Thursday, July 05, 2007

Game Theory, Cooperation, and Crime

A couple of interesting reports today on research studying the evolution of cooperation and how the "free riders" who live off others' cooperation get brought into line. (Need a better description of what we're trying to do with corr sent?) The traditional analyses show how cooperation can thrive even when there are strong incentives to "defect" against those with whom we interact because, while short term cheating may win out, long term those cheaters develop reputations and begin to lose. That's sort of the logic behind each of these studies.

If you've got a bunch of people cooperating and they can't keep you from benefitting from their production, then why should you cooperate, too? Enjoy learning the tin whistle and hop in when they're done. Of course, when everyone sees how they can benefit without working, no one works and the system fails. Pursuit of individual self-interest gains you right now but bites you tomorrow (the failure of economists to never really grasp well the full difficulties and paradoxes of self-interest still boggles my mind, but let's not upset their little crystal worlds). So how do you get things to succeed?

One study says don't just have cooperators and free riders in your group. If you can get enforcers in there before the free riders control things, then you can have people who don't cooperate AND don't get the good stuff. This lends support, of course, to a crim just system, although it only works when the cooperation choice is voluntary. Make it compulsory and the incentives and outcomes change. (h/t Scientific American) The other study emphasizes that having durable goods that have to be maintained pushes the free rider dilemma into future generations as well, meaning instability can be forever.

Okay, yes, this is primarily a description of how and why laws and law enforcers get rolling in any starting community, although our modern forms are really just, well, modern. Things like commandments engraved on stones worked in ancient times. But what's really most interesting to me in this is the way that both studies point to how a system can get into a cycle of dominance first by the cooperators, then by the defectors, and so on. This indicates again what we talk about here that we are really dealing with chaos, complexity, and nonlinearity, "predator-prey" and tipping point relationships when we try to handle crime and its waves. That again says that our traditional ways of looking at the ability of policy to "control" crime and at criminological study still need serious rethought. Anyway, a couple of interesting articles that someone a lot smarter than I am may turn into part of the new and better paradigm for crim just study in the future.

No comments: