The Myth of the Rational Voter: Why Democracies Choose Bad Policies

by Bryan Caplan

More interestingly, errors with drastic real-world repercussions can be cheap for the individual who makes them. How? When most or all of the cost of the mistake falls upon strangers. One person messes up, but other people live with the aftermath.

the typical citizen drops down to a lower level of mental performance as soon as he enters the political field. He argues and analyzes in a way which he would readily recognize as infantile within the sphere of his real interests. He becomes a primitive again. —Joseph Schumpeter, Capitalism, Socialism, and Democracy

most people reject the view that pushing up wages increases unemployment. When I teach intro econ, linking unemployment and excessive wages frequently elicits not only students’ disbelief, but anger: How could I be so callous? But irrationality about labor demand is selective. What happens when my outraged students reach the “Salary Requirements” line on job applications? They could ask for a million dollars a year, but they don’t. When their future rides on it, students honor the economic truism that labor demand slopes down.

The bulk of available evidence suggests that people in all societies tend to be relatively rational when it comes to the beliefs and practices that directly involve their subsistence. . . . The more remote these beliefs and practices are from subsistence activities, the more likely they are to involve nonrational characteristics. —Robert Edgerton, Sick Societies

Above all, the steps should be conceived as tacit. To get in your car and drive away entails a long series of steps—take out your keys, unlock and open the door, sit down, put the key in the ignition, and so on. The thought processes behind these steps are rarely explicit. Yet we know the steps on some level, because when we observe a would-be driver who fails to take one—by, say, trying to open a locked door without using his key—it is easy to state which step he skipped.

A few Muslims sacrifice their lives for their faith, but a billion do not.

Stalin was not so concerned about the condition of agriculture—he tolerated, after all, a desperate famine in the Ukraine in 1947—and so it may not have mattered very much to him whether Lysenko was a charlatan or not. The nuclear project was more important, however, than the lives of Soviet citizens, so it was crucial to be sure that the scientists in the nuclear project were not frauds.

Similarly, in many other areas of the Soviet economy, Marxism fostered reluctance to motivate workers with material rewards for success. In the atomic project, however, Stalin dumped Marxist dogma in favor of bourgeois horse sense: Stalin said also that he was anxious to improve the scientists’ living conditions, and to provide prizes for major achievements—“for example, for the solution of our problem,” Kurchatov wrote. Stalin “said that our scientists were very modest and they sometimes did not notice that they live poorly . . . our state has suffered very much, yet it is surely possible to ensure that several thousand people can live very well, and several thousand people better than very well, with their own dachas, so that they can relax, and with their own cars.”

Maybe Stalin covertly scoffed at the inanities of Marxism, but a more plausible interpretation is that he was rationally irrational. Marxism-Leninism was important to his sense of identity, but his preference was not absolute. As the price of illusion went up, he chose to be less fanatical and more objective.

Want to bet? We encounter the price-sensitivity of irrationality whenever someone unexpectedly offers us a bet based on our professed beliefs.59 Suppose you insist that poverty in the Third World is sure to get worse in the next decade. A challenger immediately retorts, “Want to bet? If you’re really ‘sure,’ you won’t mind giving me ten-to-one odds.” Why are you are unlikely to accept this wager? Perhaps you never believed your own words; your statements were poetry—or lies. But it is implausible to tar all reluctance to bet with insincerity. People often believe that their assertions are true until you make them “put up or shut up.” A bet moderates their views—that is, changes their minds—whether or not they retract their words.

How does this process work? Your default is to believe what makes you feel best. But an offer to bet triggers standby rationality. Two facts then come into focus. First, being wrong endangers your net worth. Second, your belief received little scrutiny before it was adopted. Now you have to ask yourself which is worse: Financial loss in a bet, or psychological loss of self-worth? A few prefer financial ...

If you listen to your fellow citizens, you get the impression that they disagree. How many times have you heard, “Every vote matters”? But people are less credulous than they sound. The infamous poll tax—which restricted the vote to those willing to pay for it—provides a clean illustration. If individuals acted on the belief that one vote makes a big difference, they would be willing to pay a lot to participate. Few are. Historically, poll taxes significantly reduced turnout.65 There is little reason to think that matters are different today. Imagine setting a poll tax to reduce presidential turnout from 50% to 5%. How high would it have to be? A couple hundred dollars? What makes the poll tax alarming is that most of us subconsciously know that most of us subconsciously know that one vote does not count.

For practical questions, standard procedure is to acquire evidence before you form a strong opinion, match your confidence to the quality and quantity of your evidence, and remain open to criticism. For political questions, we routinely override these procedural safeguards.

Incentives often improve performance on tasks of judgment and decision. People “spend” hypothetical money more freely than actual money; they are much more likely to say they will buy something than to actually do so.

