Transcript from an interview with Professor Daniel Kahneman and Vernon Smith on 12 December, 2002. Interviewers are Profesor Karl-Gustaf Löfgren and PhD Anne-Sophie Crepin.
Welcome. I'm Karl-Gustaf Löfgren, Professor of Economics at the Umeå University. Next to me is Anne-Sophie Crepin who holds a PhD in Economics from Stockholm University. Together we have the pleasure of interviewing the winners of the 2002 Bank of Sweden Prize in Economic Sciences in memory of Alfred Nobel. The winners are Professor Daniel Kahneman from Princeton University and Professor Vernon Smith from George Mason University. Professor Kahneman is a psychologist who has changed the way economists think about decision making and, in particular, decision making under uncertainty. Professor Smith has taught economists how to conduct and learn from experiments like in the sciences and I'll ask you Anne-Sophie to ask the first question. Please go ahead.
Well Professor Kahneman, what made you decide to become a scientist?
Daniel Kahneman: I wanted to be a philosopher actually and I decided, I became interested in psychology as a substitute for philosophy, as a way of answering questions about the human condition but answering them by looking at facts rather than by discussing words. So that's how I became a psychologist.
And Professor Smith, your answer too?
Vernon Smith: Actually I was interested in philosophy at one point. This is before I really was seriously into my undergraduate training, so it was just a reading interest of my own and I read Sir James Jeans' Physics and Philosophy and Sir Arthur Eddington's books and Burton Russell; I was interested in science and philosophy and originally I expected to study science and I went to Caltech for that reason and I did study physics and would have taken my degree in physics, my undergraduate degree, except for one hurdle.
To take the degree in physics you had to take Smyth's course, a famous course that was very, very hard to get through. By taking my degree in electrical engineering I didn't have to take Smyth's course but I took everything else in physics. So I took electrical engineering because I chickened out and I wanted to finish on time. But I got interested in economics as a senior; I took a course in economics and at the time I thought, well you know, this just looks like physics and little did I know how deceptive that was. But anyway that was my naïve view at the time.
Well if we turn to your research. I mean one of you has shown that people are frequently irrational in their economic decision making and the other has shown that the market mechanism works efficiently, at least in the lab. How would you reconcile these seemingly different views, I mean both of you?
Vernon Smith: Question to me?
To both of you.
Vernon Smith: Well, let me begin. I think it's fairly simple. In experimental economics we have 3 classes of results. We have situations in which people do better for themselves as individuals and as a group and is predicted by economic theory. Ok. I call that super rationality. We have a second class of results where the predictions of economic theory do very well and they conform and I'll come back to what those are shortly and thirdly we have results which people do not do as well as predicted by theory. The first is in 2 person extensive form games. We use the term personal exchange there. Too many people cooperate relative to the predictions of the economic, not everyone in single play games but up to half and so we have to come to terms with that. The experiments where the theory of markets does extremely well is where we're talking about production and consumption markets; flows. We think of consumers as daily or regularly enjoying the value from what they buy in markets and producers regularly incurring a cost to produce that. Those markets, if you study them in the lab are remarkable efficient and although their ability to converge to the equilibrium predicted by the theory varies with the institution, they basically all function quite well but the third class of phenomena where the predictions are not good is in what we call asset markets or capital markets and there's an inherent uncertainty in stock markets and we capture that in these laboratory games and those do not converge quickly and easily to what you might call rational expectations equilibrium. If you hold the environment constant for kind of 3 times back, and now we're talking about 6 hours of experiment, it gets home, ok, but that's not very inspiring or encouraging because the world out there it doesn't stand still while people look for the rational expectations equilibrium. So does that help to clarify?
Yes, I think so. Professor Kahneman?
Daniel Kahneman: Well the first comment I would make is about the word irrationality as characterising the research that we have done. I never think of myself as having demonstrated irrationality. There is a definition of rationality within the contest of economic theory or decision theory more broadly, which is a completely unrealistic conception of a human agent with a complete preference order about all states of the world, with a Bayesian set of beliefs about all possible states and this defines rationality in the context of economic theory. Now as a descriptive hypothesis this is a totally implausible hypothesis and, you know, it is fairly easy to show that that hypothesis isn't true and we've been doing that, my late colleague Amos Tversky and I, and many others. It's also not particularly interesting to show that it isn't true because it's so easy to do. We have been able to show some of the ways in which people depart from this ideal of rationality but this is not irrationality. People are reasonable, they're prudent agents. It's just that the definition of rationality that is used in economic theory is, I think, a very implausible definition and it fails descriptively and we have been able to document some of these failures and explain them.
Both of you have worked in the frontier between different fields; economics and psychology and natural sciences with experiments. What kind of difficulties or what kind of challenges did that implement for you?
Vernon Smith: Well, I think if you are curious about some of the things we observe in economics and you want to better understand this phenomena, whether it's in a laboratory or the field, I really think you have to reach outside of economics because economics, although it has an incredible body of technique it's developed and the methodology that has value it's much too, I think, restrictive to embrace the range of observations. So you have to reach outside and actually if you go back to the Scottish philosophers, David Hume and Adam Smith, they were not narrowly oriented in the way that we often think of as modern economics. Adam Smith had huge breadth. He wrote the definitive History of Astronomy, 18th century. He was probably the first great post Newtonian scientist and he wrote on other aspects of human sociality besides just what has become known as economics narrowly construed.
