Tuesday, 17 May 2011

Questions about economists' favourite economists

Davis, Figgins, Hedengren and Klein have put together an interesting survey of American economics professors - asking about their favourite economists, alive and dead, and about the journals and blogs they read.

I downloaded the data behind the paper in the vain hope that this blog might be among the long tail of responses not reported in the main paper (it wasn't). But the data does provide lots to think about.

An intriguing point (from a survey design point of view, at least) is raised by one of the questions. The survey asked people to choose their most respected/admired economists from two groups: over 60 years old and under 60. I don't know about you, but I'm not sure I'd know the age of many of the economists I admire.

I speculate that the respondents might have subconsciously chosen economists who are much older, or much younger, than 60 years of age, in order to answer these questions with greater certainty. Even if aware of this potential bias, I might find those who are clearly very young or old more easily leaping to mind - the availability bias - when I attempt to answer. Are there perhaps a bunch of under-represented economists who are close to 60 years old? To take a random example, I perceive Richard Thaler to be around that age (in fact he's 65), and he is mentioned by just two people despite his prominence in recent years. Maybe that does reflect his true influence, but still...

The paper does address the question of misclassification of age - Paul Romer, for example, is under 60 but was named by three times as many people in the over-60 category - so was moved into that group for data analysis purposes. But it doesn't mention this potential bias towards extreme ages. It's perhaps an unusual kind of bias - people are more commonly biased away from extremes. But in this case, I suppose age 60 is the extremum of each question, so it does fit the general pattern of extremeness aversion.

The results in any case are very interesting and I'm particularly intrigued to see what happens when you strip out the economists who are selected by a politically biased sub-sample of the respondents (Becker or Hayek on the right, Krugman or Galbraith on the left). While this doesn't imply their contributions are any less important, the answers are likely to be partly influenced by political agreement.

This leaves a smaller number of highly-rated economists who display no strong bias either based on party affiliation or on the authors' measure of "liberalism". The top-ranked "neutrals" then are:
  • 20th century economists: Tobin and Schumpeter.
  • Over-60s: Heckman, Akerlof, Nash and Paul Romer.
  • Under-60s: Acemoglu, Bernanke, Gruber and perhaps Rabin.
I'm surprised Rabin doesn't show more of a leftward bias but with only four mentions, it's hard to draw any firm conclusions.

The blogs are noticeably politically polarised - every blog in the top 15 has a party ratio that is either 1.2 or lower (strongly to the right) or 3.6 or higher (strongly left). Only James Hamilton has a party ratio in between these two extremes - and even that may be because he shares a blog with Menzie Chinn. Do people only follow blogs which reinforce their political position? Or are these simply the blogs which have become well-known? Anyone even a little familiar with the economics blogosphere knows the implicit biases in nearly all of the popular blogs, and I guess economics professors are no more immune to this than any other reader.

Finally, the authors give three reasons for the low response rate to their survey (15% - which I don't think is that bad). Here are their suggestions:
(1) The survey was six pages and involved complex philosophical questions; (2) it asked about policy views and voting; (3) economists are growing tired of responding to surveys
In the spirit of the libertarian economics espoused by at least one of the authors - and with tongue slightly in cheek - might I propose a fourth reason for the low response rate: (4) incentives! What rational economist would spend 20 minutes answering a survey which can only help boost the career prospects of four of their competitors, with only the tiny benefit that they will slightly boost the already-high reputation of someone within whose school of thought they might work.

Lots of interesting questions raised, and I look forward to the planned future publication of more results from this survey.

No comments: