Introduction

Current models of inequality and political participation assert that changes in national-level income inequality shape individuals’ traditional political participation choices based on their income level. As a theoretical mechanism linking macro-level economic changes to individual decisions, individuals’ levels of income ‘translate’ observed changes in national-level income inequality into individual participation choices. However, this mechanism faces a fundamental challenge as mounting evidence has shown clearly that individuals struggle to correctly assess the levels and changes in national-level inequality and that income is, at best, weakly correlated with individuals’ knowledge about the level of inequality in their own country (in the US: Bartels 2008; Norton and Ariely 2011; Western Europe: Alesina et al. 2004; Kaltenhaler et al. 2008; Kumlin and Svalfors 2013; Eastern Europe: Loveless and Whitefield 2011; Tverdova 2012; Binelli and Loveless 2016; Gimpelson and Treisman 2018). Such weaknesses undermine the essential assumptions of the current approach and thus challenge our understanding of how national inequality might structure political behavior.

We propose to advance our knowledge of the relationship between national-level income inequality and individual political participation choices by incorporating the observed impact of individuals’ attitudes toward inequality on a number of political outcomes (Gimpelson and Treisman 2018; Whitefield and Loveless 2013; Loveless 2016). Many of these studies show that those who feel income inequality is an important matter or that it must be addressed tend to be more supportive of popular democratic participation and/or more active themselves. However, to the extent that democratic politics are not seen as sufficiently attentive to the problems of inequality, within system remedies—e.g., voting and political engagement—are eschewed for outside system approaches such as non-traditional forms of political participation (Wegener 2000; Verwiebe and Wegener 2000; Osberg and Smeeding 2006; Filetti 2016). In other words, despite macro-economic realities and their theorized linkages to socio-economic group membership, attitudes toward and perceptions of inequality offer us potential insight into individuals’ experiences with inequality that may better explain participation choices.

To explore this alternative explanation, we use the European Values Surveys (1990–2017) in more than 40 European states to test whether individuals’ normative attitudes about inequality explain individuals’ political participation choices and whether these attitudes contribute to explain traditional and/or non-traditional choices in political participation. The results show that when individuals’ normative attitudes toward inequality are introduced to fully specified models of traditional and non-traditional political participation, not only does the interaction between income and changes in national-level income inequality essential to current theories nearly universally disappear but normative attitudes toward inequality are also significant and consistent predictors of most forms of political participation. Specifically, individuals’ negative orientations to inequality decrease all forms of traditional political participation and increase the more ‘extreme’ forms of non-traditional political participation such as boycotting and striking.

The results draw attention to individuals’ normative attitudes toward inequality as a potentially better indicator of how individuals understand and experience inequality and in turn make choices for political participation. This is a timely study in the current environment of political and ideological battles over inequality, particularly as political participation choices can also represent a frustration with democratic responsiveness. Thus, the role of inequality as an instigator to political participation may not simply (or merely) be as policy, electoral, or campaign contests between socio-economic groups under varying inequality conditions but rather provoke deeper concerns among citizens about economic fairness and ultimately democratic performance and legitimacy.

Inequality, political participation, and political engagement

Economists and political scientists have long used relative inequality models based on individuals’ socio-economic location to predict changes in aggregate and individual-level outcomes, including political participation. Several works suggest that objective, national-level inequality influences individuals’ choices of political engagement in conjunction with the level of income of the individual (Soss and Jacobs 2009; Solt 2008; Bollen and Jackman 1985; Lichbach 1989; Goodin and Dryzek 1980; Steinbrecher and Seeber 2011; Karakoc 2013; Krieckhaus et al. 2013; Filetti and Janmaat 2018). While many of these studies show that higher income inequality reduces overall participation, the role of income in countries with varying levels of income inequality have led to results pointing in opposite directions. Whether higher national-level income inequality reduces the participation gap between higher and lower income groups (Steinbrecher and Seeber 2011) or increases that gap (Filetti 2016) is not clear and remains theoretically contested.

