Just over twenty years ago Francis Fukuyama declared liberal democracy the end of history. But history marched on, revealing rot in democracy’s roots. Around the world, from radical leftists in Venezuela and Greece to American Trump supporters, bitter voters wave their banners around populist demagogues. Nationalist movements, echoing those that lead to the first world war, are on the rise. The working classes reject globalization, immigration and economic liberalism. The United Kingdom voted to leave the European Union, and other countries may soon follow suit. In the United States, the political parties are more polarized than ever before, with the most right-wing Democrat to the left of the most left-wing Republican. As a result, the United States faces gridlock and tribal politics rather than compromise solutions.

These movements are driven by low-information voters and the politicians who serve them. The past few decades have been perhaps the best in human history, with more people around the world rising out of absolute poverty than ever before. But many Western voters, ignorant of the social sciences or even of basic political facts, see change all around them, feel left behind and neglected, and strike out in fear and resentment.

When we take a close look at the science of voter behavior, we should not be surprised to see democracy producing poor results on occasion. What’s surprising is that democracies do not fare even worse.

Democracies contain an essential flaw. By spreading power out widely, they remove any incentive for individual voters to use their power wisely. In a major election or referendum, individual voters have no greater chance of making a difference than they do of winning Powerball. They have no incentive to be well informed. They might as well indulge their worst prejudices. Democracy is the rule of the people, but entices people to be their worst.

What if there were an alternative? In my forthcoming book Against Democracy, I describe a new system of government called epistocracy. Epistocracy is meant to do what democracy does well, but guard against democracy’s downsides.

In a democracy, every citizen automatically receives an equal basic right to vote and run for office. Most modern democracies are republican democracies, containing checks and balances, with judicial review, constitutional constraints, multicameral legislatures, contestatory forums, bureaucratic autonomy, political parties and the like, all intended to slow down politics, prevent majoritarianism and protect minority interests.

Epistocracies retain such structures. The essential difference is that in an epistocracy, the right to vote is apportioned, to some degree, according to knowledge. An epistocracy might grant everyone the right to vote, but weigh some votes more than others, or more might exclude citizens from voting unless they can pass a basic test of political competence.

Democracy is the official religion of the West. Now is as good a time as any to question the faith.

Democratic Triumphalism

Most Westerners, left and right, embrace what I call Democratic Triumphalism. Triumphalism’s slogan is, “Three cheers for democracy!” It holds that democracy is a uniquely just form of social organization. People have a basic right to an equal fundamental share of political power. Participation is good for us—it empowers us, it’s a useful way for us to get what we want and it tends to make us better people.

Against Democracy attacks Triumphalism. Democracy does not deserve at least two of those cheers, and might not deserve the third, either.

I argue that political participation is not valuable for most people: it does most of us little good, and participating in politics tends to make us mean and dumb.

I argue that citizens don’t have any basic right to vote or run for office. The right to vote is not like other liberal rights. A right of free speech gives a citizen power over herself; the right to vote gives her power over others.

Democracy, I argue, is not an end in itself. It has the kind of value a hammer has. It’s just a useful instrument for producing just and efficient policies. If we can find a better hammer, we should use it. Indeed, epistocracy may be a better hammer. Perhaps a liberal republican epistocracy might outperform liberal republican democracy. It’s time to experiment and find out.

A Crash Course in Voter Behavior

Political scientists, psychologists and economists have studied voter behavior for over sixty years. They’ve conducted thousands of studies and amassed a huge amount of data. Their findings are largely uniform and depressing. In general, voters are ignorant, misinformed and biased. But there is tremendous variance. When it comes to political information, some people know a lot, most people know nothing and many people know less than nothing.

During election years, most citizens cannot identify any congressional candidates in their district. Citizens generally don’t know which party controls Congress. During the 2000 U.S. presidential election, while slightly more than half of all Americans knew Gore was more liberal than Bush, significantly less than half knew that Gore was more supportive of abortion rights, more supportive of welfare-state programs, favored a higher degree of aid to blacks or was more supportive of environmental regulation. When asked to guess what the unemployment rate was, the majority of voters tend to guess it is twice as high as the actual rate.

And so on. In general, voters in most countries can identify the incumbent chief executive, but know little else beyond that.

These kinds of surveys overstate how much knowledge citizens have, in part because they only ask easy questions, such as who the incumbents are or whether crime is falling. But democracies ask citizens to choose among political parties offering different policy platforms. To evaluate these platforms, citizens need at least some grasp of economics and political science. There’s little reason to think they are informed about these things. On the contrary, American voters, both left and right, have systematically different beliefs about the economy from professional economists, and these differences are not explained by demographic factors.

Citizens aren’t just ignorant or misinformed, but irrational. Few citizens process information with an open mind; most citizens disregard any information that contradicts their current ideology. Voters suffer from a wide range of biases, including confirmation bias, disconfirmation bias, motivated reasoning, intergroup bias, availability bias and prior attitude effects.

