Abstract
This article is part of the special cluster titled Parties and Democratic Linkage in Post-Communist Europe, guest edited by Lori Thorlakson, and will be published in the August 2018 issue of EEPS
Various forms of inter-party cooperation have important effects on party system fragmentation and stability in young democracies. However, the conceptualisation and measurement of these forms of inter-party cooperation and the examination of their consequences on party system development remain limited in the literature on parties and party systems. This research addresses this gap in the scholarship in three ways. First, we present the analytical scheme of different types of party cooperation. We argue that the forms of inter-party cooperation vary on two dimensions. The first dimension refers to their structural basis: the stability of the cooperation as captured by whether it is rule-based or, in other words, underpinned by shared rules that are mutually accepted. The second dimension refers to their scope: the number of functional areas of party life subject to cooperation. The two dimensions lead us to four basic forms of inter-party cooperation: (1) non-rule-based, functionally restricted coalitions; (2) rule-based, functionally restricted coalitions; (3) non-rule-based organization-wide mergers; and (4) rule-based organization-wide mergers. Second, we develop theoretical expectations on the frequency of these forms of inter-party cooperation in Central and Eastern Europe. Third, to test these expectations, we present empirical evidence on the number of electoral coalitions and mergers in the first six electoral periods in 10 countries in the region. The results of the analyses support our expectations: non-rule-based organization-wide mergers are rare. The other three forms of party cooperation (non-rule-based coalitions; rule-based coalitions; rule-based mergers) are fairly common in most countries in the region, although less so in the more recent electoral periods.
Introduction
The shape and dynamics of the party system is one of the most important characteristics of a representative democracy. Although there is some disagreement among scholars and practitioners on what levels of fragmentation and stability are most desirable, it is generally agreed that excessive fragmentation and instability can undermine the quality of democracy, and in some cases may even lead to its erosion or downfall. This is a pertinent issue for many young democracies, including those in Central and Eastern Europe (CEE), where the number of parties and party system instability (as measured by the levels of electoral volatility) remain high even after more than two decades of democratic experience. 1 The literature suggests that this pattern of party system development is largely a consequence of the behaviour of political elites that do not coalesce around the stable set of political parties. 2 Consequently, the emergence and electoral success of new parties has been extensively studied in the scholarship on party politics in the region. 3
However, political elites also change party supply through cooperation, such as electoral coalitions and party mergers. While the prevalence and causes of these forms of inter-party cooperation have been studied in several recent publications, 4 existing literature does not discuss in sufficient detail how they are similar and different from each other—or other forms of inter-party cooperation (e.g., government coalitions). Furthermore, different types of electoral coalitions and mergers have also not been clearly conceptualised. This is an important gap because careful conceptualisation and measurement of different forms of inter-party cooperation is important for understanding the development of party systems. For example, the complexity and diversity in the forms of inter-party cooperation even prevents the analysts from establishing clearly the number of parties in the region. 5 Moreover, while mergers and electoral coalitions may be seen as an important mechanism leading to the emergence of stable and moderately fragmented party system, 6 Kreuzer and Pettai qualify this argument suggesting that the effect of these forms of inter-party cooperation depends on whether cooperation is durable (as well as the size of parties involved in cooperation). 7 Understanding when inter-party cooperation emerges and lasts requires the systematic conceptualisation and measurement of the organizational decisions underpinning cooperation.
This research contributes to addressing this gap in the scholarship in three ways. First, we present the analytical scheme of different types of party cooperation including electoral coalitions and mergers. Second, we develop theoretical arguments about the costs and benefits of electoral coalitions and mergers for political parties. This allows us to develop expectations on the frequency of these forms of inter-party cooperation in the context of relatively young democracies in Central and Eastern Europe. Third, to test these expectations. we present empirical evidence on the number of electoral coalitions and mergers in the first six electoral periods in ten countries in the region.
