2
Framing the Impact of Enlargement
1. Introduction
Drawing on the classic literature on party organization and ideology, Chapter 2 presents a theoretical framework to understand the impact of enlargement on the Europarties. It is here suggested that enlargement has a differentiated impact on the dimension of institutionalization: largely positive for their organizational development, but largely negative for their ideological cohesion and for party competition. It is further argued - on the basis of an analogy with national parties - that if the enlarged Europarties are now more similar to catch-all or cartel parties, then the pursuit of a 'representative' function in the EU political system, pace the Treaty of Lisbon, would be more difficult.
The feasibility of theorizing on the impact of a specific external event, such as Eastern enlargement, on the process of party institutionalization could be challenged. Prima facie, institutionalization would seem to require a longue duree, historical perspective. As Randall and Svasand put it (1999, p. 10): 'Time is implicit in a study about institutionalization.' Yet, as Welfling (1973, p. 13) argued years ago: '[institutionalization is] not only a process, but also a property or a state'. While an organization institutionalizes over time, it is also possible to capture a snapshot at two specific time points and compare the degree of institutionalization of the observed organization at f0 and t\.
A further caveat lies in the difficulty of separating the effects of enlargement from broader trends, both theoretically and empirically. Thus, it has been pointed out that 'distinguishing the specific contribution of enlargement is not a simple matter' (Best, Christiansen and Settembri, 2008; also Settembri, 2007), and that 'it is difficult methodologically to discern between enlargement impact and broader
40
Framing the Impact of Enlargement 41
integration dynamics' (Blavoukos and Pagoulatos, 2008, p. 1149). My objective here is to theoretically capture the specific effect of enlargement on Europarty institutionalization. Separating empirically enlargement from broader integration dynamics will be the task pursued in Part II of this book.
Finally, it is important to stress that my propositions are specifically framed with the aim of assessing the impact of the 2004/07 enlargement. However, there are no inherent features of the theories on partisan formation and development inhibiting a broader applicability. Consequently, the framework developed here can be potentially extended to future enlargements of the EU, in order to understand their implications for political parties in the EU.
This chapter proceeds as follows. Section 2 and Section 3 concentrate, respectively, on the impact of enlargement on the structural and value dimensions of institutionalization, while Section 4 focuses on competition in the EP party system. Section 5 proposes some additional factors accounting for the variation in the impact of enlargement between the Europarties. Finally, Section 6 places enlargement within a broader developmental trajectory, and concludes.
2. Enlargement and structural institutionalization
In the domestic literature on party institutionalization, it has been demonstrated that a party increase in size, measured by the number of rank-and-file members, leads to larger bureaucracies and strengthens the party leadership. Because individual membership is not (yet) possible in the Europarties, which are still parties of parties, their size could rather be measured by the number of national member parties. Building on the literature on party organizational change, I suggest here that the broad expansion in membership produced by enlargement (an 'external shock') triggers organizational change. Additionally, on the basis of a functionalist argument, I further specify the nature of this change, which would be reflected in the empowerment of the party leadership and an increase in the complexity of the party organization.
Size and party change
In the most classic studies of parties as organizations, growth in their size has been systematically linked with changes in their organizational structure. This relation was famously postulated by Robert Michels (1999 [1910]) in his study investigating the so-called 'iron law of oligarchy'. According to Michels, there is an 'iron law' in party organizational
42  Context and Theory
Framing the Impact of Enlargement 43
development: every party eventually succumbs to a more centralized structure and to oligarchic tendencies more generally. The broad implication of this argument is that democracy cannot be reached, and any party is bound to be controlled by restricted elites. Michels' argument has strong normative implications and is deeply rooted in his distrust of mass psychology, leading to the conclusion that the masses are bound to be dominated by leaders. However, the eventual formation of an 'oligarchy' is not due merely to psychology, but also (and most importantly here) to 'technical' and 'mechanical' reasons. As an organization expands in size, democratic control - that is, control by the rank-and-file members - becomes increasingly difficult, and delegation is necessary. Forums of direct democracy, such as assemblies, lose their effectiveness, and the decision-making system is bound to centralize.
Expansion in size, however, not only promotes party oligarchic tendencies, but entails other important effects: it promotes party differentiation, in terms of both organs and functions, and it demands more formalization. As Panebianco (1988, p. 183) neatly sums up: 'growth in size is correlated with growth in internal division of labour, multiplication of hierarchical levels, and bureaucratic development'.
In Duverger's Political Parties (1967 [1951]), the relation between growth in size and organizational reform is explained again in both psychological and mechanical terms. Centralization of power (in his words: 'authoritarianism') is needed to guarantee an organization's 'efficiency' (1967, p. 171). Furthermore, the extension of the franchise and the entry of the masses on the political stage required parties to devise new organizational structures. The socialist branch was the organizational response to the need to actively include the masses in politics. The parties - with the socialist one leading the other party families (1967, pp. 24-25) -had to set up local branches, improve the vertical coordination of their central and local organs and strengthen their bureaucracies.
Hence, both Michels and Duverger postulate a causal relation between size and party organizational change. Furthermore, Duverger specifically links a change in the environment (the extension of the electoral franchise) to broad-ranging transformations in how parties are organized. In this regard, their work can be related to more contemporary theories of party change, which could be usefully applied to the EU enlargement(s).
External shocks and party institutionalization
Michels and Duverger could be taken as ideal-typically representing two positions in the current debate on party organizational change
(in general, see Harmel, 2002). Michels can be catalogued as a supporter of a 'life cycle' approach, according to which parties are bound to adopt certain organizational features in their institutional life. Change is continuous and incremental. In contrast, Duverger posits that a specific change in the environment leads to party organizational adaptation. Extension of the franchise leads parties to redefine their structures accordingly. Change is abrupt and linked to alterations in the environment (a 'discrete' approach).
The importance of environmental factors in explaining changes in party organization has been emphasized by contemporary scholarship. Thus, while accepting that parties may also change under other circumstances, it has been argued that 'the most dramatic and broadest change will occur only when the party has experienced an external shock' (Harmel and Janda, 1994, p. 265). In their theory of party change, Harmel and Janda (1994) argue that the primary source of change is when political parties experience a 'shock' impacting upon their primary goal. Thus, for those parties whose main objective is to maximize votes, an electoral defeat ('external shock') is likely to be a fundamental source of change. Although their definition of change does not encompass only party organization, but rather 'any variation, alteration or modification in how parties are organized, what human and material resources they can draw upon, what they stand for and what they do' (p. 275), their framework has been widely applied in organizational studies (e.g. Harmel, 2002; Mueller, 1997). In the abstract, the sequence of events in their model can be described as follows: first, an external event, such as a constitutional change, an electoral defeat or the rise of a new competitor, takes place. Second, the 'external event' impacts upon the primary goal, leading the party decision-makers to re-evaluate the party effectiveness and performance. Third, organizational reforms are implemented in order to adapt to the new situation.
If their model is applied to the Eastern enlargement of the EU, the sequence of events can be described as follows. To start with, enlargement takes place and produces a large expansion in size in the major Europarties: in the EPP group, 40 per cent of the national party delegations and 24 per cent of the MEPs were from the 'new' Europe; the number of national party delegations in the PES group grew by more than a third; the ALDE increased by about 34 per cent in the number of national party delegations and 22 per cent in the number of MEPs. In contrast, increase in size was negligible for some of the smaller political groups: the G-EFA included only one member from Eastern Europe, 'For Human Rights in a United Latvia', while the GUE-NGL
44  Context and Theory
Expansion in size
+ Organization
+ Autonomy
Figure 2.1   Enlargement and structural institutionalization
also recruited only one new Eastern member, the Communist Party of Bohemia and Moravia (see figures in Chapter 1). The magnitude of the effect of enlargement would obviously vary as a function of the number of countries and parties involved.
As the new members enter the political groups, the expansion in size affects their performance. The organizational rules which guaranteed effectiveness in the pre-enlargement context come under strain after enlargement. The huge increase in the number of delegations increases the transaction costs of bargaining and negotiating to reach agreements. With more members, the range of preferences represented within the party is also increasing (see below, also Hix et al, 2007). It becomes more difficult to monitor members' compliance with majority decisions, and the costs of ensuring cohesion grow (in general, see Heritier, 2007, pp. 11-22). Consequently, decision-makers need to redesign the rules of their organization, so as to deal with the new situation and preserve the organization's capacity to perform. More specifically, the imperative to take decisions and ensure that members comply with them should lead both to a re-balancing of power towards the party leadership, with more effective coordination and sanctioning tools, and to an increase in the complexity of the organization by, for instance, setting up working groups and introducing additional rules (Figure 2.1). These organizational reforms can either anticipate the expected 'shock', or follow it within a limited time. The alternative path to action and reform might lead to institutional failure.
3. Enlargement and value institutionalization
Chapter 1, introducing the value dimension of institutionalization, quoted Randall and Svasand (1999, p. 12): 'parties are not only formal instrumental organizations that potential supporters regard like any type of supermarket, but purposeful actors in which the participants
Framing the Impact of Enlargement 45
share an ideology and identify with the values of the organization'. If 'supporters' are neither voters nor rank-and-file members, but rather the national member parties, Randall and Svasand's argument can be extended to political parties at the EU level. Hence, what is required for institutionalized Europarties on the value dimension and, specifically, for cohesive Europarties is that 'the member parties share an ideology and identify with the values of the organization'. Building on a 'party family' approach, I suggest here that the inclusion of the Central and Eastern members would not neatly fit the (West European-based) Europarties. Additionally, the structure of incentives for political parties in the EP, and in the EU more generally, facilitates the aggregation of ideologically heterogeneous parties, which would combine together in 'marriages of convenience'.
Party families at the EU level
The Europarties and the EU party system are based respectively upon the parties and the party systems of the EU member states. As Schmitt and Thomassen (1999) remarked, their 'feasibility' is due to the similarity of the major lines of political conflict within the member states. If major divisions coincide with the national borders, or are different in each and every member state, it would be impossible to conceive of transnational parties and party system.
The similarity of the major lines of political conflict in West European countries has been famously described by Lipset and Rokkan (1967) in their theory of cleavages. The social cleavages - between agriculture and labour, the Church and the state, the centre and the periphery and the industrial workers and the owners of the capital - had been similarly activated by competing political parties, grouped in different party families according to the social groups they represent and their basic ideological principles, throughout Western Europe. Of the four cleavages they originally identified, the class cleavage proved to be of particular importance. Indeed, the left-right dimension, traditionally capturing the opposition between state intervention in the economy and free market, still represents the foremost reference for political parties in West European politics (Bartolini, 2000, pp. 9-11; Benoit and Laver, 2006; Budge et al., 2001).
As a consequence, when national parties aggregate together at the EU level, a socialist party has generally more in common with a socialist party from another member state than with a conservative party from its own country. Differently put, there is more ideological variation across member states than across party families. Therefore, a 'party family'
46 Context and Theory
Framing the Impact of Enlargement 47
explanation for the formation and development of the Europarties has been proposed by several observers.
For instance, David Marquand (1978, p. 444) so commented on the eve of the 1979 direct elections:
the three genuinely transnational groups - the Socialists, the Christian-Democrats and the Liberals - all contain wide divergences of attitude and outlook but, although this is true, it is not the whole truth. There are big differences between the German, British and French socialist parties... but no-one is likely to confuse Helmut Schmidt or Willy Brandt with Margaret Thatcher or Giscard d'Estaing.
A decade later, Klaus Von Beyme (1985, p. 137) shared a similar opinion:
at the level of party elites, programs are becoming increasingly similar, and it cannot be denied that the parties are moving closer together. Even without exaggerated optimism it can be said that the process is strong enough to make the classification of the famille spiritttelks less problematic now than it was before the Second World War.
Two decades later, Simon Hix and colleagues (2007, p. 181) commented: 'politics in the European Parliament is... dominated by left-right positions and driven by the traditional party families of domestic European polities'.
The party family argument has traditionally offered a valid explanation for the transnational affiliation of the national parties in the EP. The problem with the party family approach is that Western Europe constitutes its traditional and almost exclusive reference. Hence, the issue of its validity beyond the 'safe' boundaries of the old Europe needs to be further discussed.
Party families in the new Europe
The use of the party family concept beyond Western Europe is disputed. Lipset and Rokkan's classification has been said to be valid when applied to the old Europe and, within such boundaries, to suit particularly well Continental Europe and the Nordic countries. When made to 'travel' beyond the borders of Western Europe, party family categories become less meaningful, as would then be the case for post-communist countries.
In general, the literature provides different evaluations of the applicability of the party family categories, or more simply the left-right categories, to Central and Eastern Europe. As Mair and Mudde (1998, pp. 213-14) (see also Rose and Munro, 2009, pp. 29-33) put it:
On the one hand, it is sometimes argued that the major differences between the east and west in Europe have been virtually overcome, and that the political parties and party systems on both sides of the former divide now increasingly resemble one another... On the other hand, scholars have also sought to develop distinct classifications adapted to the particular features of post-communist democracy.
The most explicit denial of the validity of party family categories in Central and Eastern Europe is due to the tabula rasa theory (Lawson et al, 1999; see also Dehaene, 2001). According to this theory, years of communism had flattened the sociat and ideological landscape in Eastern Europe, forcing political elites to broadly converge on liberal and free-market values and policies. In this context, voters' choices were mostly based on who could do best the job of delivering these policies. Party competition in Eastern Europe was structured around a single liberal-authoritarian dimension, separating the 'winners' from the 'losers' of the democratization process (Kitschelt era/., 1999).
In contrast, another stream of research has shown that -notwithstanding the relevance of some peculiar divisions, often based on ethnicity - Central and Eastern members are not really different from their Western counterparts. In most of the new members there is competition among parties over economic redistribution and, in general, their cleavage structure is comparable to that of the countries without a communist past (Evans and Whitefield, 2000, p. 197). Lewis (2001, p. 56) goes as far as to argue that 'the established left-right party differentiation is indeed valid and has widespread meaning in contemporary Eastern Europe, and party families can be roughly distributed in line with analysis of the political space conducted on a [left-right] basis'.
Nonetheless, the findings of recent comparative studies cast some clouds on Lewis' positive views. Hence, Klingemann et al. (2007, pp. 24-26) found that a left-right representation of politics works well in Central Europe, but becomes less valid the more we move to the East. In a review chapter about left-right orientations, Mair (2007, p. 214) also argued that a 'somewhat confounding pattern' has emerged in post-communist Europe, where 'some of the more conventional positioning
48  Context and Theory
Framing the Impact of Enlargement 49
associated with left and right has been turned on its head'. The strongest argument against 'standard' left-right politics in the new Europe is due to Benoit and Laver (2006). As they put it: 'the classic socio-economic definition of left and right seems to be a particularly Western way of looking at things' (2006, p. 145). Working with expert data, they seek to predict the left-right placement of a party from its position on economic and social policy. What they find is that a socio-economic definition of left-right is indeed accurate for West European countries, but it fits poorly the post-communist societies (2006, pp. 132-36).
Additionally, comparative research has also found that Central and Eastern parties make a different association between left-right and the second most important dimension in EU politics, the integration dimensions. In Western Europe, opposition to the EU project is located at both extremes of the left-right continuum and is hardly found among mainstream parties. Conversely, in the new Europe, hard Eurosceptic parties exclusively populate the radical-left pole (Marks et a}., 2006), while critical attitudes and moderate opposition to the EU can also be found among mainstream or governing parties, especially towards the right of the ideological spectrum (Taggart and Szczerbiak, 2004, p. 23).
Pragmatic incentives
Despite the differences identified by (most) contributors to the comparative literature among the parties in the old and the new Europe, following enlargement the vast majority of Central and Eastern parties became members of one of the existing political groups in the EP (see Chapter 1). Neither a substantial growth in the number of the non-attached deputies, nor the formation of new regional groupings has occurred. How can this apparently contradictory evidence be justified?
Two alternative explanations can be provided. The first and obvious one is that the argument proposed inter alia by Lewis (2001) is indeed correct. Party families and left-right categories can be meaningfully applied on both sides of the former divide. Central and Eastern parties fit well the existing Europarty configuration, based on the traditional party families. Differences among the two sides of Europe have been exaggerated. A second, alternative argument is that membership in the Europarties indicates, by itself, little about the ideological fit of the members.1 Indeed, both the supply side (the Europarties granting membership) and the demand side (the national parties applying for membership) have powerful instrumental or pragmatic incentives to, respectively, include the new members or seek membership in the existing Europarties. As Hlousek and Kopecek (2010, p. 11) note, 'many
parties aspire to membership in this or that established supra-national party structure for pragmatic tactical reasons rather than political self-identification; it is simply advantageous for them, because it allows them to better advance their interests in the European arena'.
More specifically, for national parties a 'marriage of convenience' is rewarding in terms of both office positions and influence in the EP, while membership provides an important source of legitimacy to be used in national politics. The formal rules of the EP provide two alternatives for the national parties unwilling to be members of the existing political groups. The first is to constitute a new group. However, the rules make this option hard to pursue. As Article 29.2 of the official rules reads: 'a political group shall comprise Members elected in at least one-fifth of the Member States. The minimum number of Members to form a political group shall be 19' (RoP, 2004 ed.).2 It was clearly difficult for a party, new to the Parliament, to 'recruit' fellows in four other countries and overcome the threshold of 19 deputies. The second option is to remain 'non-attached'. However, the status of non-attached penalizes members in several ways. They do not have voting rights in the Conference of Presidents, the EP executive organ; they are allocated a residual fraction of the speaking time in the plenary; they are rarely allocated important reports; and, as if to represent their minor role symbolically, they sit at the very back of the plenary room, Indeed, due to their minor status, the 'non-attached' are, often, either individual members that are experiencing problems with their national party, or extreme parties regarded as 'pariahs' in the chamber.
In contrast, by joining an existing group, parties enjoy the advantage of having a large bureaucratic apparatus at their disposal, with an experienced staff and financial resources they can draw on. In particular, it is crucially important for the governing parties to gain membership in either the EPP (group) or the PES (group) (Fabbrini, 2006). There are several rules in the EP rewarding the larger, and penalizing the smaller, political groups.
First, votes in the Conference of Presidents are weighted for the number of MEPs each group has, thus favouring the largest of them in all important decisions about the EP legislative planning and organization (art. 24, 2004 ed.).3 Second, the D'Hondt method is generally employed for the distribution of office positions. Although it belongs to the family of proportional formulas, it tends to better reward the largest parties. Thus, when committee or inter-parliamentary delegation chairmanships are distributed, all the most coveted positions are distributed to the largest groups. As illustrated by Corbett era/. (2011, p. 147); 'in 2009, for
50  Context and Theory
example, the EPP group had the right to the first, third, fifth, seventh, eleventh, thirteenth, fifteenth, eighteenth and twenty-first choices, the Socialists to the second, fourth, eighth, twelfth, sixteenth, twentieth and twenty-second choices, ..the EFD group to the nineteenth choice'. The distribution of legislative reports in committees is also generally based on a points system which privileges the largest political groups. It has been noted that 'the auction-like points system means that only the PES and the EPP can normally afford the most expensive reports' and that the largest political groups are 'over-represented' in the share of reports they get (Mamadouh and Raunio, 2003, pp. 346-47). Even though other considerations might also matter (such as a member's seniority or her/his policy experience), only two minor committees -Petition and Budgetary Control - do not apply any points system at all (Bressanelli et al., 2009).
If self-interested incentives to join the existing political groups are strong, there are also some instrumental reasons to seek membership of the extra-parliamentary parties. The Central and Eastern parties were granted the possibility to become members of the Europarties ahead of formal accession. With membership, they had the power, albeit limited, to influence or, at least, participate in the EU policy-making. More important, however, is the 'use' of Europarty membership as a source of legitimacy (Pridham, 2005). For parties without long political biographies, membership in the Europarties could offer a significant competitive advantage vis-á-vis potential competitors. Furthermore, not only did the Central and Eastern parties have strong incentives to seek membership, but the EU extra-parliamentary parties also had equally powerful incentives to grant them membership, without being too strict on ideological compatibility. According to the regulation on political parties at European level,4 the quota of financial resources the extra-parliamentary parties get is strictly proportional to their number of seats in the EP.5 Inclusion pays (literally) more also for the political groups. The parameters for the distribution of resources among them are based on both their size (the sheer number of MEPs) and their diversity (the number of member countries represented). Consequently, the larger and more diverse the group is, the more staff and financial resources it will get. Again, failure to include newcomers might substantially weaken a group vis-á-vis its competitors in the EP.
In sum, pragmatic incentives for parties with different ideological positions to seek membership in the existing Europarties are strong. In turn, the latter have significant material advantages in pursuing a strategy of inclusion, where ideological motives play a minor role.
Framing the Impact of Enlargement 51
4. Enlargement and competition
As argued in Chapter 1, a democratic EU not only needs institutionalized parties; it also needs some degree of competition between the parties at the 'core' of its party system. However, if enlargement weakens the cohesion of the Europarties, this is not good news for the competitiveness of the EP party system either. The close conceptual and empirical relation between the ideological cohesion of the Europarties and their 'distinctiveness' has been identified by Jacques Thomassen (2002, pp. 21-23). He argues that the ideological difference between the Europarties and their internal cohesion is not independent, but the former is a function of the latter. Hence, when the Europarties are poorly cohesive, their ideological differences also tend to be blurred.
Furthermore, the ideological diversity between the Europarties at the 'core' of the party system is not only relevant as an indicator of their potential competitiveness, but also because it is a strong explanatory factor for the competitive behaviour of the political groups in the EP. Studying coalition formation in the Parliament, Hix et al. demonstrated that variation in the ideological distance between coalition partners is the main factor underpinning the creation of a particular coalition. As they explain: 'as the policy positions of the two main parties in the European Parliament diverged, they voted together less. Similarly, as the distance between the EPP and the liberals increased, the liberals started to vote more with the socialists and less with the EPF' (2007, p. 158). Therefore, the EPP and the Socialists vote increasingly together as their ideological positions become more similar and undifferentiated, while their votes diverge more as their differences grow larger.
Thus, less internally cohesive Europarties tend to be less different Europarties. In turn, such Europarties are less capable to play an 'expressive' function and provide alternative policy agendas to the European voters, thus weakening the competitive potential of the EP party system. When ideological differences are less stark, the voting behaviour of their members also becomes more similar, with more collusion, rather than competition, between the major parties of the enlarged EP party system.
5. Factors affecting the impact of enlargement
My argument on the impact of enlargement has been framed as if enlargement had a homogeneous impact on all the Europarties. Yet, in this section, 1 suggest that two further factors need to be taken into
52  Context and Theory
Framing the Impact of Enlargement 53
account to explain the variation in the impact of enlargement between them: one has to do with characteristics of the demand side (the new parties accessing the EU) and the other one with features of the supply side (the Europarties themselves).
Starting with the former, there is a different level of boundedness of the 'left' vis-a-vis the 'right' of the ideological spectrum. In Western Europe already, the 'left' would be more cohesive and, for the party scholar, easy to define than the 'right'. As Peter Mair (2007, p. 213) put it: 'while the term "left" is easily associated with a specific class and with a broad set of policy alternatives, the term "right" is far less clearly delineated'. In other words, while the social and ideological referents of the left are relatively easy to identify, the boundaries of the non-socialist area are more uncertain. A similar, but stronger, argument can be made for Eastern Europe. On the left side of the ideological spectrum, where the socialist, social democratic and labour parties are traditionally placed, the communist successor parties largely transformed themselves into social democrats (Paterson and Sloam, 2005). In organizational terms, they relied on the structures of the former communist parties. In policy and ideological terms, they moved towards the centre left.
On the contrary, the right in Eastern Europe is populated by a wider and more heterogeneous set of parties. Hence, it has been argued that the 'liberal category is quite a broad one in Eastern Europe and ranges quite far to the right' (Lewis, 2001, p. 56); that what is 'right' is more difficult to identify among the many 'nationalist, conservative, Christian, liberal and populist groupings' (Vachudova, 2008, p. 388); that the right and the extreme right are less clearly distinguishable than in the West (Mudde, 2001). As Hanley (2004, p. 23) commented: 'In contrast to the limited range of post-communist successor party types, a diverse range of parties has emerged on the new Eastern and Central European centre-right. These, moreover, seem to lack any single, identifiable, common point of origin.'
Moreover, besides ideology, national parties in the old and the new Europe further differ in their level of institutionalization. A strong Europarty organization requires a stable membership and, with the inclusion of the less institutionalized parties of Central and Eastern Europe (see Rose and Munro, 2009, pp. 47-54; also Bartolini, 2005, p. 330), the organizational boundaries of the Europarties became less clear. Obviously, a Europarty which has national affiliates disbanding or splitting is less institutionalized than another one constituted by national parties which have shown the capacity to persist. In Central and Eastern Europe, the transformation of most communist parties
into social democratic successor parties gave them more stability and stronger organizational bases (Hanley, 2001). For both ideological and organizational reasons, their inclusion into the Europarties would therefore be less problematic than for parties in the liberal or conservative families.
The other factor which needs to be considered has to do with the level of institutionalization of the Europarties before enlargement. The literature on party organizations has demonstrated that 'over a certain threshold (which can be hardly specified) each further growth in size does not impact automatically on the level of complexity and on the degree of bureaucratization' (Panebianco, 1988, p. 348). Following this reasoning, it is here suggested that the largest and most institutionalized Europarties would be sufficiently equipped to accommodate the new members without implementing any further reform. Strong organizations would only require adaptation to the margins in order to adapt themselves to their membership expansion.
6. Conclusion: Enlargement in the process of institutionalization
This chapter has proposed a theoretical framework to understand the impact of enlargement on Europarty institutionalization. On the one hand, I suggested that the imperative of organizational performance in a context of higher transaction costs leads to broad organizational changes, with an increase in complexity and more centralization in decision-making. Party change would be conducive to the structural institutionalization of the Europarties. On the other hand, the different socio-political structure of the new members and the pragmatic rewards offered by membership are strong incentives for 'marriages of convenience', undermining the ideological cohesion of the Europarties. Consequently, competition on policy alternatives would become more difficult, and voting at the core of the EP party system more similar.
Placing enlargement within the broader developmental trajectory of the Europarties, more general reflections on their nature and the functions they (could) perform in the EU political system obviously arise, and party models could be useful heuristic devices to better understand them. In the light of the argument developed in this chapter, it is tempting to compare the enlarged Europarties with catch-all or cartel parties. Indeed, the expansion in membership beyond the boundaries of the traditional party identity and ideology, with the broad appeal to heterogeneous parties, might resemble the transition to catch-all or cartel
54   Context and Theory
parties in national political systems, when ideologically based parties were gradually replaced by ideologically loose parties, seeking support well beyond their classe gardee (Katz and Mair, 1995, pp. 13-14, 18). In turn, this transition led to a waning of the ideological differences between parties and to a containment of their competition on policy and programmatic alternatives. As a consequence, parties with a convergent ideological outlook started to collude in public office, forming 'cartels' and excluding non-mainstream parties. Cartel parties have limited incentives to compete and powerful ones to cooperate, in order to guarantee their collective organizational survival, distributing public resources and controlling the key positions of power. The similarity with the enlarged Europarties is appealing.
Moreover, the structural institutionalization of the Europarties could also be interpreted as functional to the objectives of a pragmatic and self-interested membership. Cartel parties strongly rely on public subsidies to survive. In national political systems, the shrinking of membership figures demanded for other sources of party financing, which were largely found in a state's public provisions. By the same token, the organizational consolidation of the Europarties could be instrumental to the needs of the member (cartel) parties, looking at the EU for additional resources and patronage positions (Bardi etctl., 2010, pp. 92-95). In other words, strengthening party structures at the EU level could be seen as part of a pragmatic strategy by the national parties, rather than as a step in the development of genuine Europarties performing more traditional representative functions.
Overall, for the prospects of party democracy in the EU, the consolidation of cartel parties would not be good news. As is well known in the literature (Katz and Mair, 1995, pp. 22-23; 2009, pp. 762-63), efficient as they are in managing financial resources to satisfy their cadres, cartel parties are not interested in providing the link between civil society and government. If the Europarties are approximating the cartel-party type, representative (party) democracy in the enlarged EU would be more distant,6 and the pursuit of a channelling, or expressive, function - pace the Treaty of Lisbon - would be what its parties are neither capable, nor willing, to do.
Part II
Empirical Analysis
Left-Right Confrontation or Grand Coalition? 149
8
Left-Right Confrontation or Grand Coalition?
1. Introduction
This chapter shifts the analytical focus from individual parties to the EU party system, tackling the question of the competitiveness of the EP party system after the 'mega' enlargement of the Union. As I argued in Chapter 1, competition is a key feature of a democratic political system. As Randall and Svasand (2002, p. 6) put it: 'To the extent that the process of party system institutionalization is seen as contributing to democratic consolidation, the implication is that the type of party system must entail a certain level of competition.' They illustrate the case of an 'institutionalized', but not really 'democratic' party system with reference to Mexico, where political parties are strong but, until recently, no other party was really able to challenge the hegemonic Institutional Revolutionary Party.
Party competition is central to the mechanisms of representation. When parties compete, citizens can choose among rival leaders and policies and punish or reward parties for their performance when in government. This is one of the central elements of the responsible party government model (Dalton, 2008, p. 226). But even when there is no party government but, rather, a separation-of-powers system - as in the political system of the EU - parties can still effectively act as representatives, and seek to translate citizens' preferences into enacted legislation. When parties collude rather than compete - as in Katz and Mair's famous 'cartels' - the traditional linkage function is relegated to the background. Parties collude in order to guarantee their collective survival (with state funding, the distribution of patronage positions,...), and competition over alternative policies is replaced by party cooperation in public office. Political programmes become increasingly similar
and parties come to resemble one another, not least in terms of the policies they want to pursue (Katz and Mair, 1995, 2009).
Traditionally, observers of politics in the EU have stressed its consensual character. The EU has typically been described as a consensus type of polity, where accommodation and compromise are sought at all levels. Specifically, consensual politics have been observed in the EP. For instance, Westlake (1994) described the party system in the EP as a 'market oligopoly', as the two largest groups in the assembly - the EPP and the PES - vote frequently together and divide among themselves the top parliamentary offices. This description of politics in the EP is by no means outdated. In a recent study on representation in the EU, Richard Rose (2013; also Rose and Borz, 2013) has pointed to the persistence of what he labels the 'black-red' agreement. Indeed, aggregate votes show that these two parties, ostensibly opposing each other, vote together in about two-thirds of the times a roll-call vote is cast (VoteWatch, 2013), and in more than half of the recorded votes the super-majority also includes the ALDE.
Collusive practices among the two parties at the core of the EP party system have not failed to attract critical remarks. In particular, it has been argued that a 'cartel' between the EPP and the PES makes more difficult for the European citizens to understand what is at stake in the EU and contributes to the decline in turnout for the EP elections. In contrast, by presenting different policy programmes to the European voters, and by pursuing their programmatic objectives in the EP, the Europarties would reduce the distance between the people and the EU institutions and, consequently, contribute to the democratization of the EU.
Moving from the assumption that preferences matter for coalition formation, this chapter assesses the argument that, by integrating parties from the ten new member countries from the post-communist region in 2004/07, the ideological differences between the Europarties would be lessened and the enlarged Parliament would then witness a 'return of the Grand Coalition' (Hix, 2009), with a shift away from the 'culture of competition' that some scholars observed in the early 2000s (Kreppel and Hix, 2003). In terms of data, the chapter relies on a large sample of roll-call votes, and compares the result of two logistic regressions for the pre-enlargement (1999-2004) and post-enlargement (2004-09) legislatures.
This chapter develops as follows. Section 2 reviews the debate on cartelization in the EP party system in the light of the most recent findings. Section 3 presents the theoretical arguments explaining the formation of a 'black-red' coalition. Section 4 introduces the data and
148
150  Empirical A nalysis
Left-Right Confrontation or Grand Coalition? 1S1
variables. Section 5 develops the empirical analysis. Finally, Section 6 discusses the findings and concludes.
2. A cartel party system?
To argue that the parties and the party system at the EU level fail to represent the European citizens is nothing new. On the one hand, when Reif and Schmitt (1980) proposed their 'second-order' theory to describe elections to the EP, they referred to the absence of a truly European dimension for electoral competition. Elections were contested by the national parties, on the basis of national issues, in each of the nine member countries of the then EC. European elections served as a test for the popularity of the incumbent government, rather than for choosing alternative courses of action over Europe. On the other, Europarties failed to coordinate national election campaigns and propose coherent policy platforms. In the electoral arena, an EU party system did not materialize with the introduction of direct elections for the EP. As Mair (2000, p. 38) bluntly put it: 'there is certainly no real sense in which we can speak of a distinct and autonomous European party system'. A party system at the EU level would require stable and patterned interactions among parties, and, at best, what there is in the EU is a 'basket of parties' (pp. 38-39).
Nevertheless, elections are not the only arena where parties can form a system of interactions. National parties aggregate into transnational party groups in the EP. Moving from Sartori's (1976) classic definition of a party system - 'a system of interactions resulting from inter-party competition' - there can be little doubt that the political groups put into place a system of interactions. After the Treaty of Lisbon, the EP de facto is the lower chamber of a bicameral legislature, and its parties need to constantly interact with each other to build up the necessary majorities and thus shape EU legislation. It is precisely for this reason, however, that the notion of competitive interaction among the parties could be challenged.
The consensual nature of EU politics, and the search for consensus and compromise have been traditionally observed in the EP. From Westlake's (1994) characterization of the EP party system as a 'market oligopoly' to Rose's (2013) description of the patterns of interaction between the People's Party and the Socialists as a 'black-red' cartel, the mechanics of the EP party system have been described as collusive rather than competitive. Rose (2013, pp. 127-28) goes as far as to claim that 'consensus in the European Parliament is repressive; a cartel that fails to
represent the divisions of Europe's citizens'. His argument is that the collaboration between the two major political groups to enact laws comes at a high price: the inter-party compromise significantly reduces the extent to which the votes of the MEPs match their voters' preferences. Consequently, the output of the decision-making process - disregarding the need to negotiate with other institutional actors in the EU political system, but simply observing the policy process inside the EP - is not representative of the citizens' preferences.
This argument bears important implications for party democracy in the EU. If a grand coalition replaces ideological confrontation at the core of the EU party system, however effective parties would be in organizing the decision-making process in the EP, representation is doomed. What the Lisbon Treaty emphatically states in its Article 10.1 on the EU as based on the 'principle of representative democracy' would likely remain written on paper. Representative (party) democracy needs (at least some) competition. Citizens must be able to choose among different policy programmes, which the legislative parties will seek to transform into concrete policies. This is what cartels are unwilling to do.
3. Explaining the grand coalition Ideological preferences
Explanations for the cartel-like behaviour of the two largest parties in the EP are to be found either in their ideological preferences or in the institutional rules of the game (see Kreppel, 2002; Kreppel and Hix, 2003; Hix etal, 2003). The argument that preferences matter to explain coalition formation in the EP has been most strongly developed by Hix and colleagues (2007, pp. 147-60). Using all roll-call votes from July 1979 to June 2004, they test the proposition that parties will tend to form coalitions with the parties that have the closest positions to theirs. On the basis of their analysis of the dimensionality of the ideological space, they argue in particular that the political groups will try to minimize the distance with their coalition partners on the left-right dimension. Their notorious finding is that ideological (that is left-right) distance matters, as one standard deviation change in left-right distance between two parties implies a 16 per cent standard deviation change in the frequency that these parties will vote the same way. As Hix et al, (2007, p. 157) put it: 'This result gives us a strong indication of the crucial importance of left-right politics in the EP.' Observing in particular the voting behaviour of the EPP and the PES, they argue that when the two major political groups voted together to a greater extent - as they
Left-Right Confrontation or Grand Coalition? 153
152 Empirical Analysis
did in the late 1980s and early 1990s - it was so because they shared position on many issues on the parliamentary agenda such as, for instance, the regulation of the common market, or environmental protection. When they voted differently - as in the 1999-2004 legislature (Kreppel and Hix, 2003) - it is so because their left-right preferences moved apart.
Hix' arguments are not undisputed. First, the idea that left-right conflicts within the EP are so dominant 'seems to fly in the face of direct experience and received knowledge' (Westlake, 2007, p. 342). This is difficult to believe because many votes in the EP concern regulatory and technical matters, where ideological divisions are unlikely to emerge and super-majorities should rather form. Second, votes on issues related to the integration dimension, where the groups have similar preferences (see Chapter 5), should result in nearly unanimous outcomes (excepting the fringe parties on the left and the right). As several votes in the EP regard EU integration issues, the importance of left-right divisions should not be exaggerated. Third, it is argued that the institutional rules of the game and strategic incentives are powerful enough to make the political groups collude even when their left-right preferences diverge.
Institutional rules and strategic factors
Three main arguments of institutional nature have been used to explain the coalition between the EPP and the PES. A first argument has focused on the absolute majority requirements in the parliamentary decisionmaking procedures. In the second reading of the co-decision or (after Lisbon) the ordinary legislative procedure to reject or amend the Council's common position, in budgetary matters and in the assent procedure, the EP decides by an absolute majority of its members, regardless of whether or not all are present. After the 2004 elections, the new EP had 732 seats: the absolute majority was set at 367. After the enlargement to Bulgaria and Romania in 2007, the EP had 785 members and the absolute majority moved to 393 votes. Taking into consideration that in the 2004-09 EP the average rate of participation in the plenary sessions was about 85 per cent, a grand coalition could be regarded as the only viable option for passing the absolute majority requirements. Thus, while a coalition between the EPP and the ALDE was also slightly above the absolute majority threshold - counting 376 members - the average attendance rates made this option far less secure than the alternative EPP-PES alliance (amounting in June 2004 to 488 seats, with about 414 members, on average, casting their vote).
A second perspective focuses, instead, upon the collective institutional interest of the EP to show itself united vis-a-vis the other EU institutions.
Here, it is argued that party competition and ideological battles can only be fought when they are not detrimental to the interests of the EP as a whole. Thus, while in the early round of a co-decision procedure, as well as for votes over individual amendments, ideological divides are likely to emerge, bearing no costs for the Parliament, compromise solutions are rather sought in later rounds of co-decision, as well as when a final text is put to a vote (Kreppel, 2002, pp. 153-67). Indeed, the EP has a strong institutional interest to present itself united vis-a-vis the Council to expand its bargaining power. Thus, while the EPP and the PES can fight ideological battles more when they vote over individual amendments and in the earlier rounds of the legislative procedure, they need to build a compromise on a final text, or in later rounds of a co-decision (i.e. ordinary legislative) procedure.
Finally, grand coalitions are more likely to emerge on internal parliamentary matters, such as reforming the Rules of Procedures, changes in the parliamentary agenda, and similar issues. Indeed, as described in Chapter 2, there are a number of formal and informal rules in the EP that tend to favour the larger vis-a-vis the smaller political groups: from the allocation of the committee chairmanships to the points system for the attribution of legislative reports, from the voting rules in Conference of Presidents (Corbett et al, 2011, p. 110) to the allocation of the speaking time in the plenary. As it has been demonstrated by Kreppel with a thorough analysis of the EP Rules (2002, pp. 102-22), the two largest groups have made a strategic use of each round of reform of the Rules to advance their own interests. The cooperation (or cartel-like behaviour) between the EPP and the PES has effectively minimized the power of the smaller groups in the EP organizational structure over the last 20 years. Hence, controlling for other factors, the EPP and the PES can be expected to more often vote together on internal parliamentary matters than elsewhere.
4. Research design
In order to test the impact of enlargement on competition in the EP party system, I rely on two random samples of roll-call votes for the fifth (1999-2004) and sixth (2004-09) legislatures. For both periods, I extracted 10 per cent of the votes from the database created by Hix and colleagues (http://personal.lse.ac.uk/hix/HixNouryRolandEPdata.htm), stratifying the sample by year.1 Hence, my data set counts 570 roll-calls for the earlier legislature, and 620 votes for the later one. There are no restrictions on the votes: they cover legislative and non-legislative
154  Empirical Analysis
resolutions, internal and budgetary matters, and so on and so forth (see Table A.7 in Appendix for details).2
What I focus on in this chapter - my dependent variable - is the formation of the 'grand coalition' between the EPP and the PES. The coalition between the two groups is considered to have formed when a majority of the members of the two voted the same way. Hence, when a majority of the members of the EPP and the PES attending a plenary session together voted 'Yes' or 'No' or abstained, this variable was coded 1; 0 was coded otherwise.
The key independent variable - ideology - has been operationalized with seven dummy variables, indicating the policy area of a vote (Hix et al., 2007, p. 123).3 'EUJnstitutionaV captures votes on constitutional and inter-institutional affairs, such as the agenda of the European Council, or the discussion on the reform of the treaties. 'External' refers to votes on foreign policy, security and defence as well as commercial agreements with third parties. 'Economic' includes votes on competition policy, on monetary policy and on the common market. 'Social' is coded 1 when workers' rights, pensions or other social provisions are under discussion, as well as when specific funds (Cohesion and Structural funds) are mentioned. 'Environment' captures the votes on environmental protection and health. Finally, 'Civil Liberties' includes the votes on human rights, individual freedom, justice and gender equality. All the other votes were subsumed under the category 'Other'.
Furthermore, I used the following variables to operationalize the institutional factors mentioned above. To begin with, the dummy variable AbsJAajority distinguishes between the votes where the absolute majority requirement applies (second reading in codecision to reject or amend the Council's common position, budgetary and assent procedures), which are coded 1, and those where a simple majority suffices, coded 0. Second, in order to capture the collective interest of the EP, two dummy variables have been extracted from Hix et al.'s (2007) database. A first binary variable - Whole - is coded 1 when the vote is on a text as a whole (be it a resolution or a legislative dossier), while it is 0 when the vote is on an amendment. A second dummy - Final - is coded 1 when the vote refers to the final reading of a legislative text (first reading in consultation and assent procedures and third reading in co-decision), 0 otherwise. Finally, the variable Internal_EP controls for the internal nature of the vote; it is coded 1 when the vote is on procedural or organizational matters such as the reform of the Rules of Procedure, changes in the parliamentary agenda or elections.
In order to analyse the formation of the grand coalition, given the binary nature of my dependent variable, I use a simple logistic
Left-Right Confrontation or Grand Coalition?  155
regression.4 The statistical analysis proceeds in two steps. In a first step, all observations - all the votes in my samples from July 1999 to May 2009 - are pooled together. By including a dummy variable for the sixth legislature {VI_EP), 1 conduct a first preliminary test of the impact of enlargement on the formation of the 'grand coalition'. In a second step, I run two separate models for the fifth and sixth legislatures. Here, I expect smaller coefficients and weaker effects for the ideological variables associated with left-right in the enlarged Parliament. Before running the statistical models, 1 present some descriptive data.
5. Empirical analysis Descriptive evidence
Aggregate data from the sixth and seventh EP suggests that the grand coalition is indeed more frequent than in the EP-15. VoteWatch (2013) calculated that the EPP-PES alliance occurred in about 70 per cent of the votes in the 2004-09 EP and in about 72 per cent of the votes in the 2009-14 EP. Contrariwise, the grand coalition formed 'only' in about 65 per cent of the votes cast in the last legislature before enlargement (see Table 8.1).
It is also in the highly symbolic vote for the presidency of the EP that the grand coalition appears to be back in full strength. Since the late 1980s, the two major political groups agreed to 'share' among themselves the presidency of the EP. This agreement broke out after the 1999 EP elections, when the French EPP Fontaine became president for the first half of the term and the Irish Liberal Cox was appointed for the second half. According to some commentators, this passage indicated the
Table 8.1   Coalition frequency in the 1999-2004 and 2004-09 legislatures
Political group	GUE-NGL	G-EFA	PES	ALDE	EPP	UEN	IND-DEM
GUE-NGL	_	0.79	0.69	0.55	0.42	0.46	0.59
G-KFA	0.74	-	0.72	0.62	0.47	0.45	0.56
PES	0.62	0.70	-	0.73	0.65	0.53	0.53
AI.DF.	0.51	0.62	0.75	-	0.68	0.55	0.52
EPP	0.41	0.50	0.70	0.77	-	0.71	0.52
UEN	0.44	0.48	0.63	0.71	0.81	-	0.63
IND-DEM	0.36	0.34	0.35	0.40	0.46	0.48	—
Key: Above the off-diagonal the proportion of times the majorities of any two political groups voted the same way in all roll-call votes In the 1999-2004 parliament (Hix et al., 2007, p. 151); below the off-diagonal, In the 2004-09 EP (VoteWatch, 2013).
156 Empirical Analysis
Left-Right Confrontation or Grand Coalition? 157
o
er
a o O
• EPP-UEN
• EPP-ALDE
PES-ALDE.X
GR-GUE" ,
PES-GR
ALDE-UEN \ * *EPP-PES
• PES-UENN .
•\«PES-GUE ALDE-GRX
ALDE-GUE •
• UEN-GR
• EPP-GR
UEN-GUE
• EPP-GUE
-1-
4
Distance
Key: Each dot represents the coalition frequency between two political groups. The j^axis does not start at zero.
Figure 8.1   Ideological distance and coalition frequency in 2004-09
shift to a 'new culture of competition' (Kreppel and Hix, 2003). However, the EPP and the PES agreed to divide the presidency among them again in 2004. The Spanish Socialist Borrell was appointed president for the first half of the legislature and replaced by the German Christian-Democrat Pottering in the second. In the 2009-14 legislature, a similar agreement was established, with the Polish EPP Buzek voted president for the first half of the legislature, and the German Socialist Schulz for the second.
Yet, other data convey the impression that ideology has not become meaningless in the enlarged EP. Figure 8.1 displays a bivariate regression of coalition frequency and ideological distance on left-right between any two political groups. Each group's ideological position has been measured, from the Euromanifestos, as the weighted average position of its member parties. Distances have been computed as the absolute value of the difference between the positions of any two groups.
Clearly, as the left-right distance between any two political groups grows, the coalition frequency decreases. Hence, the lowest coalition frequency is between the EPP and the GUE-NGL, the most distant
groups, while the highest frequency is between the EPP and the conservative UEN, which are very close in the ideological space. As for the grand coalition, it is somewhat more frequent than the ideological distance between the two groups would let one expect. Nonetheless, knowing only their ideological position, it is possible to make a fairly accurate prediction about their coalition frequency.
The descriptive evidence is, therefore, far from conclusive. In order to provide a more stringent test about the occurrence of the grand coalition in the enlarged EP, and assess if ideological conflicts ceteris paribus are less prominent than before enlargement, a multivariate analysis is needed.
Statistical analysis
To begin with, Table 8.2 displays the result of a logistic model with pooled data. The results of the model are informative insofar as they provide a first test about the existence of any 'enlargement effect'. Indeed, the results seem to indicate that some difference between the two periods exists. The enlargement variable is significant, albeit weakly {p < 0.1), and - as discrete changes (not reported) indicate - it makes the formation of the grand coalition more likely by about 6 per cent. Variables are in the expected direction. Agreement between the EPP and the PES is more likely when an absolute majority is required, when a vote is in the final reading of a legislative procedure, when the political groups vote on internal parliamentary matters and when the whole text, rather
Table 8.2  'Grand coalition' and enlargement effect
	Coefficient	SE	z	p<\z\
Absolute Majority	0.302	0.229	1.31	0.189
Final Vote	0.096	0.201	0.48	0.634
Whole Text	1.201	0.209	5.76	0.000*"
Internal_EP	0.329	0.589	0.56	0.576
Enlargement	0.247	0.134	1.84	0.066*
Constant	0.429	0.104	4.11	0.000***
Observations = 1,106 Prob. > chi2 = 0.000 Log-likelihood = -659.632 Pseudo-Rz = 0.037
Key. The sample for the analysis ln= 1,106) is smaller than the whole sample (n= 1,190) due to missing data which could not be retrieved from the EP website. *"p < 0.01, "p < 0.0S, *p<0.1
158 Empirical Analysis
than a single amendment, is put to vote. Nevertheless, only the variable Whole is significant (p < 0.01) - suggesting that ideological battles between the political groups are more likely to be played over individual amendments, while consensus is more likely to emerge on the final text.
