The reality of representation in Europe: the mode of leader selection in political parties

ABSTRACT

Political parties's choices of who leads them can have a major bearing on politics. Recent research shows that selection procedures exhibit considerable variation, even among parties in broadly comparable European parliamentary democracies. The most common analytical approach is to focus on the ‘official story’ – that is, what the parties’ statutes say that they do when selecting a leader. This, in turn, implies a heavy emphasis on the final stage of the selection procedure, in which the decision about who will lead the party is made by the ‘selectorate’. Yet this, the ‘official story’, is only a part of the process, and quite often not even the most important part. In this article, we seek to make the classification of selection processes more manageable and meaningful. We propose a typology of the ‘mode’ of selection, in which the emphasis is on the management of competition for the leader's position before the decision reaches the selectorate. We identify five modes of competition: open, enclosed, filtered, enclosed and filtered, and managed.

KEYWORDS:

• Political party
• leadership selection
• Europe
• steering agent
• sets

Introduction

National politics, in Europe especially, can be understood primarily in terms of political parties. When the highest political authority comprises a collective organ, such as a parliament, and when that organ comprises individuals chosen periodically through public elections, parties become indispensable for organizing political competition and decision-making. Parliamentarism, which applies in most European democracies, revolves around parties (Müller 2000; Strøm 2003, 67–73).

Parties are purposive actors (Katz 2002a). The action of each one is influenced by its core electoral support, but also by much else besides – including the preferences and strategies of the party leader. Indeed, the choice of leader is arguably the most important decision that a party makes. It expresses a collective choice about the party's immediate trajectory. Furthermore, the manner in which a party selects its leader is crucial in determining who it ends up choosing – and there is intriguing variation in selection procedures in European democracies. Open, cut-throat competition between contenders for the leader's job usually occurs within British parties (Quinn 2012). By contrast, closed, negotiated leadership transitions still often arise in the north of the continent. Most parties make their own decision on who leads them – that is, they limit the decision to members only. However, parties in Southern Europe often allow mass participation, extending beyond the party membership.

We think it is unlikely that variation across parties arises randomly. In fact, our core argument is that the procedure used for choosing a party leader reflects the structure of power within the party. This obviously implies that power structures also vary across parties. For example, the power distribution within some parties may favour the politicians elected under its banner. As politicians are often thought to value relatively highly the selective benefits derived from holding public office, this internal power structure might incline the party to behave in ways that prioritize the pursuit of public office, perhaps through programmatic flexibility and compromise with other parties. One of the basic claims of the venerable ‘cartel thesis’ (Katz 2022c; Katz and Mair 1995, 2018) is that more and more parties had drifted into this office-seeking behavioural category.

Yet some parties sometimes clearly prioritize goals other than government office. They may instead strive for short-term vote-maximization or the promotion of certain policies (Harmel and Janda 1994). In such cases, it is reasonable to suggest that leader selection in particular, and intra-party power more generally, may be weighted in favour of other sections of the party organization (Strom 1990). Selection procedures can thus have very considerable political significance. For instance, it is plausible to suggest that the result of the Brexit referendum in 2016, and the hard form of Brexit that was eventually implemented in 2020, might have both been very different had the two main British parties chosen different leaders than those they appointed in 2015 and 2019. Both parties’ selection procedures had recently changed. What looked like arcane procedural tweaks may well have had enormous ramifications.

Valid measurement of a phenomenon is a vital first step towards understanding and explaining it. The aim of this paper is to make this task of measurement, and thus of comparative analysis, more meaningful and manageable. It proposes a typology of selection procedures in which, in contrast to previous approaches, the identity of the ‘selectorate’ – the organ of the party that is formally designated by the statutes as that which makes the decision – is not a main feature. Instead, much more emphasis is placed on pre-selection phases. These phases may still leave the selectorate with a major decision to make, or they may disempower it by presenting it with only one candidate to rubber stamp. In other words, the pre-selection phases often determine who will win – or at least which sort of aspirant has a chance.

