Abstract
This article provides a dynamic framework through which factionalism can be examined and the circumstances of individual parties compared in multi-level contexts. We discuss the interaction between factionalism and party structure by setting out a model of factional organization dependent on the tolerance of host parties to dissent and their degree of vertical integration, their combination yielding four possible strategies for opposition factions: centralized, inter-layered, multi-layered and decentralized. We also consider what implications there are for the party’s dominant coalition in episodes of high factionalism. These act as a catalyst for the modification of party rules that regulate dissent and vertical distribution of power. The hypotheses developed are tested on four Spanish political parties that differ on the autonomy of regional branches and factions, the competitive position in the party system and factionalism type – more policy or more patronage-oriented.
Introduction
Factionalism has been poorly developed in the study of political parties and it is still orphan of conceptual approaches and hypotheses. This is so despite most parties passing through periods of unity and dissent depending on particular external or internal circumstances, such as electoral performance or changes in leadership (Hine, 1982: 48). Indeed, factionalism is ubiquitous in political parties (Harmel et al., 1995: 7). The study of factionalism is thus not confined to the understanding of party politics but to the comprehension of politics at large (Belloni and Beller, 1976: 549).
Parties are composed of sub-coalitions or groups of activists who might disagree over the goals and how to pursue them (Kitschelt, 1989: 47). The group that controls skills and knowledge, recruitment, finances and communications is the dominant coalition (Panebianco, 1988: 37). However, it is often the case that dominant coalitions are fragmented. When dominant coalitions are heterogeneous they become coalitions of factions made up of governing and opposition factions (Key, 1956). The party’s governing or dominant faction is then the coalition component that is better equipped to impose its ideological or organizational preferences (Harmel and Tan, 2003: 411).
Hine (1982: 38 f.) identifies three dimensions of intra-party group conflict: the issues dimension, coverage and organizational solidity. Regarding the first, conflict consists of either a policy or a strategy axis, or a combination of the two. Yet, party conflict can also be motivated by patronage incentives such as power, career and spoils (Sartori, 1976: 76). Usually the two types of factionalism interrelate (Janda, 1993). Second, coverage comprises the number of arenas in which conflict occurs. Katz and Mair (2002) suggest that in the modern catch-all party conflict is mainly played out in the central office, so factions are expected to settle in this face (Carty, 2005: 5). As for organizational solidity, scholarly disagreement denotes the existence of a continuum from highly organized factions (Rose, 1964; Zariski, 1960) to loose and non-cohesive tendencies (Hine, 1982: 39; Key, 1952).
Intra-party actors, including both the dominant and the opposition factions, perform cost-benefit calculations on a continuous basis, making strategic decisions designed to fulfill their basic objective (Laver and Shepsle, 1996: 20). Strategies are shaped by different institutional and organizational incentives (Boucek, 2005: 6). In multi-level contexts, two types of organizational incentives come into play. Besides party rules regulating internal dissent, political decentralization provides new incentives for intra-party actors (Duverger, 1964; Hennessey, 1968: 12; Houten, 2009; Panebianco, 1988). State decentralization processes impact on the most relevant dimensions of party organization, namely distributional conflicts over resources, the definition of platforms and electoral and governing strategies (Biezen and Hopkin, 2006; Chhibber and Kollman, 2004; Deschower, 2006; Downs, 1998; Hopkin, 2003).
The degree of ‘self rule’ that regional branches have and the type of ‘shared-rule’ between party levels determine a party’s vertical integration, that is the degree of organizational and programmatic autonomy allocated to sub-state branches and the strength of joint decision-making structures within party organizations (Detterbeck and Hepburn, 2010). Different combinations of both ‘rules’ might determine from what territorial level the opposition faction strategies are built. Alternatively, the central dominant coalition might have a minority status at one or various regional branches. So, factional conflict is potentially aggravated by territorial conflict between party levels.
Therefore, we argue that in compounded states party factionalism should be examined with these two types of organizational incentives taken into account: party vertical integration and the regulation of intra-party dissent. These are precisely the incentives dominant coalitions will try to act on in moments of high factional conflict in order to counteract opposition factions. Abstracting out different characteristics and forms of organization, we will build on broadly inclusive definitions, considering factions as intra-party groups that act collectively in order to achieve common goals (Belloni and Beller, 1978: 447; Boucek, 2009: 14; Zariski, 1960: 33).
