nicholas aylottniklas bolinPolitical parties across Europe have a wide variety of procedures for selecting their leaders. While some hold formal leadership contests with all members casting a vote for their preferred choice, other parties can end up with a single candidate being presented to the party following negotiations behind closed doors. Nicholas Aylott and Niklas Bolin outline a model for understanding how these decisions are made by focusing on the role of so called ‘steering agents’ who help to determine which candidates are able to successfully navigate the selection process.

Party politics is about leadership. Yet the ways in which parties select the individuals who lead them have been oddly neglected by political science. Part of the reason for this relates to methodology: leadership selection processes do not naturally lend themselves to quantitative study. Time and local expertise are also required to study and interpret these relatively rare events and the often clandestine procedures that surround them. Case studies are better suited to this kind of research, but this makes generalisable conclusions more difficult to draw.

In a recent study, we have explored the topic and some of the approaches that have been pursued in studying leadership selection. Several authors have employed what might be termed a ‘middle way’ in approaching the problem, where a medium-sized number of national leadership selection processes are assessed using a balance of detailed description of political events and more systematic comparisons. We essentially suggest that the next step should involve, empirically, veering a little back towards case studies and, conceptually, developing new process classifications. Specifically, we want to complement our knowledge of what parties’ statutes say about how they select their leaders, the “official story”, with more insight into what happens in practice, the “real story”.

Research has concentrated on the “selectorate”, the party unit that enjoys the formal right to select the leader. This focus is necessary and interesting, not least because selectorates have grown more inclusive and decentralised. Party members, for example, are now often directly involved in leader selection. Sometimes even non-members get in on the act. But we suggest that prior stages in the process, in which candidates are filtered before the choice reaches the selectorate, are at least as important in determining who becomes leader.

When is a contest not a contest? When it’s a coronation

Our argument has been inspired by observing the selection of party leaders in Sweden, a country that is often regarded as being scrupulously democratic. Yet few non-Swedes would consider Swedish parties’ methods of selecting their leaders to be especially democratic. The congress, each party’s sovereign body, is the selectorate, which sounds fairly customary. With rare exceptions, however, the congress has little scope to decide anything, because it usually has only a single candidate to vote for – which it nearly always does.

In fact, single-candidate non-contests are fairly common. In a recent anthology, Ofer Kenig and his co-authors report that such coronations are actually the norm in the established democracies of continental Europe. Yet it is implausible that the supply of potential leaders in any party consists, at a given moment, of a single person. The question is, then: when, how and by whom is the supply winnowed down to one person – or, at least, to a small selection of candidates?

In answering this question, we have examined what we call “precursory delegation”, which takes place before the involvement of the selectorate. It is performed by a “steering agent” – an individual or group that is charged with overseeing the process of selecting the leader. The task involves actively encouraging, discouraging and perhaps blocking particular candidates, and may also involve controlling information about the preferences of various actors. Delegation to this “steering agent” can occur formally, according to statutes, or informally, even secretly. It can determine the selectorate’s eventual decision, or be fairly marginal to it. And this process can be mapped empirically by classifying any leader selection on two dimensions: the intra-party source of the steering agent’s authority; and the degree of control that the steering agent imposes on the process.

Steering leader selection

In the two main British parties, the Conservatives and Labour, the steering agents comprise the respective parliamentary groups. The distribution of preferences within these groups determines which candidates proceed to a decisive ballot of the selectorate, the party membership. In Labour, this filtering role is relatively weak, which can produce surprises. No one predicted that its endorsement of Jeremy Corbyn’s long-shot candidacy in 2015 would spectacularly mobilise his hitherto latent support among party members and sympathisers. By contrast, the Conservatives’ comparable but more restrictive rules induced – for the second occasion among the four on which they have been used – a coronation when Theresa May became leader in July this year.

Sweden offers unusually clear examples of steering agents in action. Swedish coronations are not just the result of informal, behind-the-scenes power struggles that shape the contest before the selectorate gets a look-in. Rather, each party’s rules and customs are shaped by the unvarnished aim of anointing one candidate (or, in the Greens, a pair of candidates) for the congress to confirm. The steering agent comprises a very Swedish institution, the “selection committee” (valberedning). All the parties have one. Yet we see interesting variation in how it operates.

The Greens’ version, for instance, is an agent primarily of the party members and activists, the “party on the ground”. The committee ultimately endorses a single duo for the party’s leadership positions, but it exercises a relatively light touch in doing so; other candidates are sometimes seriously considered by the congress. The Social Democrats’ selection committee, by contrast, is more an arena for bargaining between the party’s various power centres, while the control that it exercises over the selection process is almost total. No aspiring leader is permitted to make a pitch for the job, or even to acknowledge that they want it, before the committee’s public endorsement. Figure 1 compares these two Swedish parties.

Figure 1: Two-dimensional classification of two Swedish parties’ leader-selection procedures

sweden party leadership

Note: For more information, see the authors’ longer journal article.

We think that the identification of its steering agent, and the conditions under which it operates, says a lot about a party’s character, and particularly about its most important power centres. This sort of classification might also tell us something about the ways in which parties behave – what they prioritise, how they interact with other parties, how they change. Indeed, the framework is one part of a developing institutional take on the whole business of understanding political leadership.

Please read our comments policy before commenting.

Note: This article gives the views of the author, and not the position of EUROPP – European Politics and Policy, nor of the London School of Economics. Featured image: an election debate televised by the Swedish broadcaster STV. Credits: De rödgröna (CC BY 2.0)

Shortened URL for this post:


About the authors

nicholas aylottNicholas Aylott – Södertörn University
Nicholas Aylott is a Senior Lecturer (docent) in political science at Södertörn University in Sweden. His primary research interest is comparative European politics and, especially, political parties.


niklas bolinNiklas Bolin – Mid Sweden University
Niklas Bolin is a Senior Lecturer in Political Science at the Mid Sweden University. His research interests include political parties and party systems.

Print Friendly, PDF & Email