September 16, 2014
Last week, the 7th Circuit Court of Appeals upheld lower court rulings striking down anti-marriage laws in Indiana and Wisconsin. Even those of us who aren’t legal scholars can find good reading in Judge Richard Posner’s written opinion, which skewered the states’ arguments against marriage equality.
As a social scientist, I was pleased that his legal analysis was informed by data from social and behavioral research. And I was gratified that he referenced some of my own work.
Early in his 40-page decision, Judge Posner wrote,
“We begin our detailed analysis of whether prohibiting same-sex marriage denies equal protection of the laws by noting that Indiana and Wisconsin … are discriminating against homosexuals by denying them a right that these states grant to heterosexuals, namely the right to marry an unmarried adult of their choice. And there is little doubt that sexual orientation, the ground of the discrimination, is an immutable (and probably an innate, in the sense of in-born) characteristic rather than a choice. Wisely, neither Indiana nor Wisconsin argues otherwise.” (p. 9, my emphasis)
The evidence he cited in support of this assertion included materials from the American Psychological Association and a paper on which I was the lead author, describing findings from a survey I conducted with a nationally representative sample of lesbian, gay, and bisexual adults.
This blog post is about the research and the context in which I conducted it.
* * * * *
Early research on sexual minority subcultures in the United States tended to focus on gay men. And the researchers often reported that most gay men felt they hadn’t chosen their sexual orientation. For example, in his 1951 book, The Homosexual in America: A Subjective Approach, sociologist Edward Sagarin (writing under the pseudonym Donald Webster Cory) wrote:
“This does not mean that sexual inversion [homosexuality] is voluntary, and that one need only exercise good judgment and will-power in order to overcome it or to choose some other pathway. Not at all. It is entirely involuntary and beyond control, because one did not choose to want to be homosexual.” (p. 183)
And psychologist Evelyn Hooker, in her 1965 paper, Male Homosexuals and Their “Worlds” (in Judd Marmor’s edited book, Sexual Inversion: The Multiple Roots of Homosexuality), reported from her ethnographic observations of gay male communities:
“One of the important features of homosexual subcultures is the pattern of beliefs or the justification system. Central to it is the explanation of why they are homosexuals, involving the question of choice. The majority believe either that they were born as homosexuals or that familial factors operating very early in their lives determined the outcome. In any case, homosexuality is believed to be a fate over which they have no control and in which they have no choice.” (p. 102)
In recent years, religious conservatives have strongly disputed this view, and the argument that homosexuality is a sinful choice has achieved considerable prominence in their public rhetoric. In the 1990s, they mounted media campaigns promoting the notion that people can and should stop being gay. The director of one of these ex-gay campaigns told the New York Times that its goal was to strike at the assumption that homosexuality was immutable and that gay people therefore need protection under anti-discrimination laws.
Not surprisingly, public opinion reflects this dimension of the culture wars. Heterosexuals’ attitudes toward lesbians and gay men are reliably correlated with their beliefs about choice. Antigay heterosexuals are likely to assert that homosexuality is a choice, whereas unprejudiced heterosexuals are likely to believe that sexual orientation is inborn or otherwise not chosen. (As discussed below, the question of whether heterosexuals choose their orientation is rarely asked.)
* * * * *
In the 1990s, I was surprised to discover that, despite all the debate and heated rhetoric, relatively little empirical research had directly examined how people perceive their own sexual orientation.
Indirect evidence for a lack of choice was available. For example, most participants in the Kinsey studies of the 1940s and 1950s reported they had experienced sexual attraction only to one sex (men or women) throughout their entire lives; but the Kinsey team did not ask directly about perceptions of choice.
Illuminating research was conducted by sociologist Vera Whisman, who set out to study lesbians and gay men who said they had chosen their sexual orientation. However, as she reported in her book, Queer By Choice, most of her sample did not experience their patterns of sexual attractions as a choice. Those who were “queer by choice” were typically referring to choosing their sexual behaviors and the labels and identities they adopted for themselves.
Otherwise, anecdotal and autobiographical accounts were available and a few studies reported relevant questionnaire data from small samples. But as best I could tell, no large-scale studies had asked people whether they perceived their own sexual orientation (whether hetero-, homo-, or bisexual) as a choice.
* * * * *
This lack of data prompted me to begin asking about choice in my own research.
Based on the available evidence, I expected to find that many – probably most – gay men didn’t perceive their sexual orientation to be a choice.
For women, however, I thought the pattern might be different. Many feminists argued that lesbianism is a choice women can (and should) make for themselves. And in some early studies, gay men tended to report having been aware of their homosexuality at an earlier age than lesbians, which might be evidence of a gender difference in the experience of choice.
These and other patterns led me to tentatively hypothesize that lesbians would be more likely than gay men to report they experienced some degree of choice about their sexual orientation.
In an exploratory study during the 1990s with a relatively small community sample that included 125 gay and lesbian adults, these hypotheses were supported. My colleagues and I found that most of the gay men (80%) said they had no choice at all about their sexual orientation. The proportion of lesbians who said they had no choice was smaller, but still a majority (62%).
While these findings were interesting, the sample was small. I subsequently decided to ask a similar question in two survey studies with larger and more representative samples that also included enough bisexual women and men to permit meaningful analyses of their responses.
In the first of those surveys, we collected questionnaire data from 2,259 gay, lesbian, and bisexual adults in the greater Sacramento area. One questionnaire item was, “How much choice do you feel that you had about being lesbian/bisexual?” [for men the wording was "gay]/bisexual”]. The 5 response options were “no choice at all,” “very little choice,” “some choice,” “a fair amount of choice,” and “a great deal of choice.”
The results weren’t dramatically different from those we obtained in the pilot study: 87% of the gay men reported they experienced “no choice at all” or “very little choice” about their sexual orientation. Once again, women perceived having more choice than men. Even so, most lesbians (nearly 70%) reported having little or no choice.
