Conflict regulation in political parties

Abstract

Independent party tribunals (i.e. intra-party courts) can be used by both the party leadership (e.g. to discipline members) and rank-and-file members (e.g. to challenge the leadership overstepping its authority). Thus, their study offers broad insights into party conflict regulation we know little about. Integrating the literatures on party organization, intra-party democracy and judicial politics, we propose two theoretical rationales to account for tribunal decision-making (whether a case finds tribunal support): tribunal decision-making can be theorized as shaped by elite-member divisions or, alternatively, by how verdicts affect the tribunal’s own position in the organization and organizational stability generally. We test hypotheses derived from these rationales using a new data set covering 243 tribunal decisions made over the life spans of three German parties. While both rationales are empirically relevant, the ‘organizational stability rationale’ proves particularly insightful.

Keywords

conflict regulation intra-party conflict intra-party democracy new parties party tribunals German parties

Introduction

Whether or not political parties can cope with internal diversity and, more crucially, handle conflict is important for their ability to protect their organization’s integrity and to successfully pursue their goals. Organizational theory has long emphasized that all complex organizations – such as political parties –need to be able to reconcile internal differences on a day-to-day basis (March, 1988; Rahim, 2002). Yet even within works conceptualizing parties as organizational actors, the study of intra-party conflict has remained a marginal theme. As van Haute recently puts it in an insightful piece on the neglect of member discontentment: ‘the literature on party membership often takes for granted that members are happy, loyal and love and support their party’ (2011: 170). We also know little about how elites handle divisions in their organization. Data collection is difficult, not least because intra-organizational turmoil is considered a sign of weakness, that is, a reflection of the inability of parties to keep ‘their own house in order’. As a result, parties have an interest in downplaying or, even better, hiding internal conflict. The capacity to maintain or – at least – publicly display internal cohesion is considered as an important performance indicator, especially for new parties that have not yet proved themselves in parliament or in government (e.g. Art, 2011; Burchell, 2001; Poguntke, 1993). Consequently, the scenarios that are most intensively studied tend to be those that are least representative of intra-organizational conflict regulation: spectacular cases of disintegration, where conflict escalates and is displayed in public – be it by leading figures, rank-and-file or both – regardless of the damage done to the party’s image or functioning.¹ Similarly, if existing work deals with intra-party conflict and its regulation, it does so in association with specific events (e.g. organizational reform), in particular arenas (e.g. as triggered by coalition dynamics and government participation) or in specific types of parties (e.g. heavily factionalized parties).² Day-to-day conflict regulation in party organizations, that is, the procedures and strategies to manage internal differences and thereby assure parties’ ongoing functioning are more often assumed to exist than being examined empirically (see, for exceptions, Jungar, 2013; Smith and Gauja, 2010: 768–771).

Against this backdrop, this article makes a first attempt to theorize and empirically examine conflict regulation in political parties with a focus on the decision-making of party tribunals. In particular, this article theorizes and examines the factors driving the outcomes of tribunal decisions, with the dependent variable capturing whether cases were won or lost from the perspective of the initiator. These intra-organizational tribunals – that correspond (if they are formally independent from other party organs) to court structures within democratic states – represent only one mechanism of conflict regulation as detailed below. Yet their study is particularly insightful; first, they can be initiated and thus used both by the party leadership (e.g. to discipline party members or subunits perceived as doing damage to the party) and by ordinary rank-and-file members (e.g. to challenge leadership decisions that fall outside the remit of the latter’s authority). They are not a mechanism of either top-down or bottom-up control that procedurally privileges one group of internal actors over the other. Second, and closely related to the first point, unlike other hierarchical mechanisms of conflict regulation (that allow for the imposition of solutions on actors against their preferences), tribunals deal with a wide range of issues: disagreements over the interpretation of party rules, the monitoring of actors’ compliance with party procedures or the sanctioning of damaging behaviour by members, party officials or party organs. Consequently, a framework on factors driving tribunal decision-making – whether those initiating cases find support of the court or not – can generate broader insights into patterns of conflict regulation, processes that otherwise are widely hidden.

Integrating research on intra-party democracy, party organizations and other complex organizations as well as judicial politics, we propose two rationales to theorize tribunal decision-making. While these rationales are not mutually exclusive, they rest on different assumptions of what the main (not necessarily the only) motivations of tribunal members might be when making a decision. They thereby allow us to formulate systematic hypotheses on a so-far undertheorized phenomenon. While we derive a set of distinct hypotheses from each, we also consider their intersection. One can theorize tribunal decisions as shaped by internal lines of conflict, most notably by divisions between members of the party elite (a group defined by the holding of status functions in the organization) and rank-and-file members (without such responsibilities) which we call the ‘elite partiality rationale’ – a rationale that stresses the importance of who (elite or ordinary member) initiated and who is targeted by a case. Alternatively, one can theorize tribunal decisions as shaped by the assessment of tribunal members (the ‘judges’) regarding how their decisions affect the tribunal’s own position in and the stability of the organization it forms part of – called the ‘organizational stability rationale’ stressing the importance of the type of case (appeal against prior verdicts vs. new initiative; cases asking for the exercise of sanctions vs. the correction of procedural errors) and the maturity of the organization when a case is dealt with in a party’s life cycle.

The following section locates party tribunals among other mechanisms of conflict regulation and thereby identifies them as a suitable analytical focus to study party conflict regulation generally. Then we develop the two theoretical rationales on tribunal decision-making and present out hypotheses. We test the hypotheses using a unique data set of 243 tribunal decisions, covering all national tribunals’ verdicts made in the course of three German parties’ life cycles: the Greens (1980–2013), the Left Party (2007–2013) and the Pirate Party (2008–2013). The results suggest that the maturity of the party at the time of decision-making is particularly important, in line with the ‘organizational stability rationale’. Simultaneously, elite initiatives have a higher chance of finding support as long as they do not appeal against earlier decisions, thus, ‘elite partiality’ plays a role yet only if it does not clash with a tribunal’s endeavour to protect the court system stabilizing conflict regulation in the organization. The article concludes with the broader repercussions of the findings and avenues for future research.

Conflict regulation in political parties

Following Rahim, we define ‘conflict regulation’ as (a) the prevention of diversity and differences in opinion in an organization from translating into conflict and (b) once conflict materializes the capacity to reduce or suppress it (2002: 206). Modes of conflict regulation can rely on deliberation/persuasion, on negotiation/bargaining or on hierarchy/sanctions. The main difference between hierarchical mechanisms and non-hierarchical ones such as persuasion and negotiation is that the former allow one actor or set of actors to impose a decision on others against the latter’s preferences.³ Hierarchical sanctions seem counterintuitive in voluntary organizations that actors are free to leave at any time (Wilson, 1973), which is why they are little studied in the respective literatures. Still, hierarchical sanctions are common in political parties and tend to range from warnings over temporary removal from internal status functions such as party or public office, members’ ineligibility for such status positions (implying the withdrawal of special rights or privileges usually available to members) to expulsion (permanent removal from the organization). While most parties use both hierarchical and non-hierarchical mechanisms, political parties differ in how easily or by whom these mechanisms can be used. While in hierarchical party organizations the expulsion of members might be the prerogative of the executive or even single leader, in parties that aspire to intra-organizational democracy this mechanism can require extensive approval, for example, by the party congress, a membership poll or an independent tribunal (or a combination thereof). Figure 1 links the nature of conflict regulation (hierarchical vs. non-hierarchical) from the viewpoint of those affected by the outcome of decisions with whether a mechanism is controlled by the leadership or not, reflecting a leader-centred or member-centred approach of conflict regulation, respectively (Smith and Gauja, 2010: 769).

Figure 1.

Procedures of conflict regulation. The procedures classified are examples of basic configurations and do not give an exhaustive overview of mechanisms.

Party tribunals are formally independent and can be considered as equivalent to ‘state courts’ (Figure 1, grey quadrant) if neither members of the party in public office nor in central office (the two groups that constitute the party leadership) are allowed to take over a formal role in them. In other words, party tribunals need to be composed of members of the party on the ground, that is, of ordinary rank-and-file members (Katz and Mair, 1995). In order to ensure the independence of tribunals, members are regularly elected and not appointed by the party leadership. Formal independence is crucial since it puts rank-and-file members – the party on the ground – and members in leading positions on an equal footing when it comes to getting a positive verdict from tribunals. Only under the conditions of formal independence, we would realistically expect that both groups of party actors are likely to get a verdict in their favour.⁴ Consequently, we would expect our theoretical approach to hold for decisions made by such tribunals.

This set-up further suggests that independent party tribunals deal with a wide range of issues such as the violation of party procedures (by leaders or members alike), conflicts between party organs, as well as calls for the expulsions of the full range of party actors – be it activists, public representatives or whole local units – whose behaviour is considered by the plaintiff as damaging to the organization. Given these basic characteristics of formally independent party tribunals, the analysis of their decisions offers us broad insights into factors driving conflict regulation in political parties more generally.

Theorizing tribunal decision-making

Tribunal verdicts as shaped by ‘elite partiality’

Ideal-typically, party tribunals set up as independent intra-organizational ‘courts’ ought to assure equal treatment of party members within the organization, whether they hold status functions or not. Thus, they are in charge of assuring the neutral and fair resolution of disputes. In the by-now prolific literature on intra-party democracy, however, their capacity to protect ordinary members’ ability to express dissent and to defend their intra-organizational rights is contested. As Rahat, one leading contributor in this debate, points out:

Parties in democracies can have judicial or semi-judicial organs that deal with complaints concerning breaches of the freedom (and fairness) of their internal elections. Yet, these organs are suspected (often with reason) of being partial. (2013: 141)

While tribunals ought to prevent the abuse of privileges or rule violations on behalf of leading personnel in public and in central office, the manipulation of tribunal decision making is considered likely. Similar perspectives we find in various classical works on party organization. They have long stressed how organizational leaders dominate members despite formally democratic structures, most famously expressed in Michels’ ‘iron law of oligarchy’ (Duverger, 1964; Michels, 1962: 64, 70; for more recent portrayals of elite-member relations echoing this tradition, see Katz and Mair 1995; Katz 2001). Even if membership in a party’s executive or holding public office is incompatible with membership in a party tribunal, tribunal members, once in office, become part of the elite, that is, of those party members holding status functions, a group of that simultaneously encompasses the party leadership. As far as this creates an affinity and informal affiliation between these two groups that share – unlike rank-and-file members – special responsibilities in their organization, neutrality of tribunal decision-making could be undermined. Thus, assuming that elites (encompassing tribunal members and leadership alike) and rank-and-file members hold distinct interests (Maor, 1998: 135; Tolbert and Hiatt, 2009: 178), we can formulate the following hypotheses:

Hypothesis 1 (Elite Initiative Hypothesis): Cases initiated by members holding status functions, that is, belonging to the party elite, are more likely to be won than those initiated by ordinary rank-and-file members not holding any such positions.

Hypothesis 2 (Elite Target Hypothesis): Cases directed against members holding status functions, that is, belonging to the party elite, are less likely to be won than those against ordinary rank-and-file members not holding any such positions.

Tribunal verdicts as shaped by organizational stability considerations

Unlike the ‘elite partiality’ rationale, this rationale does not start out from the notion that tribunal decisions are shaped by intra-organizational divides between groups with distinct interests, undermining the impartiality of decision-making. Instead, it theorizes decisions as being shaped by the assessment of the tribunal of how its verdicts affect the position of party tribunals within the organization and – being part of the elite with special responsibilities – the stability of the organization overall (Farazmand, 2002: 181; Rahim, 2011). Following this rationale, we can identify three alternative factors as relevant to tribunal decision-making.