In order to identify more precisely the factors behind the 'grand coalition' after enlargement, Table 8.3 reports the estimates of two logistic models run separately for the 1999-2004 and 2004-09 periods. In general, what is striking when comparing the estimates in the two models is the overall similarity of results.5
For the fifth legislature, all coefficients are in the expected direction. The voting rules in the EP and the inter-institutional nature of a vote make the formation of the grand coalition more likely. However, the only strong and significant effect (p < 0.01) is when a resolution or a legislative text as a whole is put to vote. What really matters for the formation of the EPP-PES agreement is, however, the ideological domain of a vote. In constitutional and inter-institutional issues, the EPP and the PES vote more frequently together. On economic, social, environmental and socio-liberal policies, instead, the EPP and the PES are significantly more likely to diverge. The effects are strong and highly significant (only Social is below the 5 per cent threshold) - confirming that left-right competition was a key feature of the party system in the fifth Parliament (see Hix et «/., 2007). Moving to the estimates for the 2004-09 legislature, results are broadly similar. All coefficients are in the expected direction. Additionally, all the ideological variables have the expected sign and are significant at 1 per cent (with the exception of EUJnstitutional). This important result indicates that left-right competition is a central feature of the enlarged party system, as it was in the EU-15.
The comparative effect of the policy variables can be better appreciated by looking at discrete changes (Table 8.4). 'Discrete change' is used in logistic models to measure the change in the predicted probability of an outcome (here, the formation of the EPP-PES coalition) when an independent variable moves from its minimum to its maximum value (for dummies, obviously, from 0 to 1), keeping all the other variables constant (for instance, at their mean values; see Long and Freese, 2006). Other things being equal, the ideological domain of a vote is not of minor importance in the enlarged EP. Looking at the discrete changes for each variable, the EPP and the PES became more competitive after enlargement on external, social and environmental policies. For instance, compared to Other as reference category, a vote on employment or social policy made the grand coalition less likely by about 30 per cent in the enlarged EP - about 5 per cent more than in the previous
p.
.a
T3
5
C
g
□
—
»1 ad
3
j ' I I I ! 5 ! S?S38SSS8g
dddddddodd
u-> — oc *o M d ^ d
<r oo n n a n *o
r4 ro tvj ^
I    I    I    I I
©dddooddcJd
I I I I I
c
'B
2,
Sg
1 n
C
5
~ -a 1 §
■a al
c 1
J= .S > "J  X  (J  O  G «
£ o s
S 8
—	LT.			N			
	,—i		-t	[--.	Q	N	
		o		00	9	O	O
d	d	d	d	d	d	d	d
		1-	c				m
©			-t	—*		N	
d	d	~~	c	d	-t	N 1	to 1
c	1/3	•c	-t			N	vO
N		m	N				
		n:	t		PA	-*	
d	d	d	d	d	C	d	d
2.1
3 >
£ £ E
a> ^ R ■=
S3 ?
o d
3 t
3§'
ii ° ,
5 £ £ ™ s u ~ K
O t J e.
i-h VO ** CM
dd-Jdon »h i-5 -j i-<
I I I I I
II o
.2 J5
c
o
II
2 i
« : c o- oJ
"o i j- e « .3 ™
S z
Si.
60 g
160 Empirical Analysis
Table 8.4  Ideological competition before and after enlargement
1999-2004
2004-09
	Sign	Discrete	Sign	Discrete
	coefficient	change	coefficient	change
EUJnstitutional	( + )	0.042	( + )	0.063
External	(-)	0.014	(-)	0.191
Economic	(-)	0.338	(-)	0.195
Social	(-)	0.2S4	(-)	0.301
Environment	(-)	0.273	(-)	0.285
Civil Liberties	(-)	0.381	(-)	0.283
Key: 'Discrete change' indicates the effect of a change In each independent variable on the propensity to form a 'grand coalition', holding all other variables constant. Estimation from the models in Table 8.3. Values for other variables are not reported. Reference category is Other.
legislature. True, the effect of the variable Economic is smaller in the new Parliament - and voting on issues like the internal market, or consumer protection (compared to Other) made the EPP-PES coalition less likely by about 20 per cent in the 2004^09 Parliament, compared with 34 per cent in the EP-15.
All in all, ideology remains a key factor behind the competitive/collusive behaviour of the two political groups at the core of the EP party system. Voting on a whole text - the other strong effect in the model - increases the chances of the grand coalition by about 23 per cent in the EP-15, and by about 19 per cent in the EP-27.
Finally, two differences between the two legislatures are worth a mention. The first is the effect of the absolute majority requirement. While it bears practically no effect in the EP-15, it becomes significant (p < 0.05) and moderately strong in the enlarged Parliament. Second, external policy issues are significantly more divisive in the enlarged Parliament than before. The concluding section will speculate on the reasons behind these changes.
6. Conclusions
The main conclusion emerging from the analysis of voting behaviour of the groups at the core of the EP party system is clear: after enlargement, competition on left-right issues has not been buffered away by a return to more consensual practices. More generally, extending to the
Left-Right Confrontation or Grand Coalition?  161
post-enlargement period the findings of the literature on party competition in the EP (for instance Kreppel, 2002, p. 170; Hix et al, 2007, p. 158), this chapter has shown that the policy area of a vote is a crucial factor factor behind the formation of the 'grand coalition' between the EPF and the Socialists. Thus, on economic, socio-liberal and environmental policies, the two major political groups tend to vote in a different way, while on EU issues their votes tend to converge. The inter-institutional game played by the EP also tends to affect the propensity of the EPP and the PES to coalesce, but neither the rules of the game nor the stage of the procedure is strong enough to force cooperation among them when their preferred ideological outcomes strongly differ.
These findings, based on voting behaviour, nicely mirror the findings of Chapters 5 and 6 based on preferences. In the previous chapters, I have shown that the EPP and the PES have different preferences on left-right and correlated dimensions, while they have similar preferences on the EU integration dimension. Preferences do translate into behaviour: the political groups are competitive on the former dimension, and collusive on the latter. In turn, this finding bodes well for the responsiveness of the system. As Peter Mair has argued (2005; also Mair and Thomassen, 2010, pp. 27-29), political representation with regard to a specific issue dimension needs to be realized at the same level where decisions on this issue are taken. Clearly, the EP has strong decision-making powers on left-right issues - which are largely decided under the ordinary legislative procedure - but is weak on EU 'constitutional' issues. In other words, even if parties were competitive on the latter issues, their voting behaviour would hardly have any policy impact.
If ideology largely accounts for coalition - and, specifically, 'grand coalition' - formation before and after enlargement, there are some differences between the two periods which are worth discussing in more depth. The first regards the effect of the absolute majority requirements, which only matters in the enlarged Parliament. Since the absolute majority rule is applied in the second reading of the codecision procedure to amend or reject the Council common position, this change might reflect a trend to conclude 'early' in the co-decision procedure (Corbett eta/., 2011, pp. 240-45; Rehefa/., 2013). Conciliation is becoming rare and, in the fewer cases when the legislative procedure reaches second reading, the PES and the EPP might wish to signal their dissatisfaction with the Council about the outcome of the informal negotiations. In other words, the informalization of policy-making might also have an effect on competition in the EP, and empirical analyses should incorporate these new developments in the EU decision-making process.
162 Empirical Analysis
The second difference regards competition over external affairs. In the enlarged EP, the EPP and the PES have become more competitive regarding foreign policy. The reasons behind this change need to be carefully investigated, but the high 'intergroup solidarity' on external matters of the late 1990s seems to have vanished.
Generally, on the basis of the empirical evidence presented here and in previous chapters, it can be concluded that institutionalization and competition go hand in hand. The political groups are becoming organizationally stronger and more autonomous from the national member parties (Chapter 3), have clear ideological boundaries (Chapters 5 and 6), behave cohesively (Chapter 7) and compete on left-right ideological issues. True, the findings exclusively regard the political groups in the European Parliament. Competition in the electoral arena remains embryonic. Nonetheless, after and, somehow, despite the mega enlargements of the EU, a more active role of the Europarties in the EU political system does not appear out of reach.
Conclusions: Europarties' Prospects beyond the 2014 EP Elections
1. Summary of the findings
This book theoretically framed and empirically assessed the impact of enlargement on the Europarties. Drawing on the literature on comparative party politics, and developing an analogy with party development at the national level, it suggested that the enlarged Europarties would be more similar to catch-all or cartel parties. The structural differences between Western and Eastern members were expected to dilute their ideological coherence while powerful pragmatic incentives pushed for marriages of convenience among ideologically heterogeneous parties. Additionally, the consolidation of the Europarty organizations could also be understood as functional to the interests of the member (cartel) parties, rather than conducive to the development of genuine EU-level parties.
Yet, the empirical analysis demonstrated that this scenario fails to capture the reality of the enlarged Europarties. After enlargement, the Europarties both have stronger organizations in the EP and are formed by ideologically cohesive members. All in all, there is a clear potential for the Europarties to play a representative role in the EU political system, as the Treaty of Lisbon prescribes. I summarize below, chapter by chapter, the path through which this book arrived at this (unexpected) conclusion.
In Chapter 3, the empirical analysis began with the organization of the political groups. On the basis of interviews and the groups' official documents, it found that enlargement has been an important catalyst for reform. Enlargement was perceived as a 'membership shock' by the party leadership, pushing it to reconsider the rules of the party organization. In order to maintain a high level of cohesion
163
164 Europarties after Enlargement
among their members, the groups centralized decision-making powers and introduced new mechanisms of coordination between the committee and the plenary levels. Such reforms were implemented especially by the PES and the ALDE, but less by the EPP group. This latter had introduced important changes to its organizational structure during the 1990s already, and enlargement no longer required them. Interestingly, this chapter demonstrated that cohesion is built through bargaining and deliberation, rather than being produced by the sanctioning tools of the party leadership.
