September 2022

Sun Mon Tue Wed Thu Fri Sat
        1 2 3
4 5 6 7 8 9 10
11 12 13 14 15 16 17
18 19 20 21 22 23 24
25 26 27 28 29 30  
Blog powered by Typepad

« The Next Ten Years of Research Among Elite Economists | Main | Gaus on Liberal Justice and the Liberal Virtue of Tolerance »


Feed You can follow this conversation by subscribing to the comment feed for this post.


Right on, Pete. Some of this stuff is like a lot of modern art: What it is about is itself, not anything outside of itself.

Thank you for this courageous post.

We need a more substantive account of what the mistake of "Scientism" really is in relation to the phenomena & how & why it kills genuine scientific achievement than we currently have.

Simply charging Sargent & Sims with using math, data & mathematical statistics isn't enough.

This is a burden which Hayek nor any contemporary Austrian has adequately discharged.

Unfortunately economics today tries to be physics but misses one important point: physicists admit, that it is impossible to precisely calculate a system with more than a handful of particles interacting. Statistical methods will only be applied for gaseous states, where all particles are identical in a homogenous surrounding without structure (a feature which is certainly not true for human society). Economics try to calculate a multi-particle problem with the methods used for gaseous states, an effort doomed to fail.
So returning to more descriptive methods (which most physicists will apply to multi-particle problems) or to admit the limits of knowledge might be an interesting alternative

Its not obvious to me that Sargent has killed analytical diversity and scholarship in economics. I like the fact that he has given attention to economic history and history of thought. In addition,some of his work, such as The End of Four Big Inflations, uses narratives, graphs and tables to support the argument. Perhaps he has acted in other ways to restrict the scope and method of economics, but I don't see it in his work. On the other hand, if graduate students are led to believe that his Macroeconomic Theory is the sum of all knowledge that could be a problem.

I would agree, Pete, that Sims is largely a technician who got it for seriously developing VAR, whatever one thinks of that.

The real question is Sargent, who has without question been on the A list of those likely to get the award at some point (and, well, Sims has been on that list as well, if a bit behind Sargent). Unfortunately, I think you have things really botched up with regard to him.

You say he was a good guy in his early work but then somehow went bad. Huh? In his early work he was the hardlining rational expectations advocate, declaring it to be axiomatic that economists must assume it in all their work, harder line than Lucas. This is what you approve of, Pete, axiomatic ratex?

Then about 20 years ago, he blew it off. In his book on Bounded Rationality and Macroeconomics, 1993, Clarendon, he said that it was he fall of the USSR and its system that made him give up on ratex. Well, it is still in his models, but as an an asymptotic goal that the system tends to if agents have good learning mechanisms in their real world of boundedly rational adaptive expectations.

The hard fact that nobody has said (and I seem to be unable to post on my blog from Florence to do it), is that in a way this Nobel is another one for bounded rationality. Even though the Fin Times and many others view it as a prize for ratex, it is not. Sims is into "rational inattention," and I have already described what Sargent is into, which you seem to disapprove of for some unclear reason, Pete. In any case, I think Sargent's later work is far superior to his earlier work, and I have no problem with this prize, despite the multiple misinterpretations of it, including by gonzo Minnesota types viewing it as another one for rational expectations, which the FT agrees with without a shred of evidence from those in Stockholm making the award.


You are reading my post the wrong way. Analytically and methodological I disagree with Sargent in the old days and over the past 20 years.

And my impression of his attitudes is based not just on reading his work and reading his methodological pronouncements, but from personal interactions with him during my year at Stanford where he was at as well. He is no doubt an absolutely brilliant mind, and within the given rules of our profession he made amazing contributions and transformed the discipline. But he is also the first person I ever heard in person make the distinction between "scholarship" and "science" in economics, where he exalted science, and denigrated scholarship. I know he has done historical research -- in fact that year at Stanford one of the most original insights I heard was from Sargent on the fiscal origins of hyper-inflation. He is VERY smart.

But his rhetorical stance is always to privilege "science" over "scholarship".

He is the quintessential model and measure economist. He transformed the way we model and measure, but we are still modeling and measuring and if you don't do that, then you are NOT doing economics.

That is what I am objecting to --- I am NOT objecting to the prize within the existing structure of economic argument, in fact, I would wholeheartedly agree that it is richly deserved. Just as Samuelson's prize was richly deserved or Arrow's, or a host of others.

But at another level -- a Mises/Hayek paradigm perspective -- this formalist and positivist direction of economic research is not an advance. As Kenneth Boulding wrote when reviewing Samuelson's Foundations for the JPE, the flawless precision of mathematical economics may prove impotent in comparison with the literary vagueness of economic sociology with respect to understanding the messy world we occupy.

So it is not about axiomatic ratex that I am objecting, it is SCIENTISM and the methodological (as opposed to substantive) definition of economic understanding.

Maybe I'm wrong, but what I have read from Sargent tells me he didn't offer anything new, but merely translated old ideas into math. Can someone point me to a new idea he had?

