Party Preferences as a Polarizing Factor During the COVID-19 crisis? The Evidence From Panel Survey

Abstract

This article examines the impact of party preferences on societal polarization in Czechia during the COVID-19 pandemic, particularly in response to anti-pandemic measures. Using panel survey data collected throughout the pandemic, we assess how varying levels of stringency in safety measures influenced polarization within the population. The results show that polarization was most pronounced among a small group of voters supporting a radical right populist party, whose politicization and opposition to the measures intensified over time. Notably, when populist parties were part of the governing coalition, their voters demonstrated increased compliance with anti-pandemic measures. These findings underscore the complex interaction between party loyalty, ideological positions, and political context in shaping societal responses during crises, offering new insights into the dynamics of polarization under stress.

Introduction ¹

The topic of polarization has recently gained significant attention in both scientific research and public debate (Chen et al., 2021; Gonthier and Guerra, 2023; Kluknavská et al., 2023a; Rodriguez et al., 2022; Schedler, 2023; Wang, 2014). This article aims to contribute to existing research by providing a novel analysis of how long-term political attitudes shape citizens’ responses to potentially politicized and polarizing issues. Specifically, we examine public perceptions of various anti-pandemic measures implemented during the COVID-19 pandemic, using data from a unique panel survey conducted in Czechia between 2020 and 2022.

Thematically, political polarization is typically studied from the perspective of the bipartisan system of the United States (Druckman et al., 2020; Hegland et al., 2022; Rodriguez et al., 2022; Tokita et al., 2021). Our study, however, explores how polarization functions in a multiparty system, where multiple parties must cooperate and often reach challenging compromises.

Simultaneously, research on polarization often focuses on populist political actors who use polarized debates to challenge or destabilize their opponents in government (Koch et al., 2023). Our study offers a fresh perspective by examining both a populist party in government and in opposition, an unusual dynamic in polarization research that adds complexity to the discussion. Furthermore, the study adds to research on how populist parties adapted their narratives and agendas to the new situation and how successful they were in politicizing public debate over anti-pandemic policies (see Ringe and Rennó, 2022: 290).

Many studies on pandemics, particularly surveys and large comparative studies, highlight important findings but are limited either to aggregated data or to a single point in time when the data were collected (Engler et al., 2021; Guasti, 2020; Guasti and Bílek, 2022; Heinisch and Werner, 2023; Heinzel and Liese, 2021; Jennings et al., 2021; Kübler, 2024; Lu et al., 2021; Ward et al., 2020). In contrast, our analysis spans over 23 months, from March 2020 to February 2022, allowing us to track the evolution of public opinion and examine whether and how polarization developed over time. The uniqueness and availability of the data for such a long period are one of the chief reasons underpinning our decision to select the Czech case for our analysis.

In addition, our article focuses on a relatively under-researched area, central-eastern Europe (CEE), in the context of polarization. This region, comprising mostly post-communist countries, presents an intriguing comparative opportunity due to its unique historical and political trajectory.

In multiparty systems, the implementation of anti-pandemic measures can intensify existing political divisions, as parties may have divergent views on the necessity and effectiveness of such measures (Jørgensen et al., 2021). Populist movements, which often thrive on anti-establishment sentiments, can further polarize public opinion by portraying these measures as either excessive or insufficient. The debate surrounding anti-pandemic measures evolves over time, shaped by the changing scientific advice, economic conditions, and social values. This dynamic interaction between scientific expertise, political ideologies, and public sentiment highlights the polarizing potential of anti-pandemic measures in modern societies.

For our quantitative analysis, we rely primarily on two data sources. At the aggregate level, we use the COVID-19 policy stringency index developed by Hale et al. (2021) to contextualize the pandemic and its progression. For our main analysis, we use unique panel survey data collected over nearly 2 years of the COVID-19 pandemic (Prokop et al., 2024). This individual-level data allows us to estimate the extent of polarization among supporters of different political parties and track its changes over time.

Our study reveals that stringent government safety measures significantly polarized public opinion, with compliance differing based on political affiliation. While voters of governing parties and centrist opposition were more likely to comply with anti-pandemic measures, supporters of right-wing populist parties were less compliant, especially as the issue became politicized. These findings underscore the importance of considering temporal dynamics and the specific context of multiparty systems in understanding public responses to crises, thereby contributing to broader debates on political polarization that extend beyond the focus on bipartisan systems.

