This is an open access article distributed under the terms of the Creative Commons Attribution License (Voting is the cornerstone of democracy. By electing official representatives, one has the possibility to take part in important collective decisions, in the formation of legislative bodies, and in shaping policies that will affect healthcare, public transportation, and many other issues of importance for the community or the country. Yet, in recent decades there has been a trend of declining voter turnout in many democratic countries (e.g., Solijonov, 2016). Given the importance of voting in democratic societies, understanding the factors that may encourage individuals to participate in elections has become a pressing concern. In the present research, we focus on national identification as an explanation for the motivation to vote. Patriotic appeals have become a frequent rhetorical feature of many electoral campaigns around the world. In 2016, for example, Donald Trump promised to ‘make America great again’, making patriotism a central feature of his speeches. Similarly in 2020, Vladimir Putin said, ‘We are voting for the country we want to live in… and which we want to hand down to our children’, while standing beneath a statue of a Soviet soldier. Consequently, one may assume that the evocation of national identity is a successful tool to motivate people to vote, but studies on the links between national identification and voting behaviour have produced mixed results (e.g., Rupar, Sekerdej, & Jamróz-Dolińska, 2021; Urbanska et al., 2020). The reason for these inconsistencies may lie in the aspects of national identity that people have in mind when they think of their relation to their own country (Huddy & Khatib, 2007). Thus, the right question is not whether national identification is linked to voting behaviour, but rather which of its forms may be the most relevant to consider when motivating people to vote. To answer this question, we examined the link between three forms of national identification, namely constructive patriotism, glorification, and conventional patriotism, and intentions to vote and actual voting behaviour in three types of elections (i.e., state parliamentary, EU parliamentary, and presidential elections).
National Identification and Voting
It is well established that national identification is a multidimensional concept. It may come in the form of pure love of one’s country (i.e., conventional patriotism, Sekerdej & Roccas, 2016; see also Bar-Tal & Staub, 1997; Kosterman & Feshbach, 1989) or love of one’s country accompanied by a belief in the superiority of one’s country or nation (i.e., glorification, Roccas et al., 2006; this definition may also refer to blind patriotism, see Schatz et al., 1999; Staub, 1997, or nationalism, see Kosterman & Feshbach, 1989). Apart from these two traditional forms of national identification, another relevant form of national identification has been recognized in the literature – constructive patriotism (Schatz et al., 1999; Sekerdej & Roccas, 2016). As with conventional patriotism, constructive patriotism involves a positive attachment to the country, yet in comparison to glorification it does not impose a belief that the country is flawless. On the contrary, it is accompanied by criticism of national practices as motivated by a desire for positive change (Schatz et al., 1999; Sekerdej & Roccas, 2016).
Participation in civic activities is traditionally viewed as central to national identity and embodies what is perceived as the normative behaviour of being a ‘good’ citizen (Conover et al., 2004; Huddy & Khatib, 2007; Schildkraut, 2007). Thus, it is not surprising that the existing literature shows that individuals with higher levels of patriotic sentiments are more likely to engage in activities on behalf of their country or compatriots (e.g., military service, Levi, 1997; Fordham, 2016 or military mobilization in general, Reicher & Hopkins, 2001), or to behave in a manner that is civically responsible (e.g., paying taxes, Gangl et al., 2015). Moreover, recent studies show that different forms of national identification are particularly relevant in understanding the role of national identification in citizens’ civic engagement (Rupar, Sekerdej, & Jamróz-Dolińska, 2021) or group practices that may benefit the country’s citizens (Rupar, Jamróz-Dolińska, et al., 2021). Since voting is considered to be one of the most traditional forms of civic engagement, one would expect it to be positively linked with national identification. Yet, studies on the relationship between national identification and voting have yielded mixed results. One of the reasons for this may lie in conceptualizations of national identity that take into consideration only one aspect of it, such as attachment to the country (Urbanska et al., 2020), belief in national superiority (Shook et al., 2020), or national pride (Rooij et al., 2012). Yet, even studies that considered different forms of national identification and voting behaviour in the same study (e.g., Huddy & Khatib, 2007; Rupar, Sekerdej, & Jamróz-Dolińska, 2021) did not provide consistent results. The reason for these studies’ inconsistent findings may be due to methodological issues, such as the fact that participants were asked to recall past voting behaviour. Moreover, these studies focused only on one country (i.e., Poland or the United States), and their results cannot be generalized to other contexts. In the present research, we address these limitations by examining intentions to vote in upcoming elections, as well as actual voting behaviour in those elections across different European Union (EU) countries.
