Abstract
This paper uses a new set of questions to analyse the impact of populist attitudes on party preferences and voting behaviour in the 2015 Polish parliamentary elections. At these elections, voters faced a choice between two broad blocs: parties that accepted the “liberal-orthodox” model of post-communist politics, and those that rejected this model and the political elites associated with its implementation. I find that there is a coherent set of populist attitudes among the Polish electorate, and that it correlates with economic and cultural attitudes in ways consistent with the supply-side divide between liberal and anti-liberal parties. Analysis of the individual and combined impact of these attitudes on voting behaviour reveals that populism plays a significant role both in structuring the sentiments of voters towards particular kinds of political parties and in determining how they cast their vote.
Introduction
Populism is a blossoming research area that has generated numerous articles and monographs in recent years. Yet aside from ongoing attempts to refine and redefine theories of populism, the literature is mainly dedicated to populism’s supply-side manifestations, with single-case and comparative studies of the emergence of populist parties, the broader ideologies of these parties, their participation in government, their interaction with other parties, and their impact on political and party systems. There has been comparatively little work that examines populist attitudes among the electorate and the impact of these attitudes on voting behaviour. In a political environment increasingly shaped by the clash between a liberal-democratic mainstream and the insurgent forces of anti-liberal populism, it is important to understand whether party preferences and vote choice are driven by populism as well as the traditional ideological attitudes that most research to date has focused on.
With this article, which analyses the attitudes and behaviour of voters at the 2015 Polish parliamentary election, I aim to contribute to an emerging body of work that addresses this gap in the literature. The absence of adequate measures has hitherto made it difficult to operationalise populist attitudes, with proxy variables that tap attitudes to democracy and specific policy issues such as immigration and European integration proving inadequate to the task. Yet recent years have seen more systematic attempts to distil the essence of populism into a battery of survey questions, resulting in studies of populist attitudes and voting behaviour based on direct operationalisation of populism (see, e.g., Akkerman et al., 1 Elchardus and Spruyt, 2 Hawkins et al., 3 and Stanley 4 ).
The availability of data sets using these questions opens new possibilities for the literature on populism. These data allow us to ascertain whether populist attitudes cohere as a distinct set of dispositions, which is important in its own right, since the presence or absence of those attitudes on the part of the voting public may help to explain the successes and failures of parties that make populist appeals to the electorate. The creation of index variables of populism facilitates research into how populist attitudes correlate with other sets of attitudes at the demand side, as “thin ideological” theories of populism suggest they ought. 5 The measurement of populist attitudes also allows us to identify who the populists are: what socioeconomic status they have, what personal and psychological characteristics they possess, and how they differ from those who reject populism. Finally, these data allow us to address the issue of how populist demand interacts with populist supply: the extent to which populist attitudes among the people determine the electoral successes and failures of populist parties.
My overall aim in this study is to determine whether the supply-side divide between populist and non-populist parties that has emerged over the last decade in Poland is reflected at the demand side. First, I derive a set of hypotheses from a discussion of the emergence of this populist divide and its relationship to a “liberal-orthodox” model of post-communist political development. I then conduct a series of statistical analyses to identify whether populist attitudes exist as a coherent set of dispositions, whether these attitudes are correlated with other ideological issues, whether populist attitudes help to determine attitudes towards political parties, and the extent to which these attitudes explain voting behaviour.
My analysis confirms the persistence of certain long-established patterns in Polish voting behaviour, in particular the importance of cultural values and the relatively minor role played by economic attitudes. However, it goes beyond these findings to show that populist attitudes play an important role in Polish politics, independently differentiating preferences for political parties and intensifying the impact of cultural attitudes on party preferences and voting behaviour.
From Supply to Demand: Framing the Research Questions
In this paper, I address two research questions about the demand side of Polish populism:
• Do populist attitudes correlate with economic and cultural attitudes?
• Do populist attitudes predict voting behaviour for populist parties?
To develop hypotheses to answer these questions, it is necessary first to discuss the contribution populism has made to structuring the Polish party system from the supply side. I follow the definition of populism as a “thin ideology” which holds that politics is about the antagonistic relationship between a corrupt, illegitimate, and usurping elite and an authentic, morally upstanding, and sovereign people. 6 Populism, thus conceived, is “thin” because it does not offer a comprehensive set of answers to the major issues with which modern democratic politics deals. Rather, it is an ideology whose main purpose is to aggregate the various discontents of individuals and social groups around a narrative of blame and moral condemnation. Populists tell voters that the problems they are experiencing are the fault of out-of-touch and unaccountable elites, and they propose to restore genuine popular sovereignty, either in the form of instruments of direct democracy or via a more “authentic” representation of the interests of the people.
While populism may stand alone conceptually, to achieve concrete political goals populist parties tend to require “thicker” ideologies to fill out populism’s limited set of concepts with a fuller set of principles and policies. Populism’s simplicity and flexibility make it adaptable to a wide variety of political contexts. In Western Europe, the dominant form over the past few decades has been the populist radical right, with the target of populist criticism evolving from the overweening state in the 1970s to the cosmopolitan, trans-national liberal elites of the era of globalisation. In Central and Eastern Europe, populists either have followed their Western European counterparts in attacking the liberal elites responsible for the disruptive impacts of post-communist transition or have articulated a “centrist populist” critique of the corruption and incompetence of incumbent political parties. Recent years have seen convergence on the anti-liberal populist model across Europe, as a beleaguered European elite appears increasingly vulnerable to a potent combination of nationalism, economic interventionism, and opposition to multiculturalism.
