esp | eng

New Family Structures Research and the “No Differences” Claim
Ana Samuel

The widely circulated claim that same-sex families are “no different” from intact, heterosexual families is not settled science. New studies suggest that children do best when they are raised by their biological parents in a stable, intact marriage.


Two peer-reviewed research articles in the social sciences, released June 10, 2012, challenge the claim made by same-sex parenting researchers over the last decade that parents engaged in same-sex relationships do just as well as other parents at raising children. This claim, that there are “no differences” in outcomes between the two kinds of parenting, is undermined by new evidence that these papers present.

Article by Professor Loren Marks

The first article, published in Social Science Research, July 2012, can be found here, along with responses from peers, and a final reply by the author. In that manuscript, “Same-Sex Parenting and Children’s Outcomes: A Closer Examination of the American Psychological Association’s Brief on Lesbian and Gay Parenting,” Professor Loren Marks of Louisiana State University’s School of Human Ecology reviews the 59 studies referenced in the 2005 American Psychological Association brief that came to the conclusion that there are “no differences.” Marks concludes that “not one of the 59 studies referenced in the 2005 APA brief compares a large, random, representative sample of lesbian or gay parents and their children with a large, random, representative sample of married parents and their children. The available data, which are drawn primarily from small convenience samples, are insufficient to support a strong generalizable claim either way.”1 Marks’s study casts significant doubt upon the older evidence on which the APA brief, and thus the “no differences” paradigm, rests.

Article and New Family Structures Study conducted by Professor Mark Regnerus

The second article, by sociologist Mark Regnerus of the Population Research Center of the University of Texas at Austin, presents new and extensive empirical evidence that suggests that there are differences in outcomes between the children of a parent who had a same-sex relationship and children raised by their married, biological mothers and fathers. This new evidence is based on the data from the New Family Structures Study (NFSS) of the University of Texas, of which Regnerus was the lead investigator and which in 2011 surveyed 2,988 young adults for the specific purpose of collecting more reliable, nationally representative data about children from various family origins: intact biological families, late-divorced families, stepfamilies, single-parent families, adoptive families, families with a lesbian mother, families with a gay father, and other family types (such as families with a deceased parent or other combinations). The NFSS has already been acknowledged by critics to be “better situated than virtually all previous studies to detect differences between these groups in the population.”2

The remaining portion of this summary will focus on this new research as featured in Dr. Regnerus’s article, entitled “How Different Are the Adult Children of Parents Who Have Same-Sex Relationships?”, which was published in Social Science Research, and can be found here. Dr. Regnerus’s article highlights the data comparing children from intact biological families to children who were raised with a parent who had same sex relationships.

Distinctive Features of the Regnerus Article and the NFSS

As Regnerus explains, the NFSS is unique among gay parenting research in three ways:

First, it compares the outcomes of children who reported having a mother who had a lesbian relationship with another woman (MLR for short) or a father who had a gay relationship with another man (FGR for short) with the outcomes of children who reported coming from an intact biological family (IBF for short). Most gay parenting research compares gay and lesbian parenting to single, divorced, and step-parent parenting, or conversely compares a select, and often socio-economically privileged, population of gay parents to a broad, representative sample of the general population.

Second, the NFSS focuses on the responses of young-adult children. Other current studies on gay parenting focus on what is going on inside the households of lesbian and gay parents at present, while the children are still under their parent’s care. Moreover, these studies most typically interview the parents for their point of view about what it is like to be parenting as a gay man or lesbian woman.3 This research does not tell us how the children turn out as adults. Indeed, no study has explored that question until now. The NFSS interviewed the sons and daughters of parents who had a same-sex relationship after they had grown up and matured into young adults (ages 18–39), and most of them had already moved out of their parent’s home. These children spoke for themselves about their experiences at home when they were younger and were able to report on how they are doing now as young adults.

