PARTY TYPES AND ELECTORAL STABILITY IN CENTRAL AND EASTERN EUROPEAN STATES INTRODUCTION The new democratic regimes in developing countries face the severe challenge of constructing and maintaining strong and stable representative institutions. The states from Latin America provide evidence in this respect with a democratic regime durability that surpassed even the most of the expectations (Remmer 1992), the instability of parties and of the party systems represents an ongoing puzzle and a source of analysis (Mainwaring and Scully 1995). The direct consequence of this phenomenon is a volatile situation at the level of representation where political identities and organizational loyalties are recomposed every electoral cycle (Roberts and Wibbels 1999: 575). This research aims to establish a relationship between parties formal organization and activity and the behavioral aspects coming from voters. Starting from the metatheoretical question regarding the incentives for parties to develop stable and deep rooted structures, we narrow down the question to which we provide an answer: What is the impact of formal and structural party organization on voters attitudes? In this respect, we conceptualize party organization as types, reflecting also different aspects from institutional theories that include the environment, structure, actors, and the state. Consequently, we differentiate among five party types (elite, mass, ethnic, cartel and catch-all), proposed in previous research, and we hypothesize that the more organized and permanently active a party is, the lower the level of electoral volatility (the more stable its hardcore voters). In other words, a high level of formal organization and political activity is correlated with a constant acceptance coming from the population. Thus, we expect to find the mass and ethnic based parties among the most stable in terms of electoral volatility, cartel parties whereas the catch-all and elite parties to be the most volatile. For cartel parties it is difficult to have any plausible expectation based on the existing theoretical and empirical evidence. The cases included in the research are the Central and Eastern European (CEE) states: Bulgaria, Czech Republic, Hungary, Poland, Romania, and Slovakia. The primary units of analysis are the parties from these countries that got in the Parliament 11
in at least half of the elections 1. The time-frame of the analysis goes from the beginning of transition (1989-1990) until the last elections. We define the party formal institutionalization on two dimensions: systemness and autonomy (Panebianco s 1988, 55). Systemness is indicated by the degree of development of the central organization, the degree of homogeneity of organizational structures at the same hierarchical level, and the degree of correspondence between a party s statutory norms and its current power structure. The indicators for autonomy are party financing and the relation with external organizations. Compared to the alternative explanations provided in the literature, this research emphasizes the central role the party plays in shaping its electoral stability. This should not be regarded only as the result of general processes as democratization and economic performance or as a result of the electoral framework. In an unstable environment as the one in the CEE states in the first decade after the fall of Communism, party types may prove decisive in enhancing electoral stability of the party. If the expectations are correct, our research contributes to the institutionalization theory, by providing an innovative explanation for variations in electoral volatility. Moreover, at practical level, it emphasizes the need for a party to get in a specific direction with respect to its organization and behavior between elections. The theoretical framework elaborates on the relationship between party type and electoral volatility, dismissing briefly the alternative explanations provided in the literature to account for the variation of the dependent variable. Despite their functionality elsewhere, these alternative explanations are not, as we will see, appropriate for the CEE space. The concept formation section emphasizes the particularities of the main variables, whereas the operationalization comes together with a close look of the current research methodology. The final two sections provide the results of our analysis and explain why specific outcomes occurred in the case of individual parties. At the same time, the last section explains the direction of our hypothesis and why the reverse is not possible in the case of the parties under observation. THEORETICAL FRAMEWORK Party organization and behavior between elections can represent both an independent and dependent variable. Its status as a dependent variable is reflected by research questions such as: what determines the level of party organization? What conditions make party organization possible and what conditions make it thrive? How can the level of party organization and activity be influenced? The answers vary according to the research design, selected cases, time period, and used methods. However, this research leaves aside these aspects and identifies party organization as independent variable, trying to answer the question: does a high level of party organization and political activity contribute to electoral stability of the party? Therefore, this section argues in favor of a linkage between the two, emphasizing an institutional explanation of electoral volatility. Citizens preferences, although manipulable on the short term, have the tendency to be highly stable on the long term (Bartolini and Mair 1990). This situation, valid for Western Europe, has all the chances to be encountered in the CEE states, after a few elections have passed and the transition is almost over. Major parties, in durable 12
