Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

1 Introduction

As of June 2012, 21 states, Washington, D.C., and many localities prohibit discrimination in employment on the basis of sexual orientation.Footnote 1 In the majority of U.S. jurisdictions, however, employment decisions based on sexual orientation are not subject to state or local legal restrictions. And, although previously proposed, no federal law exists that prohibits discrimination on the basis of sexual orientation for hiring, promotion, compensation, or termination decisions. Sexual minorities are consequently legally vulnerable in most of the United States.

Whether lesbian, gay, or bisexual (LGB) individuals are actually in need of legal protection, however, is often disputed by policymakers and the public at large. In particular, the image of the professional and highly compensated gay man pervades the media, and is the subject of targeted marketing for luxury goods and services. Magazine surveys or other surveys of convenience have provided fodder for this perception, reporting above-average incomes for gay men and lesbians (Badgett 2001; Black et al. 2000). Population-based data, however, generate a different picture of the economic situation of LGB individuals. Analyses of these data suggest that sexual orientation plays an important—but not always positive—role in several economic outcomes, including income and occupational segregation. Further, survey-based data provide estimates of the prevalence of disclosure of sexual orientation in the workplace, a variable which undoubtedly has a strong effect on income and occupational outcomes.

In this chapter, I present data from nationally representative samples to examine several labor market outcomes for LGB persons in the United States, with a focus on evaluating the evidence of inequality. Inequality in the workplace could be suggested by differences in income or occupations between heterosexual and non-heterosexual persons. Further, given that disclosure of sexual orientation in the workplace provides the opportunity for direct discrimination, data on workplace disclosure are also presented in this chapter. I conclude by presenting data on the prevalence of reported sexual orientation discrimination in the workplace.

2 Sexual Orientation and Income

According to descriptive data from the 2000 U.S. Census, employment differences are present between heterosexual and non-heterosexual individuals. Table 13.1 presents data of partnered individuals (married heterosexuals, cohabiting heterosexuals, and cohabiting gay men/lesbians) who reported being employed in 1999 and had earnings of $1,000 or more from employment. Of these individuals, married heterosexual men reported the highest mean income, followed by cohabiting gay men, cohabiting lesbians, cohabiting heterosexual men, married heterosexual women, and cohabiting heterosexual women. Similarly, the data indicate that married heterosexual men report working the most weeks in a year and the most hours in a week, with gay men following closely on both indicators. Lesbians reported working more weeks in a year and more hours per week than either married or cohabiting heterosexual women. This could suggest that lesbians must work more, on average, in order to make up for the lack of a (typically higher) male income in the household.

Table 13.1 Mean values of employment-related variables for partnered individuals who reported employment in 1999, 2000 U.S. census
Full size table

The descriptive data, therefore, suggest both that employment differences exist between heterosexual and gay individuals, and that gay men and lesbians experience an income advantage over all partner types save married heterosexual men. Many studies have been conducted to determine whether such income differences persist when other employment-related explanatory variables are introduced.

When evaluating the effect of sexual orientation on earnings, policymakers and researchers have frequently used data gathered in surveys of convenience, such as those obtained from readers of magazines and newspapers (see Badgett 2001; Black et al. 2000). Due to the biases in these data, no generalizable conclusions may be reached about the effect of sexual orientation on earnings. Beginning in the mid-1990s, however, social scientists began to use nationally representative data to quantify the earnings differences between gay and heterosexual individuals. They relied principally on two data sources: the General Social Survey (GSS) and the 1990 U.S. Census.

The GSS is a representative sample of the U.S. population and, presumably, of the gay male and lesbian populations, although the numbers of gay men and lesbians in the GSS are relatively small. The GSS relies on a behavioral definition of sexual orientation which may be problematic in earnings studies since those who self-identify on a survey, rather than those who are identified on the basis of behavior, may be more likely to disclose their sexual orientation in the workplace (Badgett 2001). In turn, individuals who reveal their identity in the workplace may be more subject to discrimination (see section on disclosure, below).

