WALKING THE LINE:  FEAR AND DISCLOSURE OF SEXUAL ORIENTATION

 IN THE WORKPLACE

One of the most important career decisions faced by gay employees is whether they disclose their sexual orientation to others in the workplace.  Most gay employees limit the disclosure of their sexual identity at work, and up to a third of gay employees do not disclose their sexual orientation to anyone at work  (cf., review by Croteau, 1996).  Given the fact that between 25-66% of gay employees experience workplace discrimination based on their sexual orientation (Croteau, 1996), the decision to "be out at work" is not an easy decision, and is often preceded by anxiety, and fears of rejection and discrimination (Friskopp & Silverstein, 1996; Woods, 1994).  In fact, the fear of negative consequences of disclosure may have a greater impact on the psychological health and well-being of gay employees than the actual decision itself, which often  brings a sense of relief to gay employees (Friskopp & Silverstein, 1996; Woods, 1994).  Since gay men and lesbians constitute between 4-17% of the workforce (Gonsiorek & Weinrich, 1991), a larger proportion than many other minority groups, it is important to gain insights into one of the most important career decisions faced by this under-studied population.

Only a few studies have examined the decision to disclose sexual orientation in the workplace.  These studies test the view that those who disclose their sexual orientation to others at work achieve a congruence in their public and private identities (Ellis, 1996), obtain a sense of psychological wholeness and well-being (Cass, 1984), and are relieved of the debilitating strain of secrecy involved with leading a double life (Fassinger, 1996).  Although this view seems reasonable, empirical tests of this idea yield inconsistent and contradictory findings.  Disclosure has been found to have positive, negative and non-significant effects on work attitudes, psychological strain and compensation (Croteau & Lark, 1995; Day & Schoenrade, 1997; Driscoll, Kelley & Fassinger, 1996; Ellis & Riggle, 1995; Schneider, 1987). 

One reason for these contradictory findings is that we know little about the processes involved with a gay employee's decision to come out at work.  In fact, the actual decision to disclose at work, while important and complex, may have less of an impact than the immediate fears preceding that decision.  Croteau (1996), for example, observes that fear and anticipation of discrimination as a result of disclosure is pervasive among gay employees, and may result in psychological distress and decreased job performance even in the absence of actual discrimination.  This suggests that work and career attitudes may be related more to the fear associated with being out at work, than the actual decision itself.  Although the fear of coming out is an important variable to consider when examining the disclosure process, this variable has been overlooked in prior research.

Along similar lines, we know little about the factors that lead gay employees to share their sexual identity with others in the workplace.  Emerging views suggest that the decision to disclose at work may not reflect the identity development of individuals as much as their immediate work environment  (Cain, 1991; Fassinger, 1995; McCarn & Fassinger, 1996).  These views challenge the assumption that gay employees who are out at work are automatically more socially well-adjusted, have stronger gay identities, and are happier at work than those who conceal their sexual orientation.  Cain (1991) notes that the individual focus that has been prevalent in the literature to date has led researchers to overlook the situational and relational factors that affect the management of gay identity in the workplace.  It is clear that we need to understand not only the consequences associated with the disclosure process, but also the factors that affect gay employees' fear of disclosure and their ultimate decision to disclose at work.

Accordingly, this study has two primary purposes.  First, we use a large national sample of gay employees to examine the antecedents that influence gay employees' fears about disclosing their sexual orientation in the workplace, as well as their actual decision to disclose at work.  In particular, we investigate the impact of the work group environment and prior experiences of workplace discrimination on fear of disclosure and the decision to disclose.  The second purpose of this study is to explore the impact of fear of disclosure and the decision to disclose on a number of work-related outcome variables,  such as work and career attitudes, psychological strain, compensation and promotion. 

Although gay men and lesbians constitute a substantial proportion of the workforce, there has been a noticeable lack of theory and research on sexual orientation in the workplace.  While diversity researchers strive to understand the unique experiences related to race, ethnicity, and gender, we know little about the unique workplace experiences of gay employees.  This study contributes to our understanding of one of the most difficult career decisions faced by gay employees, and provides a foundation for future theory building in this emerging and sometimes controversial area.

THEORETICAL AND LITERATURE REVIEW

Disclosure in the Workplace

Psychological models of gay identity development.  Although there is a lack of theory on the establishment of a gay identity in the workplace, psychologists have been theorizing about the general process of gay identity development for some time.  This theorizing has yielded a number of models of gay identity development (e.g., Cass, 1984; Coleman, 1982; Morris, 1997; Troiden, 1989).  Although these psychological models do not specifically address workplace disclosure, they provide an important foundation for understanding the complex processes and fears underlying gay employees' decisions to disclose their gay identity at work.

Most theories of gay identity development include two initial stages (cf. reviews by Gonsiorek & Rudolph, 1991; Sophie, 1986).  The first stage involves a general awareness of homosexual feelings.  This stage often includes a sense of confusion, , defensiveness, fear and denial.  The second stage involves emotional and behavioral testing and exploration.  This may involve "coming out to oneself" or the self-labeling of oneself as gay, seeking out other gay people, and establishing the first gay sexual relationship.  A common theme underlying these early stages is the fear that disclosure of a gay identity will evoke negative reactions from others.

