Left and Right in new democracies: the Baltic States and Southern Europe. André Freire (ISCTE Lisbon University Institute, and CIES-IUL, Portugal)

Transcription

1 Left and Right in new democracies: the Baltic States and Southern Europe compared André Freire (ISCTE Lisbon University Institute, and CIES-IUL, Portugal) Kats Kivistik (University of Tartu, Estonia) Paper to be presented at the: IPSA 2012 Madrid 2012, XXII World Congress of Political Science Panel: Left and Right in Long Consolidated and New Democracies: Exploring and Explaining the correlates of the divide Chair: André Freire (ISCTE Lisbon University Institute, Portugal) Co-chair: Kats Kivistik (University of Tartu, Estonia) Discussant: Oddbjørn Knutsen (University of Oslo, Norway) Left and Right in new democracies: the Baltic States and Southern Europe compared

2 Abstract (originally proposed in the IPSA website) The left-right ideological dimension is one of the most common tools for comparing political parties and mass beliefs across democratic societies all over the world, but lately scholars have started to call in question the relevance and the usefulness of the concept and that at least partially because of the not unambiguous meaning of it. For instance, research shows that left and right can be understood differently or even in reverse terms in different countries if Left is associated with support for innovative views and reforms in Western Europe, then in Eastern Europe the association is rather opposite. While studying the content of Left and Right, usually only individual level variables have been included into the analysis. The purpose of this study is to take into consideration also the context of the societies and especially the type of the precedent regime. In order to do so, the study encompass three countries with communist past (Baltic countries) and three countries with right-wing authoritarian past (Portugal, Spain, Greece) trying to clarify which are the similarities and differences between these countries and how much are the developments of Left and Right identities associated to their authoritarian legacy. Analysis is based on (national) survey data covering as long period from post-authoritarian era as possible.

3 Left and Right in new democracies: the Baltic States and Southern Europe compare Introduction The formation of ideological (and partisan) identities is a process that continues to evolve along the courses of socialization (primary and secondary) of individuals (Campbell, 1980; Converse, 1964 and 1969; Niemi at. al, 1985; Barnes, McDonough and Pina, 1985). Moreover, the formation of either of these two identities is strongly dependent upon the existence of an environment of effective partisan and ideological differentiation, associated with the existence of free political competition. In the right-wing authoritarian regimes existent until the beginning of the 1970's in Southern Europe, the political parties were a proscribed reality, and ideological differences were repressed. In the second half of the twentieth century, with the exception of Greece, partisan and ideological pluralism were only implanted there following the democratic transitions of the 1970's. In spite of partisan and ideological pluralism being repressed by the authoritarian regimes of Southern Europe, this does not mean that such pluralism did not exist at all. Regardless, it was very limited (Freire, 2006). On the other hand, during the authoritarian period, the experiences and contacts with the notions and the political organizations associated with such pluralism, whether it was by way of oppositional currents (to the

4 authoritarian regimes), or by way of contacts with foreign democratic nations, always reached limited segments of the populations of Portugal, Spain and Greece. Therefore, the conditions for the formation of ideological (and partisan) identities were severely reduced, above all when we compare these new regimes with the older democracies of Western Europe. On the other hand, The Baltic States were under USSR rule at least since the end of Second World War and thus lived under a single party totalitarian regime since at least since 1945 and until the Post WWII independence, However, contrary to the new Southern Europe, in the Baltic States there was a left-wing totalitarian regime. Moreover, also contrary Southern Europe, whose right-wing dictatorial regimes were more closed to the authoritarian prototype, the left-wing dictatorial regimes in the Baltic States were more close to totalitarian prototype (about the concepts of authoritarian and totalitarian regimes, see Linz and Stepan, 1996). Among other things, this means that in the latter (totalitarian) regimes mass political participation was more extensive, even if it was clearly top-down controlled, and the political systems were characterized by a higher level of ideological structuration. On the contrary, authoritarian regimes (like those in Southern Europe) are characterized by a comparatively lower level of ideological structuration and, additionally, by lower levels of mass political mobilization (except perhaps in exceptional periods) because the regime invests on a certain demobilization of the citizenry.

5 Moreover, the different regime heritages and the different trajectories of the democratic transition in the two sets of countries (Southern Europe vis-à-vis the Baltic States) imply that there are different types of winners and losers with the change of regime. I.e., in the two sets of countries, due to the replacement of old regime (the left-wing communist regime, in the Baltic States, and the right-wing authoritarian regime, in Southern Europe), a repositioning took place, which allowed some people to gain better position while others lost because their position grew worse suddenly (Daatland: 1997, Whitefield 2002; Zarycki 2000, Jurkynas 2003, 2004, Lie 2006; Kivistik, 2012). From these very different political contexts in the Baltic States versus Southern Europe, we derive our research question and also our research design. Our research question reads as follows: in a new democracy, does the type of authoritarian legacy (right-wing vs. leftwing dictatorship, and also authoritarian vs. totalitarian regime) has any effect on the way citizens think about and use the left-right political divide? In terms of research design, considering the very different authoritarian heritages, but also the very different historical trajectories (recent independence after long colonial rule, in the Baltic States, and at least a one and a half century independence, in Southern Europe), we of course rely on the most dissimilar systems comparative design. To answer our research question, relying on several mass comparative surveys we will analyse and compare the correlates and behaviour consequences of the left-right divide in the six countries under

6 consideration (Estonia, Latvia, and Lithuania, for the Baltic States, and Greece, Portugal, and Spain, for Southern Europe) in three major dimensions. First, in terms of LR recognition: aggregate percentage of persons in each country that are able to place themselves in the leftright scale. The major expectation here is the following: due to a higher level of mobilization of the citizenry in the Baltic States (vis-à-vis Southern Europe, except for Greece see below), and also due to higher levels of ideological structuration of those political regimes, we expect to find higher levels of LR recognition in the Baltic States than in Southern Europe, especially in the first years of the new democratic regimes because in the long run we expect some convergence across the six countries. Greece, however, should stand as a case apart vis-à-vis the other five countries, with the highest levels of LR recognition, because after WW II the authoritarian interlude ( ) lasted only for around 7 years, and several continuities existed between the democratic regimes that existed before and after the colonels dictatorship (see Clogg, 1987). Second, we will analyse the attitudes of left-wingers and rightwingers vis-à-vis the new democratic regimes, as well as their nondemocratic alternatives, and confidence in its key political institutions, in the different countries. Here, due to the different ideological leaning of the previous authoritarian regimes (left leaning in the Baltic States, right leaning in Southern Europe), and the different types of winners and losers with the regime change in the two sets of countries, we expect

7 that left-wingers are more supportive of democracy (than right-wingers) and more critical of the old regime in Southern Europe, and that in the Baltic States it is the other way around. Again, dilution of these patterns and cross country convergence is also expected with the passage of time. Third, both before and after controlling for social factors, we will analyse the value correlates (in terms of socioeconomic values, libertarian-authoritarian values, and post materialism) of the left-right divide of the left-right divide in the six countries. Here we don t have so clear and straight hypotheses because, bearing on the South European experience of great internal diversity, we believe that the substantive value content of the left-right divide will depend not so much on the type of authoritarian legacy, but more on type of democratic transition and on the type of politicization of the different issues that the political parties do since then. In next section, we will describe briefly the major social, political, and institutional characteristics of the new democratic regimes in both the Baltic States and in Southern Europe, as well as the major heritages from the old totalitarian or authoritarian regimes, and the major changes with the democratic transition. In the second section, we will review the relevant literature on the topic, and then formulate our major hypothesis. In the third section, we will describe the data and methods that will be used, as well as the concept operationalization. In the fourth section, we test our hypotheses with the empirical evidence resulting from the

