Autogynephilia is a tendency to be aroused by the thought of being a woman, dressing as a woman, or a tendency to have sexual fantasies in which one imagines oneself as a woman. This strikes some people as gay, likely through some sort of association between women, femininity, and gay men. But is it true?
You might think that this has already been disproven by the many studies (e.g. here) that find that autogynephilia and exclusive attraction to men are very strongly negatively correlated among trans women, but what needs to be taken into account is that we are conditioning on a collider here, namely seeking gender transition, possibly leading to Berkson’s paradox where the correlation between two variables is reduced (or made negative) by restricting your sample. In fact, a simple calculation will tell you that these results suggest a positive correlation between autogynephilia and exclusive androphilia in the general population, unless some other effect is in play.
Are other effects in play? I think so; we know there’s meta-attraction, social desirability bias, and some people who want to get past gatekeepers. So we need a more-direct estimate.
This is a bit tricky to estimate, though. The obvious method would be to estimate the rate of homosexuality among autogynephiles and among non-autogynephiles and compare these two numbers, but since homosexuality estimates seem to vary wildly by sample and method, it becomes very important to use a consistent approach for this. The ideal would be to simultaneously estimate AGP; homosexuality, and its overlap, rather than combining different samples in possibly unreliable ways.
Through my surveys on reddit, I regularly get data on this, and because the AGP rates on reddit are very high (~50% for any degree of AGP, no matter how small, and perhaps around 2% for high degrees of AGP), it is very easy to get clear results here: gay men are less AGP than straight men. Here’s a select set of surveys showing that it holds for a lot of different ways of asking about AGP:
|Survey||Approach||Prevalence difference||Degree difference|
|One Word On Trump, And 19 Other Questions||Picturing self as female in sexual fantasies (regardless of reason; includes those who do it for non-paraphilic reasons)||2x higher for straight men than gay men (39% vs 18%)||2.5x more strongly for straight men than gay men|
|A Survey on Gender, Sexuality and Other Things||Aroused by thought of self as woman||3x higher for straight men than gay men (44% vs 14%)||4.6x more strongly for straight men than gay men|
|Gender and Psychology Survey||Aroused by wearing women’s clothes||2.3x higher for straight men than gay men (22% vs 10%)||2.6x more for straight men than gay men|
|Porn Survey 4||Aroused by picturing a beautiful woman and imagining being her||2.3x higher for straight men than gay men (70% vs 30%)||3.4x higher for straight men than gay men|
I don’t really expect that different ways of asking or different places to sample would make much difference. As long as the basic methodology is followed, I think you’d always get this sort of result. As such, I conclude that AGP is in fact negatively associated with male homosexuality. This works well with the theory that AGP is a variation of gynephilia that is self-directed in some sense.
Since homosexuality is also associated with being feminized in some sense, this also suggests that autogynephilia isn’t a feminine sexuality. But this is something I can go into more detail with in a later post…