-shelter ratio arrived at through counts inseven cities. The ratios vary widely, by city, within city,and by season. Thus, it is difficult to accept the basicassumption of this approach--that a single ratio can beapplied nationwide.

Extrapolations from Local Estimates:Choosing the Population Base

A second, more reliable approach to estimating the sizeof the homeless population uses local estimates, whichinclude both sheltered and unsheltered homeless to calculate

21) Friedman, a. cit.

22) Unfortunately, _he only available data on the number ofhomeless people sheltered nationally, at a single point intime, are those obtained by HUD. Thus, they are the onlystarting point for national estimates using the street-to-shelter ratio.

20

TABLE 3 Variation in Street-to-Shelter Ratios

City Date of Study Ratioa

BostonChicago

Nashville(downtown area)

New York City (NBER)PhoenixPittsburghWashington, D.C.

October 1983September 1985February 1986December 1983March 1984June 1984September 1984June 1985March 19, 1983June 19, 1983June 1985

1.291.44

b

0.35,0.49"0.690.780.772.232.731.3%.2.49'

a Number of homeless on the street for each homeless person foundin a shelter. Boston, Pittsburgh, and Phoenix as reported byHUD.

b Peter Rossi, Fisher, and Willis, "The Condition of the Homelessof Chicago," University of Massachusetts at Amherst, 1986.

c Ratio is based on mission to all other. Proportion of homelessactually outdoors would be smaller. See Weingarn, "Counting theHomeless," in American Demographics December:1985.

d Friedman and Hall, "Permanent Homelessness in America," WorkingPaper No. 2013, National Bureau of Economic Research, Cambridge,Massachusetts, 1986.

e Based on a study estimate that attempted to adjust for hiddenhomeless missed by the count. The ratio based on actual,observed homeless would be 0.39.

21

rates that can be projected to the national level. Most ofthe national estimates produced to date adopt this approach,although their results vary widely. While some of thisvariation can be attributed to differences in the underlyingestimates, much of it occurs because of explicit andimplicit assumptions about the variance of rates amongdifferent types of areas.

The high national estimates tend to assume that therates of homelessness found within the central cities arealso found in the suburbs and in small towns and ruralareas. This can be criticized on several grounds. First,data developed at the local level indicate that small- andmedium-sized cities have lower rates of homelessness than dolarger cities (23). Second, in the sole case where separatecity and suburban estimates are available, suburbs showdistinctly lower rates (24). Finally, even within thecentral city, homeless people tend to concentrate inparticular areas. For example, in the Washington, D.C.study, three-fourths of the homeless street population werefound in two of the city's wards. Such concentrationreflects the fact that shelters, soup kitchensr and otherfacilities for the homeless are found only in certain areas.Furthermore, the downtown ecology, with bus and trainstations, hot air grates, abandoned buildings, and so on,offers sleeping places that are unavailable in the suburbs.Thus, procedures that assume a constant national rate, orfail to account for variations between cities and theirsuburbs, are likely to greatly overstate the size of thehomeless population.

On the other hand, it is likely that HUD's primaryextrapolation procedure understates the national totalsignificantly. As described previously, in producing itsnational estimate of 254,000, HUD asked local experts toprovide estimates of the total population in themetropolitan area, and used this population base (RMA) (25)to calculate a rate of homelessness for each. Observers

23) See HUD, op. cit. Homeless rates were 13 per 10,000population in TErg6Wetro areas and 6.5 per 10,000 in smallmetro areas.

24) For example, estimates of the homeless population inand around Washington, D.C. work out to a rate of 41 to 100per 10,000 in the city itself (depending on the use ofobserved or adjusted figures) versus an average of 9 per10,000 in the suburban counties. See "D.C. Tries to KeepPace with Homeless Rate," 112_14211121a11.91.1.91I, November 15,1987.

25) Rand McNally Metropolitan Areas.

3L

22

later criticized this approach on the basis that theestimates represented the central cities, without thesuburban areas and other cities that are inclt.ded in theRMAs.

REVISING THE ESTIMATES

The Alliance has attempted to construct a revisedestimate of the national homeless population on a singlenight. Its procedure tries to avoid typical errors bycalculating the central city and suburban homeless totalsseparately. The underlying data for this estimate areHUD's, but the Alliance has assumed that figures referexclusively to the central city of each metropolitan area.The resulting estimate is about 350,000 homeless on anygiven night as of January 1984. This is believed to be areasonable figure, but the limitations of the data cannot beoveremphasized.

First, it is assumed that HUD's national metro figure(of 210,000) is correct as an estimate of the central cityhomeless population on a given night in 1984 (26). Toestablish a suburban homeless rate, data were used from theWashington, D.C. count (the only such data available), whichshowed that the homeless were concentrated in downtown areas(27). In those wards within the District having more"suburban" characteristics (that is, most of their area wasdevoted to residential use, and more than one-third of thedwellings were owner-occupied), the rate of homelessness wasone-xhird (32 percent) of that in the District as a whole.Therefore, the Alliance assumed that the national suburbanhomeless rate is about one-third of the central city rate.This yielded approximately 145,000 homeless people outsidethe central cities, including nonmetropolitan areas, for atotal urban homeless population of 355,000. Revised ratesfor the estimated homeless population are as follows:

Homeless in Central City: 210,000 at a rate of28/10,000.

Outside Central City: 145,0W1 at a rate of9/10,000.

Total: 355,000 at a rate of 15/10,000.

26) HUD's national estimate of 254,000 homeless was basedon 210,000 in metro areas and 44,000 in nonmetro areas.

27) Robinson, off. cit.

23

While this approach relies on information from a singlecity, it is the only source available that provides detailon the geographic distribution of the homeless amongdifferent types of areas. However, the Chicago studyconfirms the underlying notion that homeless individuals areconcentrated in certain parts of the city (28). Furtherverification comes from a study of the homeless in FairfaxCounty- -the only suburban count of which the Alliance isaward. The rate obtained here was 9.7 homeless per 10,000population, almost identical to the average suburban rateused in our estimates (29).

GROWTH OF THE HOMELESS POPULATION

The local estimates used to construct the Alliance'snational estimate of the homeless population date fromJanuary 1984. However, significant changes during the lastfour years have occurred in the size and the composition ofthe homeless population.

Unfortunately, little reliable information is availableon the growth of the homeless population. The U.S.Conference of Mayors, after conducting a survey of cityofficials, suggests that the demand for shelter has beenincreasing at a rate of about 20 percent per year (30).However, increased demand may simply reflect the fact that a

higher proportion of the homeless now seek shelters. InBoston, for example, where the homeless on the streets andin shelters were counted in 1983 and 1986, the total figurewas only slightly higher in 1986 (3.5 percent), but a much

28) The Chicago researchers asked local police andorganizations, such as the Coalition for the Homeless, toassist in classifying census blocks as to the probability oflocating homeless persons. Only 6 percent of the blockswere classified as having some significant ch nee of havinghomeless persons on them in the dead of night.

29) Goplerud, "Homelessness in Fairfax County," GeorgeMason University, 1987. The 9/10,000 Fairfax rate is closeto one-third of the D.C. city rate (41/10,000) usingunadjusted figures from the D.C. study. It is one-tenth ofthe rate (100/10,000) obtained after adjusting for"concealed" homeless within city boundaries.

30) Shelter demand was reported to have increased by anaverage of 20 percent in 1986 and by 21 percent in 1987.See U.S. Conference if Mayors, The Continued Growth ofHunger, Homelessness, and Poverty InAmerica's Cities, Washington, D.C., 1986 and 1987.

3J

24

higher proportion were in shelters (31). The rate ofincrease implied by the Boston numbers world be 1 percentper year, or approximately the national growth rate for theg'neral population.

Given the limitations of available data, it isdifficult to verify the reliability of either of the twoalternative 1988 estimates--368,000 (which assumes anincrease of approximately 1 percent per year based onoverall population trends) versus 736,000 (which assumes ananneal increase of 20 percent). However, given widespreadconsensus that substantial growth has occurred in the sizeof the homeless population, the higher (736,000) figureprovides an upperbound estimate of the maximum number ofhomeless people at a given time as of early 1988.

The studies noted above also suggest important shiftsin the composition of the homeless population. In Boston,for example, the number of children increased from 138 to376 over the period of the study, or by about 40 percent peryear. Shelter providers surveyed by the U.S. Conference ofMayors also consistently report that families are thefastest growing group among the homeless population. Onaverage, local officials reported that family membersconstituted 28 percent of the homeless population in 1986and 33 percent by 1987 (32).

HOMELESSNESS OVER THE COURSE OF A YEAR

As noted earlier, estimates of the homeless populationrefer to a given point in time. Of perhaps greaterrelevance for many policy decisions is the total number ofpersons who are homeless over the course of a year. If

homelessness is a short experience for most, then the totalnumber of people homeless over the course of a year will bemany times greater than the number of homeless people at anypoint in time. For example, if every homeless person warhomeless for just 1 month, the annual total would be 12times greater than the count for any particular day.

31) City of Boston, "Making Room: Comprehensive Policy forthe Homeless," 1986. Of 2,767 homeless in 1983, 57 percentwere in shelters, as compared with 76 percent of 2,863 in1986. The study cautions that this does not necessarilymean that the total number of homeless in Boston hasstabilized during the three-year period.

32) U.S. Conference of Mayors, 22. cit.

25

In Chicago, researchers asked both street and sheltergroups how long they had been homeless. On the basis ofthis information, they estimated:

approximately 2,200 persons were homeless at any onepoint in time;

between 3,000 and 3,800 additional persons becamehomeless over the year; and

a total of between 4,000 and 6,000 persons werehomeless sometime durinrj the year.

The Chicago estimates make a number of simplifyingassumptions that tend to overstate the annual rate. Forexample, they assume that "new" entrants become homelessonly one time during the year--an assumption that is clearlyincorrect, given the observed tendency of at least somesegments of the population to move in and ou' ofhomelessness at relatively frequent intervals. However, asshown in Table 4, the available studies suggest that Chicagohas a smaller proportion of short-term homeless people thanmany other cities; and the greater the size of the short-term homeless group, the higher the annual figure. Thus theChicago overstatements may be largely self-canceling.Applying the Chicago results to a national nightly estimateof 736,000 yields an upperbound estimated annual total ofbetween 1.3 million and 2 million homeless people.

CHARACTERISTICS

A discussion is warranted of the homeless populationsubgroups and some of the demographic characteristics thatcut across the population as a whole, but a few words ofcaution are necessary. An examination of subgroups seeksonly to describe some of the broadest trends in thepopulation in order to identify particular housing needs.It is not a national profile, and in any given locality onewould expect to find a population greatly at variance withwhat is described herein. Also, these subgroupings aredescriptive only. In absolutely no sense are the factorsdiscussed to be considered causal. The causes ofhomelessness cannot be sought in the characteristics of thehomeless.

Two basic sources of information reveal characteristicsof the homeless. The first consists of local studies inwhich homeless individuals are identified and interviewed.The second source of data about the homeless population isthat of shelter providers and other key informants.

26

TABLE 4 Length of Time Homeless, by percentage

StudyaLess ThanOne Month

More ThanOne Month

Austin 11 47Boston 9 60Chicago 12 39Las Vegas 42 20Los Angeles 23 49Minneapolis 26 24New York 21 56Ohio 40 23Phoenix 31 27

aStatistics for places other than Austin, Chicago, and Ohio were

compiled by HUD. (See "Report to the Secretary on the Homelessand Emergency Shelters.") Austin data are from Baumann et al.,"The Austin Homeless," Chicago data from Rossi et al., "TheCondition of the Homeless of Chicago;" and Ohio 3aTT from Rothet al., "Homelessness in Ohio: A Study of People in Need."

The problems of compiling a description based on thesesources are substantial. Shelter provider estimates aretypically based on those homeless who use services; clearlythere are some homeless who avoid all contact with sheltersor other sources of help. While interview-based studiesoften include respondents drawn from the streets, souplines, welfare hotels, or other places that would provide a

more broad-based picture of the population, coverage (aswell as the operating definition of homelessness) variesfrom study to study.

For example, two recent studies (Chicago and LosAngeles Skid Row) (33) provide maximum coverage of streetand shelter populations, though neither includes residentialhotels and one is limited to a small portion of the downtownarea. Other studies cover shelters only (Seattle) andstreets only (Austin), and one (Ohio) includes householdstemporarily residing with families or friends (34). As aresult of the different populations surveyed, true patternsare difficult to ascertain.

Despite these problems, local homeless studies doprovide a starting point for identifying significantsubcategories of the homeless population. This kind ofinformation allows us to begin to design appropriateintervention and prevention strategies. The followingdiscussion provides a brief overview of some subgroups ofthe homeless population, followed by a louk at the generaldemographics of the homeless population. The latter isbased on 13 local, survey-based studies and 5 studies thatrely on provider information. (These data are presented inAppendix B.)

Families

Based on available information, estimates for the shareof homeless individuals in families rang. from 25 to 33percent. Estimates from some shelter providers areconsiderably higher, but families tend to be found among theshelter population rather than on the street. It alsoshould be noted that estimates of the percentage of homeless

33) Rossi et al., "fhe Condition of the Homeless ofChicago," 138.6; and Hamilton, Rabinovitz, and Alschuler,"Faces of Misery: The Skid Row Homeless in Los Angeles,"1987,

34) Bauman et al., "The Austin Homeless," 1985; KingCounty, WasWTn§Ton, "Emergency Shelter Study," 1984; Rothet al., "Homelessness in Ohio: A Study of People in Need,r7837

90

it

27

28

people in families vary considerably across cities, withproportions as high as 50 to 65 percent reported in citiessuch as Providence, Trenton, and New York City (35).Averaging provider estimates across cities, the U.S.Conference of Mayors arrived at a proportion of 28 percentin 1986 and 33 percent in 1987. Service providers almostinvariably report that families are the most rapidly growingsegment of the homeless population.

Initial reports of family homelessness focused on thedepression-era image of laid-off workers traveling withtheir families in search of a new start. The intact,middle-class household suddenly thrust into homelessness byeconomic circumstances dominated early media reports, andproviders continue to identify the "new poor" as animportant segment of their service populations.

The profile of family homelessness offered by theChicago study is of a relatively small segment (12 percent)of the population, predominantly black and found primarilyin shelters. These women and their children appear to havefled from untenable living arrangements to strike out ontheir own. According to the study, these women typicallyremain homeless for a short period of time (36).

On the other hand, a study of residents of a New YorkCity welfare hotel (37) indicates that family homelessnessmay be of longer duration. Hcre, primary reasons forcurrent homelessness are eviction (48 percent, includingeviction by a primary tenant) and displacement due tocondemnation or fire (40 percent). Eighty-seven percentreport public assistance as their main source of income.Just over half of the families have lived in a hotel for sixor more months, and 10 percent have lived in the hotel for

35) U.S. Conference of Mayors, ok. cit.

36) It is important to note that the Chicago study excludedresidential hotels and other places where additionalhomeless families might be found. Moreover, a broaderdefinition of homelessness could result in greater numbersin this group. For example, Ohio included in its survey asegment of the population defined as "resource people" whowere located in cheap hotels or found temporarily stayingwith families or friends. The largest proportion of femaleswas found in this group.

37) Simpson et al., "Struggling to Survive in a WelfareHotel," 1984.

2V

more than a year. The average length of stay, at the timeof the study, was 11 months (and is closer now to 13 months)(38).

In Bostool Ellen Bassuk has profiled homeless mothersas follows: single, black, 29 years old, with 2.4 children,several years of high school education, and poor or no workhistories. Almost all (96 percent) are supported by AFDCand 41 percent have been on AFDC for more than 4 years.More important, about one-third have been chronicallyhomeless for the last 5 years, moving between relatives orfriends, welfare hotels, and other shelters.

Mentally Ill

Many studies have attempted to establish the proportionof the homeless population evidencing mental healthproblems. Although the attempts vary, the Allianceestimates that about one-fourth of the homeless populationsuffers from severe or chronic mental problems. Some datashow that the proportion of family members observablysuffering from mental illness is low (5 percent) whencompared to the single population (32 percent) (39).

Although the Alliance has examined no data that relatelength of homelessness to mental health status, we couldexpect this group to be counted among the relatively long-term homeless population.

Veterans

Based on local studies, a substantial proportion of thehomeless are veterans. Expressed as a proportion of thesample, the range is 22 to 46 percent for the studiesreported by Robertson (40). Expressed as a proportion ofthe male sample, homeless veterans account for between one-third and one-half in most studies reported. Where Vietnam-era veterans are distinguished, this group tends to account

38) Bassuk, "The Feminization of Homelessness," keynoteaddress given at The Shelter, Inc., Benefit, June 11, 1985,

39) New York State Department of Social Services,"Homelessness in New York State: A Report to the Governorand the Legislature," 1984.

40) Robertson, "Homeless Veterans: An Emerging Problem?"The Homeless in Contemporary Society, Bingham, et al., eds.,PM/.

30

for between one-fourth to one-third of all homelessveterans.

It appears that few homeless veterans receive anybenefits from the Veterans Administration (for example, 4percent in Los Angeles). Robertson found that veterans tendto have been homeless for shorter periods of time, areslightly older than other homeless people, and report higherrates of hospitalizat,on for psychiatric reasons.

Youth

Runaway youth constitute a small portion (4 percentaccording to the U.S. Conference of Mayors) (41) of thehomeless population and one that has very specific shelterand service requirements.

Migrants

Migrant families or individuals, particularly thosebetween jobs, constitute a small, as yet unknown portion ofthe homeless population.

Disabled

A small percentage of the homeless are physicallyhandicapped or disabled. While their number is not large,they do require certain architecturally modifiedaccommodations.

Other Factors

The following characteristics of the homeless popu-lation were derived from the studies cited in Appendix B.

Sex

The majority of the homeless are male. The highestproportion of men are found in skid-row populations (forexample, 84 percent in Nashville) (42) and on the street.By contrast, women are more likely to be shelter users or tobe housed in welfare hotels. Women are also much more

41) U.S. Conference of Mayors, a. cit.

42) Nashville data reported in Weingarn, "Counting theHomeless," American Demographics December,:1985.

31

likely to be part of family groupings, with only a verysmall proportion of the overall homeless population made upof single (that is, unaccompanied) women. Based on evidencefrom local studies, and HUD's 1984 shelter survey, it islikely that single males account for roughly two-thirds ofthe population, and single females account for between 10ard 15 percent.

Age

Overall, the studies consistently portray a populationthat is predominantly made up of persons in their middleyears. The median age is reported to be around 35 in moststudies. Although elderly people are among the homeless,they are underrepresented, possibly due to their entitlementto federal Supplemental Security Income (SSI), whichprovides them with some income.

Race

Blacks and minorities are strongly overrepresentedamong the homeless. While the minority composition variesby city, whites and Asians are typically underrepresented.At the same time, some studies (for example, Chicago) showsignificant racial differences by age, with a higherproportion of whites to be found among the older (over 40)homeless population.

Income

A substantial proportion of the homeless indicated theyhad worked for pay during the last month (39 percent inChicago) or had full- or part-time jobs (11 percent inRoston). The U.S. Conference of Mayors reports that, in the48 cities it surveyed, an average of 22 percent of thehomeless are employed in either full- or part-time jobs.Nevertheless, even when employed, the income of homelesspeople is very low.

Receipt of public assistance among the homeless is alsoreported to be low in most localities. One of the highesLrates reported is Chicago (35 percent). In Chicago, thedominant form of public assi tance is General Assistance(GA) (22 percent), folloveo by AFDC (6 percent),Supplemental Security Assistance (6 percent), anddisability. While the study showed that the homeless do

32

tend to apply for benefits (43), and treat rejection ratesare low, the programs are not providing long-term support.Of those who have at one point received assistance, onlyabout half are currently receiving benefits. In Chicago,the average income of homeless persons, from all sources,was found to be 5168 per month. Half of the homeless liveon $100 per month or less.

Duration

The studies show that slightly over half of allhomeless people have been homeless for less than one year.They also show that families tend to be found among theshort-term group. It is important also tc note that manyare beginning to draw more attention to the "cyclical"nature of homelessness among poor people, and to explain itis as a part of their "makeshift" adaptation to chronichousing shortages and low disposable incomes.

Rural

Most available studies of the homeless focus oh urbanpopulations. Relatively little data are available on ruralhomelessness, however, it is an equally serious problem ofgrowing dimensions. One of the few interview-based studiesto include rural populations is that conducted in Ohio (44).As compared to urban respondents, the nonurban homelesspopulation was found to be somewhat younger and to contain ahigher proportion of females and lower proportion of backs(45). The rural population also contained fewer long-termhomeless people (defined as being homeless for two or moreyears). Rural homeless people were less likely to be foundin shelters, and more likely to reside with friends orrelatives. Data analyzed by the Housing Assistance Council(based on surveys of rural community action agency directorsand other local experts) portrayed a population consistingof numerous subgroups including migrant workers, displacedrenters, bankrupt farmers, Indian reservation families, andlaid-off workers. As in urban areas, respondents indicated

43) The study estimated that 95 percent of those ostensiblyeligible for AFDC had applied and about 70 percent of thoseeligible for GA had done so.

44) Roth, "Homelessness in Ohio: A Study of People inNeed," Ohio Department of Mental Health, 1985.

45) In the Ohio study, rt-ll and mixed urban and ruralcounties are tynically coma ned for reporting purposes.

33

substantial increases in the size of the rural homelesspopulation, with families constituting a rapidly growingsegment (46).

Substance Abuse

An estimated one-fourth of the entire homelesspopulation may suffer from problems relating to substanceabuse. Based on the Chicago study, substance abuse affectsproportionately more men than women, occurring morefrequently among older men (44 percent) but also amongyounger men (35 percent). Many substance abusers live withrelatives or friends prior to becoming homeless, and itappears that the immediate cause of their homelessness isrejection by relatives or friends no longer willing toharbor them.

SUMMARY

The face of homelessness is the face of poverty inAmerica today. Homeless people live in families and alone;have jobs and are unemployed; are educated and uneducated.In spite of all these differences, they share two things:they are very poor and they are without housing.

Section II discusses the housing crisis that has causedhomelessness and offers strategies to help house thehomeless.

46) Housing Assistance Council, "The Homeless Crisis from aRural Perspective," forthcoming.

35

II

HOUSING THE HOMELESS

A variety of factors have contributed to the increasedincidence of homelessness. Two of the primary causes are:

increasing rents and the growing number Lfhouseholds unable to afford nonsubsidized rents, and

increasing imbalances between the demand for andsupply of low income-housing.

Homeless individuals, to be sure, have special needs,but the causes of homelessness do not relate to any one typeof social or physical condition. Rather, the actual causesof homelessness relate more to declining incomes, increasingrents, and the lack in number of low-income housing units.

Increases in real rents have exceeded the growth inreal income, and gross rent as a percentage of income is atits highest level in two decades. Only one in four renterhouseholds with incomes at or below the poverty level livein public or other subsidized housing, and there is evidencethat the growing rental payment burden has contributed tothe rise in homelessness (47).

These facts, combined with the imbalances between thedemand for and supply of low-income housing, more or lessexplain the homeless condition. What they do not explain isthe rationale behind certain changes in national housingpolicy, the severe reductions in federal spending forhousing, and the lack of appropriate response mechanisms.

47) The State of the Nation's Housing 1988, Joint Centerfor Housing Studies of Harvard University, Cambridge,Massachusetts, 1988, pp. 1, 10, 14. This study estimatesthat the growing rental burden contributed to the recentrise in homeless families with children. From 1974 to 1987the number of single-parent families nearly doubled. At thesame time, the median income of young single-parent renterhouseholds declined by 34 percent to $7,721; their rentburden increased from 34.9 to 58.4 percent.

44

36

What has stirred the public conscience, madehomelessness the critical issue of the 1980s, and movedlocalities to find creative solutions to the housingproblem? It is the knowledge that the number of low-incomerenter households seeking affordable housing has outstrippedthe theoretically available supply of units (48). Americafaces a shortage of affordable housing, and for some ofthose working in the field of low-income housing,homelessness has proven to he a "radicalizing experience"(49). At present, 63 percent of all poverty-levelhouseholds are living in rental housing. As real incomesdecline, those with the least ability to compete riskfalling out of the housing market entirely (50); they becomehomeless.

The lack of affordable housing is a problem of crisisproportion. The National Alliance to End Homelessness isone of several organizations to note this fact. Therecommendations offered in this report relate to housing thehomeless: those men, women, and children living in thestreets, shelters, welfare hotels, and also on the brink.The dimensions of the nation's housing crisis can onlybriefly be addressed in this report. A fuller understandingof the issues and an in-depth examination of policyalternatives are contained in recently released reports oftwo congressionally mandated task forces: the NationalHousing Task Force, chaired by James H. Rouse, and theNational Low Income Housing Preservation Commission, co-chaired by Carla A. Hills and Henry S. Reuss (51).

48) James W. Rouse, chairman and chief executive officer,The Enterprise Foundation, Columbia, Maryland, testimony ata joint hearinc of the Senate and House HousingSubcommittees, kpril 12, 1988, p. 3.

49) Interview with Barry Zigas, president, National LowIncome Housing Coalition, Washington, D.C., March 4, 1988.

50) Thirty-two million of the 241 million people living inthe United States in 1986 lived below the poverty line. In1986, one-fourth of all the full-time jobs (24 millionpositions) did not pay enough to raise a family above thepoverty line of $11,203. See A Decent Place to Live:Re_port of the National HousinL Task Force, Washington, D.C.,March 1986, p. 5.

51) See A Decent Place to Live: Report of the NationalHousing Task Force, pp. cit. See also, Preventing theDisappearance of Low-Tncome Housing: The Report of theNational Low Income Housing Preservation Commission,a h ing on, . pri .

4L

37

Following is a summary of some of the Alliance'sobservations on elements of the housing issue; specifically,funding sources, service providers, and tenant assistancemeasures.

HOUSING: FEDERAL ISSUES

The 1949 Housing Act established a national goal ofaffordable, decent, safe, sanitary housing for everyAmerican family. Yet, on any given night, 736,000individuals are homeless; at least as many, probably more,are doubled up. Six million others remain "at risk" ofhomelessness by virtue of their economic circumstance. Manydwellings are in substandard condition (52) and about halfof the inexpensive units are occupied by higher-incomehouseholds (53). Clearly, America has fallen short of itscommitment: new initiatives are needed as well as greaterefforts through existing programs.

The homeless by definition are without permanentresidence, and the magnitude of the housing problem requiresthat the federal government become more actively involved inits solution. The policy of HUD under the ReaganAdministration has been to respond to the housing crisiswith programs that focus almost exclusively on tenant-basedsubsidies (vouchers). While such programs alleviate some ofthe rent burden, they do little to create additional supplyat a time when availability falls short of demand. Anestimated 5.5 million new units of housing are needed tohouse the poor (54), but such housing will never be builtunless certain housing policies are either modified orreverse.

52) For example, Chicago estimates that one in sevenhousing units in Chicago is in substandard condition. SeeJudith Walker, "Local Housing Policies AffectingHomelessness," September 1987. It is estimated that,overall, 5 million renters and 4.5 million homeowners livein structurally inadequate, substandard housing. See TheState of the Nation's Housing 1988, op. cit., p. 16.

53) James W. Rouse, op. cit.

54) Sandra Newman and ,nn Schnare, Subsidizing Shelter:The Relationship Between Welfare and Rousing Assistance,Report go. 1, The Urban Institute Press, Washington, D.C.,1988.

38

Public Housing

RECOMMENDATION: The federal government should provideadequate funds for the maintenance andphysical rehabilitation oTf public housingas well as its expansion.

About 1.35 million units of public housing constitutethe nation's largest single holding of housing for the poor(55). Although subject to politicization, inadequate fundsfor ,aaintenance and capital improvement, and continuing badpublicity, public housing must be recognized as a nationalasset. There is simply no way to replace these housingunits for the poor, and the periodic efforts to sell orotherwise disiltantle public housing should be resisted. In

fashioning a sensible housing bill, priority should be givento preserving, upgrading, and expanding the nation's currentinventory of permanent, publicly owned, low-income housingunits. The Alliance believes renewed attention to publichousing is in orcer. Public ownership obviously eliminatesthe continued recycling of peivately owned housing andassures long-term use of properties for low-incoiae tenants.

Preservation of Existing, Subsidized Stock

':-Er,1::;ENDATI)NS: The number of housing units currentlyavailable underre-d-eral-projectassistance and other Dousing programsshould not be reduced.

The federal government should givepriority to the provision of thoseincentives that encourage owners ofassisted properties to extend their low-income use.

Where the current owner wishes to sell,the federal government should developfinancing mechanisms by which assistedprojects can be purchased and can FETWinaffordability. Local and stategovernalents also should make everyeffort to preserve assist-0-d stockeither through purchase or ny providingthe fina-ncing for private sector projectsthat preserve iow-income use.

55) The scope of this repo does not inchAe an in-depthanalysis of public housing issues.

39

Households displaced by the loss ofhousing that cannot be retained for low-income use should 5e assisted infinding alternative a fordable, comparableDousing.

Currently, about 4 million households benefitfrom housing subsidies provided by the U.S. Department ofHousing and Urban Development. These include about 1.3million households residing in public housing and another800,000 subsidized with Section 8 certificates and vouchers.The remainder--nearly 2 million households--live in unitsthat were built under a variety of constructior andrehabilitation schemes over the past three decades (56).These privately owned, federally subsidized projects accountfor nearly half of the currently assisted housing stock.Yet, up to three-fourths of these units could be lost forlow-income use over the next 20 years. These include:

e Units eligible for prepayment. These units weredeveloped under 14U6 and Farmers Home Administration (FmHA)mortgage subsidy programs, primarily in the late 19603 andearly 1970s. The mortgages, which typically were writtenfor a period of 40 years, contain provisions for prepayment(and termination of use restrictions) after 20 years. Up to645,000 units will become eligible for prepayment over thenext 20 years (57).

e Units eligible for opt-out or with subsidy contractsslated to expire. Most of these units were developed underthe new construction and substantial rehabilitationcomponents of the Section 3 program. Early contracts permitowners to terminate them after five years. Additionally,subsidy contracts for many units will' expire during thisperiod. Altogether, 774,000 units could be lost by 2005,assuming that all eligible owners opt out of their contractsand expiring subLiiies are net replaced or renewed (58).

56) Section 8 New /Rehab /flod Rehab; Section 236, Section 221Below-Market Interest at (BhIR); FwHA Section 215; andmarket rate FHA with rental subsidies.

57) Issues Facing the Assisted Housing Stock, prepared forthe National-tor-poration for Housing Partnerships, by theUrban Institute, Washington, D.C., p. 4.

52) Ibid.

Older FHA_Rrojects. The losses previously discussedprimarily reflect removals from the low-income stock basedon more profitable uses of the property. However, a segmentof the stock is more likely to be lost because of physicaldeterioration or financial problems. The expiration ofSection 8 subsidies could be an important contributingfactor to this loss.

Overall, the potential exists to lose as many as1.4 million units through conversion to higher-income use,expiration of existing subsidies, or physical and financialdeterioration over the next 20 years. Incentives toencourage owners of assisted propertiE.s to extend low-incomeuse are listed in Section 224 of the Housing and CommunityDevelopment Act of 1987. Among the incentives are:measures to increase the rate of return on investment;revisions to the method of calculating equity; inc,-easedaccess to residual receipts accounts or excess replacementreserves; provision of insurance for a second mortgage;financing of capital imirovements; and so on.