If considerations around organizational stability are central, the maturity of the organization itself can be expected to affect how a tribunal – embedded in it – evaluates the implications that positive or negative verdicts are likely to have when facing a case. Thus, a party’s institutionalization is likely to be a central property. Drawing on Panebianco (1988), one central dimension of institutionalization is value infusion, a process whereby party followers (ordinary rank-and-file members and elites – members holding status functions – alike) start caring about the survival of their party as such separate from the goals or values the latter stands for (Levitsky, 1998: 82).⁵ The nature of this process suggests that members elected to tribunals become less likely to support cases, as the party in which they are embedded becomes more institutionalized. This is because in early phases of development, the affiliation of tribunal members (as of all other members) to the party organization as such is still weak. Responsiveness to the party’s membership is not yet counterbalanced by functional considerations of organizational stability, suggesting a broad willingness to support challenges to the organization initiated by its members. This is all the more the case since in early periods of the party life cycle the allocation of power in a party organization (including the power of tribunals) is still contested and tribunals need to consolidate their own position assuring broad approval. Once the protection of the organization as such becomes – through growing value infusion – a central goal for tribunal members and their position becomes increasingly consolidated, organizational stability becomes a more dominant concern, potentially weakened by the overturning of too many internal decisions. This leads to the following hypothesis:

Hypothesis 3 (Institutionalization Hypothesis): The more institutionalized a party organization in which a tribunal is embedded in at the time a case is decided, the less likely a case will be won.

If the stability of the organization is a core consideration, we can further expect that tribunal members assess the specific costs that supporting or turning down a particular case generates for the organization. These costs are different depending on whether the tribunal deals with expulsion or compliance cases. Expulsion cases are usually responses to behaviour perceived as severely damaging to the party and tend to be considered as last resorts, even in authoritarian parties where elites have a lot of discretion to use this mechanism. At the same time, expulsions can be essential to protect a party from intra-organizational turmoil and destructive behaviour and to sustain its organization’s integrity (Cross and Katz, 2013: 181). The decision of tribunals to reject such a case can thus have severe consequences for the party. Furthermore, the costs of verdicts are predominantly carried by the expelled, not the organization that the tribunal represents. Considered from the perspective of the initiator asking publicly for another member (or unit) to be penalized with the harshest sanction available, we can expect such cases not to be initiated lightly given that – if turned down – all participants will have to continue to work together in the same organization. This should decrease the number of expulsion cases coming forward in the first place and those that go ahead should be substantively well founded, that is, be of higher quality. Similar dynamics are less likely when conflict evolves around procedural matters, where personal stakes are lower (Pondy, 1992: 257). Such ‘compliance cases’ involve the formal challenging of internal decisions or activities that, if rule violations are confirmed, can be overturned by tribunals and require the reversal or repetition of decision-making processes, which – different from expulsion cases – generate considerable costs for the organization itself (e.g. by having to re-run internal selection processes).Consequently, assuming organizational stability to be a prior concern for tribunals, we can formulate the following hypothesis:

Hypothesis 4 (Expulsion Hypothesis): Cases that ask for expulsions are more likely to be successful than compliance cases dealing with procedural issues.

Finally, drawing on the literature on judicial politics theorizing court decision-making, especially in multi-tiered systems, leads us to another central distinction. EU scholars have argued that courts aim at consolidating and protecting their own authority as well as the authority of courts they interact with (e.g. Bier, 2008; Burley and Mattli, 2006),⁶ thereby stabilizing the legal system they are embedded in. Applying this argument to intra-party tribunals, cases that appeal against earlier tribunal verdicts should be less likely to be supported by tribunals than first-time initiatives and this should hold no matter whether an appeal challenges the court’s own decision or the decision of a lower level court. This is substantiated by the literature on courts operating in the same legal system such as the United States. Courts are disinclined to overturn earlier court decisions due to deference, thus “lean toward affirmation as the usual course” (Clermont and Eisenberg, 2001: 131). Furthermore, appeals have undergone a ‘legal check’ already and – different from new cases – the court’s task is to identify an error in the initial proceedings narrowing down conditions under which affirmation is rejected (Clermont, 1987: 1126). Thus, high affirmation rates are common, even though one might expect that only ‘shaky’ decisions of lower judicial quality are appealed to start with, which – in theory – should increase the likelihood of success. This, however, does not hold. For once, appeals are not costly (Posner, 1985: 195) and might function – for those initiating them – less as an instrument to correct ‘legal’ errors than as a means to be granted a fairer process or gain satisfaction (Barclay, 1999: 1–5). Both arguments go against a thorough ‘self-selection’ of appeals prior to submission to the court and are – if anything – more convincing in voluntary contexts where roles and procedures are less formalized, leading to the following hypothesis:

Hypothesis 5 (Appeal Hypothesis): Appeal cases are less likely to be won than new initiatives.

Tribunal members as part of the ‘court system’ and part of the overall elite

As H1–5 illustrate, we can use the ‘elite partiality rationale’ and the ‘organizational stability rationale’ as heuristic devices to identify different factors as likely drivers of tribunal decision-making, respectively. The rationales, however, might also operate simultaneously as the same cases can combine elements associated with either, especially when the mechanisms underpinning hypotheses though belonging to different rationales (i.e. ascribing different motivations to tribunal members) resemble each other as in the following: Our Elite Initiative Hypothesis (H1) points to an affiliation between tribunal members and all other holders of status functions (including the leadership) and expects elite initiatives to be more likely to find support as a form of partiality strengthening the elite. In contrast, the Appeal Hypothesis (H5) points to, more particularly, an affiliation between tribunals operating at different levels of the party organization (a subset of the overall elite including all holders of status functions) and therefore expects appeals to be rejected, thereby stabilizing specifically the ‘court system’. These affiliations co-exist as the decision-making tribunal belongs to both groups which are, in turn, overlapping. Organizational theory and social psychology alike argue that affiliations are likely to be stronger to those groups with which similarities are more pronounced, for example, whose members share the same organizational roles and responsibilities (Dovidio et al., 2009; Fiske and Taylor, 1991; Turner, 1987). This suggests that elite initiatives are likely to be favoured by the tribunal only as long as they do not appeal against earlier tribunal decisions, leading to our last hypothesis:

Hypothesis 6 (Elite Initiative vs. Appeal Hypothesis): Cases initiated by members holding status functions, that is, belonging to the party elite, are more likely to be won than those initiated by ordinary rank-and-file members, if they do not appeal against earlier verdicts.

Rationale for our selection of parties

As we find independent tribunals in democracies where party law requires parties to assure their members’ right to dissent or prescribe the type of tribunal to be established (Biezen and Piccio, 2013), German parties are suitable for our analysis. German Party Law prescribes three areas that affect intra-organizational conflict regulation directly or indirectly: ‘rights of members’, ‘party tribunals’ and ‘measures against regional party units’ (De Petris, 2000: 54). It requires tribunals to be independent, thus sitting on a tribunal is incompatible with membership in the party executive (Biezen and Piccio, 2013: 39–40). While legal regulation allows party executives to decide on the removal of members from party offices, expulsions from the organization can only be handled by tribunals.

Reflecting this legal environment, the statutes of the Greens, the Left and the Pirates (as the case in other German parties) establish tribunals whose independence is ensured by a number of intra-organizational rules⁷: incompatibility of tribunal membership with any party office,⁸ members of tribunals have to be (re)elected at least every two years by the party congress, all participants in a case – judge, plaintiff or defendant – can apply to request the replacement of a judge, given suspicion of bias and judges are obliged to report any conflict of interest they might encounter in a case and have to excuse themselves from the trial. Finally, tribunals consist of several judges – from five in the Greens to 10 in the Left – and they can make decisions only if a quorum of at least three judges (Greens and Pirates) or five judges (Left) is reached.⁹

The choice of these three parties has several methodological advantages regarding our ability to test our hypotheses due to the composition of the resulting sample of tribunal decisions and the contextual conditions in which these decisions were made (Slater and Ziblatt, 2013: 1311-13): First, the German Party Law was adopted in 1967 and all three parties were formed afterwards, which keeps constant the formal independence of tribunals and core procedural requirements regarding tribunals’ composition and operation throughout their existence, factors that might have affected tribunal decision-making otherwise. This is crucial to test our hypotheses – especially those on tribunal partiality (H1 and 2) – in an unbiased fashion. This also means that for our three parties (unlike older German parties), tribunal data are available over their full life cycle (founding date – end of 2013). This assures coverage of decisions made in early stages of party development which is important for testing hypotheses derived from our ‘organizational stability perspective’ – especially the Institutionalization Hypothesis (H3). Second, all three parties are participatory, inclusive and grant considerable rights to rank-and-file members of which the right to dissent is only one. In those parties, it is more likely that both party elites and ordinary members will actively use party tribunals enhancing the diversity of tribunal decisions covered and with it the sample’s representativeness of the types of conflicts parties have to deal with generally (see Appendix 1, Table A1). On the one hand, facing a participatory membership, referring disputes to party tribunals becomes a more important tool for elites to manage diversity than for elites in more hierarchical parties. This is because in the latter, elites are structurally stronger and are likely to possess alternative means to control dissenters. On the other hand, members in participatory parties will be more inclined to use tribunals to hold the elite accountable (members in authoritarian parties are bound to be more accepting of top-down decisions by elites, for reasons of self-selection). Third, the three parties share these – for our analysis – important properties even though they are ideologically relatively distinct. In the context of the German party system, the Left is considerably further left than any other party when positioned on a left–right scale, while the Greens are very close to the Social Democrats taking a centre-left position. The Pirates, in turn, are located further on the right between Social Democrats and Christian Democrats, closer to the latter (Bender et al., 2013: 8; see also Zolleis et al., 2010). The same argument can be made when considering party placements on the GAL–TAN dimension. The Greens and Pirates are positioned below average on the libertarian side of the spectrum, while the Left is positioned above average between the Social Democrats and Christian Democrats on the traditional-authoritarian side (Bakker et al., 2015). In sum, an ideological bias in our sample of verdicts is therefore unlikely.¹⁰

Data and first overview

Our data set contains 243 tribunal verdicts, covering each party’s overall lifespan, that is, all cases their national tribunals decided on, respectively (see Table 1). The data on the tribunal cases of the Greens are drawn from the database compiled by the Institut für Deutsches und Internationales Parteienrecht und Parteienforschung (PRuF).¹¹ Cases for the Left Party and the Pirate Party were compiled from the parties’ websites.¹² To our knowledge, such data have not been systematically analysed so far. Neither has tribunal decision-making been systematically analysed as such.

Table 1.

Tribunal decisions in three German parties.

Party name	Period covered	Cases per party
Pirates	2008–2013	86
Greens	1980–2013	77
Left	2007–2013	80
Cases in total		243

Table 1 shows that despite having existed much longer, the total number of cases that the Green Party tribunal had to deal with is lower than the number of Pirate Party cases and the number of Left Party cases, both of which were founded more than 20 years later.¹³ At first sight this might indicate that the Greens are overall a more peaceful party or that internal decisions might be reached in a more inclusive manner and are less challenged. However, the graph also shows that the Greens had conflictual periods, such as in the mid-1980s when entering a government coalition in Hessia, triggering fundamental debates about the nature of the party, which fed into conflict over party procedures such as the rotation rules for office holders. The early 1990s were marked by intense conflict between ‘Fundis’ and ‘Realos’ (ideologically driven vs. more pragmatically orientated members of the party) that nearly led to a split of the party (Doherty, 1992; Poguntke, 1993). Simultaneously, the first half of the 90s was marked by the prolonged merger between the Greens (West Germany) and Bündnis 90 (East Germany) into one party organization, officially only completed in 1993, which further increased intra-party conflict (Poguntke, 1998).

Also in the Left Party conflict was relatively low in the early years but increased sharply after the 2009 national election. The Left is a merger of two parties, the WSAG (Wahlalternative Arbeit und soziale Gerechtigkeit) and Die Linkspartei (former PDS, Partei des Demokratischen Sozialismus), that merged in 2007 to maximize electoral support (Bukow, 2010: 181; March, 2011). In the run up to the 2009 election, the newly formed party tried to portray itself as unified for campaign purposes, but once the election was over, differences between the constituent parties started to emerge in full (Neu, 2009: 199). A similar picture emerges when looking at conflict within the Pirate Party. As long as the party was electorally successful, conflict remained low; but after experiencing some major setbacks at both national and regional elections, conflict became more pronounced (Klecha and Hensel, 2013; Zolleis et al., 2010).

To sum up, the peaks in the number of national tribunal cases in Figure 2 are in line with what case study research suggests with regard to how critical events (e.g. new challenges such as government entry or electoral failure¹⁴) in the histories of the three parties have shaped patterns of conflict. This is important since it indicates that national tribunal decisions (as one mechanism of conflict resolution) provide us with a valid picture of conflict in these parties more generally.