Election results have consequences for the formation and implementation of policy and consequently lead to various long-term changes on different societal levels (e.g., health, education). By voting, citizens act as agents of these changes; thus, engagement in voting should be related to people’s motivation to bring about changes in the country. While all three forms of patriotism (constructive patriotism, conventional patriotism, and glorification) share a common aspect, namely positive identification with and feelings of affective attachment to the country (Schatz et al., 1999), their main difference lies in their relation to the country and, specifically, actions related to the well-being of the country and nation. For example, conventional patriotism, in comparison to other forms of national identification, assumes only love for and attachment to the country, which may not be enough to motivate individuals to engage in actions that benefit the country and nation (Rupar, Jamróz-Dolińska, et al., 2021; Sekerdej & Roccas, 2016). Again, in contrast to both constructive and conventional patriotism, glorification implies idealized beliefs about the nation. Individuals who highly glorify their nation may believe that significant changes in the country are unnecessary or undesirable; as a result, they may resist or oppose activities that promote change. Moreover, due to the nature of the comparison with other nationalities that is inherent in glorification, individuals who highly glorify their nation state may be more inclined to prioritize interactions and perceptions related to other groups. This focus on intergroup relations may obscure internal reflections on the nation’s own challenges, shortcomings, or areas for improvement. In comparison to conventional patriotism and glorification, constructive patriotism is more instrumental in its nature and emphasizes the belief that individual efforts should be made to actively contribute to and benefit the nation’s goals (Rupar, Jamróz-Dolińska, et al., 2021; Schatz, 2018). Moreover, people high on constructive patriotism are willing to criticize the country’s current state (e.g., in terms of education, the economy), which might motivate them to engage in political actions (Sekerdej et al., 2023). Conversely, both conventional patriotism and glorification are characterized by a tendency to overlook or justify the country’s problems (Sekerdej & Roccas, 2016; Sekerdej et al., 2023). Consequently, the inability to see the nation’s flaws may obstruct individuals’ motivation to engage in acts that can bring about change, such as voting. Yet, importantly, even if one believes that changes are not necessary and wishes for election results that would solidify the status quo, one still needs to invest time and effort to vote and make it happen. The only form of national identification characterized by the motivation to invest time and effort in order to bring about positive changes in the country is constructive patriotism (Sekerdej & Roccas, 2016). Thus, we argue that constructive patriotism will be the only form of national identification linked with greater intentions to vote and with actual voting behaviour.
The link between national identification and voting can also be looked at through the prism of traditional voter turnout theories, such as rational and informational ones. Rational theories suggest that turnout decisions are influenced by individuals weighing the costs and benefits of voting (Downs, 1957). Specifically, the decision whether to vote or not results from the relative benefits of having one party winning over the other, multiplied by the probability of casting the deciding vote and subtracting the costs of voting. According to rational theories, the probability of casting the deciding vote is relatively small, therefore voting is not the usual choice of utility-oriented citizens. However, individuals high on constructive but not conventional patriotism or glorification are especially prone to undertaking actions that require dedicating their resources and exertion to contribute to the betterment of their country (Rupar, Jamróz-Dolińska, et al., 2021; Rupar, Sekerdej, & Jamróz-Dolińska, 2021; Sekerdej & Roccas, 2016). On the other hand, individuals high on glorification act in a selfish, utility-oriented way and tend to support actions that enhance personal welfare. Therefore, it could be assumed that compared to individuals high on glorification, constructive patriots’ actions may not be led by selfish motives. They can see that voting is an additional effort that is required to benefit the country. Moreover, according to information theory, a person is more likely to vote when they become more confident about the voting procedure, and this confidence is increased by the scope of information about the candidates and the elections (Matsusaka, 1995). On the other hand, uninformed people are more likely to abstain from voting (Degan & Merlo, 2004; Feddersen & Pesendorfer, 1996, 1997). Individuals high on constructive patriotism but not glorification are characterized by higher levels of political information gathering, therefore they have greater political knowledge (Parker et al., 2009; Schatz et al., 1999) and are more likely to engage in voting. Indeed, constructive patriotism has been linked to greater endorsement of civic commitment and practices (Rothi et al., 2005). Also, constructive patriotism is underpinned by care for the country and the capability of the nation’s institutions to provide benefits to citizens, both of which are related to political knowledge as well as political participation (Schatz & Lavine, 2007). That said, both rational theory and information theory support our idea regarding the positive link between constructive patriotism and voting.
Current Research
Our research adds to the existing literature on the links between national identification and voting behaviour in four ways. First, we acknowledge the multidimensional nature of national identification and simultaneously consider its three forms as differing in content and implications: conventional patriotism, constructive patriotism, and glorification. Second, we measure voting intentions in upcoming elections as well as actual voting behaviour in these elections. Third, we investigate various types of elections: state parliamentary elections, presidential elections, as well as EU parliamentary elections. Fourth, we conducted this research within the context of four different EU countries: Poland, Spain, Croatia, and the Czech Republic.
