THAT study on same-sex parenting

With the furore over same-sex marriage gaining steam both in the United States and in New Zealand, where the issue is about to be debated in Parliament, I thought I should say a few things about that recent study that’s got some people upset – The study that said the things that we are all supposed to just know aren’t right and, more importantly, shouldn’t be said, that (stated in very general terms) children do better when they have, in their home life, a man who is their father and a woman who is their mother (as opposed to only one parent or parents of the same sex).

Mark Regnerus, the man who headed up the research project called “New Family Structures Study” (NFSS), authored “How different are the adult children of parents who have same-sex relationships? Findings from the New Family Structures Study,” in the peer-reviewed Social Science Research 41 (2012), 752-770.

I suppose I shouldn’t have been surprised by this – but for some reason I was, since I (perhaps foolishly) assumed that analysis of a study like this would be done in an at least apparently scientific, logical manner. On reflection, this was not a wise supposition. Anyway – I was surprised to see that by far the majority of negative comments that are made about this study are not about the research at all, but about the person behind it. “The Daily Beast” blog provides a sadly typical example:

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.

Look through the list of factors that David Sessions (the author of the above article) lists. These are listed as things that Gays and Lesbians should be alarmed by when seeing this study:

  • The research project was led by a person who engages in “attention grabbing” research
  • This man’s findings sometimes agree with what social conservatives think about the sexual revolution.
  • This man was once a professor at a Christian College
  • This man wrote a cover story in a Christian magazine, saying that Christians should encourage marriage at a young age.
  • This man wrote an article claiming that the sexual revolution has had some negative consequences for women.
  • This man’s latest research (the study on same-sex parenting) was funded by groups that are not socially liberal, but socially conservative instead.

It is difficult to imagine another excerpt of writing this brief that contained so many implied examples of the tried and trusted ad hominem fallacy, especially the variety known as “poisoning the well.” Of course there’s nothing rationally compelling about these factors that should cast doubt on any of this study’s findings, and they certainly should not cause anyone alarm. These factors are only mentioned at all because they stir up prejudice against the author on the part of those who do not identify with his background. But surely research should be evaluated on its merits, and not based on whether or not you like the researcher. Imagine that there was a study this large that showed on evidential grounds that in fact the children of same-sex couples do as well or better than other children (as far as I know there isn’t one, but ignore that for now). Now imagine that the Pope said that we should be alarmed by the existence of this study because:

  • The author of the findings is an attention seeker.
  • The author’s findings sometimes agree with what social liberals think about the sexual revolution.
  • The author was once a professor at a well-known liberal college
  • The author wrote an article for a gay men’s magazine, urging parents to encourage their children to explore their sexual identity before deciding that marriage is right for them
  • The author wrote an article claiming that the sexual revolution on the whole was good for women and relationships in general.
  • The author’s research was funded by left-leaning organisations who would not be described as socially conservative.

I think the response from many in this case would be a predictable “So what? We expect a scientifically minded person to reach the same conclusions as us!” Whether they would be right or not, dismissing an author like this is plainly no way to set the scene for fair criticism. If you have found flaws in a person’s research, attack the flaws. Don’t dismiss the research early by attacking the person behind it. Unfortunately however, prejudicial tell-tale signs like this are rearing their head everywhere I can find criticisms of this research. Regnerus is a conservative, or better yet, an ultra-conservative. His research is really driven by “right-wing ideology.” He has links to scary groups, by which we mean that conservative (I mean ultra conservative fundamentalist hard right wingers) are somehow associated with him. And on it goes.

Moreover, as I look around at the self-appointed blogger army of peer reviewers, it seems that almost any claim seems acceptable if it means making the research look bad. A reputable commenter at the Gay South Florida site (“Mommy Psychologist,” who is actually a practicing child psychologist) felt quite confident in telling others that “It is also important to note that his research wasn’t peer-reviewed. Big red flag there.” There’s isn’t even the slightest credibility to this claim. Social Science Research is a peer-reviewed journal, and they certainly didn’t make an exception for Regnerus. Yet people will just allow themselves to think this if it casts doubt on the study, without any evidence. Big red flag right there!

But what of the actual criticisms of the research itself? Well, after hacking through all the verbiage, yes there are actually several criticisms being made.

Getting out of the all-in-brawl that is the blogosphere (yes, including this blog, naturally) and back into the restraint of what will pass peer review (and not just from mind to keyboard), the assessment of other social scientists is somewhat more restrained, even favorable. In the July 12 issue of Social Science Review several such commenters reviewed Regnerus’ work – along with a paper by Loren Marks that was highly critical of the standards of earlier studies that had found little or no disadvantages for children of same-sex couples. Paul Amato of Pennsylvania State University acknowledges that the previous studies that Marks criticises are indeed based on “small convenience samples,” which is a disadvantage, but hard to avoid. In terms of sampling methods, he grants “The New Family Structures Study is probably the best that we can hope for, at least in the near future.” ((Paul Amato, “The well-being of children with gay and lesbian parents,” Social Science Review 41:4 (2012), 2.1)) He grants that in terms of statistical power, “the Regnerus study is better situated than virtually all previous studies to detect differences between these groups in the population.”((Amato, “The well-being of children with gay and lesbian parents,” 2.2)) The only thing he says that could be construed as a criticism was to note that “more than half” of the children of same-sex couples are from families where there has been a divorce or the end of a previous relationship. This is because more than half of the children in question were conceived naturally by one of the parents (either mother or father) in the same-sex union, with a former partner. Here, a controlling factor will always be the brute natural limitation that same-sex couples cannot reproduce with one another.

It is a little unfair, therefore, for so many of the blog army to pile up against the study and accuse it of biased selection by surveying so many children of same-sex partners where there has been a relationship breakdown. This type of history is not easy to avoid in any random survey for the simple reason that children of homosexual individuals will have experienced the breakdown of a parental relationship if that parent was in a relationship capable of producing children (namely, the child whose well-being we are now considering). Indeed, to exclude such children from this study as some seem to be saying should have been done, would be to force the study to misrepresent reality, thereby producing unreliable results. We should care less about how well children would do in some other make-believe world and more about how well they do in the real world.

Lastly, Amato raises the concern – which is not a criticism of the NFSS, that people may wish to use the helpful findings of this study “to undermine the social progress that has been made in recent decades in protecting the rights of gays, lesbians, and their children.” Fair comment. They might. Amato suggests that this is misguided for a variety of reasons. Two that stood out to me as eminently plausible are 1) The legal right to enter marriage with another person should be a constitutional one, not grounded in the range of variables that make a difference in family well-being, and 2) the children of same-sex couples, by virtue of the fact that same-sex couples are usually not legally married (since legal same-sex marriage does not exist in many states), have unmarried parents. There is research indicating that marriage between parents improves outcomes for children, so it may well be that the results of this study would have turned out differently in a society where same-sex marriage existed as a legal institution.

Regnerus, responding in the same issue of the Journal, shares this concern (a concern raised by Amato and also Cynthia Osborne), saying, “I recognize, with Paul and Cynthia, that organizations may utilize these findings to press a political program. And I concur with them that that is not what data come prepared to do. Paul offers wise words of caution against it, as did I in the body of the text. Implying causation here—to parental sexual orientation or anything else, for that matter—is a bridge too far.”1)

On the whole, then, I think Amato’s analysis of the study is that it was more comprehensive than others, and did indeed tell us about real-world outcomes for the children of same-sex partnerships, even if those outcomes might have been different under different societal conditions (conditions that do not in fact exist). It is certainly a kinder assessment than that of those whose comments have no hurdle of scientific respectability to clear before they get to be read by the public.