Change in Party System Competition and Polarization

Increasing societal polarization is closely linked to shifts in party system competition. Even before the significant events of the past 4 years—primarily the COVID-19 pandemic and, more recently, the war in Ukraine—party systems in CEE were showing signs of growing instability and fragmentation, driven by increasingly unpredictable electoral behavior (Balík and Hloušek, 2016; Bernhard and Karakoç, 2011; Casal Bértoa, 2014; Casal Bértoa and Enyedi, 2016; Casal Bértoa et al., 2017; Haughton and Deegan-Krause, 2015, 2020; Powell and Tucker, 2014; Wessels et al., 2014). In countries like Czechia, this trend was further accelerated by the electoral success of new anti-establishment populist parties (Buštíková and Guasti, 2019; Cirhan and Kopecký, 2017; Havlík and Voda, 2016; Hloušek et al., 2020). These developments led to the gradual erosion of the traditional left-right political divide, giving way to a new dynamic in which established political parties found themselves competing against anti-establishment alternatives (Cirhan and Kopecký, 2020).

As these populist parties entered the political arena, they quickly demonstrated their willingness to exploit potentially polarizing societal issues for political gain. Their use of politicization reflects their pragmatism, as they capitalize on divisive topics to shape public discourse. Through social media, populist parties help to co-create narratives around these issues, influencing the broader societal debate. Scholars refer to this process as the creation of “polarized information ecosystems,” where such parties strategically disseminate information to heavily politicize specific topics and shape future narratives (Gollwitzer et al., 2020; Tokita et al., 2021). This allows political actors, along with other influential groups and individuals, to influence public opinion and drive the societal discussion in a direction that aligns with their interests.

Populist parties typically seek to construct narratives that resonate with particular segments of society, thereby cultivating a demand for their interpretation of events, which they can then supply. In the case of the COVID-19 pandemic, however, this strategy initially faced significant obstacles. Because the crisis was exogenous, global in scope, and affected nearly everyone’s daily life, it was difficult for populists to deploy their usual framing of a clear-cut divide between “the elites” and “the people.” Over time, however, many populist actors adapted by reframing the crisis response itself as a manifestation of elite failure, such as criticizing restrictions, contesting the role of experts, or portraying government measures as disproportionately harmful to “ordinary people.” In this way, the pandemic was gradually incorporated into a broader populist narrative of grievance and polarization (Ringe and Rennó, 2022: 5).

Political elites thus play a critical role in shaping the narrative around polarizing topics. By amplifying their reach through online networks, they forge stronger connections with their supporters. This dynamic benefits populist parties because the common stance their voters adopt on polarized issues reinforces loyalty to the party. As Druckman et al. (2020: 14) note, “As polarization increases, partisans are motivated to distinguish themselves from the other party. They can do so by following the positions endorsed by their party’s elites and rejecting those endorsed by the other party’s elites.” This highlights the pivotal role parties play in driving affective polarization by politicizing specific issues, strengthening divisions between party lines, and intensifying animosity between political groups.

Polarization of Society in Times of Crises

Polarization tends to intensify during societal crises, such as pandemics, as these events present an opportunity to measure societal divisions in response to rapidly implemented policies. Differences in compliance with government measures across various social groups help us understand this relationship. Societal polarization can undermine trust in governments, and several factors can influence it, including the stringency of measures, the evolution of the crisis, and support for specific political parties. Populist parties, especially those in opposition, play a significant role in shaping the extent of polarization (Devine et al., 2021).

In this study, the hypothesized mechanism of polarization is based on the concept of attitude activation (Dilliplane, 2014). The concept refers to the process by which latent long-term political views or beliefs become suddenly more salient as a direct consequence of facing external stimuli. In the context of a crisis, anti-elitist and anti-establishment attitudes are expected to play a particularly strong role, as the political solutions to these societal-level events require a relatively high level of trust in elites and compliance with suggested policies. It is known that economic hardship has a significant effect on anti-establishment mobilization and party system changes (see e.g. Gyöngyösi and Verner, 2022; Haughton et al., 2020). Moreover, the impact of a bad economic situation is further moderated by conservative societal attitudes (Ferrari, 2021; Voorpostel et al., 2020). This specific combination is further responsible for triggering emotional response, leading to activation of political behavior grounded in latent populist attitudes (Rhodes-Purdy et al., 2021). Recently, Kim (2025) employed an experimental research design to demonstrate that partisans can adjust their policy positions in response to evidence. The hostile political environment, however, provokes resistance to political persuasion and fuels political polarization.

