Impact on Earnings

If the reports of discrimination in those surveys are based on accurate perceptions, then we would expect to see some impact on other measures of employment success. For instance, if discrimination prevents promotions or creates salary inequities, we would expect to see lower employment earnings for lesbian, gay, and bisexual people.

Contradictory surveys:

Opponents of employment protections for gay, lesbian, and bisexual people often cite marketing surveys designed to attract advertising dollars by presenting gays and lesbians as an affluent, desirable market. One such survey by Simmons Market Research Bureau states that the annual gay household income was $55,430 in 1989, far more than the typical U.S. household.¹⁷ A more recent Simmons Study released in 1996 found that 28% of gays earned more than $50,000, and 21% of gay households had incomes of over $100,000 per year. Another company, Overlooked Opinions, promotes its own survey results, showing similarly high incomes.

The methods by which Simmons and Overlooked Opinions identified gay people to survey resulted in a high income sample. In doing so, these marketing statistics are misleading and provide an inaccurate picture of gays' and lesbians' economic status.

The people who filled out the first Simmons study were readers of gay newspapers and magazines. Readers of newspapers and magazines tend have more education and, therefore, higher incomes than the average person. For instance, readers of Ebony or Jet earn 40 to 80% more than the typical African-American worker. People reading gay newspapers and magazines most likely have higher incomes than typical gay men or lesbians.

The more recent Simmons study surveyed people on mailing lists of three groups: a national gay political organization, a mail order house, and a credit card company. Because they make political contributions and qualify for credit cards, these individuals also represent the high end of the income scale.

Overlooked Opinions finds survey participants through similarly skewed means, recruiting people who attend events like the 1993 March on Washington. People who could afford to travel to Washington from throughout the U.S. and stay there for a few days are likely to be more affluent- not an economic cross-section of gay, lesbian, and bisexual people.

Appropriate data:

To create a more accurate economic portrait of a typical gay, lesbian, or bisexual person, we need data that meet two requirements. First, we need more accurate and representative data from a random sample. Several such academically respectable surveys done over the last decade have provided data to use in comparisons between gay and heterosexual people. (The definitions of sexual orientation vary in these surveys, involving different factors related to behavior and self-identity.) Second, we need data on what people actually earn from their jobs.

Overall, the findings from scientifically sound surveys clearly show that lesbians and gay men are not a wealthy subgroup of the United States population. In most cases, they earn less than their heterosexual co-workers.

Figure 1 compares the average earnings for gay and heterosexual men from three surveys, and Figure 2 does the same for lesbians. In each survey, gay men earn less than heterosexual men, with the difference ranging from 4 to 7%.

For women, the difference depends on whether only full-time workers are included in the data. In the comparisons of full-time workers from the General Social Survey and Yankelovich Monitor data, lesbians earn less than heterosexual women. Analysis of the 1990 Census data reveals that partnered lesbians work many more hours than married heterosexual women, on average.

The Impact of Discrimination:

Comparisons of average incomes still do not answer the question about discrimination. A better way to isolate the employment effect of being in a stigmatized group involves comparing people with similar income-affecting characteristics. This is the same kind of evidence used to show the harmful economic effects of other kinds of discrimination, including those based on race, sex, or disability.²¹ The underlying principle is that people who are similarly qualified should, on average, receive the same earnings. If one group systematically earns less, discrimination is a likely cause.

Next Page | 1, 2, 3, 4

NOTES:
17. Joan E. Rigdon, "Overcoming a Deep-Rooted Reluctance, More Firms Advertise to Gay Community," Wall Street Journal, July 18, 1991, p. B1.  [ Go Back ]
18. The General Social Survey is an annual survey by the National Opinion Research Center at the University of Chicago and includes questions on the sex of sex partners since the age of 18. Figures 1 and 2 present the incomes of people who have had one or more same-sex partners. See M. V. Lee Badgett, "The Wage Effects of Sexual Orientation Discrimination," Industrial and Labor Relations Review, July 1995.  [ Go Back ]
19. The 1990 United States Census of the Population allows the identification of people who report having a same-sex "unmarried partner." Lesbian, gay, and bisexual people who do not have an unmarried partner or who have an opposite-sex partner are not counted as gay. See Marieka Klawitter and Victor Flatt, "The Effects of State and Local Antidiscrimination Policies for Sexual Orientation," manuscript, University of Washington, 1997.  [ Go Back ]
20. The Yankelovich Monitor is an annual survey of attitudes and demographics and included a question on sexual orientation (identifying as "lesbian/gay/homosexual") in its 1993 survey. See Grant Lukenbill, Untold Millions, New York: Harper Business, 1995.  [ Go Back ]
21. For a good discussion of this technique, see Morley Gunderson, "Male-Female Wage Differentials and Policy Responses," Journal of Economic Literature, Vol. 27, March 1989.   [ Go Back  ]