Many of those struggling with both homelessness and mental illness refuse offers of help, often of psychiatric care, but also offers of housing that depend upon psychiatric diagnosis. At least, they refuse particular offers at particular times (1, 2). Many have a complex, ambivalent relationship with psychiatric services. They may come to the community mental health center to use the telephone but refuse their medication. They may sit every day in the waiting room of the Department of Human Services but refuse to talk to a caseworker. They may accept diagnosis-dependent housing but violate the rules of the setting and find themselves back on the street. The repeated refusal and an apparent willingness to sabotage (in clinicians’ eyes) their care contributes substantially to what often becomes a fundamentally nomadic life in which a client moves between supported housing, jail, hospital, homeless shelter, and the street, a trajectory that has come to be called “the institutional circuit” (3; see also references 4 and 5).
Why do people refuse help, particularly—as is often the case—when they insist that they do not want to be homeless? Clinicians are often tempted to ascribe such refusal to accept help to a lack of insight. This study suggests that many of these refusals are best understood not (or not only) as psychiatric symptoms but as “costly signaling”—acts so expensive to the actor as to seem irrational to the observer but that may communicate important information (from the actor’s point of view) within a particular social world. Understanding the refusals in this way should lead us to change the way we offer help to those who need it.
The research used ethnography—long-term participant observation—to determine whether there were common, salient cognitive models used by these individuals to interpret their daily lives (whether they had a “culture”). The research involved over 1,000 hours of participant observation across 3 years in a Chicago neighborhood that exemplifies what sociologists call the “service-dependency ghetto,” a concentration of homeless shelters, supported housing, and services that have become the de facto treatment setting for many persons with serious mental illness in the United States (6).
Because the aim of the research was to understand a population that is known to refuse services related to psychiatric illness, ethnographic research was based in a drop-in center that offered nonclinical services (laundry, mail, a daily meal) and required no diagnostic interview for entry. The ethnographic work reported here was largely carried out by the author, a professional anthropologist with substantial psychiatric knowledge (7). She spent most of her time in the drop-in center but also met with women she knew from the drop-in center in local shelters, parks, and restaurants. These meetings were casual, unscripted, and dominated by the subject’s concerns. More systematic interviews were done by a team of five graduate students in anthropology and psychology trained by and supervised by the author. Before conducting the interviews, each student interviewer spent one afternoon a week for 20 weeks in the neighborhood learning how to engage the women. The following year, the students carried out formal interviews in the drop-in center. The students met with the author weekly in a group throughout the 2-year period to discuss the interview process. The work reported here includes the analysis of semistructured interviews with 61 women at the drop-in center. Over the course of several months, most women who came to the drop-in center on the afternoons when the two interviewers were present were asked for an interview; most women, but not all, agreed to be interviewed in exchange for three bus passes. Each woman was then asked systematically about her experience in the neighborhood. The work reported also includes 30 follow-up interviews carried out by a third interviewer in the drop-in center in which women were asked to explain the meaning of terms (such as “crazy” and “strong”) identified as important through the participant observation and in the semistructured interviews.
Because we believed that the women would be hostile to structured diagnostic probes, we did not conduct formal diagnostic interviews. Of the 61 women interviewed by two students, 43% lived in homeless shelters, 29% lived in supported single-room occupancy housing, and 10% slept on the street, in the park, or intermittently in hotels; 95% had stayed at a shelter at least once, 55% reported being psychiatrically hospitalized at least once, and 55% had been arrested at least once. The great majority of the study group (78%) displayed obvious psychotic symptoms, reported a history of either schizophrenia or bipolar disorder, or reported disability support without obvious physical cause. Only 28% volunteered a diagnosis of schizophrenia or bipolar disorder. It is important to note that this diagnostic uncertainty will be shared by those who seek to approach this difficult population.
