Children raised by same-sex couples – a review of the evidence

What is this?

This page contains a developing summary of the more recent evidence for the outcomes of children raised by same sex couples. I have written a companion piece and a blog post on the narrower issue of children adopted by same-sex couples.  I will take each of the recent published papers, and briefly summarize it, and then review it. This will feed into a blog post on the topic. For each paper I will give a reference, so that the interested reader can follow up on the materials reviewed. If you are aware of any recent papers that I ought to cover, please let me know – by email to or by phone on 086 606 9713.

There is one good, and quite indispensable, review. It is a book, by Abbie Goldberg – ‘Lesbian and Gay Parents and Their Children: Research on the Family Life Cycle. Washington, DC: American Psychological Association; 2010’. In it she summarizes the entire literature, up to about 2008. Amazon will sell you a copy at this link for only £ST 48.95. No legitimate open access version is available. If you are seriously interested in this topic, you will need a copy.

Goldberg has also published a more recent review – Goldberg AE, Gartrell NK. LGB-parent families: the current state of the research and directions for the future. Adv Child Dev Behav. 2014;46:57–88, parts of which are accessible here. The most accessible of the recent reviews is a report for the Williams Institute at UCLA, Goldberg A, Gartrell NK, Gates G. Research Report on LGB-Parent Families, which can be found here.

There is a good review, covering the whole range of studies, by Fiona Tasker ( “Lesbian Mothers, Gay Fathers, and Their Children: A Review.” Journal of Developmental and Behavioral Pediatrics: JDBP 26, no. 3 (June 2005): 224–40), which concludes “There is a variety of families headed by a lesbian or gay male parent or same-sex couple. Findings from research suggest that children with lesbian or gay parents are comparable with children with heterosexual parents on key psychosocial developmental outcomes. In many ways, children of lesbian or gay parents have similar experiences of family life compared with children in heterosexual families. Some special considerations apply to the context of lesbian and gay parenting: variation in family forms, children’s awareness of lesbian and gay relationships, heterosexism, and homophobia”.

There is a review of the older studies on this topic, which are not considered further here, in  Anderssen, Norman, Christine Amlie, and Erling André Ytterøy. “Outcomes for Children with Lesbian or Gay Parents. A Review of Studies from 1978 to 2000.” Scandinavian Journal of Psychology 43, no. 4 (September 2002): 335–51. This review concluded that ‘The studies indicate that children raised by lesbian women do not experience adverse outcomes compared with other children. The same holds for children raised by gay men, but more studies should be done.’

Finally, there is an excellent review in Future Child (Meezan, William, and Jonathan Rauch. “Gay Marriage, Same-Sex Parenting, and America’s Children.” The Future of Children / Center for the Future of Children, the David and Lucile Packard Foundation 15, no. 2 (2005): 97–115), which critically reviews the literature, sets it in a wider policy context, and makes intelligent suggestions for taking both the research and policy forwards.

The National Lesbian Family Study

This study was started by Gartrell et al. in 1986. Families with lesbian women intending to have children by donor insemination (DI) were recruited in San Francisco (39) Boston (37) and Washington DC (8). There were 84 families included, 70 couples, and 14 single women. The context of the study was that DI was a relatively new phenomenon at the time. Previous work on children raised by lesbians, had looked mostly at children born into heterosexual relationships, and now being raised in lesbian relationships. The earlier work, which is reviewed in the first of these papers, (and in the Goldberg book cited earlier) found no pertinent difference between these children and other American children.

1) Gartrell, N., J. Hamilton, A. Banks, D. Mosbacher, N. Reed, C. H. Sparks, and H. Bishop. “The National Lesbian Family Study: 1. Interviews with Prospective Mothers.” The American Journal of Orthopsychiatry 66, no. 2 (April 1996): 272–81.

This paper describes the design of the original study, and the content of interviews with the mothers and their partners. The women were mostly well educated, and moderately affluent. They were monogamous, and had been together for an average of six years. Sixteen (of the 70) couples already had one or more children. For many of these women the desire to have a child was seen as quite a basic part of their lives. These women had mostly good support form their families of origin. About half wanted the child to have knowledge of their biological father, and half of these wanted the child to have contact with him. Major concerns were about the risk of the child being discriminated against, because of the circumstances of their birth. The majority felt that it was important for good male role models to be involved with their children’s lives.

2) Gartrell, N., A. Banks, J. Hamilton, N. Reed, H. Bishop, and C. Rodas. “The National Lesbian Family Study: 2. Interviews with Mothers of Toddlers.” The American Journal of Orthopsychiatry 69, no. 3 (July 1999): 362–69

This paper describes the second suite of interviews, done when the children were aged 2 to 3. The focus here was on the mothers and their partners. They found that child rearing was both much harder, and much more rewarding than they had expected. The children were almost all developing normally, and had only minor health problems. Most, but not all couples, felt that were sharing equally in the burdens of parenting. There were impacts on the relationships between the parents, as one might expect. Co-mothers who were also the legal guardians of their child found this very positive.

3) Gartrell, N., A. Banks, N. Reed, J. Hamilton, C. Rodas, and A. Deck. “The National Lesbian Family Study: 3. Interviews with Mothers of Five-Year-Olds.” The American Journal of Orthopsychiatry 70, no. 4 (October 2000): 542–48.

This paper reports on the interviews conducted when the children were 5. At this age the mothers reported that 87% of the children related well to peers. For 63% their grandparents fully acknowledged the family. Most couples shared child-rearing tasks. Some children had experienced discrimination because of their parent’s relationship, and many had used role play to help support their children with this.

4) Gartrell, Nanette, Amalia Deck, Carla Rodas, Heidi Peyser, and Amy Banks. “The National Lesbian Family Study: 4. Interviews with the 10-Year-Old Children.” The American Journal of Orthopsychiatry 75, no. 4 (October 2005): 518–24. doi:10.1037/0002-9432.75.4.518.

This is the first report including data collected from the children themselves. The Child Behaviour Checklist (CBCL) was completed by their mothers. There were no notable differences to US norms, apart from a lower level of Externalizing behaviour problems in girls. Generally these children scored slightly better than the US norms. The children expressed positive views about the experience of having lesbian mothers. A majority of the children had experienced critical comments, usually from other children, about their parents.

5) Gartrell, Nanette, and Henny Bos. “US National Longitudinal Lesbian Family Study: Psychological Adjustment of 17-Year-Old Adolescents.” Pediatrics 126, no. 1 (July 1, 2010): 28–36. doi:10.1542/peds.2009-3153.

This is the second report based on questionnaires completed by the children themselves, and a questionnaire (the CBCL) about the child completed by their mothers. A total of 78 children took part. The short version – adolescents who have been reared in lesbian-mother families since birth demonstrate healthy psychological adjustment. The children were compared with a set of US norms for the CBCL, and, if anything, did slightly better than their peers, with higher scores for academic and social competence, and lower overall problem scores. There was no difference in scores between those who reported that they had experienced homophobic behaviour themselves (40% of the group), but those whose mothers reported it ~(30% of the group), had higher scores for anxiety/depression, and somatic symptoms.

6) Gartrell, Nanette, Henny M. W. Bos, Heidi Peyser, Amalia Deck, and Carla Rodas. “Adolescents with Lesbian Mothers Describe Their Own Lives.” Journal of Homosexuality 59, no. 9 (October 1, 2012): 1211–29. doi:10.1080/00918369.2012.720499.

This report is based on a more detailed analysis of the data collected from the children at the age of 17. Results revealed that the 17-year-old adolescents were academically successful in supportive school environments. They had active social networks, a wide range of academic and personal interests, and close family bonds. Nearly all considered their mothers good role models. The adolescents rated their overall wellbeing an average of 8.14 on a 10-point-maximum scale. Their academic achievement was substantial with a GPA of 3.6.

The AddHealth Study

This is a US study, which is a subset of a much larger study – AddHealth. This was  a school-based study of the health-related behaviours of US adolescents in Grades 7 to 12. Participating families were drawn from a large national sample of adolescents in the United States. A sample of 80 eligible high schools was initially selected. Schools were stratified to ensure that this sample   was representative of U.S. schools with respect to region of country, urbanicity, school type, ethnicity, and school size. More than 70% of the originally sampled high  schools were  recruited. If a high school refused to participate, a replacement school within its stratum was selected. Personnel at participating schools provided rosters of their students and, in most cases, agreed to administer an In-School Questionnaire during one class period. They also assisted in identifying their feeder schools (i.e., schools that include seventh grade and send their graduates to that high school). The final sample consisted of a pair of schools in each of 80 communities, with the exception of some high schools that spanned Grades 7 to 12 and therefore functioned as their own feeder schools. A sample of children who were selected to take part in the School based study, were also eligible for the In Home study, and a total of 12,105 were recruited across the US. There were questionnaires for both parents and children.

Offspring of same-sex couples were identified through a two-step process. First families in which parents reported being in a marriage or marriage-like relationship with a person of the same sex were identified. Because no data had been collected on parents’ sexual identities, per se, families headed by gay, bisexual, or lesbian parents who did not report that they were in a marriage or marriage-like relationship at the time of data collection could not be identified. This is an important weakness of this study.

In the second step, the consistency of parental reports about gender and family relationships was examined. To guard against the possibility that some families may have been misclassified because of coding errors, only cases in which parental reports of gender and family relationship were consistent (e.g., a parent reported being female and described her relationship to the target adolescent as ‘‘biological mother’’) were retained. Any families in which parental reports of gender and family relationships did not make sense or did not fit these criteria (e.g., a parent reported being female and described her relationship to the target adolescent as ‘‘biological father’’) were discarded. This procedure was designed to ensure that, insofar as possible, only adolescents whose parents reported being involved in a marriage or marriage-like relationship with a person of the same sex were selected for further study. (description is taken from paper 1).

There were only 6 households headed by a pair of men, and 44 headed by a pair of women. The number of homes headed by  gay men was too small for further analysis, although there were no evident differences between the child outcomes in these 6 families and those headed by gay women. All reports are based on the 44 families with a female couple, and a comparison group of 44 families headed by a man and a woman. By comparison with the National Lesbian Family Study, the families taking part in this study are much less selected.

The key message of these three studies is that the quality of parent–adolescent relationships better predicts adolescent outcomes than does family type.

1) Wainright, Jennifer L., Stephen T. Russell, and Charlotte J. Patterson. “Psychosocial Adjustment, School Outcomes, and Romantic Relationships of Adolescents with Same-Sex Parents.” Child Development 75, no. 6 (December 2004): 1886–98. doi:10.1111/j.1467-8624.2004.00823.x.

This paper reports on an analysis of a wide range of measures of psychosocial adjustment, school outcomes, and romantic attractions and behaviours among adolescents. There were no differences between the two groups of families. The relationship between family variables, and child outcomes were similar in the two groups of families.

2) Wainright, Jennifer L., and Charlotte J. Patterson. “Delinquency, Victimization, and Substance Use among Adolescents with Female Same-Sex Parents.” Journal of Family Psychology: JFP: Journal of the Division of Family Psychology of the American Psychological Association (Division 43) 20, no. 3 (September 2006): 526–30. doi:10.1037/0893-3200.20.3.526.

This study reports on a range of adverse behaviours, including smoking, alcohol and drug use, and early sexual activity, in the study participants. There were no material differences between the two groups.

3) Wainright, Jennifer L., and Charlotte J. Patterson. “Peer Relations among Adolescents with Female Same-Sex Parents.” Developmental Psychology 44, no. 1 (January 2008): 117–26. doi:10.1037/0012-1649.44.1.117.

This study reports specifically on the quality of peer relationships of the adolescents in the study. There were no differences between the two groups. Peer relationships were linked with parental relationships, but the link was similar in the two groups.

California donor insemination study

Eighty families were recruited from users of one sperm bank in California. 55 of these were headed by  lesbian, and 25 by heterosexual parents. Children were born before July 1990. The average age of the children at the time of the study was 7. 108 of 195 eligible families were contacted and 80 agreed to take part. The CBCL, and its teacher completed counterpart, the Teacher Report Form (TRF) were completed for each child.

1) Chan, R. W., B. Raboy, and C. J. Patterson. “Psychosocial Adjustment among Children Conceived via Donor Insemination by Lesbian and Heterosexual Mothers.” Child Development 69, no. 2 (April 1998): 443–57.

There were no statistically significant difference between children in heterosexual families and same-sex families on either the parent-reported or teacher-reported behaviour scores. Both sets of parents were similar on scores of parental adjustment. As expected there was a correlation between the biological mother’s report of parental stress, and the child’s behaviour score.

