Monday, October 10, 2011

Three cheers for Sargent & Sims, one and a half for the "Economics Nobel"


Today, as you may have heard, the Sveriges Riksbank Prize in Economic Sciences in Memory of Alfred Nobel, more commonly called the "Economics Nobel," went to Thomas Sargent and Chris Sims. To get some info about Sims' work, see here and here, and to get some info about Sargent's work, see here.

First, what I think about today's award (as if that matters, ha). Short version: I heartily approve.

Both of these guys are data guys - empiricists, first and foremost. Specifically, both are macro data guys, which means the difficulty of their job is somewhere between Moses and Gandalf. I have nothing but respect for anyone who manages to make any progress at all in macro empirics. And furthermore, I think that good empirics is exactly what macro needs right now; the field has had a wild proliferation of theories that describe anything and everything, with far too few efforts to prune the idea tree and isolate which of the models are really describing the world. Thus, both Sims and Sargent have, in my eyes, been fighting the good fight.

This applies to Sargent especially. Sargent is famous for "taking models seriously," i.e. testing econ models like you would test a model in physics or biology. As a result, although Sargent is typically identified with the "freshwater" or "rational expectations" school of thought, his hardheaded empiricism was actually somewhat of a thorn in the side of the original RBC guys (Lucas and Prescott), who went so far as to ask him to stop testing their models! Also, his most famous work indicated that central banks, which we usually task with stabilizing both GDP growth and inflation, might not be powerful enough to do even one of those. This is a splash of cold water for both Milton Friedman-era monetarists and for modern believers in central bank omnipotence.

So, basically, Sargent is a badass. Scientists who exert serious efforts trying to knock down their own schools of thought are usually unsung heroes, and it's good to see one of them getting recognition.

As for Chris Sims, I'm much more familiar with his work, having taken a time-series econometrics course from the eminent Lutz Killian of Michigan. Sims is best known as the inventor of the Vector Autoregression, a way of looking at data that makes as few theoretical assumptions as possible. I like this approach. VARs have fallen somewhat out of favor due to their enormous standard error bands; rare is the properly specified VAR that actually makes a strong conclusion about what the data says. But I see this as a feature, not a bug! All too often, macroeconomists claim "empirical support" from the flimsiest of hand-waving (e.g. "The correlation in the data is 1.45, my model gets 1.13, so that's pretty close."). VARs are honest about what they don't know. 

So basically, I think these are great awards. However, my longstanding dislike for the "Economics Nobel" itself remains. 

What people need to understand about the "Economics Nobel" is that it is not a prize for a specific discovery, like the science Nobels (medicine, physics, and chemistry). It is more of a lifetime achievement award, like the Fields Medal in mathematics. The reason this is so is that to get a science Nobel your discovery actually has to be verified empirically, while in economics, convincing empirical verification is extremely rare. So what ends up happening is that "Economics Nobels" are generally given either for A) development of new techniques and methods, or B) theories that tell interesting stories. 

I have no problem with the development of new techniques and methods (it would be pretty dumb if I did!), but I think it's obvious that it's a lower bar to clear, science-wise, then what medicine Nobelists face. In biology, a new technique that leads to no new empirically verified discoveries (or demonstrably useful technologies) is just not that interesting. In economics, DSGE models won Nobels long before there was any DSGE model capable of making good macroeconomic forecasts (some would argue there still is no such model). The prize committee just thought DSGE was neat.

It's not just the prize committee that makes the decisions, though; it's the profession itself. Consensus is key. But in a field with as little empirical verifiability as macreconomics, consensus is something of a free parameter, subject to fads and herd behavior. In 40 years the profession may have completely abandoned DSGE, but Ed Prescott's Nobel will stand forever. 

