The Transgender Movement and Bad Stats: A Debunking Compilation

It has come to our attention, that, scattered across half a dozen posts, is debunking of a variety of statistics associated with the transgender movement. We fear they may be a little buried in some very long posts. We wondered how to fix this problem. The solution is this article.

Nothing excites readers like the Medium equivalent of a television clip show, which is exactly what this is. We’ve decided to gather all those statistics together in one handy article, so you can reference it in all your online Twitter debates, as God intended.

The US Transgender Survey is a source of many statistics about transgenderism you will find in international policy debates, arguments on the internet, and cited by LGBTQI+ activist organizations. It is run by the National Center for Transgender Equality (NCTE).

It describes its survey to the IRS with the following:


The IRS form lets us know how much that survey cost — $318,154. So, what information about the transgender community did $318,154 give us?

We took a look at the lauded NCTE survey, the National Transgender Discrimination Survey(NTDS) which is downloadable from their website, to find out about what information it can give us.

The problem is that the survey, despite its six figure costs, contains numerous methodological flaws, rendering it’s information useless. It isn’t worth discussing what the survey actually shows us, because it is a survey where the sample was built on self-selection. It isn’t random. The survey, which was run online, had as its first question ‘have you already taken this survey before?’, and warned that taking the survey repeatedly would not increase the number of entries into a prize draw (you can view a screenshot here). That meant the survey could have been taken over and over again by the same person. It was also meant to provide US-based statistics, but had no geo-location restrictions. That’s not a valid data-set. That’s not even going to pass an undergraduate statistics course. Supposedly NCTE cleaned the data-set, but I am not sure how you can clean a survey with such flaws. It should only serve as an indicator for further research at best, not a bible or a reason to bring about legislative change. It brings into question every statistic in the survey. Other criticisms were that it tried leading participants into a particular response.

It is an issue because the survey has gone on to shape public policy and be cited by numerous other organizations. The survey, which is incredibly flawed, has been cited numerous times by other associated transgender lobby organizations — the Human Rights Campaign, the Transgender Law Center, the National LGBTQ Task Force, a litany of other lobby groups and the Democratic Party, all groups that use its statistics as crucial evidence for their argument that transgender people are the most oppressed minority in America. Despite its methodological flaws, it was published, and proudly sponsored and cited by a number of corporate and philanthropic foundations. This was also used by groups funded by these organizations as electioneering material”, and for lobbying purposes — to advance an agenda. But if the survey is flawed as it is, why not try and find better statistics? And why use bad statistics to advance an agenda? And this is not the only example of bad transgender statistics.

Bad Stats Go To Prison

The assertions of many of these groups and media outlets — that trans women are stopped more frequently by police, for example, are just simply not true or lacking in evidence. For example, in this Vox article, we are told, for example, that according to a report by Movement Advancement Project and the Center for American Progress, that ‘Police often target LGBTQ people, particularly transgender women’, and told that ‘Gay, lesbian, and bisexual people are twice as likely to be incarcerated. Rates are even worse for transgender people.’:

“According to the National Inmate Survey, in 2011- 2012, 7.9% of individuals in state and federal prisons identified as lesbian, gay, or bisexual, as did 7.1% of individuals in city and county jails. This is approximately double the percentage of all American adults who identify as LGBT, according to Gallup (3.8%).

Sixteen percent of transgender and gender non-conforming respondents to the National Transgender Discrimination Survey indicated they had spent time in jail or prison, with higher rates for transgender women (21%) and lower rates for transgender men (10%). Comparatively, about 5% of all American adults will spend time in jail or prison during their lifetimes.”

The problem is, is that’s just not true. I took at look at the report they’re citing in the article. Trans people aren’t over represented in prison: but the LGB are.

The report states roughly 7.9% of inmates in state and federal prisons were lesbian, gay or bisexual. Trans isn’t included in that statistic. That’s disproportionate compared to Gallup’s count of LGB people, which is around 3.8% of Americans.

