Tuesday, January 27, 2009

A Puzzle

About a month ago, Razib at gnxp posted this image from Strange Maps.

There is an uncannily strong correlation between which side of the border a person would have fallen on and which party he votes for.

Matthew Yglesias and Andrew Gelman argue that this is due to persistent wealth differentials-- places which were richer in 1918 are still richer today.

I can think of two alternate explanations:

1) Persistent cultural differences. There are strong network effects for many ideas, political ideas included, and this can lead to path dependency. In other words, popular ideas can stay popular for long periods. It seems likely that, being in two different countries, there would be cultural differences between the two areas. These cultural differences could have persisted and shown up in voting patterns.

2) Coincidence. Given enough maps, one or two will eventually match up reasonably closely.

Or it could be some combination of the three. (Maybe cultural differences drive wealth differences, for example.)

No comments: