Thinking about AIDS and stigma: A psychologist's perspective

As Jonathan Mann observed, the problem of AIDS-- related stigma is inextricably bound to issues of health, human rights, and the law. Such stigma transfates into feelings of fear and hostility directed at people with HIV.2 It finds expression in avoidance and ostracism of people with HIV, discrimination and violence against them, and public support for punitive policies and laws that restrict civil liberties while hindering AIDS prevention efforts. Being the target of stigma inflicts pain, isolation, and hardship on many people with HIV, while the desire to avoid it deters some from being tested for HIV, seeking treatment, or practicing risk-reduction.3

AIDS-related stigma and its attendant prejudice and discrimination are significantly shaped by misunderstanding and fear of HIV disease, coupled with social attitudes toward the people who contract it and the conditions under which it is transmitted. Thus, social psychological factors play a central role in the maintenance, enactment, and experience of stigma, and a better understanding of them will permit a more effective response to stigma. In the present article, I provide a brief introduction to social psychological theory and research concerning AIDS-related stigma.

I begin by considering what stigma is and I distinguish it from two related constructs, prejudice and discrimination. Next, I briefly describe the contours of contemporary AIDSrelated stigma in the United States, using findings from my own empirical research to illustrate key patterns. Then I explore some reasons HIV is stigmatized and how this stigma is enacted in social encounters. I also discuss the distinction between stigma based mainly on fear of HIV transmission (instrumental stigma) and that based mainly on preexisting attitudes toward the groups disproportionately affected by HIV (symbolic stigma). In the final section, I reflect on various strategies for reducing AIDS stigma.4

DEFINING STIGMA

English usage of "stigma" dates back at least to the 1300s.5 The term derives from the same Greek roots as the verb "to stick," that is, to pierce or tattoo. The earliest recorded English usage of stigma referred to the cluster of wounds manifested by Catholic saints, which corresponded to the wounds of the crucified Jesus. These stigmata were said to appear regularly, sometimes bleeding, in conjunction with important religious feast days.

Religious stigmata signified holiness, but stigma more commonly has had a thoroughly negative connotation. Taken literally, it refers to a visible marking on the body, usually made by a branding iron or pointed instrument. But the "mark" could also be a nonphysical condition or attribute. A 1907 textbook of psychiatry described a form of psychopathology known as a Stigmata of Degeneration, for example,6 and the Oxford English Dictionary notes a reference in 1859 to the "stigmata of old maidenhood."7

Historically, then, the meaning of stigma has had two components. First, it refers to an enduring condition or attribute. It is a mark borne by an individual. Second, that mark is negatively valued by the larger group or society. Whether it was physical or figurative, the mark was commonly understood to signify that the individual carrying it was criminal, villainous, or otherwise deserving of social ostracism, infamy, shame, and condemnation. Thus, not only are stigmatized individuals marked, but the larger group or society in which they live is characterized by a shared knowledge about the negative evaluation associated with the mark. People might vary in their personal responses, but everyone knows how society regards those who possess a particular stigma. As Erving Goffman pointed out in what is generally considered the classic sociological text on stigma, the stigmatized and the "normal" (his term for the nonstigmatized) are social roles, and all parties to any given interaction regardless of their own status - know the expectations associated with each role.8

The two roles are not simply complementary or symmetrical. Rather, they are characterized by a clear power differential. Sometimes the power difference itself is a basis for stigma, as when the poor and members of minority ethnic groups are stigmatized. Alternatively, members of society's ingroups may lose their advantaged status as a result of developing a stigmatized condition (e.g., schizophrenia or AIDS) or identity (e.g., as a homosexual or criminal). In either case, stigmatized individuals have less power and access to resources than do "normals."

Based on these considerations, we can define stigma as an enduring condition, status, or attribute that is negatively valued by a society and whose possession consequently discredits and disadvantages an individual. This definition allows us to distinguish stigma from two other relevant constructs: prejudice and discrimination. These words have somewhat different meanings for lawyers and social scientists. I use them here in their social psychological sense.

Prejudice is a negative attitude - that is, an evaluation or judgment - toward members of a social group.9 As an attitude, it involves emotions such as fear, disgust, anger, and contempt. Whereas stigma resides in the structure and relations of society, prejudice resides in the minds of individuals. One can be prejudiced against any individual or group regardless of how they are evaluated by the rest of society. But personal prejudice is a manifestation of stigma only when it reflects society's negative judgment of the target. A member of a stigmatized minority can harbor prejudice against members of the dominant majority, for example, but it is still the minority individual who is stigmatized, not those in the majority.

Discrimination is behavior. It refers to differential treatment of individuals according to their membership in a particular group.10 Discrimination, the act, is distinct from prejudice, the attitude. Attitudes often find expression in an individual's ongoing pattern of behavior, but many intervening variables can affect that relationship.11 Establishing a causal link between a person's general attitudes and her or his behavior in a particular situation is especially difficult. Thus, the relationship between attitudes and overt behavior is complex, hard to predict, and often indirect.12 Prejudice is only one element that potentially contributes to discriminatory conduct. Whether or not an individual discriminates in any specific instance depends on her or his having the motivation and ability to enact the behavior, and on the environment facilitating or hindering action.13 Like prejudice, individual acts of discrimination are distinct from stigma. One can discriminate against members of any group, but such behavior is a manifestation of stigma only when society as a whole condones or encourages it.