8 comparative mass surveys we will be using. The paper ends with some concluding remarks. The new democratic regimes and their authoritarian legacies: Southern Europe and the Baltic States in comparative perspective Although there is significant variation between Greece, which was under Ottoman rule from since the XV century until at least the XIX century (around 1830), vis-à-vis Portugal and Spain, because the latter two countries are among the oldest independent states in Western Europe (and stayed independent for most of the time since their formation in early Middle Age), the truth is that even in Greece (the birthplace of western civilization) there is a long and deeply rooted national tradition (Bruneau, 2001; Caramani, 2000). On the contrary, with the partial exception of Lithuania, which succeeded to create an independent state during the medieval ages (although in strong connection with the Polish kingdom), the Baltic countries are very recent independent political entities: they only became independent states after the end of WW I, but lost that status by the end of the 1930s (following the Nazi-USSR pact, and the Baltic invasion by the USSR), and were under Soviet rule since the end of WW II until 1991, when they finally recuperated independence again (Plakans, 2011). Moreover, the three southern European countries were early

9 adherents to the winds of political liberalism (Constitutionalism, rule of law, separation of powers, elections to choose representatives, party pluralism, governments accountable to the people, etc.) that blew from the American and French revolutions in Western Europe (Caramani, 2000). On the contrary, during the XIX century and the beginning of the XX century the Baltic States were under Russian rule and thus the winds of political liberalism only arrived properly during inter-war independence (Plakans, 2011). But if all the three southern European countries were early adherents to the winds of political liberalism and republicanism, the truth is that their democratization trajectories were also marked by frequent reversals in the nineteen and twenty centuries, and that the consolidation of full democratic regimes only happened pretty late when compared with the rest of Western Europe. In the case of Greece, whose present borders and population structure only came into being by the 1920s, during WW II the country was marked by Nazi occupation, which put an end to the last vestiges of a multicultural society (with the killing and deportation of the Jews). After liberation a civil war arose opposing the communists to the pro western liberal forces, This civil conflict ended up with the defeat of the communists, after around deaths, and the new democratic regime, , was precisely marked by a ban on the communist party. This latter element, besides some manipulation of electoral results and political persecutions and even assassinations, made the new regime less than fully

10 democratic (Clogg, 1987). Nevertheless, only after the colonels coup (1967) democracy was turned down and a right-wing dictatorship was installed ( ). In the cases of Portugal, , and Spain, 1939, the liberal (and democratic, in the latter case) regimes were overthrown much earlier and the right-wing dictatorship lasted for much longer times, respectively until 1974 and Portugal (April 1974), Greece (November 1974), and Spain (November 1975) then inaugurated the so-called third-wave of democratization world-wide. The path through democracy was, however, significantly different in the three countries. First, in Portugal democracy was implemented following a military coup and a revolutionary turmoil which was mainly dominated by left-wing forces. Second, in the case of Greece the collapse of the dictatorship followed the lack of success of the Greek colonels to influence an extreme right-wing military coup in Cyprus. Following that collapse the democratic transition was mainly conducted by right-wing politicians who organized a referendum on the type of regime (the monarchy, associated with the colonels dictatorship as well as with other dark periods in Greek history, was abolished by then) as well as the first completely free and fair elections (the ban on the communist party was then collapsed). In Spain, democratic transition took place shortly after the death of the dictator (Franco), in November 1975, through political pacts between the liberalizing elites originated in the former authoritarian regime, which took the lead of the democratic transition, with the left-wing opposition forces (communists and socialists forces,

11 above all). The Baltic States only knew parliamentary rule and multiparty democratic competition, with constitutionalism, free and fair elections and the rule of law, during a short period in their inter-war independence (Plakans, 2011, pp ). However, shortly after the democratic regimes turned into more or less soft authoritarian solutions: both in Lithuania, 1926, and in Estonia and Latvia, 1934 (Plakans, 2011, pp ). As said before, with the Nazi-USSR Pact come Soviet invasion, the Nazi invasion, and with the German defeat in WWII come USSR rule in the Baltic States, which came to be three of the several USSR Republics. Democracy only came around 45 years later, following the openness in the USSR (with Perestroika) and the formation of the «Popular Fronts» which in all the three countries were able to win most of the seats in the assemblies following the first multi-party elections during Soviet rule (1991). Declarations of independence followed and in 1991 the three Baltic States were accepted as UN members, and independence recognized by Russia (Plakans, 2011, pp ). Despite this common framework, some significant differences occurred between Lithuania and the two other countries. First, in Lithuania the former communist party, especially some of its more liberal members, was able to adapt to the new winds earlier, still during USSR rule, and then was able to convert itself into a social democratic party (LDLP: Lithuanian Democratic Labour Party) that has been able to alternate in government with parties from the right, alone or in coalition with other left-wing or centrist parties (Rose

12 and Munro, 2009, pp ; Müller-Rommel and Hansen, 2001; Blondel and Müller-Rommel, 2001). Besides, Russian migrations to Lithuania during Soviet rule were much smaller than in the case of Latvia or Estonia, and thus the potential of conflict around these cultural-ethnic issues was severely reduced in the first country. In 1918 and in 2001, around 84% and 83% of the Lithuanian population, respectively, was composed of native Lithuanians. On the contrary, in Estonia, in 1920s the native population comprised around 88%, but in 1989 they were only around 65% of the population; in Latvia the native population passed from around 80%, in the 1920s, to around 53%, in 1989) (see Rose and Munro, 2009, pp , 157, ). Besides, this element also facilitated the affirmation of the social democratic left in Lithuania (Jurkynas, 2004), and created further obstacles to its affirmation in both Estonia and Latvia, where at least a large part of the left is associated with the Russian and other Slavic minorities (see Jerschina and Górniak, 1997). As said before, the third wave of democratization worldwide begun in Southern Europe with the collapse of the right-wing dictatorships: Portugal, April 1974; Greece, November 1974; Spain, November The new democratic regimes (two Republics, with parliamentary and semi-presidential political systems in the former cases, respectively; a Constitutional Monarchy with a parliamentary system in the case of Spain) were marked by the choice of very disproportional PR (Proportional Representation) electoral systems, in the cases of Greece

13 and Spain, which are behind the low party system fragmentation. In the case of Portugal, the PR electoral system is not so disproportional but its level of disproportionality is nevertheless above the average for West European PR systems (Freire, Meirinho and Moreira, 2008). The first two systems produced party regimes dominated by two major parties (PSOE vs. PP/AP/UCD, in Spain, and PASOK vs. ND, in Greece) which have been alternating in government alone (always in Greece, most of the time in Spain), or with parliamentary support from very small parties (regionalist or and/or leftist) in the case of Spain. In the case of Portugal, always with the same PR system, there are two phases, one marked by party system fragmentation and coalition government, ; and another marked by vote concentration in the two major parties and single party majority government (or almost single party majority government), current date (although with some right-wing coalitions). A major difference between the three Southern European countries in terms of the clarity of left-right governmental and policy alternatives is that in both Greece and Spain there is full alternation between left (PASOK, PSOE) and right (ND, UCD/AP/PP), although sometimes the major party as to rely on tiny support parties (Spain). But in the case of Portugal, there is a major difference between the left and the right because although the right can cooperate and build coalitions (PSD with CDS-PP), the left cannot (PS vs. PCP and BE). Thus, whenever the socialists win they have to rely on right-wing support to govern (either formally or informally), except when there is a single-party socialist majority (only once: 2005-