The models of gay identity development differ after the second stage, but usually posit third and fourth stages that involve identity tolerance, acceptance, commitment and integration.  Disclosure of sexual orientation to friends, family and coworkers is an integral part of these later stages of development.  However, in practice there are considerable individual differences in the degree and target of disclosure; some individuals are open about their sexual orientation to everyone, while others selectively disclose based on the trust developed in individual relationships (Cain, 1991; Croteau, 1996).

Finally, some models propose discrete and linear stages, whereas others postulate a more cyclical nature of development involving backtracking to earlier stages involving fear, defensiveness and even denial of same-sex feelings  (Rivers, 1997; Sophie, 1986).  There are also individual differences in the timing of these stages; some gay men and lesbians begin their identity process early in adolescence, whereas others may not come to terms with their identity until later in life; in some cases individuals may enter the first stage of gay identity development after they have raised a family in a heterosexual relationship (Garnets & Kimmel, 1991).

These developmental models offer some important insights for understanding the process of gay identity development in the workplace.  First, the establishment of a gay identity is an ongoing process that spills over into the workplace.  This process may not be smooth or linear, but may involve backtracking and retreats to earlier phases.  These models also suggest that employees may be at various stages in the development of their gay identity over the course of their careers; they may vary on the degree to which they are out at work, out to relatives or friends at home, and even whether they are out to themselves.  For example, gay employees in the early stages of development may deny their gay identity, even as gay and heterosexual co-workers perceive them as being gay.  Finally, an assumption underlying these developmental models is that individuals who disclose their gay identity at work are at a more advanced developmental stage, have greater identity integration, and may be more psychologically well-adjusted than those who do not disclose their gay identity at work.

Environmental models of gay identity development.  In contrast to the psychological models, environmental models propose that disclosure is due more to the environment and social relationships than the individual's psychological adjustment or level of development, and thus  challenge the notion that disclosure decisions are internally driven (Cain, 1991; Fassinger, 1995; McCarn & Fassinger, 1996).  For example, Fassinger observes that "an individual may choose to be professionally 'closeted' for important contextual reasons, but as long as the choice has been fully addressed, the individual can be as developmentally mature as one who is professionally open" (1995: 153).  Cain (1991) proposes that decisions regarding disclosure and secrecy are guided by social influences and the anticipated response of others in the social environment.  In his interview study of 38 gay men in Montreal, he found that disclosure of a gay identity was guided by the anticipation of potentially negative consequences of disclosure within a social context, and that these anticipated consequences were distinct from gay identity development (Cain, 1991).  Cain concluded that individuals who are open about their sexual orientation are not necessarily more developmentally advanced than those who conceal it; concealment in a non-supportive environment may be a necessary strategy for managing a stigmatized identity (cf. Goffman, 1974; Herek, 1988).

One important implication of this environmental perspective is that gay employees who are more open about their sexual orientation at work may not automatically be more satisfied or have less psychological strain at work than those who conceal their orientation.  Gay employees may come to full terms with their gay identity, but may choose to keep that aspect of their identity private if they believe it would hurt their careers.  As discussed below, the fear of negative consequences of disclosure may be a key factor that may affect not only the decision to disclose, but also the work-related consequences of disclosure.

Fear of disclosure in the workplace.  Although there has been little research on the fears associated with the disclosure of a gay identity in the workplace, related research indicates that such fears may be deep-rooted and prevalent (Croteau, 1996).  The learning lessons involved with disclosure may occur early in life; in a study of gay adolescents,  D'Augelli and his colleagues found that gay teenagers who disclosed to family members faced more verbal and physical abuse and were more likely to report suicidal tendencies than those who concealed their sexual orientation (D'Augelli, Hershberger & Pilkington, 1998).  Fear of rejection may be an important component in disclosure; in their study of 184 gay men and lesbians, Franke and Leary (1991) found that disclosure was predicted more by the willingness to risk the social rejection of others than by the degree to which the respondents accepted their own sexual identity.  

Interview studies suggest that fears of negative consequences of disclosure may be a key factor that affects gay employees' decisions to disclose their sexual orientation in the workplace.  Woods and Harbeck (1992), in an interview study of 12 lesbian physical education teachers, found that all of the interviewees assumed that they would lose their jobs if they disclosed their sexual orientation at work.  A study of 67 gay Harvard Business School graduates revealed that 37% of respondents reported that their careers had been hurt because people at work suspected or knew they were gay, and 12% were blackmailed or threatened with blackmail regarding their sexual orientation (Friskopp & Silverstein, 1996).  In an interview study of 70 gay men in five metropolitan cities, Woods (1994) reported that virtually all of the gay professionals in his study sought to avoid discrimination by posing as a heterosexual at some point in their careers.  Woods observes that unlike other stigmatized groups, gay employees are often assumed to be heterosexual until they reveal their stigma to others; the decision to conceal may therefore represent a "path of least resistance" for some gay employees.  Many of the interviewees in Wood's (1994) study reported that the decision to disclose was affected by the fears of negative repercussions of disclosure.  These fears ranged from social isolation and ostracism to job loss, harassment and career derailment.  Woods concluded that "… while actual discrimination is common, the fear of potential discrimination is epidemic… it is a perpetual threat, a fact of life for gay professionals who reveal their sexual orientation at work" (1994: 202).

The prediction that gay employees who fear negative consequences will be less likely to disclose at work is not only supported by interviews with gay employees, but is also grounded in Status Generalization Theory (Berger, Cohen & Zelditch, 1972; Webster & Foschi, 1988).  According to this theory, individuals hold status characteristics derived from their group memberships, such as race or gender.  These status characteristics lead to valenced expectation states, in which the individual may anticipate positive or negative outcomes associated with their status characteristic.  These expectations, in turn, lead to behavioral outcomes.  Applying this theory to gay employees, it is reasonable to expect that gay employees who expect a negative reaction following disclosure will be less likely to disclose than those who do not.  This hypothesis is straightforward and provides a foundation for our subsequent hypotheses.

Hypothesis 1: Gay employees who anticipate more negative consequences to disclosure at work will disclose to fewer people at work than gay employees who anticipate less negative consequences of disclosure.

We now turn to some of the antecedents and consequences associated with fears and disclosure of sexual orientation in the workplace.  While not an exhaustive list, an examination of these variables provides a greater understanding of workplace experiences of gay employees and builds a foundation for theory on managing stigmatized sexual identities at work.

Antecedents of Fear and Disclosure in the Workplace

Related theory and research indicates that there are a few key variables that may affect gay employees' fears of disclosure and influence their decision to disclose their sexual orientation to others in their workplace.  One key variable is the sexual orientation of the gay employee's work group.  Gay employees with gay supervisors or coworkers should be less fearful of disclosing their sexual orientation and should be more likely to disclose their orientation at work than employees who work with primarily heterosexual coworkers.  From a practical perspective, it is clear that gay employees may be less likely to fear sexual orientation discrimination and negative repercussions from disclosing their sexual orientation to gay coworkers or managers as compared to heterosexual team members.  However, the very presence of a gay supervisor or co-worker may also trigger an internal social identity process that may facilitate the employee's decision to be out at work.

According to social identity theory, individuals classify themselves into social categories that shape their personal identities and allow them to define themselves in relation to their social environment (Tajfel & Turner, 1986).  Ashforth and Mael (1989) propose that this identification process persists even if group affiliation is painful or members of the group are disliked by others in the organization, which suggests that social identity processes also occurs among stigmatized groups in organizations.  In fact, as an application of social identity theory, distinctiveness theory predicts that being a member of a relatively rare group in an organization increases group salience and identification (McGuire, 1984).  In support of this idea, Mehra, Kilduff and Brass (1998) found that the relative rarity of a group promoted group members' use of that group as a basis for shared identity and social interaction.  Although this study relied on the observable characteristics of race and gender, the presence of openly gay coworkers or supervisors may also trigger this social identity process.  Individuals who have not disclosed their sexual orientation at work, but who work with other gay employees who have,  may develop a sense of shared identity which may facilitate the development of their own gay identity at work.  It is therefore reasonable to expect that the sexual orientation of the supervisor and coworkers may affect fears of disclosure and the decision to disclose in the workplace.

Hypothesis 2a:  Gay employees with heterosexual supervisors will anticipate more negative consequences of disclosure and will be less likely to disclose their sexual orientation than gay employees who have gay supervisors.

Hypothesis 2b: Gay employees who work with a greater proportion of heterosexual coworkers will anticipate more negative consequences of disclosure and will be less likely to disclose their sexual orientation than gay employees who work with primarily gay coworkers.

As discussed earlier, emerging theory on gay identity development predicts that supportive social relationships may facilitate the self-disclosure of sexual orientation (Cain, 1991; McCarn & Fassinger, 1996).  In support of this idea, in a study of 499 lesbians, Jordan and Deluty (1998) found that social support was a key predictor of sexual orientation disclosure, and mediated the relationship between identity development and disclosure of sexual orientation.  In terms of disclosure in the workplace, some of the gay men interviewed in Woods' (1994) study reported that their decisions to disclose were not only affected by the presence of gay coworkers, but also by the presence of supportive heterosexual coworkers and managers.  It is therefore reasonable to expect that even after holding the sexual orientation of the supervisor or coworker constant, gay employees who lack supportive supervisors or coworkers will be more fearful of disclosing their sexual orientation, and will be less likely to disclose their sexual orientation than those who work in a more supportive group.

Hypothesis 3a:  Holding the sexual orientation of the supervisor constant, gay employees who lack supportive supervisors will anticipate more negative consequences of disclosure and will be less likely to disclose their sexual orientation than gay employees who have supportive supervisors.

Hypothesis 3b:  Holding the sexual orientation of coworkers constant, gay employees who lack supportive coworkers will anticipate more negative consequences of disclosure and will be less likely to disclose their sexual orientation than gay employees who have supportive coworkers.

Finally, prior experiences with discrimination in the workplace should also be included in an investigation of the disclosure decision.  Gay employees who experienced workplace discrimination in prior positions may carry those experiences with them into their next position.  For example, some of the gay men interviewed in Woods' (1994) study reported that past experiences with workplace discrimination contributed to a heightened awareness of the potential for discrimination in their current position.  In a study of 228 lesbians in New England, Schneider (1987) found that lesbians who lost a prior position due to disclosure were less likely to disclose their sexual orientation in their present position than those who were not fired because of their sexual orientation.  While losing a position due to disclosure is an extreme case of workplace discrimination, other more general forms of discrimination, such as verbal comments or  differential work treatment, may also affect both gay men and lesbians' decisions to disclose and their fears of disclosure in their next position.

