Predictive policing predicts police harassment, not crime

Originally published at: http://boingboing.net/2016/08/18/predictive-policing-predicts-p.html

2 Likes

the “Heat List” system

6 Likes

“Their decisions will ultimately be reviewed by impartial judges before punishment is delivered”. facepalm As if being arrested, incarcerated waiting for trial, and then forced into a plea bargain irrespective of factual guilt is cost-free. Geez.

13 Likes

It astounds me that a study was required to produce these results.

Every time I hear LE folks talking about their algorithm-based policing schemes it leaves me speechless. Of course the crime statistics are going to follow where the cops go. That does nothing to indicate that the cops were in the right place.

12 Likes

Found and read through this the other day - at one point it discusses policing methods as part of the statistical outliers, sort of an interesting read for those who are into stats. NYC has a big statistical impact on the data set, and the author suggests that it is policing methods that set it apart.

I would think most LEO would use some sort of targeted enforcement, you want to be where it’s happening, right? But in Ann Arbor downtown business owners have been asking for increased foot patrols in certain areas to help with nuisance crimes like graffiti, littering, public intoxication/public urination (like right on the sidewalks) and aggressive panhandling, but for a variety of reasons (mostly to do with understaffing and political pressures to deflate crime stats) we still don’t have foot patrols where they are needed most. So there are many ways to slice it, is my point. We have the crime, if you owned a business and were down here every day you would see it; but we don’t have the reporting, because we don’t have the enforcement… Does this mean if we had no police at all anywhere, statisticians would say that crime was reduced to zero - because there have been no arrests reported?

4 Likes

I don’t know if an understaffed police force is best used to police littering or public urination. I would even argue that the issues you list are hardly crimes and more vexing social and cultural issues that have entirely different cures.

3 Likes

I’m not able to review this fully right now but this sounds like a self-fulfilling (and self-justifying) prophecy generator. The police will find reasons to charge people (it’s what they do on these kind of patrols; see NYC’s stop and frisk).

Then the charges get added to the system and, voila, the loop is closed: you generated the “crime” your model predicted.

6 Likes

Some while back I was reading about how there were predictive policing models that would send police patrols to high-crime areas and did have a measurable impact reducing crime rates, but that US police simply didn’t use them. Apparently US police preferred to concentrate patrols where they believed crime should be, not where it actually was.

3 Likes

Tonry (1987)also argues that the ethical issues with using prediction in a policing context are less controversial than in other criminal justice settings because they are necessary for the cost-effective distribution of scarce resources, and their decisions will ultimately be reviewed by impartial judges before punishment is delivered.

That sounds so nice, so reassuring. Here I am, never arrested in my life, secure in my white privilege – and it’s freaking obvious even to me that just being arrested – never mind charged, much less convicted – is going to be hideously expensive. And I’m damned well off.

To someone making the median household wage, less than $500 from financial ruin [1], being arrested – just arrested – has an excellent chance of costing one’s job, one’s home, and the rest of the cascading failures that I’ve never needed to worry about. And that assumes that the charges are never filed, or dropped before the cost of legal defense forces either bankruptcy or (much more likely) a “public defender” who advises me to plead guilty to a charge that’s less than the amped-up one that the prosecutor chose to force a plea bargain.

Whoever wrote that quote above was either bullshitting or totally divorced from reality.

[1] Half of the households in the USA

5 Likes

I must be reading it wrong. I think it says that the people that were on the list still ended up committing crimes, despite the extra attention from the police. So they were targeting the right people, just not able to intervene and stop them from committing crime. I’m missing the part where otherwise blameless people were harassed.

Bad cops are a curse that must be eradicated. Same with violent gangsters. But this particular program, while it might be a poor allocation of resources, doesn’t seem particularly malevolent.

1 Like

What I read was that they ended up going after the victims, but had zero improvement in preventing the crime or rounding up the criminals.

1 Like

Oh, that’s quite different from what I got. Thanks, I’ll give it a re-read.

“Make something up.”

Accusations of racism would follow such a policy as (sorry!) night follows day.

if there is such a thing as gateway crimes these behaviors are it. Just last night I was walking by a crowd of grizzly street people who were discussing which windows to smash. One of them was telling another “all these places have insurance man it will be fixed by the morning don’t bother”

Two of my customers buildings have had broken windows or doors in the last week. Property crimes are not victimless crimes, in the high rent districts promoted by the DDA and City Council zoning schemes (designed to drive up property values and therefore increase tax $ capture) the small business owner can be put out by these continued maintenance costs, by drops in business because a bunch of greasy street-campers set up shop in front of their store every day harassing passers-by, doing drugs, letting their dogs shit in the alleys and sidewalks. It’s a real problem that our city “leadership” ignores. We are one of the most policed cities in the state, with the University of Michigan having their own police force, plus we have county state and federal presence. But anything that happens in the DDA downtown is ignored for statistical appearances… It’s bad PR for their narrative of safe streets full of happy consumers. Even the police I speak with when filing damage reports say they want to be on the street, engaging with the community, but are not assigned to it… There is some manipulation happening at the top of the heap.

This topic was automatically closed after 5 days. New replies are no longer allowed.