Those statistics are from the same 2011–2012 National Inmate Survey that Vox is citing in that article. According to the same survey, 5,000 inmates identified as transgender. (this is all in state and federal prisons). The Williams Institute estimates the transgender demographic to be 0.6%-1% of the population.

The current US prison population is 2,220,300 (according to a 2013 report), which means that 0.2% of the US prisoner population consists of transgender prisoners , which is actually under-representation for their demographic. Trans people are less likely to be imprisoned. It certainly illustrates very neatly the problem with talking about ‘LGBTQI statistics’ and treating a set of very diverse groups as a monolith and erasing the LGB in the process. It is clear from the statistics provided that discrimination exists. That discrimination is towards same-sex attracted individuals — the LGB.

Including their statistics with groups that are not same-sex attracted is simply a recipe for erasing discrimination against same-sex attracted individuals. This is not just the case with prison statistics, but others too, such as homelessness statistics. The popular figure is that 40% of homeless youth are LGBT. I have seen this re-purposed into 40% of homeless youth are trans, or are queer, etcetera. The reality is very different — the 2012 Williams Institute survey that is the source of that information says, well, I’ll quote for you:

“The findings from this survey demonstrate that many LGBT youth are at high risk of homelessness, often as a result of family rejection and abuse. The analyses offer critical insights into the challenges that these young people face when they seek help during a very difficult time in their lives,” said Laura E. Durso, Williams Institute Public Policy Fellow and study co-author.


Among the key findings:

* 94% of respondents from agencies work with LGBT youth

* 30% of agency clients identified as gay or lesbian

* 9% identified as bisexual

* 1% identified as transgender”

Yet you will see this figure repeated as ‘LGBT homelessness’. Yet of that 40%, 39% were same sex attracted. 30% of homeless youth are gay or lesbian and erased from their own statistics, which have been re-purposed for the transgender movement. Those are the figures you never hear.

Of course, there’s always more bad stats to debunk involving the criminal justice system and transgenderism. But very few of them come from valid sources. For example, you might hear that police disproportionately target the LGBT, or that that don't take hate crimes seriously. The problem is that those aren't great statistics either. They appear to come from this report from the National Coalition of Anti-Violence Programs(NCAVP). The Arcus Foundation, notorious for funding transgender astroturf, provided ‘generous support’ and funded the survey.

Even in that report, we get this statistic:

“Of the total number of survivors, 47% identified as gay, 17% identified as lesbian, and 14% identified as heterosexual”

There are two problems here. The first is that this is no where near a representative sample of the LGBT community when it only 17% of its respondents are lesbians (which would further be distorted by 'lesbians with penises' i.e heterosexual men).

The vast majority of those surveyed as victims of violence were homosexuals. That heterosexual figure is heterosexuals being mistaken for homosexual and then attacked on that basis. This agrees with FBI statistics, which recorded 1,303 hate crimes based on sexual orientation, and 131 as a result of gender identity bias. That means that, according to FBI figures, 16% of single-bias hate crime incidents were from bias against sexual orientation, and 1.7% were a result of gender identity bias. (For reference, 59.5% of single-bias hate crime instances were racial in nature, and the vast majority of hate crimes are single-bias incidents). Those FBI figures come from 16,149 law enforcement agencies — which is almost certainly a very good sample. The NCAVP survey, by contrast, surveyed thirteen local member organizations for their report. It isn’t a survey of the general population, nor representative of the general population. It’s a survey of people who engaged with the NCAVP.

That means we can say that trans women who engage with the NCAVP are six times more likely to be assaulted by police. But we cannot apply that to the general population, which is what many of these groups are doing.

This is not the only statistical error that is commonly cited when relating to trans people in the justice system. For example, you may read that trans women are more likely be assaulted in a men’s prison. This comes from a 2007 California study, which used a convenience sample of transgender women in California’s prison system, and then compared that convenience sample to a sample of the general prison population, which is to put it mildly, a statistical abortion. The amount of transgender women sampled was 39, and compared to a random sample of 322 male California state prisoners. A later 2009 study on the entire transgender population of men’s state prisons in California found that trans women prisoners were more likely to be in maximum security (Level 4 Custody in California), with 32.1% of transgender prisoners in California men’s prisoners, compared to 22.8% of the general population in California. 29% of transgender prisoners were in Level 3 Custody compared to 24.1% of the general population. They were also more likely to be sex offenders — 20.5% of transgender inmates were sex offenders, compared to 14.6% of the general population. They were also more likely to be mentally ill — while 26% of California’s male prison population had been diagnosed, 71% of transgender inmates had ‘ever had a mental health problem’, and 66.9% had had a mental health problem since their incarceration. 20% of them hadn’t presented female until their most recent incarceration which lends credence to claims that male inmates are identifying as trans to take advantage of special privileges and the ability to transfer into a women's prison, as self-ID laws remove all protections from women.

Not quite the rosy picture painted by activists, is it?


Partcipate in this debate long enough, and you’ll see the repeated claims that trans people attempt or commit suicide at extremely high rates. It’s used to browbeat opposition into submission. But where does it come from?

That 41% suicide statistic comes from a report done in 2014, based on data from 2008 in the National Transgender Discrimination Survey, from the Williams Institute, part of UCLA School of Law. Here is a link to the William’s Institute report. Of course, they debunk their own statistic on the third page of the report. How convenient for me.

“While the NTDS provides a wealth of information about the experiences of transgender and gender non-conforming people, the survey instrument and methodology posed some limitations for this study. First, the NTDS questionnaire included only a single item about suicidal behavior that asked, “Have you ever attempted suicide?” with dichotomized responses of Yes/No. Researchers have found that using this question alone in surveys can inflate the percentage of affirmative responses, since some respondents may use it to communicate self-harm behavior that is not a “suicide attempt,” such as seriously considering suicide, planning for suicide, or engaging in self-harm behavior without the intent to die (Bongiovi-Garcia et al., 2009). The National Comorbity Survey, a nationally representative survey, found that probing for intent to die through in-person interviews reduced the prevalence of lifetime suicide attempts from 4.6 percent to 2.7 percent of the adult sample (Kessler et al., 1999; Nock & Kessler, 2006). Without such probes, we were unable to determine the extent to which the 41 percent of NTDS participants who reported ever attempting suicide may overestimate the actual prevalence of attempts in the sample. In addition, the analysis was limited due to a lack of follow-up questions asked of respondents who reported having attempted suicide about such things as age and transgender/gender non-conforming status at the time of the attempt.”

Oh. It’s inflated. Because it was a binary question and may include all self-harm attempts. Studies done on those binary questions have shown that it can completely inflate your results.


Worse is yet to come though.

“ Second, the survey did not directly explore mental health status and history, which have been identified as important risk factors for both attempted and completed suicide in the general population (Lasage, Boyer, Grunberg, Vanier, Morissett et al., 1994; Suominen, Henrikssen, Suokas, Isometsa, Ostamo, et al., 1996; Harris & Barraclough, 1997; Bertolote & Fleischmann, 2002; Nock, Hwang, Sampson, & Kessler, 2010). Further, research has shown that the impact of adverse life events, such as being attacked or raped, is most severe among people with co-existing mood, anxiety and other mental disorders (Breslau, Davis, Andreski, & Peterson, METHODS AND LIMITATIONS 4 Methods — continued 1991; Kendler, Kardowski, & Presco, 1999). The lack of systematic mental health information in the NTDS data significantly limited our ability to identify the pathways to suicidal behavior among the respondents”

They don’t know why the rate is so high — so you can’t say 41% of transgender people attempt suicide because of ‘lack of acceptance’ or ‘bathroom bills or ‘Donald Trump’. Because the study didn’t ask those questions. That would be the case even if the study didn’t have major methodological problems anyway:

Third, since the NTDS utilized convenience sampling, it is unclear how representative the respondents are of the overall U.S. transgender/gender non-conforming adult population. Further, the survey’s focus on discrimination may have resulted in wider participation by persons who had suffered negative life experiences due to antitransgender bias.1 As the relationship between minority stress and mental health would suggest (Meyer, 2003), this may have contributed to a higher prevalence of negative outcomes, including lifetime suicide attempts, in the sample. These limitations should be kept in mind in interpreting the findings of our analyses.

