Short Piece on Probability Theory

6 August 2009 at 7:55 am Peter G. Klein 10 comments

| Peter Klein |

“Risk, Uncertainty, and Economic Organization” is my contribution to the Hoppe Festschrift. I got the topic idea from some blogger guy. My chapter focuses, as the title suggests, on Knightian uncertainty. Hoppe places Knight, along with Ludwig von Mises, squarely in the frequentist camp (typically associated with Ludwig’s brother Richard). I tend to agree with Hoppe although, as I discuss in the paper, there are many interpretations of Knight, and some commentators argue that subjective (Bayesian) probability theory renders untenable the Knightian distinction between insurable risk and true uncertainty (see, for example, Dick’s 1982 paper).

Ultimately, however, I don’t think the approach to the firm promoted on this blog depends on a particular interpretation of Knight. The central claim is that judgment represents a kind of decision-making that cannot be traded on the market, and that therefore requires the entrepreneur exercising such judgment to establish a firm (more specifically, to take ownership of capital resources). To put it differently, ownership of assets implies a kind of ultimate responsibility that the owner cannot delegate. I think one can be agnostic about exactly why judgment isn’t tradable — it could be a form of asymmetric information, rather than ontological differences between types of knowledge — and still buy the basic Knight-Mises-Foss-Klein approach to the firm.

Entry filed under: - Klein -, Austrian Economics, Management Theory, Strategic Management, Theory of the Firm.

The History of England and the Future of the Archive Better Coauthoring Tools in Office 2010

10 Comments Add your own

1. srp | 6 August 2009 at 8:43 pm

I like the rare clarity of this statement of the KMFK view of the firm. Now I can easily come up with counterexamples: People who give their money to mutual fund or hedge fund managers, say, or publicly managed corporations where professional managers sell their judgment to owners. These seem to violate this claim.

As these are not exactly marginal phenomena in the economy, I’m not sure how valid is the principle that “ownership of assets implies a kind of ultimate responsibility that the owner cannot delegate.” Possibly I am misunderstanding; trivially, for example, one could say that the owner is ultimately responsible for investing in a mutual fund (or hiring an advisor to pick a mutual fund or hiring an advisor to hire an advisor to…). But that doesn’t seem to be the intent of the claim. (If it is, then it is an uninteresting tautology.)

Rather, this KMFK formulation appears to be making a substantive point about an unbreakable nexus between asset ownership and direct entrepreneurial judgment (of projects, not advisors). And if it is, it seems both empirically testable and false. And since it is unlikely that the existence of mutual funds and public corporations was somehow forgotten in forming this principle, I am led back to the conclusion that I still don’t get what it is trying to say.
2. Peter Klein | 6 August 2009 at 9:20 pm

Sure, but what don’t you like about it?

Seriously, these concerns are all addressed in various papers, which you might enjoy reading. You’re simply defining judgment differently than we do. Professional managers do not sell judgment to owners; owners exercise judgment in hiring, monitoring, governing, and firing professional managers. What you call trivial and uninteresting is, we think, substantive and important. Oh well, de gustibus non est disputandum. Perhaps you take the view that the plant manager of neoclassical production theory, hiring as much capital as he wants in a perfectly competitive capital market, just as he hires labor and other factors, is the critical decision-maker in the market economy. Rather, we think that choosing output to set MR=MC is a second-order, indeed trivial and not terribly interesting (or, for managers, practical) problem. The central economic problem, in our approach, is the allocation of capital to individuals, firms, and industries. You might find Mises’s discussion of market socialism interesting in this regard. One implication, which I develop in my 2008 SEJ paper, is that the funder, not the “idea man,” is the real entrepreneur in a startup company. This has several non-trivial implications for organization, governance, competitive advantage, and so on.

As I said, it’s all in the papers, though perhaps not expressed with the clarity you desire.
3. srp | 7 August 2009 at 12:46 pm

Sorry to be so flip. I suppose it shows the need to balance clarity with precision–the clearer one seems to be the more important it is to pin down exactly what one means, otherwise misconception follows.

BTW, I think decisions like what products to launch and what technologies to pursue (and even how much capacity to deploy) are pretty important and are often delegated by owners. But I can see that with a different definition of “judgment” they aren’t covered by the KPFK claim.
4. Peter Klein | 7 August 2009 at 4:30 pm

I certainly agree that those are important decisions. We are just focusing on a different problem, namely who gets to make those decisions and how that is decided.
5. Brian Pitt | 8 August 2009 at 9:05 am

I do not plan to “steal,” but will the festschrift be available on-line.
6. Peter Klein | 8 August 2009 at 9:22 am

Yes, in its entirety:

Click to access property_freedom_society_kinsella.pdf

Free the books!
7. Rafe Champion | 9 August 2009 at 7:19 am

Surely class probability vs case probability (Mises) is the same as typical vs unique events (note 5, O’D and R). Frequency measures can be applied to classes but not to unique events. Compare the probability of a number coming up from the next role of the dice to the probability of a batter scoring in his next innings or the Red Sox beating the opposition next time out.
The propensity theory of probability has been developed to handle unique events, including quantum effects. Popper has taken this line although when he was involved with the Richard Mises and Karl Menger seminars he may have been a frequency man.
http://en.wikipedia.org/wiki/Propensity_probability

A lot of effort has been made to generate numerical probabilities for explanatory theories, I think it is genererally accepted that the Carnap line (objective probabilities) has died and now the Bayesians are making the running with subjective probabilities. I think they are doing no better.