Building on this literature, the proposed causal mechanism (see Figure 1) assumes that long-term political attitudes do not directly determine specific policy responses; rather, they remain latent until activated by external triggers. Importantly, such triggers need not be major crises themselves. Ideological polarization can intensify as the public debate becomes increasingly politicized or hostile, creating opportunities for latent attitudes to surface. In this process, political elites are expected to play a crucial role in shaping the agenda and framing societal events, thereby influencing whether and how these latent predispositions become activated.

Figure 1.

Causal Mechanism.

Based on the proposed mechanism, we suggest a set of hypotheses. National emergencies, such as pandemics, do not happen all at once but develop gradually, and governments’ reactions adjust accordingly (Cairney and Wellstead, 2021; Lebernegg et al., 2024; Neumayer et al., 2024). Thus, examining how party support impacts compliance with anti-pandemic rules requires attention to these temporal dynamics. Initially, many citizens rally behind their governments in times of crisis, a phenomenon known as the “rally around the flag” effect, which manifests as a short-term national unity (Plümper and Neumayer, 2022). Our first hypothesis, therefore, seeks to explore how responses to government measures vary over time as the pandemic evolves.

• H1. Citizens’ response to the measures adopted by the government will differ over time, depending on the epidemic situation/COVID-19 waves.

The different phases of the pandemic provide insights into how polarization develops, as political elites’ responses are tied to the unfolding crisis. As the crisis deepens and measures become more restrictive, public attitudes tend to shift, with initial support often giving way to increased skepticism and criticism (Louwerse et al., 2021). Kritzinger et al. (2021: 1222) found that the rally around the flag effect is typically short-lived, fading as the immediate threat diminishes. The presented analysis is capable not only of modeling where such an effect begins and ends (Hegewald and Schraff, 2024; Johansson et al., 2021; Kritzinger et al., 2021), but also of answering the question of what drives subsequent polarization over the topic.

In the early stages of the pandemic, citizens were more likely to support government measures, regardless of their political preferences (Kritzinger et al., 2021). However, this initial approval tends to erode over time. As subsequent COVID-19 waves emerged, skepticism and non-compliance increased, contributing to the polarization of society (Altiparmakis et al., 2021). This growing polarization often politicizes previously non-political issues, such as pandemic responses (Bavel et al., 2020; Ward et al., 2020).

• H2. More stringent measures will have a stronger polarizing effect.

Several studies highlight the role of party support in shaping polarization during societal crises, particularly in relation to compliance with government measures (Altiparmakis et al., 2021; Hegland et al., 2022; Jørgensen et al., 2021; Kritzinger et al., 2021). Polarization often deepens when political elites take radical positions, as individuals tend to align their views with those of their preferred leaders (Heinisch and Werner, 2023). Ideological differences further influence this process, with supporters of right-wing parties often prioritizing economic freedoms, while supporters of left-wing parties focus on protecting vulnerable populations (Toshkov et al., 2022). This ideological divide shapes how various groups perceive and respond to government policies, fueling polarization (Jørgensen et al., 2021).

Trust in government institutions and experts also plays a crucial role in this dynamic (Premat et al., 2024). Supporters of far-right or anti-establishment parties tend to distrust experts more than those aligned with mainstream parties, contributing to cycles of polarization (Heinzel and Liese, 2021). This distrust, coupled with the consumption of alternative media, intensifies divisions, especially in societies already experiencing political fragmentation (Altiparmakis et al., 2021). Consequently, we expect that voters of populist parties will be less compliant with anti-pandemic measures, reflecting their skepticism toward the establishment (for a detailed overview of the literature on populism during the COVID-19 pandemic, see Zulianello and Guasti, 2023).

• H3a. Being a voter of a populist party will have a negative impact on compliance with the anti-pandemic measures.

Research by Jørgensen et al. (2021) suggests that polarization is more influenced by whether individuals support government or opposition parties than by populist affiliations or left-right ideologies. The distinction between government and opposition voters often supersedes traditional ideological divides in shaping responses to crisis measures. Government supporters tend to moderate polarization, as they are more likely to support measures implemented by the ruling parties. This theory posits that compliance is divided between supporters of government and opposition parties (Rovny et al., 2022).

