André Krouwel. Handbook of Party Politics. Editor: Richard S Katz & William Crotty. 2006. Sage Publishing.
More than a century of scholarly attention to political parties has resulted in a substantial number of party models. Yet, so far all these party typologies have not accumulated into a more general theory on the genesis, development and transformation of political parties. This is caused primarily by the fact that most of the party models are seriously biased. First, most party models were developed in the context of western Europe and the United States of America, resulting in a limited ‘travelling capacity’ of these conceptualizations (Sartori, 1984) even across the Atlantic (see Ware, this volume). Secondly, most party models are very unidimensional in their approach, oftentimes focusing heavily or even exclusively on organizational aspects. Duverger (1954: xv) even argued that ‘present-day parties are distinguished far less by their programme or the class of their members than by the nature of their organization. A party is a community with a particular structure. Modern parties are characterized primarily by their anatomy.’ An anatomist, however, does his work by dissecting corpses, while party observers usually analyse political parties that are alive and kicking or are even still in their infancy. The fact that numerous scholars observed the same political parties yet only focused on a specific element at a particular stage in its development has proliferated the number of party models dramatically. Moreover, analysing parties merely by their bodily structures neglects one of the first observations, namely that a party is ‘a body of men united, for promoting by their joint endeavours the national interest, upon some particular principle in which they are all agreed’ (Burke, 1975: 113). Apparently not only organizational structures guide the behaviour of party members, but also some principle, some common goal, perspective or ideology. In addition, political parties perform many functions: they form the link between the state and civil society as they recruit and select the elite, nominate candidates for public office, form the executive or the (parliamentary) opposition to the incumbent power-holders and mobilize the people through political campaigns. Clearly, all these aspects also have to be included in party models and theories if we want to understand what a political party is, what it does and to what extent parties have transformed over time.
It is problematic that, when multiple dimensions have been used in modelling parties, often the organizational dimension is privileged over others and that additional aspects included in these typologies of parties generally refer to widely varying and inconsistent features (Gunther and Diamond, 2003). Another consequence of the large number of party models is the very low level of conceptual and terminological clarity and precision. In addition, proposed typologies are often neither mutually exclusive nor totally exhaustive. Furthermore, most of the proposed models of party do not include clear empirical indicators that would allow us to determine which parties actually do fall into each of the categories or when they have transformed into a different type (see Krouwel, 1999, 2003). Thus, we lack an effective way to classify different types of parties and consensus over indicators to determine what types of party we are observing.
Classifying and Linking Party Models
In the literature of political science basically three methods of party classification have been proposed and used. The first method is to simply list the party types and enumerate the major characteristics of each of the different models. Katz and Mair (1995: 18), for example, distinguish four party models (elite, mass, catch-all and cartel party) and then list 13 aspects on which these types of party differ. As a second method, some scholars identify ‘genera’ of party types and subsequently chart all the party types that have developed from each genus. An example of this method is Seiler (1984a, 1984b, 1993), who departs from Duverger’s distinction between the internal and external origin of parties and from these two genera groups eight party types into their respective lineages. Gunther and Diamond (2003), to take another example, develop five genera on the basis of which they classify 15 species of party. A third method of classification is based on more abstract dimensions along which parties differ. Wolinetz (2002: 161), for instance, uses the dimensions of vote-seeking, policy-seeking and office-seeking to position six party types in a triangular space on the basis of their primary goal. Pomper (1992) positions eight party types on three dimensions (breadth of focus, goal orientation and functional mode).
Although there is undoubtedly a certain path-dependency in the development of political parties, the genera method is too deterministic. Moreover, it is almost impossible to develop indisputable and consistent genera and there is no generally accepted method to determine in what lineage the different party models should be grouped. The deductive method of positioning parties along abstract dimensions is also problematic as no generally accepted indicators for each of the dimensions are currently available, so the position of each party type along the various dimensions becomes quite arbitrary.
Therefore I opt for the most parsimonious and straightforward method of differentiating parties on the basis of several crucial distinguishing characteristics. Not all party models that have been proposed are totally unique. Among the proposed models there is substantial similarity and overlap, and numerous party types that have been suggested are merely reformulations of an already existing model. On the basis of their similarities in focus and crucial features I have clustered the numerous party types into five basic species.
Many authors writing about the first modern parties that emerged in the late 19th century before the introduction of mass suffrage use various concepts basically to refer to the same phenomenon: loosely structured elite-centred cadre parties led by prominent individuals, organized in closed and local caucuses which have minimal organization outside parliament. Because of the significant overlap in characteristics I have grouped all models that refer to these first modern parties into the first cluster.
The second cluster comprises all models of mass parties. Wolinetz (2002: 146) argues that Panebianco’s mass bureacratic party is basically equivalent to Duverger’s mass party and Neumann’s party of mass integration (see also Gunther and Diamond, 2003: 179). The defining elements of this type to which numerous authors refer are: the extra-parliamentary mass mobilization of politically excluded social groups on the basis of well-articulated organizational structures and ideologies.
The third species of party is the electoralist, catch-all party type. Panebianco’s professional-electoral party is basically a respecification of Kirchheimer’s catch-all model (see Wolinetz, 2002: 146; Katz, 1996: 118; Gunther and Diamond, 2003: 185), while the rational-efficient party model proposed by Wright (1971) basically describes the same phenomenon (Katz, 1996: 118). Catch-all parties originate from mass parties that have professionalized their party organization and downgraded their ideological profile in order to appeal to a wider electorate than their original class or religious social base.
A fourth species is the cartel party. The formation of a so-called ‘state-party cartel’ was described by Kirchheimer (1954b), long before Lehmbruch (1974: 97), Lijphart (1968, 1974: 76), or Katz and Mair (1995) proposed their later versions of cartel democracy (see Krouwel, 2003). Basically this party type is characterized by a fusion of the party in public office with several interest groups that form a political cartel, which is mainly oriented towards the maintenance of executive power. It is a professional organization that is largely dependent on the state for its survival and has slowly retreated from civil society reducing its function mainly to governing.
The final cluster of party types that can be distinguished is of quite recent origin. Business firm types of party originate from the private initiative of a political entrepreneur and have, by and large, the structures of a commercial company. The image of the party leader, combined with some popular issues, is marketed by a professional organization to an ever more volatile electoral market.
As a second step, in an attempt at cumulative theory-building, I will sequentially link the five generic party models. The main reason for this is that these five clusters of party models are not isolated and unconnected species. As Lipset and Rokkan (1967: 50) argued, most of the party organizations are far older than the majority of the electorates they represent. This means that, at least in part, observers from different times have been observing and describing the same political parties in subsequent stages of their development. Since most of the models are derived from these empirical observations of the same phenomena in different periods, linking them chronologically also provides an historical overview of major party characteristics culminating in a general theory of party transformation over the last century.
Mass parties emerged as a result of the political exclusion of large proportions of citizens by the dominant elite and their cadre parties of the proto-democracies of the late 19th and early 20th centuries. Kirchheimer (1954b, 1966) departed from Neumann’s concept of the mass integration party and argued that, after the political integration of their followers had been successfully completed, these mass parties were transforming into catch-all parties in the late 1950s and early 1960s. Mass parties slowly professionalized their organizations, moderated their demands for social and political transformation and began to appeal to voters outside their original core electorate. As their party programmes became increasingly interchangeable and cooperation between former political enemies became the norm, rather than the exception, a political cartel was formed that became increasingly impenetrable for new political actors and groups. Cartel parties slowly monopolize the resources of the state and create a legal environment that favours the incumbent parties and discriminates against new competitors. As a reaction to this exclusion, political entrepreneurs who have no access to the resources of the state use the resources and strategies of the private sector, particularly the commercial mass media, to gain access to the electoral arena and executive power. As this brief chronology shows, the five models in sequence provide a tool to assess party transformation over time.
In a similar vein, Katz and Mair (1995: 6) framed the development of political parties as a dialectical process, in which each new party type generates a reaction that will lead to a new party model and a further chain of reactions. They identified different party models within distinctive time periods on the basis of the relationship between political parties, civil society and the state (Katz and Mair, 1995: 12-18). Clearly, party transformation is an ongoing evolutionary process in which parties adapt to their particular social and political context. This is also why the models of party are sequentially interconnected: observers build on existing models or reformulate an earlier model when they perceive that these models are no longer applicable to current political parties. The main concern for a comprehensive theory of party transformation then becomes to identify the specific characteristics that make the models of parties mutually exclusive. Below I propose a number of indicators that can be used to differentiate the party models from one another.
The Organizational, Electoral and Ideological Dimensions of Party Models
Since existing typologies and models of political parties usually have been developed in a specific political and social context on the basis of a limited number of observations, the models vary substantially in their focus and level of sophistication. As argued above, most models focus on organizational aspects; often the level of centralization or federalization is taken as the basic feature (Lenin, 1961; Michels, 1962; Eldersveld, 1964, 1982; Kitschelt, 1994), along with territorial penetration and diffusion (Eliassen and Svåsand, 1975). Organizational forms such as the caucus (Ostrogorski, 1902), branch, cell, militia (Duverger, 1964), nucleus (Schlesinger, 1965, 1984) or cadre (Duverger, 1954; Koole, 1996) are also used to distinguish between party types. Others have proposed to define parties on the basis of the level of professionalization, bureaucratization, institutionalization and rational efficiency of the party organization (Wright, 1971; Downs, 1957; Panebianco, 1988) or their collusion with the state (Kirchheimer, 1954b; Katz and Mair, 1993, 1995). Party models also refer to the main functions of the party organization, for example the selection of candidates (Bryce, 1929; Schumpeter, 1942) or their representational and integrational functions. An example of the latter is the distinction between ‘parties of individual representation,’ ‘parties of democratic integration’ and ‘parties of total integration’ (Neumann, 1956). Duverger’s famous distinction between the internal and external origin of parties also needs to be included in this enumeration of possible organizational classification schemes.
Some models include sociological or electoral characteristics such as the representation of social groups in terms of class, religion or ethnicity (Duverger, 1954; Kirchheimer, 1954b). Party models such as mass parties, elite parties and amateur parties (Wright, 1971) are classified by the class nature of their membership, the most active or dominant social group within the party, the level of rank-and-file participation or the type of leadership (Weber, 1925; Neumann, 1956; Kirchheimer, 1954a, 1966; Wildavsky 1959). Other party models, such as catch-all parties or ethnic parties, are typified by the width of their electoral appeal (Kirchheimer, 1966).
Concerning party classifications on the basis of ideology, Weber’s typology of Weltanschauungs or Glaubensparteien is often cited (Weber, 1925), while political scientists also frequently use ideological labels for parties such as right-wing, left-wing, extremist, protest, populist or fundamentalist. In grouping parties cross-nationally into party families, generally ideological labels such as conservative, liberal, Christian democratic, social democratic, socialist, communist, Green or environmental are used. Combining ideology with sociological aspects has resulted in party typologies such as ‘radical mass parties’ and ‘clientelistic parties’ (Rueschemeyer et al., 1992).
Party models should not be too reductionist, by emphasizing only a single dimension of political parties. Instead, parties should be regarded as complex phenomena with multiple attributes or properties that constitute one ‘bounded whole,’ and jointly constitute a pure or ideal type from which real political parties will deviate to varying degrees (Sartori, 1987: 182-5). Since there is no consensus as to which attribute or dimension should be privileged over others, I have opted for a broad range of analysis that is better able to capture the existing variation among different types of parties. This broad analysis includes first of all the genetic origin as a basic criterion guiding the classification of the different party types. The origin of parties determines to a large extent their initial format and their subsequent transformation is path-dependent on these foundational elements (Panebianco, 1988). In addition, I include three other dimensions to which earlier models refer: electoral, ideological and organizational. On the electoral dimension, the five party models can be distinguished on the basis of their electoral appeal and social support as well as the social origin of the elite they recruit. The ideological dimension comprises both the basis for party competition and the extent of inter-party competition. On the organizational dimension, the generic types are differentiated by examining the importance and status of the membership organization and the position of the parliamentary party and party in public office. The relative power balance between these three ‘faces of a political party’ is different within each of the five models. In addition, parties can be differentiated on the basis of two other organizational features: the structure of the resources that are available to the party and the type of political campaigning in which they engage. In this section I will discuss each of these nine characteristic features for each of the five party models.
