Nationalisation of Party Systems in the Baltic States and in Central Europe: A Comparative Perspective

Transcription

1 Paper for the ECPR Joint Sessions, April, Rennes Workshop: The Nationalization of Party Systems in CEE. Nationalisation of Party Systems in the Baltic States and in Central Europe: A Comparative Perspective Tõnis Saarts (Tallinn University) Abstract The paper is exploring the main differences between the party systems in the Baltic States and in Central Europe (the Visegrad countries). It is focused on the main characteristics of the party systems and on the process of nationalisation. The analysis suggests, that despite of several similarities between the Baltic countries and the Visegrad countries, the Baltic States (especially Latvia and Estonia) are different form the Central European countries in four important aspects: (1) party systems are ideologically less balanced than in the Visegrad countries, (2) two major parties obtain usually less seats and party systems have been always even more fragmented than in the Visegrad countries, (3) socio-economic cleavages are playing even less important role in the inter-party competition than they do in Central Europe, (4) church-state cleavage is almost missing, but in Central Europe it is playing a prominent role. The analysis also indicates that party system in Lithuania is quite different from rest of the Baltic States and in many aspects more similar to the Visegrad countries. Finally the analysis demonstrates that if we are looking at today s party systems in CEE, we can distinguish between three different groups in the region: consolidating Central European group, less consolidated Central European group and the Baltic group. The most nationalized is a consolidating Central European group Hungary and Czech Republic. Introduction Considerable attention has been paid to party systems and their evolution in Central and Eastern Europe (CEE). The amount of literature is growing on the field and most of the recent analyses are demonstrating that there is a stabilization process going on in the region (Lewis 2000, Jungerstam-Mulders 2006b). It doesn t mean, that party systems in CEE are going to be more and more similar to the Western Europe, but at least the most turbulent period in the 1990 s seems to be over (at least in the new member states of the EU). Most of the studies about the party systems in CEE countries have been concentrated on the CEE region as a whole (e.g. Rose & Munro 2003; Birch 2001; Lewis 2000; Kitschelt, Mansfeldova, Markowski, & Tóka 1999). Usually CEE countries are compared with the Western Europe and the conclusion is that party systems in the East and in the West are quite different. Though some authors argue, that we can see some limited convergence during the recent years (Kostelecky 2002). It is important

2 2 to compare party systems in the Eastern and Western part of Europe and to map the major divergences, but recent years have demonstrated that CEE countries are going to be more and more different from each other as well. There is no sense anymore to put party systems in Moldova and in Czech Republic together and to approach them in same way if making comparisons with the Western Europe. Comparative analyses between the different countries in CEE and even different sub-regions (the Baltic States, Balkan countries) are going to be more and more important. There are already some prominent studies, where party system were analysed within different sub-regions in CEE. For example Pettai and Kreuzer (2004, 1999) and Krupavičius (2005a) have compared party systems in three Baltic countries. We can find similar and even more comprehensive analyses about the Visegrad countries (Kostelecky 2002, Toka & Henjak 2006). But analyses what are comparing different subregions with each other (for example the Visegrad countries and the Baltic States, the Balkan countries and the Visegrad countries, etc) are really rare or almost absent in the field. In this proposed paper we intend to fill this gap. The research objective of the paper will be to compare the main features of the party systems in Central European countries and in the Baltic States and also to focus on the process of nationalization. The research question is: If we are talking about the main features of the party systems and nationalisation of party systems in CEE, are the party systems in the Baltic States different from the Central Europe and in what aspects? Central Europe is specified as the Visegrad countries: Czech Republic, Slovakia, Poland and Hungary 1. The Baltic States are: Estonia, Latvia and Lithuania. The author of the paper is convinced, that these regions are historically and geographically distinct and therefore this kind of comparison is justified and not arbitrary. The nationalisation of party systems refers to the integration of electoral behaviour, issue dimensions, cleavages and party organizations (see: Caramani 2004). Therefore it is quite a wide concept, what comprises the main aspects important for analysing and comparing party systems. In proposed paper the author will concentrate on three dimensions related to general aspects of the party systems and nationalisation: 1 Slovenia is excluded, because we have already too many countries included to this analysis. Slovenia is not a Visegrad country anyway and has also a different background compared with them.

3 3 The main characteristics of the party systems. For comparing the party systems we will use an analytical scheme provided by A. Ware (1996) and most of attention will be paid to the number of parties and share of votes. The main variables and indicators would be the following: effective number of parliamentary and electoral parties (see: Toka & Henjak 2006; Biezen & Caramani 2007), share of votes won by two largest parties, statistics about splits and mergers. Electoral behaviour. Here the main indicators would be electoral volatility and voter turnout (see: Sikk 2005; Rose & Munro 2003; Kostadinova 2003). Cleavages structures and divisions. We will map the main cleavages constellations in countries under investigation and analyse how they have changed during last years. It would help us to make some cross-subregional comparisons. If we find some prominent differences, we will attempt to search for some explanations. We exclude party organizations from the analysis, because this is a too bulky theme and demands for a separate article. But at same time we will pay an extra attention to the first and to the third dimension: the main characteristics of the party systems and cleavages constellations. Because the main differences between the Baltic States and the Central European countries would be found especially related to these aspects. Now we will give a short overview about the structure of the article: In the first section we will give a short overview about the main features of the party systems in CEE and make some comparisons with the Western Europe. In the second section we will concentrate on the main characteristics of the party systems in the countries under investigation (a brief comparative overview about the origins of the main parties, number of parties in the party systems and other variables mentioned earlier). In the third section we will focus on electoral behaviour and in the fourth section on the cleavage structures. Finally we will conclude our results and try to find an answer for our main research question: are the Baltic States different from the Visegrad countries? And we also try to group different countries according to the results of our analysis regarding nationalization of the party systems.

