Jan 20, 2011

A problem for majoritarian democracy

Does the mere fact that a democratic majority chooses a policy provide citizens of the democracy with a reason for endorsing or following the policy, or must that justification come from the fact that the policy embodies certain goods like justice or equality? I will argue that there is a strong reason to reject the first option, a reason that has surprisingly little to do with political philosophy at all.

Let's give the view under attack a name.
Majoritarianism: The mere fact that a policy is democratically selected by a majority gives citizens a reason to endorse or follow the policy.
When I say the policy is democratically selected, I mean to imply that the ordinary democratic requirements of universal and equal suffrage are met, that the election was not rigged, and so on.

To show the problem with majoritarianism, I need first to introduce a result that Philip Pettit discusses in his 2006 Analysis article, "When to defer to majority testimony—and when not": majorities can endorse a set of inconsistent propositions, even when none of the individuals in the majority holds an inconsistent set of propositions. This is easy to see from an example Pettit gives (p. 83):
“Take three people, A, B and C, and consider how the majority judgements might go on three simple propositions, p, q, and p&q. A and B may support p, with C against; B and C may support q, with A against; and so, if they are individually rational, A and C will reject p&q, with only B in favour. A majority will support p, a majority q, and a majority not-p&q. A majority will support a semantically inconsistent set of propositions. And this will be so, despite the fact that each of the members of the group has a perfectly consistent set of beliefs.”

When we translate Pettit's example into the political sphere, we conclude from majoritarianism that each citizen has a reason to endorse or follow policies p, q, and not-p&q, when those policies are democratically selected by a majority. This is contradictory. The same result does not affect supermajoritarian views, such as this one:
Supermajoritarianism: The mere fact that a policy is democratically selected by a sufficiently large supermajority gives citizens a reason to endorse or follow the policy.

One complication with the case is that legislatures don't usually enact contradictory policies. (This may be true at one time but obviously false historically. Prohibition, for instance, was enacted and then repealed.) But this really doesn't matter. Majoritarianism cannot plausibly limit itself to what is actually democratically selected (which could be thwarted by some catastrophe that prevents a vote from taking place), but rather what a majority of the legislators actually favors. As we learn from Pettit, this can be contradictory.

A further complication: Pettit suggests that the problem affects only beliefs that are connected to others. Political beliefs are presumably paradigmatic examples of embedded beliefs, however, so this is not a point that undermines my argument.

Finally, one might object that the contradictory cases are rare or non-existent in practice. This is irrelevant, however, since the question concerns what normative significance they would have were they to occur.

The moral is that majoritarianism, as I've defined it, fails. An instrumental version of majoritarianism would escape the argument, since presumably contradictory positions cannot instantiate any substantive goods like justice. Supermajoritarianism is probably wrong as well, but that's a separate question.