The Emergence of Right-Wing Partisanship in Poland, 1993–2018: Reconciling Demand-Side Explanations of the Success of Illiberalism

Abstract

Scholars studying the electoral breakthroughs of right-wing illiberalism have arrived at two general conclusions: while they largely rejected the hypothesis that this phenomenon is grounded in voters’ attitudinal shift, they have shown that those voting for the illiberal Right have distinguishing socio-economic and attitudinal characteristics. My analysis reconciles these two sets of findings by documenting the gradual emergence and transformation of the right-wing electorate in Poland in the period 1993–2018 and points to the consolidation of a right-wing partisanship as an organizing factor of the “illiberal moment.” Using the POLPAN panel dataset I find that populist and authoritarian attitudes indeed emerge in Poland in the twenty-first century to distinguish those supporting the Right more and more centered around the PiS party. These attitudes, however, have been incorporated in the context of partisan rivalry—right-wing voters, for example, are more supportive of limiting democratic procedures but only when the Right is in power. In the first decade of the twenty-first century PiS also politicized the lack of partisan consensus on the expansion of the welfare state. PiS incorporated this demand in its stance legitimizing the expansion of the welfare state through what was available in its ideological repertoire: national solidarity, national victimhood, and the idea of a sovereign nation-state joined under the umbrella of Catholic symbolism. This post-consensus polarization and asymmetrical political radicalization resembles the “illiberal moment” in Western Europe that followed the convergence between center-left and center-right parties but they lack a crystallized class-based political identity and social-democratic understanding of political economy to build on.

Academic discussions on the recent electoral successes of right-wing illiberal politicians tend to emphasize the novelty of this phenomenon. Often centered around the concepts of populism (Mudde 2016a; Meyer and Wagner 2020) and democratic backsliding (Bermeo 2016; Haggard and Kaufman 2021), these discussions frequently make use of the metaphors of “rise” and “surge” to describe the advance of illiberalism in Europe and the United States. While these metaphors are convincing when used to illustrate the growing support for specific types of parties and changes in countries’ democracy scores across time, analogous arguments about changes across time on the demand-side of politics are much more contested. On the one hand, there is a general consensus that there are attitudinal and socio-economic differences between those voting for the illiberal Right and the rest, even though these differences do not present identical patterns in every country (Berman 2021, 75–76). On the other hand, while there is also an agreement about the direct link between attitudinal shifts in society and changes in public policy (Wlezien 1995; Rehm 2011; Stimson 2015) the existence of this connection in the explanations for the “illiberal moment” is disputed. Some works trace it in the changes across time in aggregate attitudes, as well as the transformation and emergence of social classes (Goodwin 2019) whereas others do not find any demand-side explanations (Bartels 2017).

By building on and joining both approaches—that is, tracing changes in the attitudinal and socio-economic profiles of different parties’ electorates across time—I attempt to explain the recent advance of right-wing illiberal politics. To this end, I test on the case of Poland two major demand-side explanations present in the literature on right-wing populism and the radical right: socio-economic explanations pointing to changes across time in relative differences in the economic and social status of individuals in the context of crises and job market transformation (Gidron and Hall 2017; Goodwin 2019), and explanations focusing on changes across time in authoritarian and populist attitudes (Bonikowski 2017; Golder 2016). I link the results of my analysis to wider scholarly discussions on the economic and cultural drivers of political radicalization, the ways in which right-wing parties capture economically vulnerable populations—that is, traditionally left-wing voters—and the transformation of the anti-establishment, populist rhetoric of incumbent parties.

The Polish Panel Survey (POLPAN) (Słomczyński and Tomescu-Dubrow 2021), which has been conducted every five years since 1988, allows me to trace said changes in the last three decades with cross-sectional and panel models. My stress on understanding the current predicament through the lens of continuity has its basis in the argument that the politics of those classified as “national” populists (Brubaker 2017) builds on the right-wing identity proposition of the sovereign nation rooted in tradition (Art 2020) and constitutes an attractive security project in the face of globalization (Bauman 2007; Giddens 2002). To account for some of these long-term cultural and political transformations, I also include in my analysis variables measuring the attitude toward the Catholic Church and the European Union.

I focus on Poland not only because it has been an important case in the literature on democratic backsliding and right-wing populism, but also because the combination of continuity and change in the Polish case tends to confuse observers. In Poland, similarly to the United States and Hungary, the recent success of illiberal politics was not concomitant with the emergence of formerly marginal political parties. Poland has been ruled by the United Right (ZP), a right-wing coalition comprised of Law and Justice (PiS), United Poland (SP), and the Agreement (Porozumienie) since 2015. ¹ PiS, a parliamentary party since 2001, has been one of the two main actors on the Polish political scene for over 15 years, alternating in power with centrist liberal Civic Platform (PO). Right-wing politics, also in the form of governments led by right-wing parties, is hardly a novelty—it has been a constant feature of Poland since 1989. The novelty after 2015 consists in the radicalization—along traditionalist and nationalist lines—of policies pursued by the political actors present on the Polish democratic political scene since 1989. For these reasons, Poland is a perfect case for studying the role of both continuity and change in the recent political transformation.

My findings show that across the last three decades, many of the socio-economic and attitudinal features exhibit a general stability among voters on the Right at the aggregate level, while they change for the rest of the electorate (except for the opinion about the EU, where the situation is reversed). This partisan profile signals, after 2003, a newly emerging right-wing identity centered around the idea of a dominant, protectionist, and sovereign nation-state, a partisanship that also incorporates paradoxical elements: relative populist and authoritarian attitudes that depend on which party is in power. Supporters of radical right parties, therefore, regard the anti-elite politicians they vote for as members of the elite. I also identify the sources of the current polarized political system, tracing them to the early 2000s. I show that the unevenly spread risk of unemployment in the face of growing market competition at that time in Poland resulted in the contentious lack of consensus about the expansion of the welfare state (Rehm 2011) which fueled political polarization as the issue of redistribution was instrumentalized and acted upon by the Right. The lack of appropriate language to communicate the salient demand for a more expansive welfare state has its roots in the 1990s transition consensus. What was available to right-wing politicians as a legitimization for expanding social programs was national solidarity, national victimhood and the idea of a protective and sovereign nation-state joined under the umbrella of Catholic symbolism.

In the following sections I first discuss 30 years of right-wing politics in Poland, focusing on political parties and their voters. The goal of this section is to show the constant presence of, as well as continuities and ruptures in right-wing politics in Poland after 1991, that is, the first fully democratic parliamentary elections following the transition to democracy. I then describe the structure of the variables used in statistical analysis and present my hypotheses, results, and conclusion.

Three Decades of Right-Wing Politics in Poland’s Third Republic (1991–2019): Parties and Their Electorates

The Right as a Constant Part of the Public Sphere in Poland

Contrary to Western Europe, Polish society has the experience of living in an authoritarian single-party socialist state established in the end of the Second World War, and known in 1952–1989 under its official name as PRL (Polish People’s Republic). As a consequence, the political field controlling the distribution of power during the PRL—with a dominant position of PZPR (Polish United Workers’ Party—Polska Zjednoczona Partia Robotnicza) at the helm of the state, with ornamental roles played by its satellites: ZSL (United Peasant Party—Zjednoczone Stronnictwo Ludowe) and SD (Democratic Party—Stronnictwo Demokratyczne), and the growing anti-communist dissident movement taking clear shape in 1980 around the Solidarity workers’ union—strongly informed the post-1989 constellation of the democratic partisan landscape (Markowski 1997; Kitschelt et al. 1999; Shabad and Słomczyński 1999; Ekiert 2011; Buštíková and Kitschelt 2009). This post-communist specificity resulted in right-wing parties and politicians being explicitly anti-communist and emphasizing their roots in Solidarity (Gwiazda 2008, Koczanowicz 2008).

As the Right has been an inherent part of the public sphere after 1989, ZP is not the first right-wing government in post-communist Poland. Between 1991 and 1992, Poland was governed by three short-lived consecutive coalitions of mostly center-right and right-wing parties. Two of them, PC ² and ZChN, together with two centrist liberal parties—KLD and SD—formed a government with Jan Krzysztof Bielecki, who kept his position as prime minister for less than a year, until the end of the parliamentary term in 1991. The first government during the next parliamentary term survived for approximately seven months and comprised PC, ZChN and PSL-PL with Jan Olszewski as prime minister who was replaced for one month by Waldemar Pawlak from PSL. Hanna Suchocka from UD was prime minister between July 1992 and October 1993 at the helm of the consecutive government coalition composed of right-wing ZChN, and PSL-PL, as well as PChD, KLD, UD, and PPPP. The Polish Right came back to power in 1997–2001, after four years of government led by the social-democratic SLD, with the AWS-UW coalition, and again in 2005–2007 when PiS was first a minority government and afterwards (2006–2007) ruled in coalition with the radical right LPR and radical agrarian SRP.

The presence of right-wing parties in Polish politics goes beyond their roles in governing coalitions. This is clear from table 1, where I listed all parties and electoral alliances that I identified as being right-wing and radical right, as well as agrarian, centrist-liberal, and socialist. They are listed according to POLPAN waves, presented in the table in pairs, to flesh out the general diversification and the changing number of right-wing parties across three decades. I identified these parties and electoral alliances on the basis of two comparative datasets: the Manifesto Project Dataset (Volkens et al. 2021) and the 1999–2019 Chapel Hill Expert Survey (Jolly et al. 2022), works that use programmatic and policy-driven criteria in classification of the political parties and alliances in Poland (Kowalczyk 2014; Shabad and Słomczyński 1999; Lewandowski 2016; Jasiewicz 2008a; Minkenberg 2013; Buštíková 2018; Heinisch et al. 2021; Fijał 2022; Pankowski 2010), as well as those basing their classifications on respondents’ self-placement and party placement on the left-right scale (Flis and Kwiatkowska 2018; Kwiatkowska et al. 2016; CBOS 2021). Following most of these works, I do not treat economic policy as a decisive factor in the classification. Conceptualized in this way, right-wing politics in Poland has only partially resembled its Western counterparts by putting emphasis on national sovereignty and on the importance of upholding the national tradition and identity (Buštíková and Kitschelt 2009, 462) being classified closer to the Traditional/Authoritarian/Nationalist extreme on the GALTAN axis (Vachudova 2021, 484). I also accounted for the change occurring in party stances: PO started as a center-right, Christian-democratic party to subsequently, after 2005, moved to more explicit centrist-liberal positions whereas PiS, more or less simultaneously, moved toward radical right-wing positions (Buštíková 2018; Rydgren 2018) emphasizing more decisively the importance of traditional values, national myths, and cultural identity, as well as the need for economic protectionism.

Table 1 Parties and electoral alliances according to POLPAN waves

Note: *Classified as right-wing only in the 2003 wave.

For more discussion on the classification refer to note 1 in the online appendix.

Acronyms of parties and party alliances (1991–2018) not explained within the text are listed in the Appendix following the References section.

The partisan representation of the Right was, therefore, dispersed among many parties in the beginning of the 1990s to largely converge into a loose center-right coalition in 1997 in the form of AWS, an instable coalition in 2006–2007, and in 2015 and 2019 when there was one dominant right-wing party, PiS, with its two coalition satellites, as well as right-wing K’15 (8.81% of the votes and 42 seats in the parliament) in 2015–2019 and radical right KORWiN (6.81% of the votes and 11 seats) since 2019. The period during which PiS was at the lead of ruling coalitions is also the time when the position of the country’s president, who is elected directly and holds inter alia the veto power, was occupied by right-wing politicians: Lech Kaczyński (2005–2010) and Andrzej Duda (since 2015 until now).

