The personalisation of electoral campaigns appears obvious to professional and even casual observers of politics yet political science research suggests that this, as yet unestablished, phenomenon is debateable at least. Particularly questioned is the impact of individual leaders on the vote. Studies have struggled to show an impact of leaders, which seems unintuitive given the prominent role leaders play as ‘the face of the party’.
Some research, particularly by Garzia, suggests that this might be a problem with model specification; that the effects of partisan attitudes are overestimated because it is (erroneously) assumed they are exogenous to vote choice. Optimal solutions are not available with cross-sectional data, but in this paper we observe that most studies treat all voters as equally likely to be influenced by leaders.
We study the 2016 general election in Ireland, one which given the high levels of electoral volatility in it and the preceding 2011 ‘crisis’ election, would suggest makes Ireland a most likely case for the impact of leadership on the vote. We partition our dataset to look at the impact of leaders on those who rank lower on partisan attachment, again making this an easy test for the ‘personalisation’ hypothesis. It was also an election that media outlets regarded as a leadership driven election. If we do not see a leadership effect here, then it might be unlikely that one exists.