Figure 2.

Party tribunal verdicts by party (1980–2013).

Operationalization of explanatory variables

To test our hypotheses, we created a new data set of tribunal decisions based on the tribunal documentation as initially released by the parties. Thanks to the high level of formalization of these party procedures and the clear documentation of each tribunal case, coding decisions, guided by a codebook capturing all the core variables, were straightforward and unambiguous.¹⁵ To assure inter-coder reliability each case has been coded twice by separate coders.

The dependent variable – whether the initiator is supported by the tribunal or not indicating the winning or losing of a case – was coded based on the section of the documentation presenting the tribunal decision trial (always starting with the words ‘the court decided …’). This section presents either the initiator’s claim as justified or his or her challenge as granted (coded 1) or dismissed by court (coded 0).

Cases initiated by an ordinary rank-and-file member or, alternatively, a member of the elite (H1) are clearly distinct as ordinary rank-and-file members are only listed by their initials (e.g. A.B.), while members holding status functions are identified with their initials and the position or unit their represent (e.g. C.D. chair of local party branch or C.D. representing the national executive). This variable was coded as a 1 if the initiator formed part of the elite and as a 0 if he or she was an ordinary rank-and-file member instead. In order to identify the target of a case, the same unambiguous coding strategy could be applied (H2). Furthermore, allowing for the clear-cut coding of the initiator and the target variable, the role of the individuals within the court procedure is clearly labelled in the case protocol as either plaintiff or defendant. In all three parties elite initiatives are less numerous than member initiatives, ranging from 10% in the Pirate Party to just under half in the Green Party with 49%. Clearly, tribunals are actively used by members who do not hold any status functions. The particularly low number of elite initiatives in the Pirate Party echoes the extreme structural weakness of elites, especially on the national level, that characterizes this organization even when compared to membership-oriented parties such as the Left and the Greens (Niedermayer, 2013). Furthermore, all parties had to deal with more cases targeting elites than those targeting members, with 65% in the Left, 69% in the Green Party and 89% in the Pirate Party, reflecting their inclusive intra-organizational processes and participatory credentials.

We measure the institutionalization of the party when a case was decided (H3) by computing the difference between the date of the party’s foundation and the day when the case was decided. We divide the number of days by 365 to get as precise a measure in years as possible. The coding of this variable is straightforward as in all cases the date of the court ruling is included in its protocol and parties’ founding dates could be drawn from earlier studies (Bukow, 2010; Klecha and Hensel, 2013; Poguntke, 1993).¹⁶ The expulsion variable (H4) and the appeal variable (H5) were coded based on the protocols that document each individual tribunal case in detail. Expulsion cases (coded as 1) could be identified whenever protocols – describing the case at hand – explicitly refer to sections in a party’s statutes regulating expulsion as one penalty for particular types of behaviour in the context of the organization. Also, when dealing with expulsion cases the subject of the court proceeding is already included in the protocol’s title. Other cases were coded as 0. They refer to (usually procedural) violations of intra-party rules. The Pirate Party has the lowest rate of expulsion cases with 8% compared to 35% in the Left Party and 30% in the Green Party. Unsurprisingly, compliance cases dealing with procedural issues formed the majority in all three parties. Also the coding of the appeal variable was straightforward: the first lines of the protocols of appeal cases make reference to the initial case (including case number and date) in which the issue at hand was initially discussed and decided. In some cases, the status as appeal is even included in the title of document. Furthermore, appeal protocols outline – with references to the party’s statutes – whether all procedural condition for an appeal were met or not. In all three parties appeals are similarly frequent: 47% of the cases in the Pirate Party were appeals, 64% in the Left Party and 65% in the Green Party. For a complete overview of case characteristics, see Appendix 1 Table A1.

Empirical analysis

Our analysis seeks to identify the factors driving the outcomes of tribunal decisions, with the dependent variable capturing whether cases were won (N = 97) or lost (N = 146) from the perspective of the initiator. Since the dependent variable is a dummy, we fit the model with a logistic regression. Table 2 presents the results in log odds. Model 1 (M1) includes the main effects of all our explanatory variables, while model 2 (M2) adds the interaction between Elite Initiative and Appeal as theorized in H6. As we use cross-sectional pooled data, we apply a fixed-effects technique including party controls (the Left is the comparator category).

Table 2.

Factors shaping tribunal decision-making: Logit models.¹⁷

DV: The case is won by its initiator	M1	M2
Ivs
Elite initiative hypothesis (H1)	0.391 (0.35)	1.497* (0.59)
Elite target hypothesis (H2)	0.075 (0.43)	0.104 (0.44)
Institutionalization hypothesis (H3)	−0.104** (0.04)	−0.101** (0.04)
Expulsion hypothesis (H4)	0.190 (0.42)	0.296 (0.43)
Appeal hypothesis (H5)	−0.403 (0.29)	0.037 (0.34)
Elite initiative × Appeal hypothesis (H6)		−1.698* (0.68)
Constant	0.078 (0.43)	−0.218 (0.45)
Controls
Pirates	−0.186 (0.35)	−0.247 (0.35)
Greens	0.932 (0.49)	0.706 (0.5)
Overall model evaluation
Nagelkerke R ²	0.087	0.12
BIC	354.74	353.71
Log likelihood model	−155.4	−152.14
Log likelihood null model	−163.5	−163.5
Hosmer and Lemeshow test	χ ²(8) = 11.23, p = 0.19	χ ²(8) = 6.84, p = 0.55
Observations	243	243

^aStandard errors in parentheses.× stands for interaction.

*p < 0.05; **p < 0.01.

The overall model evaluation suggests that M2 which includes the interaction effect fits the data better than M1. Indeed, BIC statistics for M2 is lower; simultaneously the log likelihood for M2 in comparison to the null model is less than the same measure for M1. Also, Nagelkerke pseudo R ² is higher for M2. That means we find less unexplained variation in M2 than M1. Overall, M1 is significant at 5% level, while M2 is significant at 1% level. As M2 fits the data best, we interpret the results of M2.¹⁸ Tests for validation of the predicted probabilities (see details below) show that M2 predicts more than 63% of the cases observed in the data set rightly.