Discrimination and overt expressions of prejudice are not necessary for stigma to affect its targets. Many stigmatized individuals regulate their own behavior to avoid others' hostility and abuse. Recognizing this fact, British sociologist Graham Scambler distinguished between enacted stigma (overt acts of discrimination) and felt stigma (a stigmatized person's internal sense of shame and fear of persecution).14 Felt stigma motivates individuals with a stigmatized condition to attempt to pass as members of the nonstigmatized majority. Successful passing reduces their likelihood of actually being the target of enacted stigma, but also significantly disrupts their lives and is likely to increase their psychological distress. A valuable insight provided by Scambler's model is that actual discrimination need not occur for stigmatized people to suffer as a result of their status. They may restrict their own behavior and limit their own opportunities out of a sense of vulnerability resulting from felt stigma. This feeling of vulnerability is likely to persist in the absence of overt stigma. Even a few dramatic enactments of stigma (e.g., violence, discrimination, or ostracism directed at one individual) can heighten the sense of felt stigma among all members of the target group.

AIDS AND STIGMA

Having AIDS or being HIV-positive meets the criteria described above for a stigmatized condition: HIV infection is an enduring characteristic that relegates an infected individual to a socially recognized, negatively evaluated category. AIDSrelated stigma is manifested in prejudice, discounting, discrediting, and discrimination directed at people perceived to have AIDS or HIV and at the individuals, groups, and communities with which they are associated.15

AIDS stigma is a worldwide phenomenon. People with HIV in different nations are subjected to varying degrees of personal rejection, social ostracism, discrimination, violence, and, in many countries, laws and policies that deprive them of basic human rights.16 In the United States, press accounts and personal testimony since the first cases were detected in 1981 have told stories of people with AIDS - as well as those merely suspected of having the disease - being evicted from their homes, fired from their jobs, and shunned by family and friends.17 Although the prevalence of enacted AIDS stigma over the years has not been precisely determined, these anecdotal reports have been corroborated by experimental research and questionnaire and interview studies with people with HIV18 as well as reviews of litigation.19 In addition, public opinion surveys conducted in the early years of the epidemic revealed widespread fear of AIDS, lack of accurate information about its transmission, and willingness to support draconian public policies that would restrict civil liberties in the name of fighting the disease.20

After more than two decades of AIDS, we might expect stigma to have diminished to undetectable levels. Unfortunately, this is not the case. For example, my own national surveys throughout the 1990s showed that although support for quarantine and public identification of people with AIDS - the most punitive aspects of AIDS stigma - had diminished considerably by the end of the decade, one-fifth of those surveyed in 1999 still feared people with AIDS and one-sixth expressed disgust toward them or supported their public labeling. In addition, the surveys revealed that more covert forms of stigma persisted. Even as recently as 1999, roughly one-fourth of respondents felt uncomfortable having contact with a person with AIDS. Nearly one-third of respondents said they would avoid shopping at a neighborhood grocery store whose owner had AIDS.21

The surveys also revealed troubling signs that the sorts of beliefs and opinions that provide a foundation for AIDS stigma continue to be widespread. The proportion of adults believing that a person infected with HIV through sex or drug use deserves to have AIDS increased over the decade, peaking in 1997 at 28 percent. In response to a question phrased in less harsh language, approximately one-half of respondents said that people with AIDS are responsible for their illness.22 As elaborated below, when people are perceived to be personally responsible for having an undesirable condition, they generally are subjected to greater stigma than if their situation is perceived to be beyond their control.23

Of further concern is the fact that although members of the public understand how HIV is transmitted, they are much less knowledgeable about how it is not transmitted. In my surveys, the proportions of respondents overestimating the risks posed by some forms of casual social contact were actually higher at the end of the decade than in 1991, with between 42 percent and 51 percent believing in 1999 that HIV can be spread through sharing a drinking glass, being sneezed on, or using a public toilet.24 Similar results have been reported by other researchers.25 Those who believe that HIV can be spread through casual social contact are probably more likely to fear contact with those who have HIV and may be more willing in the future to support punitive policies that violate their human rights under the guise of protecting public health. Such fears may partly account for the widespread support that my surveys recorded for mandatory testing of various groups, including pregnant women, immigrants, and people judged to be at risk for getting AIDS. Although such support declined somewhat between 1997 and 1999, mandatory testing continued to he favored by most respondents.26

The social psychology of AIDS stigma

Why AIDS is stigmatized

In the field of social psychology, we have generally focused more on how stigma affects face-to-face interactions than on why a particular mark is stigmatized.27 Nevertheless, past theory and research highlight three characteristics that seem particularly relevant to understanding AIDS stigma.28

First, as noted earlier, the stigma attached to an undesirable condition tends to be more intense when that condition is understood to be the bearer's responsibility. An illness is likely to be stigmatized if it is perceived as having been contracted through voluntary and avoidable behaviors, especially behaviors that evoke social disapproval. Such an illness tends to evoke responses of anger and moralism rather than pity or empathy.29 Because the common routes of HIV in the United States include male-male sex and behaviors associated with injecting drug use, it is not surprising that people with HIV are regarded by a significant portion of the public as responsible for their condition and consequently are stigmatized.30 Indeed, research in the United States consistently reveals hierarchies of stigma associated with HIV according to who gets infected and how. Gay and bisexual men with AIDS are more stigmatized than heterosexual women with the disease. People who contracted AIDS through sex with multiple partners or sharing needles receive the most hostile reactions, with the least negative evaluations going to those who were infected through receiving contaminated blood products.31

Second, greater stigma is associated with conditions that are lethal and incurable. AIDS has been widely perceived to be a fatal condition since the earliest days of the epidemic.32 Although new drug regimens now offer realistic hope that HIV disease may be transformed from a fatal malady to a chronic illness, these regimens are not effective for everyone who takes them and many with HIV cannot tolerate them or do not have access to them. For the foreseeable future, therefore, most of the U.S. public will probably continue to perceive AIDS as invariably fatal.