New Construction

The radical cutback in funding authorizations forassisted housing and the adverse effects on low-incomehousYng of the -1,x Reform Act of 1986 have led to a nearcessation of new construction for low-income housing. From1976 until 1982 approximately 1 million units of low- andmoderate-income housing were constructed under the Section 8program; in 1988, it is unlikely that more than 25,000 newunits will be built.

The National Housing Task Force has made a series ofrecommendat;ons for the development of a new housing policy,and the initiation of new construction approaches. TheAlliance strongly supports these recommendations.

Reassessing Welfare Allotments

RECOMMENDATION: The level of public housing assistancethrough HUD, 1-mHA, and the U.S. Departmentof Health and Human Services DNS) shouldbe reassessed to take into account incomelevels (including benefits-Iand reahousing costs (including utility costs).Housing benefit levels should then moreclosely match these real cost s. Inaddition, BUD, HHS, and other agenciesshould be strongly encouraged to ImproveieircoorcfL9gttinationinprovidinhousin

assistance.

49

40

41

Unde the HUD voucher programs, households arerequired to pay no more than 30 percent of their incomes forrent (59). HUD pays the difference between this amount andthe Fair Market Rent (FMR) for the unit. However, becausesuch programs are costly, only a small proportion of thehouseholds eligible for assistance can be served. Amongvery low-income households, for example, only about 17percent receive HUD subsidies.

A second, and often overlapping, source of subsidy is,hat provided under public assistance programs. Accordingto recent research, the major welfare programs essentiallyguarantee that program recipients will live in substandardhousing and similar individuals in different locations willnot be treated equally. Nationally, AFDC recipients receivehousing allotments that average only about 49 percent of thecost of a modest unit. SSI and GA recipients receiveallowances that cover 65 and 67 percent, respectively (60).The fact that a significant proportion of the homelessinclude families supported by AFDC should provide someindication that benefits under this program are inadequateto cover reasonable housing costs. Moreover, benefits varywidely from state to state, with housing allotments in someareas falling as low as 29 percent of the rent for a modestunit.

A coordinated housing program appears essential. Forexample, in one fiscal year, a total of more than $21billion was allocated for housing assistance under variouspublic assistance and housing programs administered by HUD,NHS, and other governmental agencies.

HOUSING: STATE AND LOCAL ISSUES

Single-Room Occupancy

RECOMMENDATIONS: Municipalities should be encouraged tosave tnat portion of their housingre resented in single-room occupancySul1dIngs.

59) Under the voucher program, households may elect to paya greater proportion to live in a unit that rents above theFair Market Rent for the area.

60) Newman and Schnare, cit.

50

42

The eligible use of Section 8 fundsfor residents of SRO housing should beexpanded.

Another diminished category of housing is the single-room occupancy hotel. SROs traditionally accommodate singlemen and women, including those with substance abuse problemsand the mentally ill. Considered to be substandard housinguntil recently, many of these hotels were demolished orconverted as a result of downtown expansion andredevelopment.

Between 1970 and 1983, the number of SROs in New YorkCity has declined by 89 percent; and in Atlanta by 88percent. Overall, between 1974 and 1983, 896,000 SROs havebeen lost (61). According to Andy Raubeson, director of theSRO Housing Corporation in Los Angeles, "Displacement fromSROs will surely lead to homelessness. There is nocushion."

SRO-type housing can provide a viable housing optionfor individuals with extremely low incomes and, according toRaubeson, good management and a sound tenant selectionprocess can make SROs the housing of choice (62). As aresult, several large cities have developed SRO preservationprograms and are using HUD and local rehabilitation funds toupgrade existing structures.

Well-maintained and managed SRO housing has severalbenefits: it is conveniently located, usually neartransportation hubs; it is economical; its rents aregenerally affordable to those whose incomes are derived frombenefits; and it is well-suited to special needs populationssince services can be located within the facility and thezoning constraints that face community-based facilities areabsent. SRO housing generally consists of clusters ofsingle rooms (with or without partial baths) with sharedbathrooms and lounge areas, and community kitchens. Oncelost, SRO housing is difficult to replace.

61) "Help for the Homeless," Newsweek, April 11, 1988, pp.58-59. The Reagan AdministratiFi7T17Toposed 1989 budgetdoes not include requests for funding for either Section 8assistance for SROs under the McKinney Homeless AssistanceAct or the Section 8 SRO Mod/Rehab Program. However,Congress may authorize up to $400 million for the Section 8SRO Mod/Rehab Program.

62) Interview with Andy Raubeson, director, SRO HousingCorporation, Los Angeles, California, March 8, 1988.

51

43

Zoning

RECOMMENDATIONS: States and localities should beencouraged to establish zoningtechniques that encourage rather thandiscourage mixed-income use of land. Indeneral, exclusionary zoning should beiscouraged and where- appropriate,

inclusionary zoning, designed to promoteOle development of low-income housing,should be encouraged.

States should be encouraged to pass lawsthat facilitate establishment of grouphomes, including the designation ofcommunity residences as single- familydweTlings for all local zoningpurposes.

Exclusionary zoning practices can be used to curtailestablishment of multi-unit or multi-family housing units ofthe type necessary to provide affordable housing to low-income and homeless people. While states and localitiesplan land use through zoning, this process should not beused to exclude a reasonable amount of housing for homelessand low-income people.

Design and Regulatory Changes

RECOMMENDATIONS: Federal government-sponsoredenterprises engaged in housing financeshould discourage lenders, includingstate and local housing finance agenciesrom mposing un u y res r c Ye

construction standa-Ti as part of thefinancing process.

Local building codes, while assuringsate housing, should be designed insuch a way as to take into account low-income housing needs, preservation oft±o_t_Isir__dttjeactuaT state of

is___e)ng,occA model code for rehabilitation isnee e to roviae ui ance to localau on es. uc a o

repairs andsuss at-4fti on n. the case

e 0

54,

44

of repairs or modest rehabilitation, thecode should only prohibit alterations thatproduce a nef decrease in the safety andlivability of the unit. For subftritialreKabilitation the code should onlydictate thi requirements for constructionactually performed and should not mandafechanges in other as_Rects of-The unit.

There should be coordination among thestandards for existing housing, forrehabilitation grant programs, and forbuiTding codes.

Building codes °nd other housing quality regulationsare enacted to protect the health and safety of the publicbut, by raising the cost of housing and constraining thehousing options available, they may cause some people at thebottom of the income scale to find housing less availableand more expensive. The result may be, ironically, lessadequate housing and increased homelessness. On the otherhand, housing quality standards are an important and hard-won eleil*.ent of the health, safety, and quality of lifeenjoyed in the United States. They should not be sacrificedfor some short-term cost reductions. Continuing attentionneeds to be focused on more affordable housing types and onconstruction and design approaches that reduce unit costs.

HOUSING PROVIDERS

RECOMMENDATION: The recommendations of the National Housingtask Force and National low 171655e HousingPreservaTIZT117mmission should be implemen-ted as they relate to the housing needs Er-6T-111;=Income families and e Tiome ess.

Role of States and Financing Agencies

To the extent that supply problems are localized, it isappropriate that states and localities initiate action ontheir own behalf. States and localities have responded byadopting a more aggressive role in the development ofsupply-oriented programs, including low-income housing fundsto finance construction and rehabilitation.

Almost all commissions and groups analyzing housingproblems have concluded that states and localities must playan even greater role in stimulating housing change.Coordinated partnerships between the federal government and

45

state and local entities are needed. The National HousingTask Force gives suggestions as to how such partnerships canbe forged and implemented.

Role of Nonprofits

The provision of housing alone will not satisfy thecomplex social needs of the existing homeless population.The issue is one of effective service, and nonprofitorganizations have been better able to provide combinedhousing and support services. Such programs may not alwaysbe as efficient, in terms of scale, but they almost alwaysresult in well-targeted allocations to the poor (63).