Brussels-Leiden donor insemination study

This is a small study of a cohort of children, 30 with lesbian mothers (Brussels), and 38 with heterosexual mothers, all conceived by donor insemination, and 30 naturally conceived children, also with heterosexual parents (both of these groups came from Leiden). There are baseline differences between the three groups.

1) Brewaeys, A., I. Ponjaert, E. V. Van Hall, and S. Golombok. 1997. “Donor Insemination: Child Development and Family Functioning in Lesbian Mother Families.” Human Reproduction 12 (6): 1349–59. doi:10.1093/humrep/12.6.1349.

This is the first report from the cohort. The children were aged between four and 8. There were differences in CBCL scores, with higher scores for those children conceived by DI from heterosexual families. There were no differences in the child’s views of their relationships with their parents. Co-mothers were more involved with the children, than either set of fathers. Parents were recruited through personal contacts of the researchers, and from a maternity hospital.

The US Adult outcome study

This is a study of 91 adults aged from 18 to 62, recruited via four methods: (1) snowball sampling of personal acquaintances, (2) advertisements to gay, lesbian, and bisexual family support groups in person and on the Internet, (3) community advertisements in restaurants and shops with prominently gay or lesbian clientèle, and (4) a pool of introductory psychology students. Recruitment materials targeted adults who grew up with at least one parent who openly identified as gay, lesbian, bisexual, or queer. In the final sample, most participants reported hearing about the study from a friend or family member (53%), while others heard about it from a gay, lesbian, or bisexual family organization (31%), a university participant pool (6%), or other means (11%). Data was collected from eligible participants on-line.

Lick, David J, Charlotte J Patterson, and Karen M Schmidt. “Recalled Social Experiences and Current Psychological Adjustment among Adults Reared by Gay and Lesbian Parents.” Journal of GLBT Family Studies 9 (2013): 230–53.

There was no comparison group in this study, but as almost no studies have looked at older people raised in gay/lesbian families, it is of some interest. The analysis was within the sample, looking at differences in social and psychological adjustment by age, gender, family type, and  the age at which they learned about their family. Despite great diversity of experience, there was little difference in long-term psychological adjustment between the various subgroups.

The UK cohort study

This is a study of 30 lesbian mother families, 42 families headed by a single heterosexual mother and 41 two-parent heterosexual families. The lesbian mothers were recruited through contacts by the researchers, the heterosexual single mothers by press advertising, and the heterosexual couples from a local maternity unit. The aim was to investigate the quality of parent-child relationships, and study the psychological and emotional development of the children.

1) Golombok, S., F. Tasker, and C. Murray. 1997. “Children Raised in Fatherless Families from Infancy: Family Relationships and the Socioemotional Development of Children of Lesbian and Single Heterosexual Mothers.” Journal of Child Psychology and Psychiatry, and Allied Disciplines 38 (7): 783–91.

The children in this study were aged between 3 and 9. The results show that children raised in fatherless families from infancy experienced greater warmth and interaction with their mother, and were more securely attached to her, although they perceived themselves to be less cognitively and physically competent than their peers from father-present families. No differences were identified between families headed by lesbian and single heterosexual mothers, except for greater mother-child interaction in lesbian mother families.

2) MacCallum, Fiona, and Susan Golombok. 2004. “Children Raised in Fatherless Families from Infancy: A Follow-up of Children of Lesbian and Single Heterosexual Mothers at Early Adolescence: Children Raised in Fatherless Families from Infancy.” Journal of Child Psychology and Psychiatry 45 (8): 1407–19. doi:10.1111/j.1469-7610.2004.00324.x.

By the time of this study, the children were aged between 11 and 13. Twenty-five lesbian mother families and 38 families headed by a single heterosexual mother and 38 two-parent heterosexual families were followed up. Children in fatherless families experienced more interaction with their mother, and perceived her as more available and dependable than their peers from father-present homes. However, there were no group differences in maternal warmth towards the children. Mothers raising their child without a father reported more severe disputes with their child than did mothers in father-present families. The children’s social and emotional development was not negatively affected by the absence of a father, although boys in father-absent families showed more feminine, but no less masculine characteristics of gender role behaviour. No major differences in parenting or child development were identified between families headed by lesbian and single heterosexual mothers.