For example, even restrictive policies like lockdowns do not necessarily alienate right-leaning citizens if their parties are in government (Jørgensen et al., 2021). Government supporters are typically 20% more likely to approve of such measures compared to opposition voters (Jørgensen et al., 2021: 1141). This partisan alignment reflects a broader trend where supporters of ruling parties positively evaluate government policies during crises, while opposition voters express skepticism (Altiparmakis et al., 2021). Studies from France and Austria similarly show that opposition party voters are less trusting and supportive of government measures (Kritzinger et al., 2021).

Trust in government is also a significant factor. Higher levels of trust correlate with more favorable evaluations of pandemic responses in both health and economic domains (Altiparmakis et al., 2021). In the context of Czechia, where the ANO party led the government, this trend is notable. The region’s experience with populist parties in government during the pandemic provides a unique case to test this hypothesis.

• H3b. Being a voter of the government party will have a positive impact on compliance with the anti-pandemic measures.

Data and Methods

Case Selection

The theoretical framework above highlights several societal factors influencing public reactions to the pandemic measures and the potential for polarization. Key variables identified include the stringency of government policies, the evolving epidemic situation, and the party support. In particular, the prior research emphasizes the role of populist party supporters and the moderating impact of whether a party is in government or opposition. Accordingly, populist parties tend to adjust their communication strategies when they move from opposition to government office (Swanson and Gherghina, 2023).

Czechia presents a highly relevant case for this study for several reasons. First, it allows for an analysis of how a populist party in government (ANO) and a populist party in opposition (Freedom and Direct Democracy, SPD) influence polarization levels. Second, as a CEE country, Czechia offers an understudied context for polarization studies, especially regarding the COVID-19 pandemic.

Simultaneously, SPD and ANO may be perceived as increasingly more similar populist parties, making these two parties highly comparable. Existing research on ANO demonstrates how the party pragmatically abandoned its initial technocratic managerial style of politics (Cirhan, 2024; Havlík and Voda, 2016) and was replaced over time by a nativist approach, embodied for instance by anti-immigrant rhetoric (Naxera, 2024; Stulík and Naxera, 2021). Similar anti-immigrant stance, together with the emphasis on direct democracy, such as referendums, defined the staple of SPD’s political communication, as is the case for other populist parties elsewhere (Gherghina and Pilet, 2021). Regarding the demand side similarity, the regression analysis comparing electorates of both parties shows a relatively low level of differences (see Appendix A6).

The Czech party system has undergone significant changes over the past decade. Prior to the COVID-19 pandemic and the war in Ukraine, Czech politics was already marked by growing instability and fragmentation. Historically, the country had a stable party system, with clear ideological divides between the liberal conservatism of Civic Democrats (ODS) and the welfare-state-focused Social Democrats (ČSSD) (Berglund and Dellenbrant, 1991; Hanley, 2008). These two parties formed the main ideological spectrum, supported by smaller actors such as the Christian Democrats (KDU-ČSL) and the Communists (KSČM) (Cirhan and Kopecký, 2020). However, this status quo began to shift after the 2010 “electoral earthquake” when new anti-establishment parties broke through, leading to the fragmentation of the Czech party system (Cirhan and Kopecký, 2024; Hájek, 2018; Hloušek et al., 2020; Kopeček, 2016; Šárovec, 2019; Stauber, 2019). This trend has continued with each subsequent election. New political forces emerged, such as the populist ANO 2011 (ANO), Freedom and Direct Democracy (SPD), and new liberal alternatives like Mayors and Independents (STAN) and the Pirate Party (Maškarinec, 2017). Voters of these parties play a crucial role in this study, as their responses to government measures during the pandemic are central to understanding the dynamics of polarization.

Regarding the timeline of the pandemic in Czechia, it can be divided into three main phases. The first phase, referred to as solidarity and mobilization, was marked by collective action and adherence to government policies (see Buštíková and Baboš, 2020: 500–501; Guasti, 2020). The second phase involved a relaxation of restrictions during the summer of 2020, which gave way to the third phase, characterized by politicization and polarization. This latter stage was marked by growing distrust and the rise of an anti-vaccine narrative promoted by far-right political parties and public initiatives (Guasti and Bílek, 2022; Hartikainen, 2021; Havlík and Kluknavská, 2022; Kluknavská et al., 2023b).

Finally, Czechia is not only theoretically relevant but also regionally informative. Similar patterns of populist governance and crisis management have been observed in other CEE countries, such as Poland, Slovakia, and Hungary (Buštíková and Baboš, 2020; Stanley, 2022). The Czech case thus provides a valuable lens to understand the broader dynamics of populist governance and public opinion during crises in multiparty CEE systems, while also allowing for cautious inference about patterns that may generalize to comparable contexts.

Data Description

The core data source for our study comes from a unique panel survey Life during Pandemics (Život během pandemie) of attitudes toward the pandemic of COVID-19 (Prokop et al., 2024). This collection of panel data in Czechia was gathered between March 2020 and February 2022 in 38 waves during the COVID-19 pandemic using the CAWI method. The panel study we use for our analysis is based on an initial sample of 2025 respondents (for sample sizes across research waves and descriptive statistics, see Appendices A1 and A2). The sample is representative of the adult population of the Czech Republic. ²

The data for contextualization of the development of the pandemic situation in Czechia are computed using a comparative dataset by Hale et al. (2021) consisting of stringency policies’ indicators and information on new COVID-19 cases by day. For our analysis, we aggregate the observations on a monthly level.

Estimation Strategy and Operationalization

Regarding the stringency of measures, we employ three indicators, each focusing on a specific measure with increasing severity of impact on citizens’ lives: washing hands, wearing a facemask, and finally, the intention to vaccinate. Concerning the impact of one’s party preference on one’s likelihood of ideological polarization, we analyze the relationship between party support and the willingness to comply with the adopted measures, with a focus on populist and governmental parties.

First, we focus on the question of the extent to which people react to the changing pandemic situation. Following the first hypothesis, we expect that a higher number of COVID-19 cases will lead to a change in behavior regarding willingness to follow different anti-pandemic measures. To analyze the correlation between two time-varying variables, we apply a Granger Causality statistical test (Rivieccio et al., 2021).

Next, to answer the question of how the adoption of different anti-pandemic measures influences political polarization, using repeated generalized linear regression models, we examine the change of explained variability between models, including simply for socio-demographic variables (age in years, gender, education (4 categories), and economic status (7 categories)) as predictors and models controlling for political party respondents voted in the parliamentary election preceding the pandemics (2017). In this design, the higher the difference in classification performance (log-loss), the more polarizing the issue or policy is for different groups of voters. To capture the expected non-linear trend over time, we use a generalized additive model (GAM) summarizing the development of political polarization.

The log-loss function is a measure of goodness of fit in the logistic regression, quantifying the discrepancy between predicted probabilities and the actual class labels in binary classification. It calculates the average difference between the actual labels and the predicted probabilities, penalizing incorrect predictions more heavily, especially those that are confidently incorrect. The log-loss is sensitive to the quality of the predicted probabilities, providing a nuanced measure of model performance.

To operationalize the polarizing effect of political attitudes, we estimate two logistic regression models. Model 1 includes predictors such as age, gender, economic status, and education. Model 2 extends Model 1 by incorporating the party voted for in the 2017 election as an additional predictor. The polarizing effect of the restrictive measure is then quantified by comparing the log-loss values of the two models. Specifically, we calculate the difference in the improvement of the fit. This difference captures the extent to which including the party voted for in 2017 enhances the model’s predictive power, thereby indicating the measure’s polarizing impact (see equations below).

• Model 1:

$$\begin{array}{l}Y={\beta }_0+{\beta }_{age}*age+{\beta }_{gender}*gender\\ +{\beta }_{economicstatus}*economicstatus+{\beta }_{education}*education\end{array}$$

• Model 2:

$$\begin{array}{l}Y={\beta }_0+{\beta }_{age}*age+{\beta }_{gender}*gender+{\beta }_{economicstatus}*economicstatus\\ +{\beta }_{education}*education+{\beta }_{partyvoted2017}*partyvoted\in 2017\end{array}$$

• Polarizing Effect of the Restrictive Measure:

$$\Delta logloss=\left(1-loglossModel2\right)-\left(1-loglossModel1\right)$$

Dependent Variables

To test our hypotheses, we use a set of three dependent variables that measure respondents’ willingness to comply with anti-pandemic measures of varying stringency levels. First, the respondents were asked if they were washing their hands more often to prevent infection. Second, the dependent variable operationalizes respondents’ willingness to wear facemasks/respirators to protect themselves or the surroundings from disease. Third, to measure the level of agreement with vaccination, our operationalization combines two responses to two questions: (a) If a vaccine against COVID-19 is available, would you be willing to get vaccinated? (b) Are you vaccinated yet? This particular approach was applied because the availability of vaccines varied over time, from a hypothetical option (before the vaccine had even been invented and distributed), through partial access limited to certain parts of the population (e.g. medical personnel or elderly people), to unlimited access in later stages of the pandemic situation.

