June 28, 2003
Calculating the Irrational in Economics
hen the Federal Reserve Bank of Boston invited the leading behavioral economists to a Cape Cod golf resort this month to make their case, it was plainly a signal moment. "It has the feeling of being summoned by the king," said Colin F. Camerer, a star behaviorist who teaches at the California Institute of Technology. "Sort of like: `I understand you're the finest lute player in the region. Will you come and play for me?' "
Until the last few years, behavioral economics — which blends psychology, economics and, increasingly, neuroscience to argue that emotion plays a huge role in how people make economic decisions — was an extremely tight-knit group. It had little influence and few practitioners. One economist at the bank's conference recalled an Alaska kayaking trip that Mr. Camerer took with another prominent behaviorist a decade ago. "If that kayak had flipped," he said, "half the field would have been eradicated."
But the field has grown, as has its influence. In 1996, Alan Greenspan's warning of "irrational exuberance" acknowledged, as the behaviorists do, that the average investor is hardly the superrational "homo economicus" that mainstream economists depict. In 2001, the young behaviorist Matthew Rabin won the John Bates Clark Medal; last fall, the psychologist Daniel Kahneman, a forefather of behavioral economics, was awarded the Nobel in economic science.
And so it was that the Boston Fed summoned the behaviorists to the Wequassett Inn in Chatham, Mass. The conference was given the quaint title "How Humans Behave," as if monetary policymakers had suddenly realized that, lo and behold, on the other end of all that policy are actual people. The collection of mainstream economists and central bankers would be the highest-level audience the behaviorists had ever enjoyed, the best chance yet for their new thinking to hit the bloodstream.
From the outset the mood was civil, especially considering that the behaviorists are essentially calling for an end to economics as we know it. (As one economist grumbled, "What you have to understand is that behavioral economics is attacking the foundation of what welfare economics is built on.") So it was not surprising that some Fed elders seemed wary, as if they were at a family reunion and welcoming a distant cousin about whom they had heard only puzzling rumors. But with the economy stuck in a condition between dismal and desperate, the behaviorists' timing could not have been better.
"All our models and forecasts say we'll see a better second half," Cathy E. Minehan, president of the Boston Fed, said in her opening address. "But we said that last year. Now don't get me wrong: mathematical models are wonderful tools. But are there ways this process can be done better? Can we inform the policymaker from 50,000 feet with wisdom gained on the ground, in the human brain, or in the way humans make decisions and organize themselves? I hope so."
Ms. Minehan and her colleagues were particularly hoping to learn why Americans save too little, acquire too much expensive debt and perform such achingly self-destructive feats of portfolio management. The behaviorists, for their part, were put in a tight spot: eager to prove themselves but leery of overpromising. "Virtually everyone doing behavioral economics agrees we should go slowly in advocating policy change," Mr. Camerer wrote in the paper he presented. "Our thinking was also not designed to precisely answer questions about welfare and policy, but this is a good time in the intellectual history of the field to say something."
Eldar Shafir, who teaches psychology and public affairs at Princeton, began with a behavioral economics primer. It was full of the anomalies the field is known for, including the popular "6 jam-vs.-24 jam" experiment. In an upscale grocery story, researchers set up a tasting booth first with 6 jars of jams, and later with 24 jars. In the first case, 40 percent of the customers stopped to taste and 30 percent bought; in the second, 60 percent tasted but only 3 percent bought. The point is that too many options can flummox a consumer — and if 24 jars of jam pose a problem, imagine what 8,000 mutual funds can do. Standard economics would argue that people are better off with more options. But behavioral economics argues that people behave less like mathematical models than like — well, people.
Among the behaviorists, there is the common sentiment that economics has been ruined by math. "Neoclassical economists came along in the mid-19th century and wanted to mathematize the new science of economics," said George Loewenstein, a professor at Carnegie Mellon University. "They couldn't include `the passions,' or emotions, in their models, because they were too unruly, too complex. But they also thought that the emotions were unknowable."