The Elite Party Model
One of the first scholars to describe a political party was Edmund Burke, who, writing in 1770, defined a party as a group of parliamentary representatives who agreed to cooperate upon a certain principle (Burke, 1975). These first political parties emerged in proto-democratic systems with suffrage limited to a small privileged class of the more propertied male population. An extra-parliamentary party organization was practically non-existent and the coordination between its members, a small elite from the middle and upper classes, was loosely structured. Wolinetz (2002: 140) describes this type of party as closed caucuses of prominent individuals. Distinguishing between internally and externally created parties, Duverger (1954) characterized these first parties by their emanation from groups of parliamentary representatives (see also Kirchheimer, 1954b). According to Duverger, these internally created parties are commonly led by a small cadre of individuals with high socioeconomic status, who have only weak links with their electorate. Clearly the defining sociological characteristic of elite parties is the high status of their members, who already had obtained politically powerful positions before the advent of an extra-parliamentary party organization. The emergence of these ‘modern’ extra-parliamentary parties, under the influence of the extension of the suffrage, was analysed by Mosei Ostrogorski (1902). He compared these organizations in Britain and the United States and, with the latter having a more extended electorate, concluded that power became increasingly concentrated in local party ‘machines’ that aimed at winning elections through an extensive system of patronage and clientelism.
At the organizational level, elite parties have basically two layers: in the constituencies and in parliament (Ostrogorski, 1902: VIII-IX; Katz and Mair, 2002: 114). The extra-parliamentary party is weakly articulated or even absent, and each constituency is able to provide its own resources so that central authority and control are weak. Katz and Mair (2002: 115) argue that the elite party is basically an agglomeration of local parties consisting of ‘a small core of individuals with independent and personal access to resources able to place either one of their number or their surrogate in Parliament as their representative’ (see also Ostrogorski, 1902: i). Such a picture of the elite party is also sketched by Duverger (1954: 1-2, 62-7) who characterized the caucus party by its local and embryonic organizational structures that were exclusively aimed at recruiting candidates and campaigning for them during the election period. In a similar vein, Neumann (1956) identified the earliest political parties as parties of individual representation, which are characteristic of a society with a restricted political domain and only a limited degree of participation. They articulate the demands of specific social groups and their ‘membership activity is, for all practical purposes, limited to balloting, and the party organization (if existent at all) is dormant between election periods. Its main function is the selection of representatives, who, once chosen, are possessed of an absolute “free mandate” and are in every respect responsible only to their own consciences’ (Neumann, 1956: 404).
Not much is said by the various authors on the ideological character of elite parties. What can be assessed is that, although the different groups of parliamentarians may have held ‘widely diverging views’ of what the national interest was (Katz, 1996: 116), competition between parties was relatively limited. Since all parties consisted of members of the higher echelons of society and only represented a limited section of the population, political conflict centred on the extent of unification and centralization of the state, the level of local autonomy and the level of state intervention in the economic process (primarily taxes and tariffs).
The Mass Party Model
Whereas political power preceded the formation of the elite party, the mass party is the mirror image of the latter in that the formation of the party organization precedes the acquisition of power. Typically, mass parties are externally created and mobilize broad segments of the electorate previously excluded from the political process (Duverger, 1954; Kirchheimer, 1966). These parties have been typified by Neumann (1956) as parties of social integration, as they seek to integrate these excluded social groups into the body politic. Since they aim at a radical redistribution of social, economic and political power, these parties demand a strong commitment from their members, encapsulating them into an extensive party organization that provides a wide range of services via a dense network of ancillary organizations. In the words of Neumann (1956: 404):
Modern parties have steadily enlarged their scope and power within the political community and have consequently changed their own functions and character. In place of a party of individual representation, our contemporary society increasingly shows a party of social integration. It demands not only permanent dues-paying membership (which may be found to a smaller extent within the loose party of representation too) but, above all, an increasing influence over all the spheres of the individual’s daily life.
The extra-parliamentary origin, in addition to the fact that mass parties represent and mobilize a particular and clearly defined social, religious or ethnic segment of society, influences their ideological and organizational character. In order to organize a politically excluded group, the mass party needs a coherent vision of a better and different world that has to be communicated in a compelling manner. As Panebianco (1988: 264) pointed out, the stress is on ideology, and ‘believers’ play a central role within the organization. Paradoxically, these ‘parties of the excluded’ attempt to integrate their followers by insulating them from possible counter-pressures (Katz, 1996: 118). This insulation is achieved by a distinct ideology that is ingrained in the minds of the members through propaganda, the party press and party-organized activities in all spheres of life (Neumann, 1956: 405). Ancillary organizations were created in the field of education, labour, housing, sports, banking, insurance and so on, so that all social, economic and cultural activities were consistent with the ideology. The ideological vision of a better world becomes visible and materializes within this social niche. Needless to say, the ideologies of these mass parties differ from the already powerful groups, but they also differ from various ideologies of other mass parties. The result is fierce and principled competition among parties. Among mass parties themselves there is substantial variance in ideology and (consequently) in organization.
Duverger (1954: 63-71) distinguishes between branch-based mass parties and cell-based devotee parties, the latter being more totalitarian in ideology and organization. This distinction is also found in Neumann, who separates the party of social integration from the party of total integration. A party of total integration is ‘all inclusive’ and ‘demands the citizen’s total surrender. It denies not only the relative freedom of choice among the voters and followers but also any possibility of coalition and compromise among parties. It can perceive nothing but total seizure and exercise of power, undisputed acceptance of the party line, and monolithic rule’ (Neumann, 1956: 405). Lenin (1961: 464-5) describes such a party as a small and cohesive party of professional and totally committed revolutionaries that lead huge masses of uncritical followers.
The mass party can also be found in a religious variant, the denominational mass party (Kirchheimer, 1957a: 437, 1966), which Kirchheimer differentiated from the totalitarian party and the democratic mass party (Kirchheimer, 1954b). Both the denominational and the democratic mass party try to appeal to a maximum of voters to take over the administration and carry into effect a definite programme (Kirchheimer, 1954b). They are, however, still limited in their appeal and only aim to mobilize a specific social class or religious group. According to Gunther and Diamond (2003: 180-3), the mass party can also be found in nationalistic and fundamentalist variants, which are more proto-hegemonic in their ideology and tend towards the militia type of organization.
In terms of organization, all mass parties share the characteristic of extensive and centralized bureaucracy at the national level. The democratic variants of the mass parties are characterized by an elected and representative collegial leadership, often combined with formal powers for a national congress with representatives of the membership (Wolinetz, 2002: 146). Formally, mass parties are democratic organizations, but the ideological rigidity and the internal processes of training and recruiting members of the elite (through extensive socialization in the local branches and the internal educational system) make real competitive intra-party elections unlikely. Observing one of the first mass parties, Michels (1962) noted the bureaucratic rationalization within mass parties in which a small and unrepresentative elite gains control over the resources and means of communication. Michels thought that in any large organization power-concentration into the hands of an oligarchy is inevitable.
