SHELTERING HOMELESS PERSONS IN CAMBRIDGE AND SOMERVILLE:

Homeless people have become a familiar sight across the United States--in cities large and small, in suburbs and in rural areas. A growing body of research provides a portrait of the homeless across these diverse locations: most are extremely poor, many are plagued by serious health problems, all confront a severe housing shortage; the average age among adult individual homeless persons is about 34 and many have completed high school, but families with children are also represented increasingly among the homeless.

In response to the needs of this large array of problems and needs, a wide variety of service arrangements have developed. Sometimes, the local response has been relative disinterest. More commonly, traditional charities or city agencies have expanded their shelter beds. In many communities, local churches and other concerned groups have relied on their financial contributions, volunteer labor, and warmth and concern to operate small shelters or meals programs. Support services often develop with little coordination and few resources.

This report describes the efforts of two communities in the Boston area to respond to the needs of homeless people. They are not average communities nor were their responses to homelessness typical: Both communities developed exceptionally comprehensive, coordinated shelter and service systems for homeless persons. But although they can hardly be said to represent most communities, even in the Boston area, the experiences of these two communities can help to inform efforts to help the homeless throughout the United States.

This report highlights the experiences and attitudes of shelter staff, the characteristics of the homeless people they serve, the relations between shelters and service agencies, and innovative approaches to meeting homeless persons' needs.

Studying Sheltering

Information on services for homeless persons in Cambridge and Somerville was collected from four sources: shelter staff, shelter directors, agency service personnel and service network administrators. Staff were surveyed with a twelve-page self-administered questionnaire, while shelter directors and agency personnel were interviewed in-person with a structured interview schedule. These different data sources are indicated in the following tables; in some tables, staff responses are aggregated to the shelter level in order to highlight inter-shelter diversity. The two network administrators were interviewed with a less structured set of questions. Information they provided is incorporated throughout the report.

The goal of the study was to interview every shelter director and a representative of every agency important for homeless services in Cambridge and Somerville. This goal was nearly achieved: interviews were conducted with representatives from twelve of fourteen Somerville agencies and all eight Cambridge agencies, while only one of thirteen shelters in the two communities (nine were in Cambridge) declined to participate.

Nine shelters participated in the staff survey (two shelters for battered women were not included in the staff survey). In eight of the shelters, questionnaires were distributed to all full-time staff; the other shelter relied exclusively on volunteer staff, so all regular volunteer staff were surveyed. Of 91 shelter staff given questionnaires, about sixty percent (54) returned them to the university in self-addressed, stamped envelopes.

Describing Shelters and the Service Network

Overview

Cambridge, Massachusetts is bordered on its south by Boston (and the Charles River) and on its north by Somerville. Cambridge and Somerville differ in many respects, but their social service systems are intimately related. Cambridge is home to two of the world's greatest universities, Harvard and MIT, and includes some of the highest-priced real estate in the already expensive Boston area. But Cambridge also is a city of over 100,000 with many neighborhoods of poor and working class families living in typical Boston multi-family homes. Somerville is a largely white working-class suburb that is both smaller and more homogeneous than Cambridge.

Somerville

Somerville provides four shelters for homeless persons: the shelters for single adults and for women and men with dependent children were included in each component of the study (a shelter for homeless teens was not included in the study, while the shelter for battered women was included in the interviews but not in the staff survey; a city-operated apartment for two families was not considered to be a shelter and was not included). The adult shelter opened in February, 1986; the family shelter, in December 1987. Both shelters are operated by the Somerville Homeless Coalition for the Homeless with oversight provided by a community advisory board.

A group of concerned church members and other community residents formed a task force in Somerville early in 1985 to explore ways of helping the local homeless population. One local church narrowly approved the task force's recommendation to establish the adult shelter in its basement, but a number of church members and neighbors expressed their opposition. In part due to this opposition, the task force agreed that on a number of restrictions on shelter operations: prospective guests would be screened by local service agencies and would be escorted to the shelter, people who were severely behaviorally disordered by mental illness or actively abusing alcohol would not be allowed, case management would be provided to help all guests, guests would have to be motivated to resolve their homelessness, and the shelter initially would be open for just six months.

The plan worked. Problems with neighbors dissipated and in October, 1986 the shelter opened on a year-round basis with two additional beds. Three-fourths of the congregation expressed its support for the expansion and many church members provided meals and financial support. And some shelter guests continue to visit the shelter long after successful rehousing--just to keep in touch.

Members were selected for the shelter's initial advisory board to represent different constituents in the community. Now, board members also are recruited for their expertise in areas of board concern: fundraising, bookkeeping, computers, personnel, social work, nursing. Meetings are now scheduled monthly; initially they were held weekly.

Many decisions concerning shelter operations are made by the director of the coalition. Operations have been improved by a policy of hiring shelter managers based on their experience and human relations skills.

Positive relations with local service agencies, in both Somerville and Cambridge, have aided Somerville shelter operations. Good community relations and a spirit of volunteerism have also helped to ensure the success of several annual fundraising events--a walkathon, a comedy night and a Christmas party. Several companies, agencies, fraternal organizations and schools also have donated goods or provided volunteers.

Cambridge

At the time of the survey, Cambridge had two family shelters and six for adult individuals (one other shelter for battered women was included in the interview study but not in the staff survey). Many services for homeless persons in Cambridge were provided at the Cambridge Multi-Service Center. The Center was started, in July 1987, to reduce two major problems experienced by service agency staff in responding to the needs of homeless persons: homeless people had to walk between multiple agencies in order to meet their needs (as a result, many did not receive critical services), and staff in different agencies had to have special meetings in order to coordinate responses to their homeless clients.

The Multi-Service Center provides emergency casework, social work, housing assistance, medical, mental health and teen services and referral services in a central location. Frequent staff meetings and interchange of staff with local shelters and agencies facilitate effective service delivery.

Some of the wide national variety of shelter approaches was represented among Cambridge shelters. One shelter was operated by a nationally-recognized recovery program for alcoholics; it drew most of the homeless alcoholics in the city and provided them access to a wide variety of rehabilitation services (Schutt and Margelis, 1986). Another Cambridge shelter was run by a church near Harvard Square in cooperation with students at Harvard University. Student volunteers provided all the shelter staff and operating funds were raised on campus. Another prominent Cambridge shelter for adult individuals received substantial funds from the private sector, relying on a board that included representatives from some of the city's large corporations. This board also funded a family shelter in Boston; both shelters used a proactive approach to rehabilitation. In general, shelters in Cambridge varied in their emphasis on social services and in their accommodations: from furnished bedrooms in a multi-family house to cots in a church basement.

