Nudge Nudge, Wink Wink
Behavioral economics—the governing theory of Obama’s nanny state.
Apr 19, 2010, Vol. 15, No. 29 • By ANDREW FERGUSON
The premise of behavioral economics is “predictable irrationality.” (Another catchphrase—you have to get used to them.) We all know we do dumb things. But the behavioralists say they’ve discovered that we do dumb things systematically; we act against our own best interest (eating pie, failing to save for the future) with a consistency that smart people can observe, catalogue, anticipate, and exploit. If you as choice architect, for example, know about the “status quo bias”—people are disinclined to alter their immediate circumstances even in the face of a clear long-term benefit—you’ll switch the default option on the 401(k). A list of the irrational quirks, or cognitive biases, that behavioral science claims to have uncovered would be endless. In addition to status quo bias, there’s delusional optimism, loss aversion, the representativeness heuristic, the law of small numbers, disaster myopia, the availability heuristic, the planning fallacy, the mere-measurement effect, the mere-exposure effect, even the “yeah, whatever heuristic,” so named by Sunstein and Thaler, who have a bias for whimsy, often fatal.
This grounding in the real world, confirmed by social science, is supposed to make behavioral economics superior to traditional economics as a guide to regulating human activity. Traditional economics—rational choice economics, or neoclassical economics—gets a rough going over from behavioral economists. By their reading, its gravest error is to accept homo economicus, the notion that man is a rational economic actor who is acting always and everywhere in his own best interest, however conceived. Traditional economists don’t really believe this, at least not with the dogmatic insistence they’re accused of, but pretending that they do allows behavioral economists to position themselves as hard-headed realists trying to correct the airy abstractions of out-of-touch dreamers—a clever reversal of the cliché that usually makes liberals out to be the softies and right-wingers the no-nonsense types. Behavioral economics, wrote a smitten correspondent for the New York Times, “is the study of everyday life as it actually happens, not as some textbook says it should.”
It’s been 15 months now since behavioral economics was enthroned as the administration’s reigning regulatory philosophy. If it does indeed break with a century of conventional wisdom in economics, as its partisans claim, then we should be seeing its effects already.
“It’s all over the place,” Thaler told me. “It’s hard to find a domain where you don’t see aspects of this way of doing things.” He mentioned a recent proposal to require all employers to enroll their employees automatically in retirement accounts, drawing on the opt-out model championed in Nudge. The nudge given to employees, however, comes only after Congress levels an unnudgey mandate on employers. Thaler also pointed to Michelle Obama’s public campaign against obesity, in which she has delivered stern lectures to grocers, food processors, parents, and schools about how fat their customers, kids, and students are. Yet Mrs. Obama’s pestering is just an example of the bully pulpit—government officials and first ladies have never required behavioral science to pound the podium.
Sunstein himself, in an OMB report issued earlier this year, listed several administration proposals that had been touched by the insights of behavioral economics. One would build on the behavioralists’ notion of “social norms”: “Individual behavior is much influenced by the perceived behavior of other people.” So President Obama issued an executive order that banned texting in government cars, “to help promote a norm” that would discourage private citizens from driving while distracted. The Family Smoking Prevention and Tobacco Control Act of 2009 required that warnings on cigarette packages be greatly enlarged and simplified—another insight based on the behavioral concept of salience, which tells us that people pay more attention to images that are hard to ignore. And when the administration designed the 2009 “middle-class tax cut,” it hearkened again to the wisdom of behavioral science: Experiments proved that taxpayers would be more likely to spend the extra money if it was dispensed in increments, through adjustments in paycheck withholding, than if it came in a lump sum, as a rebate.
Some with high hopes have found these small-bore results unexpectedly disappointing. Only a year after heralding the invention of the “nudge state,” the Wall Street Journal’s economics writer followed up this March with a story headlined “Economic Policy ‘Nudge’ Gives Way to ‘Shove.’ ”
“Some of the biggest proposals of last year have disappeared without a trace,” the reporter wrote. In financial reform, for example, the Treasury Department had proposed requiring mortgage lenders and credit card companies to offer “plain vanilla” contracts—another idea popularized by Nudge—written in simple language and providing straightforward terms free of fine print. Those proposals have been shelved. Yet the Journal article quoted administration officials who insisted that behavioral economics was still alive. They cited a plan to give cut-rate loans or tax incentives to landlords to encourage them to upgrade their properties with energy efficient appliances. Once again, though, the influence of the behavioralists is hard to credit. Such a proposal operates according to traditional economics—landlords will rationally pursue their economic self-interest by grabbing a tax break—rather than to the “predictable irrationality” that the behavioralists believe they can correct through regulation.