I've only read Sargent's textbook, Dynamic Macroeconomic Theory (and not his other work) and my comments are limited to his contributions to "real-life" economics.

Sargent's models have the advantage of being based on micro-foundations and being explicit. However, like Keynesian models, they are extremely simplistic and of limited practical use. Any attempt to simplify the institutional and informational richness of real economies is bound to fail, and will have effectively no predictive power. Like climate models which simply can't mimic the richness of the real climate, none of the mathematical models of economists have the capacity to mimic the richness of ANY real economy.

The result is that when it comes to managing real economies, the world's operational economists (people with economics degrees who work in governments) are forced to resort to their hunch about what might be happening.

It is at that point that ideological foundations (either pro- or anti-liberty) of the policy maker take effect.

By now I've forgotten almost all the mathematical complexity of Sargent's models because these did not have any realistic relationship with the real world.

What remains behind from my five years of PhD studies are key insights which have been later buttressed by further reading. In that regard I find the richness of thinkers like Hayek and Mises to totally over-ride the flimsy deductive (NOT scientific!!!) models of Sargent and Co.

Consider Part 2 of Sargent's text cited above. A set of deductive gymnastics more irrelevant to the real world of money and currency can't be imagined. None of that can prevent "global financial crises" and the lot for it has not the slightest mention of the basic institution of the central bank or the concept of government failure.

Mathematical fiction is being rewarded with Nobel prizes these days. I, on the other hand, prefer to think through issues on my own by considering ALL possible human incentives and institutions that are at work in a particular situation.

At the end of the day, modern macroeconomics has become entirely irrelevant.

Barkley, You could find quotes from Hayek that directly stake out very nearly exactly this same position:

"Rational expectations is still in his models, but as an an asymptotic goal that the system tends to if agents have good learning mechanisms in their real world of boundedly rational adaptive expectations."


If you think this is about the use of math in economics, then you are confused on this issue. Look at my description of pre-WWII economics, math and stats are in the mix.

It is about methods and methodology, but not as you are thinking about it.

Fortunately, nobody cares what Barkley Rosser, Pete Boettke, Greg Ransom, or Sanjeev Sabhlok thinks about Sargent and Sims. Your position as a fringe group of crackpots is secure, guys, keep up the pointless work.

Really nice nobel price :D Hard to get it !!

Just keep working it's great job !

Sanjeev Sabhlok is absolutely correct. As to the point that "nobody" cares what we think. This is absurd. If you seek the evidence, look around you!

In any event, I do not hold my beliefs because they are (or even will become) popular but because they are the result of my honest thinking. I am more interested in my integrity than in anyone's else's.

I share some of Sanjeev Sabhlok's sentiments, especially in regards to graduate studies in economics.

But my opinion is that mathematical modelling of the economy has its place, as an alternative (though not necessarily superior) methodological approach. It is unlikely these models will be useful from a quantitative public-policy perspective (ala Hayek's 'Pretense of Knowledge' syndrome), but their conclusions can still be *qualitatively* useful.

Mathematical approach to economics can be grouped into two categories, though I should stress that both are inter-related:

(i) simplistic mathematical models (e.g. Solow growth model, Cobb-Douglas functions, etc.)

(ii) econometrics a.k.a. statistical interpretation of data

I once suggested on this blog for economics to be taught in two components:

(i) from a philosophical perspective - Adam Smith, Hayek, Keynes, etc.

(ii) mathematical modelling of the economy - which is to translate the philosophical, economic framework into mathematical language

Like it or not, mainstream macroeconomists tend to reason via mathematical models (and not always conscious of the limitations of these models; or that their conclusions are sometimes very contingent on the flawed assumptions they make). Students learn these flawed habits, which is quite easy to demonstrate: just try asking someone with a bachelor in economics as to their fundamental economic convictions, they will struggle to articulate them.

Anyway, to engage with mainstream economics, there is a necessity to learn the mathematical modelling 'game' - and beat them at it!

Unfortunately Austrians have failed in doing this. Rather, they have largely remain mired in methodological criticism - which is not entirely invalid, but it's easier to criticize than to build.


I just wanted to underline my suggestion of having economic teaching being separated into (i) philosophy; and (ii) mathematical modelling.

I think this is a particular good time to stress this point - when most economists are particularly aware of the limitations and failings of mathematical models, and yet struggling to find answers.

A good indication of this 'struggle' would be MIT's Caballero's paper - "Macroeconomics after the Crisis: Time to Deal with the Pretense-of-Knowledge Syndrome". In his paper, I think he misinterprets what Hayek meant by 'pretense of knowledge', but at least he's honest about the necessity for seeking a new approach to macroeconomics (both research & teaching).

All I'm saying is that the relation between these -- and the mistake made by Sargent & Sims & the rest -- has not been satisfactorily shown or explained, esp. directly and in detail addressing the work of folks like Sargent and Sims.

If I've overlooked something new in the literature which does this, point me to it.

Pete writes,

"If you think this is about the use of math in economics, then you are confused on this issue. Look at my description of pre-WWII economics, math and stats are in the mix.