All the reported models and analyses were repeated for the three dependent variables. This serves two purposes, that is, (a) as a robustness check and (b) as an indicator of the variability of effect size for the independent variables under analysis.

A potential concern in this study is endogeneity, which arises when unobserved factors influence both the independent and dependent variables, potentially biasing the estimated relationships. For example, individuals’ underlying political attitudes, trust in institutions, or personal risk perceptions could simultaneously affect both their partisan alignment and their compliance with pandemic measures. Unfortunately, these variables are not explicitly present in the survey data and could be addressed only indirectly.

Reverse causality is another possibility, as attitudes toward the specific policies may reinforce partisan loyalties over time. To avoid this issue, we mainly work with the 2017 electoral choice. On the other hand, one could argue that voters are not loyal to political parties. To investigate this, we analyzed vote transfers between elections 2017 and 2021 (see Appendix A8). It should be noted that most of the observed volatility can be attributed to transfers within ideological blocs.

While our analytical approach, including panel data and temporal ordering of variables, helps mitigate some of these concerns, we acknowledge that residual confounding may remain. Consequently, the results should be interpreted as indicative of associations rather than definitive causal effects.

Independent Variables

Main variables of interest

• H1. Monthly aggregated mean of daily new cases of COVID-19 based on Hale et al. (2021).

• H2-H3b. Political party respondents voted in 2017.

Control variables used in regression models are age, gender, economic status, and the highest level of education achieved. All variables are taken from wave 1, which was surveyed on 03/2020. For a detailed overview of descriptive statistics, refer to Appendices A1 and A2.

Descriptive Evidence

In the initial phase of the pandemic, starting in March 2020, Czech society displayed a strong sense of solidarity and mobilization. The government’s swift response was widely supported, with strict measures being met with the public’s compliance. During this period, the emphasis was on a collective responsibility, protecting public health, and ensuring the healthcare system was not overwhelmed. Public initiatives, such as groups and institutions spontaneously sewing and distributing facemasks, symbolized this collective effort. The period from March to June 2020, which marked the first wave of the pandemic, may have generated overly optimistic expectations about future developments (see Figure 2).

Figure 2.

Descriptive Analysis of Trends in Compliance With Policies.

During the summer of 2020, with the decline in COVID-19 cases, there was a growing sense that the worst of the pandemic had passed. Restrictions were relaxed, and the public began returning to normalcy, enjoying the summer months with a sense of relief. However, this temporary reprieve was short-lived, as by the fall of 2020, cases began to rise once again, marking the beginning of the pandemic’s third phase.

The Granger causality test analysis shows no significant effect between hand hygiene practices and new COVID-19 cases (see Appendix A3). This suggests that respondents do not adjust their hand hygiene in reaction to the pandemic situation. Similarly, the number of new cases does not predict the adoption of wearing facemasks, nor does it influence vaccination rates. Because various mutations of the virus with very different mortality levels (e.g. the Omicron variant with high incidence but low mortality) occurred over the time of the pandemic, we repeated the analysis with the number of deaths as an independent variable. The results remained unchanged. These findings imply that compliance with anti-pandemic measures is not primarily driven by the epidemiological situation. Instead, we propose that political polarization is a more likely explanation for the observed behaviors. The evolving political debate and party affiliation appear to play a crucial role in shaping individuals’ responses to the pandemic rather than the immediate number of COVID-19 cases.

With the rise in cases in the fall of 2020, as the pandemic situation worsened, the public debate became increasingly politicized and polarized. The government’s response was heavily criticized, and there was a growing sense that the measures being taken were too extreme and unnecessary. Despite the government gradually adopting some of the most restrictive measures, which included a complete ban on free movement from one’s district of residence, the country was among those most severely hit by the virus. This is evident statistically as the Czech case ranked 9th in the worldwide comparison in regards to the number of deaths by COVID-19 per capita (over 40,000 deaths for a population of 10 million). The stringency of adopted measures not only heavily politicized the pandemic-related debate but also contributed to a rise in anti-vaccination sentiment, which further fueled the polarization in society. Some politicians and media outlets began to promote anti-vaccination views, contributing to the erosion of public trust in the government’s response. The Czech Republic clearly followed the same trend as many other governments in the region, taking the wrong lessons from the initial success during the first months of the pandemic.

