Labeling Mental Illness: The Effects of Received Services and Perceived Stigma on Life Satisfaction

by Sarah Rosenfield
Labeling Mental Illness: The Effects of Received Services and Perceived Stigma on Life Satisfaction
Sarah Rosenfield
American Sociological Review
Start Page: 
End Page: 
Select license: 
Select License


Sarah Rosenfield

Rutgers University

Labeling theory proponents and the theory's critics have different views of stigma and thus differ on the consequences of labeling for people with men- tal illness. The labeling perspective posits that because of stigma, ofJicial labeling through treatment contact has negative consequences for mental patients. In contrast, critics of labeling theory claim that stigma is relatively inconsequential. Instead, they argue that because labeling results in receiv- ing needed services, it provides significant benefits for mental patients. Thus fa< no study has tested the relative positive and negative effects of labeling. I examine these views by comparing the importance of perceived stigma ver- sus the receipt of services for the quality of life of persons with chronic men- tal illness. Results show that both stigma and services received are signifi- cantly associated with quality of life, but in opposite ways. These findings

have important implications for interventions for mental illness.

Aconflict between proponents of the labeling perspective on mental illness and their critics has continued for several de- cades. Labeling theorists examine mental ill- ness as a form of deviance: The label rather than the behavior per se shapes the fate of mentally ill persons, by creating chronic mental illness or by compromising the life chances of those so labeled (Link 1982, 1987; Link et al. 1987; Link et al. 1989; Palamara, Cullen, and Gersten 1986; Scheff 1966, 1974). In contrast, from what is often called the psychiatric perspective, critics of labeling theory view mental illness as a form of individual pathology. The fate of people

*Direct all correspondence to Sarah Rosenfield, Institute for Health, Health Care Policy, and Aging Research, Rutgers University, 30 College Avenue, New Brunswick, New Jersey 08903 ( This research was supported in part by the Ittleson Foundation and the National Institute of Mental Health MH4478-02. Support was also provided by the Center for the Organization and Financing of Services for Severe and Persistent Mental Illness at Rutgers University. I thank Naomi Gerstel, James Jasper, Allan Horwitz, Mary Clare Lennon, Bruce Link, David Mechanic, and three ASR reviewers for their helpful comments on an earlier version of this paper. I also thank the staff at ASR for their excellent editing.

with mental illness depends primarily on the severity of their illness and their treatment rather than on extra-illness factors, such as labels (Huffine and Clausen 1979; Kirk 1974; Lehman, Possidente, and Hawker 1986; Schwartz, Myers, and Astrachan 1 974).

A pivotal difference between these per- spectives involves the importance of stigma. From a labeling perspective, the stigma at- tached to the illness is a central problem. A psychiatric label sets into action cultural ste- reotypes and negative images about mental illness that are applied to the person by oth- ers and by the person to himself or herself (Link 1987; Link et al. 1987; Link et al. 1989; Thoits 1985). These images devalue those with mental illness and result in dis- crimination-persons who have mental ill- nesses are evaluated as "not quite human" (Goffman 1963:5). Originally, labeling theory held that the expectations attached to the label perpetuate the mental illness (Scheff 1966, 1974). The theory was modi- fied later to claim that the devaluation and discrimination created by the label interfere with a broad range of life areas, including access to social and economic resources and general feelings of well-being (Link et al. 1987; Link 1982, 1987; Link et al. 1989).

Atnerican Sociological Review, 1997, Vol. 62 (August:660-672)

Critics of labeling theory, however, ques- tion the claims of both the original and the modified labeling approaches. Perceptions of stigma among mental patients are seen as subjective and untrustworthy or, at the ex- treme, as distortions resulting from the pa- thology (Crocetti, Spiro, and Siassi 1974). Other people are seen to be reluctant to label and stigmatize those with mental illness (Gove and Fain 1973; Huffine and Clausen 1979). Thus, stigma is deemed by labeling theory critics to be relatively inconsequential for the mentally ill (Gove, 1970, 1975, 1980, 1982; Killian and Bloomburg 1975). In con- trast, critics emphasize that being labeled mentally ill allows people to receive needed treatment. High-quality treatment provides persons suffering from psychiatric disorders with a range of services to improve symp- toms, expand functioning, and enhance well- being (Gove and Fain 1973; Linn 1968).

In sum, the contrasting views of stigma offered by labeling theory and its critics im- ply opposite effects of psychiatric labels: La- beling theorists predict destructive outcomes, while psychiatric theorists claim beneficial results. Past research has found evidence for both positive and negative effects of labeling. However, this evidence comes from indepen- dent bodies of research that reflects the often adversarial relationship between the propo- nents of labeling and those with more tradi- tional psychiatric concerns. Because no em- pirical study has examined the joint effects of services and stigma, the relative impor- tance of each for the mentally ill is unknown (Link and Cullen 1992).

I directly compare the effects of the receipt of services versus perceptions of stigma on the subjective quality of life for people with chronic mental illness. Quality of life is in- creasingly emphasized as an outcome of par- ticular significance for persons with chronic illnesses-illnesses that medicine has a lim- ited ability to cure (Lehman 1983). Subjec- tive quality of life or life satisfaction is a critical component of well-being, and well- being is damaged by stigma according to the modified labeling approach, but is improved by services according to the psychiatric view (Bradburn and Caplovitz 1965; Gurin, Veroff, and Feld 1960; Veroff, Kulka, and Douvan 1981). Thus, along with symptoma- tology and objective quality of life or func- tioning, subjective quality of life is a crucial outcome for both the labeling and psychiat- ric perspectives.