Resource theory suggests that ‘the rich’ are able to engage more fully because of their being rich (i.e., generally more educated, higher incomes, etc.). Increasing inequality antagonizes income groups’ different participation responses resulting in higher political participation among higher income groups and lower political participation among lower income groups. Conflict theory, on the other hand, simply states that rising national levels of income inequality increases engagement across all income levels. That is, in order to engage in the debate politically, ‘the rich’ and ‘the poor’ seek to address inequality through engagement, the former in order to restrict redistributive policies and the latter to enact them (see also Brady 2004).

However, in general, the Relative Power theory does best in explaining the effect of income inequality, controlling for individual- and national-level economic and political factors (Solt 2008),Footnote 1 and provides the fundament of the current literature on the relationship between national-level income inequality and individuals’ choices of (traditional) political participation. While there are variations, nearly all current theories rely on the linkage between national-level inequality and individuals’ participation choices via their socio-economic location. Therefore,

H1: Relative Power Theory

Political engagement decreases with increases in national-level inequality, affecting the poor more than the rich, i.e., ‘relatively.’

This is an appealing and intuitive assumption as inequality tends to be more costly to ‘the poor,’ while ‘the rich’ tend to either avoid suffering or even benefit from inequality. ‘The rich’ see inequality as benign or even positive (economically and politically) for themselves, while ‘the poor’ have no choice but respond with deference to the system fatalistically. It is easy to therefore expect differences in political engagement. Yet, this literature leaves us without an efficient micro-level explanation that goes meaningfully beyond (social) conflict theory or socio-economic explanations (see Verba et al. 1979; Verba et al. 1995).

We suggest a reconsideration of the assumption that individuals’ income levels are the best means to understand what national-level inequality means to them and in turn their political choices. Are individual’s incomes the most salient dimension along which individuals consider inequality and its importance or meaning to them? Specifically, this assumption requires all individuals to be able to (1) correctly assess the level or changes in national inequality, (2) respond in accordance with their socio-economic group, and (3) ignore any intensifying or attenuating effect derived from personal experience with or normative orientations to inequality.

The first (1) has been shown to be an unsustainable assumption in the US and both Western and Eastern Europe such that individuals struggle to consistently and correctly identify levels or changes in national-level income inequality (Bartels 2008; Norton and Ariely 2011; Alesina et al. 2004; Kaltenhaler et al. 2008; Kumlin and Svalfors 2013; Loveless and Whitefield 2011; Tverdova 2012; Binelli and Loveless 2016; Gimpelson and Treisman 2018). For (2), individuals responding to changes in income inequality in accordance with their socio-economic group assume that individuals feel or evaluate objective inequality universally and uniformly within, and universally and differently between, socio-economic groups. While members of similar income groups undoubtedly share many common political, social, and economic orientations, can we assume that every member of ‘the rich’ and ‘the poor’ not only evaluate inequality accurately (ibid.) but also derive the same meaning (both economically and politically)?

Finally, we take direct issue with (3). Developments in related literature show that individuals’ attitudes toward inequality matter to their social, economic, and political attitudes and choices, regardless of actual levels of national inequality (Gimpelson and Treisman 2018; Tverdova 2012; Whitefield and Loveless 2013; Loveless 2016; see also Anderson and O’Conner 2000). This makes individuals’ attitudes toward inequality applicable to political participation choices for two reasons. One, attitudes, beliefs, and perceptions, however objectively (in)accurate, represent how individuals ‘see’ macro-economic or political performance indicators. This is not a radical notion. It corresponds to existing work is several other fields of investigation, as but one example, perceptions and beliefs play and important role in collective action and social movements studies. Much of the literature rests on a foundation of perceptions. Individuals are motivated by perceived disparity (Gurr 1970), groups respond to pressure at perceived moments of structural opportunity (Tarrow 1998), and individuals participate in groups if they believe that the group can succeed in obtaining its goals (Tilly 1978). This does not imply that the disparities, opportunities, or successes are not real, rather the internal or subjective mechanism through which individuals process the incentives and context and groups ‘see’ opportunity can be highly relevant to collective action.

