Party adaptation and change and the crisis of democracy

Introduction

This special issue is a collection of papers written and presented at a conference held in September 2012 at the European University Institute (EUI) in honour of Peter Mair in celebration of his memory and intellectual legacy. The conference inevitably focused on the themes central to Peter's interests and scientific output – parties and party systems – as well as on their importance for the functioning of contemporary democratic systems. This initiative was met with great enthusiasm by many of Peter's colleagues and collaborators, which in practice meant many of the most influential scholars in the field. The number and the quality of the contributions presented at the conference induced the convenors, Luciano Bardi, Alex Trechsel and Stefano Bartolini, to organize them for publication in two special issues of leading political science journals: Party Politics and West European Politics, both of which were very close to Peter's heart. The West European Politics special issue focuses on themes that had become of fundamental importance to Peter in the latter part of his career: those linked to the nature and causes of the tension between responsiveness and responsibility and to the impact such tension had on the role of parties in democracy. This special issue of Party Politics, on the other hand, includes papers that are relevant to what was for a long time Peter's most important and to some extent most original research interest: party and party system adaptation and organizational change.

Peter Mair was one of the most accomplished scholars in this field. Indeed, the question of party and party system change can be seen as the unifying concern of Peter’s intellectual contribution (see also Bardi et al., 2014b). For most of the past thirty years, the study of political parties has focused – also in part due to Peter’s influence – on aspects related to the transformation of the parties themselves. The growing scepticism surrounding the role of parties in contemporary democracies has induced many party scholars to progressively shift their focus from the study of the functions parties allegedly perform in democratic systems to that of the changing role of parties and their consequent adaptation. ¹ This tendency, very well documented in the articles in this issue, is implicitly or explicitly visible in several empirical works (Katz and Mair, 1992, 1994; Dalton and Wattenberg, 2000; Webb et al., 2002), as well as in contributions containing new conceptualizations and theoretical reflections (e.g. Katz and Mair, 1995; Luther and Müller-Rommel, 2002; Carty, 2004).

Peter Mair was one of the initiators and protagonists of this trend. Most importantly, he was also one of the most active scholars among those responsible for the resurgence of theoretical reflection on parties. The point of departure of this resurgence was probably at the beginning of the 1980s, with his first contribution (Mair, 1983) to what can cumulatively be described as the theory of party organizational change. The stimulus underpinning such reflection certainly came from his desire to understand the reasons for the crisis of parties, but even more importantly it stemmed from the need to explain the causes of the evident transformations parties were undergoing, from the catch-all party onwards. Party adaptation to changing macro-economic and societal conditions was the central problem of the theory, which unfolded through many other works based on, or inspired by, the cartel party model that Richard Katz and Peter Mair had developed together (1995). From the cartel party onwards, the attention of scholars was directed toward the position that parties occupy between society and the state, a crucial aspect that had been almost completely overlooked up to that point. Katz and Mair (1993) were also responsible for a review of the stratarchic party model (Eldersveld, 1964). Their conceptualization, which is based on the three organizational faces of parties, inspired later works on the internal diversification of party organization, thus favouring the completion of a process that has produced the replacement of the original hierarchical/monolithic model with alternatives such as the multi-faceted franchise party (Carty, 2004) or the hierarchy/stratarchy/federation organizational tripartition (Bolleyer, 2012). The interest in party organizational change also stimulated Peter Mair's theoretical and conceptual efforts on the topic of party systems (Mair, 1996; Bardi and Mair, 2008). The emergence of new types of parties – be they a consequence of new organizational models or of new cleavages – can in fact be seen as the ultimate manifestation of party change, but it is also certainly one that has powerful potential implications for party systems as well. ² The articles in this special issue all draw from and contribute to the further development of the debate on party and party system crisis, adaptation and change, as well as to the positioning of parties between society and the state from both theoretical and empirical perspectives.

