Some complexity in thinking about the rates of AAP

In an earlier post, I talked about the rates of AAP among transmascs. One troubling point here is that the AAP (and general paraphilia) rates in my surveys are usually far above the baseline, so there’s an argument to be made that the rates in FtMs should be smaller, to about 20% and 25% respectively, assuming we just drop the AAP rate 10x.

Formalizing the above a bit more: about 11% of the natal females in my surveys are trans, and about 50% of the natal females in my surveys are AAP. It is probably not the case that 40% of women are AAP; in reality it’s probably more like 5%. It’s also clearly not the case that 11% of natal females are trans. However, suppose we split the 11% up into 8% who are AAP, and 3% who are not AAP. We then see that about 8%/40% = 20% of AAP natal females are trans, and 3%/60%=5% of non-AAP natal females are trans. If we then readjust the AAP ratio to be 95% non-AAP, we should get 5%*20%=1% of natal females who are AAP and trans, and 95%*5%=4.75% of natal females who are non-AAP and trans. This yields an AAP rate among transmascs of 17%, which is close to the 20% I gave before.

Now, the reason I made the kinda-ridiculous calculation above explicit is because there’s clearly something very wrong with it. According to this, 5.75% of natal females would be trans, which is still clearly not true. So what’s going on here?

Probably a mix of things. Most likely, some of those who report not being AAP actually are AAP. This creates a flaw in adjusting for the high paraphilia rates, where the rate of false negative AAPs isn’t reduced. It’s not too unbelievable that 28% of AAP FtMs might fail to realize they’re AAP in my opinion (this is pretty close to the same rate for MtFs), but it might be a bit of a stretch. reddit is probably also unusually trans due to some as-of-yet unknown factor. I think the implication here is that the non-AAP trans rate should be reduced as well, which would probably yield a higher rate of AAP among transmascs.

Really, I should be declaring some degree of ignorance here, but I think we can continue considering this for a bit more. About 20% of AGP trans women misreport their experiences as including no AGP. If we assume the same holds for FtMs (dubious? they might not see as much stigma from it, and so feel less need to misreport…), this means that about 72%*1.2=86% of the transmascs in my surveys are AAP. We can then repeat the previous calculation with the new numbers:

9.5% of natal females in my surveys are then AAP trans, while 1.5% are non-AAP trans. This means that 9.5%/40%=24% of AAP AFABs are trans, while 1.5%/60%=2.5% of non-AAP AFABs are trans. With a modified AAP rate of 5%, this means that 1.2% of natal females in general are AAP trans, while 2.4% of natal females in general are non-AAP trans. This implies an AAP rate of 33%.

Now, there’s a lot of extra ways we can manipulate the parameters. For example, if we assume a general population AAP rate of 10% rather than 5%, we get a transmasc AAP rate of more than 50%. If we instead assume that the non-AAP group is also overrepresented, say 2x, we get a transmasc AAP rate of almost 50%. There’s endless variations that can be done, but that gives endless possibilities for manipulating the data in whichever way you want.

Can we do something empirically to determine the answer here? Well, really it would require asking a representative sample (e.g. a clinical one) to make progress, but there are some hints we can get. Smith (2005) found that among non-exclusively-gynephilic FtMs, 73% reported having never experienced arousal while crossdressing during adolescence, 14% did not answer the question, 9% did not consider the question applicable to themselves, and 5% had sometimes experienced such arousal. I think there is an argument to be made that the missing data is an attempt to avoid admitting arousal from crossdressing, which would imply that we can say that 18% had experienced transvestic arousal. If we instead exclude the missing data, we can instead find that 1/19 = 5% experienced transvestic arousal. The study did have a very limited sample size, though, so there’s a lot of uncertainty in this estimate. Regardless, how do these 5% to 18% numbers compare to what I see in my samples?


This suggests that transvestic fetishism will dramatically underestimate the rates of AAP. In fact, this result says that TV fetishism will only be present in 13% of AAP cases, which would imply that somewhere in the range of 40% and 100% of Smith’s sample is AAP. And of course there’s infinitely extra uncertainty that I’m not even taking into account here.

I don’t know that this is particularly useful. Maybe it is, since it does help justify the possibility that AAP might be genuinely common. Another point to be made is the fact that there’s a lot of case studies with clearly-AAP FtMs, but these might not generalize to modern times.

A study that is also worth looking into is Bockting (2009). They find:

Half of participants elaborated on a shift in the image of themselves in their sexual fantasies as they became more comfortable with their bodies as a result of going through reassignment. They explained that, in time, they relied less on sexual fantasies of themselves as male to affirm their gender identity and were more likely to fantasize about themselves as transgender.

This implies that at least half the participants imagined themselves as male before they changed genders, but there’s good reason to expect that a large fraction of the other half also pictured themselves as male. On the other hand, the study also states:

An alternative interpretation of the transgender sexuality found to be emerging among our participants is that this is a form of autoandrophilia (sexual arousal to the thought or image of oneself as a man), the female analogue of autogynephilia believed by some (e.g., Blanchard, 1989) to be a core component of nonhomosexual (defined as not being attracted to the same natal sex) gender dysphoria (Chivers & Bailey, 2000). An exploration of autoandrophilia was not the focus of our study. However, more than two-thirds of the female-tomale participants did not report any history of transvestic fetishism (almost always found among transsexuals with autogynephilia) or any evidence of an erotic target location error (in this case, the target of eroticism being the thought or image of oneself as a man rather than another human being). Rather, as in the Dutch study (Coleman et al., 1993), femaleto-male participants described genuine sexual attractions to, and intimate relationships with, other men that went beyond satisfying their curiosity about male sexuality (as suggested by Devor, 1993, 1997).

Now, we already know that transvestic fetishism doesn’t clearly co-occur with AAP, so this doesn’t matter much. It’s not clear what the authors meant when they said they found no evidence of an erotic target location error. Autoandrophilia was not part of their interview guide, and there’s no rule that you can’t both be AAP and alloandrophilic.

I want to say that so far, the evidence supports the position that most of the FtMs in question are autoandrophilic, but it doesn’t seem inconceivable that a study could come around and totally change the situation.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s