The Myth of the Rational Voter Review: When Democracy Gets What It Wants
In The Myth of the Rational Voter: Why Democracies Choose Bad Policies, Bryan Caplan attacks one of democracy’s most comforting explanations for failure: the public wanted something sensible, but lobbyists, politicians, bureaucrats or the media intercepted it. His more unsettling claim is that democratic systems often produce bad economic policy because they respond to voters rather than betray them.
Bad policies are not always imposed on a reluctant electorate. Some remain popular because they fit intuitions people find morally satisfying: imports destroy prosperity, labour-saving technology destroys wealth, prices mainly reflect greed, and economic change is usually decline in disguise. Democracy aggregates preferences, but it also aggregates beliefs about how the world works. When those beliefs are systematically mistaken, faithful representation does not guarantee a sensible result.
Caplan’s argument is deliberately abrasive, and sometimes too neat. It places great confidence in economists, gives less attention to elite error than voter error and treats democracy more as a policy-selection mechanism than as a system of legitimacy and peaceful conflict. Even with those limits, the book identifies a pathology that democratic romanticism prefers not to see. Public opinion can be sincere, public-spirited and wrong at the same time.
Voters Are Wrong in Patterns
I picked up The Myth of the Rational Voter partly because I had enjoyed several of Caplan’s other books, and partly because I already suspected that political ignorance was not the whole problem. Ignorance is absence; error has architecture. A voter who knows little about trade may remain uncertain. A voter who knows that imports impoverish the country, that automation destroys prosperity or that high prices prove exploitation already possesses a model of the economy. The room is not empty. It has a tenant.
Caplan tries to describe that tenant through four recurring biases. Anti-market bias underestimates the coordinating role of prices, profits and voluntary exchange. Anti-foreign bias sees trade and migration chiefly as competition in which outsiders take something from insiders. Make-work bias confuses the preservation of labour with the creation of value, treating effort as an economic end rather than a cost that productive societies try to reduce. Pessimistic bias assumes that economic conditions are worse, more fragile and more steadily deteriorating than the evidence warrants.
These categories are memorable because they describe intuitions that appear repeatedly in political argument. Protectionism feels like solidarity because its beneficiaries are visible and its wider costs are dispersed. Automation can look like destruction because the displaced job is easier to see than the lower cost, new product or different work that may follow. Housing shortages can coexist with opposition to building because abundance is desired in general while each concrete change acquires a local opponent. These extensions are not all evidence from Caplan’s book, but they show why his framework travels.
The empirical core is much narrower than the rhetoric sometimes suggests. Caplan relies heavily on the 1996 Survey of Americans and Economists on the Economy, which compared responses from 1,510 members of the American public with those of 250 economists specialising in domestic policy. The two groups answered many of the same questions, allowing Caplan to examine whether disagreement was random or moved in consistent directions.
In a 2002 analysis of the survey, Caplan argued that the public–economist gap persisted after accounting for income, political affiliation, ideology and general education. That is stronger than simply observing that experts and non-experts disagree. It is not conclusive proof that economists are always right, nor does one American survey from the 1990s establish a universal psychology of democracy. The dataset is old, national rather than global and filtered through the questions its designers chose to ask. The book is most persuasive when it treats the survey as evidence for patterned disagreement, not as a licence to convert every public preference into a cognitive defect.
Why Error Is Cheap in Politics
Caplan’s most useful concept is “rational irrationality.” The phrase describes a difference in incentives rather than a claim that voters lose the ability to reason when they enter a polling station. People exert more effort to correct beliefs when error imposes a direct and visible cost. A bad mortgage, an unreliable contractor or a neglected warning light eventually sends a personal invoice. Feedback may be delayed or noisy, and markets do not eliminate delusion, but many private decisions give the decision-maker a reason to notice failure.
Political belief works differently. One ballot is almost never decisive, while the emotional rewards of a satisfying worldview arrive immediately. The voter can enjoy the belief that foreigners are stealing prosperity or that a painless policy can deliver incompatible goals without personally determining the outcome. If the policy is adopted, its costs are distributed across millions of people, entangled with other causes and often delayed beyond the next election. The individual receives nearly all the psychological benefit of the belief and only a minute share of the resulting damage.
This mechanism does not require crude selfishness. Caplan emphasises that voters may be sociotropic: they may vote for what they sincerely believe will help society rather than for their own material advantage. That can make the problem worse rather than better. A selfish voter at least knows what private interest is being pursued. A public-spirited voter with a false theory of prosperity can support large sacrifices on behalf of an imagined common good.
The account is compelling because it explains why additional information may not be enough. If error were merely a shortage of facts, education and better reporting would gradually solve it. Rational irrationality suggests that some beliefs survive because they are gratifying. They supply villains, preserve moral innocence or make complicated trade-offs resemble acts of betrayal. Correction then threatens more than a factual proposition. It threatens the story through which the voter understands loyalty, fairness and personal virtue.
When Popular Error Becomes Policy
Caplan’s theory overlaps with public-choice economics but changes the emphasis. The familiar public-choice explanation begins with concentrated benefits and diffuse costs. A small group has a strong incentive to defend a subsidy, tariff or regulation, while each member of the public loses too little to organise against it. Lobbying succeeds because beneficiaries care intensely and everyone else remains rationally inattentive.