It is organization which gives birth to the dominion of the elected over the electors, of the mandataries over the mandators, of the delegates over the delegators. Who says organization, says oligarchy. (Michels, 1962: 365)
Inevitable or not, mass parties are hierarchical in their structure as all activities of the ancillary organizations and the local party branches are coordinated by the extra-parliamentary leadership. In contrast to the elite parties where local caucuses voluntarily form a national organization, the central office of the mass party has a top-down approach. Local branches and cells are founded in order to increase the level of penetration of the party. Characteristic of mass party development is the establishment of an extra-parliamentary office that precedes the formation of a party in public office. As a consequence, the party in public office is controlled, disciplined and supervised by the extra-parliamentary leadership as all representatives are considered to have the same mandate (Katz, 1996: 118). The party in public office is simply instrumental to the implementation of the party’s ideology (Katz and Mair, 2002: 118). These strong vertical organizational ties (Panebianco, 1988: 264) are needed to amass and pool resources at the central level of the extra-parliamentary party (Katz and Mair, 2002: 117). The mass party derives its name from the mass of members that form the core of the organization. Membership levels and the extent of involvement and participation of members in inner-party activities and electoral campaigning are part of the defining characteristics of mass parties (see Ware, 1985, 1987, 1996). Beyond the voluntary work members are expected to do for the party, they are also the main source of income. Membership fees are used to finance the central bureaucracy and the campaigning activities of the mass party. Other sources of income for mass parties derive from the activities of the ancillary organizations and their own party press.
Electoralist Catch-all Parties
Mass parties in Europe have been very successful in integrating their followers in the body politic and in replacing their ancillary organizations with full-blown welfare states at the national level. Coupled with high levels of economic growth, the maturation of welfare states resulted in the emergence of a substantial new middle class made up of skilled manual workers, white-collar workers and civil servants. Their interests converged and became indistinguishable from those of the old middle classes. According to Kirchheimer, this diminished social polarization went hand in hand with diminished political polarization as the doctrines of mass parties slowly became interchangeable. Mass parties gradually transformed into ideologically bland catch-all parties, and this process culminated in a waning of principled opposition and a reduction of politics to the mere management of the state (for a comprehensive version of Kirchheimer’s theory of party transformation, see Krouwel, 2003). Kirchheimer distinguished the catch-all party from the Weltanschauungs-party and argued that the modern catch-all party was now forced to think more in terms of profit and loss of electoral support and policy (Wolinetz, 2002: 145-6). He asserted that political parties had been reduced ‘to a rationally conceived vehicle of interest representation’ (Kirchheimer, 1957b: 314-15). Although catch-all parties still functioned as intermediaries between elements of formerly united groups, the working class accepted these parties only because they promised to give priority to their material claims, not because of their social vision. Catch-all parties were reluctant to perform the role of opposition, as this would seriously diminish their success in realizing group claims. This transition from the ideologically orientated mass party to the interest-group-oriented catch-all party is indicative of the erosion of principled opposition.
Kirchheimer’s development of the catch-all thesis is a good example of how erratic theory-generating processes are concerning party transformation. Kirchheimer formulated his catch-all thesis on the basis of only a limited number of observations, in particular the Italian Democrazia Cristiana, the German Sozialdemokratische Partei Deutschlands, the British Labour Party, the French Union pour la Nouvelle République and the German Christlich-Demokratische Union (Kirchheimer, 1966). He hypothesized that the catch-all development witnessed in these cases was likely to be prevalent in many countries in Western Europe and led to a more or less generalized transformation of party systems. Kirchheimer was also fairly categorical in identifying the properties of this new party -including its ideological, organizational and electoral dimensions – which is why there still remains substantial confusion in the contemporary literature regarding precisely what a catch-all party is and precisely which parties can genuinely be regarded as catch-all (see Dittrich, 1983; Wolinetz, 1979, 1991, 2002; Schmidt, 1985, 1989; Smith, 1989; Krouwel, 1999).
As early as 1954, in an analysis of the West German political system, Kirchheimer (1954a: 317-18) first introduced the concept of the catch-all party. Over a period of at least 12 years the somewhat loosely specified notion of the catch-all party was continuously altered (Kirchheimer, 1957a: 437, 1957b: 314, 1959: 270, 274; 1961: 256; 1966: 185). In none of his essays does Kirchheimer develop an exact definition of this new type of political party and at no time did he ever provide a clear and coherent set of indicators as to what precisely constituted a catch-all party. Confusingly the catch-all party is sometimes referred to as the ‘catch-all people’s party’ (Kirchheimer, 1966: 190), at other times as the ‘catch-all mass party’ (Kirchheimer, 1954a: 250, 1966: 191), the ‘conservative catch-all party’ (Kirchheimer, 1954a: 250), the ‘Christian type of catch-all people’s parties’ (Kirchheimer, 1959: 270) and, in still another version, as the ‘personal loyalty variant of the catch-all party’ (Kirchheimer, 1966: 187, n. 12). Indeed, 12 years after its first introduction, Kirchheimer (1966: 190) had still only formulated a very cursory definition of the catch-all transformation, a process which he then conceived as involving five related elements:
a) drastic reduction of the party’s ideological baggage. b) Further strengthening of top leadership groups, whose actions and omissions are now judged from the viewpoint of their contribution to the efficiency of the entire social system rather than identification with the goals of their particular organisation. c) Downgrading of the role of the individual party member, a role considered a historical relic which may obscure the newly built-up catch-all party image. d) De-emphasis of the class-gardée, specific social-class or denominational clientele, in favour of recruiting voters among the population at large. e) Securing access to a variety of interest groups for financial and electoral reasons.
Yet earlier versions list different characteristics as the key features of catch-all development (1964b; 1965). Kirchheimer (1964a: 16) included a feature dealing with the extra-parliamentary party and argued that the change towards catchallism involves: ‘Further development of a party bureaucratic apparatus committed to organizational success without regard to ideological consistency.’ In later versions, this element is formulated more generally, now referring to the relative power of the entire party leadership while dropping the idea that catch-all parties will develop more elaborate bureaucratic apparatuses (Kirchheimer, 1966: 190). Over the years, substantive alterations were also made in Kirchheimer’s argumentation as to what factors influence the catch-all development in different European countries. At various stages Kirchheimer added arguments about the particular social structures that determine the success of a catch-all strategy, as well as an explanation as to why only major parties in the larger European countries could hope to appeal to wider electoral clienteles. Kirchheimer also reformulated his thesis with respect to the expressive and the aggregative function. First, he argued that the expressive function migrated from parties to other political institutions, while this claim is later reformulated in that catch-all parties continue to function as expressive institutions but are limited by widely felt popular concerns. Another late addition to his theory is that the loose-fitting structure of the catch-all party and its disconnection from society will considerably limit its scope for political action.
On the basis of Kirchheimer’s entire oeuvre, his personal archive of unpublished papers, his lecture notes as well as the references he cites with the various elements of the catch-all thesis, it is possible to reconstruct Kirchheimer’s original ideas (Krouwel, 1999; 2003). Thus, for example, concerning party transformation at the organizational level, Kirchheimer (1966: 190) cites Lohmar (1963: 35-47, 117-24), Pizzorno (1964: 199, 217) and Lipset (1964: 276). These references suggest that Kirchheimer regarded the downgrading of the role of party members as a multifaceted process, including a stagnation in the size of party memberships, a loss of attendance at party meetings and of readership of party newspapers, a transformation towards a more balanced social profile, and a reduced importance of membership fees in overall party revenue. Additionally, the role of active party members with regard to the selection of the party leadership is also in decline, which erodes the members’ function as mediators between the electorate and the political leadership. Party leaders are co-opted into the leadership group on the basis of their technical and managerial qualities rather than because of their ideological orientation or class origin. Moreover, with reference to Duverger, Kirchheimer (1966: 178, 182, 193, 199; 1954b: 246, 259) also argues that citizens are increasingly excluded from political participation, in that catch-all parties offer less and less opportunity for membership activity, particularly as they disconnect themselves from formerly affiliated organizations. Catch-all party organizations become increasingly professional and capital-intensive, and depend increasingly on state subsidies and interest-group contributions for their income, and on the commercial mass media for their communication needs (see also Panebianco, 1988: 264-6). This political professionalization, in which experts and managers with specialized tasks replace the old party bureaucracy, is also emphasized in Panebianco’s (1988: 222-35) model of the electoral-professional party. Catch-all parties also use their connection with interest groups as a source of policy ideas (in the absence of a coherent and independent policy platform) and implement policy proposals originating from organized interests in exchange for financial resources and electoral support.
On the ideology of catch-all parties, Kirchheimer (1962: 3, 1966: 195) assumed that catch-all parties will adopt similar policy positions in the centre of the political spectrum and that they will emphasize similar issues. Concerning this centripetal political competition, Kirchheimer refers to Lipset (1964) and Duverger (1964), who argue that most major parties make a trans-class appeal, with programmes spearheaded by a commitment to collective bargaining and moderate political and socioeconomic changes. Parties on both the left and the right had amicably resolved the class conflict in an acceptance of social democratic ideology, since rightist parties had accepted the welfare state and economic planning and leftist parties had moderated their ideas for revision of capitalism. Alternation in cabinet composition no longer leads to a change in government policies. All political parties and their leaders cooperate closely with one another, thus leaving little room for political opposition. With reference to the Downsian ‘multi-policy party,’ essentially equivalent to Kirchheimer’s catch-all concept, it is suggested that catch-all parties sacrificed their former ideological position and the interests of their core electorate in order to maximize their electoral appeal (see also Mintzel, 1984: 66). Parties, however, are limited by the fact that voters will not vote if all parties stress totally identical programmes and will therefore compete with candidates and remnants of traditional loyalties, reducing politics to individual personalities. This pre-eminence of the public representatives of the party, personalized leadership and candidate-centred campaigns are also crucial characteristics of the electoral-professional party of Panebianco (1988: 266).
On the third and crucial electoral dimension, which gives the catch-all party its name, Kirchheimer argued that catch-all parties attempt to bridge the (already declining) socio-economic and cultural cleavages among the electorate in order to attract a broader ‘audience’ (Kirchheimer, 1966: 184). This wider electoral ‘catchment’ of parties transformed the European mass parties into American-style catch-all parties that appeal to all social classes (Kirchheimer, n.d.: 27). Denominational mass parties were transforming into interdenominational catch-all parties, appealing to all voters except convinced anti-clericals, and social democratic parties were attracting voters far beyond the core working-class supporters. In sum, a catch-all party is characterized by an indistinct ideological profile, a wide electoral appeal aimed at vote maximization, a loose connection with the electorate, a power balance in favour of the party elite vis-ă-vis the party members and a professional and capital-intensive organization (Krouwel, 1999: 59).
In the United States, Eldersveld (1964, 1982) and Schlesinger (1965, 1984) had also pointed towards parties that became primarily oriented towards the recruitment and selection of candidates for public office and organizing election campaigns. The representation and mobilization of specific social groups in the United States is also organized through professional interest organizations that contribute, financially or otherwise, to the election campaigns of individual politicians. Eldersveld (1964) sketches a picture of local candidate organizations that function almost autonomously without substantive coordination or support from a national party organization. He called it the stratarchy party model: parties with limited levels of formal organization and high autonomy. Parties have a ‘porous nature’ and easily absorb anyone willing to work for them, run as a candidate or support them with a donation or vote. The party is merely an alliance of coalitions at the various levels (substructures) with little or no hierarchy. Similarly, Schlesinger (1965, 1984) describes parties basically as local candidate organizations: a nucleus mainly devoted to capturing public office. All party activities are specifically linked to an individual candidate and the different nuclei of the same party can even be in competition with each other for resources and votes. Nuclei have no members, only contributors of all sorts – in financial terms, in time spent on campaigning or by voting for a candidate. All these models stress the autonomy of political actors, but in Europe observers see an opposite development towards more state-dependent parties.