Throughout these ebbs and flows, the support for the Right in parliamentary elections to the Sejm never fell below 29% since 1991, as I show in figure 1. The constant presence of the Right in the public sphere therefore made the 2015 double electoral success of ZP a surprise only with respect to the presidential election, where the incumbent Bronisław Komorowski, supported by PO, was defeated. As for the parliamentary elections, the surprise was associated with an unexpected consequence of the elections, that is, a parliamentary majority for ZP without the need for a post-electoral coalition, rather than the Right’s relative advantage over the weakened incumbent PO. Figure 1 also shows that support for the Right was a sinusoid since 1991, growing in one election to fall afterwards. This trend was broken twice: first in 2011, when support failed to rise after it had dropped in 2007, and in 2019, when the 2015 level of support was largely maintained by the Right.

Figure 1 Share of votes and seats in the Sejm held by right-wing parties, and support for the Right according to two POLPAN questionnaire items

Source: PKW 2023; Sejm 2023; Słomczyński and Tomescu-Dubrow 2021.

*PO as right-wing.

The Right Turning the Page: The End of the Transitional Consensus

The 1990s and the first half of the 2000s were marked by a general consensus among post-Solidarity and post-communist political elites. This post-communist “transitional culture” (Kennedy 2002) extended over the second decade of the twenty-first century and referred to economic and foreign policies (Markowski 1997, 223). In the economy, it consisted in the lack of divergence from the Polish post-communist variant of the “embedded neoliberal” regime, characterized by the partial continuity of welfare institutions from the state-socialist period and dependence on foreign capital and actors, including later the EU, in the construction of free market economic policies (Bohle and Greskovits 2012; Jasiecki 2017). The reasons for the foreign policy concord were different on each side of the political divide. The engagement in mainstream European politics in the beginning of the 1990s by SLD, an indirect successor of PZPR, was in large part perceived as protection against the surge of radical right-wing politics (Zioło 2009; Jasiewicz 2008a). On the other hand, the mostly post-Solidarity Right perceived European integration as a civilizational goal, a final break with the communist legacy, linked with corruption as well as social, political, and economic stagnation. After 2001, when the radical right-wing anti-EU LPR party won 36 seats in the lower chamber of parliament, Eurosceptic discourse found a space, if still marginal, in the mainstream, acquiring cultural and moral dimensions, striving to disengage from integration policies originating from the “decadent” West, and promoting the idea of a Europe of Nations (Shibata 2013). The presence of hardline Euroscepticism in the political mainstream was a clear sign that the cultural weight of this consensus left a lot of space for identity politics (Ost 2005; CBOS 2000). If we define the neoliberal project in globalist terms, consisting in abolishing all barriers to free trade (Slobodian 2018), then the Polish right-wing security project of restricting globalizing change, both cultural and economic, through national-level policies, had a very fertile base of support. Furthermore, it was easy for the Right to take ownership of economic grievances in Poland, because SLD, the main left-wing party, had been deeply engaged in transition politics and carried the political weight of PZPR while dealing after 2002 with a very consequential Rywin affair (Mencwel 2009).

The gradual end of this consensus, first on the level of discourse, then in policy, was concomitant with the emergence of a new main divide on the Polish political scene in late 2005, with the failed coalition negotiations between PiS and PO (Słomczyński and Shabad 2003; Bustikova and Kitschelt 2009; Żerkowska-Balas, Lyubashenko, and Kwiatkowska 2016; Matthes 2016). These two post-Solidarity parties that put out joint electoral lists in the local elections to 14 out of 16 voivodship councils (sejmiki wojewódzkie) in 2002 (Grabowska 2004, 290), have been the main rivals on the Polish political scene since 2005. The petrification of this divergence can be traced to the events and aftermath of April 10, 2010, the date of a catastrophic crash of the Polish presidential airplane with 96 passengers on board, among whom were president Lech Kaczyński, as well as members of the Polish parliament and military (Cześnik 2014). The trajectory of right-wing politics in Poland resembles in this way the “populist polarization” in Hungary in its paradoxical institutionalization of anti-establishment politics and polarization in the form of relatively stable governments, as well as a stable party system (Rovny and Polk 2017)—for almost 20 years—in which citizens vote for systemic alternatives instead of for parties (Enyedi 2020).

In consequence, after taking power in 2015, the ZP decisively upended the post-communist status quo in Poland. The most internationally visible move was to undermine the political independence of the judicial system, which resulted in a conflict with the European Commission over infringements of the rule of law (Sadurski 2020). Other policies pursued by the government consisted in a steady course towards centralization characterized by various pressures, including financial ones, being imposed on local governments (in urban areas mostly controlled by the opposition), thus weakening one of the fundamental reforms of post-communist Poland, aiming to establish and develop local and regional governance (Wojnicki 2020; Flis and Swianiewicz 2021). Through centralization of public financing, ZP has been marginalizing civil society organizations that were not acting in line with its political agenda, and pulling previously independent organizations closer to the state, thus leading to a heightened cultural war, with restrictions on the right to abortion in 2020 as its pinnacle (Kotwas and Kubik 2019; Bill 2022). The takeover of a large proportion of local and regional newspapers and journals, as well as attempts at taking control over liberal-leaning media, played an important part in this process (Klimkiewicz 2021).