The 1990, 2000, and 2010 U.S. censuses and the American Community Surveys (ACS) provide another source for examining earnings differentials based on sexual orientation. Unlike the GSS, the censuses and ACS provide large samples that measure sexual orientation via identity, i.e., membership in a same-sex partnership. But the census data are limited in that only persons who choose to identify themselves as same-sex unmarried partners, and who are residing in the same households as their partners, are enumerated. Single gay men and lesbians are, therefore, uncounted.

Several researchers have used the GSS or U.S. census data to examine earnings differences between gay and heterosexual individuals, and have obtained varying results (Table 13.2). Although the magnitude of the effect of sexual orientation differs from study to study, the results suggest overall that gay men experience an earnings penalty in employment while lesbians experience an earnings advantage.

Table 13.2 Summary of results of prior research regarding the economic cost of being homosexual
Full size table

According to past findings, gay men experience an earnings penalty that ranges from a statistically insignificant 0.15–26%, depending on the heterosexual comparison group (married men, unmarried men, or both), the dataset, and the model employed (see Table 13.2; Baumle and Poston 2011; Baumle et al. 2009; Black et al. 2003; Blandford 2003; Berg and Lien 2002; Allegretto and Arthur 2001; Klawitter and Flatt 1998; Badgett 1995). An earnings penalty of some magnitude, therefore, has typically been found for gay men.

The penalty is, however, greater for gay men in comparison to married heterosexual men. For example, in Baumle and Poston’s (2011) multilevel analysis using 2000 U.S. Census data, the results indicated that gay men experience a far greater earnings penalty compared to married heterosexual men (12.5%) than when compared to cohabiting heterosexual men (an insignificant 0.15%). This suggests that a large portion of the earnings difference between partnered gay and heterosexual men could well be attributable to marital status. Prior research has demonstrated that marriage results in an earnings benefit for men (Waite and Gallagher 2000). Although cohabiting heterosexuals are also disadvantaged by their unmarried status, these individuals have the option of entering into a legal marital union and gaining the benefits associated with marriage. This option is not available to most gay men and, as a result, could well be contributing to the earnings differential between partnered gay men and married heterosexual men.

Some of the earnings benefits derived from marriage for men, however, may be attributable to traditional gender roles, such as men benefitting from women’s care of their homes and children. Whether these benefits would play out in the same manner for same-sex couples is uncertain. Other studies have suggested that employers tend to engage in discrimination in favor of married men, believing married men to be more dedicated workers and more deserving of pay raises (Waite and Gallagher 2000). If employer discrimination plays a role in the marriage premium, then gay men should experience an earnings penalty regardless of whether their own marital relationships adhere to traditional gender roles.

The findings regarding sexual orientation and earnings are less consistent for women than are those for men (Table 13.2). Research suggests that lesbians’ earnings are either not statistically different from those of heterosexual women (Klawitter 1998; Klawitter and Flatt 1998; Badgett 1995), or that lesbians have an earnings advantage that ranges from 2.1 to 30% (Baumle and Poston 2011; Baumle et al. 2009; Berg and Lien 2002; Black et al. 2003; Clain and Leppel 2001). For women, therefore, the effect of sexual orientation on earnings seems to be unclear. Regardless, sexual orientation does not appear to have a detrimental effect on the earnings of lesbians, as it does for gay men.

2.1 Income, Sex, and Sexual Orientation

Several explanations have been offered for the different effects of sexual orientation on earnings for men and women. Badgett (1995) hypothesized that gay men might face greater discrimination than lesbians because of associations with HIV and AIDS (see also Berg and Lien 2002). Also, lesbians may be more readily accepted into male-dominated professions as “one of the guys,” permitting them to excel in areas where heterosexual women are barred (Berg and Lien 2002; Badgett 1995). Some data suggest that heterosexual men evince more hostility toward gay men than towards lesbians, indicating perhaps more of an acceptance of female than male homosexuality (Kite and Whitley 1996; Herek 1991).

