how think tanks could buy themselves a study

Generally, when researchers studied children in same-sex households and measured how well they fared in terms of education, criminal records, and income, then compared them to heterosexual households with the same criteria, they found that as long as the kids get love, support, and attention, they tend to grow up happier and go on to do generally well in life. Obviously, bullying and discrimination are a bigger problem for kids who grow up with same-sex parents, but they usually bounce back from such episodes and in some communities that’s not even an issue since more parents and adults in charge have a life-and-let-live policy towards gay or other nontraditional couples with children.

These findings mean that when a study purporting that children of gay parents fare worse in every possible life outcome, from being subject to more sexual abuse, to ending up requiring food stamps, to considering suicide more, it better have really, really good evidence behind it. In the light of a recent much-covered retraction of a paper that justified therapy for homosexuals, you’d think that anyone studying gay couples wouldn’t simply conduct the study his donors wanted. But he did just that.

You see, the problem is not in the fact that Mark Regnerus’ study was contrary to current scientific consensus because science corrects itself based on new data and were he to have good data, the consensus has to be reconsidered in light of new evidence. And it’s not even in the fact that the study was funded by two think tanks which promote social conservatism as the solution to all our problems, and one of which hosts a forum that questions the separation of church and state.

Good scientists follow the data rather than the donors, which is why a study funded by the Koch brothers and trumpeted by global warming denialists as the imminent end to the current scientific view on global warming, ended up agreeing with the prevailing consensus and was ultimately denounced by its backers, who up until the last week or so before publication were praising it as an exemplary undertaking in climate science.

No, the problem with Regnerus’ study is that it was just weird and has all the obvious traits of cherry-picking data to suit his goal, repackaging the results of previous studies of children in broken homes as a survey of kids from same-sex households. He essentially did the equivalent of studying the likelihood of having your car getting damaged on any given date based almost entirely on data he collected from visiting body shops until he felt he had enough data points to make his conclusions.

After hiring a survey company to contact some 15,000 people over several months, Regnerus got a sample of only 255 subjects who said that one of their biological parents had a same-sex relationship. Which is odd. I’m pretty sure that if you’re studying how well children fare after being raised in same-sex households, you would build your sample out of those who live in committed same-sex relationships rather than people who left their partners for someone of the same sex for an unspecified period of time. And what’s more bizarre is that we’re also told that 58% of the respondents had a biological mother in a same-sex relationship leave and this figure is said to show higher than average instability in the family, but what average we’re talking about is not specified or cited.

What about instability in heterosexual households in which infidelity caused a split? Regnerus spent no time trying to control for such situations in heterosexual families and simply barreled ahead, calling many previous studies obviously unreliable because they don’t represent “how gay and lesbian parents really are,” apparently meaning that because in the sample he collected parents who had a same-sex relationship even once are now officially gay, and therefore, this is an accurate picture of how same-sex couples function.

Obviously the respondents came from broken homes and we know that in such situations, kids are very often neglected, which makes them more prone to do badly in school, get worse jobs, feel depressed, and expose them to sexual predators or inappropriate behavior by their parents. Less than 1% of the respondents had an established same-sex couple as parents and were planned either through IVF or adoption, so the sample of people studied to figure out what effect same-sex parenting have on kids are barely a blip on Regnerus’ radar and he justifies his omission of them by declaring that he captured real same-sex households rather than an idealized convenience sample.

But in reality, he didn’t separate bisexual experimentation from a committed gay relationship, or consider that his sample showed broken homes in which one parent decided to pursue a same-sex relationship once rather than leave for a heterosexual one. We know kids raised in a broken home without constant contact and support from their families are prone to all the negative outcomes that the study found and that’s not what we’re studying. We don’t use broken homes in which both parents had strictly heterosexual relationships as the benchmark for straight parenting. Why would we use broken homes where a homosexual relationship occurred at least once as a benchmark to evaluate gay parenting?

Oh, and one more thing, the sample of the study considered only people between 18 and 39, which would very easily miss a period in which more and more gay couples were settling down and planning families, omitting many functioning gay relationships from consideration. Odd, huh? Maybe we should consider Regnerus’ very generous donors, the Witherspoon Institute and the Bradley Foundation, who gave him $785,000 for his work as you can see on page 5 of his CV.

In fact, a cursory glance by The New Republic finds that Regnerus has a penchant for espousing the kinds of views that you’d hear from Focus On The Family or Heritage, views such as declaring that consensual premarital sex between adults is a male’s way of avoiding a proper marriage to a woman he finds attractive, and that women better net themselves a husband quickly or face spinsterhood in their 30s and 40s. So why wouldn’t two socially conservative think tanks read a few of his papers, and offer to give him some money to study life outcomes of children in gay couples? And they were extremely generous in their donations, especially Witherspoon, which gave $640,000 towards the venture.

Curious minds really want to know how that money was spent since his surveying would cost $20,000 to $25,000, tops, and his sample could be easily analyzed in Excel by one person. A budget of $50,000 would more than cover this study even if he bought an equipped a new office especially for it. Are these grants for more than one study?

Here’s what irks me. If scientists take money from private companies or organizations, it’s not proof of bias in their studies by default because it all comes down to data. The issue is that the data and conclusions have to make sense. Were a scientist accept a million dollar grant from say, Sugar Inc., and publish a study that a lot of sugar in one’s diet does not in fact cause diabetes, red alarms should be going off in your head because a lot of existing work shows that it very much does. Likewise, when Big Pharma pays doctors to sign off on very, very laudatory papers about their latest blockbuster drug, we should thoroughly question any off-label use this paper would recommend.

So when a social scientist takes money from wealthy conservative groups and then espouses their views in a study in which data screams of cherry-picking and spends a lot of time denouncing the existing body of literature as being just plain wrong while packaging those who grew up in broken homes as the end products of a “representative sample” of gay parenting, we should cry foul. Right wing groups saw literature that found few differences between gay and straight parents and they needed a study that supported their talking point. So they found a sympathetic researcher and bought one. It’s just that simple.

See: Regnerus, M. (2012). How different are the adult children of parents who have same-sex relationships? Findings from the NFSS Social Science Research, 41 (4), 752–770 DOI: 10.1016/j.ssresearch.2012.03.009