Who Belongs?: An Analysis of Ex-Mental Patients’ Subjective Involvement in the Neighborhood

Abstract

What causes people with psychiatric disabilities to feel they belong in their neighborhood? This article examines predictors of belonging for a sample of former psychiatric patients in community settings. The authors consider what differentiates ex-patients who feel they belong in their neighborhoods from those who do not. For the sample as a whole, belonging primarily results from satisfaction with the dwelling. It depends neither on reception by neighbors nor on whether they live in sheltered care. Furthermore, there is nothing about sheltered care (i.e., a supervised residence) that makes people feel less belonging. For long-term sheltered care residents, belonging depends on neighbor relations and ease of arranging activities with house residents.

Deinstitutionalization’s failings have been amply documented. Mental patients were never given the community support and acceptance necessary to their full reintegration into the larger community (Dear & Wolch, 1987; Nirje, 1976; Wolfensberger, 1972). As noted by Bloom (1977), deinstitutionalization was grounded in a naive notion of community openness. Instead of being easily accepted, many ex-patients were ghettoized, residing in single room occupancy hotels, nursing facilities, and board and care residences or on the street in poor, inner-city areas. Much of the segregation of ex-mental patients resulted from their low levels of housing and income support. However, community resistance is also to blame. Surveys show the general population holds negative attitudes about ex-mental patients and newspaper and other reports document the frequency of neighborhoods organizing when sheltered care facilities attempt to open (Link, Cullen, Frank, & Wozniak, 1987; Segal, 1987; Baron & Piasecki, 1981;Rabkin, 1980; Piasecki, 1975).

While much attention has been paid to neighborhood resistance to ex-mental patients, less notice has been made of neighborhood acceptance. Yet, Sherman, Newman, and Frenkel (1984) find that 87 percent of ex-mental patients in their sample of foster family care homes feel very much at home in their neighborhoods. Here, the authors look at what differentiates ex-patients who feel they belong in their neighborhoods from those who do not.

We look at a mentally disabled population that had been residing in sheltered care in 1973, but in 1985 lived in a variety of community residential settings. We look at the extent to which ex-patients feel they “belong” in their neighborhoods and at the situational and affective explanations for why they differ in their feelings of belonging.

There is one caveat to our analysis. Any discussion of neighborhood attachment risks assuming that it is a desirable goal—either that people wish to feel they belong in the neighborhood or that it contributes to their well-being. Certainly, it is assumed that neighborhood attachment is a desired goal in its own right for ex-mental patients and that it helps “normalize” them and to live normal lives. These normative expectations surrounding neighborhood involvement are part and parcel of the longstanding mystification surrounding the notion of community.

People’s supportive networks are rarely based in the neighborhood (Fischer, 1982). Because of changes in communication and transportation technology, community has been freed from a territorial basis—people can select ties based on preferences rather than geographic proximity. People can also vary in the extent to which feeling they “belong” in their neighborhood is important to them (Merton, 1968). In corroboration, surveys of the general population show that a non-trivial number of people have no wish to be involved with neighbors (Fischer, 1982).

A survey of ex-patients shows that they prefer to live in “regular” housing rather than other settings (Tanzman, 1993; see also Carling, 1992). However, we have little information on their desires for social integration in the larger neighborhood. Indeed, some people in this sample said that the best thing about their neighborhood was the fact that people left them alone. None-the-less, when a group is stigmatized, a feeling of local acceptance and thus belonging may assume greater importance than it would for people whose potential acceptance is secure. Certainly, Tefft, Segall, and Trute (1991) consider “belonging” to be one criterion of success of residential integration.

Belonging and Local Community Involvement

What does it mean to “belong” in a neighborhood? Does it mean that a person has a pleasant, yet more than nodding acquaintance with neighbors, uses local facilities or simply finds it a pleasant environment? For this population, which may experience rejection, does it mean they live in a place where neighbors at least accept them? Or does it mean no more than satisfaction with a housing arrangement?

Lacking a well-developed literature in the area, we consider two possibly alternative, possibly complementary, sets of hypotheses. We hypothesize that an ex-patient’s feelings of belonging will be conditioned by experiences in the neighborhood. Experiences include both actual involvement and subjective evaluation. In terms of actual involvement, we look at use of neighborhood facilities and services and involvement with neighbors. In addition, we look at the ex-patient’s satisfaction with his or her dwelling and the extent that activities with other house residents create a feeling of belonging within the house.

