More on Supermajority Rules: May’s Theorem

Recently, Adrian Vermeule had a review of the new book by Melissa Schwartzberg on supermajority rules – Counting the Many: The Origins and Limits of Supermajority Rules. I haven’t read Schwartzberg’s book yet, but Vermeule does discuss a number of arguments against supermajority rules.

One such argument is based on May’s Theorem. According to the Theorem, only majority rule satisfies four conditions on group decisionmaking. Two of the key ones are anonymity (the group decision treats each voter identically) and the neutrality (the group decision treats both outcomes the same, in that reversing the preferences reverses the group decision).

Supermajority rules satisfy the anonymity condition. Under ordinary supermajority rules, each voter receives one vote. But supermajority rules violates the neutrality conditions. Under a 3/5 supermajority rule for passing new spending, if a majority is opposed to the new spending, it gets it way and the spending does not pass, but if a majority is in favor of the new spending, it doesn’t get its way and again the spending does not pass. Clearly, this 3/5 supermajority rule privileges in a sense the decision against new spending.

Is this a problem? Advocates of May’s Theorem certainly treat it as one. But all that the violation of the neutrality condition shows is that there is a privileging of a decision (as compared to majority rule) and that this privileging needs a justification.

This requirement of providing a justification is unexceptionable – supermajority rules should be justified. But there are many justifications. In my arguments (authored with John McGinnis) in favor of certain supermajority rules, we always provide a justification. For example, we argue that spending laws tend to be disproportionately favored by special interests and therefore a supermajority rule should be required to offset the effects of these special interests. As another example, we contend that constitutional provisions (which are designed to constrain government and to last for long periods and therefore can only be repealed by a supermajority) need to be of higher quality than ordinary legislation that can be more easily eliminated. The supermajority rule enactment process helps to ensure that the constitutional provisions that pass have higher quality than ordinary legislation.

Of course, strong advocates of May’s Theorem would regard its prescriptions as either absolutely binding or not to be easily overridden. But it is not clear that this approach is correct. Democracy is a process that needs to be justified based on its results. Some contestable and abstract notion of equality is not a good enough reason to employ an inferior system.

In other areas, we do not accept the notion that all decisions should follow neutrality. For example, we might make decisions case by case on whether to eat a food — treating the decision whether to eat it neutrally.  But some of us recognize that we are prone to overeating and therefore we place additional burdens on the decision to consume food. Such violations of neutrality in our individual decision are the essence of reasonableness, not something to be criticized.  The same can hold true for collective decisions.