Thursday, June 21, 2012

Outcomes for Adult Children of Homosexual Parents

The largest ever (and only decently sized) random sample study of the adult children of homosexual parents was published in a peer reviewed journal a couple weeks ago. The media has fed us a steady diet of studies of the children of gay households based on non-random samples with subjects recruited from advocacy groups, so it will come as a shocker for many that a truly random study doesn't agree with earlier "snowball" studies. Honestly, given the media conditioning we've received in the last few years, even I was shocked at the magnitude of some discrepancies in the well being of adults with a parent whose had a gay relationship. 

Some highlights...
a) 31% of adult children with lesbian mothers have been forced to have sex against their will vs. 8% from intact biological families.b) 23% of adult children with lesbian mothers have been sexually touched by a parent or adult caregiver vs. 2% from intact biological familiesc) 13% of adult children from intact biological families have had an affair while married or cohabiting vs. 41% with a lesbian motherd) 61% of adults with lesbian mothers identify entirely as heterosexual vs. 90% for adults from intact bio. families.e) only 29% of adults with lesbian mothers are currently employed full time while 28% are currently unemployed vs. 49% of adults with intact bio families having full time employment and only 8% being unemployed.f) While faring worse overall than adult children from intact biological families, adult children of gay men fare much better than adult children of lesbians. I found this incredibly counter-intuitive given the reputations women have over men as parents. 

Here are some some random reasons I've brainstormed on that last one... 
f1) I've read elsewhere lesbians tend to break up more than gay men (e.g., have a higher rate of dissolution of civil ceremonies/"marriages"). Unintuitive at first for me, until it was pointed out that women initiate more divorces.  Period.  Doubling the number of women in a relationship greatly increases the relationship standards and doubles the gender most likely to initiate.f2) the number of people with gay men as parents were much lower than the number of of people with lesbian mothers, since it's more unusual to have a parent who's a gay man and they typically have to work much harder to become a parent than women. That is, self selection for parenting plays a greater role for men.f3) The reputation women have as being more competent parents is extremely over-rated. f4) Lesbian women are more bitter than gay men, and children with bitter parents don't fare as well.  Before anyone gets mad at me for saying this, I've only heard it from people in favor of gay marriage and it honestly doesn't jive too well with most lesbians I've known.  So, get mad at other people, not me.f5) Your idea here________________

Overall, it's a fascinating read for those interested in such things.

Of course, none of this will matter, let alone get much media attention, because it doesn't agree with elite views on how society should be structured.  That and society hasn't given a rat's patootie about how divorce affects kids for the last 50+ years, so why should we start worrying now?  Damn the torpedoes.  Full speed ahead into cultural suicide!!!!

As an added bonus, here is a review of the totally inferior studies cited in the APA's brief on gay parenting.


Kevin said...

Thanks for sharing this, Doug. As you said, it is very surprising given the dominant themes in the media and academia which I have roughly adopted as representative, absent contradictory evidence. Such subtle harm is hard to see on an individual basis.

One technical thing I don't understand, though: I'd expect the bold labels for "statistically significantly different from IBFs without controls" to be a simple numeric error range comparison, but that doesn't seem to be the case. e.g. "Ever forced to have sex against will" IBF: 0.08, GF: 0.25* (not bold), Stepfamily: 0.16* (bold). Without controls, how could 0.16 be significant, but not 0.25?

In any case, the asterisked "with controls" is probably more useful anyway, it just struck me as odd. Maybe it is just something about stats that I don't know or have forgotten.

Douglas said...

Regarding how 0.16 could be significant, but not 0.25, I think it has to do with the fact that sometimes control variables can't be separated in such a way that leaves a statistically significant sample left to make the comparisons. He had pretty high standards for what he considered statistically significant, so the numbers people left in the study were probably just too low after adjusting for the control variables.

That's my guess at least. I'm just an engineer, not an expert in social science statistics and have never conducted a study like this myself.

The thing I found most odd was that 82% of kids from adopted families where said to self identify as entirely heterosexual, while 23% were said to be currently living in a homosexual relationship. That just doesn't add up. There was either a typo, or I'm missing something very big in how these statistics are calculated.

Kevin said...

Good explanation, Doug. The sample sizes are actually pretty small now that I see them.

I agree -- either 5+% claim to be both entirely heterosexual AND in a homosexual relationship, or something is wrong. `23% of people adopted by strangers are in a homosexual relationship` (23 out of N=101) seems erroneously high to me and it is not marked as statistically significant, so I think something is off there.

The author of the study, Mark Regnerus, has a website which provides his e-mail, and he also has a Christiany blog, so he seems accessible if you want to ask him your question. Or I can give it a shot, if you prefer.

Douglas said...


I had sent the question to the e-mail address on the paper. Thanks for the other links. If I don't hear back in the next week or so, I will be asking the question there.

The sample sizes may seem low, but they spent $800,000 and surveyed tons of people to obtain that data. It really is the best data out there, from what I've seen. It's a really tough group to study using random sampling methods given how few homosexuals there are and how few of them have kids. If you want to see some small studies, look at the Loren Marks review of studies which the APA used in their brief on gay parents. Most of those studies were pitifully small, on top of relying on non-random sampling methods.

Kevin said...

Great! I'll look forward to hearing what you find out.

I agree, it is still an impressive study. Regnerus explained the inherent limits well in his latest blog post.

Scott said...

Hi guys, here are a couple of articles on this topic that I came across in case you are interested:

Kevin said...

Those are excellent additions, Scott! Thanks for sharing them.

The Public Discourse Lopez's first hand account, in particular, was very compelling. It puts a real, living face on the statistics, which I think is a great boon to our understanding them.

Both of the articles provide insights into the definitions of sexuality and parenting being at the base of much disagreement in interpretation.