The Legacy of Sargent and Sims
Some thoughts, beyond my reaction of “eh.”
1. Since their work became widely circulated, one can argue plausibly that 99 percent of all macroeconomics papers published, dissertations completed, and tenure decisions awarded have been in the mold of Sargent, Sims, or both. Has anyone other than Samuelson had so much impact on what constitutes acceptable academic work in economics?
2. On the other hand, it would be hard for me to argue convincingly that the average citizen feels the impact of all of the resulting research. It has settled nothing important. For example, on monetary policy, John Taylor and Scott Sumner are poles apart, even though each claims fealty to Sargent.
3. As economic advisers, neither Greg Mankiw nor Christina Romer could do without the approach to policy evaluation that Sargent and Sims supposedly displaced.
4. What can be said for Sargent and Sims is that they provide guidance on producing models that embody rational expectations. Whether that is a worthwhile goal depends on the trade-offs relative to other considerations. Imposing rational expectations may or may not be a necessary condition for a creating a useful model. It is certainly not a sufficient condition.
5. Rational expectations in the Sargent-Sims tradition treats everyone as having the same model with which to form expectations. As Frydman and Goldberg point out in Imperfect Knowledge Economics, this assumes away the local knowledge and tacit knowledge that Hayek correctly identified as being very important in the economy.
6. Indeed, if Sargent and Sims represent a slap in the face to Keynes, they must be regarded as a knee to the groin of Hayek. Hayek coined the term “scientism” to describe the pretentious pose that economists strike when they equate mathematics with rigor. If scientism is a germ that infects economics, then Sargent and Sims were responsible for unleashing some of the most virulent strains.
7. Ultimately, your view of Sargent and Sims has to correlate with your view on the current state of macro. If you agree with Blanchard’s August 2008 assessment (“the state of macro is good”), then you should be grateful for Sargent and Sims. If you harbor serious doubts about the value of what passes for cutting-edge macro, then you should regret that the profession closed ranks so tightly around Sargent and Sims in the three decades that followed their original work.
Oct 13 2011 at 5:00pm
Even Einstein traded his wings for mathmatics.
The Hat of the Three-Toed Man-Baby
Oct 14 2011 at 10:07am
[Comment removed for supplying false email address and for rudeness. Email the firstname.lastname@example.org to request restoring your comment privileges. A valid email address is required to post comments on EconLog and EconTalk.–Econlib Ed.]
Oct 14 2011 at 10:34am
I would argue that Prescott and Kydland have had more influence on macro in the past 30 years. They create the entire RBC methodology and approach to fluctuations, as well as contributing to policy analysis and growth/development economics.
Oct 14 2011 at 11:35am
That is quite a good post. But I am afraid I can’t follow point 3. Perhaps too many negatives.
Oct 15 2011 at 7:04am
Yes on mark’s point above, could you be explicit for the undergrads in the audience? What policy approach? Otherwise thanks for the good read, I’ve been reading quite a range on the topic and this is a great breakdown.
Oct 15 2011 at 9:04am
1) Feynman expressed views similar to Hayek’s. I am not a Hayekian, and mention this both to emphasize that Hayek’s criticism is not, or is not only, a case of whining or special pleading on Hayek’s part, and to provide additional “proof by appeal to authority” for the statement.
2) The questions in comments (see mark and adlai) concerning point number 3 seem to be data points in support of Krugman’s observation on regression in the field (YMMV). The obvious answer to those questions is IS-LM &/or Keynesian economics more generally.
3) I don’t think S&S can be held “responsible for unleashing some of the most virulent strains.”
3a) Sims was responding to the project of Keynesian macroeconometric model-building that was prevalent before he arrived on the scene. It had been seriously criticized beforehand, but until he came along, it continued to thrive in academia. IIRC, he both built, & nailed shut, the coffin in which it was buried (in academia anyway). Hard to believe that even with the forecasting failures of the 70s, this would have happened without something to replace it.
3b) Sargent married or hybridized two distinct strains in macro, Lucasian theory and aggregate econometrics, and 1 more purely statistical, Box-Jenkins time series analysis. Of the 3, 2 were already very vigorous on their own, so it is not that surprising that the offspring was as well.
Oct 17 2011 at 4:52am
[Comment removed pending confirmation of email address, for crude language, and for rudeness. Email the email@example.com to request restoring your comment privileges. A valid email address is required to post comments on EconLog and EconTalk.–Econlib Ed.]
Comments are closed.