Consistent with Link and Cullen (1992), I argue that labeling has both positive and negative effects on subjective quality of life. The label itself implies stigma; the positive effects are indirect, through the receipt of services. I contend that both stigma and ser- vices are consequential because they shape central aspects of self-concept, which is con- nected to life satisfaction. I test this perspec- tive in a model treatment program for per- sons with chronic mental illness.

A sense of mastery and self-esteem are fun- damental goals that protect and enhance the self and contribute to a general feeling of well-being (Pearlin et al. 1981). Low levels of perceived mastery and self-esteem com- promise subjective quality of life by produc- ing a sense of hopelessness and a tendency to give up in difficult times. These dimen- sions of self-concept prevent individuals from trying for and getting what they want, thus diminishing the chances for life satisfac- tion. Ample research documents the relation- ships between perceptions of mastery and self-esteem and feelings of psychological well-being and subjective quality of life (Pearlin et al. 198 1; Rosenfield 1992).

Modified labeling theory holds that the stigma of mental illness is problematic be- cause it damages mental patients' sense of self-esteem and self-efficacy (Link 1987). The degree to which the stigma of mental ill- ness is incorporated into patients' self-con- cept increases with the likelihood that the ill- ness is long-term, which most chronic mental illnesses are. To the extent that stigma be- comes a master status (Becker 1963) or part of one's self-schema, it affects patients' basic evaluations of themselves. They reorganize their views of themselves and their self- knowledge in terms of what they are not and cannot do rather than what they are and what they can do (Jones et al. 1984). Thus, people with chronic mental illness often experience a profound sense of loss of the characteristics they valued in themselves and of cherished life goals and assumptions. The damage to these assumptions and ideals is seen as the


basis of demoralization (Frank 1973; Link 1987; Sullivan 1941), and by extension, such damage affects subjective quality of life.

Evidence links stigma specifically to self- esteem. Patients and their families indicate that strong perceptions of stigma are associ- ated with low self-esteem (Link 1987; Wahl and Harman 1989). When patients know oth- ers are aware that they are in psychiatric treatment, they perform more poorly, feel less appreciated, and are more anxious in their interactions compared to patients whose labels are concealed (Farina et al. 1971).

In contrast to the effects of stigma, the re- ceipt of certain services positively affects pa- tients' self-conceptions. Feelings of low self- efficacy are a particular problem among people with chronic mental illness, who of- ten feel ruled by forces beyond their control (Lamb 1982; Mendel and Allen 1978; Mosher and Menn 1978). Some researchers propose that the major goal of therapeutic in- tervention is to enhance patients' sense of mastery (Lamb 1982). Some evidence sug- gests that the receipt of services increases perceptions of control. Specifically, services that give patients greater power in terms of status or economic resources (Weber 1946) improve patients' levels of self-efficacy (Rosenfield 1992).

In sum, I argue that stigma and received services have independent and antithetical effects on quality of life through their oppos- ing influences on self-concept. Evidence for this approach comes from separate lines of research on the consequences of stigma and the effects of mental health services.

Tests of modified labeling theory demon- strate the stigmatizing consequences of psy- chiatric labels for a number of outcomes. Comparisons of individuals who are similar in psychopathology but vary in exposure to treatment show that labeling has a negative impact on income and work status. Individu- als who have been in treatment earn lower incomes and are less likely to be employed than those with similar symptoms who have not been in treatment (Link 1982).' A later

Whereas hospital treatment could reduce in- comes and employment rates by removing pa- tients from the labor force, individuals in Link's (1982) study were in outpatient treatment. Thus, the effects are clearly attributable to stigma.

study examined the effect of perceived stigma directly. Patients from several treat- ment groups (first-contact and repeat-contact or chronic patients) and untreated individu- als were matched on psychopathology. Stron- ger perceptions of stigma, defined in terms of devaluation and discrimination, were as- sociated with greater income loss and higher unemployment only for the treated groups (Link 1987). The degree of stigma perceived by labeled persons and their adaptations to the stigma shaped patients' social relations and inhibited the support available from in- dividuals outside their households (Link et al. 1989). Most important, perceived stigma was associated with greater demoralization among treated groups compared to non-labeled individuals (Link 1987).

Evaluations of services show that certain programs influence a range of outcomes, in- cluding the mental status and quality of life of persons with chronic mental illness. These model programs include "Training in Com- munity Living" in Madison, Wisconsin, and the Fountain House program in New York City. Compared to standard care, patients in these programs have fewer rehospitaliza- tion~,are more economically self-sufficient, and have more social relationships. In terms of subjective indicators, they are more satis- fied with their work lives, their living situa- tions, their relationships, and with their lives in general (Beard 1978; Beard, Malamud, and Rossman 1978; New South Wales De- partment of Health 1983; Stein and Test 1976, 1978, 1980; Stein, Test, and Marx 1975). These programs combine traditional psychiatric treatment with psychosocial and vocational rehabilitation, supervision, and guidance; they also provide basic needs like housing, financial support, and medical care. The programs tend to share an approach to treatment that some term "empowerment," an approach that stresses personal rather than professional relationships and minimizes sta- tus differences between patients and staff. Staff encourage patients' strengths and take a partnership approach to treating mental ill- ness. Empowerment is seen as an attempt to combat stigmatizing attitudes about mental patients, at least within the programs.

Recent work indicates that certain of these services improve patients' quality of life. For example, economic services, including voca- tional rehabilitation and financial support, and services providing greater status through empowerment increase life satisfaction. Fur- thermore, these services affect patients' sub- jective quality of life by increasing their per- ceptions of personal control (Rosenfield 1992).

