June 4, 2013

The Massey University Study of Transgender People

When the gender shoe does not fit (photos.com)
In my post Jaimie Veale's study of gender variant people throw new light upon crossdreamers I presented the high lights from one of the most important new studies of gender variant people. In this post I give a more in depth assessment of the findings.

The study

What makes the underlying study  from Masset University of New Zealand and her Ph.D: thesis much more reliable than similar research, is the number of respondents.

Ray Blanchard, who coined the term "autogynephilia", used the database of the Clarke Institute in Toronto for his studies of transsexuals. His ""Heterosexual and Homosexal Gender Dysphoria" paper of 1987, for instance, refers to a selection of 197 respondents, all of them patients at the university clinic.

Veale, on the other hand, is working on a sample of 2277 respondent, gender variant and gender typical, recruited via the internet.

There is still a bias towards white, "western", respondents, but she manages to cover a much wider group of gender variant people, also people who have not and will not seek out gender therapists.

In other words: Veale does to a much larger degree include non-transsexual transgender and crossdreamer people.

Gender Variant

Veale defines gender variant as "a subjective sense of not belonging  completely to the gender of one's birth-assigned sex." 

Note the word "completely". The term gender-variance is used to refer to the behavioral expression of this  identity which could range from occasionally dressing as one's identified gender to  living full-time in this gender. 

She has deliberately chosen to look into biological, psychological and cultural factors that other researchers have argued may influence the development of gender identity.

She explains this in this way:

"Previous research has found that genetics, prenatal hormone exposure, neuroanatomy, handedness and dermatoglyphics  [studies of fingerprints] (proposed to be related to susceptibility to developmental disturbances), fraternal birth order, and abuse are related to gender identity. While a number of investigators have studied these variables individually, the present study is the first know research to examine the inter-relationships between these variables." (p. 2).

Her point is that variations in gender identity and the intensity of alternative gender experiences may be the end result of an interaction between various factors -- genetic, hormonal, psychological and cultural. Moreover one variable might "hide" another influential cause of gender variance.

How to intepret results

When one read studies like this one, it is extremely important to keep in mind that she is writing about statistical aggregates, and not absolutes that apply to all people of a certain category.

For instance: Relative finger length varies between men and women on average.

Among men the the second (index) finger is more likely to be shorter than the fourth (ring) finger. This so-called 2D:4D digit ratio is believed to point to a higher exposure to "male" hormones (androgens) in the womb. Among women we will find that the fingers are more likely to be of equal length.

I have seen male to female transgender people enthusiastically proclaim that they must be female, because they have a "female" 2D:4D ratio. That is mixing the macro level with the life story of the individual. A lot of women born women have a "masculine" ratio. That does not stop them from being real women.

This graph can serve to illustrate the fact that we are talking about relative small, although significant, differences. Note the huge overlap between men and women.
Digit ratio visualization according to data from Bailey and Hurd
Blue line represents men, green line women. Click on image to enlarge!

With the exception of body height and sex organs practically all differences between men and women are less significant than this.

It is clear from this figure that if there is a causal relationship between exposure to androgens in the womb and the digit ratio, the effect must be masked in many, otherwise "manly", men.

This tells me, at least, that we should be very careful to interpret such findings in a dogmatic way. Androgen exposure alone cannot explain the existence of all male to female gender variant people.

Her main findings

Since Veale has only studied a limited number of potential factors influencing the development of gender identity, she finds that somewhere between 7 and 23 percent of the variance found can be accounted for from the biological  and psycho-social factors listed above.

I will go through the variables, one by one.

Number and ratio of siblings

Male to female (MTF) transsexuals are more likely to have older brothers than non-transgender males, according to Veale's data. This difference is not found to the same effect in other gender variant MTF gender variant people:

"Birth order and sibling sex ration did not differ between birth assigned males with other gender variant identities [as opposed to transsexual] and males with gender-typical identities in spite of previous research suggesting that cross-dressers are more likely to be first-born males."

I guess someone would argue that this proves that real transsexual women are completely different from non-transsexual transgender male bodied persons. However, the difference  is a matter of degree, non-transsexual MTF transgender taking an intermediate position between MTF transsexuals and non-transgender men. This finding could be used to support a spectrum model of gender variance.

