Patronage as Party Resource - Selective Incentives and the Generation of Organizational Support -

Patronage as Party Resource - Selective Incentives and the Generation of Organizational Support - Nicole Bolleyer European University Institute Via dei Roccettini 9 50016 San Domenico (FI), Italy Nicole.Bolleyer@iue.it Paper Prepared for the 3 rd ECPR Joint Session in Nicosia, Cyprus 25-30 April 2006 Workshop: Political Parties and Patronage Directors: Petr Kopecký & Peter Mair Draft Version Please do not cite without the permission of the author! Recent research on party democracy argues that parties have begun to establish tight linkages to the state, while functioning less and less as voluntary associations for citizen representation. The increase of patronage observed in several European systems can be read as one symptom of this overall phenomenon. In order to conceptualize patronage as a resource which parties use for their organizational self-maintenance in modern party democracy, the paper puts forward the following arguments: The readiness of parties to use patronage needs to be understood as a resort to one type of selective incentive to assure individual loyalty where the capacity to do so through collective incentives declines due to several mutually reinforcing processes. Since political engagement has become a matter of choice more than of belonging, parties need to assure loyalty on the basis of selective benefits. Citizens who are politically active for ideational reasons increasingly use channels other than parties which has implications for the characteristics of newly incoming party members, a process supported by the changing recruitment styles of party elites. While as a consequence of vote-maximizing strategies party-voter linkages become less, party-member linkages become more particularistic leading to a growing demand for party patronage. An increasing demand is one part of the story, the problems to handle patronage as a resource to satisfy these demands is another. The second part of the paper looks at the defining features of patronage exchanges and the challenges springing from them when patronage goods are used to maintain organizational support and leads to the formulation of a set of hypotheses on how the use of patronage might develop over time and might impact on party organization. 1

I. Introduction: Membership Decline and the Rise of Party Patronage In recent party research, the role of membership for the maintenance of parties in Western democracy is the subject of an on-going debate. More and more frequently, works on party change point to the decreasing need and the increasing costs for party leaders in modern democracies of recruiting and retaining members (Blyth/Katz 2005; Katz/Mair 1995; Mair 1997). Other studies on party membership draw a more mixed picture. Scarrow (1996), for instance, points out that the motives of party leaders to maintain a stable membership might have changed, away from the provision of funds towards the legitimation of their party as an organization, while refusing the claim that membership has become a liability rational party elites will try to free themselves from. In Germany and Great Britain Scarrow observes changes in the mode of recruitment, while leaders still try to maintain a membership basis. The findings on the development of party membership, in contrast, speak a fairly clear language. There is an observable trend of membership decline across Western democracies over time (Mair 1997; Mair/van Biezen 2001). Furthermore, studies on party membership have shown that not only membership decreases in number but changes qualitatively. On the one hand, members age due to the considerable recruitment problems among young citizens, on the other, the members remaining in the party become less active (Cross/Young 2004; Gallagher/Marsh 2004; Hooghe et al 2004; Pedersen et al. 2004; Salgie/Heidar 2004; Sundberg 2003). The consequence that the bulk of passive members shrinks and with it the amount of membership fees they contribute is problematic. Still, parties seem to be able to cope with this development by acquiring access to state funding (Katz/Mair 1995; van Biezen 2000; 2001; 2004; Sundberg 2003). At the same time, however, the number of active members shrinks as well - those core members who are essential to maintain the party on the ground as an organization (Panebianco 1988: 26). As Seyd and Whiteley point out, in bad electoral times the existence of a core of supporters is essential to a party s survival and recovery (2004: 360). Furthermore, the goal of organizational self-maintenance presupposes the smooth replacement of leaders and candidates over time again presupposing that a pool of office aspirants is recruited in the first place. To control this process is not only in the interest of the party elites who want to maintain their own position in the party machinery but also adds to the outside credibility of parties able to deliver programs, policies and personnel in the long run. Thus party elites need an organization which helps to channel, monitor and preselect potential successors and keeps them waiting without threatening the current leadership. 2

