Why transphobia hurts gender non-conforming cis people

Content note: This post will discuss transphobia, homophobia, misogyny, police brutality, some sources will include *phobic statements or arguments

Recent transphobic discourse has focused on claiming that trans people, by virtue of our mere existence, prop up patriarchal gender norms or that transition is a form of highly convoluted anti-gay conversion therapy by parents who are deeply homophobic but curiously accepting of the idea of having a trans child. This is obvious nonsense, it also begs the question of exactly what the effect of transphobia is on gender non-conforming people.

Gender norms are the set of social expectations placed on us by society based on our presumed gender, such as the expectation that women should be nurturing, shy and submissive while men should be strong, outgoing and dominant, the expectation that women will be exclusively sexually and romantically attracted to men (and vice versa) or the expectation in modern day Western societies that only women should wear dresses or skirts. Within the dominant heteropatriarchical view, gender is seen as binary, essential and invariant with respect to history and culture. Within this ruling perspective, gender is seen as innately linked to “biological sex” and reproductive capacity in a deterministic fashion. This view of gender is reflected in the recent leaked memo by the Trump administration, which seeks to write such a definition of gender into law as a means to deny civil rights to transgender people, such a move will also likely cause serious harm to intersex people. Interestingly, trans exclusionary radical feminists (TERFs) who consider themselves of the left, such as recently resigned Labour councillor Ann Sinnott, have indicated their support for this attack on transgender people by a vehemently right wing government.


sinnot trump
Former Labour councillor Ann Sinnot links to a Christian Post article on Trump’s attack on trans people, commenting “Strange times, strange bedfellows. So be it.”

People who do not easily fit within these heteropatriarchical norms (whether because of their presentation, their identity, their sexuality or their bodies) can be termed “gender non-conforming” (GNC). While most GNC people are cisgender, all trans people can be considered to some extent GNC because our existence is necessarily in defiance of dominant gender norms.  GNC people are, to some extent, a spanner in the works of heteropatriarchy; a challenge to the perceived natural order. We are difficult to subsume within the heterosexual family, which is seen as the basic unit of society.

Transphobia, therefore, can be seen as a means of policing gender presentation and enforcing gender conformity.  Within this context, it is unsurprising that the recent hysteria about the alleged dangers of allowing trans people to use the toilet of our choice (despite a total lack of evidence of any such danger) has often had a detrimental effect on GNC cisgender women, with high profile examples of cisgender women, often butch lesbians, being challenged and discriminated against while attempting to use women’s public toilets, (see the Walgreens and Strike Bowling cases for examples). While it is obviously the case that such policies are explicitly targeted at trans people, it is also undeniably the case that GNC cis women are just as often the victims as trans people, due to their greater numbers in the wider population. The misogynistic nature of toilet policing can be particularly clearly illustrated in the Strike Bowling case, where the wife of the woman targeted was aggressively told on social media that her partner was to blame for dressing in an insufficiently feminine manner.

The idea of a strict divide between minority sexualities and gender variance is a relatively recent invention. In the 19th century, queer people were often considered as being of a “third sex”, with the term “Uranian” being coined by German writer Karl Heinrich Ulrichs to describe what we would now consider to be the LGBTQ+ umbrella, in this period a man who sexually desired other men and a person assigned male at birth who wished to live and be recognised as a woman, what we would today consider a gay/bisexual man and a trans woman, respectively, were considered to be subsets of the broader third sex that all GNC and queer people belonged to. Indeed, the Stonewall riots, considered now to be the origin of the modern LGBTQ+ liberation movement, were as much about gender non-conformity as they were about sexuality. At the time anti-gay legislation frequently targeted people wearing clothes of the “opposite sex”, the gay communities of the time, therefore, were a broad alliance of all gender and sexual minorities. The riot began when, during a police raid on the Stonewall Inn, butch lesbian Stormé DeLarverie was brutalised by the cops and fought back, self-described transvestites, Marsha Johnson and Sylvia Riveria, were amongst those who took the initiative to stand with Stormé. The history of Stonewall is a history of GNC people of colour fighting back against police brutality side by side.