The homeless population constitutes the poorest ofdUerica's poor. Given existing conditions, includingchances in the 1936 Tax Reform Act, it is unlikely that for-profit developers will find the present rate of return oninvestiGent sufficient to engage in large-scale activities onbehalf of the homeless. Current subsidies are consideredtoo meager to attract private capital. Until existing lawchanges, nonprofits, by necessity, will be one of theprimary providers of low-income housing (64).

While generally supportive of nonprofit organizations,several commentators have noted that some nonprofits, forall their good intentions, are often poorly mananed, devoidof necessary capital, and unable to deal with theefficiencies of scale. Efficiency, however, becomes less ofan issue if one accepts that the mission of a nonprofit isto provide effective, comprehensive services to the homelessand other low-income groups. Additionally, nonprofits arecommitted to maintaininq developed units for low-income usein perpetuity.

Role of the Private Sector

It is, of course, unrealistic to assume that 5.5million new units of low-income housing can be built by

63) Sarah Becker and Donna Glenn, Off Your Duffs 8 Up theA$Set$: Common Sense for Moo Profs[ T4anagers, FarnsworthPublishing Co., Inc., Rockville Center, New York, 1985, p.215.

64, H.R. 3891 was introduced on February 2, 1988 in aneffort to encourane increased nonprofit sponsorship of cost-effective, community-based prograAs for housinorehabilitation and development. Such programs, to includeunions and churches as service providers, are intended toassist low- and moderate-income families.

46

nonprofits alone. Even the strongest advocates ofnonprofits recognize that nonprofits can supply only a smallnumber of annual units required, and that conclusion appearsas correct today as 20 years ago. The Kaiser Commission in1968 stated that to meet national housing goals, theinvo)vement of the private sector was required. The privatesector must be viewed as an ally in building and maintaininglow-income housing.

SERVICE FOR TENANTS

There is an ever increasing concern that displacementcould swell the ranks of the homel. ,s well beyond currentestimates. For example, the General Accounting Office (GAO)estimates that as many as 1.7 million units of federallyassisted housing could be lost by the year 2005 (65). Ofparticular concern is the existing stock of Section 236 andSection 221(d)(3) housing.

The National Low Income Housing Preservation Commissionhas found that, in the next 15 years, over 80 percent of theSection 236 and Section 221(d)(3) housing units will be losteither through default or prepayment of mortgages. Ninetypercent of those displaced by such actions have annualincomes below 80 percent of their locality's median income(66). These tenants may find it difficult to absorb themoving costs and are, therefore, "at risk" of homelessness.

The National Housing Preservation Task Force hasrecommended that "in the event of prepayment, the tenantshould be permitted to remain in place at no increase inrental cost to the tenant, except for regulatory reasons,until replacement housing of comparable cost and quality isavailable" (67). While there are compelling arguments infavor of such recommendations, owners and others feel theirimplementation would be in violation of the owner's,]ntractual rights.

65) Barry Zigas, "Homelessness and the Low Income HousingCrisis," National Low Income Housing Coalition, Washington,D.C,, 1987, n 1.

66) Statement of Carla A. Hills and Henry S. Reuss, co-chairs, National Low Income Housing Preservation Commission,Washington, D.C., April 27, 1988.

67, The Preservation of Low and Moderate Income Housing inthe United States, The NiTiona Housing Preservation Taski-orce Washing57, D.C., 1988, p. 4.

Jo

47

Eviction Assistance

RECOMMENDATION: States should be encouraged to establishcost-etfect-!..e E if ion asses anceprograms.

For many people, homelessness begins with eviction.For example, in New York City, 500,000 eviction cases areinitiated in the Housing Courts each year and, on average,25,000-27,000 households are displaced through warrantedevictions. any of those evicted are public assistancerecipients. Since 1969 the number of low-income householdsdisplaced through the eviction process has increased by 93percent (63).

States, such as Ne,1 Jersey and Massachusetts, havefound that programs that prevent eviction, through financialand social service assistance, are highly effective in humanand econo,Jic terms. In Massachusetts, the state will pay upto four months rent and utility arrears, providelandlord/tenant mediation and offer counseling in order toprevent eviction. The Massachusett: Department of Public1:elfare has found that the average rent arrearage payment is

This compares to an average of $6,000 to house ahomeless family in a hotel or ilotel. By keeping low-incomepeople in existing, affordable units, states can reducetheir costs and prevent homelessness.

Housing Counseling

RECMiENDATION: The federal government should institute a

program of support for housing counselingagencies (existing 1AD-certified agenciesand others) with strong state and localparticipation in design and monitoring.

68) Brent Cnaruiin, 5 Minute Justice, Report on theMonitoring Subcommittee Citywide Task Force on HousingCourt, November 1986. At present, the average number ofhouseholds evicted each year is 25,000 - 27,000. Dependingon the size of the household, the actual number of peopleevicted is even greater. By comparison, the average numberof households evicted in 1969 was 14,000; in 1973, 17,000;and in 1974, 18,500, with the peak in 1982 of 30,700. Since1982, the number has fluctuated between 25,000 and 27,000households.

48

As of 1984, there were approximately 600 public andprivate nonprofit housing counseling agencies.Approximately one-third of these received HUD funding toprovide services to HUD program participants. Many moresocial service agencies provide housing counseling as partof their case caanageiiient or other services. To theextent that they deal with tenants (most people becomehomeless after having rented, rather that through mortgagedefault), such agencies can provide a valuable service bypreventing eviction and potential homelessness.

SPECIAL NEEDS OF THE HOMELESS

RECONiiENDATIONS: It is reco2mended that those who worki+with the homeless adopt a casemanagement system. Such a system wouldrecognize the complex needs of clientsand could deliver, in an integratedway, a full range of assistan6FUFTchincludes day care, job training, jobreferral, health care, housing a-sistance,counseling, rehabilitative services,assistance in obtaining benefits, andTegai

CoinComprehensive services should beprove e to ose families andindividuals who reside in-transitionalhousing.

HUD should work more closely withmental health agencies to design andfund housing units in which adequateservices are easily and effectivelyde;ivered to the mentally ill.

Federal, state, and local governmentsshould work togetT---T§irnreaequaertoetefunding for community -based mentalhealth services that are required bythe deinstitutionalized and never-institutionalized.

Federal, state, and local governmentsshould be encouraged to expand theirfundinq programs for residential andoutpatient detoxification and treatmentcenters for homeless substance abusers.Operation of such facilities should betargeted to nonp:ofits.

49

Housing, while the fundamental issue of homelessness,is not the only need of homeless people. The majority ofhomeless people need an array of social services to helpthem stabilize. For those who previously have maintainedsuccessful households, but lost their housing through someexternal cause, such services are temporary. Others, suchas the mentally ill, require immediate assistance and apackage of permanent services to maintain them in long-termliving situations. Between these extremes lies the wholerange of the homeless population, struggling with housingshortages and low incomes. The homeless require more thanjust emeroency relief.

BENEFITS

RECOMMENDATIONS: Federal, state, and local governmentsshould improve their coordination ofbenefit and service delivery to thehomeless. This could includeeliminating paperwork and other barriersto t 'he receipt of benefits (such asaddress or Mirth certificaterequirements-) and coordinating theeligibility standards for the variousweTfreprograms. The Alliance encouragesthe ormatio-h-of sate coordinatingcouncils (Stith as those in Ohio andMassachusetts) that encourage and monitorin eragency cooperation to solve theproblems of the- homeless.

Efforts should be directed to makingstafe and Comprehensive HousingAssistance Plans (CHAPS) both morereflective of shelter needs and responsiveto the coordination of existing resources.Existing intergovernmental vehicles, suchas Slate human service agencies, should 'beutilized to facilitate communication andTITNTFEEthe tocarEapacity to addressfunding gaps and-service overlaps.

