A clear instance of a collective member of a political nature is that of the party of the Catalan socialists (PSC) within the Spanish socialist party (PSOE). In this case, the basis of the articulating linkage between the collective member and the overall party is territorial. Firstly, the PSC is not a sub-organization of the PSOE. Apart from some (substantial) mutually agreed connections, the PSC is a fully sovereign political party. In particular, the establishment and maintenance of the agreement between the PSC and the PSOE is a decision of both parties. The PSC could break the “contract” without altering its structure. The fact that this would almost certainly lead to a split in the PSC lies in the sphere of the political consequences of political decisions.

Secondly, the entitlement to power of the PSC in the PSOE qualifies the former as a collective member of the latter. The relation between the PSC and the PSOE is governed by the “Protocol” (1978); a document agreed upon on the occasion of the merger of the Catalan branch of the PSOE and two Catalan socialist parties of the time (PSC-C and PSC-R) to create the PSC-PSOE.3 In this agreement, the direct participation of the members of the PSC in the Federal Congress through delegates elected in the same way as those of the PSOE was established (thereby through the local branches, and in practice the decisional structure of the PSC is kept), as well as a representation of the PSC (proportional to the number of its members) in the Federal Committee and a participation in the Executive Commission. After the Federal Congress elects the General Secretary, he or she proposes to the Congress the members of the new Executive Commission. According to the Protocol, the representatives of the PSC in the Executive Commission will be “proposed or endorsed where appropriate” by the PSC delegates in the Congress. Needless to say, this splendidly ambiguous wording has usually led to last minute bargaining between the recently elected General Secretary and the leaders of the Catalan delegates in the Congress. Roller and van Houten (2003) speak of “the PSC PSOE’s dual nature as a regional party and an affiliate of a national party.”

In the words of Verge and Barberá (2009), “the relation between the PSOE and the PSC has been characterized by a considerable ambiguity” (our translation). Initially, the PSOE and the PSC formed separated parliamentary groups in the lower house of the Spanish parliament, although “under common discipline of speech, action, and vote,” with decisions being made after deliberation in common assembly and joint meeting of both permanent committees. Eventually a common parliamentary group was constituted, in whose executive organs the presence of PSC parliamentarians has always been above their proportion in the whole group; the vote discipline has been very strict (Colomé 2003). As in the Spanish Constitution, the President of the Government is elected by the lower house of Parliament, and due to the ancillary role of the upper chamber in the legislative procedure, the common action of the PSOE and the PSC in the lower chamber brings about a common action in national politics. In fact, until very recently, most Spaniards have perceived both parties as “the same thing,” at least at the national level.

The incorporation of the Catalan Esquerra Unida i Alternativa (EUiA) in the statewide Spanish United Left (IU) is also an instance of party political, territorially-based collective membership. This incorporation is asserted in Art. 84 through 94 of the Statutes of EUiA, while both the independence of EUiA and its representation in the organs of IU are clearly stated in Art. 113 of the Statutes of IU.

A third instance of this type also concerns Catalonia. Convèrgencia Democràtica Aranesa-Partit Nacionaliste Aranès (CDA-PNA) is a political party of the Aran Valley, in the Pyrenees. By virtue of an agreement between CDA-PNA and the Catalan party Convergència Democràtica de Catalunya (CDC), the former has representation rights in the governing bodies of the latter and can be considered a collective member.

By contrast, the cases of the Italian Northern League, the British Conservative and Liberal parties, and the CDU-CSU fail to meet our criteria for collective membership. At first sight, the Northern League (Lega Nord) seems to be based on an articulating agreement between different territorial organizations. Its Statutes (Art. 1) indicate that the party is a confederation of different political organizations (national sections according to Art.2.). National sections have the right to participate in the decision processes of the whole party, in so far as the secretaries of each national section are members of the highest political body in the League, the Federal Council (Consiglio Federale) (Art. 12). However, national sections lack true political and institutional autonomy and are subjected to the decisions and rules adopted by the whole party. This situation resulted from the formation process of the party in the late 1980s and early 1990s. The party emerged out of electoral alliances among different local political organizations and took shape first in 1989 as a confederation of parties (Lega Lombarda, Liga Veneta, Lega Emilia-Romagna, Union Ligure and Movimento Autonomista Piamontese) within the Northern Alliance (Alleanza Nord) (Gómez-Reino 2002:95-95). In 1991, however, the Northern League was launched as a unified and even centralized party. In Gómez-Reino’s words (2002:96), “what was devised as a confederation evolved into a single centralized party under Umberto Bossi’s command.” The 1991 statute thus stated that the Northern League had “legal ownership” over all the different regional labels (Gómez-Reino 2002:97). This centralized outcome resulted from the hegemony of the Lombard League and its leader (Bossi), from the internal divisions suffered by several other local leagues, and from the electoral opportunities opened up by the crisis of the Italian party system (Gómez-Reino 2002:96-97).

In Great Britain there are some similarities between the structures of the Conservative and Liberal Democratic parties and the territorial types of collective membership, but none of these parties meet the criteria we defined in our classification. The association of the British Conservative Party with its Scottish organization could seem similar to that between the PSOE and the PSC. In its Constitution, the party considers as members not only the individual members (classified into “party members” and “Scottish party members”), but also the constituency associations and the “Scottish Conservative and Unionist Party” as such. But in fact, the Scottish party cannot be considered a fully self-governing organization. All the constituency associations (also the Scottish ones) are submitted to the British party and they are also the basis of its articulation. And the Scottish party members participate like the other members in the election of the key organs of the party.4 There are two substantial characteristics that separate the Scottish Conservative Party from the Catalan PSC. Firstly, the “constitutional power” lies in the (central) Conservative Party, which has the right to define the relation between the Scottish section and the global party. Secondly, the constituency associations in Scotland belong to the (global) party as such associations (although their members are “Scottish members”), and thus the supervision in the selection of candidates to the Parliament in Westminster corresponds to the global party. All in all, the Scottish Conservative Party can be considered an organization within the articulation of the global Conservative Party rather than a collective member as defined above.

As for the British Liberal Democrats, they have an explicitly federal structure, and the Federal party is integrated by the state parties of England, Wales, and Scotland. However, these state sections lack full autonomy and are bound by the Constitution of the Party and by the rules made by it (Art. 2.2 of the Party Constitution).5 In addition to that, the Federal Party reserves to itself exclusive control on several crucial policy areas (Art. 2.3 of the Party Constitution).

Finally, another instance which is close to an articulating agreement between political parties is that of the CDU-CSU. However, the relation between the CSU and the CDU in Germany is that of a coalition and therefore the CSU is not a collective member of the CDU. In a coalition, none of the parties receives an entitlement to decision power in the organizations of the other parties. Here it is necessary to distinguish between the decisions taken by the organization of the party and those made by their members in the institutions of the state. Sometimes this distinction is not easy, and this is particularly the case with parliamentary groups. In the British tradition, parliamentary groups are part of the organization of the parties and play a role in their internal decisions (McKenzie 1963). In continental political parties, parliamentary groups are not generally part of their organization, although they are arguably part of their decision structure. Then it may be the case that the parties in a coalition join in a single parliamentary group. This is the case of the CDU and the CSU in the German Bundestag. In general, it is difficult to say how far the party allegiances of the members of such mixed parliamentary groups shape the political alignments in the discussions and how a common position is reached, both among the members of each party and in the whole group, either by majority voting or negotiation. As for the CDU and the CSU in the Bundestag, there is one organized subgroup for each party. According to the current agreement between these parties, “fundamental political decisions of the parliamentary group CDU/CSU are to be taken only in agreement between both subgroups” (our translation).

