Same-Sex Parenting: Getting the Story Straight

How well do children fare when raised by a pair of same-sex parents? Whenever this question arises—and the issue has loomed large in the debate over same-sex “marriage”—homosexual activists have brandished dozens of sociological studies apparently demonstrating that children of same-sex couples do just as well or better than peers being raised by two biological parents.  A growing body of new research, however, raises doubts about the reliability of those studies—and so intensifies concerns about the well-being of children in the care of same-sex couples.  The latest addition to this body of research comes in three new studies conducted by sociologist D. Paul Sullins of the Catholic University of America.

In one of these studies, Sullins highlights indications of “strong bias resulting in false positive outcomes” in many studies of same-sex parenting. These indications emerge as Sullins carefully scrutinizes five studies of same-sex parenting, all using the same survey to assess child well-being—two focused on random samples and three with recruited samples. In addition, Sullins draws data from two other random-sample studies of same-sex parenting relying on other metrics of child well-being.  

Sullins stresses the importance of the studies involving recruited samples, noting that research on same-sex parenting “has been dominated by small convenience samples recruited from GLBT [Gay, Lesbian, Bi-sexual, Transgender] interest, advocacy or support groups.” Such recruitment practices, as Sullins notes, can easily lead to “volunteer bias.” That bias is all too likely to grow in a field in which it is very rare for “the aim of the study [to be] unknown to participants.” In-the-know participants may deliberately skew the data toward outcomes they favor.  

It should come as no surprise, then, that as Sullins compares the random-sample studies of same-sex parenting with the recruited-sample studies, he limns a very suspicious pattern. In the three recruited-sample studies, when researchers compare children in same-sex-couple households with children in opposite-sex-couple households on various survey measures of well-being, almost four-fifths of the item-by-item comparisons (79.3%) yield favorable outcomes for the children of same-sex couples. In sharp contrast, in the two random-sample studies using the very same survey, when the researchers compare children in the care of same-sex couples with peers in the care of opposite-sex couples, they find “no favorable comparisons” for the children in the care of same-sex couples.

Results favorable to same-sex parenting look even more dubious when Sullins turns to two other random-sample studies of same-sex parenting. Using measures of child well-being other than the survey employed in the other five studies in view, the authors of these two studies conclude that for psychological outcomes (“depression, self-esteem, and anxiety”) and for behavioral outcomes (“delinquency . . . such as binge drinking, illegal drug use [and] risky sexual behavior”), “all of the [study’s] measures were less favorable for S[ame]S[ex] children in the sample” than they were for peers in the care of opposite-sex parents.  

Sullins’ study makes it impossible to ignore the “strong evidence of substantial bias understating the psychological difficulties of children with same-sex parents . . . in studies using recruited convenience samples.” Driving home the point, Sullins remarks, “To date, no representative-population data have found lower emotional problems among children with same-sex parents. Every random-sample [study] has observed higher emotional problems among such children; where the sample was large enough, those differences were statistically significant.”

Understandably, Sullins stresses that policymakers, academics, and ordinary citizens need “accurate, unbiased information” if they are to “best serve the health and welfare of all children involved.” 

And in two other new studies of same-sex parenting, Sullins provides such accurate and unbiased information.

In the first of these studies, Sullins investigates the prevalence of Attention-Deficit Hyperactivity Disorder (ADHD) among children in the care of same-sex couples. To measure this prevalence, he parses data collected between 1997 and 2013 from a nationally representative sample of 195,240 children, including 512 children in the care of same-sex couples. These data provide unmistakable indications of disadvantage for children with same-sex parents. Using a statistical model that compensates for differences in the sex, age, and ethnicity of the children and the education of the parents, Sullins calculates that “children with same-sex parents were 2.4 times more likely to have ADHD than were children with opposite-sex parents.”

Among children living with same-sex parents, Sullins finds that “the well-documented co-occurrence of ADHD with general emotional problems and attitudinal defiance appears to affect children with same-sex parents at about the same rate as those with opposite-sex parents.” But, of course, since children with same-sex parents are almost two-and-a-half times as likely to suffer from ADHD as children with opposite-sex parents, they are also much more likely to develop the related emotional problems and attitudinal defiance. 

What is more, some problems incident to ADHD seem particularly acute among children living with same-sex parents: the data indicate that while children who suffer from ADHD are “over three times [Odds Ratio of 3.2] more likely to be stigmatized or bullied (than are those without ADHD)” in the general population, they are “over seven times [Odds Ratio of 7.5] more likely to be bullied” if they are the children of same-sex couples.

In light of his findings, Sullins believes it appropriate to “ask whether higher ADHD risk for children with same-sex parents results from impaired parenting or a direct biological link.”  Sullins is inclined to interpret his data as evidence that “genetic rather than environmental factors influence child ADHD with same-sex parents, perhaps through a direct association, that is, a correlation between same-sex attraction and parent ADHD.” Still, Sullins admits to some difficulty reading the data in strictly genetic terms. After all, in households headed by same-sex couples, we see “the presence of at most one biological parent and the corresponding existence of an absent, most likely heterosexual, parent.” 

But it is not just genetic calculations that become dicey when a biological parent is absent from a child’s home. Sullins indeed demonstrates just how much that absence matters in a third new study, one focusing on the type and prevalence of emotional problems among children living with same-sex couples.

In this third study, Sullins compares the incidence of emotional problems among children living with same-sex parents to that of children living with opposite-sex parents. To perform this comparison, Sullins uses a slightly enlarged version of the same nationally representative data set he uses for his ADHD study, a version comprising information for 207,007 children, including 512 in the care of same-sex couples.  

