Out of Sight but Not out of Mind: Managing Invisible Social Identities in the Workplace Author(s): Judith A. Clair, Joy E. Beatty and Tammy L. Maclean Source: The Academy of Management Review, Vol. 30, No. 1 (Jan., 2005), pp. 78-95 Published by: Academy of Management Stable URL: https://www.jstor.org/stable/20159096 Accessed: 27-01-2019 21:22 UTC

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at https://about.jstor.org/terms

Academy of Management is collaborating with JSTOR to digitize, preserve and extend access to The Academy of Management Review

This content downloaded from 75.108.117.50 on Sun, 27 Jan 2019 21:22:27 UTC All use subject to https://about.jstor.org/terms ? Academy of Management Review 2005, Vol. 30, No. 1, 78-95.

OUT OF SIGHT BUT NOT OUT OF MIND: MANAGING INVISIBLE SOCIAL IDENTITIES IN THE WORKPLACE

JUDITH A. CLAIR Boston College

IOY E. BEATTY University of Michigan-Dearborn

TAMMY L. MACLEAN Suffolk University

Invisible social identities influence social interaction in distinct ways and create unique dynamics in terms of identity management. We integrate research from the sexuality, illness, and racial literature, as well as the stigma, disclosure, and identity literature, to create a generalized model of invisible identity management. We focus specifically on revealing and strategies of identity management and conclude by discussing the implications of invisible differences for diversity research.

Demographic diversity, defined as any char ment. As a way to ease into the conversation, the acteristic that serves as a basis for social cate chairperson mentions that it looks like the job gorization and self-identification, comes in two candidate has recently gone on a vacation, types: visible and nonvisible (Tsui & Gutek, since his skin looks tanned. In fact, strangers 1999). Visible characteristics usually include frequently make this comment because the job sex, race, age, ethnicity, physical appearance, candidate is multiracial and appears to be Cau language, speech patterns, and dialect. Nonvis casian,1 except for his tanned-appearing skin ible characteristics usually include differences tone. The job candidate must decide whether or like religion, occupation, national origin, club or not to reveal his multiracial background and social group memberships, illness, and sexual risk offending the chairperson by making her orientation. Most organizational scholarship on appear naive. In a second example, a woman diversity in the workplace has focused on such with multiple sclerosis (MS) is asked to take visible social identities as age, race, and notes on a flipchart in front of a group of people (Williams & O'Reilly, 1998), but left relatively she doesn't know at a work-related conference. unexplored are the dynamics of invisible social On this day her MS symptoms are flaring. Her identities. hands are shaky and uncoordinated, and it is To understand the importance of invisible so quite difficult for her to write quickly and legi cial identities in the workplace and how these bly. Instead of declining the request or explain identities can influence social interactions, con ing her situation, she muddles through the awk sider the following workplace scenario. A job ward task and wonders what the others think of candidate is interviewing for a tenure-track po her (Beatty, 2003). Finally, a woman seeks time sition with the chairperson of a history depart

1 We recognize that the language we use to refer to par We presented an earlier draft of this paper at the 2002 ticular social identity groups impacts shared understand annual meeting of the Academy of Management, in Denver, ings of that group. We also acknowledge that people have a Colorado. We improved the paper with helpful suggestions variety of opinions about which terms are appropriate or and feedback from Ron Dufresne, Yahuda Barach, Paula inappropriate when referring to certain social identity Caproni, and Ellen Ernst Kossek. Additionally, we extend groups. We choose to draw from popular usages at the time our gratitude for guidance provided by Arthur Brief and this paper was written to refer to particular social identity three anonymous reviewers. groups. 78

This content downloaded from 75.108.117.50 on Sun, 27 Jan 2019 21:22:27 UTC All use subject to https://about.jstor.org/terms 2005 Clair, Beatty, and Maclean 79 off from work for the birth of her child, which personal information to avoid stigma interferes should be a routine request. But she must ex with one's authentic self-presentation (Creed & plain her situation because she is a and Scully, 2000; Moorhead, 1999; Reimann, 2001). her child is being carried by her partner. Thus, In addition, individuals with an invisible so the woman must "out" herself in order to receive cial identity may also struggle to acquire and standard parental benefits (Reimann, 2001). maintain legitimacy within social interactions These examples illustrate how invisible social at work. For example, research on invisible ill identities influence and complicate workplace ness and disability suggests that individuals interactions. Such situations force individuals to may need to perform their tasks in different make quick strategic decisions about self ways?for example, by sitting down or taking presentation regarding an invisible identity. more frequent breaks. They must either explain Invisible social identities are common in or these differences or bear the negative reactions ganizations. One study suggests that about one from others who think they are strange or simply in every fifty people in the workplace is , a lazy (e.g., Charmaz, 2000). statistic that rises to one in every ten to twenty Those with invisible differences also struggle persons in major metropolitan areas (Michael, to balance conflicting relationship demands. Gagnon, Laumann, & Kolata, 1994). Chronic, dis For example, individuals in mixed-race relation abling conditions cause major limitations in ac ships may seek to keep this aspect of their lives tivity for more than one in every ten Americans hidden at work so as to avoid stigmatization and (Centers for Disease Control, 2002). Chronic ill . These individuals fear negative ness is often invisible and is the leading cause stigma associated with interracial dating or of disability in the United States, accounting for marriage, which others may see as deviant or 44 percent of disabled people age 15 and older. abnormal. However, these individuals can iso Yet many people with chronic illnesses are ac late themselves as well as their partners in an tive participants in the workforce (e.g., Pinder, attempt to keep their relationship hidden in the 1988; Vickers, 2001). Finally, 6.8 million people event of work-related social events that include reported more than one race on the 2000 U.S. family and significant others. Census (Jones & Smith, 2001). Employees from a mixed-race heritage can have an invisible or The consequences of stigma in the workplace ambiguous racial identity, because they may are readily quantified: the literature on organi zational diversity documents how women, racial not "look like" someone with a particular racial background to others. minorities, older workers, and others bearing a While organizational scholars who study stigmatized identity have suffered job loss, lim workplace diversity have mostly overlooked in ited career advancement, difficulty finding a visible social identities, there have been studies mentor, and isolation at work (Cox, 1993). To of particular types of invisible social identities avoid negative consequences, those with stig within the literature on race (Leary, 1999), on matized invisible social identities may struggle (Herek, 1996; Reimann, 2001; with whether or not as well as how to reveal Woods, 1994), on chronic illness (Pinder, 1995; their difference in an organizational setting, Schneider & Conrad, 1980; Vickers, 2001), and on causing them stress and anxiety in workplace disabilities (Matthews & Harrington, 2000). social interactions. For this reason, invisible so These studies suggest that individuals with cial identities invoke some distinct issues that stigmatizing invisible social identities have dif cannot be easily collapsed under traditional or ferent interaction experiences at work than ganizational diversity research that focuses on those with visible differences. visible differences. In particular, management of information To advance research, in this article we use about the invisible social identity is a central identity, disclosure, and stigma theories, as well issue for these people. The focus on information as the gay, illness and disability, and ethnora management is shaped not only by the threat of cial diversity literature, to draw attention to the stigmatization but also by concerns of authen influence of invisible social identities on work ticity and legitimacy. People experience a feel place interactions. We highlight workplace ing of authenticity when they can be fully challenges faced by those with stigmatizing in "themselves" in public. Conversely, concealing visible social identities. Additionally, we reveal

This content downloaded from 75.108.117.50 on Sun, 27 Jan 2019 21:22:27 UTC All use subject to https://about.jstor.org/terms 80 Academy of Management Review January how these individuals manage their invisible relationships by straight colleagues can put the social identity in workplace interactions. gay or lesbian person into a situation where he We organize the article as follows. First, we or she must reveal his or her sexual orientation. introduce the concepts of invisible social iden Conversely, an individual may wish to reveal a tity and stigma. Second, we discuss two strate difference at work but feel that he or she lacks gies people use to manage their invisible social control over this choice. For example, Clinton's identities at work: passing and revealing. Third, "Don't ask, don't tell" policy in the U.S. military we explore antecedents influencing whether a made it difficult, if not impossible, for gay sol person decides to pass or reveal in a workplace diers to openly reveal their sexual orientation. social interaction. Fourth, we consider outcomes An individual's choice to reveal information for the individual who chooses to pass or reveal. may also be constrained by relationships with Finally, we consider implications of our ideas significant others if socially damaging informa for research on diversity at work. tion compromises or violates relationships with Before moving forward, we wish to highlight these persons (Bernstein & Reimann, 2001). For several important assumptions made in this pa example, people with an invisible illness may per. First, we distinguish our work from recent face resistance from family members about re diversity research focused on "deep-level" di vealing potentially stigmatizing conditions such versity characteristics (e.g., Harrison, Price, & as epilepsy (cf. Schneider & Conrad, 1980). Bell, 1998; Harrison, Price, Gavin, & Florey, 2002). Deep-level diversity characteristics are less vis INVISIBLE SOCIAL IDENTITY AND STIGMA ible psychological features of individuals, such as personality traits, values, and attitudes (Har A person's social identity is derived from the rison et al., 2002). Here we explore those social groups, statuses, or categories that the individ identities that are derived from such individual ual is socially recognized as being a member of differences as race, chronic illness, and sexual (Rosenberg, 1997). Social categorization (Tajfel, orientation, which are invisible. However, 1981) and social identity (Turner, 1982) theories though not the focus of this article, we acknowl assert that individuals classify themselves and edge that deep-level diversity characteristics others into social categories using salient and are also likely to be invisible. available characteristics such as age or race Second, while we assume that invisible social (Triandis, Kurowski, & Gelfand, 1994). Others identities can be a basis for status as well as easily recognize visible differences in social in stigmatization, we focus here on the challenges teraction (Stangor, Lynch, Duan, & Glass, 1992; associated with stigmatized invisible identities. Williams & O'Reilly, 1998). Thus, the way people Third, we avoid taking a normative stance in think about identity and differences (including this article regarding whether a person should many scholars) is largely formulated along vis be "out" at work. We believe that a person's ible lines. choice to reveal or pass at work is contingent on As Schlossberg explains: a range of factors, a number of which we con Theories and practices of identity and subject sider in this article. In fact, while scholars some formation in Western culture are largely struc times implicitly or explicitly support being "out" tured around the logic of visibility_We are at work (i.e., Creed & Scully, 2000), threats of subjects constituted by our visions of ourselves and discrimination offer significant and others, and we trust that our ability to see reasons for a person with an invisible stigma and read carries with it a certain degree of epis temological certainty (2001: 1). tizing social identity to avoid disclosure in the workplace (i.e., Schneider, 1987). We seek to ex For example, Pisares (1999) notes that skin color, pose these factors influencing disclosure deci facial features, dress, and hair are used in com sions. bination with other factors, such as precon Finally, we focus on situations where individ ceived notions of race, immediate context, a per uals have at least some free choice to make a son's name and accent, and his or her apparent disclosure decision, even though in practice in social class or education, to determine a per dividuals may not always have full discretion to son's racial identity. In general, people take pass or reveal at work. For example, Sullivan each other's social identities at face value (2001) asserts that inquiries about one's personal based on cultural norms of expected behavior

This content downloaded from 75.108.117.50 on Sun, 27 Jan 2019 21:22:27 UTC All use subject to https://about.jstor.org/terms 2005 Clair, Beatty, and Maclean 81

(Pinder, 1995) and assume membership in par et al., 1984). This traditional perspective of ticular identities in the absence of visual or be stigma focuses on the social reactions of others havioral cues that would alert them otherwise. to a person with a stigma and on the fact that Goffman (1963) has called these expected char invisibility helps the stigmatized individual acter attributes "virtual social identity." Hetero avoid problematic social interactions that may sexuality and able-bodiedness are examples of occur because of the stigma. Yet for people with common virtual social identities. invisible differences, issues arise prior to social Visible cues become problematic in social in interaction. These issues are psychological, oc teractions when they lead to a person's stigma curring within the individual as he or she con tization, as pointed out by Goffman (1963) and siders how to manage his or her stigma in pub subsequent researchers. "Stigmatized individu lic. The dyadic social interaction, which has als possess (or are believed to possess) some been the focus of most stigma research, comes attribute, or characteristic, that conveys a social into play after the psychological issues have identity that is devalued in a particular social been addressed (although not necessarily re context" (Crocker, Major, & Steele, 1998: 505). solved). In particular, given the possible costs Stigma is a relative phenomenon; thus, certain associated with being stigmatized, people with characteristics considered stigmatizing in one these invisible social identities are likely to culture are not considered so in another. For think strategically about whether, when, and instance, in the United States and old how to reveal their difference (Frable, Black age are devalued and stigmatized, yet in other stone, & Sherbaum, 1990). Thus, the decision to cultures they are revered. Stigma is also collec reveal or conceal information about the self is tively defined and recognized, meaning that central and significant in interpersonal interac people within a culture tend to agree on stigma tions at work. tized characteristics. In the next section we consider the dilemma Being stigmatized is harmful for targeted in people with an invisible social identity face in dividuals because it leads to stereotyping, sta social interactions at work regarding whether tus loss, and discrimination (Link & Phelan, they should pass or reveal their difference, and 2001). Stereotyping influences social interac how they should do so, given risks associated tions by shaping what we notice in an encoun with being stigmatized as a result of revela ter, how we evaluate ambiguous information, tions. and how we remember an event or person (Crocker & Lutsky, 1986). We negatively evaluate TO PASS OR REVEAL AT WORK others who bear a stigma, regardless of the truth or merit of the negative evaluation. Based on our review of the literature, in this In the workplace, stigma can stifle an individ section we explore two basic strategic choices ual's advancement and personal development in social interactions at work that individuals opportunities and can lead to social isolation with stigmatized invisible social identities can (Cox, 1993). It can interfere with the development make: to hide or to disclose their invisible social of relationships, which are critical for network identity. We call the first choice "passing" and ing and career advancement (Day & Schoen the second choice "revealing." A person must rade, 1997). These issues can lead to poorer job make decisions about whether and how to re performance (Stephan & Stephan, 1985). Some of veal over and over again across many social the most stigmatizing differences also lead to interactions (Friskopp & Silverstein, 1995; difficulties in being hired and maintaining a Schneider, 1987). For instance, an individual job. For instance, job loss is a common concern who has already revealed that he has epilepsy for gays and (Woods, 1994; Woods & to a coworker or to a project team will have to Harbeck, 1991), people with epilepsy (Iphofen, decide again and again whether to reveal his 1990; Schneider & Conrad, 1980), and people with chronic illness to new people and, if so, how and AIDS (Leary & Schreindorfer, 1998; Weitz, 1990). when to do so. Visibility is a key dimension of stigma (cf. The degree to which an individual reveals or Crocker, Major, & Steele, 1998; Jones et al., 1984), passes at work translates into an overall state of and scholars assert that people with conceal being totally "in the closet" (i.e., he or she tells able differences are relatively "better off" (Jones no one about his or her difference), being "par

This content downloaded from 75.108.117.50 on Sun, 27 Jan 2019 21:22:27 UTC All use subject to https://about.jstor.org/terms 82 Academy of Management Review January

tially out" or revealed (i.e., he or she tells only a People use different tactics to pass. Herek select number of others, who keep the secret (1996) identifies three: fabrication, concealment, among themselves), or being entirely "out of the and discretion. Differences among these three closet" or revealed (this person indiscriminately approaches are subtle; however, we believe that reveals his or her difference). each represents a distinct passing tactic. Fabri cation occurs when a person deliberately pro vides false information about himself or herself to others. In other words, one creates a new Passing identity through deception by presenting fabri Passing is "a cultural performance whereby cations about one's life. For example, Woods one member of a defined social group masquer (1994) describes how gay men actively construct ades as another in order to enjoy the privileges heterosexual identities in the workplace by afforded to the dominant group" (Leary, 1999: 85). bringing platonic, opposite-sex friends as dates Passing leads a person to be classified incorrectly to organizational functions. by another person as someone without a discred Concealment involves actively preventing ited or devalued social identity; for example, in others from acquiring information about oneself the workplace, chronically ill people pass for (Herek, 1996). Those who use concealment hide healthy, gay people pass for heterosexual, and personal information in order to pass as a dom people from a multiracial background (e.g., inant group member. For example, during the white and Asian-American) pass for white. The Jim Crow era, multiracial individuals who were above definition suggests that passing is inten light skinned and appeared white evaded dis tional, but passing may also occur unintentionally crimination by hiding personal information, (Conyers & Kennedy, 1963). For example, passing such as where they were raised, what churches occurs when someone is the subject of another's they attended, or where their relatives lived, so mistaken assumptions. Bowman (2001) describes that their African-American heritage would not how physical features of individuals who have be revealed. white and African-American heritage?such as Finally, discretion occurs when a person straight hair, light skin, or blue eyes?may lead avoids queries related to the invisible identity them to pass every time they walk into a room, or social group membership in question. Differ step into a cab, or are with a new group of ences between discretion and concealment are coworkers and choose not to announce their ra subtle, but important nonetheless. Those who cial background. use discretion do not actively hide personal in In this article we focus on intentional passing formation, as with concealment, but instead tactics. The ability to intentionally pass de "dodge the issue" (Woods, 1994). Individuals pends on both the degree to which one can con avoid opportunities where they would need to ceal a social identity and on its ambiguity. One hide or share such information, and they "elude must be physically and culturally able to fit into personal questions, talk in generalities or another social identity group and to hide reveal change the focus of conversation" (Chrobot ing information about oneself from coworkers. Mason, Button, & Declimenti, 2001: 323). For in For example, a person with an invisible chronic stance, many gay people reveal nothing at all illness can pass at work if symptoms are mini about their sexual identity to their coworkers, mally intrusive. As Goffman states, "He who thus appearing asexual. No matter what tactic is passes will have to be alive to aspects of the used, passing requires a person to live a "dou social interactions which others treat as uncal ble life" and to adopt different personas within culated and unattended" (1963: 88). The inten and outside of work. tional passer must be vigilant of social cues in their work environment and take care to avoid situations where markers of his or her real so Revealing cial identity become evident to others. For in Those who reveal disclose an identity that stance, gay people must be careful not to dis would otherwise be invisible or unrecognizable play personal photographs of their partner, and to others. is a term commonly used chronically ill people must carefully hide intru to describe when one reveals one's homosexu sive symptoms while at work (Thorne, 1993). ality. The term is also used to describe how