Further, some have theorized that the lesbian earnings advantage could be partially explained by work and family differences (Baumle 2009; Berg and Lien 2002; Badgett 2001). If lesbians are more hesitant to interrupt their careers to have or to raise children, or if employers perceive them to be so, then their earnings should be higher than those of heterosexuals. Potential parenting differences, therefore, have tended to dominate the possible explanations of the lesbian wage advantage.

There are many reasons to believe that parenthood could play an important role in explaining the wage difference between lesbians and heterosexual women. Recent research examining the effect of motherhood on employment outcomes has found that parental status is an important predictor of women’s earnings. Findings suggest that there is approximately a 3–8% wage gap between mothers and childless women, after controlling for other relevant characteristics (Anderson et al. 2003; Budig and England 2001; Crittenden 2001). In fact, Budig and England (2001) found that the majority of the gender gap in wages can be attributed to lower earnings by employed mothers. Further, Peplau and Fingerhut (2004) conducted a study where subjects rated job applicants on measures of warmth and competency. Their findings show that parents received higher ratings on measures of warmth, regardless of sex or sexual orientation. In terms of competency, however, motherhood resulted in a lower rating for heterosexual women, but did not affect competency ratings for lesbians.

Using 2000 U.S. Census data, I examined the effect of parenthood on income for lesbian and heterosexual women (Baumle 2009). Employing Ordinary Least Squares regression analysis to estimate the effect of having a child present in the household on income, I found that the motherhood penalty is experienced primarily, if not solely, by heterosexual women. In fact, lesbians appear to experience a motherhood advantage that increases their wages by approximately 20%. Further, results from a Blinder-Oaxaca analysis support the notion that lesbians receive different returns to the presence of children in the household than do heterosexual women. Approximately 35% of the wage differential between lesbians and heterosexual women is attributable to differences in returns to childrearing. This indicates that some of the lesbian wage advantage is attributable to heterosexual mothers being treated economically differently than both lesbians and childless heterosexual women. Concomitantly, lesbian mothers receive treatment that differs from heterosexual women when they avoid some, or all, of the wage penalty associated with motherhood.

The results do not, however, provide an explanation for this pay differential. Prior research has indicated that lesbians are more likely to be in the labor force and to have dual-wage earners in the household (see e.g. Baumle et al. 2009). This suggests that there might be some truth to the notion that lesbians are dependent on having both partners employed and, consequently, could be less likely to exit the labor force to raise children. If this is the case, then the fact that they do not experience a motherhood penalty would be understandable, as they would be more likely to have a stable employment history. Nonetheless, past research would suggest that employer stereotypes do play a strong role in the employment outcomes of lesbians, with lesbian mothers not experiencing a decrease in competency ratings in controlled experiments as do heterosexual mothers (Peplau and Fingerhut 2004). This suggests that assumptions are made even at the time of hiring regarding the effect that motherhood will have on the careers of women—and these assumptions differ based on sexual orientation.

2.2 Contextual Factors Affecting Income Differences

In addition to sex and the presence of children in a household, other individual and contextual factors affect the earnings difference between gay and heterosexual individuals. Baumle and Poston (2011) employed a multilevel analysis to examine the role of both individual- and state-level factors in the effect of sexual orientation on income. Prior studies had considered primarily individual-level factors in predicting earnings. Baumle and Poston found that, while most variation occurs at the individual level, nonetheless state-level factors had an effect on earnings outcomes. Findings regarding two state-level factors were of particular note: antidiscrimination laws that prohibited discrimination based on sexual orientation, and the presence of other individuals in same-sex partnerships. These two variables tend to influence the relationship between sexual orientation and earnings (i.e. display cross-level effects).

Baumle and Poston found that the presence of a state-level antidiscrimination law decreases the earnings gap between gay men and married heterosexual men by approximately 2.5%. Research using the 1990 data indicated that antidiscrimination laws had no statistically significant effect on earnings (Klawitter and Flatt 1998), but it is possible that these laws, many of which were passed during the 1990s, have now had time to exert a positive influence on the earnings of gay men.

In addition to antidiscrimination laws, Baumle and Poston examined the manner in which residence in a state with a higher concentration of same-sex partners interacts with the effect of sexual orientation on earnings. Living in an area with a high concentration of same-sex partners could well result in the following: provide additional pressure on employers and politicians to provide domestic partner benefits and other employment perks; indicate a more liberal climate and greater acceptance (Baumle and Compton 2012); provide a needed support network which could improve mental health and work performance; and provide social contacts and business opportunities (Collins 2004), thereby increasing earnings. Findings indicate that a higher concentration of same-sex partners in a state consistently resulted in a decline in the negative effect of orientation on earnings for gay men. For lesbians compared to heterosexual women, a higher concentration of same-sex partners in the state resulted in a decrease in the positive effect of orientation on earnings. Baumle and Poston hypothesize that this sex difference could be attributable to differences in the types of enclaves that develop for men versus women, with men tending to reside in large, higher-income urban areas and women in low-income rural areas.

2.3 Income: Summary

Overall, findings suggest that gay men experience a wage penalty (particularly in comparison with married heterosexual men), and lesbians experience a wage advantage compared to heterosexual women. These findings of an earnings differential between gay and heterosexual individuals could be due to discrimination (both against gay men, and, perhaps, in favor of lesbians). The findings could also suggest differences in occupation (Baumle et al. 2009; Baumle 2004). If gay men, for instance, tend to work in occupations that pay less than those of heterosexual men, their earnings disparities could be based on these occupational differences. Income analyses often include some measure of occupational difference, but these variables have typically been broad categories or occupational status indicators. In the following section, I consider more detailed data on occupational segregation based on sexual orientation.

3 Occupational Segregation

Limited data have been available to examine occupational segregation based on sexual orientation, given that the small sample sizes of LGB persons on most surveys prevented a detailed analysis of occupational differences. With the availability of large sample sizes of same-sex partners on the census and ACS, however, researchers can now take a closer look at the role of sexual orientation in occupational segregation. In this chapter, findings from occupational analyses using a variety of representative data sources are reviewed.

3.1 Occupational Segregation and Sexual Orientation

Analyses of occupational differences between gay and heterosexual individuals have been conducted using both occupational categories, and finer comparisons across specific occupations. Using the 1989–1991 GSS data, Badgett (1995) found that lesbians and bisexual women are less likely to work in managerial or clerical/sales positions, more likely to work in craft/operative and service positions, and about equally as likely as heterosexual women to work in professional/technical occupations (Badgett and King 1997; Badgett 1995). Significantly, half of the lesbians and bisexual women in her sample fell into the craft/operative and service occupations, which are the lowest paying occupations. Blandford (2003), drawing on 1989–1996 GSS data, found a similar overrepresentation of lesbian and bisexual women within the service occupations.

Most research, however, indicates that LGB persons are overrepresented in professional and service occupations. Badgett (1995), using 1989–1991 GSS data, found that gay and bisexual men were less likely than heterosexual men to be in managerial or blue collar occupations, and more likely to be in professional/technical and service occupations. She concluded that her “results suggest that gay/bisexual men are in higher-paying occupations but earn less than heterosexual men within these categories” (1995: 736). Similarly, Blandford (2003) used 1989–1996 GSS data and found that gay and bisexual men were concentrated in managerial and professional organizations. Klawitter (1998) used 1990 census data and also found that gay men and lesbians were more likely to be in the highest paid occupations, such as managerial and professional positions, and less likely to be in technical/sales or operator/fabricator positions.

Taking a closer look at specific occupations, Baumle and colleagues (2009) used the 2000 Census data to analyze the largest professional occupations (see Table 13.3 for list of occupations). Table 13.3 displays the index of relative advantage for each profession; this index compares how over- or under-represented same-sex partners are in relation to partnered heterosexuals, controlling for the differences of each group in the labor force overall.