We envisage five ‘modes’ of selection procedure. Each one reflects a different type of competition between individuals who aspire to the leader's position. The nature of the competition is determined by the party's rules and practices, which – we suggest – are themselves shaped by historical patterns of intra-party power relations. Note that the typology is not itself sufficient to identify the location of intra-party power in any particular party. That requires further analysis. The typology is rather a way of structuring the investigation of where power lies. It tells us where to look.

The paper unfolds as follows. First, we briefly summarize the state of research on party leader selection. We then discuss and define the basic concepts that we use in our typology, which comprise the subordinate categories that we work with. Among the subordinate categories are the important notions of precursory delegation and, connected to it, the steering agent. After that comes the empirical part. We introduce the cases on which we have data, and we allocate them to the five target categories in our typology. Finally, we conclude and sketch some of the ways in which the typology might be deployed.

Previous research – and our approach

Comparative research on party leadership was oddly neglected until the last couple of decades. Since then, pioneering work like that pursued by Bynander and ‘t Hart (2008), Campus, Switek and Valbruzzi (2021), Cross and Pilet (2015), Kenig (2009a), Lago and Astudillo (2023), Pilet and Cross (2014), Sandri, Seddone- and Venturino (2015), Schumacher and Giger (2017) and So (2012) and has taken our knowledge about leader selection considerably further. It became clear, for example, that the procedure had over time become more ‘inclusive’ in many parties, in the sense that the selectorate expanded. In many parties, the selectorate is no longer an exclusive group, such as the parliamentary caucus or the organizational executive, but rather a bigger section, such as the party congress, the entire party membership or even a self-chosen set of sympathizers within the broader electorate. Some call this the ‘democratization’ of leader selection (Kenig 2009b; LeDuc 2001).

Yet the standard interpretation of this ‘democratization’ has, according to some, missed crucial aspects. For one thing, it might give undue weight to what has been called the ‘official story’ – that is, what the parties’ statutes say that they do (Katz and Mair 1992, 6–8). This, in turn, implies a heavy emphasis on the final stage of the selection process, in which the decision on who will lead the party is made by the selectorate. Indeed, one of the arguments in this paper is that the decision by the selectorate, the culmination of the process, is only part of the story. Quite often, it is not even the most important part.

Take the occurrence, quite frequent in some countries (Cross and Pilet 2015), of what are sometimes called ‘coronations’. In a coronation, the selectorate finds itself with only a single candidate to choose – or rather to confirm. A survey of selections in 1965–2012 found considerable stability in the rate of coronations (Kenig, Rahat, and Tuttnauer 2015, 60). In such a scenario, the identity of the selectorate, and rules by which it decides, are obviously unimportant. The far more interesting and significant steps in the selection process will have been taken at an earlier stage. Unless a party genuinely contains only a single aspiring leader (which is probably rare, even in tiny parties), some ways will have been found, prior to the coronation, to filter out all but one aspirant. These steps may have been more or less open, according to agreed procedures; or they can be clandestine, in which power is exercised by individuals or factions in informal, unregulated ways. Whatever, these steps are generally hard to measure validly and reliably, which is presumably a big reason why it is seldom tried within the field of comparative politics, particularly in large-N studies.¹

There are also potential implications for debates within political science about the development of parties generally. Some argue that parties have become more inclusive in their decision-making to try to forestall membership loss and maybe attract new (or lapsed) members (Kosiara-Pedersen, Scarrow and Haute 2017). Others suggest that ‘democratization’ is essentially a sham: incumbent elites have not only retained their intra-party power, they have also enhanced it. Elites have achieved this, it is argued, by assuming extended powers of agenda-setting and scattering the mid-level activists who did most to restrict the power of the elites in previous eras (Ignazi 2020; Katz 2022b; Mair 2013).

Which interpretation of greater inclusivity is closer to the truth? In seeking an answer to this question, we think that the ‘real story’ is at least as revealing as the official story. In this paper, we seek to describe meaningfully the ‘real’ process of leader selection, since description is a necessary step before we can explain anything (Gerring 2012). In doing so, we follow a ‘classic’ style of concept-building (Mahoney 2021, 83–87). Membership of what we call target categories, which are the five modes in the typology, are determined by set relations between subordinate categories. Membership of a certain subordinate category may be necessary or sufficient or both for membership of a target category.

