The oft-cited assertion that there are "no differences" in outcomes between children of same-sex parent households and those of intact biological families may not be accurate, according to a new study published Sunday in the journal Social Science Research.
Adult children of parents who have been in same-sex relationships are different than children raised in intact biological families on a number of social, emotional and relationship measures, according to research from the University of Texas at Austin.
Among other things, they reported lower income levels, poorer mental and physical health and more troubled current romantic relationships. The study found 25 differences across 40 measures.
The research does not address why the differences exist. It doesn't predict if changing attitudes that are more accepting of same-sex relationships will mean that children growing up today with same-sex parents will one day fare better in similar analysis. It doesn't address stigma or whether the difference is not the sexual preference of the parents but rather how stable the home life was, lead investigator Mark Regnerus, associate professor of sociology at University of Texas Austin's Population Research Center, told the Deseret News.
"Nor does the study tell us that same-sex parents are necessarily bad parents," he said in a written statement. "Rather, family forms that are associated with instability or non-biological parents tend to pose risks for children as they age into adulthood."
His study does challenge long-held assertions that there are no outcome differences between children raised in intact biological families and those with same-sex parents.
A question of bias?
A separate analysis in the same journal edition by Loren Marks, associate professor at Louisiana State University, more directly challenges previous same-sex parenting studies as inadequate, biased and unreliable. He lists seven concerns with the science, including the fact that "well-educated, relatively wealthy lesbian couples have been repeatedly compared to single-parent heterosexual families instead of two-parent marriage-based families." Single-parent families typically have poorer child outcomes across several measures, so it's easier to look better against them, he said.
W. Bradford Wilcox of the National Marriage Project at the University of Virginia said biological married families are the gold standard for better outcomes for children.
How children fare in different family structures is a timely question because those on both sides politically of the gay marriage issue said the very studies Marks criticizes are often referred to by judges and legislators as authoritative in showing that lesbian mothers, in particular, are as good or better parents than biological, married parents. The American Psychological Association asserted as much in a brief in 2005, citing that research.
"The claims in that brief seem premature and overstated," Marks said. "I'm not trying to say the truth is 180 degrees in the other direction. That would be premature and overstated as well. But we need high-quality science on this topic that informs decisions with more valid and generalizable data."
Attempts to interview a spokesman from the APA or a primary author of the brief were not successful.
The research published Sunday is certain to be controversial, Marks said, adding he is not affiliated with either political party. "I never wanted to be co-opted on either side as someone to hate or as a campaign manager for anybody."
A broad-based sample
Regnerus used data from the New Family Structure Study (NFSS) to see how adults ages 18 to 39 who were raised by same-sex parents do on various outcomes compared to those raised by married biological parents, co-habiting adults, a single parent, step-parents or adoptive parents, among others. NFSS has data from more than 3,000 adults, including 175 who said their mother had a same-sex romantic relationship and 73 who said their father did.
Regnerus said his findings were more valid on lesbian-mom households than gay-father households because they included more families and also because those studied were far less likely to have actually lived in gay-dad households. A cursory look might lead some to conclude incorrectly the study found gay dads were better parents than lesbian moms. The sample wasn't large enough to draw strong conclusions about the men.
Data on display
Regnerus plans to make his data public. "In a piece like this that is overturning conventional wisdom, the onus is on me to be very up-front about how I reached my findings." He said he will post the research design, codebook and statistical analyses online Monday.
The study said lesbian mothers compare most favorably to step-families and single parents, not to intact biological families. He noted the step-, single and lesbian mom families structures all clearly included some upheaval.
The study does not say same-sex parenting is responsible for the outcome differences. "Causality would mean I ruled out other plausible explanations. I didn't. I can't say something about being a lesbian is particularly pernicious for young adult outcomes." But if he can't explain the outcomes he found, he said, neither can earlier studies, which were less robust and found the opposite.
He eliminated socioeconomics, age, politics, gender, geography, race and bullying as explanations for the gaps he found between family structure types.
Is it the stigma the parents felt? He doesn't know. "We didn't talk to parents, and I can't measure stigma." Single-parent and step-families have, much like same-sex parents, "a higher degree of instability" compared to intact biological families, he said. It's probably not just having a man and woman, either, since step-families have those and the kids don't fare as well.
"I think what good research does is open the door to replication and explanation," said Cynthia Osborne, associate professor in the LBJ School of Public Affairs at the University of Texas at Austin, who wrote one of the commentaries for the journal. "We're going to need to see if what we're finding here in his study will be replicated through lots of robustness checks, looking from various angles, cutting the data in different ways and coming up with the same thing. That's the next step."
Social science research seldom explains causation, said Osborne. And finding differences takes a back seat to understanding their cause and importance. A mom in a same-sex relationship may have had lots of different family structures, from a failed marriage to same-sex dating and cohabitation. "We know that change in family structure matters a lot in children's outcomes," Osborne said. With the variations, "comparison of family structure to partnership choice is an interesting comparison, but it doesn't take us far enough."