Party Types and Electoral Stability in Central and Eastern European States democracies, try to fulfill minimum requirements in order to meet the expectations of the potentially stable electorate. Thus, continuity and solidification on established basis appear to be the main goals of parties that want to keep, if not increase, a hard core of voters every election. Dramatic changes are risky because they can destabilize voter s perceptions on the party and, thus, the communication process between party and citizen is strongly affected (Colomer 2001: 136). The stability and continuity, registered on various dimensions, starting with the ideological and ending with the formal and structural one, can be observed in the research of Castles and Mair (1984), Laver and Hunt (1992), Huber and Inglehart (1995) and Knutsen (1998). Thus, the CEE parties, in their attempt to win votes and offices (Downs 1957), pursue the same goals of continuity and stability but on weaker organizational and electoral bases (Rivera 1996; Rose 1996). Aware of no iron law of parties development and organization, Panebianco (1988: 18) emphasizes that parties tend to go from an initial period characterized by certain needs to a subsequent period with different prevailing needs. This process represents the transition from a phase where collective incentives (formation of organization s identity) and dominance of the environment prevail to a phase where selective incentives (development of a bureaucracy) and the adaptation to environment prevail (Panebianco 1988: 19). Parties present different combinations of organizational elements in the first phase and these initial organizational differences contribute to organizational differences in the third phase (Panebianco 1988: 19). The strength of political organizations and procedures varies with their scope of support and their level of institutionalization where scope refers simply to the extent to which the political organizations and procedures encompass activity in the society (Huntington 1965: 394). How and why do party organization, political activity and electoral volatility connect Having these in mind, we bring three main arguments to justify the relationship between party organization and the level of electoral volatility. First, party organization and political activity influences electoral volatility through complex mechanisms where causality changes its direction. Initially, electoral volatility and the erosion of partisan affiliation lead to the need to reform and stabilize party organization (Mair et al. 2004: 8). Thus, the initial answer of parties to the electoral stimuli is represented by a reform of their internal organization meant to increase the autonomy of the party, the concentration of power to the centre and increasing the interdependence of the local party organizations. By doing so, the reformers hope, most of the times, to increase the continuity and stability of the party (Scarrow 2004: 98), making it attractive to voters on the long-run, in order to earn stable electorate. Thus, organizational stability and continuous political activity occurred from the need to diminish electoral volatility and are meant to fulfill this goal. Second, the stability of the electorate is increased when perceiving the continuity and stability of the party in terms of ideology and policies. If, on the contrary, the position of the party in economic, social and political issues changes from one election to another, their credibility among the stable electorate decreases significantly as well 13
as their chances to be identified with a certain position on the political space axis. Thus, parties that are able to situate themselves on a stable position on the left-right axis and manage to promote consistent policies in every election are favored. In doing so, parties have to develop solid and stable organizations and to make sure that their position is perceived similarly by voters at any moment in-between elections. This, corroborated with a highly institutionalized environment, discourages electoral volatility by closing off the electoral marketplace, narrowing the range of viable alternatives and socializing voters to embrace established identities (Roberts and Wibbels 1999: 578). Third, the specificities of the CEE region and the transition period determined a different environment for parties to develop their organizations. Four main specificities can be emphasized and for each of them it is relevant to notice that formal development in terms of organization and political activities between elections appears to be valid solutions for electoral stability. First, parties in CEE confront more open and more available electorates in terms of party allegiance than the ones in Western countries (Mair 1997: 179). Thus, the electorate is more volatile and uncertain, the average volatility being much higher than in Western cases, while party membership is much lower (Mair 1997: 182, 186). In such conditions, parties that institutionalize have greater chances to attract a constant number of voters and to increase the incentives for becoming a party member through its bureaucratic mechanisms. Second, there is a context of competition, where the number of actors changes constantly and the parties with a higher level of continuity have, most of the times, an advantage in front of the new-comers. Third, the competition pattern in CEE states, characterized by conflict and adversity, induces an uncertainty status the parties cannot predict their success or they failure according to different policies promoted in the medium and long-run (Mair 1997). Thus, they increase the chances of diminishing volatility by a focus on shortterm policies. However, the short-term policies have to be constant and presented to the voters in successive elections so that the party can be perceived as continuously engaged. The fulfillment of this ongoing policy elaboration process, at party level, requires was conceptualized as systemness (Panebianco 1988). Fourth, party formation is different from the cleavage structure identified by Lipset and Rokkan (1967) as a basis for party creation. There were only isolated cases, Poland and Hungary, where one or two dimensions of the cleavages proposed by the two researchers can be encountered. Parties emerged mainly in two phases: in the beginning, mostly for the first elections, there were forums, unions, fronts, and alliances that grouped the opposition forces or the followers of the communist parties; afterwards, the split of these entities lead to the creation of new parties for the following elections. In this respect, institutionalization plays a key role the party did not occupy from the beginning one position in the political space, but had to find and consolidate on a certain position so that the electorate gets used with this position in consecutive elections. Environment of party institutionalization in CEE states In order to show the complexity of party existence and organization in the CEE space, it is necessary to analyze the environment and to observe the sources of isomorphism. Party development happens in a competitive institutional environment and not in an 14