This sort of set-theoretical approach (Mahoney 2021; Ragin 2000), combined with a focus on pre-selection phases, offers a far more rounded and realistic way to understand variation in leader selection than does a search for correlations between variables, which tend to rely on many observations, simplistic operationalization and a focus on the selectorate. The typology is illustrated by cases drawn from a comparative, international project on leader selection.

Concepts in the analysis of party-leader selection

There must be lots of aspiring contenders when a vacancy arises at the top of a political party. Should all of them have a chance to run? Probably not – at least from the party's perspective. Such radical openness would make the process slower and more costly than it needs to be. Ideally, fringe or long-shot aspirants should be weeded out early.² But how to define such aspirants? How to weed them out? And how do coronations – in which all but one aspirant is weeded out – occur? In this section, we define some of the most important concepts that we employ in this hitherto under-studied process (see Aylott and Bolin 2023a; Bolin and Aylott 2021).

There are essentially two types of filter through which the field of aspirants must navigate. The first type is eligibility requirements – the formal, impersonal, standing rules that restrict the range of eligible candidates. These rules are ‘always on’: they apply before the process starts, even before there is any vacancy. Some parties require that candidates are party members.³ Some allow only parliamentarians to run. Beyond these, however, most European parties have few formal barriers to entry into leadership contests (Pilet and Cross 2014, 230–231).

The second type of filter is nomination requirements. These are formal and, unlike eligibility requirements, they involve decisions in an active contest about real candidates. To fulfil nomination requirements, a candidate must secure an expression of support from individuals (such as signatures from a specified number of members) or units (such as endorsement from a certain number of local or regional branches). Rather like eligibility requirements, nomination requirements are, with some striking exceptions, generally not that difficult to meet. This, we argue, is often because there are more informal means by which some aspirants are blocked from continuing their attempts to win the party leadership.

Some of these informal means involve the normal stuff of politics. Candidates, their agents and their supporters will campaign and canvas with an eye to the selectorate, whatever the selectorate is in a particular party. The candidates may well also signal or bargain with each other, which may lead to deals. A candidate may withdraw, for example, if she receives a persuasive offer from a rival to join the latter's team. It may thus be an anarchic process, in the sense that, once the candidates are established, their competition is unregulated and a settled order is established only by agreement among one or more of them – at least until the selectorate has made its decision.

However, in some parties, order in the competition between candidates is established in a different way. There may instead be an act of precursory delegation, in which a ‘steering agent’ is given the task of filtering them so that only one ultimately makes it to the selectorate. A steering agent is an intra-party actor – a person or a group. Its status can be formal or informal, ad hoc or long-established. This agent has the mission and, at least to some degree, the capacity to set the agenda for the selectorate through promoting one aspiring leader at the expense of others. As an agenda-setter, the steering agent's own decision-making needs to be nimble, sometimes confidential and usually consensual – all of which puts natural constraints on its size. A steering agent's endorsement of a particular candidate to the selectorate can be seen as a non-binding decision. Yet it is a decision that often shapes the result of the selection more than anything in the preceding or subsequent phases.

And who or what is the steering agent's principal? From where does it derive its mandate? If we are interested in mapping the distribution of power within a party, this is perhaps the essential question. For now, though, we can leave it aside. We content ourselves here with noting that the crucial, defining characteristic of a steering agent is that it monopolises its role. Its very existence reflects either active involvement in, or more passive acceptance of, the act of precursory delegation on the part of the main power centre or centres within the party.⁴

However, if a power centre becomes seriously unhappy with the arrangement, either before or when the decision reaches the selectorate, the foundations of precursory delegation may erode and collapse altogether – possibly thus reflecting an important shift in the party's internal power structure.⁵ A party may attempt to control competition for the leadership through a steering agent. However, if that breaks down, another categorization of the case will be required.⁶ That said, precursory delegation is compatible with a ‘near-coronation’, in which the opposition to the steering agent's favoured candidate is marginal, or a ‘crypto-coronation’, in which marginal opposition to the favoured candidate is promoted by the steering agent itself, in order to legitimise the selection. In such cases, the steering agent's authority is not threatened and precursory delegation retains the support – or, at least, the acquiescence – of the party's main power centre or centres.

