Q & A

What are you thinking?

Robert J. Shiller — October 2009

Decades of economic research have assumed people pursue their goals in a rational manner, discounting the effects of emotion, bias, error, and other irrational forces. Robert Shiller argues that economists need to take a closer look at how people make decisions.

Q: How important is it to understand what people are thinking and feeling when you are trying to understand the economy as a whole?
That's been a controversial question in economics for a long time. Milton Friedman wrote a collection of essays in 1953 called Essays in Positive Economics, in which he argued that you shouldn't try to infer what people are thinking because people really can't tell you what they're thinking. If you ask people why they did something, they will give you a conventional answer or mislead you. The idea was that the essence of economics is to look at the constraints that people have and assume that people are behaving rationally, subject to those constraints, and interpret economic data as reflecting that rational behavior. That is the defining characteristic of economics as a discipline — as opposed to psychology as a discipline — that, in understanding something as massive as the economy, it's best to look at people's actions, not their ostensible reasons. There is some appeal to that. I just wish it were more right.

I can get enthusiastic talking about this theory because, in some respects, it is good. To give an example, suppose you are trying to understand the seasonality of food prices — why they go up in the winter and down in the summer. Well, it's pretty obvious that it has something to do with the weather as a constraint, but you better think it through, because we live in a global economy, and when it's winter up here, it's summer down south. Obviously they'll ship food from one hemisphere to another. That puts a limit on seasonality. This is pure economics, and I'm sure it's right, because the seasons occur year after year after year, and you have people whose job is to ship fruits and vegetables and food around. They're going to find the best pattern of shipping, given all the costs. It wouldn't make a lot of sense to ignore that. Thinking that people get emotional in the summer, or something like that, would probably be wrong.

The problem is that positive economics has been carried too far. Part of the reason is an institutional specialization problem in academia. Academics have to specialize in some method of research; you can't know it all. This positive economics took over, and I think there's an institutional reason why it got overemphasized.

Fortunately, there's a lot of enthusiasm among some people for behavioral economics now. It's obvious that there are certain phenomena that don't fit into Friedman's prescription of positive economics, and those include speculative bubbles, for example. They're kind of hard to pin down, but they really are there.

Q: Some economists say that the positive approach gets you 99% of the way there and behavioral is about that last 1%.
I've heard that before.

Q: Where do you put the number?
I think that the failure to predict this financial crisis had something to do with failing to understand behavioral economics. So I don't think that it's 99% there. I think it's kind of the other way around. Positive economics is great for explaining the seasonality of fruit and vegetable prices, but if you want to know why we go through financial crises, I think you need behavioral economics.

Q: The disciplinary problem you describe in economics seems to have a very direct effect on economic activities outside of academia.
Right. This world now has over six billion people. We couldn't have sustained this number of people 100 or 200 years ago. In order to sustain them, we need modern, big business activities going, and the framework for that is the financial framework. We have to recognize that billions of lives depend on this. Finance, and economics more broadly, is a technology that has brought us to a very different world, and it's a necessary technology now.

Q: You mentioned speculative bubbles as an example where you might need a behavioral explanation. Can you explain your thinking behind that?
I think that there was a huge error in overstressing efficient markets. The idea developed that the stock market was smarter than any individual, because it pools the wisdom of all individuals. And so it becomes an oracle. In ancient times, when a snake was found on the steps of the Roman Senate, or something like that, it would be considered a hugely important omen and all the wise men would be asked to interpret it. Well, we're doing the same thing with the stock market. There is this theory that when it moves up, that that's revealing some information. Even now, for example, the general interpretation of the recovery that we've had is that the markets saw some good news coming and that the stock market, after March 2009, started to react to good information. Later information confirmed that we were coming out of this crisis. But I don't think that that's the right interpretation. This is what I think, though I can't prove it: There was the beginning of a turnaround in confidence and the market going up stimulated people to be more optimistic, and then there is a feedback loop. The market is not really responding to any information. I've argued in books I've written that these are social epidemics. Ideas have a contagion, just like a disease has a contagion, and a certain idea may spread only because, for some cultural reason, the contagion rate is high.