The crisis of parties and the decline of the hierarchical/monolithic model

It could be argued that the crisis of political parties is not a new phenomenon. Historically, political parties have been favourite targets for criticism, as Piero Ignazi notes in his contribution (this issue p. XY). This is probably due to the generally negative perception citizens have of politics – which is often seen as something corrupt or difficult to understand – and of its protagonists of more than a century, the parties. This attitude, which according to Ignazi precedes the birth of modern political parties, prevails irrespective of the efficacy of parties, or rather of their capacity to serve politics effectively. For centuries, citizens, commentators, scholars and critics have accepted political parties as a necessary evil for the functioning of democracy.

The literature is dense with significant evidence of the growing distance between citizens and parties; decline in electoral turnout, collapse in party membership figures and member disenchantment are also very well documented in several contributions in this issue. The link between citizens and parties, however, was weakened as a result of a two-way process: on the one hand, post-war economic growth and widespread increases in levels of education and secularization produced significant changes in social stratification, thus weakening the privileged relationship between specific groups of citizens and their corresponding parties; on the other hand, political parties in response attempted to broaden their appeal to more than one group in society. The result of these tensions between citizens and parties was the catch-all party model (Kirchheimer, 1966), which conceptualized the ongoing transformation of already existing parties, which were usually formally respectful of the mass party model. According to Kirchheimer, party change was first determined by societal transformations, which in turn modified the attitudes and behaviour of citizens and, consequently, their demands toward and their relationship with parties. These societal transformations had obvious consequences for party organization and for the performance of party functions. Kirchheimer thus portrayed the transition of the decline of industrial society and its impact on the prevailing party model of the time, the mass party.

As such, the effects of the transition were not accurately perceived initially insofar as parties maintained strong grassroots organizations characterized by large memberships. In other words, party change occurred more in terms of the quality of the organization than in terms of its strength. This largely concerned the inversion, at least at the formal level, of the relationship between party members and the leadership, wherein the latter progressively became more autonomous and powerful. At the same time, the membership's role declined as parties resorted to different sources of party financing, but it also declined because of the increase in the professionalization of the organization.

The theme of party adaptation and change was studied in depth by Peter Mair, also in joint efforts with other scholars. Mair was the first to identify party organization as one of the reasons for what was being described with increasing frequency as the crisis of parties (Daalder, 1992). In an elaboration on the observation he made with Stefano Bartolini in their work on the stabilization of European electorates (Bartolini and Mair, 1990), Mair developed the conviction that increases in electoral volatility were not the consequence of a weakening of ideologies – a position that was not congruent with the relative stability of right-wing and left-wing blocs – but rather were caused by the specific characteristics of individual parties, which augmented or diminished their ability to attract electors within their respective blocs (Mair, 1990: 177–184). Volatility was thus the result of an instability of individual parties caused by differences in leadership, in specific policy positions – albeit within the broad ideological orientations of party blocs – and, lastly, in strategies and organizational styles. Peter Mair considered this last factor to be of the greatest potential importance. Individual electors generally remained attached to the same broad party family, tendency or bloc, but their individual connections to specific parties could be eroded by the weakening of their organizational links with the party (Mair, 1990: 179). This situation resulted in an electoral market that was only partially open and shaped by the political supply within each bloc. According to Mair, this could explain the simultaneous manifestation of party instability and the persistence of tendencies. With this analysis, Peter Mair admirably linked the transformation of parties to that of party systems and provided a more articulate and convincing explanation for the latter than could be found in the literature up to that point, wherein party system instability was attributed to simple if not clearly motivated changes in the voting orientations of electors.

In the stabilization of party systems, the mass party's remarkable organizational strength, reflected in the encapsulation of electoral/societal groups through the inculcation of specific party identities, was an even more important factor than the structuring of the social cleavage system. Hence, as party systems became progressively unstable, the hypothesis that this was caused by an organizational decline of the parties seemed plausible. Contrary to most scholarly reactions to the catch-all party, which primarily addressed the implications of this last for the decline of party ideology, Mair's reflection concentrated on the more numerous, and in his view more important, elements of organizational change between the catch-all party and the mass party model, i.e. the strengthening of the party's top leadership; the weakening of the role of the individual member; the loss of the classe gardée's centrality; and the opening of the party to a variety of interest groups. According to Peter Mair, not only were all of these elements factors of change, they were also indicators of party adaptation to very broad transformations that had taken place in the economy and in society and which were beyond party control.