Joint planning and networking is facilitated in both Cambridge and Somerville by monthly meetings of the Cambridge/Somerville Committee on Homelessness--a group of all service providers, the Greater Boston Adult Shelter Alliance--an organization of all shelter directors that plans common initiatives, the Massachusetts Shelter Providers--comparable in function to the GBASA, and the Massachusetts Coalition for the Homeless--an advocacy group. The Cambridge Fund for the Homeless receives funds from local businesses and maintains a meal program with volunteer and institutional help from local churches and Harvard University.

Operations

Interviews with shelter directors and agency liaisons provided additional information on the operation and financing of Cambridge and Somerville shelters. Internally, shelter operations were often formalized. Of 22 possible types of records, charts, manuals and written rules, half of the shelters used at least 15.

The operations of most agencies were even more formalized. Of 22 possible types of records, charts, manuals and written rules, half of the agencies used at least 17. Two-thirds of the agencies kept an intake census or other records on their homeless clients.

Over half of the non-shelter service providers represented in the interviews were public agencies; about one-third were private nonprofit vendors who received some public funding. Only one agency did not receive any public funds. Many agency respondents did not know what particular percentage of their funds came from particular government sectors. Federal and state government provided funds to two-thirds of the agencies, while city government, private groups, and special fundraisers each provided funds to one-third (Table 1). A few agencies received funds from national foundations and churches.

Table 1

Funding Sources of Service Agencies,

Agency Responses

Source Percent Receiving Some Funds

State 73%

Federal 67%

City 33%

Private Donations 33%

Fund Raising Events 33%

Other Local Group 33%

United Way 27%

Charities 27%

Churches 13%

Local Gov't 13%

Ford Foundation 13%

R. W. Johnson Fdtn 7%

Community Support 0%

All but one of the agencies reported coordinating with other agencies or coalitions on a regular basis. When referring homeless mentally ill clients to other agencies, follow-up was provided by seven agencies. Five were aware of regulations on follow-up and eight were aware of regulations on referrals. The types of information often used by agencies in deciding where to refer homeless mentally ill clients were client-provided information, demographic information, and records.

Summary

Overall, services for homeless persons differed markedly between Somerville and Cambridge. Somerville's service system was relatively small and the shelters were relatively exclusive, screening out potentially troublesome clients. The small size of the shelters was associated with a relatively homey atmosphere and ongoing supportive contact with former shelter guests; the walls of one shelter were decorated with artwork by "graduates." Nonetheless, shelter operations had become increasingly professionalized, with greater attention to including management and human service experts on the board of the Somerville Homeless Coalition.

Cambridge, a larger and more diverse city, had a larger and more diverse shelter system that included within it several approaches to sheltering. In two important respects, however, the Cambridge shelter system was very centralized: most alcoholics in the system relied on one large shelter, while most services for needs other than substance abuse were available in the city's innovative Multi-Service Center. In both communities, relations between the community and the shelters were mutually supportive and in both, the shelters depended largely on state and federal funds.

Shelter Clients

The homeless population is strikingly diverse, ranging from older single men and women to young mothers with children; from recently unemployed workers to chronic substance abusers and schizophrenics. Shelter operations and staff experiences can only be understood in terms of the type of homeless persons with which they work.

Staff Description

Shelter guests can be afflicted with a range of health problems. According to the staff, the most common of these health problems is alcoholism or alcohol abuse, reported by Somerville staff for one-third of their guests and by Cambridge staff for 45 percent of their guests. Drug abuse was identified among about one in five guests by staff in both cities. Chronic mental illness and physical illness or injury were identified for only about one in ten guests in Somerville, but for one in five guests in Cambridge. (Figure 1)

Although alcohol abuse was the most commonly reported health problem, it was also distributed most unequally between shelters (Chart 1). Staff at two shelters reported less than 20 percent alcoholic guests, while staff at Cambridge's shelter for alcoholics had more than 80 percent alcoholic guests. Most shelters identified less than 20 percent of their guests as having problems with physical or mental illness or drug abuse.

Chart 1

Percent of Guests with Health Problems, by Shelter

Aggregate Staff Responses

Percent Alcohol Physic Mental Drug

80-99 │E

60-79 │

40-59 │BD E

20-39 │ACHI I DEI BCDE

0-19 │FG ABCDFGH ABCFGH AFGHI^*

^*Each letter represents one shelter.

Staff opinions varied concerning the kinds of people and procedures that are best suited for working with the homeless. On average, staff were undecided about the proposition that those who work with the homeless should have experienced poverty; although the Cambridge staff leaned more toward disagreement. Staff opinions were split when asked whether those who work with the homeless should be professionally trained, rather than being just a caring individual. Responses were also mixed when considering the proper procedures for dealing with the homeless. Most of the staff did not agree that intrusive procedures such as intake interviews and formal assessments should be avoided but were undecided when asked whether people who work with the homeless should try to remain somewhat detached. (Figure 2)

When the staff from Somerville and Cambridge were questioned concerning their opinions of homeless people and the kinds of services they need, the most widely held belief was that "the homeless need a lot of social services to get back on their feet; just giving them financial benefits is often not enough." The staff also, on average, disagreed that the homeless are responsible to some extent for their homelessness and felt that they are somewhat the victims of social problems. The prevailing sentiment was that, "in general, the homeless are just regular people, like the rest of us; they're not so different." (Figure 2)

Agency Description

Most agency representatives were able to estimate the percent of their homeless clients according to race, gender, age, family status and geographic mobility. The median percent black clients among the agencies was 20; most homeless clients were 18-40 and few were elderly; the median percent female was 50; single homeless people and single parents with children were represented almost equally among homeless clients of the agencies; on average, only ten percent of the homeless clients seen at these agencies recently had moved to the local area. (Table 3)

Table 3

Representation of Subgroups Among the Homeless at Agencies,

Agency Responses

Subgroup^* Median

Black 20%

Hispanic 5%

Asian 3%

White 60%

Under 18 3%

Age 18-40 60%

Age 40-60 30%

Over 60 10%

Female 50%

Single 50%

Single with Child 60%

Couples 5%

Recent Migrant^** 10%

^*Categories of race and marital status are neither mutually exclusive nor exhaustive, so percents do not add to exactly 100 within clusters.