Going along this route of systemic change in the realm of economic policy as well, in 2015 ZP headed the first government that broke with free-market subtexts by implementing, as its flagship policies, expansive social programs such as the “500 plus” child allowance and regular raises of minimum wage and pensions, as well as by undoing the PO-PSL government’s reform which had raised the retirement age. In this way, ZP expanded and strengthened the distinction, articulated by PiS since 2005, between their “solidarity-based” vision of Poland and PO’s “liberal” politics focused on securing the narrow interest of the winners of the transition (Cześnik and Kotnarowski 2011). ZP reversed policies of all consecutive Polish governments, including its own policies during its first time in power in 2005–2007, as well as the political choices of Jarosław Kaczyński’s first party, PC, which had not shied away from coalitions with economically liberal parties in the beginning of the 1990s. Importantly, the post-2015 change also extended to incorporate the rejection of closer EU integration (CBOS 2017). One of its symptoms has been PiS’s engagement with right-wing Eurosceptic political parties including the Hungarian Fidesz, the French National Front, the Spanish VOX, and the Italian League, accompanied, e.g., by common declarations of tighter cooperation in the future, including the idea to establish a new faction in the European Parliament (Le Pen 2021). This path for ZP was in Central and Eastern Europe (CEE) to the large extent paved by Viktor Orbán’s governments which also curtailed the media freedom, as well as carried out a thorough replacement of cultural, political, civil society, and business elites in Hungary (Enyedi 2016).

Right-Wing Voters

Since 1989, the majority of Poles have been able to place themselves on a left-right scale (Grabowska and Szawiel 2001, 251; Godlewski 2008). There is also a significant relationship between self-identification on the left-right scale and voting for parties placed as such (Kwiatkowska et al. 2016).

In terms of substantive attitudes, the right-wing electorate was characterized by a set of stable features in the last thirty years. One of them has been anti-communism. In the 1990s and 2000s, it translated into support for lustration and related attempts at excluding former functionaries of the communist party from public life (Grabowska and Szawiel 2001, 238–253; Killingsworth 2010). Previous research also suggests that people who are more attached to the institution of the Catholic Church, who regularly attend church services and declare that traditional Catholic values are inseparable from the Polish national identity, are more likely to support right-wing parties (Pankowski 1997; Grabowska and Szawiel 2001; Szawiel 2002; Jasiewicz 2002; Kwiatkowska et al. 2016; Żerkowska-Balas, Lyubashenko, and Kwiatkowska 2016). It is, therefore, not an accident that all right-wing parties emphasized the importance of the Catholic Church for Poland’s independence. In addition, in 2005 church attendance and self-placement on the right were already predictors for being a stable voter (Cześnik 2009, 117–120). Accordingly, the nature of populism in Poland, where it is connected almost solely with right-wing politics, has also been charged with cultural factors, lately by those concerning liberal cultural policies and trends politicized in partisan debates on European integration (Fitzgibbon and Guerra 2010; Fomina and Kucharczyk 2016; CBOS 2017). Authoritarian conservatism is often displayed by individuals supporting right-wing policies and parties in Poland and abroad (Swindal 2011; CBOS 2012; CBOS 2015; Freire and Kivistik 2016). These cultural and historical identities, rather than economic factors, were argued to be the most efficient weapons of the Right on the political battleground (Markowski and Cześnik 2002; Skarżyńska and Henne 2011; Kwiatkowska et al. 2016), and were accordingly emphasized in the explanations of its recent political success in 2015 (Matthes 2016).

Class boundaries and status differences, at least until the 2010s, were not regarded as characteristics dividing voters between those supporting the Right and the rest. Class distinctions, for example, have been used to identify those who were losers rather than winners of the economic transformation: a post-communist parallel to the losers of globalization discussed in the literature on Western Europe and the United States. The early 1990s brought disillusionment among workers and farmers with basic education, with the unfolding new economic regime not providing economic protection and stability comparable to that offered by state socialism (Kurczewski 1994, 418). This explicit resentment was not translated into any specific partisan support during the first decade of the Third Republic of Poland, with the vote of the working class (“skilled and unskilled workers” and “farmers and agricultural laborers”) being divided between social democratic SLD and UP and post-Solidarity center-right AWS. In contrast, the intelligentsia were much more likely to support liberal UW, and its predecessors UD and KLD (Domański 2008, 175–178).

Explanations of the Illiberal Turn

Since the right-wing electorate exhibited largely stable features over the last 30 years, explanations of the recent electoral success of illiberal politics in Poland must build on—but also go beyond—the issues discussed thus far. While some comparative works precisely point to the lack of radical shifts in public opinion that could have served as a convincing demand-side explanation for the recent surge of illiberalism (Bartels 2017), other authors consider the fact that populist discourse has been more ubiquitous in CEE lately (Engler, Pytlas, and Deegan-Krause 2019), focus on systemic crises and opportunities that opened up for “ethnopopulist” political entrepreneurs (Vachudova 2020) providing various security projects to their voters (Blyth 2016) through intensification of culturally salient discourses (Kotwas and Kubik 2019) or point to long-term processes of identity formation leading to the abandonment of social democratic parties by workers (Goodwin 2019). The question remains how these anti-establishment political actors keep their appeal when they become the elite. Attempts at solving this problem included the elaboration of “paternalist populism” consisting of “switching” public’s attention from the national political elite to the “real elite”—international institutions (Enyedi 2020) and polarizing national politics along the GALTAN axis as it was done by Andrej Babiš in the Czech Republic (Hanley and Vachudova 2018).

Some studies associate this electoral change with the 2008 economic crisis (Blyth 2016), that could have made the economic factor much more salient. Some works draw a straight line from the economic crisis to the radicalization of politics, pointing to Central and Eastern Europe suffering a “forgotten crisis” (Tooze 2019, 220-238). Analyses of voting behavior indeed show that economic assessments were an important factor shaping the vote after 2008. In 2011, a positive assessment of the condition of Poland’s economy was associated with a higher probability of voting for the incumbent PO-PSL, whereas those dissatisfied with the economy tended to vote for PiS (Kotnarowski and Markowski 2014). Other studies emphasize the importance of the socioeconomic factor while underlining that it became relevant already around 2005, with the beginning of the newly emerged PiS-PO divide (Jasiewicz 2008b) leading to a clear-cut partisan divide between the right-wing coalition and centrist-liberal voters (Tworzecki 2019), with the former being also more populist (Stanley 2019). ZP’s success was also perceived as a paradox by some, which is supported by the fact that its achievement of obtaining a majority of the parliamentary seats in 2015 was in large part a consequence of other electoral results: almost 16% of the vote was gathered by two parties and an electoral coalition that did not enter the parliament ³ (Markowski 2016). In addition, Poland did not experience high unemployment, austerity, or recession (Markowski and Kwiatkowska 2018). That said, the Right did increase its vote in 2015, reaching its 2005 level of support, and ZP got a larger share of the right-wing vote, which remained stable overall, in 2019. What is more, some authors, by focusing on the process of economic transformation and the fact that economic growth did not benefit all social groups equally, find the “illiberal moment” as less of a contradiction (Bernhard 2021).