Since Stonewall, an increasing emphasis on queer taxonomy has become de rigueur in LGBTQ+ movements, with frankly tedious arguments about whether bisexual women have an equal claim to the term “lesbian” or whether historical figures such as Dante Gill can be strictly considered trans often dominating queer discourse to the detriment of solidarity and taking action to support one another. Ultimately the oppression of LGBTQ+ people relates to the fact that we so often defy easy categorisation within heteropatriarchical hierarchies, as seen in arch-conservative Peter Hitchens claims that transgender activists are “destroying truth itself”. Our identities are messy and alien to the simplistic taxonomies we impose on top of them. The emphasis on self-identification in queer movements, therefore, can be seen as much useful heuristic as it is metaphysical claim. From this point of view “Anybody who says they are a woman is a woman” (for example) is not a statement of causation (“I am a woman because I say I am”), but an admission that the categories of gender are ultimately unstable and cannot truly encompass all of human gender and sexual diversity (“Identity as a woman is as good a way of distinguishing women from non-women as any other”).

Currently we are in the midst of a global backlash against LGBTQ+ liberation, with trans liberation seen as a key battleground and wedge issue. Attacks on our civil rights in the US, the rising fascist movement in Brazil and the ongoing transphobic moral panic in the British press will harm not only trans people, but all LGBTQ+ and GNC people and we have to stand together against it. Transphobia, homophobia and all the other *phobias we experience are aspects of heteropatriarchy defending itself, we have to get better and sticking together while standing against it.


An analysis of a Fair Play For Women crowdfund

The TERF group Fair Play For Women (FPFW) has seen considerable press attention after placing an ad in the Metro claiming that reform of the Gender Recognition Act (GRA) would place (cis) women and girls in danger. This nonsense claim has been debunked at length many times in the past and will not be the subject of this post. Instead, I will be discussing claims about the funding of TERF groups, using data scraped from a FPFW crowdfund page as a case study.

Trans inclusive feminists such as Casey Explosion and Zoe Stavri have raised some interesting questions in relation to this, pointing out the considerable cost of placing a full colour, whole page ad in the Metro (around £45,000) and questioning where the funding for this relatively opaque group comes from.

Edit: Shortly after publishing this, a commenter on Twitter pointed out that the quoted cost for the ad may be incorrect, as the ad may have been published as a “classified” rather than “display” ad, at a presumed cost of £17,300 including VAT. Still no small chunk of change.

Allegations and suspicions that TERF groups are at least in part funded by conservative interests are common, driven in part by documented instances of TERF groups being funded by conservative evangelicals in the US, right wing evangelicals talking openly about using TERF talking points to undermine the LGBTQ community in general, prominent TERFS signal boosting anti-LGBTQ organisations and the willingness of conservative news sites like Breitbart to defend TERF fundraising efforts amongst many other instances of close relationships between right wing anti-LGBTQ groups and individuals and TERF organisations and individuals. It is, however, difficult to prove with any certainty direct collaboration in most cases, TERF groups are typically small organisations that are not required to keep public accounts and often refuse to disclose their funding sources.

In response to a TERF scoffing at the notion that FPFW’s funding came from anywhere but money raised by “ordinary people” who “donated a few quid to protect women’s rights”, Zoe Stavri produced a Twitter thread suggesting that the success of FPFW’s recent fundraising efforts was suspicious. In order to investigate this further, I decided to scrape FPFW’s most recent and still active crowdfunding page to investigate further. An anonymised version of the resulting raw data (given in reverse chronological order) can be found here.

The resulting data is interesting, to give a brief summary, a considerable proportion (around a third) of the money contributed to the crowdfund came from anonymous donations of £100 or more and that of the money contributed by named individuals around a fifth came from individuals who had donated £100 or more. While we cannot, by definition, state with any certainty where the anonymous contributions came from, it is interesting that such a large chunk of the money comes not from “ordinary people” chucking in “a few quid” but large donations (some in the region of £500-£1000) of unknown provenance.