^**Migrated from other states or counties within last 3-6 months.

Most agencies reported seeing homeless clients of both sexes, with and without children (Chart 2). Few reported elderly homeless clients. Each of the major health problems were represented among the agencies' homeless clients.

Chart 2

Number of Agencies with Particular Types of Homeless Clients,

Agency Responses

Client N of Agencies

Men XXXXXXXXXXXXXXXXXXX

Women XXXXXXXXXXXXXXXXXX

Families XXXXXXXXXXXXXXXX

Children XXXXXXXXXXXXX

Mother+Child XXXX

Elderly XXXX

Mentally Ill XXXXXXXXXXXXXXXXX

Substance Ab. XXXXXXXXXXXXXXXX

Ment. Ill+S.Ab.XXXXXXXXXXXXXXX

Physically Ill XXXXXXXXXXXXXX

Situationl Hml XXXXXXXXXXXXXXXX

Estimates of the prevalence of health problems are suspect, of course, if they are not based on acceptable assessment methods. In this study, methods of assessing mental illness were identified. The methods used most often were less formal--behavioral observation and social history; a history of psychiatric hospitalization was used somewhat less often, while the most formal assessment methods--psychiatric or psychological or medical evaluation and clinical diagnosis were the least often used methods.

Chart 3

Kinds of Information Used to Determine Mental Illness,

Agency Responses

Evaluation Method N of Agencies

Behavioral Observation XXXXXXXX

Social History XXXXXXX

Other XXXXXX

History of Hospitalization XXXXX

Psychological Evaluation XXXX

Psychiatric Evaluation XXX

Medical Evaluation XXX

Clinical Diagnosis XXX

Summary

According to shelter staff, alcoholism was the most common health problem among shelter users; it also was the health problem most concentrated in one shelter and most variable in its representation between the other shelters. This variation reflects different shelter policies: some would not accept people who were actively drinking, while others would. The two Somerville shelters required that alcoholics have abstained from drinking for at least three months prior to entering the shelter. According to staff, serious mental illness and drug abuse occurred in only a few shelters among as many as one in twenty shelter guests; few staff identified serious physical illnesses among their guests. Staff believed strongly in the need of homeless persons for social services, but they had mixed feelings about the need for professionals in shelter work. Most shelter staff rejected ideas that blamed homeless persons for their problems.

According to agency respondents, their homeless clients tended to be white women in their 20s and 30s. Most agencies also saw some homeless persons who were members of minority groups and who were mothers with children; almost all the agencies reported homeless clients who were mentally or physically ill or substance abusers, as well as persons who were homeless due to situational factors. However, few agencies conducted a formal mental health assessment.

Shelter and Agency Services

Services were provided to homeless persons in Cambridge and Somerville both directly by shelter staff and indirectly, through referrals to other agencies.

Philosophy and Inclusion Criteria

Shelters in Cambridge and Somerville provided a range of services for their guests, but the shelters varied in their specific service approaches and experiences.

Shelter service philosophies emphasized the worth of homeless persons: "Provide a caring and home-like environment, with long-term support"; "provide fair and compassionate service, and treat others as you want to be treated"; "encourage empathy toward guests, empowerment of women, validate guests' feelings and help guests with resources"; "enhance the dignity and self-esteem of guests"; "help guests obtain housing"; "provide emergency housing"; "enhance family life and self-sufficiency."

Most of the shelter directors said that they preferred a specific type of guest; usually, the preference was for guests who were not substance abusers. Some of the shelters sought primarily to serve a particular portion of the homeless population. In addition, all of the shelters had barred some guests from subsequent admission. The most common reason for barring was substance abuse, particularly when in combination with mental illness or violent behavior.

Nine agencies indicated criteria that homeless persons must meet to be eligible for the agency's services: mentally ill (2 agencies), former residence within the geographic area (2 agencies), homeless through no fault of their own (2 agencies), non-disruptive (2 agencies), in danger of becoming homeless (2 agencies), other criteria (2 agencies).

Service Delivery

Shelters.

Shelter directors identified food and shelter as the service most often provided, of course (Chart 4). Most shelters also advocated for their guests in efforts to secure services needed from other agencies. Half of the shelters provided case or resource management for their guests and almost half provided some type of social skills training and leisure activities.

Few health services were provided in-house by the shelters, but screening for infectious diseases was provided at a hospital, by a local clinic, or, in Somerville's family shelter, by a nurse practitioner from the hospital. Several shelters provided regular in-service training on health problems for both guests and staff. Almost half of the shelters monitored psychiatric medication, but few provided therapeutic or rehabilitative services for mental health or substance abuse problems. In at least one shelter, the local state mental health office across the street provided assistance as needed.

Chart 4

Services Provided In-house,

Shelter Director Responses

Service Number of Shelters

Housing

Food/Shelter XXXXXXXXXXXX

Trans. Housng X

Perm. Hsng Ast X

Physical Health

Physician

Nursing

Phys. Therapy

Dental Srvce

Speech/Hrng Tpy

Nutrition Cnsl X

Mental Health

Acute MH Trt XX

Crisis Stblztn XXX

Diagnosis XX

Monitoring Med XXXXX

Self-Med XXXXX

Psychotherapy XXX

Alc. Rehab XX

Drug Rehab XX

Rehab and Vocational

Social Skills XXXXX

Life Skills XXXX

Leisure Actvty XXXXX

Job Cnslng XX

Job Develop. XX

Job Placement XXX

Sheltered Wrk XXX

Trans Work XXXX

Social Services

Fmly Cnslng XX

Case Manag. XXXXXX

Resrce Manag. XXXXXX

Advocacy XXXXXXXXX

Shelter staff must respond in some way to whatever problems their guests report, so they were more likely than shelter directors to report providing services directly for a range of problems.[1] Those services most often provided directly were food and beds and assistance with personal crises. Direct help with family, alcohol, physical and mental health problems was provided somewhat less frequently. Help with child care, financial benefits, and job training/placement was less likely to be provided directly. (Figure 3)

Staff working in Somerville were somewhat more likely to provide direct help for family problems and child care than were Cambridge staff. Cambridge staff provided more direct help with food and beds. (Figure 3)

The median number of agencies used weekly by shelters was five, with a range from zero to seven. In both cities, referrals were provided less often than direct service for all types of problems except financial benefits and job training/placement. Referrals were most common for drinking problems and physical health problems and, in Cambridge, for food and beds. (Figure 3)

The average frequency of providing help was three, on a scale from 1 to 7; no shelters deviated from this average by more than one point. Most of the shelters also received an average score of three on overall frequency of referrals, but two shelters seemed almost never to refer guests (Chart 5).