Partisan States: The Cartel Party Model
Analysing the functional transformation of parties, Kirchheimer (1954b, 1957b) identified several types of political collusion. The first is an inter-party cartel of centrist catch-all parties that try to maintain their power position in public office. As a result of the disappearance of a goal-oriented opposition, combined with consensus on most important policy issues, genuine political competition is almost completely eliminated. The combination of vanishing political opposition with a shift of power from parliament to the executive resulted in a firm inter-party cartel, from which political competitors, particularly more radical parties, were increasingly excluded. A second type of collusion is the formation of a state-party cartel, where parties disconnect themselves from their social foundations and become amalgamated with the state, reducing politics to mere ‘state management’ by professional politicians (Kirchheimer, 1954b, 1957b). This extensive collusion of political parties with the state and the severing of the societal links of party organizations evidence a power shift from parliament to political parties. Kirchheimer alleged that the parliamentary party and the central party organization became highly interwoven at the personal level, resulting in an ever growing discipline of the parliamentary party. A third type of collusion, closely related to the catch-all development, is the tripartite power cartel consisting of political parties, the state and powerful interest groups. According to Kirchheimer, political parties try to ‘close the electoral market’ by seeking the loyalty of large groups of voters not on the basis of their ideology but through their interest organizations. Parties are increasingly subsidized by interest groups, which are also their main channels of communication with the electorate. At the same time, the party on the ground is neglected and parties display an increasing aloofness towards civil society. Finally, Kirchheimer predicted further collusion between the executive, the leadership of the major political parties and the judicial powers (the courts), indicating an ongoing process of diffusion of state powers.
These distinctions by Kirchheimer are useful when we look at later versions of the cartel thesis. The most widely cited is Katz and Mair’s (1995) cartel party thesis, in which the cartel is defined in terms of a state-party cartel: ‘colluding parties [that] become agents of the state and employ the resources of the state [the party state] to ensure their own collective survival’ (Katz and Mair, 1995: 5). To ensure this collective organizational survival, parties allocate substantial state support to themselves and regulate the activities of parties through the state. This state-party collusion is a reciprocal process in which, on the one hand, parties increasingly extract state resources and ‘colonize’ the institutions of the state and, on the other, the state increasingly regulates party political organizations and activities through law (Katz, 1996; Krouwel, 2003). Colonization of the state is evidenced by the fact that political parties become increasingly dependent on the state, allocating state resources to their organizations while disengaging from their former resources within civil society. Within this oligopolistic cartel, a vast portion of the state’s resources and institutional assets is accrued in the hands of the elites of the major parties. Politicians make increasing use of public institutions such as ministerial bureaucracies (to which they appoint spokesmen, media and policy advisors) and other state agencies and public utilities or quasi nongovernmental organizations (quangos) and the state-owned media for party-political purposes and electoral campaigning. What seems to be occurring is a symbiosis between political parties and the state, a weakening of the democratically crucial institutional differentiation of civil associations and formal state institutions. The state becomes ‘partisan’ as political elites weld party organizations and state institutions together to such an extent that citizens can no longer distinguish between them. While party organizations are formally considered as part of civil society in most constitutions, in reality parties are ‘colonizing’ the state through extensive processes of patronage and overlapping functional linkages. More evidence of this development can be found in the fact that politicians often simultaneously perform formal functions within political parties as well as formal roles in the state (civil servant or minister). This symbiosis of a supposedly ‘neutral’ state bureaucracy and a professional political class is advanced as in most European countries political recruitment has to a large extent been narrowed to the state-employed civil servants. As Puhle (2002) has pointed out, this structural proximity and overlapping of state institutions and party organizations leads to serious democratic problems, as political parties cease to be ‘intermediary’ and ‘representative,’ and also can lead to more patronage, clientelism and corruption.
Through increased formal regulation of party activities, established political parties seek to monopolize the route to executive office. In order to ensure these privileges, party elites obviously prefer to have them enshrined in law. Although political competition cannot be totally eliminated, cartel parties attempt to block competition from political ‘outsiders’ by using legal means to their political advantage. Both processes of state dependency and ‘self-regulation’ increase and intensify the reciprocal linkages between political parties and institutions of the state, colluding into a ‘partisan state’ (Krouwel, 2004).
Later specifications of the cartel thesis by Katz and Mair also include an argument concerning inter-party collusion. Cartel parties are seen to limit and carefully manage the level of inter-party competition through informal agreements and by sharing office. The cartel is largely implicit and entails the gradual inclusion of all significant parties in government. The range of acceptable coalitions is widened and the politics of opposition is abandoned (Katz, 1996: 119-21; Mair, 1997: 137-9; Katz and Mair, 2002: 124). This common goal has transformed apparent incentives to compete into a positive motivation not to compete (Katz and Mair, 1995: 19-20). Outside challengers are not formally excluded from electoral competition by the allocation of disproportionate state resources to the incumbent parties, they are simply excluded from executive office as long as possible and can only enter the cartel through absorption and adaptation (Katz and Mair, 1996: 531). Inter-party collusion creates its own opposition. Exclusion from executive power offers challengers ammunition to mobilize against the cartel parties (Katz and Mair, 1995: 24). Favourable conditions for the development of party cartels are a tradition of strong state-party relations, patronage and a political culture of inter-party cooperation.
In sum, what distinguishes cartel parties is that,
in contrast to more entrepreneurially oriented catch-all parties, cartel parties appeal to an even broader or more diffuse electorate, engage primarily in capital-intensive campaigns, emphasise their managerial skills and efficiency, are loosely organised, and remote from their members. Even more important, rather than competing in order to win and bidding for support wherever it can be found, cartel parties are content to ensure their access to the state by sharing powers with others. (Wolinetz, 2002: 148)
At the organizational level the relation of the cartel party to the state is central as the state provides the institutional environment and the resources by which cartel parties can retreat from society. Long periods in government transform the internal structure and power balance within parties as they enhance the status of the party in public office (Katz and Mair, 2002: 124). State resources are progressively accumulated by the parliamentary party and the party in public office becomes increasingly independent from the membership party on the ground and its central office (Katz and Mair, 2002: 123). The organization of the cartel party becomes characterized by a stratarchical relation between the various levels of the party: both the local office-holders and the central party are to a certain extent autonomous (Katz and Mair, 1995: 21).
A second feature is increasing professionalization, accumulation of financial and human resources in terms of staff at the parliamentary face of the party, eventually leading to a domination of the party in public office (Katz and Mair, 2002: 123). This domination is visible in an increasing presence of representatives of the party in public office appointed to the party central office (Katz and Mair, 1993). Concerning ideology, competition focuses increasingly on the managerial skills, competence and efficiency of the party in public office (Wolinetz, 2002: 148). In response to criticism by Koole (1996: 517) that it was not clear what this ‘toning down of competition’ exactly entails, Katz and Mair argued that this has to be seen as convergence of parties on the left-right scale, an expansion of coalition combinations and the increasingly circumscribed scope of policy innovation. Cartel parties display high levels of ‘symbolic competition’ (Katz and Mair, 1996: 530). Not much is said about the width of the electoral appeal, but cartel parties seem to campaign for the support of diffuse groups of voters that have weak links – or none at all – to the party.