What’s a convenience sample? How is that a methodological flaw? Simply put: the results of a survey of a convenience sample are only relevant to that particular sample. How?

Say I asked ten friends about whether they liked purple hats. As it turns out, all those friends like purple hats. I cannot then go and say ‘one-hundred per cent of people like purple hats’. I only asked my friends — maybe we all belong to the Purple Hat Club. Convenience sampling introduces too much bias for results to be meaningful outside of the sample itself.

In fact let’s have this paper in Developmental Review explain it better than I can, because you can’t use a convenience sample like that:

“Regarding its disadvantages, results that derive from convenience sampling have known generalizability only to the sample studied. Thus, any research question addressed by this strategy is limited to the sample itself. The same limitation holds true for estimates of differences between sociodemographic subgroups. As another disadvantage, convenience samples typically include small numbers of underrepresented sociodemographic subgroups (e.g., ethnic minorities) resulting in insufficient power to detect subgroup differences within a sociodemographic factor or factors. Moreover, although small in number, these underrepresented sociodemographic subgroups introduce modest amounts of variation into the sample, enough variation to produce statistical noise in the analyses but not enough variation to harness or control statistically. Indeed, the widespread use of convenience sampling may be partly responsible for the host of small and inconsistent effects that pervade developmental science, why sizes of effects often vary depending on the variables considered, and why research shows links between particular setting conditions and outcomes for some, but not other, groups”

That 41% stat is bogus. As is everything else in the 2008 National Transgender Discrimination Survey. I’ve officially debunked it. Well done me. Unfortunately, that will not stop the effects of citing its statistics for a decade even though it used a convenience sample and you can’t generalize those statistics to the broader transgender population.


The most disturbing thing about all of this? You can find the description of the study’s methodology on page three. It literally takes some basic curiosity and five minutes, to find out that 41% statistic is statistical noise and not representative of the transgender community. To find that out, I Googled the statistic, and the report was the first result. I then read the PDF.

That is all the effort it took to point out that this statistic is a load of crap. It didn’t stop the media citing it, or activists citing it to policymakers though.

Instead, those groups use it frequently, despite the fact it’s a statistical artifact. It’s been cited to policymakers — and its horseshit. The suicide statistic is a false number used to deceive and scare people.

Yet another scare stat — like the fake murder epidemic.

Fake News And Murder Statistics

We’re often told that there’s a transgender murder epidemic: that trans women are the most likely demographic to be murdered, and therefore the most oppressed, and so on.

I decided to verify this, and GLAAD (which is, remember, an acronym now devoid of meaning. How’s that for symbolism?) gave me the answers. How many homicides of American transgender people were there in 2016? I am sure you are waiting with bated breath for some kind of titanic, earth shattering number that will have you click ‘exit tab’, and bitch about my bullshit article on Twitter. Okay, here it is:


That’s not a typo. It really is 27. The number of total murders in the US in 2016? 17,250, and disproportionately trending black and male. 27 is 0.15% of murders in the US. In terms of figures, the Williams Foundation did a survey and estimated the number of trans people at 0.6% of the US population. The US population is estimated at 325 million at time of writing, which results in a figure of 1.95 million trans people across America.

We’ll take 1.95 million Americans. If we figure how many trans people are victims of murder a year as a percentage, that figure is 0.0013%. Per capita,that’s a ratio of 1.3 trans people murdered per 100,000. The murder rate of women in the US is triple that, and of men, quadruple. Even with an extremely conservative estimate of 0.1% of the US population (or 325,000 trans people), we have a murder rate of 8.3 per 100,000. The murder rate of Chicago is twice that conservative figure at 16.02 people murdered per 100,000. In terms of gross numbers — that’s 11,535 murders of male Americans, and 3,292 murders of female Americans in 2017. 27 is small potatoes. That is not a murder epidemic — in fact it’s a murder rate per capita lower than Canada. It certainly doesn’t mean that there’s an ‘epidemic of transphobic violence’. That’s not something to campaign about — you’ve got it better than literally everyone else. Even if we use the Human Rights Campaign estimate of 750,000 trans people, which is half the 0.6% number, we get a murder rate of 2.7 per 100,000. That’s not a high murder rate. That’s lower than every other demographic in the US.