To do so, party workers (of which future party elites are a part) are needed to uphold the basic organizational infrastructure. To motivate them to continuous work quite naturally creates costs for the party leadership. However, it is argued that these costs are still lower than the costs created by, to mention the most radical solution, the introduction of primaries which would externalize the selection problem and reduce parties capacity to steer recruitment considerably. This is the case first, because party elites would give away one core resource, the control over candidate selection and with it an important instrument to discipline actual party MPs and party members hoping for future candidacy. This would lead to a considerable loss of control when coordinating government and parliamentary MPs. Second, with the increasing professionalization of politics, the expectations towards candidates in terms of skills as well as the burden of political life are increasing. As one consequence, office aspirants are expected to rise slowly through the ranks to acquire professional experience (King 1981: 277) which functions as a mechanism of ex ante screening of candidates used by party leaders (Müller 2000a). Again, the advantage of such intra-organizational career paths is the one of control. Potential candidates have to move up the internal hierarchy before having access to valuable positions in terms of money and status, while current elites profit from aspirants long-term engagement and gather information about the latter s qualification. Starting from the claim that modern party elites still need the party on the ground - although rather a core nucleus of qualified activists than a broad mass of passive members - this paper discusses, first, the growing challenges for party elites to reconcile electoral and organizational goals and, second, in how far patronage as selective incentive can help to cope with these challenges. Selective in contrast to collective incentives gain value because they are attributed to only a sub-set of members and are not accessible through simply entering the organization (Wilson 1971: ch. 6). Based on this very basic distinction, the paper argues that while abandoning the party on the ground as Blyth and Katz (2005) hypothesize, is not a feasible option, the capacity to maintain it through collective goods declines. To understand this development, the arguments pointing at leaders increasing efforts to gain programmatic flexibility for vote-maximizing purposes is only one side of the coin (Mair 1997), the increasing difficulties to attract members through collective benefits due to societal transformation is the other. As electoral and organizational challenges are intertwined, so are parties responses to them. To the extent that political leaders become non-discriminating service providers for the electorate as a whole, they weaken their capacity to provide 3

collective benefits to followers. While parties relations to voters and to members (both belonging to the same social segment) were complementary in the past, more recently, linkages to these two groups diversify: While party-voter linkages become less, party-member linkages become more particularistic. With this, strategies to maintain both are increasingly difficult to reconcile. Arguing that material incentives in contrast to purposive incentives - provide the greatest flexibility to set and to seek organizational goals as well as to choose the instruments to achieve them (Schlesinger 1965: 768), an increasing relevance of the former is a likely outcome. As Sundberg showed in his analysis of party organization in Scandinavia, while mass party membership has eroded, parties invest more material resources to educate and reward their active membership holding public office than ever before (2003: 149-50). Simultaneously, the capacity to generate a stable ideological affiliation wanes. Hence, while the loss of passive party members might have a limited impact on parties as organizations due to increasing affiliation to the state, they get under pressure on a smaller scale in having to compensate for the declining impact of purposive incentives through selective incentives to those core members essential to maintain the party infrastructure. This problem is all the more essential given that programmatic appeals more often induce activists to enter a particular party and become active than help to sustain this activity and take over additional burdens (Conway/Feigert 1965: 1172). This said, party elites strategies to provide selective incentives to their core members and the trade-offs involved in doing so when simultaneously providing different types of incentives (collective- selective; material-immaterial) become a core issue for the debate of party change and transformation as triggered by the debate around the cartel party model (Katz/Mair 1995; see for a critique Kitschelt 2000a). Since the bonds linking members constitutes important variables for assessing a party s interrelationship with its environment (Crotty 1970: 276), it is worthwhile to approach this debate from the side of party-member relations since elites use of particular incentives and member motivations are naturally linked. The use of patronage represents one strategy available to parties to satisfy the demands for selective incentives. Empirically, the increase in party patronage observed in several European democracies of fairly heterogenous institutional make-up (Blondel 2002; Müller 2000b) can be read as one symptom of these more fundamental transformations of Western democracies. While party elites resort to patronage - along with their increasing dependency on state funding or their deliberate dissociation from affiliated organizations - has been 4

identified as one indication of party change (Mair 1997: 142-3), its intra-organizational triggers, its interplay with party elites electoral strategies and its societal roots have not been spelled out systematically. Consequently, while empirical work on patronage shows that patronage is used to compensate party activists (Müller 2000b: 157), this paper goes one step back and discusses the organizational implications of the changing supply of members as a result of societal changes on the one hand and due to party elites changing relations to voters and members on the other. Further, it asks in what way the professionalization of parties and the demands of the electoral game changes leaders demands for members of a particular type shifting focus from ideologically motivated activists directed towards purposive benefits to pragmatic professionals directed towards selective ones. The main argument of this first major section is that these trends are complementary in reinforcing the pressure on party elites to provide selective benefits. While patronage itself represents a rather neglected topic in party theory, it evidently touches upon very old question of how organizations cannot only be formed but maintained based on the provision of different sets of incentives (Olson 1965; Wilson 1973; Panebianco 1988). At the same time, the impressive literature on party membership, organization and change allows to embed this specific phenomenon into a wider context. This is all the more necessary since existing approaches on party patronage tend to conceptualize patronage as the use of government resources to obtain votes or to maintain support in general (Key 1936; Shefter 1977; 1994; Warner 1997; Manow 2002) irrespective of the very different roles of active and passive members, of voters and supporters. By focusing instead on the increasingly demanding and therefore less stable relationship between core members and party elites in face of changing party-voter relations, it becomes comprehensible why party elites resort more frequently to patronage as an instrument to assure intra-organizational loyalty. On the basis of this general claim, in a second major section the challenges of using patronage are discussed. There are three conditions to efficiently use patronage as resource or, more precisely, as selective incentive for stimulating membership activity. These conditions are the access to patronage resources that cannot be provided exclusively within organizational boundaries, the control of patronage resources which refers to the control over the distribution of spoils and finally the feedback of using patronage to be expected from voters and from non-rewarded or ideologically motivated members. All three problem dimensions refer to organizational challenges within parties and to the institutional context in which they 5

try to pursue their goals since the type of incentives used intra-organizationally has implications not only for the internal functioning of the party itself but for its relationship to the political system and of society of which it is part (Conway/Feigart 1968: 1060). Since existing approaches on party patronage centre around the question of access and of feedback, the second section of the paper focuses on the problem of control over distribution. Analytically speaking, in the first part the incentives for parties as organizations to provide selective benefits to members will be discussed as a response to societal changes which impact upon parties in Western industrial democracies similarly. It delivers arguments why party elites should increasingly resort to selective incentives such as patronage in party democracy in general. In the second part, we move down to the meso and micro level and shift from the implications of systemic factors to the implications of patronage exchanges themselves when trying to generate member support. Ironically, while the first part stresses the need for more selective benefits to generate organizational stability, the second part stresses the difficulties to do so through patronage whenever parties do not deliberatively counteract the tendencies of patronage exchanges prone to generate instability in member support in the long run. Finally, the last part returns to the system level where the paper started out. It makes a first step towards a necessary specification of the overall claim that there is more pressure to use selective incentives and discusses how this pressure might comparatively vary across majoritarian and consensus democracies. II. Patronage as Party Resource: Defining Patronage as Selective Incentive Most generally, party patronage denotes particularistic and direct exchanges between party politicians or functionaries and recipients in which resources the party as patron controls are exchanged for some kind of support (Müller 2006: 189). While this paper focuses on members as recipients, likewise voters or party supporters outside the party organization can be involved in these exchanges. Patronage relations are distinguishable from programmatic linkages in that they are personalized. However, they can be directed towards different functions: While service patronage involves a non-simultaneous exchange of goods or positions against supporters loyalty, power patronage refers to the partisan distribution of crucial positions in the civil service to assure the access to patronage goods in the first place (Eschenburg 1961). In contrast to a distribution of posts on the basis of merit, hence, based on impersonal criteria and technical competence, patronage can flourish especially in professions, or in our case organizations, providing few impersonal mechanisms or objective 6

criteria of advancement and in which patronage goods can be monopolized by a group of superiors (Abercrombie/Hill 1976: 423-4). Beyond applying personal instead of generalist criteria for the selection of recipients, patronage denotes an unequal and non-simultaneous exchange crucially different from market-relations: A patron controlling a desired good has the choice to reward certain clients and exclude others which gives him a stronger position in the exchange. Hence, the client might provide support only in the expectation to be rewarded; vice versa, the patron distributes a good in the expectation that loyalty will be maintained in the future. Since both cannot be taken for granted patronage relations can be expected to be the less stable the less the patron controls the patronage good as is particularly true when discussing party patronage. Since patronage denotes a mode of exchange it is difficult to pin it down with regard to the type of good which is distributed in the course of these exchanges beyond the statement that such a good needs to be divisible and exclusive in order to gain value. Accordingly, patronage goods can range from appointments and posts in the civil service over positions in public-sector firms to pork-barrel legislation as common in the U.S. and micro-policies of distributive character which favour particular clients as has been a common practice in Italy (Cotta/Verzichelli 1996: 197). Facing this range of goods, this paper has a particular focus on the distribution of posts and appointments following a patronage-logic since it immediately affects the functioning of parties as organizations: moreover, if internal posts form part of the available patronage goods the nature of patronage practices has an immediate impact on how parties organize: Despite the usual use of patronage referring to the partisan penetration of bureaucracies and state institutions (e.g. Manow 2002; Warner 1997) and its usual separation from appointments of, for instance, cabinet posts (Müller 2000b: 141; Blondel 2000), patronage relations can be equally found within parties, since positions can be easily distributed on the basis of personal criteria to assure party workers on-going commitment and activity by rewarding their loyalty (Wilson 1971) as will be specified later. Furthermore, works on party democracy have shown that the distribution of appointments can supplement patronage practices (Blondel/Cotta 2000: 214). Especially government posts can be a resource for party elites to prolong and strengthen their career within the leadership (Cotta/Verzichelli 1996: 185). Strictly speaking, patronage 7

practices are easier to implement in the sphere of political or organizational posts since access to civil service can be restricted by a merit system resting on general performance criteria. 1 Within parties, as Hooghe et al point out, [n]etworks tend to be important for any kind of political recruitment, and so we might expect that those strongly integrated in the party fabric [ ] will have a better chance of acquiring leading political positions. (2004: 202). On the way towards these positions, patronage relations are likely to play a considerable role. Note that this does not indicate a structural affinity between patronage practices and parties with strong personalistic leaders since the latter tend to marginalize their party and prefer more immediate links with followers (Gordin 2002: 541-2). As argued later, from a theoretical viewpoint, exactly due to the personal link these exchanges rest upon, patronage practices seem to be more efficient in generating support within decentralized structures. It is important to note that patronage goods need not be material (Müller 2006: 189; 2000; for a contrasting conception see Warren 1997) but might include immaterial decorations. 2 In order to conceptualize patronage as party resource in more general terms, it can be characterized best as a selective incentive, including both selective solidary and selective material incentives. Selective in contrast to collective incentives gain value because they are attributed to only a sub-set of members and are not accessible through simply entering the organization. Collective incentives can be either material or purposive, while collective purposive benefits in particular are generated by a member s affiliation to a shared cause the organization stands for. Material incentives refer to money or equivalent resources being either collectively or selectively available. Selective solidary incentives refer to status and power related to organizational functions and positions, while collective solidary incentives are generated by positive personal relations among the members in general (Wilson 1971: ch. 6). Selective solidary and material incentives are often interconnected insofar as positions involving a certain status are considered stepping-stones to access higher positions also valuable in material terms. Evidently, the greatest conflict referring to these two types of incentives evolves around their distribution with solidary incentives being able to generate at 1 This restriction exists only in some countries, while even there promotion follows a partisan rationale given equal qualification of several candidates. 2 One example would be nominations for the British House of Lords as a compensation for financial contributions by party supporters. Süddeutsche Zeitung, Finanzkrise bei Labour, Partei will sich mit Verkauf ihrer alten Zentrale behelfen, 23. März 2006, p. 6. 8

least as severe conflicts as the distribution of material goods because it is often hard to compensate for personal rejection through side-payments (Wilson 1971: 36-45). To conceptualize the use of patronage as the handling of selective incentives to assure organizational loyalty allows specifying the difficulties to do so alongside with providing non-material collective benefits for the group of members as a whole be it related to policy or to the distribution of decision-making rights between passive members and activists. While Wilson assesses organizational challenges of parties which rest on one type of incentives material, solidary or purposive (1971: ch.6), regarding patronage, problems resulting from selectiveness itself are particularly crucial, just to mention the most fundamental one, the resulting inflationary tendency: the more selective benefits are provided, the less support can be generated on their basis. After having laid out the general perspective on patronage, section III points at several developments which allow to hypothesize that the pressure on party leaders to provide selective incentives among them patronage - is growing due to processes of selection and self-selection of active members resulting from changing member supply and demand. III. Member Demand and Supply: Patterns of Selection and Self-Selection In his critique of the - in his view - overly state-focused cartel party model introduced by Katz and Mair (1995), Kitschelt has emphasized the necessity to take a more society-centred perspective on parties role and position in order to discuss the relationship between parties and supporters, mainly voters, more adequately. In particular, he points at a rather increasing responsiveness of parties instead of a growing distance of parties from voters as claimed by Katz and Mair (Kitschelt 2000a: 162). However, this critique is based on the assumption that responsiveness to voters and distance from society cannot coexist, a claim which demands qualification. Given that parties responsiveness boils down to their programmatic assimilation responding to changes in the electorate in which patterns of social division play less and less a role, while issue- and personalized voting gain importance, this undifferentiated responsiveness is fully compatible with organizational dissociation. And this dissociation becomes visible through declining linkages to a broader base of supporters and changing party-member relations. In fact, the two processes can be seen as corresponding and mutually reinforcing. That responsiveness is no longer directed towards a particular social segment, with which party members and supporters are affiliated, decreases the collective benefit 9