This content downloaded from 75.108.117.50 on Sun, 27 Jan 2019 21:22:27 UTC All use subject to https://about.jstor.org/terms 2005 Clair, Beatty, and Maclean 83

persons reveal other stigmatized differences, tempts by an individual to establish, maintain, such as religion (Delbecq, 2002) and social class or pretend to be living as "normal" an existence (Rothenberg & Scully, 2002). as possible (Joachim & Acorn, 2000; Register, As with passing, people rely on different tac 1987; Royer, 1998; Strauss, 1975). Royer (1998) as tics to reveal an invisible social identity. Three serts that normalization helps a person preserve tactics include signaling, normalizing, and dif emotional balance, cope with possible alien ferentiating. Signaling straddles the line be ation, and adjust his or her expectations to in tween going public and remaining private corporate the limitations of an invisible identity. about one's invisible social identity. Individuals For example, a lawyer with MS who experi who signal disclose their invisible social iden ences vision and balance problems that affect tity by sending messages, dropping hints, and his courtroom performance can normalize his giving clues. Signals may involve the use of situation by saying that others get ill in the ambiguous language, specific conversational courtroom too. He notes that although illness is topics, and various symbols or nonverbal cues. an unusual event, there are general procedures For example, a person from a particular ethnic in place to cope with them in the courtroom: or religious heritage may choose to retain a dis lawyers leave temporarily and return to com tinctly ethnic name to signal this heritage, or plete the trial when they are well (Beatty, 2003). gay people may use conversations about partic Thus, the lawyer can cope with his illness on ular books, music, and vacation spots as ways of days when it flares and not feel that he must disclosing their homosexuality. These clues "in hide it. vite speculation ... encouraging their peers to Research suggests that gay people also nor read between the lines" (Woods, 1994: 176). malize by presenting information about an un People can choose signals that are meaning familiar sexuality and lifestyle in familiar, het ful among "insiders" but that can be ambiguous erosexual terms (Woods, 1994). Creed and Scully and innocuous to others. For example, devout cite an example of a lesbian woman who works Christians place "Jesus fish" plaques on their for a Fortune 50 company and normalizes by cars to signal their affiliation or beliefs to oth saying, "Lots of other people have had troubles ers. By using this signal, people make their in with their teenage sons. We've got a problem. visible social identity visible to knowledgeable Think of something mundane. We mow the others. Further, even if others do know the grass, we call Roto-Rooter for plumbing" (Creed meaning, signals are ambiguous. For example, & Scully, 2000: 23). She shares these everyday does a rainbow bumper sticker mean that one is difficulties with others to underscore the com gay or that one is a heterosexual person who monality of their life experiences. supports gay rights? Some signals are more Finally, differentiating occurs when people widely known than others, so people can choose highlight their invisible social identity and how strong signals or weak signals, depending on it differentiates them from others. Those who the context and motives in their situation. Sig differentiate seek to present an identity as naling minimizes the risk of explicitly revealing equally valid (rather than stigmatized) and may an invisible stigmatizing social identity and engage in an effort to change the perceptions provides an interim step to see if it seems safe to and behavior of the groups, organizations, and reveal more. institutions that might stigmatize them. For peo Normalizing is a second approach to reveal ple with stigmatizing illnesses, this strategy in ing one's invisible social identity. Individuals volves speaking up when others are dismissing employ normalizing by revealing their invisible them or treating them poorly, voicing their rights social identity to others and then attempting to to fair employment, and threatening legal op make their difference seem commonplace or or tions based on employment laws like the Amer dinary. Those who normalize attempt to assim icans with Disabilities Act. For multiracial indi ilate into local organizational culture, behave in viduals, this may mean publicly claiming their accordance with the norms of the organization, multiracial background, rather than allowing and may deny that their invisible difference others to categorize them into one racial group matters. The invisible social identity is subtly or another. The stigmatized person reclaims and acknowledged, but its significance and stigma redefines the way in which his or her invisible are minimized. Thus, normalizing represents at social identity is understood at work, preserving

This content downloaded from 75.108.117.50 on Sun, 27 Jan 2019 21:22:27 UTC All use subject to https://about.jstor.org/terms 84 Academy of Management Review January

its marginality and transforming traditionally work: (1) organizational diversity climate, (2) in stigmatized differences into assets. dustry and professional context, and (3) legal Differentiators' very act of revealing and protections. Additionally, we explore features of claiming their social identity may drive change interpersonal relationships that influence a per at both the individual and organizational level. son's decision concerning whether or not to re Researchers describe a strategy of "identity de veal. We suggest that people interpret these ployment," where people reveal differences in contextual conditions as "signals" or symbolic ways that challenge perceptions, values, prac indicators of whether support for or stigmatiza tices, and perspectives of the dominant group or tion of their invisible difference will occur from organizational culture vis-?-vis stigmatized oth disclosure. ers (Bernstein, 1997; Creed & Scully, 2000; Taylor Organizational context. First, organizational & Raeburn, 1995). These encounters can be as context influences the decision to reveal as in simple as individuals sharing personal stories dividuals assess the social norms of their work and allowing others to ask questions. place. Tsui and Gutek (1999) term this the diver sity climate, and scholars such as Cox (1993) and ANTECEDENTS TO PASSING AND REVEALING Ely and Thomas (2001) discuss how diversity cli mate influences a person's experience of being Stigma and social identity theories lead us to different within that social context. Supportive expect that individuals with stigmatizing invis and affirming work climates are associated with ible social identities will pass in the workplace more openness (Chrobot-Mason et al., 2001; to avoid negative consequences and to maintain Driscoll, Kelley, & Fassinger, 1996). For example, positive social images. However, research indi individuals are more likely to reveal their sex cates that people do reveal their stigmatizing ual orientation if they perceive they have sup social identities to those with whom they inter portive coworkers (Cain, 1991; Ragins & Corn act at work. In the following sections we elabo wall, 2001b) and managers (Day & Schoenrade, rate on our proposed model (see Figure 1), which 1997). We suggest that organizational policy and outlines antecedents to the decision to pass or procedures, transparency of decision making, reveal, the outcomes of that choice, and the pro and the presence of other "out" individuals in posed iterative relationship between these ele fluence an organizational diversity climate. ments. Organizational policies are likely to influence Disclosure research illustrates that environ whether a person passes or reveals. Nondiscrim mental and individual factors influence the ination policies, either legally mandated or vol choice to reveal or pass. Thus, we divide ante untarily adopted by an organization, establish cedents into two categories: (1) contextual con organizational accountability so that a person ditions, which include organizational, profes has some measure of protection against dis sional, industry, and legal environments, as crimination and feels safer revealing a differ well as interpersonal relationships, and (2) indi ence. In addition, policies that are voluntarily vidual factors, which include individual differ adopted by an organization, such as domestic ences (i.e., personality characteristics and/or so partner benefits, signal that an organization cial identity group memberships) and personal seeks to support employees with particular in motives. These factors are unlikely to operate visible social identities. Employees are more independently. Consequently, a person is likely likely to reveal in organizations that exceed to experience conflicting pressures to reveal minimal compliance standards, because these and to conceal stigmatizing information. We dis efforts signal that the organization values these cuss each of the antecedents separately, how employees and seeks to create a welcoming en ever, to identify their unique contributions to a vironment. person's propensity to pass or to reveal. However, policies and procedures designed to ensure a discrimination-free environment will Contextual Conditions be interpreted as empty promises if they are not enforced. Research on organizational compli In this section we consider how three contex ance with securities regulations suggests that tual conditions influence a person's propensity when training related to compliance is infre to reveal or to pass in social interactions at quent, punishment for compliance rule violators