We would also like to emphasize the importance of studying not only actual voting behaviour but also intentions to vote. In the light of conditional choice theory, which aims to understand election participation (Rolfe, 2012), the decision to vote is not individualistic but depends on the attitudes of co-citizens. Individuals who have intentions to vote may express their ideas and attitudes about these intentions, a process that may exert an influence on other individuals’ actual voting behaviour. Similarly, one’s voting behaviour may depend on knowledge of whether others took part in the voting process in the past. In other words, this process may contribute to the creation of social norms regarding voting behaviour. Consequently, such norms may convey group standards concerning expected practices and may contribute to the sense that one should vote. Indeed, beliefs regarding the frequency of others’ voting have been found to be related to people’s intentions to vote (Gerber & Rogers, 2009; Glynn et al., 2009). Moreover, even a small set of people who feel strongly about voting in elections may contribute to changes in the turnout rate of elections (de Matos & Barros, 2004). Thus, investigating both intentions to vote and actual voting behaviour is important in order to better understand electoral turnout.
In Study 1, we focus on state parliamentary elections in Poland and Spain, while in Study 2 we focus on presidential elections in Croatia. All these elections were held at roughly the same time, hence the potential influence of various external factors (e.g., sudden political changes, unexpected international or local incidents) was minimized. In Study 3, we focused on EU parliamentary elections in Poland, the Czech Republic, and Croatia. We decided to focus on different types of elections because voter turnout is higher for some elections (e.g., national parliament) than for others (e.g., European, regional, or local parliaments, Marsh, 1998; Reif & Schmitt, 1980). One of the reasons for this difference may be the greater importance that society assigns to some types of elections over others. Given that individuals with high constructive patriotism have broad awareness of what can contribute to the country (Sekerdej & Roccas, 2016, Study 3) and thus could consider all elections to be highly important, we assumed that constructive patriotism would be linked to voting intentions and actual voting regardless of the type of election.
Research has shown that identification with Europe is positively related to intentions to vote in EU elections (e.g., Mazzoni et al., 2018). Thus, we chose to conduct research in three countries whose length of EU membership differs. While Poland and the Czech Republic have been EU member states for almost two decades, Croatia is the youngest EU member state. Moreover, in order to provide more insight into the role of national identification, in Study 3 we also included a measure of European identity.
Some additional factors may be linked to voting behaviour. Younger people and the elderly are less likely to vote (e.g., Blais, 2000; Wolfinger & Rosenstone, 1980). Women are more likely than men to cast a vote, yet this may be true only for second-order elections (e.g., European or local parliament elections, Kostelka et al., 2019). Educated citizens vote more than uneducated citizens (e.g., Lewis-Beck et al., 2008). Finally, people hold different political views that may not be in line with the current political system in the country, which could motivate them to vote with the aim of change. Also, recent literature has suggested that political ideology may underlie current operationalizations of patriotism and nationalism (Hanson & O’Dwyer, 2019). Thus, we additionally ran all our models whilst also accounting for age, gender, political views (all studies), and education (only in Study 1).
In all studies, we planned to recruit at least 250 participants, given that correlations tend to stabilize with a sample size of approximately that number (Schönbrodt & Perugini, 2013). In all studies, we excluded participants who did not identify as majority members of respective countries and were less than 18 years old. We also removed participants in the second and third waves of the studies in case we could not match the answers of these participants with their answers in the previous waves. If a given participant filled out the questionnaire twice, only the first entered data was kept. Some of the studies were still running after the elections. If the questionnaire was completed on or after election day, participants were removed from the study. We performed all analyses using the statistical software R and the statistical package Lavaan (Rosseel, 2012). All data and materials are available as Supplementary Materials (see Rupar et al., 2021S, 2024S). The questionnaire was designed in the Polish language and translated to Spanish, Croatian, and Czech with the help of native speakers.
Study 1
In Study 1, we examined the link between constructive patriotism, conventional patriotism, and glorification and intentions to vote in the state Parliamentary elections in Poland and Spain that took place on October 13th and November 10th, 2019, respectively. The study was conducted approximately one week before the election date. In Poland, we also examined actual voting behaviour by conducting a second survey one week after the elections took place.