Table 13.3 Indexes of relative advantage for gay individuals compared to heterosexual individuals in the 33 largest professions, U.S., 2000
Full size table

Non-representative surveys have typically reflected an overrepresentation of gay men and lesbians within highly paid, professional occupations. Baumle and colleagues’ analyses confirmed some of these stereotypes, indicating that same-sex partners are overrepresented in the professions as a whole relative to heterosexuals. Specifically, they are 10% more likely to be in the professions compared to partnered heterosexuals (Table 13.3). However, when results are disaggregated by sex, the data reveal that lesbians are underrepresented in the professions relative to heterosexual women, being 6% less likely to be employed in the largest professions. Gay men, on the other hand, are overrepresented in the largest professions relative to heterosexual men; they are 26% more likely to be in the largest professions than are partnered heterosexual men.

Further, Baumle and colleagues’ (2009) analyses suggest that same-sex partners are distributed differently within the professions than are partnered heterosexuals. Relative to partnered heterosexuals, same-sex partners are overrepresented in professions concerned with physical or psychological difference and disability (e.g. psychologists, counselors, physicians, special education teachers), those connected with the computer industry, those that could be seen as focusing on effecting change (e.g. lawyers, social workers), and those connected with creative expression (e.g. designers, artists, writers, or architects) (Table 13.3). Same-sex partners are most underrepresented, relative to heterosexuals, in the engineering and teaching professions, excluding postsecondary teaching.

Lewis (2010) similarly finds that same-sex partners might be overrepresented in occupations focused on service or social work. According to his analysis of the 2000 Census data, individuals with same-sex partners are more likely than those with different-sex partners to work for nonprofit organizations. Some of this propensity can perhaps be explained by the smaller differences in pay between gay and straight men within the nonprofit sector, and/or the ability to afford the lower pay typical of the nonprofit sector due to the greater likelihood of having an employed partner and lesser likelihood of having children. Lewis notes, however, that his findings suggest “a strong desire to serve others is an important factor” in generating the observed differences.

Although findings indicate that sexual orientation plays a role in occupational outcomes, the effect that occupational differences have on income is still unclear. Antecol and colleagues (2008) used the 2000 U.S. Census data in a Oaxaca-Blinder decomposition and found that occupational sorting explained little or no variation in the wage differences between gay and heterosexual individuals. Using the GSS data, however, Badgett (1995) found that differences in occupational categories did account for some of the income difference between lesbians and heterosexual women. And Black and colleagues (1997) concluded that much of the observable pay difference between gay men and heterosexual men can be attributable to the occupational choices of gay men. Thus, the effect of occupational segregation on income differences is an area requiring additional research, particularly as data with fine occupational categories are increasingly available.

3.2 The Role of Sex, Gender, and Sexual Orientation in Occupational Segregation

Analyses of occupational categories not only reveal differences by sexual orientation, but also significant sex and gender differences. Gay men and lesbians tend to cross gender barriers in employment more so than their heterosexual counterparts. (Badgett 2001; Blandford 2003). Drawing on census data, Black and colleagues (2007) found that the average man in a same-sex partnership is employed in an occupation that is 47% female; this is compared to heterosexual men who work in occupations that are 39% female. Similarly, the average woman in a same-sex partnership is employed in an occupation that is 55% female, as compared to heterosexual women working in occupations that are 60% female.

Baumle and colleagues (2009) reported similar results from their analysis of the 2000 U.S. census data (see Table 13.3 for details). Their analysis of data on professional occupations revealed that gay men are significantly more likely to work in female professions than are heterosexual males, although they are still underrepresented in female professions as a whole. For example, gay men are much more likely to be teachers than are heterosexual men; data show that gay men are 16 times more likely to work as a preschool or kindergarten teacher than are heterosexual men (Table 13.3; Baumle et al. 2009). This finding supports the notion that the underrepresentation of gay men within the teaching profession appears to be a consequence more of their sex than their sexual orientation. Rather, their sexual orientation actually makes them less underrepresented as teachers than their heterosexual counterparts.