Indeed, although the claims about the comparatively good shape of the Polish economy during the global financial crisis are correct on the aggregate level, if we look at the level of unemployment across time according to the level of education, it is clear that those with lower education suffered comparatively more; 2009 marks the beginning of growing unemployment for all education categories. However, whereas it grew by 1.2% (from 3.8% to 5%) between 2008 and 2010 for people with higher education, it grew by 3.75% on average for other categories during the same period, e.g. from 11.9% to 17.3% for “Lower secondary, elementary and lower secondary” (figure 2). The difference between those with the lowest education and those with higher education is especially striking in 2013, when the unemployment rate peaked, reaching 20.3% for the former and 5.7% for the latter population.

Figure 2 Rate of unemployment according to education level among professionally active population in Poland, 1995–2020

Source: GUS (Local Data Bank).

The fact that this situation took place after a period of declining unemployment could have left an imprint of authenticity on the PiS narrative, claiming that only the affluent part of Polish society benefitted from the PO-PSL government policies. The second half of the 2000s, therefore, was a period during which Poland witnessed a set of compounding factors, making the field for the right-wing security project more fertile. First of all, the demand for a qualified workforce has been actually growing in Poland, signaled also by the fact that in the mid-2000s it gained the label of “high-income country” (World Bank Group 2017). This phenomenon tends to leave people with basic education behind: they often lose their jobs or suffer lower wages, a process that can result in wage polarization without adequate investments in human capital (Rozelle et al. 2020). The risk of unemployment in Poland has, therefore, been unevenly distributed, a phenomenon that often leads to a lack of social consensus for generous unemployment benefits (Rehm 2011, 288) and for expansion of the welfare state in general. The result was that the less educated groups—in contrast to highly educated—experienced the opening of the economy as a risk, and turned to parties that could minimize this threat (Walter 2010). We can even expect that this wage polarization and uneven unemployment risk were experienced also as social and political polarization—those with much safer jobs perceived, for example, the raising of retirement age by PO-PSL government as a much more positive perspective opening the possibility to earn more money before retirement. For those with lower education, more often working in blue collar jobs, it meant more time spent in physically taxing work. This polarization has likely been built through inter alia personal interactions between, for example, family members.

Hypotheses

I expect that voting for the Right is associated with some constant characteristics, as well as some emergent ones. Attachment to the Catholic Church and national sovereignty (emerging in Euroscepticism), as well as anti-communism are expected to always be associated with support for the Right across the period studied. I also hypothesize that the support for the Right, due to the consolidation of its parliamentary representation in the form of one dominant political party on the Right and the structuration and growing stability of the whole political field (Gwiazda 2016, 112), resulted in consecutive statistical models better explaining support for right-wing parties. I expect that people with lower education are increasingly less likely to be included in the qualified labor force. To measure being “left behind,” my analysis includes household income per capita.

Building on these expectations, I hypothesize that the success of the Right is rooted in the changing reasons for switching to the Right, that is, the emergence of the economic driver that started to accompany the culturally rooted predictors. I expect that this shift occurs between 2003–2008, that is, the period of the political collapse of SLD, EU accession, the end point of the period of falling unemployment, as well as emergence of a new post-Solidarity divide and Eurosceptic LPR, later absorbed by PiS. During this change authoritarian and populist attitudes will transform to legitimize the takeover by the radical right.

Data, Variables, and Methods

Data and Variables

The dependent variable I use in the analysis is either dichotomous or categorical with three values, and contains information on partisan support at the time of the interview. The coding of the dependent variable follows the division of parties and electoral alliances as proposed in table 1. In the dichotomous variables used in logistic regressions the value of 1, therefore, always implies support for a right-wing party, whereas 0 can indicate either explicit lack of support for any party combined with the support for a non-right-wing party (first variant of the dependent variable), or just support for a non-right-wing party (second variant of the dependent variable). In multinomial regressions the outcome 0 stands for supporting right-wing parties, 1 means the lack of support for any party and supporting a non-right-wing party in the first variant, or the support for a non-right-wing party in the second, whereas value 2 stands for the support for a radical right party. I used two base outcomes in my analyses: 0 and 1. I compiled these variables using the questionnaire item referring to prospective voting. Depending on the POLPAN wave, this item takes one of two versions. The first version was used in 1993, 1998, and 2003 and contains two consecutive questions: “If the elections for Sejm and Senate happened in the near future, would you participate in them?” followed with a question on prospective party preference (“What party would you vote for?”). In 2008, 2013, and 2018, these two questions were replaced with a single item (“If there were elections next Sunday, which party would you vote for?”). Since I follow change across time in individuals, the measure of prospective voting is much more reliable in statistical models than the information on retrospective voting, which is also available from POLPAN (refer to table 1 in the online appendix). Using the questions about party support rather than about voting also makes the dependent variable closer to conventional measures of partisanship, which rely on different survey instruments that all aim to capture a concept different from voting itself (Bankert, Huddy, and Rosema 2017). The relation between partisanship and voting, however, is very strong (Holmberg 1994, 2007). My choice of “partisanship,” rather than voting, as the phenomenon of interest was also motivated by studies based on interviews with ZP’s supporters (Gdula, Dębska, and Trepka 2017) suggesting that consequent voting for PiS is very often concomitant with stronger identification with this party relative to other electorates (Sadura and Sierakowski 2019, 6–9).

