Suchergebnisse

This paper considers basic constitutional questions about how to elect a legislature. Electoral systems that require blocks of voters to coordinate their votes create a need for pre-election leadership and raise barriers to entry against new parties. Such barriers to entry can rigidity the political systems and decrease the incentives for established political leaders to serve the public honestly and effectively. So we consider an axiom of coalitional straightforwardness, which asserts that an electoral system should minimize the need for pre-election coordination of voters who share simple dichotomous preferences. Axioms of nondivisiveness, neutrality of party labels, responsiveness, and homogeneity (or coalitional autonomy) are also formulated. It is shown that only two kinds of electoral systems satisfy these axioms: winner-take-all approval-voting systems (AV), and single-divisible vote-proportional-representation systems (PR). Thus, AV and PR are seen to be uniquely compatible in terms of the incentives for party structure that they create, although AV and PR may differ in their incentives for party positioning in issue space. Possible use of AV and PR together in a bicameral legislature is discussed.

Even though voters can have complex preferences over multiple candidates or parties, all extant electoral systems provide political representation based solely or primarily on voters' first preferences. I present a new concept of proportional representation that takes account of voters' preferences over the full list of alternatives—fully proportional representation (FPR)—and schemes for its implementation. I outline a "pure" FPR scheme, but because this scheme would have several undesirable features when used by real voters, I also discuss modifications that account for these difficulties. Although there are a variety of interpretations of the role played by voting in democracy, several can be shown to suggest FPR as a normative ideal. Fully proportional representation provides us with new ways to conceptualize existing electoral systems, a new standard against which alternative systems can be evaluated, and several feasible alternatives for approximating this new ideal.