4 4 1. Party systems in CEE and in the Western Europe There is a growing amount of literature where the parties and party systems in the Western Europe are compared with the party systems in post-communist countries (see: Jungerstam-Mulders 2006a; Rose & Munro 2003; Lewis 2000). Because the discrepancies between the West and the East are not the main focus of this article we will give only a short overview about the main features of the party systems in CEE, comparing them with the Western Europe: Party systems are more fragmented than in the Western Europe. For example number of effective parliamentary parties and electoral parties is higher than in the West (Biezen & Caramani 2007; Jungerstam-Mulders 2006a). The picture is changing constantly: new parties are emerging, older ones are splitting or merging, changing the names, etc. The whole party system is in flux. The linkage between the voters and parties is quite weak. Because there was no multi-party system before the transition in CEE, voters don t have previous party affiliation and they have only vague sense of group belonging or political identities (Jungerstam-Mulders 2006a). Because this, volatility is much higher than in Western Europe and even higher than in Southern Europe at an equivalent stage of early transition (Lewis 2000, Sikk 2005). The average volatility in the Western Europe was 12% in the 1990 s. In CEE countries the corresponding figure was 28 30% (Biezen & Caramani 2007: 16, Jungerstam-Mulders 2006a: 15). Ideological stances of the parties are not so well-defined, like in the Western Europe and party names are often confusing. It is sometimes hard to classify parties on the political spectrum and to distinguish between different party families in post-communist countries (Lewis 2000; Jungerstam- Mulders 2006a). For example conservative parties can be former communists in some CEE countries and right-wing neo-liberals in another. Liberal camp is often overpopulated and very broad: you can find there either nationalistic populists or Western-type neo-liberals, etc (Lewis 2000). Cleavages (socio-economic cleavages, church-state cleavage, urbanrural, etc) are absent or very weak in CEE (see: Lawson 1999, Kitschelt 1995). Or if you want to analyse cleavages in CEE you need to find differ-

5 5 ent labels for them than in the West (see: Kitschelt et al. 1999). The linkage between social cleavages and party support is also quite blur and not so well-accentuated, like in the West (Whitefield 2002, Kostelecky 2002). Party organizations are less democratic and not so institutionalised than in the Western Europe. They are usually created according to topdown model, dominated by small group of elite (Biezen & Caramani 2007). This is also a reason why linkage between parties and voters is weak, because parties have not emerged as grassroots-level organizations, a part of the civil society, but they have been rather established for realizing the power ambitions of the elite. They are rather part of the state, not rooted in civil society (Biezen 2003). Though in some countries party organizations are already quite strong and developed, they are still in developing phase (see: Jungerstam-Mulders 2006b) and are usually not so institutionalised like in Western Europe. Party membership is lower than in the Western Europe. In Western Europe approximately 5% of electorate is belonging to the political parties. In CEE countries it is below 3% (Mair & Biezen 2001). Voter turnout in CEE countries is lower than in the Western Europe. In the old EU member states the average voter turnout has been around 83 % (IDEA 2004) in CEE countries it has been around 70% since founding elections in 1990 up to 2004 (Rose & Munro 2003). If we compare old and new EU member states in , the figures are the following: 76, 5% vs. 64,9% (Jungerstam-Mulders 2006a: 14). Trust in parties is lower compared with the Western Europe. Though trust in parties is also decreasing in the Western Europe (Dalton & Wattenberg 2002) it is even lower in CEE (Lewis 2000). Some scholars have called CEE countries, because their strong anti-party sentiments, as anti-party systems (Arter 1996). Of cause we can find even more feature for characterising party systems in CEE, but above mentioned aspects were the most striking ones and also useful to know from the point of view of our following analysis.

6 6 2. The main characteristics of the party systems: the Baltic States vs. the Central European countries. In this section we will concentrate on the main features of the party systems in the Visegrad countries opposed to the Baltic States. This is also important because analysing nationalization of the party systems it gives an overall picture about the nature of the party systems analyzed. There are many different ways how to analyse the party systems and different classifications of the party systems (Wolinetz 2006). In this analysis we will use an analytical scheme proposed by A. Ware (1996). For analysing and comparing different party systems Ware was introducing 4 different dimensions: The extent to which parties penetrate society. The ideologies of the parties. The stance of the parties towards the legitimacy of the regime. The number of the parties in the system. We will briefly analyse each of these dimensions, but the most attention is paid to the last dimension, because this is maybe the most important feature, making the difference between two sub-regions The extent to which parties penetrate society. The extent to which parties penetrate society is weak in all CEE countries. In Ware s approach this dimension is not very well operationalized and because this, it is very hard to compare different CEE countries according to this aspect. Should we take into account recognition of the party labels among population, trust in parties, party membership, voter turnout? All these indicators are directly or indirectly connected with the parties penetration into the society, but there are no references in Ware s approach how to measure this. Therefore we just put this dimension aside Origins of the parties Before we are looking at the ideologies of the parties, we should at first talk briefly about the origins of the parties in CEE countries. Kostelecky (2002) suggests that we can find 5 different types of parties in CEE if we are looking at the roots of them: (1) communist successor parties, (2) former satellite parties (existed legally during the