It’s no surprise that most voters are ignorant, misinformed and biased. Our individual votes make no difference. When it comes to politics, smart doesn’t pay, and dumb doesn’t hurt.

An individual vote for the worst possible candidate produces the same results as a vote for the best possible candidate. Abstaining from voting produces the same results as voting. A well-informed vote produces the same results as a badly informed, misinformed or irrational vote. An individual vote after careful deliberation produces the same results as voting after flipping a coin or dropping acid.

Information matters. Which policies people prefer depends in part on how informed they are. Even controlling for the influence of sex, race and income, highly informed citizens have systematically different policy preferences from ignorant or misinformed voters. For instance, high-information voters favor free trade, globalization, immigration and civil libertarianism. Low-information voters, regardless of their demographics, favor the opposite: they tend to favor Trump’s platform.

Political Liberty: Who Needs It?

The democratic faith holds that the right to vote is the most important right of all. On reflection, it’s a strange view. Consider: your rights to choose an occupation, to control your sex life, to choose what and when to eat, or to buy and sell as you desire, give you significant control and autonomy over your own life. In contrast, your right to vote does you little good.

Most people talk as if the right to vote has major instrumental value. They say your right to vote allows you to consent to government, to control and shape political outcomes, and to protect yourself from being dominated by others.

But none of this withstands mathematical scrutiny. How we vote matters; how any one of us votes does not. Casting an individual vote has roughly the same power over political outcomes as praying to Jupiter or blowing one’s nose. Democracy empowers the majority, but it does not empower any of the individuals who form that majority.

The probability that your individual vote will change the outcome of a major election or referendum is roughly on the order of the probability you will win the Powerball. Winning the lottery is worth hundreds of millions, but it still doesn’t make sense to buy a ticket. So it goes with voting. Imagine Trump promises to pay you $10 million if he’s elected. Though his victory would net you $10 million, it’s not worth the effort to vote for him, any more than it’s worth buying a Powerball ticket.

Many people understand that individual votes matter little. They instead invoke the symbolic value of the right to vote. In Western democracies, we treat the right to vote as a metaphorical badge of dignity and equality. We imbue people with the equal right to vote in order to express that they are full and equal members of the national club. Many philosophers believe that democracy necessarily expresses that all citizens have equal worth.

This widely held view is odd. Democracy is not a poem or a painting. Democracy is a political system. It is a method for deciding how and when an institution claiming a monopoly on legitimate violence will flex its muscles. Government is supposed to protect the peace, provide public goods and advance justice. It’s not in the first instance an institution intended to boost, maintain or regulate our self-esteem.

Political theorist and British MP Auberon Herbert said, “The instinct of worship is still so strong upon us that, having nearly worn out our capacity for treating kings and such kind of persons as sacred, we are ready to invest a majority of our own selves with the same kind of reverence.” In feudal times, we regarded the king, in virtue of holding power, as possessing a kind of majesty. In a democracy, we instead imagine every voter, in virtue of sharing what was the king’s power, as possessing that same majesty. But there’s no obvious reason why we should think that way.

Instead of viewing a president or prime minister as majestic, we could just regard her as the chief public-goods administrator. Instead of viewing the right to vote as signaling that a person is an equal and valued member of society, we could regard it as possessing no more status than a plumbing or hairdressing license. Or instead of considering that such rights signify membership in the national club, people could just regard these rights as licenses—no different from driving, hairdressing or plumbing licenses.

Here’s the dilemma: suppose epistocracy tends to perform better—to produce better, more just, more efficient outcomes—than democracy. We could conclude that, nevertheless, epistocracy “expresses” contempt, and so have deal with suboptimal government in order to protect people’s feelings. Or we could conclude that treating the right to vote as a badge of dignity is silly, and instead pick the system that works better.

The Injustice of Incompetent Government

Democracies do not just choose mundane things like flag colors or national anthems. They decide matters of peace and war, prosperity and poverty, growth or stagnation.

When a democratic majority picks a policy, this is not akin to you picking a sandwich from a menu. When the majority chooses, it chooses not only for itself, but for dissenting voters, children, foreigners, nonvoters and others who have no choice but to bear the consequences.

Ample empirical research shows that voters are systematically ignorant, misinformed and irrational. That’s not just a bad thing. It might be an injustice.

As an analogy, suppose a jury were deciding a capital murder case. But suppose instead of carefully considering the evidence, the jury found the defendant guilty out of caprice or malice. Suppose a third of jurors paid no attention to the evidence, and just decided, by coin flip, to call the defendant guilty. Suppose another third decided to find the defendant guilty because they dislike his skin color. Suppose the final third paid attention to the evidence, but found the defendant guilty not because the evidence suggested he was, but because they subscribed to a bizarre conspiracy theory.

If we knew a jury behaved that way, we’d demand a retrial. The defendant’s property, welfare, liberty and possibly life are at stake. The jury owes the defendant and the rest of us to take proper care in making its decision. It should decide competently and in good faith.