Method
Participants and Procedure
In the first survey, when testing intentions to vote in the upcoming elections, the sample comprised 1,160 Poles (68.4% female, Mage = 21.33, SD = 3.67) and 110 Spaniards (54.5% female, Mage = 34.73, SD = 8.38), all of whom completed an online questionnaire. Participants filled out informed consent forms before data collection in this and all subsequent studies. In the second survey, when testing actual voting behaviour in Poland, the sample consisted of 523 Polish participants (30% male, Mage = 21.74, SD = 4.14). In Poland, we recruited participants by posting links on various Facebook students’ groups for universities across Poland. In Spain, we additionally recruited participants by the snowball sampling method.
Measures
Scales ranged from 1 (not at all) to 6 (very much), unless otherwise indicated.
Constructive Patriotism
We measured constructive patriotism with five items (e.g., ‘I oppose some Polish/Spanish policies because I care about my country and want to improve it’ and ‘I express my attachment to Poland/Spain by supporting efforts to achieve positive change’, adapted from Schatz et al., 1999). Cronbach’s α = .77 for Poles and α = .86 for Spaniards.
Conventional Patriotism
We used five items to measure conventional patriotism (e.g., ‘I feel attached to Poland/Spain’ and ‘I love my country’, adapted from Sekerdej & Roccas, 2016 and Roccas et al., 2006). Cronbach’s α = .87 for Poles and α = .86 for Spaniards.
Glorification
To measure glorification, we used eight items (e.g., ‘Criticizing your nation is a sign of disloyalty’ and ‘Other nations can learn a lot from us’, adapted from Roccas et al., 2006). Cronbach’s α = .87 for Poles and α = .86 for Spaniards.
Voting Behaviour
We measured intentions to vote with one item: ‘Do you plan to vote in the upcoming elections for the state Parliament?’ Answers were given on a scale from 1 (definitely not) to 7 (definitely yes). We measured actual voting behaviour in Poland one week after the elections took place by asking participants whether they had voted or not in the recent elections for the state Parliament. There were 454 participants who had voted and 69 who had not.
Results
The means and correlations of the measured variables for both groups are displayed in Table 1.1
Table 1
Means, Standard Deviations, and Correlations of Constructive Patriotism, Conventional Patriotism, Glorification, Intentions to Vote in State Parliamentary Elections in Spain and Poland (Study 1)
| Variable | 1 | 2 | 3 | 4 | M | SD |
|---|---|---|---|---|---|---|
| 1. Constructive patriotism | – | .48** | .12** | .27** | 4.83 | 0.74 |
| 2. Conventional patriotism | .61** | – | .62** | .10** | 4.24 | 1.05 |
| 3. Glorification | .40** | .78** | – | -.07* | 2.75 | 0.86 |
| 4. Intentions to vote | .11 | -.05 | -.12 | – | 5.34 | 1.12 |
| M | 4.40 | 3.54 | 2.45 | 5.37 | ||
| SD | 1.18 | 1.48 | 0.96 | 1.34 |
Note. Results for Poland are presented above the diagonal, and for Spain below the diagonal.
*p < .05. **p < .01.
We tested the links between constructive patriotism, glorification, and conventional patriotism and intentions to vote in the Polish and Spanish samples using multigroup structural equation modelling. We treated all forms of national identification as latent variables. Intentions to vote were treated as a manifest variable because they were measured with one item. In order to allow for comparisons between groups, we first investigated measurement invariance for the scales of national identification in the Polish and Spanish samples. This is done by sequentially equating parameters between countries. If the model fit is not significantly worsened by introducing new constraints on parameters, a more restricted model is preferred (Ferrer et al., 2008; Milfont & Fischer, 2010; Vandenberg & Lance, 2000). Then, configural equivalence is contrasted with metric invariance, and metric invariance is contrasted with scalar invariance. We aimed to establish support for configural and metric invariance, which allows the structural paths in the structural model between groups to be compared (Byrne et al., 1989; Davidov et al., 2018). To establish measurement invariance, we compared configural and metric invariance in the two samples using the thresholds of ΔCFI < −.010; ΔRMSEA < .015; and ΔSRMR < .030 (Byrne & Stewart 2006; Chen, 2007). The basic configural model, which included testing of the invariance of the overall structure of national identification, had an acceptable fit, χ2(258) = 1066.54, p < .001, CFI = .924, RMSEA = .070, 90% CI [0.066, 0.075], SRMR = .062 (see Hu & Bentler, 1999). The metric model, which included restricting of factor loadings to be equal across the two samples, yielded a good model fit, χ2(273) = 1112.23, p < .001, CFI = .921, RMSEA = .070, 90% CI [0.065, 0.074], SRMR = .068. Importantly, the metric model did not deviate from the configural model according to all indicators, ΔCFI = -.003, ΔRMSEA = .000, ΔSRMR = .006, thus supporting the notion that the structural paths in the Polish and Spanish samples could be compared.
We then defined and compared two models: one in which all paths were freely estimated, and another in which all paths were constrained. The results of the chi‐square difference test were not significant, meaning that the regression paths were invariant for both groups, Δχ2 = 5.98, df = 3, p = .113. The model with constructive patriotism, conventional patriotism, and glorification as predictors predicted intentions to vote in the parliamentary elections (R2 = .12; for factor loadings of the national identification items, in this and subsequent studies, see Table 1 in the Supplementary Materials). Constructive patriotism positively and glorification negatively predicted intentions to vote in state parliamentary elections: B = .55, SE = .09, p < .001 and B = -.18, SE = .09, p = .045, respectively. Conventional patriotism was not related to intentions to vote, B = -.03, SE = .08, p = .666.
In this and subsequent studies, we also ran the models while controlling for age, gender, education and political orientation. We also measured political orientation with one item: ‘On a scale from 1 to 7, where 1 means left-oriented and 7 right-oriented, please indicate your political orientation’. Entering these covariates into the models did not change the association between the different forms of national identification and intentions to vote or actual voting behavior in any of the studies. Overall, across studies, the internal meta-analysis showed that the link between political orientation and all forms of national identification was positive (random effects model; constructive patriotism: r = .18, 95% CI [.03, .32]; conventional patriotism: r = .42, 95% CI [.30, .53]; glorification: r = .46, 95% CI [.29, .59], and the link between political orientation and intentions to vote was negative (r = -.08, 95% CI [-.12, -.04]; for details on the results, see Supplementary Materials). These results suggest that individuals with more conservative or right-wing views have lower intentions to vote in elections. However, having such beliefs does not influence the link between national identification and voting intentions.
Finally, to test the links between different forms of national identification and actual voting behavior in the Polish sample, we employed structural equation modelling with a DWLS estimator. Again, we treated all forms of national identification as latent variables. Voting behavior was treated as a manifest variable because it was measured with one dichotomous item. The model2 with constructive patriotism, conventional patriotism, and glorification as predictors significantly predicted the likelihood of voting in state parliamentary elections in Poland, χ2(3) = 32.295, p < .001; R2 = .19. Constructive patriotism was linked to greater likelihood of voting in Parliamentary elections in Poland, B = .52, SE = .16, p = .001. Conventional patriotism and glorification were not associated with the likelihood of voting in Parliamentary elections in Poland: B = -.07, SE = .14, p = .617, B = -.11, SE = .18, p = .565, respectively.
Overall, these results support the idea that constructive but not conventional patriotism is positively related to intentions to vote and to actual voting behavior in state parliamentary elections. However, surprisingly, glorification was negatively linked with intentions to vote, but not with actual voting behavior.
Study 2
In Study 2, we investigated the link between constructive patriotism, conventional patriotism, and glorification, and intentions to vote in the Croatian presidential election that was held on December 22nd, 2019. As no candidate received a majority of votes, a second round of elections took place on January 5th, 2020. The study was conducted approximately one week before the first round of elections took place. We conducted the second survey one week after the first round of elections were held and asked the same participants about their actual behavior in the first round of these elections and their intentions to vote in the second round. Finally, we conducted a third survey asking participants about their actual voting behavior in the second round.
Method
Participants and Procedure
In the first survey, when testing intention to vote in the upcoming election, the sample comprised 640 Croats (76.4% female, Mage = 23.47, SD = 4.90), out of which 186 took part in the second survey, where we tested actual voting behavior in the first round of the election (72.6% female, Mage = 23.59, SD = 4.72). In the third survey, when testing the actual voting behaviour in the second round of the election, the sample consisted of 119 Croat participants (73.9% female, Mage = 23.88, SD = 5.42). The recruitment procedure followed the one in Study 1.
Measures
All measures were the same as in Study 1, apart from the fact that the measure of conventional patriotism did not include the ‘I like Croatia’ item due to translation issues, and voting questions referred to the presidential elections (‘Do you plan to vote in the upcoming presidential elections/the second round of presidential elections?’ and ‘Did you vote in the recent presidential elections/second round of presidential elections?’). The reliability of all scales was acceptable: constructive patriotism, α = .73, conventional patriotism, α = .83, and glorification, α = .81. There were 149 participants who voted and 37 who did not vote in the first round of presidential elections; also, 91 participants voted and 28 did not vote in the second round of presidential elections.
Results
The means and correlations of the variables are presented in Table 2.