It is about methods and methodology, but not as you are thinking about it."

And as I've always insisted, their is a legacy of a deeply false view of science and explanation which comes with the very language of "method" and "methodology" which is directly tied to JS Mill and the failed nomological deductive model of explanation and science, which views "science" as made up of "laws" justified by the successful prediction of particular events suing 2 variable linear equations.

The math does not exist to understand economics. We are only now getting some computational models that can do some things -- but these are only really proving everything the Austrians already said.

The challenge has to be a choice between a sound explanatory frame and a failed explanatory frame ie it has to be between successful science and scientific deficiency and failure. This has to be laid out and directly shown the way Galileo did it or the Continental driff guys did it.

It will NEVER work if it can be dismissed without consideration as a debate over methodology -- which economists don't consider "science" but dismiss without a second consideration as "philosophy".

By the way you can expose how the failed explanations and failed science of the mainstream is motivated and conceived out of its philosophical roots in a failed and debunked conception of science and knowledge -- but exposing this fact has to be a secondary endevour in addition to the direct exposing of the failure and logical unsoundness of the explanatory effort.

From the "methodological" point of view current at the time, Newton was aware that what he was doing wasn't scientific knowledge as that was understood at the time: it didn't satisfy the methodological demands of his time. It wasn't demonstrative knowledge. And the math itself had similar issues.

Darwin thought he had to satisfy Whewell's image of "science" -- he tried to follow Whewell's official methodological dictates for "science".

Ultimately, Darwin realized his explanatory problem along with his explanatory mechanism really didn't much involve Whewell's"general laws or laws of causation. And his explanation really couldn't be proved in any ordinary sense of the word.

Galileo, similarly, knew he was doing something outside the bounds of what was understood to be science. He justified himself in part by rehearsing many of the well known problems with the ruling picture of things.

In each case, Newton, Galileo and Darwin ultimately were fighting against a demonstrative image of "knowledge" and science inherited from Euclid via Aristotle, and variations on that paradigm evolving across in a whole variety of different forms and incarnations.

But Newton, Galileo & Darwin didn't fight this fight on "methodological" grounds, they let their explanatory strategy speak for itself, matching it up against the alternative in terms of could be accounted for -- and pointing out only here and there how things things fit or how things differed with the official prescriptions for producing "knowledge" or science evolving in the tradition descendent from Euclid & Aristotle.

Out of the box, many of the great scientific achievements didn't fit the picture of science and knowledge held by the leading scientists and academics.

It was only after the scientific community had switched allegiances that the old school picture -- e.g. ether physics, or Cartesian physics, or Aristotelian physics, or intelligent design biology or vitalism -- could be belittled as "scientism", or "philosophy", or "anthropomorphism".

Similar examples could be identified in the history of mathematics, e.g. all the debates over trying to make sense of the infinitesimals of the calculus, non-Euclidean geometry, set theory, the cognitive status of mathematical entities, etc.

You don't win these battles by debating the one true "methodology" .. you win them by setting up alternative rivals side by side and pointing out the strengths and weaknesses of each, and the hidden or not so hidden pathologies embedded in the stated assumptions -- and unstated background assumptions -- of the rivals.

This is what Galileo - famously -- did.

It is in part what Hayek tried to do in his essays and books of the 1930s and 1940s.

It's what Hayek sought to do with his book on global brain theory, battling associationism, positivism and behaviorism.

It's also what Chomsky sought to do with language & the mental battling behaviorism; and what Wittgenstein sought to with language,logic & mathematics battling the Frege/Russell/Carnap formalist project.

It isn't easy, but it doesn't start with laying down a set of methodological stipulations and then battling over "methodology".

The battle has to be between real explanatory examples and exemplars, and the "methodological" lessons fall out of what is learned from the examples.

It doesn't work the other way around.

I once got Lawrence White to understand and acknowledge this point years ago on my old Hayek-L email list.

But its a point so important that it can't be rehearsed enough.

What you have to have is "inference to the best explanation" via head to head comparisons and evaluations of explanatory rivals - the examples teach far more and deeper lessons than can be articulated in any simple set of "methodological" stipulations derived from the examples or developed in the process of explicating the explanatory rivals.

Here, Roger Garrison's _Time and Money_ is exemplary.

But we must acknowledge also that the domain of what is going on in economists is more vast even than what Roger has conquered in his excellent explications of the differences between Hayek and Keynes and Friedman.

Don Lavoie's _Rivalry and Central Planning_ is another example of the power of placing explanatory rivals into head-to-head comparison.

It was "Game Over" for the rivals to Mises-Hayek on central planning after the publication of Lavoie, where prior to Lavoie Mises-Hayek was the universal loser, a scientific FAIL as far as the profession was concerned.

I will say that blogging made me realize that I wasn't interested in people who judged me or made me feel bad for being who I am. It's a huge relief getting rid of those people. Like you said, the awesome team is who you want around you.

Nice post, really interesting, keep doing it !!

The comments to this entry are closed.

Our Books