Thus, we propose that individuals’ normative attitudes are strong candidates as the mechanism for political participation. There is a body of evidence that individuals’ perceptions of inequality are driven by a variety of attitudes of fairness and justice as it relates to the distribution of income and other public goods in society (for Europe, see Silagadze et al. 2023; Kaltenhaler et al. 2008; Dickes et al. 2010; Kreidl 2000; for the US: Osberg and Smeeding 2006; Bartels 2008; Kluegel et al. 1995; Norton and Ariely 2011). On this basis, prior work proposes a theoretical basis linking individuals’ normative attitudes toward inequality and their political participation choices as a means of (democratic) government to remedy economic distortions (inequality).

In examining 13 central and eastern European countries, Loveless (2016) demonstrates that citizens who see high, undesirable levels of income inequality demand popular democratic participation more than those who do not. He argues, “[n]ormatively, democracy is designed to function in a roughly egalitarian manner and is the vital institutional mechanism available for most citizens to influence other aspects of their society …. Thus, … political democracy can be perceived as an important means to contest perceived market inequalities” (emphasis mine, Loveless 2016, p. 1004). This is consistent with previous research that has demonstrated that citizens’ attitudes about inequality are tightly—and inversely—linked to their views about the performance of democracy (Jackman 1975; Bartels 2008; Whitefield and Loveless 2013; Gimpelson and Treisman 2018).

This may not be particularly novel to other literature such as redistribution preferences (Corneo and Gruner 2002; Kenworthy and McCall 2008; Rehm 2009; Finseraas 2012). From an inter-disciplinary perspective, there is widespread evidence of the misalignment of ‘really existing’ macro-economic realities and individuals’ perceptions of those same realities (Mols and Jetten 2017). This is not to provoke expectations in line with “What’s the Matter with Kansas” in which constituents display attitudes and behaviors incongruent with stated values and goals (Frank 2004). But rather, as other work has suggested, that such incongruencies facilitate separating realities in which individuals come to see the world in the way they prefer it (Norton and Ariely 2011; Gimpelson and Treisman 2018) and in turn ‘see’ their place in it (Burgoon et al. 2019; Rooduijn and Burgoon 2018; Gidron and Hall 2017; Kurer and Van Staalduinen 2022). Above all, political participation has been demonstrated to be directly related to individuals’ perceptions and misperceptions about social equality, fairness, and entitlements (Mols and Jetten 2017).

However, what is novel is the specific linkage of individuals’ normative attitudes toward inequality and their political participation choices ‘directly.’ ‘Directly’ in the sense that regardless of actual levels of income inequality in a country or an imposed process of translating that level of inequality via individuals’ socio-economic location, individuals’ normative attitudes toward inequality and their political participation choices move together most consistently and predictably. That is, individual perceptions of reality, rather than reality itself, are prime movers of many individuals’ choices and behaviors. The mechanism of this theoretical innovation is the belief that if democracy is perceived as the means to remedy perceived excessive inequality, it follows to engage in politics.

There is reason to accept this as, in states with strong democratic political institutions, citizens see these as a bulwark against market-generated inequalities (Reuveny and Li 2003; Szelenyi and Kostello 1996; Bollen and Jackman 1985). For example, Whitefield and Loveless (2012) find that while individuals’ perceptions of inequality are supportive of more forceful control of the economy, these same perceptions do not generate support for anti-democratic leadership or non-democratic institutions. This corresponds to long-standing work such that individuals who exhibit dissatisfaction with the functioning of democracy want more—rather than less—democracy (Dalton 2004; Norris 1999). However, if these political institutions fail to offer solutions, they can be perceived to be misused by public representatives for their own rather than the public’s benefit (i.e., corrupted, see Loveless 2011; Letki 2004).

Therefore, we propose a theoretical update in which individuals’ political participation choices are driven primarily by individuals’ normative attitudes toward inequality; are, at the same time, uncoordinated with objective national-level levels of inequality; and fail to reflect coordination with their socio-economic group in the context of changes in national-level levels of inequality. Each part of this theory comports with previous empirical findings in which individuals cannot consistently identify national levels of inequality, individuals’ socio-economic location weakly conditions individuals’ perceptions of income inequality, and individuals’ normative attitudes toward inequality have been correlated with demands for popular democratic action.

