There is good news hiding inside the statistics on violence against transgender Americans: fewer of us are showing up on annual lists of homicide victims than during the worst years of the past decade.
Good.
Seriously. We should be thrilled whenever fewer transgender people are murdered.
But a September 22 Boston Globe opinion column takes that encouraging statistic considerably further. Writer David Mastio argues that violence against transgender Americans is declining and that the numbers undermine what he describes as the prevailing narrative about an “epidemic,” “onslaught,” or “genocide” of anti-trans violence.
There’s just one problem. The numbers don’t actually answer all of those questions. They tell us something important. They don’t tell us everything.
A Count Isn’t Automatically a Rate
Mastio points to Advocates for Trans Equality data showing documented violent transgender deaths falling from 53 in 2023 to 27 during its 2025 reporting period. Human Rights Campaign tracking similarly shows fewer deaths than during the 2021 peak.
Those numbers shouldn’t be dismissed simply because they’re inconvenient. Documented transgender homicides have declined.
But these are counts of identified victims, not a comprehensive census of every transgender homicide in America.
A4TE explicitly describes its own 2025 report as “almost certainly incomplete,” citing underreporting and the continued misidentification of transgender victims by families, law enforcement, and news organizations.
There is also a larger statistical problem. America as a whole became dramatically less homicidal during the same period.
According to the FBI, murder and non-negligent manslaughter dropped an estimated 18.1% from 2024 to 2025. Robbery fell 18.5%, aggravated assault fell 7.2%, and reported hate-crime incidents overall declined 7%.
So if transgender homicides declined while homicide generally was plummeting, the decline alone cannot tell us whether something specifically changed about the danger faced by transgender people.
Correlation remains stubbornly unwilling to become causation just because it makes a good headline.
Zoom Out and the Picture Changes
The Globe column mentions the decline in FBI hate-crime reporting from 2024 to 2025. That’s true. Now zoom out.
A 2026 Williams Institute review found the FBI recorded 184 anti-transgender and anti-gender-nonconforming hate crimes in 2018. By 2024, there were 527.
Gender-identity bias also grew as a proportion of reported hate crimes, from 2.2% in 2018 to 3.9% in 2024.
That doesn’t prove anti-trans hatred alone caused the increase. Changes in reporting, identification, and law enforcement participation can influence hate-crime statistics.
But that’s precisely the point. These statistics require context.
One year of declining hate-crime reports doesn’t prove anti-trans hostility is disappearing any more than one year of increases would prove every transgender person is suddenly in mortal danger.
Murder Isn’t the Only Violence
There’s another problem with using homicide counts as a stand-in for transgender safety.
Most violence doesn’t end in murder.
National Crime Victimization Survey data provide a much broader picture because the survey captures violent victimization rather than only homicide.
A Williams Institute analysis of 2022-2023 data estimated transgender people experienced violent victimization at 93.7 incidents per 1,000 people, compared with 21.1 per 1,000 non-LGBT people.
Earlier peer-reviewed analysis of the 2017-2018 NCVS produced a similar disparity: 86.2 violent victimizations per 1,000 transgender people versus 21.7 among cisgender people.
That doesn’t mean every one of those attacks happened because someone was transgender. It does mean transgender people were experiencing violent victimization at substantially higher rates.
That’s rather important information in any discussion about whether transgender people experience disproportionate violence.
What About the Claim That Trans People Aren’t Murdered at Higher Rates?
Here’s where intellectual honesty matters.
There actually is peer-reviewed evidence supporting part of that argument.
A 2017 study examining transgender homicides from 2010 through 2014 concluded that the overall transgender homicide rate was likely lower than the overall cisgender homicide rate.
We shouldn’t pretend that study doesn’t exist.
But read further.
The study also examined risk by age, race, and gender rather than treating every transgender person as belonging to one identical demographic group. It found particularly elevated homicide risks among some young Black and Latina transfeminine populations compared with corresponding cisgender female populations.
That matters because transgender homicide victims aren’t randomly distributed across our community. A4TE’s 2025 report found that 63% of its documented violent deaths were Black transgender women.
Saying “transgender people” have one homicide rate can therefore obscure enormous differences within the population.
Now Let’s Talk About Intimate-Partner Violence
This is where the argument becomes particularly interesting.
One response to claims about anti-trans violence is that many transgender homicide victims aren’t killed by strangers screaming slurs. They’re killed by partners, acquaintances, friends, or family members.
That’s true.
HRC has previously found intimate-partner violence to be an important component of transgender homicide. Its decade-long review through 2022 found that 19% of victims with a known killer were killed by an intimate partner and another 10% by friends or family members.
But there’s a logical leap hiding here: Killed by a partner does not necessarily mean transgender stigma had nothing to do with the killing.
“Who killed her?” and “Why did he kill her?” are different questions. And researchers have specifically studied this distinction.
The Boyfriend Nobody Was Supposed to Know About
Research involving transgender women of color has documented something many trans women will recognize immediately: men who privately date or have sex with transgender women while desperately trying to ensure nobody else finds out.
Researchers found cisgender male partners sometimes concealed their relationships because they feared being exposed as attracted to transgender women and anticipated rejection or ridicule from friends, relatives, or the broader community.
More troublingly, participants connected that relationship concealment and stigma to physical violence, including homicides of transgender women by intimate partners.
Separate research examining severe intimate-partner violence among transgender and gender-diverse people reached a similar conclusion. Researchers identified gender disclosure and relationship stigma as potential risk factors and specifically discussed partners hiding relationships because of the transgender person’s identity.
That doesn’t mean every transgender woman murdered by a boyfriend was killed because she was transgender. It means you cannot automatically classify intimate-partner violence as evidence against anti-transgender stigma playing a role.
Sometimes the categories overlap.
Dime Doe Shows Exactly Why
We don’t even have to keep this hypothetical.
Dime Doe was a Black transgender woman murdered in South Carolina in 2019 by Daqua Ritter, a man with whom she had a sexual relationship.
Ritter tried to keep their relationship secret. Evidence showed he became angry when other people discussed it. After rumors spread through the community, Ritter lured Doe to a remote location and shot her three times in the head.
Federal prosecutors proved beyond a reasonable doubt that Ritter murdered Doe because of her gender identity.
He was convicted of a federal hate crime and sentenced to life in prison. Think about what happens if we insist upon simplistic categories.
Was Dime Doe killed by someone with whom she had an intimate relationship? Yes.
Was Dime Doe the victim of a gender-identity hate crime? Also yes.
The two facts happily coexist, inconvenient though that may be for a tidy spreadsheet.
“Not Classified as Hate” Doesn’t Mean “Proven Not to Be Hate”
There is another important distinction.
The FBI defines a hate crime as an offense motivated wholly or partly by bias. But investigators need sufficient evidence to conclude that bias motivated the crime before reporting it that way. The FBI itself acknowledges that determining motivation can be difficult because motive is subjective.
That creates three possibilities, not two: A crime can have evidence establishing bias. A crime can have evidence establishing another motive without meaningful evidence of bias. Or investigators simply may not know enough to determine what role bias played.
That middle ground matters enormously. Absence of evidence proving a hate motive is not evidence proving the absence of one.
The Bottom Line
We can say documented transgender homicides have declined from earlier peaks. We can say reported hate crimes overall declined in 2025. We can say America’s overall homicide rate also plunged.
Those are all welcome developments.
We can simultaneously say that broader victimization data continue to show transgender Americans experiencing substantially higher rates of violence than comparison populations, that longer-term FBI data show reported anti-trans hate crimes rising considerably between 2018 and 2024, and that Black transgender women remain dramatically overrepresented among documented transgender homicide victims.
And we can acknowledge that intimate-partner violence accounts for some transgender homicides without pretending intimate relationships exist in a magical bubble untouched by anti-trans stigma.
The mistake isn’t reporting that transgender homicide counts have fallen.
They have. That’s genuinely good news.
The mistake is turning that fact into a conclusion it cannot carry: that declining homicide counts settle the much broader debate about anti-transgender violence, discrimination, and safety.
Numbers don’t become more accurate when they confirm what we want to believe. Neither do they become less accurate when they don’t.
Our community deserves better than fearmongering. But we also deserve better than someone finding the most reassuring number on the spreadsheet and declaring the problem solved.
Sometimes good news is simply good news. It doesn’t have to become something the data never said.
