By Jo Phoenix
September 26, 2023
Many of us have argued for several years that accommodating prisoners according to gender identity is a problem. It undermines very hard fought for women’s rights to single sex spaces. It creates a policy induced set of risks to women in women’s prisons estate. In the face of arguments like this, women like me were told that we were wrong to cast transwomen prisoners as sexual predators and that there was no evidence that transwomen prisoners were any more or less sexually risky than women prisoners or the general population. Our response was always the same: we have no evidence to indicate that gender identity overrides what we know about male pattern offending and the male prison population.
We weren’t wrong. We now have, for the first time, factual evidence that confirms what we suspected. The majority of transgender and gender diverse prisoners are male, are serving long sentences and that a significant proportion of transwomen prisoners do, in fact, have histories of violent and sexual offending – including homicide. The breakdown of these individuals’ offending histories makes for some very grim reading. In comparison no trans identifying female prisoners had any history of sexual offending.
Where does this evidence come from? Correctional Services Canada (CSC) own study (“Examination of Gender Diverse Offenders” (R-442), published in 2022. The aim of the report: to provide a complete profile of the transgender and gender diverse prisoner population in Canada’s federal prison system from December 27, 2017 thru March 13, 2020. It is thorough and describes the demographics of gender diverse prisoners in the federal prison system, sentencing and offence information, criminogenic risks and needs, prevalence of abuse and trauma. For those that have committed sex offences, the report documents the nature of their crimes; as well as the ‘institutional adjustments’ (known as individualised protocols) made for gender diverse prisoners and their post-release outcomes (i.e. rates of rehabilitation and recidivism).
Rather than provide a general overview, my summary highlights what the report tells us about demographics, sentence length, and offenses of a violent or sexual nature. These aspects warrant special attention because they speak directly to widespread concerns about self-identification policies in prisons.
The report has its methodological strengths. On the plus side, the study captured the total population of individuals in the federal prison system who informed CSC that they were either trans, gender fluid, gender non-conforming/nonbinary, intersex, two-spirited or an unspecified gender. In other words, this is not a representative sample drawn from the population – a method which is only suitable where one can describe the population’s characteristics.
Nor is it a convenience sample, meaning that inclusion in the study is reliant on people coming forward to participate. Convenience sampling has, to date, been the most common methodology used by researchers seeking to understand the challenges and experiences of gender diverse prisoners. Whilst convenience sampling is a very useful method for eliciting qualitative data, no generalisations about group characteristics can be made using this method. Generalisations about group characteristics can only safely be drawn from a full census of the relevant population.
There were only 99 gender diverse prisoners in a total federal prison population that fluctuated between 12,000 to 14,000, comprising less than 1% of all prisoners. The study classified the prisoners into three groups: transwomen, transmen or other (gender fluid, gender non-conforming/non-binary, intersex, two-spirited or unspecified).
As noted above, transwomen make up by far the largest cohort of the gender diverse prison population. Nearly two-thirds of all gender diverse prisoners (61 of 99) were transwomen. Of the remaining 38 prisoners, 21 were transmen and 17 were other.
The breakdown of these categories is messy because the authors lump together biological sex and legally recognised gender. Just to remind the reader: legally changing gender in Canada does not presuppose that someone has had any medical, hormonal or surgical treatment (often referred to as ‘medically transitioning’). For most provinces in Canada, one need only apply to change gender on their foundational documents (birth certificate, passport etc). A few provinces require supporting documentation from a psychologist or doctor. So,
of those 61 transwomen, 13 had a biological sex or legal gender of female and 48 were recorded as biologically or legally male. My interpretation of this data is that the 13 ‘female’ transwomen were those who had gone through the process of legally changing their gender. I can think of no situation natal female would become transwoman. A biological female who changes their legal gender would, by definition, not be a transwoman but a transman. Hence, for all intents and purposes, the 61 incarcerated transwomen in the study were natal males.
By this logic, of the 18 of the 21 transmen classified as biologically female and 3 had changed their legal gender to male. Regarding the 17 ‘others’, it is impossible to disaggregate this group because the researchers had conflated biological sex with legally recognised gender. So, of this ‘other’ group, 7 were either biologically or legally female and 10 were either biologically or legally male. This renders the ‘other’ category analytically meaningless when relating gender diversity among prisoners to what is already known about sex-based patterns of offending.