It is perhaps not surprising that bisexuals reported feeling they had more choice about their sexual orientation. Nevertheless, nearly 59% of bisexual men and 45% of bisexual women said they experienced little or no choice. Another 15% and 20%, respectively, said they had only “some choice.”
This study’s sample was large but it wasn’t a probability sample, i.e., one that is representative of the population at large. We had recruited the participants mainly through Northern California lesbian, gay, and bisexual community organizations and at community events, most of them in the Sacramento area. People who weren’t active in the community or weren’t open about their sexual orientation were probably underrepresented.
I subsequently had the opportunity to assess how well these findings fit the population as a whole when I surveyed a nationally representative sample of self-identified lesbian, gay, and bisexual adults. We asked them “How much choice do you feel you had about being lesbian?” [Or gay or queer or bisexual or homosexual, depending on the term they had previously said they preferred for describing themselves.] Four response options were available: “no choice at all,” “a small amount of choice,” “a fair amount of choice,” and “a great deal of choice.”
The responses of gay men and lesbians were strikingly similar to those we obtained from the Sacramento-area community sample: 88% of the gay men reported “no choice at all” about being gay, with another 6.9% saying they experienced “a small amount of choice.” Only 5% reported they experienced “a fair amount” or “a great deal” of choice. Among lesbians, 68.4% reported no choice, and another 15.2% reported experiencing a small amount of choice; only 16% experienced a fair amount or a great deal of choice.
Thus, 95% of gay men and 84% of lesbians reported experiencing little or no choice about their sexual orientation. This is the finding Judge Posner cited last week in his opinion.
In contrast to the community study, a majority of bisexuals in the national sample reported having little or no choice about their sexual orientation, although they were less likely than gay men and lesbians to say they experienced no choice at all. Among bisexual men, 38.3% said they experienced no choice, and another 22.4% experienced a small amount of choice, a total of 60.7%. Among bisexual women, the numbers were 40.6% and 15.2%, respectively, a total of 55.8%.
None of these surveys explicitly defined the term choice, so we don’t know whether respondents interpreted it as referring to their pattern of attractions, their sexual behaviors, their identity, or some other facet of sexual orientation. Based on Vera Whisman’s research, cited above, it seems likely that most were referring to the amount of choice they experience in their sexual attractions and desires.
* * * * *
What about heterosexuals? Do they perceive their sexual orientation as a choice?
To the best of my knowledge, no published research based on a probability sample of heterosexual adults reports data that directly answer this question. I intended to ask it in a national survey I conducted in the 1990s, but was dissuaded from doing so by other members of my research team. They convinced me the question would create problems during data collection because most heterosexuals simply wouldn’t know how to answer it.
This asymmetry in who can answer the choice question can be understood as a reflection of sexual stigma. One manifestation of stigma is the widespread assumption that heterosexuality is normal and unproblematic. Few heterosexuals are ever asked what made them straight, and most have probably never thought about the origin of their own attractions to the other sex.
Homosexuality, by contrast, is viewed as problematic. Nonheterosexuals are routinely asked what made them “that way” and, in the course of coming out, they often ask themselves this question. Even when a scientific study evenhandedly examines the origins of all sexual orientations, its subject matter is typically characterized as what causes people to be gay or bisexual.
In this context of stigma, it is perhaps not surprising that I encountered some raised eyebrows when I initially shared my findings about perceptions of choice with other researchers – not so much because of the numbers, but simply because I had asked the question.
Some assumed that documenting how people perceive their sexual orientation would be the basis for arguing that gay, lesbian, and bisexual people shouldn’t be persecuted because “it’s not their fault” – they never chose to be “that way.” This argument is perceived (often correctly) as implicitly suggesting that (a) being lesbian, gay, or bisexual is a defect, and (b) if people did choose to be anything other than heterosexual, they would deserve to be discriminated against.
* * * * *
But although Judge Posner’s opinion takes up the question of choice – as did Judge Vaughn Walker, who cited the same research in his decision overturning California’s Proposition 8 – he doesn’t treat homosexuality as a defect. Nor does he suggest that gay, lesbian, and bisexual people would deserve to be persecuted if they freely chose their sexual orientation.
However, Judge Posner recognizes that lesbian, gay, and bisexual people constitute an identifiable minority group defined by an immutable characteristic that is irrelevant to a person’s ability to contribute to society. Consequently, any attempt by the state to discriminate against them must serve some important government objective.
And, as he concluded, the rationale offered by Wisconsin and Indiana for their laws denying marriage rights to same-sex couples, “is so full of holes that it cannot be taken seriously…. The discrimination against same-sex couples is irrational, and therefore unconstitutional…” (pp. 7-8).
* * * * *
Here are the bibliographic sources for my studies, described above.
Herek, G. M., Cogan, J. C., Gillis, J. R., & Glunt, E. K. (1998). Correlates of internalized homophobia in a community sample of lesbians and gay men. Journal of the Gay and Lesbian Medical Association, 2(1), 17-25.
Herek, G. M., Gillis, J. R., & Cogan, J. C. (2009). Internalized stigma among sexual minority adults: Insights from a social psychological perspective. Journal of Counseling Psychology, 56, 32-43.
Herek, G. M., Norton, A. T., Allen, T. J., & Sims, C. L. (2010). Demographic, psychological, and social characteristics of self-identified lesbian, gay, and bisexual adults in a U.S. probability sample. Sexuality Research and Social Policy, 7, 176-200.
A brief introduction to sampling terminology and methods is available on my website.
August 25, 2014
In 1957, Dr. Evelyn Hooker’s groundbreaking study documented that, despite the conventional psychiatric wisdom of the day, gay men were not inherently maladjusted. More studies followed that similarly failed to find differences in psychological functioning between heterosexuals and nonheterosexuals.
Eventually, this body of research provided the scientific foundation for the American Psychiatric Association’s 1973 decision to remove homosexuality as a diagnosis from the Diagnostic and Statistical Manual of Mental Disorders, and the American Psychological Association’s strong endorsement of that declassification.
