What explains how citizens living in young democracies feel about their authoritarian past? While the impact of autocratic legacies on support for democracy and left–right placement has been thoroughly studied, we know less about the determinants of attitudes toward the past in post-authoritarian democracies. This study relies on survey data collected in Southern and Central European countries ten years after their transitions to democracy in order to test context-dependent variance in the relevance of ideology and party identification on citizen attitudes toward the past. The results show that classical factors such as regime type and mode of transition are not the main determinants of the politicization of attitudes toward the past and that the existence of a strong authoritarian successor party is associated with stronger politicization of the past.

Authoritarian regimes leave an imprint on societies. In post-authoritarian democracies, up to 20% of citizens tend to express positive attitudes toward the previous regime long after its collapse or overthrow (Pop-Eleches & Tucker, 2017; Santana-Pereira, Raimundo & Costa Pinto, 2016; Gherghina & Klymenko, 2012; Ekman & Linde, 2005); however, little consensus exists as to what these positive evaluations of democracies’ predecessor regimes represent. Ekman and Linde (2005) have interpreted it as an expression of nostalgia and a desire to return to the status-quo-ante; Pop-Eleches and Tucker (2014) and Barnes (1972) have seen it as an expression of the regimes’ ability to indoctrinate their citizens; Szczerbiak (1999) has looked at it as an expression of citizens’ ideological preferences. Most scholarship is focused on the impact these legacies may have on democracy, particularly on political culture. Specifically, they tend to focus on the causal relationship between authoritarian/communist legacies and support for democracy, satisfaction with democracy, and the left–right self-placement (Neundorf, 2010; Pop-Eleches & Tucker, 2011; Dinas, 2017; Freire & Kivistik, 2016; Neundorf, Gerschewski & Olar, 2020).

Few studies have looked into the determinants of positive and negative attitudes toward the past. Among them, Ekman and Linde (2005) find a significant impact of age in post-communist Europe; other studies have found that these attitudes are rooted in ideology. Huneeus and Martín (2000) suggest there is a strong relationship between voting for a particular party and the evaluation of the regime in post-authoritarian Chile, while Santana-Pereira et al. (2016) show that the effect may not be the same in different time frames, as found in the case of post-authoritarian Portugal, suggesting that politicization of the past may be context-dependent.

In this article, we examine the relationship between attitudes toward the past and ideology and party identification as potentially moderated by contextual factors. Concretely, we examine to what extent different regime types, transition modes, and the historical baggage of the main political parties influence the levels of politicization of the past in consolidated democracies. By politicization we mean the existence of strong correlations between attitudes toward a specific object (in this case, the authoritarian past) and ideological and partisan preferences. Our research relies on three streams of literature. The first is the literature on regime legacies, which has argued that the “cultural, political, economic, and social conditions inherited from the past” could have an enduring effect on democracy (Beverly & Lijphart, 1997). A large segment of this literature focuses on communist legacies, helping perpetuate the idea of a “post-communist exceptionalism,” that is, that post-communist legacies are significantly different from other authoritarian or historical legacies (Pop-Eleches & Tucker, 2017). In testing the impact of regime type, we include democracies preceded by communist regimes as well as democracies preceded by right-wing civilian dictatorships. We exclude from our analysis democracies preceded by military dictatorships because we assume that they have a lower ability and impetus to indoctrinate their citizens (Dinas & Northmore-Ball, 2020). We are aware of the country differences in culture, historical traditions, and religiosity, among other things, but we believe these differences should not substantively impact the relationship we explore in this article. Second, we rely on the democratization literature (Gunther & Montero, 2004; Schneider, 2009). According to this literature, new institutions and practices are partly conditioned by the mode of transition to democracy (Mahoney & Snyder, 1999): that is, the levels of rupture or continuity are expected to influence the way democracy works. We test this assumption at the attitudinal level by looking at citizen attitudes toward the past in regimes that transitioned from authoritarianism to democracy in different ways. Finally, we rely on the literature on successor parties, looking at the impact of parties connected to the former ruling elite and parties connected to the former regime opposition. We follow earlier studies that suggested that identification with such parties may be linked to specific political attitudes (Gunther, Montero & Puhle, 2007). We analyze six third-wave democracies that experienced distinctive types of autocratic regimes and of transition modes, and display different patterns in terms of the existence of successor and opposition parties, including two Southern European post-authoritarian democracies and four Central European post-communist democracies.

In the next section, we identify the mechanism linking distinctive regime types with different anchors of attitudes toward the past; we examine the classical contributions in democratization studies and propose a return to the origins of transitology, that is, a dichotomous classification of transition modes; then, we look into the party systems in post-communist and post-authoritarian democracies; each subsection presents a specific hypothesis or set of hypotheses to be tested. In the third section, we present the methodology used in this study. Finally, we communicate our main results and discuss their implications to the understanding of how the context impacts the ideological and partisan anchors of attitudes toward the past.

Regime Legacies

The literature on regime legacies emerged after 1989 and developed particularly within the field of post-communist studies. One of its precursors was Jowitt (1992), for whom the experience of communism was bound to have a deep effect on societies. In the following years, several types of potential legacies associated with the communist experience were identified: cultural, social, political, national, institutional, and economic (Beverly & Lijphart, 1997). As democracies thrived in certain parts of the region, other more nuanced interpretations emerged. Some talked about an important distinction between the countries that were part of the Soviet Union and those that were not. Others talked about the role of incentives to Western integration as well as several structural conditions (Pop-Eleches, 2007). In both cases, a distinction was made between the Visegrad countries1 and all others.

Later, Hite and Cesarini (2004) proposed a definition of authoritarian legacies that could travel across regime types: “those rules, procedures, norms, patterns, practices, dispositions, relationships, and memories originating in well-defined authoritarian experiences of the past” (p. 4). Yet very few studies have empirically tested the impact of legacies in post-communist and post-authoritarian democracies. Neundorf et al. (2020) go beyond the classical regime-type distinction and offer an analysis of the impact of autocratic regimes on regime support based on autocracies’ level of political and socioeconomic inclusion, that is, on whether they “incorporate various social, economic, and ethnic groups into their power structure” or have a “narrow social basis that excludes from power most social, religious, and ethnic groups” (p. 1894). This is an interesting way of comparing autocratic legacies at the attitudinal level, but it does not allow us to test the effect of indoctrination and ideology, which implies the distinction between communist and right-wing dictatorship. This could have a powerful effect on whether citizens express positive attitudes toward the prior regime, as suggested by Dinas and Northmore-Ball (2020).

Inspired by this stream of research, our first hypothesis is that the type of authoritarian regime will have an impact on the levels of politicization of the past, which will be stronger in post-communist than in post-authoritarian democracies (Hypothesis 1). Our expectation is rooted in the conviction that communist regimes may have been stronger in terms of indoctrination and politicization than right-wing authoritarian regimes (Neundorf, 2010): in the former, citizens were expected to be active and vocal party members; in the latter, they were encouraged to leave “politics” in the hands of an elite. These different degrees of indoctrination may be linked to different ways the authoritarian past is assessed in current ideological and partisan preferences.

Transitions to Democracy: The Role of Negotiations

Hite and Morlino (2004) suggest the mode of transition is an important variable to take into account when analyzing the effect of regime legacies. In the literature, there is an abundance of typologies of modes of transition into democracy. Linz (1978) was among the first authors to distinguish between two modes: reforma and rupture. Later, almost every conceivable typology of transition modes involved a third type, including Linz’s revised version (Linz & Stepan, 1996; Karl, 1990; Share & Mainwaring, 1986; Huntington, 1991; Munck & Leff; 1999). Yet, the proliferation and complexification of typologies resulted in some degree of overlap and redundancy. Huntington’s transformation type is similar to Linz’s reforma, Linz and Stepan’s transition by pact, Munck and Leff's extrication, or Share and Mainwaring’s transaction: a process that occurs when reformers emerge from within the old regime and maintain power simply by transforming their political beliefs. This was the case in Hungary (1989) and Spain (1975); Linz and Stepan add Poland (1989) to the set of examples. Transplacement, Huntington’s second type, is similar to Share and Mainwaring’s extrication, Linz and Stepan’s transitions by “collapse,” and Munck and Leff’s transaction. These are transitions that occur when the old regime starts to move toward democracy, yet loses both initiative and legitimacy and is quickly outpaced by opposition reformers who proceed to take power, possibly resulting from “rigidity, ossification and loss of elite responsiveness” (Linz & Stepan, 1996, p. 322). This was the case in Poland, Czechoslovakia, and East Germany in 1990. Finally, Huntington’s replacement type (equivalent to Share and Mainwaring’s collapse or breakdown, and Linz and Stepan’s and Munck and Leff’s transitions by rupture) occurs when there is an overthrow and complete replacement of an unrepentant and unreformed government by the opposition. This was the case of East Germany, Portugal, Greece, and Romania.