In the grander areas of public policy, in the environment, financial reform, and health care, the administration’s hoped-for libertarian paternalism is nowhere to be found. In place of gentle pokes and prods and nudges, the administration is hoping to levy taxes and bans, impose mandates and caps, set prices and restrain trade to make people behave properly—all the command-and-control methods from the Old Kind of Democrats’ handbook. Removed from the nurturing environment of the university, soft paternalism stiffens up considerably.
What’s happened? It’s not yet clear how pertinent the science of behavioral economics is to the real world, even though the real world is supposed to be its specialty, as the Times man said. Certainly it shows no advantage in predictive power. No behavioral models foresaw the fiscal collapse of 2008; behavioral economists were as surprised as traditional economists when the housing bubble went pffft. Projecting their principles into the future, behavioral economists can be as goofy as the rest of us. Like many Americans, many behavioralists were against President Bush’s surge in Iraq in 2007. Unlike many Americans, however, the behavioralists could pretend that their skepticism was rooted in science rather than political disposition.
As the surge was being debated, the behavioralist Daniel Kahneman published an essay that was intended as a rebuke to Bush’s warmongering. Kahneman pointed to “several well-known laboratory demonstrations” proving that “hawkish beliefs and preferences . . . [are] built into the fabric of the human mind” and hence not entirely rational. A hawk’s irrationality takes many forms, upon each of which the behavioralists have bestowed a complicated name. He mentioned “reactive devaluation” and “illusion of control” and “the fundamental attribution error” and much else. Unchecked, these cognitive biases might lead a nation, or at least its leader, to escalate a war foolishly, based upon nothing but reptilian instinct.
In hindsight, of course, Bush’s decision doesn’t look irrational at all. And it didn’t seem irrational to lots of reasonable people at the time. Kahneman’s decision to cast the prudential question of the surge as a contest between reasonable science and blind biological urge was silly at best, sinister at worst.
Aside from being wrong—and unreasonable, to boot —the Kahneman essay illustrated one of the salient tendencies among behavioral economists. Their definition of “irrational” is slippery. It can apply to any opinion or style of behavior they disagree with on political grounds. Consider the landlord initiative mentioned above. It’s telling that the Obama regulators consider this a case for behavioral economics. If a landlord chooses to waste energy with inefficient appliances, traditional economics would give him the benefit of the doubt and search for reasons why he might do that. His rationality, that is, would be assumed. But the Obama regulators presume the landlord’s behavior is irrational and ripe for a correction based on their behavioral insights. And why is the landlord being irrational? Because wasting energy has social effects (global warming, increased dependence on foreign oil, and so on) that the behavioralists dislike and the landlord discounts. Such behavior, in their view, is irrational on its face, the symptom of a cognitive bias—“myopia,” maybe, or the “endowment effect.”
The behavioralists are often caught smuggling in a normative and political judgment under the cloak of disinterested science. A hidden assumption is easy to conceal because the science that the behavioral economists draw upon is highly elastic, not to say flimsy. One cognitive bias that the behavioralists don’t mention, though its lure seems irresistible, is the bias that makes human beings swallow uncritically the declarations of social science. The bias deters the layman from snooping around to see if the science makes sense. This is the well-established “chump effect,” a name I just made up. It accounts for the breathless reception given to the books by Gladwell and the other popularizers of sociological and psychological research. “Findings reveal . . .” “Scientists have uncovered . . .” “Research has shown that . . .” And we swoon.
But what does “research show”? What do “findings reveal”? Usually much less than the behavioral economists want to believe. And they do want to believe. They burrow through stacks of boring journals and come upon an article describing a new experiment with a deliciously provocative conclusion and looking up from the page they can hear the cry: “Generalize me, big boy! Make me relevant!” Skepticism flies off, and the economists never stop to consider the fishy process by which those provocative conclusions were reached.
The vast majority come from behavioral experiments that are completely artificial in their construction. Most take place in labs at elite universities, where graduate students and professors pay undergraduates a pittance to sit for varying periods of time and fill out questionnaires of varying length. Sometimes the subjects are asked to interact while the grad students watch them, other times the questionnaires alone suffice to produce the data. “Behavioral economics,” Thaler likes to say, “is the study of humans in markets.” Actually, it’s the study of college kids in psych labs.
An example: In his recent OMB report, Sunstein insists that regulators take account of a cognitive bias called “probability neglect” in finding ways to impose their soft paternalism. Probability neglect is defined like this: “When emotions are strongly felt, people may focus on the outcome and not on the probability that it will occur.” Which is to say, when you really want something you tend to be unrealistic about your chances of getting it. Surely that’s true for all of us sometimes, and always true for some of us. But is it a universal pattern of behavior, one reliable enough to enshrine in a one-size-fits-all government regulation?