7 7 communist period, though they were absolutely marginal), (3) historical precommunist parties (banned by communists, but were re-established after the fall of the regime), (4) parties having their roots in dissidents movements, (5) new parties (completely new parties). Looking at the Baltic States and the Visegrad countries we can see that all 5 types are represented, except former satellite parties in the Baltic States, because under the Soviet regime all parties were banned in USSR, except Communist Party. This is the first major difference between the Central European countries and the Baltic States. Though in Central Europe former satellite parties have never played a major role after 1989, they are still represented on the political landscape in some countries like in Czech Republic and in Poland (e.g. Christian Democratic Party in Czech Republic as former Czechoslovakia Peoples Party; Polish People's Party as former United People's Party). The second striking difference is that in the Visegrad countries we can see more historical pre-communist parties than in Latvia and in Estonia. In the beginning of the 1990 s there were several attempts to restore former inter-war parties in Estonia and Latvia (e.g. Latvian Social Democratic Workers' Party, Latvian Farmers Union, Farmers Assembly in Estonia), but finally all these experiments failed and reestablished parties merged with the larger new ones (see: Toomla 2005, Runcis 2005). In Latvia some historical parties, mentioned before, were even successful in some elections in the 1990 s, but finally disappeared or merged with the other parties. Hence we don t see any independent historical party on today s Latvian or Estonian political landscape. Lithuania is an exception. In Lithuania former historical parties got a stronger foothold. Lithuanian Social Democratic Party, Lithuanian Christian Democrats, Lithuanian Peasant Party - all are historic parties and have been represented almost in every parliament, though their support has been rather moderate and in some cases quite low (Krupavičius 2005). Despite of this in elections 1992 and in 1996 historical parties got all together 23-27% of votes (ibid.). In recent elections they have been less successful, but they are still on political landscape (except Social Democrates what merged with Democratic Labour Party of Lithuania). We can see same situation in Vishegrad coutntries, where some historical parties were restored and have been more and less successful. For example Hungarian Smalloholders Party and Czech Social Democrates (Kostelecky 2002). Also Polish

8 8 People's Party considered to be more a historical party than a satellite party, because it is stressing rather its inter-war identity than former communist roots. Polish People's Party and Czech Social Democrates have been also the most successful among historic parties in CEE. Only Visegrad country where historic parties are missing or have disappeared, like in Latvia and in Estonia, is Slovakia. The third difference between the Baltic States and Visegrad countries is that in the Baltic States former Communist parties didn t survive or didn t transform to social democrates. Lithuania is exceptional again. In Estonia and in Latvia the legitimacy of the former communist parties was very weak and almost 50% of the party members were Russian-speakers and hard-line communists (see: Toomla 2005, Runcis 2005). It was impossible to transform them to Western-type social democrats. In Latvia Socialist Party has been the only successor party of the Communist Party of Latvia. But it represents Russian minority and has never got more than 5% of votes, so it has been really marginal. Now it is united with Harmony Centre, what is the major party representing Russian-speakers in Latvia (Millard 2004). In Estonia the only communist successor party was Estonian Labour Party what was also totally marginal and has disappeared as an independent party for now. In Lithuania the situation is pretty different. Todays Social Democratic Party of Lithuania (leader A. Brazauskas) has its roots in former Communist Party. In the beginning of the 1990 s the party transformed itself to parliamentary party and has been the major political party in Lithuania since 1992 (Krupavičius 2005, Romanaite 2006). We can see the same situation in Poland (Democratic Left Alliance - today Left and Democrats), and in Hungary (Socialist Party). In these countries leading left-wing political parties having their orgins in former communist parties (Markowski 2002, Bozoki 2002). Even in Slovakia, despite several splits and mergers and crises, former communists survived and communist successor party, Party of the Democratic Left, was quite successful in several elections (Fisher 2002). Now it is merged with Smer (Direction Social Democracy) the major leftist party (Rybar 2006). In Czech Republic reformist elements were destroyed within Communist Party and therefore the party wasn t able to transform to a Western-type social democratic party (Kostelecky 2002). Despite of this, basing on local party organization and quite strong

9 9 (even pre-war) communist tradition, Communist Party of Bohemia and Moravia managed to survive (Hanley 2002). Since 1996 they have won constantly 10-12% votes (2002 even 18%) and today it is the third major party in the country. To conclude this section we can argue that two Baltic States, Latvia and Estonia are quite different compared with the Central Europen countries, because they are only countries in CEE where communist successor parties didn t manage to survive and also historical parties failed to get sufficent support. Most of the parties in Latvia and in Estonia having their roots in dissident movements or are complely new parties. But Lithuania is different and more similar to Central Europe: former communist party transformed successfully to a social democratic party and historical parties got stronger foothold than in other Baltic countries Ideologies of the parties, party families. Now we should turn back to Ware s analytical scheme and focus on the ideological stances of the parties. We don t have much space here to analyze ideological profiles of every single party, we will rather concentrate on the main tendencies. According to Ware we can distinguish nine families of party ideologies: (1) Liberal and Radical, (2) Conservative, (3) Socialist and Social Democratic, (4) Christian Democratic, (5) Communist, (6) Agrarian, (7) Regional and ethnic, (8) Right-wing extremists, (9) Ecological movement (Greens). As menioned earlier, in CEE countries it is not easy to classify different parties according to this scheme, because the party system is still in formation and ideological labels and policies pursused by the parties are not often corrseponding to each other (see: Lewis 2000). Therefore in CEE contexts party ideologies could have different meanings we are used to see in the West. If we are comparing ideological profiles of the parties in Central Europe and in the Baltic States the first aspect what catches our eye is that in Estonia and in Latvia social democratic parties are almost missing or are quite weak. At same time in Lithuania and in the Visegrad countries we can find quite strong social democratic parties in every country and because this party systems are more ideologically balanced, than they do in Latvia or in Estonia. In Latvia social democratic parties are actually missing on political landscape (see: Pabriks & Štokenberga 2006, Runcis 2005). Despite several attempts in the 1990 s to establish a strong social democratic party in Latvia, all these aspirations finally failed