We are told, that supposedly, 1,700 trans people have been murdered worldwide over the last seven years in this article on Buzzfeed. Your eyes immediately drift to the ‘1,700’ figure, and don’t see the 7 years, do they? That’s why I bolded it. It’s fairly obvious statistical sleight of hand. If we take the 0.6% estimate of trans people in the US and apply it globally to a population of 7 billion people, we get 42 million people. 1,700 divided by seven years gives us a grand total of 242 murders a year. That amounts to 0.003 murders per capita of trans people, worldwide, every year. That’s definitely not an epidemic. In fact, that’s a global murder rate lower than every other category on earth. The murder rate per capita of unicycle-riding clowns is probably higher. To #StopTransMurders would be to eliminate homicide for an entire group of people, which no nation has been able to accomplish. Ever.

Maybe you think 42 million trans people on this earth is too big a number. So, we’ll make the figure 5 million people. A murder rate of 242 per year of a group of 5 million people is still a per capita rate of 4.84 — roughly similar to the US overall murder rate of 4.7. And that’s with a hugely conservative number that I literally pulled out of thin air. All these figures say the same thing — there is no trans murder epidemic — and philanthropic groups and their funded organizations supporting trans rights and sympathetic media have to perform statistical sleight of hand to even make such a proposal look even the slightest bit true. For comparison, the highest murder rate in the world belongs to Honduras, which had 90.4 homicides per 100,000 in 2017. That’s a violent epidemic. To add — the majority of those 27 killed? Black prostitutes. No middle-aged white trans women were killed at all (though some did commit murders) yet they are the ones bleating about #StopTransMurders and working in activist organizations. And the sex-work and transgender lobby does not seem to care about those vulnerable prostitutes, beyond using their names and deaths as a political prop.

If you’re wondering about the gay and lesbian side of things, rather than the transgender epidemic that doesn’t exist, yet is talked about so heavily, the FBI reports in its latest Hate Crimes report state that 16.7% of hate crimes were motivated by sexual orientation. 1.7% were motivated by gender-identity bias. Of the 1,255 victims targeted by sexual orientation, 62.7% were anti-gay male, 21.6% were LGBT (mixed group), and 11.7% targeted towards lesbians. There were 131 victims of ‘gender identity-bias’, 20 of whom were simply ‘gender-non-conforming’. In terms of hate crimes (which is criminal offenses carried out motivated by bias, not necessarily violent) gay men are disproportionately over-represented among the LGBT. That’s an actual disproportionate epidemic of violence — rather than the trans murder epidemic that doesn’t exist.

When was the last time you saw that on BuzzFeed?

But this is not the only ‘transgender murder epidemic’ article on Buzzfeed. The author of that article on Buzzfeed I just linked you, Dominic Holden, wrote a feature entitled Why Are Black Transgender Women Getting Killed In Detroit that uses the same sleight of hand, saying that the murder rate has ‘doubled’ yet doesn’t give you a number. I looked into the source it cited, and the murder rate doubled from 12 murders to 24 murders. That figure comes from the National Coalition Of Anti-Violence Programs, counting between Transgender Days of Remembrance. That’s an even lower figure than the GLAAD data! And it is not just Buzzfeed doing this. A quick Google search leads to more repeating of the ‘trans murder rate is so high it’s an epidemic’ meme . In fact, I googled ‘trans murder epidemic’ and got 535,000 results from Wikipedia, to the Human Rights Campaign, to ‘America’s transgender murder epidemic: why is nothing being done?’ from a UK website called ‘Blasting News’. It even appears as the beginning of the National LGBTQ Task Force’s 2016 annual report, highlighting its #StopTransMurders campaign. But the facts and figures say there isn’t an epidemic. Rather the opposite — trans people have the world’s best murder statistics, as a group. The ‘epidemic’ is easily debunked using LGBTQI+ groups’ own statistics. The trans murder rate is a false meme worthy of inclusion in a late 90’s chain email promising you the truth about Bill Clinton and Whitewater. It’s literal fake news.

