Marieka M Klawitter. Handbook of Lesbian and Gay Studies. Editor: Diane Richardson & Steven Seidman. 2002. Sage Publication.
The emergence of sexual orientation as a component of personal identity was closely tied to changes in the organization of economic activity in the past century. Industrialization and urbanization allowed many men and women to separate from farm-based families and from prescribed productive and reproductive roles (D’Emilio, 1983). Urban wage labor markets provided the economic means for men and women to create lives apart from heterosexual marriage and to create sexual and affectional identities (Matthaei, 1997). Throughout the world, the ability to live as gay or lesbian is still greatly affected by the level of economic development and by personal wealth.
Economic relationships have also been an enduring target for advocates for and against gay rights in industrialized countries. In the 1950s and 1960s, gay bars, bathhouses and other businesses were targets for those seeking to wipe out homosexuality (Escoffier, 1997). Patrons of those businesses risked losing their jobs if arrested in a public raid or if otherwise identified as homosexual. Since then, attitudes about homosexuality have changed and gay businesses have greatly expanded in number and variety. Many cities have developed gay neighborhood business districts and directories of gay businesses. Similarly, some workplaces have been transformed from dangerous places for known homosexuals to gay family-friendly institutions. Some private organizations and governments have created antidiscrimination policies that protect gays and lesbians from employment discrimination based on sexual orientation.
The incredible changes in gay rights and social acceptance, though not uniform, have allowed many gays and lesbians to openly participate in the economy. In this chapter, I will explore how sexual orientation continues to affect economic participation as workers and consumers. The first section analyzes the recent evidence of discrimination in labor markets. It also considers how sexual orientation affects employment through choices in education, location, occupation, and work time. The second section reviews the effects of sexual orientation on consumption behavior and activities.
Gays and Lesbians as Workers
Earnings from employment are the most important source of income for most individuals and families. As such, earnings are an important determinant of economic well-being and work is an important economic activity. The amount written on a paycheck is determined by a complex set of decisions and outcomes negotiated by individual workers, their families, and employers. Differences in average earnings could be a reflection of discrimination (different treatment or opportunities), different choices, or both. Are gays and lesbians different than others in labor market opportunities and choices that affect their paychecks? Here we’ll explore the determinants of earnings, the concept of discrimination, evidence of sexual orientation discrimination in labor markets, and the possibility of different employment choices made by gays and lesbians.
Is There Discrimination in Labor Markets?
Labor market discrimination is differential hiring, firing, promotion, or wages for individuals based on their real or perceived group membership. The past history of employment discrimination based on sexual orientation by government and private employers is well documented in court cases (Harvard Law Review, Editors, 1989: 44-74). However, some have questioned whether gays and lesbians are still subject to labor market discrimination (Hewitt, 1995). Widespread discrimination should lower average earnings by closing off job and promotion opportunities and encouraging gays and lesbians to trade off high wages for more supportive working conditions (a ‘compensating differential’).
The best quantitative evidence in the discrimination debate comes from econometric studies that use national random samples. These studies use multivariate statistics to assess the effects of sexual orientation on average earnings, after disentangling the effects of other factors. Overall, the econometric evidence of sexual orientation discrimination is mixed. There is stronger evidence of discrimination against gay men than against lesbians.
Badgett (1995) used data from the US General Social Survey from 1989 to 1992 to study annual earnings. She estimated that men who had same-sex sexual experiences earned 11 to 27 per cent less, on average, than did other men, after controlling for age, education, occupation, marital status and region of residence. Average earnings were also lower for women who had same-sex sexual experiences, but the differences were not statistically significant. Two newer studies replicated Badgett’s work with additional years of data and alternative definitions of sexual orientation (Blandford, 2000; Black et al., 1999). Both studies found that gay men earned less than heterosexual men and that lesbians earned more than heterosexual women.
My co-author, Victor Flatt and I (1998) used the 1990 US Census data to compare average earnings for members of different-sex couples (married and unmarried) and same-sex couples. After controlling for differences in individual characteristics and labor market conditions, we found that men in same-sex couples earned about 26 per cent less than married men, but about the same as men in unmarried different-sex couples. Women in same-sex couples earned significantly more than women in married couples and women in unmarried different-sex couples. However, the higher earnings for women in same-sex couples were almost completely explained by the greater hours and weeks worked by that group. Beyond the effects of sexual orientation, per se, men and women in same-sex couples had higher average levels of education and were more likely to live in urban areas. Both of these factors contribute to higher levels of earnings.
These studies show mixed evidence of earnings discrimination against gays and lesbians. In contrast, there have always been large earnings gaps between people of color and whites and between women and men that were not explained by other characteristics. Is it possible that labor market discrimination might not show up in the average amount of paychecks?
There are two primary explanations for the lack of consistently large differences in average earnings: the ability to pass as straight and differences in work effort based on gender roles. If gays and lesbians are not ‘out’ at work, then they may face less discrimination and avoid lower wages. As I discuss below, most workers are out to some, but not all co-workers. The second explanation is based on the gender of the worker and his or her partner. The pattern of gay men earning less than married men and lesbians earning more than married women (before accounting for hours of work) is consistent with workers making adjustment to their own work effort based on potential earnings from a partner. Because women still earn significantly less than men on average do, a lesbian can expect significantly less income from a potential partner than can a gay man. She may increase her work effort (e.g., number of hours and weeks of work, investment in job training) to compensate. I’ll discuss this possibility more when I consider intrahousehold work decisions below. The effects of gender on work could offset lower earnings due to sexual orientation discrimination for lesbians.
Other evidence of employment discrimination is available from surveys of the general public and of gays and lesbians. Data from national random samples show that Americans have increased general support for equal job opportunities for gays and lesbians from 56 per cent in 1977 to 84 per cent in 1996 (Yang, 1998: 6). However, in 1996 only 55 per cent of Americans thought that ‘homosexuals’ should be hired as elementary school teachers and only about half thought being lesbian or gay should not disqualify someone from being president of the USA (ibid.: 10-11). Sherrill (1996) found that US public attitudes (on a ‘feeling thermometer’) toward gays and lesbians were worse than those toward any other group except illegal aliens. These statistics suggest that public support for employment discrimination is decreasing, but still not negligible.
Surveys of gays and lesbians have generally found that 25 to 30 per cent believe that they have experienced sexual orientation discrimination (Badgett, 1997b). In addition, most respondents were ‘out’ to some but not all co-workers which suggests significant fear of discrimination. Similarly, Woods (1993) describes the incredible efforts by some gays to hide their sexual orientation from co-workers and employers. These studies point to perceptions of discrimination, but are not from random samples and therefore may not be representative of the experience of the general gay and lesbian population.
Some countries and many states, provinces, or cities have laws that prohibit discrimination based on sexual orientation in public or private employment (International Gay and Lesbian Human Rights Commission, 1999). Many European nations adopted antidiscrimination legislation in the late 1990s and South Africa became the first country to include sexual orientation as a protected category in its constitution in 1996.
As of 2000, the USA does not have federal legislation that bans private employers from discriminating as it does for race, sex, religion, national origin, age, and disability. However, this type of legislation has been introduced in the US Congress as the Employment Non-Discrimination Act. There are currently 11 states and many local governments that do have antidiscrimination legislation for private employment (van der Meick, 2000: p4; General Accounting Office, 1997). Local policies have usually been adopted in urban areas with higher than average levels of education and numbers of non-family households (Klawitter and Hammer, 1999; Haeberle, 1996; Wald et al., 1996). These characteristics, not surprisingly, also predict lower levels of anti-gay feeling and higher levels of average earnings.