This line of reasoning applies even more strongly to the electorate as a whole. Political decisions are high stakes. The outcomes—including all ensuing laws, regulations, taxes, budget expenditures, wars, and so on—are imposed upon us involuntarily. These decisions can and so harm us, and can and do deprive many of us of property, liberty and even life. At first glance, we should think that voters, like jurors, have a moral obligation to vote in a competent and morally reasonable way. But when we look at actual voter behavior, it seems like they systematically violate this obligation.

Forms of Epistocracy

In a democracy, every citizen receives an equal basic share of political power. It’s a small share indeed. In an epistocracy, some citizens have greater voting power than others. Each individual citizen at most receives only small share. What makes epistocracy different—and why it might perform better—is that it reduces the power of the least informed.

Democracy tends to prevent citizens from dominating one another because it spreads out power widely. But this requires that literally every citizen have equal power. An epistocracy could produce the same results so long as it avoids concentrating power in just a few hands.

Don’t confuse epistocracy with technocracy. When people talk about technocracy, what they usually have in mind is a cadre of experts who use government to manage the citizens and engage in massive social engineering projects. Technocracy is not so much about who rules but about how they rule and what they do. Many democrats advocate technocracy, and an epistocrat can reject it.

Don’t confuse epistocracy with totalitarianism. Totalitarianism isn’t about who rules, but what they rule. Totalitarian governments stick their noses in everything. Liberal governments leave many issues off the political bargaining table.

Any reasonable form of epistocracy will spread power out among rather than concentrate it. Any reasonable form will retain all the republican checks and balances. No modern epistocrat advocates the rule of philosopher-kings. Instead, the reasonable forms of epistocracy, those worth considering, include:

Restricted Suffrage: Citizens may acquire the legal right to vote and run for office only if they pass a test of basic political knowledge.

Plural Voting: As in a democracy, every citizen has a vote. However, some citizens, such those who who pass a test of basic knowledge, or who meet some other criteria correlated with political competence, can acquire additional votes.

Epistocratic Veto: Just as in a democracy, all laws are passed by a democratic legislature elected through universal suffrage. However, an epistocratic body with restricted membership retains the right to veto rules passed by the democratic legislature. Just as judges can veto legislation for being unconstitutional, so, perhaps, a board of economic advisors might have the right to veto legislation (such as protectionist policies) that violate basic economic principles.

Weighted Voting: During the election, every citizen may vote, but must at the same time take a quiz concerning basic political knowledge. Their votes are weighted based on their objective political knowledge, all while statistically controlling for the influence of race, income, sex and/or other demographic factors. With such data (which will be made public), any statistician can then calculate or estimate, with a high degree of certainty, what the public would want if only it were informed. The epistocracy does what the informed public would want, rather than what the uninformed public in fact wants.

The big question, of course, is what counts, and who decides, political competence or basic political knowledge. I’m less troubled by this question than many. We could just use the type of questions we’ve been using on the American National Election Studies. We could use the questions we’ve been using on the American citizenship exam. These are easy, objective, easily verified questions, but we have good grounds to think that the capacity to answer them is correlated with the kind of social scientific knowledge that really matters.

One somewhat paradoxical-sounding, but surprisingly reasonable, idea is that we could use democratic procedures to choose a public definition of political competence, which we in turn use to selected epistocratic voters. For instance, imagine that to vote for president, one must pass a “voter qualifying exam,” but then imagine that this exam itself was selected through a democratic vote. This may seem strange—if democracies are competent to choose a legal definition of competence, why aren’t they also competent to choose a president? But there are two reasons why this is less paradoxical than it sounds. First, the problem with democracy is not that citizens fail to understand, in the abstract, what counts as a good president. Rather, they have good abstract standards, but they are bad at applying their standards, at selecting a person who meets them. Second, the question “What counts as political competence?” is a much easier question than, say, “Should we have free trade or protectionism?” The latter question requires social scientific knowledge most voters lack, but the former question does not.

Conclusion: The Better Hammer

There’s no doubt that in the real world, any epistocratic system would suffer government failures and abuse. But the same goes with democracy. In the real world, special-interest groups would try rig both systems for their benefit at the expense of everyone else. In the real world, both epistocracy and democracy will be imperfect and flawed. The question we should ask is which system would work better

Governments are like hammers, not poems. The point of a government is to produce good outcomes. Democracy has had a good run. But it has an endemic design flaw. It’s time to experiment with a new system, to see if we can improve upon the design.

Jason Brennan is the Robert J. and Elizabeth Flanagan Family Associate Professor of Strategy, Economics, Ethics and Public Policy at the McDonough School of Business at Georgetown University. This article was adapted from his new book Against Democracy (Princeton University Press), released September 7, 2016.

Image: Voting booths in Cleveland Heights, Ohio, for the general election held on November 3, 2015. Flickr/Tim Evanson