Starting the New Year Off Right: Welcome to 2009

Kenneth Boulding, Thomas Schelling, Albert Hirschman, and Jon Elster have longed topped my list of social scientists that I recommend to my students to read and learn from.  What they can learn from these thinkers is almost limitless in terms of communication skill, style of argument, breadth of perspective, and insight.  In each case I also have fundamental disagreements with aspects of their argument about the nature of economic processes and the relationship between the citizen and the state.

So I recommend to all to read Elster's latest paper "Excessive Ambitions" (a new working paper released in 2008).  It is a timely critique of formalistic models of maximizing behavior and perfect competition as well as standard statistical analysis in the social sciences.  Unfortunately, I think Elster (who is an expert in rational theory) is not subtle enough in his critique and often uses the term "rational choice theory" to capture not only the conceptual use of this analytical framework, but its formal structure, its predictive power, and its normative use.  But one can be a critic of models of hyper-rationality and omniscient agents without denying the value of rational choice framework in economics, politics, law, sociology, and history.  As Lin Ostrom once put it, we can have a rational choice theory as if the choosers were human.  In fact, that is what Mises's laid out for us in the first 100+ pages of Human Action.

Elster argues in this paper that the excessive ambitions of rational choice theory and misguided statistical analysis represents not only a waste of intellectual resources (something Kenneth Boulding actually argued back in the 1970s), but that it produces real harm in the societies where people come to believe the information provided by these models and analysis.  As an example, he cites the case of Long Term Capital Management.

Elster argues that social scientists curtail their ambitions and admit our ignorance.  Of course, readers of this blog should remember that in the 1940s Hayek embarked on his "Abuse of Reason" project, in which he adopted Hume's dictum to "use reason to whittle down the claims of reason." The Counter-Revolution of Science contains a critique of excessive formalism, excessive aggregation, confused statistics, and excessive ambitions fueled by excessive arrogance.

Is Elster merely rediscovering Hayek, or is his argument completely different?  If similar in intellectual argument, why so different in ideological interpretation?

Comments

[Mario Rizzo]

I must read Elster's paper! However, I have been thinking about this recently in connection with the critique that Edmund Phelps and others have made of those who incorrectly, or so it seems, evaluated the risks of mortgage-backed securities based on sophisticated financial models. His critique involved not understanding the difference between Knightian uncertainty and risk. The problem emerged when the market participants believed these scientistic models. Few, it seems, had the courage to argue against the modelling establishment and say -- hey, we don't really understand the uncertainties involved, etc. Is this a defect of the market? Or is it the consequence of a spirit of scientism that the market cannot have been expected to eliminated ex ante? Were the market participants insufficiently rational or hyper rational? A great puzzle to solve.

[Matt]

I cannot find the paper you cite, would you please provide a link?

[knapp]

Maybe Austrians need a rethink on their embrace of Knightian (Keynesian?) uncertainty.

The concept has not proven itself to be a very good marketing tool for capitalism. In fact (as you know) Post Keynesian Institutionists use it as a justification against capitalism!

The positive case for capitalism needs to be built on the stronger epistemological foundation of "certainty", not omniscience, but certainty in the proper contextual sense.

If this is indeed what Knight meant by his definition of uncertainty then great, just rename it Austrian (contextual,endogenous) Certainty and change your advertising campaign.

[Greg Ransom]

It struck me that there is no good sophisticated Hayekian book or paper targeted specifically at the economic and explanatory and empirical mistakes of the economists.

Is there a good, in depth, sophisticated Hayekian book or paper on the mistakes of econometrics?

Is there a good, _in depth_, modern Hayekian book or paper on the mistakes of the macroeconomists? There's Garrison, but Garrison really mostly skims the surface, for the purposes of presenting his own positive economics.

Why aren't there any Austrian Jon Elster's?

[matthew mueller]

Greg Ransom,

The best "good, in depth, sophisticated Hayekian book" available is Don Lavoie's "National Economic Planning: What is Left?". His chapter entitled "The Knowledge Problem" is exactly what you are looking for. I revisit that chapter whenever I encounter a sophisticated liberal who advocates positive government legislation.

Knapp,

The Post Keynesian / Institutionalist defense of radical uncertainty is in no important way connected to their political views, much in the same way that one need not be a libertarian to appreciate uncertainty. G. L. S. Shackle is the best source to consult on this question, as he recognized that legislation is helpless in the face of uncertainty. (Paul Davidson fails to recognize this point.)

Also, I disagree with Pete Boettke's assertion that one can jettison "hyper-rationality" while maintaining the assumption of rational choice theory. There are just too many problems. One example (by no means conclusive) is that in order to behave "raionally", one must know what information to exclude; such an analysis presumes (implicitly, at the very least) that the agent possess complete information, even if accessing it is costly.

An interesting literature to explore for those interested in rationality is the book and related works connected to Adorno and Horheimer's "Dialectic of Enlightenment." They conceive of rationality as power, as it is concerned with the domination of nature and man. The elimination of all contingency (the claim of reason) is inherently totalitarian, and libertarians should, I believe, take these arguments very seriously.

[Peter Boettke]

Matthew,

You might want to look at Lin Ostrom's presidential address to the APSA and also John Searle's Rationality in Action as well as revisit the first 100 pages of Human Action. I wasn't making an assertion, I was pointing to a literature that argues that discussing human purposive action is not identical to discussing agent behavior in an formal structure.

There are issues, but we don't need to have single exit models and we don't need prediction or prescription to realize the conceptual benefits of rational actor models.

[Brian Pitt]

I have to read Elster's paper. Thanks for the heads up.