Populist Entrepreneurialism and Political Backlash
As Kriesi and his collaborators 7 illustrate, the populist radical right has put itself at the forefront of cleavage change in Western Europe over the last few decades by articulating a set of responses to globalisation: in particular, the economic liberalisation brought about by neoliberalism; the cultural challenges of immigration; and the hollowing out of national-level politics as a result of political integration. This appeal has begun to resonate with cohorts of relative “globalisation losers” who have lost out—or, at least, perceive themselves to have lost out—as a result of these processes. The populist element of this discourse indicts the mainstream liberal-democratic elites as responsible for the plight of those left behind by globalisation. Yet it also provides a clear expression of a new structural conflict in Western European politics: between liberal-capitalist, cosmopolitan, and Europhile “winners”; and “financial-nationalist,” 8 culturally monist, and Eurosceptic “losers.”
As Kriesi 9 has recently reflected, the comparative lack of institutionalisation of party systems in Central and Eastern Europe makes them “even more susceptible” to these processes. Transition to democracy was based on a set of predicates that were imitative of the Western European “winning formula”: the “liberal-orthodox” model of transition mandated the creation of capitalist economies, the building of liberal-democratic institutions, and the establishment and protection of the rights of the individual. In the first few years of transition, Kitschelt 10 anticipated that the clash of interests between relative winners and losers of transition would result in competition between pro-market, cosmopolitan, internationalist parties and interventionist, nationalist, and anti-integrationist parties.
It should be emphasised that there is no necessary connection between populism and the anti-liberal politics of economic interventionism, nationalism, and opposition to political integration, and indeed parties in both halves of Europe have opposed the key tenets of cosmopolitan liberalism without employing populist appeals. However, there are obvious complementarities between the anti-elite critique offered by populism and opposition to the prevailing political orthodoxies. If the assumption of a populist aspect by the Western European radical right has been a gradual process, in Central and Eastern Europe it was evident from the beginning of transition that post-communist politics was “an elite project driven by small groups at the apex of politics, business, academia, and officialdom.” 11 As Auer 12 has argued, the rise of populism across Europe can be attributed to a “growing disconnect” between this exclusive coterie of enlightened decision makers and an “ordinary people” ever less inclined to accept their claims to wisdom. Populism gives political entrepreneurs an effective means by which to broaden their appeal from specific niches to a heterogeneous range of popular resentments by pointing to the common cause that lies behind them all. This is an increasingly effective political strategy both in the dealigning party systems of Western Europe and the magmatic party systems of Central and Eastern Europe.
Anti-Liberal Insurgency: The Supply Side of Polish Populism
Populism has not always been a significant feature of the Polish party system. During the first decade of post-communist democracy, populist parties were minor and mostly extra-parliamentary, or part of broader coalitions whose ideological appeal they struggled to influence. The party system was dominated by a “regime divide” between post-communist and post-Solidarity formations. This divide mainly reflected different historical identifications and varying levels of religiosity, but there was a broad consensus about the goals and principles of liberal-democratic transition.
This picture changed in 2001, when an upsurge in disenchantment with the reforms of transition and the political class responsible for their implementation led to significant party system change, with four parties entering parliament for the first time. Two of these parties, Self-Defence (Samoobrona Rzeczypospolitej Polskiej, SRP) and the League of Polish Families (Liga Polskich Rodzin, LPR) were not allied with either side of the regime divide. Instead, they articulated a populist critique of the transition elite as a whole, contesting the legitimacy of the current political system and decrying the economic inequalities it generated and the threats they felt it posed to Polish traditions and identity.
The process of transition was essentially brought to completion with Poland’s accession to the European Union in 2004. However, the post-communist coalition that ruled between 2001 and 2005 did not reap the benefit of this, haemorrhaging support as a consequence of numerous corruption scandals. The chief beneficiaries were the other two parties that emerged in 2001: Law and Justice (Prawo i Sprawiedliwość, PiS) and Civic Platform (Platforma Obywatelska, PO). Both owed their origins to different currents of the post-Solidarity elite, and built their popularity primarily on the promise to deal with the pathologies of transition. As such, they were expected to form a governing coalition in the next parliament. However, the dynamic of the 2005 dual presidential-parliamentary elections resulted in conflict between PiS and PO to the extent that coalition between the two parties became impossible. 13 Instead, after a period of minority rule, PiS invited SRP and LPR to form a government.
Given the mercurial character of the minor parties, few had anticipated the formation of this “populist coalition.” Yet at an ideological level, the three parties offered critiques of transition liberalism that may have varied in focus and sophistication, but together offered a comprehensive and compelling alternative narrative for those who rejected the liberal orthodoxy (see Stanley 14 for a description of the coalition’s platform). During its short and controversial period in office, PiS succeeded in marginalising its radical coalition partners and absorbing a significant proportion of their electorates, but in doing so became significantly more radical and populist in its own right.
