Thread by @AndrewMCrespo

Thoughtful article about 911 reform from @voxdotcom. But it makes an empirical claim that made me scratch my head: Basically, it suggests that 911 calls account for half of all incidents where the police approach civilians.

Looking under the hood, I don't think that's right.

https://twitter.com/voxdotcom/status/1292790320900509698

Here's the claim in the article. The implication is that half of "officer-civilian interactions" that might conceivably end in "aggression" or "violence" are "the result of citizen-requested police services, usually through an emergency call number."

But that's not what the cited source says. It takes as its denominator people who had *any* contact with law enforcement... including *calling* 911.

People who call the police account for half of all law enforcement interactions.

People who are "approached" make up the other.

The cited report only surveyed civilians, so it never asked police departments "how many of your officer-initiated interactions with civilians are based on 911 calls?" And when it asked civilians if the police gave a reason for a stop, "investigating 911 call" wasn't an option.

So, the report doesn't actually tell us the extent to which 911 calls drive officer-initiated interactions with civilians.

But in a different study, researchers observed 442 officer shifts (spanning 3,536 hours) and coded every minute. "Patrol" by far took the lion's share (26%). In fact, "crime calls" didn't even merit their own category, but instead got lumped in with all "crime incidents" (10%).

All of which is to say that while 911 reform is surely important, it might not be the place to "start" if we "want to fix policing." vox.com/2020/8/10/2134…