The final stage in the selection process, which interests us least here, is the management of the decision in the selectorate – whatever the selectorate might be.

The mode of selection: a typology

The subordinate categories in the typology have now been explained. They can now be pieced together in order to construct the target categories in what we call the mode of leader selection (see Figure 1). The cases allocated to each category could be individual leader-selection processes or – perhaps more meaningfully, but more demanding empirically – the tendency within a particular party over a number of selection processes.

Figure 1. Modes of competition in leader-selection processes.

In this section, we illustrate each target category with empirical descriptions of individual selection processes of which we have relatively detailed knowledge. To this end, we use data from 30 cases in parties from nine European countries. They were coded by the basis of country experts’ accounts, which were based on official party documents, such as statues and internal rule books, elite interviews with key party informants, and media reports (Aylott and Bolin 2021).

We focused on cases in which the outcome of the process was not given in advance. That is, we exclude routine renewal of an incumbent leader's mandate, which some parties require regularly. While our sample is not representative of the broader population of leader selection, our ‘diverse’ collection, typical of descriptive research (Gerring 2017, 60), offers a variation in geography, culture, history of democracy, electoral system, legislature-executive relations (majority or minority governments), party-system dynamics (bloc politics or more fluid inter-party relations) and – not least – rates of reforms to increase inclusiveness in intra-party decision-making (apparent especially in Britain, Estonia and Italy, less so elsewhere). There are also parties from the main European party families, the conservatives, Christian democrats, social democrats, liberals, greens and radical right (coded according to Europarty affiliation). Some parties are often in government; others rarely or never are.

The parties in our survey are categorized according to the typology in Table 1. In the following sub-section, we comment on these classifications.

Table 1. Classification of parties.

Country	Party	Selected leader(s)	Year(s)	Open	Enclosed	Filtered	Enclosed and filtered	Managed	Selectorate	Coronation
Estonia	Centre	Ratas	2016	1					D
Estonia	Reform	Pevkur	2017	1					M
Finland	Greens	Niinistö	2011			1			D	1
Finland	National Coalition	Stubb	2014	1					D
Finland	Social Democrats	Rinne	2014	1					D
Germany	Social Democrats	Schultz	2017					1	D	1
Germany	Christian Democrats	Kramp-Karrenbauer	2018			1			D
Germany	Social Democrats	Nahles	2018			1			D
Germany	Social Democrats	Walter-Borjans, Esken	2019			1			D
Italy	Five Star Movement	Di Maio	2017					1	M
Italy	The League	Salvini	2017	1					M/D
Italy	Democratic Party	Renzi	2017			1			S
Latvia	National Alliance	Dzintars	2017			1			D	1
Latvia	Unity	Ašerādens	2017	1					D	1
Latvia	Farmers Union	Krauze	2019			1			D
Lithuania	Order and Justice	Paksas	2014			1			D
Lithuania	Order and Justice	Žemaitaitis	2018			1			D
Poland	People's Party	Kosiniak-Kamysz	2015					1	C
Poland	Law and Justice	Kaczyński	2016					1	D	1
Poland	People's Party	Kosiniak-Kamysz	2016					1	D
Poland	Civic Platform	Schetyna	2016			1			M	1
Poland	Civic Platform	Tusk	2013			1			M
Sweden	Social Democrats	Juholt	2011					1	D	1
Sweden	Liberals	Sabuni	2019					1	D	1
Sweden	Greens	Fridolin, Romson	2011					1	D
UK	Conservatives	May	2016				1		M	1
UK	Conservatives	Johnson	2019				1		M
UK	Liberal Democrats	Swinson	2019				1		M
UK	Labour	Corbyn	2015				1		S
UK	Labour	Corbyn	2016				1		S

Selectorate.