Btw, In an interview conducted by Swedberg for "Economics and Sociology: Redefining their Boundaries," he is quite subtle in his critique of "rational choice," and pretty much exhorts sociologists, historians, and political scientists to brave the fatiguing climb of learning it.

[knapp]

Greg,

I'm not certain, but I think you make a radical departure from orthodoxy when you claim that the Post Keynesian defense of radical uncertainty is in no important way connected to their political views. It's true i'm relying on Davidson for this perspective, but who else has done the heavy lifting within PK on the inherent uncertainty of the market and the need for political institutions as a stabilizing agent.

The sooner Austrians go back to their Mengerian/Aristotelian methodological roots and dump Shackle and his radical uncertainty, the better. Otherwise, just go all the way and designate Austrian Economics a mere free market branch of Post Keynesianism.

[matthew mueller]

Hey knapp,

I address the question of the relationship between Paul Davidson's political views and his theory of uncertainty here:

http://post-austrianeconomics.blogspot.com/2008/12/in-response-to-paul-davidson.html

I am convinced that Paul Davidson's own theory of uncertainty (ontological uncertainty) doesn't permit him to positively advocate economic regulation. And I would argue that it is precisely in Menger that we see the defense of uncertainty (there is a reason why Lachmann believed he was following Menger in his emphasis on disequilibrating market tendencies). I think that it is the influence of the libertarian dogma that has pushed us away from these important questions of uncertainty and coordination.

[Alex]

Perhaps a relevant point to keep in mind is that Jon Elster has been engaged in discussions on Nassim Taleb's ideas recently -- even before the financial leveraging hit the fan.

Taleb makes the point --in his strident style-- that reliance on dubious mathematics (he calls it charlatanism) leads to overconfidence in one's risk estimation, hence to greater leverage and further, it leads to inducing systemic risk in finance.

Taleb praises Hayek and Shackle but he does not seem to be economically literate beyond that and he has a low opinion of economists. Taleb is a friend of Paul Wilmott who makes similar points in a less radical garb.

It is not surprising that financial academics would not be enthusiastic in making the point that their models share a good part of the blame in the current mess -- but it seems a point well worth making.

[knapp]

Matthew,

First of all, great job on your blog, I especially like the variety of your topics.

Of course, Paul Davidson refutes you in your comments section which I think supports my point that Greg's viewpoint was certainly unorthodox.

I'm starting to feel like a skunk at the garden party on this blog with my bashing of Shackle, uncertainty etc.. Feels like the Lachmanian wing of the Austrian Party. That's OK. I could use the refresh. But at this point, I'm still inclined to feel that the Austrian focus on radical uncertainty and disequilibria is a poor tactic if the goal is untimately free market policy persuasion.

To continue my skunky behavior, I think Ayn Rand's focus on certainty and contextual knowledge also dovetails well with Menger but with a focus on the positive forces of human cooperation and the market process.

[Greg Ransom]

Matt, I'm well familiar with this very good book. I recommend it as well. But the book really doesn't so much address my earlier post (indeed, Lavoie's more technical book on central economic planning -- which I believe was his dissertation -- more directly addresses the detailed and specific ins and out of good economic explanation and bad economic explanation).

You write:

"The best "good, in depth, sophisticated Hayekian book" available is Don Lavoie's "National Economic Planning: What is Left?". His chapter entitled "The Knowledge Problem" is exactly what you are looking for."

[Bill Stepp]

Taleb makes the point --in his strident style-- that reliance on dubious mathematics (he calls it charlatanism) leads to overconfidence in one's risk estimation, hence to greater leverage and further, it leads to inducing systemic risk in finance.

Alex,

We don't need to invoke reliance on math models to understand why lots of balance sheets got overleveraged. Home builders, to name one industry, didn't get overleveraged because they were using math models--they were not.
On the other hand, Moody's, a rating agency, used very little debt, yet built dodgy models of the likelihood of default of various bonds, including mortgage-backed bonds.
As an anarcho-capitalist friend who works there (and who is known to lots of other a-cs in the libertarian community) pointed out, they forgot that housing prices could ever fall, so the only sign in the housing price variable was +. Oops. (He didn't work on their models.)

[Steve Horwitz]

Greg,

I can't find an email address for you, but I wanted to thank you for pestering the "Curious Capitalist" blogger at Time about reading the Austrians. Shoot me an email if you have a sec.

[michael webster]

@Peter;

I have not read Elster's new paper, but Long Term Capital blow up was the result of uncorrelated bonds becoming correlated and moving together. This is a fairly serious problem for any actuarial theory of risk, whether based on probability or some other theory of uncertainty. I am not sure how much progress on this problem will be made by throwing up ones hands and calling it "unquantifiable".

@Mario;

If you read all of Phelp's op-ed piece, http://online.wsj.com/article/SB120545520548735247.html?mod=opinion_main_commentaries, you see that he constantly switches back on forth between a critique of certain assumptions of the rational choice model and then use some version of rational choice to explain/predict why the Fed cannot maintain low interest rates for very long.

His actual critique of the subprime models was: there was no sufficient past history to model, and the history was changing in any event. But that is no more than to point out that smart people tried to fool themselves about the power of their model - it isn't a radical critique of calculable risk versus some other type of risk.

[Greg Ransom]

Has any Austrian written on the Nassim Nicholas Taleb vs. Robert Merton feud?

Taleb is on record stating the the #1 thing he would do if allowed to do anything which might improve the world is to eliminate every every academic economist in the world -- except for the Austrians (Open University lecture, 2008)

[matthew mueller]

Greg,

A good question. Taleb does speak favorably of Hayek. I know that Jeffrey Friedman plans to publish a composium of Taleb's work in Critical Review. Maybe we can try to get an Austrian to contribute a piece?