The Great Illinois Violent Crime Disappearance
Discovering the why behind a big drop in Illinois reported violent crime.
The FBI’s Crime Data Explorer can be difficult to use, but there can be good data in there if you know where to look. One of the features that I like is the Data Discovery Tool which allows you to see reported crime at an agency, state, or local level back to 1986.
Doing so for Illinois violent crime, however, is also instructive about the possible pitfalls of our crime data reporting system. Behold, the state of Illinois’ violent crime trend since 2010:
That’s a big drop in 2021 followed by a sizable increase in 2022 and 2023 albeit at substantially lower levels than before the drop. But us violent crime in Illinois really well below 2020 levels now?
Well, no. This mystery, unfortunately, has a relatively easy answer that can be summed up as: Chicago.
One of the challenges to working with the Crime Data Explorer is that it doesn’t tell you when data is missing. So Chicago only reporting 6 months of data in 2021 due to the NIBRS transition is not evident in the CDE’s Data Discovery Tool. CDE does provide an ability to look up monthly data and the lack of data between January and June 2021 in Chicago is evident there.
So the drop in 2021 is easily explainable as a NIBRS transition quirk. But what about the lower level in 2022 and beyond?
That appears to be a change in how Chicago was counting aggravated assaults.
If you look at Chicago’s crime in the Real-Time Crime Index you don’t see any obvious changes in how crime in Chicago is being counted. Here’s aggravated assaults in Chicago rolling over 12 months which makes up a huge share of Illinois violent crime. There's no huge drop, indeed aggravated assaults in Chicago have been up a bit post-COVID compared to pre-COVID levels.
Here, by comparison is the monthly crime counts of aggravated assault in Chicago from 2018 to 2023 in the CDE.
There is no data for the first 6 months of 2021 followed by a huge drop in aggravated assaults. The difference between the CDE and RTCI data is that the former comes from aggregated totals submitted by the Illinois State Police while the former stems from Chicago police’s open data portal.
I started writing this piece a few weeks ago and grabbed data from the Illinois State Police to show the differences in monthly aggravated assaults were between the FBI, Illinois State Police, and CPD open data. There’s no state police data prior to July 2021, but one can clearly see that aggravated assaults in Chicago remained relatively steady overall from 2018 through 2024 in the city's open data. The huge drop in the number of aggravated assaults in Chicago was a function of far fewer offenses being reported by the city to the state police and then to the FBI after NIBRS was implemented.
Here is where I would ordinarily put the concluding thoughts on the challenges of working with crime data and what it means for analyzing violent crime trends in Illinois. National crime estimates are just that and we should assume uncertainty when discussing changing national trends. In addition, I’d note that Chicago is not alone in this problem and point out how “human error” in Oakland rose the US violent crime rate by 1 percent in 2023.
All that is well and good, but a weird thing happened when I sat down last week to finish this piece: Chicago fixed the problem.
Whereas the number of aggravated assaults in Chicago was around 300 per month in the Illinois State Police data a few weeks ago, now it is well over 1,000 per month. So much so that the revised figures are back to mostly matching what’s being reported by CPD (sorry for all the colors on this graph!):
The 2021 figures reported by the ISP still aren’t great, but 2021 was a disastrous year for crime data reporting everywhere. The 2022 to present data looks spot on which is great to see.
So what comes next?
Chicago’s figures should be revised into the FBI’s next Uniform Crime Report release in the Fall though that process is opaque so it’s hard to say for certain. The underreporting in Chicago and overreporting in Oakland should more or less cancel out (though not quite), so I wouldn’t expect an enormous increase in violent crimes nationally. But, then again, one could get fixed or neither, or there could be another problem that hasn’t come to light yet.
Either way, this is a good example of the living nature of crime data. There are enough mistakes and enough uncertainty that it is never complete, even when it's complete. And, finally, it's a good lesson in looking for data reporting issues whenever you find a reported count that seems too good (or bad) to be true.
Recently we in Chicago were told that science says Gang Crime in Chicago is at a two decade low. If you talk to any cop, they will tell you this is ridiculous. When the police backed off, the gangs took over and now run the neighborhoods. The only reason gang crime appears low is 4 years ago Chicago passed a law saying it was unlawful discrimination for the police to retain lists/records of the 12,000 known criminal gang members. Now there is no way to easily tie a crime to a gang. So gang crime appears to be at a 2 decade low. Its the same logic with the FBI statistics which show a ridiculously low amount of violent crime in Chicago. Check Shot Spotter before they stopped it and rationalize the 3.5 million shots fired (each a felony in Chicago) versus the FBI 35,000 violent crimes reported annually. Police response is 6 times slower than it was. The streets are aware now you can beat up people, rob, mug, and steal less than $1000 daily with virtually no legal consequence. Sales people and public employees are told to not interfere with theft or crime as policy. Everybody knows there are minimal arrests and far fewer prosecutions now for violent crimes. That's a fact. Very few victims and fewer witnesses will ever step forward and face certain criminal retaliation (unless you plan on moving your family out of the neighborhood). It is unbelievably cruel and depressing to people who live the daily hell of fear for personal safety, who can't sleep at night, or send their kids to school, or take public transportation to work because of constant violent crime, that they are unnecessarily concerned because crime has dramatically decreased. If in doubt, survey local business and store owners, or the people who actually endure the daily hell of violent crime. Is comparative reported crime ever adjusted for any of these very real factors? Please help.
Nice work. I appreciate it.