Table 2 shows that, in line with H3, the coefficient for the institutionalization variable, capturing the age of the party at the time a case was decided, is negative and significant at 1% level, and is much larger than 2 standard errors away from 0, indicating a clear tendency in the data. Keeping in mind that the ‘logistic curve is steepest at its center’ (Gelman and Hill, 2007: 82), we divide the coefficient by 4 to interpret the average effect size. We find that the probability of a case to be won decreases by approximately 2.5 percentage points with every additional year in a party’s life cycle, supporting our Institutionalization Hypothesis (H3). This suggests that concerns about organizational stability on behalf of the ‘judges’ are relevant and intensify with a party’s growing age as a reflection of enhanced value infusion in the course of increasing institutionalization (Panebianco, 1988). This process counterbalances an initially higher responsiveness of the tribunal to individual initiators’ claims and demands – as expressed through the cases brought to the tribunal. Thus, over time, tribunals are increasingly likely to prioritize the public display of cohesion by limiting the number of cases that overturn internal decisions or that highlight heavy internal conflict by publicly exercising organizational sanctions. Figures 3 and 4 show that the effect of institutionalization holds regardless of the political party and the type of case (appeal vs. new initiative) concerned.

Figure 3.

Predicted probabilities for cases to be won: Appeals vs. new initiatives.

Figure 4.

Predicted probabilities for cases to be won: By political party.

Moving on to our second main finding, the coefficient of the interaction between the appeal and initiator variable is significant at 5% level. This supports H6 which theorized how tribunal members mediate between their potential affiliation to the party elite generally and to other tribunals (a subset of the elite), integrating our two theoretical rationales. The chances that initiatives by party elites are more successful depend on whether a case appeals an earlier tribunal decision or is a new initiative. The difference of the means in probabilities to win a case comparing elites appealing a tribunal decision and elites bringing forward a new initiative is approximately 43 percentage points. In other words, elites initiating a new case have higher chances to win ( $\hat{p}$ = 0.704, SE <0.10), but this does not hold for appeals ( $\hat{p}$ = 0.333, SE <0.08). Consequently, ‘elite partiality’ plays a role, yet only if it does not clash with a tribunal’s endeavour to protect the court system as derived from our ‘organizational stability rationale’. To visualize this finding and respective effect sizes, Figure 3 shows the predicted probabilities of elite initiatives to be won conditional on the level of institutionalization.

Discussion and conclusion

This article has theorized and empirically examined factors shaping the decision-making of independent party tribunals to gain insights in party conflict resolution, a phenomenon that is notoriously difficult to study. To do so, we developed two theoretical rationales as heuristic tools to derive hypotheses on the conditions under which the initiator of a case is likely to gain tribunal support, that is, likely to ‘win’ a case. On the one hand, tribunal decision-making was theorized as shaped by ‘elite partiality’ inspired by Michels’ ‘iron law of oligarchy’ (1962): tribunal members – the ‘judges’ – once taking on their role in the tribunal become part of the elite themselves and (irrespective of the formal independence of the tribunal) might favour elites when making decisions. On the other hand, tribunal decision-making was theorized as shaped by ‘stability considerations’, specifically verdicts’ implications for the tribunal’s own position in and the stability of the organization, assuming tribunal members – as part of the elite – primarily care about the maintenance of the organization rather than the interests of particular intra-party groups (Farazmand, 2002; Rahim, 2011). While H1–5 theorized factors associated with each of the two rationales separately, H6 theorized their interplay.

Our findings suggest that both rationales are relevant to account for patterns of conflict resolution through independent party tribunals. While the institutionalization of the party when a case is decided as derived from our ‘organizational stability rationale’ had a significant negative effect (H3), new initiatives are more likely to be won when brought forward by members of the party elite (H6), suggesting that the ‘elite partiality rationale’ is relevant as well. This highlights advantages of elites even when facing formally independent tribunals, a finding that echoes Rahat’s concerns about the possibility of court partiality in organizational settings (2013: 141). However, our findings simultaneously point to the limits of such advantages as they do not apply to cases that appeal earlier tribunal decisions. Thus, ‘elite partiality’ plays a role yet only if it does not clash with a tribunal’s endeavour to protect the position of tribunals in the organization in line with our ‘organizational stability rationale’.

In order to test the full range of our hypotheses, the focus on tribunal cases decided over the full lifespan of three parties with participatory credentials was – methodologically speaking – particularly advantageous. For once, for other (older) parties, the coverage of tribunal decisions in the early years of party evolution would have been impossible. Furthermore, in participatory parties it is more likely that both party elites and ordinary members will actively use party tribunals, enhancing the diversity of cases covered and thus their representativeness of the range of conflicts parties have to deal with more generally. At the same time, our parties position differently on the left–right spectrum. And finally, all three parties were formed post-1967 – after the German party law had been adopted – which assures the presence of independent party tribunals throughout, preventing tribunal decisions to be affected by procedural differences as established by (changing) organizational rules within or across parties.

Despite these advantages, the generalization potential of our findings needs to be considered carefully. There is no theoretical reason why the distinction between appeal cases and new initiatives as derived from research on the decision-making of state courts (significant in conjunction with our ‘initiator variable’) should not be relevant in other parties with formally independent tribunals, be these adopted due to legal constraints or voluntarily as an expression of intra-organizational values. The same goes for the role of organizational maturity (institutionalization), as far as we are able to cover decisions across the various stages of parties’ life cycles in the sample of decisions studied. However, moving from the analysis of (relatively) new parties to parties in which the power of party elites has been entrenched for many decades already and tribunals are more likely to take organizational stability for granted, ‘elite partiality’ might play a relatively bigger role in tribunal decision-making, especially in later phases of party development. Thus, future research has to explore in greater detail the scope conditions under which factors associated with each rationale (or combinations thereof) shape tribunal decision-making in different phases of parties’ life cycles. And returning to where we started from, future research also needs to explore how such tribunals as one type of conflict resolution are used in conjunction with different or less formal mechanisms (Figure 1).

