Study Debunks Claim that Kids of Same-Sex Parents Do Less Well

Anne here…

The Southern Poverty Law Center alerts us to a new study out of Australia about children growing up in households led by same-sex parents:

In the latest blow to anti-gay forces, a major study conducted at the University of Melbourne in Australia has found that children of same-sex parents do as well or better than children raised by heterosexual parents. The Australian Study of Child Health in Same-Sex Families collected data on 500 children up to the age of 17 who grew up in gay or lesbian households across the nation.

An interim report on the study released earlier this week noted no statistical difference between children of same-sex couples and the population at large on such indicators as self-esteem, emotional behavior, and time spent with parents. Children of same-sex parents, however, scored higher than the national average for overall health and family cohesion.

Dr. Simon Crouch, the study’s lead researcher, theorized that because same-sex families have to deal with homophobia and bullying, they may be more willing to communicate about these and other issues, resulting in closer families.

Unsurprisingly, University of Texas sociology professor Mark Regnerus, author of a widely discredited study that claims children actually do worse in same-sex families, expressed doubt about the Australian study’s sample. He criticized the study for using what’s called a convenience, or “snowball,” sample to find its same-sex families, like reaching out to same-sex parenting E-mail lists or organizations, something that Regnerus said will bias a study toward positive results. He went on to say that “nonrandom samples are not a representative reflection of the population as a whole.”

However, Regnerus’s study, which has been trumpeted by the anti-gay right as proof that LGBT people are a danger to children, used a marketing firm (which paid participants) to gather his data, something for which he has been criticized. And his sample failed to include children raised in self-identified same-sex households, as the Australian study did. Only three people in Regnerus’ sample said they’d lived with a parent who was involved in a long-term same-sex relationship. Regnerus has publicly admitted that he didn’t know if the parents of those people self-identified as lesbian or gay and he has also stated that his data didn’t include the number or variety of people with a gay parent that he would have liked. Regnerus claims that his study is not about saying gay or lesbian parents are inherently bad. Nevertheless, anti-gay groups continue to use it as a weapon against marriage equality and LGBT people and Regnerus continues to court those same groups.

The Australian study could have wide-reaching effects in that country and beyond. Former Australian Prime Minister Kevin Rudd recently switched his position to support marriage equality, explaining that concern about the welfare of children of same-sex couples had been the primary obstacle for him earlier. Although the Australian Senate rejected a marriage equality bill last September, the issue is expected to come up again after this September’s general elections.


Now that more same-sex parents are raising more children, more studies like this one will be possible. Data good!

About Anne

Writer. Voracious reader. Lawyer. Jack’s mom. Irreverent. Coffee drinker. Cat owner. Grudging dog owner. Chief cook and bottle washer. Over-educated. Irish-Italian. Irreligious. History buff. Paleontology freak. Science fiction fan. Political junkie. Part-time avenging angel. Tea lover. Music nut. Tale spinner. Movie addict. Opinionated. Wordy.

Got a legal question related to religion? Contact me at

  • Mackinz

    We can try all we like to bring facts to the light, but the fundies will plug their ears and regurgitate the same arguments endlessly.

  • Emmet

    So is there anything wrong with using a “snowball” sample or not?

    • Zinc Avenger

      Snowball sampling has problems, such as the fact that the initial sample may bias the nature of the set.

      However, snowball sampling is often used because the population under investigation is hard to approach either due to small numbers of potential participants which are hard to locate among a population that has no bearing on the question, or the low likelihood of response if the participants are not referred by people they trust, both of which are criteria applicable in this case.

      Do you have an alternative methodology you’d like to share? Or would you just like to ask some more leading questions?