Neither An In-Law Nor An Outlaw Be: Trends in Americans’ Attitudes Toward Gay People

Oscar Wilde once quipped, "Fashion is a form of ugliness so intolerable that we have to alter it every six months." In the wake of the 2004 presidential election, it looked as if the gay and lesbian movement—which had been enjoying a string of unprecedented successes—had suddenly fallen terribly out of fashion.

American gays and lesbians began the twenty-first century in the largely fawning spotlight of popular TV shows like Will and Grace, which centers on the relationship between a gay man and the straight woman who is his best friend, and Queer Eye for the Straight Guy, in which a team of gay experts on fashion, culture, and interior design each week "makes over" the life of a different couth-challenged heterosexual male. In 2003, gay people celebrated a historic legal victory overturning sodomy laws at the U.S. Supreme Court, and a year later they won the right to wed one another legally in Massachusetts. But on November 2, 2004, no fewer than eleven states—ranging from traditional Utah to more liberal Oregon—overwhelmingly passed referenda adopting state constitutional amendments banning same-sex marriage. (Two additional states—Louisiana and Missouri—enacted similar amendments earlier in the year.)

And as a series of largely inaccurate stories were published indicating that "moral values" had played an unexpectedly large role in turning voters out to reelect George W. Bush, much of the blame for the defeat of John F. Kerry and the Democrats was heaped upon LGBT people. Straight Americans, who had once appeared perfectly comfortable with (or at least titillated about) the idea of gay people peeking into their hall closets and underwear drawers, were now shooing them out the door as if they were houseguests who had overstayed their welcome.

Are these most recent events the dawn of a new era—or merely a terribly ugly fashion of the political moment? While only time will tell for sure, here we explore this question by documenting trends in Americans' attitudes regarding their feelings toward gay people and gay and lesbian issues—including their feelings about the morality of homosexual sex and gay relationships, their opinions on various policies affecting the status of gays, and some of the political ramifications of these attitudes.

While Americans' warmth toward gay people has increased since surveys began regularly seeking attitudes toward them in the early 1980s, a striking fact remains true even in 2004: Many Americans dislike gay people, and they aren't reluctant to say so to survey researchers. As we shall see, any efforts by the gay rights movement to promote policies favorable to gay people are handicapped by this deep-seated antipathy, which is shared (to varying degrees) by Americans of all races, backgrounds, and ages.

The "feeling thermometer," on which survey respondents are asked to locate their feelings toward a group of people on a scale ranging from zero for the coldest possible feelings to one hundred for the warmest, is a standard measure of affect. The National Election Studies, conducted by the Center for Political Studies at the University of Michigan, used the feeling thermometer on eight occasions between 1984 and 2002 to measure the distribution of Americans' feelings toward gay men and lesbians, among other groups.

We calculate the public's affect toward gay people first in terms of the mean feeling-thermometer score given to gays and lesbians by respondents. The mean score gives us a sense of exactly how warm or cold feelings are toward a group. These data tell us that between 1984 and 2002, the American public moved from feelings that best can be described as icy (a mean score of 30) to a temperature just a shade below neutral (46).

(Click for larger view of Figure 1a-1b.)

A second measure is the proportion of respondents who assigned gays and lesbians the lowest score of all the social groups included in that year's survey (which average about twenty per year).

This measure gives us a better sense of the place of gay people with reference to other social groups, allowing us to determine the extent to which those who do not dislike lesbians and gay men still like gay people less than anyone else. In 1984, a majority of Americans—including a majority of virtually every demographic group—ranked gay men and lesbians last among groups included in the study. (This was no mean feat; on average, respondents rated twenty-three different groups on the feeling thermometer that year.) By 2002, in only one major demographic group—African Americans—did a majority rate gay people last on the feeling thermometer.

African Americans are distinctive as the only group to have shown no significant increase in warmth toward gays and lesbians during this time period. "Blacks' current attitudes" toward gays are all the more interesting because they represent a dramatic reversal of their pattern of increasing warmth toward gay people, beginning at 36 in 1984 (warmer feelings than those of all other groups with the exception of those with college degrees), reaching a peak in 1998 with a mean score of 50, and followed by mean scores of 42 in 2000 and 38 in 2002—lower than any other group in that year. We can only conjecture about this reversal, and leave the explanation for it to future research.

How do these figures help us answer whether we are witnessing the "end of an era" of gay rights? We believe that LGBT people can take particular heart from the age cohort data: Americans of all ages are becoming more—not less—tolerant as they grow older. And older, colder, Americans are being replaced by citizens who express more warmth for gay people. But gays and lesbians have two reasons to be nervous: One is blacks' lukewarm feelings toward gay people (which would appear to present an opportunity to drive a wedge into the Democratic civil rights coalition); the other is a slight cooling trend toward gays that appeared across all demographic groups between 2000 and 2002.

