Did Richard Thaler Really Shift the Paradigm in Economics?

Richard Thaler deserves the Nobel Memorial Prize in economics. But media treatments of Thaler’s work, and of behavioral economics more generally, suggest that it provides a much-deserved comeuppance to conventional microeconomics. Well . . . Not quite.

It may sound snarkier than I intend it, but economists often win the Nobel Prize for pointing out obvious things that most other economists (and other social scientists often enough as well) have a habit of ignoring.

For example, James Buchanan won the Nobel Prize for pointing out that government officials have individual-level incentives — higher pay, bigger offices, greater personal power — that can distort how they implement government policies. Well, duh. James Madison points out as much in The Federalist No. 51. But before Buchanan (and Tullock) began pointing out the obvious, economists (and political scientists, and journalists and others) had black-boxed government administration and implementation, often effectively assuming government officials neutrally implement precisely what policymakers intend to implement when they make policies. Recognizing that government employees respond to incentives in their work just like the rest of us was an important corrective to the way much modern economics was being done.

Or take Elinor Ostrom. She published scores of studies from all over the real world showing that people can often (but not always) solve collective action problems without government intervention. Well, we knew of that possibility as a theoretical matter before her studies. But economists and political scientists often ignored the theoretical options, rushing instead to recommend government action whenever someone observed a collective action problem. Or think of Robert Lucas, who argued that microfoundations of macroeconomic models need to be consistent with the way people behave individually. Well, duh again. Yet mainstream macroeconomics had ignored that obvious insight for generations prior to Lucas point it out.

But here Thaler’s work intersects our other “well, duh” moments. Lucas, and Ostrom, and Buchanan and Tullock worked consistently within the “rational choice” framework of modern economics. Thaler’s work challenges the breadth of application of that framework. Economists, and rational choice theorists more generally, have a blind spot, he argues, for just how often their assumptions about human behavior are inconsistent with real human behavior. That’s an important point.

Yet here’s where spin matters: Does Thaler provide a correction to previous economics, underscoring something everyone always knew but just ignored as a practical matter, or is Thaler’s work revolutionary, inviting a broad and necessary reconceptualization of standard microeconomics?

Much of the commentary on Thaler, and on behavioral economics more generally, hints that he overthrows the basis for standard, modern microeconomics. Umm. Eh. Well . . . No. He has built a career by correcting a blind spot in modern academic economics. But his insight provides us with a “well, duh” moment rather than a “we need totally to rewrite modern economics” moment that some of his journalistic (and academic) supporters suggest it provides. And to say this takes nothing away from the contribution that Thaler has made to economics, or the merit of the award.

Let’s start with the rationality postulate. The problem is that economists (and rational choice theorists more generally) use the word “rationality” more narrowly than we use the word in everyday conversation. That causes confusion when we hear rational choice scholars talking about rational choice models. In everyday usage, to say someone is “irrational” means that the person is crazy; the person who thinks he or she is Napoleon, for example. Irrational people deny fundamental aspects of reality.

Rational choice theorists employ a much thinner notion of rationality. Rationality for economists basically means two things: First, given a choice between two outcomes (call them outcome “A” and outcome “B”) a person can rank those two outcomes like this: I prefer A to B. I prefer B to A. Or I am indifferent between A and B. Rationality means that people can compare two outcomes. Secondly, for economists, rationality means that a person’s choices are transitive: If I prefer A to B, and if I prefer B to C, then I prefer A to C. (So, too, if I am indifferent between A and B, and am indifferent between B and C, then I am indifferent between A and C.) To be sure, this description needs to be complicated, particularly when we move to probabilistic outcomes, or when we turn preference relations into mathematically-specified “utility functions.” But it’s a good enough place to start.

Note, then, that a person who is certifiable crazy by every-day standards can be entirely rational by the narrower definition economists use. A crazy person who believes himself to be Napoleon can nonetheless be entirely rational given the thin notion of rationality.

At the same time, these are assumptions, assumptions made to simplify reality to make economic understanding and explanation easier. For example, experiments exist in which people do not act consistently with the assumption of transitivity. That’s o.k. That result is not a huge theoretical problem for microeconomics if most people act consistently with the postulate much of the time.

Given the press devoted to Thaler’s work, however, one might think that behavioral economics has discovered that when people prefer cherry pie to blueberry pie, and prefer blueberry pie to apple pie, they nonetheless consistently prefer apple to cherry pie rather than cherry to apple. But it’s not true that most people’s preferences are mostly intransitive.

While it’s important to know that people’s preferences are not always transitive, we need to know a lot more than that before this observation becomes a threat to standard microeconomics.

To be sure, there are examples of consistent error derived from things like framing effects (for example, most people consistently choose something framed as providing a 95 percent chance of living over something framed as providing a 5 percent chance of dying, even those they’re mathematically identical odds). That’s fine as well. Those results no more imperil the usefulness of standard microeconomic analysis than does the existence of abnormal psychology. It just means the social scientists toolbox needs to be larger than rational choice theory. But, duh. It’s like observing that it doesn’t help to have a map of New Mexico when traveling in Arizona.

Thaler’s work underscores that the economist’s rationality postulates cannot account for all human behavior. That’s an important point. But I don’t know that many, or even any, economists very much believed the opposite in any serious way. But it was a blind spot in standard economic treatments of human behavior. Thaler’s work underscores that other social sciences can matter, and alternative expressions of rationality can matter.

The limitation of behavioral economics, at least to date, is its inability, so far at least, to generate theories that apply much beyond highly particularized cases. (And, no, “nudge” is not a theory. But that’s for another column.) The theoretical contribution of behavioral economics to date has been to pile up examples of inconsistencies between economic theory and observed human behavior. That’s not unimportant. Nonetheless, that is insufficient to cause (in Kuhn’s overused and often misapplied phrase) a “paradigm shift” in economics, or in rational choice theory more generally.