Party Types and Electoral Stability in Central and Eastern European States institutional vacuum or tabula rasa (Kitschelt 1992). Political parties compete, in the attempt to fulfill their functions, with several institutions in societies: the institution of personal connections built on friendship and nepotism (Peng 2004), the institution of direct democracy (Katz 1987), and the system of pressure and lobby groups (Katz 1978: 19). For CEE states, some of the competing institutions are older than political parties and difficulties occur in party organization and structuring due to isomorphism. All three types of isomorphism identified in the literature coercive isomorphism (formal and informal pressures exerted on party by other organizations on which the party is dependent and by cultural expectations in the society), mimetic isomorphism (the modeling of the party according to other organizational models) and normative pressures (based on professionalization) (Di Maggio and Powell 1991: 67-72) have a certain influence on parties types and on their organizational features. Besides these competing institutions, the environment has to provide a series of institutional prerequisites for appropriate party development. Briefly, these conditions are: the democratization of institutional procedures (their speed was different in the six countries under observation), the clear separation of powers and delimitation of competences, parliamentary rules of party activities, legal bases for party formation, and the legal regulations of party activities. These aspects are, most of the times, taken for granted in democratic environments, their roles being more obvious in environments that lack them, such as those years in the beginning of the transition period in the analyzed countries. Alternative explanations of electoral volatility Understanding the complex environment in which party types develop and pursue their goals represents a necessary step in linking organizational aspects of parties with electoral stability. The following step is to show that other explanations used in other contexts to explain the variation of electoral volatility do not work in the observed CEE states. Three of the main alternative explanations provided in the literature are briefly dismissed in this sub-section, a larger part being dedicated to the electoral system, considered to be the key variable in explaining the electoral volatility variation. The electoral system The choice of electoral system is among the most important and, arguably, the most important of all constitutional choices that have to be made in democracies (Lijphart 1992). Parties are the main actors to pursue the electoral reform resulting in the modification of the electoral system. In this respect, by using a single-case study on Hungary, Montgomery (1999) emphasizes the electoral effects on party behavior and development. Besides this studied linkage, there is no support for the relationship between the electoral system and volatility. There are three main reasons, two theoretical and one empirical, for which the electoral system is dismissed as being the cause for parties electoral stability. First, electoral volatility might be a cause for which electoral reform to be initiated, however not the only one. In CEE states, causes can be represented by: representation of minorities, a better representation of interests, and stabilizing new institutions. Second, the overall acknowledged effect of electoral systems has been to diminish the parliamentarian fragmentation and to stabilize the competition 15
between parties by introducing the thresholds (Bakke and Sitter 2005: 251). There is no direct effect on the electoral volatility of individual parties; it may affect, at most, the gross electoral volatility of the system. Third, the results of electoral reforms are two-faceted, often generating important and unanticipated political consequences that are not all the time in the advantage of the designers (Kaminski 2002: 350; Grofman et al. 1999; Moser 2001; Stein 1997). At empirical level, individual party volatility calculated for CEE states (Gherghina 2006) reflects no variation according to the electoral reform in these countries. The economic performance Another alternative explanation used in the literature for the variation of electoral volatility is the economic performance. Roberts and Wibbels (1999) show that for Latin America, the economic performance is poorly correlated with the electoral volatility, the stronger variable that correlates being party institutionalization. No similar study has been made for CEE states. However, at a glance, the economic performance cycles cannot provide an indicator of how the electorate votes due to the fact that within the same cycle more elections were held and more parties got into government. Moreover, if we take a look at the individual volatility for the years where economic cycles occur there are no significant differences between in the variation of the electoral volatility during economic development or backslashes. The level of democratization Democratization is one of the most difficult concepts to be measured and operationalized. For reasons that are connected with time and space, we consider Freedom House indicators, indicating the level of civil liberties and political rights, as eloquent for level of democratization although they are not identical. Starting from the study of Bartolini and Mair (1990), it can be assumed that in a democratic country, the variation of electoral volatility is smaller in comparison with countries that initiate the process of democratization. Mair s figures confirm this aspect (Mair 1997: 182). However, if we take a look at the electoral volatility and democratization trends within CEE region we notice that, in many cases, countries that performed better on the Freedom House scale have a higher level of electoral volatility (e.g.: Poland has a higher individual electoral volatility than Romania, Czech Republic is more volatile than Bulgaria). THE CONCEPTUAL FRAMEWORK The first methodological step in a research, usually the most difficult and influential, resides in appropriately defining and conceptualizing the terms and concepts to be used. Building on the literature from three fields and sub-fields of Political Science institutionalism, political parties and electoral systems the current section deals with the two main concepts of this research electoral volatility and party types providing a brief introduction that reveals the elements to be taken into account in making the typology of parties organization. 16