Click on image to enlarge! (Figure from Veale's thesis)

There is no difference for the number of older brothers among birth assigned females.

You might ask what all of this has to do with the price of eggs.

Well, it turns out Ray Blanchard has argued that homosexual men are more likely to be effeminate if they have a larger number of older brothers, while we will not see this effect among non-homosexual men (and by implication MTF transgender attracted to women). This argument has been  used to "prove" that gynephilic transwomen are "less feminine" than the ones loving men.

Veale finds no significant difference between androphilic (man-loving) and non-androphilc (woman-loving and bisexual) MTF transsexuals in her data.

Sexual orientation

This conclusion, i.e. that there is no significant differences between male to female transgender who are attracted to men and those that are attracted to women, applies to the whole data set as well.

Blanchard argues that there are two distinct types of male to female transgender and transseuxals: the "homosexual transsexual", who is a extremely feminine man, and the "autogynephile" who is a non-homosexual man suffering from an erotic target error. He is basically a pervert in love with the idea of himself as a woman.

On the Blanchard scale of perversions, the man loving trans woman is slightly above the "autogynephile" on the mental illness scale. The fact that Blanchard also considers homosexuality a mental disorder, doesn't leave much respect for any of the two categories, however.

Veale, of course, knows this theory very well, and has very carefully checked her own data to see if they support Blanchard's dichotomy.

"Blanchard's theory predicts that there would be improvement in the model if biological and psychosocial variables' prediction of gender variance in this study were allowed to vary between androphilic and non-androphilic birth-assigned males. However, when these restrictions were relaxed, no significant improvement in the model fit was observed."

Anne Lawrence and J. Michael Bailey had tried to shoot down earlier papers by Veale and her colleagues, by arguing that she had no real androphilic trans women in her sample. Because of this she and her friends had imposed strict criteria for being categorized as androphilic. Veale argues strongly that the sample size is large enough to detect group differences.

There are no such differences. Sexual orientation does not determine different types of gender variant people. The autogynephilia theory is wrong.

The role of gender variant relatives

Transsexuals and gender variant respondents were much more likely to report gender variant relatives than respondents with gender typical identities.

Veale explains:

"Given the methodology of this research, it is not possible to distinguish whether this concordance is the result of a genetic or a social learning effect. However, it has been reported elsewhere that usually these individuals are not aware that their relative is gender variant until they reach adulthood, suggesting a genetic explanation is more likely." (p. 110)


Handedness has been used to determine hormonal effects in the womb.

"The study found non-righthandedness was significantly related to adult gender variance in birth assigned females, but not birth-assigned males. The results are in accordance with previous findings for both birth-assigned females and consistent with the hypothesis that developmental instability has a role to play in the development of gender identities." (p.111)

I am not sure how she comes to this conclusion given that she finds no such effect among the birth assigned males.

She provides a methodological discussion that is simply to complex for me to popularize it here. See chapter 8.1.4 in the thesis.

Other studies have shown, however, higher levels of left-handedness among MTF transsexual women, and it is the fact that this study does not confirm these findings that causes confusion.


Veale finds an increased level of emotional, physical and sexual abuse among transsexuals and participants with other gender variant identities.

It is harder to decide whether the abuse is causing gender variance or whether it is caused by it:

"The SEM [Standard Error of Mean] results showed that recalled childhood gender-variance partially meditated the relationship between emotional abuse and adult gender-variance among both birth assigned genders. This suggest that as well as having a direct influence on adult gender variance, emotional abuse was also the result of gender variance being the target for abuse. (p. 112)

In other words: Veale believes abuse is caused by gender variance, but that it will also influence its development. This study does not provide a final answer to this enigma.

Finger length ratio

As explained above finger length ratio (2D:4D) is believed to be a marker for prenatal androgen exposure (i.e.variations in the presence of "male" hormones in the womb).

Veale finds no differences between gender typical and gender variant respondents or between the different types of gender variant people.