This content downloaded from 75.108.117.50 on Sun, 27 Jan 2019 21:22:27 UTC All use subject to https://about.jstor.org/terms 2005 Clair, Beatty, and Maclean 85

FIGURE 1 A Conceptual Model of the Decision to Pass or Reveal

Individual differences Choice to pass or to reveal Individual and * Propensity toward interpersonal risk taking costs and benefits * Self-monitoring of revealing * Developmental stage * Motives or passing

Interpersonal and environmental context

* Organizational diversity climate * Professional and industry norms * Legal protections * Target relationships and characteristics

is weak or nonexistent, and monitoring for rule Finally, the presence of other individuals who breaking is ineffective, the compliance function have publicly revealed an invisible difference is itself is marginalized and undermined in the likely to facilitate the decision to reveal, espe eyes of organization members. Consequently, cially if the "out" others do not seem to have policies are perceived as symbolic in nature? suffered any negative consequences as a result not to be taken seriously (MacLean, 2003). This of their revelation (Ragins & Cornwall, 2001b). suggests that the original intent of the policy The existence of recognized groups of revealed will be undermined, and those with invisible others, such as a Gay-Straight Alliance, simi differences will be discouraged from revealing larly will act as a facilitating factor for reveal themselves if policies designed to protect stig ing. Conversely, individuals are more likely to matized groups are not enforced. pass at work if it appears that others with invis The degree to which organizational decisions ible identities are also hiding their differences. regarding hiring and promotion are transparent Industry and professional norms. Industry and and appear to be based on merit is also likely to professional norms for passing or revealing set influence whether a person feels safe revealing. the context within which employees may feel Transparency will be increased when individu pressure to tailor their behaviors to "fit in." For als have access to information about the bases example, referring to gay employees, Friskopp for hiring and promotion decisions. For exam and Silverstein note that people who "work for ple, tenure and promotion decisions for profes companies associated with the defense indus sors are more transparent when specific criteria try, that serve children, or that are connected for promotion are available to everyone, those with conservative or fundamentalist religious under consideration for promotion have access groups face special challenges in the workplace to materials collected for making promotion de ... [and] almost all are completely " cisions (i.e., recommendation letters), and peo (1995: 158). The military also encourages pass ple receive specific feedback about why they ing, as publicized in the "Don't ask, don't tell" were (or were not) granted a promotion. Trans policy of the 1990s (cf. Herek, Jobe, & Carney, parency and the appearance of meritocracy in 1996). People are more likely to be stigmatized if these decisions should make an individual feel they reveal their differences to others in organi safer revealing rather than passing at work. zations situated in industries or containing jobs

This content downloaded from 75.108.117.50 on Sun, 27 Jan 2019 21:22:27 UTC All use subject to https://about.jstor.org/terms 86 Academy of Management Review January

where passing for "normal" is implicitly or ex intimate information (Miller, Berg, & Archer, plicitly encouraged. In these situations, passing 1983). People may also share personal informa is seen as the most viable strategy to avoid tion with "wise others" (Goffman, 1963), who discrimination and to secure equal cultural, so seem knowledgeable about and sympathetic to cial, economic, and recreational advantages ward the particular difference. People seek out (Conyers & Kennedy, 1963). and prefer to interact with such individuals, Conversely, revealing may be encouraged in whom they believe support and validate their other industries or occupations, if doing so is identity (Schlenker, 1984). In particular, the sa consistent with shared values and meaning sys lience of the stigma in social interaction is re tems. For example, human services work relies duced when a person interacts with someone on delivering help to clients through a medium who shares the stigma (Gibbons, 1986). For ex of honest personal relationships, which requires ample, in their study of gays and lesbians, Corn an attitude of social tolerance (Halmos, 1970), so wall and Ragins (2001b) found that having gay professional norms in that occupation will pro supervisors or coworkers diminishes individu mote disclosure to "practice what they preach." als' fear of discrimination, thereby making dis Legal protections. Individuals who enjoy legal closure easier. In sum, we expect a person to protection from discrimination against an invis reveal more readily to a target who seems ible social identity at the local, state, or federal knowledgeable about, sympathetic toward, or level may be more likely to reveal than those similar to oneself because he or she also shares who are not afforded such protection (Ragins & the invisible difference. Cornwall, 2001b). The existence of this protec Research also suggests that gender-based be tion will vary by the nature of the invisible dif liefs can affect disclosure (Derlega et al., 1993). ference and by location. For instance, federal Some people may be less likely to reveal to law protects individuals from employment dis males, believing that men are more embar crimination on the basis of differences such as rassed or upset talking about intimate topics. race and disability status, but it does not protect Conversely, an individual may be more inclined people on the basis of their sexual orientation. to reveal to women, because, as suggested by Interpersonal context. The target (i.e., the per gender research, women's communication son with whom the holder of the invisible iden styles engender revealing personal information tity is interacting) influences passing or reveal (Tannen, 1990). ing behaviors. We highlight two features of the target: (1) the relationship an individual be Individual Differences lieves he or she has with the target and (2) the dispositional and demographic characteristics Early disclosure researchers found that indi of the target. viduals vary in their propensity to reveal per The specific nature of the relationship be sonal information (Jourard, 1971), suggesting tween parties shapes the decision to reveal. Re that disclosure behavior is at least partially de search suggests that people share more per termined by such individual differences as per sonal information about themselves with those sonality characteristics and social identity others whom they feel close to and trust. Reveal group memberships. In this section we consider ing is related to trust, and more trust leads to some of these differences. more personal intimate disclosure (Derlega, First, research suggests that individuals differ Metts, Petronio, & Margulis, 1993; Jourard, 1971). in the extent to which they are willing to take For example, close, trusting relationships have risks in decision making (Bromiley & Curley, been shown to positively influence the choice to 1992). Someone who has a higher propensity to reveal AIDS (Derlega, Lovejoy, & Winstead, ward risk taking is more likely to make riskier 1998), sexual orientation (Cain, 1991; Herek, 1996; decisions than someone with a lower propensity Ragins & Cornwall, 2001a), and illness (Greene, (Sitkin & Weingart, 1995). As we mentioned ear 2000). lier, someone who reveals information about a In addition, dispositional and demographic potentially stigmatizing social identity faces characteristics of a target can shape an individ stigma risk, or the possibility that he or she may ual's willingness to reveal. Some people are bet be stigmatized as a result of his or her revela ter at placing others at ease and at eliciting tions and suffer real, negative work-related con

This content downloaded from 75.108.117.50 on Sun, 27 Jan 2019 21:22:27 UTC All use subject to https://about.jstor.org/terms 2005 Clair, Beatty, and MacLean 87

sequences. We expect that individuals who are as high self-monitors, but they may not be as dispositionally less risk averse will be more good at managing impressions because they likely to reveal a stigmatizing invisible social are likely to be less effective at tailoring their identity. Conversely, an individual who has a image in ways that benefit their self-interest lower risk-taking propensity will be more likely (Gangestad & Snyder, 2000; Snyder & Copeland, to choose to pass at work. 1989). Thus, a low self-monitor in an interracial Second, an individual's self-monitoring ten relationship may be unskilled at effectively dencies may also influence disclosure. Self making his or her relationship appear "normal" monitoring refers to the degree to which an in to others. dividual observes, regulates, and controls how Third, models of adult development also sug well he or she is fulfilling the social expecta gest that one's developmental stage may influ tions of his or her role within a particular con ence one's propensity to reveal or pass. For ex text (Snyder, 1987). Those individuals who are ample, models of racial identity development high self-monitors closely monitor how they are (i.e., Tatum, 1997) suggest that racial minorities perceived in their interactions with others and at the highest stage of development fully em are more likely to behave in socially acceptable brace their racial heritage as an equally valid ways. In contrast, low self-monitors are not as identity, even though society denigrates their concerned with the social appropriateness of racial background. Highly developed individu their behavior and will act in a manner more als will tend to freely express their racial heri "true to themselves" (Kilduff & Day, 1994), with tage, although it may not be visible to others. expressive behaviors reflecting their inner atti Scholarly research on gays and lesbians tudes and emotions (Gangestad & Snyder, 2000). (Cass, 1984; Moorhead, 1999; Warren, 1974), as We expect high self-monitors to have the larg well as illness (Adams, Pill, & Jones, 1997), also est repertoire of passing and revealing behav implicitly and explicitly suggests that those in iors, because they are likely to be sensitive to dividuals who are "out" rather than closeted are situational demands; whether and how a person more sophisticated and self-assured than those passes or reveals may depend on perceptions of who choose to pass. However, new models of the social environment. Further, the high self biracial identity (self-views about "who I am" of monitor should be more effective at judging those from biracial heritage) suggest that this whether he or she will suffer negative conse idea may be too simplistic. For example, one quences from revealing an invisible difference study reveals that individuals with biracial at work. For example, Anderson and R?ndlet backgrounds express their identities in a variety (1993) assert that high self-monitors are more of ways, including altering their behaviors or skilled at assessing whether revealing that they mannerisms to suit the situation (i.e., a person are gay or lesbian will lead to negative conse changes vernacular and body language to fit quences. the situation when with black versus white col Self-monitoring theory also predicts that low leagues; Rockquemore & Brunsma, 2001). Re self-monitors are more likely to reveal, because search on biracial identity suggests that this doing so expresses "who they really are." Yet behavior is a true expression of a person's mul the additional consideration of stigma theory tiple identities, rather than an attempt to pass may mute this revealing effect. Stigma theory for black or white (Rockquemore & Brunsma, suggests that people within a culture commonly 2001). This suggests that predictions of passing agree on which characteristics are considered versus revealing behavior may be more com stigmatizing, implying that low self-monitors plex than suggested by traditional models of are likely to perceive the risks of stigmatization adult development. and other possible negative outcomes (i.e., Finally, those individuals with additional vis workplace discrimination) associated with re ible stigmatizing social identities may be less vealing their invisible difference. For example, likely to reveal an invisible difference because we expect low self-monitors to know that society they already face possible bias in the workplace stigmatizes individuals in an interracial rela from a visible stigma. Friskopp and Silverstein tionship. For this reason, we expect low self (1995) illustrate how lesbians in top manage monitors with such a potentially stigmatizing ment were aware that they already faced possi invisible difference to attempt to pass as much ble gender bias and consequently were less