Belgian cultural pluralism might seem prone to the structures of party-political, articulating collective membership, but the divisions of socialist, Christian-democratic, and liberal parties along linguistic lines from the 1960s onwards excluded this possibility. In Portugal, the possibility of creating this type of collective members has been excluded by national legislation, which directly prohibits the creation of regional parties (Art. 9 of the Portuguese Law of Political Parties). This exclusion may have played an important role in the development of the party system of Portugal’s Atlantic islands.

We can find instances of a non-articulating linkage between political parties in the small parties embedded in the UMP, the Christian-Democratic Party (Parti Chrétien-Démocrate) and, until 2011, the Radical Party (Parti Radical) under the framework of Art. 3 of the Statute of the French UMP (Union pour un Mouvement Populaire) which contemplates the possibility that legal persons (political parties or not) are associated to the party.6 The roots of this organizational structure are political-ideological rather than territorial and relate to the historical pluralism of the “non-Gaullist moderate right” (Knapp 2004:199), in which three main types of forces were traditionally present: the Christian Democrats, the Conservatives, and the ideologically oscillating Radicals (Knapp 2004:232). These three currents were also present in the now extinct Union for French Democracy (Union pour la Démocratie Française), which was defined in its statutes as a “federation of political parties” (Knapp 2004:216) and which worked in fact as a confederal organization of three main independent forces (the Christian Democratic Center of Social Democrats, CDS; the conservative Independent Republicans, RI; and the Radical Party), all of which enjoyed equal representation rights in the UDF national bodies (Knapp 2004: 216-7). To some extent, these parties of notables (Knapp 2004:217-18) continue the long-standing Orleanist tradition of the French right (Knapp 2004:201; Rémond 1954; Moreno Recio 1918:149). Despite their residual character, they have also held important positions of power at the local, regional, and national levels. For instance, “they enjoyed a constant majority in the upper House until 1980” (Knapp 2004:214 and 217). Given the symbolic, strategic, and institutional importance of the groups of the non-Gaullist right (Knapp 2004:221), it was clearly justified for the UMP to create specific organizational frameworks in order to absorb these parties (2007:219).

For any agreement between political parties, articulating or non-articulating, the question arises as to whether one of the parties becomes a collective member of the other, or there is merely a coalition between them. The relation between an associated legal person and the UMP is governed by the association agreement, a “contrat de droit privé” according to the French Law of Associations of 1901. It is usual to keep most clauses confidential. The agreement between associated parties and the UMP usually stipulates that the joint candidate to the presidential election is to be chosen by the UMP. Also the parliamentarians elected to the National Assembly typically integrate into the common parliamentary group. As for the meaning of this to guarantee common political action, the personalized character of French politics and the fluidity of its party landscape are to be taken into account. Typically the members of the associated legal person become members of the UMP if they wish, a representation in the National Council of the UMP is agreed (under the conditions fixed by the Political Bureau of the UMP), and a subsidy is paid by the UMP to the associated legal person. Thus the associated parties are collective members of the UMP.

The parties integrated in the Belgian liberal party Reformist Movement (Mouvement Réformateur, MR) also fit our criteria of collective membership. The MR is integrated by the Reformist Liberal Party (Parti Réformateur Libéral, PRL), the dominant component in the party and which in fact is reported to have a “fictitious existence” (Dedecker 2011:149); the Party for Freedom and Progress (Partei für Freiheit und Forschritt, PFF), which represents German-speaking citizens; and the Movement of Citizens for Change (Mouvement des Citoyens pour le Changement, MCC), a party of Christian-Democratic origins that split from the PSC and first joined an alliance with the PRL in 1999. In 2002 this alliance gave rise to a new party, the MR.

