the Evil Line

Several things about this post by Professor Bainbridge are interesting, but what struck me are these statistics on the political leanings of American faculty which (he says) are from the LA Times:

The Democratic advantage by department in 1999 English: 35 to 1 History: 17.5 to 1 Biology: 4 to 1 Engineering: 3 to 1 Computer science: 2 to 1 Chemistry: 1.5 to 1 But in agriculture, Republicans held a 1.3-1 edge.

What makes scientists more conservative than their colleagues?