This content downloaded from 75.108.117.50 on Sun, 27 Jan 2019 21:22:27 UTC All use subject to https://about.jstor.org/terms 88 Academy of Management Review January

willing to reveal their sexual orientation. Simi ships, and individuals are likely to reveal lar examples can be found addressing age, race, personal information as their relationships ma and ethnicity, in which multiple potentially stig ture (Derlega, 1984). Individuals may choose to matized identities interact with each other (e.g., reveal an invisible social identity to a target in Creed & Scully, 2000; Reimann, 2001). the normal process of relationship building to gain general (Derlega et al., 1993; Personal Motives Rubin & Shenker, 1978). Additionally, informa tion is typically revealed in a reciprocal way Personal motives are also likely to influence (Derlega, Harris, & Chaikin, 1973; Taylor, 1968). disclosure decisions. We address four motives: Thus, people may feel compelled to disclose to (1) maintaining self-esteem and , (2) minimize social awkwardness (Herek, 1996). building or preserving relationships, (3) arrang As we mentioned earlier, a person's decision ing accommodation, and (4) creating social to reveal or pass can also be motivated by an change. While these motives drive disclosure interest in preserving or enhancing existing so decisions, they do not guarantee actual out cial relationships. For example, a person may comes obtained. For example, an individual choose to pass for straight at work to protect his may reveal that he has early Alzheimer's dis or her closeted partner or other family members ease to a colleague as a form of coping; how who work at the same organization from stigma ever, the colleague may respond with shock and tization. Conversely, a person is more likely to disgust. As we discuss later, a person is likely to reveal if passing compromises a significant re avoid future revelations if previous disclosure lationship. For example, a person is more likely has led to these negative outcomes. to reveal that he is gay if his partner is offended People reveal an invisible difference to main by being excluded from company functions. tain a coherent sense of self (Harry, 1993; Moor Another motive to reveal is so that one's dif head, 1999). Declaring one's difference allows ference can be accommodated or so that one can one to be a complete and integrated person receive a benefit within the workplace. Creed (Friskopp & Silverstein, 1995) and obviates the and Scully (2000) discuss how gay employees difficult task of constantly being engaged in in may reveal their sexual orientation in order to formation management to conceal aspects of request domestic partner benefits. Similarly, a oneself (Charmaz, 1991). A person reveals stig person must disclose a disability in order to matizing information to maintain control over receive workplace accommodation legislated the way in which others learn about the stigma under the Americans with Disabilities Act. This tizing difference and to positively frame the in presents a dilemma for people whose difference formation (Herek, 1996). Similarly, passing can would otherwise be unknown to others (Mat be an attempt to create one's own life narrative thews & Harrington, 2000). People also mention and to control the process of signification itself their difference to explain or justify unusual or (Schlossberg, 2001). Individuals may seek to unexpected behaviors. Sometimes a difference avoid acknowledging perceived flaws and can be a bargaining tool to achieve preferential weaknesses because it lowers their self-esteem treatment?a phenomenon Strauss (1975) calls (Rosenfeld, 1979). In either case, the choice to "secondary gain." In secondary gain, a person pass or reveal is linked to individuals' identity uses his or her difference to claim a benefit?for and their goals for self-presentation. example, being relieved of an unpleasant work Revealing is also a coping strategy for deal task. ing with stressful situations. People are moti Finally, revealing can be motivated by a de vated to share personal information to release sire to educate others and to change social con pent-up concerns and feelings to one another ditions at work. Some people feel a responsibil (Derlega et al., 1998). Therapeutic disclosure re ity to reveal or to educate others about the duces negative feelings that accompany secrecy special nature of their difference. Bernstein (Herek, 1996) and has both emotional and phys (1997) has illustrated that gays and lesbians re ical benefits (Jourard, 1971; Pennebaker, 1997). veal to explain gay identity and to change het A person may also reveal to gain social close erosexual people's perspectives. This motive al ness. Revealing personal information is a nor lows individuals to assert their identity and to mal and integral part of developing relation challenge social norms that lead to discrimina

This content downloaded from 75.108.117.50 on Sun, 27 Jan 2019 21:22:27 UTC All use subject to https://about.jstor.org/terms 2005 Clair, Beatty, and MacLean 89

tion. The civil rights and disability rights move decision to pass may compromise personal re ments grew out of this approach to revealing lationships outside of work, if a significant oth and claiming stigmatized identities (Anspach, er's existence is denied or his or her importance 1979; Fine & Asch, 1988; Scotch, 1988), and it has minimized. been a powerful force for social change. Revealing presents different types of risk. The person who chooses to reveal his or her invisible OUTCOMES FROM PASSING AND REVEALING social identity opens himself or herself to possi ble stigmatization, prejudice, and discrimina We contend here that people with invisible tion at work. Upon revealing, this person may social identities are aware of the risks, costs, worry about how others will react to his or her and benefits associated with revealing their revelations and whether these revelations will stigmatizing difference. As we depict in Figure lead to negative personal consequences. How 1, we predict that individuals engage in a type ever, the benefit of minimizing dissonance be of cost-benefit analysis in deciding whether to tween who one "really is" and the image that pass or reveal. In this section we discuss some one projects to others may be worth the risk of of the outcomes of passing and revealing. stigmatization (Shallenberger, 1994). This indi The risks of passing occur at individual and vidual may also benefit from building closer interpersonal levels. For the individual, passing social relationships, as we discussed above, by is a "sequestering of the self," which raises is revealing personal information about the self. sues of authenticity (Leary, 1999: 86). Conse Revealing may allow an individual to acknowl quently, passers may experience psychological edge significant others and family members, strain from feeling like a fraud (Goffman, 1963) rather than deny their existence or minimize because of a need to construct credible and con their importance. Finally, an individual can af sistent fabrications about their lives and to en fect social changes within the organization by sure that colleagues and clients remain un strategically deploying his or her invisible dif aware of their invisible difference. As Smart and ference (Creed & Scully, 2000). Wegner note, "Concealing a stigma leads to an People with an invisible social identity will inner turmoil that is remarkable for its intensity inevitably need to decide whether, when, and and its capacity for absorbing an individual's how to reveal their invisible, potentially stigma mental life" (2000: 221). Ragins and Cornwall tized social identity in social interactions at (2001b) found that fears associated with the de work. They are likely to expend energy and to cision to reveal sexual orientation were more experience increased stress in every new social deleterious than disclosure itself, with respect to relationship at work that requires a decision to job attitudes and psychological strain. pass or reveal. In interpersonal relationships, the passer may find that he or she is more isolated from col INFLUENCE OF OUTCOMES ON FUTURE leagues and less effective at working with cli BEHAVIORS ents because of the need to keep others at a personal distance in order to hide a difference. Figure 1 includes a feedback loop, which sug Work colleagues expect to have some degree of gests that a person's outcomes from revealing knowledge about each other's personal life. This will influence his or her future choices to pass or knowledge helps establish effective work rela reveal. Not surprisingly, we expect that a person tionships (Kronenberger, 1991; Schneider, 1987). is more likely to reveal in the future if he or she Relationships can become strained in the work believes that prior revealing led to positive out place when interaction partners become suspi comes. For example, a person will be more cious because of the lack of personal disclosure likely to reveal that he or she has asthma to (Herek, 1996). Associations with others thus be additional colleagues at work if prior revealing come an ordeal for the passer and something to led to workplace accommodations. Conversely, be avoided (Jones et al., 1984). Also, passers may we expect this person may pass in the future if have more limited networks and mentoring re he or she encountered a negative outcome, such lationships (Day & Schoenrade, 1997), resulting as being denied promotion. in long-term deleterious effects on career ad These predictions arise from theories of social vancement. Further, as already noted, a person's cognition, which suggest that past social en