As Köllner and Basedau (2005: 19) suggest, it is often unclear whether factionalism results from a particular way of organizing or whether the causality runs the other way round. In other words, factionalism is intrinsically endogenous and must be studied in a dynamic perspective (Boucek, 2009: 25). Although one has to be cautious when drawing connections between factionalism and intra-party organizational arrangements, we suggest that this relationship can be established if factionalism is simultaneously examined as both dependent and independent variable and if external factors (i.e. formal institutions, party system characteristics, socio-economic, cultural and political features) are kept constant.
This article provides a framework through which factionalism can be examined and the circumstances of individual parties compared in multi-level contexts. It aims to identify what factional strategies can arise in these settings according to certain party organizational arrangements. We start by exploring in the first section the dynamic connections between factionalism and party structure and set out a model of factional organization according to the degree of party vertical integration and the tolerance of host parties towards dissent. We also consider the dominant coalition’s strategic choices in episodes of intense factionalism. The second section applies the hypotheses developed for four Spanish political parties in which the autonomy of their regional branches and factions differ. The third section discusses the main findings and the last section concludes.
Factionalism and party organization in multi-level settings
In order to contextualize the development of factionalism and to analyse its course of action we need to study it both as dependent and independent variable. In the first case, we need to focus on the determinants of factional activity which stem from a particular party organization, whereas in the second case we consider how factionalism might produce party change.
When considering factionalism as a dependent variable, we look at how party rules encourage factions to develop particular organizational strategies. Belloni and Beller (1976: 548) defined the organizational characteristics that may invite the formation of factions as the ‘permissiveness of the parent body’. As Hine (1982: 38) also noted, ‘a group’s need for organisation will depend on the rules and conventions governing party life’. In multi-level contexts, two organizational variables determine what shape internal conflict will adopt: the autonomy of factions and the autonomy of regional branches.
Factions, be it formally structured or loosely articulated, help individuals in different party levels to coordinate decisions (Carty, 2004: 15). The capacity to do so depends on the ‘internal market for consent’, i.e. party pluralism, party sharing and the representation of sub-party interests (Hine, 1982: 42 f.). Hence, intra-party democracy might maximize the influence of opposition factions (Boucek, 2005: 21).
Multi-level dynamics add complexity to the inherent (permanent or temporary) lack of unity within parties. The balance of power between the central and the territorial levels of the party over organizational resources, the definition of platforms and electoral and governing strategies can be measured through a continuum that defines its vertical integration, ranging from extremely weak linkages to highly integrated structures. From weakest to highest, the categories included in the continuum are: truncated, bifurcated, confederal, integrated and unitary parties (see Carty, 1988; Dyck, 1991; Smiley, 1987; Thorlakson, 2001, 2006). Vertical integration is thus defined by the degree of sub-national party influence (the representation of party branches at the central level) and sub-national party autonomy on candidate selection and on the management of the party structure and discipline (Thorlakson, 2009).
In previous research it has been stated that centralized parties are likely to discourage factionalism and to induce opposition groups to concentrate at the central level (Beller and Belloni, 1978; Duverger, 1964; Köllner and Basedau, 2005; Zariski, 1960). We argue that further theoretical refinements are required. We assume that factions will use all the organizational resources available to deploy their strategies without putting in jeopardy their own survival, regardless of whether their motivation is more policy or office-seeking. Given that organizational resources are conditioned by party rules, different combinations of these will lead factions to follow distinct strategies. Hence, our first hypothesis claims that in multi-level contexts both vertical integration and factional autonomy will come into play, their combination yielding four likely strategies for opposition factions, as Table 1 illustrates.
Likely opposition factions’ strategies
(i) Centralized strategy. If factions are tolerated by party rules but regional branches have limited autonomy (i.e. the federal bodies have the capacity to impose decisions to the lower levels), concentrating factional activity at the sub-national level is a hazardous strategy, as it could easily be counteracted by the centre. It is then likely that in highly integrated parties opposition factions will put their efforts at the national level and compete with the dominant coalition for party patronage.
(ii) Interlayered strategy. Factions are not tolerated or restrictively recognized and the centre is very strong, so it can easily intervene in any regional branch. Possibilities for organization and access to party resources are limited in this case. Minority factions cannot exclusively develop at either the national or the sub-national level. Instead, they are likely to be loosely organized, with leaders having to retain power at the two levels in order to survive. Without one of these pillars, there will not be much chance of mounting an effective opposition.
(iii) Multilayered strategy. When party rules admit factions and regional branches enjoy high levels of autonomy, factionalism can develop at either the central or the regional level. Given that factions may need to follow distinct sub-national strategies, they will most likely be heterogeneous at the central level. They might present different degrees of coordination, and alliances are likely to shift at different levels. Territorialized strategies might be more relevant in some cases, whereas in others centralized factional activity may prevail.