Studies of the federal legislation banning discrimination based on race and sex suggest that these laws contributed to labor market gains for blacks and for women (Burstein, 1985; Donohue and Heckman, 1991; Gunderson, 1989). However, we did not find similar gains for members of same-sex couples covered by antidiscrimination policies (Klawitter and Flatt, 1998). Using multivariate analysis, we estimated the earnings effects of living in a state or local area that had sexual orientation in its employment non-discrimination policy. Our results showed that these policies were adopted in places that had higher average wages for people in both same-sex and different-sex couples, but there was no differential effect for same-sex couples. We also did not find that the policies were generally adopted in places with relatively higher wages for people in same-sex couples (after accounting for the effects of education and urban location). These results could reflect the lack of enforcement of non-discrimination policies, an unwillingness of gays and lesbians to initiate legal action, or lack of clearly identifiable employment discrimination.
Employment benefits make up another significant part of compensation for employment and gays and lesbians often do not have the same access to these benefits. Overall, benefits average almost 30 per cent of all compensation for work (Economic Benefits Research Institute, 1999). Many married workers can obtain health insurance for spouses who are self-employed, working in jobs without health benefits, or not in the labor market. Although a growing number of government and private employers offer benefits to domestic partners of unmarried employers, this coverage remains scarce. A 1999 study found that only 12 per cent of workers worked for an employer who offered benefits for same-sex domestic partners (Economic Benefits Research Institute, 2000). Access to health benefits for domestic partners would allow families a wider array of opportunities for combining employment and home work.
Discrimination may also affect other job characteristics such as the ability to socialize with co-workers and to participate in the workplace community. Ellis and Riggle (1995) found that gays and lesbians who were out were more satisfied with their relationships with co-workers and those working for companies with antidiscrimination policies were more satisfied with their jobs. An early study found that lesbians were more likely to be out in jobs where they worked in small organizations, in female-dominated human services occupations, when they did not work with children, and when their incomes were low (Schneider, 1987). Schneider did not find that being out had a direct effect on the chances of socializing with co-workers outside of the workplace. The choice of strategy for ‘managing’ sexual orientation at work may dictate the level of social integration. Woods (1993) found that some gay professional men chose to ‘play straight’ at work or to maintain strict boundaries between their personal and professional lives. These strategies might minimize possible discrimination, but would also exact a toll by limiting social interaction at work.
Do Gays and Lesbians Make Different Choices in the Labor Market?
Individuals and families make choices about schooling, work, and home that affect wages and earnings. The effects of gender and sexual orientation on these choices could explain some differences in labor market outcomes for gays and lesbians. Also, real or potential discrimination in the labor market could feed back to affect these choices.
Economists use the term ‘human capital’ to refer to skills and knowledge that are valuable (and therefore rewarded) in the labor market. Job experience and schooling are the most common forms of human capital.
Several studies have found higher levels of education for gays and lesbians than for others, among both men and women (Blumstein and Schwartz, 1983; Klawitter, 1998; Black et al., 2000). Black et al. (1999) found that the higher level of education for gay men was not likely the result of reporting self-selection because the education levels for fathers of gay men were similar to the levels for fathers of other men. If men from higher socio-economic backgrounds were more likely to identify themselves as members of a same-sex couple, then we would expect to see higher average education levels for their fathers.
Gay men and lesbians might get more education to offset expected earnings effects of discrimination in employment. Alternatively, educational settings may be relatively gay-friendly and encourage longer stays for gays and lesbians. Either way, education is one of the strongest predictors of earnings and higher levels push earnings up for gays and lesbians. This is one of the reasons that econometric studies that account for the effects of education are important in assessing earnings discrimination.
Gays and lesbians are more likely to be located in urban areas than are other people (Badgett, 1995; Black et al., 2000). Indeed, Black et al. (2000: 148) found that 59 per cent of men in same-sex couples were clustered in the twenty cities with the largest gay population (compared to only 26 per cent of the general US population in those cities). Lesbians were somewhat less clustered; 45 per cent were in the top twenty cities (with 25 per cent of the US population).
Gays and lesbians may move to urban areas to find a supportive gay community and to find tolerant labor markets. Because wages (and the cost of living) are generally higher in urban areas than in suburban and rural areas, this pattern should lead to higher earnings for gays and lesbians. Consistent with this, our findings showed that same-sex couples lived in areas with higher average wages (Klawitter and Flatt, 1998). However, we did not find that those labor markets were so much more tolerant that average wages were differentially higher for gays and lesbians after accounting for other characteristics.
