by Carolyn Moynihan. July 10, 2014.
“Kids from same-sex families fare as well as peers – or better” claims the heading of an article by an Australian academic repeated in the New York Times, Washington Post, the Guardian and many other media outlets this week. Children with same-sex attracted parents are “doing well in terms of their overall health”, their “families are getting along really well”, says Simon Crouch. Actually, they are doing “better” than average on those points, but they do “experience stigma”, which explains any “lower scores” on certain measures. So we have to get rid of that stigma somehow…
It’s a familiar story, and if this one seems particularly deja vu it is because we heard it all a year ago, when interim results from the Australian Study of Child Health in Same-Sex Families (ACHESS)were published with similar upbeat headlines. Dr Crouch, himself an openly homosexual man with twin boys aged about 5, and his University of Melbourne team, couldn’t wait to tell the good news to a public debating same-sex marriage and adoption and surrogacy.
Now, with elections looming in his home state of Victoria and same-sex adoption an issue, he is keen to show why that move would be justified. But, as many have pointed out before, the ACHESS study can provide no grounds for such a policy — or any other concerning same-sex couples and the children they want or already have.
It might be the “largest study of its kind”, but it is not the kind of study that can tell us how well children fare in the care of same-sex parents. That’s because it does not ask a representative sample of parents. It does not ask children independently of their parents. And it uses a method that is bound to show same-sex parents “on their best behaviour”.
What kind of study is ACHESS?
It is based on a self-selected sample of people who filled in questionnaires online. Some 315 same-sex attracted adults responsible for 500 children aged 0 to 17 completed the survey some of whom could also fill in a questionnaire. Ninety-one of the children had a male “index” (or main) parent and 400 had a female index parent. The remaining few identified as other genders.
The adults were recruited in the first place through the email lists of organisations such as Gay Dads Australia and the Rainbow Families Council of Victoria, and then through the “gay and lesbian press, flyers at gay and lesbian social and support groups, and investigator attendance at gay and lesbian community events …” “Snowballing” was also used, with first responders recruiting others.
It’s true that this produced more participants than in similar studies — in particular more males – but the size does not make it representative of same-sex families. It’s what is called in the trade a “convenience sample” – easy-to-find people and their friends.
Compare this with the New Family Structures Study by US sociology professor Mark Regnerus. The NFSS began with the random screening of over 15,000 Americans aged 18-39 to identify any who had experience of a parent in a same-sex relationship while growing up. This produced 175 who said this was true of their mothers and 73 said the same about their fathers. The key research questions were not revealed until the interviews with this group actually began.
ACHESS participants, by contrast, were well aware of the nature of the study and its political significance from the outset. The core volunteers came from organisations which, no doubt, have political as well as community agendas. Why would their accounts of their children’s welfare not suffer from what’s known as “social desirability bias” – putting the best spin on things?
How do we know about the children?
Remember, these studies are not about homosexual adults but about the wellbeing of the children they have in their households. To find out what the children experienced Dr Regnerus asked them after they had grown up. ACHESS asked the parents of mostly young children. Which method sounds more objective? They certainly had different results. (In both cases there were comparison groups.)
In ACHESS children aged 10 through 17 could fill out a questionnaire but only with parental consent, and in any case the data from them does not seem to be included in the current study. Note that the median age for the children of male parents in the study is 2, and of female parents, 4. So most of the children would not be old enough to make their own report, even if the parent consented. The authors say:
There is no evidence to suggest that any group of parents would systematically respond in a particular way on any given scale, however this cannot be discounted entirely. Future research will report on child-reported measures of health, as well as a contextual analysis of qualitative data drawn from family interviews, in order to draw out any bias that parental reporting might have.
A slightly less rosy picture would add credibility.
What ACHESS does tell us about the children
The most interesting and useful part of the Australian study is its demographic information about the children of the self-selected group. It is important to note, here, that we are getting a picture of children in relation to an “index parent”, that is, the main parent responsible for the child or children – not a couple. Over 90 percent of these parents were “in a relationship” at the time of the study, but we don’t know whether that means living under the same roof. A same-sex “family” could be a single female raising a male child—the sort of thing the NFSS study was criticised for.
Of the 91 children in the care of a male index parent, 46 were his biological child, 17 a partner’s biological child, a few were fostered and two were adopted. What “biological” means has to be interpreted in the light of these statistics: only 18 of the children were conceived through sexual intercourse, while 75 of the children were conceived using surrogacy (involving either artificial insemination or assisted reproductive technology) and 33 of those involved donors – 23 unknown.
Given that commercial surrogacy is illegal in most of Australia it is not surprising to find that 56 of these children were born overseas where it is legal – 36 in India and 20 in the US. This means legal parenthood is uncertain and explains why adoption is a big issue for same-sex males. It also means, almost certainly, that these children have no chance of knowing their biological mothers.
Of the children whose index parent is female, all but a handful were born in Australia; 64 percent were the index parent’s biological child and 20 percent a partner’s biological child. Here also “biological” means conceived through sexual intercourse only in 20 percent of cases, nearly all the rest involving “home insemination” or ART – the latter, mostly with an unknown sperm donor. For over a third of these children, then, their biological father is unknown.
While the NFSS study of grown children found that family instability was a major theme, especially in relation to a lesbian parent, a prominent claim of ACHESS is that “family cohesion” is significantly better (6 percent) than in the general population. However, for around 30 percent of the children in ACHESS the index parent is not in the same relationship as when the child was born, fostered or adopted. The current “family” may be “getting along really well” (at the moment) but, given that half of the children are still very young, that may not tell us very much about its stability.
In any case, is it really fair to compare the family cohesion of this tightly-knit group of specially recruited LGBT parents with the general population? Why not compare it to a “convenience sample” recruited at Mormon temples in Salt Lake City? That would make an interesting comparison.
Bottom line: any bad stats are due to stigma
The ACHESS report does hint at some not-so-good results but puts them down to “stigma”:
In spite of doing well, many children did experience stigma, which was linked to lower scores on a number of scales. Stigma can be subtle, such as letters home from school addressed to Mr and Mrs. Or it can be overt and very harmful, in the form of bullying and abuse at school. The more stigma the family experienced, the greater the impact on the social and emotional well-being of the children.
In other words, it is highly unlikely that anything could be wrong with the same-sex family itself; it is society that is at fault and must change. The main recommendation for further study concerns stigma.
So there you have it: a study that is “large” but not “strong” though the life chances of many children may be affected by it. As Mark Regnerus has pointed out, given what’s at stake, nothing but the effort to do a really strong study is good enough, although this requires hard work and, above all, expense.
Convenient and cheap may be good enough keywords for the family’s weekly shopping, but not for deciding the fate of children.