Important Correction: On Violence From and Against Police

Hmm… your wording is still somewhat opaque to me, to be honest. But I think it’s kind of a moot point; this metric still doesn’t seem meaningful. Let me expound on that for a sec:

So we’re comparing the ratio of "civilians killed by cops" to "cops killed by civilians" in a year. That ratio is 242:40, or approximately 6:1. The purpose of any metric is to be useful for something. For example, it would be useful if this metric could be used to set a threshold for what ratio of civilian deaths to cop deaths is too much.

Well, imagine three different scenarios:
1. In the future, police officers are outfitted with cheaper armor. The number of police officers killed goes up, and the number of civilians killed stays the same. All other variables also stay the same.

Now the civilian death:cop death ratio is smaller, but this is not because anything got better. In fact, things got worse.

1. In the future, cops receive better training in defusing dangerous situations. As a result, fewer conflicts escalate to the point where a threatening civilian is (justifiably) shot. The number of police deaths, however, remains the same, because civilians who were planning on shooting an officer do not care about defusing the situation.

As a result, the civilian death:cop death ratio got smaller, and this time it is a good thing.

1. In the future, cops receive better marksmanship training. They only shoot at violent criminals who are a legitimate threat, but when they do shoot, they are better able to hit their target. As a result, they justifiably kill more threatening civilians. All other variables stay the same.

Now the civilian death:cop death ratio is larger than before, and this is arguably an improvement. Before, cops were shooting and missing; now they’re shooting and hitting.

Given the above hypotheticals, the civilian death:cop death ratio seems to be orthogonal to whether things are getting better or worse. The ratio can be up or down, and – depending on the reason for the change – this can represent something good or bad.

Also, maybe the increase in dangerous civilian deaths causes a decrease in innocent civilian deaths. The metric above doesn’t take into account what’s important.

Lastly, it’s the job of a cop to go where the dangerous people are. So… yeah, they’re going to end up confronting and possibly shooting more people per cop than a civilian shoots per civilian. But I’m not sure what that means or how I should feel about it.

So what’s the point of the metric? Even to describe it accurately seems to engender unnecessary confusion.