The shifting winds of consensus lead to the politicization of the economics award. This has been noted and criticized by prize recipients Gunnar Myrdal and Friedrich Hayek. The "Economics Nobel" confers a level of respect to the theories of the recipients that is almost on a level with that accorded by the science Nobels, but the decision process is sort of halfway between the science prizes and the Nobel in Literature. Like economics itself on its worse days, the prize is always in danger of putting a sheen of "science-y-ness" on matters of opinion.

This year, of course, the politics are fairly obvious, and I happen to agree with them completely. Macroeconomies all over the world are doing some really weird stuff. There are lots of ideas, but little consensus, about which theories we should be using to understand events like the financial crisis, Little Depression, and current relapse, not to mention globalization and China. In times like these, it is best to look to the data first. Pruning the idea tree is more important now than ever. If today's award to Sargent and Sims has a political message, it is that.

But in the long run, I think it would be better if the Bank of Sweden adopted more stringent criteria for the awarding of the prize, more akin to the criteria for the Nobel in Medicine. That may mean fewer winners - perhaps only one a year, or even some years of "no recipient." It will certainly tilt the award toward microeconomists. But so be it. If you're going to call it the "Prize in Economic Sciences," then my opinion is that you should back that up.

In the meantime, though, congratulations to today's winners. Three cheers for better macro empirics!

18 comments:

  1. Anonymous8:52 PM

    "What people need to understand about the "Economics Nobel" is that it is not a prize for a specific discovery, ... It is more of a lifetime achievement award, like the Fields Medal in mathematics."

    Or like the Nobel for literature?

    ReplyDelete
  2. Good post. It would be interesting if we got to an empirical macro world... I hope I will see it in my professional lifetime (hasn't started yet, really).

    Do you have any comments re: Jeffrey Sachs' recent articles or his just published book, "The Price of Civilization"? It's bound to be rather topical at the moment. He recently posted an article called "A New Direction for American Economic Policy" on the Huffington Post. Interesting read, basically making an argument for increased public investment in things like education and infrastructure and... strangely enough the idea of paying for them through taxes!

    I made a blog post partially in response at http://socialmacro.blogspot.com/

    ReplyDelete
  3. This comment has been removed by the author.

    ReplyDelete
  4. The Fields is a 'lifetime achievement' award only if you assume life stops at 40. I think you might mean the Abel Prize.

    ReplyDelete
  5. "In 40 years the profession may have completely abandoned DSGE, but Ed Prescott's Nobel will stand forever."

    All right, now I'm depressed.

    ReplyDelete
  6. @nkh: It's not me who thinks mathematicians die at 40, it's the Fields Medal. Which is why the Fields Medal is dumb. And the science Nobels are dumb for not having any kind of posthumous version of the prize, meaning that Emmy Noether, Chien-Shiung Wu, and Lise Meitner will forever be relegated to second-class status by the sexism of the early 20th Century.

    Oh, and all prizes are dumb, period.

    ReplyDelete
  7. DealWithIt1:54 PM

    This is an awful selective overview of a prize for a man who went a great deal of the way towards burying Keynes, or at least the Keynes of the 1960s. This was very, very clearly a prize for freshwater macro.

    http://i.imgur.com/oxTYX.gif

    "Sorry, Art, but aside from the foolish and intellectually lazy remark about mathematics, all of the criticisms that you have listed reflect either woeful ignorance or intentional disregard for what much of modern macroeconomics is about and what it has accomplished. That said, it is true that modern macroeconomics uses mathematics and statistics to understand behavior in situations where there is uncertainty about how the future will unfold from the past. But a rule of thumb is that the more dynamic, uncertain and ambiguous is the economic environment that you seek to model, the more you are going to have to roll up your sleeves, and learn and use some math. That’s life."
    -Thomas Sargent, in an interview at History's Most Evil Regional Fed Bank

    ReplyDelete
  8. DealWithIt said...

    " This is an awful selective overview of a prize for a man who went a great deal of the way towards burying Keynes, or at least the Keynes of the 1960s. This was very, very clearly a prize for freshwater macro."