The literature firmly states that voting for the Right in Poland is strongly associated with attachment to the Catholic Church, in the institutional and ideational sense. I constructed two variables measuring such commitment: frequency of attending church services, and opinion about the institution of the Catholic Church vis-à-vis politics. The first variable, named “Church attendance” has four values: 1 stands for “Never,” 2 for “Less frequently than once a month,” 3 for “At least once a month but less frequently than once a week,” and 4 for “Once a week or more frequently.” ⁴ The second variable, “Catholic Church and politics,” contains information on the extent to which the respondent agrees with the following statement: “The Catholic Church has too much influence on political decisions of the state.” I coded “strongly disagree” as 5 and “strongly agree” as 1. The remaining statements were coded in the following way: somewhat agree as 2, neither agree nor disagree as 3, and somewhat disagree as 4.

“Opinion about the EU” measures the general opinion about and support for Poland’s membership in the EU. For 1998, it is based on the question regarding the future vote in the referendum (“If the referendum regarding Poland’s entry into the European Union were to take place now, would you vote …”). It is coded 3 for voting against the accession, 1 for voting for it, and 2 for “difficult to say.” For the 2003 wave, when the outcome of the referendum was already known, the variable is based on the question about the predicted impact of EU membership on Poland (“Do you think that Poland’s entry into the European Union will have a positive or negative impact on your life, or will it have no impact at all?”). It is coded 1 for “positive (or somewhat positive),” 3 for “negative (or somewhat negative),” and 2 for “no impact at all.” In 2008, 2013, and 2018, this variable measures the general opinion about the impact of EU accession on Poland. In 2008, it is coded from 1 to 3 and follows the logic of the variable compiled for 2003. For 2013 and 2018, it has values from 0 to 10: the higher the value, the more negative the opinion about the impact of the EU accession on Poland.

I compiled two sets of variables suggested in the relevant literature as measures of respective populist and authoritarian attitudes. The first variable, “Conflict between the elite and the people” from the populism set, records the respondent’s opinion on the severity of conflict between the elite, referred to in the questionnaires as “those in power,” and the rest of society. The second one (“Minority yields to majority”) contains information on the respondents’ opinion about the statement “Democracy means that after elections the minority must yield to the majority.” I do not put the category of populism in the center of my analysis, treating it as a performative tool or repertoire (Bonikowski and Gidron 2016) rather than a distinctive characteristic of a party family (Müller 2016; Mudde 2007, 2016b).

The set of variables measuring support for authoritarian politics is organized around the notion of the centralization of power in the hands of a single party or leader. Each variable contains information on the opinion of respondents regarding two separate statements: “It would suffice if a single good political party ruled in Poland. Other parties would be superfluous then” (One good party is enough), and “If a wise person governs the country, he or she need not obey the law” (Good leader can be above the law).

These three variables measuring respondents’ opinions on characteristics of political regimes are coded from 1 to 5 (1 for Strongly disagree, 2 for Rather disagree, 3 for Neither agree nor disagree, 4 for Rather agree, 5 for Strongly agree) and the remaining “Conflict between the elite and the people” from 1 to 4 (1 for No conflicts, 2 for Rather weak, 3 for Strong, 4 for Very strong). The last variable measures respondents’ opinion about the period of state socialism (“Opinion about state socialism”). It ranges from 1 to 5 and follows the logic of the four variables mentioned earlier.

For factor analysis, ⁵ I used the following issue position questions: “Large differences in income are necessary to assure the prosperity of the country”; “The state should assist children from poor families in facilitating their access to higher education”; “The state is responsible for reducing differences in people’s incomes”; “The state should provide jobs for everyone who wants to work”; “The Catholic Church has too much influence on political decisions of the state”; and respondents’ opinion on the EU. I coded the last two variables as described previously. The rest is coded from 1 to 5 according to the following logic: the higher the value, the higher the support for the redistributive policies. Independent variables in multinomial logistic regressions were constructed on the basis of said factor scores.

“Gender” is coded 1 for male and 0 for female, whereas the “Age” variable specifies the age of each respondent at the time of the interview. I also included respondents’ “Place of residence” coded 5 for rural, 4 for municipalities with less than 20,000 inhabitants, 3 for those with 20,000–99,999 inhabitants, 2 for 100,000-499,999, and 1 for population over 500,000. The level of education was coded as 0 for above primary and 1for primary or lower.

A separate battery of variables focuses on the financial and social status of respondents. I included binary variables indicating whether the respondent is unemployed or not. For financial status, I included standardized household income per capita, and for social status, “Self-assessed social position” was compiled on the basis of a questionnaire item that recorded respondents locating themselves on the social position scale—the higher the number, the higher the position.

Methods

The second set of variables traces changes across POLPAN waves in some of the aforementioned measurements. Two panel variants of the dependent variable trace respondents who switched toward the Right across two consecutive POLPAN waves. The logic of their structure follows the way the main dependent variable was compiled. Therefore, for example, value 0 in variables tracing this change in support between 1993 and 1998 stands for not supporting the Right in 1993 and 1998, whereas value 1 indicates not supporting the Right in 1993 but supporting it in 1998. Depending on the variant, 0 stands either just for party support or party support combined with abstaining from any support.

Two remaining independent variables measure the change between standardized household income per capita and standardized self-assessed social position across two consecutive waves.

Results

Support for and switching to the Right in Poland had some constant and some emergent characteristics. Accounting for the changes that occurred in specific years, there were three independent variables characterizing the right-wing electorate: “Church attendance,” “Catholic Church and politics,” and “Opinion about state socialism.” As for the attitude toward the Catholic Church and church attendance, right-wing supporters stayed on the same level whereas the remaining respondents lowered their church attendance and started having a deteriorated opinion about its political role since 2000s. Therefore, it is not about a surge on the right but a decline for the rest (figure 7).