This proposed theoretical shift draws together these disparate strands in the literature on inequality and participation to examine the degree to which individuals’ normative attitudes toward inequality shape their political participation choices. Unlike more socio-economic approaches, our psychological approach better reflects the growing body of literature around how individuals’ political participation choices are driven by inequality. Thus, for the first time, we can determine whether ‘how people think about inequality’ is a superior and consistent predictor of political participation choices (Loveless 2016; Gimpelson and Treisman 2018).

H2:

Individuals who are more normatively averse to income inequality are more likely to engage in traditional forms of political participation than those who think inequality serves as an incentive for individual effort.

Types of political participation and inequality

Given the normative nature of these attitudes, there is potential salience to non-traditional political participation as well. Some forms of political engagement require a higher level of commitment. Traditional forms of political participation, e.g., voting and having political discussions, are relatively low cost both individually and socially. Other less common forms—such as strikes, boycotts, and demonstrations—have individual and social obstacles that must be overcome in order to engage in. Verba et al.’s (1979) typology of political participation runs from conventional to unconventional participation (and on to illegal acts) simultaneously ranking these from the easiest to perform, to the most difficult.

A number of recent studies, growing out of the far-right populist literature in Political Science and Sociology, indicate that support for non-mainstream parties is driven by socio-economic distress, whether through a lower positional status (Burgoon et al. 2019; Inglehart and Norris 2016) or their perceived relative change—specifically, declines in social status (Rooduijn and Burgoon 2018; Gidron and Hall 2017; Kurer and Van Staalduinen 2022). That is, individuals’ perceptions of their own assessment of ‘status discordance’ (Kurer and Van Staalduinen 2022) are meaningful to political choices.

Similarly, work has also shown that strong normative orientations against inequality are correlated with lower system support (Simpson and Loveless 2017; Whitefield and Loveless 2013) and lower democratic values (Loveless 2013, 2016). Particularly for more ‘extreme’ varieties, (mis)perceptions about macro-economic inequalities can as well motivate searches for a ‘strong leader’ (Sprong et al. 2019). Thus, as alternatives to traditional venues for participation grow increasingly difficult to perform—by increasing the likelihood of social sanction and exacting a higher personal costs—a strong impetus is required. Given the basis of the theory relies on the power of values, individuals’ normative attitudes toward inequality may push individuals to overcome the ‘participatory hurdle’ for such choices. This leads to the following hypothesis:

H3:

Individuals who are more normatively averse to high levels of income inequality are more likely to engage in non-traditional forms of political participation.

Methodology

We use the European Values Surveys (EVS) collected in the waves covering 1990–2017 in 42 European states for three important reasons.Footnote 2 First, we want our analysis to be both comparable to existing work as well as extend the analysis across different time periods. Second, the EVS also gives us a set of cases that vary both politically and economically but qualify as European and thus share some historical and cultural similarities. Third, several waves of the EVS have all of the questions at the individual level necessary to fully specify these models.

Dependent variables

To assess individuals’ levels of traditional and non-traditional political participation, respondents were asked a series of engagement questions. The first three correspond to existing and long-standing work on political participation:

  • Vote (1 Yes, 0 No; mean: 0.85, sd: 0.34, N = 40,098)

  • What is your level of interest in politics? (1 not at all interested, 2 not very interested, 3 somewhat interested, 4 very interested; mean: 2.54, sd: 0.92, N = 70,937)

  • How often you talk about politics with friends and neighbors? (1 Never, 2 Occasionally, 3 Frequently; mean: 1.97, sd: 0.65, N = 70,937)

While there is a slight tendency toward being ‘somewhat’ or ‘very’ interested in politics and a majority have ‘occasional’ or ‘frequent’ conversations about politics, the voting variable is problematic. More than 80% of the respondents report voting, which is incompatible with European national averages. This is likely a function of the inconsistency in questions throughout the EVS waves. The earlier rounds of the EVS (1981–2009) asks if someone will vote tomorrow in an election. The more recent round (2017) asks if they ‘usually’ or ‘always’ vote.Footnote 3 Given the limitations with the voting variable, we did not constrain the interest or discussions to the much lower vote number of observations although we did standardize the sample for political interest and political discussion (N = 70,937).Footnote 4

For non-traditional political participation, respondents were asked whether they 1 ‘would never participate,’ 2 ‘would participate,’ or 3 ‘had participated’ in