There are no instances of articulating linkages between a party and functional organizations. Perhaps the closest case to this combination of circumstances is that of the Austrian Popular Party (ÖVP) and its interest organizations. These organizations have participation rights in the political structure of the ÖVP (Müller 2006:343) and also enjoy very high levels of autonomy, to the point that experts have considered the ÖVP as an indirect party (Müller 2006:349). In fact, there have been different attempts at curtailing the autonomy of the professional organizations of the ÖVP. The concern about intersectoral conflicts and the need to assert the leadership of the federal party can be perceived in the 1979 Organization Statute.7 Although something has been done since then in this direction, the ÖAAB, ÖBB and ÖWB are still a key factor in the dynamic (and the static) of the party. However, although these organizations enjoy high levels of autonomy, they do not meet the criterium of being fully self-governed, and are subjected to the rules and political decisions of the ÖVP.

In addition to the formal rules that specify representative rights in the party federal bodies (as in the Bundesparteitag; Müller 2006: 343), there are informal criteria for the sharing out of power among the three professional organizations. In this sense Müller (2006: 349) writes: “When the composition of some body of the party is not to be decided through delegation of the professional organizations, but through election, there is an implicit apportionment (Proporz) among them, which today reaches still up to the leading bodies of the party, and earlier even included the top positions (leader of the party, general secretary, leader of the parliamentary group)” (our translation).

The ÖVP has a double articulation: territorial (Landesparteien) and sectorial, through interest sectors (Teilorganisationen). It is an archetypal case where the exclusivity of the territorial articulation does not hold. Both territorial and sector organizations have juridical personality. There are as many territorial parties as regions, and six sector organizations for blue-collar and white-collar workers (ÖAAB), farmers (ÖBB), business people (ÖWB), women (ÖFB), young people (JVP), and senior citizens (ÖSB). Even if it is possible to be a member of the ÖVP without also being a member of one of the sector organizations, this is rare in practice.8 The three latter sector organizations have correlatives in many other parties, and respond to the idea of “natural groupings” for collectives supposed to need their own channels of representation and the defense of whose interests is not considered particularly “divisive.” In contrast, the three former ones (ÖAAB, ÖBB, and ÖWB) are supposed to represent all occupational sectors, to which are (were or are to be) bound, directly or indirectly, all citizens: they constitute a further articulation of the party, parallel to the territorial one. Certainly the fact that ÖAAB, ÖBB, and ÖWB represent all occupational sectors does not mean that the membership of the ÖVP is a mirror image of Austrian society. Farmers, public servants, and business people are overrepresented in the ÖVP. In the case of the ÖVP, this overrepresentation has reached some crystallization through the professional organizations.

In the German (Bavarian) CSU there are also organizations (Arbeitsgemeinschaften) for employees, business people, and farmers (art. 27 of the Statute). The same can be said of the CDU, where there are organizations (Vereinigungen) for employees and business people (art. 38 of the Statute). However, this panoply of organizations is not intended to articulate the party, but to look after specific groups by targeting the message of the party and channeling impulses and suggestions. Furthermore, these functional organizations also lack the level of autonomy of the other collective members discussed above. In both respects, the occupational organizations are far from being structurally essential in the case of both the CDU and the CSU (Mintzel 1978).

The trade unions affiliated to the British Labour Party are a clear instance of a functional non-articulating collective member. The origins of this relationship lie in the formation process of the Labour party itself, which was initially an extension of the trade unions (Scarrow 1996:24). The non-Marxist character of British social-democracy may have also favored the persistence of this type of organizational linkage (Bartolini 2000:70-72). The British Labour Party admits not only individual members in its structure, but also “affiliated organizations,” among which the trade unions play a very important role in the life and governance of the party. Half of the total voting entitlement in the Party Conference corresponds to the affiliated organizations (in practice, mostly the trade unions), and half of the elective members of the National Executive Committee (NEC) are appointed by the trade unions. In the election of the party leader, the same voting entitlement (one third each) is apportioned to the affiliated organizations, to individual members, and to members of the Commons and the European Parliament. The election procedure to elect their delegates in the Party Conference is up to every union (these delegates choose in turn the representatives in the NEC), whereas for the election of leader of the party it is laid down that all union members paying the “political levy” should participate on a one-person-one-vote basis.