Popular images of gay men as hairdressers and decorators and lesbians as construction workers convey the notion that sexual orientation affects occupational choice. Occupational choice is a complex process tied up with schooling decisions, lifestyle choices, and work opportunities.
Badgett and King (1997) studied occupational choice by gays and lesbians and how closely it followed the occupational patterns of anti-gay sentiments. They hypothesized that gays and lesbians might choose occupations based on the potential for harassment (anti-gay sentiment), the ability to hide sexual orientation (level of social requirements on the job), and compatibility with family structure (less frequent child-rearing for gays and lesbians). They found that levels of anti-gay sentiments did seem to vary by occupation, but that urban location, education levels, and age explained much of the variation. Although hampered by very small sample size, their results also suggest that gay men may cluster in more tolerant occupations, but that lesbians do not. In my work, I found that women in same-sex couples were more likely than women in different-sex couples to be in managerial or professional jobs and much less likely to be in technical/sales or operator/fabricator jobs (Klawitter, 1998). Blandford (2000) found a similar pattern by sexual orientation for women in technical occupations, but a different pattern for those in service and operator occupations. For men, Blandford found that unmarried gay men were more likely than heterosexual men to be in managerial and technical occupations, and less likely to be in production or operator occupations.
On the whole, grand patterns of occupation choice by sexual orientation have not materialized. However, it does seem that occupational choice may affect the wage gap by sexual orientation. Badgett’s (1995) estimates of sexual orientation discrimination were larger for men and smaller for women when she controlled for occupational choice. I found that occupational choice partially accounted for higher wages for women in same-sex couples (Klawitter, 1998). Similarly, when Blandford (2000) used more detailed occupational information he found that less of the difference in average earnings was attributable to sexual orientation for both men and women. Further work on the interrelations of occupational choice and sexual orientation might help sort out the patterns.
Earnings are greatly affected by the amount of time spent working. Having a partner or having children can change the incentives to work in the market and the amount of time devoted to work for wages. In traditional husband/wife families, marriage and child-rearing increase work time and earnings for husbands and decrease them for wives.
Black et al. (2000) estimated that about 40 per cent of lesbians and about 20 to 30 per cent of gay men had partners living with them. In contrast, more than half of all adults are married and another 3 per cent cohabit with different-sex partners (Bureau of Census, 1993: Table 16). Black et al. also estimated that there were children living with 22 per cent of lesbian couples, 5 per cent of gay male couples, 36 per cent of unmarried different-sex couples, and 59 per cent of married couples. They estimated higher levels of child-rearing for all gays and lesbians (not just couples): 28 per cent for lesbians and 14 per cent for gay men.
Economists and sociologists have generally assumed that family members jointly make decisions about work. There is a large literature on intrahousehold allocation of home work and labor market work for heterosexual couples and a smaller amount on allocation for homosexual couples. Economists have viewed the allocation within families as reflecting the comparative advantage of partners in home and labor markets (Becker, 1973). This model suggests that the partner with the greatest opportunities in the labor market (highest potential wage) will spend more time on a job and less time doing home work. The other partner will then ‘specialize’ in home production—using time and market goods to cook, clean, and make a happy home life. The economic approach is very useful in recognizing some incentives for household allocation (for example, wage rates and overall income), but generally ignores the importance of gender roles or other cultural influences.
The effects of gender are at once diminished and amplified in same-sex couples. For same-sex couples, gender roles are less useful as a tool for assigning work tasks within the family. Gender may matter less for an individual’s labor market decisions (and a partner’s gender may matter more). For example, women with female partners may work more in the market (and thereby take on a less traditionally feminine role) to offset lower expected partner earnings (due to gender). Similarly, men with male partners may work less because they expect male-sized earnings from their partner. However, gender may be amplified by the presence of two people with similar gender socialization and by an effort to maintain gender identification while doing work that is normally reserved for the other gender. For example, Carrington tells how he found that partners seemed to portray their partners’ work as gender appropriate by mis-remembering or mis-reporting who had completed household tasks (1999: 51-2).
