Interview transcript

On behalf of the Nobel e-Museum, I have the pleasure of welcoming this year’s winners of the Bank of Sweden Prize in Economic Sciences in Memory of Alfred Nobel. The winners are Professor George Akerlof from the University of California Berkeley; Michael Spence, Stanford University; and Professor Joseph Stiglitz from Columbia University. They were awarded the prize for their analysis of markets with asymmetric information. And I am Karl-Gustaf Löfgren, a Professor of Economics at Umeå University and sitting beside me is Anne-Sophie Crepin who is a grad student from Stockholm University. And I’ll ask Anne-Sophie to ask the first question.

Yes, and the first question would be how did you decide to study economics? And maybe we could start with George?

George A. Akerlof: How decide to study economics? Well I think I always wanted to be an economist, if there was such a thing. I think I didn’t know that there were really economists until I went to college. And I knew that there was such a thing. I was always interested in economics, for a very long time. And I remember at the age of 10, I asked the following question: If one person loses their job and then they stop spending, and I was a little boy, so I decided, one father, remember this was the old days, so if one father lost his job, wouldn’t that cause that family to stop spending their money and that would cause another family, another father to lose his job and that would cause that family to stop spending.

And so I was worried that the economy would have a very bleak reaction to this. And it wasn’t until I went and took freshman economics that I learned the answer to that, which is that the family would only stop spending, let’s say, three quarter of it’s money, it would save a quarter, and so the multiplier wouldn’t be that great. So I think I had a reasonable number of such questions long before I knew that economics existed.

What about you Dr Spence?

A. Michael Spence: I was not thinking about infinite multipliers when I was 10. But I did have a father who was a PhD in commerce and finance and an intellectual man. And so I had a feeling, probably about the time I went to college, that I would try to be a scholar and teacher, but I didn’t know which field. And I picked economics at the end of my undergraduate time because it seemed to be a really nice combination of theory, including mathematical theory on one hand, and things that are quite practical that you can touch and see and feel. So I picked it and I consciously thought of it as an experiment to see if I liked it. And it worked.

Finally, Dr Stiglitz?

I love mathematics, but I decided I really wanted to work on problems of society …

Joseph E. Stiglitz: I had always been interested in economics and social problems when I had been young, maybe a little older than George, in high school. But when I went to Amherst College I studied physics and math. And then towards the end of my third year, my junior year, I decided that I was more interested in working on social problems, problems with society and using the mathematics I’d learned and combining that with my interest in history and society, to work on economic problems. And so that was really the decision. I love mathematics, but I decided I really wanted to work on problems of society.

George A. Akerlof: So I think in addition, I felt that the one thing that you could do to make people better off and able to lead self-fulfilling lives was if people have more money then they’ll have fewer constraints on their lives, and so they can make more of themselves and lead happier lives. So that was actually another reason for studying economics.

What about the asymmetric information? I always wondered why things happened in the ‘60’s and early ‘70’s, why didn’t it happen earlier or why did it happen at this time?

George A. Akerlof: I can speak for myself, I think as far as I’m concerned it was an outgrowth of the work on quality in growth theory. That in growth theory, especially Joe’s and my thesis advisor, Robert Solow, he worried and made models of different qualities of capital. So vintage capital and capital in which one could choose the capital labour before the capital was built, but not afterwards. And so that made it possible, I think, because he had figured out how to model different qualities that made it possible to make models with asymmetric information, in which the key variable was the quality of the goods. Prior to that time I think we didn’t have enough mathematical ability to deal with what happened, not only when price varied, but when quality varied, especially in some continuous way.

Joseph E. Stiglitz: I think it was in a slightly different origin in my case. I was persuaded that models that we were being taught didn’t make a lot of sense for describing lots of the problems, and that there were a number of key problems, lack of perfect competition, lack of perfect information. Work had been done on the consequences of lack of perfect competition, theories of imperfect competition like Joan Robinson and Chamberlain. And the next natural question was what to do about imperfections of information? I think the key thing was that there were some very specific questions that at least I began to address, posed when I went to Kenya about how much should they be investing in education, that lead to what is the role of education? Was education just human capital, which had been the older view? Or whether other issues of education, like credentials or providing information. I think that was critical in the development.

There had been people working on generic equilibrium models with imperfect information. In other words, bringing in information into very abstract models. But they did it in such an abstract way that the questions weren’t posed in ways that lead to interesting answers. Beginning on the other side, what was a very specific issue and you started thinking about just a simple thing, assuming that two abilities, one low ability and a high ability, how do you sort them out? How do these people who are more able convince others that they’re more able? And by taking the simplest possible problem of information and thinking about how you solve that and then building up from that, rather than the abstract and try and deduce it. I think that was the critical breakthrough and I think it’s what all three of us had in common in our work, beginning with a very concrete problem and then generalising it.

A. Michael Spence: That’s very accurate. I think in addition there may have been some very interesting work in game theory, or at the application of game theory. Once again, as Joe said, not highly mathematical, most general game theory. But a game theory that was used to deal with deterrents problems and what not. And so every time you turn around there was a question about information, where it resided, how it was communicated, who knew what and when? And very bright people like Tom Schelling, for example, had started to write about it. When I started doing this, I’ll just add this one thing, there wasn’t anything there.

There was some writing by Bill Vickrey, and he was the one who I think firstly pointed out that information is a very unusual commodity and that when Joe gives it to me, when he possesses it we both have it. And I got quite intrigued. But I was sent up a whole lot of blind alleys. There are people who sent me off to study signal processing theory, because that was what they called information theory. I learned absolutely nothing useful except what the capacity of a fibre optic cable is. So there was all that going on at the same time.

Joseph E. Stiglitz: One of the key things that came out of that earlier work is the importance of how to make inferences. It wasn’t the mathematics of how you make inferences, but the conceptual process of what are the signals, what are the things from which we make inferences about what somebody else is or what the world is like? And I think that, not at the mathematical level, but at the conceptual level of this process of making inferences is really very important I think and is a common element in all three of our work.

… the reaction from the editors, and possibly also from the referees, was this was not economics and therefore should not be accepted …

George A. Akerlof: I think one interesting aspect of this is when this work was initially done it wasn’t considered to be economics. So I submitted the market for lemons to three separate journals before it was finally accepted at a fourth. And the reaction from the editors, and possibly also from the referees, was this was not economics and therefore should not be accepted. And I think that’s because in fact its methodology was different, that this was a different way of looking at price theory.

It took some time before people saw that you could do price theory this way. And I think what was so different was that instead of arguing from the top down, from taking some general principles about how markets work and pricing systems work, instead we argued from the bottom up. So we took a look at examples of such things as insurance markets and education markets, and credit markets, and market for used cars. And then argued from the way we thought those specific examples work to how markets should look.

Joseph E. Stiglitz: We looked for general principles.

George A. Akerlof: From the particular.

Joseph E. Stiglitz: And it turned out that these general principles applied very broadly, but looking at ways in which they applied differently in different markets also gave you a lot of insight into the general principles.

A. Michael Spence: And the editors that rejected George’s paper have since been fired.

But this is probably not very unusual. When Black and Scholes came up with their formula, I mean there were some problems there to publish it too, if I remember it correctly.

What are you doing now? What kind of research are you doing now?

Joseph E. Stiglitz: One strand of research is a continuation of the problems of economics of information. It has gone into areas such as macro-economics, organisation theory, the insights of information economics has lead to theories of corporate finance, how the firms finance themselves, that has lead it to theories of firm behaviour, the theory of the risk averse firm. That has lead in turn to macro-economic theories of how the aggregate behaviour of the economy behaves. And that in turn has lead to, as one example, monetary theories, money rather than just being, monetary theory used to focus on transactions, the role of money. But in fact most transactions today use credit. And what is credit? It’s ascertaining who is credit worthy, which is an information issue. So it’s really reformulating monetary economics on the basis of theories of asymmetric information.

There’s another line of research that I’ve been very heavily engaged in which grows out of my work at the World Bank, and that is issues of development, issues of strategies for economies and transition from communism to a market economy. Both of those have a usage of the ideas that have come out, but broaden other ideas as well. Just to give you one example, a key aspect of the doctrines in development economics was a set of ideas called the Washington consensus, which was based on the belief in market fundamentalism. That markets by themselves lead to efficient outcomes. And that’s based on a belief of markets with perfect information, that set of ideas doesn’t work very well in developed economies, but in less developed economies it’s absolutely abysmal theory. And trying to think through how markets in developing countries are affected by the lack of information, as an example, and how that affects development strategies, is one of the key issues with which I’ve been concerned.

… I think the internet actually has moved the parameters, informational structure parameters and the number of markets …

A. Michael Spence: I had this somewhat unusual career and I stopped in mid 1987, became an academic administrator so that I, in the language of venture capital, this is kind of a restart. But I think the thing that I’m going to focus on when I know that I have the time again to do research is I think the internet actually has moved the parameters, informational structure parameters and the number of markets, in ways that, and I’m not sure of this, may require us to really look at the models. So I don’t think that means throwing the whole lot that we have out, but I think it probably does mean looking again. For example, and just to take one that was mentioned briefly this afternoon, if most people post prices and they’re accessible on the internet, the search cost that George Stigler did in some interesting early work, have simply disappeared.

So the naive conclusion I guess is that, you know, there isn’t any other place to hide based on that kind of search cost. But probably the correct answer is that some smart young economists, or maybe even some older ones like us, ought to take a look at the question of whether as a matter of strategy people are going to post prices anymore. And if not, what is actually going to happen in the market? That kind of thing.

Joseph E. Stiglitz: Or we’ll come back to the kinds of issues the quality, you can’t post quality.

A. Michael Spence: No they can’t post quality.

Joseph E. Stiglitz: And that is at the heart of a lot of what our work was concerned with. And that will never be well described on the internet, or perfectly described on the internet.

George A. Akerlof: I think beyond working on asymmetric information, what I’ve sort of been doing and making a career of is thinking about what assumptions or not in economics, which would make a difference. So I think the next thing I did is I worked on what happens when prices are not made at the same time, not set at the same time, and wages are not set at the same time. That’s called staggered wages and staggered prices. Then I worked on what happened when there was some band in which people were indifferent in holding their money, so you didn’t decide to do anything about your money until you had more than some threshold level or less than some threshold, so that’s called target threshold. Then I’ve worked on bringing in anthropology and sociology into economics. Which is again the same thing, it’s sort of seeing what assumptions could be in economics, that should be there but aren’t.

So I’ve worried about what happens when there’s reciprocity, especially in the employment relationship. What’s happening when fairness is an issue? And I’m currently working on the importance, especially to the labour markets and to education, of peoples’ self-concept. I think that probably the most important decision that anyone makes in their life is who they think they are and who they want to be. And economists tend to think of price as being the most important decision that they make. I think working on the asymmetric information said that another aspect of markets was quality.

But then I think there’s this third aspect that hasn’t yet been incorporated, which is who we think we are and who we want to be. And I think that this is the root cause of poverty in developed countries. That if people think that they can do something, they think they can be anything, then in fact they will and so there will be much more social mobility than they have. And I think this is the leading cause of poverty in the United States, that there are people who just don’t see the potential and just don’t have the right self-concept.

So that’s your recent work on identity?

George A. Akerlof: That’s my recent work on identity, yes.

To what extent have you been involved in implementing your research results?

Joseph E. Stiglitz: I spent basically from 1993 through 2000, first as President Clinton’s chief economic adviser, then as chief economist of the World Bank. And in those jobs I had to deal with a wide range of issues, not just the ones related to my own work. But an anecdote may illustrate, when I first came into the White House, I went to a meeting in which a big issue on the agenda was health insurance, reforming the health care system in the United States. And I remember going to a meeting and just sitting in the back of the room while they were discussing, and they were talking about moral hazard and average selection as the key ideas, the key problems that had to be resolved in reforming the health insurance market. And it gave me a lot of pleasure to see how quickly some of the ideas that we had developed had gone from theoretical research into being taught at graduate schools and by this point, were just common tools that everybody, you wouldn’t begin that discussion of insurance reform without the concept of moral hazard and average selection.