The early 2007 elections confirmed the realignment of the main line of competition in the party system. PO, the winner of the election and main party of government thereafter, was now the chief defender of liberal-democratic transition, while PiS, the main party of opposition, remained adamantly opposed to the constitutional settlement. The two main parties also followed very different political strategies: PO emphasised consensus and caution, while PiS sought to deepen the divide between the liberal elites of transition and the popular majority of “real Poles.”
PiS’s trajectory can be compared to that of the Hungarian party Fidesz, whose successful return to power after eight years in opposition PiS leader Jarosław Kaczyński sought to emulate. As Batory 15 observes, Fidesz’s evolution was unusual for the region, in that it transformed itself into a stronger political force by moving away from the moderate ideological mainstream rather than towards it. Indeed, the party’s ideological shift was an explicit response to the hegemony of the liberal consensus. If transition reformers viewed a rights-based constitution, open markets, cultural liberalism and European integration as a panacea for the ills of communism, Fidesz regarded this liberal-democratic political system as responsible for the ills of post-communism. 16 They sought to replace the system of checks and balances with one in which the executive branch was supreme, with clear restrictions on the capacity of unelected bodies to control the actions of the government. In defiance of free-market orthodoxy, they pursued financial nationalism, seeking to curb the influence of foreign capital and currencies and undermine the independence of the national bank. 17 They legislated to enshrine a particular conception of Hungarian national identity and religiosity in the constitution, replacing the inauthenticity of liberal democracy with a “value-based democracy.” 18 Finally, they undertook all these changes in open defiance of the European Union, which was portrayed as the institutional embodiment of a cosmopolitan political, economic, and cultural establishment whose priorities and values were alien to those of ordinary Hungarians. 19 This alternative model clearly resonated with PiS’s diagnosis of the problems of transition, and provided a blueprint for action when they regained power.
With Poland one of the only countries in Europe to come through the Great Recession relatively unscathed, populism remained largely in abeyance during the two parliamentary terms between 2007 and 2015, re-emerging only at moments of particular emotional tension, such as in the aftermath of the Smoleńsk plane crash of 2010. However, in the months prior to the 2015 parliamentary elections, it became clear that voters were disillusioned by a PO-led government whose cautious, prudent style now looked like complacency, and whose long period in office had bred cynicism and corruption. As a party promising politics in a radically different mode, PiS benefited beyond its own expectations, winning both the May 2015 presidential elections and the October 2015 parliamentary elections by convincing margins.
The 2015 election also saw the emergence of two minor parties whose profiles largely corresponded to the divide between the two major parties. The Modern (Nowoczesna) party gained the support of many disenchanted PO voters by advocating a return to the economic liberalism PO had left behind. While Modern kept relatively quiet about social values, their liberalism was clearly more compatible with the mainstream consensus, and, while critical of the way in which PO had governed, they did not seek to delegitimise the political, economic, and cultural elites of the Third Republic. The same could not be said for the Kukiz Movement (Kukiz ’15), an organisation which leveraged the unexpectedly good performance of rockstar and social activist Paweł Kukiz in the presidential elections to enter parliament as the third largest party. Above all, the Kukiz Movement constituted a populist attack on the established political system and parties, depicting the party system as a closed cartel of professional politicians which had become entirely detached from the needs and concerns of ordinary Poles. The movement consciously sought to avoid defining itself ideologically, arguing that manifestos were a symptom of the problem with party politics. However, Paweł Kukiz’s social conservatism and the willingness of the movement to populate its electoral list with representatives of numerous minor radical right parties and movements bore witness to a strongly nationalist tendency, and its hostility to the European Union also suggested that it was congenial to the anti-liberal insurgency.
As PiS’s electoral list 20 won a majority of seats, they were able to avoid having to govern with the mercurial Kukiz Movement. Nevertheless, while the divide of government and opposition did not entirely correspond to the key line of ideological difference, the party system was defined by two essential camps: those who accepted the liberal-democratic status quo and those who sought to overturn it. 21
Populist Attitudes and the Liberal-Orthodox Model: Hypotheses
In summary, the last decade of Polish party politics has seen the crystallisation of a line of competition between populist parties that contest the liberal model and non-populist parties that broadly accept it. The political discourse of PiS and Kukiz links populist arguments in particular to a set of cultural attitudes that embrace traditionalism, nativism, and a scepticism towards supranational integration, and in PiS’s case to a more market-sceptic and interventionist set of positions on economic policy. On the other hand, PO and Nowoczesna avoid populist rhetoric and arguments, and have a broadly liberal stance on economic and cultural issues. This context is conducive to the articulation of populist attitudes among the electorate.
