Bryan Caplan  

Debate: Does Democracy Work?

Atkinson and Krugman on Tax Ra... Parental Economics and Risk: A...

Last week I did an online debate for Learn Liberty with philosopher Helene Landemore.  The topic: Does democracy work?  Here's my opening statement.

Democracy clearly works if you set the bar low enough.  Is democracy better than dictatorship?  Of course.  Does democracy allow most people in the First World to live long, comfortable lives?  Sure.  But we now hold most of our social institutions to far higher standards.  If 90% of women survived childbirth, we wouldn't say "Medicine works."  We'd expect doctors to use everything they know - and constantly strive to learn more.  And if mothers were dying because doctors stubbornly clung to superstitious treatments, we judge the doctors very harshly indeed.

So what would we conclude if we held democracy to analogous standards?  Do democracies use everything we know?  Do they constantly strive to learn more?  Do they at least avoid acting on sheer superstition?  I say the answer is no across the board.  When we actually measure voters' policy-relevant beliefs against reasonable proxies for the Truth, voters do poorly.  Democracy's defenders often insist that these errors will harmlessly balance out, but the facts of the matter is that voter errors are usually systematic.  Voters err alike. 

What "reasonable proxies for the Truth" do voters fare poorly against?  They fail against objective statistics.  For example, American voters vastly overestimate the share of the budget that goes to foreign aid and welfare, and vastly underestimate the share that goes to Social Security and health.  They fail against better-informed voters.  Voters who score higher on tests of objective political knowledge favor very different policies than otherwise identical voters who score poorly.  And they fail against expert consensus - even after you adjust for possible biases in the experts' judgments.

I'm an economist, so I've focused a lot of my attention on the public's economic illiteracy.  Despite popular stereotypes about economists' failure to agree with each other, there are many contrarian conclusions that economists across the political spectrum share.  Regardless of ideology, for example, economists have a strong tendency to support free trade, oppose price controls, and favor more open immigration.  The main reason is that they actually know surprising, unpopular facts about trade and markets: Specialization and exchange enrich the world, even if one trading partner is better at everything; minimum prices create surpluses; maximum prices create shortages.  Economists are often distressed by how little economics college students learn and remember; but the average college student knows far more economics than the typical voter.

I do think that democracies are very responsive to public opinion.  Given the facts about public opinion, however, responsiveness is greatly overrated.   When the public holds systematically biased beliefs, politicians who want to win have to pander to popular error.  And that's precisely what happens in every major election: Pandering.

Couldn't we solve this problem with better education?  I'd like to believe that, but the facts once again get in the way.  "Educating" people out of their policy beliefs is very hard.  Why?  In large part, because error is, selfishly speaking, free.  If a voter is intellectually lazy, what happens to him?  The same thing that happens to people like you who voluntarily attend online debates on "Does Democracy Work!"  This contrast is easy to see when you offer to bet someone about his policy views: Even passionate ideologues usually decline to back up their extravagant claims with cold hard cash.  As I explain in The Myth of the Rational Voter, we shouldn't think of democracy as a market where people buy the policies they like.  We should instead think of democracy as a common well where people throw their intellectual garbage, heedless of the fact that we all drink the water.

What can be done?  For starters, we need to learn from Alcoholics Anonymous: The first step is admitting that you have a problem.  The fact that most people vote for X doesn't make X a good idea.  We also need to take a more favorable view of constitutional checks on democracy.  But above all, the evidence is a strong argument for relying less on government and more on markets.  Economists may know a lot of ways for government to improve on market outcomes; but thanks to misguided public opinion, actual government policies tend to make market outcomes worse than they'd be if government left well enough alone.

COMMENTS (24 to date)
Chris Koresko writes:

@Bryan Caplan: That was very well spoken. Your argument about the perverse incentives faced by voters (I heard you make it in more detail on EconTalk a while back) was persuasive and frankly rather disturbing.

I'd like to hear how you elaborated on some of the ideas in your opening statement, and especially on some of the results you cite on correlation between voters' preferred policies and their knowledge of objective facts.

Is there a link to a video? Thanks if so.

Besides increasing the role of markets, consider demarchy as a representative system. I'm quite convinced it wouldn't be worse than democracy.