At the same time, other processes were unfolding that had a direct impact on the electoral alignments of parties and created disturbances in the party–electorate relationship. The most important of these processes concerned the blurring of class distinctions and, with the creation of the post-materialist social cleavage, the emergence of new value systems. Created in response to these challenges, the new party model increased, or at least reflected, the electoral vulnerability of parties. This transformation in organizational approach, which was based on essentially top-down decision-making processes as opposed to the bottom-up flows of the past, resulted from: (a) changes in party financing, whereby parties were increasingly reliant on the replacement of membership fees and other forms of grassroots-based revenue with state contributions; (b) changes in political communication between the party leadership and the electorate, wherein communication through the organization networks of parties was progressively eclipsed and substituted by direct communication through the mass media; (c) changes in policy-making that came from the use of marketing and opinion polling techniques in the formulation of policies at the expense of the demands and suggestions that came directly from the rank and file. These new options were not per se causes of the transformation of mass parties into catch-all organizations, but they made this transformation possible when changes in social stratification and in value systems required adequate responses. This analysis was based on considerations suggested by Kirchheimer's catch-all model, which was explicitly defined through its differences to the mass party. As a result, the mass party remained the principal, if not only, reference model. It was only with the theoretical debate generated by the cartel party model (Katz and Mair, 1995) that the foundations of a new benchmark were created.

Theories of party organizational diversification: From the cartel party to the franchise party

The long-standing and sustained collaboration between Richard Katz and Peter Mair produced numerous and very relevant contributions for party research and literature – from the project on party organization (1992, 1994) to the conceptualization of the three faces of party organization (1993). However, their best known, and in all likelihood most influential, contribution is the article (1995) that introduced the first party model – the cartel party – since the introduction of Kirchheimer's catch-all party model and its subsequent elaboration by Angelo Panebianco (1988) as the electoral professional party. The article represents a synthesis of reflections made by the scientific community on party organization and change in the last decades of the 20th century. It is important to note that the two authors do not consider the new model to be a complete replacement of the previous model, as had Duverger (1951) with regard to the transition between the cadres and the mass party, but rather they see it as a continuous process of adaptation to changing environmental conditions; moreover, this model is in perfect harmony with and constitutes the perfect continuity of Peter Mair's previous work. ³

The cartel party is ‘characterized by the interpenetration of party and state, and also by a pattern of inter-party collusion’ (Katz and Mair, 1995: 17). Like the catch-all party, the cartel party was not conceived in strictly organizational terms. Both models reflect changes in society and in the state that modified their respective relationships with political parties. The underlying hypothesis of the Katz and Mair model is that the social, cultural and, most importantly, political changes that have been experienced by contemporary democracies have resulted in the progressive ‘anchoring’ of the parties to the state, this last being the means through which parties obtain the necessary resources for their survival and maintenance of their power. How parties are positioned between society and the state becomes a crucial element in the definition of the model and in the way in which parties pursue their organizational imperatives. As parties move away from society and occupy the state, they can more easily take advantage of their position as legislators and approve norms in their own favour, particularly those regarding privileged media access and party financing by the state.

Precisely because of the importance they attach to the position of parties between society and the state, the Katz and Mair model is more dynamic than its predecessors. This allows them to entertain not only the idea that differences in the organizational solutions chosen by individual parties are possible, but also that the process of party change may be gradual. When faced with particular environmental conditions, parties all tend to position themselves in the same way with respect to society. ⁴ The relationship between parties and the state varies, however: the so-called government parties are more inclined to conform quickly to the cartel party model. The same cannot be said about new parties or about those that are relegated to opposition roles most of the time and for those that are defined as anti-system or anti-politics parties. Thus, according to Katz and Mair, ‘the development of party organizations reflects a stimulus-response dynamic’, the sequence of which can be observed but does not ‘necessarily characterize the trajectory of every specific party’. As a result, no ‘single evolutionary model with a single dominant party type’ exists; each model can then be seen to be composed of a number of separate organizational solutions that can be assembled in various ways by parties according to their individual needs. In their view, ‘there are in fact no fully fledged cartel parties, just as there are, or were, no fully fledged mass parties, catch-all parties, or cadre parties’ (Katz and Mair, 1996: 532). Thus, Katz and Mair offered more than a new model, they offered an approach to use in the study of party organization; many of the contributions in this issue make full use of the intuition they had.

The other important contribution Katz and Mair have brought to the study of party organization comes from their identification of the three faces of party organization: the party on the ground, the party in central office, and the party in public office (1993). This conceptualization is complementary to that of the cartel party and suggests the need for an even more explicit deconstruction of party models. The links among the various organizational faces go in very different directions depending on: the different phases of the electoral cycle (campaign, election, legislative term, end of legislature); the political-institutional positions of parties (government, opposition); and the party organizational level (national, regional, local and, we could add, European). These variations allow for a very high number of possible organizational combinations. Consequently, the statement that no party is or ever was fully mass, catch-all, or cartel appears plausible; moreover, one could surmise that within each party the different prevailing characteristics (mass, catch-all, cartel) will vary between faces and, above all, between levels.

Kenneth Carty (2004) exploited the potentialities of the more flexible conception of party models that emerged out of the cartel party debate. He was the author of a framework for the analysis of stratarchic tendencies that he had originally observed in his native country, Canada, and that are also present in other political systems. The framework is based on the definition of franchise parties; it departs from Katz and Mair's tripartition and builds on their observations concerning trends in party organizations from hierarchic to stratarchic models. According to Carty, ‘what is required is a framework for thinking about party structures that identifies the autonomy of their various parts as a defining feature while recognizing the integral character of the organization as a whole, and allows us to explore how individual parties operationalize and institutionalize the stratarchical imperative in form and practice’ (2004: 7). The model thus captures the organizational flexibility of contemporary political parties, a feature that perhaps always existed but was never recognized.

Underpinning Carty's reasoning is the conviction that hierarchy and stratarchy are not alternatives, but coexist within the multiple configurations of the interdependent links among the three faces of party organization. Naturally, this implies that the franchise model allows for at least partial autonomy of the party's peripheral structures vis-à-vis the central organization. Carty clearly separates his position from that of Rud Koole (1996: 518), who prefers to describe party organizational change as a ‘federalization’ of political parties, because this vision ‘underestimates the real autonomy that constituent elements in federal systems enjoy as well as not recognizing that, in such systems, local units can operate quite independently in areas of activity that transcend the merely local’ (Carty, 2004: 7). In reality, even if they can no longer be considered monolithic structures, political parties present diversified modes of internal power distribution that can be classified on a continuum that ranges from hierarchy to federation, upon which stratarchy – intended as a form of balance between local unit autonomy and central control – occupies an intermediate position that, in the great majority of parties, can be observed in its many variations.

Thus, the franchise party must be understood as an integrated structure with components that enjoy a degree of autonomy and, therefore, ‘is constrained by a fundamental interdependence of parts’ (Carty, 2004: 8). The model provides a framework for the study of the relationships between the central and the peripheral organizations of parties; it reveals how internal flows of resources, personnel and information vary according to an internal intra-organizational pact (Webb, 2000). This organizational transformation, owing to its superior flexibility and adaptability, necessarily provides more effective responses to the aforementioned changes in society, communications and political-institutional settings than the traditional party models. As such, Carty considers stratarchic transformations in party structures to be an imperative.