First, I examine the relationships between populism and the “thick-ideological” issues of cultural and economic liberalism. Given the association of non-populist currents with the “liberal-orthodox” model of post-communist political development, and the association of populist currents with opposition to that model, I also expect that the linkage of populist attitudes and opposition to liberalism has been shaped not only by the current supply-side configuration of parties but also by longer-term processes of economic, cultural, and political transformation over the last twenty-five years of post-communist democracy. I expect that the more an individual’s views accord with populism, the more likely they will be to oppose free-market capitalism, espouse traditional values, and hold a more sceptical view of European integration. As discussed in the following section, the stances of the Polish electorate on the thick-ideological issues that pertain to the liberal–anti-liberal divide are two-dimensional in character. The first dimension consists of economic issues. The second, which I shall refer to as the cultural dimension, comprises attitudes to moral values, the nation, and European integration. On this basis, I advance the following hypotheses.
Hypothesis 1: Identification with populist attitudes increases as:
(a) support for economic liberalism decreases,
(b) support for cultural liberalism decreases.
The existence of a set of populist attitudes does not necessarily mean that those attitudes are politically active, even if they resonate with those thick-ideological divides that are known to drive political preferences and choices. The second step of the analysis involves ascertaining whether populist attitudes influence sentiments towards political parties, whether separately or in combination with economic and cultural attitudes. If the supply side of populism is echoed at the demand side, then populist attitudes should distinguish those who have a positive opinion of PiS and Kukiz from those who have a positive opinion of PO and Nowoczesna. Given the linkage of populist attitudes with cultural and economic issues at the supply side, and also the expectation that demand-side populist attitudes will correlate with these issues, I expect that the effect of populism on sentiments towards political parties will be strengthened by economic and cultural attitudes.
Again, it is necessary to preface this hypothesis with the caveat that no assumptions about causality can be made here. Recent comparative research has suggested that there may be some justification in regarding populism as a latent set of attitudes that characterises the electorates of modern democratic polities, and which is “activated” by populist mobilisation in some contexts, but remains dormant in others. 22 Yet in the absence of appropriate data covering at least the last decade, we cannot determine whether populist attitudes have persisted throughout. It is therefore impossible to address the question of whether the persistence of demand-side populist attitudes has cued parties to represent the concerns of the electorate, or whether populist attitudes among the electorate have been generated and shaped by political entrepreneurs. In the absence of causal arguments, the hypotheses should be understood as correlational in character.
Hypothesis 2: As attitudes become more populist:
(a) the probability of having a positive opinion of PiS and Kukiz increases;
(b) the probability of having a positive opinion of PO and Nowoczesna decreases;
(c) the impact of these attitudes on (a) and (b) is reinforced by cultural and economic attitudes.
The third set of hypotheses concerns the impact of populist attitudes on voting behaviour. While populist attitudes may find expression in the form of positive and negative sentiments towards particular political parties, these attitudes are not politically salient unless they help determine voting choices. If populist attitudes are salient as well as active, then sections of the electorate who are more populist in their outlook will be more likely to vote for PiS or the Kukiz Movement, while those who are less populist will be more likely to vote for PO or Nowoczesna. Again, I expect that cultural and economic attitudes will strengthen the effect of populism on vote choice.
However, it is also possible that populist attitudes are active, but not salient; they help explain the preferences of voters for particular kinds of political parties, but do not explain actual voting behaviour. Indeed, if the populist-minded voter’s disdain for the political elite is more salient than support for populist parties, populism should be associated with an increasing propensity to abstain from voting. If the effect of populism on voting for specific parties is less significant than expected, this may be attributable to disproportionally high rates of abstention among those with populist attitudes.
Hypothesis 3: As attitudes become more populist:
(a) the probability of voting for PiS or the Kukiz Movement increases,
(b) the probability of voting for PO or Modern decreases,
(c) the impact of these attitudes on (a) and (b) is reinforced by cultural and economic attitudes,
(d) the probability of being a non-voter increases.
Methods of Analysis
To test these hypotheses, I use data from the 2015 Polish National Election Study. 23 First, I model the predictors of populist attitudes using ordinary least squares regression. The dependent variable is an index of populist attitudes (populism), derived from several variables that tap different aspects of populism (see Table 1). This operationalisation is rooted in the aforementioned thin-ideological conception of populism. While the available survey questions do not cover each of the core concepts of populism as comprehensively as I would like, they capture the fundamental antagonism between the people and the elite at the heart of populism, the notion of politics as a simple choice between moral absolutes, and the sense of conspiracy and crisis that accompanies this mode of understanding politics.
Survey questions comprising the populist attitudes index
Note: PGSW = Polish National Election Study; SD = standard deviation.
Two questions operationalise the sense of an us versus them divide at the heart of politics. Firstly, respondents were asked to what extent they agreed with the statement “In Poland, a few people have taken control of powers that should rightfully be exercised by the people.” This captures the populist argument that actual political power is wielded by a small elite instead of the people, and that this is illegitimate. 24 In similar vein, respondents were asked whether they agreed with the following statement: “It isn’t the government that rules over us; the people who are really controlling things are unknown to us.” As well as hinting at conspiracy against the people, this statement also captures the notion that the people have been deprived of genuine, authentic representation. 25 A further statement makes more explicit the sense of elite conspiracy: “Those who claim that there are powerful hidden forces conspiring against Poland are in many senses correct.” While belief in conspiracy theories is not of itself a core concept of populism, there is, as Castanho Silva et al. 26 have noted, both a theoretical complementarity and an empirical correlation between the populist’s mistrust of elites and the conspiracy theorist’s mistrust of malevolent groups that control events in their own narrow interest.