14 2009). These differences tend to blur the distinction between left and right, in the case of Portugal, or to underline them, in the cases of Greece and Spain. In the Baltic States, democratization began with Perestroika ( ), and was achieved with full independence from USSR (1991) and the first free and fair elections to the new parliaments (and presidential offices), All the three new democratic regimes are Republics, but although both Estonia and Latvia have parliamentary systems, Lithuania established a semi-presidential political system with a reasonably strong president (Blondel and Müller-Rommel, 2007; Rose and Munro, 2009). In terms of electoral systems, both Estonia and Latvia have PR systems with a high level of proportionality, although more the latter than the former due a higher average of district magnitude. Lithuania on the contrary established a mixed parallel system with seats in each component. All the three systems have fragmented party systems with a strong probability of producing coalition governments, although more Estonia and Latvia than Lithuania (see Blondel and Müller-Rommel, 2007: p. 27; Rose and Munro, 2009: pp. 12 and 19). A major difference between the three Baltic countries in terms of the clarity of left-right governmental and policy alternatives is that both Estonia and Latvia have had only coalitions dominated by centre or centre-right parties, even if sometimes including left of centre parties, thus not allowing the voters to know clearly the difference between left and right in terms of government outputs. On the contrary, Lithuania has known alternation

15 between centre-left dominated governments (either single party or coalition) and centre-right dominated governments (coalitions) (see AA.VV., 2012; Blondel and Müller-Rommel, 2007; Rose and Munro, 2009; Müller-Rommel and Sootla, 2001; Müller-Rommel and Norgaard, 2001; Müller-Rommel and Hansen, 2001). Thus, these differences tend to not allow the voters to know clearly the distinction between left and right, in the cases of Estonia and Latvia, or to let them know those distinctions more clearly, in the case Lithuania. When competing for voters support, parties present different packages of public policies, each with different levels of priority. Both the packages of public policies and their relative priority are related to the issue dimensions of partisan conflict (Lijphart, 1999: pp ). Lijphart emphasizes the need to distinguish between the dimensions of policy competition and the characteristics of the voters that parties represent (Lijphart, 1999: p 78). This author refers to seven issue dimensions of policy completion ( socioeconomic, religion, urbanrural, cultural-ethnic, post-materialism, regime support, foreign policy ). Two relevant questions here are: what are the relevant dimensions of policy competition in each one of the six countries?, and how far can they be subsumed in the left-right distinction? In the case of Spain, the relevant dimensions are socioeconomic, religion (low salience), cultural-ethnic, and post-materialism (low salience). Except for the cultural-ethnic dimension, which cuts across the left-right divide splitting both left and right parties in terms of center and

16 periphery, all of them can more or less be subsumed in the left-right divide. In the cases of Greece and Portugal, the relevant dimensions are socioeconomic, religion (low salience), post-materialism (low salience), and regime support / foreign policy (low salience after the end of the cold war). Besides, although some of them ( regime support / foreign policy ) cuts across the left-right divide (separating the communists and the radical left, more anti-capitalist and critical of international bloc alignments, from the socialist and the bourgeois parties, pro capitalist, pro EU and pro NATO), that political cleavage subsumes the other issues. In the Baltic States, regime issues related with attitude towards the previous communist heritage (communist vs. anti-communist split), as well as issues related to foreign policy (namely in terms of the type of relationships to be established with Russia and with the West), have been among the most salient, especially in the first years after the democratic transition. The cultural-ethnic dimension (concerning policies related with citizenship and native majorities rights and interests vis-à-vis the Slavic minorities rights and interests) is also particularly salient in Estonia and Latvia, but less so in Lithuania. Additionally, the socioeconomic (dividing policy orientations more pro-market or more pro-state, namely in terms of the Welfare State and privatization of state companies) has also been salient in these countries. Nevertheless, these issue divides can be more or less subsumed in the left-right divide. The left parties are less critical towards the communist heritage, more prone to favor close

17 relationships with Russia, more prone to grant citizenship rights to ethnic minorities, and more pro-state in terms of state intervention in the economy. On the contrary, right-wing parties are more prone to be more critical of the communist heritage, less prone to favor close relationships with Russia, less prone to grant citizenship rights to ethnic minorities, and are more pro-market in socioeconomic issues (Plakans, 2011, pp ) Insert Table 1 around here Insert Table 2 around here Insert Table 3 around here In Tables 2 and 3, we present the average location of both South European and Baltic citizens (Table 2) and MPs (Table 3) on the left-right scale. We can see that Baltic citizens, especially in Latvia and Lithuania, are usually to right of South European electorates, especially in the case of Spain. Moreover, if we take dispersion around the mean as an indicator of polarization, we can conclude that the south European electorates are clearly more polarized than the Baltic ones. In terms of MPs (Table 3), the differentiation between Southern Europe and the Baltic States is even more clearly with the former more to the left and more polarized than the latter (although Lithuanian MPs are to the left and more polarized then Estonian and Latvian parliamentarians, contrary to the case of voters). Finally, if we consider voters perceptions of parties locations in the left-right scale, we can clearly see that Spain is the more polarized country of the six (Table 1). Greece and Portugal, the

18 latter to a lesser extent, occupy a position in between Spain and the Baltic States when we consider the index of polarization (which takes into account all the relevant parties in the system), but are the less polarized then the Baltic States when we consider only the two major parties. Literature Review, theory and hypotheses The importance of the left-right divide in democratic mass politics is well known among the political science community. After the French Revolution, the idea of a left-right (LR) divide gained great importance in mass politics. At the individual level, the division between left and right functions as an instrument to reduce the complexity of the political universe; at the systemic level, it functions as a code of communication (Fuchs and Klingemann, 1990). Moreover, new democracies in Europe have provided extensive opportunities to study the formation, development, and institutionalization of political cleavages, namely cleavages of more ideological nature like the left-right divide. And as a matter of fact there are several case studies about the formation, development and content of ideological cleavages in new European democracies, namely about the left-right divide (Evans and Whitefield, 1998; Freire and Belchior, 2011; Jou, 2010a and 2010b; Jurkynas, 2004; Kivistik, 2012; Neundorf, 2009). Additionally, there are several studies that take groups of new European democracies (either from Central and Eastern Europe or from Southern Europe) and either compare them