Hypothesis 4: Gay employees who faced sexual orientation discrimination in prior  positions will anticipate more negative consequences of disclosure and will be less likely to disclose their sexual orientation in their current position than gay employees who did not face prior discrimination.

An interesting research question would be to compare these three antecedent variables.  Do current experiences in a supportive work environment override past experiences of a discriminatory workplace, or does prior discrimination create a deep and resilient fear that overshadows current experiences?  Does the support of the work group matter more than its sexual orientation?  On the one hand, a work group that is composed of primarily gay coworkers provides a relatively safe haven for gay employees to disclose their sexual orientation.  On the other hand, a supportive heterosexual work team may not only support disclosure, but could also send the message that other heterosexuals in the organization may also support the gay employee's decision to disclose.  Moreover, if the organization is not supportive of gay employees, a supportive heterosexual work team could provide more of a buffer to the gay employee than a primarily gay work team.  Since there is no research or theory to guide these competing perspectives, we explore this issue as a research question:

Research Question 1: Which antecedent variable has the greatest impact on the anticipation of negative consequences of disclosure and the decision to disclose sexual orientation in the workplace?

We now turn to an examination of work-related attitudes and outcomes associated with fear of disclosure and the decision to disclose in the workplace.

Outcomes Associated with Fear and Disclosure in the Workplace

A fair number of studies have examined the relationships between the degree of disclosure of sexual orientation in the workplace and work-related attitudes, outcomes and psychological strain at work.  A basic premise behind these studies is that gay employees who are open about their sexual orientation at work essentially avoid the conflict and stress involved with the concealment and management of dual-identities.  Accordingly, these studies predicted that employees who disclosed more at work would also be more satisfied with their jobs, experience less psychological strain, and have more positive work-related outcomes than those who concealed their sexual orientation from others at work. 

Although these predictions make intuitive sense, the results of these studies yielded inconsistent and even contradictory findings.  Some researchers found that gay employees who disclosed their sexual orientation to more people at work had more positive job attitudes (Day & Schoenrade, 1997; Ellis & Riggle, 1995), while other studies found no relationship between the degree of being out at work and work attitudes (Croteau & Lark, 1995; Driscoll, Kelley & Fassinger, 1996).  In fact, gay employees who disclosed to more people at work had lower continuance organizational commitment (Day & Schoenrade, 1997), earned less compensation (Ellis & Riggle, 1995; Schneider, 1987), and had less pay satisfaction (Ellis & Riggle, 1995) than those who concealed their sexual orientation from others in the workplace.  The decision to disclose should be strongly related to work-related stress and psychological strain, but research on these variables has also produced inconsistent findings; some studies found a significant relationship between disclosure and reduced general anxiety (Jordan & Deluty, 1998), but other studies found no relationship between disclosure and work-related stress (Day & Schoenrade, 1997), psychological strain or occupational coping (Driscoll, Kelley & Fassinger, 1996).

As discussed earlier, the degree of disclosure of sexual orientation, while important, may have less of an impact than the immediate fears that precede this important decision.  Environmental models of gay identity development hold that gay employees who are more open about their sexual orientation at work may not automatically be better adjusted, more satisfied at work, or have less psychological strain at work than those who conceal their sexual orientation (Cain, 1991; Fassinger, 1995; McCarn & Fassinger, 1996).  Individuals may choose to conceal their sexual orientation at work if they believe that disclosure would result in negative repercussions involving social isolation, discriminatory job treatment, or even job loss.  It is therefore important not to treat the decision to disclose in a vacuum, but to examine simultaneously the relative effects of disclosure and fears of disclosure on work-related attitudes, psychological strain and career outcomes.  In fact, as noted earlier, the fears regarding disclosure may have an even stronger impact on outcome variables than the actual disclosure decision.

In support of this idea, Meyer (1995) theorized that minority groups experience psychological distress that is related to their group stigmatization, and that a key process of minority stress for gays is the fear of rejection and discrimination.  Meyer (1995) found support for his model of gay stress in a study of 741 gay men living in New York City; expectations of rejection and discrimination were not only significantly related to the mental health and psychological stress of his respondents, but these effects were independent of actual discriminatory events.  Along similar lines, in a study of 533 Belgian gay men, Vincke, DeRycke and Bolton (1999) found that commitment towards a gay identity did not buffer the psychological strain that accompanies being gay.  Vincke and his associates observed that openly gay males experienced "daily hassles and conflicts" that resulted in gay stress.  Although these studies were not conducted in the workplace, they do suggest that irrespective of the actual decision to disclose, the very fear of negative consequences of disclosure may significantly influence work-related attitudes, psychological strain at work and career outcomes.

Hypothesis 5: Gay employees who anticipate more negative consequences of disclosure will report greater psychological strain at work, and less positive work and career attitudes, work environments, and career outcomes than those who anticipate less negative consequences of disclosure.