Party Types and Electoral Stability in Central and Eastern European States Grounds for Organization Scott s definition (2001) and conceptualization of organizations, with the emphasis on stability and pillars, provides useful ground for analyzing parties, through the emphasis put on organizational aspects that incorporate norms, values and regulations. The three pillars reflect diverse party features that are relevant in the organizational process. The regulative pillar refers to constraints and regulations coming from the institution towards its members. Through these, the party establishes rules, monitors activities, and sanctions behavior. The normative pillar includes both values and norms 2 on which the party was built and wants to be perpetuated or adjusted. When norms and values are attributed to certain specific actors: these institutionalized positions become roles (Scott 2001: 55). The cultural-cognitive pillar implies the shared conceptions that constitute the nature of social reality and the frames through which meaning is made (Scott 2001: 57). However, the internally attributed meaning is shaped by external cultural frames: myths and ceremonies (Meyer and Rowan 1992). When speaking about types of parties (see the following sub-section), it is helpful to distinguish between internal and externally related aspects of the party development process. Internal aspects refer to developments within the party itself (organization, structure); external aspects have to do with the party s relationship with the society in which it is embedded, including other institutions. Within each of these aspects there will be a structural and an attitudinal component (Randall and Svasand 2002: 12). We follow only the structural aspect both internal and external translated into Panebianco s concepts of autonomy and systemness. In the life of an organization there are three phases: genesis, institutionalization and maturity (Panebianco 1988: 19). The particular combination of organizational factors from these phases influences the type of parties we encounter in our analysis. Bluntly put, we are interested in how its organization relates to existing centers of power in the community, how the internal sub-units are linked with each other, from what strata of society is its leadership drawn and how this affects policy; and how it justifies its existence ideologically. Electoral volatility Electoral volatility represents an indicator of the net support the party has in consecutive elections. Thus, it represents a reliable indicator of party stability in relationship with the electoral potential of the party. It reflects the differences in votes obtained by a party from an electoral cycle to another (Taagepera and Shugart 1989; Lane and Ersson 1999: 127). A more precise definition of the electoral volatility sees it as a measure of the net electoral change between two consecutive elections (Bartolini and Mair 1990: 19). There are various ways to calculate the electoral volatility and, being interested in the dynamic evaluation of individual party development in the CEE states, we use the individual volatility because it allows me to weight differently the parties, according to the votes they receive (for calculation and details, see concept operationalization section). The formula we use for individual electoral volatility is an adjustment to the formula proposed by Bartolini and Mair (1990), the only difference being that I calculate it relatively to the mean party support. The formula for individual volatility is: 17
Volatility = ½ * [(V t0 V t1 / (V t0 + V t1 ) / 2] * 100 Where V t0 is the share of votes gained by the party in the first elections of the electoral cycle and V t1 is the share of votes gained by the party in the second elections of the electoral cycle. Without calculating the mean party support and taking into account only the share of votes obtained at the first election, a party that increases from 2% to 4% and another one that increases from 30% to 60% have, according to the formula I use, the same level of electoral volatility (33%). Party Types The literature dealing with parties includes many typologies that try to capture the diversity of organizations that emerged worldwide in the last decades (Duverger 1954; Kircheimer 1966; Kitschelt 1989; Koole 1996; Gunther and Diamond 2003). Ranging from four to fifteen different types, an important shortcoming of these typologies is that many are Western based that do not take into account developments in the newly emerged democracies starting the 70s. Gunther and Diamond (2003: 168) acknowledged the risk of having conceptual stretching when including new parties into old typologies. Aiming to test the relationship between party type and electoral stability, we do not provide a new typology of parties as this requires extensive space and research. Consequently, we focus on five types, resulted mainly from Gunther and Diamond s typology, drawing analogies, when possible with other types identified in the literature. Thus, we ended up with five main types of parties: elite based, mass, catch-all, cartel, and ethnic 3. 1. Elite based parties: have minimal principal organizations, based upon established elites within a specific geographic area. Deference to the authority of these elites is a basic feature and the national-level party structure is based on alliances among locally based elites. In the literature the following types of parties were conceptualized as elite based: parties of parliamentary origin (Duverger), parties of individual representation (Kircheimer), modern cadre parties (Koole). 