On the Cutting-Room Floor
Mises Review 11, No. 1 (Spring 2005)
THE CHANGING FACE OF ECONOMICS: CONVERSATIONS WITH CUTTING EDGE ECONOMISTS
David Colander, Richard P.F. Holt, and J. Barkley Rosser, Jr., Editors
University of Michigan Press, 2004, x + 358 pgs.
For the editors of The Changing Face of Economics, “cutting edge” is more than a phrase. They have an elaborate theory of how change takes place in economic theory. Brilliant young economists, who have secured teaching positions in the “best” departments, develop new ideas. These new views do not deviate completely from the dominant view. The new economists remain within the mainstream and do not form new schools of their own, institutionally isolated from more conventional departments. Quite the contrary, the ideas of these “cutting edge” figures eventually become the new dominant paradigm.
The editors have conducted interviews with a number of economists they classify as cutting edge: these include Deirdre McCloskey, Matthew Rabin, Robert Frank, Duncan Foley, and several others. The book also includes interviews with two demigods of neoclassicism, Kenneth Arrow and Paul Samuelson.
To the editors, complexity theory is the wave of the future, and this view has guided their choice of cutting-edge figures. This poses for me a severe problem. I can make very little of complexity theory; and, what is more, the editors’ account of theory change interests them much more than it does me or, I suspect, readers of these pages. Why then review the book? Several points discussed in the interviews raise important issues for those who believe that Austrian economics, rather than the latest mathematical fad, should be at the cutting edge.
The economists of this book on occasion address Austrian themes, but they have little patience with Austrian solutions. Neoclassical economists no longer take for granted models in which everyone has perfect information, and the economics of information is today a key topic. Austrians will at once call to mind the key work of Friedrich Hayek: has he not shown how the price system coordinates a mass of information that far exceeds the capacity of any central planner to grasp? Hayek famously claimed that much of this information could not be fully articulated. Only local actors, aware of their particular circumstances, grasp this information; and much of their knowledge is tacit rather than explicit. Nevertheless, by reacting to changes in prices, everyone can make use of this local knowledge.
Kenneth Arrow is not impressed. He remarks: ‘I read that paper [apparently, “The Use of Knowledge in Society”] carefully a few years ago and think it is incoherent because he [Hayek] stresses the idea people can only know their local thing. He doesn’t ever answer why the system’s price conveys the correct global information. In the Walras/Pareto story you do wind up with the correct information—subject to the appropriate assumptions—but Hayek doesn’t like that story” (p. 302).
Arrow’s comment rests crucially on a single word: “correct.” Only if prices meet various requirements that Arrow imposes will he recognize market prices as efficient. If, e.g., Pareto superior moves—changes that will make at least one person better off without making anyone else worse off—are possible, then the market has failed to generate socially optimal prices.
Arrow fails to see that the Austrians reject altogether the view that the market is to be subject to external standards of correctness. Rather, the Austrian point about information is that participants in the market use prices to calculate the benefits and costs of their proposed courses of action. Changes in prices reflect local conditions; since market actors use prices in their calculations, they necessarily take account of locally available knowledge. Hayek endeavors to show how the market makes use of information: unlike Arrow, he does not weigh that use of information in the balance, and find it wanting. As usual, Mises states the key issue with exemplary clarity: “the prices of the market are the ultimate fact for economic calculation. It [calculation] cannot be applied for considerations whose standard is not the demand of consumers as manifested on the market. . . . Economic calculation in terms of money prices is the calculation of entrepreneurs producing for the consumers of a market society. It is of no avail for other tasks” (Mises, Human Action, Mises Institute, 1998, p. 217).
Robert Axtell falls victim to the identical misunderstanding. Like the Austrian approach, Axtell’s parallel processing view emphasizes “decentralized, local information. Everybody has his or her own idea of how the world will unfold. That seems to be a natural way to think about how markets really work, as opposed to the centralized view of the auctioneer where there is one entity that clears all the markets and knows all the prices and knows exactly what utility functions everybody has. It’s very close to an Austrian view” (p. 282).
Unfortunately, not close enough. Axtell rejects the Austrian assumption that the market price is the “socially right” price. “It may be the socially right price, it may not; it may depart from the fundamentals; it may not” (p. 282). Like Arrow, Axtell wrongly thinks that Austrians judge the market by an external standard of social rightness.
For many neoclassical economists, all economic behavior is assumed to be strictly governed by a narrowly interpreted standard of self-interest. Behavior that appears not to conform, such as donations to charity, must be explained away. Austrians take an altogether broader view. No “rationality” requirements constrain the choice of ends; Austrian economists accept peoples’ preferences as they find them.
Matthew Rabin also finds the standard view problematic. He considers the “ultimatum game.” By the terms of this puzzle, someone will receive ten dollars if he can get a second person to agree to his proposal to distribute the money. If the second person rejects the proposal, neither gets anything. Does this game have a rationally unique solution?
According to some proponents of neoclassical rationality, it does. The first person will offer one dollar to the other, retaining nine for himself; and the offer will be accepted. (We assume that the money cannot be divided into smaller units.) Both are motivated by self-interested reason: the first person, who has sole authority to pick the distribution, chooses the most he can get, given that he must offer the other person an incentive to accept the agreement. Since the other person prefers something to nothing, it is rational for him to agree. Kenneth Binmore, a leading game theorist interviewed in the book, finds this analysis entirely right.
Rabin does not. In experiments, many people reject the “rational” split as unfair. Rabin comments: “Many economists wanted to see turning down money in the ultimatum game as irrational behavior. The level of energy economists have spent . . . in believing that people turning down money . . . isn’t rational behavior or that it indicated that people didn’t understand the experiment, just seems to me bizarre” (p. 150). Austrians would agree.
Robert Frank points out another case in which the neoclassical view of rationality is inadequate. Suppose that you eat in a restaurant to which you never plan to return. There is no self-interested reason to leave a tip; but “people seem puzzled by the fact that the model would predict that behavior [not leaving a tip] , and they don’t seem to think that they’d want to do that” (p. 117). Like Frank, Austrians see no need to assume that rationality requires consumers to be cheap.
To some neoclassical economists, the Pareto criterion is the unchallengeable linchpin of welfare economics. Duncan Foley remembers “telling one senior professor [at MIT] that I doubted the relevance of the concept of Pareto efficiency to actual policy and his responding that if I felt that way I should get out of economics altogether” (p. 193).
Since the appearance of Murray Rothbard’s great paper “Toward a Reconstruction of Utility and Welfare Economics” in 1956, Austrian economists have been well aware of the problems of the conventional view. In particular, the standard view takes as given a given distribution of wealth and income. From this baseline, the question arises whether Pareto superior moves can be made. If not, the situation is “Pareto optimal.” By this criterion, a system of slavery can qualify as Pareto optimal; since freeing the slaves will make the owners worse off, emancipation is not Pareto superior. Rothbard stressed the need for an ethical system to solve questions of economic justice; Paretian welfare economics is inadequate to the task. It is thus encouraging to find that Paul Samuelson’s “personal credo mandates: Never think about Pareto Optimality without at the same time contemplating differential effects on different peoples’ well-being” (p. 311).
Unfortunately, Samuelson appears to have nothing better on offer. He calls for a tradeoff between equality and efficiency but spares us the details. In like fashion, Robert Frank recognizes that standard cost-benefit analysis takes for granted a given distribution of wealth. It determines which projects should be undertaken by asking whether supporters are willing to pay more than opponents. Concerning this, Frank says: “A general criticism of cost-benefit analysis is that the willingness to pay is biased in favor of the rich” (p. 129). His solution will not commend itself to Austrians. He suggests using progressive taxation to reduce inequality. Those included in this book’s “cutting edge” would gain much from a careful study of Rothbard’s The Ethics of Liberty.1
1 I have noticed the following error, surprisingly made by Kenneth Arrow: Kakutani’s fixed-point theorem is a clarification of a theorem of Brouwer, not von Neumann (p. 304).
