Wednesday, 24 October 2012

Economic history pays

at least in the UC system... Keefe and Wang have a new paper on faculty salaries in economics departments of the UC system, and find that there is a 13-17% premium for economic history, after accounting for campus, seniority, gender, and publications. My guess is that economic history is a tougher field to get hired in (more of a "luxury" item); the scholars that do get in are way better than the marginal candidate in more mainstream fields. Eventually, quality shows and a good university system rewards them accordingly.

Amongst other interesting findings is how much higher pay at UCLA is than at Berkeley... the authors also have a hugely misspecified regression where they claim that they can back out the value of extra publications in different journals... after controlling for rank. This is a classic "controlling for an intermediate outcome" mistake, where an important outcome influenced by what you are interested in - publications - goes on the right hand side, and gives you completely the wrong results. For an interesting example, see the discussion by Andrew Gelman of "Engineers have more sons". 

No comments:

Post a Comment