19 among themselves (Jerschina and Górniak, 1997; Markowski, 1997; Barnes, 2002; Regt, Mortelmans and Smits, 2011; Whitefield, 2002) or compare them with long consolidated democracies (Freire, 2006a; Pop- Eleches and Tucker, 2010). What we don t find in the literature about the left-right divide is comparisons of different types new democracies in terms of their authoritarian legacies, namely we don t find contrasts of new democracies with a right-wing authoritarian legacy (for example like those in Southern Europe) with democracies with a left-wing authoritarian legacy (like the post-communists democracies in Central and Eastern Europe). However, we believe that such a comparison can be illuminating about the formation, development and content of the left-right divide in new democracies. Namely, it can help us to answer the fundamental research question in the present paper: in a new democracy, does the type of authoritarian legacy (right-wing vs. leftwing dictatorship, and also right-wing authoritarian vs. left-wing totalitarian regime) has any effect on the way citizens think about and use the left-right political divide? In terms of research design, considering the very different authoritarian heritages, but also the very different historical trajectories, we of course rely on the most dissimilar systems comparative design. As we said in the introduction, the formation of ideological (and partisan) identities is a process that continues to evolve along the courses of socialization (primary and secondary) of individuals (Campbell, 1980; Converse, 1964 and 1969; Niemi at. al, 1985; Barnes,

20 McDonough and Pina, 1985). Considering the very different regime heritages and the different trajectories of the democratic transition in the two sets of countries we will be comparing, Southern Europe (Greece, Portugal, Spain) vis-à-vis the Baltic States (Estonia, Latvia, Lithuania), our research question will be answered in three major dimensions of analysis. First, we will be analysing left-right (LR) recognition in the two sets of countries, and how it evolved over time, Second, we will consider that in a transition from left-wing and right-wing authoritarianism there are different winners and losers, namely in terms of ideological orientations, and thus for the two sets of countries we will analyse what is the type of relationship between left-right self-placement with both current regime support and evaluations of the previous regimes, and how those relationships evolved over time, Third, for the two sets of countries we will investigate what is the nature and strength of the relationships of the major value/issue conflicts (in each country) and leftright self-placement, and how they evolved over time, In the remaining part of the present section we will present the hypotheses concerning each dimension, and the rationale underlying each one of them. On the one hand, in the right-wing authoritarian regimes existent until in Greece, Portugal, and Spain, the political parties were a proscribed reality, and ideological differences were repressed. Therefore, the conditions for the formation of LR identities were rather reduced. On the other hand, Estonia, Latvia, and Lithuania lived under a single party

21 totalitarian regime at least since 1945 and Thus, the conditions for the formation of LR identities, usually related with party competition and ideological pluralism, were also severely reduced here. However, contrary to the new Southern Europe, in the Baltic States there was a leftwing totalitarian regime. Moreover, also contrary Southern Europe, whose right-wing dictatorial regimes were more closed to the authoritarian prototype, the left-wing dictatorial regimes in the Baltic States were more close to totalitarian prototype (Linz and Stepan, 1996). In totalitarian regimes mass political participation is usually more extensive, and the political systems are characterized by a higher level of ideological structuration, when compared with merely authoritarian regimes. Namely, the South European authoritarian regimes were characterized by a comparatively lower level of ideological structuration and, additionally, by lower levels of mass political mobilization, because these regimes invested on a certain demobilization of the citizenry, vis-àvis the totalitarian regimes in the Baltic States, where both ideological structuration and extensive mass political participation (even if top-down controlled) were the norm. As we said before, Greece should stand as a case apart with the highest levels of LR recognition: after WW II the authoritarian interlude ( ) lasted only for around 7 years, while for the other five countries it lasted for at least around years, and several continuities existed between the Greek democratic regimes that existed before and after the colonels dictatorship (Clogg, 1987). Thus, the hypotheses concerning our first dimension of analysis are

22 the following. First, we expect the highest levels of LR recognition to occur in Greece, especially in the first two decades of the new democratic regime Hypothesis 1. Second, for the remaining countries, we expect to find higher levels of LR recognition (percentages of people that are able to place themselves on the LR scale) in the Baltic States (BS) vis-à-vis the new Southern Europe (SE), especially in the first decade of the new regime Hypothesis 2. Second, due to political learning in a democracy and also due to generation replacement, we expect those differences between the Baltic States and SE to vanish with the passage of time Hypothesis 3. Concerning our second dimension of analysis, the following considerations should be taken into account before we formulate our hypotheses. The different regime heritages and the different trajectories of the democratic transition in the two sets of countries (SE vis-à-vis BS) imply that there are different types of winners and losers with the change of regime. I.e., in the two sets of countries, due to the replacement of old regime (the left-wing communist regime, in the Baltic States, and the right-wing authoritarian regime, in Southern Europe), a repositioning took place, which allowed some people to gain better position while others lost because their position grew worse suddenly (Kitschelt, 1995; Daatland, 1997, Whitefield 2002; Zarycki 2000, Jurkynas 2003, 2004, Lie 2006; Kivistik, 2012). Scholars interested in the study of political transitions from communist rule, dominated by state controlled mechanisms for resource allocation, to liberal democratic regimes,

23 dominated by market mechanisms to the same for resource allocation, have defined different types of winners (more prone to support the new regime) and losers (less prone to support the new regime). Kistchelt (1995: pp. 459) argues that the more educated (i.e., with more adequate skills to successfully adapt to these changes), the younger cohorts ( with more motivational and emotional capacities to stomach change and typically fewer family responsibilities that make individuals more risk averse ) and urban residents (which are exposed to social networks and flows of information that enhance the opportunities for market adaptation) are among the winners in the post-communist regime transitions. Of course, the less educated, the older cohorts, and the rural residents are among the «losers», i.e., those less able to adapt to the social and political conditions of the new regime. The new ethnic majorities in the Baltic States are also among the winners in these countries, contrary to the Slavic minorities (Kistchelt, 1995: pp ). Except for the case of ethnic groups, we could also say that for similar reasons the more educated, the young, and the urban residents are among the winners in the democratic transitions in the new southern Europe. But in this paper we are especially interested in the politicalideological winners and losers. In this respect, the change of regime from state controlled communism to liberal democratic capitalism clearly meant that left-wing people are among the ideological losers in the Baltic States, even after controlling for social status. On the contrary, we expect the ideological winners in the South European

24 democratic transitions are those individuals located on the left (even after controlling for social status) not only because of the profile of the previous regime (right-wing authoritarian), but also because the transition to democratic rule also meant the implementation of the Welfare State (practically non-existent before democracy in SE) and thus not only civil and political rights were newly granted for all citizens but also social rights were granted for everybody. Considering all that, our hypotheses here are the following. First, we expect that (after controlling for social status), negative evaluations of previous regimes, democratic regime support and trust in current political institutions are higher among right-wing than among left-wing people in the Baltic States; the opposite is expected for SE, i.e., that negative evaluations of previous regimes, democratic regime support and trust in current political institutions are higher among left-wing than among right-wing people Hypothesis 3. Second, for the same reasons as for dimension one, we expect those differences between the Baltic States and SE to vanish with the passage of time Hypothesis 4. Concerning the third dimension, i.e., the value correlates (in terms of socioeconomic values, libertarian-authoritarian values, and post materialism), we state the following. Bearing on the South European experience of great internal diversity (great anchoring in Spain and Greece, low anchoring in Portugal: Freire, 2008; Freire and Belchior, 2011), we believe that the substantive value content of the left-right divide will depend not so much on the type of authoritarian legacy, but

