Against Democracy

I don't talk a lot about democracy on this blog. Frankly, that's because I do not much care about democracy. Let me explain why.

Ideally, democracy is a way of aggregating preferences. It is a way of combining the preferences of a bunch of people into one group preference: Democrat over Republican, or vice versa.

(There are many theoretical problems that arise when you try to combine preferences this way. While these problems are important, my concerns here are more practical in nature.)

Group preferences are thought to have normative significance, above and beyond the significance of the preferences of an individual. It is supposed to matter that more of the people (who voted) prefer candidate X over candidate Y. The will of the people -- as it is called -- is supposed to tell us something about "the way things ought to be."

Now, I find this way of thinking to be completely absurd; not just wrong, but ridiculously wrong. A person prefers X to Y. This may tell us something about the way things ought to be for his own life. Suppose he prefers to eat peaches over bananas. We might say: then he ought to eat a peach and not a banana. His preference for one alternative over another provides him with reasons to pursue some courses of action over others.

One might think that democracy simply extends this reasoning. Surely, if 60% of the people prefer eating peaches, that ought to have some significance. And it does: it provides sellers of fruit with reasons to stock more peaches than bananas (perhaps.) But ought it to have any significance for the way the other 40% of people live their lives? Not obviously. For the minority has its own preferences, and it ought to pursue them. Why should the minority care that the majority prefers peaches to bananas?

The point I am trying to make is the fact that 60% of people prefer peaches does not make peaches better than bananas, in any objective sense. It does not provide anyone with a reason to eat peaches he did not have before. The will of the majority is normatively impotent.

There are two kinds of responses to the criticism I have raised. One response is that, yes, sometimes, the fact the majorty prefers X over Y does make X better than Y. The other response is that, while majority preference doesn't make X better than Y, it does provide evidence that X is better than Y. I will deal with these responses in order.

1. Does majority preference for X over Y make X better than Y?

Answer: no. One consideration in support of this answer is that X's properties do not change just because 50%+1 find X preferable to Y. If X is truly better than Y, it should have something to do with the nature of X and Y. But majority approval is an exogenous factor with respect to X and Y. If there really is good reason to support X instead of Y, that reason exists independently of majority preference.

Another consideration in support of a negative answer to the question: democracy aggregates preferences. But it says nothing about the basis of those preferences. The democratic process is blind to the reasons why people prefer X over Y. Indeed, in principle, people could prefer X over Y for no reason whatsoever, or for reasons that are (again) exogenous with respect to X and Y.

Bottom line: in itself, the democratic process cannot make X better than Y, if the former isn't already better (and thus, more desirable) than the latter.

2. Does majority preference provide evidence that X is better than Y?

Answer: no. In this context, democrats usually cite Condorcet's jury theorem. But the jury theorem has little application to this issue for several interelated reasons, including:

(a) The jury theorem requires that the people on the jury have a better chance of being right than being wrong. If the reverse is true -- if they're more likely to get it wrong than get it right -- then the theorem works in reverse, and the group as a whole will almost certainly make the wrong decision.

(b) Informational cascades: people don't make up their minds independently of one another in a democracy. Bad information gets spread and the usual methods of filtering it out do not function adequately (more on this below.) For example, in the last election, rumors about Obama being a Muslim were widely spread and believed through email. While this didn't tip the election the other way, it is an instance of bad information poisoning the deliberative process.

(c) Rational ignorance: normally, people have good incentives to filter out bad information. False beliefs can quickly ruin lives and reduce a person's welfare. However, in a democracy, these incentives either don't exist or don't exist to the same degree. If X is inferior to Y, and people choose X over Y, the costs of X will be spread around even to those who didn't choose X over Y. When the costs of making a bad decision are dispersed and transfered in this way, most individuals will have little incentive to gather the information necessary to make the right decision. Taking the steps necessary to filter out bad information would cost more to most individuals than making the wrong decision would, so why would they do it?

(d) Ideological considerations: here I mean ideology in the Marxist sense, i.e. a set of ideas (usually false) promulgated by the ruling classes to keep the hoi polloi under their control. Liberal democracies are not immune to this kind of ideology, and politicians have done a good job instilling in people the ideas necessary to perpetuate their own power. But if the people are voting based on ideological considerations, they aren't voting based on the intrinsic merits of the alternatives. That's another reason to suspect that democratic decision-making provides no evidence for the superiority of one alternative over another.

Incidentally, Jan Narveson wrote a great paper on the role of ideology in liberal democracies...

Thus, I submit: the fact that a majority of voters (not even a majority of "the people") prefer X over Y tells us little or nothing about whether X is better than Y.

3. So what use is democracy? Or: democracy versus liberalism.

Democracy is instrumentally valuable, in that holding an election is one way of preserving liberal institutions. But I do not think this works the way most people seem to. It isn't that, given the choice, the majority will support liberalism over the alternatives. It's that, giving people a choice (or the illusion of choice) keeps them from overturning liberal institutions and imposing their own view of the good on everyone else.

Democracy is like icing on the cake of liberalism. It makes the cake taste better to those who are going to eat it, and prevents them from tossing aside the cake in favor of some less savory alternative.

However, this is a minor benefit, and conditional on the existence of strong institutions to keep the majority in check, e.g. courts, constitutions, etc. I would take liberalism over democracy any day.

And, no, I don't think voting is rational. Make sandwiches for homeless people if you truly desire to make the world a better place.