Via John H:
Background here: https://pjmedia.com/matt-margolis/2026/05/22/the-left-tried-to-destroy-kash-patel-crime-data-just-vindicated-him-n4953137
I’ve also heard it suggested that Trump is faking the FBI data. But that would be easy for the cities to catch when FBI reports did not match what the cities reported.
Crime stats are pretty easily manipulated, especially once the prosecutor has had a chance to plea bargain Attempted Murder down to Aggravated Jaywalking 2 so he gets a “conviction”.
Violent crime victimization rates are harder to cover up. You’re got a body or an emergency room record, and unless you can substantiate that they did it themselves, the assumption is going to be somebody else did the injurin’, maimin’ or murderin’.
From Sir Robert Peel’s principle number nine: “To recognise always that the test of police efficiency is the absence of crime and disorder, and not the visible evidence of police action in dealing with them.”
How are crime stats different from “violent crime victimization rates?” If you had said convictions instead of “crime stats” then I understand. FBI UCR statistics, which I think is what is being talked about, reflect crimes reported to police, not convictions. They measure law‑enforcement activity, not judicial outcomes.
Crime stats are reported by the police departments to the FBI.
A very quick search for department manipulation of reported stats by NYPD lead me to this Reuters story:
https://www.reuters.com/article/world/us/nypd-report-confirms-manipulation-of-crime-stats-idUSBRE828187/
Apparently, if you try to report the manipulation, the NYPD will declare you an “emotionally disturbed person” and throw you in a mental institution against your will for six days.
So, that’s why I am disinclined to trust stats from the bureaucracy that may find those stats less than salubrious concerning their performance. I’d prefer to take stats accumulated from third parties that cannot gain from the manipulation, in this case, bodies and emergency medical calls. Now, if you’d like to riposte that ER stats can be skewed to cover their double- and triple-billing of Medicare… also proving to be true across the country, but they don’t get paid more if the injury is alleged to be due to crime or not.
I beg forgiveness for my terseness and less research than usual; on a phone.
To repost
Murder is the really easy one to talk about, because there’s a body, but “crime victimization” is similar, just less undeniable.
For a significant majority of murders, there’s a body that, forensically, says loudly, “I’ve been killed by another human being”.
When you have “crime victimization” states, you are looking at evidence like that, but for other crimes. It’s a bit like the difference between the household survey for employment vs the employer stats for employment – neither is perfect, of course, but their failures are different, so both can be useful.
There are arguments that there is some level of over-reporting in victimization stats. Certainly, rape has been over-reported in those stats for political reasons for the last few decades – multiple studies have been done to reach that conclusion, STRONGLY… but the rate of rape is still higher than the official “crime stats” would indicate. (Yes, BOTH of our best measures on that are notably and provably wrong in opposite directions. Phun times, eh?)
Back in the 1980s, colleges were caught falsifying or simply not reporting crimes. This led to the passage of the Clery Act in 1990, mandating reporting. Enforcement is spotty and it is still probably going on. There are technical problems, like many campuses not having police departments and depending on local government jurisdictions and the inclusion of public areas adjacent to campus which are not within the college’s jurisdiction even if they do have police departments.