The study of conflict resolution in parties is an important theme for future research, as to date empirical research has faced fundamental challenges of access. The by-now large literature on intra-party democracy has rightly emphasized the need to understand intra-party dynamics including conflict as intra-organizational rights of members and even supporters expand (Cross and Katz, 2013). Yet it has tended to focus on those intra-party mechanisms that correspond to elections on the state level (e.g. leadership selection, party primaries), while paying only little attention to those mechanisms mirroring the legal foundations of democracy and attempts to reproduce them in organizational contexts (see for exceptions Bolleyer et al., 2015; Rahat, 2013; Smith and Gauja, 2010). To transplant ‘court structures’ into organizational contexts raises the same fundamental questions as does the transplanting of elections. Preconditions for the working of such mechanisms are different in a state which citizens find difficult to exit compared to organizations that can be left easily (Rahat et al., 2008). However, this does not mean that hypotheses developed in the state context – such as on judicial decision-making – cannot be helpful when theorizing conflict resolution in voluntary contexts as our findings illustrate.

Footnotes

Acknowledgements

We thank all participants of the ECPR General Conference in Glasgow as well as Patricia Correa Vila and Gregor Zons who commented on earlier versions of this article. We are further grateful to the referees and the editors of Party Politics for their helpful suggestions to improve the final manuscript. Many thanks also to Mariana Skirmuntt, Veronique Wavre and Serik Beimenbetov for their research assistance.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research has received funding from the European Research Council under the European Union’s Seventh Framework Programme (FP7/2007–13)/ERC grant agreement 335890 STATORG) and profited from the HASS Project Development Fund ‘The Management of Intra-Organizational Dissent’ provided by the University of Exeter. Furthermore, this research profited from a scholarship granted by the Konrad Adenauer Foundation. This support is gratefully acknowledged. All remaining errors are solely ours.

Notes

Appendix 1

Table A1.

Tribunal decisions in three German parties and their characteristics.

Party name	Period covered	Cases per party	Number of elite initiatives	Number of cases targeting elites	Number of appeals	Number of expulsion cases
Pirates	2008–2013	86	9	77	40	79
Greens	1980–2013	77	38	53	50	54
Left	2007–2013	80	19	52	51	52
Cases in total		243	66	182	141	185

References

Art

(2011) Inside the Radical Right: The Development of Anti-Immigrant Parties in Western Europe. Cambridge: Cambridge University Press.

Aylott

(1995) Back to the future: the 1994 Swedish election. Party Politics 1: 419–429.

Bäck

(2008) Intra-party politics and coalition formation. Party Politics 14 (1): 71–89.

Bakker

Edwards

Hooghe

. (2015) 1999-2014 Chapel Hill Expert Survey Trend File. Version 2015.1. Chapel Hill: University of North Carolina, Chapel Hill. Available at: www.chesdata.eu (accessed 6 December 2015).

Bale

Blomgren

(2008) Close but no cigar? Newly governing and nearly governing parties in Sweden and New Zealand. In: Deschouwer

(ed) New Parties in Government: In Power for the First Time. London: Routledge, pp. 85–103.

Barclay

(1999) An Appealing Act: Why Parties Appeal in Civil Cases. Evanston: Northwestern University Pres.

Bardi

(1996) Anti-party sentiment and party system change in Italy. European Journal of Political Research 29(3): 345–363.

Bender

Bianchi

Hohl

. (2013) Die Policy-Positionen der Parteien bei der Bundestagswahl 2013. School of Governance Report. Essen: University Dusiburg-Essen. Available at: http://www.uni-due.de/imperia/md/content/politik/aktuelles/bender_u.a._dwi_btw_2013.pdf (accessed 22 October 2015).

Bier

(2008) The European court of justice and member state relations: a constructivist analysis of the European legal order. College Park: University of Maryland. Available at: http://www.gvpt.umd.edu/irconf/papers/bier.pdf (accessed 22 March 2015).

10.

Biezen

Piccio

(2013) Shaping intra-party democracy: on the legal regulation of internal party organizations. In: Cross

Katz

(eds) The Challenges of Intra-Party Democracy. Oxford: Oxford University Press, pp. 27–48.

11.

Bolleyer

(2013) New Parties in Old Party Systems: Persistence and Decline in Seventeen Democracies. Oxford: Oxford University Press.

12.

Bolleyer

Little

von Nostitz

(2015) Implementing democratic equality in political parties. Organizational consequences in the Swedish and the German pirate parties. Scandinavian Political Studies 38(2): 158–178.

13.

Boucek

(2009) Rethinking factionalism: typologies, intra-party dynamics and three faces of factionalism. Party Politics 15(4): 455–485.

14.

Boucek

(2012) Factional Politics: How Dominant Parties Implode or Stabilize. Basingstoke: Palgrave Macmillan.

15.

Burley

A-M

Mattli

(2006) Europe before the court: a political theory of legal integration. In: Eilstrup-Sangiovanni

(ed) Debates on European Integration: A Reader. Houndsmills: Palgrave Macmillan, pp. 226–252.

16.

Bukow

(2010) Die professionalisierte Mitgliederpartei. Wiesbaden: Springer.

17.

Burchell

(2001) Evolving or conforming? Assessing organisational reform within European green parties. West European Politics 24(3): 113–134.

18.

Clermont

(1987) Procedure’s magical number three: psychological bases for standards of decision. Cornell Law Review 72: 1115.

19.

Clermont

Eisenberg

(2001) Appeal from jury or judge trial: defendants’ advantage. Cornell Law Faculty Publications. Paper 223.

20.

Cross

Katz

(2013) The Challenges of intra-party democracy. In: Cross

Katz

(eds) The Challenges of Intra-Party Democracy. Oxford: Oxford University Press, pp. 1–10.

21.

de Lange

Art

(2011) Fortuyn versus wilders: an agency-based approach to radical right party building. West European Politics 34: 1229–1249.

22.

De Petris

(2000) Die Verwirklichung innerparteilicher Demokratie und die Transparenz innerparteilicher Entscheidungsprozesse fuer die Oeffentlichkeit. MIP 10 Jg: 52–63.

23.

Deschouwer

(ed) (2008) New Parties in Government. London: Routledge.

24.

Doherty

(1992) The Fundi-realo controversy: an analysis of four European green parties. Environmental Politics 1(1): 95–120.

25.