Q: How do you go about making an idea like that into something that's testable?
It seems to me that we have to read the whole array of psychological research and sociological research and try to put it together into a view of the world. And conventional econometrics plays a role. Another direction is neuroscience, which has started to change our ways of thinking.

The problem with science, any science, is that it's hard to move linearly forward. You have some breakthrough, and you explore that to the end. And then you feel like you're wasting your time and you have to go back and reflect about something different. And it may be very different. It may be something that's really off the track that you were on before. And then we gradually grow into an understanding. I think that mathematical models play a role in this. But they're not the only thing; they're just part of the story. I'm thinking of great breakthroughs in science: Darwin's theory of evolution — there's very little mathematics in it, but it's a very powerful theory. I think behavioral economics is a bit like that.

Economics is describing a very complex phenomenon. You can't reduce it to a simple formula. We have to understand that what concerns us about the economy is a bit psychological and sociological. That's just reality.

Q: I imagine that simple models or simple explanations are well received, though.
It's also a challenge that academia rewards teaching. It's good that professors want to be successful at teaching, of course. But sometimes that leads them to oversimplify. You don't want to say “on the other hand” too many times in your lecture, because then it starts to feel unsatisfying. So there is a tendency to want to have some core theory or framework that students will memorize. And it has to be iconic. That may be the problem with behavioral economics: there is no simple paradigm. There's no supply and demand curve or IS-LM model that is considered the deep insight of the theory. In my new book, Animal Spirits, written with George Akerlof, we talk about Keynes's theory and why it was such a revolution. And we think it was partly because it was early behavioral economics. It recognized some realities that weren't modeled.

Q: I wanted to get to some of the specific points in your book. Could you start with why you chose the term “animal spirits”?
Well, Akerlof and I thought that the most essential thing that we were saying was that traditional macroeconomics leaves out something very fundamental. It leaves out what's driving people's changing patterns of thinking. Economists like to think that people are rational and that means they never change their mind about anything. That doesn't accord with the reality that I know.

So we thought of our book as reflecting a different approach to economics. “Animal spirits” represents this, at least for people who know what it means. The term goes back to ancient Rome and the physician Galen, who talked about spiritus animalis, meaning the animating spirit that drives people. And it's used in one of the most important passages in Keynes's works. In that passage, in the General Theory of Employment, Interest, and Money in 1936, he stresses the fundamental lack of knowledge about the future — you just don't know what's coming. If people were completely rational, they might be paralyzed by the uncertainty. But people have a natural urge to action. That's what we mean by animal spirits. They're going to do something. Maybe it's dangerous, but you're not going to stay home and just stay in bed all day. Or maybe you would, and that's an economic depression. A depression is a period when the animal spirits are beaten down. I remember reading an article from around 1933 by a psychiatrist and he said that what struck him about the Great Depression was that people were dejected. He said that they would go home at night — or maybe they'd be home all day because they couldn't find a job — and just listen to the radio. He said a slave mentality had returned. That's an extreme form, but more moderate variations in our animal spirits are driving the economy.

We thought that that phenomenon really deserves the central place in macro theory. And it's not even mentioned. How can you have a theory that doesn't mention the driving force that makes everything move? It's like astronomy without gravity.

Q: You see confidence as a major part of animal spirits. We hear the word all the time, but is it well understood or measured now?
Keynes talked about it in the 1930s and lamented that nobody studied it. The first people that studied it, as far as I know, were at the University of Michigan more than 50 years ago, when they started the Consumer Sentiment Index. But academic economists have generally ignored it. There's only been a little bit of research on it, and I think that it has been encouraging. It's been shown that consumer confidence at an individual level drives human actions, that some people are very worried about the future and then they tend to behave differently. I think there's enough evidence that it's important.

We have been producing the Stock Market Confidence Indexes here at SOM. But the problem with collecting data as a research agenda, by the way, is that you can do it for 20 years without anyone paying much attention, unless you have enough data that it starts to correlate with historic events. It isn't a very rewarding career strategy to start collecting data and wait 20 years.