Third, greater stigma is associated with a condition when it is perceived to pose a risk to others. More negative attitudes are directed at a stigmatized person to the extent that others believe they can be physically, socially, or morally tainted by interacting with him or her.33 Perceptions of danger and fears of contagion have surrounded AIDS since the beginning of the epidemic,34 and are evident in Americans' continuing overestimation of the transmission risks posed by casual social contact.35

Stigma and social interactions

What happens when a person possessing a stigmatized condition interacts with others? When addressing this question, it is important first to specify the type of individual with whom the stigmatized person interacts. When both parties possess the stigmatized characteristic (for simplicity, let us consider the case of a dyad), it may have little effect on the interaction. Minority status may be stigmatized in most of society, but typically not in gatherings of individuals from the minority group itself. Shared stigma can create a bond between individuals when they perceive themselves to be members of a group that is set apart from the rest of society. A consensus may even emerge whereby members of the majority group are disparaged. It is important to recognize, however, that such micro-climates of stigma - in which the stigmatized turn the tables on "normals" - cover extremely small areas of the social map.

When only one party to an interaction possesses the stigmatizing characteristic, her or his "normal" counterpart may not know about it. For example, being gay or asymptomatically infected with HIV is usually concealable, which permits the stigmatized individual to pass as a "normal." When the stigmatized member of a dyad successfully passes, the "normal" party believes that the interaction is occurring between two "normals." However, the individual who is passing knows that her or his stigmatized condition might be exposed at any time with an ensuing change in status. In Goffman's terms, she or he is "discreditable."36 The stigmatized person's central concern in such interactions, according to Goffman and others, is the management of information, that is, preventing the "normal" from learning about the mark.37

Like other stigmatized people, those with HIV who are passing as uninfected may divide the world into two camps: a large group to whom they disclose nothing about their serostatus and a small group of individuals who know about their diagnosis and are relied upon for help in keeping the secret. Sometimes these intimates put themselves in the role of protecting the person with HIV from prejudice or rejection by others and, in the course of filling this role, become acutely sensitive to the diagnosis and its attendant problems. People with HIV may also find themselves having to rely for help in protecting their secret upon others who, although they are not known personally, are able to detect their condition. These might include others with HIV as well as individuals whom Goffman labeled "the wise,"38 e.g., healthcare professionals and lay individuals active in the AIDS community. The wise are "persons who are normal but whose special situation has made them intimately privy to the secret life of the stigmatized individual and sympathetic with it, and who find themselves accorded a measure of acceptance, a measure of courtesy membership in the clan."39

Once the "normal" knows that the other party is marked, the latter individual is "discredited" (again using Goffman's terminology).40 The discredited no longer need to worry about passing, but they face different problems, the most obvious of which are overt expressions of stigma in the form of rejection, discrimination, or violence. Even if such enactments of stigma do not occur, however, they must manage the "normal's" impression of them. They may feel uncertain as to what others are really thinking of them and may perceive that they are under closer scrutiny than nonstigmatized individuals in comparable interactions. Their minor accomplishments may be considered too remarkable, whereas minor failings may be interpreted as a direct expression of their illness. Others might stare at them, especially if they manifest lesions, lipodystrophy, or other physical signs of illness. Strangers may feel free to strike up personal conversations about AIDS or offer unwanted help.

Of course, most people with HIV move through multiple social worlds in the course of a single day. Their stigma may be known in some of those worlds while they pass as uninfected in others. Hence, at various times they are likely to find themselves in both the discredited and discreditable roles. Even when the stigmatized person's status is known, the stigma's effect on the social interaction may vary according to how much the stigmatized condition actually disrupts the interaction or is perceived by others as repellent, ugly, or upsetting. For example, it is relatively easy to overlook the condition of an HIV-infected person who is asymptomatic, whereas the physical manifestations of advanced AIDS are readily evident in social interactions and often evoke distress from observers.41

The partners, family members, and close friends of those with HIV, as well as the professionals and volunteers who work with them or provide AIDS services or advocacy, often experience what Goffman called a "courtesy stigma."42 They are stigmatized through their close association with AIDS, those with HIV, and the many outgroups associated with HIV, including gay people and drug users. This secondary form of stigma creates challenges for those who are subjected to it. The loved ones of people with HIV, for example, may find that they are isolated from others. Caregivers and advocates may find that their work is more difficult and stressful as a result of stigma, or they may be deterred from working with those with HIV.43

Instrumental and symbolic stigmas

Given that AIDS is generally regarded as a degenerative and fatal disease, that HIV is transmissible, and that HIV infection is widely perceived to be the infected person's responsibility, AIDS probably would have evoked stigma regardless of which social groups it first affected. In the United States and many other countries, however, AIDS-related stigma has another layer of meaning because of the widely perceived association between HIV and disliked sectors of the population, especially gay and bisexual men and injecting drug users (IDUs). Recognizing this fact, many social scientists have found it useful to differentiate between two types of reactions to people with HIV

