Bryan Caplan's book could only have been written by an economist. This has a good side and a bad side. On the good side, Caplan makes an important contribution to the public-opinion literature by analyzing the American people's economic views—an area overlooked by most previous researchers. And he asks a crucial question: Are the public's economic views wrong?

The heart of the book is Chapter 3, where Caplan organizes 1996 survey data on economic attitudes into four “biases”: antiforeign bias (opposition to free trade); antimarket bias (suspicion of the profit motive); make-work bias (favoring public-works programs); and pessimistic bias (underestimating how well the economy is doing). By “biased,” Caplan means that the public's views are systematically erroneous. His data do show systematic—nonrandom—opinions. But when he tries to prove that these opinions are erroneous, his economism works against him.

Caplan's astonishing metric for the truth about free trade, and so on, is the consensus view on these matters held by people with economics PhDs. By showing that they disagree with the public, Caplan claims to have established that the public is wrong. This is an argument from authority, logically invalid on its face: It runs into the obvious rejoinder that even economists can be biased.

The author easily dispatches weak versions of this rejoinder: that economists are biased because they are rich, or because they are conservatives. But he does not mention a long line of critics who hold that economists are biased by a methodological dogma: rational choice theory (RCT).

Still, bracketing the question of whether economic “experts” are right, one should certainly stipulate that the public could be systematically wrong—about economics or anything else. The question to which Caplan then turns is: What might cause systematically erroneous opinions? He thinks that his biggest contribution is his answer to that question: his theory of “rational irrationality” (Chapter 5). Caplan wants to make room within RCT for the rational choice of “irrationality”—a choice that he views as responsible for voters' economic errors.

The usual RCT explanation for voter error is “rational ignorance” (Chapter 4). The average voter supposedly knows that a lone vote probably will not determine an electoral outcome; so the voter chooses not to invest resources in becoming politically well informed. A rational-ignorance theorist would similarly attribute voters' economic errors to their low incentive to become economically well informed. Total ignorance of economics, however, would produce a random scattering of economic views, not systematic patterns such as those displayed in Chapter 3. To explain these nonrandom “errors,” Caplan contends, we need another independent variable: not ignorance but irrationality. Thus, according to the author, the public's economic opinions are not merely uninformed; they are the product of “emotions or ideology” (p. 14 and passim). Indeed, they are like religious dogmas, held so “fervently, dogmatic[ally], and fanatical[ly]” (p. 15) that attempts to dispute them provoke anger.

But Caplan provides no evidence that the public is fanatical, or even stubborn, about its economic opinions, nor that the opinions held by the public “feel” better (p. 2) than the opinions economists hold. Nor does he show that voters know the high odds against individual votes mattering (nor does he explain why, if they knew that, they would vote anyway), nor does he show that they therefore (rationally choose to) “turn off” their “rational faculties,” and indulge their “emotions or ideologies,” because they “don't care about the truth” (p. 2).

Instead of proving that (unspecified) emotions or ideologies explain the public's economic views, Caplan simply assumes it. Why else would the public be so “crazy” (p. 69) as to disagree with economics “experts”? His only actual argument for mass economic irrationality, however, is that it can provide an RCT explanation of nonrandom mass opinion. He gives no reason to believe that rational irrationality “exists”—except as a solution to a problem in the theory of rational choice.

However, the fact that Caplan fails to prove that rational irrationality exists in the real world does not mean that there is nothing in the real world to be explained. Emotion and ideology, not to mention error, are plentiful in politics, amply deserving the attention he pays them. But to explain them as rational choices, he has to assign them to the mass level, where they can be seen as corollaries of the average voter's insignificance.

Caplan references, but does not seem to have understood, Philip Converse's “The Nature of Belief Systems in Mass Publics” (David E. Apter, ed., Ideology and Discontent, 1964), which showed that ideology tends to be the province of the well-informed few, not the ignorant masses. More recent political-psychology literature on “motivated reasoning” suggests that people cling dogmatically to their beliefs—by using their knowledge to fend off discrepant information. Thus, “ideological constraint” would be likelier among relatively knowledgeable elites than relatively ignorant masses—as Converse found.

Elite dogmatism, however, fits uncomfortably with RCT. The politically knowledgeable few would tend to have greater incentives than typical voters to avoid stubbornly clinging to what might be mistaken ideas, because all things being equal, the politically knowledgeable would likelier be politicians, bureaucrats, political activists, terrorists—even economists or political scientists—who influence politics much more than average voters do. Caplan would predict that influential elites are less ideological than average voters, yet at least with knowledgeable elites, the opposite holds. They seem inclined toward dogmatism regardless of the incentive to keep an open mind (cf. Tetlock, Expert Political Judgment, 2006).

But then, the world they are trying to understand is complicated. In confronting complexity, people err inadvertently (and nonrandomly, for cognitive and cultural reasons): not by choice, but by accident. And the more misleading information people inadvertently accumulate, the easier it is to fend off discrepant information. Here, dogmatism is unintentional. In RCT, though, everything—even “irrationality”—is intentional. Extracting from this hyper-rationalism a plausible theory of mistake may simply be impossible.

These days, understanding political error and dogma is especially crucial. But judging from The Myth of the Rational Voter, that objective might best be advanced by leaving RCT behind—no matter how dogmatically economists cling to this particular error.