3) Golombok, Susan, and Shirlene Badger. 2010. “Children Raised in Mother-Headed Families from Infancy: A Follow-up of Children of Lesbian and Single Heterosexual Mothers, at Early Adulthood.” Human Reproduction 25 (1): 150–57. doi:10.1093/humrep/dep345.

At the time of this study the children were aged 18 to 20 years. Twenty-seven families headed by single heterosexual mothers (solo mothers) and 20 families headed by lesbian mothers  were compared with 36 two-parent heterosexual families as the child entered adulthood. The female-headed families were found to be similar to the traditional families on a range of measures of quality of parenting and young adults’ psychological adjustment. Where differences were identified between family types, these pointed to more positive family relationships and greater psychological well-being among young adults raised in female-headed homes. Children raised by solo heterosexual mothers or lesbian mothers from infancy continue to function well as they enter adulthood.

The Dutch Lesbian Planned Family Study.

This is a cohort study of 100 families headed by a lesbian couple, and 100 headed by a straight couple. The two groups were recruited separately, but each had to have at least one child aged between 4 and 8.

For the lesbian headed families recruitment was carried out using four different strategies. We used the patients’ files of the Medical Centre for Birth Control (MCBC), a centre providing assisted reproductive services to clients regardless of sexual orientation or having a relationship. In addition, families were recruited with the help of the largest interest group for gay and lesbian parents. Furthermore, lesbian families were selected with the help of individuals with expertise in the area of gay and lesbian parenting. We also placed an advertisement in a lesbian magazine. A letter of invitation was sent to 178 lesbian families. These families were selected because they met our criteria for participation. Of the 178 lesbian families contacted, 43 came from the MCBC, 60 from the interest group and 75 from experts in the area of gay and lesbian parenting. The total response rate was 99 (55.6%), for the medical centre 18 (41.9%), for the interest group 47 (78.3%) and for the experts 34 (45.3%). One family responded to the advertisement.

The group of heterosexual-parent families was drawn from the population register of two cities that have a level of urbanization comparable to the cities in which the participating lesbian-parent families lived, as well as through schools and referrals from the participating lesbian-parent family group. The response rate for the population registry office was 42 (17.3%), 49 (24.1%) for the schools, and 9 (38.7%) for the referrals. The total response rate was 251 (21.4%). Of these 251 families, 100 were matched with the lesbian-mother families, based on the degree of urbanization and on age and gender of the target child. There were slightly fewer children, on average, in the lesbian couple headed households (1.87) than in the heterosexual couple households (2.03).  (Description drawn from paper 1 and paper 2).

In the study each parent completed a separate questionnaire, covering a range of demographic and social issues, and the CBCL was completed for the target child. Parents were provided with diaries of household activities to fill in, and a home visit was done, including videotaping each parent interacting with the child separately.

1) Bos, H. M. W., F. van Balen, D. C. van den Boom, and Th G. M. Sandfort. “Minority Stress, Experience of Parenthood and Child Adjustment in Lesbian Families.” Journal of Reproductive and Infant Psychology 22, no. 4 (November 1, 2004): 291–304. doi:10.1080/02646830412331298350.

This paper only reports on the lesbian headed families. The lesbian mothers in this sample generally described low levels of rejection, they perceived little stigmatization, and they also manifested low levels of internalized homophobia. However, minority stress was significantly related to experiences of parenthood. Lesbian mothers with more experiences of rejection experienced more parental stress, and appeared to defend their position as mother more strongly (e.g. parental justification). Furthermore, mothers with higher levels of perceived stigma and internalized homophobia felt significantly more often that they had to defend their position as mother. Finally, mothers who reported more experience of rejection were also more likely to report behaviour problems in their children (as measured by the CBCL).

2) Bos, Henny M. W., Frank van Balen, and Dymphna C. van den Boom. “Child Adjustment and Parenting in Planned Lesbian-Parent Families.” The American Journal of Orthopsychiatry 77, no. 1 (January 2007): 38–48. doi:10.1037/0002-9432.77.1.38.

This paper is a comparative study of the two groups of families. The main outcome is child behaviour as measured by the CBCL. There are no differences between the two family types. Major differences are found, as expected, between boys (higher) and girls (lower). Of more interest, a range of parental variables were associated with better child behaviour scores – including satisfaction with their partner as a co-parent, and the emotional involvement of the parent with the child, and their supportive involvement with the child.