With the advantage of hindsight, we can see that debates about Dr. Hooker’s work and that of later researchers — and, more broadly, about the status of homosexuality as a pathology — often conflated questions about homosexuality’s classification as a mental illness with questions about the prevalence of psychological disorders in a particular population. It was inappropriately assumed that if lesbian, gay, or bisexual people had higher rates of psychopathology or psychological distress than heterosexuals, homosexuality itself must be an illness.
We now recognize that sexual stigma in its many forms is a significant stressor that can affect an individual’s physical and mental health. Thus, it is not surprising that large-scale studies of the US population have revealed that, while most lesbian, gay, and bisexual people are functioning well, some are not. And, as a 2011 report by the Institute of Medicine documented, a substantial array of health disparities exist between the population at large and sexual and gender minorities.
Against this backdrop, newly released data from Gallup reveal that US adults who identify as lesbian, gay, bisexual, or transgender (LGBT) report lower levels of well-being than their non-LGBT counterparts. Comparing the self-reports of 2,964 LGBT research participants with those of 81,134 other respondents, and controlling statistically for relevant demographic variables, Dr. Gary Gates found that the latter group reported less well-being in all five areas covered by the index.
The disparities were especially pronounced among women respondents. Sexual and gender minority women scored substantially lower than other women on measures of financial, physical, social, and community well-being, as well as a measure of having a sense of purpose in life. Among men, disparities were observed for financial and social well-being.
The initial report, which is available on the Gallup website, doesn’t separate the well-being scores of lesbian/gay, bisexual, and transgender respondents. Comparing these groups will be important insofar as past research has revealed important differences among them. (From the perspective of scientific research, a problem with combining the groups under the “LGBT” initialism is that it tends to obscure these differences.)
While reading the tables in the report it’s also important to keep in mind that, because the sample sizes are so large, relatively small differences between groups (i.e., 1 or 2 percentage points) can be statistically significant without having much practical importance. But the differences highlighted by Dr. Gates are generally larger than this.
As Dr. Gates concluded,
“These disparities associated with sexual orientation and gender identity highlight the ongoing need for the inclusion of sexual orientation and gender identity measures in data collection focused on health and socio-economic outcomes. Availability of better data that identify the LGBT population will help researchers, healthcare policymakers, and healthcare providers craft better strategies to understand and prevent well-being disparities associated with sexual orientation and gender identity.”
June 26, 2009
President Obama is under increasing pressure to begin dismantling the military’s “Don’t Ask, Don’t Tell” policy.
Last month the Palm Center at the University of California, Santa Barbara, released a new report showing how the President can legally suspend discharges of gay and lesbian personnel in advance of congressional action to permanently repeal DADT. (Full disclosure: I’m a coauthor of that report; my main contribution was to the section on how to effectively implement a new, nondiscriminatory policy.)
Earlier this week, 77 members of Congress sent an open letter to the President, urging him to suspend investigations and discharges of service members in the Armed Forces because of their sexual orientation. “By taking leadership of the important issue of Don’t Ask, Don’t Tell,” said Rep. Alcee Hastings (D-FL), “President Obama would allow openly gay and lesbian service members to continue serving their country and send a clear signal to Congress to initiate the legislative repeal process.”
And Wednesday, the Center for American Progress issued a five-step plan for repealing DADT that begins with an executive order suspending discharges.
What would be the political costs for the President and Congress of eliminating DADT?
In terms of public approval, not much. In fact, repealing the policy might score points with voters. Poll data indicate that opposition to DADT has been steadily increasing in the years since it was first enacted, and that Americans now solidly support getting rid of it.
Public Opinion About DADT: Then and Now
Just before the 1992 Presidential election, by a margin of 54% to 40%, respondents to a national Harris survey supported “allowing gays and lesbians to serve in the military.” But the following year, as public debate about military policy took center stage, most polls showed public opposition to gay service members. (Also in 1993, pollsters began asking about service by openly gay and lesbian personnel.)
By 1994, however, after DADT had been enacted, public opinion again began to shift and a growing majority wanted to allow openly gay people to serve. As shown in the graph, this trend has continued to the present day. In the most recent poll – conducted last month by Gallup – 69% favored allowing openly gay people to serve. Only 26% were opposed.
This graph only includes polls that asked “Do you favor or oppose allowing openly gay men and lesbian women to serve in the military?” (or used similar wording). Other surveys that asked the question in different ways have obtained similar results.
In some surveys over the years, for example, the question about military service has been posed in terms of equal employment opportunities. With the question framed this way, the public has been even more likely to endorse military service by gay and lesbian personnel. As early as 1977, 51% of Gallup respondents said gay people “should be hired” for the armed forces, and that majority grew steadily over time. By 2005 (the last instance I could find when this question was posed in a national poll), 79% favored having the armed forces “hire” gay men and lesbians, while only 19% opposed it.
Other surveys have asked respondents to select from among three options: (a) allowing gay men and lesbians to serve openly, (b) allowing them to serve but without openly acknowledging their sexual orientation (the DADT approach), or (c) banning them entirely from military service. Over the past 15 years, relatively few respondents have advocated completely banning gay people from the military – never more than roughly 1 in 5. In July, 2007 (the most recent poll that asked this version of the question), only 15% said they wanted a total ban on gay personnel. In that same poll, 36% endorsed the DADT approach, whereas a plurality of 46% favored allowing gay people to serve openly. Compared to a similar poll in 2000, the 2007 results represented an increase of 6 points in favor of openly gay personnel, and a decrease of 7 points in support of a total ban.
Thus, regardless of how the question is asked, the long-term trend has been toward increasing support for openly gay service members.
The Unit Cohesion Argument: The Public Isn’t Buying It
Americans also have grown skeptical of the main rationale that has been offered for retaining DADT. In a 1993 Gallup/CNN/USA Today poll conducted shortly after Bill Clinton’s inauguration, 51% agreed that “admitting gays to the military would undermine discipline and morale” (46% disagreed).