(iv) Territorialized strategy. If factionalism is not allowed but vertical integration is weak, minority factions will be more effective if they organize at the regional level, being able to fight for organizational, political and economic resources in areas that the centre has difficulty controlling. Hence, factions are likely to act under the umbrella of the territories trying to increase their power within the party structure. This will ensure their survival as well as offer them valuable means by which to confront the central (dominant faction’s) leadership.
It should be noted that this framework only applies to parties where cross-level decision-making and strategies exist, thus excluding truncated, bifurcated or extremely stratarchical parties.
We have presented how rules might shape the opposition faction’s strategies. However, we should also analyse the dominant faction’s possible responses. More precisely, how factional conflict can act as a catalyst for the modification of party rules. In this case, factionalism is considered as an independent variable.
Although factions’ and territories’ autonomy might provide structural incentives for intra-party conflict, these are not static factors. Our second hypothesis posits that dominant factions may make a strategic use of the party’s rules to counteract opposition factions and thus retain control of the party. In other words, the party organization may become a device aimed at restraining factionalism.
Regarding regulation of dissent, extreme factionalism will lead the dominant faction to revise intra-party democracy (i.e. the voting system to elect decision-making bodies) and formal recognition of dissent, including access to party resources.
With respect to the party’s vertical integration, the dominant faction’s responses are likely to vary according to the level that opposition factions concentrate at. When factionalism is found at the central level, the dominant faction will tend to loosen vertical integration, so that allied regional branches can implement the goals which remain blocked at the central level by opposition factions. This possibility will depend on the rules for leadership selection, the support dominant factions can orchestrate among the regional branches, and the selective incentives the former can offer in exchange. Conversely, when critical factions organize their opposition from the territories, the dominant faction will seek to increase the party’s vertical integration in order to reduce the influence of hostile regional branches. This strategy is more likely to be implemented the more homogeneous the federal dominant faction is.
An empirical evaluation of multi-facetted factionalism: The Spanish case
Spain’s decentralization and democratization processes occurred concurrently. The previous unitary state gave way to a multi-level polity divided into 17 regions, namely Autonomous Communities, each with its own representative and executive institutions (Aja, 2003). Both established and recently formed political parties had to adapt their organization to the new territorial dimension.
Whereas political decentralization has clearly shaped parties’ electoral and governing strategies and they all organize according to the structure of the state, their internal decentralization has remained to a great extent isolated from the broad statewide process of institutional change. Indeed, the degree of political decentralization of the state is greater than intra-party decentralization (Montero, 2005: 68). Alternatively, variation on the degree of intra-party decentralization is explained by incumbency at the central and regional levels, ideological tradition, party competitive strategies and party dissent. Generally speaking, governing parties feel more pressure to centralize, left-wing parties define themselves as advocates of federalism and some parties may see electoral payoffs in presenting their organization as the only one granting coherence across territories. Yet, the presence of intra-party struggle may produce either more decentralization or more centralization (Fabre and Méndez-Lago, 2009: 115 f.).
Four Spanish political parties have been selected for the analysis on the basis of their different levels of vertical integration and tolerance of dissent. They also diverge on their ideologies, competitive position in the party system and type of factionalism experienced (policy or patronage-oriented).
The Spanish Socialist Workers’ Party [Partido Socialista Obrero Español, PSOE] and the Popular Party [Partido Popular, PP] are the largest state-wide parties and have alternated in government at the federal level since 1982. Although the PSOE was founded in 1879, it was strongly restructured during the transition to democracy. Organizationally, it became a successful catch-all party with a rapid increase in size, resources and members. Ideologically, the party abandoned Marxism and fully embraced social democracy (Méndez-Lago, 2000). The PP was born in 1976 (Popular Alliance [Alianza Popular] at the time) as a coalition of seven minor conservative parties whose members had been linked to the Francoist regime. The party stood very radical in its right-wing stances until the mid-1980s when it gradually moved to the centre-right (Montero, 1988). By then, all the founding parties along with other liberal and Christian-democratic parties had dissolved into a single entity. The PP led the central government between 1996 and 2004. United Left [Izquierda Unida, IU] is the third largest state-wide party. It was formed in 1986 as a coalition led by the Communist Party of Spain [Partido Comunista de España, PCE] along with other minor left-wing parties as a means by which to counteract the electoral crisis of communism (Ramiro-Fernández, 2004). Because of a majoritarian twist in the electoral system, IU has little governing potential at the national level. The Basque Nationalist Party [Euzko Alderdi Jeltzalea–Partido Nacionalista Vasco, EAJ–PNV] is a centre-right non state-wide party founded in 1895 and based in the Basque Country and Navarre. From 1980 until April 2009 it governed the Basque Country (Pablo and Mees, 2005).