10 10 (Runcis 2005). In today s Latvia social democracy and leftist ideology is mainly connected with the parties representing Russian minority. Ethnic Latvians are not usually connecting themselves with social democracy as a Russian ideology and they are rather voting for right-wing parties like Peoples Party, New Era Party, etc. Russian parties have never been in any government and despite the fact they have regularly got 20% of votes, they are still playing rather marginal role in Latvian politics (Millard 2004). We should also take into account that Russian parties in Latvia (e.g. Harmony Centre) are not usually declaring, that they are social democrats, they are rather suggesting that they are just left-wing parties, but actually they are even more typical ethnic minority parties. The situation is better in Estonia. There is Social Democratic Party in Estonia, but compared with major parties in the Estonian party system (Reform Party, Union of Pro Patria and Res Publica) what are clearly right-wing parties, social democrats have been rather weak. The party has been in several right-wing governmental coalitions, where it hasn t been usually very powerful and assertive enough to stress their own left-wing agenda. Therefore the Estonian party system is rather unbalanced and rightwing parties are dominating (see also: Mikkel 2006; Krupavičius 2005a) If we are looking at cabinets than almost all cabinets in Estonia (except one) and all in Latvia have been with centre-right dominance (Krupavičius 2005a). This is a quite different picture compared with Lithuania and Central Europe. In Lithuania conservative right-wing Homeland Union has been always balanced by Social Democrats (Romonaite 2006, Novagrockiene 2001). Also in Hungary, Czech Republic and Poland we can see, that one of the major parties is usually a social democratic party, what is opposing to conservatives or liberals (in Czech Republic Civic Democratic Party vs. Social Democratic Party, in Hungary Fidesz vs. Socialistic Party; in Poland Left and Democracy vs. Civic Platform or populist Law and Justice). Even in Slovakia Smer (Direction Social Democracy) has became the major competitor for rightist Christian Democrats or populist Meciar s party (Movement for a Democratic Slovakia). All these countries have seen both left-wing and right-wing governments, not only predominately right-wing governments like Estonia and Latvia. Hence, Estonia and Latvia are the only countries in CEE where we can see an unipolar party competition, where the competition between the major parties is going on mostly on the right-wing axis of the political spectre (See: Mikkel 2006, Kru-

11 11 pavičius 2005a). Lithuania is following rather the Central European pattern, where party competition is rather bipolar and the party system is ideologically quite well established (see also: Krupavičius 2005b). During the last years there has been ideologically rather unipolar party competition also in Poland: two right-wing parties (Civic Platform and Law and Justice) have dominated and the main leftist party, Democrats and Left, has not obtained more than 15%. But at same time in elections 2001 Democratic Left Alliance won 47% and it was quite successful throughout the 1990 s. In two Baltic States we have never seen the situation like this. Of cause in CEE context we should take into account that not all social democratic parties are representing social democratic ideology. For example if we are looking at nowadays Hungarian socialists we can see that the policy they are pursuing is much more liberal and rightist than expected from a mainstream social democratic party. But still most of social democratic parties in Central Europe are emphasizing their social democratic identity, however in government they can act sometimes differently. In Latvia and in Estonia social democracy as an ideology is looked at with some suspicion, because in peoples mind it is connected with communism and Russia. In Central Europe, and even in Lithuania, associations like this are not playing so important role. The second striking contrast between Latvia and Estonia and the Visegrad countries is that in these two Baltic countries Christian democratic parties are almost missing. There is a Christian-Democratic Union in Latvia, but it was able to get 5 seats in parliament only in the 1993, after that they have been absolutely marginal. In Estonia Christian democratic parties have never managed to won any seats in the parliament. This is different from Visegrad countries and from Lithuania. Christian Democrats have been quite successful especially in Slovakia, but also in Czech Republic (see: Kopecky 2006, Rybar 2006). In Lithuania and in Hungary they were moderately successful in the 1990 s, now they are almost extinct and have merged with the larger parties. However leading conservative parties in Lithuania and in Hungary (Fidesz in Hungary, Homeland Union in Lithuania) are still emphasizing Christian values, moral issues and religion. This kind of rhetoric is almost missing in Latvian or Estonian political debate. Also in Poland, where Christian Democratic parties are officially absent, various conservative parties (e.g. Law and Justice) are powerfully stressing reli-