50

ACRONYMS

AFDC Aid to Families with Dependent ChildrenBMIR Below-Market Interest RateCCNV Community for Creative Non-ViolenceCDC Community Development CorporationCHAP Comprehensive Housing Assistance PlanEA Emergency AssistanceFHA Federal Home AdministrationFmHA Farmers Home AdministrationFMR Fair Market RentFY Fiscal YearGA General AssistanceGAO General Accounting OfficeHHS U.S. Department of Health and Human ServicesHUD U.S. Department of Housing and Urban DevelopmentNBER National Bureau of Economic ResearchRMA Rand McNall:: AreaSRO Single-Room OccupancySSI Supplemental Security InsuranceVA Veterans Administration

51

APPENDIX A

ESTIMATES OF THE HOMELESS POPULATION BY CITY

53ESTIMATE OF THE MOSER OF HOMELESS

CentralCity Metropolitan

Local (1) Local Pep. (4) Pop. (4) HomelessCities Estimates Counts (000,$) (000's) rate

ammangsamensamenessmassnessmssasammensamsamormassammaxasassmsamsaammosassimusLarge Metropolitan Arm*:

Elizabeth, NJ 300 (3) 107 17677 0.0028Yonkers, NYNew York, NY 29

1300 (2) 1917165

1767717677

0.00600.0040

Newark, NJ 3000 (2) 314 17677 0.0096Los Angeles, CASanta Monica, CA

32550900 (2)

3097al

1237312373

0.01050.0102

Chicago, IL 19850 2992 8035 0.0066Philadelphia, PP 3600 1646 5755 0.0022San Jose, CA 1000 (3) 686 56BS 0.0015San Francisco, CA 8250 712 5685 0.0116Detroit, MIMonroe, MI

750023

108663

4577457,

0.00690.0004

Boston,MA 3200 2113 571 4027 0.0056Brockton MA 250 (3) 96 3695 0.0026Mauston, TX 6350 1706 3566 0.0037Washington D.C. 4700 6454 623 3430 0.0075Dallas-Fort worth, T 7300 (2) 1386 3348 0.0050Riad, FL 5950 372 2799 0.0160Cleveland, ON 410 547 2708 0.0007St. Louis, MO 500 (2) 429 239B 0.0012Atlanta, GA 2000 (2) 426 2380 0,0047Pittsburgh, PAAnnapolis, MO

88873

857 40332

23722245

0.10220.0023

Baltimore, MD 690 763 2245 0.0009MinnampotisiSt. Paul 1010 624 2230 0.0016Seattle, WA 3175 488 2208 0.0065San Diego, CA 3000 (2) 960 2064 0.0031Temps, FL 625 275 1811 0.0023Denver, CO 2500 (2) 504 1791 0.0050Phoenix, AZ 1075 2477 853 1715 0.0013Cincinnati, ON 875 370 1674 0.0024Milwaukee, WIKansas City, MO

1370000 (2) 621443

15681477

0.00160.0008

Portland, OR 1550 366 1341 0.0042New Orleans, LA 1600 (2) 559 1319 0.0029Norfolk, VABuffalo, NY

300 (3)500 (3)

280339 1205

1261 0.00110.0015

Providence, RI 250 (2) 156 1095 0.0016Charlotte, NC 275 331 1031 0.0008Hertford, CT 1200 136 1030 0.0088Salt Lake City, 0T 538 165 1025 0.0033

Totals 159129 32975 147239664 (5) 0.0048

medium Metropoli tan Areas:Rochester, NY 13$ 243 989 0.0006Louisville, KY 575 290 963 0.0020Dayton, ON 295 181 930 0.0016Birmingham, AL 575 280 895 0.0021Orlando, FLRichmond, WA

4001173

(3) 137219

824746

0.00290.0054

Jacksonvitto, FL 300 (3) 578 795 0.0005Notetton, PA 1 3 26 727 0.0005Scranton, PA 70 84 727 0.0008Tula*, GL 43 374 726 0.0001!mouse, NY 375 164 650 0.0023Grand Rapids, MI 350 183 626 0.0019Raleigh, NC 213 169 609 0.0013Durham, NC 45 102 609 0.0004Tucson, AZ 1000 (2) 365 595 0.0027Fresno, CA 600 (3) 267 565 0.0022Baton Rouge, LA 175 369 538 0.0005Lae Vegas, NY 1100 183 536 0.0080Springfield, MA OTO (3) 150 516 0.0052Little Rock, AR 510 170

4730.0030

Charleston, SC 83 89 0.0012Mummer**, NM 1000 (2) 351 449 0.0028Worcester, MA 1700 160 445 0.0106Davenport, IA 360 102 381 0.0035Colorado Springs, COFort Warne, IN

98725

278165

349348

0.00040.0044

Lompoc, CA 29 323 0.0012Charleston, WA 300 (2) 267 0.0048linghemton, NY 53 55 264 0.0010

61

r4

ESTIMATE OF THE NURSER OF HOPELESS

CentralCity Metropolitan

Local (1) Local Pop. (4) Pop. (4) NceeleseCitis. Estimates Cams (0000s) (000's) rate

isatosiussammusaa. missusssevesseswismassounsuvosasusgaumwatavimmuursaluivai

Totals 13033 5806 59226211 (5) 0.0023

&esti Metropolitan Areas:Reno, NV 313 106 212 0.0038LincoW, NE 133 180 203 0.0007Fell River, MA 33 92 157 0.0004Tyler, TX 3,0 (3) 73 145 0.0048Jackson, NI 48 209 145 0.0002Monroe, LA 65 56 143 0.0012Athens, GA 43 137 0.0010Pueblo, CO 46 100 125 0.0005Burlington, VT 100 (2) 37 121 0.0027Sioux City, IA 55 82 118 0.0007Danville, VA 65 45 111 0.0014Columbia, NO 65 63 106 0.0010

Totals 131$ 1086 23483718 C5) 0.0012

Non Metropolitan Areas:Lewiston, NE 28 I 39 85 I 0.0001Pittsfield, MA 50 (3) SO I 82 I 0.0010

Totals t8 89 3775376 (5) 0.0009UMWM*SPOUVIMMUMLIMUSIMO MMMMMM USSIMMUMINNUMBINIMMAMMICOMMIMMUUMIMSWISSOURIBUMOMINMS**

Total: 173608 39956000Homeless rate: 0.00434

Sources:1. All estimates from the Department of Mousing and Urben Development, 1984,

unless otherwise noted.2. Estimates obtained from Tucker, 1987.3. Estimates from published date.4. U.S. Bureau of the Census, Department of Commerce, 1986. "State and

Metropolitan Date gook." Washington, D.C. Data are for 1934.5. Bureau of the Census, Department of Commerce, 1987. "Statistical

Abstracts of the united States.' Washington, D.C. 1985 percentages Weapplied to the total U.S. population in 1984.

Notes:1. Homeless rate local estimate divided by total central City population.

55

APPENDIX B

CHARACTERISTICS OF THE HOMELESS: RESULTS OF LOCAL STUDIES

CHAR6CTIOLISTIC21 OF= 111111ELESS POIREATION

Mental Substsoco

Sox EPPA11,69--2

AK! RateEducettoo lantk YajjArt.....tae

Length of

IPPEAmtkmaHealth

r1.0.1.

Abuse

HSBloats Gtanis Ono More

CUILIML-21-r3.1111

PopulationCfrerfst II

1

BI9I2 fooals

1

'10

1

19 39

i

40.591

60.

Median(?vet-

Ropro

Rented

Under-

Repro-

Ropted

WakedFor Pay

1.0ot No

Tear Thanor On.

Yost

1. emmemeLiasia

D.C. (1944) Street, shelter 2,562 6918) 13lol 111 of pops-

lotion

asholllo (1984) Strata. shelter 821 84 16 5 NIA 8/A 24(Skid low only)

Boston (19D3) Street, shelter 2,767 7,04)20C8) 42 of 1.1,4-

lotion

11. Iskarvimmilftlit

wanness

Baltimore 11985/ Sluilters, scup

lines. wise' Arts

236 60 40 8/8 72 NIA N/A 33 3r 22.24

chum's* (190t) Street, avatar 722 75 25 11° 39 30 19 39 VIM 34 61 39 15.22 33

Boston (1985) Street, ,.halter,

mental health tit

328 81 19 111 33 38 33 35 yre 11 38-45 30

Seattle (1984) Shelter 455 N/A NrA 361 of 22Houdoholde

841tinammh Co., OR(1984)

L.A. (UCLA, 1985)

Streets. abettors,

pow lines. SROS.camps,

Missions, soup limes.

