The Interaction Between Electoral Preferences and Party Leaders in Britain

理论空间投票模型通常假设政党是单一行为者，并采取选举前的政策立场，以最大化选举人票（或席位优势）。最近，首席研究员和他的同事们在一些欧洲国家和以色列建立了选举反应的多项概率（MNP）模型。虽然这些模型准确地捕捉了选举行为，但很明显，政党通常不是选票最大化者。形式分析认为，在比例代表制政治中，政党采取选举前政策是为了有利于选举的联合议价。然而，这种模式并不能解释英国选举前的政党选择，在英国，明确的政府联盟是罕见的。本研究以1992年和1997年的选举为研究对象，构建了“多数主义”英国政体中选举和政党行为的综合模型。为了构建英国选举的MNP模型，需要选民偏好立场和政党声明的数据。埃塞克斯档案馆提供的调查数据用于生成选民偏好的样本分布。政党宣言的计算机编码被用来估计政党的立场。选民意图被用来估计MNP模型。额外的月度盖洛普“受欢迎程度”数据，连同经济变量和回顾性/前瞻性汇总数据，用于完善选举人票的MNP模型。如此构建的选举模型将有助于深入了解“战略性”选举行为的程度。通过对选区划分模式的模拟，可以估计出选举操作者将政党立场与席位优势联系起来。通过问卷调查获得英国立法者的政策偏好。这些数据与政党积极分子态度的调查数据一起，用于分析各政党内部对政党领导人的选择和选举宣言的构建的决策问题。相关的理论工作主要集中在三个主要政党在选举前的战略操作的性质上。对英国的理论和实证模型的整合，以及对欧洲“多党制”政策的并行研究，提高了对多数选举制度和比例选举制度之间“结构性”差异本质的理解。宏观政治模型与微观选举分析相结合，有可能对理解不同体制安排下的民主决策作出重大贡献。

Theoretical spatial voting models typically assume that parties are unitary actors and adopt pre-election policy positions so as to maximize electoral vote (or seat strengths). Recent work by the principal investigator and his colleagues has constructed multi-nomial probit (MNP) models of electoral response in a number of European countries, and Israel. While the models accurately capture electoral behavior, it is also quite obvious that parties are generally not vote maximizers. Formal analysis has proposed that parties in polities based on proportional representation (PR) adopt pre-election policies with a view to pros-election coalition bargaining. However, such models do not explain pre-election party choices in Britain, where explicit government coalitions are rare.This research builds an integrated model of electoral and party behavior in the "majoritarian" British polity, focussing on the 1992 and 1997 elections. To construct the MNP model of British elections, data on both voter preferred positions and party declarations are required. Survey data, available from the Essex Archive are used to generate sample distributions of voter preferences. Computer coding of party manifestos are used to estimate party positions. Voter intentions are used to estimate the MNP model. Additional monthly Gallup "popularity" data, together with economic variables and retrospective/prospective aggregate data are used to refine the MNP model of electoral vote. The electoral model so constructed will provide insight into the degree of "strategic" electoral behavior.Simulation of the operation of the pattern of districting (into constituencies) permits an estimation of the operator relating party positions to seat strengths. The policy preferences of British Legislators are obtained through questionnaires. These, together with survey data on attitudes of party activists, are used to analyze the internal decision problem within each party over the choice of party leader and construction of election manifestos.Related theoretical work focuses on the nature of strategic maneuvering by the three major parties prior to elections.The integration of the theoretical and empirical models on Britain, together with parallel work on European "multiparty" policies improves understand of the nature of the "structural" differences between majoritarian and proportional electoral systems. The combination of macro-political models and micro-electoral analysis has the potential to contribute substantially to the understanding of democratic decision-making under different institutional arrangements.