Is Alfred Marshall Passe?

How a young Harvard economist is challenging the traditionalists in his field.

Dana Linden, Forbes magazine

Three years ago the Macarthur Foundation awarded economist Sendhil Mullainathan one of its $500,000 "genius" grants. Only 29 at the time, Mullainathan celebrated by buying himself a new pair of $49 Allen Iverson sneakers.

That's it? "The luxury part is, I already had a pair of sneakers," he explains. To the lanky, understated professor, his purchasing decision made sense: Even though he really didn't need a new pair of sneakers, he bought them anyway. But is that rational? That question goes to the core of Mullainathan's specialty--behavioral economics, an emerging branch of the field that seeks to integrate psychology with economics. He's one of perhaps four or five of his generation leading the overhaul of the traditional economic model and its underlying assumption--that people always behave rationally and make the right decision, unfettered by emotion or impulse. Says economist Richard Thaler of University of Chicago Graduate School of Business: "I'd put him with a small group who are where the future is."

What's rational, for instance, about the sports fan who refuses to pay more than $200 to buy a Super Bowl ticket, but won't accept less than $400 for one he already owns? Or travelers who leave generous tips at restaurants they expect never to visit again? Or gamblers who, at the end of a losing day, bet on a long shot?

It's one thing, though, for the psychoeconomists to say that people make foolish choices, quite another for them to turn this observation into a theory that can explain consumer decision making, stock market movements or economic cycles. Says ardent rationalist Eugene Fama of the University of Chicago: "They've uncovered lots of interesting details about behavior, but will anyone show this has an impact on prices and markets?"

That's where Mullainathan comes in. "He can think conceptually," says Thaler, "but he has very strong empirical skills and an interest in real data."

Born in a small farming village in India, Mullainathan lived there for seven years while his father moved to the U.S. to go to graduate school. On his fifth birthday, his father sent him a three-piece suit. On the way, via oxcart, to have his photo taken, his uncle and grandfather spent the whole time arguing about whether the vest went over or under the jacket. In the photo a beaming Mullainathan proudly wears the vest on top.

After moving to Los Angeles in 1980, Mullainathan left high school without graduating and went to Clarkson University, and then on to Cornell, where he took graduate-level courses in math and computer science. He studied under Thaler, then at Cornell, who was injecting psychology into economics.

Mullainathan was hooked and his research has become increasingly more empirical and less ivory-tower. He went to graduate school at Harvard, then became a junior faculty member at MIT. Harvard snatched him back last year with tenure.

One of his most provocative papers, just submitted for publication, suggests that tiny psychological effects can have potentially enormous impact on demand, more of an impact than price. In 2003 he and several coauthors worked with a bank in South Africa that sent out letters offering short-term loans. They varied the interest rate and also varied a number of cues designed to trigger psychological responses, such as a smiling photo in a corner of the letter and tables that provided more--or less--information and choice. The sample was large, more than 50,000 letters, and the study was randomized and controlled.

The impact of some of the small, nonfinancial cues surprised even the study's authors, though it probably wouldn't have been a shock to creative types on Madison Avenue. It turned out that having a wholesome, happy female picture in a corner of the letter had as much positive impact on the response rate as dropping the interest rate by four percentage points. Says Eldar Shafir, professor of psychology and public affairs at Princeton and a coauthor of the study: "I didn't know that, nobody expected that."

The practical takeaway is that an insurance company can probably sell more auto policies by featuring Reese Witherspoon in its brochures than by slashing margins and sending out letters that scream: "Unprecedented Low Rates!"

Says Matthew Ryan, executive director at Boston ad agency Arnold Worldwide, whose clients include McDonald's, Radio Shack and Fidelity Investments: "If behavioral economists can quantify more, that's exciting. The ad industry will be looking at those things closely."

Conventional economic theory says people accurately calculate costs and benefits, present and future. If they don't save much, it's because consuming more today makes them feel better. If they smoke, it's because the happiness they get from smoking outweighs the costs.

Behavioral economists disagree. They argue that a hardwired lack of self-control prevents people from making rational choices. Mullainathan and Jonathan Gruber of MIT tapped a huge trove of data--surveys of self-reported levels of happiness conducted since 1973 by the U.S. and Canada. What did they find? The population was happier after excise taxes on cigarettes rose. The taxes, Mullainathan suggests, helped some smokers drop an unwanted habit.

Another application of behaviorism that has gotten much attention has to do with 401(k) plans. If you want workers to save more, help them with their self-control by getting them to commit in advance to increase their savings as they get future raises. Make it, in other words, more difficult to spend the entire raise. In 2004 Vanguard rolled out an automatic-contribution plan designed with Thaler and UCLA's Shlomo Benartzi. If it meets expectations, savings should better than triple. More than 300 companies have the plan in place.

A study in Malawi by one of Mullainathan's students is looking at whether what psychologists call a "channel factor"--a stimulus that leads you to act immediately on your intentions--might increase the number of people who pick up the results of their AIDS tests. A puny incentive, equivalent in monetary worth to two cans of Coke, boosted the response from 30% to 70%.

"Let the data build up; let's see who's right," says Mullainathan. "If I'm wrong, I'll be wrong. And that's good. Everyone will see."