Distributed Mind

Gladwell on Power-Law Problems

Jordon Cooper points to a fascinating and useful article ("Million-dollar Murray") by Malcom Gladwell in The New Yorker about power-law social problems, such as homelessness and car emissions. A power-law problem is one where contributions to the problem follow a power-law distribution (instead of say, a normal distribution, which, as Gladwell points out, seems to be what we usually expect). So, for example, Gladwell points to the finding that a small fraction of Los Angeles police officers were responsible for much of the problems with police abuse. In the case of homelessness, there is a similar problem, where a handful of chronically homeless people cost an inordinate amount to the system as a whole, and with auto emissions, a small fraction of cars are responsible for much or most of automobile pollution.

This article points to a lot of things I've realized in the last few years but don't get talked about very much and that I hadn't really thought of how to articulate. One of the things that comes out is that it is sometimes worth it to fix some problems, even if people don't "deserve" to have the problem fixed - such as some of the worst chronic homeless, in whose case it would be cheaper to just give them an apartment and such than to have to pay their medical bills everytime they end up in the hospital (a solution I thought of a long time ago in reference to specific homeless persons, but I never thought anyone would buy it at a political level - although apparently some people have already tried it!). The problem of course, as Gladwell points out, is that this violates our sense of fairness, along with some practical problems (one has an incentive, say, to be worse, not better). Gladwell says, "Power-law problems leave us with an unpleasant choice. We can be true to our principles or we can fix the problem. We cannot do both." Gladwell is more optimistic, as we might be, when writing about the problem of automobile emissions, where he can point to a practical and efficient approach to helping solve the problem. This suggests that not all "power-law solutions," as Gladwell calls them, are really as similar as he would lead us to believe. But, some, such as the proposed solution to the homelessness problem are classic examples of unfairness (in a positive sense - giving people what they don't deserve), but unfairness that we might want engage in for the benefit of everyone. It occurs to me that by lumping all of this together Gladwell may have mixed together two separate issues: That it might to help to check if problems fit power-law distributions, and that the solutions to some social issues are "unfair." But, whether these are separate issues or the same thing, they are both useful observations that we need to think about when approaching social problems.

posted at 02:23:59 on 02/09/06 by ben - Category: Politics


No comments yet

Add Comments

Want to comment? You'll need to create an account first.