In the following empirical analysis we are interested not in providing an exhaustive diachronic description of factional activity, but in examining those episodes of deep factionalism that the four parties have gone through in the recent democratic period under a decentralized state structure. A permanent tension between unity and dissent exists, but only occasionally we see the dominant and opposition factions clearly fighting each other for control over policy, strategy or party resources. These episodes help us illustrate the opposition factions’ strategies defined in the previous section and the response provided by the dominant faction. We have carefully analysed party statutes in order to trace relevant changes affecting both recognition of dissent and vertical integration. Thus, we are interested in concluded factional episodes that allow us to examine factionalism from a dynamic perspective identifying the variables which explain it and considering it as an explanatory variable for party change.
Centralized factionalism
From 1979 until the mid-1990s, the PSOE was dominated by Felipe González (secretary-general) and Alfonso Guerra (deputy secretary-general). They managed to end internal dissent and the 1982 landslide victory ultimately favoured cohesion (Gillespie, 1995). González became Spain’s Prime Minister and, although Guerra occupied relevant offices, including that of deputy Prime Minister, he focused on expanding control over the party. This triggered one of the best-known episodes of factionalism in recent Spanish party politics.
At the time, the PSOE combined tolerant rules towards factionalism with low levels of regional autonomy. Although informal factions had always existed, factional activity was only formally admitted in 1984. Factions could not have their own symbols, organizational structure and vote discipline, although the latter two aspects could easily be achieved through informal arrangements. They should restrict their activity to fostering debate and providing input on platforms and policy within the party. Access to party documents and communication means was limited. Rules concerning the election of party boards were modified to help minorities: All groups reaching 20 percent of the vote were entitled to 25 percent of offices (Verge, 2007: 340).
The regional party structure was developed by the centre after the transition to democracy. So, the party organization very quickly matched the state structure. Regional branches were allocated low influence and autonomy. The federal bodies monitored, controlled and could veto the decisions adopted by regional branches regarding candidate selection, manifestos, finances, platforms and alliance choice. Coordination of public policy across territories was assigned to the federal level. Besides, the presence of regional leaders within the highest executive boards was insignificant during the 1980s (Méndez-Lago, 2000).
Although informal factionalism has been common since the mid-1980s and the 1990s, factions have often been reluctant to apply for formal recognition – with the exception of the Socialist Left/Izquierda Socialista, which has never been a threat in terms of effective opposition and membership support.
The factional episode we examine here initially opposed the party in public office, whose economic policies were inspired by orthodox liberal ideas and developed by a group of ministers directly appointed by González (later on called renovadores), and the party in central office led by the left-winger Guerra and his followers (guerristas). However, the reason was not only ideological, since guerristas’ power was based on an extensive patronage network. Although guerristas never directly attacked the popular leader González, they criticized the liberal ministers’ economic policies and accused them of having provoked the 1988 general strike. Liberal ministers were also accused of populating political posts with technocrats from outside the party. Rising levels of division finally resulted in a clear, and sometimes public, deterioration of party–government relations (Gillespie, 1989).
Guerra and his supporters used the central party as a stronghold. Delegates in party conferences were elected by provincial delegations, whose local leaders were mostly controlled by the party apparatus. In fact, as provincial delegations voted in block in party conferences, guerristas enjoyed a great deal of influence within the decision-making bodies vis-à-vis the renovadores (Maravall, 2003: 145). The strict way in which Guerra managed the party, along with a limited regional autonomy, clearly favoured the position of his faction at the centre.
Internal conflict hastened during the 1990s. Tensions emerged between regional power-holders, increasingly supportive of the renovadores sector (Gangas, 1995: 169), and the party in central office, in the hands of guerristas. Simultaneously, as the country’s decentralization process consolidated, regional leaders had access to more resources, several of them governing the regional institutions with a wide majority.
González tried to reduce the power of the guerristas by giving more room to the regional leaders. Although the guerrista sector controlled the central office, González used his position as Prime Minister to hold regular meetings with the regional leaders and presidents of the regions governed by the PSOE who had attained a prominent role owing to the enhanced position of the Autonomous Communities, which had been transferred a significant number of competences from the state. According to some party officers, these meetings became more relevant than executive bodies (Méndez-Lago and Orte, 2010).