25 more on the type of democratic transition and, especially, on the type of politicization of social and political issues that the political parties do since then. Moreover, two recent comparative studies (Eijk et al, 2005; Freire, 2008), analysing six and thirteen West European countries, respectively, have shown that the clarity of policy alternatives offered by political parties to the electorates do have an impact on the nature and strength of left-right self-placement at the mass level. First, the higher the clarity of policy alternatives, the more structured by values left-right selfplacement is at the individual level (Freire, 2008). Second, the higher the clarity of policy alternatives the stronger is the impact of individual leftright self-placement on the vote (Eijk et al, 2005; in a similar vein, but with much more countries and testing also the impact of electoral institutions, see Wessels and Schmitt, 2008). Other studies have also shown the importance of partisan supply (Neundorf, 2009) and social and political elites (Steen, 2007) in the structuring of mass political attitudes, namely in terms of left-right ideological reasoning. Additionally, as we have shown in the previous sections both voters and political elites are much more polarized in Southern Europe (especially in Spain and Greece) than in the Baltic States. The picture in terms of parties locations in the left-right scale is more mixed, but points in a similar direction. But there is a major difference within both Southern Europe and the Baltic States. In the first group there is clear alternation in government between left and right in Spain and Greece, but not in Portugal (in the latter polity, often the centre-left has to rely on rightwing parties to govern). A similar divide

26 occurs between Lithuania (where there is a clear alternation in government between left and right) and both Estonia and Latvia (where the left and the centre-left are either weak or excluded from government or diluted in right-wing dominated coalitions). Considering jointly the data on polarization and the type of leftright government alternation, our hypotheses here are the following. We expect the level of anchoring of left-right self-placement on values to be highest in Spain and Greece (high polarization, and clear left-right government alternation), and to be lowest in Estonia and Latvia (low polarization, and not clear left-right government alternation) Hypothesis 5. Both Lithuania (low polarization, clear left-right government alternation) and Portugal (middle polarization, not clear left-right government alternation) are expected to be cases in between Hypothesis 5. Data and Methods To answer our fundamental research question (does the type of authoritarian legacy has any effect on the way citizens think about and use the left-right political divide?), and considering the very different authoritarian heritages, but also the very different historical trajectories (Baltic States vs. Southern Europe), we of course rely on the most dissimilar systems comparative design. Besides the contextual information already presented, we will focus mainly on citizens left-right self-placement (LR recognition) and

27 the correspondent correlates in terms of regime support and value conflicts. Thus, in terms of data we will rely mainly on comparative mass surveys. To increase comparability and due to the fact that most of the items to test our hypotheses are present in the European Value Study, , we will rely mainly on these surveys to compare the six countries. However, despite their advantages, these data sets have several problems. First, LR was only asked for the Baltic States from 1999 on. Second, Greece and Portugal only integrate the network since 1999 and 1990, respectively. Thus, for the six cases we to use also other data sets temporally more close to the democratic transition in each country. For that reason, we will use also the following datasets. First, we will use the Political Culture in Southern Europe, a Four Nation Study, 1985, for the Mediterranean countries. This dataset will allow us analyse the two first dimensions under scrutiny for these countries in 1985, i.e., around ten years after the democratic transition. Although far from perfect, this is the best available and we believe can furnish us adequate conditions to test our hypotheses. Additionally, for the Baltic States we will use also the special edition World Values Survey, , for the Baltic States and Central and Eastern Europe. Thus, this data set will allow us to analyse all the relevant dimensions under scrutiny around 5 or 7 years of the democratic transition. But to evaluate ideological divides in terms evaluations of the previous regimes, we will use also the study Values and Elections in Estonia, Lithuania and Latvia, Additionally, to calculate the aggregate levels of LR recognition and also LR

28 behavioural in terms of vote, across countries and across time, we will use also several other datasets (all available and with the relevant variables for the period : Central and Eastern Barometers, Political Culture in SE A Four Nation Study, European Social Survey, European Value Study, European Election Studies). As we already mentioned, in terms of research design, we will rely on the most dissimilar systems comparative design to compare LR attitudes and its behaviour consequences in the Baltic States (new democracies with a left-wing and totalitarian legacy) and Southern Europe (new democracies with a left-wing and authoritarian legacy). Moreover, we will use percentages, correlations and OLS regressions to test our hypotheses. But for the passing to regressions, when analyzing the two last dimensions (regime support and values) we first reduce the set of indicators of regime support and values through Principal Component Analysis (PCA) see Appendix 1 and 2. LR recognition In the present section we test the impact of the type of authoritarian legacy on LR recognition by citizens in each country. With the exception of Greece (already explained), in all the remaining five countries, due to repression of political competition, the conditions for the formation of LR identities were rather reduced. However, in the BS there was a left-wing totalitarian regime; on the contrary in SE the right-wing dictatorial regimes were more closed to the authoritarian prototype. In totalitarian

29 regimes mass political participation is usually more extensive, and the political systems are characterized by a higher level of ideological structuration, when compared with merely authoritarian regimes. Thus, our hypotheses are the following. First, we expect the highest levels of LR recognition to occur in Greece, especially in the first decade of the new democratic regime Hypothesis 1. Second, for the remaining countries, we expect to find higher levels of LR recognition in the BS vis-à-vis SE, especially in the first decade of the new regime Hypothesis 2. Second, due to political learning and generation replacement, we expect those differences between the BS and SE to vanish with the passage of time Hypothesis 3. Insert Figure 1 around here Insert Figure 2 around here Looking at Figures 1 and 2, we clearly see that (especially in the second decade of the new democratic regime: data only available for the second decade in the Greek case) Greece stands as case apart with the highest levels of LR recognition: above 80%: 85. On the contrary, in the first decade of the new regimes, the remaining five countries have levels of LR recognition that are between 65% and 73%. Using the test for the difference in proportions across independent samples, we concluded that those differences are indeed statistically significant. Thus, this confirms H1, i.e., the reduced length of socialization in a democratic regime clearly explains lower levels of LR recognition. As for Hypothesis 2, concerning the differences between SE (Spain

30 and Portugal) and the BS, the distributions are rather similar across the two sets of countries, except perhaps for Portugal in the 1970s (with the lowest levels of recognition of all five countries, slightly above 60%: 65, while the others have values around 68%, in Estonia, and 73%, in Spain). But although the differences are small, they are all statistically significant: in the first decade of democracy, all the Baltic States have higher levels of LR recognition than Portugal; on the other hand, contrary to expectations all the Baltic States have lower levels of LR recognition than Spain. Thus, we have mixed results concerning H2 and no firm conclusion can be extracted from here. To put it another way, the nature of the authoritarian regime (totalitarian and more ideologically structured/bs vs. authoritarian and less ideologically structured/se) does seem to have a clear impact upon LR recognition, at least for the set of five countries under scrutiny. As for H3, which predicts convergence with the passage of time, to begin with let us underline that there are two separate trends between the first and the last decade of the new democratic regime. First, for Greece (from 85% to 82%) and Lithuania (from 69% to 65%) where there is a slight decrease in LR recognition between the first and the last decade under scrutiny. Second, in the remaining four countries there is a growth in LR recognition from the first to the last decade of the new regime. But to test the hypothesis of convergence (H3) we need to see if the differences between the Baltic States and the new SE do vanish with the passage of time. Using the appropriate tests, we

31 conclude that except for the contrast Latvia (77%) vs. Portugal (78%), all the other differences are statistically significant (with higher levels of LR recognition for SE than for the BS). Thus, we do not confirm H3. Nevertheless, this might be due to the fact that, in the 2000s, SE is in its three and a half decade of democracy; in the BS in the 2000s the second decade of democracy was still going. «Winners» and «Losers» of regime change and the LR correlates of attitudes towards the democratic regime and institutions Our second dimension of analysis is about the relationships between the very different authoritarian regimes heritages, which might generate different types of winners and losers, and regime support by left-.wing and right-wing individuals. Let us recall our hypotheses. First, we expect that democratic regime support are higher among right-wing than among left-wing people in the Baltic States, even after controlling for social status (age, education, income, church attendance, and habitat) and party voting (in government vs. not in government); the opposite is expected for SE, i.e., democratic regime support is higher among left-wing than among right-wing people Hypothesis 4. Second, we expect those differences between the BS and SE to vanish with the passage of time, namely due to generational replacement and also political learning Hypothesis 5.

32 When analyzing regime support, we have several indicators (several items for trust in political institutions, several items in terms of evaluations of the previous regimes, several questions for democratic values, several indicators for the shortcomings of the democratic regimes, and numerous indicators for authoritarian alternatives for democracy), which will be our vectors of variables to be regressed on left-right self-placement (plus social factors and party voting). Due to the fact that we are dealing with secondary data, the same items are not always available, but the ones that are do allow us to measure the different dimensions in different periods of time. In any case, we first reduced the attitudinal data on regime support through Principal Components Analysis (PCA) (Tables A1.1 to A1.8 in Appendix 1), and then used the factors extracted as the dependent variables to test our Hypothesis 4 and 5. Of course, due to the availability of data, we have different solutions in the different sets of countries in the several years. In any case, the PCA solutions were always found for either the three BS taken together or the three SE countries taken together. Moreover, the different items were always coded so that higher values mean more trust in current political institutions, more support for democratic values, more rejection of authoritarian alternatives, or more critical stances vis-à-vis the previous authoritarian regime. So, we expect to find negative relationships with LR in SE, and positive relationships with left-right selfplacement in the BS. As for the control variables, in terms of social factors we have that higher values mean older people, more education, higher

33 levels of income, rural residence and more church attendance. Additionally, we also control for voters party preference in terms of governing parties (1) vs. non governing parties (0), because we know that the level of institutional trust, of satisfaction with democracy and also, to some extent, of regime support is somewhat higher for those that have their preferred partisans controlling the governmental institutions than for those who voted for opposition (Anderson and Guillory, 1997). Insert Table 4.1 and 4.2 around here Looking at Table 4.1, we can see that due to data availability in SE we have only one (Trust in Political Institutions), for 1990 (and only for Portugal and Spain), or two factors (Trust in Political Institutions and Evaluation of Political Regimes, especially authoritarian alternatives, namely the previous regime), for Here we can see that we have only partial support for H3. First, we find that rightwing individuals trust more political institutions of the new regime than left-wing ones. Of course, one could argue that this due to indicators that measure trust in 1985 and 1990 which relate mainly to hierarchical institutions (Police, Military, Judiciary; Parliament and Civil Service only appear in 1990: see Tables A1.1 and A1.3 in Appendix 1), which are usually more supported by right-wing authoritarian people. But we believe this is not the case because the same pattern is found in 1999 and 2008 when the trusted institutions are much more diverse and the relevant factor also includes satisfaction with democracy. Where we find support for H3 is in the evaluation of the authoritarian alternatives for the democratic regime,

34 and in the evaluations of the authoritarian past: right-wing individuals are less critical of the previous authoritarian regime and also less critical of authoritarian alternatives for the current regime, than left-wing individuals. And in both cases, the relationships between individual LR ideology and regime support remain significant even after controlling for several indicators of social status and party voting. Moreover, a similar pattern is found for both 1999 and 2008 when we have three (Trust in Political Institutions and Satisfaction with Democracy, Democratic Regime Principles, and Democracy Shortcomings) or four factors (the latter three plus evaluation of Authoritarian Regimes) see Table 4.2. In the case of current regime support (Trust in Political Institutions and Satisfaction with Democracy), we find that rightwing individuals are more supportive (but this is only significant in Spain and Greece). However, for the case of the Evaluation of Authoritarian Alternatives, Democracy Shortcomings and Democratic Principles, the opposite is true: rightwing individuals are more critical of democracy and more tolerant vis-à-vis authoritarian alternatives. Again, these relationships remain significant (although more often in Greece and Spain than in Portugal) even after controlling for social status and party voting. Insert Table 5.1 and 5.2 around here In the case of current regime support (Trust in Political Institutions and Satisfaction with democracy), we find that also in the BS right-wing individuals are more supportive. Thus, since in this sub-dimension there are no differences between BS and SE, we can say that H4 is partially not

35 confirmed (see Tables 5.1 and 5.2). But in the cases of evaluation of Democracy Shortcomings, Democratic Regime Principles, and Authoritarian Alternatives, whenever the relationships are significant they point in an opposite direction vis-à-vis SE. This is because right-wing individuals give less emphasis to Democracy Shortcomings, are more supportive of Democratic Principles and more critical of Authoritarian Regimes than left-wing individuals in BS. And although the relationships are not always significant, when they are they always point in the expected direction (even after controlling for social status and party voting). Thus, taken this into account, as well as the differences vis-á-vis SE, we can say that we partial confirm H4. Hypothesis 5 predicts convergence between SE and BS with the passage of time, due both to generational replacement and political learning in the new regimes. One way to test this hypothesis can be to evaluate the strength of the impact on regime support, as measure by the net impact (R 2 ) of LR (i.e., after controlling for social status and party voting) on regime support between the first and last surveys under scrutiny. In two sets of countries, although more clearly in SE, there is a decline of the net impact of LR ideology on regime support, thus the evidence do seem to confirm H5. The value correlates of the LR divide In this final section, we concentrate on our third dimension of analysis,

36 i.e., the value correlates (in terms of socioeconomic values, libertarianauthoritarian values, and post materialism) of the left-right divide in the six countries. As we said before, bearing on the SE experience of significant and determined internal diversity, we consider that the substantive value content of the LR divide will be more dependent on the type of democratic transition and, especially, on the type of politicization of social and political issues that the parties do since then, and not so much on the type of authoritarian legacy. Here we recalled that previous studies have shown that the clarity of policy alternatives have a positive impact on the nature and the strength of anchoring of the LR divide at the mass level, and that there is a significant impact of partisan supply and elites leadership in the structuring of mass LR reasoning. Thus, we stated Hypothesis 6 as follows: the level of anchoring of LR self-placement on values will be highest in Spain and Greece (high party LR polarization; clear left-right government alternation), and it will be lowest in Estonia and Latvia (low polarization; no clear left-right government alternation). Lithuania (low polarization; clear left-right government alternation) and Portugal (middle polarization; not clear leftright government alternation) are expected to be cases in between. To analyze the relationships of value orientations with the LR divide, using either the European Values Study or the World Values Survey, we selected the indicators that, first, let us cover the major political-value conflicts in present day modern societies and, second, that were present in all the surveys so that we can increase the possibilities of

37 comparability. Since we selected six indicators of new politics values (attitudes towards homosexuality, abortion, euthanasia, suicide, and divorce; post materialism index based on 4 items) and two indicators of socioeconomic values (attitudes towards governmental vs. individual responsibility; attitudes towards market competition), we reduced information using PCA (see Appendix 2: Tables A2.1 to A2.6). The reader should take into account that we coded all the variables so that higher numbers in each item means more authoritarian, more materialist, or more economic right-wing attitudes, thus we expect positive correlations between the Factors and LR self-placement. Looking at the Factor solutions, we have usually a first and more important Factor labelled Social conservatism that includes new politics issues, except for postmaterialism (not relevant in SE; loaded in the socioeconomic factor in the BS, except in 2008: separate factor in BS). And then we also have a second factor that includes only as relevant variables (loadings superior to 0.4) either only the socioeconomic items, in SE, or the socioeconomic items and post materialism, in BS (except for 2008: ditto). Thus, the major difference in factor solutions between SE and BS is that in the former set of countries post-materialism does not appear relevant in factor loadings, and in the latter set of countries post-materialism does appear relevant but it is related with old politics, even if in an unexpected way (post-materialism orientations associated with pro market attitudes). Only in 2008 does post materialism appear alone in a third factors, for BS; remaining not relevant for Factor solutions in SE.