When women refused services, they often did so publicly and on the grounds that they were not “crazy”—despite a well-articulated interest in being housed. It seemed to be common knowledge that one could get housed in a local single-room occupancy through a psychiatric diagnosis. Women could and did tick off the options for housing on their fingers: you had to be “crazy” or “addicted” or “have a job.” To become eligible on the basis of psychiatric disability, a woman had to make and keep an appointment with a mental health professional who would conduct a diagnostic interview. Refusals often hinged on this requirement. “They even wanted me to go to [mental health services] to qualify for the housing. Whatever it was, I did not want it. Why should I say I’m not competent?” Often, women used the word “crazy.” “I would not to go to [this single-room occupancy hotel]. They’re all crazy.” “I can’t do that; I’m not crazy.”
The ethnographic evidence suggests that the word “crazy” did indeed elicit a shared cognitive schema specific to this social world (8). The word appeared to have a highly consistent prototype: when asked who was “crazy,” women invariably pointed to or described someone who was flagrantly psychotic and openly talking to unseen voices. At the drop-in center or in a shelter, there was almost always at least one such woman present. The women’s use of the word was consistent with the following underlying cognitive model: that flagrant psychosis arises when a woman is not strong enough to cope with the difficulties of homelessness, that the condition is permanent, and that only those who give up the struggle to get out become flagrantly ill.
These three features—the social cause of psychosis, its permanence, and the belief that the strong and determined will withstand but the weak and feckless will become crazy—occurred spontaneously in the women’s conversations. “She’s been on the street too long,” women would say about someone else, twirling their fingers or rolling their eyes to show that the person that they were talking about was “crazy.” “Reality is so overwhelming for them,” one woman explained. “It is like a powerful explosion; they have to go within themselves; they have to create a safer ground. They cannot understand what’s happening, and it is the only way they can exist because they would otherwise just wither and die.” Another commented, “Some people cannot handle the pressure….They break and become mentally ill.” A woman whose husband had shot himself in front of her some months previously said, “I didn’t think anything was wrong with his head because he was a strong man. I just thought he was this strong man, that that would not ever happen to him, you know, he would never be crazy, he would never be actually crazy because he was a strong-minded person, strong-minded man, strong, so it would not happen to him. But I was wrong because it did.” Yet another remarked that “crazy” was “something that would never be fixed.” And another woman explained,
It [being crazy] is something you absolutely cannot control. And a lot of them don’t even take medication. They have retardation, and there’s nothing you can do about it. Alcoholism you can do something about. You can stop drinking. Smoking, you can stop smoking. You can do those things and thereby reverse your situation, but someone who appears mentally ill can’t do that.
Women repeatedly spoke about mental illness as retardation. As one woman put it, “Half of these people slow up here—you know what I’m saying—half of them got a little problem. They don’t think that well.”
In the follow-up interviews, 21 women were asked what other women at the drop-in center meant when they used the word “crazy” or when they said that the “street drives people crazy.” Five gave conventional psychiatric answers or refused to answer, but 16 readily provided a dynamic model of psychosis in which the social experience of being on the street caused illness in people who were weak or who gave up. For example, see the following:
There’s a couple of girls come up here that talk to themselves. That’s because they let the streets take over them…a lot of women have been raped by the men here, and [those girls] just can’t deal with it, so that just made them go haywire.
The street, it will drive you to the brink of—it comes back to being mentally strong….I’m not gonna let that happen to me. It happens because there are women in the shelter when they gave up. That’s why I say crazy now because you gave up.
They down and out, and you don’t want to be like that. You go in there [the shelter], and right away you feel the aroma.
They all weak….I seen women in my family being abused and I seen them be strong and stand up….But it’s bad when you on the street and a man is jumping on a woman real bad, and no one’s trying to help….I’ve seen these streets do a lot of things to people. I really seen a lot of them go to the hospital. So I guess it will drive them crazy. They get tired of being abused and stuff they just get tired.
You don’t have the willingness, the open-mindedness to change your life and your lifestyle, it’s gonna get so crazy it might kill you….The streets will kill you, they’ll kill you. You go out there and you experience eating out of garbage cans and panhandling, asking people for a [bus] transfer to get to and from so you won’t have to sleep in alleys and stuff; that shit will drive you crazy. That’s what they mean.