As a result of the coordinated action between González and the regional leaders, along with increasing dissatisfaction with how Guerra managed the corruption scandals involving party members, by 1994 the presence of renovadores and regional leaders in the federal executive bodies had increased enormously and the guerristas had been cornered within the government. Finally, the indirect block vote by provincial heads of delegation was substituted by an individual and secret vote by all delegates, thus eroding one of Guerra’s main tools for controlling the party (Méndez-Lago, 2000: 117). Besides, norms regulating the election of party boards were modified too: the most voted list in the party conference would be entitled to the majority of seats, with the rest allocated proportionally among any other candidature reaching 20 percent of the delegates’ vote (Verge, 2007: 341).
Inter-layered factionalism
The organizational penetration of the PP began with the sub-regional level, the provinces, which are the electoral constituencies in national and regional elections. The regional organization was only fully set up in the mid-1980s. Nevertheless, the party structure was designed in an extremely centralized and hierarchical fashion (the president of the party accumulating numerous functions), and factionalism was forbidden and sanctioned with expulsion (Gangas, 1995: 156). All sub-national branches were obliged to follow the instructions of the national level, which approved the lists of candidates for all elections. Owing to the party’s deep presidentialism, coordination was often carried out between the sub-national and national leaders. As regional branches increased their competences, new bodies were created to enhance control over them, such as a national electoral committee and a body in charge of coordinating the activity of the party in public office at all tiers of government. As opposition factions could only be organized informally, they had to settle at both levels of the party organization in order to survive, so they were encouraged to deploy an inter-layered strategy.
The factional conflict we analyse arose between 1982 and 1989 within the national leadership as well as between national and sub-national leaders. It can be considered mainly as a career-oriented case of factionalism. On the one hand, the once uncontested party leader Manuel Fraga was increasingly seen as an electoral liability by the secretary-general and many sub-national leaders, especially after the strategy of focusing exclusively on right-wing voters collapsed in 1986 when it became clear that the median voter was located in the centre-left (López Nieto, 1995). Fraga thought that open lists for the election of decision-making bodies would help soften the conflict, but they empowered provincial leaders by facilitating their coordination to obtain more central posts (see García-Guereta, 2001: 139).
On the other hand, there was a latent conflict between the regional and the central organization. The federal level aimed to renew the provincial and regional executive boards and candidates to make the party more competitive by introducing moderate and reformist leaders with a view to gaining a hold on more segments of the electorate and to help forge coalitions with moderate regionalist parties in various Autonomous Communities. Many of the newcomers belonged to the liberal party that the PP had formed a coalition with (the Democrat Popular Party/Partido Demócrata Popular, PDP). Sub-national organizations were very unhappy with what they considered an excessive representation of coalition partners, and many sub-national leaders felt they had been displaced (Gangas, 1995: 168).
Eventually, Fraga resigned and called a conference to decide on his succession. Exceptionally, in the party’s past and future history, two candidates competed for the post: Miguel Herrero de Miñón, the leader of the parliamentary group and the man Fraga had transferred the presidential powers to until a new leader was elected, and Antonio Hernández Mancha, the president of the regional branch of Andalusia. The latter won the ballot with the vote of 72 percent of the party delegates, thanks to the support he managed to galvanize among the sub-regional (provincial) delegates of the largest regions, one sector of the federal leadership and the party’s youth organization (Baón, 2001: 775).
The defeated faction then exerted pressure at both the regional and national levels, fighting for the inclusion of their candidates in the party lists for the local and regional elections of 1987, while opposing the new leader from the party in public office, which they controlled (Gangas, 1995: 211).
In order to shut the conflict down, Fraga finally announced his nomination as party leader in 1989 – finding no competition. Once back, he commanded a transition which led to the party’s refoundation which was completed by José María Aznar, elected party president in 1990 (see García-Guereta, 2001). The party changed its ideological position towards the centre, augmented the president’s competences 1 and culminated in the definition of the territorial structure in two directions: one, the provincial branches were put under the strict control of the regions, then adjusting to the state territorial architecture; and, two, although the party apparently decentralized towards the regions, the latter became strictly supervised by the federal level. Electoral and disciplinary committees are only found at the regional level and are monitored by the centre. Since then, party delegates have been elected at the local level, diminishing too the power of the provinces. Besides, it is the national executive board, and not party delegates, which ratifies the highest party officers (secretary-general, president of the electoral commission, etc.). Finally, open lists have never been used again.
Between 1989 and 1991, nine regional leaderships were removed from the centre, and by 1993 all regional branches except Galicia, which Fraga controlled, had been renewed. Regional party leaders did not become members of the national executive board until 1999, although presidents of the regional governments ruled by the party were incorporated in 1989. Nevertheless, regional representatives account for only 25 percent of this board (Astudillo, 2010). The disciplinary regime is also controlled by the centre, which can dissolve unruly regional branches. All sub-national conferences must be held in cascade following the federal conference, thus installing stronger control over the territories. Since 1989, regions have been unable to request the celebration of an extraordinary conference. They can establish their rules, but candidate lists for all elections require federal approval. Severe incompatibilities between territorial and parliamentary representation were introduced too, limiting regional influence on the party in public office at the federal level (Verge, 2007: 313).
Multi-layered factionalism
Since its creation in 1986, IU has gone through two significant episodes of factionalism. The first occurred in the period 1991–7 and the second from 2000 to 2008. On both occasions, factions developed at the two levels of the party organization and mounted opposition to the dominant faction through a multi-layered strategy fostered by policy and strategy differences.
By the beginning of the 1990s, IU was suffering from increasing internal strife. The multi-party left-wing coalition dominated by the Communist Party of Spain (PCE) was divided on several issues. Regarding organizational politics, although most of the leadership had once supported PCE’s dissolution into the new party, when some leaders claimed in 1991 that the moment had arrived, the majority of communist leaders opposed, fearing members’ rejection and arguing that electoral progress was diluting pressures to get rid of the party (Ramiro-Fernández, 2004: 18). Confrontation over the degree of policy radicalism the party should defend was also present. Finally, alliances and strategy towards the PSOE generated the most acute division.
In 1992, founding parties ceased to enjoy representation quotas in executive boards and conferences. Internal conflict was also regulated by admitting the creation of currents of opinion at the federal level as far as these were established in at least four regional party branches – they could also form within regional branches. Factions could not impose vote discipline on their members or present candidates’ tickets in internal elections. Opposition factions soon emerged and one of them, New Left/Nueva Izquierda, obtained the status of federal faction. Increasing divisions forced the party to be more generous with factional activity. Constraints for their creation were eased as the required presence at the regional level was reduced to three regions. Factions were also granted access to party publication and communication means. In 1994, factions were allowed to present their candidates in internal elections, but the pure proportionality principle was modified: Seats in decision-making bodies would be allocated by consensus among the competing lists obtaining at least 10 percent of the delegates’ vote at the party conference.
Whereas New Left was the minority faction at the federal level, it managed to become the dominant faction in seven regions. This way it could deploy a multi-layered strategy combining its opposition in the federal party boards with coordination with the regional branches it controlled. The regional branches led by New Left challenged the IU leadership by establishing local coalition governments with the PSOE, which totally confronted the party’s strategy to radically oppose the main left-wing party.
However, the dominant faction’s advantage was control of the party’s territorial structure due to federal vertical integration. It used its pre-eminence in the federal bodies to increase control over the territories. In 1997, when one of the party branches reached an agreement with the PSOE for the regional election, the national dominant faction created a new regional branch. The other regional units that had supported the unruly branch saw their executive boards dissolved and their membership purged (Paniagua and Ramiro-Fernández, 2003).
The dominant leadership gradually tightened up on control of the territorial structure. In 1994, the statutes established that the federal level should ratify the candidate lists and the platform for legislative elections. It was also specified that alliances with state-wide parties or with regional branches of state-wide parties were a federal competence. In 1997, the statutes clarified that the competence to decide on alliances at any level affecting the coherence of the federal project corresponded to the centre. Besides, the capacity to sanction regional branches if they contravened federal decisions was strengthened by reducing the required votes (Pérez-Nievas and Ramiro-Fernández, 2011). Although still far from the levels of the PP and PSOE, vertical integration clearly increased. With regard to rules of dissent, the PCE, the main actor within IU, prohibited its members from joining factions that contravened the party’s position.
The second episode of high factionalism emerged soon after the previous one had just been resolved. The main actor behind internal conflict was not one of the recognized factions but the very same PCE which has been increasingly acting as an opposition faction at both levels of the party organization since 2000.
The severe electoral defeat in the 1999 local and European elections and in the 2000 legislative elections led to major change within the dominant coalition. To recover votes, the new leadership sought to de-radicalize the traditional communist discourse by re-launching the organization as a post-communist eco-socialist party. A more pragmatic position towards the PSOE and towards office was introduced too. Since 2003, several regional branches have reached government coalitions with a broad array of parties, including the PSOE, and, later in March 2004, when the PSOE went back into a minority government, IU accepted supporting the government’s investiture, and its most important legislative initiatives, in exchange for certain policy concessions.