Chart 5

Overall Frequency of Helping Guests,

Aggregate Staff Responses

Through

Directly Referrals

Often 7│ 7│

6│ 6│

5│ 5│

4│AEF 4│

3│BDGHI 3│ADEFGI

2│C 2│C

Never 1│ 1│BH

The primary shelter mission in Cambridge and Somerville was to provide food and beds; but many shelters also created opportunities for leisure-time activities and monitored medications. In many shelters, case management services assisted homeless persons to secure services from outside agencies. In a few shelters, staff also maintained supportive contacts with guests after they left the shelter for a stable residence.

Agencies.

Agencies in Somerville/Cambridge provided services in each of the areas relevant to the needs of homeless persons Chart 6). Food and clothing were the most common services provided, by twelve of the nineteen agencies, with counseling and housing services a close second. Mental health, substance abuse, and other medical services were provided by between eight and five agencies; seven agencies offered employment-related services.

Chart 6

Services Provided by Agencies,

Agency Responses

Service Number of Agencies

Food & Clothing XXXXXXXXXXXX

Counseling XXXXXXXXXXX

Housing XXXXXXXXXXX

Mental health XXXXXXXX

Employment XXXXXXX

Substance Abuse XXXXXX

Medical XXXXX

Other XXXXXXXXXXXXXXX

After listing all services their agency provided, representatives were asked to identify the two services that were most important to their agency. Housing was mentioned most frequently, by seven agencies; general counseling and "other" were mentioned by five (Chart 7). Four agencies considered either medical or mental health services most important, while only two identified substance abuse or food and clothing. Two agencies also mentioned legal services or financial services, while three simply identified "referrals."

Chart 7

Two Most Important Services Provided by Agency,

Agency Responses

Service Number of Agencies

Housing XXXXXXX

Counseling XXXXX

Other XXXXX

Medical XXXX

Mental health XXXX

Referral XXX

Substance Abuse XX

Food & Clothing XX

Legal XX

Financial XX

Other XXXXXXXXXXXXXXX

Homeless mentally ill individuals were referred to the nineteen agencies for a range of services. The most common service for which they were referred was transitional housing--11 agencies had received such referrals (Chart 8). Six agencies had been sent homeless mentally ill clients needing financial benefits, while between two and four agencies had been sent clients needing some type of health service, counseling or food. Only one agency reported having been referred homeless mentally ill individuals for drug abuse services. Five agencies (of 14 responding to the question) reported that more than ten percent of homeless persons referred to their agency had been referred for services they did not provide.

Chart 8

Kinds of Services for which Homeless Mentally Ill Individuals Referred to the Agency,

Agency Responses

Kind of Service N of Agencies

Transitional Housing XXXXXXXXXXX

Financial Benefits XXXXXX

Mental Illness XXXX

Physical Health XXXX

Food XXX

Alcohol Abuse XX

General Counseling XX

Legal Counseling XX

Referral XX

Drug Abuse X

Overall, service agency representatives reported few referrals from homeless shelters (Chart 9). Six of the fourteen agencies answering the question reported no referrals of homeless persons from shelters; but six agencies reported that one-quarter of more of their homeless clients were referred by shelters. The median percent of homeless clients who were self-referred was 50 among these 14 agencies; of the self-referred, or "walk-in" homeless clients, about one-third were currently residing in a shelter for the homeless.

Chart 9

Source of Referrals of Homeless Clients,

Agency Responses

% of Clients Number of Agencies

Source: Shelter Walk-In

0 XXXXXX

1-10 X XXXX

11-19 X

20-29 X

30-39 XX X

40-49 X

50-59 X XXX

60-69 X

70-79 X

80-89 X X

90-100 XXX

Summary

Homeless persons enter shelters in Cambridge and Somerville with numerous service needs. The shelters responded readily to the most urgent of these needs--for food, beds and crisis management--but few shelters were equipped to help with mental or physical illness, substance abuse, or family problems. There were some exceptions--most alcoholic single homeless persons relied on a shelter connected with extensive substance abuse rehabilitation services and women with children used one of the three family shelters. But to find help for many problems, homeless persons had to rely on local service agencies.

The network of service agencies in Cambridge and Somerville seemed to provide a wide range of services, but many shelters were not actively involved in that network. Most homeless clients were seen at agencies on a walk-in basis, without referral from shelters, and shelter staff reported relatively low rates of referral to agencies. Housing was the most common basis for referral.

Assessing the Service Network

Shelter staff and agency representatives were asked a range of questions to determine their satisfaction with shelter and agency operations.

Service Effectiveness

Both shelter staff and agency representatives rated the delivery of services by the shelters and by the agencies.

Staff Evaluation

Staff at both the Somerville and Cambridge shelters were highly satisfied with their shelters efforts toward their guests. The staff reported being "very satisfied" with their shelters' efforts at providing food and beds for their guests and were more than moderately satisfied in nearly all other areas, from providing help with physical problems to helping with family problems and child care. The staff were least satisfied with job training/placement efforts for their guests. (Figure 4)

Staff satisfaction with shelter service efforts varied widely among the individual shelters. This inter-shelter variation exceeded the variation between Cambridge and Somerville (Chart 10).

Chart 10

Overall Staff Satisfaction with Shelter Service Efforts,

Aggregate Staff Responses

Very satisfied 7│

6│AG

5│BDE

4│FHI

3│C

2│

Not at all satisfied 1│

Staff perceived some problems in the network of agencies delivering services to the homeless. Those homeless persons requiring multiple services were seen as falling into cracks between agencies, and most staff perceived service gaps--services that were not being provided by any agency. Staff also identified a lack of coordination among agencies at the direct service delivery level. To a lesser extent, staff indicated that there were "some" disagreements within the network of service agencies and that these differences were worked out; Cambridge staff were more likely to report that differences were worked out. On average, staff perceived "little" competition among service providers. (Figure 5)

There was little variation in perceptions of service network functioning between shelters (Chart 11).