Two-thirds of transwomen (41) were housed in men’s prisons. This means that fully one-in-three incarcerated transwomen were accommodated in the women’s estate. The 17 prisoners categorized as ‘other’ were split evenly between women’s and men’s prisons. 20 of the 21 transmen stayed in the women’s correctional facilities while only 1 cohabitated with the male prison population. The significance of these numbers ought not be underestimated. They constitute strong empirical evidence that the adoption of self-identification policies by CSC is having a disproportionate impact on the composition of women’s prisons. 1 transman was accommodated in one of the 53 federal men’s prisons. 20 transwomen were accommodated in one or more of the 6 CSC women’s facilities.
One of the surprising characteristics of gender diverse prisoners in Canada’s federal prisons is that they are disproportionately from indigenous ancestry. Slightly less than half of all gender diverse prisoners were indigenous (46 of 99). Whereas indigenous peoples make up around 4% of the total Canadian population, they make up a quarter of all federal inmates.
But there is nuance needed. We have known for some time that indigenous women are more over-represented in the female federal prison population than indigenous men are in the male population. In 2018, 40% of women incarcerated in Canada were of indigenous ancestry compared to 32% of men serving time in federal prisons. Looking at the gender diverse population, more than half of the transmen (12 of 21) were of indigenous ancestry whereas 44% of the transwomen (27 of 62) were of indigenous ancestry. This is broadly consistent with the gender skew in the Indigenous inmate population.
The CSC does not, at present, disaggregate prisoner data by sex and sentence length, making it difficult to compare sentences handed to male and female prisoners. We can, however, draw inferences from peer nations. In the UK, in 2021 for example, 17% of incarcerated women, and only 7% of incarcerated men, were serving sentences of less than a year.
The pattern from the UK is consistent with the sentences described in the report about gender diverse prisoners in Canada. The CSC report evidences that 30% of transwomen (18 of 61) were serving a sentence of 6 years or less. The remaining 70% (i.e. 43 of 61) were serving sentences of 6 years or more or had indeterminate sentences. By comparison, over seven in 10 transmen (15 of 21) were serving sentences 6 years or less. Put simply, transmen committed less serious crimes that resulted in shorter sentences.
Violent offenders and those with a history of sexual offences
If transwomen prisoners, as a category, retain what might be called ‘male pattern offending’ (i.e.: committing more violent and sexual offenses than women), this would introduce three interrelated problems into women’s prisons: the need for increased security to deal with higher risk prisoners; the increased risk of sexual violence against female inmates; and the de-prioritisation of biologically female prisoners’ needs with respect to safety, security and possible re-traumatization – particularly among those who have previously experienced domestic or sexual violence. Hence, knowing the offending profile of transwomen is critical for the safety of both female prisoners and female prison staff. This report outlines exactly this.
More than 90% (55 of 61) of transwomen prisoners were incarcerated for violent offences. Of the group, nearly half (25) had a most serious offence that was homicide related and a third (18) had a most serious offense that was sexual in nature. In comparison, fewer than three-in-10 (6 of 21) transmen were convicted of homicide related offences. This proportion of transwomen incarcerated for sexual and homicide-related offences is extraordinarily high compared to the general female prison population. This raises a number of serious questions about the relationship between gender diversity for women, violence and criminalisation.
One telling insight provided by the report is that 27 transwomen prisoners (44%) have a history of sexual offenses compared to 0 transmen. It is worth noting that 6 of the 17 ‘other’ category also had a history of sexual offenses; 10 of whom were either biologically male or legally changed their gender to male. Alarmingly, nearly all gender diverse prisoners with a history of sexual offences committed those sexual offences before identifying as trans or non-binary (31 of 33). 85% (28) committed a schedule 1 offence (that is caused death or serious harm to their victims). Over half (18) had exclusively victimized women, slightly less than half (16) committed sexual offences against minors. Just 2 had exclusively victimized men.
Earlier in this article, I noted that 22 transwomen prisoners were accommodated in women’s federal prisons in Canada. 26 transwomen had no history of sexual offending. Strictly on numbers, it is conceivable that none of the transwomen with histories of sexual offending have been sent to the women’s federal prisons. This, however, is unlikely given the anecdotal stories that have come to light since CSC introduced its self-identification policy (see rights and wrongs).