Nonetheless, leadership replacement was only partial, as the newly elected leader received just one more vote than his rival in the internal ballot and the minority faction rejected this more policy-seeking strategy. The PCE considered the new eco-socialist discourse as a threat to the organization’s core ideological identity and a betrayal of the party’s communist past, and strongly rejected the party-of-government role at the regional level and support to the PSOE in the national lower house (Verge, 2010).
As the distribution of seats in the highest party executive board is strictly proportional to the votes obtained by competing lists, the losing candidate who represented PCE’s majority faction (actually PCE’s secretary-general) retained considerable influence within IU’s dominant coalition, thus making it difficult to implement party change. 2 Besides, as the dominant faction did not command the support of three-fifths of the vote at the executive board, it risked losing votes affecting conflict between the federal level and unruly regional branches.
However, the multi-level setting allowed the dominant coalition to turn the regional arena to its advantage and use the party meso-level to circumvent factionalism. Whereas the opposition faction was still strong at the federal level, its control over regions had weakened. The dominant faction used this decline to orchestrate a coalition of support at this layer which came along with an empowering of regional leaders in the national leadership selection process – regional leaders were enfranchised as members of the selectorate. 3 Regional party branches also obtained an increased autonomy in sub-national issues, including the capacity to unilaterally decide whether to enter an electoral or a government coalition. Those regional party branches facing stronger confrontation from the opposition faction were also granted looser integrative linkages, establishing a confederal vertical integration for those territories (Ştefuriuc and Verge, 2008).
Decentralized factionalism
Factionalism in the Basque Nationalist Party (henceforth PNV) has traditionally divided those who pursue a pragmatic nationalism, which demands devolution but compromises with Spanish politics, and those who defend a more radical nationalism, which advocates the independence of the Basque Country and excludes collaboration with non-nationalist forces (Pablo and Mees, 2005). However, it has often been claimed that career-seeking motives go hand in hand with the ideological claims of factions in the PNV (Pérez-Nievas, 2002). Although both ‘souls’ have always coexisted, occasionally there have been intense factional conflicts. Of the two consecutive conflicts that originated between the 1970s and the 1980s, we focus only on the second, as the first does not come under our period of analysis. Party rules forbid members from belonging to any political or socio-political association, trade union or entity at odds with the party ideology. All members are required to obey the party stances and discipline at both federal and regional level and non-compliance can lead to expulsion.
As the PNV is a non-state-wide party, the party federal level corresponds in this case to the central leadership of the Basque region and the regional level to the sub-national party structure, namely the three provinces which constitute the Autonomous Community of the Basque Country (Alava, Guipuscoa and Biscay) and Navarre, which is a separate region. 4
In the period under study, regional branches enjoyed high levels of autonomy, matching the high degree of decentralization which is found within the Basque Country. For historical reasons, the Basque provinces enjoy high levels of autonomy and hold competences on issues such as tax collection, healthcare and infrastructure. Not only was the central executive branch elected by the regions, it was also subject to territorial mandate, regional branches deciding on their delegates’ vote. Moreover, the party defined itself as a confederation, and its central bodies were exclusively formed by regional delegates. All territories had the same share within the federal bodies, and the president of the executive board was a mere spokesperson, a primus inter pares. Regions elected their leaders and candidates and autonomously decided on electoral platforms and coalitions within their constituency. They had their own disciplinary bodies, and although appeals to a federal disciplinary commission were possible all its members except the president were regional appointees. It is therefore not surprising that the most belligerent factional conflict was territorialized. Organized dissent was strictly punished and only the regional umbrella could provide minorities with some room for manoeuvre.
After the first factional conflict of the 1970s, the dominant faction tried to calm down internal dissent by giving some space to the opposition. Thus, one of its leaders, Carlos Garaikoetxea was appointed president of the federal executive board in 1977. In 1980, with the triumph of PNV in the regional elections, Garaikoetxea became the president of the Basque Country. However, owing to the strict internal ban on holding multiple posts, he was replaced as president of the executive board by Xabier Arzalluz, a loyal member of the majority faction. Once in government, Garaicoetxea tried to avoid party discipline, which was interpreted as a challenge to the power of the party in central office (Arzalluz, 2005: 213) and, consequently, to the dominant faction. As a result, the executive board blocked the approval of several laws in the regional parliament (Pérez-Nievas, 2002). Garaikoetxea then turned to the regional branches of Navarre and Guipuscoa to reach his goals, which increased territorial disputes between both factions (Garaikoetxea, 2002: 214).