Objective Involvement in the Neighborhood

Feelings of belonging, we hypothesize, will depend upon the extent to which the neighbors accept the ex-patient by casually neighboring—chatting on the street—or forming friendships. We further look at the importance of situations commonly thought to generate neighbor hostility. We hypothesize that residents who live in sheltered care, in contrast to those living in other forms of community housing, should feel a lower sense of belonging. They live in a situation where their disability is more readily known and where the behavior of other facility residents may reflect negatively upon them, thus making rejection by neighbors more likely. This should be particularly true when the neighbors directly complain about the facility. However, as noted by Dear and Wolch (1987), people in neighborhoods of mixed residential types or mixed residential and commercial uses should object less to the presence of sheltered care facilities in their midst. Therefore, residents in neighborhoods consisting of any use other than simple single family homes should experience less rejection and feel an accompanying higher amount of belonging.

We also hypothesize that individuals who make greater use of neighborhood services and amenities should feel more belonging in that neighborhood as should those who have lived there longer. In both cases, they are more familiar with the area and have more attachments to it. Also, those who do not feel they belong might move if they have the capacity to do so.

Subjective Evaluation of Neighborhood

Irrespective of actual behavior in the neighborhood, people create subjective evaluations of the acceptability of the neighborhood and use that in constructing their feelings of belonging. Thus, people who are more satisfied with individual aspects of their environment such as safety and appearance will feel more belonging.

The Dwelling: Objective Involvement and Subjective Evaluation

As an alternative or complimentary hypothesis, we look at satisfaction with dwelling and activities that foster a sense of belonging in it. People living in supervised arrangements can have activities planned that foster a sense of engagement and thus belonging within their house. We hypothesize that residents in houses that encourage activities among the residents will feel greater belonging. We further hypothesize that satisfaction with the dwelling will affect sense of belonging.

Disability

It is not clear whether the respondent’s level of disability affects belonging independently or through its importance for the type of residence obtained, the ability to use local services, the amount of neighboring and the like. Therefore, we include the amount of disability in our models to see if it has an independent effect on belonging, once the other variables are considered.

Methods

Sample

We analyzed data gathered in a 1973 longitudinal study of residents who had been living in sheltered care facilities. Facilities included in the sampling frame included all family care, board and care, and halfway houses in California serving at least one resident with a mental disability.

The original sample was obtained by dividing the state into three master strata: Los Angeles County, the nine Bay Area counties, and all other counties. A sample of facilities was drawn from Los Angeles and the Bay Area with probability proportionate to bed space. Two additional counties were randomly selected from the north of the state and two from the south, again with probabilities proportionate to total capacity and a sample of facilities drawn within each county. Finally, residents were randomly selected from facilities. For 1973, the final sample of 397 residents was representative of all 18- to 65-year-old formally psychiatrically hospitalized sheltered care residents.

Between 1983 and 1985, Segal and his research team attempted to contact all members of the 1973 sample. Of these, 90 were found to have died. An additional 33 individuals could not be located. The sample analyzed here includes 195 respondents. These are people, not living in institutions or nursing homes in 1985, who answered the questions used to construct the dependent variable. Of these, 55 were living in the same sheltered care facility as in 1973. Additionally, the operators of all facilities that were still open in 1983 were interviewed about their facilities. We therefore can only analyze operator-provided information for those 55 residents who lived in the same facility as in 1973. Yet, this data potentially can add to our understanding since the operator can provide independent information on neighbor receptivity and neighborhood characteristics. To take advantage of this, we present data for both the entire sample and then the subsample of people living in the same sheltered care facility as in 1973.

Analysis

We use logistic regression to look at the predictors of belonging in 1985 for the follow-up sample. The first set of models uses the entire sample; the second includes only those who lived in the same sheltered care facility in 1973 and 1985. All variables come from the 1985 interviews. We cross-sectionally evaluate the relative contribution of neighborhood and residential factors in predicting belonging. We then recompute the models, taking account of the level of psychological disability.

Measures

(Note: standardized measures were used for all non-dichotomous data)

1. Belonging. — Respondents were asked, “Do you feel that you really belong to this neighborhood, that you are a part of it?” Only 8 percent felt they definitely did not belong in their neighborhoods; an additional 19 percent felt they did not belong and 12 percent were not sure. In contrast, 35 percent said yes to the question and 26 percent definitely yes. Thus, 61 percent of the sample felt they belonged in their neighborhood. To simplify analysis, the variable was collapsed to compare the 61 percent to the rest of the sample.

2. Neighborhood Reaction.—We measure both neighbor acceptance and indicators of neighbor concern.

   1. Neighbor acceptance created from measures of (a) chatted with neighbors, (b) had more than casual discussions with neighbors, and (c) been invited over to a neighbor’s house for dinner. One point was added to the index for each behavior performed at least sometimes. Forty-five percent of the residents had a score of zero and 17 percent a score of three. The alpha for the index was .73.
   2. Possible neighbor concern:

      1. Whether the resident lived in a half-way house, board and care home or group living arrangement or not.
      2. (Only for those in long-term sheltered care) whether the operator reported there had been any complaints about the facility by neighbors.
      3. (Only for those in long-term sheltered care) whether the neighborhood was primarily single family dwellings or not.

3. Use of Neighborhood Facilities. — How often residents made use of local facilities including: a local restaurant or coffee shop, local shopping area or store as well as the amount of time the respondent spent outside the house. An additive, normed, index was created from these items. The alpha was .72. [This is a subindex of Segal’s External Social Integration scale (cf. Segal & Aviram, 1978)].

4. Satisfaction with the Neighborhood. — The mean of the responses to seven items evaluating satisfaction with the neighborhood including: safety, appearance, public transportation, recreational activities, restaurants and shopping. The alpha for the index is .94.

5. Time in Neighborhood. —The number of years they had lived in the present dwelling.

6. Satisfaction with Dwelling.—The mean of 15 items evaluating various aspects of the resident’s housing environment including: general adequacy, comfort, food, rules, cleanliness, privacy, safety, helpfulness, recreational activities, friends, satisfaction of needs, and satisfaction with influence and amount of participation. It should be noted that some items are not applicable to people who live alone and potentially not applicable to those not living in sheltered care. The mean was only taken for those items answered by the respondent so the score was not penalized for the inappropriateness of certain items. The alpha is .97.

7. Possible Belonging in the House.— People who lived with others might feel belonging because the housing environment was structured to permit participation. Residents were asked how easy it was to arrange social activities or trips to sport events with other house residents as well as vocational training, religious services and therapy at the house. (Alpha equals .72) [This is a subindex of Segal’s Internal Social Integration scale (cf. Segal & Aviram, 1978).] Since the scale is really only applicable to people living in sheltered care, it will be considered only for that subsample.

8. The extent of the resident’s disability was measured by the Brief Psychiatric Symptom Rating Scale (BPRS) (Overall & Gorham, 1962).

Results

The unstandardized means and correlations of the items included in the analysis with belonging are shown in Table 1. It should be noted that the sheltered care subsample is no more or less satisfied with their individual residences than the entire sample and only very slightly, albeit nonsignificantly, less satisfied with the neighborhood (Means comparing this subsample to the rest of the sample on these variables are non-significant).

Table 1.

Means/Range of Independent Variables and Bivariate Correlation with Belonging

	Entire sample (N = 195)				Those who in 1986 were in same sheltered care facility as in 1973 (N = 55)
Variable	Range (lo/hi)	Mean (SD)	Corr.	p	Range (lo/hi)	Mean (SD)	Corr.	p
ZBPRS	−1.4/5.03	−0.245 (1.099)	−0.2296	.001	−1.4/2.25	−0.236 (0.988)	0.0121	n.s.
Neighboring	0/3	1.058 (1.151)	0.1708	.017	0/3	1.255 (1.158)	0.3250	.026
Complaints about facility					0/2	1.698 (0.463)	0.2938	.048
Single Family Homes					0/1	0.302 (0.463)	−0.1148	n.s.
Use of nbhd facilities	−2.2/3.25	0.34 (1.092)	−0.0237	n.s.	−2.2/2.19	−0.252 (1.167)	0.0071	n.s.
Ease arranging activities with house residents					−2.3/2.99	1.06 (1.324)	0.2165	n.s.
Satisfaction with residence	1.07/3	2.493 (0.438)	0.2995	.000	1.27/3	2.481 (0.431)	0.1578	n.s.
Satisfaction with nbhd	1.17/3	2.329 (0.440)	0.2214	.002	1.67/3	2.383 (0.402)	0.0702	n.s.
Time in nbhd	0/10	5.191 (3.885)	0.1750	.018
Time in sheltered care	0/1	0.309 (0.463)	0.0659	n.s.
Belonging	0/1	0.61 (0.49)			0/1	0.73 (0.45)

Table 2 shows the four logistic regression models, two for the entire sample and two for the subsample presently living in the same sheltered care residence as in 1973. Each equation is computed twice—once without symptomatology and once with. Since few of those living in the same sheltered care facility had not been living in their residence continuously for 10 years, time in residence was not considered for inclusion in the latter groups of models. The data show that, although living in sheltered care does not significantly predict belonging, systematically different factors distinguish the long-term sheltered care residents from the entire sample. For the ex-patient population as a whole, feelings of belonging stem from their satisfaction with their individual dwellings and the length of time they have lived there. BPRS is also a significant predictor with those evincing fewer symptoms showing a greater degree of belonging. The actual use of neighborhood facilities/services does not predict belonging in the multivariate models. Satisfaction with the neighborhood and neighboring, while showing a moderately strong association in the bivariate associations, are not significant in the multivariate model because of its collinearity with satisfaction with the dwelling. Therefore, ex-patients seemed to be more affected by their satisfaction with their particular dwelling than their objective experiences in, or subjective evaluation of the neighborhood.

