Electoral Accountability

先前关于立法机构联合组建的研究主要集中在一般假设上，以得出关于哪些政党会同意组建政府的预测。这种普遍性意味着预测缺乏独特性，阻碍了它们作为研究前一阶段相互作用的基础，即立法机构政党的实际选举。本研究的目标是建立一个适用于所有类型的联合政府的模型，而不考虑政府组建过程中不同的制度安排（例如，哪个政党最初试图组建政府）。这个项目的动机是对联合政府形成的制度结构有更深的理解。这些制度结构通常为政党在政策、内阁组合等方面向其他政党提出建议和让步方面提供战略可能性。通过采用非合作博弈论的技术，研究者将对联盟的形成和立法结果作为现有制度规则的函数进行预测，从而允许比较不同结构的效果。这些技术还将允许调查人员将分析“回溯”到选举阶段，在选举阶段，选民将被建模为具体的预测，如果他们投票给这样或那样会发生什么。通过这种方式，研究者希望将多党制政府中选举竞争和联盟形成的研究统一起来。非合作博弈论的最新进展使这些更复杂的模型得以发展。这些进步之一是放松了先前工作的信息假设，允许一些参与者拥有既有价值的信息，因为它会影响其他参与者的回报，又有隐私的信息，因为没有其他人拥有它。了解如何才能说服另一方同意组建政府，就是此类信息的一个例子。结合这些先进的技术将产生一个模型，该模型不仅为参与者提供了丰富的战略可能性，而且比之前的模型更现实。这将加强上述选举和联盟形成的理论统一，并使人们更好地理解这种政治环境中参与者的动机和行为。

Previous research on coalition formation in legislatures has focused on general hypotheses to derive predictions concerning which parties would agree to form a government. This generality implies a lack of uniqueness in the predictions, inhibiting their use as a basis for studying the preceding stage of interaction, namely that of actual election of the parties to the legislature. The goal of this research is to develop a model applicable to all types of coalition governments regardless of differing institutional arrangements (e.g. which party initially attempts to form a government) that structure the process of government formation. This project has as its motivation a deeper understanding of the institutional structures within which coalition governments form. These institutional structures typically generate strategic possibilities for the parties in terms of the proposals and concessions to make to other parties, on policy, cabinet portfolios, and so on. By employing the techniques of non-cooperative game theory, the investigators will generate predictions of coalition formation and legislative outcomes as a function of the institutional rule in place, thereby allowing for the comparison of the effects of different structures. These techniques will also allow the investigators to "roll-back" the analysis to the electoral stage, where the voters will be modeled with specific predictions of what will happen if they vote one way or another. In this way the investigators hope to unify the study of electoral competition and coalition formation in multi-party systems of government. Recent advances in non-cooperative game theory permit the development of these more sophisticated models. One of these advances has been toward relaxing the informational assumptions of previous work and allowing some participants to possess information that is both valuable, in that it effects payoffs to other participants, and private, in that no one else has it. The knowledge of exactly what it would take to convince another party to agree to a government is an example of such information. Incorporating these advanced techniques will result in a model that is both rich in strategic possibilities for the participants, and more realistic than its predecessors. This will enhance the theoretical unification of elections and coalition formation described above, and lead to a better comprehension of the motivations and behavior of participants in this type of political setting.