ACHESS – The Australian Study of Child Health in Same-Sex Families

So far, I have only found two published papers, one describing the methods to be used in this project, and one giving the first set of results. More are expected. the study is intended to be a cohort study. Initial recruitment will involve convenience sampling and snowball recruitment techniques that have been successful in other survey-based Australian studies of same-sex attracted populations including the Work, Love, Play Study and the Lesbian and Gay Families Study. This will include advertisements and media releases in gay and lesbian press, flyers at gay and lesbian social and support groups, and investigator attendance at gay and lesbian community events. Discussion pieces and interviews with mainstream media outlets will help target families not engaged with the gay and lesbian community, as well as rural and remote families. Primarily recruitment will be through emails posted on gay and lesbian community email lists aimed at same-sex parenting. This will include, but not be limited to, Gay Dads Australia and the Rainbow Families Council of Victoria. Any parent over the age of 18 years, who self-identifies as being same-sex attracted, lives in Australia, and has children under 18 years of age will be eligible to participate in the study. Children aged ten years or over will also be asked to complete a questionnaire. 390 parents were identified, and invited to take part in the study. A total of 315 parents, with 500 children, responded.

1) Crouch, Simon, Elizabeth Waters, Ruth McNair, Jennifer Power, and Elise Davis. “ACHESS – The Australian Study of Child Health in Same-Sex Families: Background Research, Design and Methodology.” BMC Public Health 12, no. 1 (2012): 646. doi:10.1186/1471-2458-12-646.

This paper describes the intended design of the study in detail.

2) Crouch, Simon R, Elizabeth Waters, Ruth McNair, Jennifer Power, and Elise Davis. “Parent-Reported Measures of Child Health and Wellbeing in Same-Sex Parent Families: A Cross-Sectional Survey.” BMC Public Health 14 (June 21, 2014): 635. doi:10.1186/1471-2458-14-635.

The 500 children recruited had a median age of 4 years. The parents had higher levels of education and income than the general population. Across a range of child health measures the outcomes for the children were similar to those for contemporaneous controls living in Victoria, after adjustment for a range of social variables. There was evidence for better outcomes in the study group on the Child Health questionnaire (used for the older children)  in the domains of  general behaviour, general health and family cohesion. Within the study group, higher levels of perceived stigma reported by the parents was associated with lower scores on several health measures.

Australian three contexts study

This is an older study, outside my time horizon, which I include because it’s results are different to those of most of the other studies. I have not (yet) been able to get a copy of it. I base my account on a detailed review of the paper by Herek (UC Davis Law Review 48, no. 2 (December 2014): 583–622.). Herek describes other studies which I have been able to read, and describes them accurately. Pending securing a copy myself, I will use his account. The sample consisted of 174 Australian children of primary school age who were living in households headed by one of three different types of couples: married heterosexual parents, cohabiting heterosexual parents, or cohabiting lesbian and gay parents (n= 58 children in each group). The children were matched on age, sex, and year in school, as well as their parents’ occupation, employment status, and levels of education. Children were assessed primarily by their teachers, on  multiple variables related to school and classroom performance, social involvement, and personality characteristics.

Sarantakos, Sotirios. “Children in Three Contexts: Family, Educational, and Social Development.” Child Australia 21 (1996): 23–31.

The study’s main findings were that children of homosexual couples  were rated significantly lower than other children on most variables: language skills, mathematical abilities, sports, sociability and popularity, attitudes toward school and learning, and parents’ school involvement. These children were not born during the relationships studied. They were born in a previous relationship (presumably a heterosexual relationship), and brought into the new relationship, after a breakdown of the earlier relationship. None of the children of heterosexual couples had had this experience. There was also a high reported prevalence of homophobic bullying, reported to be notably more severe than in other studies I have read. It is not clear to what extent the teachers themselves were prejudiced against these children.

New Family Structures Study

This is a study based on retrospective data collected from a panel survey of a quota sample of 3,000 young adults in the US. The group studied were those who reported that either their father (73) or their mother (163) had had a same-sex relationship. The hypothesis tested was that here were no differences between the outcomes of young adults from such households, and those in the comparison group.

Regnerus, Mark. 2012a. “How Different Are the Adult Children of Parents Who Have Same-Sex Relationships? Findings from the New Family Structures Study.” Social Science Research 41 (4): 752–70. doi:10.1016/j.ssresearch.2012.03.009.