Table 2.

Logistic Regression Models Predicting Ex-patients’ Feeling of Belonging

	Entire sample (N = 191)				Those who in 1988 were in same sheltered care facility as in 1973 (N = 55)
Variable	Odds ratio	p	Odds ratio	p	Odds ratio	p	Odds ratio	p
ZBPRS			0.70 (−)	.03			2.72	n.s.
Neighboring	1.38	n.s.	1.32	n.s.	15.25	.05	22.61	.04
Complaints about facility					55.91	.03	59.68	.04
Single family homes					0.02 (−)	.04	0.04 (−)	.03
Use of nbhd facilities	0.76 (−)	n.s.	0.70	n.s.	1.06	n.s.	1.054	n.s.
Ease arranging activities with house residents					4.37	.04	5.90	.03
Satisfaction with residence	2.04	.00	1.86	.01	1.87	n.s.	1.95	n.s.
Satisfaction with nbhd	1.13	n.s.	1.13	n.s.	0.10 (−)	.06	0.10 (−)	n.s.
Time in nbhd	1.49	.03	1.43	.05
In sheltered care	2.09	n.s.	2.26	n.s.
% Correctly classified	71%		69%		86%		88%

In direct contrast, those in long-term sheltered care arrangements were more directly affected by their neighbors and neighborhood. Those with higher neighboring scores and those in facilities that did not garner complaints from neighbors felt more belonging. Similarly those in the mixed residential neighborhoods, hypothesized to be more accepting of the ex-patient, feel a greater sense of belonging. Subjective evaluation of the neighborhood approached significance before level of disability was entered. Objective activities in the house also mattered. Those who could more easily arrange activities with house residents felt more belonging.

BPRS does not show a significant association either the bivariate correlations or in the multivariate model for long-term sheltered care residents. Belonging, for this subsample, is more a function of environmental receptivity than it is of individual disability.

Can the model describing the long-term sheltered care residents be generalized to those who were in sheltered care in 1985 but not in the same facility? Whether the neighborhood contains primarily single family homes and whether neighbors complain is not available for this sample. When the model with the reduced number of variables is run for everyone currently in sheltered care, it resembles the entire sample, not those in long-term placement. It is clear, then, that there is something different about the experiences of those who establish longterm residences.

Conclusion

The results of this study challenge conventional wisdom in theories upon which research is based and in social policy. The absence of association between neighboring or use of neighborhood facilities and services and a sense of belonging in the entire sample should give us pause. It argues that community integration cannot be measured in unidimensional terms; being a part of the community can have many unrelated components.

Another surprising result concerned the absence of an effect of sheltered care residence on a sense of belonging. Sheltered care residence is often seen as disliked by its residents. It is clear from the regression results that residents are evaluating their housing situation more than their neighborhood in forming their opinion. Yet, whether they live in sheltered care does not predict belonging.

The results for the entire sample indicate the importance of acceptable housing and secondarily of a satisfactory neighborhood. The policy implication then is that the important variable for the ex-patient population as a whole is their happiness with their dwelling, not their integration into the local community. Further, there is nothing about sheltered care— supervised residence versus independent living—that makes people feel they do not belong. To the extent trade-offs must be made, efforts to assure adequate residential environments of all types will better benefit ex-patients in respect to their feelings of belonging.

More specific policy recommendations are possible from the model for the long-term residents of sheltered care. They are affected by their larger community and by actions taken by their operator to create a community in the house. Siting of facilities in neighborhoods consisting of uses other than only single family dwellings contributes to a sense of belonging. We should note that our previous work (Segal & Silverman, 1993) has also shown that facilities in such neighborhoods are also more likely to survive. Those who have operators more willing or able to arrange activities with house residents feel a greater sense of belonging as do those who are involved with neighbors. Actions undertaken by the operator to involve residents in the larger community clearly can affect their sense of being a part of the community. We know elsewhere that operators can help foster neighbor involvement by becoming engaged in local affairs (Sherman, Newman, & Frenkel, 1984). Doing so makes their residents feel more at home in the neighborhood.