Thread by @DrPhilGoff

1. So, while we’re in the midst of a public safety data crisis, we might lose data from 25% of law enforcement agencies. Here’s how and why that matters. A thread.

FBI crime data could go away for one in four police agencies. newsy.com/stories/fbi-cr…

2. Data from police departments on reported crime are terrible. Just awful. The quality is bad. Standardization is trash. I often argue they are the worst governmental data in the world. The system for reporting them to the FBI is known as the Uniform Crime Report: UCR.

3. The UCR format sucks. It has no category for some important crimes (e.g., identity theft) and does not give the kind of granular information necessary for some basic analyses.

4. So, a bunch of criminologists came up with a new system that tracks closer to what happens at the incident level. This is called the National Incident Based Reporting System: NIBRS (we say NIGH-burs).

5. To be clear, NIBRS also sucks. But it is meaningfully better than UCR. Still, police departments were reluctant to adopt it. Why? Two main reasons.

6. First, institutions fear change.

7. Second, they also feared that crime numbers would go up...because they were capturing more crimes. Because we (try to) hold police accountable (sometimes) for the crime rate, there was just no percentage in making the switch. So many resisted.

8. So, this year, the FBI just decided it was fed up. So they said, “That’s it. You don’t give us NIBRS, we will stop taking the data.” How many won’t be able to make the switch by next year? About one in four.

9. But this may not be the worst thing in the world. DOJ stats folks (Bureau of Justice Statistics) plan to estimate rates of crime...which might not be much worse than what they’ve been getting. And those non-reporting departments? They are likely among the smaller ones.

10. Keep in mind that roughly 75% of the roughly 18,000 law enforcement agencies across the country are 25 officers or fewer. So we may not be missing out on THAT much data.

11. Still, that we, as a nation, can’t get it together to count reports of bad things happening (disproportionately in Black and Brown communities)...the neglect is THICK.

12. And, like the census, scientists rely on these data in order to understand basic questions about crime, safety, and policing. So, if you want to make data-driven arguments about shrinking the size of police, you probably need these data.

13. That our movement forward (NIBRS > UCR) is accompanied by such a foreseeable move backward (25% fewer departments report) communicates that we are STILL not taking data on public safety seriously.

14. The result is likely to be that slogans and bromides will dominate in some places where data are absent. And...it just doesn’t have to BE like that!

15. Last, and maybe most importantly, is this: Part of the reason we have a hard time imagining different forms of public safety is that we lack the data to describe ALL the ways our current system is broken.

16. That means, if we don’t invest in the data infrastructure, we risk having the same pointless arguments for years to come. fin.