But in a Quinnipiac University national poll this April, 58% disagreed that “allowing openly gay men and women to serve in the military would be divisive for the troops and hurt their ability to fight effectively.” In that same poll, a solid 56% favored repealing DADT whereas only 37% said it shouldn’t be repealed.
Support For A New Policy Is Broad-Based
A striking feature of the May 2009 Gallup poll is the breadth of support it reveals for a new military policy. Not surprisingly, self-described liberals are almost unanimous in supporting a nondiscriminatory policy (86%). But a solid majority of conservatives — 58% — also support openly gay and lesbian service members, as do 58% of Republicans and 60% of those who attend religious services weekly. Support has increased substantially in these groups since Gallup asked the same question in 2004 — by 6 points among Republicans, 11 points among regular churchgoers, and 12 points among conservatives.
Support is even higher among Democrats, Independents, and the less religious. And openly gay service members have solid support among both men and women, in all regions of the US (including 57% in the South), and in all age groups (even among those who are over 65, of whom 60% support a new policy).
The Bottom Line
This is a policy area in which the public is ahead of Congress and the President. There will certainly be an outcry from some far Right groups when President Obama suspends DADT and Congress overturns it permanently. In contrast to 1993, however, there appears to be a solid base of public support for a new policy that allows lesbian and gay Americans to serve their country without having to lie about who they are.
* * * * *
A slightly different version of this post appears on the Palm Center’s blog, Blueprints.
June 16, 2009
Late last month, Gallup released findings from a new poll demonstrating that opposition to marriage equality is higher among American adults who say they don’t know anyone who is lesbian or gay.
The survey, which was conducted earlier in May, found that Americans oppose legalizing marriage between same-sex couples by 57% to 40% . That margin hasn’t changed notably since a previous Gallup poll about a year ago.
When the May sample was split into those who said they have a gay or lesbian friend, relative, or coworker (58% of the sample) and those who didn’t (40%), the differences in marriage attitudes were striking.
The latter group registered overwhelming opposition to marriage equality — 72% opposed it whereas only 27% favored it. Within this group, 63% said legalizing marriage for same-sex couples will change society for the worse, compared to six percent who said it will change society for the better. 30% believed it won’t have any effect on society.
By contrast, respondents reporting personal contact with a gay man or lesbian were almost evenly split — 49% supported marriage equality and 47% opposed it. They were also divided over whether marriage equality will change society for the worse (39% believed it will) or will have no effect (40% believed this). Only about one in five said it will change society for the better, but that percentage was more than three times higher than the comparable figure for respondents without a gay or lesbian friend, relative, or coworker.
Consistent with past research, the poll found that attitudes toward marriage equality are linked with a person’s political ideology, and that liberal respondents were more likely than their conservative counterparts to personally know gay people. But Gallup found that the correlation between personal contact and opinions about marriage remained significant, even when political ideology was statistically controlled.
But Why Only 49%?
The Gallup report prominently characterized the survey as showing that “Opposition to gay marriage [is] higher among those who do not know someone who is gay/lesbian.”
But we might well ask why there wasn’t greater support for marriage equality among poll respondents with gay or lesbian family or acquaintances. Why did only about half of that group support marriage rights? After all, research conducted over the past two decades has consistently shown that heterosexuals are less prejudiced against gay people if they know someone who is gay, and such prejudice is closely associated with opposition to marriage equality. (Data are lacking on how knowing a bisexual man or woman affects sexual prejudice among heterosexuals, but there’s reason to believe that the pattern is similar.)
My own reading of the research literature suggests that the strength of the correlation between prejudice and mere contact has diminished in recent years. A decade ago, knowing whether a heterosexual had a gay or lesbian friend or relative provided a very good indication of that person’s attitudes toward gay people in general. Today, personal contact remains a good predictor of prejudice, but it’s not as reliable as it once was.
I believe this diminution of the predictive power of mere contact may provide insight into what it is about contact that links it to nonprejudiced attitudes. My hypothesis is that the key variable isn’t — and never was — whether heterosexuals simply know a gay man or lesbian. Rather, what’s always been critical is the nature of that relationship. Perhaps the central variable is whether or not heterosexuals have talked with their gay friend or relative about the latter’s experiences and, in the course of those discussions, developed a better understanding of and more empathy for the situation of sexual minorities.
My hunch is that in the past, when most gay men and lesbians were highly selective about coming out, it was sufficient for researchers to simply ask heterosexual survey respondents whether they knew gay men or lesbians. If they had a gay friend or relative, more likely than not, they’d found out directly from that individual about her or his sexual orientation. Or, subsequent to finding out through some other means, they talked about it with her or him.
Today, by contrast, gay men and lesbians are more publicly visible. Many more heterosexuals probably have the experience of knowing that a relative, friend, or (especially) a coworker or neighbor is gay without ever having discussed it directly with that individual. Thus, knowing the details about a heterosexual person’s contact experiences is more important today than it was a few years (or decades) ago.
This hypothesis is partly supported by data I collected in a 2005 telephone survey with a representative national sample of more than 2100 adults who identified as heterosexual. Along with questions about the nature and extent of their personal relationships with lesbian and gay individuals, respondents were asked a series of questions about their general feelings toward gay men and toward lesbians, their comfort or discomfort around both and, using a standard psychological attitude scale, their general attitudes toward them.
For purposes of analysis, I divided the sample into three groups: (1) those who said they had no gay or lesbian friends, acquaintances, or relatives as far as they knew, (2) those who knew at least one gay or lesbian person but hadn’t ever talked with that individual about being gay, and (3) those who had talked with a gay or lesbian friend or relative about the latter’s experiences as a sexual minority.
Compared to Group 1, Group 2 had more positive feelings, less discomfort, and generally more favorable attitudes toward gay men and lesbians. But Group 3 had significantly more positive views of lesbians and gay men than either Group 1 (those with no personal contact) or Group 2 (those with personal contact but no open discussion).
