You can read up on blog reactions to Overkill here.

So far, I’ve gotten really positive feedback from people who have read the paper. The only criticism I’ve seen thus far has come in the comments sections of a couple of blogs from people who haven’t read beyond the executive summary.

I do plan to respond to criticism if/when it emerges. But I will require that the critics actually have read the paper first.

With one quick exception. A few people have employed some back-of-the-envelope statistics, and concluded that 300 or so botched raids in a country of 300 million isn’t anything we ought to be worried about. Or that 300 mistakes drawn from 40,000 raids per year is a pretty good record.

There are a number of problems with that analysis, most of which are addressed in the paper. But I’ll quickly rehash them, just this once:

1) The map doesn’t claim to be, and is nowhere near comprehensive. I suspect I’ve only found a small fraction of the actual number of botched raids out there. There’s some evidence for this in the fact that when high-profile raids in which someone is killed happen in large cities, they’re inevitably followed by dozens of people who come forward and say “this happened to me, too.” Typically, these new cases were never reported. Which leads one to believe that there are many, many more like them. It took me just a few hours in Mississippi, for example, to find another victim of a mistaken no-knock raid by the Pear River Basin Narcotics Task Force. I suppose that could be purely coincidental. But my hunch is that if I had stayed down there another week, I’d have found several more.

One other thing — many of the raids on the map aren’t single incidents, but incidents in which police raided entire housing units, city blocks, or neighborhoods.

Finally, the map is biased toward more recent events. I’ve found 20 botched raids in 2006 alone, for example. That’s because I’ve been more closely monitoring the news for these incidents over the last few months. I’ve been watching Lexis and Factiva, getting Google News alerts, and you all of have been sending me incidents over email.

Raids before late 2005 were culled from research services and searches. Which means they’ll be limited to larger newspapers, and to raids that received enough coverage to trigger search engines to the keywords I’m looking for. As I’ve mentioned before, I continue to find new accounts of botched raids — even cases that ended in the deaths of innocent people — from the 1980s and 1990s. And since the paper was released on Monday, several people have already sent me accounts of raids I’ve missed.

2) The 40,000 raids per year is in itself a big problem. That a decreasing percentage of these raids end in death, injury, or the terrorizing of innocents doesn’t mean there isn’t a problem, or that we shouldn’t be concerned. When you start to send SWAT teams after people who aren’t violent, you shouldn’t be surprised when fewer SWAT raids end in violence. To use a reductio argument, we could begin sending SWAT teams after people who don’t pay parking tickets on time, too. And the percentage of total SWAT raids that end in violence would drop even further. That doesn’t mean that sending SWAT teams after parking ticket scufflaws is a good idea.

My argument is not only that too many SWAT raids violate innocent people and/or cause needless violence. It’s that sending SWAT teams to break down the doors of nonviolent offenders is in itself a bad idea, and not a police tactic free societies ought to tolerate.

3) If you’re going to argue that there isn’t a problem here because “only” 40 or so innocent people have been killed in paramilitary police raids since 1985, I wonder, what is a good number? At what point do you begin to get concerned? Seems like an odd sort of consequentialism. If we could keep the drug supply down (not that these raids are actually doing that — which is another matter, also addressed in the paper) by having the government randomly execute one innocent person every six months, and randomly terrorize an innocent family once every ten days or so, would that be okay?

That’s essentially what we’re doing.

This entry was posted on Wednesday, July 19th, 2006 at 9:08 am by Radley Balko and is filed under Police Militarization. You can follow any responses to this entry through the RSS 2.0 feed. Both comments and pings are currently closed.