It follows quite clearly that the degree of stratarchic characterization of a party is simply an empirical fact to be ascertained in each individual case. Within a set of internal arrangements, whereby each component enjoys a degree of functional autonomy, the effective relevance and distribution of tasks can vary considerably from case to case without the franchise nature of the party being called into question.

The organization of parties in multi-level systems: Hierarchy, stratarchy and federation

The franchise party model is also the basis upon which a new approach has been developed for the study of parties as multi-level organizations, which thus far have mostly been objects of descriptive studies (Carty, 2004; Moon and Bratberg, 2010; van Houten, 2009). If we consider them, as the model suggests, as organizations with internally diversified structures characterized by varying levels of interdependence between their components, we must also recognize the need to study them in terms of the functions and power relations of such components. ⁵

Further progress in the study of parties in multi-level settings comes from Nicole Bolleyer, who departs from Eldersveld's original discussion of stratarchy (1964: 9) ‘characterized by the interplay of organizational units that act, to varying extents, autonomously, but in the end remain interdependent. The central party needs the local units because voters have to be mobilized on the ground, while the central party ensures the overall integration of the organization’ (2012: 316). This configuration had induced Eldersveld to define political parties as structures characterized by ‘reciprocal deference’ between centre and periphery. His analysis was based on his observation of American parties, which are notoriously characterized by the substantial inability of the central organization to sanction excessively autonomous peripheral structures; this led him to believe that tendencies towards the dispersion of power were prevailing, even if he admitted that this might be different in different systems. Bolleyer, accepts Carty's warning about the error made by those who dismiss the hierarchical elements present in stratarchic party organizations. However, hierarchical elements, which consist of centralized conflict resolution mechanisms, centralized control and distribution of financial resources, as well as of top-down modes of communication usually prevail, or at least remain prominent, when the party is formed through a top-down process prompted by the central organization's need to consolidate on the ground electoral support for the party. On the contrary, if the party is built with a bottom-up process resulting from a common decision by local, regional or state structures to create a national or supranational structure in order to better pursue their interests, there will be a tendency for such units to maintain considerable short-term autonomy and to minimize the potential for long-term centralizing trends (Bolleyer, 2012: 318).

Bolleyer continues her analysis by identifying three party types – hierarchies, stratarchies and federations – which are ‘defined by (a) the way competences and resources (e.g. candidate selection, conflict resolution, finances) are allocated to the national and regional or local level and (b) the extent to which national interest representation follows territorial lines (visible in the composition of central party organs)’ (Bolleyer, 2012: 319). This approach not only clearly and systematically expresses positions that are implicit in the franchise party model, it also provides an articulated analytical framework for the analysis of the three party types and for the placement of individual parties on the hierarchy–federation continuum on the basis of a number of variables, such as procedures in candidate selection, local autonomy, procedures for conflict resolution and financial control (Bolleyer, 2012: 320). As Ken Carty has already asserted, discussing the nature and the features of the party models that prevail in given historical periods is no longer useful insofar as it would inevitably lead to a sterile comparison between the new model(s) and the original hierarchic/monolithic benchmark.

It would seem that we have now come full circle; the long path taken more than 30 years ago by Peter Mair with his studies on party change and organizational adaptation to the pervasive party crisis that had stemmed from the inability of parties to perform their traditional functions has been continued with the debate generated by the cartel and franchise party models and has reached a provisional conclusion with the abandonment of integrated models as privileged analytical instruments for the study of parties. This should favour a return to a more empirical study of parties, especially in terms of their internal dynamics and of the way parties and their internal components perform their functions. The sustained global economic crisis has placed contemporary democracies under an inordinate amount of strain. As a result, scholarly attention has focused on how parties perform crucial functions for democratic institutions. Interest in the tension between responsibility and responsiveness, or rather on how the representational and governmental functions are performed, reflects this very well.