Combined with other survey findings that I’m still analyzing, these data suggest it often isn’t enough for heterosexuals to simply know that a member of their family or immediate social circle is gay or lesbian. In order for the experience to reduce their sexual prejudice, they also must communicate directly with their friend or relative about what it’s like to be gay.
But although such discussions probably play a key role in reducing sexual prejudice and increasing support for the civil rights of sexual minorities, they can be difficult. Not surprisingly, they don’t occur often enough. In a separate study (which is not yet published), I’ve found that most gay men and lesbians say they are out to their immediate family and close heterosexual friends, but many aren’t out to their extended family, coworkers, or heterosexual acquaintances. And many of those who are out never discuss their experiences with their family or friends.
These findings highlight the importance of assisting gay, lesbian, and bisexual people in having conversations — giving them support and helping them find the best way to talk with their heterosexual friends and family members about their lives and how they’re affected by issues like marriage equality. The Tell 3 Campaign is one strategy for promoting such discussions. If the marriage equality movement is going to succeed in changing public opinion, it will have to devote more resources to Tell 3 and other programs like it.
* * * * *
More information about my 2005 survey can be found in the following chapter:
Herek, G. M. (2009). Sexual stigma and sexual prejudice in the United States: A conceptual framework. In D.A. Hope (Ed.), Contemporary perspectives on lesbian, gay and bisexual identities: The 54th Nebraska Symposium on Motivation (pp. 65-111). New York: Springer.
May 31, 2009
Now that the California Supreme Court has upheld Proposition 8′s constitutionality, some marriage equality supporters are ready to begin collecting signatures for a new ballot measure to overturn it in next year’s election.
Instead, I hope Californians who support marriage rights for same-sex couples will take a deep collective breath and engage in level-headed strategizing about how best to achieve the long-range goal of marriage equality.
There are at least two good reasons not to put an anti-Prop. 8 measure on the 2010 ballot.
First, such an initiative stands a strong chance of losing. Highly respected statewide polls, such as those conducted by Field and the Public Policy Institute of California (PPIC), indicate that support for marriage rights for California same-sex couples hasn’t increased noticeably since November. In a February Field Poll, for example, fewer than half of registered voters said they would support a new ballot measure to legalize same-sex marriage, and about the same percentage would oppose it. Only a 49% plurality said they generally support “California allowing homosexuals to marry members of their own sex and have regular marriage laws apply to them.” And a March PPIC survey found that the state’s likely voters oppose marriage equality by a 49-45% margin.
These numbers don’t bode well for a 2010 ballot campaign to overturn Prop. 8. Just over a year ago, the Field Poll found that more than half of likely voters opposed changing the state constitution to define marriage as between a man and a woman. PPIC surveys similarly revealed a widespread reluctance to enact Prop. 8. Yet that solid majority evaporated during the final months of last fall’s campaign. Launching a new initiative with support from less than half of the electorate is ill advised. And if the next campaign fails, it’s highly unlikely that the necessary resources will be available anytime soon to mount yet another ballot fight.
Second, win or lose, another initiative campaign will exact a substantial psychological toll. Research shows that marriage amendment campaigns have negative mental health effects on the people whose lives they target. A recently published nationwide study, for example, found that during the months leading up to the 2006 November election, psychological distress increased among lesbian, gay, and bisexual adults living in states where an antigay marriage measure was on the ballot, but not among their counterparts living elsewhere. By Election Day, sexual minority residents of the states with antigay ballot measures had, on average, significantly higher levels of stress and more symptoms of depression than their neighbors in other states.
Comparable research on the 2008 election isn’t yet available but the limited data I’ve seen, supplemented by my own observations, lead me to believe that the Proposition 8 campaign had a similar, negative effect on many Californians. Perhaps the psychological fallout of another statewide campaign will be tolerable if Prop. 8 is repealed. But without a strong likelihood of succeeding, it is irresponsible to subject lesbian, gay, and bisexual Californians to another prolonged period of daily attacks on the legitimacy of their relationships and families.
It has become almost a cliché to assert that time is on the side of the marriage equality movement. Younger voters support marriage rights for same-sex couples more strongly than their elders (although the strength of support among young voters shouldn’t be overstated). That view will eventually achieve majority status in California, perhaps even by 2012. But almost certainly not by next year.
I’m not suggesting that marriage equality supporters should sit on their hands. There’s much work to be done to create a solid majority of California voters who feel they have a personal stake in overturning Prop. 8.
For example, heterosexuals who support marriage rights for same-sex couples can become agents of change by making their opinions known to their spouse, family, neighbors, and coworkers.
And it’s critically important for lesbian, gay, and bisexual Californians to speak directly with their straight relatives and friends about their own experiences, to explain how measures like Prop. 8 personally affect them. In the wake of the November election, the American Civil Liberties Union and other groups launched the Tell 3 Campaign to encourage and assist sexual minority adults in telling their stories to the people who love and respect them. Having such conversations is one of the most potent strategies for changing attitudes. Yet, according to my own research, they occur all too infrequently.
Last week’s Supreme Court decision has rightly evoked strong feelings among gay, lesbian, and bisexual Californians and their heterosexual supporters. That emotion can be harnessed to build a successful movement for marriage equality in California. But it shouldn’t push us prematurely into a ballot campaign that poses a significant risk not only of losing, but also of ultimately harming many lesbian, gay, and bisexual Californians.
* * * * *
A briefer version of this essay appeared in the Sacramento Bee on Sunday, May 31, 2009.
November 8, 2008
Was the passage of Prop. 8 always a foregone conclusion, despite poll results throughout the summer and early fall showing most likely voters opposed it?
Or were the major polls correct, and the sentiment of California voters actually shifted in the weeks leading up to Election Day, from opposition to support?
Some Prop. 8 supporters maintained throughout the campaign that survey results consistently understated support for their side because many respondents wished not to appear bigoted to a pollster. They cited a “study” they had conducted with polling data from previous state anti-equality campaigns to support their argument.