The boring maths-y bit

The mean donation to the fundraiser was £31.88 (95% CI -128.16-204.55), the mean donation for anonymous and named donors was £38.19 (95% CI -128.16-204.55) and £23.47 (95% CI -15.70-62.64), respectively Summary statistics for the donations are given in the table below:

FPFW summary stats

This might initially suggest no statistically significant difference between anonymous and named donations, however, when, out of interest, I sorted the donation data by amount, something interesting happened. Nearly all of the largest donations were anonymous!

FPFW big anon

If we define a “large” donation as one of £100 or more (considerably more money than most ordinary people would feel able to chuck at a crowdfund in one go), there are 37 such donations in the data, 32 of which are anonymous. These large anonymous donations total £6,970, a little over 30% of the total money raised.

Of particular interest are the five donations of £500 or more (two of £1,000 and three of £500). All of these donations were anonymous. In one instance a donation of £500 was immediately followed by a donation of £1000 two minutes later, possibly indicating a single anonymous donor contributing £1,500. A very “ordinary” thing to do indeed.

There were 275 named donors. Of named donations, larger donors (£100 or more total donations) contributed £1,430 (19% of total named contributions). One off donations of £50 or more totalled £13,190 (57% of total raised). It appears to be the case that a considerable proportion of the funding for this crowdfund comes from relatively large single donations, often extremely large amounts from anonymous donors.

While, again, we cannot know with any certainty the exact source of money going into this crowdfund, it is reasonable to assume that the large donations that comprise a considerable chunk of FPFW’s income come from wealthy individuals or groups, many of which for some reason prefer to remain anonymous to the public.

Tip jar

Contrary to some odd allegations I have seen doing the rounds, most trans women are not, in fact, in receipt of large amounts of money from billionaire George Soros. The work I do to research and produce my blog posts is done entirely on my own time. If you’d like to be my shady, unidentifiable source of funding, why not drop the price of a pint in my paypal tip jar?

Why I don’t care whether or not I was “born this way”

Anybody who has spent any time around mainstream, liberal LGBTQ+ discourse will be at least passingly familiar with the idea that queer folks are “born this way”, that is to say that our sexualities and genders are fixed from birth and unchangeable. This is often presented as an alternative to the idea that queer sexualities and genders are in some sense chosen and changeable. This can be considered an essentialist view of sexuality and gender – both are seen as fundamental to a person and the idea that any person could be anything other than the gender and sexuality they were born with is rejected.

It’s not hard to see why this is such an attractive idea. Much of our moral thinking rests upon the idea that “ought implies can”; this principle, often attributed to Kant, states that a moral imperative (a statement about what a person ought to do) assumes that it is possible for that imperative to be carried out. Such a principle, applied to born this way rhetoric, suggests that a moral imperative to be cisgender and/or heterosexual is untenable since LGBTQ individuals are fundamentally incapable of carrying it out. In other words, if we accept “born this way” as true, then demands for LGB folks to become heterosexual and for trans folks to become cis are clearly unreasonable.

Indeed evidence exists that belief amongst straight people that sexuality is a choice is correlated with homophobia. However, correlation is not causation. It is impossible to say with any real certainty that a person’s homophobia is caused by their beliefs about the origins or sexuality. Such an assumption is, perhaps, far too charitable to queerphobes of various sorts, granting that their prejudices rationally precede from some set of beliefs about the world, a difference of opinion, rather than the far more depressing reality that people’s beliefs about the world are often formed post hoc to justify their prejudices. The idea that convincing people of a particular theory of the origins of sexuality is a useful strategy for reducing the overall levels of homophobia in society is fundamentally misguided.