Third, the NFSS drew from a large, random sample of the U.S. population of young adults ages 18–39. This third point is a significant strength of the NFSS because, to date, there is only one other gay parenting study that draws from a large, random sample, that of Michael Rosenfeld of Stanford University, who relies upon 2000 U.S. Census data. Every other gay parenting study thus far relies upon small or non-probability samples, which do not allow for generalization and are thus inadequate for drawing conclusions about the population at large.4 For example, the National Longitudinal Lesbian Family Study used a “convenience” sample and recruited respondents entirely by self-selection (from announcements posted in lesbian newspapers, women’s bookstores and lesbian events in Boston, Washington, and the San Francisco areas).5 While these types of studies are valuable for gathering information about the specific lesbians who demonstrate those kinds of behaviors (that is, who attend book stores, read newspapers, and are “out” enough to attend lesbian events) they are problematic when the goal is to generalize to the general population of lesbians, some of whom may not have the social, economic, or behavioral patterns of the former group. Any claims about the general population that are based on a group that does not represent it will be defective, because the sample will be less diverse than what a truly representative sample would reveal.

This problem is compounded when these studies compare data from the small convenience samples of gay parenting with data on heterosexual parenting from large, population-level samples. Although researchers usually note this limitation of their studies, the media almost always fail to transmit that limitation to the public at large, so the overall impression is that gay parents are doing just as well (or better) when compared to a large selection of heterosexual parents. By contrast, the NFSS drew a large, random sample from the general U.S. population of children from each family structure and compared these children using the same source data and methods so that comparisons are fair and representative.

Additionally, the NFSS gathered data on an ample range of children’s outcomes that are of interest to research on gay parenting, covering the social, emotional, and relational well-being of the respondents. The one other gay-parenting study of large, random samples of children—Rosenfeld’s—measured only one such outcome: children’s educational achievement. The NFSS looked at how the children fared in 40 different areas, a few of which will be highlighted here.

Qualities and Qualifications of the NFSS

Before detailing the results of the NFSS, two important points must be made: First, the results do not claim to establish causality between parenting and child outcomes. In other words, the results are not a “report card” on gay parenting, but a report on the average condition of grown children from households of parents who had a same-sex relationship versus those from IBFs. So, for instance, when the study finds that children who had a parent in a same-sex romantic relationship are much more likely to suffer from depression as young adults than the children who came from intact biological families, it does not claim that the gay parent was the cause of the depression in his or her child, but simply that such children on average had more depression, for reasons unidentified by the study. The goal was simply to identify average differences among the groups of children and to test just how strong the groups’ differences were.

That said, the study did control for many other variables, like age, gender, race or ethnicity, level of mother’s education, perceived household income while growing up, the degree of legislative gay-friendliness of the respondent’s home state, and experience of being bullied as a youth. Controls help sociologists eliminate alternative explanations for a given outcome, making the causal link between parenting structure and children’s outcomes more likely when the results are statistically significant after controls.

Second, the kind of gay parenting identified was rarely planned by two gay parents. Although there is speculation that today’s gay men and lesbian women are more likely to plan for children and seek out IVF, surrogacy, and adoption to have them, the study suggests that the children who were raised by a gay or lesbian parent as little as 15 years ago were usually conceived within a heterosexual marriage, which then underwent divorce or separation, leaving the child with a single parent.6 That parent then had at least one same-sex romantic relationship, sometimes outside of the child’s home, sometimes within it.

To be more specific, among the respondents who said their mother had a same-sex romantic relationship, 91% reported living with their mother while she was in the relationship, and fewer (57%) said they had lived with both their mother and her partner for at least four months at some point prior to age 18. An even smaller share (23%) said they had spent at least three years living in the same household with their mother’s romantic partner. This is to say that out of 2,988 respondents, only 40 children reported living with two lesbian women for three years or more, which is not a long time. Only 2 out of the 15,000 screened spent a span of 18 years with the same-sex relationship spent a span of 18 years with the same two mothers. Among those who said their father had had a same-sex relationship, 42% reported living with him while he was in the relationship, and only 24% reported living with him and his partner for at least four months. Only 1.1% of children whose father had a same-sex relationship spent at least three years together in the same household with both gay men.7

This strongly suggests that the parents’ same-sex relationships were often short-lived, a finding consistent with the broader research on elevated levels of instability among same-sex romantic partners. For example, a recent 2012 study of same-sex couples in Great Britain finds that gay and lesbian cohabiting couples are more likely to separate than heterosexual couples.8 A 2006 study of same sex marriages in Norway and Sweden found that “divorce risk levels are considerably higher in same-sex marriages”9 such that Swedish lesbian couples are more than three times as likely to divorce as heterosexual couples, and Swedish gay couples are 1.35 times more likely to divorce (net of controls). Indeed, sociologists Timothy Biblarz and Judith Stacey, two of the most outspoken advocates for same-sex marriage in the U.S. academy, acknowledge that “preliminary data hint that their [lesbian] relationships may prove less durable.”10