This content downloaded from 75.108.117.50 on Sun, 27 Jan 2019 21:22:27 UTC All use subject to https://about.jstor.org/terms 90 Academy of Management Review January counters frame in one's mind what such encoun that allow for increased creativity and innova ters "should" entail, thereby shaping future ex tion. The second perspective hypothesizes that pectations and perceptions (Fiske & Taylor, increased diversity hinders group performance, 1984). Thus, positive coming out experiences be because social categorization processes and get more revealing behaviors, whereas negative concomitant skewed attributions and stereotyp experiences are likely to generate future pass ing undermine communication and cooperation. ing. This is likely to be true even though prior Both of these perspectives rely on individuals experiences may not accurately predict out in workgroups being able to identify differences comes of future social interactions. in each other. What happens when differences aren't visible? If individuals reveal their invisi DISCUSSION ble social identities to group members only after establishing positive relationships and trust, do This article illuminates the role of invisibility these groups experience higher levels of group in organizational diversity dynamics. Although cohesion and performance than groups with vis organizational scholars acknowledge that de ible diversity, whose observable composition mographic differences can be visible as well as may have precluded the formation of close invisible, in most diversity research scholars fo bonds? cus on observable differences. The literature on race, on sexual orientation, and on chronic ill ness suggests that those with invisible social Group Dynamics identities face a choice that visibly different oth ers do not face: whether or not to reveal their In the same vein, there is a need to develop an invisible social identity or to pass as a member understanding of the process by which invisible of the (nonstigmatized) majority. By synthesizing social identities affect group dynamics. The way these bodies of literature and explicitly ac a group deals with revelations of those with knowledging the unique dynamics associated invisible social identities is likely to influence with invisible social identities in the workplace, the group's future interactions. Groups that fail we accomplish several things: (1) we describe to respond sensitively to a person's revelations how individuals manage invisible social identi about a stigmatizing invisible social identity ties by passing and revealing in organizations, may suffer from decreased cohesion among as well as discuss the complexities inherent in their members. Research about this will not only those strategies; (2) we explore individual and enrich diversity literature but also will benefit contextual factors that affect an individual's practicing managers, who may be aware of in choice to pass or to reveal; (3) we suggest con visible differences within groups they manage sequences associated with both revealing and but uncertain about managing group responses passing; and (4) we open up new avenues of to them. inquiry for organization studies researchers. We address some of the potential areas for further research below. Trust In addition to considering what will be dis Group Performance closed to whom, individuals with invisible iden Given a tradition of diversity research at the tities must consider when to disclose. Our model group level of analysis (Williams & O'Reilly, suggests that trust and intimacy in coworker 1998), one significant area of research is to in relationships are precursors to revealing. Trust vestigate the impact of the visibility of social develops over time, in response to earlier inter identities on group performance. Williams and actions, increasing or decreasing based on past O'Reilly (1998) categorize this group-level diver interactions, as indicated in the feedback loop of sity research into two competing and conflicting our disclosure model. In the future, other schol perspectives. The first perspective suggests that ars may explore the decision-making process increased diversity improves group perfor through which individuals assess whether the mance through enhanced information networks threshold of intimacy and trust necessary for and more and varied perspectives on the world disclosure has been reached.

This content downloaded from 75.108.117.50 on Sun, 27 Jan 2019 21:22:27 UTC All use subject to https://about.jstor.org/terms 2005 Clair, Beatty, and MacLean 91

Social Identity which may dissuade an individual from disclos ing unless trust in that person is extremely high. There are likely to be specific factors unique to particular social identities that influence how people manage these identities in social inter Organizational Change actions at work. As an example, invisible chronic illnesses are different from other social Finally, this work opens the door for research on how the decision to pass or reveal may affect identities because of the potential intrusiveness the organization. Individuals with invisible so of periodic symptoms. Those with an invisible cial identities often conceal their differences un chronic illness must find ways to explain or hide til they have built a network of mentors, friends, symptoms from others. As another example, a colleagues, and supporters within their organi pregnancy slowly moves from invisible to visi zation. Revealing invisible social identities to ble. How do women manage social reactions to an established, strong, and supportive network the changes in their bodies during pregnancy at work? How does the movement from invisible to may influence the way the revealed difference is viewed by members of the organization and visible during pregnancy affect passing and re may have far-reaching effects on the overall cli vealing behaviors? How do women make deci mate of the organization. Similarly, passing sions about when and how to reveal, knowing may simply further institutionalize stigmatiza that their pregnancy implies they will be absent tion of particular social identities in given orga from work or that their performance may suffer nizational contexts. in the future if their child care responsibilities conflict with work demands? The possible destigmatizing effects of reveal ing strategies present a new area for organiza As these examples illustrate, from a practice tional change researchers to explore (cf. Creed & perspective, managers may need to be aware of Scully, 2000). This particular area provides fer the range of different issues important for dis tinct invisible identities as well as the common tile ground for exploring the macro-micro link in organization studies. Case study researchers alities among them. Grounded qualitative re should attempt to trace the evolution of the mi search is one method to build rich insights on crosocial actions of one or more individuals who particular types of invisible social identities reveal an invisible difference in an organization and to create advice useful to managers for spe that translates into macrosocial changes in the cific types of invisible social identities. climate of that organization.

Social Networks Measurement Issues

The importance of a target relationship in the We close by briefly considering measurement decision concerning whether to reveal suggests issues. One of the particular challenges for that social network theory may offer clues re scholars is that this social identity is invisible. garding how individuals manage their invisible In other words, scholars who conduct field re social identity. In social network theory it is search on invisible social identities rely on assumed that people are embedded within a those with such invisible differences to reveal network of relationships that create opportuni their social identity. Samples may be systemat ties for and constraints on individual action ically biased, because those who agree to par (Granovetter, 1985). Our model raises questions ticipate in such research (i.e., who reveal a dif for social network researchers, including ference to the researcher) may also be higher in whether the strength of a tie with a target influ risk-taking propensity or willingness to trust ences an individual's passing or revealing be others. However, scholars may mitigate bias by haviors. Given the potential stigma associated ensuring confidentiality, thus attracting sub with revealing, another research question con jects who are less likely to reveal an invisible cerns whether or not individuals consider the social identity to strangers. Scholars may also network relationships of others when deciding benefit by drawing subjects from interest groups whether to pass or reveal their difference. A rather than from one particular organization. For person with many social ties and high centrality example, Beatty (2003) drew participants with has the ability to "spread the word" to others, invisible chronic illnesses from support groups

This content downloaded from 75.108.117.50 on Sun, 27 Jan 2019 21:22:27 UTC All use subject to https://about.jstor.org/terms 92 Academy of Management Review January