In order to provide a comprehensive assessment of the disclosure process, we also need to examine the direct relationship between disclosure and the outcome variables.  As discussed earlier, prior research examining the relationship between disclosure and various outcome variables has yielded inconsistent and contradictory results.  We observed that one explanation for these inconsistent findings is that these studies did not capture the full picture, and that we need to examine the perceived consequences of disclosure as well as the actual decision to disclose.  However, another explanation for these inconsistent findings is that most of these studies employed small, regional samples composed of either gay men or lesbians, but not both.  In order to provide clarity on this issue, we would like to replicate and extend prior work that examined the relationship between disclosure at work and work-related attitudes by examining a full array of 15 outcome variables in one large-scale investigation using a national random sample of both gay men and lesbian employees.

Hypothesis 6:  Gay employees who disclose to more individuals in the workplace will report less psychological strain at work, and have more positive work and career attitudes, work environments, and career outcomes than those who disclose to fewer individuals in the workplace.

Finally, a central question to the present investigation is whether the fear of disclosure accounts for more of the variance in these outcome variables than the actual decision to disclose.  Since there is a lack of research or theory to guide us, we pose this as a research question.

Research Question  2:  Does fear of disclosure of sexual orientation in the workplace account for more of the variance in these outcome variables than the degree of disclosure?

METHOD

Procedure and Respondents

Sampling procedure.  As part of a larger national study on heterosexism in the workplace (reference withheld to preserve anonymity of review process), surveys were sent to a national random sample of 2,919 members of three national gay rights organizations in the United States.  Specifically, we sent surveys to 1,488 members of one of the largest gay civil rights organization in the nation.  In addition, we sent an additional 681 surveys to members of a national gay Latino/a organization, and 750 surveys were sent to a national gay African-American organization.  A stratified random sampling technique was used whereby equal numbers of men and women were selected by geographic area.  The surveys were completely anonymous; there was no identifying information on the surveys that would connect them to the respondents.  Two reminder postcards and a reminder letter were sent to all respondents.  A total of 334 surveys were returned unanswered for various reasons, the primary reason being undeliverable mail (283);  51 surveys were returned unanswered because respondents were retired, unemployed, self-employed, heterosexual or deceased.  Completed surveys were returned by 768 respondents, yielding a response rate of 30%.

Respondents.  Because this study investigated gay employees' fears and decisions to disclose their sexual orientation in the workplace, surveys from the following groups were not used in the analyses: those indicating they were heterosexual (20) or unsure of their sexual orientation (3), those who were self-employed (99) or employed by a gay, lesbian, or bisexual organization (51), and those who were not employed in paid positions (i.e., retired, unemployed or volunteers) (61). 

The final sample therefore consisted of 534 respondents.  The sample consisted of 168 women and 363 men; 3 did not report their gender.  The majority of the respondents considered themselves to be gay or lesbian (92.9%), as compared to bisexual (7.1%).  The racial and ethnic background of the respondents was 67.6% white (n=361), 15.2% black (n=81), 12.2% latino/hispanic (n=65), .7% Asian (n=4), 1.1% multiracial (n=6) and 1.1% other (n=6); 11 (2.1%) did not report their race.  The average age of the respondents was 41 years.  The respondents were highly educated, the majority held bachelor's degrees (38.6%), and many had master's (28.2%) and doctoral (17.9%) degrees.  Forty-one percent of the sample earned between $26,000 and $50,000 a year, and 24% earned between $51,000 and $75,000 a year.  The average tenure in their current organization was 9.3 years, and the average tenure in their current position was 6.1 years.  The majority of respondents held professional or technical jobs (68.5%) and managerial jobs (19.7%), the remainder of the sample was employed in clerical or sales positions (4.9%), service or craft (6.4%) or agricultural (.4%).  Respondents also came from a large range of industries, such as education (24.2%), health (17%), government (14.8%), service (12.2%), manufacturing (9.2%), finance/insurance (6.8%), arts/entertainment (4.8%), advertising/publishing (3.1%), travel (2.0%), human services (2.0%) and design/fashion (.9%).  Respondents were also employed in organizations that varied in size: 30.3% were employed in companies with over 10,000 employees, 26.4% were at companies with 1,000 - 10,000 employees, 20.7% were at companies with 100-999 employees, and 21.5% were employed at companies with less than 100 employees.

Measures

The survey was developed and pretested on a separate pilot group of 28 gay employees across the nation.  The pretest was used to ensure clarity, refine instruments and select items.  Staff from the gay rights organizations that provided the mailing lists also critiqued and approved the pretest and final survey.

Degree of disclosure of sexual orientation in the workplace.  Disclosure was measured with the following question: "At work, have you disclosed your sexual orientation to: (Please check one option): (1) no one  (2) some people (3) most people (4) everyone."  These four options were modified from similar "out at work" scales used by Croteau and Lark (1995), Levine and Leonard (1984) and Schneider (1987).  Higher values represent greater disclosure of sexual orientation at work.

Fear of disclosure of sexual orientation in the workplace.  A review of the literature revealed no measures of this variable, so we developed a 12-item scale based on existing research and qualitative reports of anticipated consequences of disclosure in the workplace (Friskopp & Silverstein, 1996; Schneider, 1987; Woods, 1984). The items are listed in Appendix A.  The instrument uses a 7-point likert scale, with responses ranging from completely disagree (1) to completely agree (7).  Higher values therefore represent greater fears of disclosure.  A principal components factor analysis yielded a single factor with an eigenvalue of 7.78 accounting for 64.9% of the variance. The coefficient alpha for the 12-item scale was .95.

Sexual orientation of work group.  Two items were used to measure the sexual orientation composition of the work group.  First, respondents were asked about the sexual orientation of their coworkers. The options for this item were: most coworkers were heterosexual (coded as 1), workgroup about equally balanced (coded as 2), most coworkers were gay or lesbian (coded as 3), and did not know (coded as missing).  Next, respondents were asked whether their supervisor was the same sexual orientation as them (heterosexual supervisors were coded as 0, gay supervisors coded as 1, unsure of sexual orientation coded as missing).  Higher scores on these items therefore represent a greater proportion of gay coworkers and supervisors.

Social support from coworkers and supervisors.  The Caplan, Cobb, French, Harrison and Pinneau (1975) measure of social support at work was used in the study.  The scale asks respondents to use a 4 point scale ranging from 1 (not at all) to 4 (very much) to rate the support they receive from work relationships.  Example questions include:  "How easy is it to talk with each of the following people?  How much can each of these people be relied on when things get tough at work?"  Each of the 4 questions asks respondents to first rate support received from their supervisor and then from other people at work.  The scale therefore involves 4 items for supervisory support and 4 items for coworker support, with higher values indicating greater social support at work.

Prior experiences with workplace discrimination.  Seven items were used to assess prior workplace discrimination.  The items and responses are displayed in Appendix B.  For example, respondents were asked to indicate whether they were verbally or physically harassed in prior positions because of their sexual orientation, whether they were terminated from a job because of sexual orientation discrimination, and whether they left their last job because of such discrimination. The options for these items were: yes (coded 2), unsure (1) and no (0).  The items were summed to create an overall scale of prior discrimination.  Higher values therefore represented greater reports of past workplace discrimination.

Work and career attitudes.  Established instruments were used to measure all career and job attitudes.  Six attitudes were assessed: job satisfaction, satisfaction with opportunities for promotion, organizational commitment, organization based self-esteem, career commitment and turnover intentions.  Job satisfaction was measured with the Quinn and Staines (1979) five-item measure of job satisfaction.  Satisfaction with opportunities for promotion was appraised with the nine-item promotion subscale of the Job Description Index (Smith, Kendall & Hulin, 1969).  Organizational commitment was measured with Mowday, Steers and Porter’s (1979) fifteen-item scale.  Organization based self-esteem was measured using the ten-item Pierce, Gardner, Cummings and Dunham (1989) scale.  Career commitment was measured with Blau’s (1985) seven-item Career Commitment scale.  Turnover intentions were measured using a two-item scale from the Michigan Organizational Assessment Questionnaire (Nadler, Jenkins, Cammann & Lawler, 1975).  Higher values on all attitudinal scales represented more positive career and job attitudes and less turnover intentions.

Psychological strain at work.  Four established instruments were used to measure psychological strain at work  (cf. Begley & Czajka, 1993; Burke, Brief & George, 1993).  The first psychological strain at work measure used was Caplan et al.'s (1975) 10-item somatic complaints at work scale.  This scale asks respondents whether they experienced specific physical stress-related symptoms on the job during the past month.  Respondents used a 3-point scale ranging from never (1), to once or twice (2), to three or more times (3), to report the occurrence of such stress-related symptoms as dizziness spells, shortness of breath, insomnia, and upset stomach.  In addition to the somatic complaints scale, we used three other established scales to measure psychological strain at work.  Respondents were asked to describe how they feel when they think about themselves and their jobs, and use a 4-point scale that ranged from "never or little of the time" (1) to "most of the time" (4) in their ratings.  The three scales included:  Caplan et al.'s (1975)  6-item work-related depression scale (example item: "I feel sad"), Caplan et al.'s (1975) 4-item work-related anxiety scale ("I feel nervous"), and Cobb's (1970) 3-item work-related irritation scale ("I get irritated or annoyed").  Higher values on these four scales represent greater reports of psychological strain at work.

Work environment.  Three established measures were used to assess the general work environment.  We used Caplan et al.'s (1975) 3-item role conflict scale (example item: "Persons equal in rank and authority over you ask you to do things which conflict."), their 4-item role ambiguity scale (example item: "How often are you clear about what others expect of you on the job?"), and their 3-item workplace participation scale (example item: "How much do you take part with others in making decisions that affect you?").  These likert scales were coded so that higher values represent more positive work environments. 

Career outcomes. Promotion rate and compensation were used to measure career outcomes.  Promotions were defined as involving two or more of the following criteria: significant increases in salary, significant increases in scope of responsibility, changes in job level or rank, or becoming eligible for bonuses, incentives and stock plans.  Given the above definition, respondents were asked how many promotions they received over the last ten years.  Respondents were also asked to check the income category that reflected their current annual compensation, which included salary, bonuses, commissions, stock options and profit sharing.

Control variables.  We started with a large list of potential organizational and demographic control variables that may be related to disclosure at work and have been employed as control variables in other studies.  We considered organizational size, industry, respondent's job, rank in organization, and tenure in the organization and position as potential organizational covariates.  For demographic covariates, we considered variables that may be related to gay identity formation, such as age, gender, race and education (Gonsiorek & Rudolph, 1991).  In order to preserve power, we selected covariates that had a significant relationship with fear of disclosure, disclosure decisions, and the outcome variables, but had low intercorrelations.  We used independent variable interaction terms to test for homogeneity of the regression assumptions fundamental to covariance analyses.  On the basis of these criteria, existing theory, and prior research, we used three control variables in the present study: respondent's age, education, and organizational size.