Dovidio

Saguy

Shnabel

(2009) Cooperation and conflict within groups: bridging intragroup and intergroup processes. Journal of Social Issues 65(2): 429–449.

26.

Duverger

(1964) Political Parties: Their Organisation and Activity in the Modern State. London: Methuen.

27.

Farazmand

(2002) (ed) Modern Organizations: Theory and Practice. Westport: Praeger.

28.

Fiske

Taylor

(1991) Social Cognition. New York: McGraw-Hill.

29.

Gelman

Hill

(2007) Data Analysis Using Regression and Multilevel/Hierarchical Models. Cambridge and New York: Cambridge University Press.

30.

Giannetti

Benoit

(eds) (2008) Intra-Party Politics and Coalition Governments in Parliamentary Democracies. London: Routledge.

31.

Gunther

Hopkin

(2009) A crisis of institutionalization: the collapse of the UCD in Spain. In: Gunther

Montero

Linz

(eds) Political Parties: Old Concepts and New Challenges. Oxford: Oxford University Press, pp. 191–232.

32.

Jungar

A-C

(2013) Keeping the party together: party leadership and cohesion in the true Finns and the Sweden Democrats. In: The EUSA Conference. Baltimore, USA, 9 May 2011.

33.

Katz

Mair

(1995) Changing models of party organization and party democracy: the emergence of the cartel party. Party Politics 1: 5–28.

34.

Katz

(2001) The problem of candidate selection and models of party democracy. Party Politics 7(3): 277–296.

35.

Klecha

Hensel

(2013) Zwischen digitalem Aufbruch und analogem Absturz: Die Piratenpartei. Opladen: Barbara Budrich.

36.

Levitsky

(1998) Institutionalization and peronism: the concept, the case and the case for unpacking the concept. Party Politics 4(1): 77–92.

37.

Maor

(1992) Intra-party conflict and coalitional behaviour in Denmark and Norway: the case of highly institutionalized parties. Scandinavian Politics Studies 15(2): 99–116.

38.

Maor

(1998) Parties, Conflict and Coalitions in Western Europe. London: Routledge.

39.

March

(1988) Decisions and Organizations. Oxford: Wiley-Blackwell.

40.

March

(2011) Radical Left Parties in Europe. Oxon: Routledge.

41.

Michels

(1962) Political Parties: A Sociological Study of the Oligarchical Tendencies of Modern Democracies. New York: Free Press.

42.

Mudde

(2007) Populist Radical Right Parties in Europe. Cambridge: Cambridge University Press.

43.

Müller-Rommel

Poguntke

(eds) (2002) Green Parties in National Governments. London: Frank Cass.

44.

Neu

(2009) Die Linke: eine Volkspartei? In: Kronenberg

Mayer

(eds) Volksparteien: Erfolgsmodell für die Zukunft? Konzepte, Konkurrenzen und Konstellationen. Freiburg: Herde, pp. 190–229.

45.

Niedermayer

(2013) Organisationsstruktur, finanzen und personal der piratenpartei. In: Niedermayer

(ed) Die Piratenpartei. Wiesbaden: Springer, pp. 81–100.

46.

Panebianco

(1988) Political Parties: Organization and Power. Cambridge: Cambridge University Press.

47.

Poguntke

(1993) Alternative Politics: The German Green Party. Edinburgh: Edinburgh University Press.

48.

Poguntke

(1998) Alliance 90/The greens in East Germany: from vanguard to insignificance? Party Politics 4(1): 33–55.

49.

Pondy

(1992) Reflections on organizational conflict. Journal of Organizational Behavior 13: 257–261.

50.

Posner

(1985) The Federal Courts: Crisis and Reform. Cambridge: Harvard University Press.

51.

Reuter

(2009) Rechtspopulismus in Belgien und den Niederlanden. Wiesbaden: VS Research.

52.

Rahat

Hazan

Katz

(2008) Democracy and political parties: on the uneasy relationships between participation, competition and representation. Party Politics 14(6): 663–683.

53.

Rahat

(2013) What is democratic candidate selection? In: Cross

Katz

(eds) The Challenges of Intra-Party Democracy. Oxford: Oxford University Press, pp. 136–149.

54.

Rahim

(2002) Towards a theory of managing organizational conflict. The International Journal of Conflict Management 13(3): 206–235.

55.

Rahim

(2011) Managing Conflict in Organizations. New Bruswick: Transaction Publishers.

56.

Scarrow

(2005) Political Parties and Democracy in Theoretical and Practical Perspectives: Implementing Intra-Party Democracy. Washington: The National Democratic Institute.

57.

Seeleib-Kaiser

(2010) Socio-economic change, party competition and intra-party conflict: the family policy of the grand coalition. German Politics 19(3–4): 416–428.

58.

Sandri

von Nostitz

(2015) Are they in or are they out: the diversity of forms of political participation and its regulation across party organisations. Paper presented at APSA Annual Conference, 3–6 September, San Francisco, CA, USA.

59.

Slater

Ziblatt

(2013) The enduring indispensability of the controlled comparison. Comparative Political Studies 46(10): 1301–1327.

60.

Smith

Gauja

(2010) Understanding party constitutions as responses to specific challenges. Party Politics 16(6): 755–775.

61.

Tolbert

Hiatt

(2009) On organizations and oligarchies: Michels in the twenty-first century. In: Adler

(ed) The Oxford Handbook of Sociology and Organisational Studies: Classical Foundations. Oxford: Oxford University Press, pp. 174–199.

62.

Turner

(1987) Rediscovering the Social Group: A Self-Categorization Theory. Oxford: Basil Blackwell.

63.

Van Haute

(2011) Who voices? Socialisation process and ideological profile of discontented party members. In: Van Haute

(ed) Party Membership in Europe: Exploration into the Anthills of Party Politics. Brussels: Editions de l’Université de Bruxelles, pp. 169–186.

64.

Wilson

(1973) Political Organizations. Princeton: Princeton University Press.

65.

Zariskia

(1965) Intra-party conflict in a dominant party: the experience of Italian Christian democracy. The Journal of Politics 27(1): 3–34.

66.

Zolleis

Prokopf

Strauch

(2010) Die PiratenparteiHype oder Herausforderung für die deutsche Parteienlandschaft? Hans Seidel Siftung: Munich.