Daniel Potter on Gay Parenting

You wait ages for a study on gay parenting to come and then loads appear round the corner at once.

Daniel Potter has published a paperin the Journal of Marriage and Family examining the issue of gay parenting from the perspective of academic gain in early years education. By using data from the Early Childhood Longitudinal Study—Kindergarten cohort Potter was able to examine various educational gain during the first few years of a child’s school life and see if there were any differences related to family setup. It’s an interesting study and has challenging outcomes. Here’s Potter’s description of what he did.

In this study, I used a dynamic model of the family to explore the importance of the timing and accumulation of family structure transitions for shaping children’s academic achievement in same-sex parent households. Specifically, I addressed two research questions: (a) What is the association between living in a samesex parent family and children’s academic achievement as measured by their mathematics assessment scores and (b) how does this association compare with the outcomes of children from other nontraditional families? To answer these questions, I applied a hierarchical linear modeling (HLM) growth curve model to data from the kindergarten through eighth-grade waves of the Early Childhood Longitudinal Study—Kindergarten cohort (N = 19,043) and examined the association between living in a same-sex parent family (n = 158) and children’s math assessment scores.

Got that? The idea is to build models that look at family setup as impacts on child educational attainment.

The first models were simple regressions against outcome, looking at various input factors. Model 1 simply regressed family setup at kindergarten entry to mathematics outcome at eighth grade. The outcomes are very clear – every single family setup that is not the biological parents produces poorer outcomes for mathematics for their kids. Same-sex parents perform slightly worse then divorced parents or step-families but not as bad as single parents or co-habiting parents. Model 2 introduces factors such as age and sex of parent, race, parents educational attainment etc. Now the coefficients for family setups are not so powerful. In particular the same-sex family coefficient halves and moves out to a significance level ( 0.05 < p < 0.1) where we would probably reject it in a confidence test. Finally in the third model cumulative family transitions are included and suddenly most of the non “nuclear family” setups become insignificant.

What does this tell us? Potter describes the phenomenon perfectly.

Differences in academic achievement associated with living in traditional and non-traditional families are largely reflective of the transitions and changes that accompany the formation of such households and less clearly indicative of any inherent deficiencies in these family structures. On the basis of the findings presented herein, the academic achievement of children living in same-sex parent families conforms to this pattern, in that baseline disparities in children’s assessment scores were accounted for by the transitions that children experienced. These results contrast with those of earlier studies that found no evidence of worse performance by children in same-sex parent families, even at a baseline level (Bos et al., 2007; Gartrell & Bos, 2010; Wainright & Patterson, 2006, 2008; Wainright et al., 2004), but the dynamic model of the family used in this study aligns these results with other research on opposite-sex parent nontraditional families (Cavanagh et al., 2006; Fomby & Cherlin, 2007; Sun&Li, 2009, 2011). Same-sex parent families are often created through a series of changes to and transitions in children’s family structure; therefore, the view of the family incorporated into this study provides a more realistic account and reflection of the experiences of these children (Stacey, 2006; Telingator & Patterson, 2008). Nonetheless, not all same-sex parent families are created from dissolved opposite-sex relationships, and other research, in particular, the U.S. National Longitudinal Lesbian Family Study, continues to offer invaluable information regarding the outcomes of children in same-sex parent families and the benefit that a consistent family structure can have on the development of children regardless of parents’ sexual orientation (Gartrell & Bos).

These are challenging findings for conservatives but they need to be taken seriously. Any differences between nuclear families and same-sex families on mathematical achievement of children are down to family structure transitions and not the fact that the alternative family is gay. Kids in gay families still perform worse than kids in nuclear families, but they perform no worse then any other child in another kind of non-nuclear family. This finding is confirmed by repeating the models with same-sex families as the base structure, not nuclear families. It is not having a mother and father that is important, it is having the same mother and father today that conceived you that is the magic factor in educational attainment.

Now, the usual caveats apply – this study would need to be repeated several times to be confirmed, and the data doesn’t tell us whether same-sex parents are likely to have more unstable family setups (leading to more family structure transitions) then other non-nuclear families. But what is important is that, unlike the Regenerus Study, Potter went to great lengths to ensure that the “gay families” he studied were proper monogamous stable couples. In that sense this is a better study then Regenerus, but it still delivers the same broad outcome, that children in gay families have poorer attainment then those born in nuclear-families, but no different then other non-nuclear families.

10 Comments on “Daniel Potter on Gay Parenting

  1. Obviously a thorough and objective study. However, if someone had said to me, after my very disrupted childhood, “What effect do you think it had on your maths performance?”, I’d have thought they were insane.

    • Yes, I don’t really understand why this is being measured in school performance. One reaction to difficult family life in childhood is to retreat into school work. It doesn’t mean the teenager is happy!

      • It’s looking at educational attainment because it’s looking at educational attainment! If we want to explore whether (or even argue that) different family structures have effects on children then we need to quantify that in some form. Hence mathematics (and literacy) attainment.

        • OK. Just as long as we’re aware of the limitations of that. It seems that same-sex couples have been particularly good at adopting kids with special needs. Can we measure their success in maths scores?

          • I guess you would have to find a way to quantify it. Also, I suspect that the “monogamous gay couple adopting disabled child” sample will be very small indeed and not lend itself to any meaningful comparison with other groups. Potter’s (and Regenerus’) work looked almost exclusively at biological children.

  2. This fits in pretty well with what I have observed as a teacher – it’s instability at home that is the real issue. What would be interesting to see is a study comparing children adopted by same-sex couples as opposed to those adopted by heterosexual couples.

    • Hi Justin,

      Yes, that’s exactly the kind of study that would flesh out the real differences. Trouble would be finding enough gay adoptive families to do it (you would need ideally 150+) on a random basis.


Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.