Sunday, 11 January 2015

Ideology is Back!


"Anyone who says to you: 'Believe me, I have no prejudices,' is either succeeding in deceiving himself or trying to deceive you." Joan Robinson.

The subject of ideology is becoming hot once again in the econo-blogosphere. The story so far:

Dec 8, 2014, Zubin Jelveh, Bruce Kogut, and Suresh Naidu (JKN henceforth) published a brief note at 538.com, summarizing their working paper on ideology:
"So we set out to test the idea of nonpartisan economics on a large scale. In a recent paper, we researched whether economists' political leanings were associated with their professional work. The answer: yes."
They presented a chart for one among several tests, purportedly showing economists' susceptibility to ideology.

Mother Jones' Kevin Drum replied the same day:
"[I]t sure looks to me as if … economists as a whole are remarkably unbiased. I mean, look at that chart again. I would have expected a much steeper line. Instead, what we see is just the barest possibility that ideology has a very slight effect on economists' findings.
"If these results are actually true, then congratulations economists! You guys are pretty damn evenhanded. The most committed Austrians and the most extreme socialists are apparently producing numerical results that are only slightly different. If there's another field this side of nuclear physics that does better, I'd be surprised."
The next day Matthew Martin found that the regression line included in the original chart in the 538.com was too sensitive to two outliers and agreed with other bloggers that "an effect of this [small] size cannot explain a noticeable fraction of the total variation in economists' result" (link).

In a comment to Martin's post, a few days later, Z. Jelveh explained that:
"Just wanted to point out that do [due to?] a communication error the incorrect chart was posted on 538. The correct version is now up and is from Figure 9 in the paper."
The JKN paper is fairly long (80 pages: 38 of text, plus 42 of references, tables, figures, and appendix); it's also very complex, statistically and IT-wise (Jelveh, a Ph.D. candidate, seems to be the computer guy in the JKN team). Frankly, I would be surprised these early and lightning-fast replies reflected an in-depth appraisal of the paper and probably refer only to the 358.com summary.

At any event, Drum's article gives away why his reaction was so quick: the paper casts doubt on the impartiality of economists! That's a serious no-no.

I am no competent to emit any pronouncement on the statistics and computer stuff involved. However, I did find something that makes me somewhat skeptical of the paper, without denying the authors' innovative efforts.

The paper's Introduction, as motivation, cites two news stories involving two prominent economists nominated for the public office. One was "blocked by the Senate, with one senator remarking that he was 'an old-fashioned, big government Keynesian'"; the other withdrew his candidacy "after opposition from liberal groups over prior support for deregulation and relationships to the financial sector".

This was supposed to demonstrate that economists are broadly seen as susceptible to ideological influences (I wholeheartedly agree with them on that).

But their examples admit another reading: who are the ideologues in those examples, the economists or the politicians who opposed them? Maybe both? If this alternative reading seems unclear to the authors -- as it is apparent -- then one is unfortunately left with doubts about their results: how can one quantify something as tricky as ideology, when one is not clear what it is?

This suspicion only gets worse, as the authors apparently conflate "American political party membership" with "ideology" (revealed even in their choice of subindexes: d for "Left" and r for "Right"). After all, considering fiscal policy, a Republican President, like Dwight Eisenhower, would probably be classified as "Left" when compared to a Democratic President, like Bill Clinton (in other matters, Clinton would probably be to Eisenhower's "Left"). Both would be considered "Right" compared to senator Bernie Sanders, and "Left" compared to the KKK or to the senator Joseph McCarthy. Are economists immune to this?

In connection to Drum's remark above: are there any "committed Austrians and the most extreme socialists" in JKN's database? It's hard to tell, based on the list of top ngrams JKN offered.

----------

Robert Vienneau identified another strand of debate, unconnected to the JKN paper and featuring Paul Krugman, Simon Wren-Lewis and others. As usual, the whole thing revolves around the Keynesian family feud (although the illegitimate side of the family uses the "heterodox" to diplomatically refer to the legitimate children, inviting thereby the other non-Keynesian heterodox to join the fray against the "orthodox").

Vienneau, however, missed this post by prominent Austrian economist Peter Boetke, complaining the other mainstream economics is too Left (SRSLY???). Anyway, Boetke's appraisal is another no-no for the other mainstream economists: aren't they supposed to be "inhumanly impartial"? At least Austrians are clearly not "impartial": they side with the other side. Credit for the honesty, I guess.

Matías Vernengo shirtfronts da Man, da Krug-Man, that is.

Chris Dillow (maybe you could say an Analytical Marxist; I call them neoclassical Marxists, pdf) intervenes "to take a little heat out of the orthodox vs heterodox debate". Good luck with that.

No comments:

Post a Comment