In my work using US census data (1995), I found that same-sex couples were much less likely to have one member specialize in the labor market than were married couples. (Unmarried different-sex couples were in-between.) Partners in gay male couples were the most likely to work similar numbers of weeks in a year and hours in a week, followed by lesbian couples. Earnings for partners was the most similar for lesbian couples (r = .42) then for gay male couples (r = .29) and unmarried different-sex couples (r = .26). Married couples had the lowest correlation in earnings for partners (r = .12). High correlation shows a lack of labor market specialization, but it may also reflect the matching gender for partners—hence matching gendered labor market opportunities. However, gay and lesbian couples are less closely matched on age, education and race than are married couples (Klawitter, 1995) and this could work to lower the earnings correlations for same-sex couples.
Several studies have found that gender and sexual orientation affect the division of home work (Blumstein and Schwartz, 1983; Kurdek, 1993). Kurdek found that lesbian couples tended to share home work by alternating tasks or doing them together, while gay male couples tended to split up the tasks. In married couples the wife usually did most of the household tasks. Based on an extensive qualitative study of San Francisco area same-sex couples, Carrington (1999) argues that equality in household work in gay and lesbian households is a myth. It seems that gay and lesbian families share market and home work more equally or fairly (not solely on the basis of gender) than do married couples, but the allocation is not simply 50-50.
Specialization may be economically efficient, but it is a risky strategy for those specializing in home production because a break-up would leave them without income (England and Kilbourne, 1990). This is probably one of the reasons that unmarried different-sex do not specialize as much as married couples (Blumstein and Schwartz, 1983; Klawitter, 1995). Similarly, Lundberg and Rose (1999) found that specialization was much less likely in married couples that later divorced. Same-sex couples (like other unmarried couples) lack the legal protections of marriage that might allow home specialists to share assets after separation. For example, Carrington (1999: 180-4) tells the story of a gay man who experienced a huge decrease in income after leaving a high-earning partner. The man did not expect any compensation or a share of the property acquired as a couple despite having done a majority of home work during the relationship.
To summarize, families make decisions about household work and labor market work and those decisions follow patterns based on sexual orientation and legal marriage. The implications for earnings are that we might expect women in same-sex couples to work more in the labor market than other women because of the need for more income, the lesser effect of gender roles in the household, and the lack of legal protections. Similarly, gay men might work less than married men because they have a male partner’s earnings and lower gender role expectations in the home, but this might be offset by the lack of legal marriage protections. These influences may partly explain the patterns of earnings by sexual orientation found in the quantitative studies. The remaining influence of gender roles on behavior may mitigate these effects by encouraging gay men to work in the labor market and lesbians to work in the household.
The other side of the family balance sheet involves expenditures—using income to buy goods and services. In a simple economic model, consumers buy goods and services to increase their ‘utility,’ subject to the constraints of prices and income. But in the world we live in, people shop for entertainment, consume to create identity, and are categorized and targeted by marketers seeking to influence consumption decisions. In this section, I will review research on how sexual orientation affects consumption decisions and experiences. Do gays and lesbians make different consumption choices than others? Do they face discrimination in consumption markets? How are gays and lesbians targeted by marketers and what are the implications for consumers and for the gay community?
Few studies of consumption patterns distinguish between consumers by sexual orientation. Some marketing studies have suggested that, compared to heterosexuals, gays and lesbians are more likely to spend money, spend more on luxury goods (travel, entertainment, expensive clothing), and are more brand-loyal and fashion conscious (Penaloza, 1996: 26). These studies generally reflect the tastes and habits of a non-representative sample of gays and lesbians and are used to bolster claims of a lucrative and under-tapped market.
Other sources of information confirm the idea that gay and lesbian consumers at least partly create their sexual identity through their consumption choices—what they buy and where they shop. Interviews conducted by Freitas et al. (1996) support the notion that some gays and lesbians use clothing and jewelry to help define themselves as part of a queer subculture. Carrington (1999: 154) described how some of the families in his study consciously shopped in gay-owned businesses in spite of paying higher prices—they labelled it the ‘gay tax.’ Especially in areas with gay-identified neighborhoods, the act of shopping may become part of the performance of sexual orientation.