Even though each typology highlights distinctive aspects of democratic transitions, one may question the utility of applying all these distinctions. For instance, Huntington (1991) acknowledged that “the line between transformation and transplacement is fuzzy, and some cases might be legitimately classified in either category” (p. 124).

We believe the issue of negotiations deserves closer attention. In particular, the role played by negotiations during democratization processes will impact the mechanism that links the transition to certain attitudinal predispositions: legitimation. The way in which democratic elites search for their justifiability constitutes the mechanism through which the features of the past and/or of the transition transform into enduring mnemonic legacies. Here we draw first on the notion of “inverse legitimation” as coined by Valenzuela, to describe how, in certain contexts, elites legitimize the new regime by highlighting and even exaggerating the negative aspects of the previous one (Valenzuela, 1992). This is very distinctive from what Di Palma calls backward legitimation. This refers to a strategy through which the democratic elite does not condemn the previous regime but rather searches for distant sources of legitimation. This notion was also used by Huntington, who argued that “backward legitimacy had two appeals and two effects: it legitimated the new order because it was a product of the old, and it retrospectively legitimated the old order because it had produced the new” (Huntington, 1991, p. 138).

Whether or not the path to democracy needs to be negotiated with the outgoing elite will determine how the authoritarian past will emerge in political culture. This is what democratization scholars have also argued (Munck & Leff, 1999), and it also overlaps with transitional justice patterns as identified by democratization scholars: transitions by negotiation do not allow for an immediate process of coming to terms with the past and a complete delegitimation of the previous regime because the outgoing elite will always demand immunity before relinquishing power, whereas in transitions without negotiation the window is rapidly opened and punishment often takes place together with a process of stigmatizing the old regime (Welsh, 1996; Williams, Fowler & Szczerbiak, 2005; Kim, 2012).

In line with the above, we propose a dichotomous view of transition modes including “transitions by negotiation” and “transitions without negotiation.” We expect higher levels of politicization in the cases of transition by negotiation, because in such a context the former elite is not forced to reject the past, nor is the new democratic elite. In this sense, the electorate that continues to support the previous elite will feel less coerced to reject that past and to evaluate it negatively, whereas the elite that does not support it (admittedly quite different in ideological and partisan terms from the former) will blatantly reject it. On the contrary, in transitions by rupture followed by processes of inverse legitimation, the zeitgeist may be so unfavorable to positive assessments of the undemocratic past that we can assume that people with different ideological positions will differ less in the way they address the previous regime also due to the emergence of a “new discourse driven by political elites striving to bolster their legitimacy by emphasising their break with the past” (Dinas & Northmore-Ball, 2020, p. 1962). In concrete terms, we expect ideological self-placement and party identification to explain a lower degree of variation in attitudes toward the past in countries that experienced a process of transition without negotiation (Portugal, Slovakia, and Czech Republic) than in those whose transition was negotiated (Poland, Spain, and Hungary) (Hypothesis 2).

Authoritarian Successor Parties and Opposition Forces

The extant literature on post-authoritarian and post-communist democracies shows that new party systems always include old and new parties, but the levels of continuity appear to vary. We argue that continuity is likely to affect attitudes toward the past in two ways, First, the former ruling elite is likely to survive and to play a role in the new democratic setting. As Grzymala-Busse (2020) explains, “the influence of authoritarian ruling parties does not end with the collapse of their regimes. Even when their official monopoly on rule ends, their subsequent exit from power and their transformation (or failure to transform) into moderate democratic parties continue to structure political party competition” (p. 27). This pattern has been identified across the world, regardless of the type of regime or mode of transition. Not long after 1989, Ishyama (1997) identified a relationship between the success of ex-communist parties, previous regime type and the strength of leftist competition, but his focus was exclusively on post-communist countries; therefore, his distinction was between types of communism and not between regimes of a different ideological nature.

In an effort to overcome the tendency to analyze legacies of communism and legacies of right-wing dictatorships separately, the term authoritarian successor parties (ASPs) has been adopted to define “parties founded by high-level incumbents of authoritarian regimes that continue to operate after a transition to democracy” (Loxton, 2015, p. 2). Loxton and Mainwaring (2018) find that their presence in post-authoritarian societies is the pattern, not the exception. In order to account for the variation among different cases, Loxton distinguishes between two subtypes: the former authoritarian ruling parties—the party that ruled under the authoritarian regime—and the reactive authoritarian successor parties—parties created by high-level figures of the former authoritarian regimes.

In Central European countries, these parties are more often referred to as “post-communist parties” and the most well-known are former ruling parties. Most of these parties did not simply survive but actually thrived in the new democratic context. In the most successful cases they have won elections (Poland); in other cases they have integrated coalition governments’ and yet in other cases they were able to achieve the status of second- and third-largest opposition parties (the Czech Republic). Authors disagree as to what explains different levels of success—whether it is the resources they carry with them or how they adapt to the new democratic setting (Morse, 2020; Grzymala-Busse, 2020). Autocratic ruling party strength appears to be connected to their ability to lead a democratization process (Kavasoglu, 2021), but it is not necessarily linked to these parties’ ability to keep their advantage after democratic consolidation. Although ASPs that emerge out of negotiated transitions are likely to maintain some leverage compared to ASPs that do not, this relationship is yet to be shown.

Authors also disagree as to the consequences of successor parties breaking with the past or not (Grzymała-Busse, 2002; Bozóki & Ishiyama, 2002). While earlier studies suggested that in order to obtain popular support, they had to distance themselves from the past and even condemn it (Grzymała-Busse, 2002), more recent contributions have shown how successor parties that mobilized these narratives during electoral campaigns have been electorally successful (Loxton, 2015). Where these parties exit power and reinvent themselves, they promote the rise of free, institutionalized, and robust party competition.

This type of successor party is less typical of post-(right-wing)-authoritarian democracies, despite the persistence of parties with some institutional and leadership ties with the former elite. In democracies preceded by military and civilian regimes, the most frequent type is the reactive successor party, “created in reaction to a transition to democracy, either by high-level authoritarian incumbents in anticipation of an imminent transition, or by former high-level incumbents shortly after a transition” (Loxton, 2015, p. 159). It is also among these democracies that we find a higher number of cases of absence of successor parties. It is important to note that post-authoritarian democracies are also those in which non-negotiated forms of transition exist in higher numbers. Argentina, Greece, and Portugal, three countries where the transition to democracy occurred without negotiation, constitute examples of democracies in which there are no successor parties (Loxton, 2015; Raimundo, 2015).

Second, although to our knowledge there is no equivalent concept to define the other side of authoritarian regimes, we believe parties that are deeply connected to the regime opposition should not be excluded from the analysis. We name them opposition successor parties (OSPs)—parties that existed during an authoritarian regime without any ties to that regime, mostly created and/or headed by well-known individuals who opposed, and were persecuted by, the regime. In post-communist Europe, low levels of trust in parties made it difficult for new party structures to blossom in the first years (Lewis, 2001); hence low levels of party system institutionalization in these countries are more often connected to splits and mergers than to the emergence of parties that are completely new to the political system. Post-authoritarian democracies, on the other hand, could count on historical parties created in the beginning of the 20th century, regardless of whether or not they attempt to claim that heritage and despite some difficulties in establishing themselves in the new democratic setting (Biezen, 2003). Hence, OSPs in post-authoritarian democracies were more clearly defined in ideological terms than OSPs in post-communist countries, where they were most diverse and tended to split and form new parties and coalitions.

We therefore expect politicization of the past to be stronger in contexts in which there is a strong ASP with low or average levels of detachment from the past (Hypothesis 3). This pattern is most clearly observable in Spain and the Czech Republic than in the other countries, as we will see in detail in the next section. Our last hypothesis is that the politicization of the past will be impacted by the integrity and ideological congruence of OSPs, being that in contexts where opposition movements or parties disintegrated and/or adopted a new ideological baggage, politicization is bound to be lower (Hypothesis 4). As in the case of ASPs, the existence of a strong representative of the fight against authoritarianism in the democratic party system is understood to be a necessary element for party attachments to matter in terms of attitudes toward the past.

The goal of this article is to assess the relative impact of partisan and ideological stances on the way authoritarian regimes of the recent past are assessed by the citizenry in different macro-level contexts. We expect the relative contribution of these variables to be impacted by the nature of the previous regimes (Hypothesis 1), by the mode of transition (Hypothesis 2), and by the specificities of the party systems in place during the first decade after the transition (Hypotheses 3 and 4). These hypotheses are competing, in the sense that we do not expect all to be confirmed by the empirical analysis; the first stresses the potential role of nondemocratic legacies, the second merely suggests differences due to the macro impact of the strategies of legitimation after different transitions, and the third and fourth acknowledge that in similar transitions we may still find different patterns due to different behaviors from successor and opposition parties. Figure 1 displays the six countries under analysis (four post-communist countries from Central Europe and two post-authoritarian countries from the Iberian Peninsula) and how they are placed according to these variables.

Figure 1.

Distribution of the six cases under analysis in the four theoretical dimensions.

Figure 1.

Distribution of the six cases under analysis in the four theoretical dimensions.

In order to warrant full comparability of the data and models computed, we chose time points that represent more or less the same period in these third-wave democracies. For Portugal and Spain, the data used are from 1985, that is, about one decade after the end of the right-wing authoritarian regimes of Salazar/Caetano and Franco, respectively. These data were retrieved from the Four Nation Study, implemented by a team of social scientists in Portugal, Spain, Italy, and Greece (Sani & Santamaría, 2005 [1985]). In turn, the data on the Visegrad countries (Poland, Czech Republic, Slovakia, Hungary) were collected by the European Values Survey of 1999, that is, also about ten years after the processes of transition to democracy began. These six cases were selected because they allow for variation in context in terms of the variables we are interested in, while keeping constant other relevant dimensions, such as geography (all are European democracies) and length of the nondemocratic regime (about four decades in the six countries: Portugal 1933–74, Spain 1939–75, Czechoslovakia 1948–89, Hungary 1949–89, Poland 1944–90). East Germany is not included due to lack of survey data for this specific segment of the unified Germany in the late 1990s.

Our individual-level dependent variable is the attitudes held by citizens regarding the political regime in place in their home countries before the democratic transition—authoritarian right-wing regimes in the Iberian Peninsula and communist regimes in Central Europe. In the Four Nation Study, the Spanish and Portuguese respondents were asked: “On the basis of what you know or remember about Franquismo/Salazarismo, do you think that (a) All considered it was good; (b) It was in part good and in part bad; (c) It was only bad?” In turn, the Central European citizens interviewed for the European Values survey were asked “Where on this scale would you put the political system as it was under communist regime?” and offered a 10-point scale in which 1 meant “bad” and 10 “very good.” To foster comparability and taking into account the goals of this article, this later scale was reduced into a three-point scale similar to that used in the Iberian countries (Table 1).2

Table 1.

Variables Used in Ordinal Logistic Regression Models

Variable StatusVariable NameScale
Dependent Attitudes toward the previous authoritarian regime in the country 3-point scale (1 = all considered it was bad; 2 = it had good and bad things; 3 = all considered it was good) 
Independent Ideology (left–right self-placement) 10-point scale (1 = left; 10 = right) 
Party Identification
Portugal
PCP (Portuguese Communist Party)
 PS (Socialist Party)
 PRD (Democratic Renewal Party)
 PSD (Social Democratic Party)
 CDS (Democratic Social Centre)
Spain
PCE (Spanish Communist Party)
 PSOE (Spanish Socialist Workers’ Party)
 CDS (Democratic and Social Centre)
 CiU (Convergence and Union)
AP (People’s Alliance)
Poland
SLD (Democratic Left Alliance)
PSL (Polish People’s Party)
UW (Freedom Union)
AWS (Solidarity Electoral Action)
 ROP (Movement for the Reconstruction of Poland)
Hungary
MSZP (Hungarian Socialist Party)
 SZDSZ (Alliance of Free Democrats)
Fidesz (Hungarian Civic Alliance)
 FKgP (Independent Smallholders, Agrarian Workers and Civic Party)
 MIEP (Hungarian Justice and Life Party)
Czech Republic
KSCM (Communist Party of Bohemia and Moravia)
 CSSD (Czech Social Democratic Party)
 KDU-CSL (Christian and Democratic Union – Czechoslovak People’s Party)
US (Freedom Union)
ODS (Civic Democratic Party)
Slovakia
SDL (Party of the Democratic Left)
 SOP (Party of Civic Understanding)
 SDK (Slovak Democratic Coalition)
 SMK (Party of the Hungarian Community)
HZDS (Movement for a Democratic Slovakia)
 SNS (Slovak National Party) 
Dummy (1 = the respondent would vote for this party) 
Control Age Continuous 
Gender Dummy (1 = Female) 
Education 3-point scale (1 = primary complete or less; 2 = secondary complete or incomplete; 3 = university degree, complete or incomplete) 
Religiosity 5-point scale (1 = never attends religious services; 5 = attends religious services more than once a week) 
Union membership Dummy (1 = respondent belongs to a union) 
Habitat (community size) 3-point scale (1 = fewer than 2,000 inhabitants; 2 = from 2,001 to 10,000 inhabitants; 3 = more than 10,000 inhabitants) 
Satisfaction with democracy 3-point scale (1 = not/not very satisfied; 2 = rather satisfied; 3 = very satisfied) 
Media use (newspapers, radio, TV) 5-point scale (1 = never; 5 = every day) 
Variable StatusVariable NameScale
Dependent Attitudes toward the previous authoritarian regime in the country 3-point scale (1 = all considered it was bad; 2 = it had good and bad things; 3 = all considered it was good) 
Independent Ideology (left–right self-placement) 10-point scale (1 = left; 10 = right) 
Party Identification
Portugal
PCP (Portuguese Communist Party)
 PS (Socialist Party)
 PRD (Democratic Renewal Party)
 PSD (Social Democratic Party)
 CDS (Democratic Social Centre)
Spain
PCE (Spanish Communist Party)
 PSOE (Spanish Socialist Workers’ Party)
 CDS (Democratic and Social Centre)
 CiU (Convergence and Union)
AP (People’s Alliance)
Poland
SLD (Democratic Left Alliance)
PSL (Polish People’s Party)
UW (Freedom Union)
AWS (Solidarity Electoral Action)
 ROP (Movement for the Reconstruction of Poland)
Hungary
MSZP (Hungarian Socialist Party)
 SZDSZ (Alliance of Free Democrats)
Fidesz (Hungarian Civic Alliance)
 FKgP (Independent Smallholders, Agrarian Workers and Civic Party)
 MIEP (Hungarian Justice and Life Party)
Czech Republic
KSCM (Communist Party of Bohemia and Moravia)
 CSSD (Czech Social Democratic Party)
 KDU-CSL (Christian and Democratic Union – Czechoslovak People’s Party)
US (Freedom Union)
ODS (Civic Democratic Party)
Slovakia
SDL (Party of the Democratic Left)
 SOP (Party of Civic Understanding)
 SDK (Slovak Democratic Coalition)
 SMK (Party of the Hungarian Community)
HZDS (Movement for a Democratic Slovakia)
 SNS (Slovak National Party) 
Dummy (1 = the respondent would vote for this party) 
Control Age Continuous 
Gender Dummy (1 = Female) 
Education 3-point scale (1 = primary complete or less; 2 = secondary complete or incomplete; 3 = university degree, complete or incomplete) 
Religiosity 5-point scale (1 = never attends religious services; 5 = attends religious services more than once a week) 
Union membership Dummy (1 = respondent belongs to a union) 
Habitat (community size) 3-point scale (1 = fewer than 2,000 inhabitants; 2 = from 2,001 to 10,000 inhabitants; 3 = more than 10,000 inhabitants) 
Satisfaction with democracy 3-point scale (1 = not/not very satisfied; 2 = rather satisfied; 3 = very satisfied) 
Media use (newspapers, radio, TV) 5-point scale (1 = never; 5 = every day) 

Note: ASPs are presented in bold and OSPs in italic.

In terms of individual-level independent variables, we rely on the survey respondents’ ideological self-placement and proximity (measured through the proxy “vote intention”) to the main political parties in each country. For each country, we selected the political parties that were able to elect members of parliament in the parliamentary elections that were closest to the survey interviews. The full list of parties included for each country is presented in Table 1—ASPs are presented in bold and OSPs in italic—but some additional information on them is necessary. By 1999, post-communist parties in the Visegrad countries were former ruling parties that underwent different levels of rupture with the past: minimal (KSCM in Czech Republic),3 medium (SDL in Slovakia4 and SLD and PSL in Poland),5 and very high (MSZP in Hungary).6 In Southern Europe, ASPs are practically nonexistent and the few existing ones are of the reactive successor party type. This is the case of the Spanish party People’s Alliance (AP), the predecessor of the current major right-wing party.7 In Portugal, the residual role of the single party during the dictatorship and the absence of negotiations during the transition contributed to its disappearance after 1974. There was no ASP of any type in the 1980s as there is no such party today (Lisi, 2015).8 As for OSPs, in post-communist Europe, the strongest actors in the transition were movements that, with the noteworthy exception of Hungary, ended up giving birth to myriad new parties of various ideologies, from liberals to extreme right.9 In Spain and Portugal, the two communist parties, PCE and PCP, were key players of the regime opposition and resisted transforming into catch-all types after the breakdown of both dictatorships.10

Lastly, we control for the effect of other variables that previous research has found to be significant. These variables include age (older respondents are more likely to hold positive views of the past), gender, education, religiosity (more-religious respondents are more likely to have a positive view of the right-wing authoritarian regimes/a less positive view of the communist regimes), union membership, community size, satisfaction with the way democracy is working, and media use for political information (average frequency of TV watching, newspaper reading, and radio listening), a proxy for interest in politics and political engagement. The full list of control variables, as well as their scales, is presented in Table 1.