Thus, past research has found both posi- tive and negative consequences of labeling. In simultaneously investigating the impact of stigma relative to services, I build on these findings. I predict that perceived stigma and received services shape the subjective qual- ity of life through their effects on dimensions of self-concept. This hypothesis is examined in a model treatment program for people with chronic mental illness.
Research Strategy

Ideally, to evaluate the relative effects of stigma and received services, the full range of treatment and stigma experiences would be studied, including detailed measures of services provided and a uniform measure of stigma in diverse settings. However, the practical problems of such an investigation are formidable. Model treatment programs offer an opportunity to assess whether re- ceived services and stigma have simulta- neous effects in opposite directions. Based on their clients' low rehospitalization rates and high quality of life, several model pro- grams offer high quality treatment in a con- text specifically designed to combat stigma (Beard 1978; Rosenfield 1992; Stein and Test 1978). Labeling theorists, who gener- ally downplay or ignore the positive effects of treatment, would expect such positive ef- fects to emerge only when contrasting the extremes of treatment, that is, comparing in- dividuals receiving very high quality treat- ment to those receiving very low quality treatment. Similarly, the critics of labeling theory, who downplay the importance of stigma, would expect to find any effects of stigma only when comparing those individu- als who are highly stigmatized to those ex- periencing very low stigma. Thus, if services have positive effects and stigma has negative effects in a sample of participants in a model program, labeling theorists would have to acknowledge positive effects of treatment even when gross variations in treatment quality are not present, and the critics of la- beling would have to acknowledge that stigma has effects even when large variations in stigma are not present. Consequently, sup- port for the predicted outcomes should be convincing to labeling theorists and their critics.

Labeling need not be defined dichoto- mously (i.e., by comparing people who are officially labeled and thus have contact with both services and stigma to those who are not labeled and have contact with neither services nor stigma). Rather, labeling can be conceptualized as a continuum that exposes people to the possibility of stigma and the possibility of services: Patients can experi- ence varying degrees of services and stigma. Indeed, there is substantial variation in per- ceptions of stigma and in the receipt of ser- vices among persons who are labeled (Link 1987; Rosenfield 1992). The services re- ceived as well as perceived stigma vary sub- stantially even within one program. My strat- egy examines the effects of such variations in experiences and perceptions among indi- viduals who are officially labeled.
Research Site

This research was conducted at The Club/ Habilitation Services program, which was funded in 1973 as a model program for people with chronic mental illness. It is a unit of the University of Medicine and Dentistry of New Jersey-Community Mental Health Center in Piscataway, New Jersey. This program offers ongoing, intensive rehabilitative care for mentally ill adults within the psychosocial clubhouse model, based on the Fountain House prototype (Beard 1978; Lenoil 1982).~ The program serves all patients with chronic mental illness in the Community Mental Health Center's catchment area. In an evalua- tion, this program's rehospitalization rate (20 percent over a year period) was comparable to that of effective model programs (Rosenfield 1992).

Individuals are eligible for The Club if they have had psychiatric disorder of two or more years duration with significant impairment in so- cial or vocational functioning.

In a day-program format, The Club offers vocational rehabilitation; supervision and guidance; services providing basic needs; and training in psychosocial skills including daily living skills and social skills. The pro- gram also arranges for psychiatric treatment and provides social contacts and a structure of activities. The Clubhouse model holds an empowerment approach to treatment termed "mutual empowerment." This approach em- phasizes the development of independence, decision-making skills, and individual re- sponsibility as well as personal, supportive relationships between staff and patients. All program activities are by mem- bers and staff working together, and mem- bers participate in most aspects of decision- making within the program.

The data were collected from face-to-face interviews with patients conducted by psy- chiatric social workers and psychiatric nurses who were independent of the program and unaware of the research hypotheses. Inter- views averaged one and one-half hours. The population included all patients active in The Club program for at least one entire month over a one-year period (1988-1989). The to- tal number of respondents in the sample was 157, or 93 percent of the eligible patients.

Subjective quality of life is assessed using Lehman's measure of life quality developed for people with chronic mental illness (Lehman 1983; Lehman et al. 1986; Lehman, Ward, and Linn 1982). Two items ask pa- tients to rate their levels of satisfaction with life as a whole. Patients are also asked about their satisfaction with specific areas of their lives, including their living arrangements, family relations, social relations, leisure ac- tivities, financial situations, employment sta- tus, safety, and health. The scales combining these items-as well as measures described below-are discussed more fully elsewhere (Rosenfield 1992; Rosenfield and Neese- Todd 1993). The reliability of each scale, us- ing alpha coefficients, is .85 or above. I present results for overall satisfaction with life, noting consistencies or discrepancies with results for the specific life areas.

To measure received services, patients were asked whether they received specific


services in the last month, including psychi- atric services, services for daily living skills, social skills, vocational rehabilitation. struc- tured activities, social contacts, and basic needs. For example, vocational rehabilitation services include learning work responsibili- ties, expectations, and skills. Services for de- veloping a structure of activities involve help from staff with planning time on week nights, weekends, and holidays. For most of the scales measuring program services, the reliabilities are .60 or above. Exceptions are services for daily living skills (.57), for plan- ning time (.56), and for social skills (.49). ~or~these

areas, I analyze separate items.

Mutual empowerment is measured using a scale of 21 items focusing on patients' deci- sion-making power and supportive interac- tions. For example, patients were asked how much influence they felt they had in the pro- gram, how much their opinions and ideas counted, how much they felt cared about, and how much the staff accepted them the way they are (alpha = .82).