RESULTS

The correlations, means, standard deviations and coefficient alphas for the study variables are displayed in Table 1.  The coefficient alphas for the study variables were all acceptable, and ranged from .70 to .95.

An initial review of the independent variables reveals some interesting insights into the workplace experiences of gay employees.  Specifically, as displayed in Appendix B, 37% of our respondents encountered discrimination in prior positions because others suspected or assumed that they were gay, and 11% left their last position because of sexual orientation discrimination.  A full 22% of our sample reported being verbally harassed and 10% were physically harassed because of their sexual orientation in prior positions.  In terms of sexual orientation of work group, 89.3% of  respondents (n = 474) reported that most of their coworkers were heterosexual, 6.6% (n=35) reported that their work group was about equally balanced,  2.3% (n=12) reported that most of their coworkers were gay or lesbian,  and 1.9% (n=10) did not know.   The majority of respondents had heterosexual supervisors (85.8%, n=452); 8.9% (n=47) reported having gay supervisors, and 5.3% (n=28) reported not knowing their supervisor's sexual orientation.  Our sample varied on the degree to which they disclosed their sexual orientation at work: 11.7% reported that they were out to no one at work, 37% reported being out to some people, 24.6% reported being out to most people, and only 26.7% reported being out to everyone at work. 

Antecedents of Fear and Disclosure in the Workplace

Hierarchical regression analyses and bivariate correlations were used to test Hypotheses 1-4 and Research Question 1.  In support of Hypothesis 1, the significant negative correlation     (r =-.45, p<.0001) between respondents' fears of disclosure and disclosure at work indicated that gay employees who anticipated more negative consequences to being out at work disclosed to fewer people than employees who anticipated less negative consequences to disclosure.

Tests of the impact of the antecedent variables on the dependent variables of fear of disclosure and the decision to disclose were made using a series of hierarchical regression analyses in which respondent's age, education, and organizational size were entered as control variables in the first step of the hierarchical analyses.

As displayed in Table 2, Hypotheses 2a and 2b were supported; employees with work groups composed of primarily gay coworkers reported less fear of disclosure and were more likely to disclose than employees with heterosexual coworkers.  The presence of a gay supervisor was significantly related to fear of disclosure, and marginally related to the disclosure decision (p=.059).  When holding the sexual orientation of the supervisor constant, supervisor support was significantly related to fear and the decision to disclose.  Similarly, when controlling for the sexual orientation of the work team, employees with supportive coworkers reported less fear and were more likely to disclose than those with less supportive coworkers. Hypotheses 3a and 3b were therefore both supported.

We found mixed support for Hypothesis 4.  As predicted, employees who had experienced sexual orientation discrimination in prior positions reported greater fears of disclosing their sexual orientation in their current position than employees who did not experience prior discrimination.  In contrast to our predictions, those experiencing prior discrimination were more likely to be out in their current positions than those who did not experience prior discrimination.

Our first Research Question asked which of these various antecedent variables would have the strongest impact on fears of disclosure and the decision to disclose.  To answer this question, all of the antecedent variables were entered together in a second step of the hierarchical regression analysis, following the control variables.  As displayed in Table 2, different antecedent variables predicted fear of disclosure as compared to the actual decision to disclose.  The most significant predictor of fear of disclosure was the presence of a supportive supervisor, followed by the presence of supportive coworkers and prior experiences with discrimination.  In contrast, the key predictor of actual disclosure was the presence of gay coworkers, followed by supportive coworkers and prior discriminatory experiences.

Outcomes Associated with Fear and Disclosure in the Workplace

We first examined the intercorrelations among our outcome variables: the six work and career attitude scales, the four psychological strain scales, the three work environment scales, and the two career outcomes of compensation and promotion.  Since we found moderately high intercorrelations among scales reflecting work and career attitudes, psychological strain, and work environment, we used multivariate analysis of covariance (MANCOVA) in analyses involving these variables.  The career outcome variables of compensation and promotion were not highly intercorrelated, so we used analysis of covariance (ANCOVA) to assess relationships involving these dependent variables.  Since analysis of variance requires categorical independent variables, we used a mean split to transform the fear of disclosure variable from a continuous to a categorical variable.  The adjusted means for the dependent variables are displayed in Table 3.