Incentives also lead subjects away from “favorable self-presentation behavior toward more realistic choices.”80 Furthermore, a recent paper finds that people get less overconfident when they have to bet real money on their beliefs.

“Useful cognitive capital probably builds up slowly, over days of mental fermentation or years of education rather than in the short-run of an experiment (1–3 hours) . . . [I]ncentives surely do play an important role in inducing long-run capital formation.”83 This claim is consistent with the growing literature on field experiments: Economic actors in their “natural habitat” look considerably more rational than they do in the lab.

Camerer and Hogarth also admit that experiments slight the power of incentives by relying on volunteers, whose “intrinsic motivation”—desire to do well for its own sake—is unusually high.85 Money cannot spur greater effort in those who are already trying their best.

A common summary of the experimental literature is that incentives improve performance on easy problems but hurt performance on hard problems.86 As Einhorn and Hogarth argue: Performance . . . depends on both cognition and motivation. Thus, if incentive size can be thought of as analogous to the speed with which one travels in a given direction, cognition determines the direction. Therefore, if incentives are high but cognition is faulty, one gets to the wrong place faster.87 What Camerer and Hogarth highlight, however, is that the difficulty of a problem falls if you have more time and flexibility to solve it. Hard problems naturally decay into easier problems. Once they are easy enough, incentives work like they are “supposed to.”

Even if we trust only experimental evidence, rational irrationality is a credible explanation for the public’s biased beliefs about economics. Experimentalists admit that incentives help for relatively easy questions. Antimarket, antiforeign, make-work, and pessimistic bias all qualify. These are not subtle errors, but knee-jerk reactions. In non-political contexts, people routinely overcome them. How many refrain from buying appliances because it “destroys jobs”?

Experimentalists also emphasize that incentives help less when there is intrinsic motivation to get things right. In economics, there is intrinsic motivation to get things wrong. If you think the right answer, you feel insensitive and unpatriotic; if you say the right answer, you feel like a pariah. There is about as much intrinsic motivation to understand economics as there is to take out the garbage.

Nearly all economists assume that people vote instrumentally; that is, they vote to get the policies they prefer. What else would they do? Brennan and Lomasky point to the expressive function of voting. Fans at a football game cheer not to help the home team win, but to express their loyalty. Similarly, citizens might vote not to help policies win, but to express their patriotism, their compassion, or their devotion to the environment. This is not hair-splitting. One implication is that inefficient policies like tariffs or the minimum wage might win because expressing support for them makes people feel good about themselves.

The bigot who refuses to serve blacks in his shop foregoes the profit he might have made from their custom; the anti-Semite who will not work with Jews is constrained in his choice of jobs and may well have to knock back one she would otherwise have accepted. To express such antipathy at the ballot box involves neither threat of retaliation nor any significant personal cost.

Brennan and Lomasky do not merely draw the moderate conclusion that political decisions, like market decisions, depend on expressive as well as instrumental concerns. Their conclusion is instead the radical one that—unlike market decisions—political decisions depend primarily on expressive concerns:

The key difference is the mechanism. In expressive voting theory, voters know that feel-good policies are ineffective. Expressive voters do not embrace dubious or absurd beliefs about the world. They simply care more about how policies sound than how they work. The expressive protectionist thinks: “Sure, protectionism makes Americans poorer. But who cares, as long as I can wave the flag and chant ‘U.S.A.! U.S.A.!’ ” In contrast, rationally irrational voters believe that feel-good policies work. The rationally irrational protectionist genuinely holds that protectionism makes Americans richer. If he must deny comparative advantage, so be it.

“Individual voters may, each of them, be entirely rational in voting for war—even where no one of them would, if decisive, take that course.”

Political behavior seems weird because the incentives that voters face are weird. Economists have often been criticized for evading the differences between political and market behavior.98 But this is a failure of economists rather than a failure of economics.

A jaded old statehouse reporter noticed my astonishment and offered some perspective on the unruly behavior of the elected representatives. “If you think these guys are bad,” he said, “you should see their constituents.”

You can’t do serious economics unless you are willing to be playful.

Democracy aggregates preferences. Members of a group want things done. Democracy combines their wants and stirs to get a group decision. This process is terribly confusing because humans almost never completely agree. So what happens? On every issue, democracy must either impose a compromise, or favor one side over its rivals—which is another way of answering, “Who knows?”

Irrationality makes society as a whole better off as long as the psychological benefits minus the material costs are positive: Psychological Benefits – Material Costs > 0. Irrationality makes the individual better off under far less stringent conditions: Psychological Benefits – p * Material Costs > 0. where p is the probability of casting the decisive vote. If p = 0, irrationality is utility-maximizing as long as there are any psychological benefits: Psychological Benefits > 0

With identical voters, most of the biases from the SAEE readily map into foolish policies. Antimarket bias boosts price controls and shortsighted redistribution. Antiforeign bias pushes for protectionism and immigration restrictions, and against trade agreements. Make-work bias recommends labor market regulation to “save jobs.”