A further two questions tap the populist call for the restoration of power to the people and their legitimate representatives. Respondents were asked to what extent they agreed with the statements “Poland is in need of someone who will be strong enough to change our system of government entirely, and introduce a new, just order,” and “Solving the problems our country faces is very easy; it is simply necessary to give power to those who want to do this.” Although these questions do not explicitly raise the issue of popular sovereignty, they tap populism’s sense of the ideal political order as one in which power is exercised without the unnecessary intermediation of political institutions. 27 A further question captures the emphasis populists place on the need for radical and decisive change in circumstances that require immediate remedies, 28 asking to what extent respondents agree with the statement: “This is the last moment to rescue Poland in the face of a looming crisis.” Finally, the crucial moral dimension of populism 29 is tapped by a measure of agreement with the assertion that “everything in politics is either good or evil; the choice is a simple one.”
Polychoric factor analysis of these variables (see Figure 1) shows that they form a single factor accounting for 36 percent of the variance. As a further test of the reliability of this index, I calculated the ordinal alpha of these variables using the procedure described in Gadermann et al. 30 The result of 0.79 confirms that the index is robust, and that dropping items would decrease its reliability. Accordingly, I took the factor score as the measure of populism, and standardised it to a 0–10 scale for ease of interpretation.
Populist attitudes are regressed on two tranches of independent variables in successive ordinary least squares models. 31 The first of these models contains only socio-demographic controls, comprising age (age), gender (gender), region of residence (region), level of education (edlevel), religiosity (relig), and income (income). Further information about these variables is given in Table 2. While there are reasons to expect that some of these variables will be predictive of populist attitudes—in particular, level of education—for the purposes of the present analysis they primarily serve as control variables rather than variables of interest, and form the baseline model. The second set of independent variables operationalises the two key ideological dimensions. I created an economic attitudes index (economic) by taking the standardised factor score of attitudes to tax policy, social policy and privatisation, and an index of cultural attitudes (cultural) from questions measuring respondent positions on European integration, the church–state divide, foreign policy, and abortion. The ideological variables are entered into the second model, and their interaction is entered into the third model. The interaction allows us to ascertain the extent to which populist attitudes are predicted by a combination of economic and cultural attitudes.
Operationalisation of the independent variables
For the second hypothesis, I model attitudes to political parties by using variables that measure the degree to which a respondent likes or dislikes each of the four parties of interest (likePiS, likePO, likeKukiz, likeNowo). As these variables are ordinal scale, I use ordinal logistic regression models. The aforementioned socio-demographic variables are entered into these models as controls, and in successive models I add the populist index as an independent variable, the two ideological index variables, and an interaction of the ideological index variables.
For the third hypothesis, the dependent variable measures vote choice (including the decision not to vote), so I use multinomial logistic regression to estimate these models. 32 The independent variables are the same as those used in the ordinal logistic regression models.
Analysis of Findings
Populist Attitudes and Their Thick-Ideological Correlates
As the baseline model in Table 3 shows, some of the socio-demographic control variables have a substantive impact. Level of education has a strong relationship with populist attitudes: having higher education lowers the level of populist attitudes by around 1.5 points compared with those who have little or no education, while having secondary education lowers it by just over 0.5 points. Those who live in large towns and cities are also less likely to espouse populist attitudes than those who live in rural areas. Finally, increasing income also has a negative impact on populist attitudes, although in this case only a slight one. The impact of these variables diminishes slightly when ideological variables are added to the model, but they remain significant.
Regression of Populist Attitudes on Socio-demographic and Thick-Ideological Variables
Data Source: PGSW 2015 (2015).
Note: Standard errors are in parentheses. *p < 0.05; **p < 0.01; ***p < 0.001.
The baseline model explains 15.6 percent of the variance in populist attitudes. Adding cultural and economic attitudes increases the explained variance to 23 percent. As Figure 2 shows, both economic and cultural attitudes correlate significantly with populist attitudes. In both cases, the more liberal a respondent is, the less populist they are. The average respondent with strongly antiliberal attitudes on cultural issues scores just short of 7.5 on the populism index, while those with anti-liberal attitudes on economic issues score just under 7. In both cases, this declines to just over 5 among those with strongly liberal attitudes.
Polychoric factor analysis of populist attitudes
The relationship between “thick ideological” attitudes and populist attitudes
Interacting cultural and economic factors does not raise the explained variance appreciably, but helps elucidate the nature of the relationship between cultural and economic attitudes and populist attitudes. The plot of the interaction in Figure 2 shows how the impact of cultural attitudes varies at different levels of economic attitudes. If we hold economic attitudes at 0 (the solid line), then the level of populism does not vary over the range of cultural attitudes, as arrayed along the x-axis. If economic attitudes are held at the midpoint of 5 (the dashed line) then there is a clear decrease in the level of populism as cultural attitudes become more anti-liberal. If economic attitudes are held at 10 (the dotted line), then the impact of cultural attitudes is significant: respondents who have anti-liberal attitudes on both cultural and economic issues score nearly 8 on the index of populism, while those who have liberal attitudes on both dimensions score just over 3.