Figure 3 Characteristics of support for the Right across six consecutive POLPAN waves with two variants of the dependent variable – 0=would not vote and would not vote for the Right (●) and 0=would not vote for the Right (■)– with panel versions of Household income per capita and Self-assessed social position variables. Graphs present the odds ratios with Robust S.E.

Notes: ^ Difference between standardized scores for Household income per capita and Self-assessed social position; The difference always between the specified and antecedent POLPAN waves

Figure 4 Predictors of switching to the Right across two consecutive POLPAN waves with two variants of the dependent variable: 0=would not vote and would not vote for the Right (●) and 0=would not vote for the Right (■). Graphs present the odds ratios with Robust S.E.

Note. ^ Standardized

In “2008 => 2013” and “2008 => 2013 [with ch. attendance]” Education is omitted because 0 predicts failure perfectly

Figure 5 Switching from supporting agrarian, socialist, and centrist-liberal parties, the lack of support for any party and by retirees to right-wing and radical right parties across adjacent POLPAN waves

Figure 6a Multinomial logistic regression of a categorical variable: 0=would vote for a right-wing party, 1=would not vote or would vote and not vote for a right-wing party (●), 2=would vote for a radical right party (0=would vote for a right-wing party, 1=would vote and not vote for a right-wing party (■), 2=would vote for a radical right party). For figures 6.1, 6.3, 6.5, 6.7, and 6.9, the base outcome is 0, for 6.2, 6.4, 6.6, 6.8, and 6.10, the base outcome is 1. Dependent variable in figure 6.11 is binary (0=would not vote and would vote and not vote for a right-wing party/ would vote and not vote for a right-wing party, 1=would vote for a radical right party); scores for relative risk ratios

Figure 6b

Figure 7a Populist and authoritarian attitudes, church attendance, Catholic Church and politics, opinion about state socialism, opinion about the EU: average scores for those supporting right-wing, radical right, socialist and non-right-wing parties, as well as for the whole population (those with a specified partisan support and those abstaining from partisan support)

Figure 7b

The fact that in 1998 “Opinion about the EU” was either significant or not significant but always with a negative impact supports the claim about the existence of “transitional consensus” and means that at the time the impact of the EU (only potential at the time) could be regarded nationwide as a potentially positive move which, after a long period of state socialism, could facilitate the transformation into a modern and economically efficient state. Higher odds values for models including only the respondents who specified their support for a party indeed shows that this issue has been among the sources of political identity and has been organizing the divide between the Right and the rest since the early 2000s.

The character of support for the Right has to be put in the context of policy issues that have been salient in Polish society. Factor analyses and multinomial regressions show that support for a protective state existed since 1993 and was always a more salient issue for the radical right than for the right-wing. Since 2003 radical right parties were also more traditionalist. We can, therefore, say that there was always a more radical—although smaller during the first decade—political alternative to the moderate Right that supports the process of mainstream parties borrowing ideas and practices from their more extreme counterparts (Enyedi 2020). What is more, since 1993, support for the redistributive policies was almost always exhibited by those with lower income, lower self-assessed social position, and basic or lower education (refer to table 2 in the online appendix).

Interestingly, variables designed to measure populist and authoritarian attitudes are either non-significant or relative to whether a government was right-wing or not. Therefore, for instance, the probability of supporting the Right in 2018, that is, when the ZP was in power, grows as the perception of the degree of the conflict between the elite and the people declines. In contrast, in 2003 and 2013, when Poland was governed by SLD-led and PO-led coalitions respectively, this association had the opposite vector: the higher the perceived severity of this conflict, the higher the likelihood of supporting the Right.

Analysis of the average level of selected measures of attitudes, populist and authoritarian, across time in Poland only partially supports Bartels’s (2017) claim on the lack of explanation for the illiberal political change on the demand side. The changes were not observable on the aggregate level but appeared with time and were relative—they depended on which party was in power. There is also no specific proof for the economic crisis being the spark that ignited the fuse of illiberalism—the Right began to attract lower-educated Poles before unemployment started to grow. The post-2008 income polarization, however, definitely strengthened PiS’s political communication.

The relative character of variables measuring authoritarian and populist attitudes sheds light on the situation in which a party incorporating populist appeal, that is, going against the established elite to bring security to the people and bring moral balance in society, becomes the elite. The previously discussed explanation labeled as “paternalist populism” (Enyedi 2020) suggests that supporting the radical right incumbent from anti-establishment positions consists in “omitting” those in power and focusing on the “real” international elite instead. It can, however, imply, that the governing populists are not treated as an elite. The relative character of said variables, however, shows that an anti-elite in power is regarded as an elite by its supporters—it just is the “right” elite. My analysis, therefore, supports the results obtained in biographical interviews conducted with ZP supporters by Gdula, Dębska, and Trepka (2017), which show that PiS supporters identify with the party’s political narrative. Their right-wing partisanship has a very important emotional dimension consisting in strong attachment to the leader (Jarosław Kaczyński), through whom one can take part in the “political drama” as “a victim, a proud member of the national community or a person of strong moral principles.” The emotions are also inevitably negative, however, as they convey a strong negative attitude toward the EU establishment, as well as toward liberal and post-communist parties and politicians (Gdula, Dębska, and Trepka 2017). “Pisowiec,” a derogative term used by those opposed to Kaczyński’s party, is even used by its supporters as a term of endearment (Sadura and Sierakowski 2019, 6–7).

The kernel of my argument is the stress on the partisan identity, that is, a consequent attachment to a party or a politician even though their positions, scope of power, and even the policies they support might change. As I show in figure 8, the consequent attachment to the Right has been growing steadily since 2008, which suggests that Poles were staying with the same party more and more often, a result that largely contradicts works showing a trend of comparatively higher instability in Poland (Enyedi and Casal Bértoa 2018). One, therefore, cannot understand the role of populist and authoritarian attitudes and their connection with voting for radical right parties without acknowledging a strong relationship between the party and the voter. The question is what is the limit of partisan loyalty; my analysis suggests that becoming part of the establishment with all its beneficial consequences does not constitute this barrier.