We argue that the forms of inter-party cooperation vary on two dimensions. The first dimension refers to their structural basis, that is, the stability of the cooperation as captured by whether it is rule-based—underpinned by shared rules that are mutually accepted—or not. The second dimension refers to their scope, distinguishing cooperation that is functionally restricted, that is, ensures cooperation in some areas (e.g., electoral competition) but maintains separate organizational identities of the constituent parties, from cooperation that implies a complete fusion of two or more separate party (membership) organizations across all core functional areas of party life. In this latter configuration, the constituent parties, as separate units, cease to exist. The two dimensions lead us to four basic forms of inter-party cooperation: non-rule-based, functionally restricted coalitions (e.g., electoral coalitions if functional cooperation refers to the joint running of individual elections, 8 negotiated ad hoc prior to specific elections), rule-based, functionally restricted coalitions (e.g., electoral coalitions if the putting together of joint lists is organized through established rules and structures), non-rule-based organization-wide mergers and finally rule-based organization-wide mergers.
These different forms of inter-party cooperation that are rarely explicitly distinguished are likely to shape party systems and their dynamics in different ways and have severe repercussions for the functioning of democracy. The greater the number of functional areas (e.g., running elections, governing) in which parties cooperate, and the more rule-based their cooperation is, the more the constituent parties are able to function as a single party in terms of ensuring internal cohesion in legislative and executive decision making. In essence, legislative party system fragmentation depends on the way in which parties cooperate, which is why we need to move beyond a conceptualization of parties as either completely separate actors that compete in multiple areas, or, as the opposite extreme, fully merged parties that do not compete at all.
The next step in the article involves developing theoretical expectations on the prevalence of these four forms of inter-party cooperation in CEE democracies. While electoral coalitions and mergers do occur in established democracies, their electoral and office benefits to parties tend to be higher in younger democracies where a number of parties have not established (or lost) an independent capacity to win legislative representation. Previous research also shows that electoral coalitions and mergers are more prevalent in younger CEE democracies as opposed to older democratic systems in Western Europe. 9 Given greater importance of electoral coalitions and mergers in CEE countries, developing theoretical expectations on the empirical prevalence (and change in it over time) of different types of coalitions and mergers in this region is a pertinent question. We address it by mapping out various costs and benefits of (non-)rule-based coalitions and mergers for political parties.
Our empirical analysis provides a test of these expectations by investigating the empirical frequency of four forms of inter-party cooperation in ten CEE countries in the period between 1990 and 2014. The results of the analyses support our expectations: non-rule-based organization-wide mergers are rare. The other three forms of party cooperation (non-rule-based coalitions; rule-based coalitions; rule-based mergers) are fairly common in most countries in the region, although less so in the more recent electoral periods. Furthermore, non-rule-based coalitions are somewhat more frequent than rule-based coalitions and mergers. Empirical analysis also demonstrates that the electoral strength of parties that engage in these three forms of inter-party cooperation is quite substantial. Thus, it is likely that they have important implications for party system development. The concluding section, besides summarising our findings, discusses these implications in greater detail.
Four Forms of Inter-party Cooperation: An Analytical Map
In this section, we distinguish four different forms of inter-party cooperation. We understand party cooperation as a continuum. One end of this continuum is represented by completely separate party organizations that only interact ad hoc, on a temporary basis and whose interaction is not structured by any commonly agreed rules. Thus, cooperation is absent if parties do not in any way work together—not even temporarily—towards the pursuit of a shared goal. The other end of the cooperation continuum is represented by the fully merged new organization that is integrated across all areas of party organizational life. More specifically, inter-party cooperation varies on two dimensions: the structural basis of cooperation (non-rule based / ad hoc cooperation vs. rule-based cooperation) and the scope of cooperation as regards to core areas of party life.
The first dimension captures the presence or absence of mutually agreed and accepted (sets of) rules that structure the cooperation in the medium and long term, thereby enhancing the stability of the cooperation. The second dimension distinguishes cooperation that is functionally restricted to specific areas only (e.g., electoral competition, through the formation of joint lists, or within parliaments, through joint parliamentary groups), while in other areas (e.g., member recruitment) parties still compete. Importantly, in whatever area we find functionally restricted cooperation, the separate organizational identities of the constituent parties remain intact. Such functional cooperation is distinct from organization-wide cooperation that implies a complete fusion of two or more separate party (membership) organizations into one.
Combining the two dimensions, this leads to four forms of inter-party cooperation: (1) functionally restricted, non-rule-based coalitions; (2) functionally restricted, rule-based coalitions, (3) organization-wide, non-rule-based merger parties; and (4) organization-wide, rule-based merger parties. Table 1 sums up the four categories.
Forms of Inter-party Cooperation
Before explaining these categories in greater detail, it is important to discuss how they relate to existing definitions of merger and pre-electoral coalition in cross-national research. Looking at merger parties as one of several types of new formations (next to splinter parties and newly born parties built from scratch), mergers (or in Mair’s case “fusion”) denote a minimum of two organizations forming a new one, 10 which constitutes one end of our continuum of inter-party cooperation. Lijphart discusses five cases of closely allied parties that form highly stable forms of cooperation falling short of creating a unified party organization (CDU-CSU in Germany, the Liberal and National parties in Australia, and the pairs of Christian Democratic, Social Democratic, and Liberal parties in Belgium), and concludes that they are “all genuinely somewhere in between two parties and one party.” 11 However, he does not distinguish between qualitatively distinct forms of inter-party cooperation that are neither full mergers nor mere ad hoc coalitions. Golder preferred the concept of the “pre-electoral coalition” in her work, defined “as a collection of parties that do not compete independently in an election either because they publicly agree to coordinate their campaigns, run joint candidates or joint lists, or govern together following the election.” 12 Both strands share a preference for one concept over the other and orient their empirical work towards it, instead of looking at them either as alternative forms of inter-party cooperation or as forms of cooperation that evolve in sequence to each other, that is, pre-electoral coalitions as a first step that can facilitate a merger. 13 Golder’s definition covers a variety of ways to cooperate, that is, is inclusive as to the purpose of the coalition but presupposes that we have a collection of separate parties to start with that give up their independence in the context of an election. As such, her definition corresponds to our category of non-rule-based functional coalition. Her definition leaves open whether, if independent competition is given up in the longer term and such cooperation is open-ended and rule-based (rule-based-functional), it would still be captured by her concept of pre-electoral coalition, because, in fact, the coalition would remain intact after the election as well. 14 The definition of merger as a fusion of several organizations answers the question. It indicates that even if functionally restricted cooperation is rule-based, we do not speak of the formation of a “merger party” but of a coalition of two still separate organizations.
Non-rule-based coalitions are regularly revisited/renegotiated. For example, the constitution of shared lists is renegotiated before every election and the return to inter-party competition is regularly considered as an option. Similarly, a coalition agreement before one election is reconsidered at the next. Thus, we need to look at the establishment of shared rules in specific functional areas (e.g., the nomination of candidates on shared lists) to differentiate between non-rule- and rule-based coalitions. While cooperation can fail at any point for a variety of reasons, rule-based cooperation is in principle open-ended. However, in both cases cooperation is restricted to a particular purpose or area of party life and does not encompass the organizations as a whole. The constitutive parties involved in the cooperation arrangement remain intact as separate organizational units. Organization-wide cooperation presupposes the dissolution of constitutive units in a shared infrastructure (including a fusion of the membership organization), which presupposes cooperation that is not restricted to certain areas such as electoral competition (e.g., a committee to put together shared candidate lists) or parliamentary decision making. This is when we talk about a merger party. Yet, in a similar way as a coalition, a merger party can be rule-based or not. That is, the presence of cooperation across all core areas that defines a merger (organization-wide cooperation) does not necessarily have to be rule-based across all these areas, which underpins our distinction between non-rule-based mergers (which are less likely to be stable) and fully rule-based mergers (which are more likely to be stable).
Distinguishing Forms of Cooperation: How to Capture Their Scope and Nature
The classification of four types of party cooperation is—at first glance—relatively straightforward. The allocation of actual cases to the categories, however, can be challenging. To identify the scope of cooperation, we need to identify the core areas of party activity to start with. To gain a sense whether an arrangement is rule-based we need to gain a sense whether and which rules are followed in core areas of party life. Katz and Mair differentiate between three faces of parties: the party in public office, the party in central office, and the party on the ground (i.e., the party membership organization). 15 Naturally, depending on the party we look at, these different “faces” might have different weight and relevance. While many functional forms of cooperation have implications for more than one face, since the three are inevitably intertwined, 16 the trio allows us to locate where functionally restricted cooperation can take place and in which areas we need to find cooperation to speak of merger (i.e. organization-wide cooperation). At the same time, while the formation of a merger, in practice, presupposes the adoption of common rules in some core areas (so far we discussed [1] the presence and absence of cooperation (scope) and [2] whether cooperation is structured by rules or not), the nature of these rules give us an indication of the commitment of the constituent parties to the shared arrangement. Furthermore, the nature of the common rules also indicates how likely the merger is to be stable. The nature of these rules can be characterized in terms of the extent to which they assure the integration of core operations within the three units of the party.
Functionally restricted cooperation (see grey shadings in Table 2) usually focuses on areas that relate to the party’s role in public office (parliament or government) or its endeavour to gain access to it (elections). Policy-coordination might be perceived as a core organizational matter, important to the rank-and-file. However, party policy is usually formulated by elites, especially public representatives (even in relatively participatory parties) and its coordination is often central to party elites’ efforts to ensure meaningful cooperation in elections, parliament, or government. We therefore subsume this “area of coordination” under “the party in public office.” Organization-wide cooperation covers these areas but also encompasses intra-organizational matters that relate to either the party on the ground (e.g., the regulation of membership or the organization of territorial or functional subunits) or the party in central office (e.g., collective leadership arrangements). As long as two parties still maintain separate leaderships and separate organizations (non-shaded categories in Table 2), even if cooperation was rule-based in all four “functional” areas, we would consider the arrangement as a coalition, not a merger, since cooperation remains restricted in scope. 17
Mapping the Scope of Rule Integration Across Functional Areas of Party Life
To sum up the analytical rationale of Table 2, we would speak of organization-wide cooperation—a merger party—when finding at least some cooperation (rule-based or not) across all eight functional areas (and thereby across all three “faces of a party”). For instance, we would consider the Union of Democratic Forces (UDF) in Bulgaria as a merged party only since 1995, when the individual membership in both the constituent parties and the UDF as a whole was permitted. 18 This is the case despite that until then the constituent parties cooperated in most other areas outlined in Table 2; specifically, they presented joint candidate lists in elections, formed a unified parliamentary group in the Bulgarian legislature, and participated in the national government together; the coalition also had the permanent leader and statutes. In contrast, following Table 2, the Solidary Electoral Action (AWS) in Poland has never become a merged party because some of its constituent parties did not integrate their membership organizations. 19 Non-rule-based mergers are also likely to exist in the early stages of the formation of merged parties, when constituent parties may not yet have been able to achieve full rule integration across all areas of party life. For example, integrating local organizations might take several years after the formal merger (often a joint congress of constituent parties) took place.
Sometimes partial integration is an intermediate step to build up trust before adopting fully integrated rules. Such sequencing may be planned in advance (parties may consider that partial integration is required before full integration can be achieved) or as a natural course of cooperation (e.g. the successes and/or limitations of partial integration induce full integration). Similarly, often cooperation starts out in some areas and is broadened successively, moving a cooperation arrangement from a functionally restricted coalition to a merger.
Frequency of Inter-party Cooperation in CEE: Theoretical Expectations
Having presented our classification of different forms of inter-party cooperation, in this section we develop theoretical expectations on the frequency of four forms of inter-party cooperation (as listed in Table 1): non-rule-based electoral coalitions, rule-based electoral coalitions, non-rule-based mergers, and rule-based mergers. Both electoral coalitions and mergers are under-studied forms of inter-party cooperation (in comparison to, e.g., government coalitions). Their relevance for party system development and empirical frequency justify our focus on them in this study. Nevertheless, arguments below can be expanded to other forms of inter-party cooperation, such as legislative or government coalitions.
Our theoretical expectations build on the still limited literature on electoral coalitions and mergers as well as the research on government coalitions, 20 in particular the strand of it focusing on the formation of coalition agreements between parties. 21 Both bodies of research suggest that parties’ decisions to form coalitions and mergers depend on their benefits and costs. Developing the arguments in these studies further, in Table 3 we present expected benefits and costs for each of the four types of inter-party cooperation, which in turn allows us to develop theoretical expectations on their frequency.
Expected Costs and Benefits of Coalitions and Mergers in CEE Countries
The benefits of electoral coalitions and mergers may include higher chances of winning electoral and governmental office. 22 While they vary substantially depending on such factors as party size and ideological or policy proximity, in most contexts there are potential electoral coalitions and mergers that could provide electoral and/or office benefits to their members. 23 This is particularly the case in younger democracies in Central and Eastern Europe. Many smaller parties in the region border the threshold of electoral viability, thus providing very strong incentives to cooperate with other parties to win legislative representation. The structure of party competition over government formation is relatively unstable, which creates incentives also for larger parties to participate in electoral coalitions and mergers to increase their chances of the participation in government (even if for larger parties expected benefits are likely to be lower than for small, marginally viable parties). 24 The evidence of numerous electoral coalitions and mergers in the region also suggests that they can provide substantial electoral and office benefits for parties. 25
The absence of systematic evidence on the electoral success of electoral coalitions prevents us from making strong arguments on whether parties expect to obtain higher benefits from rule-based or non-rule-based variants of these coalitions. It is possible that rule-based coalitions are more effective in mobilising voter support due to the more developed mechanisms of cooperation. However, such coalitions may also be less flexible in responding to voter demands than non-rule-based coalitions. In a similar vein, rule-based mergers might be more or less beneficial to parties (in comparison to electoral coalitions). At the same time, there are strong grounds for arguing that parties expect substantially lower electoral and office benefits from non-rule-based mergers than from the other three forms of cooperation. Specifically, merged parties based on informal cooperation and/or partial rule integration across all eight functional areas listed in Table 2 are likely to suffer from organizational dysfunction that would prevent them from competing effectively with other parties in electoral, parliamentary, and governmental arenas.
While coalitions and mergers may benefit parties, they also entail several types of costs. First, parties have to compromise on their ideological and policy positions. Even if parties in CEE countries are described as vote- and office-seekers, ideological proximity is an important explanation of the formation of both electoral coalitions and mergers in the region. 26 Ideological and policy differences may also reduce the expected electoral benefits of cooperation as voters are less likely to support ideologically heterogeneous formations. The costs of ideological and policy compromises are highest for rule-based mergers, which demands constituent parties to largely give up their ideological identities. On the other extreme, non-rule-based electoral coalitions permit constituent parties to preserve a substantial degree of ideological and policy autonomy. Parties campaign largely autonomously, thus allowing them to focus on their own ideological and policy offer; and they do not commit to cooperation in the inter-election period or for the elections in the future. The other two types of cooperation impose moderate to high ideological costs. While constituent parties have to make substantial compromises by, for example, agreeing to coordinate closely their election campaigns (rule-based coalitions) or having to reach ideological and policy agreements across all functional areas, these costs still are lower in comparison to full-blown rule-based mergers.
Second, reaching and implementing agreements is costly. Parties need resources (such as time and personnel) to negotiate agreements and implement them. They may also experience audience and reputational costs from making concessions to partners. 27 All of these costs are likely to be higher for agreements with broader scope (i.e., covering more functional areas) and rule-based structural basis (as opposed to non-rule-based agreements). Thus, we argue that decision and implementation costs are highest for rule-based mergers and lowest for non-rule-based electoral coalitions while rule-based electoral coalitions and non-rule-based mergers are likely to fall in between these two extremes.
Third, inter-party cooperation also raises the issue of credible commitments. Opportunism by potential partners due to the uncertainty over the future decisions (and the circumstances they would have to be made under) is a key explanation for parties making formalised government coalition agreements. 28 Similarly, electoral coalitions and mergers can also be marred by the problems of opportunistic behaviour. Smaller and/or weaker partners may exit the cooperative arrangement if they feel that they may be able to survive on their own (or if a more favourable cooperation opportunity appears on the horizon). 29 Such opportunism could hurt their larger partners by, for example, reducing their chances to win governmental office. Larger and/or stronger partners on the other hand may use their size advantage by reneging on the terms of agreement (e.g., by taking the higher share of electoral and office benefits than originally agreed).
Narrower scope of agreements and the ad hoc/informal basis of cooperation make terminating cooperation easier and therefore decrease the credibility of commitments by smaller/weaker parties. Conversely, forms of inter-party cooperation with broader scope and higher rule integration make commitments by larger/stronger parties less credible. If these parties renege on their earlier commitments, their smaller/weaker partners are less likely to be able to exit.
We expect that the first type of commitment problems is more important. Given that the political careers of the elites of smaller/weaker parties are often dependent on such agreements, the expected benefits of cooperation in the short term would make the credibility of commitments from their partners of secondary importance. This in turn implies that the costs of credible commitment are highest in the case of non-rule-based electoral coalitions, lowest for rule-based mergers and moderate for the other two forms of inter-party cooperation.
Finally, forms of inter-party cooperation also differ with regard to legal costs. Most CEE countries are characterised by substantial regulation of political parties and electoral process. 30 This regulation varies across individual countries. For example, several countries formally prohibit electoral coalitions or apply very high electoral thresholds. 31 Similarly, the regulation on party mergers, while less examined in the scholarly literature, is likely to vary across countries. Most importantly, however, all countries have rules on how parties should be organised internally, which might increase the costs of running unwieldy non-rule-based mergers. For example, laws may stipulate individual membership through either the constituent parties or the merger party. To comply with legal regulations, such non-rule-based merger entities might be pushed towards further rule integration in the functional areas related to the party on the ground.
The summary of the benefits and costs of the four types of inter-party cooperation suggests quite clear implications on their frequency. Non-rule-based mergers are least likely, mainly because their expected benefits are low and legal costs are likely to be relatively high. In contrast, the other three forms of inter-party cooperation may provide substantial benefits and score similarly in terms of costs (even if they differ with regard to the type of costs). Still, non-rule-based electoral coalitions, by imposing only moderate constraints on parties’ ideological and policy identity, are overall somewhat less costly than rule-based coalitions and mergers. Thus, our first key expectation is that that non-rule-based electoral coalitions will be somewhat more frequent than rule-based coalitions and mergers, and all three of these forms of cooperation will be much more frequent than non-rule-based mergers.
The frequency of inter-party cooperation is also likely to change over time given that the costs and benefits of inter-party cooperation change as democratic systems mature. While the benefits of non-rule-based mergers are likely to remain low regardless of democratic age, for the other three forms of cooperation they are likely to decrease. At least some parties establish electoral viability, making cooperation with other parties less necessary. Furthermore, as parties age, they are to develop policy reputations, which makes entering cooperation with parties representing different policy or ideological positions riskier electorally. Similarly, clearer ideological identities increase the ideological costs of cooperation, especially for the ones with higher degree of rule integration (rule-based mergers and electoral coalitions). Non-rule-based electoral coalitions, however, remain a more attractive option given that constituent parties are better placed to preserve their ideological identity in such cooperation arrangements. Last but not least, legal costs of complying with state regulation are also likely to increase given that many countries in the region have over time strengthened the extent to which they regulate parties and the electoral process. Thus, we expect that the number of non-rule-based electoral coalitions will decline moderately, the frequency of rule-based coalitions and mergers will decrease substantially, and the number of non-rule-based mergers remains stable at a low level.
The Empirical Study of Inter-party Cooperation: Electoral Coalitions and Mergers
Operationalization and Data
To capture distinct forms of inter-party cooperation empirically, we measure the two dimensions—scope and structural basis of cooperation (Table 1)—in the following manner. To capture forms of cooperation that vary in the given scope of cooperation, we distinguish between (functionally restricted) electoral coalitions and organization-wide mergers, based on whether a cooperation arrangement involved the full incorporation of the organizations of constituent parties (including the integration of local party organizations and of party leaderships) or whether it was restricted to the formation of joint lists and candidates (see Table 2). 32
As a proxy for the structural basis of cooperation arrangements, we measure their stability indicating the extent to which a cooperation arrangement is likely to be based on shared rules (stabilizing the arrangement), assuming that more rule-based cooperation (be it coalitions or mergers) is less likely to be terminated by their constituent parties. More specifically, for electoral coalitions we examine the stability of their party composition in two consecutive elections. A stable electoral coalition is a coalition whose composition remains unchanged in two consecutive parliamentary elections. 33 An unstable coalition is intact for one election only. 34 We code a merger as unstable if within the time period covering the parliamentary term in which the merger occurred, as well as the next two parliamentary terms, the merged party experiences a split, and the splinter party can be traced back to one or several of the constituent parties that formed the merged party and as stable if not. This follows the definition and operationalisation of merger terminations by Bolleyer et al. 35 Thus, the instability of the merged party merger does not simply suggest the loss of support and resources through the defection of some (however constituted) faction or a range of individual defectors splitting off. It, more specifically, entails to a large extent the reinstatement of a situation of competition between the constituent parties that initially decided to go together. We deliberately chose different “stability thresholds” for the two forms, since while electoral coalitions are focused on the upcoming elections (even if they might be repeated later on), mergers are, in principle, directed towards assuring permanent integration. Furthermore, while only 8 electoral coalitions outlive two terms (which makes the survival of one inter-electoral period a sensible threshold to distinguish stable from unstable coalitions), only less than one in ten of the mergers in our sample did not outlive one term.
To pin down the broader relevance of these different forms of inter-party cooperation for party system fragmentation and stability, we measure their electoral support. Specifically, we use the share of the vote obtained by the electoral coalitions in the first of the two consecutive elections under examination. For stable mergers, we use the total share of the vote received by the constituent parties in the last election before a merger. For unstable mergers, we use the total share of the vote received by the constituent parties in the last election before the merger was terminated and, if the merger was terminated after one or two general elections, the vote share of the merged party in the last election before its termination.
Our analysis of party cooperation in ten CEE democracies focuses on, as mentioned above, the first six electoral periods in each country. Thus, in total our analysis covers sixty parliamentary terms in these countries. We identified electoral coalitions from electoral records (as provided by the Project on Political Transformation and the Electoral Process in Post-Communist Europe, 36 the Parlgov database and the websites of the electoral commissions of individual countries) on the basis of their names, which frequently combined the names of individual parties. The electoral labels of a number of other coalitions, however, differed from those of their member parties. Yet other coalitions ran under the name of one of their member parties. The coalitions in the latter two categories, as well as their member organizations, were identified on the basis of a number of sources, such as handbooks on political parties or various country-specific sources. 37 Cooperation arrangements that involved party cooperation in the inter-electoral period (e.g., through permanent bodies of representatives of member parties) were generally coded as coalitions provided that the members of these coalitions still had autonomous membership organizations, usually as a result of being legally registered as an independent entity. We included electoral coalitions with at least 1 percent of the vote in our analysis. Similarly, party mergers were identified by tracking for each pair of consecutive elections those parties that obtained at least 1 percent of the vote in the first election but did not run in the second one under the same name, and also the parties that did not participate in the first election but did run in the second one. Secondary sources on political parties in CEE countries were then examined in order to identify whether these discontinuous parties were a result of party mergers.
The Frequency and Importance of Electoral Coalitions and Mergers
The results of our empirical analysis broadly support our theoretical expectations on the frequency of different forms of inter-party cooperation. Table 4 demonstrates that in the ten CEE countries considered, the number of electoral coalitions that were dissolved or underwent significant changes by the next election (used as a measure of non-rule-based coalitions) was 1.0. This is somewhat higher than the frequency of stable coalitions (0.6) that capture rule-based coalitions and also the number of stable mergers (0.9) that we used to operationalise the number of stable mergers. Finally, the number of unstable mergers (a proxy for non-rule-based mergers) was only 0.1. The frequency of the four types of inter-party cooperation varies substantially in individual countries, but this is expected given that the costs and benefits of cooperation differ considerably.
Number and Stability of Electoral Coalitions and Mergers per Country
The empirical patterns of the change in the number of electoral coalitions and mergers also provide substantial support to our expectations (Figures 1 and 2). One important exception is that inter-party cooperation has been fairly limited in the first electoral period. This may be a consequence of the contexts with a large number of new parties and high levels of electoral uncertainty, thus creating incentives for individual parties to try to establish electoral viability on their own. Following this, we observe a substantial number of coalitions and mergers in the second, third and fourth electoral periods, and a moderate decline in the fifth and sixth periods. In line with our expectations, decline is more pronounced for stable electoral coalitions and mergers than for unstable coalitions. This evidence supports our expectation that programmatic and ideological stabilisation of parties and party systems reduces rule-based cooperation (both coalitions and mergers) by increasing their ideological and policy costs. 38 This is less of an issue for non-rule-based coalitions, which do not require parties to largely give up their individual ideological identities. Last but not least, the frequency of unstable mergers remains very low throughout the time period studied here. This again supports our theoretical expectations.
Number of stable and unstable electoral coalitions and mergers per electoral period
Vote share of stable and unstable electoral coalitions and mergers per electoral period
Very similar patterns emerge also when we analyse the electoral support of parties that were involved in different forms of cooperation. Table 5 demonstrates that the average vote share of unstable coalitions (15.4 percent) was moderately higher than that of stable coalitions (9.4 percent) or stable mergers (11.4 percent) and much higher than the support of unstable mergers (1.9 percent). The electoral significance of the first three forms of party cooperation was the highest in the second, third, and fourth periods, and has declined moderately after that. Still, throughout the time period considered here, unstable coalitions and stable coalitions and mergers were characterised by relatively high levels of electoral support. This emphasises their importance for understanding party system development in Central and Eastern Europe.
Vote Share of Stable and Unstable Electoral Coalitions and Mergers per Country
Conclusion
This article presented an analytical scheme of four different forms of inter-party cooperation and illustrates the prevalence of some of these forms of cooperation empirically for ten countries in Central and Eastern Europe. Specifically, we differentiate forms of cooperation based on two dimensions. The first dimension captures whether inter-party cooperation is functionally restricted to one or several areas of party activity (e.g., elections, parliament, government) or whether it encompasses all core areas of party life leading to a complete fusion of different organizations. The second dimension captures whether the cooperation between parties is based on commonly agreed and accepted rules or not. The classification therefore provides four types of inter-party cooperation: functionally restricted, non-rule-based coalitions; functionally restricted, rule-based coalitions; organization-wide, non-rule-based merger parties; and organization-wide, rule-based merger parties. These types of inter-party cooperation have different implications for party systems, especially in young democracies, where they tend to be fragmented and unstable.
In our empirical analysis, we capture the extent to which party cooperation is functionally restricted by differentiating between coalitions that are restricted to one arena only, cooperation in the electoral area (electoral coalitions), on the one hand and organization-wide mergers that lead to the complete dissolution of constituent parties on the other. For measuring the extent to which party cooperation is based on rules, we use the stability of the cooperation arrangement (i.e., electoral coalition and party merger) as a proxy. Consequently, as a second step, we seek to map out the prevalence and electoral importance of all four types of inter-party cooperation in Central Eastern Europe in the period between 1990 and 2014.
We find that the prevalence and electoral significance of three forms of inter-party cooperation (non-rule-based electoral coalitions, rule-based electoral coalitions, and rule-based mergers) are comparable to each other in CEE democracies, and each of these is more frequent and electorally significant than non-rule-based mergers. While we used the stability of electoral coalitions and mergers as a rather rough proxy for their structural basis, these results still support our theoretical expectations derived from the systematic analysis of the costs and benefits of these four forms of cooperation for parties involved. Specifically, non-rule-based mergers are unlikely to be beneficial and impose relatively high costs in terms of ideological compromises, reaching and implementing agreements, credible commitment of parties to these agreements, and compliance with the state regulation of parties and elections. The other three forms of cooperation, while also imposing some of these costs, are more likely to be beneficial to parties, thus explaining their greater empirical frequency. These results also imply that mergers, coming in predominantly rule-based format, are more likely to contribute to the stabilisation of party systems in CEE countries than electoral coalitions, as only some of these are stable (our proxy for them being rule-based). Last but not least, our findings showing the relative decline in the number of instances of inter-party cooperation over time provides some support to the accounts in the literature suggesting that instability in party organizations is characteristic to early stages of democratic development. 39
The large-N analysis presented in this article gave a valuable overview of two different forms of inter-party cooperation and their implications for party system development across a range of CEE democracies. This empirical analysis had—because of its scale—to inevitably restrict itself to measuring particular aspects within the analytical scheme presented rather than studying all forms of cooperation across all functional areas of party life as laid out earlier (Table 1 and 2). To make full use of the analytical map provided (Table 2), future qualitative research needs to look into the nature of cooperation arrangements parties engage in across the range of functional areas in greater detail (e.g., elections, parliament, government, organizational leadership). This would allow a comparison of the nature of rules adopted to ensure cooperation and the cooperation dynamics resulting from them within each of these areas (whether cooperation is set up as temporary or supposed to be lasting) and—similar to our comparison of electoral coalitions and mergers—explore how cooperation in one area might feed into another. This would be particularly useful as it would generate linkages between the literatures on party behaviour in electoral contests, in parliamentary settings and coalition government, each fruitful areas of work that do not speak enough to each other. Finally, returning to a systemic perspective, our findings of the prevalence of non-rule-based electoral coalitions, rule-based electoral coalitions, and rule-based mergers call more systematic research on how party coalitions and mergers affect party system fragmentation and volatility, as well as legislative party cohesion, compared to each other. Such analyses, similar to the study of the relationship between the number of new parties and electoral volatility, 40 could provide further insights into an important role that electoral coalitions and party mergers have during party system formation and stabilization.