The conclusion to be drawn here is that cultural and economic attitudes reinforce each other in their relationship with populism. The more a respondent holds anti-liberal views on both cultural and economic issues, the stronger their adherence to populist views. We therefore find that hypothesis 1 is fully supported by the analysis. It should be emphasised that this association does not imply any causative claim about the impact of populist attitudes on thick-ideological positions, or vice versa. For the purposes of this analysis, it suffices to conclude that they co-vary.
Populism, Liberalism, and Attitudes towards Parties
The second set of models allows us to analyse the impact of populist attitudes on attitudes towards the key political parties. I shall first discuss each of the parties in turn, and then synthesise the findings.
As the baseline model in Supplementary material Table 4 shows, attitudes to PiS are influenced by a number of the same socio-demographic variables that are also related to populist attitudes. Those who live in a large town or city are less likely than those who live in rural areas to like PiS, while disliking PiS rises with levels of education. In addition, there is also a significant relationship between levels of religiosity and positive assessments of PiS: the less frequently someone attends church services, the less likely they are to think positively of PiS. If we then add populist attitudes to the model, we can see that there is a significant positive relationship between populism and liking PiS. However, when thick-ideological attitudes are added in successive models, the impact of populism diminishes and becomes insignificant.
This can be better understood by looking at the plots of marginal effects in Figures 3 and 4, which are derived from the full model. Both of these plots show the probability of being among those who strongly dislike PiS (0 on the 0–10 scale of the dependent variable) and the probability of being among those who strongly like PiS (10 on the 0–10 scale), given a particular value of populist attitudes and cultural attitudes respectively. The effects of economic attitudes are not shown, as these are statistically insignificant in almost all cases. At the lowest level of populism, the probability of disliking PiS is 0.24, and the probability of liking PiS is 0.09. In both cases, there is little change as levels of populism increase. However, as Figure 4 shows, the impact of cultural attitudes is substantial. The probability of strongly disliking PiS increases from almost 0 among those with strongly anti-liberal attitudes on this dimension, to 0.58 among those with strongly liberal attitudes. Correspondingly, the probability of liking PiS declines significantly.
The impact of populist attitudes on attitudes to parties
The impact of cultural attitudes on attitudes to parties
Attitudes towards PO are only weakly related to socio-demographic variables. However, there is a significant negative relationship between having populist attitudes and liking PO. The precise nature of this relationship is best illustrated by the plot in Figure 3. At low levels of populism, there is an approximately equal probability of liking or disliking PO. However, as levels of populism increase, the probability of liking PO declines only slightly, while the probability of strongly disliking PO rises to 0.29 among those who are strongly populist. Adding cultural values to the model, we find that liberal attitudes on this dimension are positively related to liking PO. As Figure 4 shows, however, it would be more accurate to say that liberal cultural attitudes are primarily related to not disliking PO. While the probability of liking PO increases from almost 0 among those with strongly anti-liberal attitudes on cultural issues to only 0.08 among those with strongly liberal attitudes, the probability of disliking PO declines from 0.52 to just under 0.05.
Analysis of the minor parties reveals a similar set of findings. While socio-demographic variables do not have an impact, there is a positive relationship between populist attitudes and attitudes towards the Kukiz Movement, while cultural attitudes have a negative impact on these attitudes. As Figures 3 and 4 show, the effect is again primarily related to the probability of strongly disliking this party. The probability of strongly disliking Kukiz is 0.34 among those with low levels of populism, and this declines to 0.19 among those with strongly populist attitudes. Conversely, while the proportions of those who strongly like and strongly dislike Kukiz are approximately equal among those with anti-liberal cultural attitudes, the probability of disliking Kukiz rises to 0.44 among those with the most liberal cultural attitudes.
The opposite is true of attitudes to Nowoczesna. Again, socio-demographic variables are of little importance, but both populist attitudes and cultural attitudes are significant predictors of attitudes towards this party. In this case, economic liberalism also has a positive impact, but its substantive effect is smaller. As in the previous cases, the effect of both variables is primarily related to the extent to which respondents strongly dislike the party. Among those with low levels of populism, the probability of liking or disliking Nowoczesna is approximately equal. However, among those with strongly populist attitudes, the likelihood of disliking Nowoczesna rises to 0.29. Cultural attitudes are also a strong predictor of the degree to which respondents dislike Nowoczesna: among anti-liberals, the probability of this is 0.39, but this declines to 0.09 among those with the most liberal cultural attitudes.
These findings give broad support to hypotheses 2a and 2b, although it is necessary to add the caveat that they primarily concern the extent to which the parties in question are disliked by respondents. Firstly, it is clear that populism is associated with more positive attitudes towards PiS and Kukiz, as predicted in hypothesis 2a. However, the effect of populism is diminished when cultural attitudes are added to the model; this is particularly the case for attitudes towards PiS. For PO and Nowoczesna, the relationship is clearer: in both cases, as populism increases, the probability of liking both parties declines.