The goal of the ordinal logistic regression models is to measure the relative strength of partisanship and ideology in different contexts. Because of the small number of cases we are able to include in our analysis, due to both data availability and theoretical motives, we could not pool the data, directly insert our contextual variables of interest in the model, and run a multilevel regression analysis.11 We therefore opted to contribute to this discussion by analyzing the differences in the explaining power of the exact same model, operationalized in a comparable way, in six different contexts, and analyze whether differences between those contexts are congruent with the cleavages discussed above: nature of the nondemocratic regime, mode of transition, and existence of specific types of parties in the democratic party systems.

Analysis of Single-Country Regression Models

Before contrasting our hypotheses with the empirical evidence produced, by identifying major trends in the levels of politicization of the past in the six countries under study, we describe the impact of ideology and party identification in each nation. For each country, two sets of ordinal logistic regressions were computed. The first models include the control variables, whereas in the second, ideology and party identification dummies were added. By doing so, we are able to assess the amount of variation in the dependent variable explained by ideological and partisan factors in each country, as well as, via information criteria, the relative quality of the models containing our variables of interest vis-à-vis those merely containing control variables. Also, we computed average discrete changes for each factor in the second models, which allow us to assess its relative contribution to explain the variation in the dependent variable. The results are displayed in Table 2 (Portugal and Spain), Table 3 (Czech Republic and Slovakia), and Table 4 (Poland and Hungary).12 In general terms, ideology was found to be significant in the six countries, as well as proximity with different political parties.

Table 2.

Determinants of Assessment of Previous Regime in Portugal and Spain: Ordinal Logistic Regressions

PortugalSpain
Model 1Model 2Average discrete changes Model 2Model 3Model 4Average discrete changes Model 4
Threshold = 1 –.46 (.52) –.37 (.80) – –.93* (.39) 1.05 (.58) – 
Threshold = 2 2.13*** (.53) 2.64** (.81) – 1.77*** (.40) 4.29*** (.59) – 
Age .01* (.001) .01 (.01) .11 .01** (.001) .01* (.004) .10 
Gender (female) –.30* (.13) –.27 (.19) .04 –.10 (.09) –.06 (.13) .01 
Education .18 (.11) –.10 (.15) .03 –.23** (.07) –.24** (.09) .07 
Religiosity .26*** (.05) .17* (.08) .10 .57*** (.04) .32*** (.06) .15 
Union membership (yes) –.70*** (.16) –.16 (.23) .02 –.78*** (.18) –.38 (.21) .05 
Habitat –.01 (.01) .03 (.11) .01 .20* (.08) .19 (.11) .05 
Satisfaction with democracy –.33* (.14) –.33 (.20) .10 –.62*** (.08) –.34** (.12) .09 
Media use –.01 (.08) –.11 (.12) .06 –.17** (.06) –.20* (.08) .10 
Ideology  .17** (.06) .22  .40*** (.05) .39 
PCP  –1.64*** (.39) .26    
PS  –.31 (.31) .05    
PRD  –.33 (.41) .05    
PSD  .29 (.33) .04    
CDS  .89* (.38) .12    
PCE     –.77* (.33) .12 
PSOE     .04 (.18) .01 
CDS     .94** (.29) .10 
CiU     –.48 (.31) .07 
AP     1.45*** (.24) .15 
1024 593 – 1972 1258  
Nagelkerke R2 7.9% 29.3% – 22.5% 44.6%  
AIC 1912.1 1022.8  3460.4 1977.6  
BIC 2000.9 1154.3  3560.9 2131.7  
% of correct pred. observations 53.8% 62.9%  58.4% 63.8%  
PortugalSpain
Model 1Model 2Average discrete changes Model 2Model 3Model 4Average discrete changes Model 4
Threshold = 1 –.46 (.52) –.37 (.80) – –.93* (.39) 1.05 (.58) – 
Threshold = 2 2.13*** (.53) 2.64** (.81) – 1.77*** (.40) 4.29*** (.59) – 
Age .01* (.001) .01 (.01) .11 .01** (.001) .01* (.004) .10 
Gender (female) –.30* (.13) –.27 (.19) .04 –.10 (.09) –.06 (.13) .01 
Education .18 (.11) –.10 (.15) .03 –.23** (.07) –.24** (.09) .07 
Religiosity .26*** (.05) .17* (.08) .10 .57*** (.04) .32*** (.06) .15 
Union membership (yes) –.70*** (.16) –.16 (.23) .02 –.78*** (.18) –.38 (.21) .05 
Habitat –.01 (.01) .03 (.11) .01 .20* (.08) .19 (.11) .05 
Satisfaction with democracy –.33* (.14) –.33 (.20) .10 –.62*** (.08) –.34** (.12) .09 
Media use –.01 (.08) –.11 (.12) .06 –.17** (.06) –.20* (.08) .10 
Ideology  .17** (.06) .22  .40*** (.05) .39 
PCP  –1.64*** (.39) .26    
PS  –.31 (.31) .05    
PRD  –.33 (.41) .05    
PSD  .29 (.33) .04    
CDS  .89* (.38) .12    
PCE     –.77* (.33) .12 
PSOE     .04 (.18) .01 
CDS     .94** (.29) .10 
CiU     –.48 (.31) .07 
AP     1.45*** (.24) .15 
1024 593 – 1972 1258  
Nagelkerke R2 7.9% 29.3% – 22.5% 44.6%  
AIC 1912.1 1022.8  3460.4 1977.6  
BIC 2000.9 1154.3  3560.9 2131.7  
% of correct pred. observations 53.8% 62.9%  58.4% 63.8%  

Notes: For each model, in the first and second columns, values are unstandardized coefficients, with standard errors in parentheses; in the third column, values are the changes in the probability of holding different attitudes caused by shifting from the minimum to the maximum value of the independent variable, when all the other variables are held at their means. Significance: ***p < .001; **p < .010; *p < .050. Multicollinearity diagnostics give negative results: no VIFs higher than 2.

Table 3.

Determinants of Assessment of Previous Regime in the Czech Republic and Slovakia: Ordinal Logistic Regressions

Czech RepublicSlovakia
Model 5Model 6Average discrete changes Model 6Model 7Model 8Average discrete changes Model 8
Threshold = 1 –2.20*** (.33) –3.12*** (.47) – –3.59*** (.41) –3.81*** (.50) – 
Threshold = 2 –.03 (.33) –.97* (.46) – –1.13** (.39) –1.11* (.49) – 
Age .02* (.003) .01 (.01) .07 .01** (.004) .01 (.01) .05 
Gender (female) .18* (.10) .45*** (.13) .07 .05 (.13) .01 (.13) .001 
Education –.64*** (.08) –.72*** (.10) .21 –.29** (.11) –.21 (.12) .06 
Religiosity –.32*** (.06) –.07 (.08) .04 –.09* (.04) –.01 (.05) .001 
Union membership (yes) .25 (.16) –.19 (.20) .03 .11 (.15) .11 (.17) .01 
Habitat –.25*** (.06) –.17* (.08) .05 –.20** (.07) –.08 (.08) .02 
Satisfaction with democracy –.81*** (.10) –.26* (.13) .08 –.85*** (.13) –.47** (.15) .14 
Media use –.08 (.05) .03 (.07) .02 –.21*** (.05) –.15* (.07) .07 
Ideology  –.44*** (.04) .50  –.27*** (.04) .31 
KSCM  1.40*** (.23) .22    
CSSD  .14 (.17) .02    
KDU-CSL  –.31 (.28) .05    
US  –.97** (.29) .13    
ODS  –.97*** (.21) .14    
SDL     .31 (.22) .04 
SOP     .25 (.26) .03 
SDK     –.74*** (.21) .11 
SMK     .04 (.25) .01 
HZDS     .61** (.19) .08 
SNS     .19 (.27) .02 
1789 1429 – 1203 948 – 
Nagelkerke R2 16.3% 48% – 9.6% 19.3% – 
AIC 2841.6 1871.0  2340.2 1790.9  
BIC 2940.7 2029.0  2431.9 1946.2  
% of correct pred. observations 60.6% 70.4%  53.5% 58.5%  
Czech RepublicSlovakia
Model 5Model 6Average discrete changes Model 6Model 7Model 8Average discrete changes Model 8
Threshold = 1 –2.20*** (.33) –3.12*** (.47) – –3.59*** (.41) –3.81*** (.50) – 
Threshold = 2 –.03 (.33) –.97* (.46) – –1.13** (.39) –1.11* (.49) – 
Age .02* (.003) .01 (.01) .07 .01** (.004) .01 (.01) .05 
Gender (female) .18* (.10) .45*** (.13) .07 .05 (.13) .01 (.13) .001 
Education –.64*** (.08) –.72*** (.10) .21 –.29** (.11) –.21 (.12) .06 
Religiosity –.32*** (.06) –.07 (.08) .04 –.09* (.04) –.01 (.05) .001 
Union membership (yes) .25 (.16) –.19 (.20) .03 .11 (.15) .11 (.17) .01 
Habitat –.25*** (.06) –.17* (.08) .05 –.20** (.07) –.08 (.08) .02 
Satisfaction with democracy –.81*** (.10) –.26* (.13) .08 –.85*** (.13) –.47** (.15) .14 
Media use –.08 (.05) .03 (.07) .02 –.21*** (.05) –.15* (.07) .07 
Ideology  –.44*** (.04) .50  –.27*** (.04) .31 
KSCM  1.40*** (.23) .22    
CSSD  .14 (.17) .02    
KDU-CSL  –.31 (.28) .05    
US  –.97** (.29) .13    
ODS  –.97*** (.21) .14    
SDL     .31 (.22) .04 
SOP     .25 (.26) .03 
SDK     –.74*** (.21) .11 
SMK     .04 (.25) .01 
HZDS     .61** (.19) .08 
SNS     .19 (.27) .02 
1789 1429 – 1203 948 – 
Nagelkerke R2 16.3% 48% – 9.6% 19.3% – 
AIC 2841.6 1871.0  2340.2 1790.9  
BIC 2940.7 2029.0  2431.9 1946.2  
% of correct pred. observations 60.6% 70.4%  53.5% 58.5%  

