Four questions for Vernon Smith, experimental economist
Nobel Laureate Professor Vernon Smith is often referred to as one of the founding fathers of experimental economics, which uses scientific experiments to test economic questions.
In 2002, he shared the Nobel Prize in economic sciences with Daniel Kahneman, which recognised his ground-breaking work in developing methods that helped carry out real-life experiments in economics, notably his theory of induced valuation.
Professor Smith, who has co-authored more than 200 articles on capital theory, finance and natural resource economics, also pioneered his own concept of how ecological rationality – when rational decision-making matches the structure of the information in the environment it is being made - relates to financial markets.
He will be in conversation with renowned political philosopher Gerald Gaus, James E. Rogers Professor of Philosophy at the University of Arizona, in a special public event hosted by Monash University's Philosophy Department in collaboration with Monash Business School, “Learning the Most from Proving Yourself Wrong: Three Examples".
He answers some questions from Monash Business School’s Department of Economics.
1. What are some useful insights from experimental economics for people with no formal economics training?
A: "Here are three of the most prominent economic insights; none require formal economic training—parts of which is what went wrong in each case. But that is not an indictment of formal economic training, which I heartily endorse.
We have been using experimental economics in summer high school student workshops for a couple of decades. People need no formal training to learn supply and demand by first experiencing such a market with strictly private dispersed information. These are markets for non-durables, bought for consumption and are not re-traded (and they are large, 70-75% of private product). They illustrate efficient outcomes that are ecologically rational.
Workshop students also learn from experience how unstable a market can be when we put them into an asset trading market (re-trade-able durable good). And how easy it is to get a bubble with complete common information on true fundamental value, which they ignore until they acquire experience. What is missing in asset markets (also in new product consumer fads) is enough independence to dampen self-fulfilling and unsustainable expectations of rising (or crashing) prices.
In two-person trust games they learn that a surprising number of people follow ecologically rational rules of engagement in personal (even anonymous) interactions, that enable them to make more money on average than if they followed the traditional 'rational Max U' model applied to that game. People do not have to receive formal economic training to have learned that kindness begets kindness. This is a principle in Adam Smith’s The Theory of Moral Sentiments (1759), a part of classical economics lost to modern economic, but recently discovered.
None of these three results from experiments were anticipated by formal economic analysis, but they changed how we economists think, and even some of the models."
2. What are two topics within experimental economics (apart from field experiments) that are likely to lead to Nobel prizes?
A: "The Nobel Committee is much too independent to make forecasting easy or obvious. When Elinor Ostrom was announced the internet was alive with “Elinor who?”
Themes in behavioral economics were recognized in Danny Kahneman’s work, and in field experiments in my work by the Nobel committee in 2002.
Mine took the form of using the lab to test bed and prepare for implementing new markets in the field. One of the most important was in Australia that ended in the creation of a national electricity market in the 1990s. That would never have happened if Australians had not been open to new approaches. See our paper. Further such recognition seems likely in behavioral and field experimental economics in addition to Al Roth and Elinor Ostrom.
If I were to name only one, it would be in Information Markets, prominently pioneered at the University of Iowa, beginning with the Iowa Presidential Stock Market in 1988. This initiative has spread throughout the democratic election world, but has also significantly impacted business practice and forecasting."
3. Experimental economics has grown substantially in the past decade. What are some of the most exciting developments in this period?
A: "So many that I am not up to speed in most of them."
4. How does the ecological rationality notion handle a spontaneous order with sub-optimal welfare?
You are known for your concept of 'ecological rationality', have criticised the standard global rationality notion, and clearly cherish the idea that a spontaneous order arises from the limited actions of agents making decisions at their local level.
But in Phishing for Phools, Professors George Akerlof and Robert Shiller argue that firms manipulate consumers into buying goods that do not enhance the consumers’ wellbeing. Firms simply take advantage of behavioural biases (one example is selling more cars by stressing lower monthly payment rather than the life of the loan).
Consequently, we have 'sub-optimal' welfare, very similar to other cases of market failure. But in this case, the market failure arises from limited rationality, very much as the rationality of the agents in your concept of ecological rationality. So, how does the ecological rationality notion handle a spontaneous order with sub-optimal welfare?
A: "Spontaneous orders muddle through many twists, turns and failures. Free people make mistakes, including great scientists who made lasting contributions. Looking for flaws has long been a human preoccupation, beginning with Eden. But economic historians tell us that despite all these documentable foibles, per capita income and wealth started to take off in the 18th century in Northern Europe and has spread west like a tide in reducing poverty, mortality and poor health in spite of the huge growth in population. No one argues that it was due to intelligent planning, or fully informed agents.
Finding flaws is much easier than finding policies that increase realized welfare. Single dimensional policies that improve welfare is a risky enterprise of unintended consequences. Alcohol is clearly a consumer product that can cause great harm if not rationally used. Hence, alcohol prohibition in the US seemed like a good policy, and was so popular that we got a constitutional amendment in 1920. The experiment lasted 13 years, but the consequences were so unpopular that it was repealed in another amendment. We need, but do not have, better ways of trying it before flying it.
Science now tells us that we have deficiencies in health due to eating sugar and carbs, not meat and animal fat as earlier science and government agencies falsely claimed. There is evidence to suggest that the sugar interests influenced science policy: see this JAMA article. But it does not follow that prohibiting sugar and carbs will not do more harm than good.
A policy that attempts a general solution to the behavioural bias imperfection is “Truth in Labelling.” But that means you have to know what the truth is, and considering that in the history of science all truth is later supplanted by a greater truth, you have to accept sub optimality; it might even be ecologically rational."
How can we learn from failed economic experiments?
Professor Vernon Smith will sit down for a conversation with renowned political philosopher Gerald Gaus, in a special public event hosted by Monash University's Philosophy Department and Monash Business School on Wednesday 5 October.