In recent decades, parliamentary elections have become more and more leader-centric, with the competition between the leading candidates of main political parties for prime ministerial office, their campaign appearances and TV debates, drawing much attention by the media. Likewise, parties increasingly rely on leaders to mobilise voters and convince them to cast their ballot for the party. This change in electoral behaviour has been described as personalization of parliamentary elections and is also an important feature in the presidentialization of parliamentary systems.

While research into the personalization of parliamentary politics has provided crucial insights into long-term changes in advanced democracies, the rising influence of leaders has also raised concerns over the rationality of voting behaviour in parliamentary elections, in which voters should optimally reward or punish incumbent parties and secure the accountability of the parties in office. In contrast, the rising influence of popular leaders suggests that leaders’ charisma and personality are becoming more dominant drivers of voting behaviour. This may enable political parties to avoid accountability in parliamentary elections by selecting charismatic leaders as prime ministerial candidates.

In my recent article published in Politics, I address these concerns and test whether electoral accountability for government performance can be extended to leaders who have obtained office and serve as prime minister. I argue that prime ministers will be held personally accountable for government performance by voters, because they are influential agenda-setters in parliamentary governments. Therefore, control over the prime ministership serves as heuristic for voters to decide to whom they should assign responsibility for outcomes created under joint-decision making in cabinets. This leads to an attributional assessment in which voters’ evaluation of the prime minister depends on voters’ perception of government performance; even if government outputs are created by joint decision-making in cabinet. To answer my question I study voting behaviour in British, Danish and German parliamentary elections between 1983 and 2017. In addition, I exploit an as-if-random prime ministerial replacement in the German state of Lower Saxony to assess whether reverse causality could be at work and voters’ evaluation of the government’s performance could depend on their evaluation of the prime minister.

My findings show that prime ministers are indeed held accountable for their government’s performance by voters. This also means that the influence of prime ministers on vote choices is likely to be overestimated when government performance is omitted from statistical analysis. Furthermore, my findings on Danish elections indicate that those prime ministers who tend to enjoy less influence over decision-making in their government, will also be held personally accountable for government performance to a lesser extent. However, more definite conclusions on this issue are subject to further research.

In conclusion, it appears that personalisation does not endanger the function of parliamentary elections as a tool of accountability for performance in office, because prime ministers’ influence on vote choice is not only a result of their charismatic relationship with the electorate, but also driven by the performance of prime ministers in office.