Stop and Frisk

An article quoted on the Schneier on Security weblog says this:

"Interesting data from New York. The number of people stopped and searched has gone up fivefold since 2002, but the number of arrests due to these stops has only doubled. (The number of "summonses" has also gone up fivefold.)

Good data for the "Is it worth it?" question."

Gosh, I don't know. The implication is, no, its not. But my gut feeling is that you can't just expect that the numbers will go up proportionately. In a way its like packet sniffing on a network; if you look at every single packet, its going to take much more effort than looking at a subset, but in return you're much more likely to find packets that you wanted to find. If you stop every single person, you're much more likely to find people that you wanted to find.

We wouldn't stand for every single person being stopped, though, so we accept a subset. I'd like to think that the subset is the ones that the people doing the stops think 'the most likely'. If you like that idea, you call it statistically compelling. If you don't, you call it profiling. Either way, if it results in more people being arrested who should be arrested, and the effort is sustainable, I'd think that its worth it.

I know that some people would say that agreeing to this is a slippery slope, to borrow the NRA's favorite incline. But so long as I think I'm getting more value from the extra policing than I'm losing by not having as much ability to walk down the street without being stopped, I'm going to support it.

And yes, I know that that is essentially the logic that Bush uses -- that greater enforcement equals greater security -- and I don't believe it when he says it. It comes down to whom you trust. Generally, I trust the cops. Generally, I don't trust Bush.