Stability and Change in Voting Behaviour


The presence of electoral volatility, and the fact that voters can change parties from one election to another is essential for the well functioning of representative democracies. Only if voters are open to switching parties can changes in public opinion be translated into alterations in which parties are elected, how many votes they obtain, and who governs. It is also by changing parties that voters can express dissatisfaction about their representatives or incumbents. Additionally, the mere possibility of voters switching parties is thought to incentivize parties and politicians to act responsibly and to govern well. Whether those high expectations with regard to volatility are fulfilled however, depends to a large extent
on the characteristics of volatile voters and what motivates them to switch parties. In this dissertation, the evolution and determinants of volatility in advanced democracies are investigated. To that end shifts in election results (the macro level) as well as changes of individual voters (the micro level) were investigated.
Election results in Western Europe have become significantly more volatile since 1950. The Pedersen index an indicator of net volatility has almost doubled between 1950 and 2013. The mean Pedersen index of elections before 1960 was about 7 and has increased to 13 for elections since 2000. This trend of a clear evolution towards more volatility can also be observed when the individual level is focused upon. In the Netherlands for example, one of the countries in Western Europe with the strongest increase of net volatility over time, 13% of the voters reported to have changed parties in 1971. In 2010, this proportion of voters reporting to have switched has increased to 47%, or about 1 on 2 voters.
Undeniably, voters who switch parties from one election to another are dissatisfied about the party they previously voted for. This observation, however, does not imply that volatile voters are frustrated about politics. The results in this dissertation indicate that voters who change parties are still satisfied with democracy, in contrast to voters who decide to no longer turn out to vote. The results that are presented hence suggest that party switching should be considered a positive choice. Volatile voters do not turn away from politics or the party system; instead, they are looking for an alternative for the party they previously voted for. It seems as if voters are using their vote to express their evaluation of parties performances. Similarly, the analyses at an aggregate level suggest the presence of a reward-and-punishment mechanism, as evident from the fact that levels of net volatility are systematically higher in times of economic downturn.
The results presented in this dissertation furthermore invalidate the claim that volatility would be the result of voters with little interest in politics and with a low level of political knowledge switching parties. A first indication therefore lies in the fact that, in general, the probability to switch parties is highest among voters with a middle level of interest in and knowledge of politics. A second indication is offered by the fact that even though low politically sophisticated voters are more likely to switch parties in the course of an election campaign, their impact on changes in election results is compensated by the fact that high politically sophisticated voters already switched before the start of the election campaign. The increasingly large shifts in election results from one election to another are hence not the result of a large group of low politically sophisticated voters floating from one party to another. Switching parties requires a minimum level of political sophistication.
Voters who change parties are not switching haphazardly; their vote choices are strongly ideologically constrained. When a voter changes parties, s/he does so to a party that is ideologically close to the previously preferred party. This is clear from the analyses on the distance of party switching that are presented in this dissertation. In advanced democracies, the mean ideological (left-right) distance bridged by voters who switch parties is only 1.3 on a 0-to-10 scale. Additionally, the results presented point out that merely a higher number of parties at o er does not increase the probability that voters switch parties. What incentivises voters to change parties is not the number of parties at o er, but instead how ideologically close those parties are. Voters only switch parties when they find a good and ideologically close alternative to their previous party.
A large number of voters switch parties from one election to another and this number has strongly increased over the past couple of decades. This observation leads to uneasiness about the quality of democracy and its future. The results presented in this dissertation invalidate a too pessimistic account of volatility in a number of ways. First, party switching cannot be considered an expression of frustration about politics in general, but is a way to voice dissatisfaction about ones previous party. Voters are not loyally voting for the same party election after election, but evaluate the performance of the party they voted for. Second, volatility is not only the result of a group of low politically sophisticated voters changing parties. The high politically sophisticated as well switch parties from one election to another, even though they tend to do so before the start the election campaign. Finally, ideology strongly constrains party switching and volatility appears to be mainly driven by voters changing between ideologically close parties. In sum, volatility is crucial and important for the well functioning of representative democracies.

This content has been updated on 15 October 2016 at 12 h 47 min.