Lying with statistics is as American as PopTarts and energy drinks. The once-prestigious Federal Bureau of Investigation (FBI) used to gather crime statistics through something called the Uniform Crime Reporting Program. This gold standard of crime data was used for over a century and it collected data from every law enforcement agency that the FBI kept track of. It was so good, in fact, that the FBI decided to replace it with the new and completely unproven National Incident-Based Reporting System.
The old Uniform Crime Reporting Program gave what statisticians call “topline data.” It stated the total number of arsons, burglaries, car thefts, murders, and simple numbers. The newer National Incident-Based Reporting System provides details on each and every crime, so that you learn details about the racial and ethnic variables among victims and perpetrators of various arsons, burglaries, car thefts, murders, and so on. The new system has been described as “granular.”
The new system rolled out in 2021 with a learning curve as steep as Space Mountain. Feeding different statistics to the FBI on things defined different ways and including much more detail using an entirely new method turned out to be difficult and time-consuming. So difficult, in fact, that many law enforcement agencies just didn’t do it.
This would not have been a problem except there are no redundancies built into the transition period. We just went from System A to System B by tossing System A into the trashcan.
Last year, 6,000 law enforcement agencies did not turn in their crime statistics. Just to give you perspective, we have approximately 18,000 police agencies in this country and 6,000 have not reported anything. Even if you’re not good at statistics, you can see we have 1/3 of the data missing. That’s a lot. It’s what statisticians call a “data gap” and it pretty much invalidates whatever results you do get. (Think of it this way: if I have to report on all of murders committed in 18 cities and I only get data from 12 of those cities—I cannot say that my data from 12 cities reflects a true picture of what went on in those 18 cities.)
To be fair, when the FBI realized that transitioning from one system to the other cold turkey was problematic (great work, Sherlock), they temporarily revived the old system briefly, but skuttled it again. Some states like Illinois agreed to comply, in fact they’re all in with the new system—as soon as they can figure out how to actually use it.
When Democrats, including Karine Jean-Pierre speaking for Biden at the White House, announce that crime is down, she fails to tell us that we have no crime statistics at all from such places as the New York or Los Angeles Police Departments. Those happen to be the two largest police agencies in the country and both are experiencing surges in crime. But they are not in the federal data.
What she should have said was: “If we don’t count one-third of all law enforcement agencies and do not incorporate data from the crime capitals of New York, Los Angeles, or Chicago, crime isn’t that bad!”
The way most local police agencies work is that they sent their crime statistics to the state, which then processes the reports and shares the data with the FBI. However, not all states are compliant. Do they have sinister motives? Who knows? I just know there does not appear to be any7 penalty for not reporting crime statistics, and my suspicion is that the crime data for New York, Los Angeles, and Chicago, don’t look too good. So why not withhold it?
Using only two-thirds of available information has made it impossible for anyone with any sense of math and ethics to be able to say if crime has gone up or down or stayed the same. If you rely on the current data (with its gaping data gap), it looks like crime has gone down. That’s what we’re hearing, particularly on the campaign trail as Democrats want to look good. It’s sort of like Biden saying Bidenomics has increased prosperity. He can say it all he wants, he can even rig the numbers and lie about how much he decreased inflation, but people know their own experience. Same with crime.
And federal agencies who lean left are also quick to cite our decreasing crime rates based on these statistics. It’s a Biden miracle! Crime has gone down under Biden, because he’s such a competent president!
The FBI (the guys who led the raid on Mar-A-Lago) announced that violent crime decreased by 1.7% in 2022 versus 2021 nationwide. But they failed to say they were not playing with a full deck. They added murder went down 6% and aggravated assault was down 1%. You know that robbery is bad because even juking the numbers, national robbery statistics still show a 1% increase nationwide.
Another problem with crime statistics is what crimes get reported. When you collect statistics you have to define categories and it helps if you include all relevant categories. This does not happen with our crime statistics. For instance, the FBI does not report on drug-related crimes, which are actual crimes. So if you cross drug crimes off the list, crime is really way, way down.
And while this should go without saying, the FBI does say it: the FBI only reports on crimes that have been reported. So people who have been robbed or raped or assaulted and do not report it are not included in the data. Perhaps we should also say that imaginary crimes are not reported, either, which is why E. Jean Carroll’s alleged rape never got included in the official statistics.
There are clearer data from Pew Research (not a particularly conservative outfit) that says that both Democrats and Republicans are increasingly agreeing with the statement, “The President and Congress must make crime reduction a priority.” Republicans are more concerned about reducing crime than Democrats, but both groups are more concerned about crime this year than last year.
We also have an outfit called the Bureau of Justice Statistics (BJS) that tracks crime by surveying people and asking if they had been a victim of a crime in the past six months. They survey people ages 12 and over. The idea is that even if a crime was not reported, it might get captured in the survey method. The problem here are a few intellectual ones:
If a person was raped and did not to report the crime to the authorities (which happens), why would that person disclose the rape to the BJS? Why would you open up to a guy with a clipboard or an anonymous caller?
The other problem with the BJS is how does it report murder? It can’t exactly interview murder victims.
And what about crackpots and kids who might get surveyed and make up a bunch of baloney?
There is also something in statistics called “survey bias.” People who take surveys are different than people who do not. And that means that you’re getting crime stats from the survey-friendly people rather than a true cross-section of Americans, most of whom do not like to be bothered with a bunch of nosy strangers.
Democrats, intellectuals, and other liars claim that using the FBI’s new data and the BJS data together allow for a good 360-degree view of American crime. Interestingly, when you blend statistics in other sciences you blend similar types of data. You don’t take numerical data, ethnic data, and then survey data and throw them in the crockpot to cook up uniform crime statistics. You can certainly use multiple sources, but you can’t synthesize unlike things. And to make matters even more convoluted, the crimes that BJS asks people about are not the same ones the FBI tracks.
Surprise, surprise, “property crime” has gone down dramatically since 1990, the murder rate is down, and Pew Research says that Americans are so dumb, “Americans tend to believe crime is up, even when official data shows it is down.” What that means is that when the government agencies and politicians lie to you about crime statistics and you question them, it means you’re too stupid to understand. It’s a common Democrat defense and often used against so-called “conspiracy theories.” If you don’t believe what you’re told, well, you’re dumb.
It is hard to know why the FBI decided to opt out of an old established data collection system successfully used for over a century to abruptly transition to a newer and more convoluted one. All I can report is the timeline. The system switched over January 1, 2021 or roughly when Biden moved into the White House. And the switch did not allow for two redundant systems to run in parallel for a while till the transition was complete. It was binary: off with the old way, on with the new way. We have not had completely honest crime statistics (or at least crime statistics that could be fairly compared to crime statistics prior to 2021) since Biden took office.
Gee, that doesn’t seem suspicious.