12 12 gious agenda and have taken virtually their place. In Estonia in Latvia there are neither Christian Democrats nor any other parties stressing religion or Christian values Stance of the parties towards the legitimacy of the regime. The third dimension in Ware s approach is the stance of the parties towards the legitimacy of the regime. In CEE countries we can t find any parliamentary party what puts the legitimacy of the regime under question. Even communists in Czech Republic and in Slovakia are not propagating the regime change or restoration of the old communist order. So in CEE context this analytical dimension is not very useful Number of the parties in the system, types of the party systems. The last dimension in Ware s scheme is the number of the parties in the system. There are several ways how to classify party systems according to this variable (see also Siaroff 2000, Wolinetz 2006, Sartori 2005), but we are using here again the classification proposed by Ware. If we take into account the number of relevant parties (parties what have received 3% or more of the seats in the legislature) we can classify both Hungary and Czech Republic as systems with 3-5 parties and all other countries under investigation are representing party systems more than with 5 parties (see Table 1). If we take into account the results of the recent elections we can say that Hungary is moving even towards two party system (Enyedi 2006) and Poland probably towards 3-5 party system. In Lithuanian party system was also a system with 3-5 parties, but not anymore, because during recent years there has been a remarkable de-stabilization. But in post-communist context where party systems are quite fragmented anyway, above-mentioned classificatory method (taking into account only the number of relevant parties) is not sufficient and doesn t make much difference. Ware is proposing an alternative approach where relative size of the parties is taken into account. In this case the share of the votes of the party are compared with the share of the votes of their opponents. Here we can distinguish between 6 different party systems: Predominant party system more than one relevant party, but one party ever controls the legislature and obtains usually more than 50% of votes (e.g. Japan).

13 13 Table 1. Party systems in the Visegrad countries and in the Baltic States Founding elections I elections II elections III elections IV elections V elections Hungary Seats won by the first party (%) 42,5 (Democr. Forum) 54,1 (Socialist Party) 34,7 (Socialist Party) 48,7 (FIDESZ-MDF) 48,2 (Socialist Party) Seats won by the second party (%) 23,8 (Free Democrats) 18,4 (Free Democrats) 38,3 (FIDESZ-MPP) 46,1 (Socialist Party) 42,5 (FIDESZ-KDMP) Seats won by 2 major 66,3 72, ,8 90,7 parties together (%) Number of relevant parties Type of the party system One large party, several smaller Pre-dominant party system Two large parties Two-and-a-half or two-party system Two-and-a-half or two-party system Czech Republic Seats won by the first party (%) 67 (Civic Forum) 38 (ODS a ) 34 (ODS) 37 (ODS) 35 (Social Dem. P.) 40,5 (ODS) Seats won by the second party (%) 15,3 (Communist P) 17,5 (KSČM b ) 30,5 (Social Dem. P.) 32 (Social Dem. P.) 29 (ODS) 37 (Social Dem. P.) Seats won by 2 major 82,3 55,5 64, ,5 parties together (%) Number of relevant parties Type of the party system Pre-dominant party system One large party, several smaller Two large parties Two large parties Two large parties Two large parties Slovakia Seats won by the first party (%) 41,9 (Public Against Violance) 49,3 (HZDS c ) 40,7 (HZDS) 28,7 (HZDS) 24 (HZDS) 33 (Smer) Seats won by the second party (%) 20,2 (Christian Dem. Movement) 19,3 (Democratic Left) 12 (Democratic Left) 28,0 (Slovak Dem.. Coalition) 18,6 (Slovak Dem & Christian Union) Seats won by 2 major 62,1 68,6 52,7 56,7 42,6 53,6 parties together (%) Number of relevant parties Type of the party system One large party, several smaller One large party, several smaller One large party, several smaller Even party system Even party system 20,6 (Slovak Dem & Christian Union) Even party system Poland Seats won by the first 13,5 37,2 43,7 47,2 33,7 45,4 party (%) (Dem Union) (SLD) (Solidarity Elec- (SLD) (Law and Justice) (Civic Platform) toral Action) Seats won by the second 13 28,7 35,7 14,1 28,9 party (%) (SLD d ) (Peasant Party) (SLD) (Civic Platform) (Civic Platform) Seats won by 2 major 26,5 65,9 79,4 61,3 62,6 81,5 parties together (%) Number of relevant parties Type of the party system Even party systetieties Two large par- Two large par- One large party, Two large parties several smaller Lithuania Seats won by the first party (%) 51,7 (Dem Labour P) 49,6 (Homeland Union) 31,9 (Dem Labour P) 27,6 (Labour Party) Seats won by the second party (%) 19,8 (Homeland Union) 8,5 (Dem Labour P) 24,1 (Liberal Union) 21,9 (Social Dem + New Union) Seats won by 2 major parties together (%) 71,5 58, ,5 Number of relevant parties Type of the party system Two-party systetem Two-party sys- Even party system (Table continues) a ODS Civic Democratic Party b KSČM Communist Party of Bohemia and Moravia c HZDS Movement for Democratic Slovakia (Meciar s party) d SLD - Alliance of Democratic Left Even party system 36,1 (Law and Justice) Two large parties