Therefore, although it would have been helpful to compare the children of IBFs to the children of committed and intact gay or lesbian couples, this was attempted, but was not feasible. Despite drawing from a large, representative sample of the U.S. population and despite using screening tactics designed to boost the number of respondents who reported having had a parent in a same-sex relationship, a very small segment reported to have been parented by the same two gay or lesbian parents for three years or more, an insufficient number to make reliable comparisons between these groups and IBFs.11 Although there is much speculation that today there are large numbers of same-sex couples in the U.S. who are providing a stable, long-term parenting relationship for their children, no studies based upon large, random samples of the U.S. population have been published that suggest this to be true, and the above-cited studies of different nations suggest that on average, same-sex couple relationships are more short-lived than those of opposite-sex couples.

Despite the lack of empirical evidence for the claim that today there are large numbers of stable, two-parent gay households, for the last ten years, contemporary gay parenting research has nevertheless claimed that there are “no significant differences” (and some benefits) to being raised by same-sex parents. Therefore, Regnerus analyzed the new NFSS data to verify this claim. In the end, he found the claim to be more plausible when comparing the grown children of parents who had a same-sex relationship to the grown children of divorced, adopted, single-parented, or step-parented arrangements. The data suggest that the claim is false if one compares the grown children of a parent who had a same-sex relationship to those from IBFs. While the study has been criticized for “comparing apples to oranges,” Regnerus’s work studies the reality of the population of children who were raised by parents who had same-sex relationships. As the next sections illustrate, there were clear and, in most cases, very unfortunate differences between the children of parents who had a same-sex relationship and those from biological families of still-married parents.

Findings on Differences in Social Outcomes12

Public perceptions and stereotypes of children of gays and lesbians usually assume them to be white, upper-middle-class members of society. However, in response to questions about race, 48% of the respondents with an FGR, and 43% of the respondents with an MLR indicated that they were either black or Hispanic, a number much higher than previously suggested by studies based on convenience samples.13 On economic outcomes, grown children of an MLR were almost four times more likely to be currently on public assistance than the grown children of IBFs. As young adults, they were also 3.5 times more likely to be unemployed than the grown children of IBFs.

On criminal outcomes, the children of FGRs showed the greatest propensity to be involved in crime. They were, on average, more frequently arrested and pled guilty to more non-minor offenses than the young-adult children in any other category. The children of MLRs reported the second highest average frequency of involvement in crimes and arrests, and in both categories, the young-adult children of intact biological families reported the lowest average frequency of involvement in crimes or arrests.

Contrary to recent and widely circulated conclusions that there is no sexual victimization in lesbian households, the NFSS found that, when asked if they were ever touched sexually by a parent or an adult, the children of MLRs were eleven times more likely to say “yes” than the children from an IBF and the children of FGRs were three times more likely to say “yes.” The children of IBFs were the least likely of all family types to have ever been touched sexually: only 2% reported affirmatively (compared to 23% of MLRs who replied “yes”). When asked if they were ever forced to have sex against their will, the children of MLRs were the worst off again, four times more likely to say “yes” than the children of IBFs. The children of FGRs were three times more likely to have been forced to have sex than the children of IBFs. In percentages, 31% of MLRs said they had been forced to have sex, compared with 25% of FGRs and 8% of IBFs. These results are generally consistent with research on heterosexual families; for instance, a recent federal report showed that children in heterosexual families are least likely to be sexually, physically, or emotionally abused in an intact, biological, married family.14

Regarding physical health, when asked if they had ever had a sexually transmitted infection (STI), the young-adult children of FGRs were three times more likely to say “yes” than those of IBFs. Children of MLRs were two and a half times more likely to say “yes,” followed by the children of stepfamilies, who were two times more likely to have had an STI than children of IBFs. Children of IBFs and children from “other” family types were the least likely of all to have had an STI. When asked to report upon frequency of marijuana use, the young-adult children of divorced parents were the worst off, reporting to use marijuana on average one and a half times more frequently than children of IBFs; next came the children of MLRs, followed by the children of single parents, and the children of FGRs. The children adopted by strangers (people unrelated to them) and the children of IBFs reported least frequent marijuana use as young adults. When asked about frequency of smoking, the young-adult children of MLRs reported highest frequency, followed by the children of FGRs, and the children of IBFs ranking lowest frequency of all family-of-origin types.