Mr. Loewenstein described how he and his colleagues want to prove otherwise — that not only are emotions not unknowable but that when it comes to money, they may be more powerful than math. This is why Mr. Loewenstein studies how people make financial choices while they are experiencing various degrees of sadness, hunger and sexual arousal. This is why Colin Camerer has become a student of brain imaging, trying to identify where a subject's brain lights up when, for instance, a lowball offer leaves him disgusted.
But the most radical idea presented at the conference belonged to Richard H. Thaler. His paper, written with the legal scholar Cass R. Sunstein, was called "Libertarian Paternalism Is Not an Oxymoron." Leonine and youthful at 57, Mr. Thaler, who teaches at the University of Chicago, is widely considered the founder of behavioral economics (and some say, its next Nobel winner). He is more confident and, accordingly, more prescriptive than his younger colleagues.
"Behavioral economics offers powerful tools to achieve policy goals," he told the conference. "And libertarian paternalism is an attractive approach to solving policy problems. What else? I think the only other alternative is inept neglect."
Mr. Thaler has concluded that too many people, no matter how educated or vigilant, are poor planners, inconsistent savers and haphazard investors. His solution: public and private institutions should gently steer individuals toward more enlightened choices. That is, they must be saved from themselves. Mr. Thaler's most concrete idea is Save More Tomorrow (SMarT), a savings plan whereby employees pledge a share of their future salary increases to a retirement account. In test cases, the plan has proved remarkably successful.
"This was not pulled out of thin air," Mr. Thaler said. "It was done using what I call first-grade psychology. We knew this was going to work, no question." Indeed, the SMarT plan takes advantage of behavioral economics' basic tenets: "loss aversion" (people fear loss because it causes them far more pain than the pleasure they receive from gain; but since the SMarT plan covers a future raise, they never feel its loss); "status-quo bias" (since people are reluctant to change, the change can be made for them); and "mental accounting" (people have a pressing need to direct different streams of money into different "accounts").
Mr. Thaler was followed by David I. Laibson, a young Harvard behaviorist who also endorsed a paternalistic approach. "There are two enormous travesties in the financial services industry," Mr. Laibson said. "One, people have too much of their own company's stock, and two, mutual-fund management fees are too high." His solution to the first problem: an automatic asset reallocation to keep an employee from holding more than 20 percent of his portfolio in company stock.
"People could opt out," he said. "If you're crazy enough to do that, fine, that's your right, but we'd certainly push them down." His solution to the second problem: warning labels about management fees, modeled after the surgeon general's cigarette warning.
Mr. Laibson's and Mr. Thaler's proposals were warmly received by the bankers and mainstream economists. If this is behavioral economics, what's not to like? The proposals seemed to be sound and not particularly invasive solutions to Americans' troubling money habits. All the earlier talk of sexual-arousal studies and brain imaging may have left them flat; but here were some real action items.
The behaviorists, most of whom are hardcore empiricists, even felt comfortable enough to declare their own research wish lists. Dan Ariely of the Massachusetts Institute of Technology trolled the room for a good contact at the Internal Revenue Service (he wants to study the psychology of tax cheats). Duncan J. Watts, a sociologist at Columbia University, half-jokingly requested access to the phone and e-mail records of all Federal Reserve employees (he is looking for good data to better understand how organizations behave).
The warm reception didn't mean wholesale conversion. When Jeffrey C. Fuhrer, the Boston Fed's chief economist, was asked about Mr. Camerer's desire for a new Economics 101 textbook, one that puts behaviorism at the center and mathematical modeling on the fringe, he responded: "Yeah, that's his `I Have a Dream' speech. I think that's still weeks off."
Still, Mr. Fuhrer, who organized the conference, was delighted with its outcome: "I think we would have been crazy to expect we'd walk out of this conference and say, `O.K., we're going to our next meeting, and now we know what to do because these guys told us.' But having had these conversations, where people look at economics from a different viewpoint, will we think a little differently about how we do research and exactly which people we might talk to? You bet we will."
They may have little choice. As Frederick S. Breimyer, chief economist of the