I guess he was a philosopher of moral philosophy wasn't he?
Vernon Smith: Well, yes and of course we didn't have a clear delineation of the fields like we have now. Well I think a lot of the work on experimental economics points us back to that period and a need to pick up on some of the inspiration that was behind people like Smith and Ferguson and Hume and others and I hope in fact my work will help to encourage that, not because they had it right, it's because we know a lot more, obviously, after 200 years but there's certain themes there that we've lost, that we've sort of abandoned and that's unfortunate.
Daniel Kahneman: Amos Tversky and I started working together on the field of judgment and decision making and we were just doing psychology. So we were not intending to talk across the disciplines but eventually, and that came as a bit of a surprise to us, the work that we did was, to some extent, influential in economics. I mean this is of course why the prize is given because of the
influence. Now what is remarkable to me about this is actually both the ease and the difficulty of communicating across disciplines. In our case I think we've had a very easy time. You know, it's not that economists have flocked to the ideas that we're bringing, you know, behavioural economics is a minority movement and not everybody is convinced of its value but by and large I would say that, you know, I have been quite surprised by, you know 20 years is a short time for ideas to have an effect and, you know, our ideas have had whatever influence they've had relatively quickly. What impresses me is how chancy this is, that is this is entirely accidental, that is the communication if we had published exactly the same paper, which is cited in the award prospect theory, if we'd published that word for word in Psychological Review, in the Journal of Psychological Theory there would have been no Nobel Prize for this work today. So it was because it happened to be published in Econometrica that this happened.
But also the most cited paper in Econometric ever.
Daniel Kahneman: Yes, but in part it is a very highly cited paper. I should add, you know, this doesn't make it all that influential in economics because most of the people who cite the paper are not economists. So I don't know how many economists cited it; it's a well known paper in economics as well but if it had been published in another discipline it would simply not have had an impact and I think this is in part what Vernon is saying, economists do not spontaneously look outside the discipline. So if you look at the journals that are cited in the economics literature they tend to be economics journals. So we were quite lucky, you know, in the sense that publishing in Econometrica and we did that because it was the prestige journal for decision theory at that time. We thought we had a good paper and so we sent it to the best journal that would publish it and we were then lucky in a completely different way in that a young and very brilliant economist, Richard Thaler, read about our work actually before it was published and was influenced by it and he really, not we, developed behavioural economics. So it was his doing and it was through his work that our work became known and he deserves a great deal of the credit for, you know, what's happened since.
Ok. Before we leave the research side of things, I mean you published your award winning papers before 50. Is there a life in research after 50 or is economics a young man's game like mathematics?
Vernon Smith: Well I think that's probably not unusual across most of the prizes although I'm just conjecturing that that's probably true. Certainly Albert Einstein had some of his basic instincts before he was of age and particularly I think that's true in physics but it's young minds that tend to get inspired. I think it also has to do with the sociology of professional work. It's young people who chart new courses that change things. It's not the existing scientific community that suddenly has a transformation and, you know, my friend Paul Samuelson points out that science progresses funeral by funeral.
Daniel Kahneman: Well, I mean to your question of whether there is life after 50, I certainly hope so. I think Vernon has done some of the work that, you know, has done wonderful work in very recent years so there may be life after 70 and, you know, to some extent this again is accidental and is self selection, that is if you are going to have people who are going to have important ideas, they may have them fairly early in their career and then they spend the rest of their careers elaborating on these ideas and so it looks as if, you know, people have their best ideas early on but that's because they spend the rest of their lives working on them and I think, you know, this happens to most people who have one important theme to develop.
Vernon Smith: Well I think young people are sort of maybe more likely to make technical or mythological breakthroughs but just looking at my own history I didn't really appreciate the full ramifications at the time I was doing that work. I didn't come close to it and I think that can come with maturity so that the contribution you can make after aged 50, and I hope after age 75, is perhaps a different type of contribution than one makes when you're younger.
Daniel Kahneman: Yes, I would echo that. It takes a long time to understand what you've been saying. So, you know, you say it first and then, over the years, you understand what it was that you really meant because you don't know that immediately. In my particular case I was fortunate because there were two of us and that process of understanding what we said worked faster because we understood each other but when you work alone, understanding what you're saying is a long drawn process.
Vernon Smith: Well and I recognise, understood early that a component of what I was doing had to do with institutions and rule systems but I didn't begin to understand it the way I now do and a lot of that understanding came from interacting with other people over the years, Charlie Plott is a prime example, and also Martin Shubik, he and I are exactly the same age. I think I've known Martin for over 40 years, I think about 45 years and we often exchanged ideas, having to do with institutions and Martin was very interested in and that interchange was very valuable to me even though we never worked together and of course I worked with Charlie Plott and a lot of our insights came working together.
What kind of advice would you give to young people considering a research career?
Daniel Kahneman: Oh, you know, there are lots of pieces of advice. I tell my graduate students that if they don't fall asleep thinking about work, you know, then they're not working hard enough actually, you know, so that's the first piece of advice that you would give them and the second one is to try not to get trapped in uninteresting problems just because you began them. So one of the important things to have in science is to avoid the sunk cost fallacy, just to keep going with something just because you began it and have made an investment. So the ability to just make a quick turn when you've had an idea that looks better and drop everything else and follow the best idea that you have at the moment, that is certainly one of the thing that I think worked for me and I think it may work for other people as well.
Vernon Smith: Yes, I would echo that. Don't follow the path of least resistance. Be prepared to break the informal rules. You have to of course live in your environment and so there's a limit to how far you can go in breaking the rules. I was interested in experimental economics long before I could really make a living at it so I did other things and I didn't get tenure doing experimental economics, I got it doing other things but I returned to experimental economics and one of the problems can be that in getting tenure you develop all these bad habits and then you can't get out of doing the bad habits, which is doing what's easy, you know, following your sunk cost and doing trivial kind of extensions of that.
Daniel Kahneman: And trying to salvage failures, that is when something is not working there are people who spend a lot of time trying to get something out of an experiment just because they did it. I'm sure that's good advice to avoid that.
Vernon Smith: I think young people in economics are well advised to read widely outside of economics too and I think fairly narrowly within economics is good enough because the theories tend to be very similar anyway. Once you sort of get that basic model it's better to… I read Science and I read Nature, those 2 magazines, they're weeklies. I can't follow everything in them but I do find things in there that intrigue me.
Daniel Kahneman: This is unusual, you know, in economics. 18 years ago when we first came to Berkeley, George Akerlof invited me to co-teach a course with him and we taught a course on economics and psychology and there are two things to be said about that course. One is, he didn't get credit for teaching it because it was considered a frivolous thing to do and the other one was that he kept advising and warning the students not to be seduced by it because he thought it could ruin their careers if they followed that path. So he would tell them, you had better stick to what he called meat and potatoes economics and, you know, you can afford to do those strange things after you get tenure.
Vernon Smith: Too late, he'd formed all his bad habits.
Well how important has it been to you? I mean your scientific result can be applied to, what did Nobel say, to the benefit of mankind? I mean you followed your track, you found something that was interesting and pursued that?
Daniel Kahneman: I would say that the conscious sense of doing something that could be truly beneficial, well I had that early on when I thought that people could be educated to think, you know, more closely and these efforts of mine have not been rewarded. In recent years I'm consciously trying to do something for the benefit of mankind and this is to develop new and better measures of human welfare and human wellbeing that could be applied as another measure of how society is doing and I'm doing that with collaborators, including an economist at Princeton, Alan Krueger, and that is truly with the idea of trying to do something that could be useful to policy making.
You have something to say Vernon?
Vernon Smith: Well I've become more interested, particularly in the last 20 years roughly as against the first 20 years, in utilising what we've learned about markets to do a better job of helping to design markets in new areas where people don't have any experience, any field experience and I think the laboratory is a very useful tool for allowing people to get experience and it doesn't provide the final experience and the final answers but the point is it is experiential and it gives people an opportunity to try out and test bed new rule systems, practitioners for example, and we found business and government, in some situations, very receptive to that. In governments, particularly in New Zealand and Australia, with respect to the liberalisation of electric power and I think we've seen in the California fiasco how bad things can be if you don't think about some fundamental issues and furthermore, in that case, a lot those issues had long been studied in the laboratory and in fact influenced New Zealand and Australia but we didn't have an opportunity to have that much influence in California. It isn't that we didn't communicate with any of the people that might have made a difference, it's just that we couldn't convince them or influence them enough. And also it turned out to be far worse than even we would have imagined as to the consequences of really not getting some elementary features of these markets right.
Daniel Kahneman: They need some help.
Daniel Kahneman: Many years ago when we were studying failures of rationality in judgement and decision making I thought, you know, that there was a contribution to be made, for example to government decision making or to making political decisions but 25-30 years ago that was a period when the discipline of decision analysis looked extremely helpful and hopeful and it seemed as if the combination of psychology and decision analysis could be very, very useful. It has not been. I mean by and large I think this has been a failure and the reasons are quite interesting. The reasons are that the leaders do not want the help, that is the people who make decisions, important decisions, by and large do not want the kind of help that decision analysis or decision aids have to offer that we would think, you know, would certainly improve the quality of decision making. So there is a great deal of resistance and that, by itself, is quite interesting, that it's been 30 or 40 years, you know, since decision analysis was first proposed and, by and large, very little has happened.
Well, after this interesting word I would like to thank you on behalf of the Nobel Foundation for giving this interview. Thank you very much both of you.
Daniel Kahneman: Thank you very much.
Vernon Smith: It's a pleasure to be here.
Did you find any typos in this text? We would appreciate your assistance in identifying any errors and to let us know. Thank you for taking the time to report the errors by sending us an e-mail.