Chart 11

Perceptions of Fragmentation and Conflict, by Shelter,

Aggregate Staff Responses

Fragment Conflict

Very much 5│

Much 4│ADEFHI

Some 3│BCG AFHI

Little 2│ BCDEG

None 1│

For the most part, staff judged relations with the local agencies concerned with the various needs of the homeless as cooperative. Cambridge staff reported somewhat less cooperation by agencies concerned with physical health, housing, drug abuse, jobs and job training and child care. (Figure 6)

Staff gave mixed responses when asked to rate the adequacy of resources available in their city for meeting the specific needs of the homeless. In Somerville, resources for physical health and alcohol abuse needs were seen as fairly adequate, and resources for dealing with drug abuse, mental health and family counseling needs as only slightly less so. Less adequate resources were available in Somerville in the areas of jobs, financial benefits, child care and transitional housing. Cambridge staff rated resource adequacy as only moderate in most areas, with slightly higher ratings given to physical health, alcohol services and family and child care resources. (Figure 7)

There was little variation in perceptions of inter-agency cooperation or resource adequacy between shelters (Chart 12). The level of cooperation was rated more highly than was resource adequacy.

Chart 12

Inter-Agency Cooperation, Resource Adequacy, by Shelter,

Aggregate Staff Responses

Cooperation Resources

High 7│

6│BCFG

5│DEHI B

4│A CDFG

3│ AEHI

2│

Low 1│

Agency Evaluation

Service agency representatives were asked to evaluate service delivery procedures in several ways: in terms of relations with shelters, in terms of major service problems, and in terms of the adequacy of resources provided by the agency to meet homeless clients' needs.

One-third of the service agencies reported a need for improvement in communications with shelters; several sought more or quicker access to information about guests and their needs. Only three agencies expressed a need for improvement in their relations with shelters in terms of effectiveness, cooperation or competence. Problems were mentioned in shelter staff training or attitudes toward homeless persons and in some shelter regulations.

The major problem identified most often by agency representatives in servicing homeless mentally ill clients was a lack of money. Lack of government involvement and lack of affordable housing also were frequently mentioned. Only two agency representatives mentioned a lack of shelters or beds.

Agency representatives rated the adequacy of the resources provided by their own agency for meeting the needs of homeless clients on a seven-point scale in each of several areas (Chart 13). Average resource adequacy was highest for family counseling (4.5), with physical health only slightly lower (4.2). Alcohol abuse, mental health and financial benefits received average adequacy ratings at about the half-way point (3.7-3.9), with education and training and job opportunities receiving slightly lower scores (3.3-3.4). The resources felt most often to be inadequate were transitional housing (3.2) and child care (2.8).

Chart 13

Adequacy of Resources Provided by Agencies,

Agency Responses

Area of Need Average Adequacy

Family counseling ***************

Physical health *************

Alcohol abuse ************

Mental health ************

Financial benefits ************

Education/Training **********

Job opportunities **********

Trans. housing *********

Child care ********

_________________________

1 2 3 4 5 6 7

Summary

Most shelter staff expressed high levels of satisfaction with their own shelter's service efforts, although service satisfaction varied markedly between shelters. Staff evaluation of the service network was not so positive. Many found that there were gaps or coordination difficulties, and that resources were inadequate for some problems. In general, staff in Cambridge evaluated their city's service network less positively than did staff in Somerville. Understandably, shelter staff in both towns were least satisfied with the availability of transitional housing.

Service agency representatives confirmed staff evaluation of service delivery: many reported problems in communication and informal interviewee comments also suggested some conflicts between agencies. Few agency representatives identified problems in the emergency services provided by shelters or in physical health care, but transitional housing and child care resources were widely perceived as inadequate.

The Commmunity Context

Shelter directors reported good relations with the surrounding community. In fact, five of the twelve reported "very good" relations with the community and half reported "good" relations; just one shelter director felt that relations with the community were "mixed." These positive relations are particularly impressive, since they were achieved in several instances after an initial period of conflict.

Agency representatives also reported good community relations. Two-thirds reported "very good" relations and almost ninety percent said their relations with the surrounding community were "good" or "very good." Many representatives elaborated on the positive quality of relations with their community: "The community has come a long way --everyone works together"; "people are good at coming through on short notice"; "people work well together on common problems"; there is "good cooperation"; "the agency is respected throughout the community"; "there's a high level of commitment and genuine caring"; it's a "great collaborative effort."

Those who perceived problems in agency-community relations tended to focus on difficulties in maintaining order: "there are "complaints about families not controlling their children"; "women sit with their boyfriends on the stairs outside at night"; there's "yelling and fighting on the street -- neighborhood stuff." Two representatives identified problems in attitudes of some community members toward the homeless or their advocates: "there sometimes are people who hate advocates"; there's "discrimination toward the homeless--fear and hatred."

The staff were asked to reflect upon the changes in the number of homeless persons in their city and on the availability of services for them over the past year. Nearly all of the staff in Somerville, and most in Cambridge, felt that the number of homeless people in their city had increased during the preceding year while housing opportunities had worsened. Staff in both cities tended to believe that opportunities for the homeless to receive social service and financial benefits, as well as opportunities for receiving physical and mental health care had remained "about the same" over the past year. Many staff in both cities agreed that publicity concerning the plight of the homeless had improved. (Figure 8)

Shelter directors were also asked whether their shelter was different than when it began and whether it would be different two years in the future: seven felt their shelter was different than when it began and seven felt their shelter would be different in two years.

When service agency representatives were asked the same questions, they indicated somewhat more change: thirteen of fourteen answering said their agency was different than when it first started and eight thought the agency would be different in two years.

Agency representatives believed that government should play the major role in solving the homeless problem. Forty percent believed that the federal government should have primary responsibility, while an equal percent believed some combination of the federal and local government and the private sector should primarily be responsible.

Summary

Survey respondents' feelings about their communities were very positive, but many perceived a worsening of the problem of homelessness even though service agencies seemed to be functioning as well as in the preceding year. Many staff reported changes in their shelters, but judging from their satisfaction with their shelters, this change likely was for the better. The federal government was perceived by many as having primary responsibility for responding to the problem of homelessness.