Party organization and functions: Between society and the state

As has clearly emerged from the analysis thus far, parties can best be characterized as organizations that are situated between society and the state. Depending on how they position themselves between the two, parties develop different organizational structures and strategic objectives as well as different instruments for the performance of their functions. It is true, as Zsolt Enyedi states in his article in this issue, that the society–state dichotomy could perhaps be seen as an artificial academic construction and that in reality what we call society and state are indeed a whole that cannot easily be separated if not for analytical purposes. Analytically, however, the distinction has proved to be extremely useful and capable of stimulating important theoretical and conceptual advances. Be it as it may, the seven articles in this issue can best be classified in terms of their analytical point of departure, i.e. in terms of how they view the position of parties vis-à-vis society and the state.

The articles by Piero Ignazi and Ingrid van Biezen with Petr Kopecky clearly focus on the anchoring of parties to the state; two other articles, those of Richard Katz and Zsolt Enyedi, attempt to consider both aspects, and the remaining three – respectively by Ingrid van Biezen with Thomas Poguntke, Cas Mudde, and the collaborative work of Michael Alvarez, Ines Levin, Peter Mair and Alexander Trechsel – more or less directly privilege party positioning in society. These different viewpoints provide a combined mapping of structural and procedural strategies developed by parties to accomplish their organizational and functional objectives. Piero Ignazi's classic overview of party organization – Power and the (Il)legitimacy of Political Parties: An Unavoidable Paradox of Contemporary Democracy? – privileges the progressive encroachment of parties on the state. Consistently with the cartel party analysis, Ignazi believes that contemporary party organizations are shaped by their reliance on the resources and benefits resulting from the parties' occupation of the state, a manoeuvre aimed at ensuring their survival. He then takes the cartel party analysis one step further in his discussion of some of the internal organizational implications of this move. Party organizations have become more powerful by the exploitation of state resources, but they have also experienced a loss of legitimacy. Paradoxically, they are now better equipped to make policy decisions that could even be in the objective best interest of the citizens they represent, but they have lost the confidence of the citizenry as a result of their detachment from society. Attempts to bridge the gap at least with the bottom ranks of their organizations through such measures as the attribution of a more important role to ordinary members in leadership selection procedures have only backfired because they have been perceived as no more than a form of lip service to internal democracy. Parties may resemble leviathans in their might, but they are built on very unsteady foundations.

Ingrid van Biezen and Petr Kopecky's focus on parties and the state is even more explicit. Their approach, however, is very different from Piero Ignazi's and the conclusions they reach diverge somewhat from his. Their article – The Cartel Party and the State: Party-State Linkage in European Democracies – is based on an empirical analysis of three dimensions that they consider to lie at the basis of the party–state linkage and to be crucial to the cartel party thesis: the dependence of parties on the state, the management of parties by the state, and the capture of the state by parties. They build on a framework developed in their previous research on this theme and exploit original cross-national data on party regulation and party patronage. The latitude of the available data makes for rich, if not always univocal, analysis. In general, the party–state relationship has become stronger across European countries in recent years, at least as measured by the first two dimensions. With regard to state capture by parties, however, the evidence is less clear, mostly because of differences in the patronage levels and practices of the countries studied. In particular, where stratarchic tendencies are most evident, as in the case of Italy, central party organizations are often bypassed by party power elites in the peripheries. They explain this result by pointing out that the very complex regulatory systems developed by parties, ostensibly to their advantage, may have instead rendered them subject to tighter controls and may have diminished their hold on the state and their resources. Stratarchic solutions, however, can offset some of the shortcomings of the central organization.