The statistics aren’t the only fake news on this matter. Journalist Andy Ngo recently wrote about the queer community faking hate crimes in Portland, including a trans woman who said they had been attacked by thugs with bats, and started a GoFundMe to support themselves, but later turned out to have fallen over while intoxicated:

“Last month, Sophia Gabrielle Stanford was at the center of a fundraising campaign. The GoFundMe page described the trans activist as a victim of a “brutal and aggressively blatant hate crime” in which assailants had beaten her unconscious with a bat in southeast Portland.

The campaign and shocking story went viral. However, the police reports raise questions about what happened that night.

In the early hours of Sunday, Feb. 10, emergency services received a call about a woman, identified as Stanford, found on a sidewalk with scrapes on her face and knuckles, claiming that she may have been assaulted. The responding officer, Edgar Mitchell, noted that Stanford smelled of alcohol.

“I asked [Officer Zachary Roe] what happened,” the report states. “Roe said the individual admitted to being intoxicated, and Roe believed the person fell and hit her head.”

Stanford either could not or would not state her name to the police. The responding officer was unable to discover Stanford’s name and claims that she made a threat: “If you don’t treat me right, my people will get you,” she said, according to the report.

The report also states that Stanford lost a pistol and bag she was carrying at the time of the alleged attack. A local resident found both items and flagged down another officer, Cuong Nguyen. When Nguyen attempted to return the gun to Stanford at nearby Emanuel Hospital, where she had been transported, she was already discharged.

The GoFundMe page stated that Stanford had suffered a “serious concussion” and would need intensive physical therapy, CT scans and counseling”

Using fake statistics only fuels things like this — GoFundMe scams and wasting police time. I don’t have time for graft. Do you?

Oh, I forgot, and fueling a fire that doesn’t need to be fueled — all this sort of thing does is promote an unnecessary moral panic.

Bad Stats Help Nobody

The thing I always ask myself when I see these sorts of things, is well, why? Why overstate, misinterpret, or flat out lie? Why not have a survey with a rigorous, objective method of data collection?

The trans movement has a data problem. It has a data problem with its medicine, and it has a data problem with its prison statistics, its ‘murder epidemic’, and even its suicide statistics. Where it doesn’t have data, it appropriates the statistics of gays and lesbians, as we see with homelessness statistics — where 40% of homeless youth being ‘LGBT’ slowly turns into ‘40% of homeless youth are trans’, despite the fact that 39% of that 40% aren’t transgender at all. Talk about gay and lesbian erasure — they don’t even get to have their own statistics any more, even when they’re not very good ones!

There are a variety of hypotheses that come to mind for me when I attempt to answer ‘why?’. One is that an actual study that wasn’t a complete mess would show statistics that the transgender movement wouldn’t like to hear, or have the public see. They could undermine their narrative and send their movement reeling. Actually, come to think about it, this is the only hypothesis I have. It certainly explains the flat out dishonesty of the transgender movement. It is not as if the transgender movement is lacking in money. Surely the over $300,000 spent on the NCTE survey could have been put to better use than an online survey and a glossy brochure.

Because if we look at the statistics we do have, the transgender movement and it’s claims are clearly undermined — and that’s a problem. Unless they’re hideously oppressed, demands to completely undermine the concept of ‘sex’ in the legal system might be sensibly ignored by policymakers. Instead, because they’re so oppressed, because the ‘statistics’ say so, the transgender movement gets a seat at the identity politics table.

It needs to stop. Now. These claims aren’t backed by hard data. They’re barely backed by any data at all. Before we make changes that could permanently sterilize children and erase gays and lesbians from their own movement, we should really have a better idea of why the transgender movement wants those things.

Written by

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store