38 Insert Table 6 around here Insert Table 7 around here To measure the impact of values on LR self-placement, we use two types of measures (see Tables 6 and 7): first, «the gross impact of values», i.e., the impact of the value Factors on LR without any controls; second, «the net impact of values» on LR, i.e., the added value in R2 by introducing the values after taking the impact of social factors in consideration. Let us now pass to hypothesis testing. Overall, we find the Spain and Greece always have the strongest anchoring of LR on values. Thus, this clearly confirms H6. Also confirming H6 is the position of both Estonia and Latvia (with the lowest levels of anchoring of LR selfplacement on value orientations) and of Lithuania (a case in between Spain-Greece and Estonia-Latvia), the latter except on 2008 (all BS with the lowest levels). The deviant case here is Portugal: the levels of anchoring of the LR divide on values (either gross or net impact) put the country close to Estonia and Latvia, not Lithuania. The Portuguese situation is not easy to explain, but perhaps it is due to a highest level of importance of the lack of a clear left-right alternation, which locates Portugal closer to Estonia and Latvia, vis-à-vis party polarization, which locates the former country closer to Lithuania. On the changes in the Lithuanian case, in 2008, this might be due to an increase in party system fragmentation and populism in the 2000s (Blondel et al, 2007), which might be blurring the differences between left and right. As for type of relationships between values and LR, we see that in all the six countries

39 both social conservatism and pro-market socioeconomic values are positively correlated with right-wing self-placement. The only deviations are, first, that in some countries/years some values don t have a significant impact, and, second, that since in the BS post materialism is associated with pro-market attitudes, we can say that it is also connected with right-wing attitudes. Overall, the data presented in the present section show that in terms of the value content of the LR divide, the type of authoritarian legacy is not a relevant factor, what is relevant is the type of democratic transition and the level of politicization of issues by the parties in competition. Concluding remarks Studies about the formation and development of LR identities on new democracies have focused either on homogenous groups of new democracies (for example, in Central and Eastern Europe), on case studies, or on comparisons of the former groups with long consolidated democracies. However, new democracies are heterogeneous, namely in terms of the type of authoritarian heritage, and if the type of authoritarian heritage does have a significant impact on the formation and development of LR identities on new democracies, then we had a lacuna in the literature. It was precisely this lacuna that we intended to fill with the present paper where we compared very different new democracies, with contrasted authoritarian legacies (left-wing

40 totalitarian in the case of the Baltic States; right-wing authoritarian in the case of the new Southern Europe). And we have three major objectives with the present paper. First, we analyzed the impact of the type on LR recognition. Considering that left-wing totalitarian regimes are characterized by higher levels of mass mobilization and higher levels of ideological structuration, vis-à-vis the more depoliticized authoritarian prototypes, we expected to find higher levels of LR recognition in the BS than in Spain and Portugal. This was indeed the case for the latter country, but not for Spain. Thus, in terms of LR recognition other relevant factors (vis-à-vis the authoritarian legacy) might be more important. In the first section, however, we also found that the length of socialization in an authoritarian regime does imply lower levels of LR recognition: that is way Greece, with a much shorter authoritarian interlude ( ) in post WWII, has higher levels of LR recognition than each one of the other five countries, which had long authoritarian legacies (around 45 years or more). These differences remain today, even if they are now less pronounced. As for our second dimension, considering that different types of authoritarian legacies can generate different types of «winners» and «losers», we predicted that left-wing people would tend to show more democratic regime support (support for democratic principles, rejection of authoritarian alternatives, less criticism vis-à-vis democratic shortcomings, trust in institutions and satisfaction with democracy) in

41 Southern Europe, and right-wing people would tend to show more democratic regime support in the Baltic States. Firstly, we found this not to be the case in terms of current regime support (trust in institutions and satisfaction with democracy): in both BS and SE right-wing people are more prone to trust current political institutions, and to be happier with democratic performance. The reader should recall that these results are rather robust because they were found after controlling for potential social «winners» and «losers» (age, education, income, habitat, etc.) and party voting. On the one hand, these results might be due to the fact that in the set of political institutions considered there are several that are hierarchical institutions (police, armed forces, public administration, etc.) usually more trusted by right-wing (authoritarian) people. On the other hand, this means that dissatisfaction with current democracies come more form the left, but it is not dependent on ideological «winners» and «losers». What we found according to our expectation is that the evaluation of the authoritarian alternatives to current democracy, of previous authoritarian regimes, and in terms of reservations to the performance of democracies, authoritarian legacy does have a significant and rather robust (i.e., after controlling for social «winners» and «losers», and party voting ) impact. In the latter case, criticisms come more from the right in the Baltic States, and more form the left in Southern Europe. Besides showing the importance of the present study, this result has an implication: in the case of erosion of democratic support, availability for alternatives has different possible avenues in BS

42 and SE. Finally, we analyzed the level of anchoring of the LR divide in value orientations. Here we contended that what was important was not the authoritarian legacy, but the type of authoritarian transition and, above all, the level of politicization of issues by the parties. Evidence confirmed our expectations with the highest level of value anchoring in Spain and Greece, the more polarized party systems and a clear LR government alternation, and the lowest level of value anchoring in Estonia and Latvia, the less polarized party systems and a with a less clear LR government alternation. Lithuania is a case in between. Portugal is closer to Estonia and Latvia, although according to the polarization characteristics it was expected to be closer to Lithuania. In any case, considering that the connection between LR and substantive policy preferences (and the underlying values), then the present study shows that even in new democracies increasing the quality of political representation is in the hands of parties: they just need to politicize the relevant social and political issues that can give content to the LR divide. References AA.VV. (2012), Baltic Voices, CSPP: Centre for the Study of Public Policy, University of Strathclyde, Anderson, C. J., and Guillory, C. A. (1997), Political institutions and satisfaction with democracya cross-national analysis of Consensus and Majoritarian systems, American Political Science Review, 91 (1), pp. 66-

43 81. Barnes, S. H. (2002), Left and right in old and new democracies, Central European Political Science Review, 3 (7), pp Barnes, S. H., P. McDonough, and A. L. Pina (1985), The development of partisanship in new democracies: the case of Spain, American Journal of Political Science, 29, pp Blondel, Jean, Müller-Rommel, Ferdinand, and Malová, Darina (2007), Governing New European Democracies, Hampshire, Macmillan. Bruneau, Thomas C., et al (2001), Democracy, southern European style, in P. Nikiforos Diamandouros e Richard Gunther (orgs.) (2001), Parties, Politics, and Democracy in New Southern Europe, Baltimore, The John Hopkin University Press pp Caramani, Daniele (2000), Elections in Western Europe since Results by Constituencies, London, Macmillan Reference, pp (Greece), (Portugal), and (Spain). Clogg, Richard (1987), Parties and Elections in Greece. The Search for Legitimacy, Londres, C. Hurst & Co. Converse, P. (1964), The nature of belief systems in mass publics, in David E. Apter (ed.), Ideology and Discontent, New York, Free Press, Converse, P. (1969), Of time and partisan stability, Comparative Political Studies, 2, pp Dalton, R. J. (2006), Social Modernization and the End of Ideology Debate: Patterns of Ideological Polarization, Japanese Journal of