The members of the affiliated trade unions can opt out of paying the political levy, in which case they have no right to vote in the decisions of their union as an affiliated organization of the party. But this “opting out” must be explicit. In contrast to this, from 1927 to 1946 the acceptance to pay had to be explicit (otherwise the political levy was not exacted), and the result was a marked fall in the number of trade unionists paying the political levy.

When collective members are not political parties, they normally represent interest groups. If a party has such collective members, rival parties may claim to defend the national interest and not that of particular groups. In a time of falling partisanship (Dalton and Wattenberg 2009), parties trying to represent a particular section of society risk missing decisive electoral targets. Certainly it has been affirmed that the trade unions have historically been moderate in applying their clout in the Labour Party: “Restraint has been the central characteristic of the trade union-Labour Party relationship” (Minkin 1992:26). At any rate, after the “winter of discontent” in 1978-79, the prevailing perception among the British electorate was that the Labour Party was too dependent on the unions, and that the interests of the unions did not coincide in several substantial issues with the national interest. A succession of electoral defeats (until the “New Labour” victory of 1997) prompted the party to gradually reduce the power of the unions qua unions in its organization (Quinn 2004). As Kitschelt (1994) has shown, the organizational structure of the British Labour party contributed decisively to the party’s lack of adjustment to new social demands and to its poor electoral performance in the 1980s.

As in the case of the British Labour Party, the Swedish Social Democratic Worker’s Party (SAP) has historically had a strong trade union basis (Verney 1957). It has also been characterized as an indirect party (Duverger 1958:28). From 1909 to 1990, the Swedish Social Democratic Party offered the trade unions a “system of facultative collective affiliation tempered by an opt-out clause”, as characterized by Fusilier (1954:65). Two features distinguished this system from that of the British Labour Party. Firstly, the Swedish collective affiliation operated at the local level: the decision to join the party was not up to the national unions as in Britain, but each local section of each union had to decide. Secondly, in the British case, the members of the affiliated unions become “indirect members” and have a different status from that of “direct members” (and the most committed members of the Unions can also opt to be direct members), whereas in Sweden there was only a category of party members and the “indirect” members were indistinguishable from the “direct” ones. For this reason, in the Swedish Social Democratic Party one can speak of “collective affiliation”, but not of “collective membership,” as defined here. Although the Swedish Trade Union Confederation has a representative on the party’s executive committee, and the relation between unions and the party is strong (Anthonsen et al. 2011), there is nothing that can be described as participation in shared-rule of the unions in the party.

In this article, we have examined different types of collective membership in contemporary political parties. Our exploration has focused on organizations that enjoy full self-government rights and have entitlements to power in the political party in which they are embedded. Our analysis has proposed a basic descriptive typology of collective members based on their nature (party political or functional) and their pattern of articulation to the political party they are embedded in (whether they help to articulate the whole party structure or not).

Although our typology had a mainly descriptive character, the comparative examination of the cases we discussed above makes it possible to explore the determinants of this type of political organization. It is clear, in the first place, that national political institutions do not exercise a direct influence on the emergence of collective members. The electoral systems of the countries considered here include proportional representation (with closed and blocked lists as in Spain or with preferential vote as in Belgium) and relative or absolute majority representation in uninominal districts (Britain and France). Moreover, their political-territorial structures cover very different degrees of regional autonomy, ranging from federal (or quasi-federal) structures to unitary patterns of territorial articulation. National political institutions may have shaped some specific characteristics of collective membership in these parties, but they certainly did not give rise to this type of organizational linkage in these cases.