And Lorri L. Jean, executive director of the Los Angeles Gay & Lesbian Center, asserted in a September letter to marriage equality supporters that:
There is typically a 7-10 point difference between what people tell the pollster about their views on LGBT rights and how they really vote. In other words 7-10% say they believe in equality but actually vote against us.
The existence of a racial Bradley effect — i.e., a pattern in which the polls’ accuracy is affected by significant numbers of racist Whites lying to pollsters and saying they would vote for a Black candidate — has been widely disputed, and wasn’t evident in polling this year.
But was there a gay Bradley effect in California?
In a September article in The Advocate, political scientist Patrick Egan answered this question in the negative, concluding that voters were largely telling pollsters the truth. “If any such reluctance [to tell the truth] exists with regard to same-sex marriage initiatives,” Egan said, based on his examination of polling data from previous campaigns, “it is small — about two points on average since 1998. In 2006 it was effectively zero.”
Now that the election results are in, we can compare the actual vote tallies to the last Field poll before Election Day. My own reading of the data is that they reveal no evidence that survey respondents said they would vote No when they actually supported the measure.
The final Field Poll, conducted about one week before the election and released on October 31, indicated that 49% of likely voters opposed the measure. The margin of error was +/- 3.3 points, so the poll’s estimate was that opposition in the population of California voters ranged between 46% and 52% at that time.
The official tally so far (which still excludes some absentee and provisional ballots) puts the actual No vote at 48%, well within the range of the Field Poll’s estimate.
The same Field Poll found that 44% of likely voters supported Prop. 8 a week before the election, and another 7% were undecided. Factoring in the margin of error, the poll’s estimate of the actual population proportions ranged as high as 47% for the Yes side, with anywhere from 4-10% undecided.
These numbers fall short of the final Yes tally, but it’s not difficult to construct a scenario whereby they are consistent with the Prop. 8 win on Election Day.
First, if most of the undecideds ultimately voted Yes (the pattern that apparently occurred with antigay Prop. 22 in 2000), the result would have been close to what happened on Tuesday.
Add to this the fact that the No vote was trending downward in the weeks leading up to the election. Polls by Field and the Public Policy Institute of California (PPIC) estimated that opposition was around 55% in September. But it declined to 52% in an October PPIC poll and 49% in the final Field Poll a week later. To the extent that those declining numbers indicated that “soft” No voters were in the process of switching to the Yes side, it would help to account for the election day outcome.
Finally, turnout in this election was unusually high, especially among groups that historically haven’t voted in large numbers. This created unique challenges for pollsters in identifying likely voters. To the extent that the criteria used by the various survey organizations to estimate turnout in advance of Election Day were inaccurate — especially among key voter groups — their figures would have missed the mark.
In the final Field Poll, for example, African Americans were projected to constitute about 6% of likely voters, and a plurality of about 49% said they supported Prop. 8; another 9% (roughly) were undecided. These figures were derived from interviews with a fairly small number of Black respondents, so the margin of error was substantial (perhaps as much as +/- 12 points) and generalizing from them is risky. If we simply take them at face value, they suggest that Blacks’ contributions to the total vote a week before the election was about 3 points on the Yes side, and slightly less on the No side. Taking the margin of error into account, however, Blacks’ support for Prop. 8 could have ranged as high as 60%. And the undecideds could have subsequently added even more to that total — especially if they were persuaded to vote Yes by appeals from the pulpit on the Sunday before Election Day.
Exit polls were consistent with the latter scenario, finding that about 70% of Blacks ultimately voted Yes. Moreover, they constituted 10% of voters — not 6% — making the impact of their opposition considerably stronger.
It’s important to remember that exit polls — like any survey based on a sampling of the population — have a margin of error associated with their estimates. And the margin can be large for relatively small groups. In the case of Blacks’ votes, the exit poll’s error is probably +/- 5 to 6 points, and there remains the bigger question of whether the specific precincts that were sampled yielded an accurate reflection of African Americans statewide.
Nevertheless, it seems safe to assume that Blacks ultimately provided substantial support for the Yes side — perhaps enough to account for the election outcome. Most likely, there are other groups for whom turnout projections were also incorrect and, in combination with the downward trend in No voters and last-minute decisions by undecideds, these factors can probably account for the disparities between pre-election polling and the actual outcome.
Thus, it’s difficult to conclude that significant numbers of Prop. 8 supporters lied to pollsters and said they were planning to vote No. Perhaps some Yes voters disingenuously told researchers they were undecided, but it’s equally plausible that most undecideds truly didn’t make up their minds until late in the campaign, perhaps not until Election Day.
At any rate, the so-called “study” that Prop. 8 supporters posted on their web site in mid-September and promoted to the media as evidence of the “gay Bradley effect” can’t really be taken seriously. Even a cursory examination of the polling data they used reveals several glaring problems:
- They ignored the polls’ margin of error. In more than 40% of the polls cited in the “study,” the discrepancy between the poll estimate and the actual vote was 5% or less. For many statewide polls, this is within the margin of error.
- They only noted undercounts in the anti-equality vote, suggesting that all discrepancies resulted from voters telling pollsters they supported the right of same-sex couples to marry but then voting against marriage equality. But in many of the polls listed in their spreadsheet, the actual vote counts against the anti-gay measures also were higher than the polls’ estimates. How can this be? The answer lies with the undecided poll respondents. They had to make a decision in the voting booth and they tended to favor the winning side — which was anti-equality in all cases except the 2006 Arizona campaign.
- They included polls that were conducted weeks (in some cases, months) before the election. As all pollsters know, surveys are usually more accurate to the extent that they’re completed close to voting time. But the “study” included polls that were published more than a month before election day.
- They were selective in which polls they picked. For example, in the 2004 Arkansas election for Amendment 3, the “study” used an October Zogby poll, which indicated that 65% of respondents supported the amendment. But an Opinion Research Poll released in late October found that 77% of Arkansas voters supported Amendment 3, slightly more than actually voted for it.