HM writes:

Maybe it would be a good idea to replace some of the representative aspects of democracy with a jury system. An assembly of 1000 randomly selected people could achieve an acceptable degree of representativeness while still preserving some visibility and leverage for single individuals (or at least groups). On the incentive part it would definitely be an improvement.

Matthew Gunn writes:

To answer whether Democracy works or not requires defining what Democracy is or isn't supposed to do.

Democracy has been a fantastically successful solution to a classic coordination problem: how do you keep the hegemon that provides security from taking everything?

By investing decision making authority in the majority, Democracy puts a type of lower bound on the worst case policy. Convincing 50 percent of the population of something stupid may be commonplace, but selling a plurality on a truly awful policy that shocks the conscience is difficult!

By this measure, Democracy is an unrivaled success. But if one looks to Democracy for a kind of oracle, expecting public opinion polls or voting to deliver God like truth, then Democracy is a colossal failure. As you argue in the Myth of the Rational Voter, voters make mistakes. And these mistakes stem from inherently poor incentives.

My view:
Democracy is about avoiding awful decisions, not making good ones.

You want welfare maximizing, efficient markets to make decisions wherever possible, but in some cases (eg. administration of police) where there is the potential for fantastic abuse of power, you need Democracy.

Greg G writes:

As always, democracy works better than any of the real world alternatives but not nearly as well as a fantasy libertarian world.

Libertarians rightly object when they are caricatured as thinking markets will solve all problems. Their real argument is simply that it (their particular version of libertarianism) is better than the alternatives. Let's apply that same logic to constitutional democracy.

Compared to all other forms of government in human history, constitution democracy repeatedly provides the highest levels of freedom and prosperity. That is not such a "low bar" despite the many problems that democracy fails to solve.

The argument in this post also relies on some special pleading about "conclusions that economists across the political spectrum share." In fact, very few highly educated economists are anarcho-capitalists.

Finally, I couldn't help noticing the "we need to " advise in the final paragraph. I often hear libertarians denounce this concept of a collective need. Except when they agree on what that need is.

Daublin writes:

Excellent article and a point well worth making.

I'd quibble with part of your opener: "And if mothers were dying because doctors stubbornly clung to superstitious treatments, we judge the doctors very harshly indeed."

That doesn't ring true with how I view popular opinion, or for that matter the behavior of experts. It's oddballs that get hamemred by public opinion. Going with popular superstitions is a realtively safe bet.

RPLong writes:

Democracy works just fine, but it isn't a magic bullet. I think the main issue is that people are lazy and don't want to have to consider that there is no "one solution" to anything. Every problem requires multiple angles of approach.

For society to flourish, the majority of us need to be well-educated, virtuous, engaged, hard-working, and trustworthy. Any political system is viable, given those prevailing attributes, but Democracy is uniquely well-suited to them. But it's not a replacement for them.

Unfortunately, there is no political system that solves human character flaws. We have to apply ethics to those problems.

Terry Hulsey writes:

Without a limitation on the scope of democracy, this Republic will fail. Only my Amendment to Randomize Elections will accomplish this:

My nice, scholarly name for this political system is "klerotocracy":,%20subs.

R Richard Schweitzer writes:


Read The Servile Mind by Kenneth Minogue.

Democracy is a process, not a condition.

Does the process "work?"

In this discussion it seems that the answer should turn on whether or not the process achieves an objective.

Another perspective might be whether or not the process is effective in selecting objectives.

So the question really turns on what kind of "work" is to be expected of the process.

Andre Mouton writes:

The role that religious, and then scientific expertise has played in this country, would suggest that the public isn't qualified to vote on most of the issues that come before Congress, and knows it. Our representatives aren't any more capable, and in the end it's experts and functionaries who write the law -- the same scholastics who would have written it under a monarchy. Classical liberalism "worked"; democracy rode its coattails.

Philo writes:

Your positive proposal seems to be democracy *constitutionally limited to a small range of issues*, with all other collective decisions made through markets. (You don’t specify the range of issues to be decided democratically.) But in view of the misguided views of most people your proposal seems impracticable: it couldn’t be made sufficiently appealing to most people to have a chance of being implemented, or of being long continued if (miraculously) it were implemented.

Cimon Alexander writes:

Dictatorship seemed to work pretty well for Singapore and Chile. Both countries would be much worse off if leaders had to cater to the whims of uneducated mobs. Moreover, Singapore is an ethnically diverse country, and democracies in ethnically diverse countries always fall into pandering to ethnic prejudices.

Glen S. McGhee writes:

Still doesn't address the scaling-problem of governance.

How can we justify a yeoman-based, small-group democracy -- with its short turn-around times, short enough to see the results of decisions that YOU helped to made -- for modern bureaucratic life?

Only by ignoring the problem, I think.

It is not a matter of "what works" and "what doesn't work," but how and why a system of governance devised so long ago would still be applicable.

In fact, you could argue that what we have now bears little or no resemblance to what Founding Fathers conceived. The system we have dwarfs the individual; the individual no longer matters on the scale it did earlier. All that has dramatically changed.

MOF writes:

Markets work no better and no worse than democracy. That is because markets are democracy. That is why Walmart is a winner in retail, millions of Americans love it. That is why we had a subprime crisis. Millions of Americans loved cheap rates on nice houses.

Terry Hulsey writes:

I am reminded of the opening scene of The Last Meeting of the Knights of the White Magnolia, in which a black janitor overhears a problem stated by the old white men in the room, and suggests a patently obvious and perfect solution. They don't acknowledge his existence, and keep talking.

I can't imagine why this comes to mind just now. Also I think to myself, "Have I flossed recently?" and "Did I feed my guppies today?" and "Do have have any guppies?" and "That cloud up there: Does it look more like a hawk or more like a handsaw?" A dog barks in the distance: A happy, greeting kind of bark.

MingoV writes:
Is democracy better than dictatorship? Of course.
Show your proof. Some benign dictatorships functioned better than democracies. A dictatorship can be the most efficient form of government. My preferred government would be libertarian, and the government would be run by an elected dictator with a ten year term. The dictator could be removed at any time if 75% of citizens vote to depose. A minimalist libertarian government with the low bureaucratic overhead of a dictatorship could be the best government.
Julien Couvreur writes:


That is because markets are democracy.

No. In the market, there is no assumption of a monopoly provider, whereas politics presumes a uniform policy.
Instead of letting Republicans shop at Republican Co and Democrats shop at Democrats Inc, they have to fight to steer a single provider in the direction that fits their preferences.

Moreover you could shop at Walmart for food, but get clothes from somewhere else. But government forces a bundling (you can only steer food and clothes together, not independently).

R Hamel writes:

@Matthew Gunn:

The lower bound is rather low indeed, when one looks at the democratically-elected regime of Adolf Hitler among others.

Alas, for the most part it seems that the stupider the idea, the more convincing it is to 50% of the electorate.

John writes:

Voters don't decide policy so we have no way of to link outcome to decision-making process using voting or voter preferences. The other thing to note is that, at least in the USA, it's fair to say that nearly all of the representatives and certainly the president are elected by, at best, a large minority of the eligible voting population.

Also, since we don't really have a good metric for measuring performance against some defined output -- the output is very complex and in some cases mutually exclusive options across the evaluating population -- that the exercise is pointless. We would be able to reject ALL political arrangement including anarchy.

This is no different than the problems claiming social welfare optimality based on economic efficiency -- we have to assume a constant marginal utility of a unit of money and the equality of that unit of utility. It just doesn't exist in the real world.

Brent writes:

In other words, "political systems" don't matter. It is the idea of using an agency of force to run everyone's lives that gets us into trouble. A great first step would be to make the size of those agencies much smaller, both geographically and population-wise.

steve writes:

Until you have a viable alternative, democracy is the best we have. A noun, a verb and "markets" does not make an argument.


Danny writes:

You can't have a democracy without constituents but you can have an economy without economists.

Carl writes:

Greg G wrote:

The argument in this post also relies on some special pleading about "conclusions that economists across the political spectrum share." In fact, very few highly educated economists are anarcho-capitalists.

You didn't read Caplan's piece properly. He did not claim that all economists shared his political philosophy.

Here's what he actually said:

Regardless of ideology, for example, economists have a strong tendency to support free trade, oppose price controls, and favor more open immigration.
A Country Farmer writes:

An amazingly powerful summary of the problems with Democracies. Well done.

Comments for this entry have been closed
Return to top