———. 2012b. “Response to Paul Amato, David Eggebeen, and Cynthia Osborne.” Social Science Research 41 (4): 786–87. doi:10.1016/j.ssresearch.2012.05.003.

———. 2012c. “Parental Same-Sex Relationships, Family Instability, and Subsequent Life Outcomes for Adult Children: Answering Critics of the New Family Structures Study with Additional Analyses.” Social Science Research 41 (6): 1367–77. doi:10.1016/j.ssresearch.2012.08.015.

There were considerable differences, mostly adverse, between the lives of these young adults, and the remainder of the sample. The later two papers respond, briefly, and then at more length, to various detailed critiques of the first paper. The concluding paragraph of the first paper gives a good summary of its contents –

“Do children need a married mother and father to turn out well as adults? No, if we observe the many anecdotal accounts with which all Americans are familiar. Moreover, there are many cases in the NFSS where respondents have proven resilient and prevailed as adults in spite of numerous transitions, be they death, divorce, additional or diverse romantic partners, or remarriage. But the NFSS also 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, and especially when the parents remain married to the present day. Insofar as the share 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.”

Amato, Paul R. “The Well-Being of Children with Gay and Lesbian Parents.” Social Science Research 41, no. 4 (July 2012): 771–74. doi:10.1016/j.ssresearch.2012.04.007.

This is a brief response to the Regnerus paper, and a separate paper, not discussed here, by Loren Marks, criticizing an earlier APA statement on lesbian and gay parenting. It includes a very clear review of the methodological problems in all the literature to that date, including the studies discussed above. On the Regnerus paper, he finds that the effect of being in the study group, i.e. a child whose parent had had a same-sex relationship, was moderate (0.26 to 0.28 of a standard deviation across all measures), and similar to that of children whose parents had divorced for other reasons. Just over half of the Regnerus study sample had experienced divorce, so this result is not very surprising. He cautions against using the Regnerus data to make decisions on same-sex marriage, both on general policy and constitutional grounds, and because of the difficulty of interpreting the data.

Eggebeen, David J. “What Can We Learn from Studies of Children Raised by Gay or Lesbian Parents?” Social Science Research 41, no. 4 (July 2012): 775–78. doi:10.1016/j.ssresearch.2012.04.008.

Eggebeen also responds to the papers by Regnerus and Marks. He reviews the methodological issues raised by Marks,  and the great difficulty of overcoming these in feasible studies. He notes the small sample size of the Regnerus study, and calls for caution in interpreting it.

(The reference for the Marks paper is “Marks, Loren. “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 41, no. 4 (July 2012): 735–51. doi:10.1016/j.ssresearch.2012.03.006″.)

Perrin, Andrew J., Philip N. Cohen, and Neal Caren. 2013. “Are Children of Parents Who Had Same-Sex Relationships Disadvantaged? A Scientific Evaluation of the No-Differences Hypothesis.” Journal of Gay & Lesbian Mental Health 17 (3): 327–36. doi:10.1080/19359705.2013.772553.

This is a detailed critique of the Regnerus paper. In summary they argue that the sample was inadequately characterized. The way in which the study groups were defined is unreliable, and does not reflect any common understanding of a couple. Most of the families studied could not be described as containing a child raised in a household headed by a gay/lesbian couple. Many of these families started after a divorce, which is a well known risk factor for poorer long term outcomes amongst children. Overall. they do not accept that Regnerus’s conclusions are well founded.

 US and Canadian census sample studies

The studies described above have used a range of methods to identify households, headed by lesbian or gay couples, with children. All of these methods have weaknesses, and may lead to samples which are not representative of the whole population. These issues are well understood by workers in the area, and every study listed above discusses the issues of sampling. Two recent studies have used a different approach taking anonymized US and Canadian census records, respectively.

1) Rosenfeld, Michael J. “Nontraditional Families and Childhood Progress through School.” Demography 47, no. 3 (August 2010): 755–75.