D = vote by congress delegates.

M = membership ballot.

S = ballot of sympathizers.

C = party council.

* = no membership ballot was actually needed.

Source: Bolin and Aylott (2021).

Open competition. This first target category contains cases in which there are no significant eligibility or nomination requirements. There is, moreover, no discernible system of precursory delegation. In other words, non-membership of the subordinate categories significant eligibility requirements, significant nomination requirements and precursory delegation is necessary and jointly sufficient for membership of the target category open competition. In more everyday language (to which we revert in most of the rest of this section), it means that there are few formal restrictions imposed by the party on who can run for its leadership; and nor is there any attempt, conducted by one or more of the party's power centres, to pick a winner before the selectorate is involved.

Half a dozen cases in our survey are in this set. Two Finnish ones – the National Coalition and the Social Democrats (both 2014) – were clear-cut members. In those parties, precursory delegation used to occur, but collapsed in the early 1990s. Two Estonian parties offered interesting additional cases. In the Reform Party (2016), the usual pattern of precursory delegation broke down, leading to a surprise outcome. In the Centre Party (2017), a longstanding leader, whose management of opposition had usually involved reprisals after beating off challenges to his authority, was finally toppled in an internal revolt.

Enclosed competition. This target category contains cases in which there are significant eligibility requirements, but no significant nomination requirements and no precursory delegation. In other words, the party is restrictive in who it allows to run for its leadership; but once aspirants have met the formal, always-on demands, which everyone knows in advance, they proceed to compete freely with each other.

In our survey, we found no cases that could be attributed to this category. However, the main British parties could have been placed there before the organizational reforms that began in the 1980s. Their leaders were selected exclusively by their parliamentarians from among their own number (Quinn 2012). Competition among them was thus relatively unstructured, especially in the Conservatives before 1965. This procedure, enclosed among parliamentarians, is still in practice applied by the Democratic Unionist Party in Northern Ireland (Matthews 2016, 906).

Filtered competition. This target category involves a party without significant eligibility requirements or precursory delegation, but with a demanding set of nomination requirements. This was the category most populated by the cases in our survey, 11 of them. Two of them are further instances in which precursory delegation was attempted but broke down. In one of them, the German Social Democrats (2018), the outgoing leader tried to assume the same steering-agent role by which his own candidacy for the leadership had previously been promoted by his predecessor. Although his nominee did become party leader, his attempt provoked a grass-roots revolt, which resulted in a contested vote within the selectorate.

Indeed, the category of filtered competition appears to have become more populated in recent years. This may reflect a trend in which the hands of the formal selectorate are, in practice, tied by the encroachment of inclusive pre-selection processes, which we classify as nomination requirements. In the bigger German parties, for example, precursory delegation has been abandoned and advisory ballots have been used increasingly often to allow the party membership to signal a preference for one candidate (or sometimes a duo of candidates for joint leadership positions) before the formal decision is made by the party congress, as federal law demands (Jun and Minas 2023). Much the same applies in the Finnish Greens.

Enclosed and filtered competition. A few parties have both significant eligibility requirements and significant nomination requirements, but do not employ precursory delegation. The main British parties provide the cases in this category. They have unusually tough eligibility and nomination requirements. The Conservatives, indeed, demand not only that a leader is a member of its parliamentary group, but also that the person is one of the group's two favourites, a status that is determined by a series of elimination ballots. Only then can an aspirant become a candidate and be subject to the decision of the selectorate, which is the party membership.

It is worth noting that in some Conservative selections over the years – namely, that of 2016, which was in our survey, but also those of 2003 and 2022, which were not – the nomination requirements turned out to be so demanding that only one aspirant ultimately fulfilled them. This coronation disempowered the selectorate entirely: there was not even a ceremonial vote among members. Still, in most British cases, the candidates, once they get past the gatekeepers, compete openly. (A coronation, then, is not sufficient to indicate the occurrence of precursory delegation. Put another way, precursory delegation is not necessary for a coronation.)