Q: How much do those confidence surveys tell you?
They give you a snapshot of what people are thinking. It's fundamental to my understanding of what has been going on in the stock market and the housing market, but I don't try to publish this in scholarly journals. They just don't respect it. They want to be scientists. That's good, but in practice it often seems to mean to them that they will not try to figure out what motivates people. I think that leaves out a whole world of things that are obvious, and have to be studied to be understood.

For example, Akerlof and I talk about the changes in corruption and bad-faith behavior. The capitalist economy is forever vulnerable to declines in business ethics. We live in a society with rules, and enforcement of rules that attempts to bring people back when they deviate from what we call ethical behavior. But we rely on their good will. That's part of animal spirits, that people have a sense of good will and tolerance toward others. It fluctuates with time. That's what matters. Look at this financial crisis that we got into — a big part of it was the emergence of the subprime lending business, which had somewhat lower standards of business ethics. People didn't appreciate that that was happening. The mortgage-broker industry was evolving and it wasn't regulated, and so we started to see some processes set in motion that would eventually lead to disappointment. And now we're there. And you won't find this in a macroeconometric model.

Q: The idea of fairness is something else you say has been left out. How has that happened?
I did a survey comparing economists with the general public. I asked the same questions of professional economists and the general public, and I found a totally different world view in terms of fairness. Economists don't rank fairness as important for their thinking about the economy.

One of the questions asked how “the effects of general inflation on wages or salary relates to your own experience and your own job.” And Choice 1 was, “My employer will see no reason to raise my pay.” Choice 2 was, “Com-petition among employers will cause my pay to be bid up.” And Choice 3 was, “A sense of fairness and proper behavior will cause my employer to raise my pay too.” We found that 60% of the economists tended to pick 2, and only 11% of the general public picked that. They thought about their relationship with their employer. Either he was a good person who would raise their wage, or he was just an evil person. That's a totally different worldview.

Keynes talked about how important fairness is. In particular, it is hard to cut wages in an economic downturn, partly because somebody has to cut them first. If I get a wage cut and other people don't, then I'm very upset. So there's a coordination problem. If we could all agree at once to cut wages, it would be easier, but there doesn't seem to be any such agreement. So you end up laying people off.

In times like the Great Depression, when prices were falling, this really inhibited the labor market, and that was part of the reason for the massive unemployment then. In the Depression, you can find references to the obvious fact that if employers cut wages, workers could still buy as much, because everything was getting cheaper. But that feeling was limited to an enlightened few, and many others were mistrustful and cynical, and assumed that their employers were trying to take advantage of them.

Truman Bewley, a professor here in economics, wrote a book titled Why Wages Don't Fall During a Recession. He interviewed people who set wages and salaries in a company, and it really confirmed Keynes's story: issues of fairness and morale dominated. You don't want to do something that disrupts the morale of your labor force.

Q: Could you explain how stories influence animal spirits?
In Animal Spirits, we emphasize the work of Robert Abelson, who wrote with Roger Schank a series of papers about narrative-based thinking. The idea that they pushed forward was that the human brain has certain built-in organizational patterns, and one of them is to organize memories around stories. We build our lives into a story. And I think that each story evolves through time, and it ultimately changes our economic behavior — it has to.

In the boom period of the 1990s, for example, a different story about our lives developed. It was the internet boom. Why did the boom end up affec-ting all of the economy, beyond its direct application in technology? I think it had a big psychological impact because it changed people's assumptions about what kind of life they were leading, and it made people worry about being left out of the story of the time. The boom was associated with stories about entrepreneurial success, and it pushed aside stories of labor solidarity and hard work and accumulation. We've had many scientific advances in history, but they all have a qualitatively different story. I think that the internet and the associated communications revolution was fundamental because it was intrusive in our lives in a new way. Compare it with other major technological advances, like nuclear power. Right now, the state of Connecticut gets most of its electricity from nuclear power. But that really doesn't have much psychological impact. In fact, most people don't even know that. The internet and communications revolution led to different stories because it changed what we were doing for a good fraction of our time. More people got computers and spent time searching for something on the web. You'd see these new businesses that were populated by young people who became millionaires. It made the old story seem worn and outdated. I think it encouraged the development of interest in finance and business, and encouraged the development of business media outlets, and it ultimately led to the real estate boom as well.