  • Signing a petition (1–3; mean: 2.17, sd: 0.82, N = 70,583)

  • Attending a demonstration (1–3; mean: 1.84, sd: 0.77, N = 70,583)

  • Joining a boycott (1–3; mean: 1.55, sd: 0.67, N = 70,583)

  • Joining a strike (1–3; mean: 1.35, sd: 0.59, N = 70,583)

One can see the declining mean level of participation from signing a petition (2.17), to attending a demonstration (1.84), to joining a boycott (1.55), to joining a strike (1.35) suggesting not only a wariness of non-traditional forms but also that these are less likely casually done and require some intent.

Independent variables

Individuals’ attitudes toward inequality are based on the response to the question: “On this card you see a number of opposite views on various issues. How would you place your views on this scale? 1 ‘incomes should be made more equal’ to 10 ‘there should be greater incentives for individual effort.’” We reversed the responses so that supporting ‘equality’ is the highest value. The distribution is indicative of normative attitudes in that it is somewhat flat, reflecting early observations that many citizens are less likely to be indifferent about the issue (Osberg and Smeeding 2006; Bartels 2008).Footnote 5

Fig. 1
figure 1

Coefficient plots

Full size image

We point out that the survey question for normative orientation to inequality asks the respondent to state whether income should be made ‘more equal.’ One might argue that this is contingent on the status of inequality in one’s country. That is, even if someone is in favor of addressing income inequality, she may answer that she is not strongly so because she feels that income equality is well-addressed in her country. Alternatively, if someone with the same views on income inequality is asked the same question in a country with very high inequality, he might feel—and thus respond—more forcefully that incomes should be made more equal. While the question used here clearly does not ask the respondent to reflect on the current situation and then apply their normative orientation, the possibility of contextual influence lingers.

Secondly, why ‘normative orientations’ rather than ‘subjective perception’? One, we know that most people in the US and Europe have surprisingly inaccurate inequality perceptions. And two, there is no good way of asking people to estimate inequality (although an unsteady consensus has been to rely on asking about change, which has its own limitations). Individuals’ normative orientations, however, are not limited by the indicators. People can tell us how they feel about inequality. Right or wrong, individuals’ feelings about inequality—as best they can articulate them—are an accurate assessment of how they feel about inequality. Thus, in replacing a social-economic explanation with a socio-psychological process, we might forfeit marginal operational clarity but in doing so gain strongly intuitive concepts and needed theoretical linkages.

Using the EVS (1990–2017), the correlation between individuals’ attitudes about inequality and (1) change in national-level income inequality is r = − 0.03 (Fig. 2 in the online Appendix) and (2) actual levels of income inequality is r = 0.00 (Fig. 3 in the online Appendix).Footnote 6 That is, we find almost no correlation between individuals’ normative orientation to inequality and the status of inequality in one’s country, corresponding to previous work (Norton and Ariely 2011; Kaltenhaler et al. 2008; Kumlin and Svalfors 2013; Loveless and Whitefield 2011; Tverdova 2012; Gimpelson and Treisman 2018).

Crucially, in order to have an independent effect on political participation choices, normative attitudes toward inequality need to be relatively independent from other variables that are likely determinants, such as income. The literature on income inequality suggests that objective, national-level income inequality influences political engagement, such that, all else equal, those in lower socio-economic locations are less likely—and even less able—to act in their own political interests than those at the top (Soss and Jacobs 2009; Solt 2008; Bollen and Jackman 1985; Lichbach 1989; Goodin and Dryzek 1980). Yet, individuals’ attitudes toward inequality are often unconditioned by individuals’ levels of income (Alesina et al. 2004; Bartels 2008; Kaltenhaler et al. 2008; Tverdova 2012; Kumlin and Svallfors 2013). However, again using the EVS data, I regressed income on normative attitudes toward inequality, controlling for individual ideological location, countries, and years (Table 3 in the online Appendix).