Who knows? Behavioral economists trace their detailed understanding of probability neglect to a study from 2001. In three separate experiments, a pair of graduate students from the University of Chicago Business School asked undergraduates from Chicago and Rice University to complete questionnaires. The 40 students from Rice, in Texas, were asked whether they would prefer to receive $50 in cash or “the opportunity to meet and kiss your favorite movie star.” The methodological details aren’t worth describing here—we can stipulate that the experiments were conducted with the utmost rigor and elegance. What’s notable is that the experiments were thereafter assumed by social scientists to have established “probability neglect” as a consistent principle guiding human behavior in the marketplace. All thanks to 40 kids from Texas, filling out a form in 2001.
Likewise, when administration officials designed the 2009 tax cut—delivered by withholding less from paychecks instead of by making single lump payments to taxpayers—they were operating on the basis of another U.C. study. Two researchers brought undergraduates to a lab and handed each $50. Half the students were told the extra money was a “tuition rebate,” the other that it was a “bonus.” The experimenters followed up with questions by email a week later. The bonus group spent more money than the rebate group. From this result the administration felt confident in predicting how 140 million taxpayers would spend their tax cut. No one knows whether it was more effective in increments than it would have been in lump payments, of course; the tax cut was too small to trace in an economy so vast. It probably wouldn’t have worked either way. But at least the administration had an experiment on its side.
Asked about behavioral economics in an interview recently, the neoclassical economist Gary Becker summed up his reservations. “There is a heck of a difference between demonstrating something in a laboratory, in experiments, even highly sophisticated experiments, and showing that they are important in the marketplace,” he said. “Economics theory is not about how people act in experiments, but how they act in markets.”
Other prominent skeptics, among them Joshua Wright of George Mason Law School and Gregory Mitchell of the University of Virginia, have begun dismantling the behavioralists’ conceit more systematically. “Even if you discover a real cognitive bias,” Wright said last month, “there will be a good deal of variation within the population, based on cognitive ability and personality traits. And if the bias varies from person to person, you can’t assume that the bias will just ‘scale up,’ in a generalized way, when it’s in the marketplace. Thaler and Sunstein will take a single study of a hundred Duke undergrads and say, ‘Here’s what we found—and here are the public policy implications.’ That’s not scientific. That’s just sloppy.”
Mitchell cut even deeper. He has discovered what he calls a “citation bias within psychology that favors pessimistic accounts of decision making.” Experiments designed to demonstrate irrationality tend to find it. Even the most ingenious experiment can’t replicate how individuals behave in the real world. We change and adapt over the course of months and years, reflect and learn, and call on the help of friends and family. These vital and unpredictable improvisations won’t happen in the vacuum of the college psych lab, with a besmocked Ph.D. student hovering close by.
Behavioral economists deny any ideological intent in their work. The closest I’ve seen any of them come to conceding a political point of view was when Thaler, in a recent interview, said, “If there’s a regulatory philosophy in behavioral economics, it’s that we should recognize that people in the economy are human and that there are people out there trying to take advantage of them.” In this sense, behavioral economics is just conventional 1960s liberalism—and conventional 1960s economics, too—that assumes the free market itself is a kind of unending con game, with the smart guys exploiting the saps. As an advocate for the market’s hapless victims, the government has the responsibility to undo the con, a task that will require only the smartest administrators operating according to only the latest scientific research and making the most exquisite moral judgments.
You can see how useful the notion of irrational man is to a would-be regulator. It is less helpful to the rest of us, because it runs counter to every intuition a person has about himself. Nobody sees himself always as a boob, constantly misunderstanding his place in the world and the effect he has upon it. Surely the behavioral economists don’t see themselves that way. Only rational people can police the irrationality of others according to the principles of an advanced scientific discipline. If the behavioralists were boobs too, their entire edifice would collapse from its own contradictions. Somebody’s got to be smart enough to see how silly the rest of us are.
Traditional economics has always been more modest. Assuming the rationality of man was a device that made the discipline possible. The alternative—irrational people behaving in irrational ways—would complicate the world beyond the possibility of understanding. But the modesty wasn’t just epistemological. It was also a democratic impulse, a sign of neighborly deference. A regulator who always assumed that man was other than rational was inviting himself into a position where he could exert a control over his fellow citizens that wasn’t proper for a true democrat. Self-government demands this deference. It won’t work otherwise.
“Ultimately,” the economist Brian Mannix wrote not long ago, “we insist that our regulators start from a presumption of rationality for the same reason that we insist that our criminal courts start from a presumption of innocence: not because the assumption is necessarily true, but because a government that proceeds from the opposite assumption is inevitably tyrannical.”
Well, maybe not inevitably. Those behavioralists may be smart, but they’re not quick. It’s been 15 months since President Obama gave them 100 days to explain how to use behavioral economics in government regulation. They’re still working on the report.
Andrew Ferguson is a senior editor at The Weekly Standard and the author of Crazy U: One Dad’s Crash Course in Getting His Kid Into College.