for these illnesses. This approach is likely to Bromiley, P., & Curley, S. 1992. Individual difference in risk make a closeted participant feel safer, because taking. In J. Yates (Ed.), Risk taking behavior: 87-132. New York: Wiley. he or she can avoid being revealed at work. Another research challenge rests with opera Cain, R. 1991. Stigma management and gay identity devel opment. Social Work. 36: 67-73. tionalizing revealing and passing behaviors. Cass, V. C. 1984. Homosexual identity formation: Testing a For example, scholars need to consider what it theoretical model. Journal of Sex Research. 20: 143-167. means to "signal" one's invisible difference to a Centers for Disease Control. 2002. Chronic disease overview. target person. The signaler may believe that he http://www.cdc.gov/nccdphp/overview.htm, August 30. or she has signaled an invisible difference to another in a social interaction, even though the Charmaz, K. 1991. Good days, bad days: The self in chronic illness and time. New Brunswick, NJ: Rutgers University target fails to perceive this signal. Is it sufficient Press. to say that behaviors constitute signals even if Charmaz, K. 2000. Experiencing chronic illness. In G. L. Al the target fails to recognize them as such? Ad brecht, R. Fitzpatrick, & S. C. Scrimshaw (Eds.), Hand ditionally, differences among the passing be hook of social studies in health and medicine: 277-292. haviors of fabrication, concealment, and discre Thousand Oaks, CA: Sage. tion (Herek, 1996) are subtle. Scholars will need Chrobot-Mason, D., Button, S. B., & Declimenti, J. D. 2001. to demonstrate that these differences hold up to Sexual identity management strategies: An exploration standards of empirical measurement. These are of antecedents and consequences. Sex Roles. 45:321-336. just some of the measurement issues that schol Conyers, J. E., & Kennedy, T. H. 1963. Negro passing: To pass ars are likely to face when studying invisible or not to pass. Phylon. 24: 215-223. social identities at work. Cox, T. 1993. Cultural diversity in organizations. San Fran While theoretical and empirical research on cisco: Berrett-Koehler. invisible social identities to some extent consti Creed, W. E. D., & Scully, M. A. 2000. Songs of ourselves: tutes uncharted territory, we believe that those Employees' deployment of social identity in workplace who undertake the challenge will be richly re encounters. Journal of Management Inquiry. 9: 391-413. warded by new insights regarding invisible as Crocker, J., & Lutsky, N. 1986. Stigma and the dynamics of well as visible social identities within the work social cognition. In S. C. Ainley, G. Becker, & L. M. Coleman (Eds.), The dilemma of difference: A multidisci place. plinary view of stigma: 95-121. New York: Plenum Press. Crocker, J., Major, B., & Steele, C. 1998. . In D. T. REFERENCES Gilbert, S. T. Fiske, & G. Lindzey (Eds.), Handbook of social (4th ed.): 504-553. New York: McGraw-Hill. Adams, S., Pill, R., & Jones, A. 1997. Medication, chronic illness, and identity: The perspective of people with Day, N. E., & Schoenrade, P. 1997. Staying in the closet versus asthma. Social Science and Medicine, 45: 189-201. coming out: Relationships between communication about sexual orientation and work attitudes. Personnel Anderson, L. R., & R?ndlet, L. 1993. Self-monitoring and life Psychology. 50: 147-164. satisfaction of individuals with traditional and nontra ditional sexual orientations. Basic and Applied Social Delbecq, A. 2002. Hidden holy people. Paper presented at the Psychology, 14: 345-361. annual meeting of the Academy of Management, Denver. Anspach, R. 1979. From stigma to identity politics: Political Derlega, V. J. 1984. Self-disclosure and intimate relation activism among the physically disabled and former ships. In V. J. Derlega (Ed.), Communication, intimacy, mental patients. Social Science and Medicine, 13A: 765 and close relationships: 1-10. Orlando: Academic Press. 773. Derlega, V. J., Harris, M. S., & Chaikin, A. L. 1973. Self disclosure, reciprocity, liking, and the deviant. Journal Beatty, J. E. 2003. Chronic illness disclosure in the workplace. Unpublished doctoral dissertation, Boston College, Bos of Experimental Social Psychology. 9: 277-284. ton. Derlega, V. J., Lovejoy, D., & Winstead, B. A. 1998. Personal Bernstein, M. 1997. Celebration and suppression: The strate accounts of disclosing and concealing HIV-positive test results. In V. J. Derlega & A. Barbee (Eds.), HIV and social gic uses of identity by the lesbian and gay movement. interaction: 147-164. Thousand Oaks, CA: Sage. American Journal of Sociology, 103: 531-565. Bernstein, M., & Reimann, R. 2001. Queer families and the Derlega, V. J., Metts, S., Petronio, S., & Margulis, S. T. 1993. Self-disclosure. Newbury Park, CA: Sage. politics of visibility. In M. Bernstein & R. Reimann (Eds.), Queer families, queer politics: Challenging culture and Driscoll, J. M., Kelley, F. A., & Fassinger, R. E. 1996. Lesbian the state: 1-20. New York: Columbia University Press. identity and disclosure in the workplace: Relation to occupational stress and satisfaction. Journal of Voca Bowman, E. A. 2001. Black like who? Young authors stir up tional Behavior, 48: 229-242. the melting pot of racial diversity. Black Issues Book Review. 3: 24. Ely, R. J., & Thomas, D. A. 2001. Cultural diversity at work:

This content downloaded from 75.108.117.50 on Sun, 27 Jan 2019 21:22:27 UTC All use subject to https://about.jstor.org/terms 2005 Clair, Beatty, and MacLean 93

The moderating effects of work group perspectives on Jones, E. E., Farina, A., Hastorf, A. H., Markus, H., Miller, D. T., diversity. Administrative Science Quarterly, 46: 229-273. & Scott, R. A. 1984. Social stigma: The psychology of marked relationships. New York: Freeman. Fine, M., & Asch, A. 1988. Disability beyond stigma: Social interaction, discrimination, and activism. Journal of So Jones, N. A., & Smith, A. S. 2001. The two or more races cial Issues, 44(1): 3-21. population: 2000. Census brief No. C2KBR/01-6. Washing ton, DC: U.S. Census Bureau. Fiske, S. T., & Taylor, S. E. 1984. Social cognition. New York: Random House. Jourard, S. M. 1971. The transparent self (2nd ed.). New York: Van Nostrand Reinhold. Frable, D., Blackstone, T., & Sherbaum, C. 1990. Marginal and mindful: Deviants in social interaction. Journal of Per Kilduff, M., & Day, D. V. 1994. Do chameleons get ahead? The sonality and Social Psychology, 59: 140-149. effects of self-monitoring on managerial careers. Acad emy of Management Journal. 37: 1047-1061. Friskopp, A., & Silverstein, S. 1995. Straight jobs, gay lives. New York: Touchstone. Kronenberger, G. K. 1991. Out of the closet. Personnel Jour nal. 70: 40-44. Gangestad, S. W., & Snyder, M. 2000. Self-monitoring ap praisal and reappraisal. Psychological Bulletin, 126: Leary, K. 1999. Passing, posing, and "keeping it real." Con 530-555. stellations. 6: 85-96.

Gibbons, F. X. 1986. Stigma and interpersonal relationships. Leary, M. R., & Schreindorf er, L. S. 1998. The stigmatization of In S. C. Ainley, G. Becker, & L. Coleman (Eds.), The HIV and AIDS: Rubbing salt in the wound. In V. J. Der dilemma of difference: A multidisciplinary view of stig lega & A. P. Barbee (Eds.), HIV and social interaction: 12 ma: 123-144. New York: Plenum Press. 29. Thousand Oaks, CA: Sage. Goffman, E. 1963. Stigma. Englewood Cliffs, NJ: Prentice Link, B. G., & Phelan, J. C. 2001. Conceptualizing stigma. Hall. Annual Review of Sociology. 27: 363-385. Granovetter, M. 1985. Economic action and social structure: MacLean, T. 2003. How structural decoupling facilitates or The problem of embeddedness. American Journal of So ganizational misconduct. In D. H. Nagao (Ed.), Proceedings ciology, 91: 481-510. of the Academy of Management: CD (ISSN 1543-8643). Greene, K. 2000. Disclosure of chronic illness varies by topic Matthews, C. K., & Harrington, N. G. 2000. Invisible disability. and target: The role of stigma and boundaries in will In D. O. Braithwaite & T. L. Thompson (Eds.), Handbook ingness to disclose. In S. Petronio (Ed.), Balancing the of communication and people with disabilities: 405- 421. secrets of private disclosures: 123-135. Mahwah, NJ: Law Mahwah, NJ: Lawrence Erlbaum Associates. rence Erlbaum Associates. Michael, R. T., Gagnon, J. H., Laumann, E. O., & Kolata, G. Halmos, P. 1970. The personal service society. New York: 1994. Sex in America. New York: Little, Brown. Schocken. Miller, L. C, Berg, J. H., & Archer, R. L. 1983. Openers: Indi Harrison, D. A., Price, K. H., & Bell, M. P. 1998. Beyond rela viduals who elicit intimate self-disclosure. Journal of tional demography: Time and the effects of surface- and Personality and Social Psychology. 44: 1234-1244. deep-level diversity on work. Academy of Management Moorhead, C. 1999. Queering identities: The roles of integrity Journal, 41: 96-107. and belonging in becoming ourselves. Journal of Gay. Harrison, D. A., Price, K. H., Gavin, J. H., & Florey, A. T. 2002. Lesbian, and Bisexual Identity. 4: 327-343. Time, teams, and task performance: Changing effects of Pennebaker, J. W. 1997. Opening up: The healing power of surface- and deep-level diversity on group functioning. expressing emotions. New York: Guilford Press. Academy of Management Journal. 45: 1029-1045. Pinder, R. 1988. Striking balances: Living with Parkinson's Harry, J. 1993. Being out: A general model. Journal of Homo disease. In R. Anderson & M. Bury (Eds.), Living with sexuality. 26(1): 25-39. chronic illness: 67-88. London: Un win Hyman. Herek, G. M. 1996. Why tell if you are not asked? Self Pinder, R. 1995. Bringing back the body without the blame?: disclosure, intergroup contact, and heterosexuals' atti The experience of ill and disabled people at work. So tudes toward lesbians and gay men. In G. M. Herek, J. B. ciology of Health & Illness. 17(5): 605-631. Jobe, & R. M. Carney (Eds.), Out in force: Sexual orien Pisares, E. H. 1999. Daly City is my nation: Race, imperialism, tation and the military: 197-225. Chicago: University of and the claiming of Pinay/Pinoy identities in Filipino Chicago Press. American culture. Unpublished doctoral dissertation, Herek, G. M., Jobe, J. B., & Carney, R. M. (Eds.). 1996. Out in University of California at Berkeley. force: Sexual orientation and the military. Chicago: Uni Ragins, B. R., & Cornwall, J. M. 2001a. Pink triangles: Ante versity of Chicago Press. cedents and consequences of perceived workplace dis Iphofen, R. 1990. Coping with a "perforated life": A case crimination against gay and lesbian employees. Journal study in managing the stigma of petit mal epilepsy. of Applied Psychology. 86: 1244-1261. Sociology, 24: 447-463. Ragins, B. R., & Cornwall, J. M. 2001b. Walking the line: Fear Joachim, G., & Acorn, S. 2000. Living with chronic illness: The and disclosure of sexual orientation in the workplace. interface of stigma and normalization. Canadian Jour Paper presented at the annual meeting of the Academy nal of Nursing Research. 32(3): 37-48. of Management, Washington, DC.