The bottom line is that the outcome of unique events, including the decisions of entrepreneurs, are irretrievably conjectural. That does not eliminate the capacity to improve the chances of the desired outcome by planning and calculation to weigh up the options (especially to eliminate bad options). That is the approach used by the people who do maintenance on aeroplanes and it is also the function of risk managemenet for major ventures (think up everything that could go wrong and see if we can minimise the risk).

Buchanan and Vanberg have taken ideas from Popper and Prigogine to show how entrepreneurs can make creative decisions that cannot be predicted by equilibrium models or teleological (deterministic) models of the world. See Chapter 18 of “The Philosophy of Economics” ed Hausman.
http://books.google.com.au/books?id=xxo10t2x1mkC&pg=PA315&lpg=PA315&dq=Buchanan%2BPrigogine&source=bl&ots=CKbqH0WOuc&sig=AqXYV6i-yErObUkgqpGCuuzPoK0&hl=en&ei=87p-StfkPI-0sgPj4I3vCg&sa=X&oi=book_result&ct=result&resnum=1#v=onepage&q=Buchanan%2BPrigogine&f=false
8. Rafe | 9 August 2009 at 8:16 am

A bit more from Buchanan and Vanberg’s paper.

“Our purpose here is to relate the new orientation in the natural sciences [indeterminism] to a particular nonorthodox strand of thought within economics. [This] involves a shift of perspective from the determinism of conventional physics (which presumably inspired the neoclassical research program in economics) to the nonteleological open-endedness, creative and nondetermined nature of evolutionary processes.”

The practical concern in the background was the failure of central planning and they considered three views of economic events.

1. Allocation of resources.
2. Discovery of opportunities to make a profit.
3. Creation of opportunities.

“We have suggested that a perceptual vision of the market as a creative process offers more insight and understanding than the alternative visions that elicit interpretations of the market as a discovery process, or, more familiarly, as an allocative process.”

The most basic (level 1) critique of the system pinpointed the defective allocation of resources as a consequence of the incentive structure which did not reward good decisions (or punish bad ones). Kirzner followed Hayek to add another element to the critique because even if the incentive problems were solved there would still be the problem of dispersed knowledge and the need for entrepreneurs to see better ways to do business (level 2). Then Buchanan and Vanberg advanced the crtique another notch to level 3.

“We suggest that the critique, even as extended, falls short of capturing an essential element in any comparative assessment of the market and the planning alternatives. The teleological feature remains to be exorcised.”

The teleological element is the idea that the system as a whole has a specific destination or end point in view, a position which depends on the assumption of determinism. That is their critique of equilibrium approaches in general, these all make the illicit assumption of deterministic end points unless equilibrium models are clearly recognised as analytical devices like the frictionless pulleys and weightless strings of the physicist. The key to the creative process is the idea that something can come from nothing which is an important corollary of indeterminism in physics and the world at large.
9. Ludwig van den Hauwe | 30 August 2009 at 1:17 am

Ludwig von Mises himself says very explicitly that he does not belong to the frequentist camp in the philosophy of probability.
How is it possible then that this misunderstanding goes on thriving?
See also my (2008) paper _John Maynard Keynes and Ludwig von Mises on Probability_, Procesos de Mercado, Vol. V, N° 1, 11-50.
10. Ludwig van den Hauwe | 31 August 2009 at 1:25 pm

Peter Klein writes: “There are many interpretations of Knight.”
Which interpretations of Knight do you know?
The only correct interpretation of Mises is that he very explicitly rejected the frequency view, which, in the version of Richard von Mises, is an objective view, bound up with indeterminism in physics, and with the mainstream interpretation of quantum physics, all of which Ludwig von Mises explicitly rejected.
This is not debatable. Sorry for Hoppe. Klein should have known better since he is familiar with my paper. The relationship of Ludwig von Mises´ view to either objective or subjective Bayesianism is perhaps debatable. For Bayesians all probabilities are quantifiable. With respect to unique events Mises would object that this is metaphorical. Still one could argue that the use of metaphors is legitimate in science. In any case both Ludwig von Mises and the Bayesians adhere to an epistemic interpretation of probabilities.