Findings on Differences in Emotional and Mental Health

Respondents were asked to report their sentiment about their family experiences while growing up. The children of MLRs reported the lowest levels of perceived safety in their childhood home, followed by children of FGRs, with the children of IBFs reporting the highest levels of perceived safety. When asked if they were recently or currently in therapy “for a problem connected with anxiety, depression, relationships, etc.,” the children adopted by strangers reported receiving such therapy the most, followed by the children of MLRs. The children from IBFs were least likely to report receiving therapy.

On the CES-D depression index, an eight-measure survey of respondents’ happy-to-depressed thoughts over the previous seven days, the young-adult children of MLRs and FGRs reported statistically significantly higher levels of depression than young-adult children from IBFs. The young-adult children of FGRs were two times more likely to have thought about suicide in the previous 12 months than the children of MLRs, and almost five times more likely than the children of IBFs to have thought about the same.

Findings on Differences in Relational Outcomes

The study asked questions about the history and current status of the young adults’ relationships. When asked to rate their current relationship quality, the children of FGRs reported the lowest quality, followed by children adopted by strangers, the children of stepfamilies, and then the children of MLRs. The children of IBFs reported the highest levels of relationship quality.

When asked about the number of times they thought that their current relationship was in trouble, the children of FGRs reported the highest numbers again, followed by the children of divorced parents. The children of IBFs reported that they deemed their relationship to be in trouble the least often.

When asked about infidelity, children of MLRs were three times more likely to say they had had an affair while married/cohabiting than children of IBFs, followed by children from stepfamilies (who were two and a half times more likely to have an affair than IBFs) and children of FGRs (who were twice as likely to say they had had an affair).

The NFSS asked respondents to identify their sexual orientation, and found that children of MLRs were more open to same-sex romantic relationships, bisexuality, and asexuality, than any other group. Daughters of MLRs reported an average of just over one female sex partner and four male sex partners in their lifetimes, in contrast to daughters of IBFs who reported an average of only 0.22 female sex partners and 2.79 male sex partners in their lifetimes. Daughters of MLRs were also most likely to self-report asexuality, “not sexually attracted to either males or females” (4.1% of females from lesbian mothers compared to 0.5% of females from IBFs). Children of FGRs were the next least likely to identify as fully heterosexual. Children from IBFs were most likely of all to identify as entirely heterosexual.


Taken together, the findings of the NFSS strongly suggest that there are differences between children raised by a parent who had a same-sex relationship and children raised in an intact, biological, married family when it comes to social, emotional, and relational outcomes. By drawing from a large, random sample of the American population rather than a small convenience sample, by interviewing the children in their young-adult years rather than their parents, and by comparing them to the children of intact, biological families rather than only to children of divorced, stepparent, adoptive, or single-parent families, the NFSS found important and wide-ranging differences between young adults raised by their own, biological, married parents and young adults who reported having a mother in a same-sex relationship, and to a lesser degree, those who reported having a father in the same.

On 25 out of 40 outcomes evaluated, there were statistically significant differences between children from IBFs and those of MLRs in many areas that are unambiguously suboptimal, such as receiving welfare, need for therapy, infidelity, STIs, sexual victimization, educational attainment, safety of the family of origin, depression, attachments and dependencies, marijuana use, frequency of smoking, and criminal behavior. On 11 out of 40 outcomes, there were statistically significant differences between children from IBFs and those who reported having an FGR in areas such as thoughts of suicide, STIs, being forced to have sex against their will, safety of the family of origin, depression, relationship quality, frequency of smoking, and criminal behavior. There were important differences in both comparisons, but the young-adult children of MLRs exhibited the least favorable outcomes in a wider array of categories when compared to the children of IBFs and fared worse in more categories than the children of FGRs. The “no differences” claim is unsound and ought to be replaced by an acknowledgement of difference.

