Why is William Saletan Apologizing for Slate's Mistake?

Two days ago, I wrote that Slate's editors should be ashamed of having published Mark Regnerus's propagandistic tripe about his "study" comparing how children fare under intact families versus how they fare when their biological parents have a rocky time because one discovers or accepts that he or she is lesbian or gay.

I'm honored that William Saletan has taken my criticism seriously enough to reply, naming me along with the major LGBT groups that took aim. As you may recall, I wrote that Slate clearly knew it was publishing dangerous nonsense, because right before Regnerus's article, they put the link to Saletan's analysis tearing it apart. Nevertheless, here's what Saletan writes about the responses:

Wow. Regnerus' paper certainly has flaws. But before we all go get our stones, pitchforks, and kerosene, may I suggest an alternative? Trust science. Don't bury this study. Embrace it. The evidence Regnerus collected can help all of us rethink our ideas about sexuality and marriage. It can enlighten the right as well as the left. In fact, it's already doing that.

Saletan knows better. I don't know about the LGBT groups, but I wasn't taking aim at Slate for the underlying research. What Slate should be ashamed of is publishing Regnerus's sleight-of-hand interpretation of his results. Saletan's first-take analysis was correct: the study did not measure what Regnerus said it measured. Here's what Saletan wrote originally (emphasis mine):

In his journal article, Regnerus says it "clearly reveals that children appear most apt to succeed well as adults-on multiple counts and across a variety of domains-when they spend their entire childhood with their married mother and father." In Slate, he notes, "On 25 of 40 different outcomes evaluated, the children of women who've had same-sex relationships fare quite differently than those in stable, biologically-intact mom-and-pop families, displaying numbers more comparable to those from heterosexual stepfamilies and single parents."

These findings shouldn't surprise us, because this isn't a study of gay couples who decided to have kids. It's a study of people who engaged in same-sex relationships-and often broke up their households-decades ago.

But that's not what Regnerus wrote in his Slate article, or even in his journal article. He wrote that the conventional wisdom is false, and that lesbian and gay parents are bad for kids:

The basic results call into question simplistic notions of "no differences," at least with the generation that is out of the house. On 25 of 40 different outcomes evaluated, the children of women who've had same-sex relationships fare quite differently than those in stable, biologically-intact mom-and-pop families, displaying numbers more comparable to those from heterosexual stepfamilies and single parents….

Why such dramatic differences? I can only speculate, since the data are not poised to pinpoint causes. One notable theme among the adult children of same-sex parents, however, is household instability, and plenty of it. The children of fathers who have had same-sex relationships fare a bit better, but they seldom reported living with their father for very long, and never with his partner for more than three years.

The statement I bolded is flatly false. Those aren't the children of same-sex parents. Those are the children of different-sex parents, one of whom later has some sexual relationship with someone of the same sex, however brief or sustained. The gay dads he's writing about? Those are men who finally get an adolscence, late in life, after they've lied to themselves or others to try to fit in socially because people like Mark Regnerus told them being gay is bad. In our world, those men should never have married women. A healthy society would let them come out young and, if they wanted children, have children with a male partner with whom they could happily remain.

Regnerus knows what he did. He set up a study that would make it seem that anyone who ever slept with someone of the same sex hurts their children by doing so. Instability is well known to be harder on children than stability. Decades of research are clear on this: Children do better with parents who stay together and have relatively low-conflict relationships than they do in high-conflict structures. The new parenting studies that are trying to measure whether the gender of your parent's partner matters are following families where same-sex parents are together from the beginning – and comparing them with families whose different-sex parents are together from the beginning. That's how you tease out the effect of gender from the effect of instability. Regnerus did the opposite.

Regnerus is smart enough to know this. He did one thing while purporting to do another. He compared fidelity with adultery. He compared stability with instability. Then, in Slate, he said he was comparing different-sex parenting with same-sex parenting-conflating the effect of family explosion with the effect of parental sexual orientation. Saletan spotted that crap right away, and called it out in detail. In fact, in his article yesterday, he goes into his critique in more detail. An excerpt from what he writes:

The numbers don't add up, and the subset is too small to generalize, but you get the picture: Kids of gay parents, like kids of straight parents, did better in stabler families. And this fits the pattern of all those studies the gay-rights groups are citing against Regnerus: Children raised by committed, financially secure gay couples turn out fine.

This is where Regnerus made his second mistake: He pitted his study against prior studies that found happier outcomes in gay families. He attributes his findings to "better methods." But there's no contradiction between his study and the others. The prior studies simply targeted and featured the stablest, most educated gay couples. They were too narrow. Regnerus, by using the "did your parent ever have a gay relationship" question, captured all the messed-up families that had been left out. But his net was too broad: It yielded a sample dominated by kids who had scarcely lived in a same-sex household.

Yes, as Saletan writes, "There's nothing evil about the data set." True. But there's something evil about the propagandistic distortion of that data set. Most Americans aren't nerds like us who look at the underlying questionnaire and drill down to expose the flaws. They're ordinary people who are following what's important in their own lives. By the time they hear this news, what they'll hear is some TV anchor saying, briefly, "A new controversy emerged today about gay parents. Are they bad for kids?" That suspicion is what will trickle down into the debates in Maine, Maryland, Washington, and Minnesota, where ordinary folks are voting on whether two women or two men can marry. And the fear that same-sex marriage hurts children is a highly volatile one; raising the suspicion is enough to set back the vote. Preventing their parents from marrying, or letting the falsehood float around that their parents are bad for them, is what will hurt those children.

So why is Saletan apologizing for Regnerus's crap? I stand by what I wrote: Slate's editors should be ashamed.