generated by party membership and regular vote-support in material and in ideological terms. Simultaneously, parties loose the capacity to attract ideologically motivated members who have available to themselves a range of more specific channels outside party which are considered more adequate to articulate political positions. Being inside organizational boundaries looses value for ordinary members since the party as organization looses profile. It also increasingly frees voters to switch their party preference from one election to the other. Wilson saw the weakness of political organization in general in the need to be inclusive which debases the value of solidarity as one important collective benefit which is linked to organizations exclusivity (1971: 96). While in the European context parties roots in social milieus have long been a remedy to this problem, due to individualization processes, this remedy no longer functions. As Katz points out, under these conditions, membership becomes increasingly content free (2001: 289). If, as a simultaneous outcome, party membership increasingly follows a rational and individualist logic, intra-organizational relations need to become more particularistic and to be based to a stronger degree on particularistic incentives. Hence, broad responsiveness and weakening linkages to wider parts of society indeed correspond. The strategy of elites in some parties to give the broad member basis more extensive decision-making rights in order to enhance ordinary membership status can be regarded as an effort to increase the collective benefit of membership. However, it has the intra-organizational repercussion to reduce the incentives for activists the middle strata of the party whose status is devalued. An alternative interpretation points at the intention of the party leadership to enhance their own status to disempower the activists by empowering the broad membership basis, while at the same time being able to proclaim a democratization of intra-party relations (Katz/Mair 1995; Katz 1997). Facing a trade-off, it is not possible to decide which strategy is per se the most beneficial from the viewpoint of organizational stability. Yet one might argue referring to party characteristics is that smaller parties which have a smaller recruitment pool might be less inclined to reduce incentives for these core members because they have bigger problems to attract qualified personnel than bigger ones. Given that small parties tend to be more specialized in the core issues they programmatically stand for one might also argue that they are comparatively less affected by problems to generate collective benefits to the membership as a whole than big catch all parties. Interestingly, empirical studies on such reform endeavours more clearly point at the negative impacts on the middle ranks than at the positive effects on the broader membership likely to follow: A study that analyzes the attitudes of passive and active members as well as 10

of party delegates towards procedural decentralization of candidate selection within Norwegian parties shows that party delegates since they have most to lose do not favour internal democratization reducing their own privileges (Saglie/Heidar 2004: 401). Interestingly, ordinary members equally were not eager to enhance their status within the organization. Not only did they not support a shift of candidate selection away from the party congress to the basis, but preferred (in contrast to the higher ranks) the party to pay more attention to voters (Saglie/Heidar 2004: 397-9). Indeed, the smaller group of middle-rank members perceives the trade-off between increasing the weight of ordinary members and of voters and their own intra-organizational weakening. This, however, is not the case for the much bigger group of ordinary members who do neither gain nor lose a lot by such changes. Returning to the Kitschelt critique of the cartel party approach (2000a), the claim of growing distance to society implying that parties substitute linkages to society by linkages to the state (Mair/Katz 1995; Mair 1997) demands qualification as well. While indeed capturing a general tendency, it draws attention away from the increasingly diversified character of linkages to society which parties still try to maintain. Facing a rationalization of voting as well as of votemaximization strategies in Western democracies, a parallel rationalization of membership selection and self-selection is likely. This interpretation gains leverage when social changes and their impact on the supply side of membership is taken into consideration. While it is an overall trend that party members increasingly age and parties capacity to recruit young members declines, unconventional participation is on the rise. Participation becomes a conscious choice since standard social structural differences - compared to motivational interest, the role of agencies and social networks - loose importance to explain patterns of party membership on the individual level (Norris 2002: 133-34). Linking these changes in the predominant triggers for party membership to the growing incapacity of parties to provide solidary and purposive collective benefits, it comes as no surprise that citizens engaging in unconventional participation tend to be comparatively younger (Della Porta 2005; McAdam 1988; Norris 2002). Social movements, as one example for unconventional participation, tend to be more issue-specific and provide a stronger platform for identification than parties with an increasingly blurring profile do. Since modern societies become increasingly differentiated systems far beyond the segmentation along class lines, a high specialization of alternative voluntary organizations for citizen representation and their greater appeal to young purposedriven members with a high degree of enthusiasm comes as no surprise. Accordingly, single issue political movements have been considered as a response to the quest for broadening 11