Certainly, as Duverger already saw apropos of indirect parties, social democratic (particularly of a non-Marxist kind) and Christian democratic political ideas can have facilitated the existence of functional collective members. However, the party political collective members analyzed here lacked these facilitating ideological conditions, even if their existence could be loosely linked to specific idiosyncratic political-ideological traditions (such as the federal tradition of the Spanish left). Furthermore, the Christian democratic and non-Marxist social democratic ideological traditions were present in other parties that did not even approach the development of collective members, functional or of a party political nature.

By contrast, in all positive cases, the creation and persistence of collective members can be related to the heterogeneity of the political and organizational landscapes in which the political parties evolved. The character of that heterogeneity may lie in the cultural-territorial domain (as in the case of the Catalan PSC and EUiA, as well as of the Belgian PFF), in the ideological diversity of a particular political camp (as in the cases of the French and Belgian center right), or in the strength of trade unions before and during the formation process of the British Labour Party.

Furthermore, in all these cases, collective members were created exogenously to the political party they joined. In all these respects, the emergence of these arrangements can be linked to the starting moments of political parties and their underlying social and political conditions. This comparison reveals the importance of path-dependent historical trajectories. The role played by exogenous organizations at starting moments can also be grasped in a negative way, since we have not identified any instance in which a party made a top-down move towards collective membership. In those cases in which a party moved towards a structure based on overarching collateral organizations (as was the case of the Austrian ÖVP), such organizations did not attain the levels of political autonomy that distinguish collective membership.

The formation processes of collective members resemble those of federal and confederal systems (as identified by Riker 1964). Collective membership agreements implied exchanges between the leaders of organizations placed at different levels. But agreements in this case were dominated by electoral and policy-making considerations, instead of the military factors that Riker identified in the birth of federations. Formative agreements between exogenous organizations worked as a necessary but insufficient condition for the existence of forms of collective membership. Balances of power among collective members and national political parties, which are conditioned in turn by different factors such as changes in social and political conditions and individual leaderships, may also lead to the extinction of this type of political structure. The case of the Northern League, which emerged as a political confederation and moved immediately towards a centralized party under the strong leadership of Bossi, shows that exogenous origins do not guarantee the persistence of this type of organizational linkage. And as Christiansen (2012) has shown, changes in socio-economic conditions and public attitudes can also affect the benefits that political parties and interest groups extract from their cooperation agreements, and lead to the weakening or even demise of such privileged links.

Current social and political conditions may have become adversarial to the development of these types of organizational linkages. The decline of socio-political cleavages, rapid changes in social conditions resulting from globalization, the growing importance of new communication technologies, and intense preferences for direct participation all conspire against the development of the comparatively rigid structures of collective membership. In this respect, the diminishing importance of party organizational bases and structures of belonging (Puhle 2002:79-80; Wolinetz 2002) poses some limits on the continuity of this specific type of organizational linkage.

One way to assess the perspectives of this type of organizational linkage consists in examining recent instances of party creation in which a high degree of preexisting political and organizational pluralism was present, and in which party creation did not strictly obey a top-down organizational logic, that is, in cases in which the political conditions were in principle favorable to the formation of structures of collective membership. These are the cases of several new organizations in the political space of the left and radical left in Southern Europe, such as Syriza in Greece, the Nouveau Parti Anticapitaliste and the Front de Gauche in France, and Podemos in Spain. In all these cases, notwithstanding the high degree of pre-existing political and organizational pluralism, new organizations avoided the creation of collective members.

In the case of the parties of the French radical left, their attempts to deal with the preexisting political-organizational complexity of their ideological space did not lead to the formation of collective members. Mélenchon’s candidacy in the 2012 presidential elections, which received more than 11% of the vote in the first electoral round, was endorsed by a classical electoral coalition, the Front de Gauche. As for Mélenchon’s party, the Parti de Gauche (PG), its statutes defined it as a party of militants integrated in circles (Art. 8) and they also forbade its members to belong to any other party organization (Art. 2). The party statutes accepted the possibility of enlarging its Bureau National in order to include representatives of other parties, but this circumstance was restricted to parties merging with the PG (Art. 13).

