The effect of observable party cohesion on voter choice in liberal democracies

Abstract

The goal of this thesis is to assess the electoral impact of party infighting. It addresses the key research question of what is the impact of party infighting on the electoral fortunes of that party. In answering this question by confirming that party infighting does affect parties a number of interrelated questions emerge. Why does this effect occur? Is it simply a function of the way that voters receive news about the party? Is it something intrinsic about the events of disunity themselves? Is it the disunity that causes the change in party support, or might the reverse be true? This topic is approached in a number of chapters. In Chapter One the topic is introduced. It explains why this is an important area of research and outlines the structure of the thesis. Chapter Two gives a theoretical overview and outlines what is already known from the literature on this topic. It concludes by observing that the literature is very uncertain on this area and that while there are many plausible explanations for whether and why cohesion should matter, these are often contradictory. Chapter Three is the first of four empirical chapters. This chapter answers the question of whether party cohesion affects the votes of the public. Using election studies in a variety of countries it establishes that it does and negatively. Chapter Four assesses whether this effect derives from extensive coverage of party infighting in the media. It established that this was not the origin of the effect using the case of the Canadian Liberal Party from 1996 until 2006. Chapter Five investigates whether any effect observed was related to the events of disunity themselves. Using a new database of newspapers it establishes that it did, but positively and not negatively. A number of explanations are outlined for this but a temporary underdog effect is seen as most plausible. Chapter Six explores the causal mechanism. It is concerned with the causal direction of the relationship between infighting and the vote and what causes the relationship for voters. Using new opinion poll data it shows that party infighting being a cause of poor results is much more plausible than inverse relationship. Furthermore it links party infighting to leadership evaluations. Finally, Chapter Seven concludes the project by showing how these findings link together and establishing their significance for future research and the wider world. The impact and importance of this thesis stems from its deep probing into the important issue of party cohesion for vote choice. It goes beyond establishing a correlation between cohesion and the vote. It uses a variety of different sources to establish how and why that correlation comes about. Accordingly, it can make a plausible case for showing causality. Following the establishment of the relationship in Chapter Three each subsequent chapter is concerned with showing how this relationship comes about. This means that it thoroughly examines not only the existence of this relationship but also what drives it. The findings of this thesis are broadly applicable to the area of voter choice. Party infighting, by affecting policy, leadership and performance in office touches upon virtually all major strands of opinion as to what primarily drives voters. This thesis finds that infighting mostly bears a relationship to opinions on leadership to voters. However, it also finds that voters largely think of it as separate from virtually every other issue and it interacts differently than leadership in their decision-making process. It does demonstrate, though, that regardless of what primarily drives voters, cohesion demonstrates a smaller but separate effect. It is worth accounting for.