14 14 (Table continues) Founding elections I elections II elections III elections IV elections V elections Latvia Seats won by the first party (%) 36 (Latvian Way) 18 (Saimnieks) 24 (Peoples party) 26 (New Era) Seats won by the second party (%) (LNNK) (Peoples Movem.) (Latvian Way) (Human Rights) Seats won by 2 major parties together (%) Number of relevant parties Type of the party system Even party systetetem Even party sys- Even party sys- Even party system Estonia Seats won by the first party (%) 28,7 (Pro Patria) 40,6 (Coalition party) 27,7 (Centre Party) 27,7 (Res Publica) Seats won by the second 16,8 18,8 17,8 27,7 party (%) (Safe Home) (Reform party) (Pro Patria) (Center Party) Seats won by 2 major 45,5 59,4 45,5 55,4 59,3 parties together (%) Number of relevant parties Type of the party system Even party system One large party, Even party sys- Even party system several smaller tem e LNNK Latvian National Independence Movement Sources: Kostelecky 2006, Jungerstam-Mullers 2006, Lewis (Peoples party) 18 (Greens&Farmers) Even party system 30,6 (Reform party) 28,7 (Center Party) Even party system Two-party system one of two parties wins at least 50% of seats (e.g. UK and USA). Two-and-a-half party system no one party wins 50% of seats and two larger parties usually win at least 80% of seats, while the third party wins sufficient seats to hold the balance of power (e.g. Germany). System with one large party and several smaller ones one large party usually wins at least 45%, but not 50% (e.g. Sweden). System with two large parties two larger parties usually obtain more than 65% of seats, but no other party obtains more than 14% (e.g. Italy). Even party system the larger party obtains less than 45% of seats and the two larger parties obtain less than 65% of seats. If we are looking at first the share of seats obtained by two largest parties (see Table 1) we can see that in Hungary, Czech Republic and even in Poland (despite of several fluctuations) it has been usually over 65% after the mid 1990 s. It has been higher than in the Baltic States and in Slovakia, where the corresponding figure has been usually 50-60% (in Latvia even lower). If we want to analyse the evolution of the party systems in the countries under investigation, we are proposing to detect the type of the party system after every elec-

15 15 tions (see: Table 1). Of cause the party systems are supposed to be more stable phenomenon, but in post-communist context they are constantly changing and therefore this approach would be more useful. On the following pages we will give also a brief overview about the party systems in countries under investigation, because Table 1 is probably not providing sufficient information for understanding some specific aspects. We can see that from the 2000 s Hungary has moved towards two-and-a-half party system and since the last elections we can even talk about two-party system. We should take into account that during all elections since 1998 the two major parties have been the same: Fidesz and Socialist Party, and no new relevant parties haven t entered to the political arena since So Hungarian party system has been quite stable and characterized by fierce struggle between two major parties, ideologically in opposite camps (Enyedi 2006). Czech party system has also demonstrated remarkable stability compared with other CEE countries. Since the mid 1990 s we can see that there was taking shape a two large party system. These two large parties are Civic Democratic Party (ODS) and Social Democratic Party both are usually winning more than 30% of seats. The third position has been always occupied by Communist Party of Bohemia and Moravia and the fourth one by Christian Democrats. But both above-mentioned parties have almost never obtained more than 14% of seats. There have been also some newcomers in Czech politics (Greens in 2006, Freedom Union in 1998), but they have always won less than 10% of seats in the legislature and haven t much changed the overall picture. In Poland the situation has been more complicated. The Polish party system has been less stable than in Hungary and in Czech Republic: major parties have merged, split and changed the names several times during the recent decade (see: Szczerbiak 2006, Jurgerstam-Mulders 2006b). But during the recent years we can see that two major parties, Civic Platform and Law and Justice, have constantly obtained over 60% of seats and other larger parties (Democratic Left Alliance or Self Defence) less than 14%. It seems that Polish party system is probably also moving towards two large party system, but maybe it is still too early to make such kind of conclusions. Anyway, the party competition in Poland hasn t been characterized by so clear ideological bipolarity like in Czech Republic and in Hungary. The last mentioned countries are the only ones in CEE representing clearly polarized or bipolar party competition (see also: Toka & Henjak 2006)

16 16 If there started some stabilization in Hungary and in Czech Republic in the end of 1990 s, Slovakia and Lithuania have moved to opposite direction: the former, already quite stable party systems, literally blew up in the end of 1990 s or in the beginning of 2000 s and were replaced with more unstable configurations. In Slovakia Meciar s party (Movement for a Democratic Slovakia) was dominating and there was a party system with one large party with several smaller ones (some scholars e.g. Lewis 2000 called this even hegemonic party system). The situation changed in 1998 when an anti-meciar coalition was formed (Slovak Democratic Coalition), what also won elections. Before elections 2002 there were several splits and merges and some older smaller parties lost their support (Rybar 2006). The party system became even more fragmented. Since 1998 we can talk about the Slovakian party system as an even party system with more than 5 parties. Party system in Lithuania was considered to be quite stable before 2000 and some scholars were talking even about the two-party system (see: Novagrockiene 2001; Romanaite 2006, Jurkyans 2004). Dominant parties were conservative Homeland Union and social democratic Labour Party of Lithuania. Voters were supporting one of them and punishing severly the other one, and in next elections on the opposite way. Elections in 2000 brought the major change, because the center was empty and both big parties had finally disgredited themselves for the voters (Jurkyans 2004). Some successful new comers emerged (e.g. New Union) or some older smaller parties got remarkable support (e.g. Liberal Union). In 2004 elections the new populist Labour Party (not the same as former Labour Party of Lithuania) was the most successful party and this demonstrated very clearly that Lithuanian party system has destabilized in quite a big extent compared with 1990 s. In the we could speak about two-party system or at least about two large party system, but since 2000 about even party system like in other Baltic States. Krupavičius (2005a) is also suggesting that during recent years Lithuanian party system has become more similar to Latvian and Estonian party systems than it was before. But despite of this he points out two aspects what still have made Lithuanian party system a little bit more stable compared its Baltic neighbours: the strength of the largest parliamentary party (share of seats obtained by the largest party) has been usually higher in Lithuania than in Estonia and in Latvia (see also Table 1) and the share of new parties in parliament has been also much lower (in Lithuania in average 37% of parties have been new ones in the parliament, in Latvia 61% and in Estonia 51%). If we are looking at number of