It is not entirely logical. It's all about a story. They make a movie about Superman, and then millions want Superman memorabilia around for their kids. There's nothing logical about that. It follows the dictates from the story.

Q: You're saying that the logic of the story shapes people's expectations. The story of the internet entrepreneur becomes widespread and all of a sudden you have different expectations about what you should be doing next week or next year.
And then we try to reinvent ourselves along those lines. I was just talking about Superman. I happened to read that he appears to wear colorful tights. But Superman was invented in the 1930s, and such tights were not available yet, because they hadn't invented Spandex yet. It was invented in 1959 by Joseph Shivers at DuPont. So what happened was our inventors invented the sort of clothes, Spandex, that athletes wear now. So the story becomes reality.

Q: And athletes tend to look a little like Superman…
Somehow, the story became reality. You see bicycles going by, and the riders look like Superman out there.

Q: How does putting animal spirits in a primary place in your thinking about the economy affect your view of the role of government? If the economy functions rationally and well on its own, that seems to lead to one view of what government should do…
I think we have to redefine our views. The efficient markets revolution had a profound impact on people's thinking. It led people to think that the government is mostly incompetent and a nuisance. This is one of Milton Friedman's legacies, as well — that most of the things a government does are veiled attempts to grab something. In his 1962 book Capitalism and Freedom, he talks about licensing. You can't, for example, be a dentist or any number of things without being licensed. And that usually means passing an exam to demonstrate competence. Friedman said that, in fact, licensing is just a veil for monopoly power. He gave as an example the licensing of barbers, and argued there is no need to license barbers because all that would happen if we had unlicensed barbers is you'd have one bad haircut and you wouldn't go back. Maybe he's right about licensing of barbers. But I think that they carried it so far that they missed the essence of our economy. The essence of our economy is that we have elaborate rules that make our competition work to the benefit of the general public. Otherwise, competition would be vicious and destructive. And so it's the rules that matter. The rules have to be complicated, because the regulations that the government makes have to work around all the complexity of human nature. And you have to keep adjusting past structures to create a congenial environment for entrepreneurship and business.

Q: In Animal Spirits you compare government to a parent, who has to be neither too strict nor too lenient. The metaphor itself seems to put government in a role of having a lot of influence.
We maybe shouldn't have used the parent/child metaphor, because that metaphor sounds, well, paternalistic. It's really setting our own rules as a society. There is no parent. The basic truth is that human reproduction produces millions of highly competent people, and there is no one who is super-competent. Even Einstein — he was asked to be president of Israel in 1952 but he turned it down, I suppose because he knew his limits. He's not able to run all of our society. He's not that smart. That's the problem.

So I like to think about regulations of industry and finance as not imposed by the government, but chosen by us. The government responds to suggestions by people who are doing business as to what kind of rules would be good. Sometimes there's a conflict between different groups. And sometimes businesspeople don't like the rules imposed on them. We also have self-regulatory organizations and self-appointed trade groups and the like that impose voluntary rules on their members. That is all part of rule making. It's part of what makes the economy work.

Q: One criticism I've heard is that policymakers are just as flawed and act as irrationally as anyone else.
We all make rules for ourselves. I'm going to exercise every day. Or I'm not going to eat that whole box of chocolates. And we try to enforce that rule through some psychological mechanism. As soon as children are put together on the playground and they're playing some game, they'll make rules. And they'll enforce rules. That's part of human nature. The regulator doesn't have to be smarter than us. I think of it as being like the player in a sports event who may shout at the referee, but ultimately wants the referee there, because otherwise the game would be intolerable. People would be getting hurt.

The criticism is often worse than what you just said. The people who are opposed to regulation tend to have a view that the regulators are not only not smarter, but that they're idiots, because they are people who couldn't make it in the real world. And I think that is a pernicious view. The regulators I've met, who have reached important positions in the government, seem to be perfectly thoughtful and intelligent people. I think that's a place where we want to place some of our graduates, as regulators, and I'm just as proud of them as the ones who make huge fortunes.

Interview conducted and edited by Jonathan T.F. Weisberg.