Figure 8 Percentage of Poles staying with the Right across adjacent waves; according to prospective vote item and mixed prospective and perspective vote items; I – models with 0=not supporting/not voted for the Right, II – models with 0=not supporting/not voted for the Right, no party support/not voted

The fact that Church Attendance is a very robust predictor of switching to the Right apart from models (figure 4.3) explaining switching to the Right from 2003 to 2008 is also very telling. Both models presented in figure 4.3 point to a drop in household income per capita as a highly significant factor. In addition, in the same models those with primary education or lower had around two and a half times higher odds of switching to the Right (from supporting a different party or not supporting any) in 2003, and five times higher odds of switching their support to a right-wing party than those with above primary education, other things being equal. However, 2008 is the year when unemployment was still falling, and the opposite trend started only in 2009. As mentioned earlier, however, unemployment among those with basic or lower education was comparatively much higher.

According to these results, the 2003–2008 period seems to be a crucial moment for the emergence of a right-wing political identity, which will later be centered on PiS and ZP. For that period, the drop in income becomes the main cause for this decision to vote for the Right, which, as we can hypothesize, is a sign that the Right was successfully able to incorporate in its discourse the importance of a protective state. The obtained results also support the theory about the unevenly spread risk of unemployment in the face of globalization (Rehm 2011). This situation—the vulnerable population demanding the welfare state and those in favor of globalization being against it—fueled political polarization in Poland as the issue of redistribution was politicized and acted upon by the Right. The importance of the 2003–2008 period is also emphasized in other works that point to the 2005 presidential election as a crucial moment for the emergence of a new liberal versus solidary divide. Whereas in the beginning the social-solidary versus liberal divide was still overshadowed by the old post-communist versus post-Solidarity dimension of political competition and was not based on clear differences in socio-economic preferences or the demographic characteristics of voters (Szczerbiak 2007), these characteristics became explicit 10 years later (Szczerbiak 2017, 417). The 2003–2008 period is also the time when voters in Poland witnessed the rise and fall of populist parties (LPR and SRP) whose electorate was mostly absorbed by a more institutionally stable PiS which, together with PO, and to a lesser extent SLD and PSL, molded and adapted to the new rules of party competition, now more populist in character (Casal Bértoa and Guerra 2018).

It is difficult to say whether the Right, increasingly consolidated around PiS, moved consciously toward explicitly statist policies in 2005 and after losing power in 2007, or was forced into it in the face of PO’s success accompanied by the depoliticization of economic and social development. On the basis of the obtained results, we can see that at some point this positioning became consciously cultivated, but this was not necessarily the case initially. As is demonstrated in this analysis, as well as in other already-mentioned works (Kurczewski 1994, 418), there was no consolidated or self-identified working-class or pro-distribution electorate that could have been taken over by the right-wing coalition. These voters were scattered among many groups. As a result, the common argument that the growing prominence of right-wing parties is a consequence of the collapse or weakening of social democracy (Goodwin 2019) finds little support in this case. The abandoning of the working class by social democracy cannot be said to have happened in Poland, because there was no general “embracement” of the working class and economically vulnerable populations by social democracy, which was mostly post-communist to begin with. The “abandonment” by the Left in Poland consisted rather in the inability of the older left-wing parties like SLD and younger ones like Together and Ruch Palikota to construct a political proposition that would appeal to the traditionally defined left-wing electorate. As figure 5 suggests, ZP gathered their votes across the board but mostly from agrarian and socialist parties, as well as from those voters previously not declaring any support. That said, the reason for switching from socialist parties to the Right does not seem to be economic, but also connected which Church attendance (see table 4 in online appendix).

Conclusion

The changes occurring in Polish politics can be regarded as a process consisting in capturing voters who would be expected to support left-wing parties. In contrast to the West, however, the working class and the unemployed did not have a mainstream party representing their economic interests in Polish politics in the 1990s and early 2000s. The main party on the left after transition, the post-communist SLD, did not have an economically left-wing, redistributive agenda.

We can assume that the decision by the Right to break with the consensus on economic policy was largely due to the decline of the post-communist versus post-solidarity divide and the demise of the SLD. This economic realignment is one of the aspects of PiS-PO polarization post-2005. The right-wing identity started to emerge as the old political divide collapsed and a new one started to form. By rejecting the remnants of the transition consensus on economic policy, the Right became the first to capture the longing for a protective nation-state. My analysis suggests that, paradoxically, Poland in the 1990s experienced partisan consensus on liberal economic policies and the salience of the redistribution issue in society. Already then, this demand was more characteristic for the traditionalist and mostly marginal radical right than it was for the moderate right. There was, therefore, no appropriate mainstream political language that would articulate the demand for redistribution. What was available to politicians was Catholic symbolism joining national solidarity and, in its radical versions, national victimhood and chauvinism. PiS leaders, who were in the same milieu with more radical right-wing politicians, could build on a powerful framework that had already been tested, if by marginal actors, throughout the 1990s and 2000s (Pankowski 2010). This post-consensus polarization and asymmetrical political radicalization largely resembles the “illiberal moment” in Western Europe that followed the convergence between center-left and center-right (Kitschelt 2007) but lacks universal class-based political identity and social-democratic understanding of political economy to build on.

Supplemental Materials

Please consult the online appendix for further discussion on the classification of right-wing political parties, tables presenting the information on the overlap in percentage between voting and support for the Right across time, figure presenting the percentage of Poles switching to the Right across adjacent POLPAN waves, linear regressions of predictors on factor variables used in this article, logistic regression models (support for the Right with non-panel versions of Household income per capita and Self-assessed social position variables and switching from socialist parties to the Right across time), as well as full regression tables referring to figures in the article.