Notes: For each model, in the first and second columns, values are unstandardized coefficients, with standard errors in parentheses; in the third column, values are the changes in the probability of holding different attitudes caused by shifting from the minimum to the maximum value of the independent variable, when all the other variables are held at their means. Significance: ***p < .001; **p < .010; *p < .050. Multicollinearity diagnostics give negative results: no VIFs higher than 2.

Table 4.

Determinants of Assessment of Previous Regime in Poland and Hungary: Ordinal Logistic Regressions

PolandHungary
Model 9Model 10Average discrete changes Model 10Model 11Model 12Average discrete changes Model 12
Threshold = 1 –1.78*** (.42) –2.17*** (.58) – –2.05*** (.42) –3.11*** (.58) – 
Threshold = 2 .29 (.41) .12 (.58) – .71 (.42) –.23 (.56) – 
Age .02*** (.004) .02*** (.01) .23 .01** (.004) .01* (.01) .09 
Gender (female) .42** (.13) .40* (.16) .06 .31* (.14) .31 (.17) .03 
Education –.40*** (.10) –.42*** (.12) .14 –.33** (.10) –.37** (.12) .08 
Religiosity –.10 (.06) –.05 (.08) .04 –.25*** (.07) –.22* (.09) .10 
Union membership (yes) –.15 (.21) –.14 (.25) .02 –.14 (.26) –.27 (.29) .03 
Habitat –.18* (.07) –.08 (.09) .03 .12 (.09) .16 (.11) .03 
Satisfaction with democracy –.62*** (.12) –.55*** (.15) .18 –.32* (.13) –.16 (.18) .03 
Media use –.04 (.06) –.13 (.08) .08 –.07 (.06) –.19* (.07) .09 
Ideology  –.18*** (.04) .25  –.13** (.05) .12 
SLD  .88*** (.22) .14    
PSL  1.38*** (.28) .19    
UW  .20 (.29) .03    
AWS  –.52* (.27) .09    
ROP  .13 (.44) .02    
MSZP     .55** (.20) .06 
SZDSZ     –.72 (.44) .09 
Fidesz     –.86*** (.24) .10 
FKgP     .11 (.39) .01 
MIEP     –1.66** (.62) .24 
930 677 – 893 621 – 
Nagelkerke R2 13% 29.1% – 6.5% 19.7% – 
AIC 1800.2 1230.3  1670.5 1156.9  
BIC 1887.24 1365.8  1756.8 1289.9  
% of correct pred. observations 49.6% 57%  56.9% 59.3%  
PolandHungary
Model 9Model 10Average discrete changes Model 10Model 11Model 12Average discrete changes Model 12
Threshold = 1 –1.78*** (.42) –2.17*** (.58) – –2.05*** (.42) –3.11*** (.58) – 
Threshold = 2 .29 (.41) .12 (.58) – .71 (.42) –.23 (.56) – 
Age .02*** (.004) .02*** (.01) .23 .01** (.004) .01* (.01) .09 
Gender (female) .42** (.13) .40* (.16) .06 .31* (.14) .31 (.17) .03 
Education –.40*** (.10) –.42*** (.12) .14 –.33** (.10) –.37** (.12) .08 
Religiosity –.10 (.06) –.05 (.08) .04 –.25*** (.07) –.22* (.09) .10 
Union membership (yes) –.15 (.21) –.14 (.25) .02 –.14 (.26) –.27 (.29) .03 
Habitat –.18* (.07) –.08 (.09) .03 .12 (.09) .16 (.11) .03 
Satisfaction with democracy –.62*** (.12) –.55*** (.15) .18 –.32* (.13) –.16 (.18) .03 
Media use –.04 (.06) –.13 (.08) .08 –.07 (.06) –.19* (.07) .09 
Ideology  –.18*** (.04) .25  –.13** (.05) .12 
SLD  .88*** (.22) .14    
PSL  1.38*** (.28) .19    
UW  .20 (.29) .03    
AWS  –.52* (.27) .09    
ROP  .13 (.44) .02    
MSZP     .55** (.20) .06 
SZDSZ     –.72 (.44) .09 
Fidesz     –.86*** (.24) .10 
FKgP     .11 (.39) .01 
MIEP     –1.66** (.62) .24 
930 677 – 893 621 – 
Nagelkerke R2 13% 29.1% – 6.5% 19.7% – 
AIC 1800.2 1230.3  1670.5 1156.9  
BIC 1887.24 1365.8  1756.8 1289.9  
% of correct pred. observations 49.6% 57%  56.9% 59.3%  

Notes: For each model, in the first and second columns, values are unstandardized coefficients, with standard errors in parentheses; in the third column, values are the changes in the probability of holding different attitudes caused by shifting from the minimum to the maximum value of the independent variable, when all the other variables are held at their means. Significance: ***p < .001; **p < .010; *p < .050. Multicollinearity diagnostics give negative results: no VIFs higher than 2.

In Portugal, the full model is able to explain a modest amount of variation in the dependent variable (less than 30%), but the insertion of ideological and partisan variables tripled this value vis-à-vis the model comprising only the controls (information criteria metrics are also remarkably better; Table 2). Looking at the predicted probabilities of holding different opinions on the past across the ideological spectrum (Figure 2), computed based on the regression analysis results, we see that the average left-wing citizen is almost as likely to assess the previous regime neutrally or negatively (and only 4% likely to say that the previous regime was good), while the average right-wing respondent is much more likely (64%) to express a neutral assessment than a negative or positive one. In average terms, attitudes toward the past are more positive among right-wing than among left-wing citizens.

Figure 2.

Effects of ideology on attitudes toward the past: Predicted probabilities of holding different opinions on the previous regime across the ideological spectrum. The y axis represents a probability scale ranging between 0 and 1. The x axis compares the ideology scale in which 1 means left and 10 right. Each line represents the probability of expressing a positive, negative, or neutral assessment of the authoritarian regime for each level of ideological self-placement, holding all the other variables at their mean values. Source: Own elaboration of the Four Nation Study and European Values Survey data.

Figure 2.

Effects of ideology on attitudes toward the past: Predicted probabilities of holding different opinions on the previous regime across the ideological spectrum. The y axis represents a probability scale ranging between 0 and 1. The x axis compares the ideology scale in which 1 means left and 10 right. Each line represents the probability of expressing a positive, negative, or neutral assessment of the authoritarian regime for each level of ideological self-placement, holding all the other variables at their mean values. Source: Own elaboration of the Four Nation Study and European Values Survey data.

In addition, feeling close to the communist party PCP increases the odds of assessing the previous regime negatively from 29% to 68% and reduces the likelihood of neutral evaluations to an equal degree (Table 2; Figure 3). To the right, identification with CDS reduces in half (from 37% to 19%) the odds of assessing the previous authoritarian regime negatively (Figure 3).

Figure 3.

Effects of partisan attachments on attitudes toward the past: Predicted probabilities of holding different opinions according to party proximity. The y axis represents a probability scale ranging between 0 and 1. The x axis compares party attachment groups (attached vs. not attached) only in the case of party dummies found to be statistically significant in the ordinal logistic regressions. Each line represents the probability of expressing a positive, negative, or neutral assessment of the authoritarian regime for each dummy value, holding all the other variables at their mean values. Source: Own elaboration of the Four Nation Study and European Values Survey data.

Figure 3.

Effects of partisan attachments on attitudes toward the past: Predicted probabilities of holding different opinions according to party proximity. The y axis represents a probability scale ranging between 0 and 1. The x axis compares party attachment groups (attached vs. not attached) only in the case of party dummies found to be statistically significant in the ordinal logistic regressions. Each line represents the probability of expressing a positive, negative, or neutral assessment of the authoritarian regime for each dummy value, holding all the other variables at their mean values. Source: Own elaboration of the Four Nation Study and European Values Survey data.

Similar results are observed in Spain, with a few differences worth noting. First, our model is able to explain much more variation in our dependent variable in Spain (almost 45%) vis-à-vis Portugal, which is to a great extent due to a stronger impact of the control variables in the former (Table 2). Second, the average discrete changes show us that the impact of ideology is also greater in Spain than in Portugal (Table 2). In Spain, the average left-wing respondent is, as in Portugal, very likely to hold a negative view of the past and not likely to assess it positively. However, unlike their Iberian neighboring counterparts, the average Spanish right-wing respondents are as likely to express neutral as positive views of the past, while the odds of holding a negative view of the Francoist regime are negligible (Figure 2).Third, proximity to the Communist Party is relevant, but less so than in Portugal, since even those close to the PCE are more likely to express neutral than negative opinions of the previous authoritarian regime (Figure 3). Fourth and last, proximity to right-wing parties plays more of a role in Spain than in Portugal, also because in the latter there is no significant impact of proximity to PSD. In Spain, proximity to AP (but also CDS)13 increases the odds of positive attitudes in more than 10 percentage points (Figure 3).

Traveling east, we observe that our full model is able to explain a higher amount of variation in attitudes toward the past in the Czech Republic than in Slovakia, Hungary, and Poland, but this time it is only partially due to differences in the impact of the control variables (Tables 3 and 4). In the four cases, information criteria show that the full model is of higher quality than the models merely containing the control variables.

In the four post-communist democracies, ideology is significant, although more in the Czech Republic than in Slovakia and even less so in Poland and Hungary (Tables 3 and 4). Taken as a whole, and unsurprisingly, attitudes toward the past are more negative among right-wing than among left-wing citizens in these countries. Among the Czechs, the odds of assessing the previous communist regime negatively range from negligible (15%) if there are average left-wing respondents to overwhelming (90%) if they are placed in the extreme right of the ideological continuum. In turn, in Slovakia, Poland, and especially Hungary, moving from the left to the right end of the ideological spectrum, we see more modest increases in the odds of being negative about the previous regime (Figure 2).

Partisan attachments are also important. In the Czech Republic, being close to the post-communist KSCM decreases the odds of being negative about the previous regime (from 64% to 31%), while proximity with US and ODS14 has the opposite effect, albeit more modest (Figure 3). People who are close to these latter parties are very likely to be negative about the communist regime (Figure 3). Interestingly, identification with CSSD, technically a post-communist party (arguably the heirs of a communist faction ostracized by the dominant party during the communist regime) but not a successor party, is not statistically significant (Table 3). In turn, in Slovakia, partisans of SDK15 are slightly more likely to be negative about the previous communist regime, while people willing to vote for HZDS16 are slightly less likely to be so (Figure 3). The fact that proximity with the successor party SDL is not statistically significant is remarkable and a distinct feature of Slovakia within the Visegrad countries. Lastly, in both Poland and Hungary, proximity to post-communist parties (and, in the case of Poland, its allied PSL) is linked to higher odds of expressing positive views about the communist regime, while proximity with AWS (composed of “post-Solidarity” elites; Higley, Kulberg, and Pakulski [2002]) in Poland and the right-wing parties Fidesz and MIEP in Hungary is accompanied by a higher likelihood of expressing negative assessments.

Hypotheses Testing

As seen above, in the six countries under analysis, both ideology and partisan attachments are relevant predictors of attitudes toward the past. In the Iberian democracies, right-wing citizens are less negative about the past than left-wing citizens; the opposite is true in the case of the Visegrad countries. In regard to proximity to political parties, we saw that closeness to parties that had direct or indirect links to the previous regimes (the two CDS’s and AP in the Iberian Peninsula, KSCM in the Czech Republic, MSZP in Hungary, SLD in Poland) was linked to a trend toward less negative assessments of the past. In turn, closeness to parties at the ideological extreme of the previous regime—which, in some cases, were strong opponents and/or controlled the path toward transition (PCP in Portugal, PCE in Spain, US and ODS in the Czech Republic, AWS in Poland, Fidesz and MIEP in Hungary)—fostered a more negative outlook of the past. As also seen, the case of Slovakia is somewhat more intricate.

In order to test our hypotheses, variation in the degree of politicization of attitudes toward the past among these six countries was necessary, and indeed it is found, since the impact of partisanship and ideology differs among the six countries. In fact, based on the analysis of figures such as the increase of pseudo-R2 after the inclusion of the partisanship and ideology variables in the regression models and the average discrete changes associated with these variables, we observe that politicization is considerably stronger in the Czech Republic and, to a lesser extent, Spain, than in Hungary and Slovakia (Table 5). In the first two countries, we observe a considerable explanatory power of ideology and partisanship, with ideology being somewhat more powerful in the Czech Republic than in Spain. In the latter, these variables are much less important, although for different reasons: in Hungary both are, on average, of very modest impact, whereas in Slovakia ideology is much more important than partisanship. Portugal and Poland are intermediate cases due to the relatively important role of ideology (vis-à-vis Hungary) or partisanship (vis-à-vis Slovakia).

Table 5.

Summary of Main Findings for Each Country

CountryIncrease in pseudo-R2 due to inclusion of ideology and party ID in modelAverage discrete change ideologyAverage discrete change party attachments (average)Degree of politicization
Portugal 20.3 .22 .10 Medium 
Spain 22.1 .39 .09 Medium-strong 
Czech Republic 31.7 .50 .11 Strong 
Slovakia 9.7 .31 .05 Weak 
Poland 16.1 .25 .09 Medium 
Hungary 13.2 .12 .10 Weak 
CountryIncrease in pseudo-R2 due to inclusion of ideology and party ID in modelAverage discrete change ideologyAverage discrete change party attachments (average)Degree of politicization
Portugal 20.3 .22 .10 Medium 
Spain 22.1 .39 .09 Medium-strong 
Czech Republic 31.7 .50 .11 Strong 
Slovakia 9.7 .31 .05 Weak 
Poland 16.1 .25 .09 Medium 
Hungary 13.2 .12 .10 Weak 

While the empirical patterns do not point to the existence of blatant cleavages, but instead suggest that we should place these countries on a continuum in terms of the politicization of attitudes toward the past, it is clear that there is no explicit post-communist vs. post-right-wing authoritarian cleavage, since these two groups display a rather high level of within-cluster diversity in terms of levels of politicization. This does not allow us to confirm Hypothesis 1. Those patterns also do not support a clear-cut division between democracies introduced after rupture vs. negotiation, since the contexts in which the levels of politicization are higher (Spain and Czech Republic) experience different modes of transition to democracy. Our Hypothesis 2 is therefore not confirmed. However, Hypothesis 3 seems to receive support from the data analysis, since higher levels of politicization are indeed observed in the countries in which successor parties that did not reject the past existed. Hypothesis 4 is also not supported by the data, since among the three countries with cohesive and strong former opposition parties (Portugal, Spain, and Hungary), politicization varies from weak to medium-strong (Table 5).

This article has examined contextual factors in explaining the politicization of attitudes toward the past. We have concluded that right-wing regimes do not lead to different levels of politicization of the past when compared to communist regimes. Also, we found no clear-cut impact of transition modes and legitimation strategies on levels of politicization of the past. The only factor that we found to make a difference is the presence or absence of an ASP (or equivalent) that has not rejected the past. What we find is that in democracies where successor parties exist and have not openly rejected the past, levels of politicization are higher. This suggests that if political parties do not openly reject the past, their electorate will feel more willing to express similar attitudes, regardless of whether the mode of transition to democracy has opened the window to address and condemn that past or not. The two cases from which we draw our conclusions are clear examples of this. In the Czech Republic the regime collapsed and the transition to democracy was marked by a strong rupture at the discourse level. Yet the post-communist party refused to reform itself and did not reject its past. In Spain, on the other hand, the regime came to an end through a negotiated process in which the old elite could openly present itself as the heir of the regime. There was no delegitimation or coming to terms with the past. In both cases, there is a stronger relationship between expressing positive attitudes toward the past and voting for the parties that have not rejected the past.

These conclusions suggest that parties do perform a role as mediators between citizens and the political regime. Ultimately, these conclusions point to the issue of congruence between parties and their electorate. It seems that regardless of how the current regime approaches the past, the way representatives do it is more important and has a more important effect. More research is needed on this aspect to consolidate these conclusions.

The authors would like to thank José Ramon Montero for insightful comments and the two anonymous reviewers for their helpful suggestions on earlier versions of this article. This publication received support from FCT-MCTES through the project PTDC/CPOCPO/30730/2017.

1.

The Visegrad Group includes the Czech Republic, Hungary, Poland, and Slovakia.

2.

Values 1 to 3 were recoded as “all considered it was bad”; values 4 to 7 were recoded as “it had good and bad things”; values 8 to 10 were recoded as “all considered it was good.” For the sake of robustness, different recodifications of the dependent variable were experimented; no substantial changes in the patterns displayed in the statistical analysis were observed.

3.

In the Czech Republic, despite some “admissions of ‘Stalinist errors,’” the leaders of the Communist Party “defended the entire communist power as a time of social progress and economic advancement for the country” (Grzymała-Busse, 2002, p. 89). Although the party became the third most important political force in the Czech Republic, KSCM never won an election at the national level, nor has it been invited to integrate a coalition government.

4.

The Slovak successor party SDL had a very distinctive profile from the Czech one. “They moved rapidly to centralize the party and to break with the past” (Grzymała-Busse, 2002, p. 89) and, unlike KSCM, SDL has integrated coalition governments on both sides of the spectrum.

5.

Similarly, the main successor party in Poland, Democratic Left Alliance (SLD), changed its name, its symbol (the flag was replaced by the flag of the Social Democratic parties), and its discourse. Also, the choice of younger apparatchiks was aimed at ensuring strong and credible leadership breaking symbolically with the past. Together with another minor post-communist party, the Polish Peasant’s Party (PSL), they won the 1993 elections, just four years after the transition. See Skiba (2005).

6.

The Hungarian Socialist Party (MSZP) was the only one in our analysis to apologize for the past at the time of the data collection and it immediately “claimed not to be an ideological successor, and stressed its institutional discontinuity by not automatically carrying over any old members.” The party has been, since the mid-1990s, one of the most important political parties in Hungary, competing directly with Fidesz. See Bozóki and Ishiyama (2002) and Tavits (2013).

7.

This was a coalition formed in 1976 by seven parties plus other small parties from the right, all of them formed by politicians and ex-ministers of the Franco regime. Although their discourse changed throughout the 1980s, in 1977 they still affirmed their rejection “of any form of ruptura” and demanded “respect for the work of a whole people over nearly half a century.” See Gunther, Sani, and Shabad (1986, p. 171).

8.

Although the new democratic right, particularly the Social Democratic Centre (CDS), absorbed part of the former elite and the semi-opposition, no party emerged that was formed by “heads of state, ministers, and key members of the security apparatus” (Loxton, 2015, p. 159).

9.

In the Czech Republic, the anti-communist movement Civic Forum disintegrated into various parties, with the most prominent being the conservatives of the Civic Democratic Party (ODS), from which the Freedom Union (US) split in 1998. In Slovakia, HZDS was the major representative of the anti-communist movement, Public Against Violence, which had formed in 1989 to help bring down the communist regime. However, HZDS has often been incoherent; their program was said to be compatible with various parties, notably the post-communist SDL. Although difficult to place in the left–right axis, the party has been defined as closer to the Christian Democrats and Conservatives. In Poland, “post-Solidarity” elites split due to differences in opinion regarding the pace of economic reform, the scope of presidential power, church–state relations and the extent of welfare rights. The Democratic Union (UD) was the biggest participant in the Round Table negotiations, becoming the largest center-right party between 1991 and 1993. In 1994, new parties emerged, particularly the Freedom Union (UW), a party that aligned with Solidarity Electoral Action (AWS) during the 1997–2001 government. In Hungary, Fidesz was initially the youth organization of the Alliance of Free Democrats (SZDSZ). They both originated from the same dissident group and were both part of the Round Table negotiations. Fidesz made a pronounced shift toward the political right, in which the idea that the country had to deal with its past and to introduce lustration measures was always an important component. Hungarian Justice and Life Party (MIEP) was a nationalist party close to the extreme right. See Haughton (2001); Higley, Kulberg, and Pakulski (2002); and Nalepa (2010).

10.

In Spain, the PCE, created in 1920, was the major opposition (clandestine) party after the civil war. Its “national reconciliation” standpoint, adopted in 1956, the abandonment of orthodox communism in the beginning of the 1970s, and the strategy to reform its former revanchist image, each contributed to its role in the transition and in the new democratic regime. In Portugal, the PCP, created in 1921, was also the major opposition party during the regime, but unlike the PCE, it kept its orthodox communist stand throughout democracy.

11.

Pooling the data, inserting country dummies and interacting them with the individual-level variables was also discarded as an option, due to several reasons. First, we would have to compute and analyze 37 interactions between individual countries and our variables of interest—ideological self-placement and party attachments—which would make the data analysis remarkably lengthy and tiresome for the reader, with no considerable gains in terms of robustness. Also, inserting country-level variables as predictors in individual-level data analysis with single-level regression techniques disrespects the hierarchical structure of the data.

12.

The regressions for the six countries under analysis are presented in three different tables for the sake of reader-friendliness. Countries were grouped based on proximity. Table 2 presents the results of the four models computed for the neighboring Iberian countries, based on data retrieved in 1985, about ten years after the beginning of the transitional processes. Table 3 groups the models computed for the Czech Republic and Slovakia, which were the same nation-state under the communist regime. Lastly, Table 4 displays the models computed for the remaining polities under analysis, Poland and Hungary.

13.

CDS was founded in 1982 after the dissolution of Adolfo Suárez’s UCD, that is, only three years before the survey was carried out in Spain. The party cannot be easily placed together with AP as it claimed to occupy the ideological center and it was founded by individuals who were even part of the regime opposition. Still, we believed that the leadership by Suárez had strong symbolic weight in the way its partisans assessed Francoism.

14.

ODS is one of the parties created after the dissolution of the anti-communist platform Civic Forum. It is a conservative party that ruled the country for most of the time preceding the survey. The US was born as a split from ODS, in 1998.

15.

A coalition composed, among others, of right-wing parties, namely, liberals (Democratic Party, DA) and Christian democrats (KDH), but also greens (SZS) and others.

16.

The results of the HZDS may be due to the fact that this nationalist conservative party adopted economic left-wing stances and fought the trend of privatization and liberalization in the early stages of the post-communist regime, being therefore somewhat connectable with the communist past.