Income does explain normative attitudes toward inequality to some extent with a positive and statistically significant coefficient, controlling for both individual ideology and both year and country dummies. However, the income model has a R2 of 0.0125 (Model 1, Table 3, in the online Appendix) and Fig. 6 (in the online Appendix) illustrates that this relationship is weak (r = − 0.11).Footnote 7 The weak relationship between normative attitudes toward inequality and income is crucial as in order to assess the potential impact of normative attitudes toward inequality on political participation choices, and given that there is no significant variance inflation in the main model, we simultaneously include income and interact it with changes in national-level income inequality.

Table 1 Normative attitudes about inequality and traditional and non-traditional political engagement
Full size table

To measure the changes in national levels of income inequality, we use the Standardized World Income Inequality Database (Solt 2019). The Gini index used is post-tax, post-transfer household disposable income. Change in national-level income inequality is the standard in the literature as it seeks a dynamic country-level explanation for mobilizing different income groups (although we include a static income inequality indicator).Footnote 8 The interaction of income with changes in national-level income inequality, if statistically significant, would indicate differential responses of various income groups to changing levels of national inequality. In particular, evidence for the Relative Power theory would be a statistically significant and negative relationship between national income inequality and both the aggregate level of political participation and the country-specific beta coefficients of individuals’ income. This would indicate a decrease in overall political engagement with stronger effect on ‘the poor’ than ‘the rich.’

In order to strengthen our evidence, the models include controls common to standard models of political participation including other individuals’ socio-demographic variables (education, gender, and age), ideology (left/right), and marital status (see Gallego 2007; also Brady et al. 1995; Verba et al. 1979; Verba et al. 1995). We also include an indicator for democratic performance to control for variation in political participation in the aggregate, using an average of the Freedom House scores for Political Rights and Civil Liberties. All variables are described fully in the online Appendix.

The voting model is run as a single-level binomial Logit, the others are run as single-level ordinal Logits appropriate to the three- and four-category dependent variables that include year and country dummies although not shown for space. Finally, we have exploited the included weighting suggested by the EVS and clustered the standard errors by country.

Results

In Table 1 are the results of the political participation models. The table contains the average marginal effects estimated from the raw coefficients for binomial and ordered Logits.

In Models 1, 2, and 3, the interaction between income and changes in national-level income inequality does not appear. Individuals’ normative attitudes toward inequality are negatively associated with all three forms of traditional political participation.Footnote 9 That is, if one wanted to predict individuals’ choice to vote, frequency of political discussions, and level of interest in politics, the interaction of income and changes in national-level income inequality would not provide an able guide. Instead, a preference for more equal incomes is correlated with less likelihood of voting, less political discussions, and lower interest in politics. These findings are boosted by the standard socio-economic model of participation such that those with higher incomes, the more highly educated, older citizens, and the married are more likely to vote, to have higher levels of political interest, and to discuss political matters more often. We also see that those who identify as right are more likely to vote but less likely to have political discussion or be politically interested.

For non-traditional forms of political engagement, from Table 1, we can also see that individuals’ income levels are independently and positively correlated with choices in non-traditional forms of political participation. At the same time, income does not appear to respond to changes in inequality contexts (except for middle income and joining a strike). However, normative attitudes about inequality guide non-traditional political participation choices including joining a boycott and joining a strike. As above, the general robustness of these models is supported in that the youth, males, those with higher education, and those who increasingly self-identify as ideologically ‘left’ are more likely to engage in all forms of this type of participation (see Stolle and Hooghe 2011 for an excellent study in Western Europe).

In Fig. 1, the average marginal effects (AME) for the highest response category of the dependent variables can be seen for each political participation outcome.Footnote 10 For political discussion, this is ‘frequently’ (versus ‘occasionally’ and ‘never’) and for political interest, this is ‘very’ (versus ‘somewhat,’ ‘not very,’ and ‘not at all interested’). For non-traditional political participation, the highest response category is the ‘have done’ response [versus ‘might do’ and ‘would never do’]. Thus, we can see the comparative strength of income and attitudes for having participated (see Fig. 1).

Using a large, cross-national, and cross-temporal dataset, we have found little support for Hypothesis 1 (Relative Power theory). We also failed to find support for Hypothesis 2 in which those who are averse to income inequality are more likely to engage in traditional forms of political participation. The evidence suggests the opposite, that aversion makes people less likely to engage in traditional forms of political participation. However, we do find some qualified evidence for Hypothesis 3 in which an aversion to inequality motivates non-traditional forms of political participation. Here, we found evidence for this in the case of the two most ‘extreme’ forms (Verba et al. 1979), joining a boycott and joining a strike.

