Last week, Ars Technica broke the news that the National Highway Traffic Safety Administration (NHTSA) quietly took down their 2022 crash data, for the rumored purpose to scrub gender data from its Fatality Analysis Reporting System (FARS). At the time, the records had simply vanished from the public database, leaving only speculation about whether the missing information would return… and if it did, whether it would be altered.
Well, the data is back. Sure enough, the folks at Ars Technica called it, it was altered.
Updated: NHTSA FARS Data Reposted with Alterations
Buried in the latest FARS 2022 Release Notes is this little note:
“02/11/2025: Update the renaming of the variable to ‘Sex’ and the replacement of attribute 3 with attribute 8-Not Reported. There were 22 records in Person and 1 record in PBTYPE (all in US and none for PR) that were updated. Removed attribute 3 in the format library.”
So, what does that mean? Attribute 3 was “Other (e.g., “X”, Non-Binary, Not Specified, etc.)“, according to Ars Technica’s reporting. The variable “Gender” was also renamed to “Sex.”
That means that the 22 crash victims who were previously recorded as non-binary (or anything other than male or female) have had that erased from the dataset. Their gender is no longer recorded in the NHTSA database at all. The government saw those entries and went, “Nope, never happened.”
Did you see our related coverage: “40+ NHTSA Employees Fired Late Friday… Elon Doesn’t See Any Conflicts of Interest.”
The Million Dollar Question: Who’s Next?
It’s not hard to imagine how politically convenient it would be to remove other uncomfortable statistics in the NHTSA database… Or other government databases.
Just last year, NHTSA’s FARS data was used to determine that Tesla is the deadliest car brand on American roads.
Now, we’ve seen that federal crash data isn’t untouchable… it can be modified, erased, and rewritten.
Today, the Executive Order and subsequent edits target LGBTQ people. Tomorrow, could they conveniently “adjust” numbers to make politically favored automakers look safer? Could we see future administrations tweak datasets to downplay safety concerns?
This change isn’t just political noise… it has real-world consequences.
Why This Is a Problem (Other Than the Obvious)
All kind of problems here. Here’s the three that pop out to me first:
A) It makes our safety data worse. Gender can be relevant in crash outcomes, used for improving safety regulations, medical research, and vehicle design. If there are trends that affect these folks specifically, we’ll never know… because the data won’t reflect reality.
B) It sets a dangerous precedent. NHTSA is supposed to be about facts, statistics, and road safety… not political B.S. If the integrity of this data can be compromised for ideological reasons, what else is lined up for partisan enshittification?
C) It’s just plain wrong. The U.S. government is now officially recording inaccurate information on purpose. We knew who these crash victims were, and we deliberately erased that knowledge. That’s dumber than a bag of hammers and infuriating any way you slice it.
Closing Thoughts: Worse Data is Worse for Everybody
This is what happens when data becomes political. People don’t stop existing just because a spreadsheet says so. Now, 22 real people, folks who really suffered in car accidents, won’t be counted right.
For some much-needed automotive joy, we’ll leave you with this: Subaru and the Ladies Who Love Them: A History.
Right, because LGBTQ folks do exist on American highways, no matter what the government says, and they always will.