The richness of these explanations, which were presented as personal opinions, and the redundancy of their content argue that such a cognitive model is easily accessible within the social world in which these women live. In this social setting, when the women say that they are not crazy, they are asserting that they are not weak, and they have not been defeated. Such a model is, of course, stigmatizing, but it is more important to understand that it arises out of a particular social world that they experience as an assault that they have to survive.
That social world is indeed a hard one. Peer-to-peer relationships are often antagonistic. Reports of violence were ubiquitous. Every woman who completed a follow-up interview spontaneously spoke of violence; a third mentioned domestic or childhood sexual abuse. The women clearly perceived police protection as limited, although not meaningless (some women slept on the steps of the local police precinct rather than sleeping in the park). Staff in the drop-in center reported that at least once a week they saw women who had been visibly beaten.
As a result, the women adopt an interaction style in which they react aggressively to perceived insults in order to persuade the assailant to back down. The style is common in settings in which law is weak and the police are unreliable. The sociologist Elijah Anderson (9) described such an aggressive style in the inner city and called it “the code of the street.” Anthropologists identify such an aggressive style associated with a “code of honor” commonly found among nomadic peoples, pastoralists, and ranchers, where individuals can lose their entire wealth (their herd) to thievery and where, because they are isolated, they have few others to help to defend them (10). In such poorly policed settings, physical survival may depend upon an ability to defend one’s turf so aggressively at the first hint of trouble that the trouble slinks away.
Homeless women have little property to protect, but what they have and what they define as theirs—their space, their seat, their position in line—they defend with an aggression that far exceeds middle class normative expectations. It is clear that a sense of personal dignity contributes to the willingness to interpret insult (for example, to be reminded that one owes another person money in front of other people may be treated as deeply insulting). It is also clear that the response is aggressive in order to persuade the offender to back down. When the author was with a woman in a corridor or in the park and men from the neighborhood approached, the woman would pull herself up, shift into a threatening, aggressive stance, and raise her voice. “They don’t mess with me,” a woman remarked on one occasion when the men left.
In the initial semistructured interviews, 70% of the women said that “people up here are not trustworthy”; 40% agreed that “people here put you down”; 47% agreed that “people up here” sometimes or usually did things to annoy them; and 88% had “seen a woman rant or rave in a shelter” at least once.
Meanwhile, women perceive themselves to be on their own and able to rely on few others for protection. In the drop-in center, women sit together in more or less predictable patterns, and they talk and joke and clearly enjoy each other’s company. But those patterns are often fragile, and they change abruptly. The use of the word “friend” to refer to people women meet in the neighborhood is remarkably limited.
In the initial semistructured interviews, women were asked, “Do you feel connected to other women at [the drop-in center]?” Forty percent replied, “No.” Two-thirds could not name a single person in the drop-in center whom they called a “friend.” When asked to describe the previous day (with a version of the day reconstruction method, reference 11), over 40% included no face-to-face interactions, despite the presence of many other women in the places where they ate, slept, and spent time.
The need to be tough and aggressive—and the importance of not being weak—was freely verbalized by the women. “If you’re going to survive, you have to smack down somebody,” one woman explained. “You have to keep your guard up at all times,” reported another. “You cannot let people think you’re weak because they’ll take advantage of you, guaranteed, they’ll take advantage of you.” Sometimes the importance of aggression was explicitly associated with honor or “respect,” as has been reported in other inner-city settings (12).
This social toughness is in direct conflict with the polite compliance expected in staff-client relationships. Institutions—shelters, soup kitchens, drop-in centers, social services, clinic services—provide the everyday needs of daily life for these women. To remain in the shelter and to continue to get fed, a woman must avoid violating the institution’s rules while negotiating the tensions of living densely packed with women she does not know. Staff expect clients to keep their voices low and to follow obediently a myriad array of rules: for example, do not wash below the neck in the sink, return coffee cups to the kitchen before dinner, and smoke only between 9:00 and 9:15 in the evening. Service settings often post such rules on long lists in visible places. One shelter supervisor asks women to recite rules out loud together before they prepare for bed. Staff understand their goal, appropriately, as establishing the safety and cleanliness of the service setting, and they enforce these rules by refusing access to the service when the rules are broken. If two women fight, even only with words, they are “barred”—dismissed—and told not to return for a day, a week, a month, even forever if the infraction is severe. Women themselves often recognize the different codes of conflict appropriate for “decent people” and “street people” and the necessity of the rules. Yet in practice, women are often caught between the conflicting demands of two different interactional codes. The same woman who scoffs at “street behavior” and describes herself as “decent” may the next moment find herself defending herself aggressively and dramatically in self-defense and get kicked out of the service setting.
