Wednesday, January 4, 2012

Megan McArdle makes me chuckle

I know, I know - I need to be more specific with a title like that!

She weighs in on the MR/Krugman spat. I found this pretty funny though: "I will note that the commenters Paul Krugman sends to other sites harbor a very strange faith in his predictive powers"

"Sends to other sites"??? What a bizarre thing to say! I love how people act like those of us who are positively disposed towards Krugman and DeLong are somehow enthralled to them. It couldn't possibly be that we're convinced by them or just think they have good points to make. No! They "send us to other sites".

One thing you won't here me talking about much on here is "predictive powers" of Paul Krugman or anyone else. That should not be surprising, since I've long noted the problems associated with forecasting complex systems. What we need is explanatory power, not predictive power - and Paul Krugman does actually do pretty well in that regard. I don't know his "predictions" in any great detail - but if they're right I think that has more to do with (1.) luck, or (2.) stating the obvious than any great skill in economics.

7 comments:

  1. I agree with you that fans of Krugman and DeLong are often NOT blind sycophants, Mr. Kuehn.

    However, what do you make of scholars who have developed tools for better predictive power? The econophysicists have developed tools of analysis that enable them to make better predictions, though of course, not all of their forecasts have been on the spot. Here's a couple of examples in links below.

    http://www.sciencedaily.com/releases/2010/05/100511092406.htm

    http://www.technologyreview.com/blog/arxiv/25269/

    ReplyDelete
  2. Cascades, complexity, agent based work etc. - the stuff that's mentioned here - is definitely important for economics. But we don't really need physicists to tell us that. Economists have been involved in complex systems work since its inception.

    Think about it this way - the article talks about work on earthquakes and analogizes it to economics. I agree, and for much the same reasons that the article points out.

    How are we at predicting earthquakes far in advance?

    Pretty bad.

    Economists know most of what econophysicists have put out there. As far as I can tell, it's not a solution to the problems we're dealing with. I feel about econophysicists probably about what physicists would feel about economists making claims about physics: "these guys seem smart and sincere but they're working off of a caricature of the science, telling us a lot of things we already know, and telling us a lot of other things that simply seem wrong".

    I'm all for interdisciplinarity and a deeper appreciation of the complexity of the economy, but I'm not sure the econophysicists have an answer to economics' problems. I'm not even sure they understand the discipline's problems.

    ReplyDelete
  3. What about the stuff that isn't in the "most", as you put it?

    While I don't agree with all of what the econophysicists are doing, I do think that they have valid points. After all, they have cited Benoit B. Mandelbrot as one of their intellectual inspirations. Mandelbrot had important criticisms to make of economics, especially with regard to the statistical methods of econometrics and the use of the standard normal distribution (AKA, the Gaussian bell curve) in financial economics.

    I've heard that you cannot get published in an outlet for econophysics like Physica A: Statistical Mechanics and its Applications without the use of exploratory data analysis or goodness-of-fit tests.

    The econophysicists have been arguing that there is a need to expand the toolbox of the economist. I think that their tools (e.g., Ito calculus and random matrix theory) can easily be incorporated into the orthodoxy.

    I'm going to expand upon this in a future post that I shall be e-mailing to you, Daniel. Above all, I think that the culture and turf wars between econophysics and economics need to be put aside to expand tools of analysis.

    ReplyDelete
  4. I dunno if Krugman does it, but some blogs are famous for doing that very thing. Sending out their minions to annoy a critical commentator.

    ReplyDelete
  5. Krugman is so busy writing snark and analysis that I can't imagine where he would find the time to recruit an underground army of snarkers to attack other blogs for him.

    He would be the first to admit that having people doing your bidding is a good thing (he wrote on a related topic just the last couple of days) but only if what they're relating is a fair approximation of the true story. Attack dogs don't need to know black from white once they have a scent, though, and digital slobber doesn't cut the mustard in a fair debate (oh man, that was some fun mixing metaphors).

    ReplyDelete
  6. "After all, they have cited Benoit B. Mandelbrot as one of their intellectual inspirations."

    Well, there you have it! I now bow before their intellectual heft.

    ReplyDelete
  7. @Gene Callahan: Mandelbrot had points to make with regard to financial markets and econometrics...

    ReplyDelete

All anonymous comments will be deleted. Consistent pseudonyms are fine.