Stigma is measured using Link's (1987, 1989) scale of devaluation-discrimination (alpha = .88). This scale consists of 12 items asking respondents' opinions about the ex- tent to which most people would accept a former mental patient as a friend, teacher, worker, or care-taker of their children. It also covers whether former patients are seen as less trustworthy or less intelligent than oth- ers. This scale focuses on perceptions of stigma rather than stigmatizing experiences, thus allowing assessments across labeled and unlabeled groups. According to modified la- beling theory, this assessment of stigma should have effects because labeled individu- als-and only labeled individuals-apply perceived stigma to themselves, as evidenced by the damage to a range of life areas found among labeled persons but not among unla- beled persons (Link 1987). Furthermore, ask- ing about stigma in this general way avoids the pain of recounting personal experiences thatcould deter candid reporting.

Self-esteem is measured by Rosenberg's (1984) scale, which surveys feelings of worthlessness, uselessness, and failure (alpha = .82).

Mastery is measured using the scale devel- oped by Pearlin et al. (1981), which covers feelings of helplessness, control over forces affecting oneself, and the ability to change or to solve problems (alpha = .73).

In the analyses, I control for race, years of education, age, and sex. I included one item to control for the influence of recent events by asking whether anything happened in the last year that upset the respondent a lot or made him or her very unhappy. I adjust for clinical characteristics using scales based on staff-reported functioning levels in areas cor- responding to the services offered (e.g., daily living skills, social skills, etc. [alphas 2 .go]), and staff-reported symptomatology (covering hallucinations, delusions, paranoia, depres- sion, social isolation, lack of motivation, and fear of success [alphas 2 .90]).

Hypotheses are tested using regression analysis. In examining the importance of stigma for quality of life, background and clinical characteristics are entered into the equation first; stigma is entered in the sec- ond step. Received services are investigated in a similar manner: Control variables are en- tered in the first step, and services are en- tered second. I note that the correlations be- tween these independent variables and the control variables are relatively low (r < .25). For the comparative test, measures of stigma and received services are entered simulta- neously in the second step. To examine self- concept as an explanatory mechanism, mea- sures of self-esteem and mastery are entered in the final stage of the analysis.
The Relationship of Perceived Stigma and Received Services to Quality of Life

How much stigma do patients perceive? A majority of patients (65 percent) feel that former mental patients are not accepted by most people as friends, that they are not seen as being as intelligent or trustworthy as other people (57 percent and 53 percent, respec- tively), and that their job applications would

Table 1. Coefficients from Regressions of Over- all Subjective Quality of Life on Per- ceived Stigma or Received Services: Mental Health Patients, 1988-1989


Independent Variable

Received Services

Vocational rehabili- tation

Financial support


Mental illness1 chemical abuse groups

Time in leisure activities

Structure for
leisure time


-.6979* (.2649)

,2008' (.0838)

.8033a (.4293)
dardized     R2
-.23     .I99
.21     ,182
.23     ,198
.7094* (.3198)     .19     ,178
.1947* (.0916)     .20     .I91
.2349* (.0846)     .29     .238
,6909' (.2797)     .22     ,199

Note: Numbers in parentheses are standard errors. Number of cases range from 141 to 144 owing to missing data. All regressions control for clinical characteristics and demographic characteristics. A separate regression is run for each independent vari- able.

a Coefficient for financial support approaches sta- tistical significance (p < .06, two-tailed test).

*p< .05 (two-tailed tests)

leaves adequate variation for analyses of the effects of stigma.

Table 1 shows the relationship of perceptions of stigma and received services to quality of life, controlling for demographic and clinical characteristics. Perceived stigma is significantly and negatively associated with patients' evaluations of life as a whole: The greater the perception of devalu- ation and discrimination among mental pa- tients, the lower their satisfaction with life in general. These results also hold in analy-

be passed over by employers (77 per~ent).~ ses of the specific quality-of-life areas (not

Overall, slightly more than one-half of the respondents believe that mental patients are stigmatized. Thus, nearly one-half of respon- dents disagree that stigma exists, which

These percentages combine responses of "agree" and "strongly agree" to these statements.


Among the specific areas of life quality, stigma is significantly related to patients' evalua- tions of their living arrangements, family rela- tions, financial situations, safety, and health. Stigma is marginally related to satisfaction with employment status and leisure activities. The


Table 1 also shows the services that are significantly and positively related to overall quality of life. Services that increase eco- nomic resources-vocational rehabilitation and financial services (which approaches sta- tistical significance at p < .06)-enhance overall life satisfaction. Services that raise patients' status-empowerment ideology and groups for mentally ill chemical abusers (which are based on an empowerment ap- proach)-also improve subjective quality of life. Finally, time spent in activities and staff help with structuring free time positively af- fect satisfaction with life as a whole.5

The central issue here, however, concerns the importance of stigma relative to services for people with mental illness. In evaluating the effects of labeling, I hypothesized that the relationships between perceived stigma and quality of life will hold even when the effects of received services are considered, and vice-versa. To test this, I examine stigma in relation to services found to be effective for overall life satisfaction. For the most part, perceptions of stigma and received services are not related to each other.6

Table 2 presents coefficients for regres- sions that include specific services that are effective for increasing life satisfaction and perceptions of stigma, controlling for back- ground and clinical characteristics. Table 2 also shows the results of a regression of stigma and services combined on quality of life. This combined measure counts the total number of services associated with quality of life that patients received. Individuals receive one point each for receiving vocational ser- vices, using financial services, obtaining ser-

only area in which stigma has no effect is satis- faction with social relations, perhaps because most friends of respondents are also in mental health programs. The stigma measure focuses on relationships in the larger community. Also, the stigma of mental patient status may be minimized among persons sharing the same label (Estroff 1981).

5Emp~wermentis also associated with most of the specific aspects of quality of life. Other ser- vices associated with aspects of life quality are closely related to the particular quality-of-life area (Rosenfield and Neese-Todd 1993).

The one exception is that greater perceived stigma is associated with less time in leisure ac- tivities.

vices for structuring leisure time, attending MICA group meetings, reporting high levels of empowerment, and spending time in lei- sure activities within the pr~gram.~

The mea- sure thus ranges from 0 to 6 and has adequate reliability (alpha = .72).

Examining received services first, Table 2 demonstrates that each of the services is sig- nificantly related to life satisfaction when stigma is controlled. That is, vocational re- habilitation, financial support, empower- ment, MICA groups, and structure for leisure time contribute to overall quality of life even when the effect of stigma is controlled. The combined services measure is highly and sig- nificantly related to quality of life, holding perceived stigma constant. Thus, examining services separately or in combination, the positive impact of services received on over- all quality of life is not reduced by percep- tions of stigma.

Table 2 also demonstrates that the impact of stigma on life satisfaction remains signifi- cant when services are included in the re- gression. Controlling for vocational or finan- cial services, empowerment, or for services structuring leisure time, stigma retains its significant negative association with global quality of life. Stigma is also significantly related to life quality controlling for these services combined.

Thus, analyses comparing the effects of services and stigma show that both perceived stigma and received services are strongly re- lated to overall quality of life, but in oppo- site directions. The results are similar for most of the specific areas of life quality (not sh~wn).~

These analyses support the predic- tion of an independent impact of stigma and received services on subjective quality of life.

7The empowerment scale, frequency of attend- ing MICA groups, and time spent in leisure ac- tivities were split at the midpoint. Scores above the midpoint added one point each to the com- posite measure.

Exceptions are the stronger effect of received services on satisfaction with social relations, em- ployment status, and to a lesser extent, leisure ac- tivities, all areas in which stigma has the least ab- solute effect. Services may be more important in these areas because the program provides these functions. Thus, stigma encountered in the out- side community may be less relevant.
Perceived Stigma, Received Services, and Self-concept

The last prediction to be tested is that per- ceived stigma as well as received services are important for patients' quality of life because both affect self-conceptions. Table 3 presents coefficients from regressions that include measures of mastery and self-esteem as ex- planatory factor^.^

Controlling for the two dimensions of self- concept reduces the relationship between stigma and quality of life shown in Table 1 to nonsignificance. The coefficient for stigma is reduced by 38 percent (-,6979 to -.4299) controlling for mastery and 53 per- cent (-.6979 to -.3257) when self-esteem is controlled. Adjusting for self-esteem and mastery together diminishes the coefficient for stigma by nearly 60 percent. Similar re- sults also obtain for most of the specific quality-of-life areas (not shown).1°

On the basis of these findings, it appears that perceptions of stigma reduce patients' happiness with life by compromising their basic sense of self-worth and self-efficacy. The lowered sense of self brought about by expectations of stigmatizing social responses severely reduces patients' chances for overall satisfaction and in most specific life areas.

The two aspects of self-concept also re- duce the relationships between most of the specific services received and overall quality of life. The coefficients for vocational reha- bilitation and financial services each de- crease by about 20 percent when self-esteem is held constant. These associations are weakened considerably more, with reduc- tions of 42 percent and 51 percent respec- tively, after controlling for mastery.

In terms of empowerment services, control- ling for mastery or for self-esteem reduces the relationship between perceptions of em- powerment and quality of life to nonsig- nificance (compared to Table 1). Each dimen- sion of self-concept diminishes the relation- ship of empowerment to life satisfaction by

These are separate regressions for each area of service and for stigma. Results are similar when stigma is entered along with each received service into regression equations.

lo Only for financial situation and living arrangements does stigma remain significantly re- lated to quality of life.

Table 2. Coefficients for Regressions of Overall Subjective Quality of Life on Perceived Stigma and Specific Received Services: Mental Health Patients, 1988-1989


Independent Unstan-Stan-Variable dardized dardized R2
Vocational rehabili- tation Stigma     .2298* (.0815) -.7372* (.2612)     .24 -.24     ,245
Financial support Stigma     .8160* (.4200) -.6815* (.2721)     .23 -.22     .246
Empowerment Stigma     ,7138' (.3223) -.6225* (.2623)     .19 -.21     ,228
Mental illness/chemica abuse groups     l ,2176' (.0894)     .23     
Stigma     -.7220* (.2627)     -.24     ,244
Time in leisure activities Stigma     ,2202' (.0832) -.6394* (.2695)     .27 -.22     ,287
Structure for leisure time Stigma     ,6865' (.2778) -.5992* (.2693)     .22 -.20     ,239
All services combined Stigma     2.5201' (.7186) -.6803* (.27 17)     .44 -.23     ,398

Note: All regressions control for clinical charac- teristics and demographic characteristics. Numbers in parentheses are standard errors.

< .05

about one-third; controlling for both dimen- sions reduces the coefficient nearly 40 per- cent. Further, adjusting for mastery reduces the relationship between attending MICA groups and quality of life by approximately 30 percent, although the relationship remains significant when self-esteem is controlled.

Finally, in terms of structured activities, the coefficient for the amount of time spent in leisure activities in the program is reduced


Table 3. Coefficients for Regressions of Overall Subjective Quality of Life on Perceived Stigma or Received Services, Controlling for Aspects of Self-Concept: Mental Health Patients, 1988- 1989

Controlling for Controlling for Controlling for for Mastery Self-Esteem Self-Esteem and Mastery

Coefficient Coefficient Coefficient

Independent Unstan- Standar- Unstan- Standar- Unstan- Standar- Variable dardized dized R2 dardized dized R2 dardized dized R2


Vocational rehabilitation

Financial support


Mental illness1 chemical abuse groups

Time in leisure activities

Structure for leisure time

Note: All regressions control for clinical characteristics and demographic characteristics. Numbers in parentheses are standard errors.

*p < .05

by one-third when both aspects of self-con- gether, it appears that services promote pa- cept are controlled, but remains significant. tients' satisfaction with life by enhancing However, self-concept does not affect the re- their confidence in themselves and their lationship between services for structuring sense of control over their lives. free time and quality of life. Possibly the need for structure is a goal in itself, indepen-

dent of the need for control or positive self- regard (Rosenfield 1992). I began with a presentation of contrasting

These findings suggest that perceptions of views about the consequences of labeling for mastery and self-esteem mediate the associa- people with mental illness. Labeling theory tion between received services and overall predicts that the stigma of being labeled as quality of life, and that their importance var- mentally ill has a significant negative influ- ies with the specific service. The influence ence on outcomes for mental patients. Oppo- of both dimensions of self-concept is also nents think of stigma as inconsequential com- supported in the analyses of received ser-pared to the positive effects of treatment and vices in relation to the specific quality-of-life receiving high quality services. Integrating areas (not shown).12 Taking the results to- these perspectives, I proposed that percep-

tions of stigma and the receipt of services are 'I An overall measure of combined services is

related to subjective quality of life through

not included here because these perceptions dif- ferentially account for the associations between evaluations of their financial status or their living specific services and life satisfaction. These dif- arrangements. Perhaps general feelings of comferent relationships would be obscured by includ- petence and worth are less relevant in areas re- ing a measure that combined the services. lated to economic status, an area in which patients

l2 Self-concept does not account for the rela- have relatively little objective control and few tionship between received services and patients' options.

their associations with self-concept, but in opposing ways. Stigma is a problem for most people with chronic mental illness, and per- ceptions of stigma have a significant negative relationship with patients' quality of life. By contrast, services have a strong positive asso- ciation with quality of life. Furthermore, the effects of stigma and received services are a result of their relationships with perceptions of self-esteem and self-efficacy.

This study provides a test of the conse- quences of psychiatric labels, using cross- sectional data. The analyses should be repli- cated using longitudinal information to dis- entangle causal direction. Perhaps greater life satisfaction reduces perceptions of stigma and increases the use of services, or use of services enhances quality of life, which in turn reduces perceived stigma. It is also pos- sible that low levels of self-esteem or mas- tery result in high perceptions of stigma rather than stigma reducing these self-percep- tions. Link (1987) found that scores on the perceived devaluation-discrimination scale resemble scores obtained from the general public. However, the self-esteem of patients is lower than that among the general public. If low self-esteem increases perceptions of devaluation-discrimination, patients' devalu- ation-discrimination scores should be higher than those among the general public. Thus, these relationships do not support this alter- native causal explanation. However, such al- ternative explanations were not formulated by either labeling theorists or their critics before this study. They are therefore of interest only after evidence from the study suggests them. That is, before this research, labeling theo- rists could assume that received services have little importance in the face of stigma, and critics of labeling theory could assume that stigma matters little compared to received services. Although alternative explanations cannot be completely ruled out without fur- ther information, the results of this research cast doubt on both of these assumptions.

In terms of selection effects, it may be pos- sible that the more highly functional patients are selected by staff to receive services that most improve quality of life. However, this possibility is addressed in the analysis by controlling for functioning when examining the relationship of services to quality of life. Another possible alternative explanation is

that patients with a greater sense of self-es- teem or mastery are better able to obtain ser- vices and to have higher life satisfaction. This reasoning assumes, however, that pa- tients with high confidence and self-efficacy must be high-functioning in order to obtain these services. In examining this point of view, aspects of self-concept and level of functioning were not significantly related. Finally, 1 used a two-stage estimating proce- dure suggested by Maddala (1987) to check for selection effects in relation to services predicting overall quality of life (Rosenfield 1992). This procedure estimates the prob- ability of receiving a program service on the basis of symptoms, functioning, and demo- graphics, and enters that probability as an in- dependent variable in the regression analysis predicting quality of life. Results using this technique show that selection effects are not significant in relation to the program services-regressions predicting life satisfac- tion that include selection effects are similar to those obtained without these effects.

Limitations in generalizability are also an issue in the analysis. The ideal test of this theoretical perspective would be to study populations that include the full range of variation in both stigma and services. Al- though the sampling problems involved in such a test are difficult, replication of these results is needed in samples with greater variability.

Finally, specific conclusions about the rela- tive importance of the received services ver- sus perceived stigma to quality of life must be made with caution. For example, the im- portance of stigma may be underestimated because of the measurement of stigma com- pared to services. Services may have more weight because the measures of services were tailored to the specific services offered by the program, while the measure of stigma as- sessed general perceptions of devaluation and discrimination. Further development of the stigma measure is needed to more accurately estimate the roles of stigma versus services. Also, I examined high quality services in re- lation to stigma. Studies of services in other treatment settings would clarify their relative significance. In sum, I conclude that both stigma and received services are important in relation to quality of life, but I do not specify their magnitudes or comparative weights.


In spite of its limitations, this research pro- vides a unique combination of data for exam- ining labeling and opposing approaches to the treatment of mental illness. This study is con- sistent with other attempts to synthesize the labeling and psychiatric approaches. For ex- ample, the differences between the original statement of labeling theory and its critics on the causes of chronic mental illness-whether it is the label or the pathology-is reconciled by a modified labeling approach suggesting that the label is a stressor that in- creases symptoms (Link et al. 1989).13 Such syntheses provide a basis for a more complex understanding of the social processes under- lying the causes and consequences of mental illness.