In summary, I don’t believe that the findings of the PPIC and Field Polls leading up to the election were wrong. Rather, I suggest we assume that a majority of typical California voters truly were opposed to eliminating the right of same-sex couples to marry throughout the summer, but their numbers began eroding by October. Among actual voters, supporters of Prop. 8 came to outnumber opponents by Election Day, albeit by a surprisingly small margin. (Recall that just 8 years ago Prop. 22 won by more than 20 points.)
Thus, we can use the PPIC and Field Poll data as a tool for better understanding how the various strategies pursued by each side between May and November ultimately affected the outcome of the election.
October 31, 2008
Just in time for Halloween, the latest Field Poll brings some scary news for marriage equality supporters. But the results might also create a well-founded sense of foreboding among those who oppose marriage equality and want to write their views into the California constitution.
The poll indicates the Proposition 8 race has tightened considerably. Support for the constitutional amendment still hasn’t reached the 50% mark, but opposition has dropped to 49%.
Moreover, the 5-point gap between the 44% of likely voters who support Prop. 8 and the 49% who oppose it is now within the poll’s margin of error of +/- 3.3 points. In other words, the true proportion of YES voters in the population could range as high as 47.3% and the true number of NO voters could be as low as 45.7%.
About 7% of likely voters are still undecided.
Ten days ago, I posted an analysis of the Proposition 8 polls conducted by Survey USA, and concluded that they probably undercounted Prop. 8 opponents but gave a more or less accurate reading of the number of the measure’s supporters. A few days later, data from a new Public Policy Institute of California (PPIC) poll supported that hypothesis. It showed that the race had tightened, but opponents still led supporters, 52% to 44%.
In statistical terms, the newest numbers from the Field Poll aren’t significantly different from the October PPIC findings — that is, when the respective polls’ margins of error are considered, their estimates of the proportion of YES voters in the California population overlap, and their estimates of the NO voters match.
When we look at the two polls in tandem, however, it appears that the NO position has eroded somewhat and now hovers around 50%. And the YES side appears to have increased its numbers substantially during the past month.
The graph below shows the trends in voters’ attitudes toward Prop. 8 since last May, using data from four statewide Field Polls (filled circles) and two PPIC polls (filled squares). In combination, the polls indicate the race was fairly stable until September, when both sides inaugurated their media campaigns (and when voter interest in the election traditionally begins to rise). At that point, support for Prop. 8 began to increase steadily (from 38% in September to 44% in October), while opposition declined (from a high of 55% to the latest figure of 49%).
Keep in mind that the data in the graph come from two different survey organizations, that somewhat different question wordings were used in different polls, and that all of these figures have a margin of error (ranging from about 2-4 points) associated with them.
Nevertheless, the new Field Poll data indicate that the race is much closer now than at any time since the California Supreme Court issued its historic ruling last May. Prop. 8 opponents have lost the comfortable lead they held throughout the summer.
On the other hand, Prop. 8 proponents have yet to register support from a majority of the public, which the amendment needs in order to pass.
Who’s Voting For and Against Prop. 8?
Compared to the September Field Poll, support for Prop. 8 appears to have increased in the Central Valley (where it’s now leading by a margin of 53-42%) , but also in Los Angeles, where it’s now losing by a plurality of only 48.5% to 43% — within the poll’s margin of error. By contrast, the September Field Poll found it to be losing in Los Angeles by nearly 25 points.
Support for Prop. 8 has increased to 75% among Republicans, but opposition among Democrats has dropped to 65%. Democrats outnumber Republicans by a margin of 44% to 34%, so this split still favors the NO side in terms of the raw vote count. But the fact that Democrats comprise about one fourth of the YES voters is a cause for concern among Prop. 8 opponents.
On the bright side for opponents, Independents and members of other parties still oppose Prop. 8 by a 2-to-1 margin, and political moderates oppose it 50.5% to 40%.
Consistent with previous surveys, women oppose Prop. 8 by a margin of 50.5% to 42%. However, men are now evenly split, 47% NO to 46% YES. The Proposition trails by margins of about 10 points among all age groups except voters who are 65 and older, among whom it is winning, 62% to 32%.
Given recent speculation about racial and ethnic differences in voting patterns, it is somewhat surprising that Latinos don’t differ substantially from non-Hispanic Whites — pluralities of both groups oppose Prop. 8. African Americans support the amendment by a plurality, while Asian Americans oppose it by a slight majority. However, corresponding to their representation among registered voters, the absolute numbers of African Americans and Asian Americans in the sample are fairly small and, in combination, they appear to cancel each other out.
A plurality of Catholics opposes Prop. 8, while Protestants support it by a margin of 60% to 33%. Within the latter group, three-fourths of self-described born-again Christians support it (59% of non-born-agains oppose it). The proposition is opposed by strong majorities of the non-religious (77%) and those affiliated with “other” religions (64%).
Most poll respondents (78%) said they personally know or work with someone who is gay or lesbian, and 51% of those respondents say they will vote NO. By contrast, 50% of respondents who say they don’t know gay people personally are planning to vote YES. Based on my own research, I speculate that opposition to the amendment is strongest among respondents who have actually discussed Prop. 8 (or other issues related to sexual orientation) with a gay or lesbian friend or family member. However, the Field Poll didn’t include a question about such conversations.
The survey also indicates that opponents of Prop. 8 are more likely than supporters to be planning to vote at their local precinct on Tuesday. Both sides are about equally represented among absentee voters, but more supporters than opponents say they’ve already mailed in their ballot. If that pattern persists, Prop. 8 opponents will begin Election Day with a deficit in absentee ballots that will have to be balanced by votes cast in the precinct polling places.
Related to this point, Obama supporters — who constitute 55% of California voters, according to the Field Poll — are disproportionately likely to vote NO on Prop. 8 (73% versus 21% who will vote YES). McCain voters are less numerous (33% of California voters) and they support Prop. 8 by an 84% to 13% margin. Thus, the Obama campaign’s get-out-the-vote effort may help to defeat Prop. 8. But Obama supporters who oppose Prop. 8 account for only about 40% of likely voters (73% of 55% = 40%).