Additional pitfalls exist for the essentialist argument against queerphobia both in terms of the ability of queerphobic organisations to adapt to new narratives and in terms of what such arguments concede.

Many organisations advocating “conversion therapy” (a set of abusive practices that seeks to coercively change a person’s gender or sexuality to more closely align with cis/heteronormative norms) focus not on changing a person’s internal desires and identity so much as changing behaviour. These organisations will claim that it is better to be celibate out of self-hatred and shame than to be open and enjoy one’s own sexuality. Such beliefs are somewhat immune to exhortations to accept that queers are born any particular way, because they want to force us back into the closet, not actually change any fundamental aspect of our personalities.

To focus on whether we can change is to concede the question of whether we should even if we could. I am not interested in being tolerated as an unfortunate aberration, but instead insist upon my freedom to exist and exercise my own bodily autonomy and consent as long as it doesn’t impinge on the autonomy and consent of others. It doesn’t matter where my gender identity or sexuality come from, it matters that they are mine.

Addendum: Shortly after publishing this post I came across a Guardian piece by Shon Faye covering the same topic published yesterday – I’m trans, and I don’t care if we were ‘born this way’. Neither should you – I hadn’t read it at the time that I wrote this but it would be remiss of me not to mention this as a much more in depth piece covering solidarity, neuroplasticity and discourse around some recently published research on trans children’s brains.

Transphobia is a class issue.

[Content warning: In addition to transphobia in the abstract, this piece discusses harassment, violence and abuse. Some sources linked to for reference purposes feature transphobic abuse and slurs.]

Transphobia is a class issue. By this I mean that in a class society that is also deeply transphobic, it is impossible to talk about transphobia in a meaningful way without also talking about class. Trans people are more likely, all other things being equal, than our cis peers to fall into the most exploited and oppressed sections of the working class and the extent to which transphobia will negatively affect any given trans person’s life will be mediated by their economic class. This article is not intended to be a comprehensive analysis of every aspect of this issue, but to contribute to an ongoing conversation around it and illustrate a class struggle perspective on transgender issues.

By transphobia I mean two related phenomena:

  1. Overt, intentional hostility to or disregard towards the wellbeing of trans people and;
  2. Social structures and systems which put trans people at a relative disadvantage to cis people within society.

These two types of transphobia are not strictly distinct and one often creates or reinforces the other.

Often when discussing transphobia popular discourse focuses on overt, interpersonal hostility and street level violent hate crime. While these are indeed real and very serious issues, this focus on the interpersonal and the overt often leads to a failure to recognise the measurable economic effects of transphobia on trans lives. This constitutes a form of hidden, endemic, systematic violence against working class trans people.

A 2015 EU report found that trans people in the EU were more likely than their cis peers to be in the bottom 25% of earners and that around a third of trans people reported experiencing workplace discrimination in the year leading up to the survey and a similar proportion had experienced discrimination while looking for housing. Unsurprisingly, given high levels of workplace discrimination and general social stigma, trans people are disproportionately more likely to experience unemployment. Emma Rundall carried out a survey of trans people as part of her 2010 PhD thesis and found that 14% of respondents were unemployed, around two and a half times the then national unemployment rate (pp 139 of thesis), this is consistent with a general trend in the literature for higher rates of unemployment amongst trans people.

Housing discrimination and high rates of family rejection and abuse also lead to higher rates of homelessness for LGBTQ people as a whole and particularly LGBTQ youth. A 2015 report by the Albert Kennedy Trust found that LGBTQ youth were “grossly over-represented within youth homeless populations”, stating that one in four young homeless people were LGBTQ, the report also found that a majority of young LGBTQ homeless people reported rejection or abuse at home as a major factor in their homelessness, with an overwhelming majority of housing providers failing to recognise the unique and specific needs of this marginalised community for housing support. Specific figures for trans people alone in the UK are difficult to find, however in Canada, a culturally similar developed nation, the research and community organisation Trans Pulse carried out a study of health outcomes in 123 trans people aged 16-24, with a view to measuring the effect of parental support. All respondents reporting “strongly supportive” parents reported being adequately housed, however, almost half of the two thirds of respondents who did not have strongly supportive parents were “inadequately housed” (homeless or in a precarious housing situation), around one third of the total sample.

