Rational expectations might really be wrong. People might make systematic errors, thinking that booms or busts will last forever. Photo / Thinkstock
One of the strongest orthodoxies in modern economics is being challenged, and there could be big implications for the state of the profession. The new, rebellious ideas might also help us understand why financial bubbles happen.
Economists have realised for a very long time that expectations are crucial to economic behaviour. If you expect to lose your job, you might consider holding off on buying a car. If you expect inflation in the future, you might consider buying things now, before prices go up. And so on.
The problem is that it's very tricky to make mathematical models that capture how people set their expectations. Do they form them from long-term trends? Do they look at recent changes? Do they make predictions based on theories, and if so, what theories do they use?
Read also:
• Noah Smith: Finance has caught on to behavioural economics
Economists wrestled with these questions in the 1960s and 1970s, and came up with a number of different answers, none of them very satisfying. Then, in a pair of papers in 1972 and 1976, University of Chicago economist Robert Lucas constructed a theory that economists would use for decades to come. This was something called rational expectations.
Lucas' idea drew on a technique developed more than a decade before, by an economist named John Muth. Rational expectations means that, on average, people make correct guesses about the future of the economy. Sometimes they make mistakes - they predict unemployment will be lower than it actually turns out to be, or inflation higher. But they don't make any systematic mistakes. Basically, the theory of rational expectations says that people's mental model of how the economy works is the correct model.
Economists flocked to rational expectations, and pretty soon it had become canon. By the 1990s, almost no one in the mainstream of economics used anything else. Rational expectations is relatively easy to use, and it felt universal. In addition, the idea was very attractive to people who like to think that the economy works smoothly and efficiently. If rational expectations prevail, there is less scope for governments to intervene and improve the functioning of the economy, or for financiers to make money off of the irrationality of the populace.
There has always been one big problem with rational expectations - it might just not be right.
But there has always been one big problem with rational expectations - it might just not be right. People really might make systematic mistakes in the way they predict the workings of the economy. If economists insist on using an incorrect assumption as the core of their models, it will force them into ever more Byzantine theoretical contortions, as the models repeatedly fail to fit the facts. Cynics would argue that this is exactly what we see happening in macroeconomics, where the prevailing theories have led to precious few correct predictions during the past decade.
Why might rational expectations be wrong? For one thing, people aren't born knowing how the economy functions. They have to learn. That learning process will always leave them with some amount of residual uncertainty about the way things work, which will make them hesitant about things like investing in stocks. Another problem is that getting things exactly right is probably too taxing for the human brain - we probably use approximations.
Economists have begun tentatively to make models using these and other kinds of small deviations from the rational- expectations orthodoxy. But in the background, a small minority of economists has always suspected that rational expectations might fail in far more dramatic ways. Human psychology might make expectations not just wrong, but dramatically and consistently wrong. This idea was advanced in 2013 by Robin Greenwood and Andrei Shleifer, two leading economists in the behavioral finance field.
Now, a trio of economists from the Federal Reserve Board has gone further, exploring whether human mistakes can explain the housing bubble. In a recent presentation for the National Bureau of Economic Research, Jesse Bricker, Jacob Krimmel, and Claudia Sahm showed some very interesting findings.
The payoff could be big if the profession devises models that successfully explain phenomena like bubbles and crashes.
The Fed threesome looked at data from the Survey of Consumer Finances, from before and after the housing crash in 2008. They found that more optimistic ZIP codes - that is, places where people had unrealistically high expectations for their own incomes - were more likely to overpay for houses in the bubble run-up before 2008. These overoptimistic people also took on more debt, and they were more likely to increase borrowing in response to rising house prices.
This looks an awful lot like a systematic mistake. If you were a hedge-fund manager in 2005 or 2006, and you had access to data showing that people were being too optimistic, you could make a lot of money shorting the housing bubble - as some hedge-fund managers in fact did. If you were the government in 2005 and 2006, you might have been able to use expectations surveys as a warning sign that told you when to restrict mortgage lending.
In other words, rational expectations might really be wrong. People might make systematic errors, thinking that booms or busts will last forever. If that's the case, then it will require the economics profession to abandon one of its strongest orthodoxies. But the payoff could be big if the profession devises models that successfully explain phenomena like bubbles and crashes.
Noah Smith is an assistant professor of finance at Stony Brook University and a freelance writer for finance and business publications.
- Bloomberg