Managed competition. This fifth and final target category involves parties that employ precursory delegation – regardless of what other rules (eligibility or nomination requirements) they also use to restrict competition. In technical terms, membership in the subordinate category precursory delegation is thus sufficient for membership in the target category managed competition.

In eight cases in our sample, precursory delegation in some form could be discerned throughout the selection process. In four of these, coronations ensued. In the other four, there were outcomes that came close to coronations. In two of these, both involving the Polish People's Party (2016, 2017), there were crypto-coronations. In this party, the different steering agents were obliged by the party statutes to find additional candidates, but they were never a threat to the overwhelming favourite. In the two other cases, there were near-coronations. In one of them, the Italian Five Star Movement (2017), careful observation reveals that the party's founders ensured, using various tactics (such as a very short campaign period), that their favoured candidate won. Some within the party grumbled; but there was no revolt.

The Swedish system of managed competition is almost unique in its degree of formalization.⁷ Parties have no formal eligibility or nomination requirements. However, like most organizations in Sweden that claim to run themselves democratically, they appoint a selection committee – in Swedish, valberedning (literally, election or choice preparation). At national level, the job of this committee is to assess and propose individuals to fill particular roles in the organization, including party leader, whom the selectorate then considers.⁸ The selectorate does occasionally reject the recommendation. But acceptance is far more common, which then usually produces a coronation of the new party leader. Membership of the valberedning is public information. So too is membership of its principal – invariably, the party congress.

The Swedish Social Democrats manage competition especially robustly. A combination of institutions and custom leaves no room for open competition between candidates. The process is managed in such an ordered manner that any public expression of personal interest in becoming leader is usually seen as tantamount to self-disqualification from the selection process. Apart from occasional, isolated expressions of dissent, the legitimacy of this heavy control has hitherto barely been questioned within the party.

As well as the Social Democrats’ selection (2011), an additional Swedish case was more equivocally one of managed competition. The Liberals (2019) reflected a trend in smaller Swedish parties in which selections have moved somewhat towards filtered competition. In these parties, the valberedning now tends to nominate a shortlist of candidates, who then campaign openly and competitively in seeking support from the selectorate, the party congress. In defining a shortlist, therefore, the valberedning utilizes an arbitrary power to set its own nomination requirements. Moreover, it still ultimately recommends a single candidate to the selectorate, which is then unanimously or at least broadly accepted – the defining feature of precursory delegation. Nevertheless, in permitting a degree of open competition, the valberedning concedes a diminution of its own discretion. If its preferred candidate is not obviously the preference of some important section of the party, which is then revealed in the open competition between them, the valberedning may find it hard to endorse that person.⁹

Conclusions and future research

What do we achieve with our typology? We show that it is possible to categorize validly the mode of selection that individual parties utilize when they choose a new leader. This, in turn, reflects the party's attitude to competition for its top job. At one end of the scale, a party may have a relaxed, even welcoming approach to the prospect of aspirants competing against each other. At the other end of the scale, the competition is tightly policed, so that it takes place almost completely behind closed doors. In between those two poles, parties may restrict open competition using various means. Eligibility requirements, ‘always-on’ rules that everyone knows in advance of any selection process, may quite drastically limit the field of potential leaders. Nomination requirements involve choices made by certain intra-party actors about aspiring leadership candidates. Such requirements can be largely symbolic or very demanding.

We can thus arrange our five modes of competition into a nearly ordinal scale (see Figure 2). At one end is an ordered power structure, in which what Panebianco (1988, 56–59) called the ‘systemness’ of the organization is firm and established. The systemness can involve autocratic or negotiated procedures, but it is predictable and introverted, in the sense of being resistant to outside influence. At the other end of the scale is an anarchic power structure, in which the structure of competition is unpredictable and there is greater scope for outside influence. The scale is only nearly ordinal because we consider enclosed competition and filtered competition as level between the poles of anarchy and order.

Figure 2. Modes of competition and the structure of intra-party power.

It is important to note two particular characteristics of any leader selection, neither of which the typology addresses. The first is the selectorate. There is no obvious link between the mode of selection and the section of the party that has the formal role of choosing its leader. In our survey, managed competition could be combined with any one of various different selectorates – a party council, delegates to the party congress, party members, even party sympathizers.