(Albert Kennedy Trust, 2015)

As well as the economic effects of transphobia itself, we can also consider the intersections of transphobia and class, i.e. the ways in which class and transphobia interact and magnify each others’ effects; the greater financial resilience of the middle and boss classes, the ability of wealthier trans people to buy their way out of some forms of transphobia, the classed nature of the bureaucracies that trans people are often forced to navigate and the elevation of privileged voices within the broader trans community as the authentic voices of all trans people.

A core component of transphobia at present is medical gatekeeping, the process by which trans people are forced to jump through semi-arbitrary hoops in order to access certain kinds of trans specific healthcare. In Sex Educations: Gendering and Regendering Women Lisa Milbank discusses real life experience (RLE), a period of time in which trans people are expected to present “full time” as their gender in order to access certain kinds of healthcare, as a form of socially enforced “breaking” in which trans women are subjected to “an experience of public freakhood, composed of constant stares, transphobic harassment and potentially violence, without access to much of the (intensely double-edged) training given to cissexual women on how to survive this”, while Milbank focuses on the experience of transsexual women in particular, this also applies to some extent to the experience of other trans people. One’s ability to pass as cis (to be read by most people as a cis person of one’s appropriate gender) will heavily influence the extent to which RLE is a dangerous and potentially traumatic experience. Since passing as cis takes the form, in part, of being able to perform conventional cis norms, which are themselves heavily classed (and racialised), a trans person’s ability to do so will be mediated by their class status. I.e. the wealthier a person is, the more likely they are to be able to afford to take additional, elective steps (extensive hair removal, specialised clothing to hide or accentuate particular gendered body traits, etc.) to increase their chance of passing as cis. In this way, middle class and boss class trans people are more easily able to navigate gatekeeping in order to access healthcare and sidestep the harmful effects of RLE in a transphobic society. Similarly, since transphobia often takes the form of institutional and economic discrimination and/or family and community rejection, an individual trans person’s financial security becomes their ability to cope with isolation financially and to remove themselves from harmful situations (e.g. a neighbourhood in which they are frequently harassed or a family home in which they are rejected or abused) is key to their ability to survive and thrive in a transphobic society. While all trans people experience and are harmed by transphobia, the extent of that harm will inevitably be strongly classed.

To live as a trans person in today’s society is to frequently find ourselves bumping against the various bureaucracies that serve as its basis, from things as theoretically simple as changing one’s legal name to navigating the complaints procedures of government departments or companies in order to secure some kind of accountability for another instance of transphobia. While this is, in theory, something anybody can learn to do, these bureaucratic institutions are complex and exclusionary by design and often function to favour middle class people. In this way, yet again working class trans people suffer an additional burden from transphobia.

So given that trans people are disproportionately more likely to live in poverty and transphobia’s worst effects are experienced most by working class people, why is this not a part of the media discourse on trans people? Why are some of the most prominent media trans voices wealthy, right wing figures like Caitlyn Jenner? Part of this is precisely because transphobia is strongly classed; as discussed above the wealthiest people will find it easiest to “pass” and meet the standards of conformity to cis-heteronormative standards expected of professional voices in the media. Equally it is the case that middle class and rich trans people are simply more likely to have the necessary connections to be a major media presence. Where it includes trans voices at all, mainstream discourse on trans issues is dominated by an unrepresentative minority of wealthy, white, middle class, trans women. It would be remiss of me not to note an obvious irony here since, while I am far from wealthy and never have been, as a white postgrad student I am myself far from representative of the majority of trans people and, in my defence, I do not claim to be.