2. Mass parties Characterized by a large number of active members, this type of parties attempted to penetrate a wide range of social life spheres. The party uses trade unions, foundations, social organizations to promote its messages and get outside the parliamentary arena. It establishes at national level networks of local party branches and develops networks for support that will allow to mobilize as many people as possible at elections. 3. Ethnic based parties: They usually lack extensive and elaborate organizations such as those present in the mass-based parties. Their political and electoral logics target a specific group of society, being narrower in their policies and claims. The main goal consists in mobilizing the votes of its own ethnic group and they are concentrating many resources in the areas where this ethnic group is concentrated. Although running candidates in other constituencies and emphasizing national issues, their focus is on regional or ethnical aspects, using the former as a thinly mask (Gunther and Diamond 2001: 23). 4. Catch all parties: characterized by shallow and superficial (thin) organizations, vague ideology, and a relatively skeletal existence in between elections (Gunther 18
Party Types and Electoral Stability in Central and Eastern European States and Diamond 2001: 25). Their maximum level of activism is reached in the electoral campaign when a wide range of techniques is used to increase the share of votes. This type of parties can be summarized as vote seeker. In this respect, they try to aggregate a variety of social interests and, thus, their ideological focus cannot be stable it shifts according to the mood of the population and to the salience of issues at stake at a particular moment in society. The direct consequence of a stable ideological background is that these parties heavily rely on their leaders features. This type of parties was also called populist in the literature tackling post-communist states. Features of this type: 1) Nominations are largely determined by the electoral potential of the candidates rather then professionalization or experience. 2) Electoral messages are transmitted through the use of media, umediated appeals coming from the candidate. 3) The campaign messages tend to focus on the weaknesses of the opponent or the abilities of the candidate, without a clear policy and ideological back-up. 4) They avoid showing links with any type of group during campaign. 5) During its ideological diffuseness, the party is willing to take part in many coalitions. 6) Limited potential for social integration of these parties due to the low levels of citizens involvement and identification with these parties. 5. Cartel parties: This party type is characterized by a fusion of the party in public office with several interest groups that form a political cartel which is mainly oriented towards the maintenance of executive power. It is a professional organization that is largely dependent on the state for its survival and has slowly retreated from civil society, reducing its functions mainly to governing (Krouwel: 251-252). Rather than competing in order to win votes, cartel parties are office seekers, being contented with getting access to the state by sharing power with others (Wolinetz 2002: 148). Long periods in government transform the internal structure and power balance within parties. State resources are progressively accumulated by the parliamentary party and the party in public office becomes independent from the membership party and its central office. RESEARCH DESIGN AND CASE SELECTION Case selection and time frame The cases included in the research are the CEE states: Bulgaria, Czech Republic, Hungary, Poland, Romania, and Slovakia. More precisely, the primary unit of analysis is represented by the parties from these countries that got in the Parliament in at least half of the elections. The reason for choosing these cases is that the beginning of party institutionalization can be easily traced back to the fall of communism and the first competitive elections. Thus, we include in the analysis 26 parties, as follows: three for Bulgaria, four for the Czech Republic, six for Hungary, six for Romania, two for Poland, and five for Slovakia. The time-period to be taken into consideration is the beginning of the transition period (1989-1990) until 2005. Thus, we no longer take into consideration the elections which took place after 2005. This dynamic analysis includes measurement of party institutionalization and electoral volatility level for every electoral cycle within this period, leading thus to comparable time series to be used in the quantitative part of the research. 19
Method In order to identify the relationship between the independent and the dependent variable, we proceed in the following way. Based on content analysis of the official documents of the parties taken into consideration and on the evaluations of local experts, we assign each party into a particular category of parties, according to the previously presented typology, capturing the changes which took place from on electoral cycle to another, where this is the case. Then, we use simple descriptive statistics to determine the level of variation of the electoral volatility for each type of party. The total number of observations on which the analysis will be conducted is 89, therefore having a solid base of analysis and allowing a good level of variation on the independent variable for conducting a thorough analysis. We then compare the main indicators for each type of party, mainly the mean and the median, taking a look at any occurring outliers and explaining them, where it is possible. This step should give the necessary information for testing the proposed hypothesis. Finally, we create two ordinal scales for the independent variable, based on two different criteria, and test for the strength of the association between the type of party and the level of the electoral volatility, in order to establish if there is a significant statistical relationship between them. Thus, this research uses qualitative methods in order to eliminate as much as possible the level of arbitrary in the process of categorization of parties, which the influences to the largest extent the subsequent analysis, which is conducted using quantitative tools. Data sources The data needed to establish the institutionalization level of each party are taken from party manifestos and status (power relations, organization of sub-groups, formal relations with external actors), electoral funding documents (reveal the fund sources), the annual monetary balance provided by parties, media and international organizations reports with respect to corruption issues in parties (I am interested only in the high corruption scandals that may indicate domination from the environment) and other official documents issued by the parties with respect to their formal organization. The electoral volatility is calculated on the basis of the electoral results available on different databases and on national country reports made by various international organizations such as Freedom House or Transparency International. Methodological problems Before continuing with an explanation on how we assign the parties included in the analysis according in the five types of parties mentioned before, we must clarify two main problems which pose risks for the validity of our conclusions, if not properly dealt with. First, many parties in the CEE, as it is natural in any party system, have the option of running in elections by themselves or form alliances. This is a potential problem for our research, because we measure volatility for individual parties. We opted for the following solution: if a coalition remained relatively stable across time and ran together in elections under the same label, even if some parties left it and others joined it, we will treat that coalition as one party. We do the same if the coalition disbands, 20
Party Types and Electoral Stability in Central and Eastern European States but there is a party which claims or is perceived to be the successor of the coalition. As an example, this is the case of the Union of Democratic Forces (UDF), in Bulgaria. It has participated under this name in all the elections starting with 1990, but in the 2005 the coalition no longer existed; however, a party bearing the same name ran in the elections as the successor of the coalition. Another example is the Party of the Hungarian Coalition, from Slovakia. The name of the party is somewhat ambiguous, because it contains both party and coalition. It is actually a coalition of the Hungarian parties in Slovakia, traditionally three parties, with a fourth one joining for a short term. It ran under this label in all the elections following Slovakia s independence and is treated as one party in our analysis. On the other side, we have parties which have been part of coalitions (the same or different) for one or two elections, but ran on their own in the rest of the elections. This is case, for instance, with the National Liberal Party in Romania. It ran on its own in 1990, 1992, then it joined the Democratic Convention coalition in 1996, then again it ran on its own in 2000, and then formed another alliance with the Democratic Party for the 2004 elections. For such cases, we approximate the share of votes obtained by a party member of the coalition, by looking at the share of seats gained by the party, based on the result of the coalition. It is not a perfectly accurate measure, but it is the best way of approximation we have. This is particularly important for the calculation of volatilities. Second and equally important, the categorization of the parties according to the typology mentioned before poses several problems. It is very difficult to assign one party into one particular type in a straightforward way. The typology is created based on theoretical considerations, mainly by looking at Western European parties. Many parties have overlapping features and it might be easy to assign one party in at least two categories. This is mainly due to the particularities of post-communist parties and party systems during the transition years, which differentiate them in important ways from the more stable and institutionalized parties and party systems in Western Europe. The organizational structures of the parties, on which we focus, have been usually unstable, influenced to a great extent by the many mergers and splits which took place within the post-communist party systems. We assign each party in a particular type based on what we consider to be the predominant features of the party, mentioning every time all the elements which lead us in the process and also what are the elements which might lead to alternative views. However, by looking at each party from one electoral cycle to the other we manage to capture the overtime changes within the party and in many cases we see that parties change their types during the transition period. There are also some more minor issues which must be taken into consideration. In the case of Hungary, where a mixed electoral system is used, we take into account the share of votes give for party lists and we disregard those for individual candidates. In the case of bicameral legislatives, we take into consideration the electoral results for the lower chamber. CATEGORIZATION OF PARTIES We now move on to briefly present each party and explain how we assign them in a particular type for each electoral cycle. 21
Bulgaria The three parties taken into consideration in Bulgaria are the Bulgarian Socialist Party, the Union of Democratic Forces, and the Movement for Rights and Freedoms. The elections taken into consideration are 1991, 1994, 1997, 2001 and 2005. Bulgaria is not as problematic as other states. The Movement for Rights and Freedoms (MRF) is clearly an ethnic party 4. It has been set up by former Turkish dissident Ahmed Dogan immediately after the fall of Communism and the acceptance of pluralism. Its electoral target is the Turkish community in Bulgaria, comprising almost 10% of the population. Also, it addresses the Pomak community, which is a Turkish speaking, but Christian Orthodox, community, though it is usually regarded as a part of the Turkish minority. The leaders of the MRF are in their great majority Turkish and their electoral support comes from the members of this minority. Thus, the MRF is assigned into the category of ethnic parties for the entire transition period. Nevertheless, it can also be regarded as a catch-all party, but with the particularity of addressing only the Turkish community. Its ideology is not very clear and it often appealed to a populist discourse in order not to lose votes to non-ethnic parties. Also, it can be regarded as an elite party, since it is based on a particular region of the country, the one where Turks are concentrated, with the organizations being led by local notabilities, in close connection with the leadership in Sofia. The Union of Democratic Forces (UDF) is a more problematic case, because it functioned as a coalition, although changing its structure from one election to the other, for the entire transition period. In 2005, however, it disappeared as a coalition, but the label has been taken over by one single party. We treat this as the successor of the coalition, although other parties, such as the Democrats for Strong Bulgaria, claim to be the real successors of the UDF. Besides this issue, however, it is not very difficult to categorize the UDF as a mass party for the entire transition period. Since it was set up in 1990, it reunited many smaller parties, which worked together at both central and local level. Thus, the total membership in the coalition was quite substantial, was supported by several foundations and had a strong organizational base throughout Bulgaria. The Bulgarian Socialist Party (BSP) is a typical mass party for the CEE. It is the formal successor of the Communist Bulgarian Party; basically, in 1990, it was the same party, with a changed name, led by the former second echelon in the Communist party. The BSP started to reform after loosing the 1997 elections, changing its head figures and adopting a pro-european discourse. However, it still benefited from a broad membership, it took over and used throughout the entire time frame the strong organizational structure of the former Communist party. It also had some features of a cartel party, during the times when it has been in power. The Czech Republic The four parties taken into consideration in the analysis for the Czech Republic are: the Communist Party of Bohemia and Moravia, the Civic Democratic Party Christian Democratic Party, the Czech Social Democratic Party, and the Christian Democratic Union Czech People s Party. We take into account three elections: 1992, 1996, 2000. 22
Party Types and Electoral Stability in Central and Eastern European States The Communist Party of Bohemia and Moravia is again a typical mass party. Unlike the other Communist parties in the CEE, which usually changed their names and tried to present themselves as Socialist parties, the Czechoslovak party remained unreformed and tried to gain its position in the new party system (although the Slovak side of the party changed its name to Party of the Democratic Left). Thus, the new Communist party of Bohemia and Moravia took over the entire organizational structure of the former Communist party and large part of its membership, as well as maintaining its ties with unions or leftist foundations. The Christian Democratic Union Czech People s Party is a mass party, with over 50.000 members, despite a declining trend in membership in the past years. Existing in illegality even during the Communist era, it had strong ties with the Catholic church and tried to present itself as a modern, German-style, Christian-Democratic party. The Czech Social Democratic Party is an interesting case. The Czech Social Democrats are not related in any way with the former Communists, unlike in the other CEE states. However, mainly due to the population s rejection of leftist ideas and to the popularity of the new center-right forces, they were unable to win the elections or at least to be co-opted in any of the first governments. That s why they adopted a populist discourse, advertising increased social protection, in order to counterbalance the reformist measures of the new right-wing governments, turning into a catch-all party. Only in 1998 they managed to win the elections and be the main force in the government. During their first experience in government, they acquired significant features of a cartel party. The party became office-oriented and tried to strengthen its position in the local branches of the executive power and to gain as much access to resources. Consequently, they won the elections again in 2002. The Civic Democratic Party Christian Democratic Party is the successor of the Civic Forum and has been in power after break of Czechoslovakia until 1998. Despite its reformist actions, it adopted a populist and anti-europeanist discourse. Also, its organizational structure was not that strong and its ideology was rather diffuse, the party gaining its authority and its legitimacy due to its central figures. In 1998, after losing the elections to the Social-Democrats, it supported the new government without claiming seats in the government. In exchange, it maintained to some extent its positions in the local administration and its ties with interest groups, acquired during the years in government, thus becoming mainly a cartel party. However, it failed again to win the elections in 2002 and it opted for the status of opposition party for the first time in its existence. Hungary For Hungary we take into consideration six parties: the Hungarian Socialist Party, the Alliance of Young Democrats (FIDESZ), the Alliance of Free Democrats, the Hungarian Democratic Forum, the Smallholders Party, and the Christian Democratic People s Party. We look at four elections: 1990, 1994, 1998, and 2002. The Hungarian Socialist Party is an exception among the successor parties throughout CEE. Unlike in the other states included in the analysis, the Socialists in Hungary benefited to very limited extent and only in the first years of the transition from the support of unions or other civil society organizations. Immediately after 23
the transformation of the old Communist party, the Socialists decentralized their structure, maintaining the strong organizational infrastructure at the local level, but choosing to leave increased power in the hands of the local elites. The level of membership decreased significantly, which actually made it easier for the Socialists to reform, having a more flexible internal organization. The activity of this party was always parliamentary oriented and argue publicly that politics should not be a part of private lives. Therefore, we treat the Hungarian Socialist Party as an elite party for its entire existence. FIDESZ started to function in illegality before the fall of the Communist regime. Its members were predominantly young people and the party maintained the youth as its main target in the first years of the transition. It even had a clear provision in its status according to which all the members must not be older than 35 years, but this was removed in 1993. This aspect makes FIDESZ a particularly difficult party to assign in a type. However, considering the ties FIDESZ started to create with the civil society in the first years after its establishment and the intention to mobilize as many young people as possible throughout the country, we will treat it as a mass party for the first Hungarian electoral cycle (1990-1994). After 1994, it acquired features of a catch-all party. It adopted a populist discourse, challenging many reformist measures of the Socialist government, promoting nationalism and having a rather diffuse ideology, switching between liberalism and conservatism. Thus, for the next two electoral cycles, we treat FIDESZ as a catch-all party. The Hungarian Democratic Forum, as FIDESZ, started its existence before the fall of Communism. It functioned as a mass party for the first electoral cycle, when it was the main force in the government. During these years, they strengthened their local organizations and collaborated closely with the civil society. In 1994, they gained only 11.7% of the votes and started to transform into an elite party, aiming to maintain the privileged position of its notabilities at the local level. After dropping to just 3.12% of the votes in the 1998 elections, they managed to join the governing coalition led by FIDESZ and remained an elite party, with decreasing membership. The Alliance of Free Democrats has been established in 1988 and functioned as a mass party for the first two electoral cycles, being the second party in government after 1994. As all the other anti-communist parties in Hungary, they collaborated closely with the civil society in the first years of the transition. Their trajectory is similar to that of the Hungarian Democratic Forum, but remained essentially a mass party also in the second electoral cycle, benefiting from its position in the government. From 1998, its share of votes also dropped dramatically, it entered the opposition and started to function as an elite party. The Independent Smallholders Party is similar to the Polish Peasant Party. It is a typical agrarian party, very active in the countryside, where it managed to mobilize the population to a large extent, until 2002, when it gained no seats in the legislative. Although it addresses a particular segment of the population, we treat it as a mass party, because it functions as such for its particular electoral target. The Christian Democratic People s Party functioned is an elite party for its entire existence. It does not have strong ties with any Hungarian church and therefore lacks the element which usually turns Christian Democratic parties into mass parties. It 24
Party Types and Electoral Stability in Central and Eastern European States relies mainly on a number of notabilities, especially at the central level. It is currently in a coalition with FIDESZ. Romania There are six parties taken into consideration for Romania: the Social Democratic Party, the Democratic Party, the National Liberal Party, the National Peasants Party Christian-Democrat, the Greater Romania Party, and the Democratic Alliance of Hungarian in Romania. The elections taken into consideration are 1990, 1992, 1996, 2000, and 2004. The main problem concerning the Romanian case is the split in the National Salvation Front in 1991. Thus, we will only take into consideration this party from 1992. The Social Democratic Party is regarded as the successor of the former Communist party, although this label has not been used anymore in the recent years, after the party started to reform. Its origins are in the National Salvation Front (NSF), created immediately after the fall of Communism, which took over basically the entire infrastructure of the former Communist party. In 1991, the Front split into what are today the Social Democrats and the Democrats. Because the main figures of the NSF remained with the Social Democrats, which also kept the largest part of the local organizations and of the members, we treat the Social Democrats as the successor of the Front. The party has been in power from 1990 to 1996, when it had access to all the resources, did nothing to fight corruption and got involved in many relations with economic interest groups, which influenced to a great extent its policies. Because of this, we treat the Social Democrats for the first two electoral cycles as a cartel party. After losing the elections in 1996, it lost its influence at the local level, though remaining very strong throughout the country due to its widespread organizational structure and the massive membership and thus we categorize it as a mass party. After regaining power in 2000, it retied its dominance over the local branches of the executive power and we assign it back in the category of cartel parties. The Democratic Party formed following a split in the NSF in 1991. After the 1992 elections it acted as a mass party, trying to strengthen its organizational structure in the territory and regain a satisfying level of membership in order to recover the lost ground following the split from the Social Democrats. It joined the new governing coalition in 1996, but its actions have been rather peculiar. Appealing to its center-left ideology, it opposed some of the reformist measures of the right-wing parties leading the coalition. It temporarily left the government and constantly used a populist discourse, despite losing ground in the opinion polls. Following the poor electoral results in 2000, it tried to stabilize itself as a social democratic party and to strengthen its local structures and its ties with the civil society, thus acting as a mass party. The National Liberal Party is a typical elite party during transition. It was reestablished after the fall of Communism by former dissidents, but it had a rather poor local infrastructure and a relatively low membership. It did not enter parliament in the 1992 elections, but then joined the Democratic Convention coalition and gained seats in 1996. It played a secondary role in the governing coalition and therefore it remained an elite party, basing its legitimacy on a clear rightist ideology and on a few notabilities at the central and local level. Only after entering opposition in 2000, it 25