participation (Kaase/Barnes 1979: 525), while parties are hardly able to reconcile the demands of the majority of passive citizens which they prioritise in order to win elections and an educated active minority which tends to resort to unconventional forms of participation (Kaase/Barnes 1979: 533-4). This does not mean that there is a simple trade-off between party membership and the engagement in alternative channels. Distinguishing four types of participants (excluding the inactives) and examining participation behaviour in five countries, Kaase and Marsh concluded that only the conformists and the protesters resorted to one type of participation only, (conventional and unconventional forms respectively), while reformists and activists used both (Kaase/Marsh 1979: 153-7). Also more recent work shows that overall participation in new social movements is positively correlated with party membership (Della Porta 2005: 25; Norris 2002: 211-12). However, given that active membership in any organization is timeconsuming, it is likely that citizens have to set priorities where to invest their time. Further, unconventional channels often profit from their opposition to state-related channels to which political parties up to now doubtlessly belong finding its clearest expression in the official exclusion of parties from movement initiatives (Della Porta 2005: 22). Overall, criticism of parties is dominant in current political and public discourse (Mair 2005) and there is evidence that political distrust prompts individuals to seek access to politics unconventional and nonpartisan means (Cain et al. 2003; Dalton 2004). In sum, one can conclude that the more programmatically open parties become, the less they can catch up with the societal trends towards individualization and an increasingly conscious choice of participatory channels which ex ante tend to privilege more specific organizations when it comes to the attraction of young and often highly educated activists. To link this to the issue of party-member relations, declining overall membership does not only lead to the downsizing of the group of core activists, parties less and less tend to be first choice for those who engage in civic activism. Correspondingly, studies of parties youth organizations indicate that they lose members more rapidly than parties in general (Hooghe et al 2004: 196). Hence, in face of an extension of active citizens participatory repertoire, parties tend to be on the loosing side. Parties have advantages on the side of selective benefits due to their high organizational development and their privileged access to state resources which they can exploit to compensate for their declining substantial or affective appeal. Collective material benefits related to party membership necessarily loose practical importance in increasingly affluent 12

societies. While parties are usually incapable to provide sufficiently high benefits to all of their members, the challenge is manageable regarding selective benefits to a clearly restricted group. Simultaneously, members are likely to respond to what is on offer and the dominant motivational drive of core members is likely to show a stronger orientation towards these types of incentives. Albeit it is certainly an oversimplification to claim that the group of active members is homogeneously driven by the same motives, it has been argued that ideologues among them are the least dominant the higher a party s chances of government participation (Kitschelt 1989: 47-8; 57). Since not only the established but new parties as well increasingly participate in government, the hypothesized internal shift towards pragmatist candidates gains plausibility. Given that the greater dependence of party elites on selective incentives is also reinforced by party leaders strategic choice in the electoral arena as argued before, we end up in a vicious circle in which growing distrust in parties increases voter volatility (Dalton 2005: 944) again increasing the pressure both to pursue vote-maximizing strategies and to establish more particularistic linkages inside. In how far party elites had or still have the chance to counteract these dynamics is questionable since they seem to express a wider process of social transformation in which not only party identification has eroded in nearly all major democracies irrespective of their institutional make-up (Dalton 2000) but individuals affiliations to societal groups in general. What is, however, crucial at this point is to be aware of increasingly contradictory demands of party-voter and party-member relations: While for instance the dissociation from affiliate organizations such as unions is considered an effort of party elites to gain more programmatic flexibility (Mair 1997), one can also argue that, by doing so, they give up resources stabilizing their organizational infrastructure resulting from the close link between the different organizational elites. Evidently, it is one thing to say that the pool of core members changes due to changed societal supply, it is again another thing to say that elites might demand a different type of member as potential aspirant for public office and higher positions in the party organization due to changes of politics as a profession. Increasing professionalization of politics has certainly raised expectations towards politicians in terms of expertise and communication skills which being already confronted with a shrinking pool of aspirants - narrows down the group of potential candidates. Moreover, corresponding to the hypothesized change in party-voter relations, recruitment patterns have been affected as well in that strong personal appearance coupled with programmatic flexibility and efficiency seems to gain weight over ideological 13

convictions. The rise of personalized campaigns is an expression of this trend. As far as the profile of the modern politician moves towards the profile of a manager of public funds with decreasing leeway of action due to domestic and international constraints, the stronger the need for pragmatists which is likely to impact upon the recruitment style by current elites. 3 To put it in Wilson s words, political elites have little interest in public policy but considerable interest in finding a candidate for the top of the ticket who will increase the chances of winning the election and at the same time be responsive to the rank and file s desire for a share in the proceeds of winning. (1973: 37). Arguing that such candidates are likely to have better chances to move up the internal party hierarchy, not only a bias against ideologues on behalf of current elites can be hypothesized, but the patterns of self-selection of potential party activists describe before is reinforced. The major goal of this section was to elaborate on several trade-off relations party elites face and have increasing difficulties to balance in modern democracies. Both on the supply side and the demand side of membership, factors tend to support processes of member selection and self-selection producing a shift in core member motivations towards more pragmaticindividualist away from purposive-collective orientations. Most fundamentally, parties increasingly less differentiated responsiveness to the broader electorate extends the collective benefit of a parties victory to a larger group of voters. Simultaneously this decreases the collective benefits of being the member of a particular party regarding material, solidary and purposive incentives alike. In particular the weakening of solidary and purposive incentives disadvantages the party compared to more issue-specific and less hierarchical channels for interest articulation and representation especially among young citizens. Vice versa, parties are more dependent on their pool of selective incentives they have broader access to than less established voluntary organizations. To increase the collective benefit for the broad member basis by decentralizing decision-making processes (as compared to mere supporters), reduces the selective benefits available to the middle ranks of the party which do most of the work. Assuming that core members are still needed, one arrives at the expectation that party elites might less readily resort to broadly based member ballots for candidate selection than proponents of the cartel theory expect (e.g. Katz 2001: 292). Finally, the changing professional profile is likely to favour the selection of pragmatists over ideologues. Along with membership research modern party member can be understood as one less emotionally 3 See for a discussion of the link between activists motivation and recruitment styles Kitschelt (1989: 49-53). 14

affiliated to a party but one who chooses to become active based on a cost-benefit calculation (Biehl 2004: 683). IV. The Use of Patronage as Resource IV.1 Problems of Access, Control and Feedback Facing this set of indications that the demand for selective incentives and with it patronage might increase in modern democracies, this section deals with the problems involved in using patronage as a resource to maintain the support of core members. Party elites need to cope with three major challenges: acquiring access to and control over patronage goods and the handling of negative feedback of patronage practices. Negative feedback refers to the reactions of party members and voters facing the distribution of patronage goods on the basis of personal loyalty instead of merit (Warner 1997; Kitschelt 2000a; 2000b). The latter is often difficult to pin down empirically because loyalty of a person towards party elites goes often hand in hand with his or her continuous work for the party. However, since valuable positions are less numerous than active members interested in them, in-depth analysis should allow to identify whether the final choice tends to be made in favour of personal protégés or not (see for instance Carty s analysis of the intra-partisan conflicts in the Canadian Liberal Party over Chrétien s succession in 2002 (Carty 2003)). There are likely to be two types of reactions, activists who are alienated because they reject the practice of patronage itself and those who feel rejected because they have not received anything themselves. If party elites are not vulnerable to electoral dissatisfaction as a result of party collusion, the threat of voters negative feedback does not provide a strong disincentive to refrain from patronage (Hopkins 2004). However, it will show later that the internal repercussions complicate the use of patronage to generate support to a considerable extent. The problem of access to patronage refers first to intra-organizational limitations to endlessly create new positions as rewards. Second, it refers to the availability of state resources which might be more or less limited across countries due to the development the bureaucracies in the respective country (Shefter 1994; Manow 2002). 4 The control of patronage resources is closely tied to access. Still, while access refers to the capacity to get hold of resources, control refers to the capacity to decide upon the mode of these resources 4 In order to account for practices of clientilism across countries and across time, Piattoni (2001: 19-24) applies the concept of opportunity costs and political opportunity structure inherent in the institutional set-up of different states (one crucial variable is the status of the bureaucracy). The broader concepts of access, control and feedback are preferred because by denote three core challenges related to the successful use of patronage. In particular, they allow including intra-organizational pressures that need to be considered simultaneously with the institutional framework in which parties make their strategic choices to understand the obstacles involved in patronage practices. 15

distribution (Müller 2000b: 154) and to the capacity to find an efficient mode of distribution which can assure organizational stability in the long run. In contrast to existing approaches of patronage which concentrate on the crucial question of access to state resources that can used for patronage or on the impact of negative feedback to these practices (Kitschelt 2000a; 2000b; Manow 2002; Warner 1997), the following discussion deals with the challenges to distributed patronage goods efficiently and points to the problems even when resources are continuously available. To discuss the use of patronage goods within party organizations links closely to the concept of the clientilist party in which party elites focus comparatively little in modes of interest aggregation and program formation and, instead, create bonds to their followers through direct and personal exchanges (Kitschelt 2000b: 849). While the following discussion profits considerably from Kitschelt s work on clientilist versus programmatic linkages between parties and voters, it deviates from his perspective in two respects: first, it narrows the focus down from keeping supporters satisfied (the perspective existing approaches on party patronage take as well) to motivating activists for their on-going work and, with that, to keeping a pool of potential office-aspirants ready as the core rationale of party elites investments in the party on the ground. Second, while the clientilist party centers around the provision of material selective benefits (Kitschelt 2000b: 849), discussing the use of selective incentives refers to both material and solidary incentives. For instance, the distribution of honours forms part of patronage practices which is plausible because social relationships directly shape intra-organizational dynamics. The attribution of recognition can function as an effective reward, at the same time, it can cause severe conflict if members feel unjustly excluded. More generally, the maintenance of the goodwill of a specific person is one form of solidarity incentive which is both specific in effect and widely employed in civic associations (Wilson 1973: 40) and directly corresponds with the personalised logic of patronage exchanges. Accordingly, in a study of Norwegian party members members and delegates likewise emphasize the relevance of personal networks over formal procedures (Salgie/Heidar 2004: 401). Including solidary incentives also makes clear that the orientation towards selective benefits may not be equated with members being materialists. However, in order to motivate activists to work more than passive members do, it seems a realistic claim that they demand something more exclusive for themselves than club goods or purposive satisfaction which is open to every ordinary member. This argument gains particular weight considering the 16