A common means of dismissing trans people’s attempts to raise issues that affect us or criticise institutions or public figures that have harmed us as a group is to dismiss us as privileged. Trans people are a bunch of middle class kids or a load of wealthy university students who are just looking for something to complain about. For example, after the well-established journalist Suzanne Moore went on a bizarre, transphobic tirade on Twitter in response to criticism over the wording in one of her articles, fellow career journalist Julie Burchill wrote a piece, initially published in the Observer but eventually withdrawn and then republished by Spiked, which while largely consisting of a series of transphobic slurs also perfectly illustrated this ideological tendency. After claiming that she and other transphobic journalists are “part of the tiny minority of women of working-class origin to make it in what used to be called Fleet Street”, Burchill goes on to depict trans people as academics with “big swinging PhDs”, attempting to silence working class cis women by arguing about “semantics” (the semantics in this case being Moore’s use of “Brazilian transsexuals”, a group plagued by particularly high levels of poverty and violence, as a throwaway pejorative). While trans academics certainly exist, we are far from the majority of trans people or even trans activists, nor are we necessarily as highly privileged as Burchill would like to suggest. By engaging in this erasure of working class trans people, transphobes are able to both trivialise the serious, material effects of transphobia as discussed above and rhetorically exclude trans people from the working class.

In her excellent 2008 essay ‘Liberal Multiculturalism is the Hegemony – Its an Empirical Fact’ – A response to Slavoj Žižek, Sara Ahmed points out that racism is often projected onto the white working class, with liberal prohibitions on overt bigotry serving merely as a means to locate bigotry in some marginalised other. We see a similar process with transphobia, bigotry against trans people is positioned as definitively working class, and thus the existence of working class trans people can be ignored as impossible by definition. A well paid Observer journalist can mock trans people en masse as middle class kids, obsessed with identity politics, because everybody knows that real working class people are white, cishet and hostile to anybody who is not white or cishet. The reality, of course, is that this image of an “ordinary” working class as the default is a fantasy, the working class is a weird, wonderful and diverse class and only a politics that recognises the many and varied ways in which we experience exploitation and oppression can allow us to build a movement to end oppression, end exploitation and ultimately abolish class itself.

Transgender self-identification and data gathering

[Content note: brief, non-graphic mentions of domestic violence, transphobic violence, discrimination, harassment]

The Trans Exclusionary Radical Feminist (TERF) organisation Women’s Place UK recently launched to campaign against proposed reforms to the Gender Recognition Act (GRA). These reforms would include changing the way in which a Gender Recognition Certificate (GRC) would be issued. A GRC allows a transgender person to have the gender marker on their birth certificate changed to match their identified gender, as opposed to the one assigned to them at birth. It does not affect a person’s ability to have their gender changed on most forms of ID, to access most services or to present however they wish. The current process is onerous and expensive, involving the submission of medical reports and various kinds of evidence to a Gender Recognition Panel, who you will never meet face to face, and a £140 fee. The proposed reform instead suggests adopting a system similar to that which has existed, without serious problems or evidence of large scale abuse, in Ireland for the last 2 years, wherein receiving legal recognition of identified gender is a matter of simply filling out a form.

Women’s Place UK has issued “five simple demands” via their social media. In this article, I want to focus on the fifth of these demands – “The government to consult on how self-declaration will impact data-gathering and the monitoring of discrimination.“ Given that the first of their demands is for a “respectful and evidence based discussion” and I have some relevant experience as a postgrad researcher specialising in statistical modelling, I felt that I could speak to the likelihood that the proposed reforms would have any detrimental effect on the monitoring of gender discrimination.

The first question to ask is whether the proposed changes would, in fact, alter the collection of data on discrimination at all. How is this data gathered? Often demographic data is obtained by self-reported questionnaires, in such data a person’s gender quite literally reflects which box they ticked. Whether trans people will typically wait until after they have obtained a GRC to accurately report their gender in data or simply tick whatever box they please I leave as an exercise for the reader. Sometimes data is instead gathered based on pre-existing gender markers on databases (e.g. the NHS database), as with the process of obtaining a passport with a new gender marker or accessing most sex specific services, this does not require a GRC and can usually be arranged by presenting a letter from a doctor or other health care professional stating that the individual concerned has gender dysphoria. In other words, as far as “data gathering and the monitoring of discrimination” are concerned we already effectively have self-identification.