Similarly, lesbians are more likely than heterosexual females to work in male professions, and are less likely to work in female professions than their heterosexual counterparts (Table 13.3). Returning to the teaching profession, Baumle and colleagues (2009) found that lesbians are underrepresented in the teaching professions when their sex would suggest that they should be overrepresented. Specifically, relative to heterosexual women, lesbians are 63% less likely to work as a preschool or kindergarten teacher (Table 13.3). For women, then, sexual orientation rather than sex is the better explanation of the representation of lesbians in these particular professions.

Findings from these studies, thus, suggest that gay men and lesbians are more likely to cross gender boundaries in occupations than are heterosexual men and women (Baumle et al. 2009; Badgett 2001; Black et al. 2000). At the same time, if one examines the representation of gay men and lesbians within occupations, rather than their representation relative to heterosexuals, gay men are overrepresented in the male occupations and lesbians are overrepresented in the female occupations (Baumle et al. 2009; Black et al. 2000). As a result, even though gay men and lesbians are more likely than heterosexuals to cross gender boundaries in occupations, it is notable that they still remain fairly segregated in sex-typed occupations. They are simply less sex-segregated than heterosexuals. In addition, Baumle and colleagues (2009) found that gay men and lesbians are more overrepresented in gender-neutral professions than are heterosexuals, providing further support that LGB individuals may be less wedded to occupations with strong sex segregation.

3.3 Causes of Occupational Segregation

Analyses of population-based data support the notion that sexual orientation plays a role in occupational outcomes. It is unclear, however, exactly how one’s orientation translates into the selection of a particular occupation. Human capital theory would suggest that LGB persons might possess, or lack, certain skills or education, leading to their being sorted into different occupations than heterosexuals. Socialization theory, in contrast, would support the notion that gay men and lesbians are socialized to believe certain careers are more appropriate for their sexual orientation and might, consequently, develop only the skills to pursue those occupations. Some studies suggest that human capital differences (especially education) explain much of the income disparities between heterosexual and gay individuals (see e.g. Antecol et al. 2008; Baumle and Poston 2011). These same human capital differences might then account for occupational variation, but the cause of the human capital differences remains unclear.

Actual discrimination or fear of discrimination could also lead gay men and lesbians to work in particular occupations. Escoffier (1975) suggests that gay individuals might choose a particular occupation where they felt comfortable disclosing their sexual orientation with few repercussions. Similarly, gay men and lesbians could be more likely to be hired into more tolerant (often lower-paying) occupations (Badgett 1995). In this way, discrimination in some occupations would act to limit available choices (Elliot 1993). For instance, past studies have shown low levels of acceptance of gay men and lesbians working as teachers, especially in elementary schools (Elliot 1993; Fassinger 1993; Klawitter and Flatt 1998). In contrast, college and university environments are more accepting of sexual minorities (Fassinger 1993). These different levels of tolerance could well encourage gay men and lesbians to teach at universities, rather than in primary or secondary schools.

Similarly, gay men and lesbians have been limited in their ability to pursue occupations in various branches of government. Legal decisions in the 1960s, 1970s, and 1980s excluded gay individuals from government positions requiring a high security clearance, citing as a justification the notion that they are susceptible to blackmail with the threat of revealing their sexual orientation (McKeand v. Laird 1973; Adams v. Laird 1969; Padula v. Webster 1987). The prior ban on gay individuals in the military also served as a deterrent to many gay men and lesbians when selecting an occupation, and its legacy continues to be a deterrent to serving (or serving openly) in the military.

Some occupations, therefore, are more friendly to gay individuals than others. Opportunities and choices hence play a significant role in the segregation of the workforce. Gay individuals may not freely choose stereotypical or lower-paying occupations, but might settle for positions that they believe will accept them.