As I said above, the report covers a lot of ground, including raising questions about the extent to which interventions and programmes developed for women are suitable for transwomen. The report also generated some distressing but critical information about the histories of abuse and trauma experienced by gender diverse prisoners. This is the one aspect of the data that confounds some of what we know about sex-based patterns of victimisation (i.e. that women in prison compared to men in prison are disproportionately likely to have experienced abuse and trauma). Analysis of Offender Management System information indicated that seven in 10 gender diverse offenders had childhood histories of abuse. This upsetting statistic points to the need for more research on the interlinkages between histories of abuse, criminal activity and incarceration – especially as it pertains to childhood traumas.
Far fewer gender diverse prisoners (1 in 4) reported abuse during adulthood. 5 transmen fell into this category, in line with the average. Given that the percentages of women in prison who have been abused as adults sits anywhere between 60-80%, this indicates that the pathway from victimisation into incarceration may not be present for transmen. Although, again, the small number of transmen in the incarcerated population makes it difficult to draw any strong conclusions.
There are some noticeable limitations in the findings presented in the report. For the most part, the authors summarise the data they presented with rather than analyse it. They do not, for example, cross-check gender diverse prisoner populations with CSC data on the general prisoner population broken down by sex. (They can hardly be faulted for this because that data, to my knowledge, does not exist). As such, any possible comparisons, at this stage, have to use findings from other countries (e.g.: the UK) as a proxy for the (non-existent) Canadian data. That said, where the authors do attempt to analyse the data, they fail to employ sound methodologies.
For now, we know substantially less about gender diverse prisoners than we do about the broader population of gender diverse individuals. The authors of the report state that US-based research finds that gender diverse individuals are over-represented in the American prison population. They then draw from this the conclusion that there must be ‘missing’ gender diverse Canadian inmates. Instead of concluding, from the available evidence, that transgender and gender diverse law breakers are disproportionately incarcerated, they opine that there must be transgender and gender diverse prisoners who do not ask for gender diverse accommodations or come out as gender diverse.
This is a rather odd conclusion. To begin with, there is no reason why the percentage of gender diverse prisoners *ought* to reflect the wider population of gender diverse individuals. All sorts of people break the law, but there are distinct processes that bring certain demographics to the attention of the criminal justice system. Other social factors, like poverty, go a long way to explaining why the prison population (for both males and females) is not in the slightest bit representative of the general population. The overwhelming majority of all people in prison come from economically marginalised backgrounds.
The authors also note in the conclusion that “almost two-thirds of the offenders in this study were serving a sentence of six years or more… compared to 47% in the general offender population” (pg32). The authors note that one could infer from these numbers that gender diverse offenders commit more serious offences than the general federal prison population. However, they contradict themselves by quoting two pieces of research (neither from Canada) that suggest that trans and gender diverse individuals are more likely to commit drug-related, sex trade or property offences.
From my perspective, this is an extraordinary claim to make given the data the report shows. The disproportionate seriousness of offenses committed by the Canadian gender diverse prison population is attributable to the fact that so many transwomen in the system have committed serious homicide related and sexual offences. This finding begs a different, more urgent, question: given that so many transwomen prisoners in this study committed sexual offences before transitioning to their current gender, what is the relationship between committing sexual offenses as a biological male and later “coming out” as gender diverse while incarcerated?
The authors further claim that the greater number of transwomen in the gender diverse population is attributable to the fact transwomen are more likely to experience verbal, physical, and other assaults while in men’s prisons and, as such, are more likely to request gender accommodations. As evidence of this claim, they draw on an article that repeats the findings of some of the key studies of transgender prisoners’ experience in the United States. This is extraordinarily weak evidence; comparing American prisons with Canadian prisons is like comparing apple pie to poutine (see Rights and Wrongs).
For me, this report indicates something more fundamental: the fact that Canadian gender diverse prisoners have offending profiles that more closely resemble their respective biological sexes means that the approach of conflating gender identity with biological sex necessarily involves denying half-a-century of research about the differences between the needs, risks and histories of men and women who end up in our corrections systems.
About the author
Jo Phoenix is professor of Criminology at the University of Reading, where she researches sex, gender, sexualities and justice, youth justice and punishment, the production of criminological knowledge and research ethics. She has studied and written about a wide variety of subjects including managerialism and ethics in the production of criminological knowledge, prostitution, prostitution policy reform, child sexual exploitation, youth penality and youth justice practice and policy. Her most recent research concerns academic freedom, politics ethics and research and sex, gender, gender identity and criminal justice policy.