Next, it is instructive to consider the likelihood that trans people self-identifying in survey data (as is already the case) would actually mask any meaningful statistical effects with respect to discrimination. There are two reasons that I would consider this to be unlikely; the small number of trans people in the general population and the reality of trans experience.

Estimates for the number of trans people in the population varying depending on method, studies using self identification (which would, of course, be the most relevant here) tend to find prevalence in Western populations between 0.1% (1 in 1,000) and 0.5% (1 in 200) and a 2016 systematic review found a 95% confidence interval for self reported transgender identity of 0.144%-0.566% with a mean of 0.355%. For perspective this would amount to, at worst, a rounding error for descriptive statistics given to the nearest percentage point, hardly likely to interfere in any serious way with data-gathering.

But what of more sophisticated, inferential statistical methods? What if there were some effect dependent entirely on assigned at birth gender, such as increased likelihood of some negative outcome for people assigned female at birth (AFAB), which went undetected by statistical investigation due to transgender self-identification? This is also somewhat unlikely and in the spirit of a respectful and evidence based discussion, I will attempt to demonstrate so empirically. We can model such a scenario and compare the ability of a standard statistical method used in social sciences (in this case the Chi-Squared test for independence) to detect a meaningful difference in an outcome for AFAB and AMAB individuals both without and in the presence of transgender self identification.

As always in statistical modelling (and particularly while carrying out simulations), we should be clear in the assumptions that we are making. Some of the assumptions I will make are, in reality, not entirely true, but match with the reality implied by trans exclusionary organisations’ stated concerns:

  1. Effects based on gender are determined entirely by the gender a person is assigned at birth, such that a trans woman and a cis man will have the same or very similar outcomes (I will demonstrate that this is simply untrue later in the article, but will allow for such an assumption for modelling purposes)
  2. 0.5% of the population will be transgender, identifying as other than the gender they were assigned at birth (this is perhaps an overestimate, but sits at the upper bound of plausible figures).
  3. We can reasonably represent transgender individuals by changing a binary gender marker (this is not strictly true, and may be understandably contentious for non-binary individuals, but reflects the binary thinking prevalent in trans exclusionary discourse, which this simulation aims to reflect in its assumptions)
  4. AMAB and AFAB individuals are equally likely to be transgender.
  5. The outcome of interest occurs at a rate of 4% for AMAB individuals and 8.2% for AFAB individuals (this was chosen to reflect the real life statistic of annual risk of domestic abuse for men and women in England and Wales, source for this statistic here)

The procedure for the simulation was carried out as follows:

  1. For sample sizes N=[50, 100, 500, 1000, 1,500, …, 10,000]:
    1. Let G be a variable representing assigned at birth (AAB) gender, sample N cases from a binary uniform distribution (0 = AMAB, 1 = AFAB)
    2. Let O be the outcome of interest
    3. For each value in G, G(i):
      1. Sample a random number from a uniform distribution between 0 and 1, R
      2. If (G(i) = 0 AND R < 0.04) OR (G(i) = 1 AND R < 0.082):
        1. O(i) = 1
      3. Else:
        1. G(i) = 0
    4. Use a chi-squared test with alpha = 0.05 to test whether G and O are independent. If test indicates independence, record as a false negative
    5. For each value in G, G(i):
      1. Sample a random number from a uniform distribution between 0 and 1, R
      2. If R<0.005, G(i) = 1 – G(i)
    6. Use a chi-squared test with alpha = 0.05 to test whether G and O are independent. If test indicates independence, record as a false negative
    7. Repeat steps 1.1-1.7 1,000 times

Matlab code for this procedure can be provided upon request and I welcome any attempts to replicate this small experiment.