This is an analysis of 3502 households with children, headed by cohabiting lesbian (2,030) or gay male (1,452) couples, who had been at the same address for at least five years, from the 2000 US Census micro-data. All of these children were biologically related to one or other of the adults in the household. Note that most of these households the children were the biological children of one parent,and their previous heterosexual partner. This is not the same population as that described in other US studies. The outcome variable used, grade retention, is defined by looking at the child’s age, and the grade (year) of school they are in. Grade was categorized as Grades 1-4 and 5-8, so grade retention can only be identified for children too old to be in Grade 4 (11 and over), and Grade 8 (15 and over) at the time of the census. The gay couples with children had slightly lower incomes than their heterosexual peers. Observed retention rates were 1.71% for heterosexual couples, 2.38% for lesbian couples, and 2.42% for gay male couples. (Actual retention rates are surely significantly higher, because of the collapsing of grades into 1-4 and 5-8). There was no evidence of a significant difference between the two groups of children in retention rates,  once demographic variables were taken into account.

2) Allen, Douglas W., Catherine Pakaluk, and Joseph Price. “Nontraditional Families and Childhood Progress through School: A Comment on Rosenfeld.” Demography 50, no. 3 (June 2013): 955–61. doi:10.1007/s13524-012-0169-x.

Allen et al. re-analysed Rosenfeld’s data, making three changes. First all analyses are presented with same sex couple households as the comparison group. Second the requirement that the family had been living at the same address is removed. Third the requirement that the child should be biologically related to one or other parent is removed. The latter two changes increase the total number of children studied very substantially. The result is that children in heterosexual households have significantly higher odds of making normal progress through school.

3) Rosenfeld, Michael J. “Nontraditional Families and Childhood Progress through School.” Demography 50, no. 3 (June 2013): 963–69. doi:10.1007/s13524-012-0170-4.

This is a reply to Allen’s re-analysis of the 200 Census data. In brief, Rosenfeld points out that the effect of relaxing the sample restrictions is to mix in a variety of family types, and (unknown) life histories for the children, which, in his judgement vitiates the comparison – in effect while purportedly studying oranges, they have tipped in a few baskets of apples as well.

To quote “Their [Allen et al.] finding of worse school performance by children living with same-sex couples is due to their conflating the initial disadvantage of children who come into same-sex couple families (a disadvantage that appears to be substantial) with the progress children experience during the time when they are actually being raised by same-sex couples (progress that is excellent).”

4) Allen, Douglas W. “High School Graduation Rates among Children of Same-Sex Households.” Review of Economics of the Household 11, no. 4 (September 26, 2013): 635–58. doi:10.1007/s11150-013-9220-y.

5) Allen, Douglas W. “Erratum to: High School Graduation Rates among Children of Same-Sex Households.” Review of Economics of the Household 12, no. 1 (November 27, 2013): 207–207. doi:10.1007/s11150-013-9231-8.

This study (and the reported error, which is worth looking at carefully) reports on an analysis of the 2006 Canadian census microdata. This is a 20% sample of the census, with about 2 million children in total. Children living with same sex parents were identified by a Statistics Canada created variable ‘‘RELATIONSHIP TO PERSON 1’’ which included ‘‘CHILD OF A MALE (FEMALE) SAME-SEX MARRIED OR COMMON LAW COUPLE’’ among the possible categories for that variable. Only children living with their parents were included. After weighting here were 423 children aged 17 to 22 living with gay parents, and 969 living with lesbian parents included. It is not clear how many households this represented. The original sample is probably around 85 with gay parents, and 194 with lesbian parents. Substantially lower graduation rates for children from same-sex households were reported, 6% and 9% lower for children in households with heterosexual married parents (graduation rates 72%).

No peer reviewed critiques of this work have been published, that I can find, but there is a detailed blog post, from a respected US social scientist, Philip Cohen.  Cohen identifies two key issues. First, the reported high school graduation rate is not quite what it seems. Allen at al. count those in the sample who have graduated to calculate a graduation rate, but that’s not what they have. What they have is the percentage of 17-22 year-olds (many of whom are still in high school) who have already graduated high school.

Second, the population studied is highly selected. Who are the people aged 17 to 22 *not* living at home? If the pattern of high school graduation in this group differs by family structure, which is unknown, but certainly possible, then the conclusions of the study cannot be maintained. So, if for example, gay and lesbian parents were more successful at launching their children from home after completing high school, in this analysis, that would be evidence of a bad family outcome. This is not a problem that can be solved with census data. Only longitudinal studies, like those reported above, can address this issue.