Perhaps just as important as the feeling thermometer in determining Americans' attitudes on a range of gay issues is the degree to which they morally disapprove of homosexuality. This disapproval has diminished dramatically in the past dozen years. Between 1973 and 1991, the share of Americans who told the General Social Survey (GSS) that sexual relations between two adults of the same sex are "always wrong" hovered between 70 and 78 percent. But shortly thereafter, it proceeded to drop precipitously, from 76 percent in 1991 to a low of 55 percent in 2002.

(Click for larger view of Figure 2.)

Remarkably, this shift cannot be due to a widespread change in traditional sexual mores. In the same period, the share of Americans saying premarital sex is always wrong stayed just about constant (ranging between 26-28 percent), and the proportion of Americans saying adultery is always wrong has, if anything, risen (from 76 to 80 percent).

(Click for larger view of Figure 3.)

Most of the change in opinion regarding gay sex appears to have occurred among those who are untroubled by premarital sex. In the GSS surveys conducted between 1988 and 1992, 48 percent of Americans said homosexual sex was always wrong but disagreed that this was true about premarital sex.

(Click for larger view of Figure 4.)

In the surveys conducted between 1998 and 2002, only 32 percent of Americans fell into this category. At the same time, those believing that neither type of sex was always wrong increased from 23 to 41 percent. In sum, Americans have not become more accepting of nontraditional sexuality in general; rather, they have become more tolerant of gay sexuality in particular.

Two concurrent phenomena may well account for much of this change—as they may also account for the changing levels of affect measured by the feeling thermometers. First, the proportion of Americans who said they had a close friend or acquaintance who is gay or lesbian more than doubled in a decade, from one in five in 1985 to a majority in 1996. Much of this, no doubt, is a function of the willingness of gay people to come out to friends and family and to put a human face onto something that had been an abstract taboo.

Second, Americans increasingly believe that homosexuals are born, not made: In 1983, only 16 percent agreed that homosexuality was something people were born with, while in 2004, 32 percent held that belief. Those attributing homosexuality to upbringing dropped from 25 percent to 14 percent, while the share of Americans saying that gayness was a choice remained steady.

(Click for larger view of Figure 5.)

This shift away from blaming someone for his or her homosexuality to viewing sexual orientation as a fault-free personal characteristic may be having a large impact on changing evaluations of homosexuality as being right or wrong. The public's changing views of gay people and of homosexuality have been reflected in increased support for gay-related policies over time. But while it is often believed that governments are particularly responsive to public opinion on "hot-button" social policies like gay rights, the actual record is somewhat mixed. The notion of equal rights for gays in employment has long enjoyed support. Today, with support well over 80 percent, the issue hardly can be termed controversial—yet there is no federal legislation against job discrimination on the basis of sexual orientation.

(Click for larger view of Figure 6.)

Similarly, while 80 percent of Americans support the right of lesbians and gays to serve in the military, the "don’t ask/don't tell" policy—under which approximately one thousand gay service members are discharged each year—remains the law of the land. Finally, adoption rights—which are implicitly granted to homosexual couples in many states, and explicitly banned in only one—have relatively low, but increasing, public support among Americans. (Only Florida explicitly prohibits adoption by homosexuals. A handful of other states have provisions in place that put some restriction on the ability of gays and lesbians to adopt children or be foster parents.)

Given that Americans' feelings toward gays and lesbians are tepid at best, what explains the public's relatively solid support for some policies offering equal protection to gays? One reason is that even many of those who dislike gays grudgingly grant them some rights; as shown in Figure 7, the proportions of those who ranked gays last on the feeling-thermometer scale of three recent combined National Election Studies who were willing to extend job protections to gays and permit them to participate in military service ranged in the 30s and 40s across the conservative-liberal spectrum.

(Click for larger view of Figure 7.)

The other reason is that among everybody else—those who do not rank gay people last on the thermometer—political ideology has a marked effect on policy attitudes. One conclusion we can draw from this is that political principles and values can blunt the effect of discomfort and dislike of gays and lesbians to lead Americans to support policies granting rights to gay people—but not among the approximately one-third of the population whose dislike is so strong that they rank gays last.

Compared to the federal level, gay rights proponents have had more success at the state level, where governments appear more responsive to public opinion on these issues. With tens of thousands of interviews conducted over a one-year span, the 2000 National Annenberg Election Study (NAES) provided the first state-by-state account of public opinion in the United States on gay issues. Table 1 below ranks states by the proportion of respondents who told the NAES that the federal government should "do more" to "stop job discrimination against homosexuals," and indicates state laws on three key gay-related policies—discrimination protection, bans on same-sex marriage, and hate crimes laws that explicitly include gay people.

Two things immediately stand out in this table. First, public opinion on gay rights in America closely follows the much-discussed "red-blue" divide. John Kerry was the victor in the ten states (and the District of Columbia) most in favor of gay rights—California, plus most of New England and the Mid-Atlantic region. George W. Bush won the eleven states least in favor of gay rights—all drawn from the South, the Plains, or the Rockies.

