Joost van Spanje
How do voters decide who to vote for? Researchers commonly assume that voters compare parties’ policy issue positions with their own ideal points and then apply a rule to choose among these parties. This assumes that the voters’ positions on these issues are truly exogenous to voters’ party affect. In other words, do voters like the party they already agree with or do they agree with the party they already like?
In order to address this question, we would ideally have the following setting. An exogenous intervention causes considerable unidirectional change in voters’ affective sentiments about the party. The intervention does so without providing any information about the party’s position on relevant issues – and without providing voters reason to reposition themselves on any relevant issue. This shock should be as good as randomly assigned to a subgroup of a representative sample of the electorate. Moreover, information about voters’ opinions and perceptions about parties’ positions on various issues should be available. This setting should refer to real parties, real elections, and real information flows.
For such a scenario, let us to take you to the Netherlands on 14 February 2002. That day, former sociology professor Pim Fortuyn, new to the political scene, founded a new political party. It was clear from the start that Fortuyn was the party, and the party was Fortuyn. He was the party’s only founder, only leader, and only spokesperson, and openly denounced the other candidates of his own party. The party was, quite appropriately, called “List Pim Fortuyn” (LPF). The LPF’s first election was the national parliamentary election on 15 May. The party quickly rose in public opinion polls, stabilizing at around 12% of the vote in April.
The exogenous intervention is a murder. On 6 May, Fortuyn was shot at close range and died. The killer chose to remain silent about his reasons for gunning down Fortuyn and could not easily be linked to any of the LPF’s policy positions. As a result of the assassination, sympathy for Fortuyn increased substantially. We know this, as the murder took place when the fieldwork for the 2002 Dutch Parliamentary Election Study was underway. Some respondents were interviewed before the event, while others were interviewed after. After matching to make the two groups comparable, we compare these two groups of voters on their affect towards the LPF, perceptions of the LPF’s position on several issues and the voters’ preferences on the same issues. We find four things.
First, the murder generated an upward shift in respondents evaluations of the LPF. Second, the assassination affected respondents’ perceived proximity to the LPF on various policy issues. Third, the reason why voters became closer to the LPF was not that they updated their own positions but that they located the LPF in a different position as soon as the murder had happened. Fourth, the increase in proximity is dependent on the saliency of the issue. When voters believe the issue is important in their electoral choices, as were the issues of asylum seekers and redistribution, they are more likely to bring the party closer to their own views. With issues which are not electorally very important, such as euthanasia, the pressure to bring the party closer is weaker. In all instances, however, the voter seems to be the magnet and the party seems to be the iron: voters do not appear to change their own issue stances significantly. They do, however, seem to change their perceptions of where the party stands. They do so both in an issue in which prior knowledge of the party’s position was unambiguous (asylum seekers) and in an issue in which little was known about where the LPF stood (redistribution). Both issues, however, were important in voters’ final electoral decisions.
Without of course challenging the significance of issue congruence on voting behaviour, these findings suggest that at least in part congruence is endogenous to non-issue based evaluations of parties by the voters. The study questions the rational choice-based interpetation of spatial voting. What often appears as clear evidence in favour of the smallest distance theory, might well conceal affect-based convergence on non-issue based grounds. Voters, in this case, seem to bring LPF closer to their own views because they felt sympathetic to the leader’s death. Doing so, and actually voting for this party seems to have been partially the equivalent of leaving a flower in Fortuyn’s coffin.