The other recently created political formation of the French radical left is the Nouveau Parti Anticapitaliste (NPA), which was founded in 2009 by political forces of Trotskyite orientation. Its main leader, Olivier Besancenot, attained more than 4% of the vote in the first round of the 2007 presidential elections, and the sum of votes for the three candidates of the Trotskyite parties was over 10% in the first round of the 2002 presidential elections (Sperber 2010). Following the Trotskyite political tradition, the 2009 statutes enabled the creation of tendencies and fractions within the party, and specifically addressed the problem of assuring voice and representation to different positions and platforms in the Conseil Politique National (Nouveau Parti Anticapitaliste 2009). However, the party chose to reject specific organizational linkages for different political groups or ideological streams. Overall, the party explicitly positioned itself in favor of direct democracy and against internal structures that filter the voice of militants.

The degree of previous political-organizational pluralism was lower in the case of Podemos, even if also in this case exogenous social movements and political organizations (in particular Izquierda Anticapitalista) took part in its formation process. Podemos avoided not only collective membership, but also specific organizational linkages for any kind of collateral organizations. Furthermore, the new party adopted the rule that no members of other national political organizations could be elected to positions of authority within the party (point XI.e of its Ethical Code) (Podemos 2014). This policy of exclusion of internal party political pluralism led to the dissolution of Izquierda Anticapitalista as a political party and to its transformation into an association. In the case of Podemos, the appeal to direct participation and the mistrust towards exogenous political structures was most likely combined with a strategy aimed at maximizing the strategic and top-down coordination of the party, as became apparent in the voting system chosen to elect the 62 members of its highest national political body, the Consejo Ciudadano, by the Asamblea Ciudadana (which is composed of all the members of Podemos). Under the voting system selected by this party, participants were able to choose in block a general list of 62 candidates (by a computer click in this case). In the 2014 internal election, the 62 candidates elected were part of the list endorsed by the party leader, Pablo Iglesias.

Finally, the exclusion of collective members is also remarkable in the case of Syriza, which was created by a wide array of social and political organizations. Syriza eluded collective membership, and chose to encourage direct individual participation and open decision processes, as well as to transform its former components into tendencies (Tsakatika and Eleftheriou 2013:12-14; Bournous 2013).

Certainly, these outcomes may have been favored by the linkage strategies adopted by the new radical left, which have emphasized direct, mass-based, and bottom-up participatory procedures, and which have also shown mistrust of bureaucratic structures (including trade unions) (Tsakatika and Lisi 2013:9-14). We may be witnessing a relatively recent trend in this respect. The previously discussed EUiA was integrated into IU in 1998 under the form of collective membership. And a few decades ago, in 1982, a radical left organization, the Galician Nationalist Block (BNG), chose to include collective members in its organization, most notoriously, the Union of the Galician People (UPG); a fully independent organization that is represented in the national bodies of the BNG and which has consistently played a decisive political role in it (Vilas Nogueira and Fernández Baz 2004).

But it is also quite plausible that the rejection of the collective membership formula by new political parties is decisively determined by the goal of achieving greater ideological and programmatic flexibility. Highly decentralized organizations can set strong constraints on the strategic flexibility of political parties, and can thus lead them (as was the case of highly decentralized parties based on socialist clubs) to electorally suboptimal strategies (Kitschelt 1994:218-253).

Both theories on the evolution of party organizations and the empirical examination of recent processes of party formation under conditions of strong political pluralism reveal the limited appeal of this type of party structure nowadays. However, given the inertial character of political institutions, its persistence cannot be discarded. Furthermore, there can continue to emerge circumstances in which this type of political structure is adopted (as in the case of the Belgian MR) or at least considered (as in the case of the Italian Northern League). All in all, the analysis of this type of organizational linkage remains important for our understanding of the determinants and implications of different types of party structures.