Second, state policy closely follows public opinion: The thirteen states that have adopted antidiscrimination laws are generally those whose residents are most supportive of gay rights, with the exception of increasingly conservative Minnesota and Wisconsin. Many of these states have also shied away from enacting a ban on same-sex marriage. Laws that explicitly punish perpetrators of hate crimes against gays and lesbians are more prevalent (they've been enacted by twenty-nine states and D.C.), but again the pattern of their adoption is predicted quite well by the state-by-state aggregate responses to the NAES question.

We now arrive at the very timely topic of same-sex marriage, and—as was demonstrated in the 2004 elections—the issue is currently a real loser for the gay and lesbian rights movement. In the 2004 National Election Pool (NEP) exit polls, 25 percent of the voters supported same-sex marriages, 37 percent opposed any legal recognition of same-sex relationships, and 35 percent favored civil unions.

Perhaps the extraordinary development here, however, is not that 37 percent of the voters opposed any legal recognition of same-sex relationships. Rather, it is that 60 percent supported some sort of legal recognition of these relationships. Civil unions, unheard of five years before, now enjoy substantial support. While the opposition is intense and vocal, the growth in support for recognition of same-sex relationships is dramatic—even if a majority of Americans is not yet prepared to call them marriages.

Public opinion researchers are familiar with the pattern noted by Samuel Stouffer and many others of support for the abstract principles of democracy combined with opposition to many specific examples of free speech, majority rule, and minority rights. Same-sex marriage presents us with the opposite pattern: We see much more support for giving same-sex couples access to many of the specific benefits of marriage than we see for same-sex marriage. Thus, a 2004 Newsweek poll conducted by Princeton Survey Research Associates found that 60 percent of the public supported both inheritance rights and health insurance benefits for gay spouses, and 55 percent supported Social Security benefits for gay spouses, while the exit poll found only 25 percent in support of same-sex marriage.

The seeming contradiction between support for the tangible benefits of marriage and opposition to calling a same-sex relationship a marriage is best explained by the symbolic meanings of marriage. While some of the opposition is simply a reaction to the demand that the longstanding definition of a word be reconsidered, much more is at stake to the opponents of same-sex marriage. The concept of marriage is deeply entwined with people's identities as men and women and as heterosexuals. The status of being married is associated with being honorable and respected, two cherished social values. In the view of opponents, expanding access to the status of being married to same-sex couples may threaten to diminish the value of that status, just as it may threaten to subvert traditional sex and gender roles. As late as 2002, 55 percent of Americans thought that homosexuality was always wrong. While that represents a substantial drop from the 73 percent who said so in 1973, it remains a majority.

The lesson to be drawn from these findings by the gay and lesbian rights movement is a difficult one. To win real recognition of same-sex partnerships, it may be necessary to advocate only for civil unions and the practical benefits—insurance, hospital visitation, taxation, and the like—associated with them. For at least another generation, it is unlikely that gay marriage will pass muster in the court of public opinion. This is a hard pill to swallow for LGBTs and their supporters, as so many lesbians and gays are still exhilarated by the images of gay marriage broadcast last year from San Francisco and Boston—and exhausted from the state-by-state battles they have fought against the constitutional amendments.

How do we reconcile the immense popularity of Queer Eye and Will and Grace—or even the daytime talk-show hit Ellen—with the continuing controversies surrounding the rights of gay people? Perhaps the answer is to be found in what is shown and what is not shown on these programs. The character Will Truman virtually never encounters discrimination as a gay man, and he rarely, if ever, uses his skills as a lawyer to protect the rights of gay people. Everyone knows that if he were ever to marry someone, his true love is Grace. The "Fabulous Five" in Queer Eye are presented as men devoid of any private lives. They may be arbiters of taste and manners, but for all we know, they live alone and have no sex lives or romances whatsoever. And despite the nearly endless revelations about her personal life that appear in the tabloid press, Ellen DeGeneres says little about her lesbian identity to the audiences of adoring, middle-aged, mostly straight women who watch her show. The gay people who are so exceedingly popular in mass culture are smart and likable—but they raise none of the issues that are at the center of the struggle for gay rights.

So when will lesbians and gays be accepted by the American public as not just entertainers, but equals? All the trends in American public opinion toward the rights of LGBT people are trends of increasing knowledge and affection. In the long run, these trends are likely to result in greater support for substantive equality. The trends are in the direction of justice—but the struggle for justice never is an easy one.

Patrick J. Egan is a doctoral candidate in political science at the University of California, Berkeley. Kenneth Sherrill is a professor of political science at Hunter College, City University of New York.

A Note on Terminology

LGBT is the standard acronym for lesbian, gay, bisexual, and transgender. The wording of exit polls and national surveys varies greatly. Some limit themselves to “gay” or “homosexual”; the National Election Pool (NEP) used “gay, lesbian, or bisexual” as its item to measure self-identification or sexual orientation in the 2004 exit polls. Some Harris Interactive polls use “lesbian, gay, bisexual, or transgender.” When referring to specific surveys, we strive to use their language. In this article, we use in the text such terms as LGBT, the lesbian and gay movement, and gay people interchangeably as best fits a decent sense of prose style, and in the data graphics whatever terms were used in the original wordings of the survey questions shown. We strive to be as accurate and inclusive as possible while remaining readable.