17 17 relevant parties we can also see that Lithuanian party system has been less fragmented than Estonian or Latvian ones. Both Latvia and Estonia are representing even party systems with more than 5 parties. In Estonia the share of seats won by two largest parties has been traditionally higher than in Latvia. Estonian party system is considered to be also more stable compared with Latvian party system (Meleshevich 2007, Jungerstam-Mulders 2006b). In Estonia there have been two to three major parties: Centre Party, Reform Party and Pro Patria (now Union of Pro Patria and Res Publica). Two most successful parties have usually obtained less than 1/3 of seats and all together around 55% of seats. It has been lower than major parties have obtained usually in Visegrad countries (except Slovakia). Estonian party system hasn t been very stable either, because there have been several splits and mergers even among major parties (Pettai & Kreuzer 2003; Grofman, Mikkel & Taagepera 2000). In 2003 for example a completely new populist party Res Publica won elections and some new comers have been successful even before (1995 Reform Party). But still share of votes and seats won by toatally new parties has been lower than in Latvia, but much higher than in Hungary and in Czech Republic (see: Sikk 2005). Party system in Latvia has been probably the most fragmented and unstable among the countries under investigation. In Latvia before every elections a new party (or several new parties) is founded and it usually wins elections and after that it dissappears or looses the support (Pabriks & Štokenberga 2006). In 1995 elections it was Saimnieks, in 1998 Peoples Party and in 2002 New Era Party and only in last elections 2006 we didn t see any new party entering to political arena. We can all this pheneomenon as a Latvian desease. The share of seats obtained by two largest parties has been the lowest in CEE (see Table 1) and the share of votes and seats won by new parties has been one of the the highest in CEE (see: Sikk 2005). If we try to compare party systems in the Visegrad countries and in the Baltic States, basing on the previous analysis, we can say that Hungary and Czech Republic and even Poland have moved towards more consolidated two-and-a-half system (Hungary) or two larger parties system (Czech Republic, Poland). But at same time there have been destabilization tendencies in Lithuania and Slovakia since 2000 s and they have moved towards more fragmented even party system. The last type has been always dominating in Latvia and in Estonia either.

18 18 But even in this case Latvia and Estonia are still different from Lithuania and Slovakia, because in two Baltic States there has never been a period where we can speak about two-party system like in Lithunia in or the party system with one dominating party like in Slovakia The strenght of the largest party in parliament has also been usually lower than in Lithuania and Slovakia. Therefore Latvian and Estonian party systems have been quite fragmented since the fall of communism, but Slovakia and Lithuania have seen at least some periods of consolidation Effective number of electoral and parliamentary parties, splits and mergers. Before to decide about the stability of the party systems, we should look at some additional indicators: effective number of electoral and parliamentary parties, statistics about splits and mergers. Effective number of electoral parties is considered to be a major indicator for measuring the fragmentation of the party system and it is basing on Laakso and Taagepera index (Laaksoo & Taagepera 1979). The index can be used both regard to parliamentary parties and electoral parties. If we are measuring the effective number of electoral parties, the calculation is basing on the share of votes that each party receives in elections. If measuring the effective number of parliamentary parties the share of seats in the parliament is taken into account. Table 2. Effective number of electoral parties in Visegrad countries and in the Baltic States Founding elections I elections II elections III elections IV elections V elections Average (Jungerstam- Mulders 2006a) Hungary 6,4 (1990) 5,2 (1994) 4,4 (1998) 2,8 (2002) 2,7 (2006) - 4,20 Czech Republic 3,1 (1990) 6,2 (1992) 5,3 (1996) 4,7 (1998) 4,8 (2002) 3,9 (2006) 4,96 Slovakia 5,4 (1990) 5,4 (1992) 5,8 (1994) 5,3 (1998) 8,9 (2002) 6,1 (2006) 6,67 Poland 13,2 (1991) 9,8 (1993) 4,6 (1997) 4,5 (2001) 5,9 (2005) n/d 4,53 Lithuania 3,8 (1992) 7,9 (1996) 5,6 (2000) 5,8 (2004) - 6,43 Latvia n/d n/d n/d n/d n/d 7,74 Estonia 8,9 (1992) 5,9 (1995) 6,9 (1999) 5,4 (2003) n/d 6,06 Source: Data about Visegrad countries from Toka & Henjak 2006: 22; data about the Baltic States from Rominaite 2006, Mikkel 2006; average ( ) from Jungerstam-Mulders 2006a

