Interview with Vernon L. Smith
Reading broadly in science provides a way of seeing in what sense economics is a science and helping to further it.
We are honored this time to interview Professor Vernon L. Smith, the 2002 winner of the Nobel Prize (the Sveriges Riksbank Prize in Economic Sciences in Memory of Alfred Nobel).
Smith was born in 1927 in Wichita, Kansas. He received his bachelor’s degree in electrical engineering at Caltech in 1949, an M.A. in economics from the University of Kansas, and a Ph.D. in economics from Harvard University in 1955.
In early 1956, Smith started his long and exciting journey to lay the foundations for experimental economics. He has developed an array of experimental methods, setting standards for what constitutes a reliable laboratory experiment in economics. In his own experimental work, he has demonstrated the importance of alternative market institutions, e.g., how the revenue expected by a seller depends on the choice of auction method.
His work shows that formal economic modeling has little to do with how subjects in the lab or participants in real markets actually perform in real-time. When he put economic theorists in an experiment, they could not articulate an explanation of their own behavior interacting with others. Moreover, their behavior is no more or less efficacious than the typical student subject in these dispersed private information markets.
Much of the research that earned Smith the Nobel Prize in Economics was conducted at the University of Arizona between 1976 and 2001.
Smith established the Smith Institute for Political Economy and Philosophy, which seeks to “reintegrate the study of the humanities and economics in the spirit of Adam Smith.” His works can be found on his website at Chapman University. An extended autobiography is available on the Nobel Prize website.
GL: You started experimental economics in 1956 when almost no one thought about it. What motivated you to pursue “the road untaken”?
VS: Two considerations:
Edward Chamberlin had published data on his experiments showing that supply and demand had no predictive power in his classroom market, and it was an obscure paper attracting little attention.
I thought the idea of doing experiments was commendable but that Chamberlin had not provided a natural means for buyer/seller learning over time nor for collective buyer/seller participation in price formation.
So, I allowed for (a) Marshall’s production flows of units into the market over time and consumption flows off the market, and (b) introduced bid/ask multilateral bargaining into the trading process.
Experiments help students acquire a participatory experience with markets that could be related to theory
Also, I felt the teaching of this theory was deficient in not relating the theory to real on-the-ground actions that people might take in a market. Experiments might, I thought, be a means whereby students acquire a participatory experience with markets that could be related to theory.
So, I resolved to do an experiment to try and address these deficiencies. I got hooked because the results worked far too well to be credible based on what we thought we knew, which turned out to be completely wrong. Also, I realized that I had stumbled quite unintentionally on a methodology for testing my beliefs about market performance.
GL: You advise your economics students to spend their spare time reading science. Why is that? Isn’t economics a science?
VS: Reading more broadly in science and also in scientific methodology provides a way of seeing in what sense economics is a science and helping to further it.
GL: In your recent book, Economics of Markets, coauthored with Sabiou Inoua, you argue (convincingly, I’d say) that “markets succeeded where theory failed.” Does that imply that economic (neoclassical) theory failed to study real-world markets well enough?
VS: It’s more that they failed to study them at all and, in particular, failed to model price formation or price discovery in market processes.
Neoclassical economics failed to study real-world markets at all
The classical economists, beginning with Adam Smith (Book I, chapter VII of The Wealth of Nations), had a rigorous narrativized model of price formation, and Sabiou (mainly) and I mathematized it in our book.
The classical model was much deeper and more incisive than what Cournot did, who was however close to seeing demand (and perhaps supply) as a distribution function of the willingness-to-pay (wtp) values (and the distribution function of the willingness-to-accept, wta, costs). But he did not develop that concept which would have built on the classical tradition.
Walras not only failed to model price discovery but gave us a mechanism that required price mysteriously to be given, then used to model price change depending on the sign of excess demand. This diverted theorists from modeling price discovery and, we believe, created the illusion of progress, but it was more appropriately considered a regress occasioned by marginal analysis, which helped not a wit to address the fundamental task.
Walras’s work diverted economists from modeling price discovery and created the illusion of progress, but it was more appropriately considered a regress
As we note in our book, Bohm-Bawerk’s Mengerian horse market and Marshall’s Country Corn Exchange were wonderful descriptions of price discovery, but they were really elaborations on The Wealth of Nations. They were about single units or batches of a good and had nothing to do with marginal utility.
GL: You shared the Nobel Prize with Daniel Kahneman, a pioneer of behavioral economics. As you are aware, there are two schools of behavioral economics: One led by Amos Tversky and Daniel Kahneman in the US, the other by Gerd Gigerenzer and the ABC Group in Germany. Herbert Simon might form a third school, although he seems to lean towards the German school. How do you see the relationship between your work in experimental economics and these schools?