Table 2
Means, Standard Deviations, and Correlations of Constructive Patriotism, Conventional Patriotism, Glorification, and Intentions to Vote in the First and Second State Parliamentary Elections in Croatia (Study 2)
| Variable | M | SD | 1 | 2 | 3 | 4 | 5 |
|---|---|---|---|---|---|---|---|
| 1. Constructive patriotism | 5.04 | 0.69 | – | ||||
| 2. Conventional patriotism | 4.31 | 1.07 | .39** | – | |||
| 3. Glorification | 2.54 | 0.78 | .18** | .64** | – | ||
| 4. Intentions to vote – first round | 5.05 | 1.27 | .21** | .08* | .03 | – | |
| 5. Intentions to vote – second round | 4.66 | 1.69 | .10 | -.06 | .01 | .45** | – |
*p < .05. **p < .01.
First, to examine the links between different forms of national identification and intentions to vote in the first and second rounds of presidential elections, we made two separate structural equation models. We treated all forms of national identification as latent variables. Intentions to vote were treated as a manifest variable because they were measured with one item. The models that tested the links between national identification and intentions to vote in the first and second round of presidential elections yielded good model fits: χ2(128) = 443.70, p < .001, CFI = .920, RMSEA = .062 [.056, .068], SRMR = .054 and χ2(128) = 244.18, p < .001, CFI = .911, RMSEA = .070 [.056, .083], SRMR = .070, respectively; R2 = .06 and R2 = 03. Constructive patriotism predicted intentions to vote in the first round but not the second round of presidential elections in Croatia: B = .59, SE = .14, p < .001; B = .44, SE = .26, p = .099, respectively. Glorification had no role in explaining intentions to vote in the first or second round of presidential elections: B = -.02, SE = .16, p = .896; B = .39, SE = .38, p = .30, respectively. In the same vein, conventional patriotism did not predict intentions to vote in either the first or second round of presidential elections: B = -.01, SE = .10, p = .922; B = -.39, SE = .24, p = .137, respectively.
Second, as in Study 1, to examine the link between the three forms of national identification and actual voting in the first and second rounds of presidential elections, we conducted two separate path analyses, thus treating both national identification and voting behavior as manifest variables due to the small number of participants who took part in the second and third waves. The model did not significantly predict the likelihood of voting in the first or the second round of presidential elections in Croatia: χ2(3) = 3.739, p = .290; χ2(3) = 0.760, p = .859, respectively.
Overall, Study 2 found that constructive patriotism was the only form of national identification that predicted intentions to vote in presidential elections in Croatia, yet this was only evident in the first round of presidential elections. In contrast to Study 1, constructive patriotism did not predict the likelihood of voting in presidential elections.
Study 3
In Study 3, we investigated the link between, on one hand, constructive patriotism, glorification, conventional patriotism, and European identity, and, on the other hand, intentions to vote in the elections for the European Parliament in three countries (Poland, the Czech Republic, and Croatia) that were held between 23rd and 26th May 2019. The survey was administered about one week before the elections took place. In Poland and Croatia, we also examined actual voting behavior by conducting a second survey one week after the elections took place.
Method
Participants and Procedure
In the first survey, where we examined intentions to vote in the upcoming elections, the samples comprised 522 Polish (77.6% female, Mage = 23.52, SD = 6.48), 135 Czech (84.4% female, Mage = 27.29, SD = 8.18), and 581 Croatian participants (78.7% female, Mage = 24.09, SD = 6.24). In the second survey, where we tested actual voting behavior, the sample consisted of 183 Polish participants (78.67% female, Mage = 23.26, SD = 6.21) and 163 Croatian participants (81% female, Mage = 24.40, SD = 6.92). Participants were recruited in the same way as in the two anterior studies.
Measures
All measures were the same as in Study 2, apart from the voting behavior measures that referred to the upcoming European elections and the measure of European identity: ‘I feel attached to Europe’, ‘I feel European’, and ‘Being European is important to me’. The reliability was acceptable in all scales across the three samples: constructive patriotism, α = .76 in the Polish sample, α = .75 in the Croatian sample, and α = .71 in the Czech sample; conventional patriotism, α = .86 in the Polish sample, α = .84 in the Croatian sample, and α = .81 in the Czech sample; glorification, α = .88 in the Polish sample, α = .84 in the Croatian sample, and α = .84 in the Czech sample; European identity, α = .90 in the Polish sample, α = .84 in the Croatian sample, and α = .87 in the Czech sample. There were 135 participants who voted and 48 who did not vote in the EU elections in the Polish sample, and 79 participants who voted and 84 who did not vote in the Croatian sample.
Results
The means and correlations of the variables are presented in Tables 3 and 4.
Table 3
Means, Standard Deviations, and Correlations of Constructive Patriotism, Conventional Patriotism, Glorification, European Identity, and Intentions to Vote in EU Parliamentary Elections in Poland and Croatia (Study 3)
| Variable | 1 | 2 | 3 | 4 | 5 | M | SD |
|---|---|---|---|---|---|---|---|
| 1. Constructive patriotism | – | .41** | .16** | .36** | .29** | 4.78 | 0.72 |
| 2. Conventional patriotism | .47** | – | .64** | .14** | .09* | 4.18 | 1.13 |
| 3. Glorification | .25** | .66** | – | -.07 | .02 | 2.71 | 0.92 |
| 4. European identity | .19** | .16** | .26** | – | .24** | 4.39 | 1.10 |
| 5. Intentions to vote | .14** | .09* | .08 | .11** | – | 4.83 | 1.42 |
| M | 5.00 | 4.17 | 2.41 | 3.81 | 4.29 | ||
| SD | 0.70 | 1.11 | 0.82 | 1.15 | 1.54 |
Note. Results for Poland are presented above the diagonal, and for Croatia below the diagonal.
*p < .05. **p < .01.
Table 4
Means, Standard Deviations, and Correlations of Constructive Patriotism, Conventional Patriotism, Glorification, European Identity, and Intentions to Vote in EU Parliamentary Elections in the Czech Republic (Study 3)
| Variable | M | SD | 1 | 2 | 3 | 4 | 5 |
|---|---|---|---|---|---|---|---|
| 1. Constructive patriotism | 4.54 | 0.65 | – | ||||
| 2. Conventional patriotism | 4.06 | 1.02 | .45** | – | |||
| 3. Glorification | 2.78 | 0.80 | .22* | .60* | – | ||
| 4. European identity | 4.59 | 1.09 | .28** | .14 | -.02 | – | |
| 5. Intentions to vote | 4.68 | 1.73 | .17* | .10 | -.05 | .19** | – |
*p < .05. **p < .01.
We examined the relationship between national identification and intentions to vote in all three samples together, using multigroup structural equation modeling. We treated all forms of national identification as latent variables. We treated intentions to vote as a manifest variable because they were measured with one item. First, we investigated measurement invariance for the measures of national identification. The basic configural model, which included testing of the invariance of the overall structure of national identification, had an acceptable fit, χ2(480) = 1209.96, p < .001, CFI = .936, RMSEA = .061 [.056, .065], SRMR = .058 (see Hu & Bentler, 1999). The metric model, which included restricting factor loadings to be equal across two samples, had an acceptable fit, χ2(512) = 1348.02, p < .001, CFI = .927, RMSEA = .063 [0.059, 0.067], SRMR = .069, and did not deviate from the configural model according to two out of free indicators, ΔCFI = −.011, ΔRMSEA = .011, ΔSRMR = .006, thus allowing for the comparison of structural paths in the structural model between groups (Byrne et al., 1989; Davidov et al., 2018).
Next, we estimated and compared two models: one in which all paths between variables were freely estimated, and another in which all paths were constrained. The results of the chi‐square difference test were not significant, that is, the regression paths were invariant for both groups, Δχ2 = 7.054, df = 8, p = .531; R2 = .06. Constructive patriotism and European identity positively predicted intentions to vote in European parliamentary elections, B = .68, SE = .13, p < .001 and B = .16, SE = .05, p = .001, respectively. Conventional patriotism and glorification played no role in explaining intentions to vote, B = -.07, SE = .09, p = .418; B = -.01, SE = .11, p = .903, respectively.
Finally, to test for links between the three forms of patriotism and actual voting in EU parliamentary elections in the Polish and Croatian samples, we conducted a multigroup path analysis, thus treating all forms of national identification and actual voting as manifest variables because multigroup analysis based on structural equation modelling was not appropriate due to the small sample sizes. We compared the model in which the regression paths between the constructs were freely estimated for each group against the model where the paths across groups were constrained. The chi‐square difference test was not significant, indicating that the regression paths were invariant for both groups, Δχ2 = 2.197, df = 4, p = .700. However, the model only marginally predicted the likelihood of voting, χ2(4) = 9.139, p = .058; R2 = .05. An inspection of coefficients showed that constructive patriotism, but not other forms of national identification, was linked to the likelihood of voting in EU Parliamentary elections (constructive patriotism, B = .26, SE = .11, p = .015; conventional patriotism, B = .06, SE = .09, p = .531; glorification, B = -.002, SE = .11, p = .224; European identity, B = -.08, SE = .07, p = .224).
In conclusion, the findings of Study 3 supported the evidence from the other two studies in this paper by showing that constructive patriotism is the only form of national identification that predicts intentions to vote. However, it only marginally predicted the actual likelihood of voting.
General Discussion
In the present research, we examined the relevance of citizens’ national identities in relation to their intentions to vote and their actual voting behavior in different types of elections across various EU countries. We found consistent support for the idea that constructive patriotism is the only form of national identification (compared to conventional patriotism and glorification) that is positively linked to intentions to vote. The results regarding actual voting behavior were not consistent across the studies.