Not only did the NFSS find clear differences between the children of IBFs on the one hand and MLRs and FGRs on the other, but it also suggests that there are notable differences within gay parenting arrangements themselves. Contrary to monolithic images of gay parents as committed couples who plan their family and adopt a child, the NFSS suggests that the children of MLRs and FGRs raised in the mid-1970s to 2000s were a diverse group: hailing from a range of racial and ethnic groups, and with parents moving in and out of relationships, experiencing a comparatively high level of relationship instability.

Acknowledging the differences between the children of IBFs and those from MLRs and FGRs better accords with the established body of social science over the last 25 years, which finds children to do best when they are raised by their married, biological mother and father.15 At the turn of the millennium, social scientists widely agreed that children raised by unmarried mothers, divorced parents, cohabiting parents, and step-parents fared worse than children raised by their still-married, biological parents.16 Although data on gay and lesbian parenting was not yet available for the reason that it was too rare to study adequately, it was difficult to imagine that gay and lesbian parents would be able to accomplish what heterosexuals in step-parenting, adoptive, single-parenting, and cohabiting contexts had not been able to do, namely, replicate the optimal child-rearing environment of married, biological-parent homes.

However, as early as 2001, social scientists working on sexual orientation and parenting began to claim just that, that there were not as many differences as sociologists would expect between outcomes for children in same-sex versus heterosexual unions, and that the differences were not negative, but favorable.17 Since then, an increase in gay parenting research over the last decade has made similar claims, so that the emergent message from social scientists working in gay parenting has gone in a different direction, to allege that there are no differences in child outcomes—and some advantages—to being raised by parents with same-sex behavior.18

By challenging these claims, the Regnerus paper, as well as the Marks paper summarized earlier, is consistent with the consensus that existed at the turn of the millennium: to be raised by an intact biological family presents clear advantages for children over other forms of parenting. In particular, the NFSS provides evidence that previous generations of social scientists were unable to gather--evidence suggesting that children from intact, biological families also outperform peers who were raised in homes of a parent who had same-sex relationships. Therefore, these two new studies reaffirm—and strengthen—the conviction that the gold standard for raising children is still the intact, biological family.19