Caplan adds a harsher possibility: the wider public may approve of the mistaken theory that supports the policy. Protectionism survives not only because a protected industry lobbies for it, but because tariffs tell a politically attractive story about national loyalty and foreign threat. A make-work programme may remain popular not merely because it employs an organised constituency, but because voters equate visible employment with prosperity. The pressure from concentrated interests is reinforced rather than restrained by popular intuition.
This helps explain why bad policy can persist after its original purpose has weakened. Political systems accumulate subsidies, licensing restrictions, trade barriers, emergency rules and symbolic programmes that few people would design from the beginning. Their costs remain spread across consumers or future taxpayers, while beneficiaries acquire organisations, narratives and identities built around continued protection. Voter bias does not create every piece of this sediment, but it can make removal feel like cruelty, surrender or economic vandalism.
The argument is strongest where the causal chain is clear: a mistaken factual belief supports a policy whose costs are obscured and whose beneficiaries are visible. It is weaker when “bias” becomes a label for any preference an economist dislikes. Citizens may knowingly accept some inefficiency to protect stability, distribution, local autonomy or security. The existence of a trade-off is not proof that voters have failed to understand economics. Caplan’s challenge is to separate factual error from a political judgment that assigns different weights to real costs.
Where Caplan Cuts Too Cleanly
The first danger is treating economists as an uncomplicated benchmark. Economic training can improve understanding of opportunity cost, incentives, trade and prices. Economists also disagree, respond to institutional incentives and make serious errors. Consensus is strongest on some descriptive questions and weaker where evidence is uncertain, models depend on assumptions or policy mixes empirical prediction with moral judgment. Expertise deserves weight; it does not confer jurisdiction over every value conflict.
Caplan also gives insufficient attention to the irrationality of elites. Politicians, civil servants, journalists, academics, business leaders and policy experts face their own incentives to preserve flattering beliefs. A bureaucracy can confuse procedure with effectiveness. A profession can protect its status. A market can sustain a bubble. A model can become elegant enough that inconvenient reality is treated as noise. Voters are not the only people able to enjoy a theory while exporting its costs.
More fundamentally, democracy is not only a machine for selecting economically efficient policies. It distributes political standing, manages conflict and gives citizens a reason to accept decisions made by people they oppose. A system that consistently tells the governed that their judgments are too ignorant to matter may improve some technical decisions while destroying legitimacy. Caplan is persuasive as a critic of democratic innocence; he is much less persuasive if interpreted as an argument that economically literate minorities should simply take command.
These objections limit the book without dissolving its central claim. Public opinion cannot be treated as a moral solvent. Popularity does not turn a false causal theory into a true one, and sincerity does not make the costs of a policy disappear. Democratic legitimacy explains why citizens must retain political power. It does not guarantee that they will use that power wisely.
The Book After the Attention Economy
The Myth of the Rational Voter was published before algorithmic feeds, influencer politics and generative AI became ordinary parts of political communication. These technologies do not invalidate Caplan’s argument. They make the distinction between ignorance and motivated belief more important. A citizen can now obtain more information than any earlier voter while remaining enclosed inside a system that selects for attention, identity and emotional response.
Generative AI sharpens both sides of the problem. It can fabricate local facts, summarise weak sources with unwarranted confidence and turn a prejudice into a polished argument. It can also lower the cost of mundane election research. In “Use AI This Election”, Scott Alexander proposes the modest version: use an AI assistant to investigate unfamiliar offices and candidates, compare positions and endorsements, and translate a complicated ballot into questions the voter can examine.
The value of that proposal depends on verification. A civic research assistant should show sources, distinguish candidate claims from independent evidence, admit when local information is missing and make uncertainty visible. Used that way, AI may reduce the cost of becoming less wrong. Used as an oracle—or prompted until it produces the preferred conclusion—it becomes another instrument of rational irrationality. Cheap information does not help when the real demand is for confirmation.
The book therefore leaves an institutional question it does not answer: how can democracy retain equal political standing while making politically satisfying mistakes harder to sustain? That question deserves more than a final paragraph, and I pursue it separately in “After The Myth of the Rational Voter: Can Democracy Be Made Less Blind?”. The review’s narrower conclusion is that information alone is unlikely to cure a problem partly driven by the rewards of belief.
The Uncomfortable Lesson
Caplan is too confident in places, too economical in his account of political legitimacy and too ready to contrast messy democracy with cleaner markets. Yet the book remains difficult to dismiss because it attacks an excuse used across the political spectrum. When voters support a harmful policy, defenders can always blame manipulation, ignorance or elite betrayal. Sometimes those explanations are correct. They are not complete.
Democratic majorities can want protection without counting its costs, security without accepting its constraints, abundance without allowing production, and reform without tolerating disruption. The resulting policy may express public opinion accurately. That is precisely why it can be hard to change.
The Myth of the Rational Voter is strongest as a refusal of democratic innocence. It asks readers to distrust not only politicians and lobbyists, but the part of themselves that enters politics looking for confirmation, villains and costless virtue. Democracy can fail because powerful institutions ignore the public. Caplan’s contribution is to show that it can also fail because they listen to beliefs voters have little reason to test.
Comments
Post a Comment