It was naughty of Winston Churchill to say, if he really did, that "the best argument against democracy is a five-minute conversation with the average voter."

Nevertheless, many voters' paucity of information about politics and government, although arguably rational, raises awkward questions about concepts central to democratic theory, including consent, representation, public opinion, and officials' accountability.

In "Democracy and Political Ignorance: Why Smaller Government is Smarter" (Stanford University Press), Ilya Somin of George Mason University Law School argues that an individual's ignorance of public affairs is rational because the likelihood of his or her vote being decisive in an election is vanishingly small. The small incentives to become informed include reducing one's susceptibility to deceptions, misinformation and propaganda.

But if remaining ignorant is rational individual behavior, it has likely destructive collective outcomes.

Somin says that in Cold War 1964, two years after the Cuban missile crisis, only 38 percent of Americans knew the Soviet Union was not a member of NATO. In 2003, about 70 percent were unaware of enactment of the prescription drug entitlement, then the largest welfare state expansion since Medicare (1965). In a 2006 Zogby poll, only 42 percent could name the three branches of the federal government.

Voters cannot hold officials responsible if they do not know what government is doing, or which parts of government are doing what. Given that 20 percent think the sun revolves around the Earth, it is unsurprising that a majority are unable to locate major states such as New York on a map. Usually only 30 percent can name their two senators. The average American expends more time becoming informed about choosing a car than choosing a candidate. But, then, the consequences of the car choice are immediate and discernible.

Many people, says Somin, acquire political knowledge for the reason people acquire sports knowledge — because it interests them, not to alter the outcome of any contest. And with "confirmation bias," many people use political information to reinforce their pre-existing views. Committed partisans are generally the most knowledgeable voters, independents the least.

The problem of ignorance is unlikely to be ameliorated by increasing voter knowledge because demand for information, not the supply of it, is the major constraint on political knowledge. Despite dramatic expansions of education and information sources, abundant evidence shows the scope of political ignorance is remarkably persistent over time. New information technologies have served primarily to increase the knowledge of the already well-informed, which increases the ability of some to engage in "rent-seeking" from the regulatory state, manipulating its power in order to transfer wealth to themselves. And if political knowledge is measured relative to government's expanding scope, ignorance is increasing rapidly: There is so much more to be uninformed about.

A better ameliorative measure would be to reduce the risks of ignorance by reducing government's consequences — its complexity and intrusiveness. In the 19th century, voters' information burdens were much lighter because important federal issues were much fewer.

Political ignorance helps explain Americans' perpetual disappointment with politicians generally, and presidents especially, to whom voters unrealistically attribute abilities to control events. But the elections of 1932 and 1980 dramatically illustrated how voters primarily control politicians — by "retrospective voting," refusing to re-elect them.

Some people vote because it gives them pleasure — the satisfaction of expressive behavior — and because they feel duty-bound to cast a ballot that, by itself, makes virtually no difference, but affirms a process that does. And although many people deplore the fact that U.S. parties have become more ideologically homogenous, they now confer more informative "brands" on their candidates.

Political ignorance, Somin argues, strengthens the case for judicial review by weakening the supposed "countermajoritarian difficulty" with it. If much of the electorate is unaware of the substance or even existence of policies adopted by the sprawling regulatory state, the policies' democratic pedigrees are weak.

Hence Somin's suggestion that the extension of government's reach "undercuts democracy more than it furthers it."

George Will's column is distributed by the Washington Post Writers Group.