There is partial support for hypothesis 2c. While both economic and cultural liberalism are negatively correlated with populist attitudes, economic attitudes do not play an important role in activating sentiments towards political parties. On the other hand, cultural attitudes are related to these sentiments in ways consistent with the relationship between populism and democracy.
Figure 5 displays the marginal effects of the interaction of populist attitudes and cultural attitudes, showing how the impact of populism varies with strongly anti-liberal and strongly liberal cultural attitudes. For PiS, the interaction is not statistically significant, confirming the overwhelming importance of cultural attitudes in determining sentiments towards this party. However, it is clear that populism and anti-liberal attitudes work in tandem to drive antipathy towards PO. Among those who have culturally anti-liberal attitudes but anti-populist attitudes, the probability of disliking PO is only 0.19. However, this rises to 0.72 among those who have both populist and culturally anti-liberal attitudes. The effect is similar, although not as extreme, in the case of Nowoczesna. In the case of Kukiz, meanwhile, the effect is reversed: among those who have strong culturally liberal attitudes and anti-populist attitudes, the probability of disliking Kukiz is 0.73; this declines to only 0.29 among those with culturally liberal attitudes and strongly populist attitudes. These findings suggest that the populist critique of the liberal-orthodox model of post-communist modernisation resonates, at least in part, with the sentiments of the electorate towards political parties.
The impact of interacted populist and cultural attitudes on attitudes to parties
Populism, Liberalism, and Voting Behaviour
We have established that populist attitudes are a significant predictor of positive and negative sentiments for populist and non-populist parties, and that they distinguish between these parties in a manner consistent with the thick-ideological attitudes of cultural liberalism. However, this does not necessarily mean that these attitudes—whether separately or in tandem—are decisive predictors of voting behaviour. To address the third set of hypotheses, we need to model vote choice.
Supplementary material Table 5 shows the results of a multinomial regression of vote choice on the same five models used in the preceding analysis. The baseline model, which contains only socio-demographic variables, explains 15 percent of variance. Adding populism to the model increases variance explained only slightly, to 17 percent. There is a more significant increase to 22 percent when adding cultural and economic variables, but the models with interactions do not add to variance explained.
The multinomial regression outputs are useful for telling us about the overall fit of the models, but to address the hypotheses, we need to ascertain the impact that specific independent variables have on the probability of voting for particular parties. Rather than discuss the relative probabilities of voting for parties in reference to a base category, it is more informative to estimate average marginal effects, which give us the actual probabilities of vote choices given specific values of the variable of interest, with other independent variables held at their values. Figure 6 displays the marginal effects of populism on vote choice for the parties of interest, and also for non-voters.
The impact of populist attitudes on vote choice
As expected, there is a positive relationship between populist attitudes and voting for PiS: while the probability of someone with anti-populist attitudes voting for this party is only 0.17, this increases to 0.32 among those with strongly populist views. By contrast, populism has a negative impact on voting for PO, declining from 0.28 to 0.11 across the range of attitudes. There is also a significant relationship in the case of Nowoczesna: among anti-populists, the probability of voting for this party is 0.14, declining to almost 0 among those with strongly populist views. However, there is no statistically significant relationship between populist attitudes and voting for the Kukiz Movement, or being a non-voter.
Cultural attitudes are of significant importance in predicting vote choice (see Figure 7). The probability of voting for PiS is 0.67 among those with the most anti-liberal cultural attitudes, while the probability of voting for PO is almost zero. On the other hand, among those with the most liberal views on cultural issues, the probability of voting for PO is 0.39, compared with only 0.04 in the case of PiS. Cultural liberalism also has an impact on the probability of voting for Nowoczesna, but there is no statistically significant relationship in the case of the Kukiz Movement. Those at either extreme are less likely to be non-voters than those who have a centrist position on issues of cultural liberalism.
The impact of cultural attitudes on vote choice
Consistent with earlier findings, economic attitudes have a much less significant impact on vote choice (see Figure 8). They are significant in only one case: the probability of voting for Nowoczesna rises from 0.02 among those with anti-liberal economic attitudes to 0.13 among those with pro-liberal economic attitudes. The interaction of populism with cultural and economic liberalism did not yield any significant findings, so these plots are not shown.
The impact of economic attitudes on vote choice
The analysis of attitudes towards parties established that cultural attitudes and populist attitudes combine to amplify the extent to which parties are liked or disliked. However, there is no clear evidence that the same effect is present in the case of vote choice, with no statistically significant changes in the impact of populism on vote choice at different levels of cultural liberalism. Instead, the two variables exert distinct effects on voting behaviour, while economic attitudes are insignificant when controlling for cultural and populist attitudes.
To summarise this discussion in terms of my hypotheses, there is partial support for hypothesis 3a: an increase in populist attitudes is associated with a rise in the probability of voting for PiS, but the same is not true of the Kukiz Movement. There is full support for hypothesis 3b, with the probability of voting for PO or Nowoczesna declining as attitudes become more populist. However, there is no evidence to support the hypothesis that cultural and economic attitudes reinforce the effect of populism on vote choice (hypothesis 3c). There is also no support for hypothesis 3d, with populist attitudes having no effect on the decision to vote or not. This lends further support to the finding that populism is primarily a marker of difference between the electorates of political parties, rather than an expression of disenchantment with the political system in general.