These results should inform mental health interventions. They suggest that life satisfac- tion is highest for those who experience little stigma and gain access to high quality services. Life satisfaction is lowest among those perceiving high levels of stigma and lacking such services. A labeling or a psychiatric per- spective alone implies different interventions for mental illness. For example, a traditional psychiatric approach without a labeling com- ponent suggests that treatment programs for the chronic mentally ill are sufficient, pro- vided the treatment services are good. In- deed, this analysis shows the highly positive effects of high quality care. But this treat- ment stands as an oasis. Within this oasis, everything is provided-even empowerment as an antidote to stigma. But an oasis implies that a larger, harsher environment surrounds it, and treatment programs exist within com- munities that for the most part are hostile to people with mental illness (Angermeyer, Link, and Majcher-Angermeyer 1987). To the extent that stigma exists in this larger en- vironment, the best treatment given by people with the best intentions is not enough, because the treatment has little power to de- crease the stigma. The stigmatized constantly refer to the "normal world" (Goffman, 1963).

l3 There are similar attempts at synthesis in other areas of deviance. For example, Braithwaite (1989) points out that labeling criminal behavior can have negative effects in terms of exclusion and positive effects in terms of reintegrative shaming.

Good adjustment for stigmatized individuals involves acknowledging the differences be- tween themselves and normals, and in this sense, stigma defines the limits of treatment. Combining the insights of labeling theorists and their critics is thus necessary to produce positive patient outcomes. Only interventions that reduce the stigma within communities and provide high quality treatment can truly improve the life chances and quality of life of people living with mental illness.

Sarah Rosenfield is Associate Professor of Soci- ology and a core member of the Institute of Health, Health Care Policy, and Aging Research at Rutgers University. Her main research inter- ests are in medical sociology where her focus is on the impact of inequality on mental health. To this end, her current research examines the con- sequences of gender stratification for mental ill- ness. In addition, she continues her investigation of the treatments and stigma associated with se- rious mental illness.

Angermeyer, Matthias C., Bruce G. Link, and Alice Majcher-Angermeyer. 1987. "Stigma Perceived by Patients Attending Modern Treat- ment Settings: Some Unanticipated Effects of Community Psychiatry Reforms." The Journal of Nervous and Mental Disease 175:4-11.

Beard, John H. 1978. "The Rehabilitation Services of Fountain House." Pp. 201-208 in Alternatives to Mental Hospital Treatment, edited by L.

I. Stein and M. A. Test. New York: Plenum.

Beard, John H., Thomas J. Malamud, and Ellen Rossman. 1978. "Psychiatric Rehabilitation and Long-Term Rehospitalization Rates: The Findings of Two Research Studies." Schizophrenia Bulletin 4:622-35.

Becker, Howard. 1963. Outsiders. New York: MacMillan.

Bradburn, Norman M. and David Caplovitz. 1965. Reports on Happiness. Chicago, IL: Aldine.

Braithwaite, John. 1989. Crime, Slzame, and Re- integration. Cambridge, England: Cambridge University Press.

Cockerham, W. 1981. Sociology of Mental Dis- order. Englewood Cliffs, NJ: Prentice-Hall.

Crocetti, Guido, Herzl R. Spiro, and Iradj Siassi. 1974. Contemporary Attitudes toward Mental Illness. Pittsburgh, PA: University of Pitts- burgh Press.

Estroff, Sue E. 1981. Making It Crazy: An Eth- nography of Psychiatric Clients in an Ameri- can Cornmunity. Berkeley, CA: University of California Press.

Farina, Amerigo, Donald Gliha, Louis Boudreau, Jon Allen, and Mark Sherman. 1971. "Mental Illness and the Impact of Believing Others Know About It." Journal of Abnormal Psychol- ogy 77: 1-5.

Frank, Jerome. 1973. Persuasion and Healing. Baltimore, MD: Johns Hopkins University Press.

Goffman, Erving. 1963. Stigma. New York: Simon and Schuster.

Gurin, Gerald, Joseph Veroff, and Sheila Feld. 1960. Americans View Their Mental Health. New York: Basic Books.

Gove, Walter. 1970. "Societal Reaction as an Ex- planation of Mental Illness: An Evaluation." American Sociological Review 35:873-84.

, 1975. The Labeling of Deviance: Evalu- ating a Perspective. New York: Sage.

. 1980. "Labeling Mental Illness: A Cri- tique." Pg. 53-109 in Labeling Deviant Behav- ior, edited by W. Gove. Beverly Hills, CA: Sage.

. 1982. "The Current Status of the Label- ing Theory of Mental Illness." Pp. 273-300 in Deviance and Mental Illness, edited by W. Gove. Beverly Hills, CA: Sage.

Gove, Walter and Terry Fain. 1973. "The Stigma of Mental Hospitalization: An Attempt to Evaluate Its Consequences." Archives of Gen- eral Psychiatry 29:494-500.

Huffine, Carol and John A. Clausen. 1979. "Mad- ness and Work: Short- and Long-Term Effects of Mental Illness on Occupational Careers." Social Forces 57: 1049-62.

Jones, Edward, Amerigo Farina, Albert Hastorf, Hazel Markus, Dale Miller, and Robert Scott. 1984. Social Stigma: The Psychology of Marked Relationships. New York: Freeman.

Killian, Lewis M. and S. Bloomburg. 1975. "Re- birth in a Therapeutic Community: A Case Study." Psychiatry 38:39-54.