4 Disclosure

The ability to disclose one’s sexual orientation on the job could play an important role both in income and in occupational choice. Without disclosure, it becomes more difficult for individuals to discriminate on the basis of orientation. Income, then, could be positively affected by failing to disclose (and negatively affected by disclosure) depending on the environment. Further, individuals might select an occupation based on the perceived ability to disclose their orientation. Even without disclosure, however, there could be economic “costs” associated with remaining in the closet. As Badgett (2001) noted, an individual attempting to remain in the closet might choose not to participate in “career-advancing social situations”, distance him or herself from co-workers, or switch jobs in order to avoid disclosure and/or the pressure of secrecy.

On the flip side, disclosure might be beneficial in some cases. Badgett (1996: 43) notes that: “Lesbians disclosing their sexual orientation could conceivably benefit by removing employers’ fears or prejudices about their likelihood of marrying and quitting to raise a family.”

Several surveys have attempted to capture the degree to which gay men and lesbians disclose their orientation to their employers. According to GSS data, over a third of lesbian, gay, and bisexual respondents indicated that they had not disclosed their identity to anyone in the workplace (Sears and Mallory 2011a). Only around a quarter of respondents had disclosed their orientation to all of their co-workers. Of those who had disclosed their orientation in the workplace, 38% indicated that they had experienced some form of discrimination in their employment within the five previous years. In contrast, only about 10% of those who had not disclosed their orientation reported experiencing discrimination within the past 5 years. These data support the notion that disclosure can be hazardous, perhaps affecting income and/or occupational choice.

Possibly due to these repercussions, Badgett (2001) found that gay men and lesbians who responded to a non-representative survey were more likely to disclose their sexual orientation when an employer had a nondiscrimination policy in place. This suggests that individuals are more likely to disclose in environments where they feel protected and safe. On a more macro level, Baumle and Poston (2011) report similar findings. Living in a more socially conservative state actually increased one’s earnings, as compared to living in a more liberal state. This suggests that individuals living in environments where they felt less protected could well choose to avoid disclosure and, paradoxically, avoid discrimination in the workplace by hiding their sexual orientation.

5 Conclusion

The research reviewed in this chapter reveals the important role that sexual orientation plays in labor market outcomes, suggesting that sexual orientation is a useful demographic variable to include in analyses of income and occupation. The degree to which the reported differences can be attributed to employment discrimination, however, is difficult to determine in the absence of direct evidence. Reports of discrimination and legal complaints of discrimination provide some guidance in evaluating the prevalence of discriminatory experiences for LGB individuals.

According to the 2008 GSS, approximately 27% of lesbian, gay, and bisexual respondents reported experiencing discrimination on the job (Sears and Mallory 2011a). This discrimination primarily took the form of workplace harassment, but 7% reported having lost a job due to their sexual orientation. As previously noted, reports of discrimination were greater for those who disclosed their orientation to their colleagues.

Complaints filed with state or local equal employment agencies also provide evidence of the prevalence of discriminatory experiences. These serve more as a baseline of discrimination, given that only a limited subset of individuals who experience discrimination go on to file a formal complaint. Research indicates that complaints are filed under sexual orientation discrimination laws at a rate similar to that of complaints based on sex, although less than those based on race or ethnicity (Ramos et al. 2008; Rubenstein 2007). Sears and Mallory (2011b) assessed whether complaints were filed at a similar rate by public and private employees. They estimated that approximately three out of every 10,000 lesbian, gay, or bisexual state or local employees filed a discrimination complaint, whereas approximately four out of every 10,000 private employees did so. All of these studies, then, indicate that experiences with employment discrimination are relatively common for LGB individuals, and that nondiscrimination laws (where available) are used at a similar rate to those protecting other categories of persons.

Data from demographers provide a representative picture of the existence of differences in employment outcomes based on sexual orientation. These findings raise further questions regarding the causal mechanisms that produce these differences, particularly the role of discrimination. In order to generate a more complete picture of the role of sexual orientation in producing employment outcomes, new data are needed that allow both individual and contextual factors to be considered in income and occupation analyses. In particular, datasets with sufficient sample sizes of LGB identified persons are needed, as well as representative data on disclosure in the workplace. These data would permit analyses that could shed light on the existence of workplace inequalities, thereby assessing the need for additional employment protection for LGB individuals.