Chapter 4 moved the focus to the analysis of ideology, introducing the Euromanifestos and the EU Profiler data. Above all, it demonstrated that their measures of the two most important dimensions in the EP ideological space, the left-right and the EU integration dimensions, are valid. Comparing these ideological dimensions in the former two sides of Europe, Chapter 5 found that the ideological fit of the parties from the post-communist region has been, in general, smooth. True, their integration appears to be more problematic for the ALDE, on the integration dimension in particular, but Western and Eastern member parties clearly did not constitute two separate sub-groups. Finally, the chapter showed that the enlarged EP party system could still be accurately described by an inverted U-curve, where extreme positions on left-right correspond to opposition to the EU.
Chapter 6 performed a more stringent test on the ideological compatibility of the parties from the post-communist countries. It asked: by only knowing the ideological position of the national parties, how well can one predict which political group they will join in the EP? The simple answer is 'very well'. Ideology is the most important predictor of political group membership, and it is so on both sides of the former divide. Pragmatism also matters - as large national parties tend to join the larger political groups to benefit from additional resources and leverage - but its effect is significantly smaller. Thus, the chapter concluded by emphasizing the strong potential for representation in the enlarged EU, as the political groups' member parties largely share values and a common ideology, and could then propose a coherent 'programmatic supply' throughout the enlarged EU.
Following Vicky Randall and Lars Svasand's (2002, p. 12) advice to study party institutionalization in terms of both 'attitudes, culture' and 'integrated patterns of behaviour', Chapter 7 used roll-call vote to study cohesion. Corroborating the findings of the empirical analysis so far developed, the chapter showed that member parties from Central and Eastern Europe did not defect more from their political group
Conclusions: Europarties' Prospects beyond the 2014 EP Elections  165
than the average West European member. On a more general level, Chapter 7 demonstrated that preference-based theories of cohesion have a strong explanatory power for the EP. Paraphrasing Randall and Svasand, I argued that the political groups behave cohesively (largely) because their members share similar ideological preferences.
Finally, Chapter 8 assessed whether competition between the two major parties at the core of the EP party system was undermined by enlargement. Specifically, it estimated the importance of ideology for the formation of the 'grand coalition' between the two groups in the fifth (1999-2004) and sixth (2004-09) legislatures. Analysing roll-call votes, it demonstrated that competition on left-right issues is as important in the enlarged Parliament as it was in the EP-15. Thus, the EPP and the PES groups tend to vote differently when their ideological preferences diverge, and distinctively on left-right issues, while they tend to vote together when their preferences converge, as on EU integration.
Bringing all the empirical evidence together, widely shared expectations about the enlargement 'shock' did not materialize. Organizationally stronger, ideologically cohesive and competitive on left-right policies: this is as much a succinct as an effective way to describe the enlarged Europarties.
2. Contribution and implications for further research
The most obvious contribution of this book is to the field of 'enlargement studies' (Pridham, 2008). By analysing the consequences of enlargement on the basis of general theories of partisan formation and development, rather than with ad hoc explanations, the book findings echo the conclusions of most research on enlargements. The modus vivendi of the enlarged institutions of the EU can best be described as 'business as usual'. They aptly managed their widening and were in general successful to adapt their working procedures and rules to absorb their expanded membership (for all, see Best et ai, 2008). The Europarties do not make an exception to this 'rule'.
More generally, this book has cast new light on the relationship between the 'widening' and the 'deepening' of the Union (in general, Kelemen, Menon and Slapin, 2014). Without embracing a normative perspective - such as that the deepening of the EU ought to follow its widening - nor assuming tout court a negative association - with widening impeding, or at least slowing down integration - the empirical analysis has shown that deepening could be the most rational response to widening. In the case of the political groups, deeper integration was
166  Europarties after Enlargemen t
necessitated by the risk of institutional failure, as they faced a sudden growth of transaction and monitoring costs.
Furthermore, this book contributes to a better understanding of both the £wroparties and European political parties. One of its key findings is that the aggregation of the national parties - from the West as well as from the East of Europe - produces coherent and ideologically bounded Europarties. Even if direct campaigning on the basis of a common manifesto is yet to come in the EP elections,1 a coherent 'programmatic supply' is de facto offered to the European citizens already. Thus, the empirical results show that the conditions for 'representative democracy' in the EU - at least when the focus is selectively placed on the 'supply side' of politics - are good enough. If competition on left-right takes place already within the EP, there is little, in terms of ideology at least, preventing real competition among the Europarties at election times. In order to make the EP elections more distinctively European, several concrete institutional reforms are currently considered: some have already been implemented, some others are on a dead path and still others were tested for the first time in the 2014 EP elections. Their implications are discussed in the final section below.
This finding is not only relevant for the Europarties, though. While comparative studies of party competition in the West and the East of Europe (Benoit and Laver, 2006; Klingemann et al, 2007), as well as more theoretical reflections on the concept of 'left-right' (Mair, 2007), are dubious on the compatibility of socio-political cleavages between the former two sides of Europe, the empirical results provided a surprisingly different picture. Party families in Eastern Europe can be distributed in line with analyses of the ideological space conducted on the left-right dimension. The ideology of the national parties is by far the strongest predictor of political group affiliation both in the West and in the East of Europe. Furthermore, the analysis of voting behaviour shows that party delegations from Central and Eastern Europe are very loyal to their political groups in the EP. My analysis corroborates, then, recent studies in the literature which, using both 'hard' (Thomassen, 2009) and qualitative data (Hlousek and Kopecek, 2010), find that the post-communist members fit well the traditional (West European) party families.
Although this book did not investigate further why ideological convergence has been reached, this is an interesting puzzle which would surely merit further analysis. A structural explanation would require an in-depth investigation of the development of the cleavages structuring party competition in the post-communist countries. Nonetheless, the role of agency should not be disregarded: what is the role of the
Conclusions: Europarties' Prospects beyond the 2014 EP Elections  16 7
Europarties (if any) in the ideological convergence of the new members? Such question has not been the object of systematic scholarly attention yet (except von dem Berge and Poguntke, 2012). The specific contribution of the Europarties in the process of ideological convergence within party families across the enlarged EU and the now candidate countries, vis-a-vis the broader influence of the EU and endogenous processes of change, still needs to be properly assessed.
Finally, this work has contributed to the debate on the sources of voting cohesion and competition in legislatures. Stephanie Bailer (2008, p. 200) has wondered about 'the puzzle of continuing party group cohesion after enlargement'. By assessing preference-based visa-vis institutional theories of cohesion, I have shown that a theory of cohesion based on ideological preferences has the strongest explanatory power for the case of the EP. On the basis of my findings, there is hardly any 'puzzle' left: the ideological preferences of the Central and Eastern members simply fit well those of the old members, and voting cohesion follows. The Importance of ideological preferences was further confirmed by the analysis of competition in the enlarged EP party system. After and before enlargement alike, ideological differences between the political groups explain different voting outcomes. In synthesis, this book has further demonstrated that ideological preferences 'matter', and largely explain party behaviour in the EP.
Staying with voting cohesion, a theme that this book has touched only in passim regards the 'socialization' of the new members in the EP. Party delegations from Central and Eastern Europe appear to become more loyal to their political group as the legislature unfolds. The adaptation of the new members could be an extremely interesting laboratory in which to test theories of socialization (see Lindstaedt et al, 2012). Do the newcomers adjust to existing norms of behaviour because they deem appropriate to do so? Or do they follow a strategy of rational adaptation to be soon recognized as 'equals' and increase their leverage in the Parliament?
Finally, one promising avenue for further research, which Chapter 8 of this book has just briefly mentioned in its conclusions, is the impact of the informalization of the co-decision procedure on party competition. Between 1999 and 2009, according to the figures presented by Reh et al. (2013, p. 3), first reading agreements raised from 22 per cent of the co-decision acts adopted in 1999 to 86 per cent in 2009. Conversely, third reading agreements dropped from 22 per cent of all co-decision dossiers in the 1999-2004 legislature to only 5 per cent in the later period (2004-09). Co-decision has become increasingly
168 Europarties after Enlargement
informal, and agreements are found through 'trialogues', closed-doors meetings between a delegation of the EP, the Council presidency and the Commission. Enlargement is one of the key factors behind the increase of informal agreements: reasons of efficiency and time-saving motivations pushed the EU legislators to 'go informal' (Reh et al., 2013). Yet, the informalization of co-decision has not been systematically associated with patterns of competition/collusion in the EP party system. Does the absence of public scrutiny increase cross-bench agreements among the political groups? To what extent would the need to secure a winning majority in the EP compel the two major groups to side together, having struck an agreement with the Council before? Clearly, these and other questions beg for more empirical research to be answered.
3. Beyond the 2014 EP elections
While this book demonstrates that the Europarties have further institutionalized after and, in part, because of enlargement, the relationship between their EU 'faces' remains somewhat unbalanced. In the parliamentary arena, the political groups decide on the allocation of the main office positions and are the key policy-makers. They have strong organizations designed to maximize consensus and smooth out disagreements among their members. Moreover, the party system in the EP is competitive: the major parliamentary parties oppose each other on left-right policies, thus seeking to affect the content of the EU legislation.
Contrariwise, the role of the Europarties in the electoral arena remains, at best, embryonic. Rather than genuine European elections, the EP elections still represent the aggregation of 28 separate national elections, contested by the national parties, within a regulatory framework established by national legislation.2 It is common wisdom that EP election campaigns are fought on national issues and priorities, and that EU themes are hardly mentioned by the national parties (Reif and Schmitt, 1980). Moreover, the visibility of the Europarties in the election campaign is little or none beyond the limited Brussels circles.
