Slanted Research?

Mark Regnerus’s Gay Parenting Study Starts a Political War

It’s a political lightning rod but unlikely to shift the debate on same-sex parenting, says David Sessions.

Markus Moellenberg / Corbis

A blockbuster new study that claims to upend the conventional wisdom on gay parenting has quickly become a political lightning rod, with LGBT political groups denouncing it and social scientists and bloggers attacking its lead researcher and his methodology.

Conservative pundits are also taking notice of the study, which is based on surveys of thousands of Americans and purports to show that those who have parents in same-sex relationships face negative long-term consequences in employment, relationship stability, and mental health. The pundits suggest it might bolster their opposition to gay marriage, provoking concern from LGBT activists, who warn that it’s an attempt to “disparage lesbian and gay parents” and exactly the sort of slanted research that plays into the hands of the Christian right. They’re right about that, but a closer look at the study’s limited scope, not to mention its obvious methodological weaknesses, suggests it won’t move the debate on gay parenthood very much. And thanks to its unusually large and diverse sample, some social scientists are saying it provides valuable data despite its author’s questionable analysis.

One look at the circumstances of the New Family Structures Study makes clear the reasons for LGBT alarm. The project was led by Mark Regnerus, a sociologist at the University of Texas known for attention-grabbing research that sometimes seems to tack along the narratives of social conservatives seeking to roll back the sexual revolution. A former professor at the evangelical Calvin College, Regnerus wrote a cover story for Christianity Today arguing that Christians should encourage their children to marry young; he also wrote a piece for Slate arguing that the sexual revolution has produced bitter fruit for women. (Both were based on research published in his books.) Regnerus’s same-sex-parenting study was funded, at a price tag of three-quarters of a million dollars—an enormous sum in social science—by two socially conservative groups: the Witherspoon Institute and the Bradley Foundation.

With such a background, the study’s agenda may seem obvious. But leanings of its funders do not necessarily affect its conclusions, and Regnerus’s work is typically taken seriously by mainstream media outlets. The data he was able to collect make this study the kind of study social scientists dream of doing; its large, random national sample earned praise from the three commenters whose remarks appear along with Regnerus’s overview in the July issue of Social Science Research. All three hail the study’s raw data as an advance over previous studies of gay parenting, which used samples that were small and usually nonrandom.

The praise, though, came with qualification: one commenter, David J. Eggebeen of Penn State, wrote that Regnerus’s data are “still far from ideal.” And Regnerus’s conclusions and other possible extrapolations of the data attracted far more criticism from the commenters, as well as other social scientists who have spoken out about the study.

In questionnaires filled out between 1971 and 1994, the study asked 15,000 18- to 49-year-olds if either of their parents had ever been in a same-sex relationship. Respondents who reported a parent in a same-sex relationship were sorted into GF (gay father) or LM (lesbian mother) categories and measured against other respondents’ outcomes, including those in “intact biological parent” families, in personal and social well-being. Very few of the respondents with gay parents lived in a two-gay-parent household; almost all of them came from broken homes of gay parents who had been married to a straight spouse or, more surprisingly, an opposite-sex gay spouse. Not surprisingly, since they came from families that had experienced upheaval, the children with gay parents were more likely to be unemployed, more likely to cheat on their spouse or partner, and more likely to be in psychological treatment for depression or anxiety.

That’s the study’s most obvious flaw: it doesn’t compare intact, two-parent gay families with similar two-parent straight families. On the contrary, the way Regnerus categorized the children of gay families ensures that they will be wildly incongruous with the straight families. Social scientists and science writers have picked apart other aspects of the study’s methodology, including its basing the categorizations of same-sex parents solely on respondents’ reports that their parent had a gay relationship of some kind and duration during their childhood. Some of the respondents fell into several categories, and to make his data workable, Regnerus had to put some respondents into the “gay father” or “lesbian mother” category when they could have fit into several of his other categories. In his words, he had to “force [categories’] mutual exclusivity for analytic purposes.” As Jim Burroway of the website Box Turtle Bulletin argues, that decision is understandable, but it suggests a weakness of data and ultimately undermines the stated goal of the study.

In his Slate essay, Regnerus preemptively defends himself against these criticisms with a limited interpretation of the results. “We didn’t have as many intact lesbian and gay families as we hoped to evaluate, even though they are the face of much public deliberation about marriage equality,” he writes. The results reflect on a previous generation of gay parenting, before the legalization of gay marriage in some states, he writes. “I’m not claiming that sexual orientation is at fault here, or that I know about kids who are presently being raised by gay or lesbian parents. Their parents may be forging more stable relationships in an era that is more accepting and supportive of gay and lesbian couples.” Regnerus says his study’s political implications are unclear and are not intended to undermine the legal parental rights of anyone.

But despite the scientific disavowal of any political agenda, there are hints that Regnerus sympathizes with a socially conservative view of the research. On top of the way he engineered his data, his promotion of the study approaches the current debate from a particular angle: as pushback to the conventional wisdom that there is no difference between gay and straight families. He implies that political correctness has, in just a decade, sparked a 180-degree turn in scholarship from affirming two heterosexual parents as the best family situation for children to believing that having two gay parents is no different. He may be correct to suspect that politics has had something to do with this shift, but portraying support for gay marriage as the result of a wave of fashionable groupthink is a common trope of right-wing Christian rhetoric.

Regnerus is also explicitly giving scientific credibility to parenthood-related arguments against gay marriage, which have played a large role in the political fight over the issue. Despite claiming that his research is descriptive of a previous generation and offers no causal explanation of the supposed discrepancy between gay and straight parents, Regnerus says outright that his study could reasonably lead to caution against legalizing gay marriage. It could lead to support for gay marriage to stabilize gay families, he writes, but it “may suggest that the household instability that the [study] reveals is just too common among same-sex couples to take the social gamble of spending significant political and economic capital to esteem and support this new (but tiny) family form while Americans continue to flee the stable, two-parent biological married model, the far more common and accomplished workhorse of the American household, and still—according to the data, at least—the safest place for a kid.”

Regnerus’s research doesn’t really “suggest” anything of the kind; he admits that it’s not clear sexual orientation had anything to do with inferior life outcomes of certain respondents in his survey or that his research has anything do with marriage at all. If we go by his own claims about the study, it has no bearing on what is at stake in the political debate over gay marriage and parenthood: whether gay couples who get married and decide to have children are any different from straight ones. But the way Regnerus is selling the research seems designed to bait pundits like Rod Dreher, who gave his post on the study the headline “Traditional Families Are Best for Kids” and touted it as evidence that “deconstructing marriage” will have dangerous consequences.

There is no doubt that, as many gay activists have warned, this kind of marketing will land the study a starring role in the arguments of gay-marriage opponents, who often use the supposed scientific superiority of heterosexual parenting to back their point of view. But responding by attacking Regnerus and exaggerating his claims, as some LGBT activists are doing, hardly seems the best strategy for a movement that has succeeded by telling the truth about the real-life experiences of gay people and their families. Social scientists seem to agree that Regnerus has collected valuable data; that his methodology is suspect does not automatically make him a “right-wing author” who wants to “disparage lesbian and gay parents.” He may be wrong and even working with an agenda, but gay-marriage supporters don’t have to paint Regnerus a hateful ideologue to mount convincing critiques of his work. And where science is concerned, they can remain confident: just as before this research was published, there remains no evidence that children in stable gay families will be worse off than others. One slanted study is unlikely to change that.