The first type, labeled "instrumental AIDS stigma," derives mainly from fear of AIDS as a communicable and lethal illness along with a desire to protect oneself from it. Instrumental stigma reflects the apprehension likely to be associated with any transmissible and deadly illness. The second type of reaction, referred to as "symbolic AIDS stigma," results from the use of AIDS as a vehicle for expressing hostility toward other groups that were already stigmatized before the epidemic began.44 It is based on the metaphorical social meanings attached to AIDS, the people who get it, and the ways in which it is transmitted. As Susan Sontag observed, illnesses such as cancer and AIDS have often been interpreted not as amoral biological phenomena, but rather in terms of good and evil, virtue and vice, punishment and innocence.45 Symbolic AIDS stigma derives its force from the association of HIV with disliked groups. The specific groups linked to AIDS in symbolic stigma vary somewhat across cultures, depending on the local epidemiology of HIV and preexisting prejudices.46 In the United States, symbolic AIDS stigma has focused principally on male homosexuality and, to a lesser extent, IDUs.

Even though gay and bisexual men constitute a shrinking portion of U.S. AIDS cases, much of the American public continues to equate AIDS with homosexuality.47 Moreover, individuals who closely associate AIDS with homosexuality harbor more negative attitudes toward gay men than those who do not.48 At the same time, some segments of society have had different experiences with the epidemic and, consequently, make different symbolic associations to AIDS. In the African-American community, for example, AIDS has affected not only gay and bisexual men, but also a substantial number of injecting drug users, with the consequence that symbolic AIDS stigma is linked to attitudes toward both groups.49

RESPONDING TO AIDS STIGMA

How can we best respond to HIV-related stigma? In the remainder of this article, I briefly offer a social psychological perspective on this question. Unfortunately, the empirical research literature on the effectiveness of various strategies for challenging AIDS stigma and its consequences is sparse. Relatively few antistigma interventions have been attempted, fewer still have been reported at professional conferences or in print, and adequate assessments of outcomes have been rare.50 My discussion, therefore, is necessarily speculative, although it draws upon social psychological theory and, when possible, empirical research.

As a starting point, Table 1 summarizes the results of follow-up analyses that I conducted with data from a 1999 national telephone survey of 1,335 English-speaking adults.sl The statistics presented in the table describe how strongly respondents' scores on a measure of AIDS-related stigma" were associated with various social, psychological, and demographic characteristics.53

The first column of data in Table 1 reports correlation coefficients, which characterize the strength of association between each individual variable and the measure of AIDS stigma.54 As shown in the table, respondents' stigma scores tended to be higher to the extent that they expressed negative attitudes toward gay men and did not understand that HIV is not transmitted through casual social contact (e.g., sharing a drinking glass, using a public toilet). In addition, respondents who expressed beliefs consistent with the psychological construct of authoritarianism - which is associated with a generalized tendency toward prejudice55 - tended to have higher scores for stigma, as did older respondents and those who did not personally know anyone with HIV Stigma also tended to be higher among respondents who held traditional opinions about sex - i.e., those who believed that it is mainly for purposes of procreation or is acceptable only between married people - and among respondents who had less formal education, described themselves as politically conservative, accorded religion a high level of importance in their life, attended religious services frequently, had a lower income, were unemployed, resided in a small town or rural area, or were male.

This pattern of correlations paints a preliminary picture of the groups of people in U.S. society who are most likely to harbor stigmatizing attitudes toward people with AIDS. While informative and thought-provoking, however, the correlations are of limited value in guiding the construction of stigma-reduction interventions. Many of them are extremely small.56 In addition, some of the variables are correlated with each other as well as with AIDS stigma. Compared with the less educated, for example, people with more years of schooling generally tend to be less prejudiced toward minority groups and also are more likely to be knowledgeable about topics such as HIV transmission. Consequently, the strong correlation between AIDS stigma and knowledge about casual contact might actually result from the fact that both are correlated with educational level. A statistical technique called multiple regression is useful here because it permits assessment of the relationship between AIDS stigma and each of the other variables while statistically controlling for all of these interrelationships.

The remaining columns in Table 1 report the end result of a series of regression analyses. Ultimately, five variables from the correlation analysis emerged as the most important predictors of AIDS stigma: beliefs about casual contact, attitudes toward gay men, authoritarianism, age, and personal contact with people with AIDS. When the variables' relationships to AIDS stigma were all analyzed simultaneously, these five proved to be the major factors underlying stigma scores for this particular sample. The standardized regression coefficients (beta, reported in the second data column) give an indication of the relative importance of each variable in explaining variation in AIDS stigma scores.57 I use these findings as the starting point for my discussion of interventions to combat AIDS-related stigma.

Information campaigns

Early in the epidemic, it became clear that simply providing information about HIV and how it is transmitted would not be adequate to stop the spread of AIDS. At the same time, knowledge about HIV transmission clearly is a necessary, albeit not sufficient, prerequisite for risk reduction. What about for stigma? Table 1 shows that one of the best predictors of stigmatizing attitudes is the belief that casual contact can transmit HIV Can information campaigns make a dent in HIV-related stigma?

I believe they can. Social reactions to a disease evolve as new information about the illness becomes available and is widely disseminated. Historically, the stigma associated with some diseases declined dramatically when the disease etiology was understood and cures or effective treatments were developed. This is illustrated in Charles Rosenberg's account of three nineteenth century cholera epidemics in the United States.58 Between the 1832 and 1866 epidemics, Americans progressed from regarding cholera as a moral punishment for sinners to understanding its connections with poor sanitation and public health practices.