When you dislike someone, you tend to see all his actions through a negative filter. When you dislike imports, similarly, it is only natural to overestimate the economic harm of imports, their quantity, the number of jobs they destroy, and the “unfairness” of other countries’ trade policies.

If voters have different endowments, then many may objectively benefit from socially harmful policies. Inequality of wealth is the simplest reason: Even if redistribution is an awfully leaky bucket, it might still enrich the majority.19 But inequality is only the beginning. The owner of a textile mill may be just as rich as the owner of a clothing store, but tariffs affect their interests oppositely.

The sign fits the stereotype: As your income rises, you are more likely to be conservative and Republican. But the effect is small, and shrinks further after controlling for race. A black millionaire is more likely to be a Democrat than a white janitor.25 The Republicans might be the party for the rich, but they are not the party of the rich.

The broken clock of the SIVH is right twice a day. It fails for party identification, Social Security, Medicare, abortion, job programs, national health insurance, Vietnam, and the draft. But it works tolerably well for a few scattered issues.32 You might expect to see the exceptions on big questions with a lot of money at stake, but the truth is almost the reverse. The SIVH shines brightest on the banal issue of smoking. Donald Green and Ann Gerken find that smokers and nonsmokers are ideologically and demographically similar, but smokers are a lot more opposed to restrictions and taxes on their favorite vice.33 Belief in “smokers’ rights” cleanly rises with daily cigarette consumption: fully 61.5% of “heavy” smokers want laxer antismoking policies, but only 13.9% of people who “never smoked” agree. If the SIVH were true, comparable patterns of belief would be everywhere. They are not.

Most voters disown selfish motives. They personally back the policies that are best for the country, ethically right, and consistent with social justice. At the same time, they see other voters—not just their opponents, but often their allies too—as deeply selfish.

When individuals contrast their own faultless motives to the selfishness of others, our natural impulse is to interpret it as self-serving bias. But the empirical evidence suggests that self-descriptions are accurate. People err not in overestimating their own altruism, but in underestimating the altruism of others. Indeed, “underestimate” is an understatement. Individuals are not just less politically selfish than usually thought. As voters, they scarcely appear selfish at all.

If people buy more altruism when the price is low, and altruistic voting is basically free, we should expect voters to consume a lot more altruism. It would cut against basic economics if the SIVH did work.

But unselfishness also lets democratic performance fall from bad to worse. Irrational unselfish voters are probably more dangerous than irrational selfish ones. If unselfish voters misunderstand the world, they can easily reach a misguided consensus. Their irrationality points them in the wrong direction; their unselfishness keeps them in marching formation, enabling them to rapidly approach their destination.

The upshot is that the failure of the SIVH makes democracy look worse. Voter irrationality is not tempered by the petty squabbling ordinarily guaranteed by human selfishness. Precisely because people put personal interests aside when they enter the political arena, intellectual errors readily blossom into foolish policies.

Strange as it sounds, a “winning platform”—a platform able to defeat any other—may not exist. Theorists often expect democratic policies to “cycle,” and wonder why real-world policies are so stable.

In formal statistical terms, political opinions look one-dimensional. They boil down roughly to one big opinion, plus random noise. Numerical ratings of “how liberal” or “how conservative” a congressman is often accurately predict his votes.

The same is true for the general public. Partisan voting is prevalent, suggesting that the public and elites use a similar ideological framework.

Overall, though, one overriding fact about public opinion is that it is far less multi-dimensional than you might guess.

All else equal, the following predict greater agreement with economists: • Education • Income growth • Job security • Male gender

Consistent with the failure of self-serving bias and ideological bias to account for the lay-expert belief gap, income level and ideological conservatism do not make the list.

Failure to vote is a major leak in the pipeline between public opinion and public policy. Politicians only need the support of a majority of people who exercise their right to vote. If they can win the affection of one voter by alienating a thousand nonvoters, competition spurs them to do so.

This would not affect policy if voters and nonvoters had the same distribution of preferences and beliefs, but voters are not a random sample. The most visible difference is that voters are richer than nonvoters. On closer examination, income is largely a proxy for education; education increases both income and the probability of voting. The other big predictor of turnout is age; the old vote more than the young.

Good intentions are ubiquitous in politics; what is scarce is accurate beliefs.

The pertinent question about selective participation is whether voters are more biased than nonvoters, not whether voters take advantage of nonvoters.59 Empirically, the opposite holds: The median voter is less biased than the median nonvoter.