By that time the dominant faction had reached an agreement with the Popular Party (PP) which involved supporting a conservative government in Navarre in exchange for support in all Basque local institutions (Pablo and Mees, 2005). The Navarrese regional branch, supportive of Garaikoetxea, strongly opposed the pact, fearing a decline of electoral support in the region. The majority faction forced an internal ballot and won. However, despite its defeat, the Navarrese branch did not vote the PP’s candidate, arguing that regions were autonomous on choice of alliance. Failure to comply with party discipline provided the dominant coalition with a strong argument to purge the leaderships of the hostile branches and forced the dismissal of Garaikoetxea as president of the Basque government. As a result, almost all members of Navarre and half the members of Guipuscoa were expelled (Acha and Pérez-Nievas, 1998), which triggered the split and subsequent creation of a new party (Basque Solidarity [Eusko Alkartasuna]) in 1987.
Right after this factional episode, the statutes underwent several modifications to increase the power of the centre vis-à-vis the regional branches. The influence of regions in federal boards and disciplinary bodies was reduced. Most members – all members in the case of the disciplinary body – are now directly elected during a national conference. New federal bodies, such as a quadrennial general conference, were introduced, and local branches were empowered with direct consultation mechanisms set up by the federal level. Regional autonomy on finances, platforms, campaigns and coalitions was cut down and the federal executive board took over the selection of candidates for all except local elections (Gómez and Pérez-Nievas, 2009).
Discussion of findings
Although previous research has claimed that decentralized states promote the formation of regionally rooted factions and that centralized parties are likely to discourage factionalism and to induce factions to concentrate at the central level, we have shown that there are more likely possibilities.
Four Spanish parties were selected to illustrate our hypotheses (Table 2 is a summary of our findings). In the PNV and IU, regional branches enjoyed notable levels of autonomy. Nevertheless, factions’ strategies were far more territorialized in the PNV, where intolerance of factional activity pushed the opposition faction to fight for the control of regions. Vertical integration was higher in the PP and the PSOE, albeit factionalism also varied. While, in the PSOE factions organized at the federal level, in the PP collaboration between regional and federal leaders granted their survival.
Opposition factions’ strategies in Spain
For each party, we have focused on specific episodes of factionalism and shown that they constitute critical junctures for the party organization. Factionalism has always been accompanied by party organizational change in two directions: When faced with strong factional activity, the dominant faction changes the rules regulating both dissent and vertical integration as a means by which to weaken the resources and the arena the opposition faction is using. Thus, party organization becomes a device. Two clear responses have been identified: When the opposition faction’s strategy is centralized, the dominant faction turns to regional branches to marginalize opposition groups (PSOE); when the conflict is territorialized vertical integration is tightened up (PNV).
However, when inter-layered (PP) and multi-layered (IU) factional strategies are found, the dominant faction’s calculus varies. Whereas in the PP the dominant faction increased the power of the centre, making it more difficult for inter-layered factions to act, in IU both strategies were applied. The first factional episode was solved by counteracting the minority at both levels, but particularly by restricting the territories’ room to manoeuvre. Conversely, on the second occasion the dominant faction loosened vertical linkages, granting allied regions more autonomy.
Simultaneously, as Table 3 shows, the regulation of dissent has also experienced important modifications through which the dominant faction has sought to limit the opposition factions’ influence.
Dominant factions’ strategies in Spain
Concluding remarks
Far from being unitary actors, parties consist of competing groups. This article has addressed the endogeneity problem that research on factionalism faces by analysing it as a dynamic process. This implies studying factionalism as both dependent and independent variable, while keeping constant external factors, which requires examining the interaction of dominant and opposition factions’ strategic choices. According to the theoretical framework we have displayed, knowing about a particular party organizational model allows us to predict what shape factionalism might adopt in a party, regardless of it being more office- or policy-oriented.
The endogeneity problem is more acute in multi-level settings as party decentralization provides extra organizational incentives to factions. To explain how factions behave in these contexts we need to take into account not only party rules regulating dissent but also a party’s degree of vertical integration. As more states are decentralizing and current decentralized states are deepening the transference of power to their regions (Rodden, 2004), more parties are expected to adjust their organization to the multi-level competition, so the vertical integration dimension cannot be neglected in the study of party factionalism.
Yet, far from being over-deterministic, we do not claim factional strategies to be static or isolated from other factors. Given that factionalism is a dynamic process, the responses crafted by the dominant faction will introduce new incentives and restrictions on rules and procedures to factional organization, which, at the same time, can give rise to factional adjustments and new episodes of conflict.