27 June 2009

Racial Segregation

Easterly has a great new paper testing the implications of Schelling's tipping point model.

The idea is a simple but powerful one, if you took a completely random distribution of "black" and "white", it only takes very small colour preferences (any individual is happy to live in a mixed neighbourhood but doesn't want to be in the minority) to lead to stark segregation. And economists love it, I've read about it in 2 pop-economics books.

Trouble is, this elegant and simple theory apparently does a terrible job of explaining actual patterns of residence and migration when you test it with longitudinal US census data.

Easterly seems to take particular glee in debunking this as an example of "tipping points" and "non-linearities". Is this a secret dig at Sachs' poverty traps?

No comments:

Post a Comment