Regnerus, far from being the fundamentalist zealot depicted by the mobs of torch and pitchfork waving bloggers, was both moderate and conciliatory in response to his three reviewers:

As each of the three explicitly or indirectly notes, family instability—whatever the sources—is often a top culprit in predicting dysfunction in the lives of children, and the data analyses in my article likewise point in this direction. In fact, the most significant story in this study is arguably notabout the differences among young-adult children whose parents who have had same-sex relationships and those whose parents are married biological mothers and fathers, but between the latter and nearly everyone else. Contexts of instability—whether in gay or straight households—appear suboptimal for children’s healthy long-term development. While much is made in the scholarly literature about “resilient” youth—those who thrive despite the odds against them and in lieu of an optimal family context—resilience is, on average and perhaps by definition, not normal. Moreover, even resilient children would likely prefer to have engaging parents who are not simply in their lives but in their households. Adults of good will, and most family scholars, typically agree on this. Whether some relationship arrangements are more systematically prone to disorganization than others is an important and empirically-testable question.

Here is where the relative instability of the family histories of the children of same-sex couples becomes more relevant. Since, under real-world conditions, it is quite understandable why the phenomenon of relationship changes has a cause in such households that is not present in households of opposite-sex parents, what is really being thrust into the spotlight, as Regnerus notes here, is the differences (if there are any) in the average long-term stability of relationships between two people of the same-sex and two people of the opposite sex, something that he doesn’t delve into in the NFSS or his reply to reviewers.

So there you have it. It’s not the bombshell revelation-to-end-all-revelations study that settles every argument that some might have hoped for, and it’s not the atrocious omg-I-can’t-believe-anyone-would-publish-this piece of trash that some are claiming it to be. What the appearance of this study has done, however, is to again allow the ugly side of intellectual policing to rear its head. Some things just shouldn’t be allowed to be said, so when they are, they must be shouted down by any disreputable means necessary.

The comparative forthrightness and honesty of Ilana Yurkiewicz is refreshing. It’s not that I agree with her about same-sex parenting and its merits. I am inclined to disagree. And yet in spite of her stance on same-sex parenting, she admits without blushing that findings like those in the Regnerus study can simply be set aside, no matter how robust the study is: “I don’t need to reject his paper to affirm that I support same-sex couples having children.”

And why not? Here’s why:

In medicine, we learn that it is valuable to order a test only if it would change our decision-making in terms of careOtherwise, the test is considered a waste. While the same litmus test doesn’t apply to science research across the board, I think it comes a bit closer in the realm of social science work, which is what Regnerus’ study was. Suppose for a moment that all the critiques of his methodology did not apply, and that his was a robust study. Would its conclusions change your opinion on gay and lesbian couples having children?

As you might guess, Yurkiewicz’s answer is a flat no. After all, if that’s how you think, then any time a couple lives in conditions that make the outcomes for children less good in some way than they would be in some other common conditions, we would have to step in (or at least someone would) to assess whether or not that couple ought to have a right to raise children, which must surely seem draconian to most of us (it certainly does to me).

My instincts (not the most scientific things in the world at times) tell me that there is likely a good deal of truth in what the Regnerus study tells us, and that under normal (real-world) circumstances, or even ideal circumstances (the children from opposite sex couples in this study surely did not all come from ideal circumstances), a healthy marriage of a man and a woman (not just any marriage, of course, but a healthy one) provides the best environment for raising children. This is (obviously) the environment in which children are naturally produced, and I suspect there’s an intuition shared by many (including me) that a man and a woman together can provide a more balanced model of character to children (since men display different characteristics than women, and vice versa).

But ultimately that’s not the reason why I think child rearing ought to take place within a family context with a marriage of man and woman as parents. As I’m sure will not shock any of my readers, I’m a Christian. By that I don’t mean to say that I’m a special or different sort of person. I mean that this is how I view what’s proper for people. God made us and calls us to live as he intended. I’m not going to delve into what I think God thinks of homosexuality (although I do intend to at some point). But if I’m right, God’s intended model for the family (yes, recognising that for a whole range of reasons we might fall short, some of these reasons being our fault and others not) is that a marriage consists of a man and a woman, and a family consists of a married couple who raise children together. Because that’s the way that our maker intends us to live, that’s what I regard as natural and normal.

My view that this is natural and normal would not be affected if a new study came out showing that there were certain benefits for children who were raised by two men (or women). I’m pretty confident that no such study would be produced, and I’m happy to admit that I would be suspicious of any such study at least in part because I think that what God wants for us is actually what is, on the whole, good for us. But that aside, how could any such study, even if correct, change whether or not God intended us to live a certain way?

So I do share Yurkiewicz’s concern to stress that whether or not we ought to change our views and practices surrounding marriage cannot be solely driven by circumstantial factors in any given case, but on more fundamental principles.

However, I have mixed views on Yurkiewicz’s warning:

In fact, to feel we need to refute unlikeable data buys into a dangerous premise. The impulse to reject findings we don’t agree with is tacitly conceding that this kind of data can legislate rights, so to make sure we maintain the ones we want, it’s best to hide the findings that might undercut them. That admission is deeply problematic – for science because it leads us down a road of stifling findings that don’t resonate with our moral preferences, and for politics because it says our nation’s values on who should be able to have children are not founded in basic rights, but instead subject to the results of a single social science study. Neither is a road we should feel comfortable treading on.

I think it’s just false to get into the mindset that having children is simply a matter of our rights as parents. This view has little or nothing to commend itself, in my view, and of course we should think forward to the consequences of our choices for those children that we think we have a right to (this is not the same as saying that those whose circumstances aren’t as good for children as somebody else has no right to raise children). The having of other people isn’t a right (otherwise anyone who could not have children has had their rights violated by… something or someone, although it would often not be clear whom or what).

But I definitely agree with Yurkiewicz that we shouldn’t feel the need to refute unlikeable ideas. Data shows what it shows (obviously), and we should be grateful for any new knowledge we receive from it. After all, if same-sex couples want a world in which they are more likely going to be permitted to have children and there is such data out there, ignoring it would be a bit like smothering the smoke alarm because we don’t want anyone to know about the potential risks that our sleeping children face. And I’m sure you don’t want to be doing that. Right?

Glenn Peoples


  1. Mark Regnerus, “Response to Paul Amato, David Eggebeen, and Cynthia Osborne,” Social Science Review 41:4 (2012 []

7 thoughts on “THAT study on same-sex parenting

  1. I’m surprised you have time to blog at all. Doesn’t it interfere with your clearly full time role of hating other people just for being different from you, Glenn? Disgraceful. Nice Christian “love.”

  2. Dalm it Martin. You have burned out the irony meter on my laptop. I had just finished the prototype.

    I can find very little in this post about hating other people – apart from the sarcastic comment about Chrisians in the comment box.

    I also get a bit frustrated with the mixed use of data by politicians and those with a political agenda. The reaction of many to this research was to immediatly assume it must be wrong, and must be biased and motivated by hate.

  3. Thanks Glenn.

    During your research did you come across any of the multi-series blog exposé on Regnerus at “Box Turtle Bulletin”? Overall it achieves a worthy critique – even though a fault may be found here or there.

    On the whole, I don’t take issue with your assessment of this study even though I do not align with your instincts. I would be happier if this study was called:
    “Compared to adult children who come from in-tact opposite sex family situations – how different are adult children who report that one or both of their parents were engaged in some form of same-sex relationship for a period of their lives, and where the majority have experienced the breakdown of an opposite sex marriage or relationship?”

    If it had been called that, I think there would have been a far lesser furore.

  4. Henry, I offered an explanation of why people in a same sex relationship who have children are inherently more likely than most people to have ended a serious relationship in the past where the children were involved, so that proposed title may be somewhat over nuanced. 🙂

  5. Yes, do you sense that the proportion of children who move from opposite sexed households to same sexed households is in flux (vs. those who do not experience a transition)? I get a whiff from the structures of feeling (Raymond Williams) that with new marriage non-discrimination laws that it is a proportion very open to change.

    1. I assume you mean non-discrimination when it comes to adoption or IVF etc.

      I don’t think this is the sort of change we can intuit. Whether proportions change so that the majority of children had by homosexual parents (single) and same-sex couples are adopted or created via clinical means rather than from an opposite-sex union is the kind of thing that will be discovered by empirical research. My gut doubts it very strongly, but time will tell.

Leave a Reply

Your email address will not be published. Required fields are marked *

 characters available