Politics Incorporated: The Business-Firm Party Model
The fifth species, the business-firm party is a recent phenomenon in Europe but not on the American continent (see Carty 2001). Basically there are two types: one is based on an already existing commercial company, whose structures are used for a political project, while the other type is a new and separate organization specially constructed for a political endeavour. Hopkin and Paolucci (1999: 320) describe Berlusconi’s Forza Italia as an example of the first type: ‘In Forza Italia the distinctions between analogy and reality are blurred: the “political entrepreneur” in question is in fact a businessman, and the organisation of the party is largely conditioned by the prior existence of a business firm.’ Hopkin and Paolucci (1999: 307) argue that business-firm parties will emerge when a new party system is created.
In terms of organization, the business-firm party generates its resources from the private sector, which differentiates it from the cartel parties that use state resources for their activities. Although business-firm parties may have (financial) support from interest groups, such groups are not their main source of income or electoral support, or their main channel of communication. This means that the extra-parliamentary party is practically useless and will not be developed on any meaningful scale. What might be developed is a mechanism for mobilizing sympathizers to appear at party conferences to cheer on the party leadership. In the words of Hopkin and Paolucci (1999: 315), business-firm parties will have only ‘a lightweight organisation with the sole basic function of mobilising short-term support at election time.’ The party on the ground will be limited to a minimum so it does not hamper the leadership in its attempt to break the mould of the party cartel. As the dues-paying membership will be small and most of the resources will be needed for campaigning purposes, most of the activities will not be assigned to party bureaucrats. ‘Party bureaucracies are kept to a bare minimum, with technical tasks often “contracted out” to external experts with no ties to the party’ (Hopkin and Paolucci, 1999: 333). This seems to be the essence of the business-firm party: all party activities and tasks are brought under formal (commercial) contract in terms of labour, services and goods to be delivered to the ‘party.’ This means that the only individuals that have a more permanent stake in the party are the ones that occupy the party in public office. ‘Grassroots membership is also limited, with a high proportion of party members being officeholders who see the party as a vehicle for acquiring political positions, rather than an end in itself (Hopkin and Paolucci, 1999: 333). As the party and its ideology are no longer goals in themselves, the business-firm party, ‘instead of being a voluntary organisation with essentially social objectives, becomes a kind of “business firm”, in which the public goods produced are incidental to the real objectives of those leading it; in Olson’s terminology, policy is a “byproduct” ‘(Hopkin and Paolucci, 1999: 311). Business-firm parties will have a flexible ideological orientation and an eagerness to attract superficial support from broad sectors of society (Hopkin and Paolucci, 1999: 315), but, unlike the catch-all party, they are not oriented towards interest groups for their policy ideas. Policy positions will be developed as products within firms: demand-oriented on the basis of ‘market research’ with focus groups, survey research and local trials to test their feasibility and popularity. These ‘policy products’ need to be wrapped in the most attractive package and will be aggressively put into the market. This explains why what seems to characterize business-firm parties more than their predecessors is their almost total orientation to the creation of ‘free publicity’ or even direct control of the media. The best wrapping for these popular policies is an attractive candidate (or even a single leader) so that the marketing of the policies can be reduced to the promotion of individuals. Not surprisingly, those best trained for this mediatized political arena are individuals working in the entertainment sectors, which explains why an increasing number of people from this sector are now finding employment in politics. As Hopkin and Paolucci (1999: 322-3) argue: ‘characteristic of the leadership of the business firm party: personal popularity, organizational advantages, and crucially, access to unlimited professional expertise in mass communication.’ Needless to say, this extreme emphasis on the individual personality leads to vulnerability of business-firm parties as well as a high degree of centralization of power around the party leader (Hopkin and Paolucci, 1999: 323).
A Rudimentary Theory of Party Transformation
In sequence, these five clusters of party models, which were derived from a mixture of empirical observation and theoretical speculation, provide a comprehensive theory of party transformation consisting of ten developmental factors. In an effort to boil down the multi-dimensional complexity which characterizes the transformation of parties in modern European democracies, and to try to make sense of what is a multi-faceted phenomenon, this final section will draw on this multi-dimensionality and sequentiality of the various party models to suggest that the ten factors can be combined into four key dimensions through which the character of parties may best be understood. The first of these is associated with the genetic origin of parties, the second dimension relates to the electoral appeal and elite recruitment of parties, the third dimension is ideological and refers to the basis and extent of party competition, while the fourth is concerned with the organizational character of parties (the balance of power between the three ‘faces’ of the party, their resource structure and type of campaigning). These four offer a more readily grasped summary of the complexity that was revealed in the description of the party models.
Changes in the Genesis of Political Parties
The basic distinguishing feature of the five party types is their genetic origin. The party models suggest two axes along which the origin of parties can be positioned: first, their proximity to state institutions or origin from civil society; and, second, the agent that initiates the party foundation, that is, an individual enterprise versus a collective initiative.
Elite or cadre parties originated from the initiative of individual parliamentary representatives of local constituencies who felt the need for more coordination of their parliamentary work and, with the emergence of the mass party, for their campaign efforts. In contradistinction, the mass party originated directly from civil society, usually emerging from a collective effort to mobilize politically excluded social groups. This extra-parliamentary origin meant that the ‘party’ was first a social movement, often in the form of workers’ unions or religious organizations relatively distant from or even hostile to the state. Their primary goal was to change political institutions, achieve universal suffrage and other political rights such as freedom of organization and expression, as well as a more inclusive electoral system.
Mass parties were very successful in their attempts at democratization and as a result they gradually transformed into catch-all parties, as their party in public office increased its linkages with interest groups and abandoned its own attempts at mass mobilization outside election time. Thus, catch-all parties result from the merger of the party in public office of the former mass party with an interest-group organization, while simultaneously disconnecting itself from the party on the ground and civil society. A next stage in party development occurs when the party in public office dissociates itself more and more from interest groups and becomes amalgamated with state structures. The party in public office of these cartel parties comes to dominate the entire party structure because it taps into the resources of the state while societal resources (from the party on the ground and the interest groups) become irrelevant to its activities and survival. As a reaction to this colonization and monopolization of state resources, new competitors emanate from the individual initiative of political entrepreneurs that use private resources for their political project. These entrepreneurs use the organizational format of business companies to structure their organization as they go about the manufacturing of politics in a similar fashion to any other production process.