The need to be tough and the awareness that one must succeed with the staff to survive the street and to exit homelessness is expressed by another of the most important cultural schemas for the women in this neighborhood: “being strong.” The word appeared in sentences such as this, spoken by a woman in the drop-in center: “If [homelessness] ever happens to them they better have a strong heart and a strong mind because when they see everything, they’re gonna need a strong heart and a strong mind. If you aren’t strong hearted, you cannot take it out here.”
As in the case of the word “crazy,” the ethnographic work demonstrates that the use of the word “strong” evokes and is motivated by a complex of meanings. One dimension is clearly being tough: standing up for yourself, being able to protect yourself, not letting other people take advantage. That meaning is expressed by comments such as these: “Being strong is like—I ain’t getting hurt protecting myself” or “You got to be strong to deal with some of these people….A dude walked up and said, ‘I could just go over there and kill everybody in the parking lot because they sleep.’ I said, ‘Baby, I’m not asleep by a long shot….’ He got into the car with somebody else and they left.” And yet “strong” also carries the meaning of being able to resist the temptation to be tough—most often in order to follow the rules of the service setting and to get access to the goods they offer, such as housing. Asked about what it meant to be “strong,” one woman explained, “There is help there. You gotta want it. If you don’t want it, I mean, it’s not gonna just fall from the sky. You gotta put the footwork in.” Women spontaneously spoke about being “strong” when they described learning to interact with counselors, case workers, and service providers.
When you really seek help, you gonna reveal to the people that you’re seeking out who you really are.…I had to talk to these case workers and reveal the grimy things I did, and I did not feel real good. I wanted to fold inside, I wanted to lash out, but I was the author of everything that was done. And I had to be strong and come in here and say, okay, but that was then.
They also spoke about being “strong” in order give up addiction. “So I was in this facility,” a woman said. “Everyone was getting high, they were drinking, they were doing drugs. And at first it didn’t bother me. It didn’t bother me for 5 to 6 months. I was real strong.”
These two meanings of strong—toughness on one hand and yet a willingness to forgo toughness in order to get needed services (often being described as being “mentally strong”)—ran throughout the follow-up interviews. For example,
You have to be strong mentally, physically, socially, emotionally.…You cannot make friends with anybody [laughs]. You really cannot because I mean you have to be tough, you know, you have to have a tough coat.…You just you have to be strong; you can’t allow other people’s opinion of you to bring you down. You can’t afford to be depressed. You cannot afford to get sick in any kind of way in the streets because somebody will come along and take advantage of you.…You stay away from people that you know are sick, no matter if it’s an emotional sickness or it’s a physical sickness. You try to stay away from those people because in a way it could rub off on you.…You have to watch out for yourself because not nobody else do it. You go to the free clinic to stay physically fit. You stay involved with your case management.…It takes a very strong mind.
In follow-up interviews, these two themes were mentioned in 25 of 28 completed interviews.
The judgment that one must be strong is not misplaced. The women see the consequence of what they take to be the failure to be strong all around them, in women beaten and raped, in women destroyed by drugs and their violence, and in women who do not leave the shelter system or the street for years. Many have stories of women who did not survive. All of them, of course, see flagrantly psychotic women, and they correctly judge such women to be more vulnerable to violence than those who are not ill (13). On the street, flagrantly psychotic women are not only physically vulnerable, but they are often jeered at and disliked.