Conclusion
If populist ideas form a coherent set of political attitudes among the electorate, then these attitudes ought to play an active role in determining political preferences and choices in contextually propitious circumstances. As argued at the beginning of this article, the Polish election of 2015 offered such a set of circumstances. Although the results of the election meant that populists found themselves both in power and in opposition, the election campaign and the period preceding it was primarily a contest between parties that embraced populist ideology and those that did not. As argued in the first part of this paper, the emergence of a supply-side divide between populist and non-populist parties is not only a product of short-term volatility in what remains a semi-consolidated party system, and not just the expression of mounting dissatisfaction with the alleged failings of a dominant political elite. Rather, it is linked to a longer-term reorientation of the major line of political competition, a line that separated two camps: those accepting the liberal-orthodox politics of transition, and those rejecting this developmental model.
The findings of this paper go a significant way to confirming that the populist divide is important at the demand side of Polish party politics as well as at the supply side. First of all, it is clear that populist ideas form a distinct set of attitudes among the Polish electorate, regardless of electoral participation. While on average Poles tend towards a somewhat more populist mindset, there is sufficient variation in populist attitudes for them to serve as a set of “political potentials” to be exploited by competing parties.
At the level of vote choice, there is a clear difference between the electorates of the two major parties, PiS and PO, indicating that the populist divide is an integral element of the key line of political competition. While there is no clear evidence that populism informs vote choice among those who opted for Kukiz and Nowoczesna, the analysis of party preferences indicates that populism plays a significant role in structuring the sentiments of voters towards particular kinds of political party. This is relevant to the question of whether political parties have inculcated and consolidated populist or anti-populist attitudes among their loyal voters, or whether they have activated latent populist attitudes among the electorate. The fact that populist attitudes correlate with preferences for new parties (positively in the case of Kukiz; negatively in the case of Nowoczesna) lends limited but nevertheless important support to the latter argument, since by definition voters for new parties have not been “shaped” by their party of preference. However, a more convincing analysis of the “populist activation versus populist creation” question will only be possible when sufficient time series data have been collected.
Yet the story is not just about the activation of populist attitudes; it is also about how those attitudes combine with and reinforce thick-ideological divides. Although my findings with respect to the interaction of populism and thick-ideological issues are somewhat messy and incomplete, they provide demand-side support to the theory of populism as an ideology that combines with and “inhabits” more complex and policy-adjacent ideologies. Firstly, they confirm a persistent truism in the study of Polish voting behaviour: cultural issues remain the most important determinants of party preferences, while economic issues are rather unimportant. Yet this analysis breaks new ground by showing that populist attitudes serve to intensify the effect of cultural attitudes. While there is no evidence to suggest that vote choice is directly determined by a combination of populist and cultural attitudes, the analysis demonstrates that populist attitudes and cultural attitudes work together to shape the context of sentiments towards parties in which vote choices are made.
Furthermore, while economic attitudes correlate with populist attitudes just as clearly as cultural attitudes do, this analysis shows that the excellent performance of populists in the 2015 election was the result of voters responding to arguments about national identity and traditional values—arguments reinforced by the populist critique of liberal-democratic elites. The dominance of cultural attitudes helps to explain why populist parties were so successful in these elections in spite of Poland’s relatively good economic performance in recent years. There is a need for more thorough research of the campaign rhetoric of PiS and Kukiz to ascertain whether these parties made a concerted attempt to frame their populism solely in cultural terms, but the low salience of economic issues at the demand side suggests that this is a credible working assumption.
Finally, this analysis also draws attention to the need to explore in more detail the relationship between populist attitudes and socio-demographics. Since socio-demographic variables explain a reasonable proportion of the variance in populist attitudes, it is plausible to assume that variables such as education and place of residence mediate the impact of populism on vote choice, in addition to the direct effects they have on voting behaviour. In the Polish case, socio-demographic differences are likely to be of significant importance given the diverging experiences of economic and cultural modernisation among social groups and the perception of the divergent fates of the “winners” and “losers” of post-communist transition. One thing is clear: populist attitudes must henceforth be incorporated into models of vote choice, as they are able to tell us something distinct about how those choices are determined.
Supplemental Material
Table_4 – Supplemental material for A New Populist Divide? Correspondences of Supply and Demand in the 2015 Polish Parliamentary Elections
Supplemental material, filename for A New Populist Divide? Correspondences of Supply and Demand in the 2015 Polish Parliamentary Elections by Ben Stanley in East European Politics & Societies and Cultures
Supplemental Material
Table_5 – Supplemental material for A New Populist Divide? Correspondences of Supply and Demand in the 2015 Polish Parliamentary Elections
Supplemental material, filename for A New Populist Divide? Correspondences of Supply and Demand in the 2015 Polish Parliamentary Elections by Ben Stanley in East European Politics & Societies and Cultures
Footnotes
Notes
Author Biography
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