VS: Neither of these schools is about the study of people acting in experimental markets, but rather about individual choice behavior, usually shorn of any social or market context, but that also characterizes neoclassical economic theory.
My primary influence came from Sidney Siegel, but also to some extent from Ward Edwards, Paul Slovik, and Sara Lichtenstein, all of whom influenced Tversky and Kahneman. Tversky was a student of Edwards and did an important study showing that paying real money made a difference in experiments, but he and Kahneman tended to disavow and reject that idea later.
Simon’s idea that economic agents should be modeled in terms of how they saw the world and acted in it influenced me, which was also the view of Sid Siegle.
Gerd and I shared a common perspective which was also that of Reinhard Selten (who I always thought deserved to share a second Nobel prize in 2002 for contributions to experimental economics).
Reinhard Selten deserved to share a second Nobel Prize in 2002
GL: Given your critique of neoclassical theory, have you ever had the chance to meet any of its leaders (e.g., Kenneth Arrow) and discuss its shortcomings vs. the experimental approach?
VS: Yes, I knew Ken Arrow pretty well from visiting at Stanford, and reviewed his book: “Essays in the Theory of Risk Bearing.” The review article was originally published in Journal of Business, volume 47, issue 1, in 1974. That review contains my praise, and also critique, of Arrow.
GL: In your work, you argue that competitive equilibrium can be quickly achieved under very reasonable experimental conditions for consumption goods, but that this is not the case for assets where speculation may substantially distort the outcomes. Economic theory seems to pay no attention to this important difference. Why is that?
VS: It is because standard theory tends to be insensitive to close observation: Item A is purchased in preference to item B if and only if U(A) > U(B), a theory that makes no advance prediction but rather concludes only that if A is bought rather than B it must have had higher utility value.
But as in Adam Smith, value concerns either use value, or “utility” to the individual, or value in exchange. Value in exchange is price. Price, P, is a market-discovered value. For Adam Smith, every buyer has a maximum willingness to pay, wtp(x) for a unit of x. The item is purchased if and only if wtp(x) > P. Smith understood auctions. He noted, e.g., that “if two collectors equally value an antique book, the one with the greater wealth would carry it.”
This shows that he knew that the highest wtp (willingness-to-pay) bidder wins at auction and that wealth matters. He generalized this concept of wtp demand to supply and demand, where willingness to accept supply was determined by the highest profitability.
So Smith related buyer market action to something that, in principle, was observable (wtp) prior to the market. That provided a thinking process that provided a theory of price discovery which was supported in the market experiments.
All economic stability arises in consumer markets, while all instability arises in asset markets for re-tradable goods
In consumer markets, buyers attempt to buy cheap, constrained by their maximum wtp private value. Sellers try to sell dear but are limited by their minimum willingness-to-accept (wta) costs.
The problem with asset markets is that they have a value-in-use like any consumer good but also a value-in-resale. This sets up a conflict that has to get resolved before an asset market can settle into any sort of equilibrium.
All economic stability arises in consumer markets, while all instability arises in asset markets for re-tradable goods. Fortunately, about 75% of private products cannot be re-traded, causing great stability.
GL: The Nobel Prize in Economics was awarded in 2012 to Lloyd S. Shapley and Alvin Roth for their work on market design. How do you see the relationship between these two branches, market design and experimental economics?
VS: They are closely and intimately related. Indeed, my work in market design was part of my recognition in 2002, and was part of my presentation in 2001 at the Nobel Conference on Experiments in Economics, the year prior to the award.
GL: You have been involved in the deregulation of markets of energy and other markets. Did it work? Do you support deregulation in principle?
VS: Yes, but both regulation and deregulation proposals should be informed by experiments, as in natural gas and electric power systems. Deregulation worked quite effectively, worldwide, in energy, including electric power. See this paper.
In the US, deregulation was stymied by the rate of return regulation in which deregulation was quite effectively opposed by the industry — they liked it too much because it guaranteed a rate of return profit and protected them from entry. So why would they want to give it up?
In gas, to some extent, the commodity has been separated from the pipes that are regulated, a separation that is complete in the state of Georgia. In power, only the wires may need to be regulated not the energy that could be provided with suppliers bidding to supply power at the lowest cost.
Electricity deregulation at the retail level has been advanced the most only in Texas, which has its own independent grid.
GL: What three important pieces of advice do you offer to young economists?
VS: Here is my advice to young economists:
- Read widely outside economics, for the breadth of scientific method and technique.
- Read narrowly within economics for important technical details. (Every application is methodologically the same, maximizing utility or minimizing costs over the action variables of each case subject to the particular technical constraints in the case.)
- Work to develop theory and find ways of testing theory in a variety of contexts both in lab and field applications.
GL: Many thanks for your insightful and invaluable answers!
VS: You’re welcome!