  1. L., Marks, “Same-sex parenting and children’s outcomes: A closer examination of the American Psychological Association’s brief on lesbian and gay parenting,” Social Science Research (2012), 748. Steven Nock, the late sociologist of the family at the University of Virginia, said comparable things in his affidavit for the Supreme Court of Ontario, in Halpern v. Canada, 2003. Stacey and Biblarz’s affidavit criticized Nock for arguing that there was no scientifically valid evidence for the claim that children with same-sex parents and children with heterosexual parents have equal outcomes.
  2. P. Amato, “The well-being of children with gay and lesbian parents,” Social Science Research (2012), 772. Eggebeen, who like Amato is critical of the paper by Mark Regnerus that studies the NFSS, nevertheless argues that the real importance of the paper is “the description of a new data set that offers significant advantages. Whether the New Family Structures Study has the possibility of unsettling previously settled questions depends in equal parts on the richness of the information collected, as well as the willingness of scholars to make use of these data” (D. Eggebeen, “What can we learn from children raised by gay or lesbian parents?”, Social Science Research (2012), 777)
  3. Fiona Tasker (2010: 36) warns against this: “Parental self-report, of course, may be biased. It is plausible to argue that, in a prejudiced social climate, lesbian and gay parents may have more at stake in presenting a positive picture….Future studies need to consider using additional sophisticated measures to rule out potential biases….” Cited in Regnerus, 3 (also citing Bos and Sandfort, 2010; Brewaeys et al., 1997).
  4. Regnerus, 753, 755, citing Nock, 2001; Perrin and Committee on Psychosocial Aspects of Child and Family Health, 2002; Redding, 2008.
  5. Regnerus, 2.
  6. Regnerus reports, “just under half of [MLR and FGR] respondents reported that their biological parents were once married. This distinguishes NFSS from numerous studies that have been entirely concerned with ‘planned’ gay and lesbian families, like the NLLFS” (757). The claim that today’s gay and lesbian couples are more likely to plan for children using IVF, surrogacy, or adoption has not been confirmed yet with data. Moreover, because IVF is expensive, it is usually restricted to persons from the upper-middle class. The NFSS suggests that children who were raised by a parent who had a same-sex relationship often came from economically disadvantaged backgrounds. See Regnerus, Table 2 scores on “Family received welfare growing up.”
  7. Regnerus, 757.
  8. Strohm, Charles Q., “The Stability of Same-Sex Cohabitation, Different-Sex Cohabitation, and Marriage.” California Center for Population Research, UCLA, 1 Feb 2012 .
  9. Andersson, Noack, Seierstad and Weedon-Fekjaer, “The Demographics of Same-Sex Marriages in Norway and Sweden.” Demography, Volume 43 Feb 2006: 79-98. “We found that divorce risks are higher in same-sex partnerships than opposite-sex marriages and that unions of lesbians are considerably less stable, or more dynamic, than unions of gay men….In Norway, 13% of partnerships of men and 21% of female partnerships are likely to end in divorce within six years from partnership registration. In Sweden, 20% of male partnerships and 30% of female marriages are likely to end in divorce within five years of partnership formation. These levels are higher than the corresponding 13% of heterosexual marriages that end in divorce within five years in Sweden.” p. 95.
  10. The demographics of same-sex marriages in Norway and Sweden, Gunnar Andersson, Turid Noack, Ane Seierstad and Harald Weedon-Fekjær 2006. And Bibliarz and Stacey, 2010: 11: “preliminary data hint that their relationships may prove less durable.” Journal of Marriage and Family.
  11. Regnerus, 756.
  12. All outcomes reported in this summary may be found in Tables 2-4 of Dr. Regnerus’s article.
  13. Regnerus, 757.
  15. Sara McLanahan and Gary Sandefur, Growing Up with a Single Parent (Cambridge: Harvard University Press, 1994). Sara McLanahan, “Parent Absence or Poverty: Which Matters More?” (1994). G. Duncan and J. Brooks-Gunn, Consequences of Growing up Poor. (New York: Russell Sage). Marquardt and David Popenoe, Life Without Father (Cambridge: Harvard University Press, 1996). Bruce Ellis et al., “Does Father Absence Place Daughters at Special Risk for Early Sexual Activity and Teenage Pregnancy?” Child Development 74 (2003), 801-821. Sara McLanahan, Elisabeth Donahue, and Ron Haskins, “Introducing the Issue,” The Future of Children 15 (2003), 3-12. Mary Parke, “Are Married Parents Really Better for Children?” (Washington, DC: Center for Law and Social Policy, 2003). Elizabeth Marquardt, “Family Structure and Children’s Educational Outcomes” (New York: Institute for American Values, 2005). Wilcox et al. (2005). Elizabeth Marquardt, Between Two Worlds: The Inner Lives of Children of Divorce (New York: Crown, 2005).
  16. Child Trends, a nonpartisan research organization, summarized the upsurge of scholarly consensus by noting that “[r]esearch clearly demonstrates that family structure matters for children, and the family structure that helps children the most is the family headed by two biological parents in a low-conflict marriage.” Kristin Anderson Moore, Susan M. Jekielek and Carol Emig, “Marriage from a Child’s Perspective: How Does Family Structure Affect Children, and What Can Be Done About It?” Research Brief, (Washington, DC: Child Trends, June 2002).
  17. Likewise, Sara McLanahan, a Princeton sociologist, and Gary Sandefur, a Wisconsin sociologist, wrote that “[i]f we were asked to design a system for making sure that children’s basic needs were met, we would probably come up with something quite similar to the two-parent ideal. Such a design, in theory, would not only ensure that children had access to the time and money of two adults, it also would provide a system of checks and balances that promoted quality parenting. The fact that both parents have a biological connection to the child would increase the likelihood that the parents would identify with the child and be willing to sacrifice for that child, and it would reduce the likelihood that either parent would abuse the child.” McLanahan and Sandefur (1994), 38.
  18. Judith Stacey and Tim Biblarz, 2001 article in American Sociological Review.
  19. Tasker, 2005; Wainright and Patterson, 2006; Rosenfeld, 2010.
  20. “[C]hildren appear most apt to succeed as adults—on multiple counts and across a variety of domains—when they spend their entire childhood with their married mother and father, and especially when their parents remain married to the present day. Insofar as the number of intact, biological mother/father families continues to shrink in the United States, as it has, this portends growing challenges within families, but also heightened dependence on public health organizations, federal and state public assistance, psychotherapeutic resources, substance use programs, and the criminal justice system.” Regnerus, 766.