Women did not always refuse services, and they did not always refuse to seek mental health care. Some did associate being “strong” with using psychiatric services. When they refused help, however, their refusal was very often framed as a denial that they were “crazy.” It is important to note that this denial is more complicated than simple stigma. The culture in the neighborhood does not simply represent psychosis as bad. The culture represents being “crazy” as the outcome of a failure to be strong enough to get off the street. The women have many such examples of apparent failure to point to in such a neighborhood: flagrantly psychotic women in the shelters who disrupt their sleep and are vulnerable to physical violence. They wish to avoid being one of them. When they say that they are not “crazy,” they are saying that they are not weak, they have not been defeated, and they can survive. Paradoxically, what it means to be identified as “crazy” comes to carry a sense of being beyond help, incurable—exactly the opposite of what most mental health providers intend.
Many people engage in acts that others perceive as highly costly and that in some concrete sense are harmful to those who perform them. They may behave with unconditional generosity, giving large feasts without expectations of reciprocity or return. They may display wealth in conspicuous, seemingly wasteful extravagance. They may risk their lives in competitive performances that bring them little direct material reward. They may invest great time or resources into acquiring finely ornamented and instrumentally useless goods. Recent anthropological theory draws on broader social scientific analysis to argue that such acts may be best understood as “costly signals,” acts that might be thought harmful from a purely materialist or individualist perspective but in fact accrue what social theorists call “symbolic” or “social” capital (14). Such acts may assert claims to status or competence that are meaningful within a competitive social world in which actors have incomplete information about each other.
This article suggests that at least to some extent, homeless women who could get housing based on a psychiatric diagnosis but who reject it with the assertion that they are not “crazy” are making such a costly signal. The signal is indeed expensive to them. The choice to forgo housing exposes them to considerable danger and discomfort. But it is a signal that asserts competence and strength in a social setting in which those attributes are highly valued. And to some extent, it is credible. Psychosis, after all, is a continuum. Women who refuse housing and are not forcibly hospitalized and then forcibly housed are not flagrantly psychotic. In fact, the very capacity to refuse housing is an important part of the signal. Signaling theory argues that expense of the signal in fact ensures the honesty of that which is being signaled—in this case, strength and competence. When the women refuse help because they are not “crazy,” stating that refusal in a drop-in center or a shelter, they are asserting strength and competence to the hostile, unpredictable community that is their social world on the street.
Understanding refusals in this way, as meaningful social signals rather than simply the result of a lack of insight, should lead us to think differently about the way the offer of housing help could be extended: that in extending offers of help to those on the street with psychotic illness, every effort should be made to avoid an explicit psychiatric diagnosis. The primary ethnographic finding reported here is that those who live on the street and struggle with psychiatric illness may reject an offer of help not (or not only) because their illness distorts their understanding but because they perceive a diagnosis as the sign of great vulnerability to predatory others.
These findings offer support to those efforts that now exist sparsely in the American mental health system in which the offer of help—and housing, in particular—is indeed decoupled from explicit psychiatric diagnosis based on an understanding that the way help is offered affects the rate at which it is accepted (15). The standard program of supported housing for those with psychiatric disability—typically called the “linear residential treatment program”—requires the client to be explicitly assessed for psychiatric disability and to agree to participate in psychiatric and (usually) substance abuse treatment. Typically, participants are required to maintain sobriety in order to maintain the housing. Also typically, clients live in a series of step-by-step programs that progress to permanent independent living (16). The alternative program places clients immediately in permanent scattered-site housing, does not require sobriety, makes minimal demands on client compliance, and minimizes obvious psychiatric evaluation and diagnosis. (Patients must still meet diagnostic criteria.) Exemplified by Pathways to Housing in New York (17), this “housing-first” client-driven approach is increasingly gaining empirical and political support. It appears to enable clients to be housed for longer stretches of time (18) and to be more satisfied with services and more engaged in treatment (19), and it costs no more than the traditional treatment approach (17).
This alternative approach may not only be more effective than the traditional model; it represents a different understanding of the client’s perspective. The housing-first approach treats client perspectives as legitimate and meaningful. That is the approach supported by this ethnographic analysis, which argues that client refusals and the insistence that one is not “crazy” can be understood as socially meaningful within the particular social and cultural world of the institutional circuit. To treat clients most effectively requires not merely medical and diagnostic skill but sophisticated cultural understanding—and the willingness to put to one side our own cultural sensibilities about how we who offer help signal our own good will to those who need it.
