Same-sex parenting and same-sex marriage

A key part of the campaign for same-sex marriage is the claim that homosexual parents are just as good as the heterosexual counterpart and that the children grow up as happy, healthy and well-adjusted as anyone could want. Such claims are based on reviews of the research literature undertaken by, among others, the Australian Psychological Society (APS) and American Psychological Association (APA). Thus, according to the APA, all studies show that the children of same-sex parents are not disadvantaged to any significant degree compared to other children (APA 2005). The APS goes further, even claiming that “research has not only indicated that the outcomes of children of same-sex parents are notpoorer, but that outcomes would seem to be … at least as favourable” (APS 2007). That is, in some respects, same-sex parents do a better job.

Such conclusions are neither as unanimous nor as definitive as these reviews suggest. To make a properly informed choice about same-sex marriage (SSM), people need to be aware of the shortcomings. Moreover, while the focus on parenting is important, it does narrow the focus, diverting attention from the broader social functions of marriage.

The parenting studies in question employ standard, uncontroversial methods, few avoid all the pitfalls to which such research is prone. Indeed, the APA review takes pains to describe the problems found thus far in the research and the lessons it alleges researchers have learnt from earlier mistakes. Having said that, however, the APA then makes a curious statement:

It is significant that, even taking into account all the questions and/or limitations that may characterize research in this area, none of the published research suggests conclusions different from that which will be summarized below.

The summary referred to in the quotation concludes that ‘Not a single study has found children of lesbian or gay parents to be disadvantaged in any significant respect relative to children of heterosexual parents’. In other words, the APA is saying that, regardless of the quality of the research, it all reaches the same general conclusion, one that is favourable to same-sex parenting.

But how can this possibly be? How can the quality of the research have no tangible effect on the conclusions reached by the research? Indeed, how can it have no impact on the status of those conclusions? In any other field, the APA’s admission would be sufficient to place the entire body of findings under suspicion. And yet the APA writes as if this is a strength, as if it somehow proves the truth of the findings. A more impartial response would be to ask whether the researchers already had that conclusion in mind before the research was undertaken.

The suspicion grows when we look back to 2004 when the APA resolved, as an organisation, to oppose ‘any discrimination based on sexual orientation in matters of adoption, child custody and visitation, foster care, and reproductive health services’. This resolution arose from the claim that there is

…no scientific evidence that parenting effectiveness is related to parental sexual orientation: lesbian and gay parents are as likely as heterosexual parents to provide supportive and healthy environments for their children.

Two observations are in order: first, this claim, which underpins the APA policy, is not actually true. As a matter of fact, there is plenty of research, some undertaken since 2004, that finds a connection between parenting effectiveness and sexual orientation. Anyone who searches the web can find it. But note that the claim specifically alleges the absence of ‘scientific evidence’ linking parenting effectiveness to sexual orientation. That phrase introduces both authority and ambiguity, a combination equally potent and opaque that can mislead general readers and scholars alike.

Without getting into a debate about the nature of science, can we specify what the APA (or anyone) means when it describes evidence as ‘unscientific’ (or for that matter, as ‘scientific’)? Does it means that the methods employed are rigorous? Probably – all social science methods are designed to reduce the incidence of error. Yet the research the APA rejects uses the exact same methods as the research it accepts. The only difference seems to be the results produced by that ‘unscientific’ research. So perhaps it mean the methods are applied in a rigorous manner. Again, probably – but how is this sort of rigor measured? Conclusion are often rejected because flaws are detected in the research design. Fair enough, but the APA’s review openly admits that a lot of the research it relies upon has the same flaws. Indeed, look closely enough and ‘flaws’ can be found in all social research – nothing is perfect. So, is it a matter of degree – the more flawed research is rejected? Perhaps – but how much is too much? Where is the threshold? Who decides?

The APA does not specify what it means by ‘scientific’. We might say its use of the word ‘scientific’ is itself very ‘unscientific’, rhetorical rather than precise. In the APA review, the word ‘science’ is little more than a boo-word for describing evidence that agrees with the APA’s 2004 policy. As already indicated, there is plenty of ‘scientific’ evidence unfavourable to SSP but the APA rejects it as ‘unscientific’. That research includes articles by Paul Amato, Mark Regnerus, Loren Marks and many other, all as qualified and as experience as those whose research the APA favours.

Here is the imbalance: any study that concludes same-sex parenting is problematic, is examined closely for errors or limitations, and these “shortcomings” then explain why these studies are dismissed. Of course, entirely “perfect” research reaching utterly “true” conclusions does not exist. To varying degrees, all studies can be criticised on similar grounds. Yet, as the APA review confessed, studies that reach conclusions favourable to same-sex parenting are accepted despite their shortcomings. This suggests the research is being evaluated according to the conclusions rather than how those conclusions are reached.