Working in Shelters

Most shelters began only recently; there are not set requirements for shelter jobs and debate continues over the most appropriate background for shelter staff. As a result, staff characteristics and staff responsibilities vary markedly between different types of shelters.

The Staff

The median age of Somerville staff was the '40s; Cambridge staff were younger, with a median age of 30-39. Just under one-third of the staff in both Somerville and Cambridge had a high school degree or less education; forty percent of the Somerville staff and almost half of the Cambridge staff had completed college (Table 4).

Table 4

Highest Grade In School Completed

Som Cam

1 to 8 Years 0% 3%

Some High School 8 5

High School Degree 23 21

Some College 23 32

College Degree 15 18

Some Graduate Work 23 18

Graduate Degree 8 3

100% 100%

(13) (38)

About half of the staff had received additional training at their shelter. Additional training at other agencies was also common in Somerville, but not in Cambridge (Table 5).

Table 5

Special Additional Training

Som Cam

At The Shelter (1) 42% 55%

At Other Agencies (2) 50 6

In Special Courses (3) 0 12

Other (4) 8 12

1 and 2 0 6

1, 2 and 3 0 3

1 and 3 0 3

1, 3 and 4 0 3

100% 100%

(12) (33)

The majority of shelter staff in both Cambridge and Somerville were women, although the proportion female was higher in Somerville (Table 6). While about two-thirds of the Somerville staff were Roman Catholic, less than one-third of the Cambridge staff were. Neither town included more than a few veterans among their shelter staff.

Table 6

Gender, Religion and Veteran Status

Som Cam

Female 69% 56%

Roman Catholic 67% 28%

Veterans 8% 5%

(13) (38)

About two in five staff in both towns were currently married, but Cambridge reported a much higher proportion of single staff (56) than did Somerville (23) (Table 7). More than half of the Somerville staff reported being divorced or separated or living together outside of marriage.

Table 7

Marital Status

Som Cam

Married 23% 21%

Divorced/Separated 31 21

Living Together 23 3

Single 23 56

100% 101%*

(13) (39)

*Sum does not equal 100 due to rounding error.

Somerville shelter staff were more active politically than those in Cambridge, although a majority of staff in both towns had voted in recent congressional, local and federal elections (Table 8).

Table 8

Voting in Elections

Som Cam

1986 Congressional Elections 83%(10) 51%(19)

Local Elections in Last 4 Years 92%(12) 66%(25)

1988 Presidential Election 77%(10) 66%(25)

Cambridge shelters used a wider variety of employment options with their staff than did Somerville shelters, although the differences largely is due to one Cambridge shelter operated entirely by volunteers (Tables 9, 10).

Table 9

Employment Status

Som Cam

Paid, Full Time 60% 44%

Paid, Part Time (<30 hrs) 40 39

Volunteer (Not surveyed in Som.) 0 18

100% 101%*

(15) (39)

*Sum does not equal 100 due to rounding error.

TABLE 10

Payment for Work

Som Cam

Paid by the Shelter 71% 62%

Paid by Another 21 18

Not paid 7 21

99%* 101%*

(14) (39)

*Sum does not equal 100 due to rounding error.

Half of the Somerville shelter staff belonged to a professional association compared to only 20% of the Cambridge staff. Membership in religious groups was twice as high for the Cambridge staff (40%) as for the Somerville staff (20%). About one-third of all staff belonged to some type of coalition for the homeless or to another advocacy group; half as many as that were members of political groups concerned with homelessness. Only a little more than 5% of the Cambridge staff belonged to a trade union and none of the staff from Somerville did. (Figure 9)

Half of the shelter directors reported difficulty in recruiting staff and shelter staff tended to have little seniority. In one third of the shelters, no more than one-third of the staff had been working for more than one year; only in one-quarter of the shelters had as many as half of the staff been working in the shelter for more than one year.

Shelter staff in Somerville differed from their Cambridge counterparts in several respects: they were more likely to be women, Catholic, older, not single, voters and full time employees. These differences could result in different reactions to their work and their clients.

The Work

Shelter staff engage in a wide array of activities. The most common activities include answering phones and completing paperwork. Somerville and Cambridge staff differed in the frequency with which they engaged in some activities. Somerville staff answered phones more frequently and engaged in more paperwork, counseling/therapy, training sessions and training of guests, while Cambridge staff were more often engaged in crisis management, contacting agencies, assessing clients and, to a lesser degree, outreach/advocacy and case consultation. (Figure 10)

Specific shelters varied in the frequency with which staff engaged in four types of activities (Chart 14). All shelters reported a moderate level of direct service activity and somewhat lower levels of networking. Administrative work, including answering phones and paperwork was the most common activity in most shelters, but some shelters had markedly less administrative activity. Staff development activities were moderately frequent in most shelters.

Chart 14

Overall Frequency of Activities,

Aggregate Staff Responses

Service Network Admin Devel

Often 7│

6│ AEF

5│ BDG E

4│EFH FH CH ACG

3│ABCDGI CEG I BDFHI

2│ ABDI

Never 1│

Staff also participated in other activities on behalf of the homeless outside of the shelter. Many staff reported reading about the homeless, attending staff meetings, checking on guests' welfare and donating to the homeless. Other extra-work activities such as taking work home or working for a political campaign were less common. (Figure 11)

Somerville staff reported more frequent participation in reading about the homeless and campaigning for the homeless, compared to Cambridge. Cambridge staff reported working through lunch breaks quite often, more than did the Somerville staff, and taking work home more frequently. (Figure 11)

The average level of participation in extra-work activities was fairly similar between shelters, with all shelters clustered between 3 and 5 on the seven-point scale (Chart 15).

Chart 15

Overall Participation in Extra-Work Activities,

Aggregate Staff Responses

Very Often 7│

6│

5│AFH

4│BCEG

3│DI

2│

Never 1│

The shelter workload was seen as moderately or very heavy by over half the staff in both Cambridge and Somerville (Table 11). Work demands were perceived as conflicting very often by about one in five staff but as conflicting at least sometimes by over two-thirds of the staff (Table 12). Shelter staff in Cambridge viewed work demands as conflicting more often than did Somerville staff.