Yet, against this background, this book demonstrated that conditions for a more active role of the Europarties in the electoral arena exist. The Europarties are not mere aggregates of heterogeneous national parties, but ideologically cohesive organizations. Transnational campaigning on the basis of common manifestos could, in theory at least, take place. How could the current Brussels-introverted organizations be closer to the European citizens?
Conclusions: Europarties' Prospects beyond the 2014 EP Elections   169
Below, I critically discuss the most important institutional and political reforms recently introduced (or, in some cases, already implemented) in Brussels, with the stated purpose to strengthen the Europarties by granting them more visibility and a clearer role in the EU political system and, distinctively, in the electoral arena.3
1. Transnational Lists. A proposal tabled by the British Liberal MEP Andrew Duff4 asks for a modification of the 1976 Election Act by introducing a transnational list to be elected in a single EU-wide constituency. Twenty-five additional MEPs would be chosen from a single list directly managed by the Europarties. Concretely, each European citizen would be granted two votes in the EP elections: a first one for the national or regional party, as it is now, and a second one for the transnational list. In Duff's words, the Europarties would be transformed, by directly managing this list, into 'real campaign organizations'.
The proposal is fully inscribed into the federalist tradition wishing a single electoral constituency for the whole of Europe. It would certainly represent a breakthrough in the system against the monopoly of the national parties in candidate selection. Furthermore, it would also make the Europarties much more visible to European citizens.
Nonetheless, there were a number of difficulties that made the introduction of a transnational list for the 2014 EP elections impossible. The most important one had to do with the opposition of a large number of MEPs who, backed by their national parties, resisted the setting up of a parallel - albeit limited - channel of recruitment for the parliamentary elites. Ultimately, in July 2011, the Duff proposal was sent back to the Constitutional Affairs Committee for further consideration, given the high likelihood of rejection in the plenary of the EP.
Moreover, Duff's proposal prescribes the setting up of a 'closed list', whose management would be a task for the Europarties. However, deciding who will top the list is likely not to be a banal choice. How would the Europarties decide? Would they open a broad consultation with civil society, leave the matter to the congress (i.e. to the member parties), or something else?
The introduction of a transnational list is certainly a brave innovation but, even in the very unlikely case that the EP and the Council would approve it, a number of significant problems will still need to be tackled.
2. Legal Personality. A new regulation on the statute and funding of European political parties would attribute legal personality according to EU law to the Europarties and their associated political foundations.5 As a report tabled by MEP Giannokou puts it,6 'strengthening European
170 Europarties after Enlargement
political parties is a means of enhancing participatory governance in the EU and finally strengthening democracy'.7 In order to achieve this goal, both the report and the new regulation posit that an important step would be granting legal personality to the Europarties.
Indeed, the legal status of the European political parties - in this book's lexicon, the EU extra-parliamentary parties - is at present equivalent to any NGOs or pressure group registered in Belgium. They are generally recognized as 'international non-profit association' by the Belgian law, having their headquarters in Brussels. The so-called 'Europarty Statute' (Regulation 2004/2003) failed to grant an EU legal status to the Europarties, making them de facto lesser actors in the EU institutional architecture. This is what the new Europarty Statute aims to address.
The full legal recognition of the Europarties would certainly enhance their status in the EU, besides having obvious advantages, in terms of salary and job security, for their employees. However, it is hard to imagine what more substantial consequences this reform will bring. As a result of its adoption, it seems unrealistic to expect the distance between parties and citizens to be lessened. While granting to the Europarties a stronger status in Brussels might be a task worth pursuing, in order to connect citizens and the EU institutions it is definitely too small a step.
3. Individual Membership, Currently, the Europarties are parties of parties, without rank-and-file members. Recently, however, the debate over providing some form of direct involvement to party activists has gained traction. In general, the extra-parliamentary parties have resisted the direct involvement of ordinary citizens, although they have, at times, devised alternative forms of participation. For instance, the PES has institutionalized the role of 'party activist', participating in electoral campaigning, debating on European politics and making the PES voice heard at the grass-roots level. The ALDE has also recently introduced 'associate members', individual members without voting rights.
The falling turnout in the EP elections, together with the relatively large amount of public funding that the extra-parliamentary parties receive, is calling for stronger citizens' involvement in the Europarties' activities. But the introduction of new modes of participation - as for the PES activists or the ALDE associate members - amounts to little if political decisions (i.e. decisions taken by the party congress) exclude the rank-and-file members. In this regard, individual membership, with full voting rights, would be a more effective way to strengthen citizens' involvement.
Conclusions: Europarties' Prospects beyond the 2014 EP Electiom 171
Nonetheless, the arguments for resisting the introduction of individual membership are not trivial. How could individual membership be regulated? Which voting rights would members be granted? It is a widespread fear among the Europarty leaders that individual membership might be used for tactical reasons: what if a party faction uses the channel of individual membership to oppose the official political line of a member party? Would the Europarties be transformed into a battlefield for regulating domestic problems? Taking everything into consideration, then, it is unlikely that individual membership would, in the short ran at least, be introduced.
4. The Citizens' Initiative. The Lisbon Treaty (art. 11.4), drawing literally from the aborted Constitutional Treaty, introduced the so-called 'Citizens' Initiative'. For the first time, citizens of the 28 member states have been given the possibility to directly ask the Commission to initiate legislation on matters where legislative action is considered to be necessary for the purpose of implementing the treaties. The Citizens' Initiative requires 1 million of signatures, collected in at least a quarter of the member states.8 Once the collection is completed and all formalities are met, the Commission has the obligation to duly consider the initiative and communicate its conclusions and actions it plans to take (if any) within three months. The Commission is not obliged to follow up with a legislative proposal, but needs to motivate its action.
The regulation disciplining the initiative states that 'entities, notably organizations which under the Treaties contribute to forming political awareness and to expressing the will of the citizens of the Union, should be able to promote a citizens' initiative' (art. 9 of Regulation 211/2011). Even if the Europarties are not explicitly mentioned by this regulation, the Citizens' Initiative appears to be an important stimulus to promote their engagement with civil society. Indeed, Bouza Garzia and Greenwood (2012, p. 252) recently commented that 'a measure of formalized organization and resources will be necessary to gather the necessary signatures', and the Europarties, together with their national member parties, might be able to supply them. The initiative offers them a concrete possibility to engage with policy issues and might increase their visibility among European citizens (see also Rose, 2010, pp. 19-20).
Nonetheless, on mapping the initiatives for which the collection of signatures is either currently open or has been closed from the Commission's 'official registry',9 it is worth noticing that in a single case only - the initiative for 'suspension of the EU Climate and Energy Package' - a political group in the EP is listed among the sponsors, giving
172 Europarties after Enlargement
funds to support the initiative. Ironically, it is the Eurosceptic EFD, asking for a suspension of an allegedly 'ineffective' EU legislation. Although this is not the only way that the Europarties have to support an initiative (they and their members could advertise and endorse it, they could offer logistic support,...), their unwillingness to financially support the existing initiatives is rather surprising. While the reasons for this lack of engagement need to be better investigated, civil society organizations and citizens have started to use this instrument largely without the active involvement of the Europarties.
5. Electing the Commission resident. Arguably, the most important innovation which could be implemented without further reforming the treaties - which is very unlikely after more than a decade of constitutional fatigue - is by explicitly linking the EP elections with the choice of the Commission president (for instance, Hix, 2008; Bardi era/., 2010, pp. 100-01). The current institutional setting already provides very favourable conditions for this development. Indeed, in its Article 17.7, the Lisbon Treaty reads: 'Taking into account the elections to the European Parliament and after having held the appropriate consultations, the European Council, acting by qualified majority, shall propose to the European Parliament a candidate for President of the Commission.'
Hence, the Lisbon Treaty suggests, even if it does not compel, the Council to take into account the EP election results. It is then very unlikely that if (say) the People's Party obtains a majority of seats in the new legislature, the Council would choose a Socialist as president of the Commission - not the least because the Parliament might reject its nominee.
If the Europarties and, in particular, the two largest families at the EU level - the EPP and the PES - choose a candidate ahead of the EP elections and campaign throughout the EU for their candidate on the basis of a common programme, then citizens would better know why they are voting and why their vote matters. Indeed, everywhere in Europe the most important elections are linked with the formation of government. Although the college of commissioners would still be formed by a large coalition of Europarties - commissioners are selected by the national governments - the partisan outlook of the Commission would be much enhanced. What is more, real and direct competition among the Europarties would be introduced into the electoral arena.
In this respect, the 2014 EP elections represented a first, but promising, start. In between November 2013 and March 2014, all Europarties (except the Eurosceptics) completed their nomination process for the
Conclusions: Europarties'Prospects beyond the 2014 EP Elections 173
president of the European Commission. Some prominent politicians -among them, Jean-Claude Juncker for the EPP, Martin Schulz for the PES and Guy Verhofstadt for the ALDE - were selected to run as the Europarty candidates for the top executive job in the EU. While the 'second-order' nature of the EP elections poses formidable obstacles to transnational campaigning, the 'break-through' of the Europarties in the electoral arena is a very significant development whose importance should not be underscored.
In conclusion, the development of political parties and of a party system at the EU level has not been undennined by enlargement and the inclusion of the ten post-communist members. Modest and realist proposals of reform - under the current institutional framework of the EU - could boost the role of the Europarties in the electoral arena. With a clearer linkage between the European citizens and the EU institutions, the process of political representation in the Union could gradually become a reality, and the Europarties' expressive function will be more than a felicitous outcome produced by the aggregation of the national parties and Party systems (see also Mair and Thomassen, 2010). The economic and financial crisis has made the EU a far more salient issue in the lives of many ordinary citizens: in the absence of a clear representative linkage, the legitimacy of the Union will be increasingly called into question by non-mainstream parties, attracting ever-growing shares of votes from the dissatisfied Europeans. This might be too high a risk to be run by the EU - and by its mainstream parties.