Here is a full statement of the theory of political parties as long coalitions, ones organized and elected to stick together on all or most legislative votes. The incentive to form, join, and elect them comes from the external cost of simple-majority voting—the central problem of The Calculus of Consent—but more fundamentally from the Paradox of Voting, or cycles of majority preference. I prove that a cycle among prospective legislative outcomes is sufficient for that incentive to be effective, and necessary too: without cycles there would be no parties. The identification of parties with long coalitions originated in a squib written years ago. The chief innovation of this paper is the proven cyclic basis of parties (and with it the absence of parties from one-dimensional voting bodies). Other innovations include extensions of the theory to minority parties, electoral parties, and subnational parties, a deeper explanation than Duvergers’ of two-party systems, and an explanation of how parties maintain their length and use it to prevent defection.

这是对由长期联盟组成的政党理论的完整说明，这些政党是经组织和选举产生的，可以通过所有或大多数立法选票凝聚在一起。组建，加入和选举它们的动机来自简单多数投票的外部成本，即同意演算的核心问题-但更根本的原因是投票悖论或多数优先权的循环。我证明，预期立法结果之间的周期足以使这种激励措施有效，也是必要的：没有周期，就不会有政党。具有长期联盟的政党身份的确定起源于几年前的发。本文的主要创新之处在于证明了政党的循环基础（因此，一维投票机构中没有政党）。其他创新包括将该理论扩展到少数党，选举党和地方政府，比杜弗格斯（Duvergers）的两党制更深入的解释，以及对政党如何维持自己的身分并防止叛逃的解释。