That, in turn, underlines what we noted at the outset of this paper: that the mode of selection does not indicate where, in any particular party, intra-party power resides. Even in the clearest cases of precursory delegation, in which the steering agent controls the competition between candidates with a heavy hand, the agent can receive its mandate either from one among the various ‘faces’ of party organization, as Katz and Mair conceived them (1993), or by a consortium of two or more faces. Moreover, the face or faces that delegate to the steering agent could be associated with the elite, or with the grass-roots, or even with actors external to the formal party organization.

Nevertheless, we argue that the mode of selection employed by a party can be one telling indicator, among others, of where power does lie within its organization. This is because the mode tells us where to look. If, for example, competition for the leadership is enclosed and/or filtered, but not managed, our first questions should be the following. Which intra-party actors and sections are inside the enclosure – and presumably seek to retain their privileged position? Which actors have the power to nominate? Do the answers to those questions suggest a powerful parliamentary group, or central office, or regional branches? Furthermore, when competition is managed, we should ask which actors or sections delegate to the steering agent. Is there one dominant power centre, keen to preserve its status? Or does a consortium of power centres have some other shared objective – perhaps to preserve the image of party unity, perhaps simply to maintain the balance of intra-party power? Answers to such questions should help us to assess whether the ‘democratization’ of leader selection – that is, greater inclusivity in the process – has disguised the maintenance of control by party elites, or whether it reflects a genuine flow of power downwards within organizational hierarchies.

Clearly, then, much empirical work and interpretation remains in the study of party leader selection. The mode of selection, however, offers an essential part of the roadmap for that endeavour, which might ultimately promote our understanding of both national and international politics.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the Foundation for Baltic and East European Studies under grant ÖSS dnr 19/18.

Notes

1 In the related field of candidate selection, Tuttnauer and Rahat (2023) suggest measures of ‘inclusiveness’ and ‘complexity’ to account for the existence of ‘multiple selectorates’. This, too, focuses on formal procedures, as regulated by statutes, because the data with which the measures are demonstrated is drawn from the Political Party Database (see Scarrow, Webb, and Poguntke 2022). Our approach aims to capture both formal and informal procedures, which, we argue, are sometimes functionally equivalent.

2 Some members of the Swedish Greens drew this conclusion in autumn 2023 when they realised that there was nothing to prevent a certain person who had only recently joined the party, and who advocated use of the death penalty to punish some crimes, from being a candidate for one of the party's two leadership positions (SR 2023). About the same time, in the Greek party Syriza, a recent member actually won the party leadership, despite running on a vague policy platform (Politico 2023). Thanks to an anonymous referee for reminding us of this latter case.

3 A decision on whether to grant an individual party membership might be very salient to that individual's chances of standing for the party leadership. In that case, the decision might be better classified as the next type of procedure, nomination requirements. But such cases are likely to be rare. Thanks to an anonymous referee for raising this possibility.

4 By power centre, we mean a group of intra-party actors that resembles Panebianco's (1988, 37–40) idea of a ‘dominant coalition’.

5 There is thus a distinction between active involvement in precursory delegation and passive acceptance. The former indicates a division of intra-party power, the latter a concentration.

6 Thus, the moment at which the case is classified is at the conclusion of the process, when the new leader is selected, rather than nearer the start, when precursory delegation might be attempted.

7 Norwegian parties have a similar procedure (Allern and Karlsen 2014). We know of no other European parties with anything quite like it.

8 Equivalent committees at lower levels propose the party's election candidates.

9 This was exactly what happened in the Swedish Liberals (2019). After the valberedning had nominated a shortlist of three approved candidates, who were then permitted to compete openly with each other, a wave of what might be called ‘wildcat’ membership ballots broke out in the regional branches of the party. These turned the tables on the steering agent. In practice, the valberedning could not disregard the results of the ballots, even though it meant endorsing a candidate that it probably would not otherwise have preferred (see also Aylott and Bolin 2023b).