waning capacity to generate collective benefits parties face as argued before and the minor relevance of the necessarily limited material benefits parties can offer their membership as a whole in the context of increasingly wealthy societies. Studies on party activism show that a variety of selective incentives play a powerful role in motivating activism. Activists perceived impact on the provision of collective goods as middle level elite and expressive motives are important factors as well (Whiteley et al.1993: 90-1). However, although the two motives are not material, one can argue that at least the former is reinforced by activists privileged status over ordinary members. IV.2 The Problem of Controlling Patronage Exchanges Section II already pointed to the core aspects which characterize the handling of patronage goods: First, patronage goods need to remain selective. Second, they are provided by a nonsimultaneous exchange whose completion is insecure in that neither patron nor recipients can be sure that the desired good (loyalty or some kind of reward) will be provided. This is especially true since albeit the party elites as patron might control a desired good the dependency of the recipient on the respective good is limited (at least as long as patronage relations do not dominate the whole political and economic system). Simply speaking, members can exit parties and a party career is not the only option, especially for those members with considerable professional skills which parties attempt to attract. Third, the mode of distribution is based on personal criteria, which are hardly clear-cut and again add to the insecurity on the side of recipients. Finally, it is often the case that a considerable part of positions functioning as patronage goods cannot be simply re-distributed whenever it is convenient for the party elites in principle controlling distribution. Each of these characteristics inherent in the nature of such exchanges creates problems regarding the capacity to maintain organizational loyalty through complicating in particular the control over patronage distribution and therefore has direct organizational implications for parties heavily relying on patronage. Starting with the most fundamental problem of selectiveness: the inflationary nature of patronage goods has been frequently noted. Most generally speaking, the more widely patronage goods are provided, the less it is a reward to receive them from the individual point of view. To use patronage resources most efficiently, the major challenge is to find an equilibrium between the multiplication of rewards while keeping them effective by 17

maintaining sufficient exclusiveness. This problem becomes most visible when thinking of non-material decorations which most easily loose value when too many people receive them in particular since status of a post is more difficult to increase than a salary is raised. When allocating material rewards, one strategy to circumvent the problem can aim at increasing the value of individual rewards over time and to set up a differentiated incentive system where one post or rewards is a requirement to access a more valuable one. This, of course, presupposes an expansion of goods shifting the problem in part from a more efficient distribution to one of increased access. In order to increase the share of available incentives internally, one can expect a party organization to differentiate, a process which faces practical limitations (Panebiaco 1988). External state sources provide an alternative pool of patronage goods (Shefter 1994; Manow 2002) which becomes all the more important as soon as limits of internal differentiation are reached. Irrespective of particular access problems which are not addressed in the paper, an increasing reliance on selective benefits tends to trigger an increasing demand because it simultaneously devalues the collective benefits free for everyone due to the formers greater exclusivity. This is particularly true for collective solidary incentives based on group belonging which are difficult to reconcile with differentiating between members. If the increased use of selective incentives weakens the effect of collective incentives (in addition to the problematic devaluation of group membership through vote-maximizing tactics mentioned before), again the likely response is to increase the use of selective benefits by extending the types of resources distributed as patronage goods. Accordingly as Müller s analysis of eight European countries (2000b) shows that in countries in which patronage is most pervasive parties use a great variety of resources to satisfy their followers. The difficulty to find an equilibrium between a extended pool of patronage recipients without inflating resources excessively creates doubts whether patronage systems can be expected to be stable in the long run. Empirical cases do not provide a clear answer: while the Italian system broke down in the 1990s, the Belgium system survived. Remaining on the theoretical level, it is not only resource inflation which points at a possible stability problem of patronage systems due to the difficulty to find the right equilibrium of patronage provision. The flipside of the coin is the demoralizing impact of exclusion that party elites might want to reduce as far as possible. The inflation argument points at a middle solution as a response, in which exclusion is reduced by multiplying rewards without devaluing them excessively. Regarding those additionally included in the groups of the rewarded, which are motivated to continue their work, this strategy sounds 18