Presented in part as a Distinguished Lecture at the 160th annual meeting of the American Psychiatric Association, San Diego, May 19–24, 2007. Received July 21, 2007; revisions received Sept. 17 and Sept. 19, 2007; accepted Sept. 19, 2007 (doi: 10.1176/appi/ajp.2007.07071166). From the Department of Anthropology, Stanford University. Address correspondence to Dr. Luhrmann, Department of Anthropology, Bldg. 50, Stanford University, Stanford, CA 94305; email@example.com (e-mail).
The author reports no competing interests.
Supported by NIH grant R21 MH090441.
The author thanks George Luhrmann, Kim Hopper, Rebecca Bird, Glynn Harrison, Cathleene Macias, David Laitin, Hazel Markus, Richard Saller, and the five student researchers: Johanne Eliacin, Amy Cooper, Kim Walters, Jim Goss, and Barnaby Reidel.
1.Koegel P, Sullivan G, Bumam A, Morton SC, Wensel S: Utilization of mental health and substance abuse services among homeless adults in Los Angeles. Med Care 1999; 37:306–317
2.Rosenheck R, Lam JA: Homeless mentally ill clients’ and providers’ perceptions of service needs and clients’ use of services. Psychiatr Serv 1997; 48:381–386
3.Hopper K, Jost J, Way T, Welber S, Haugland G: Homelessness, severe mental illness and the institutional circuit. Psychiatr Serv 1997; 48:659–665
4.Folsom DP, Hawthorne W, Lindamer L, Gilmer T, Bailey A, Golshan S, Garcia P, Unützer J, Hough R, Jeste DV: Prevalence and risk factors for homelessness and utilization of mental health services among 10,340 patients with serious mental illness in a large public mental health system. Am J Psychiatry 2005; 162:370–376
5.Herman DB, Susser ES, Jandorf L, Lavelle J, Bromet EJ: Homelessness among individuals with psychotic disorders hospitalized for the first time: findings from the Suffolk County Mental Health Project. Am J Psychiatry 1998; 155:109–113
6.Dear M, Woloch J: Landscapes of Despair. Princeton, NJ, Princeton University Press, 1987
7.Luhrmann TM: Of Two Minds. New York, Knopf, 2000
8.D’Andrade R: The Development of Cognitive Anthropology. Cambridge, UK, Cambridge University Press, 1995
9.Anderson E: The Code of the Street: Decency, Violence and the Moral Life of the Inner City. New York, Norton, 1999
10.Nisbet R, Cohen D: Culture of Honor. Boulder, Co, Westview, 1996
11.Kahneman D, Krueger AB, Schkade DA, Schwartz N, Stone AA: A survey method for characterizing daily life experience: the day reconstruction method. Science 2004; 306:1776–1780
12.Bourgois P: In Search of Respect: Selling Crack in El Barrio. Cambridge, UK, Cambridge University Press, 1995
13.Teplin L, McClellan G, Abram K, Weiner D: Crime victimization in adults with severe mental illness. Arch Gen Psychiatry 2005; 62:911–921
14.Bird R, Smith E: Signalling theory, strategic interaction and symbolic capital. Curr Anthropology 2005; 46:221–248
15.Mowbray C, Bybee D: The importance of context in understanding homelessness and mental illness: lessons learned from a research demonstration project. Res Soc Work Practice 1998; 8:172–199
16.US Department of Housing and Urban Development: Homeless Programs. Washington, DC, US Department of Housing and Urban Development, 1997
17.Pathways to Housing: Providing housing first and recovery services for homeless adults with severe mental illness: APA Gold Award description. Psychiatr Serv 2005; 56:1303–1305
18.Tsemberis S, Eisenberg R: Pathways to housing: supported housing for street-dwelling homeless individuals with psychiatric disabilities. Psychiatr Serv 2000; 51:487–493
19.Tsemberis S: From streets to homes: an innovative approach to supported housing for homeless adults with psychiatric disabilities. J Community Psychiatry 1999; 27:225–241