As displayed in Table 4, we found support for Hypothesis 5.  Fear of disclosure was significantly related to 13 of the 15 outcome variables in this study.  Gay employees who anticipated more negative consequences of disclosure reported greater psychological strain than employees who anticipated less negative consequences (Wilk's Lambda= .926, F (4,403)=8.00, p<.001).  Follow-up univariate analyses revealed that fear of disclosure was significantly related to work-related depression (p<.001), work-related irritation (p<.001) and physical somatic stress-related symptoms experienced on the job (p<.05), but only approached significance for work-related anxiety (p=.07).  Fear of disclosure was also significantly related to all of the work and career attitudes examined (Wilk's Lambda= .853, F (6, 400)=11.486, p <.001).  Gay employees who anticipated more negative consequences to disclosure also reported greater turnover intentions and less job satisfaction, organizational commitment, satisfaction with opportunities for promotion, organization-based self-esteem, and career commitment than employees who anticipated fewer negative consequences.  Similarly, fear of disclosure was significantly related to all of the work environment variables studied (Wilk's Lambda= .928, F (3,403)=10.39, p <.001).  Employees who anticipated more negative consequences from disclosing reported more role conflict (p<.05), more role ambiguity (p<.001), and less workplace participation (p<.001) than those who anticipated fewer negative consequences of disclosure.  Finally, gay employees who anticipated more negative consequences to disclosure received  significantly fewer promotions (F (1, 375) = 10.11, p=.002) and earned marginally less compensation (F (1, 407) = 2.898, p=.089) than those who anticipated less negative consequences to disclosure. 

We found no support for Hypothesis 6.  The degree of disclosure of sexual orientation in the workplace was not significantly related to 15 of the 16 dependent variables in the study.  The only exception was that employees who disclosed their gay identity at work reported more participation in their work environment than those who did not disclose.  Given these results, the answer to Research Question 2 is relatively straightforward; fear clearly had a greater impact on outcomes than disclosure.

DISCUSSION

While diversity scholars seek to understand the unique workplace experiences of various social identity groups in organizations, little attention has been paid to the experiences of gay employees. Our study addressed this gap in the literature by examining one of the most important career decisions faced by this under-studied population: the decision to come out at work.  Using a large and diverse national sample of gay and lesbian employees, we examined the antecedents and consequences of disclosure in the workplace.  We introduced the construct of fear of disclosure, and found that this fear had an overwhelmingly negative impact on the work experiences of gay employees: their job and career attitudes, their psychological strain at work, their work environment, and even their advancement.  In fact, we found that the fears associated with the decision to disclose had a greater impact than the actual decision itself.  Although fear of disclosure predicted the decision to disclose, these constructs were empirically unique and independent in their relationships with the antecedents and outcomes investigated in this study.

What predicts fear of disclosure and the decision to disclose a gay identity at work?  Our study revealed that the work group is a critical component in the disclosure process.  To start, employees with gay supervisors and coworkers reported less fear of disclosure and were more likely to disclose at work than those in a primarily heterosexual work group.  Future research can use qualitative methods or a longitudinal research design to establish whether gay employees gravitate towards gay work groups, or whether the work group itself triggers a social identity process that results in disclosure in the workplace.

The support of the work group may be even more important than its sexual orientation.  We found that when holding the sexual orientation of the group constant, gay employees with supportive coworkers and supervisors reported less fear and were more likely to be out at work than those lacking a supportive group.  This suggests that the presence of supportive heterosexual coworkers may help alleviate fears of disclosure and allow gay employees to bring their social identity to work.  We need more research that examines the specific types of work group behaviors and interactions that support gay coworkers.  From an applied perspective, organizations can help develop supportive work teams by providing training that dispels stereotypes and sensitizes heterosexual employees to the issues faced by gay colleagues.

One of the more intriguing findings of this study was that gay employees who faced sexual orientation discrimination in the past reported greater fears of disclosing their sexual orientation in their current workplace, but were more likely to disclose their sexual orientation at work than those who did not experience discrimination in prior positions.  This suggests that while prior experiences with discrimination may heighten perceptions of the risks associated with disclosure, these experiences  may not impede the development of a gay identity in the workplace, and may in fact serve to pull the individual "out of the closet."  This apparent resilience in the face of prior discrimination also suggests that the decision to disclose is not totally a function of fear, but may be a decision that is balanced by anticipated consequences on the one hand and the need to develop an authentic sense of self in the workplace on the other.  This is an area that clearly calls for more research.

The distinction between fear of disclosure and the disclosure decision became evident when examining the variables that predicted these two constructs.  While the presence of a supportive work team was the key factor that predicted fear of disclosure, coworkers' sexual orientation was the critical predictor of the actual decision to disclose.  These findings indicate that fears associated with disclosure are most allayed when working with supportive coworkers and managers, irrespective of their sexual orientation.  The sexual orientation of coworkers becomes critical, however, in the actual decision to overcome the fears and disclose.

This information can provide a foundation for theory development on the management of a gay identity in the workplace.  In particular, these findings not only point to the importance of examining the fear construct as independent from the disclosure decision, but also suggest that theory on gay identity formation needs to take environmental factors into account.  The decision to disclose may be guided not only by internal development, but also by the anticipated consequences of disclosure, the support of the work environment, and prior experiences with discrimination.  In essence, these findings support the environmental models of gay identity presented earlier (Cain, 1991; McCarn & Fassinger, 1996), and suggest that the decision to disclose must be viewed within the social and organizational context.  The psychological health and well-being of gay employees may not be determined by the disclosure decision, as much as the organizational context in which that decision takes place.

Finally, this study took an important first step towards documenting the organizational and individual costs associated with gay employees' fears of coming out in the workplace.  While the decision to disclose was generally unrelated to the outcome variables, the fears associated with disclosure had an overwhelming negative impact on nearly all of the variables studied.  It was particularly striking that respondents who reported greater fears of disclosure also reported experiencing a host of more physical stress-related