Which Arguments Are Persuasive?
In addition to estimating the proportions of YES and NO voters, the latest Field Poll asked respondents whether they agreed or disagreed with various arguments for and against Prop. 8. Roughly half of the respondents were asked about each statement.
When interpreting these data, it’s useful to keep in mind that survey researchers often observe an “acquiescence bias.” When presented with assertions about an issue, poll respondents — especially those who don’t have strong opinions or are undecided — are more likely to agree than to disagree. There are many explanations for this pattern, including the generally plausible nature of most agree-disagree statements presented in surveys, the amount of mental effort it takes to marshal counter-arguments to them, and most respondents’ desire to be polite and agreeable.
Thus, it’s not unusual to find some respondents who agree with seemingly contradictory statements in the same survey.
Illustrating this pattern, 65% of the sample agreed that “The institution of traditional marriage between a man and a woman is one of the cornerstones of our country’s Judeo-Christian heritage,” which would appear to indicate strong support for the YES side. More than two thirds of undecided respondents agreed with this statement, as did 90% of YES voters and 39% of NO voters. (Recall that undecided voters constitute only 7% of the sample, so percentages from this group aren’t very reliable.)
But majorities also agreed with the following statements that appear to support the NO side:
- 58% agreed that “Matters relating to the definition of marriage should not be written into the constitution.” (Most undecided voters agreed, as did 74% of NO voters and 41% of YES voters.)
- 57.5% agreed that “Domestic partnership laws by themselves do not give gay and lesbian couples the same certainty and security that marriage laws provide.” (Half of undecided voters agreed, as did 72% of NO voters and a 45% plurality of YES voters.)
- 57% agreed that “Extending new rights and legal protections to different peoples and lifestyles, such as gays and lesbians, benefits California and the nation in the long run.” (A plurality of undecided voters agreed, but almost as many were unsure about this statement.)
- 61% agreed that “By eliminating the right of gay and lesbian couples to marry, Prop. 8 denies one class of citizens the right to enjoy the dignity and responsibility of marriage.” (Undecided voters were closely divided on this statement.)
In light of the acquiescence bias, the statements with which most respondents disagreed are especially noteworthy:
- 60% of the sample disagreed that “If Prop. 8 is not approved, the public schools could be required to teach kids that same sex marriage is as acceptable as traditional marriage in California.” (More than two thirds of undecided and NO voters disagreed, as did a 48% plurality of YES voters.)
- 59% of the sample disagreed that “Gay rights leaders in California are moving too fast in their efforts to win new rights and legal protections for gays and lesbians.” (Most undecided voters disagreed, as did 87% of NO voters.)
And some other arguments favoring the YES side were endorsed by only a plurality or a bare majority, with a fair proportion of undecideds:
- 50% agreed that “Prop. 8 restores the institution of traditional marriage between a man and a woman, while not removing any domestic partnership rights that had previously been granted to gay or lesbian couples” (Majorities of undecided and YES voters agreed, as did 23% of NO voters; but many undecideds were unsure about this statement.)
- A 47% plurality of the sample agreed that “Prop. 8 reverses the flawed legal reasoning of activist judges who overturned the state’s previous voter-approved law defining marriage as between a man and a woman.” (While 73% of YES voters agreed, along with 26% of NO voters, most undecideds either disagreed or were unsure.)
But one argument from the NO side also evoked considerable uncertainty:
A 40% plurality agreed that “Followers of the Mormon Church are exerting too much influence on the state’s political process by underwriting an estimated 40 percent of the Yes on Proposition 8′s campaign contributions.” (While 55% of NO voters agreed, about half of undecideds and 29% of YES voters were unsure.)
Overall, these response patterns suggest that likely voters tend to endorse arguments that support their position, but many also accept arguments that appear to contradict their own stance. The NO side apparently has been effective at persuading most California voters that the YES side’s claims about schools and schoolchildren are incorrect. And more arguments in support of a NO vote receive majority support than do the arguments supporting a YES vote.
So why is the race so close?
Old Prejudices Die Hard
It’s noteworthy that this question is even being posed. Back in May, I would have been very surprised to know that Prop. 8 would be trailing in the polls just a few days before the election.
Perhaps instead we should be remarking on the fact that so many voters have proved reluctant to write antigay discrimination into the California constitution. The widespread opposition to Prop. 8, and the fact that proponents of the measure have been so careful not to publicly bash sexual minorities, are signs of a sea change in public attitudes.
Nevertheless, old prejudices die hard. Gay, lesbian, and bisexual people are still stigmatized throughout the United States and in much of California. Powerful groups — including the Mormon Church and Focus on the Family — have dedicated their vast resources to perpetuating sexual stigma. And many heterosexuals with generally enlightened attitudes are still uncomfortable thinking about same-sex relationships.
In my own research, I’ve found that heterosexuals’ opinions about marriage equality are very closely linked to their general attitudes toward gay men and lesbians. Other factors are also important — including religious beliefs and political values — but antigay attitudes are usually the strongest predictor of marriage attitudes.
It would be surprising if antigay prejudice weren’t playing an important role in the California election. Such prejudice — coupled with the barrage of pro-Prop. 8 messages that many Californians have been getting from their religious leaders — may well account for the closeness of the contest, even though so many voters say they agree with key arguments against Prop. 8.
Nevertheless, Prop. 8 still has a good chance of being defeated. But the Field Poll results highlight the importance of swaying voters who are still undecided, and turning out the vote among Prop. 8 opponents.
If you oppose Prop. 8, no matter where you live, my advice is to talk (in person, on the phone, via e-mail) with your California friends, family, roommates, classmates, coworkers, and neighbors this weekend. Urge them to vote NO and, if you’re a California voter, tell them why you’re casting your vote against Prop. 8. The No On 8 website has some tools to assist with e-mail outreach.
And then do everything you can to ensure that they vote by Tuesday evening at 8 pm.
* * * * *
The most recent survey report can be downloaded from the Field Poll website.
« Previous entries Next Page » Next Page »