Kirk, Stuart A. 1974. "The Impact of Labeling on the Rejection of the Mentally Ill: An Experi- mental Study." Journal of Health and Social Behavior 15: 108-17. Lamb, H. Richard. 1982. Treating the Long-Term Mentally Ill. San Francisco, CA: Jossey-Bass. Lehman, Anthony F. 1983. "The Well-Being of Chronic Mental Patients: Assessing Their Quality of Life." Archives of General Psychia- try 40:369-73. Lehman, Anthony F., Susan Possidente, and Fiona Hawker. 1986. "The Quality of Life of Chronic Patients in a State Hospital and in Community Residences." Hospital and Com- munity Psychiatry 37:901-907.

Lehman, Anthony F., Nancy Ward, and Lawrence Linn. 1982. "Chronic Mental Patients: The Quality of Life Issue." American Journal of Psychiatry 139: 1271-76.

Lenoil, Julian. 1982. "An Analysis of the Psychi- atric Psycho-Social Rehabilitation Center." Psychosocial Rehabilitation Journal 1 :55-59.

Link, Bruce G. 1982. "Mental Patient Status, Work, and Income: An Examination of the Ef- fects of a Psychiatric Label." Atnerican Socio- logical Review 47:202-15.

. 1987. "Understanding Labeling Effects in the Area of Mental Disorders: An Assessment of the Effects of Expectations of Rejection." American Sociological Review 52:96-112.

Link, Bruce G. and Francis T. Cullen. 1992. "The Labeling Theory of Mental Disorders: A Re- view of the Evidence." In Mental Illness in So- cial Context, edited by J. Greenley. Greenwich, CT: JAI.

Link, Bruce G., Francis T. Cullen, James Frank, and John F. Wozniak. 1987. "The Social Re- jection of Former Mental Patients: Understand- ing Why Labels Matter." American Journal of Sociology 92: 1461-1500.

Link, Bruce G., Francis T. Cullen, Elmer Struening, Patrick E. Shrout, and Bruce P. Dohrenwend. 1989. "A Modified Labeling Theory Approach to Mental Disorders: An Em- pirical Assessment." Atnerican Sociological Review 54:400-23.

Linn, L. S. 1968. "The Mental Hospital from the Patient Perspective." Psychiatry 31:213-23.

Maddala, G. S. 1987. Limited-Dependent and Qualitative Variables in Econotnetrics. Cambridge, England: Cambridge University Press.

Mendel, Werner M. and Robert E. Allen. 1978. "Rescue and Rehabilitation." Pp. 18 1-200 in Alternatives to Mental Hospital Treatment, edited by L. I. Stein and M. A. Test. New York: Plenum.

Mosher, Loren R. and Alma Z. Menn. 1978. "Community Residential Treatment for Schizo- phrenia: Two-Year Follow-Up." Hospital and Community Psychiatry 29:7 15-23.

New South Wales Department of Health. 1983. Psychiatric Hospitalization vs. Cotnmunity Treatment: A Controlled Study. Wales, UK: New South Wales Department of Health.

Palamara, Frances, Francis Cullen, and Joanne Gersten. 1986. "The Effect of Police and Men- tal Health Intervention in Juvenile Deviance: Specifying Contingencies in the Impact of For- mal Reaction." Journal of Healtlz and Social Behavior 27:90-105.

Pearlin, Leonard I., Morton A. Lieberman, Eliza- beth G. Menaghan, and Joseph T. Mullan. 1981. "The Stress Process." Journal of Health and Social Behavior 22:337-56.

Rosenberg, Morris. 1984. Society and the Adoles- cent Self-Image. Princeton, NJ: Princeton Uni- versity Press.

Rosenfield, Sarah. 1992. "Factors Contributing to the Quality of Life of the Chronic Mentally

Ill." Journal of Health and Social Behavior 33:229-315.

Rosenfield, Sarah and Sheree Neese-Todd. 1993. "Why Model Programs Work: Factors Predict- ing the Subjective Quality of Life of the Chronic Mentally 111." Hospital and Cotnmu- nity Psychiatry 44:76-78.

Scheff, Thomas. 1966. Being Mentally Ill: A So- ciological Theory. Chicago, IL: Aldine. . 1974. "The Labeling Theory of Mental Illness." American Sociological Review 39: 444-52.

Schwartz, Carol C., Jerome K. Myers, and Boris

M. Astrachan. 1974. "Psychiatric Labeling and the Rehabilitation of Mental Patients." Archives of General Psychiatry 3 1 :329-34.

Stein, Leonard I. and Mary Ann Test. 1976. "Training in Community Living: A Follow-Up Look at a Gold-Award Program." Hospital and Community Psychiatry 27: 193-94.

. 1978. "Training in Community Living: Research Design and Results." Pp. 57-74 in Alternatives to Mental Hospital Treatment, edited by L. I. Stein and M. A. Test. New York:



. 1980. "Alternative to Mental Hospital Treatment." Archives of General Psychiatry 37:392-97.

Stein, Leonard I., Mary Ann Test, and Arnold J. Marx. 1975. "Alternative to the Hospital: A Controlled Study." American Journal of Psy- chiatry 132:517-22.

Sullivan, Harry S. 1941. "Psychiatric Aspects of Morale." American Journal of Sociology 47: 277-30 1.

Thoits, Peggy. 1985. "Self-Labeling Processes in Mental Illness: The Role of Emotional Devi- ance." American Journal of Sociology 91:221


Veroff, Joseph, Richard A. Kulka, and Elizabeth Douvan. 1981. Mental Health in America. New York: Basic Books.

Wahl, Otto F. and Charles R. Harman. 1989. "Family Views of Stigma." Schizophrenia Bul- letin 15: 137-39.

Weber, Max. 1946. Max Weber: Essays in Soci- ology. Oxford, England: Oxford University Press.

  • Recommend Us