Partisan Polarizing Effect in the Context of Pandemic Restrictions

To estimate the overall polarizing effect of pandemic restrictions, we fitted a pair of generalized linear models for every wave of the panel survey where the respective questions were available. Figure 3 shows the improvement in model fit when the variable of interest, that is, the political party voted in the previous general election, is added. This helps us to understand that the political measures and their stringency stem from polarization within Czech society. As stated previously, the level of stringency is captured by a comparison of three anti-pandemic policies imposed by the Czech government in the analyzed period, namely suggestions to increase the frequency of washing hands, wearing facemasks, and vaccination.

Figure 3.

Polarization Over Time Across Policies With Different Levels of Stringency.

Our analysis commences with the mildest measure, wherein authorities emphasized the importance of handwashing among citizens. Figure 3 below provides two crucial insights. Initially, the influence of party preferences appears relatively limited, indicating consistent adherence to this advice across diverse societal groups. However, a distinct pattern emerges, revealing heightened politicization and polarization following periods of both eased and stricter measures in 2020. Notably, supporters of the SPD began challenging the government’s stance, irrespective of the severity of the measures implemented.

Shifting the focus to the willingness to wear face masks, a parallel trend emerged during the notably polarized discourse, primarily from 2021 onward. Contrasting with handwashing habits, the inclusion of political preferences in the analysis reveals even more heightened polarization following an initial phase of widespread agreement. Once again, a statistically significant impact is observed from prior support for the right-wing populist party SPD.

Moving to the most recent measure under analysis, it’s important to acknowledge that the data we’ve examined only covers the polarized phase of the public debate since 2021. While this limitation restricts our ability to compare with less controversial anti-COVID measures, what stands out is the significant contribution of party preference in explaining the variability in willingness to receive the vaccine.

Equally noteworthy is the persistent high level of polarization observed throughout the period analyzed. This enduring polarization can be attributed mainly to the intensified public debate surrounding the opportunity to get vaccinated, prompting individuals to respond based on the pre-existing stances they had formed earlier.

To further illustrate the effect size, we fitted a set of time fixed effects models across the three dependent variables. In addition, we performed analyses on reduced samples, starting in wave 20, to specifically observe whether the increasing politicization during this period led to heightened polarization. Figure 4 visualizes the predicted probabilities of compliance with pandemic measures across different voter groups. This visualization helps to better understand how political preferences influenced the adherence to public health guidelines over time.

Figure 4.

Predicted Probabilities for Voters’ Groups (Time-FE models).

Moderating Effect of Government Responsibility

The Czech case provides a unique opportunity to estimate how being part of the government can influence and moderate the positions of populist party supporters. Our analysis focuses on comparing the effect sizes of non-compliance with different anti-pandemic measures among SPD voters, a right-wing populist opposition party, and supporters of ANO, the main governing but also populist party.

The results reveal a striking pattern that closely mirrors our previous analysis (see Figure 5 below). Most of the observed polarizing effects can be attributed to a relatively small group of voters who support opposition populist parties like the SPD. This finding underscores the significant role that opposition status plays in shaping the attitudes of populist party supporters. Interestingly, the compliance levels of these voters vary substantially over time, with the most notable differences emerging during periods of heightened political debate. This suggests that the politicized environment, rather than the severity of the measures themselves, plays a critical role in influencing voter behavior.

Figure 5.

Compliance with Policies (SPD vs ANO).

Further analysis reveals that the effect of party affiliation on compliance is not static but fluctuates in response to the political context. During the initial phase of the pandemic, when public solidarity was high and the debate was less politicized, SPD voters exhibited compliance levels similar to those of ANO supporters and voters of moderate establishment parties. However, as the pandemic progressed and the public debate became increasingly politicized, the compliance gap widened, particularly among SPD supporters. This indicates that political polarization intensifies as the debate becomes more contentious, with opposition populist parties driving much of this division.

Discussion

Summarizing the presented findings, our analysis of panel survey data reveals a steady increase in polarization over time, closely correlated with the growing stringency of implemented measures. Notably, this polarization is concentrated among the voters aligned with right-wing populist parties, particularly the SPD, which consistently exhibits a statistically significant impact across regression models.