    And how has his work stood up?

    ReplyDelete
  9. OT, but I'm new here -- what did (do?) you think of Adair Turner's address to the INET conference here. He seems like the most sophisticated guy out there talking about these issues today. But, then, I hardly know them all. thanks,

    ReplyDelete
  10. Anonymous1:33 AM

    What in the world do you know aboit "testing a 'model' in biology"?

    ReplyDelete
  11. Anonymous1:38 AM

    You are talking out of the back of your pants.

    Tell that to Gerald Edelman or Charles Darwin, etc.:

    " In biology, a new technique that leads to no new empirically verified discoveries (or demonstrably useful technologies) is just not that interesting."

    ReplyDelete
  12. Anonymous1:43 AM

    The big problem here is that you have no idea what science looks like in domains of essentially complex phenomena where explanations of the prinicple are the rule and not the exception.

    I.e. you have no idea what a successful science of economics would be, because you begin with a patently false picture of what "science" is taken from a one sided dieted and a poverty of examples

    ReplyDelete
  13. It's very interesting that you think that, most recent Anonymous, given that Noah has a degree in theoretical physics. Maybe you don't think physics is a good example of what science looks like, but I'm willing to bet you don't have much company in that.

    ReplyDelete
  14. It's a very rare 'Anonymous' who has anything worth listening to.

    ReplyDelete
  15. @Noah: It's not just the Fields. Hardy thought that mathematical life ended at 40 and was quite vocal about that. Of course, he also thought that mathematics should be abstract and "useless". Which is funny since one of his most widely appreciated results is a very simple concrete application to the study of genetics. I do think that one should note the irony of the Abel prize. A career achievement award named after a man who died at 25. If only they would have settled on the Galois Prize.

    ReplyDelete
  16. Anonymous8:13 PM

    "As a result, although Sargent is typically identified with the "freshwater" or "rational expectations" school of thought, his hardheaded empiricism was actually somewhat of a thorn in the side of the original RBC guys (Lucas and Prescott), who went so far as to ask him to stop testing their models!"

    This is just false.
    Get the original interview to Sargent here: https://files.nyu.edu/ts43/public/research/SargentinterviewMD.pdf
    The quote is taken from here:
    "Evans and Honkapohja: What were the profession’s most important responses
    to the Lucas Critique?
    Sargent: There were two. The first and most optimistic response was complete rational expectations econometrics. A rational expectations equilibrium is a
    likelihood function. Maximize it.
    Evans and Honkapohja: Why optimistic?
    Sargent: You have to believe in your model to use the likelihood function. It
    provides a coherent way to estimate objects of interest (preferences, technologies,
    information sets, measurement processes) within the context of a trusted model.
    Evans and Honkapohja: What was the second response?
    Sargent: Various types of calibration. Calibration is less optimistic about what
    your theory can accomplish because you’d only use it if you didn’t fully trust
    your entire model, meaning that you think your model is partly misspecified or
    incompletely specified, or if you trusted someone else’s model and data set more
    than your own. My recollection is that Bob Lucas and Ed Prescott were initially
    very enthusiastic about rational expectations econometrics. After all, it simply
    involved imposing on ourselves the same high standards we had criticized the
    Keynesians for failing to live up to. But after about five years of doing likelihood
    ratio tests on rational expectations models, I recall Bob Lucas and Ed Prescott
    both telling me that those tests were rejecting too many good models. The idea
    of calibration is to ignore some of the probabilistic implications of your model
    but to retain others. Somehow, calibration was intended as a balanced response
    to professing that your model, though not correct, is still worthy as a vehicle for
    quantitative policy analysis."

    ReplyDelete
  17. Anonymous11:54 AM

    I'd be interested to hear how you get a model or data to have correlations greater than 1.

    ReplyDelete
  18. Thomas Sargent and Chris Sims are geniuses. I'm still blown away by the Vector Autoregression created by Chris.

    ReplyDelete