Do gays and lesbians face discrimination in consumption markets? In a small but carefully scripted study of retail stores, Walters and Curran (1996) found that straight students posing as gay or lesbian couples were treated much differently than were those posing as heterosexual couples. The straight couples received quick, courteous treatment while the gay and lesbian couples were helped more slowly (if at all) and were subjected to pointing, laughing, and rude comments. Similarly, Jones (1996) found that gay couples had less success making hotel reservations than did straight couples.
Much of the hype about the new gay market is based on claims of high income for gays and lesbians based on results of non-representative surveys from marketers (Badgett, 1997a). There is, however, some merit to the idea of gay men as having relatively high household incomes. Using 1990 census data, I found that average incomes were $58,489 for gay male couples, $47,192 for married couples, $45,756 for female same-sex couples, and $37,518 for unmarried different sex couples (Klawitter, 1995). After accounting for location and individual characteristics for partners, male same-sex couples and married couples had similar incomes; female same-sex and unmarried different sex couples had lower incomes. Combining earnings for two men leads to higher average income for gay couples, despite the estimates of lower individual earnings for men in same-sex couples (as I discussed above). Similarly, women in same sex couples earned significantly more than other women (before controlling for other characteristics), but live in households without a male-sized income. These figures do not, however, include people who are not partnered and the best estimates suggest that fewer gay men and lesbians are partnered than are heterosexuals. However, for marketing purposes, family income and family spending habits are likely to be the appropriate metric, unlike comparisons for employment discrimination.
Gays and lesbians may benefit as a ‘target market’ because it may encourage the development of products and services that affirm queer culture and better serve family consumption needs (for example his and his monogrammed bathrobes, all-women travel packages, or baby bibs with rainbow flags). In addition, having advertisements and products for gays and lesbians serves to legitimize the community as participants in the economy. However, some have worried that marketing hype has provided fodder for anti-gay campaigns that portray gays and lesbians as privileged, frivolous, and not in need of antidiscrimination protection (Gluckman and Reed, 1997). Also, there may be a tension between being a target market and being a social or political movement (Penaloza, 1996; Freitas et al., 1996). Focus on marketing images of consumption by ‘beautiful’ people may de-emphasize political goals and lead to the alienation and ostracism of people less centrally situated in mainstream culture.
My overview of the social science literature suggests that sexual orientation plays a large role in work outcomes, but that there is mixed evidence of discrimination in average earnings for gays and lesbians. The influence of gender on work in both markets and families seems to trump sexual orientation. In consumption markets, gays and lesbians have moved from being targeted by police in gay bars to being targeted by marketers. Of course these changes, like the changes in the labor market, are neither universal nor uniformly seen as progressive. Individuals sometimes face discrimination when buying goods or services in spite of being legitimized as buyers through some marketing campaigns.
The quality and quantity of research on the economics of sexual orientation have grown immensely in the past thirty years. However, many areas remain unexplored or inadequately understood. How do sexual orientation and gender separately and jointly affect family economic decisions? How do the effects of sexual orientation on employment change over the life cycle? How much of our understanding of sexual orientation will be relevant in the changing social, economic, and historical context? How do economic opportunities for gays and lesbians vary by geographic location and social context? How have public policies affected the opportunities and choices for gays and lesbians?
Changes in the social and legal context have enhanced our ability to research issues of sexual orientation. Researchers can publish work on sexual orientation, funding agencies and foundations will support the research, and respondents will answer questions about sexual behavior or identity. Yet, the availability of good quantitative and qualitative data has lagged behind our willingness to analyse the data. I hope that the new century brings new data collection efforts aimed at exploring the economics of sexual orientation.
There have been remarkable changes in how gays and lesbians live and work in the past thirty years. Today many of us participate fully and openly in society and the economy. However, some still face serious discrimination and barriers in employment or consumption. For all, sexual orientation and gender still serve as categories that determine the way families and individuals participate in the economy.