So in a way they’ve become a tool kit, part of the vocabulary that everybody uses. In the East Asia crisis in the World Bank and IMF, the issue of whether the bail outs were going to cause moral hazard, ie leave the lenders to take less precaution in making good loans was a standard part of the debate. And I thought that they mismanaged that in a number of ways, but partly because the IMF had not really fully integrated some of these ideas, that for instance, with imperfect information you begin to think about bankruptcy. And you start thinking about the impact of monetary policy on the likelihood the firms are going to go into bankruptcy.  You worry … the financial system is destroying information and the flow of credit. And so the intellectual frame that I brought to the issue, some of it had been incorporated, some of it had not fully been incorporated, and some of that represented some of the reasons that there were a lot of disputes about the appropriate ways to respond to the crises.

… you have to begin with the legal structure, don’t privatise too rapidly before you do that …

One more example, the work on asymmetry of information brought home the importance of corporate governance. That managers have much more knowledge and discretion about what to do with the resources under their control in the firm. And that they can use that discretion for their own benefit and not for the benefit of the shareholders. In the United States and in Sweden and in most advance industrial countries, we’re aware of that problem and have passed legal structures to prevent the abuse. In Russia they didn’t have that legal structure. Without that legal structure, privatisation lead more to asset stripping than wealth creation. The strategy for moving from communism to a market economy that some of us were very sensitive about these issues, said you have to begin with the legal structure, don’t privatise too rapidly before you do that. And that was again something that was ignored by the people who believed in shock therapy who didn’t understand the subtleties of a market economy.  And I think that has contributed to a large extent to the failures that have occurred there over the last decade.

A. Michael Spence: The only piece that I would add is there were two ways that you could go other than jumping into policy, you know, once these theoretical ideas emerged. One was to go and see how people try to use them or might try to use them when they are actually doing things in the world, like business people making strategy decisions.

And you are a coach aren’t you?

A. Michael Spence: I guess a little bit, but mostly I was the dean of a business school. So you’ve got to watch how this got translated. And the other point I guess I would make is that there’s an empirical component to this, you know, there are theoretical structures and there are the observations that all three of us made just looking at how markets behaved that caused us to try to create the models and theories. But there’s a serious empirical side to this as well where you actually go take the theories and then go try to verify. I started out down that route and ended up taking the practical side. The thing that I was never interested in doing was, and I think Joe and I are different in this respect, I was always interested in the sort of economic science of it and was happy to have others really do the policy part. This is entirely personal, it wasn’t my driving motivation.

George A. Akerlof: I think I’ve done some policy work. I’ve worked at Brookings Institution for some time. And I think that probably in the last decade or so, especially jointly with Janet Yellen, I’ve developed a style in which institutions are very important. And one should pay a great deal of attention to the detail of how institutions look and then specifically analyse special historical cases describing the institutions in detail. So Janet Yellen and Andy Rose and Helga Hessenius and I wrote a paper on East Germany during the transition. We looked at details of the institution. We did a similar paper explaining the reason for out of wedlock births, in which we looked in great detail at the history of legislation regarding abortion and also the availability of contraception.

… when you understand the details the markets are actually much more interesting than you would otherwise think …

And so I’ve done a reasonable number of policy pieces in which we are very careful about getting the institutions right. And I think that’s in the same spirit as to the original asymmetric information. We thought that the details mattered as to how the market was going to work itself out. So issues that in used cars, in insurance and so forth. So it’s a matter of being a little bit more careful than the economics was prior to this work. You have to really understand the details and when you understand the details the markets are actually much more interesting than you would otherwise think.

Joseph E. Stiglitz: I just want to add, one of the aspects of traditional economics, this is called neoclassical economics, was that you didn’t need to look at the details. The theory was that demand and supply determined everything. It was really conceived as institution free economics. So it was not only that they didn’t look at the details the theory said, they didn’t have to. And in a sense we began for the premise that that was wrong. And as you began you saw that they did matter and in a very concrete way.

George A. Akerlof: I think that’s actually why originally when I submitted market for lemons for publication, it was said this wasn’t economics. They said if this is economics, we don’t do it. That is what one of the referee’s reports said. If we accept this for publication, what are economists going to do?

George A. Akerlof: I just want to make a plug, because there’s a wonderful book out which is codified all of this, and that of course is Joe’s wonderful text book, his elementary book on economics. And I was in China last summer and it had sold over 1 million copies in China. And it’s a great book; I actually refer to it whenever I want to look at some detailed question.

I think our time is up so on behalf of the Nobel e-Museum, I would like to thank you for participating in this interview. Thank you very much.

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.