When the 1832 epidemic struck, neither physicians nor members of the general public understood that cholera is transmitted through bacterial infection. Physicians believed that it was caused by the introduction of poisons into the atmosphere, e.g., from decaying matter. People were thought to be predisposed to succumb to these poisons as a result of certain conditions, including excessive sexual activity, intemperance, and idleness. Consequently, prostitutes and their customers, the poor, blacks, and immigrants all were regarded as "risk groups" for cholera. Because no effective treatment was available, public attention centered on prevention efforts, which often took a highly moralistic tone (e.g., closing saloons). Most Americans viewed cholera as a direct punishment from God or a consequence of failing to observe Nature's laws. Although physicians proclaimed (incorrectly) that cholera was not contagious, many members of the public avoided the sick. Hospitals for cholera patients provoked community protest ranging from petitions to arson.59

In the subsequent 1849 and 1866 cholera epidemics, scientific understanding of the disease increased and public attitudes changed. Once it was known that the cholera bacillus spread primarily through the vomitus and excrement of infected individuals, massive public health campaigns were mounted to destroy contaminated bedding and clothing, improve sewage disposal, purify public water supplies, and clean up cities. Outmoded moralistic conceptualizations yielded to a new respect for public health and medicine as Americans became familiar with the germ theory of disease and realized that purely material practices could prevent the spread of cholera. As Rosenberg pointed out, this shift in the paradigm for cholera did not reflect the culture's decrease in piety; rather, it was based on advances in scientific understanding that made the moralistic approach to cholera increasingly irrelevant.60

The story of cholera offers some hope for combating HIV-related stigma. We do not yet have a cure for AIDS but we understand how HIV is transmitted, and this understanding should help to reduce stigma. Unfortunately, there are at least three reasons why our advanced knowledge about HIV does not have a greater impact on stigma. First, the public remains surprisingly ill-informed about the disease. Whereas scientific understanding of HIV and AIDS has increased dramatically since the epidemic began, the public's knowledge has not followed suit. As noted earlier, my own surveys show that the proportion of American adults who incorrectly believe that HIV can be transmitted through casual social contact or donating blood actually increased over the decade of the 1990s.61

The fact that much of the public has never completely understood the viral transmission of AIDS is apparent in responses to survey questions about the HN risks associated with male homosexual behavior. In four national telephone surveys conducted between 1991 and 1999, I consistently found that a disturbingly large minority of U.S. adults equated any male homosexual behavior - even sex between two uninfected men - with HIV transmission.62 In a 1997 survey, for example, nearly one-fourth of the respondents believed that an uninfected man could get AIDS from one sexual encounter with another uninfected man, even if they used condoms. Without condoms, more than four respondents in ten believed that a man could get AIDS through sex with an uninfected man.63 In that survey, respondents who incorrectly believed that all male-male sex spreads AIDS also expressed significantly more negative attitudes toward gay men, which suggests that the association between AIDS stigma and antigay attitudes was buttressed by a belief system equating AIDS with sex between men.

These data point to the importance of reinvigorating public education efforts about HIV transmission and going beyond simple slogans like, "You can't get AIDS from shaking hands." The American public needs to understand why they can't get AIDS from shaking hands, other forms of casual contact, donating blood, or having sex (even unprotected sex) with an uninfected partner.

A second reason knowledge about HIV has not reduced stigma as much as it might is that some sectors of the public simply do not believe the scientific data about HIV Mistrust of experts is especially high among African-Americans, many of whom have long believed that HIV represents a genocidal government conspiracy targeting blacks.64 In a 1991 national survey, for example, 20 percent of blacks believed that AIDS is being used as a form of genocide against minority groups, compared to 5 percent of whites. Individuals who distrusted government and scientific sources of AIDS information also were more likely to stigmatize people with AIDS, mainly by avoiding social contact with them and supporting draconian policies that would isolate them.65

A third reason knowledge about HIV does not inevitably reduce stigma is that a considerable portion of AIDS-related stigma is symbolic and, consequently, unlikely to be affected by information campaigns. (The importance of confronting symbolic stigma is discussed below.) However, instrumental stigma also plays an important role in shaping reactions to AIDS. As shown in Table 1, misinformation about casual contact is a substantial predictor of AIDS stigma, even when attitudes toward gay men and other relevant variables, such as education, are statistically controlled. It is this component of stigma that is most likely to be affected by AIDS educational campaigns.

In mounting campaigns to provide accurate and credible information about AIDS, it is important that we do not further reinforce existing stigma. I noted earlier, for example, that the proportion of adults in the United States who perceive people with AIDS as responsible for their condition and therefore, to many, blameworthy - increased over the decade of the 1990s. This trend maybe an unintended consequence of public education efforts that have portrayed HIV prevention entirely in terms of individual decision-making. The flipside of messages stressing the importance of taking control of one's personal behavior to prevent HIV is that those who do not exercise such control may deserve whatever bad consequences follow. If so, health educators face the challenge of communicating the importance of protecting oneself from AIDS without promoting increased blame for individuals who become infected.

Mindful of these limitations, I believe we will have at least a limited impact on HIV-related stigma if we reenergize our public education efforts and address the specific problems I have outlined here. In contrast to cholera in the nineteenth century, we should not expect increased knowledge about HIV transmission to largely eradicate stigma. But without such knowledge, we cannot expect much progress at all.