Addendum 8/12/17: It’s been pointed out to me that the way that I have explained parts of this is a touch opaque. It’s not my intention to blind anybody with science with this piece so, a brief sidebar on the Chi-Squared test and exactly what I have done here.

The Chi-Squared test is a standard statistical method used to test whether two categorical variables are independent of each other. So say you have people belonging to groups A and B and they can also belong to groups X and Y, and you want to know whether membership of group A or B affects probability of being in group X or Y, you might use a Chi-Squared test to determine this.

E.g. you have a sample of cats and dogs and they are all named either Fido or Felix and you want to determine whether being a cat or a dog affects how likely an animal is to have either name. You could look at raw numbers and see that, say, 70% of cats are named Felix vs. 1% of dogs (this would be descriptive statistics), you would use a statistical test like Chi-Squared to work out the probability that you’d get those numbers by pure chance (a p-value), the lower the p-value the more sure you can be that there’s a real effect going on, quite literally asking yourself “Well what are the chances of that?” A p-value of 0.05 (5% or 1 in 20) is often used as the cut off point at which we say a result is “statistically significant”.

In straightforward terms, what I have done is generated data simulating a scenario matching the assumptions I have laid out above and applied the Chi-Squared test many times, each time both with and without allowing for self-identification of gender. For each sample size (number of hypothetical people in each simulated test) I have recorded the number of times that the test failed to detect the presence of a link between gender and our outcome of interest.

Having done this, we can compare the rate of false negatives with and without transgender self-identification. As shown in the figure and table below. Broadly speaking, there is no evidence of a statistically significant difference in false negative rate using AAB and SI gender under the assumptions of this simulation. For samples of size 4,000 or greater, no false negatives whatsoever occurred under either method. The introduction of such a small amount of noise to the data simply does not seriously affect the ability of statistical methods to detect meaningful statistical effects.

Having established that self-identified gender is the status quo for data gathering on discrimination and that it is unlikely to have a detrimental effect on that data gathering in any case, it is also salient to question the underlying assumption implied by the demands made by Women’s Place UK and similar organisations; that gendered outcomes are strictly determined by AAB gender. TERF organisations fear trans women being recorded in statistics on discrimination as women because they assume that that transgender women will have similar outcomes to cisgender men and will thus skew the data and mask discrimination. The reality, of course, is that trans women (and trans people in general) are a marginalised group who experience significantly worse outcomes in several major areas relative to cis people.

A 2015 EU report on trans experience in the EU found that “trans people face frequent infringements of their fundamental rights: discrimination, violence and harassment” and were more likely on average to be in the bottom 25% of earners. Around half of trans respondents reported being harassed or discriminated against in the last year, with this rising to 7 out of 10 trans women (see the graphic, taken from the report, below). A little over a third reported discrimination in finding housing and another third reporting discrimination at work. One in seven trans people and one in six trans women reported experiencing violence or threats of violence in the 12 months before the survey.

A 2007 equalities review by Press For Change and Manchester Metropolitan university had similarly grim findings, reporting both discrimination both economic and social, with a third of trans people being excluded from family events and/or having family members break all contact with them and a fifth experiencing isolation within their community (page 17).

Housing and workplace discrimination, violence, alienation from family and isolation within the wider community. These are common themes when both examining statistics on trans people and when listening to the anecdotal experiences of trans people. The idea that, even if it were somehow impossible to account for whether people were trans within survey data by simply asking, the inclusion of such a group within any larger demographic could serve to hide discrimination against that larger group is simply absurd.

So to return to Women’s Place UK’s demand for a “respectful and evidence based discussion” in the area of data-gathering and the monitoring of discrimination; from an evidence-based perspective, there is, respectfully, no basis whatsoever to any suggestion that making the process of obtaining a GRC less bureaucratic, expensive and invasive could in any way hamper data-gathering or the monitoring of discrimination.