This content downloaded from 75.108.117.50 on Sun, 27 Jan 2019 21:22:27 UTC All use subject to https://about.jstor.org/terms 94 Academy of Management Review January

Register, C. 1987. Living with chronic illness: Days of pa zation of individuals on the basis of multiple social tience and passion. New York: Free Press. features. Journal of Personality and Social Psychology, Reimann, R. 2001. Lesbian mothers at work. In M. Bernstein & 62: 207-218. R. Reimann (Eds.), Queer families, queer politics: Chal Stephan, W. G., & Stephan, C. W. 1985. Intergroup anxiety. lenging culture and the state: 254-271. New York: Co Journal of Social Issues, 41: 157-175. lumbia University Press. Strauss, A. 1975. Chronic illness and the quality of life. St. Rockquemore, K. A., & Brunsma, D. 2001. Beyond black: Bira Louis: Mosby. cial identity in America. Thousand Oaks, CA: Sage. Sullivan, M. 2001. Alma mater: Family "outings" and the Rosenberg, S. 1997. Multiplicity of selves. In R. D. Ashmore & making of the modern other mother. In M. Bernstein & L. Jussim (Eds.), Self and identity: Fundamental issues: R. Reimann (Eds.), Queer families, queer politics: Chal 23-45. New York: Oxford University Press. lenging culture and the state: 231-254. New York: Colum Rosenfeld, L. B. 1979. Self-disclosure avoidance: Why I am bia University Press. afraid to tell you who I am. Communication Mono Tajfel, H. 1981. Human groups and social categories: Studies graphs, 46: 63-74. in social psychology. Cambridge: Cambridge University Rothenberg, S., & Scully, M. 2002. The embarrassment of Press. riches: The "coming out" experiences of the wealthy. Tannen, D. 1990. You just don't understand: Women and men Paper presented at the annual meeting of the Academy in conversation. New York: Ballantine Books. of Management, Denver. Royer, A. 1998. Life with chronic illness: Social and psycho Tatum, B. D. 1997. Why are all the black kids sitting together in the cafeteria? And other conversations about the de logical dimensions. Westport, CT: Praeger. velopment of racial identity. New York: Basic Books. Rubin, Z., & Shenker, S. 1978. Friendship, proximity, and self-disclosure. Journal of Personality. 46: 1-22. Taylor, D. A. 1968. The development of interpersonal rela tionships: Social penetration process. Journal of Social Schlenker, B. 1984. Identities, identifications, and relation Psychology, 75: 79-90. ships. In V. J. Derlega (Ed.), Communication, intimacy, and close relationships: 71-104. Orlando: Academic Taylor, V., & Raeburn, N. C. 1995. Identity politics as high-risk Press. activism: Career consequences for lesbian, gay, and bisexual sociologists. Social Problems, 42: 252-273. Schlossberg, L. 2001. Introduction. In M. C. Sanchez & L. Schlossberg (Eds.), Passing: Identity and interpreta Thorne, S. E. 1993. Negotiating health care: The social context tion in sexuality, race, and religion: 1-12. New York: New of chronic illness. Newbury Park, CA: Sage. York University Press. Triandis, H., Kurowski, L., & Gelfand, M. 1994. Workplace Schneider, B. E. 1987. Coming out at work: Bridging the pri diversity. In H. Triandis, M. Dunnette, & L. Hough (Eds.), vate/public gap. Work and Occupations. 13: 463-487. Handbook of industrial and organizational psychology, Schneider, J. W., & Conrad, P. 1980. In the closet with illness: vol. 4: 769-827. Palo Alto, CA: Consulting Psychologists Epilepsy, stigma potential, and information control. So Press. cial Problems. 28: 32-44. Tsui, A. S., & Gutek, B. A. 1999. Demographic differences in Scotch, R. K. 1988. Disability as the basis for a social move organizations: Current research and future directions. ment: Advocacy and the politics of definition. Journal of Lanham, MD: Lexington Books. Social Issues. 44: 159-172. Turner, J. 1982. Toward a cognitive definition of the group. In Shallenberger, D. 1994. Professional and openly gay: A nar H. Tajfel (Ed.), Social identity and intergroup relations: rative study of experience. Journal of Management In 15-40. Cambridge: Cambridge University Press. quiry, 3: 119-142. Vickers, M. H. 2001. Work and unseen chronic illness. London: Sitkin, S. B., & Weingart, L. R. 1995. Determinants of risky Routledge. decision-making behavior: A test of the mediating role Warren, C. A. B. 1974. Identity and community in the gay of risk perceptions and risk propensity. Academy of world. New York: Wiley. Management Journal, 38: 1573-1592. Weitz, R. 1990. Living with the stigma of AIDS. Qualitative Smart, L., & Wegner, D. M. 2000. The hidden costs of hidden stigma. In T. F. Heatherton, R. E. Kleck, M. R. Hebl, & J. G. Sociology. 13: 23-38. Hull (Eds.), The social psychology of stigma: 220-241. Williams, K. Y., & O'Reilly. C. A., III. 1998. Demography and New York: Guilford Press. diversity in organizations: A review of 40 years of re Snyder, M. 1987. Public appearances, private realities: The search. Research in Organizational Behavior. 20: 77-140. psychology of self-monitoring. New York: Freeman. Woods, J. D. 1994. The corporate closet. New York: Free Press. Snyder, M., & Copeland, J. 1989. Self-monitoring processes in Woods, S. E., & Harbeck, K. M. 1991. Living in two worlds: The organizational settings. In R. A. Giacalone & P. R. Rosen identity management strategies used by lesbian physi feld (Eds.), in the organization: cal educators. In K. Harbeck (Ed.), Coming out of the 7- 19. Hillsdale, NJ: Lawrence Erlbaum Associates. classroom closet: Gay and lesbian students, teachers, Stangor, C, Lynch, L., Duan, C, & Glass, B. 1992. Categori and curricula: 141-166. New York: Harrington Park Press.

This content downloaded from 75.108.117.50 on Sun, 27 Jan 2019 21:22:27 UTC All use subject to https://about.jstor.org/terms Clair, Beatty, and MacLean

Judith A. Clair is an associate professor of organizational studies in the Carroll School of Management at Boston College. She received her Ph.D. from the University of Southern California. Currently, her research interests include the effects of critical events, such as crisis and downsizing, on individuals and organizations and multi cultural diversity in organizations.

Joy E. Beatty is an assistant professor of organizational behavior in the School of Management at the University of Michigan-Dearborn. She received her Ph.D. from Boston College. Her research interests include diversity, the effects of chronic illness on employees and organizations, and pedagogical issues in business schools.

Tammy L. MacLean is an assistant professor of management at Suffolk University. She received her Ph.D. in organization studies from Boston College. Her current research interests include the workplace dynamics associated with invisible social identities, how the process of decoupling affects organizational behavior, and organizational misconduct.

This content downloaded from 75.108.117.50 on Sun, 27 Jan 2019 21:22:27 UTC All use subject to https://about.jstor.org/terms