Self-disclosure of serostatus

Another approach to stigma reduction has been to encourage people with HIV to disclose their serostatus to others. The hope is that uninfected individuals will be less likely to stigmatize people with HIV if they personally know someone who is infected. HIV infection is concealable, which makes it possible for many with HIV to "pass" as uninfected in routine social interactions. Controlling who knows that they are infected not only helps them to avoid the stigma associated with AIDS and related statuses (e.g., being gay or an IDU), but also allows them to preserve their privacy and maintain a sense of normalcy in their own lives. Passing, however, is often stressful. Keeping important information about oneself a secret can have negative consequences for one's mental and physical health.66 The demands of passing can disrupt relationships, requiring those with HIV to create distance from others in order to avoid disclosing their diagnosis. This may prevent them from receiving much needed social support. Moreover, hiding their diagnosis may reinforce a sense of shame about being infected, or may cause them to feel inauthentic - that they are living a lie.

On the other hand, people with HIV have legitimate concerns about personal rejection, discrimination, and even violence if others know their serostatus. It is understandable, then, if the admonition to people with HIV to disclose to others - not only sexual partners, but also family, friends, co-workers, and even strangers - meets with resistance. It is appropriate that they should want to know that "coming out of the closet" - i.e., disclosing their serostatus - will have a positive effect or, at the least, will not cause them harm.

My use of the phrase "coming out" to describe self-- disclosure of HIV status intentionally invokes the metaphor used by gay men and lesbians to describe the process whereby they reveal their sexual orientation to others. The expectation that coming out as an HIV-positive person will help to reduce AIDS-related stigma is based to a significant extent on the experiences of gay and lesbian people in the United States and other Western nations. Indeed, the U.S. public's attitudes toward homosexuality have moderated during the past few decades and one likely reason is that increasing numbers of heterosexual Americans now know that they have a gay or lesbian friend or close relative.

This effect has been explained with reference to what social psychologists call the contact hypothesis. As originally described by Gordon Allport, it holds that many forms of prejudice can be reduced by equal-status contact between majority and minority groups in the pursuit of common goals.67 Empirical data generally support the contact hypothesis, albeit with qualifications.68 Research also suggests that its tenets are applicable to heterosexuals' attitudes toward lesbians and gay men. Heterosexuals who personally know someone who is gay are likely to hold more favorable attitudes toward gay people generally.69

Because HIV in the United States was first reported among gay men, and because the epidemic here has disproportionately affected the gay community, it is not surprising that the process of disclosing one's HIV status has often been conceptualized in a manner that parallels how we think about disclosing one's sexual orientation. Indeed, many gay men with AIDS have characterized HIV disclosure as a second coming out. Moreover, consistent with the finding that heterosexuals who know a gay person have lower levels of sexual prejudice than others, empirical research has shown that uninfected Americans who personally know someone with HIV express less prejudice against people with AIDS than do uninfected individuals who lack personal contact.70 This association is demonstrated in Table 1, which shows that personally knowing someone with AIDS was an important predictor of lower levels of AIDS stigma in the 1999 survey.

It remains to be seen whether or not a similar correlation will be found in other cultures. Coming out as a person with HIV seems intuitively like a vastly different experience in the United States compared to countries with extremely high rates of HIV infection, as is the case in many African nations. In the latter societies, individuals who do not personally know someone living with AIDS or HIV are probably in the minority. Yet, anecdotal reports suggest that AIDS stigma is widespread in many of these countries. In addition, the coming out model of HIV self-disclosure may have important limitations in such societies for several reasons. Living in an area where stigma and seroprevalence are both high may create pressures to publicly prove one's own seronegativity by rejecting or attacking others who are known to be HIV-positive. In addition, to the extent that those with HIV are perceived to be similar to oneself, it becomes difficult to take refuge in the belief that "it can't happen to me." Consequently, the fact that heterosexual behavior is the predominant transmission mode for HIV in those countries may mean that members of the majority ingroup (i.e., heterosexuals) experience a greater sense of personal threat from HIV-infected individuals than is the case in societies where most people with HIV are (or are perceived to be) members of outgroups (e.g., homosexuals, injecting drug users). Yet another difference is that vast numbers of people with HIV in those societies are women, a group already lacking in power.

Thus, whether and to what extent self-disclosure of HIV status is an effective antidote to stigma in developing countries remains an empirical question. At the least, it seems likely that the social psychological processes underlying self-- disclosure (for both the HIV-infected discloser and the uninfected recipient of that disclosure) will be different from those commonly observed in the United States.

Vicarious contact with people with AIDS

A related strategy that is often advocated for reducing AIDS stigma is to encourage famous people who are HIV-infected to publicly disclose their serostatus. The underlying assumption is that members of the public will be affected by such an announcement in much the same way that they would be affected by learning that a close friend is infected. The difference between the two situations, of course, is that people have direct contact with friends but most have only vicarious contact with celebrities.

Expectations about the beneficial effects of vicarious contact with a famous person with HIV were discussed extensively in the United States when basketball star Earvin "Magic" Johnson revealed his HIV infection in 1991. The president of one AIDS organization commented on the effect of Johnson's announcement: "The main thing that raises awareness of HIV or AIDS is to know someone who has it. Now everybody in America knows someone with HIV."71 In similar fashion, a Sports Illustrated writer compared Johnson's disclosure to the shock of learning that a member of one's own family has AIDS.72 Social scientists also speculated that the effect of Johnson's announcement on those without AIDS would be similar to that of having personal contact with someone who did have the disease.73

Magic Johnson's announcement provided an unusual opportunity to study the effects of celebrity disclosure on AIDS stigma. As it happened, my colleagues and I had completed a national telephone survey on HIV-related stigma one year earlier and were preparing to field a follow-up version of that survey when Johnson made his serostatus public.