The pattern that emerges from these data could hardly be more striking: “children with same-sex parents are assessed at higher levels of distress, compared to children with opposite-sex parents, for every measure of child emotional difficulty, developmental difficulty or treatment service.”

Nor is it a small gap that separates children with same-sex parents from peers with opposite-sex parents: Sullins calculates that “emotional problems were over twice as prevalent [Risk Ratio of 2.4] for children with same-sex parents than for children with opposite-sex parents.”

Homosexual activists might argue that children with same-sex parents suffer from emotional problems only because of homophobia, which fosters stigmatization of such children. But Sullins finds no evidence to support this argument. Indeed, the data indicate that children of opposite-sex parents actually experience slightly more bullying than do peers in the care of same-sex parents (though the difference—within the range of “sampling error”—is not statistically significant). “It appears,” remarks Sullins, “that the rate of abuse and susceptibility to emotional distress due to stigmatization does not differentiate sharply between children in same-sex and opposite-sex families.”

Sophisticated statistical analysis only increases the difficulty of explaining away the gap in emotional distress separating children of same-sex parents from peers with opposite-sex parents. The raw data indicate that 12.1% of children with same-sex parents suffer from “serious emotional problems” compared to just 5.2% of children with opposite-sex parents (p = 0.001). But the gap separating the two groups of children actually yawns wider when Sullins re-assesses the same data in a statistical model accounting for differences in children’s sex, age, and race and in parents’ education and income. In this more sophisticated statistical model, 5.5% of children with opposite-sex parents suffer from serious emotional problems, compared to 14.9% of children with same-sex parents (p < 0.001).

To be sure, when Sullins compares the raw data for children with same-sex parents to the raw data for children with single parents or opposite-sex parents who are cohabiting, the gap in emotional distress falls below the threshold of statistical significance, but just barely. And the distinct vulnerability of children with same-sex parents re-emerges when Sullins enters his data into his statistical model compensating for differences in children’s age, sex, and race and in parents’ education and income. In this model, children with same-sex parents suffer from emotional problems at rates that are statistically higher than those seen among children with two opposite-sex parents in non-traditional relationships (Relative Risk of 1.87 compared to cohabiting opposite-sex parents, of 2.16 compared to opposite-sex step-parents). The vulnerability of children living with same-sex parents stands out even more starkly when the comparison is with children living with two married biological parents (Relative Risk of 3.62).

This study renders dubious even the argument that homosexual couples can give children the advantage of living with two adults who care for them. For in Sullins’ multi-variable analysis, children with same-sex parents are almost twice as vulnerable to emotional distress as are peers living with single parents (Relative Risk of 1.78).

The gap separating children of same-sex parents from other children only grows larger when Sullins adds to his analytical model statistical measures of “instability,” “parental severe psychological distress,” and “stigmatization.” “Surprisingly,” observes Sullins, “the risk due to same-sex parents is not moderated, but increases substantially, when all three of the above factors are combined.” Adding these three “confounders” to his analysis, Sullins sees the relative risk for emotional problems climb even higher among children of same-sex parents (Relative Risks of 3.08 compared to peers from single-parent homes, 2.46 compared to peers from cohabiting-couple homes, to 2.97 to peers from stepfamilies, and 4.52 from married biological-parent homes).

Sullins thus dismisses as irrelevant considerations of “instability,” “parental severe psychological distress,” or “stigmatization.” But he finds considerable explanatory power in sheer biology. 

As Sullins surveys the overall findings of his study, he sees how “the results . . . converge on a clear central finding: Biological parentage uniquely and powerfully distinguishes child outcomes between children with opposite-sex parents and those with same-sex parents.” In fact, Sullins reports that “the risk of child emotional problems is no different with same-sex and opposite-sex parents in the comparable family forms, i.e. cohabiting and step-parent families, once differences in biological parenting are equalized.” That is, the key variable in predicting children’s emotional vulnerability seems to be number of biological parents in the home.

It is unlikely that anyone will miss the implications of this study for the debate over same-sex “marriage.” But just in case, Sullins spells them out, noting that “the primary benefit of marriage for children . . . may not be that it tends to present them with improved parents (more stable, financially affluent, etc.), although it does this, but that it presents them with their own parents.” Thus, even if homosexual activists prevail in the courts by winning for same-sex couples a right to “marry,” Sullins cannot see how these activists will repeal biology. Compared to families based on traditional marriage, families based on same-sex relationships will—regardless of the law—“continue to have fundamentally different, even contrasting, effects on the biological component of child well-being, to the relative detriment of children in same-sex families. Functionally, opposite-sex marriage is a social practice that, as much as possible, ensures to children the joint care of both biological parents, with the attendant benefits that brings; same-sex marriage ensures the opposite.”

No doubt homosexual activists will continue to proffer assurances as to the beneficence of same-sex parenting. But the research they cite in giving such assurances has now been exposed as deeply flawed. New and more rigorous research now illuminates the real and sobering risks that children faced when in the care of same-sex couples.

(D. Paul Sullins, “Bias in Recruited-Sample Research on Children with Same-Sex Parents Using the Strength and Difficulties Questionnaire (SDQ),” Journal of Scientific Research & Reports 5.5 [2015]: 375-87; “Child Attention-Deficit Hyperactivity Disorder (ADHD) in Same-Sex Parent Families in the United States: Prevalence and Comorbidities,”  British Journal of Medicine & Medical Research 6.10 [2015]: 987-98; “Emotional Problems among Children with Same-Sex Parents: Difference by Definition,” forthcoming in British Journal of Education, Society & Behavioural Science, Article no. BJESBS.20YY.0XX, ISSN: 2278-0998, emphasis added.)