Table 11

Current Work Load

Som Cam

Very Light 7% 3%

Moderately Light 40 35

Moderately Heavy 47 49

Very Heavy 7 14

101%* 101%*

(15) (37)

*Sum does not equal 100 due to rounding error.

Table 12

Frequency of Conflicting Demands

Som Cam

Very Often 13% 19%

Sometimes 53 60

Not Often 33 22

99%* 101%*

(15) (37)

*Sum does not equal 100 due to rounding error.

Jobs performed at the shelters were many and complex. Staff from both Somerville and Cambridge reported their jobs "somewhat" allowed them to do a variety of things, allowed them to make a lot of decisions on their own and required them to be creative. To a lesser extent, staff judged their jobs "a little" to "somewhat" required them to do the same things over and over, to work hard and at a high level of skill, and to keep learning new things. Cambridge staff felt their jobs required them to learn new things and be creative more than did the Somerville staff. (Figure 12)

There was little variation in job complexity between shelters (Chart 16). Staff in all but two shelters reported, on average, "somewhat" complex job requirements.

Chart 16

Complexity of Job Requirements, by Shelter,

Aggregate Staff Responses

A Lot 4│

Somewhat 3│ACDEGHI

A Little 2│BF

Not At All 1│

Staff identified a need for additional training when dealing with each of the problems found among homeless persons. Cambridge staff were more in agreement about the need for such training in each area; the discrepancy was greatest in the areas of family counseling and financial benefits--Cambridge staff thought training in these areas was needed a lot, while Somerville staff perceived the need as only moderate. (Figure 13) Perceived need for training did not vary markedly between shelters (Chart 17).

Chart 17

Training Needs, by Shelter,

Aggregate Staff Responses

High 7│

6│ C

5│ EFGHI

4│ ABD

3│

2│

Low 1│

Staff in both Somerville and Cambridge described their jobs as being relatively clearly defined. Staff "often" got enough facts and information to work at their best, reported having procedures for dealing with "whatever situation arises", and generally agreed that each employee has a specific job to do. Only "seldom" did people make their own rules on the job. (Figure 14)

Job specificity was relatively uniform across the shelters; staff in most shelters reported that their jobs "often" were defined specifically (Chart 18).

Chart 18

Job Specificity,

Aggregate Staff Responses

Specific

Always 5│G

Often 4│ABCDEHI

Sometimes 3│F

Seldom 2│

Never 1│

As indicated in responses reported earlier, shelter staff spent more time providing services directly than in making referrals. "Bureaucratic" activities were the most common activity in most shelters, although shelters varied greatly in this respect. Cambridge shelter staff seemed to face a somewhat heavier workload than did Somerville staff: Cambridge staff reported a heavier work load, more conflicting demands, were more likely to work through lunch and to take work home, and were required to be more creative and to learn more new things on the job. Shelter staff in both towns, but particularly in Cambridge, felt a great need for additional training.

Control

Seven of the twelve shelters had a board of directors, while two used a staff committee and one relied on an advisory board for governance decisions. Three had an executive committee in addition to a board. Boards of directors ranged in size from four to forty; the median size was thirteen. The boards represented a diverse constituency: two-thirds included a representative of a church group and almost as many included a representative of local residents; state and local government, charities, guests, staff and service agencies were each represented on two or three boards (one-quarter or one-third).

Two-thirds of the boards met at least monthly. More than half had changed their composition in the preceding year. Four shelter directors reported that an interagency committee or coalition had been established to coordinate activities between their shelter and other agencies. However, in three-quarters of the shelters, the shelter director had ultimate responsibility for operations on a day-to-day basis; the rest of the shelters had no central authority, operating as staff-run collectives.

Responsibility for making decisions varied with the type of issue involved. Decisions about admission policies, shelter rules and procedures were made by the director and staff in two-thirds of the shelters. No one group or combination of groups--the board, the director, and/or a staff advisory group--was more often responsible than any other group for changing the services offered by the shelter, but the shelter director together with a staff group was most often responsible for making decisions about shelter referral policy.

Responsibility for decisions involving shelter administration and direction tended to be more centralized. The shelter board was most often responsible for evaluating the director, for hiring and firing the director and for board appointments. The director alone tended to make decisions about hiring staff. The board and director, alone or in combination, made decisions about long-term plans, about relations with other agencies, about large purchases, fund-raising, sponsor solicitations, the operating budget and budget allocation in almost all the shelters; staff rarely had a role in these decisions.

All of the shelter directors reported regular staff meetings. Half of the shelters had monthly meetings; meetings in the rest were more frequent. Votes were taken on policy and/or procedural decisions in staff meetings at about half of the shelters. Current problems and information about homelessness were discussed in staff meetings at almost all of the shelters.

Participation in decision-making was higher for the Cambridge staff than for the Somerville staff. In Cambridge, staff reported that they at least "sometimes" participated in decisions to adopt new programs and policies whereas Somerville staff "seldom" did. Generally, frequency of participation of staff was low when deciding upon hirings or promotions but was still higher for the Cambridge staff. (Figure 14)

Participation in decision-making varied widely between the shelters, staff in two shelters reporting that they "often" participated in decisions about running the shelter and staff in two other shelters indicating that they "never" did so (Chart 19).

Chart 19

Participation in Decisions,

Aggregate Staff Responses

Decide

Always 5│

Often 4│CI

Sometimes 3│D

Seldom 2│AEGH

Never 1│BF

The role of the board, the shelter director and the shelter staff thus varied between issues. The directors were most important in making most decisions, but the level of staff participation in decisions varied widely between shelters and issues.

Reactions to Work

About sixty percent of shelter staff in both towns were very satisfied with their jobs overall (Table 14). This high level of satisfaction also appeared in response to more specific questions about the job. More than two-thirds would recommend their shelter job strongly to a friend and would do it over again with no hesitation. Sentiment was only slightly less positive in terms of considering the current job as ideal and living very much up to initial expectations.

Cambridge staff were somewhat more likely to recommend their shelter job to a friend, to take the job again, and to report that the job was up to their expectations. Of course, these differences could reflect the different backgrounds of the staff in these towns as well as differences in their shelter jobs. In spite of these high levels of job satisfaction, half the staff in both towns were at least moderately likely to leave their job within one year.