When we reinterviewed respondents, we asked them whether Johnson's announcement had affected their opinions. We also compared their follow-up attitudes and beliefs to their initial survey responses from one year earlier.

Johnson's disclosure did not result in dramatic reversals of attitudes, but it appeared to reduce stigma in those who had expressed the most extreme negative attitudes in the initial survey. Their AIDS-related attitudes moderated somewhat in the follow-up survey, but still remained generally more negative than those of the rest of the sample. The effect of vicarious contact was considerably weaker than that of direct contact: Survey respondents who personally knew someone with AIDS or HIV manifested substantially less AIDS stigma than those who lacked such contact.

These findings derive from what is effectively a single case study. Public disclosure by other celebrities with HIV may well have a different impact. However, given Magic Johnson's prominence, the public's high regard for him, and the way in which he disclosed (e.g., the fact that he revealed his serostatus while still in apparent good health), public reactions to his disclosure may well be as positive as we are likely to see in response to any celebrity disclosure. If this is true, the potential for reducing stigma through vicarious contact with people with HIV is probably limited.

Confronting symbolic stigma

As explained earlier, AIDS-related attitudes often serve as a vehicle for expressing hostility toward disliked groups that have been linked with the epidemic. Because its roots lie in these other prejudices, symbolic AIDS stigma is unlikely to be affected by factual information about HIV Instead, such information tends to be interpreted in a manner that reinforces stigma. Consider, for example, information about transmission. Increased knowledge about how cholera was spread helped to reduce stigma in the nineteenth century. In the AIDS epidemic, however, knowledge about the routes of HIV transmission has often been used to promote antigay stigma, as when then-Representative William Dannemeyer (R-CA) cited the epidemic as a reason for reinstating sodomy laws in the 1980s.74 Indeed, since the early years of AIDS, antigay activists have promoted symbolic stigma by blaming the gay community for starting the epidemic and portraying homosexuals as ongoing dangers to themselves. They also have admonished heterosexuals to avoid homosexuals, arguing that HIV and various AIDS-related diseases could be spread through casual contact, while they actively campaigned to prevent AIDS education programs from providing explicit information to gay and bisexual men about how to protect themselves sexually from HIV.75

The continuing potency of symbolic AIDS stigma is evident in the regression analysis reported in Table 1: Attitudes toward gay men were a prime predictor of AIDS stigma in the 1999 national survey; authoritarianism - an indicator of generalized prejudice - was a secondary predictor. The strong associations between Americans' AIDS-related attitudes and their attitudes toward gay people suggest that AIDS stigma will not be eliminated unless stigma, prejudice, and discrimination based on sexual orientation are also confronted. Unfortunately, the research literature on reducing sexual prejudice is of limited scope (though not as limited as the research on reducing AIDS stigma). Most studies in this area have focused on one of two routes by which sexual prejudice might be counteracted: through exposure to information or through personal contact with lesbians or gay men.76

In one type of study, experimental designs have been used to compare the attitudes of heterosexuals (typically college students) before and after they are exposed to information about homosexuality or the gay community (e.g., through a film, a speaker, or an entire course on gender or sexuality).77 Most of these investigations have documented reductions in research participants' prejudice after the educational program. Unfortunately, they have not identified the reasons why attitudes change. We do not know whether prejudice decreases because educational programs provide factual information and refute stereotypes about gay people, because they create an opportunity for developing positive feelings toward a specific gay person (e.g., a speaker), or because they promote general social norms of tolerance. It is also possible that such programs simply cue participants to subsequently give unprejudiced questionnaire responses regardless of their own true attitudes. The conclusions from this type of study have been further limited by self-selection bias: Participants who choose to be exposed to an educational program (e.g., by enrolling in a sexuality course) are probably different in important ways from other heterosexuals. For example, they may be more receptive to new information about sexual minorities in the first place.

In the second category of studies are the previously mentioned reports showing that personally knowing a gay man or lesbian is associated with lower levels of sexual prejudice among heterosexuals.78 The link is especially strong when heterosexuals know several gay people, when they report having close friends who are gay (rather than mere acquaintances), and when their relationship has included open discussion of the other person's homosexuality.71 An important limitation is that studies of this type have been mainly descriptive. The question of causality - whether contact reduces prejudice or whether heterosexuals low in prejudice are more likely to have contact with openly gay people has not been resolved. Most likely, contact and attitudes each exert some effect on the other. Consistent with social psychological theory and with empirical research on other types of prejudice,80 it is reasonable to assume that contact reduces prejudice.

Both types of studies - experimental assessments of educational programs and surveys of interpersonal contact - have focused on attitudes rather than behavior. We have virtually no empirical data directly assessing how to reduce enactments of antigay stigma, such as discrimination or violence. Nor have systematic studies been conducted to determine that reducing an individual's antigay prejudice will result in diminished AIDS-related stigma. Heterosexuals with more positive attitudes toward lesbians and gay men would be unlikely to use AIDS as a vehicle for expressing sexual prejudice, but they might be susceptible to other forms of symbolic stigma (e.g., expressions of hostility toward IDUs).81 As with interventions for reducing AIDS stigma, this remains an area where empirical research is urgently needed.

Other responses to stigma

Consistent with my psychological perspective in this article, I have highlighted strategies for fighting stigma that focus on individuals. I do not wish to imply, however, that such strategies are sufficient by themselves. We also need structural antistigma interventions, including legal and policy initiatives and political mobilization. Although an extended discussion of macro-social strategies is beyond the scope of this article (and my own expertise), I offer a few reflections relevant to such strategies.