The Transformed Electoral Appeal of Parties
In terms of electoral appeal and support, the party models basically suggest a negative relationship between the social heterogeneity of party support and the strength of the party-voter link. Parties can opt for a broad electoral appeal, but this will coincide with weaker party-voter links, while parties with a narrower or class-distinctive social base will have supporters that are more strongly connected with ‘their’ party. The various models also refer to the sociological character of elite recruitment. At the elite level the models distinguish between parties that have an open system of elite recruitment, while in other parties the route to the top is centrally controlled and limited to ‘party apparatchiks.’
Elite parties had a very limited electoral appeal as the suffrage was extended only to the upper classes. With the extension of the suffrage, under pressure from the mass parties, elite parties had to widen their electoral appeal in order to compete with the mobilization of the class mass and religious mass parties. Initially, elite parties recruited their representatives from a small social niche of the upper social strata. While mass parties only appealed to their core electorate, they advocated and adopted a more open structure for elite recruitment, encompassing the middle classes, and even some members of the lower classes entered the political elite through the internal educational structures of the mass party. Electoralist catch-all parties have a broad appeal on both axes, broadening their appeal beyond that of the former mass parties, and also recruit their elite from a wide social spectrum, especially representatives from various interest groups. Similar broad patterns of elite access are found within entrepreneurial party types where each individual with a significant mobilizing potential is qualified to run on the party ticket and voters from all walks of life are welcomed. The cartel party, on the other hand, displays the most closed type of elite recruitment as incumbent parties seek to maintain their control of public office by narrowing the scope of elite recruitment. Control by the cartel over elite recruitment outside their own party organizations is attempted through legal and financial hurdles for potential competitors.
Fading Ideologies and Different Types of Party Competition
At the ideological level, the various party models differentiate between polarized and more moderate, pragmatic competition. Parties either compete on the basis of a coherent and principled political programme (as with the mass parties) or adopt a more flexible and strategic use of policies. The second axis differentiates parties oriented towards the representation of interests from parties oriented towards office control based on the promise of good governance by competent managers of the state.
Elite parties competed on the basis of the traditional status of their candidates, without too much emphasis on their ideological differences. Similarly, cartel parties cater to a fixed clientele that provides them with electoral support in exchange for favourable policies. Both the elites of the cadre parties and the cartel parties are primarily office-oriented almost regardless of the policies to be implemented, and present themselves as the ‘natural’ managers of the affairs of the state. Mass parties, on the other hand, were initially oriented towards the mobilization of a core electorate that they sought to represent in the state structures. The fact that mass parties each represented different social groups and competed against an incumbent elite augmented their emphasis on diverging and fundamental ideological visions of a better world. After the relative success of their mass mobilization, these parties transformed into more pragmatic and ideologically more flexible or even ideologically bland catch-all parties. Less focused on a coherent ideology and eventually also abandoning the representation of specific social groups, party competition was narrowed down to the managerial qualities of the leadership of the party in public office. Control of office has now become the main driving force of political actors and the incumbent cartel parties try to fend off political entrepreneurs who seek to replace the elites in office by campaigning on specific popular issues (not a coherent programme) and the attractiveness and competencies of the individual leaders of these business-firm parties.
Changing Power Structures and Organization of Political Parties
Since most emphasis is placed on the organization of parties in each of the models, the most complex array of changes can be seen at the organizational level. Most of the party models focus on the relative importance of the membership party the party on the ground, in relation to the party in public and central office. Other aspects that the models highlight are the income structure and the type of electoral campaigns that parties conduct. Nevertheless, this complex series of changes, described above, can be summarized in a two-dimensional model of the organizational transformation of political parties. First, the party models all refer to the internal power balance in terms of centralization of decision-making, whereby in some parties the leadership hierarchically controls and coordinates all party activities, while in other parties more horizontal, open and democratic structures dominate. Secondly, the models emphasize the difference between professional and capital-intensive party organizations and their more amateuristic predecessors that had a more voluntary character.
Over time, the party models show that political parties transformed from the amateuristic and temporary structures of the elite party, to a more permanent bureaucracy and an extensive extra-parliamentary membership organization in which volunteers performed a large number of tasks. The transformation into electoralist catch-all parties and cartel parties entailed a further process of professionalization and more capital-intensive organizational structures. Eventually the membership organization becomes almost redundant and is only seen as a pool for the recruitment of candidates. The party in central office, practically absent within the elite parties, becomes the core of the mass party from which all activities are initiated and coordinated. As mass parties come to occupy the executive more frequently and for long periods of time, power gradually shifts towards the party in public office. Slowly the party in public office comes to dominate the extra-parliamentary party, and this process is invigorated by the allocation of resources from the state that mainly accumulate in the parliamentary party. At the final stage, the party in central office is completely absorbed by the leaders of the party in public office. Within business-firm parties, capital and expertise are centralized with the party leadership to such an extent that a separate organization that could be considered a party central office cannot be detected. The resources of the elite party were basically the private wealth of each of the individual candidates, which gave them high levels of autonomy. Mass parties, on the contrary, had to accumulate their financial and human resources from the large number of followers and volunteers within the party organization. As the catch-all party tapped into the vast resources of interest groups and later, as cartel parties, the resources of the state, more professionalization and centralization in decision-making became feasible. Political entrepreneurs and their business-firm types of party seem to resemble the old elite parties with respect to their resource structure. Again private capital is used for a political project, although the capital may not be directly in the hands of the party leadership, but provided by commercial companies and media empires. In terms of political campaigning, the models show an enormous transformation of political parties. While the representatives of the elite parties could easily attempt to meet each and every voter personally, the extension of the electorate made this impossible for the mass party. A labour-intensive campaign had to be organized to convince and mobilize all of the voters from the core social group to vote for the party at election time. With substantial financial resources from interest groups (catch-all parties) or the state (cartel parties), political campaigns became more professional. Increasingly outside expertise is hired, first on a permanent basis but later in a more ad hoc, non-permanent fashion when election time approaches.
This overview of party models has shown that parties are complex multi-faceted creatures, and their patterns of transformation are neither unidirectional nor linear. What we observe is a multiplicity of features, some of which, indeed, appear to work in opposite directions to one another. Moreover, even with the broad electoral, organizational and ideological elements of parties, change, when it occurs, tends to both ebb and flow, and sometimes, even concurrently to run in contradictory directions. This attempt to bring all these elements into a more comprehensive theory of party transformation should be seen as a first step to try to make sense of the character and function of what is still one of the most crucial organizations in modern democracies.