44 Political Science, 7 (1), pp Evans, G. and S. Whitefield (1998),. «The evolution of left and right in post-soviet Russia», Europe-Asia Studies, Vol. 50(6): Freire, André (2006a), Left-Right Ideological Identities in New Democracies: Greece, Portugal and Spain in the Western European Context, Pôle Sud Revue de Science Politique de l Europe Méridionale, nº 25, II 2006, pp Freire, André (2006b), Bringing Social Identities Back In: The Social Anchors of Left-Right Orientation in Western Europe, International Political Science Review, 27 (4), Freire, André (2008), Party Polarization and Citizens Left-Right Orientations, Party Politics, 14 (2), pp Freire, André, and Ana Belchior (2011), What left and right means to Portuguese citizens, Comparative European Politics, Vol. 9, Nº 2, pp Freire, André, Lobo, Marina C., and Pedro Magalhães (2009), The clarity of policy alternatives, left-right and the European Parliament vote in 2004, The Journal of European Integration, Volume 31, nº 5, pp , special issue on the 2004 European Elections (coordinated by Hermann Schmitt). Gunther, Richard, e José R. Montero (2001), «The Anchors of Partisanship: A Comparative Analysis of Voting Behaviour in Four Southern European Countries, in Nikiforos Diamandouros e Richard Gunther (orgs.) (2001), Parties, Politics, and Democracy in New Southern

45 Europe, Baltimore, The John Hopkin University Press pp Inglehart, R. (1990), Culture Shift in Advanced Industrial Society, Princeton, Princeton University Press. Inglehart, R. and H. D. Klingemann (1976), «Party identification, ideological preference and the left-right dimension among Western mass publics», in Budge, Ian, et al (ed.), Party identification and Beyond, London and New York: Wiley. Jerschina, Jan, and Górniak, Jaroslaw (1997), Leftism, Achievement, Orientation and Basic Dimensions of the Socio-Economic and Political Attitudes in Baltic Countries versus Other Central and East European Countries, in Hood, Neil, Hilis, Robert, and Vahlne, Jan-Erick (1997), Transition in the Baltic States. Micro-Level Studies, London, Macmillan Press. Jou, Willy (2010a), Political Cleavages in Serbia: changes and continuities in structuring left-right orientations, South East European and Black Sea Studies, 10, 2, pp Jou, Willy (2010b), Continuities and changes in left-right orientations in new democracies: the cases of Croatia and Slovenia, Communist and Post-communist studies, 43, pp Jurkynas, M. (2004), «Emerging cleavages in new democracies: The case of Lithuania», Journal of Baltic Studies, 35(3): Kitschelt, H.(1995), «Formation of Party Cleavages in Post- Communist Democracies»,. Party Politics, 4: Kivistik, Kats (2012), «The concept of Left and Right in Estonia»,

46 Working Paper Nº XXX, CIES working paper series, web address. Linz, Juan J., e Stepan, Alfred (1996), Problems of Democratic Transition and Consolidation. Southern Europe, South America, and Post- Communist Europe, The John Hopkins University Press, Baltimore. Markowski, R. (1997), «Political Parties and Ideological Spaces in east Central Europe», Communist and Post-Communist Studies 30/3: Mishler, William, and Rose, Richard (2001), What are the origins of political trust?, Comparative Political Studies, 34, 1, pp Müller-Rommel, Ferdinand, and Sootla, G. (2001), Estonia, in Blondel, Jean, and Müller-Rommel, Ferdinand (eds.), Cabinets in Eastern Europe, Hampshire, Palgrave Macmillan, pp. pp Müller-Rommel, Ferdinand, and Norgaard, O. (2001), Latvia, in Blondel, Jean, and Müller-Rommel, Ferdinand (eds.), Cabinets in Eastern Europe, Hampshire, Palgrave Macmillan, pp. pp Müller-Rommel, Ferdinand, and Hansen, Hersted O. (2001), Lithuania, in Blondel, Jean, and Müller-Rommel, Ferdinand (eds.), Cabinets in Eastern Europe, Hampshire, Palgrave Macmillan, pp. pp Neundorf, Anja (2009), Growing up on different sides of the Wall A Quasi-experimental test: Applying the Left-Right dimension to the German Mass Public, German Politics, 18, 2, pp Niemi, R. G. et. al. (1985), Testing the converse partisanship model with new electorates, Comparative Political Studies, 18, pp

47 Plakans, Andrejs (2011), A Concise History of the Baltic States, Cambridge, Cambridge University Press. Pop-Eleches, Grigore, and Tucker, Joshua A. (2010), After the Party: Legacies and Left-Right Distinctions in Post-Communist Countries, Estudio/Working Paper 2010/250, June Regt, Sabrina de, Mortelmans, Dimitri, and Smits, Tim (2011), Leftwing authoritarianism is not a myth, but a worrisome reality. Evidence from 13 Eastern European countries, Communist and Post-Communist Studies, 44, pp Rose, Richard, and Munro, Neil (2009), Parties and Elections in New European Democracies, Essex, ECPR Press. Rose, Richard, and Mishler, William (1998), Negative and Positive Partisanhip in Post-Communist Societies, Electoral Studies, 17, 2, pp Sitter, Nick (2002), Cleavages, Party Strategy and Party System Change in Europe, East and West, Perspectives on European Politics and Society, 3, 3, pp Steen, A. (1996), «Confidence in Institutions in Post-Communist Societies: The case of the Baltic States», Scandinavian Political Studies, 19(3): Steen, Anton (2007), Do Elite Beliefs Matter? Elites and Economic Reforms in the Baltic States and Russia, Comparative Social Research, Volume 23, pp Van der Eijk, C., Schmitt, Hermann and Binder, Tanja (2005), Left-

48 right Orientations and Party Choice, in Thomassen, Jacques (ed.), The European Voter. A Comparative Study of Modern Democracies, Oxford: Oxford University Press: Wessels, Bernhard, and Hermann Schmitt (2008), Meaningful choices, political supply, and institutional effectiveness, Electoral Studies, 27, pp Whitefield, S. (2002), «Political Cleavages and Post-Communist Politics», Annual Review of Political Science, 2002, 5:

49 Sources: European Value Study, 1981, 1999 and 2008; European Social Survey, 2002, 2004, 2006, 2008; European Election Study, 1989, 1994, 1999, 2004, Note: Average Percentage by decade of respondents in each country that are able to place themselves in the LR scale (equals 100% minus Don t Knows and No Answers). Sources: Central and Eastern Barometers, and 2003; World Values Survey, ; European Value Study, 1999 and 2008; European Social Survey, 2008; European Election Study, Note: Average Percentage by decade of respondents in each country that are able to place themselves in the LR scale (equals 100% minus Don t Knows and No Answers).