131

238

85

7/

15

2)

3 5A

64

32

30

34 yra

35 yrn Y

6 46

37 6$ 36

19

Psalm* lot. (Primarilyfrom Skid Row)

WIC (NelEll) Strain. Shelter,

welfera hotel

516 N/A NIA 0 NIA NIA 10 Y

Austin (1985) Street on1T o0 91 7 44 5) 47

Ohio (1984) Strout*. 'bolter',hotels

9/9 81 19 63 30 a 3$ yes Y 25 37 77 23 30

Utsh (1966) Not documented 243 78 (a)151 35 Yis 70

6 ;)

Mental Substancefilets- Lienhth of Health Abuse

in Box __Lellima__ his $0541 tsun, Kit volooms ItnitiiIlms rtekkns EsMSWEL-1.11 HS

Plea, Weds One MoreOver- Under- Wet/ed Tear Then

Mepuitalde I I I 1 s I mdfon ReP2*- Repro- For Fey or One9427111/ai of Alias Smug 0 HA, IraiL2 114 12:12 4-52 t2.! _Ams.... sIsisJ mis0 kullis _.1. kiss IHA1 1

III. PaOtelder Hal3mitas

Mbrplend (1113) 04 1* 281 of papa- 22 ItiA 0/A 8

Lath=

Sew To S* 119641 366)aOa) 5U Of pew- Td 35 31 6

Lathan

S eaton 11,231 7011) 1646) 111 2 51 32 6

(Cm)S ID Ittionil Matter 661') 331e) 211

214,17 (1104)

)W.S. tartan roe 5(s 1511) 281

Ot Plynts 11916)

U.S. Cendiesnice

at !lepers (1967)

Patna

asIs) 3411)331

Sesabdcosne lus(ado individuals and Unit, sombers in most cease. Brealdneas mbicb *poly to baseless /Anglia gulp are maked ta).

1 Ago billatAMOND are aggaroularts due to exporting differences among the studies, Wbo:s different. categories v d in the local studs** aro noted

a <15, 15-29, 0-56. 550b <21, 21-34. 35-54. 55ea ell

d <12, 111-34. 31-10. 604

17.19, 20-44. 5.66, 6,,

19

22

321 *) 33")

31")

22 38

29 29

26 33

Austin:

Baltimore:

Boston,Massachusetts:

Chicago:

Los Angeles:

Maryland:

Massachusetts:

59

APPENDIX C

REFERENCES

Local Studies

Baumann, Beauvais, Gingsby and Schultz,"The Austin Homeless," Report to theHogg Foundation for Mental Health,University of Texas at Austin, 1985

Repc-t to the Baltimore City HealthDepartment, Division of Mental Health,Mental Retardation and Additions, onthe Mental Health and Service Needs ofBa.L.,re Residents in Temporary ShelterCare, 1935

City of Boston Emergency ShelterCommission, "The October Project:Seeing the Obvious Problem," 1983

Rossi, Fisher, and Willis, Social andDemographic Research Institute andNational Opinion Research Corp.,"The Condition of the Homeless inChicago," University of Massachusettsat Amherst, 1986

Robertson, Ropers and Boyer, "BasicShelter Research Project," University ofCalifornia at Los Angeles, 1985

Maryland Department of Human Resources,"Where Do We Go From Nowhere," 1986

Human Services Research Institute,Homelessness Needs Assessment Study:Findings and Recommendations for theMassachusetts Department of MentalHealth

Multinomah County,Oregon: Social Service Division, "The Homeless

Poor," Multinomah County, Oregon, 1984

Nashville:

New York State:

Ohio:

Seattle:

Utah:

60

Bruce Weigarn, "Counting the Homeless,"American Demographics December 1985

New York State Department of SocialServices, "Homelessness in New YorkState: A Report to the Governor and theLegislature," 1984

Roth, Bean, Lust, and Saveanu,"Homelessness in Ohio: A Study ofPeople in Need, Statewide Report," OhioDepartment of Mental Health

Human Resources Coalition, "EmergencyShelter Study," Housing and CommunityDevelopment Division of the Planning andCommunity Development Department, KingCounty, Washington, 1984

Task Force on Appropriate Treatment ofthe Homeless Mentally Ill, "Homelessnessin Utah," undated

Washington, D.C.: Frederic Robinson, "Homeless People inthe Nation's Capital," University of theDistrict of Columbia, Washington, D.C.,undated

General References

Bassuk, Ellen, "The Feminization of Homelessness: HomelessFamilies in the Boston Shelters," Keynote address given atthe Shelter. Inc. Benefit, Cambridge, Massachusetts, June11, 1985

Bingham, Green, and White, eds., The Homeless inContemporary Society, Urban Research-Center, University ofWisconsin and Sage Publications, 1987

Clay, Philip, "At Risk of Loss: The Endangered Future ofLow Income Rental Housing Resources," NeighborhoodReinvestment Corporation, Washington, D.C., 1987

Department of Housing and Urban Development, "HomelessAssistance 1981 -88," Washington, D.C., 1987

Department of Housing and Urban Development, "A Report tothe Secretary on the Homeless and Emergency Shelters,"Washington, D.C., 1984

DiGiovanni, Frank, "Summary of Alternative Ways to CreatePermanent Housing Using the Higher Streams Available forHousing Homeless Families," undated

61

Erickson and Wilhelm, eds., The Homeless in ContemporarySociety, Urban Research Center, University of Wisconsin andS-S-§7-177Foductions, 1987

Food Monitor, Special Issue for the Campaign to End Hungerand Homelessness, Winter:1988

Friedman and Hall, "Permanent Homelessness in America?,"Working Paper No. 2013, National Bureau of EconomicResearch, Cambridge, Massachusetts, 1986

"Homelessness: Critical Issues for Policy and Practice," TheBoston Foundation, 1987

National Low-Income Housing Coalition, "Homelessness and theLow-Incorle Housing Crisis," Washington, D.C., undated

National Low-income Housing Coalition, "Low-Income PolicyStatement," Washington, D.C., adopted 1984

Newman and Schnare, Subsidizing Shelter: The RelationshipBetween Welfare and Housing Assistance, Report No.Urban Institute Press, Washington, D.C., 1988.

Newman and Struyk, "Housing and Supportive Services:Federal Policy for the Frail Elderly and ChronicallyMentally Ill," paper prepared for MIT Housing PolicyProject, 1987

Public lechnology, Inc., "Caring for the Hungry andHomeless: Exemplary Programs," Emergency Food and ShelterNational Board Program, 1985

Redburn and Buss, "Beyond Shelter: The Homeless in theUSA," Cities February:1987

Schussheim, Morton, "Housing the Low-Income Family withChildren," Congressional Research Service, Washington, D.C.,1987

Simpson, Kilduff, and Blewett, "Struggling to Survive in aWelfare Hotel," rommunity Service Society in New York, 1984

The Stewart B. McKinney Homeless Assistance Act, P.L. 100-77[H.R. 5585, July 22, 1987

Tucker, William, "Where Do the Homeless Come From?,"National Review September 25:1987

Turner, Margery, "Housing Needs to the Year 2000: Problemsand Assistance for America's Poor," National Association ofHousing and Redevelopment Officials, Washington, D.C., 1986

62

U.S. Conference of Mayors, "The Continued Growth of Hunger,Homelessness and Poverty in America's Cities: A 25 CitySurvey," Washington, D.C., 1986

U.S. General Accounting Office, "Homelessness: A ComplexProblem and the Federal Response," Washington, D.C., 1985

Walker, Judith, "Local Housing Policies AffectingHomelessness," 1987

i

'11-11.: NA FIONA!. All,IAN('F: IAI) 110N11.1.1.4S\FSS1518 K N.W.

..fitc 206

Washinpon, J).C. 20005(202) 638-1526