Democratically Elected Aristocracies

The article suggests a formal model of a two-tier voting procedure, which unlike traditional voting systems does not presuppose that every vote counts the same. In deciding a particular issue voters are called in the first round to assign categories of their fellow-citizens with differential voting power (or weights) according to the special position or concern individuals are perceived as having with regard to that issue. In the second stage, voters vote on the issue itself according to their substantive view and their votes are counted in the light of the differential weights assigned in the first round. We analyze the formal and the philosophical reasons that support the model. ∗Department of Philosophy, The Hebrew University, Jerusalem 91905, Israel. E-mail: [email protected] †Department of Economics, Boston College, Chestnut Hill MA 02467. E-mail: segalu@ bc.edu Voting is one among many procedures of making social choices. Like some alternative procedures, such as the market, voting faces the problem of the translation of individual preferences into social policies. There are numerous issues involved in voting, e.g. articulating the issue to be decided, counting the votes and interpreting the election results. But behind all the attempts to design voting procedures lies the overall aim of representing individual preferences as loyally as possible. In this article we propose a new way in which voting can be transformed into a more sophisticated procedure, which would make the resulting social choice more sensitive to and hence more representative of the individuals’ preferences. Unlike other methods of social choice, such as the appeal to authoritarian decision by a dictator, to a committee of specialists, or to the use of lottery, voting takes seriously the preferences of individuals as they see it, and hence is committed to some form of equality. Our proposal follows this fundamental egalitarian principle but gives it a more complex interpretation. Voting is a procedure that is applied to issues that call for collective action. Its principal attraction lies in its being a decision-making procedure through which the integrity of a group can be maintained despite disagreement among its members about the correct or desirable way in which substantive issues should be settled. And unlike other procedures of collective choice, like lottery, compromise or the exercise of sheer power, majority vote is not arbitrary, ad hoc or oppressive. In its logical structure, voting cannot be fully reflexive, i.e. its procedural conditions as well as the formulation of the issue to be decided must be antecedently given rather than put to a vote. However, some, even if not all of its rules, may be decided by voting. In this article we suggest a model for such a partially reflexive application of voting, which offers a way of fine-tuning traditional majoritarian procedures. We are particularly concerned with the failure of traditional voting methods to pay tribute to the differential weight people often believe should be assigned to different voters. We therefore suggest the following formal model. Members of society are asked to rank possible subsets of society, where A Âα B means that person α prefers the subset A of individuals over the subset B to decide issue j for society. Under some assumptions we conclude that these preferences can be represented by a function V in the following way. Each member of society is a assigned a certain weight, and Since our model permits zero weights, it relates also to the question of the scope of the voting group, that is, who should take part in the vote.