Patriotism is often emphasized in public speeches and media campaigns with the aim of motivating people to vote. Country leaders often appeal to love for the country, but also to the uniqueness and often even the superiority of this country. However, even though one might expect that love of the country could be related to such traditional civic activities as voting, our results suggest otherwise, showing that only one form of patriotism may be relevant, namely constructive patriotism. Only individuals high in constructive patriotism, but not in conventional patriotism or glorification, were more likely to show intentions to vote in upcoming elections. We confirmed these results across three different types of elections: state parliamentary elections in Spain and Poland, Presidential elections in Croatia, as well as EU parliamentary elections in Poland, the Czech Republic, and Croatia. Moreover, even in the case of EU parliamentary elections, the effect of constructive patriotism on intentions to vote was beyond and above that suggested by levels of EU identification. These results extend previous research by showing that individuals high in constructive patriotism, but not in conventional patriotism or glorification, have a particular tendency to engage in activities that may bring about changes in society (Rupar, Jamróz-Dolińska, et al., 2021; Sekerdej et al., 2023). These findings suggest that love of one’s country or belief in its superiority may not be enough to motivate people to vote. It is necessary to emphasize care for the country and a willingness to work for its benefit (e.g., to invest time in voting) in order to promote the good of the country. Despite the optimistic results concerning the role of national identification in general and of constructive patriotism specifically in explaining intentions to vote, with respect to actual voting behavior we confirmed these findings only in the case of state parliamentary elections. One of the reasons for this inconsistency in the results may be the small sample sizes that were caused by the high dropout rate of participants in the second and third waves. Another explanation could be the small number of participants who did not vote. Due to such limitations, we cannot draw final conclusions. Intentions to vote are nevertheless important, given their potential to contribute to the creation of social norms regarding voting (e.g., Gerber & Rogers, 2009; Glynn et al., 2009).
The main limitation of this research is its correlational nature. Indeed, some previous research has suggested that the link between national identity and political engagement can be looked at from both sides (e.g., Richey, 2011). This issue should be addressed in future research by employing an experimental or longitudinal design. We would favor a longitudinal design, given that the levels of constructive and conventional patriotism are relatively high, and manipulating these constructs could be difficult. Future studies should investigate potential emotional and cognitive mechanisms that may drive, facilitate, or hinder the positive association between constructive patriotism and intentions to vote (e.g., belief in change, feelings of individual responsibility, sense of agency, perception of the current state in the country). One should also look into structural factors that may shape this link, such as the cultural and democratic backgrounds of specific countries. Another limitation relates to the age of the participants. Specifically, our sample is skewed and includes mostly young people. National identification could play a different role in voting behavior in younger people than in older people. Thus, in order to address this issue, our findings would need to be replicated on a more representative sample. Also, some of the samples were small in size, such as the Spanish sample in Study 1 and the Czech sample in Study 3. Also, compared to others, the data collection method in these countries included snowball sampling. Thus, results from these contexts should be taken with caution. Finally, the limitations of this research include the relatively small variance explained by the model for the main criterion variables, suggesting that other unaccounted-for factors may have contributed significantly to the observed outcomes. While the model provides valuable insights, caution should be exercised in generalizing the findings, as the modest explanatory power may indicate complexities not captured within the current analytical framework.
Overall, in the present paper, we showed that to understand the role of identity in shaping voting behavior, one needs to consider different conceptualizations of national identification. Specifically, our results from three types of elections and four different EU countries suggest that thinking about the country in terms of constructive patriotism, namely noticing the flaws of the country but simultaneously manifesting a readiness to address them, may help to motivate citizens’ intentions to vote. These results have wide implications for encouraging voting in a variety of social settings. For example, programs for promoting voter participation could be carefully crafted to appeal to citizens’ constructive patriotism, which could encourage them to critique the status quo and actively seek better solutions for their country rather than merely praising the existing conditions. Similar to this, political parties and candidates could use patriotic rhetoric based on constructive patriotism in their communications. Specifically, they could highlight the importance of critical analysis and the pursuit of improved solutions for the nation, emphasize the benefits of voting for the betterment of the country, and underscore the importance of defending democratic principles through active citizen participation. Finally, educational programs could be conducted to encourage constructive patriotic virtues and highlight how they relate to civic engagement, particularly voting. These initiatives could provide valuable knowledge about the benefits of voting, the impact of citizen participation on democratic processes, and the role of national identity in fostering a sense of responsibility towards the nation. A culture of active citizenship may be created by carefully blending these strategies while encouraging constructive patriotism.