Richard Katz's article – No Man Can Serve Two Masters: Party Politicians, Party Members, Citizens, and Principal–Agent Models of Democracy – revisits a familiar theme in the literature on parties: principal–agent models of democracy. Ordinarily, this is a theme that requires an approach to parties that considers both sides of the society–state distinction, wherein principals are normally found in more or less elevated strata of society and the agents are usually part of the party in public office, which is very close to if not actually part of the state. This is the case for this article as well, to the extent that the argument presented here takes into account the complexity of party organizational structures. The simple, standard principal–agent model does not adequately explain the relationship between portions of the electorate (i.e. elements of society) and the respective parties in public office (theoretically their agents vis-à-vis the state) insofar as the former interact only with the face of the party's organization. By disaggregating the party into its three separate faces, however, it is revealed that the party in public office has two potentially competing principals, since the party in central office interferes with the party on the ground. In this article as well, the results of the analysis produce a negative assessment of the analytical worth of the principal–agent model, based as it is on a monolithic notion of party that has proved to be inadequate; furthermore, it does not address the role or legitimacy of parties, as did Ignazi's article and to a lesser extent van Biezen and Kopecky's.

Zsolt Enyedi's – The Discreet Charm of Political Parties – engages in the discussion of competing explanations of party decline: the dealignment theory and the cartel party model. As such, his analysis, like Richard Katz's, takes into consideration both potential party anchors: society as the locus of factors that ultimately cause electoral dealignment (seen here as one of the causes of party decline); the state as the target of party cartelization, but also as the agency responsible for party detachment from society and the consequent loss of legitimacy of parties. The comparison between the two theories does not produce a clear winner, and nor was it meant to. Rather, it serves as a backdrop against which party trends in the post-Berlin Wall era can be empirically discussed and theoretically reflected upon. The article, albeit in a partially critical form, can also be seen as a tribute to Peter Mair's discussion of party decline, characterized as this was by a rather dramatic turnaround when Mair became a supporter of the decline thesis after having opposed it for many years.

Ingrid van Biezen and Thomas Poguntke's article – The Decline of Membership-based Politics – most directly addresses the problem of how parties relate to society. Their point of departure is the familiar analysis of party membership trends in a number of European countries. Not surprisingly, the data, with few exceptions, show an erosion of membership across countries and a continuing downward trend. Perhaps more importantly, the potential weakening of links between parties and society is also revealed by the similar downward trends experienced in religious orientations and in trade-union density (the latter observed in 22 OECD countries – wherein the inclusion in this sample of several non-European countries does not really affect the comparability of the data and the validity of the conclusions); this decline makes the maintenance of societal links through collateral/ancillary organizations or interest groups very problematic for parties. Overall, the analysis identifies, in its unambiguous conclusions about the waning of party–society links, the inadequacy of both the mass and the catch-all party models, which had been built on the said links. They seem to conclude that the main reason for this resides in the difficulties society, which they describe as horizontalized and see as increasingly formed by individuals who lack strong collective identities, has in accepting conventional (vertical and hierarchical) modes of politics. The answer may lie in alternative grassroots types of political mobilization and action centred on social networks and other electronic means.

Cas Mudde's article – Fighting the System? Populist Radical Right Parties and Party System Change – addresses the topic of party system change. As aforementioned, party systems undergo transformations as a result of a number of reasons, one being the emergence of new party types or families. The emergence of populist radical right parties can be seen as the result of a number of societal changes that have to do with objectively observable phenomena, such as migration and the resultant nationalist reactions, or have to do with more subtle value changes, such as authoritarian and populist responses to the perceived manifestation of societal decadence and corruption. The article surveys the ascent of parties that belong to the populist radical right in a number of European countries. The overall picture is one of a very successful new party family, the most successful in electoral terms since the end of WWII, in which some of its members, most notably the FPÖ in Austria and the LN in Italy, have participated in majority coalition governments in their respective countries. The most important and interesting finding, however, is that the relatively assertive emergence of this party family has not produced significant changes in party system structures and dynamics. This is not particularly surprising, since in some cases such parties were the ideological transformation of pre-existing moderate parties (such as the Austrian FPÖ) with scarce or no numerical consequences for the system, or were not sufficiently strong electorally to be objectively relevant and capable of changing the system's competitive structure.