19 19 Table 3. Effective number of parliamentary parties in the Visegrad countries and in the Baltic States Founding elections I elections II elections III elections IV elections V elections Average (Jungerstam- Mulders 2006a) Hungary 6,7 (1990) 5,6 (1994) 4,26 (1998) 2,4 (2002) 2,4 (2006)* - 2,83 Czech Republic n/d n/d n/d n/d n/d n/d 3,78 Slovakia n/d n/d n/d n/d n/d n/d 5,09 Poland 10,5 (1991) 3,18 (1993) 2,95 (1997) 3,6 (2001) 4,7 (2005)* n/d 3,28 Lithuania 2,99 (1992) 3,32 (1996) 4,79 (2000) 6,13 (2004) - 3,73 Latvia n/d n/d n/d n/d n/d 6,03 Estonia 5,9 (1992) 4,1 (1995) 5,5 (1999) 5,6 (2003) 4,4 (2007)* 4,77 Source: Data about Hungary and Poland from Enyedi 2006, Szczerbiak 2006; about the Baltic States from Rominaite 2006, Mikkel 2006; average ( ) from Jungerstam-Mulders 2006a * Biezen & Caramani 2007 calculations If looking at effective number of electoral parties , we can see that the least fragmented party system is in Hungary and in Czech Republic (see: Table 2). In both countries the number of effective electoral parties has constantly decreased since the beginning of the 1990 s. In Polish party system was also not very fragmented and rather similar to Czech Republic and to Hungary, but recent elections entailed a change. The picture is pretty same if we take the effective number of parliamentary parties (see: Table 3). Parliaments in Hungary, Czech Republic and even in Poland seem to be less fragmented than in the Baltic States and in Slovakia. Especially clear trend towards more consolidated party system is visible in Hungary and also in Czech Republic. This confirms our previous conclusion, that in these countries we can see a greater stabilization than in other CEE countries. Party systems are more fragmented in all three Baltic States and in Slovakia. If we take into account the effective number of parliamentary parties the Lithuanian parliament is the least fragmented. But if we are looking also at effective number of electoral parties we can t say that the Lithuanian party system is much more consolidated than in other Baltic States. Actually Estonia seems to be only country in this group where we can see clearer trend towards greater consolidation. Lithuania and Slovakia are representing rather an opposite trend: today s party system in these countries is even more fragmented than in was in the 1990 s. Latvia seems to be again a leader in negative sense, because Latvian party system seems to be the most fragmented compared with all other countries.

20 20 We should also look at splits and mergers, because these indicators are also reflecting the stability of the party system (Table 4). We can see that Hungary and Czech Republic are again the most stable examples, but also Lithuania, where we can see less splits and mergers than in other Baltic States. Estonia and Slovakia seem to be even less stable than Latvia and even Poland. The latter one is famous because its frequent fission tendencies among parliamentary parties (see: Szczerbiak 2006). In Estonia mergers are dominating, but in Slovakia there are even more splits than mergers, what makes Slovakian party system even less stable. Table 4. Splits and mergers in the Visegrad countries and in the Baltic States Splits Mergers Total Hungary Czech Republic Slovakia Poland Lithuania Latvia Estonia Source: Biezen & Caramani 2007 Pettai and Kreuzer (2003) have also analyzed splits and mergers in the Baltic States. Their conclusions have been quite similar: there were least fissions and fusions in Lithuania, in Estonia rather fusion tendencies were dominating, but in Latvia the dominant strategy was to start up with the completely new party. To conclude this section we can suggest that the overall picture is quite complicated. If we are analyzing the basic aspect of the party systems: party ideologies, origins of the parties, types of the party systems - two Baltic States Latvia and Estonia are quite distinct in several important aspects. If we take into account some more formal characteristics, like effective number of parties, splits and mergers Lithuania and Slovakia seem to be more similar to the Baltic States than to some Visegrad countries like Czech Republic and Hungary. Slovakia is even more like Estonia and Latvia. Lithuanian party system is still a bit more stable and less fragmented (look at splits and mergers and effective number of parliamentary parties) than party systems in two other Baltic countries.

21 21 2. Electoral behavior In this section we will briefly analyze two important indicators: volatility and voter turnout. Both of these indicators are very salient in the point of view of nationalization of the party system and electoral behavior (Caramani 2004). There are different types of volatilities and different formulas for calculating these (Pedersen 1979). We should be extra careful in post-communist context because the parties are constantly changing names, there are splitting and merging. Therefore we should decide, for instance, whether to regard a merged party or a party with a new name as a totally new entity or not to do this (see: Sikk 2005, Toka & Henjak 2006; Birch 2001). This is also one of the reasons why different authors are getting different results while calculating volatility in post-communist countries (see: Lewis 2000: 85). We selected the data provided by Biezen and Caramani (2007), because they have also included the latest elections into their calculations and they are using Pedersen index what is considered to be the most widely used index for electoral volatility. In table 5 we can see, that the lowest volatility has been in Czech Republic, Slovakia and Hungary and the highest in Lithuania, Poland and Latvia. Estonia is rather a country with a moderate volatility in CEE context. Jungerstam-Mulders (2006) reached to quite similar conclusions in her analysis. Table 5. Electoral volatility in the Visegrad countries and in the Baltic States 2 nd elections 3 rd elections 4 th elections 5 th elections 6 th elections Mean Hungary 24,2 34,6 20,8 7,8-21,9 Czech Republic 19,3 31,6 14,8 14,1 17,3 19,4 Slovakia 20,2 17,0 19,5 16,0 25,3 19,6 Poland 30,6 22,6 49,7 32,1-33,8 Lithuania 37,4 47,5 49, ,9 Latvia 42,9 42,1 43,4 13,4-35,5 Estonia 27,3 34,2 27,3 21,6-27,6 Source: Biezen & Caramani Note: Figures indicate total volatility without others (TVWO), including parties which obtained at least two percent of the vote. Volatility is calculated using the Pedersen index (1979): Total Volatility = Σ Pi n t Pi n t+1 / 2, where P is the percentage of votes for parties i to n in elections t and t+1. Hence there is no clear borderline between the Baltic States and the Visegrad countries: Volatility in Estonia is lower than in Latvia and Lithuania and in Poland, but it is higher than in other Visegrad countries. If we are looking at the trends, we can see that volatility has especially declined in Hungary and in Czech Republic, what again proves our previous argument that the