Author: Ioannis Loukas Vassiliadis
Publisher:
Published: 2018
Total Pages: 129
ISBN-13:
DOWNLOAD EBOOK →
"The lifecycle of governments in parliamentary democracies has been in the forefront of comparative political research for more than half a century. While the empirical analyses have, up to this day, been performed with the individual cabinet as the unit of analysis, the decision to maintain or dissolve the government ultimately rests with the participating parties. This dissertation aims to contribute to the aforementioned literature by directly focusing on political parties as the main decision-making unit in the setting of coalition governments. The decision of political parties to leave the cabinet - and thus cause the termination of the government - is a potentially costly public action, one that strategic parties would not take lightly. Here, I examine potential reasons that account for the variation in parties' decisions to terminate the coalition governments in which they participate. In particular, I explore how the calculations of coalition partners are influenced by both party- and cabinet-specific factors, including structural characteristics of parties via-a-vis the rest of the government and the parliament, the likely political developments that a government termination engenders, and a changing political context. The analysis builds on a novel classification of government terminations that I put forth in Chapter 1. In particular, I introduce a set of systematic criteria intended to capture premature party departure from the cabinet. This allows a finer identification of the actors responsible for each government termination. This classification is then applied to a party-level dataset that contains information on all governments and their participating parties in 18 European countries in the period 1945-2014. In the first Chapter, I identify what party attributes are correlated with a higher ex ante probability of leaving the government, and explore whether ideological, electoral, or office-related incentives are associated with this phenomenon. I find that ideological conflict within the cabinet and electoral competition from the opposition are significant predictors of a political party's decision to leave the cabinet. On the contrary, there is only limited support for the hypothesis that parties leave in order to pursue better options by participating in a different government. Specifically, parties' coalition potential, as measured by their size and ideological placement in the parliament, does not statistically significantly increase the probability that the party terminates the government. This analysis allows for, and is robust to, the presence of cabinet-level unobserved heterogeneity. These findings are consistent with the notion that formateurs may not be maximizing the durability of coalitions, and that political parties may be bargaining in a non-continuous agreement space. Parties' calculations on exiting the government presumably depend on whether they expect a parliamentary dissolution and new elections to follow the termination of the government, or not. With this in mind, in the second Chapter, I offer an empirical account of parliamentary dissolutions and replacement governments using the conventional cabinet-level analysis. In particular, I examine the effects of different parliament-level factors that the existing literature has found to be conducive to dissolutions. The goal of this analysis is to derive unbiased estimates of the probability of a parliamentary dissolution following the termination of the government. This is a novel approach to an underexplored question, and provides substantive inferences independently of its larger role in this dissertation. These unbiased estimates can then be used to inform the analysis of individual party decision-making, by explicitly taking into account the possibility that a premature cabinet termination may lead to new elections or a replacement government. I show that the calling of new elections is correlated with premature government terminations, and that unless this correlation is accounted for, naive estimates of the probability of parliamentary dissolutions are biased due to sample selection. I find that time remaining until the next regularly-scheduled election is the most important predictor of parliamentary dissolution, while there is mixed support for explanations based on the complexity of the bargaining environment. Armed with the results of Chapter 2, I perform a more nuanced analysis of the decision of individual parties to leave the cabinet in Chapter 3, where I examine whether changes in the political and economic environment lead coalition partners to reconsider their participation in the government. Using changes in opinion polls and the state of the economy, I am able to quantify "critical events" and thus capture the dynamic context in which governments and parties operate. I combine these with unbiased estimates of the probability of new elections following the termination of the government, obtained in the previous chapter. By properly accounting for the different outcomes of government termination, I am able to adjudicate between "surfing" and "blame avoidance" incentives. The main results show that popularity surges can lead to government terminations, but are not sufficient to cause parliamentary dissolutions: to the extent that individual parties capitalize on their increased popularity, it is by negotiating better terms within the existing parliament. This reflects the fact that in multiparty settings, as the ones in this study, individual parties benefiting from improved polls can seldom convince their laggard counterparts to dissolve the parliament. Furthermore, I find that parties react to opportunistic incentives created by improvements in the state of the economy. Finally, I find that the effects of the different environmental changes are conditional on the likelihood of new elections, should the government terminate"--Pages xi-xiv.