Lastly, the article by Michael Alvarez, Ines Levin, Peter Mair and Alexander Trechsel – Party Preferences in the Digital Age: The Impact of Voting Advice Applications – explicitly introduces a theme that was implicitly addressed in a number of the other contributions in this issue: politics, specifically voting in this case, in the digital age. The inadequacy of parties in performing their functions, especially in representing their electors, has been made painfully visible as citizens voice their dissatisfaction through electronic media. The authors of this article reverse the picture by showing how societal demands can be translated into electoral choices with the help of online Voting Advice Applications (VAAs). Such devices provide potential voters with information about the congruence of their own positions on issues and those of political parties. The degree of overlap between the two sets of data gives a measure of the potential representativeness of the party supply in any given country, which, according to the data, is less than perfect. Another interesting finding concerns the relatively low number of electors that, among those who find there may be better alternatives to their initial choices, actually change their voting preferences as a result of the information provided. This may indicate a degree of residual party attachment, if not party identification, that is strong enough to resist external inputs such as VAAs, which admittedly are not yet very powerful. As VAAs become more popular and more widely accepted, however, this may change.

Conclusions

This issue's contributions address different aspects of party adaptation and change. As Peter Mair had rightfully indicated, adaptation has mostly concerned the organizations of individual parties, but it has also produced the emergence of new party families (Greens, Populist Radical Right) or party models. The overall consensus on the existence and persistence of a party crisis that has emerged so clearly in the articles indirectly confirms the conclusion reached thus far by the theoretical discussion on party models and organization. The common objective of organizational survival that determined the demise of the mass party and the temporary affirmation of the catch-all party, both of which, with different characteristics, presented the features of integrated party models, has produced as a final result the cartel party, the specific organizational prescriptions of which are far less precise and compelling. With the franchise party and the following discussion we can say that the era of the integrated party model, capable of shaping all parties in similar ways through unstoppable contagion, is over.

In fact, contemporary political parties seem to have only one characteristic in common: their evident separation from society. The organizational imperative that pushed parties towards the state also made them openly neglect their representational function. This neglect, however, may not necessarily be revealed by blatant inadequacies of the policy decisions parties make. Relatively low levels of electoral volatility and the inability of new parties to make significant inroads in party systems confirm that there may be at least a passive and/or partial acceptance by citizens of major policy decisions. This would not be a new phenomenon, as the collective representation and permissive consensus literature indicates. Katz and Mair (1995: 22) were themselves aware that if the parties became agents of the state, as the cartel model prescribes, democratic control would occur through retrospective voting, i.e. through the manifestation of electoral behaviour that expresses dissatisfaction with policy decisions. As most parties become part of the cartel, even this means for the representation of citizens is denied.

The organizational adjustments of individual parties or even their reforms seem in fact to be dictated more by the desire to find, often through stratarchy inspired solutions, balances in terms of access to resources or in terms of power relations among the party elites at various levels, and not to better respond to citizens' demands and expectations. This, as is evident from the contributions in this issue, is always at the expense of the party on the ground, as the autonomy often given to peripheral party structures privileges the local leaders and not the membership.

Several authors agree that citizen dissatisfaction is beginning to transpire. If, as noted above, voter frustration is not necessarily caused by their dislike for policies, it is certainly provoked by their dislike for the method used to develop such policies. Citizens increasingly ‘want to be heard’ and try to accomplish this through traditional methods, such as electoral abstention, participation in social movements, and by protesting and voicing their dissatisfaction in the social media. Parties are of course aware of the potential importance of the latter and try to engage citizens through their use. It might even be possible, by using the VAA algorithms, for parties to eventually adjust their positions to, and therefore better represent, their electorate's preferences. After all, the separation of parties from society is not per se a party objective; it is only perhaps an unwanted, if necessary, consequence of their organizational imperative in a given historical phase. Nevertheless, parties have not yet been replaced by viable substitutes capable of performing what have to date been traditionally party functions. Traditional, card-carrying membership-based parties are now a thing of the past and a transformation of that which remains of party organizations by resorting to digital technology may be the only solution.