Barnes
,
S. H
. (
1972
)
The legacy of fascism: Generational differences in Italian political attitudes and behavior
.
Comparative Political Studies
.
5
(
19
),
41
57
.
Beverly
,
C
. &
Lijphart
,
A
. (eds.) (
1997
)
Liberalization and Leninist legacies: Comparative perspectives on democratic transitions
.
Research series, no. 96
.
Berkeley, CA
,
International and Area Studies
.
Biezen
,
I
. (
2003
)
Political parties in new democracies: Party organization in Southern and East-Central Europe
.
Basingstoke and New York
,
Palgrave Macmillan
.
Bozóki
,
A.
&
Ishiyama
,
J. T.
(eds.) (
2002
)
The communist successor parties of Central and Eastern Europe
.
Armonk, NY
,
M. E. Sharpe
.
Dinas
,
E
. (
2017
)
Political socialisation and regime change: How the Right ceased to be wrong in post-1974 Greece
.
Political Studies
.
65
(
4
),
1000
1020
.
Dinas
,
E
. &
Northmore-Ball
,
K
. (
2019
)
The ideological shadow of authoritarianism
.
Comparative Political Studies
.
53
(
12
),
1839
1860
.
Ekman
,
J
. &
Linde
,
J
. (
2005
)
Communist nostalgia and the consolidation of democracy in Central and Eastern Europe
.
Journal of Communist Studies and Transition Politics
.
21
(
3
),
354
374
.
Freire
,
A.
,
Costa Lobo
,
M.
,
Magalhães
,
P
. &
Espírito-Santo
,
A
. (eds.) (
2005
)
Comportamentos e atitudes políticas, 1973–2002. Comportamento eleitoral e atitudes políticas dos portugueses
[
Political behavior and attitudes, 1973–2002. Electoral behavior and political attitudes of the Portuguese
].
Lisboa, ICS
,
Imprensa de Ciências Sociais
.
Freire
,
A
. &
Kivistik
,
K
. (
2016
)
Regime transition, value conflicts and the left-right divide at the mass level: The Baltic States and Southern Europe compared
.
Communist and Post-Communist Studies
.
49
(
4
),
293
311
.
Gherghina
,
S
. &
Klymenko
,
L
. (
2012
)
Why look back? Citizens’ attitudes toward the communist regime in Belarus, Russia, and Ukraine
.
Problems of Post-Communism
.
59
(
1
),
55
65
.
Grzymała-Busse
,
A. M
. (
2020
)
Consequences of authoritarian party exit and reinvention for democratic competition
.
Comparative Political Studies
.
53
(
10–11
),
1704
1737
.
Grzymała-Busse
,
A. M
. (
2002
)
Redeeming the communist past: The regeneration of communist parties in East Central Europe
.
Cambridge studies in comparative politics
.
Cambridge and New York
,
Cambridge University Press
.
Gunther
,
R
. &
Montero
,
J. R
. (
2004
)
Attitudes toward democracy in seven countries: Dimensional structure and behavioral correlates
.
Centre for the Study of Public Policy, University of Strathclyde
.
Report number 385
.
Gunther
,
R.
,
Montero
,
J. R
. &
Puhle
,
H-J
. (
2007
) Democracy and intermediation: Some attitudinal and behavioural dimensions. In:
Gunther
,
R.
,
Montero
,
J. R
. &
Hans-Jürgen
,
P
. (eds.)
Democracy, intermediation, and voting on four continents
.
Oxford
,
Oxford University Press
, pp.
29
74
.
Gunther
,
R.
,
Sani
,
G
. &
Shabad
,
G
. (
1986
)
Spain after Franco: The making of a competitive party system
.
Berkeley
,
University of California Press
.
Haughton
,
T
. (
2001
)
HZDS: The ideology, organisation and support base of Slovakia’s most successful party
.
Europe-Asia Studies
.
53
(
5
),
745
769
.
Higley
,
J.
,
Kulberg
,
J
. &
Pakulski
,
J
. (
2002
)
The persistence of postcommunist elites
. In:
Diamond
,
L. J
. &
Plattner
,
M. F
. (eds.)
Democracy after communism
.
Baltimore
,
Johns Hopkins University Press
, pp.
33
47
.
Hite
,
K
. &
Morlino
,
L
. (
2004
) Problematizing the links between authoritarian legacies and “good” democracy. In:
Hite
,
K
. &
Cesarini
,
P
. (eds.)
Authoritarian legacies and democracy in Latin America and Southern Europe
.
Notre Dame, IN
,
University of Notre Dame Press
.
Huneeus
,
C
. (
2003
)
Chile, un país dividido: la actualidad del pasado
[
Chile, a divided country: the present of the past
].
Providencia, Santiago de Chile, Catalonia
.
Huneeus
,
C
. &
Martín
,
J
. (
2000
)
El régimen de Pinochet. Crónicas y testimonios
[
The Pinochet regime. Chronicles and testimonies
].
Providencia, Santiago de Chile
,
Editorial Sudamericana
.
Huntington
,
S
. (
1991
)
The third wave: Democratization in the late twentieth century
.
Norman
,
University of Oklahoma Press
.
Ishyama
,
J. T
. (
1997
)
The sickle or the rose? Previous regime types and the evolution of the ex-communist parties in post-communist politics
.
Comparative Political Studies
.
30
(
3
),
299
330
.
Jowitt
,
K
. (
1992
)
New world disorder: The Leninist extinction
.
Berkeley
,
University of California Press
.
Karl
,
T. L
. (
1990
)
Dilemmas of democratization in Latin America
.
Comparative Politics
.
23
(
1
),
1
21
.
Kavasoglu
,
B
. (
2021
)
Autocratic ruling parties during regime transitions: Investigating the democratizing effect of strong ruling parties
.
Party Politics
.
Available from
: https://doi.org/10.1177/1354068820985280.
Kim
,
H. J
. (
2012
)
Structural determinants of human rights prosecutions after democratic transition
.
Journal of Peace Research
.
49
(
2
),
305
320
.
Lewis
,
P
. (
2001
)
Political parties in post-communist Eastern Europe
.
London and New York
,
Routledge
.
Linz
,
J. J
. (
1978
)
Crisis, breakdown & reequilibration
.
Baltimore
,
Johns Hopkins University Press
.
Linz
,
J. J
. &
Stepan
,
A
. (
1996
)
Problems of democratic transition and consolidation: Southern Europe, South America, and post-communist Europe
.
Baltimore
,
Johns Hopkins University Press
.
Lisi
,
M
. (
2015
)
Party change, recent democracies, and Portugal: Comparative perspectives
.
Lanham, MD
,
Lexington Books
.
Loxton
,
J
. (
2015
)
Authoritarian successor parties
.
Journal of Democracy
.
26
(
3
),
157
170
.
Loxton
,
J
. &
Mainwaring
,
S
. (eds.) (
2018
).
Life after dictatorship. Authoritarian successor parties worldwide
.
Cambridge
,
Cambridge University Press
.
Lueders
,
H
. &
Lust
,
E
. (
2018
).
Multiple measurements, elusive agreements, and unstable outcomes in the study of regime change
.
Journal of Politics
.
80
(
2
),
736
741
.
Mahoney
,
J
. &
Snyder
,
R
. (
1999
)
Rethinking agency and structure in the study of regime change
.
Studies in Comparative International Development
.
34
(
2
),
3
32
.
Morse
,
Y. L
. (
2020
)
Rebuilding, rebranding, and competitive landscapes: A set-theoretic analysis of authoritarian successor parties
.
Democratization
.
27
(
8
),
1477
1495
.
Munck
,
G. L
. &
Leff
,
C. S
. (
1999
)
Modes of transition and democratization: South America and Eastern Europe in comparative perspective
.
Comparative Politics
.
29
(
3
),
343
362
.
Nalepa
,
M
. (
2010
)
Skeletons in the closet: Transitional justice in post-communist Europe
.
New York
,
Cambridge University Press
.
Neundorf
,
A
. (
2010
)
Democracy in transition: A micro perspective on system change in post-socialist societies
.
Journal of Politics
.
72
(
4
),
1096
1108
.
Neundorf
,
A.
,
Gerschewski
,
J
. &
Olar
,
R
. (
2020
)
How do inclusionary and exclusionary autocracies affect ordinary people?
Comparative Political Studies
.
53
(
12
),
1890
1925
.
Pop-Eleches
,
G
. (
2007
)
Historical legacies and post-communist regime change
.
Journal of Politics
.
69
(
4
),
908
926
.
Pop-Eleches
,
G
. &
Tucker
,
J
. (
2017
)
Communism’s shadow: Historical legacies and contemporary political attitudes
.
Princeton, NJ
,
Princeton University Press
.
Pop-Eleches
,
G
. &
Tucker
,
J
. (
2011
)
Communist legacies and political values and behavior: A theoretical framework with an application to political party trust
.
Comparative Politics
.
43
(
4
),
379
408
.
Pop-Eleches
,
G
. &
Tucker
,
J
. (
2014
)
Communist socialization and post-communist economic and political attitudes
.
Electoral Studies
.
33
,
77
89
.
Raimundo
,
F
. (
2015
)
Strategic silence as a third way. Political parties and transitional justice
.
Democratization
.
22
(
6
),
1054
1073
.
Sani
,
G.
, &
Santamaría
,
J
. (
2005
[
1985
]) A cultura política do Sul da Europa: estudo sobre quatro nações—base de dados [The political culture of Southern Europe: A Four Nation study – database]. In:
Freire
,
A.
,
Costa Lobo
,
M.
,
Magalhães
,
P
. &
Espírito-Santo
,
A
. (eds.)
Comportamentos e atitudes políticas, 1973–2002. Comportamento eleitoral e atitudes políticas dos portugueses
[
Political behavior and attitudes, 1973–2002. Voting behavior and political attitudes of the Portuguese
].
Lisboa, ICS
,
Imprensa de Ciências Sociais
.
Santana-Pereira
,
J.
,
Raimundo
,
F
. &
Costa Pinto
,
A
. (
2016
)
An ever-shadowed past? Citizens’ attitudes towards the dictatorship in twenty-first century Portugal
.
South European Society and Politics
.
21
(
2
),
197
210
.
Share
,
D
. &
Mainwaring
,
S
. (
1986
) Transitions through transaction: Democratization in Brazil and Spain. In:
Selcher
,
W. A
. (ed.)
Political liberalization in Brazil: Dynamics, dilemmas, and future prospects
.
Boulder, CO
,
Westview Press
.
Schneider
,
C. Q
. (
2009
)
The consolidation of democracy: Comparing Europe and Latin America. Democratization studies
.
Milton Park, Abingdon, Oxon, and New York
,
Routledge
.
Skiba
,
L
. (
2005
) The people, the programme and the governments of the Democratic Left Alliance (SLD). In:
Kopecek
,
L
. (ed.)
Trajectories of the Left. Social democratic and (ex)-communist parties in contemporary Europe: Between past and future
.
Brno
,
Institute for Comparative Political Research, Masaryk University
.
Szczerbiak
,
A
. (
1999
)
Interests and values: Polish parties and their electorates
.
Europe-Asia Studies
.
51
(
8
),
1401
1432
.
Tavits
,
M
. (
2013
)
Post-communist democracies and party organization
.
Cambridge
,
Cambridge University Press
.
Valenzuela
,
J. S
. (
1992
)
Democratic consolidation in post-transitional settings: Notion, process and facilitating conditions
. In:
Mainwaring
,
S.
,
O’Donnell
,
G
. &
Valenzuela
,
J. S
. (eds.)
Issues in democratic consolidation: The new South American democracies in comparative perspective
.
Notre Dame, IN
,
University of Notre Dame Press
, pp.
353
375
.
Welsh
,
H. A
. (
1996
)
Dealing with the communist past: Central and East European experiences after 1990
.
Europe-Asia Studies
.
48
(
3
),
413
428
.
Williams
,
K.
,
Fowler
,
B
. &
Szczerbiak
,
A
. (
2005
)
Explaining lustration in Central Europe: A “post-communist politics” approach
.
Democratization
.
12
(
1
),
22
43
.