In the United States, laws have been enacted to mitigate the effects of AIDS stigma, ranging from local ordinances prohibiting discrimination on the basis of HIV status to the federal Americans with Disabilities Act. In addition to their direct benefits to people with HIV, such laws are likely to lessen AIDS stigma for several reasons. They communicate the norm that AIDS-related discrimination is socially unacceptable. They motivate organizations to develop implementation plans that typically include methods for avoiding discrimination, such as guidelines for accommodating the needs of HIV-infected persons in the workplace. And, by making it safer for those with HIV to be open about their serostatus, antidiscrimination laws can increase the number of opportunities that uninfected people have to interact with those who are infected, which, in turn, may reduce prejudice.

Yet, as Scott Burris has pointed out, the benefits of legislation are limited.82 For one thing, people with HIV often remain unaware that laws protecting them even exist. For another, whereas laws may reduce enacted stigma, actual discrimination need not occur for people with HIV to suffer from the effects of AIDS-related stigma. As noted earlier, felt stigma motivates many people with HIV to constrict their own behavioral options to avoid ostracism, discrimination, and violence. Laws may have a negligible impact on felt stigma, especially in the short term.

Another limitation of relying on the law is that it can also be used as a vehicle for the expression of stigma. Throughout the epidemic, legislators and activists have repeatedly attempted to enact statutes and implement policies that are hostile toward people with HIV As Burris observed:

Our law does as much to stigmatize people with HIV, gay people, poor people, and drug users as it does to protect them. From the succession of Helms Amendments restricting HIV education to recent legislation requiring the discharge of the HIV infected from military service, there is a prostigma plank for every antistigma one in the government's AIDS platform.83

Many - perhaps most - of those prostigma planks can be understood as expressions of symbolic stigma. They translate moral condemnation of stigmatized outgroups and behavior into public policy. In doing so, they not only reinforce stigma, but have probably made the epidemic worse. For example, a considerable body of social science research on HN prevention demonstrates the importance of creating culturally sensitive HIV-prevention programs that are tailored to their target audience, imparting information that is clear and explicit, providing instruction on safe sex behavior, and enlisting social and community support for individuals who are trying to reduce their risk.84

However, federally funded AIDS prevention programs have been effectively prevented from meeting these criteria in their outreach to gay and bisexual men because they have been barred from portraying homosexuality in a positive light, promoting sexual activity, or distributing materials that federal officials judge to be obscene.85 In a similar vein, the abstinence-only curriculum legally mandated for federally funded sex education programs in the public schools ensures that male adolescents who engage in same-sex behavior (regardless of whether they will eventually identify as gay, bisexual, or heterosexual) will not receive information about how to protect themselves from HIV And whereas syringe exchange programs have been shown to be effective in reducing HIV transmission among IDUs, the federal government has opposed their implementation, largely on the premise that such programs condone injecting drug use.86

These are examples of how, as Jonathan Mann observed in the epigraph to this article, stigmatized status is linked to risk for HIV Because gay and bisexual men and IDUs belong to social outgroups, reducing their risk for HIV infection is a lower government priority than promoting moral condemnation of their behavior. These laws and policies increase felt stigma, not only among people with HIV but also among uninfected gay and bisexual people (women as well as men) and IDUs. Any effort to reduce AIDS-related stigma inevitably must confront these laws and policies.

This observation points to the necessity of locating efforts to eradicate HIV-related discrimination within the larger context of cultural power relations. As noted earlier, stigma is premised on power differentials. A group's relative powerlessness, however, does not mean that it must resign itself to being stigmatized. Indeed, recent history provides many examples of outgroups that have successfully challenged society's negative evaluation of a particular attribute or condition, such as homosexuality, mental illness, and racial and ethnic minority status. Such challenges were based on political mobilization among members of the stigmatized group as well as their "normal" allies. In like fashion, political organizing clearly has played a central role in shaping how society has responded to the AIDS epidemic.87 It will continue to be important for creating responses to AIDS stigma that challenge the structural bases of disempowerment that are so closely linked to such stigma.88

CONCLUSION

As Jonathan Mann observed, stigma and discrimination are the enemies of public health not only because they inflict suffering on people with HIV and undermine efforts to prevent the further spread of HIV, but also because they heighten vulnerability to HIV infection among individuals and groups. "For this reason," he wrote, "preventing discrimination toward HIV-infected people and people with AIDS has been made an essential part - for the first time in history - of the public health strategy to prevent and control the global epidemic."89

In this antistigma project, lawyers, health professionals, human rights advocates, and social scientists will all benefit from each other's insights and expertise. We need to change the law and public policy so that they censure AIDS stigma rather than sanction it. At the same time, we must change individual attitudes and behaviors so that people understand that HIV is a virus, not a moral judgment, and that those who contract it are human beings, not symbols. Thus, institutional and individual change must be central objectives in an antistigma project. Part of the challenge we face is to keep both goals in sight.

ACKNOWLEDGMENTS

An earlier version of this paper was presented at the conference Health, Law and Human Rights: Exploring the Connections, co-sponsored by the American Society of Law, Medicine & Ethics and the Temple University Beasley School of Law, in Philadelphia on September 30, 2001. I gratefully acknowledge the support I received while writing this paper from the University of California, Davis, and from the National Institute of Mental Health through an Independent Scientist Award (K02 MH01455). I thank Scott Burris for his insightful comments on a preliminary draft.