Table 14

Job Satisfaction and Commitment

Som Cam

Very satisfied with job 60%(9) 64%(25)

Strongly recommend to friend 67%(10) 76%(28)

Do over, no hesitation 67%(10) 95%(37)

Current job is an ideal job 64%(9) 54%(19)

Job very much up to expectations 60%(9) 78%(28)

Moderately likely to leave-1 year 50%(7) 47%(17)

Overall job satisfaction was high across all the shelters, with only three of the nine shelters having an average level of overall job satisfaction even a little below "high." (Chart 20).

Chart 20

Overall Job Satisfaction,

Aggregate Staff Responses

High 1│ADEGHI

Medium 2│BCF

Low 3│

Satisfaction with specific job facets also was high among staff in both Somerville and Cambridge shelters. The staff responses approached being "very satisfied" when questioned about the work itself, their supervisor and their coworkers. Satisfaction with promotion opportunities and salary was lower--between "moderately satisfied" and "not very satisfied," on average. Somerville staff were slightly more satisfied with their salary and with their coworkers while Cambridge staff reported being slightly more satisfied with the work itself and with promotion opportunities. (Figure 15)

Staff satisfaction with specific job aspects varied little between the shelters. Most shelters reported a moderate level of job satisfaction with most job aspects; in the case of salary and promotion opportunities, some were less satisfied, while in the case of coworkers, the work itself and supervisors, some were more satisfied (Chart 21).

Chart 21

Satisfaction with Job Aspects, by Shelter,

Aggregate Staff Responses

Salary Cowrkr Work Super Promos

Not at all sat'd 4│ F

Not very satisfied 3│BCEF F ABH

Moderately sat'd 2│ADGH CDEF ABCEF BC CDE

Very satisfied 1│I ABGHI DGHI ADEGHIGI

Staff reported generally getting along with each other and rarely being confused about who is responsible for what at the shelter. Staff generally disagreed that those with college degrees get along better with each other than with other staff. They also disagreed with the statement, especially at Cambridge shelters, that those staff without college degrees get along better with the homeless than do staff with college degrees.

Staff were not as sure about other aspects of inter-staff relations. For instance, staff responses fell, on average, between "undecided" and "disagree" when asked whether staff should consult with professionals more often before making decisions and whether staff with college degrees see things differently than do other staff. Responses also indicated staff were "undecided" about whether there is often disagreement about the best procedures for particular guests and whether they feel the staff should bar unruly guests from the shelter more often. (Figure 16)

Perceptions of inter-staff relations and orientations toward barring unruly guests varied widely between shelters. While staff in each shelter tended to agree or strongly agree that staff generally get along at the shelter, average shelter staff opinions about barring unruly guests ranged from agreement to strong disagreement. Staff perceptions that degreed staff tend to see things differently than staff without degrees also varied as widely (Chart 22).

Chart 22

Inter-Staff Relations, by Shelter,

Aggregate Staff Responses

Staff Bar Degreed

Gnrly Unruly Staff See

Get Along Guests Diffrntly

Strongly Disagree5│ H

Disagree 4│ DFI AD

Undecided 3│ ACE BEGHI

Agree 2│BCDEFH BG CF

Strongly Agree 1│AGI

Most of the staff in both cities also had positive feelings about their shelter as a place to work. On a scale of 1 to 7, where 1 represents "strongly agree," the responses fell, on average, between 1 and 2 when staff were asked if they cared about the fate of their shelter. The staff also agreed that they were proud to tell others that they were a part of their shelters, that their shelters inspired the very best in them in the way of job performance, and that they were willing to put in extra effort to help the shelter be successful. Most staff also indicated that they found that their values and their shelter's values were very similar. (Figure 17)

There was little variation in these feelings between shelters (Chart 23).

Chart 23

Feelings About the Shelter as a Place to Work, by Shelter,

Aggregate Staff Responses

Strongly disagree 7│

6│

5│

4│F

3│ACE

2│BDGHI

Strongly agree 1│

Summary

Shelter staff in Cambridge and Somerville were very satisfied with their jobs, with inter-staff relations and with their shelters. Feelings were positive particularly concerning the work itself, coworkers and supervisors, while there was some dissatisfaction with salaries and promotion opportunities in most of the shelters. Overall, job satisfaction was somewhat greater in the Cambridge shelters.

Feelings about the service network were less positive. Inadequate resources and service gaps were the major complaints. In addition, most shelter staff were interested in additional training in each of the service areas.

The indicators of dissatisfaction with the service network must be balanced against the high levels of staff satisfaction with their own jobs and shelters, but they cannot be discounted. About half of the shelter staff were thinking of leaving within the next year. It is unlikely that shelters will be able to develop staff with adequate levels of training if such a high turnover rate cannot be lessened.

CONCLUSIONS

In spite of their geographic proximity, Cambridge and Somerville had very different shelter systems. While Somerville had only two small regular shelters, Cambridge had eight, several of which provided more than 50 beds. Somerville screened guests before intake, while some Cambridge shelters were relatively open; one in particular drew almost all of the city's alcoholic shelter users. These differences help to explain some of the variation in service approach and staff satisfaction between the two cities.

Staff perceived alcoholism as common in some shelters, but mental illness as relatively infrequent. Most staff believed that homeless persons need social services, but there was no consensus about the value of professionals as shelter employees. Few shelters were equipped to provide help with severe mental or physical health problems, substance abuse or family problems; in fact, staff--particularly those working in Cambridge--were interested in additional training in each service area. This low level of services available directly in the shelters was not as consequential for alcoholics in Cambridge, since most used a shelter that specialized in their problems. However, most homeless persons would have had to rely on agencies outside of the shelters for assistance with personal disabilities.

There were several indications that improvements are needed in the service network. Although agency-shelter relations were reported as being cooperative, rates of referral to agencies were relatively low, and both staff and agency representatives reported service gaps and inadequate resources for some problems, with somewhat greater frequency in Cambridge. The homelessness problem itself was perceived as worsening in both communities, but somewhat more in Cambridge.

With respect to their own shelters, shelter staff were very positive; only salaries and promotion opportunities were the cause of some dissatisfaction. Nonetheless, many staff were contemplating leaving their shelter jobs within the next year.

[1]The list of services used in the shelter director and agency interviews was much more detailed than the list included in the staff questionnaire, so responses are not precisely comparable.