Sexual Orientation, Gender Identity, and Crime

In 2025–2026, claims about LGBTQ+ people and crime directly drive federal and state policy. President Trump’s Executive Order 14168 (January 20, 2025) aimed to prevent “men [from] self-identify[ing] as women and gain[ing] access to intimate single-sex spaces.” The Heritage Foundation’s Project 2025 recommended that educators who “purvey transgender ideology” be registered as sex offenders. The ACLU tracked 339 anti-LGBTQ+ bills by February 10, 2025 alone, with over 1,022 anti-trans bills introduced across 49 states. At least 23 states had school bathroom bills by early 2026. Use of the term “groomer” directed at LGBTQ+ people surged 406% on social media following Florida’s “Don’t Say Gay” law (HRC/CCDH).

A per-capita comparison across all groups — including heterosexual cisgender baselines — is the only methodologically honest approach. Raw prison counts without population denominators prove nothing about relative risk.

The denominator choice for trans populations is the single most consequential variable in this entire debate. ONS itself downgraded the Census gender identity data in September 2024, noting question misinterpretation among non-English speakers. Using the ONS figure (~48,000 trans women) versus the GEO tentative estimate (~200,000+) changes per-capita rate calculations by an order of magnitude.

This is the dominant empirical reality against which all other groups must be measured. Men commit the overwhelming majority of sexual offenses across every available dataset:

Any claim about a minority group’s crime rate must be understood against this baseline: heterosexual cisgender men account for 93–98% of sexual offenses by every measure.

No population-level per-capita sexual offense data exists for gay men specifically. The best evidence comes from the Netherlands:

U.S. data shows LGB people overrepresented in prison (1,882 per 100,000 vs. 612 general rate), but this reflects criminalization history — sodomy laws (enforced until 2003), HIV statutes (felonies in 28 states with 20+ year sentences in 5 states), and solicitation enforcement — not elevated violent or sexual offending. The child abuse question is addressed in Part 4.

No per-capita sexual offense data exists for this group. They are almost entirely absent from “predator” narratives, which is itself revealing: if the debate were about data, lesbian women would be analyzed with equal rigor. Their absence from the narrative confirms the discussion is about male-bodied people in female spaces, not about LGBTQ+ identity per se.

LBW are massively overrepresented in U.S. prisons (42.1% of women in state prison identify as sexual minority), but this reflects the “trauma-to-prison pipeline” — 53.7% of incarcerated sexual minority women experienced childhood sexual assault vs. 31.4% of straight women.

This is the most politicized dataset. The headline numbers are real; their interpretation is where errors occur.

These proportions are real and striking. But three factors fundamentally change how they should be interpreted. All three are addressed in Part 3.

Trans men are almost completely absent from offender data. In Canada (CSC R-442), zero of 21 trans men in federal custody had sexual offense histories. In the UK, no trans men were found in the men’s prison estate. The March 2024 UK data showed “five or fewer” trans men with sexual offense convictions (suppressed for anonymity). This near-total absence is consistent with female-pattern offending baselines: people assigned female at birth offend at female rates regardless of gender identity.

Using the 2019 UK figure (76 trans women incarcerated for sex offenses), the per-capita rate changes dramatically by denominator:

The per-capita rate ranges from below the cis male rate (at higher denominators) to 3x above it (at the ONS figure) — a 40x uncertainty range. This is not a minor methodological caveat. The ONS figure of ~48,000 trans women is not a valid denominator for this calculation. ONS formally downgraded this data in September 2024, citing widespread question misinterpretation — particularly among non-English speakers — and explicitly stated the figure should not be used for per-capita analyses of this kind. Anyone citing “158 per 100,000” derived from the ONS Census figure is either unaware of ONS’s own disclaimer, or aware of it and choosing to ignore it. The statistic is not defensible on its own terms. Using denominators that ONS does not flag as unreliable, the per-capita rate for trans women falls at or below the cisgender male rate.

A prison population at any given moment is not a crime rate. It is an accumulation weighted by sentence length. In England and Wales, sexual offenses carry average custodial sentences of 66.2 months versus 20.3 months for all offenses — an accumulation factor of 3.3x. If a group has even a small number of sex offenders serving long sentences, a snapshot will show a high proportion of that group as sex offenders — even if their actual offending rate is identical to the general population.

The UK MOJ has explicitly noted that “prisoners on longer sentences are more likely to be managed as a transgender prisoner,” creating a filter effect that concentrates the most serious convictions in the counted population.

The Correctional Service Canada’s file review (Research Report R-442, 2022) examined 99 gender-diverse offenders in federal custody. Of 33 with sexual offense histories, 94% had committed their offenses before ever identifying as transgender — while still living as their birth sex. 82% of those with sex offense histories were trans women; 0% were trans men.

This finding is arguably the single most important piece of evidence in the entire debate. It means a prison snapshot that counts someone as a “trans woman sex offender” is overwhelmingly capturing people who offended as males before they ever claimed a transgender identity. The offending cannot be attributed to trans identity — it predates it.

The study compared trans women to cisgender women, not cisgender men. Finding that trans women have higher crime rates than cis women (HR 6.6) means they have male-pattern crime rates — which is expected and does not show they offend more than cis men (HR 0.8, not significant). For the post-1989 cohort, no elevation was found at all. Dhejne herself stated: “We were certainly not saying that we found that trans women were a rape risk.”

Since UK policy was strengthened in 2019 (now requiring that trans women with male genitalia or sex offense convictions cannot be held in the general women’s estate without ministerial exemption), there have been no sexual assaults committed by transgender women in women’s prisons. As of February 2026, 7 prisoners were on E Wing at HMP Downview, the separate unit for trans women.

When Dade County, Florida passed an anti-gay discrimination ordinance, singer Anita Bryant formed “Save Our Children, Inc.” and claimed: “Homosexuals cannot biologically reproduce children; therefore, they must recruit our children.” The campaign gathered 64,000+ signatures, voters repealed the ordinance 69% to 31%, and Florida barred gay adoption for 20 years. Two weeks after the vote, Robert Hillsborough was stabbed 15 times in San Francisco while his attackers chanted “Faggot.” The 2025–2026 “groomer” rhetoric uses an identical structure.

Groth’s distinction is critical: pedophilia is attraction to prepubescent children. It is clinically distinct from adult sexual orientation. Most child molesters of boys do not identify as homosexual in their adult orientation — their attraction is to age, not gender.

Anti-LGBTQ+ bias accounts for more than one in five FBI-reported hate crimes, a proportion that has grown steadily. Anti-transgender incidents specifically roughly doubled from ~184 offenses in 2018 to 521 in 2024.

HRC has tracked 399 transgender or gender-nonconforming people killed in the U.S. since 2013. Of cumulative victims, 73.7% were Black transgender women and 84.4% were BIPOC. The 2015 U.S. Transgender Survey (n=27,715) found 47% of respondents had been sexually assaulted in their lifetime — compared to ~20–25% for the general female population (CDC NISVS). Trans people in prison are 10x as likely to be sexually assaulted by fellow inmates.