Although our analysis is based on 2017 party support, it’s noteworthy how distinct ideological voter groups responded to the pandemic with predictable patterns tied to evolving threat perceptions. Initially, the debate surrounding pandemic measures was characterized by an unusual level of consensus, even among populist party supporters. However, as time progressed, polarization intensified, driven by the evolving and often unpredictable nature of political decisions regarding the severity of measures. It seems that the availability of vaccines, together with the then ongoing electoral campaign prior to the 2021 general elections, may serve as two chief explanatory factors for such increased politicization. This trend persisted regardless of the strictness of measures, but the degree of polarization varied depending on specific measures, such as mask-wearing or vaccination.

In essence, our comprehensive analysis reveals a steady escalation of polarization in response to the changing dynamics of pandemic-related measures, underscoring the different impacts of specific measures in augmenting this polarization. These insights offer critical implications for comprehending and addressing the complexities of contemporary political polarization in times of crisis.

Interestingly, our analysis also highlights the distinct role of ANO, a populist party in government. Governance responsibilities appear to moderate the positions of ANO supporters, leading to higher compliance rates compared to SPD supporters. This suggests that governance can moderate populist rhetoric and influence voter behavior toward a greater adherence to public health measures.

Our findings lead us to conclude that governmental status is a crucial factor in shaping voters’ attitudes toward polarizing policies. The responsibility of governance generates a moderating influence on populist parties and their supporters, potentially reducing polarization. This has important implications for understanding the dynamics of political behavior and compliance in crisis situations, emphasizing the need to consider the interplay between political affiliation, government status, and the evolving nature of public debate.

Conclusion

This study aimed to answer a critical question: What determines the variability in citizens’ responses to safety policies during crises? Using the COVID-19 pandemic as a case study, we demonstrated how public opinion in Czechia was not only mobilized but also polarized in the response to government-imposed safety measures. Understanding these dynamics is crucial for effectively managing public health crises and ensuring compliance with safety policies across different segments of society.

Our findings indicate that while the hypothesis H1 predicting that citizens’ responses would vary over time depending on the epidemic situation did not hold, other hypotheses yielded more conclusive results. Specifically, H2 was supported, showing that more stringent measures had a stronger polarizing effect on public opinion. Regarding the impact of political affiliation, our results were mixed: H3a, which proposed that voters of populist parties would be less compliant, was only partially supported, primarily among the right-wing populist voters, and contingent on the politicization of the pandemic issue over specific periods. Conversely, H3b, which posited that the voters of government parties would show greater compliance, was supported by the data.

This research contributes to the existing body of knowledge in several ways. First, it highlights the temporal dynamics of public opinion during crises, often overlooked in studies limited to cross-sectional data or single time points. Second, our findings emphasize the role of party politics in shaping responses to safety policies, underscoring the need for more nuanced analyses in multiparty systems. Finally, unlike much of the literature focused on bipartisan contexts like the United States, our study demonstrates the complexity of polarization in a CEE context, where multiple parties and emerging actors continually adjust their positions.

However, there are limits to the generalizability of our findings. The specific political and social context of Czechia, characterized by a multiparty system and the emergence of new political actors, may not fully apply to other countries, particularly those with more stable party systems or different institutional arrangements. On the other hand, some of our findings resonate with broader patterns observed in the CEE region. For instance, in Poland, the Law and Justice (PiS) party faced significant challenges managing the COVID-19 pandemic as an exogenous crisis, which disrupted its typical populist strategy of “crisis entrepreneurship” and narrative control (Stanley, 2022). Similarly, in Slovakia, populist parties in government had to balance public health measures with their partisan agendas, often struggling to maintain credibility among their supporters when strict safety measures were imposed (e.g. Buštíková and Baboš, 2020). These examples illustrate that while the specific political and social context of Czechia is unique, the broader difficulty populist governments face when confronted with unforeseen crises may be a generalizable phenomenon in the region, particularly in multiparty systems where ideological polarization intersects with crisis management.

Looking ahead, future research should explore the polarization dynamics observed during the COVID-19 pandemic in the context of other crises, such as the ongoing conflict in Ukraine. Investigating how different types of crises, whether public health emergencies or geopolitical conflicts, affect public opinion and political polarization could provide valuable insights into the resilience and adaptability of democratic institutions in times of crisis.

Supplemental Material

sj-docx-1-psw-10.1177_14789299251410829 – Supplemental material for Party Preferences as a Polarizing Factor During the COVID-19 crisis? The Evidence From Panel Survey

Supplemental material, sj-docx-1-psw-10.1177_14789299251410829 for Party Preferences as a Polarizing Factor During the COVID-19 crisis? The Evidence From Panel Survey by Tomáš Cirhan and Jakub Stauber in Political Studies Review