Her conclusion is the following:

"Overall, these findings tell us little about the extent to which 2D:4D is related to prenatal exposure, or the role that prenatal androgen exposure might play in the development of gender identities."

Mental rotation and systemising quotient

Throughout the history of sexology there has been made many attempts to test for differences in male and female cognitive capabilities. We have all heard the "truism" that men are good with numbers and women are good with words.

Being a male bodied person with a preference for words, I might have taken this as proof of my inner femininity, but the existence of a significant number of successful male authors (and female engineers) has dampened my trust in such arguments.

As time goes by, and women enter more and more occupations and take over the universities, the differences reported by researchers become more and more insignificant.

In other words: I suspect that men and women are equal intellectually and mentally, and that the differences found by researchers are the results mainly of nurture rather than nature. This means that I doubt that such tests can be used to determine whether gender variance is inborn or cultural. (See my post on the differences between men and women for details.)

In Veale's scientific world  you have to look into such variables, however, and she does.

She notes that past studies have found that transsexuals are more likely to score in the same way as persons of their target sex when it comes to tests of three-dimensional object mental rotation (which is something men are supposed to excel at).  This suggest a neuro-biological explanation for gender variant identity development.

Veale gets the same results among the birth assigned males in her study. The results for the female bodied gender variant responents are less clear, however, as Veale will have to disregard the "systemising quotient" (the interaction between variables) to get the expected results.

She concludes:

"It can be concluded from these inconsistent results that there is at best limited further evidence for a neurobiological explanation of gender variant identity development among birth assigned females."

She finds similar results for the axis "systemising ability" (typicallly male according to sociobiologist Simon Baron-Cohen) and "empathizing ability" (a female enhanced trait according to Baron-Cohen). The male bodied gender variant respondents are more likely to have a "female brain", while there is no significant differences between birth-assigned females with gender variant or gender typical identities for systemizing.

I find it very hard to believe that gender variance in male bodied persons should be neurobiological and variance in female bodied persons should be cultural. There is clearly something wrong with the premise for the argument -- and the methodology.

Veal seems to believe these variables do say something about the origins of gender variance, though:

"Overall, these results are consistent with a prenatal hormone predisposition for gender-variant og gender-typical identities, although the possibility of systemising being caused by more than prenatal androgens and the role of socially desirable responding should not be discounted."

In other words: Gender-variant people may also be tempted to live up to gender stereotypes as this may make their target gender identity more understandable in the eyes of others.

This leaves us with the question of why female bodied respondents are less likely to live up to the stereotypes.  Maybe this is caused by the fact that women in general are less likely to be punished for masculine behavior than men are for feminine behavior.

Parental age

This is the first study that shows that  gender-variant birth-assigned females are more likely to have older mothers than the rest of the population:

"The more gender-variant a birth-assigned female's identity was, the older the mother tended to be [at the respondents birth]"

The sexual orientation of the respondents  was of no relevance.

Veale notes that:

"The link between autism and adult gender-variance that is discussed in Section 8.1.8 [on the systemising quotient] is also relevant here, and it is notable that it was also birth assigned females with gender-variant identities that this study found tended to score more autistic-typically (as opposed to gender-variant birth assigned males who scored less autistic-typically.)"

Please note that  Baron-Cohen  argues that autism is caused by some kind of hyper-masculine systemizing brain. Finding more autism among female bodied gender variant people might therefore indicate that a male gender identity is caused by a masculine brain development.

At the same time the finding that male bodied transgender people are less likely to be autistic, indicates that their brains are more female.

Again: I strongly advice readers to proceed with caution when reading these types of arguments. There are just too many undocumented steps leading up to the conclusion.

It is interesting to see that Veale finds fewer autistic traits among gender variant male bodied persons than among the non-transgender men, though.

In online fora you will find MTF crossdreamers arguing the opposite: that crossdreaming is somehow related to autism or Asperger among the birth-assigned males. I have suspected that it is gender confusion and shame that have made them less socially adaptable, and not some kind of autism.

Parental cohabitation

In the Veale study the chances of being gender-variant increases the earlier the father leaves the home. On the other hand, cohabitation with mother and parental death is not related to gender variance.