Another commentator, David Eggebeen, associate professor of human development and sociology at Pennsylvania State University, said data are never perfect. They range from not yielding any conclusion through data that beg caution to "really good data where you can be a lot more confident." The NFSS data, he said, "get us closer to being a representative population."
He called the Marks paper a "little disquieting" because of the disconnect between imperfect data and claims that same-sex parents are equal or superior to other parents.
Studies consistently say kids in a biological married family with both parents "are advantaged compared to any other kind of family," Osborne said. That alone raises a "conundrum" with the previous finding that same-sex couples have equal outcomes, "since that almost always implies a step-parent, a cohabiting partner — what we call a social father or social mother — divorce, adoption, at least one of those things." Some studies say those things don't disadvantage same-sex parent families.
But she also is bothered that some would use differences to bar same-sex couples from having kids without understanding what causes those differences.
"I hope people will take it on and look at related and more complex statistical questions," said Eggebeen. If Regnerus' findings don't hold up, why? I would see this as the beginning. The provocative findings get us to look at this."
Researchers on both sides say more research is needed and it could be years before the impact of changing family structure on children is clear.
It's an important public conversation, said Wilcox. "I think as a society we do value the well-being of children across the spectrum. We would like to create a context where kids could thrive."
If the findings are replicated, the whys need to be answered, he said. "How much instability, how much community stigma, how much biological ties, and how much differential access to the legal institution of marriage account for the differences the Regnerus study finds? I think if you look back historically on the impact of divorce on children, there were a number of chapters and a changing story emerged about the impact on children. I think the Regnerus study is bringing us to a new chapter. The first suggested no difference, the second will suggest there are some. The third chapter is going to be trying to figure out the differences."
Douglas W. Allen, the Burnaby Mountain Professor of Economics at Simon Fraser University in Burnaby, British Columbia, said that no matter what future studies say about same-sex families, scientific methods have to be protected so they can't be hijacked by political pressures on either side. Science must be robust and careful, a proper probability sample the basis for any claim about a broad population.
His criticism of most of the studies since 2005 mirrors Marks'. All but one was a convenience study, called that because the subjects are "convenient" to gather into the sample. You want to study lesbians, so you ask a friend you know who is lesbian if she'll answer questions. That turns to the snowball study, when you ask if she knows other lesbians you can talk to as well, he said.
Those types of studies made up most research on same-sex parenting. "That's totally fine to do if you're just interested in exploratory work," said Allen, who doesn't know Regnerus or Marks. "Often an area of research starts this way. The problem comes when you try to extrapolate what you find to the broader population. You cannot. It says absolutely nothing about the population of lesbian and gay parents, only about those particular ones."
Allen expects a lot of people to try to poke holes in Regnerus' work. But expected criticism that it was privately funded or not rigorous enough will be the "pot calling the kettle black. Other studies are off-the-charts biased bad," he said. Every researcher has biases, but when enough tackle a question, truth "sifts to the top."
However, it may take a couple of decades, he warned.
Not everyone agrees with his assessment. "The research up to this point has shown absolutely no difference of any note — and in some cases, differences that might be considered in favor of lesbian parents," said David Brodzinsky, professor emeritus of clinical psychology at Rutgers and research director of the Evan B. Donaldson Adoption Institute. He had not seen the new studies, which the reporter did not share because of an embargo. Instead, he spoke in general terms. "In 25 years of study, more than 50 studies have not found the negative outcomes that critics are concerned about. That doesn't mean all the studies are well designed. Some are not. But there are studies that are relatively reasonable sizes, nonconvenience samples and at least reasonably representative that have been used as a database for looking at questions about same-sex parenting."
Both sides told the Deseret News they believe opposing researchers approach topics from what Brodzinsky called "a political perspective."
"We find that critics continue to make claims that simply aren't supported by the research," he said. The need, he noted, is not so much tracking negative outcomes — "I think most scientists are reasonably satisfied that a gay parent can be damaging, as can a straight parent" — but finding what supports or undermines a child's positive development.
While Brodzinsky said some studies show young adults raised by same-sex parents are more likely to acknowledge same-sex curiosity and even experimentation, there is not evidence they are more likely to self-identify as gay or lesbian. "Even if they did, is that an adjustment difficulty? Not according to the APA," he said. "It does not translate to risk for depression or other kinds of adjustment problems." Research has found no different outcomes based on parental sexual preference in the cognitive development and social success of offspring, he added. The same is true for delinquency and victimization. What is a fair criticism of the literature on same-sex parenting, he said, is that it has looked almost exclusively at middle-class families. "A lot of research on single parenting is confounded by the fact they have fewer resources." People who choose to parent as singles, gay or straight, tend to be middle or upper-class and the data looks quite good for the children, he said. "Single parenting is as much about resources as it is about the quality of parenting."
The Regnerus study is unusual, as well, because it questioned the children, now adults, themselves, instead of asking the parents to report on how they thought their kids were doing.