Veale says that:

"Generally, the finding provides some evidence in favour of theories that have proposed that parental absence plays a role in the causation of gender-variance as a psychosocial factor."

Earlier studies has found this correlation among male bodied persons only. Veale's study, however, finds this correlation among birth-assigned females, and not among birth-assigned males.

I am not sure what to make of that, and nor does Veale it seems, who suggests that more research is needed.

Socially desirable responding

When Blanchard & Co find that their data does not support their theory about "autogynephiliacs", they pull up the "they are lying" card.

In their book androphilic (man-loving) trans women cannot be crossdreaming (having "autogynephilic" erotic fantasies), so the ones that do report such arousal are reclassified as "non-homosexual". They must be deceiving themselves and/or the researcher about their love for men, according to Blanchard.

Interestingly, Blanchard never seems to look equally close at the truthfulness of androphilic transwomen. Maybe they are lying about their lack of crossdreaming fantasies, trying to live up to Blanchard's fantasy of the proper "homosexual transsexual"? But such a result would undermine Blanchard's theory, so he ignores this possibility.

I have always found Blanchard's respondents to be extremely frank about their crossdreaming fantasies. If they had been less forthcoming, maybe we wouldn't have been in this mess.

That being said, it is definitely true that many male to female transsexuals have lied to the gate-keepers of hormones and surgery to get the surgery they want. If the gate keepers lives in a world of sexual stereotypes, you'd better learn to live up to them. And if the gate keeper believes proper women love men, you tell him what he wants to hear.

Veale has one big advantage compared to Blanchard. She is not a gate keeper, and in an anonymized online survey the respondents do not have the same reason to sugar coat their life stories.

Veale writes:

"Accounting for social desirability did not significantly change the results. It is possible that the anonymous nature of the survey made it less susceptible to social desirability bias than the past research conducted by Blanchard & al.... and this differs from participation in Blanchard's et al's study which assessed participant's reports of gender-variance experience. Furthermore, participants in Blanchard's et al.'s study may have believed that their access to treatment was contingent on how they responded."

Transsexual vs. no-transsexual gender variant people

Veale finds support for Richard F. Docter's theory of gender variance. He has proposed a continuum from no gender variant identity at one end (often misleadingly labelled as "normal people") to transsexuals at the other end.

Veale puts it this way:

"As outlined in Chapter 2, Docter's (1988) theory of gender variant identity development proposes a continuum from no gender variant identity to transsexualism at the ends, with persons with other gender variant identities being somewhere between these. If this proposal is correct, then it would be expected that the participants with other gender variant identities would score intermediary between transsexuals and participants with gender typical identities on biological and psychological variables."

This is exactly what she finds:

"This was the case for the majority of the biological and the psychosocial variables in the present research that showed between group differences."

The biological origins of gender variance

All of this leads Veale to conclude that biological factors play a role in what causes gender variant identities. She is even willing to consider the possibility that this has one genetic origin:

"While there is evidence for a number of biological variables, this does not necessarily imply more than one biological mechanism plays a role -- these biological factors may be related and share a common precursor. For instance, it is plausible that there is a causal pathway in which genes that cause atypical prenatal hormone levels lead to neuroanatomical differences linked to gender identity."


Does all of this mean that Veale has provided the final truth about the origins of transgender conditions?

I seriously doubt that. But she has made an important contribution to our understanding of what this is all about. What is especially important for crossdreamers, though, is that she has debunked the theories of Ray Blanchard and his followers, on his terms.

This will not make Blanchard change his mind, though. He has invested too much in his restrictive view of humanity do to so. This research will make it easier for other researchers and care takers to develop a less toxic and stigmatizing view of gender variants, though.

Veale puts it this way:

"These findings suggest sexual orientation is not a moderating factor in the development of gender-variant identities in birth assigned males. This is likely to decrease the social stigma associated with gender-variant identities."

Jaimie Veale's publications can be found at her web site.

Richard F Docter: Transvestites and Transsexuals: Toward a Theory of Cross-Gender Behavior (Perspectives in Sexuality) Springer 1988

Discuss crossdreamer and transgender issues!