We note that while the interaction between income and change in national-level income inequality is effectively missing, income is not. Income is a strong, positive, independent effect as predicted in traditional models of political participation (Verba et al. 1979, 1995). The impact of normative attitudes toward inequality is clearly less powerful in terms of their average marginal effect. However, they are statistically significant and their inclusion coincides with the absence of the expected interaction between income and change in national-level income inequality.

Robustness

In terms of robustness, other estimating procedures were used, such as multi-level models. While there are some advantages to doing so, the multi-level ordinal model results were substantively the same and a great deal more complicated to discuss in the context of seven dependent variables. Yet, despite best efforts, this analysis does have limitations. Large, cross-national, cross-temporal datasets do give us a number of advantages in investigating different elements of individual-level relationships. However, one issue is the availability and cross-national, cross-temporal nature of the voting variable which limits our ability to make clear inferences within this analysis as well as comparatively. One could contend that other parts of the analysis, in particular the non-traditional forms of political participation, do provide compelling evidence for the argument posited here.

Second, surveys potentially provide less than ideal indicators of actual political participation as respondents often misrepresent their level of participation, often quite a great deal (Bernstein et al. 2001; Karp and Brockington 2005). There are unfortunately few means to solve this with these data. Finally, these cross-sectional data do not allow us to control for reverse causality. For instance, one might argue that people who are active politically could be more likely to have a specific or strongly held view on inequality in their country. While this could potentially be true for some individuals, the literature sees inequality perceptions as a reason to engage rather than a result of having been politically active (Gimpelson and Treisman 2018; Whitefield and Loveless 2013; Loveless 2016). Although, to the best of my knowledge, this secondary linkage has not been directly addressed. Again, I am unable to resolve this issue in this analysis and suggest future work should seek to examine the potential causal primacy of inequality perceptions or political activity.

Finally, although we found very little coordination in these EVS data, further attention could be given to differing inequality contexts. Individuals’ normative attitudes, mainly of two varieties: a preference for equality or inequality as an incentive, could be differently mobilizing in national contexts of low or high inequality. Of all combinations, those with a preference for equality in high inequality countries might be the most mobilized.Footnote 11 In the online Appendix, we show that none of the interactions between normative attitudes about inequality and the changes in national-level oncome inequality are statistically significant for political participation choices. This is underscored by the lack of correlation between views on inequality and levels of national-level inequality found here in the data as well as elsewhere (in online Appendix, see Figs. 2–5).

Perhaps this is not so surprising as it corresponds to earlier work. A substantial body of work suggests that individuals’ perceptions of the distributions of income and social goods in society do not necessarily correspond to the actual distribution of them (Kaltenhaler et al. 2008; Bartels 2008; Alesina et al. 2004; Loveless and Whitefield 2011; Loveless 2016). What individuals may regard as inequality may have little to do with inequality per se but depend on whether the economy as a whole provides high living standards and dynamic economic development (Lipset 1959; Jackman 1975). Where citizens are thriving or flourishing relative to their previous economic situation, individuals may disregard objective levels of inequality. The extent that states have relatively wealthy economies, or economies that have relatively high rates of growth, inequalities will be perceived as legitimate and acceptable, autonomous of objective levels of inequality, particularly in conjunction with robust democratic political institutions (Bollen and Jackman 1985). Although there are controls for income inequality levels and democratic performance, there may be more subtle relationships that have escaped detection in this analysis.

Discussion

The argument here is not meant to separate individuals’ motivations for participation into either social preferences or utility maximization. It is an attempt to better define how individuals experience and, in turn, respond politically to inequality. Unlike previous studies that impose upon individuals’ differing sets of motivations based on membership in different socio-economic locations, we find little evidence for individuals’ responses to changes in national-level income inequality constrained by their income level. Instead, we find that individuals’ normative attitudes toward inequality are a more consistent predictor of political participation choices and, interestingly, non-traditional ones. This is a timely and important finding as our failure to understand this process clouds our understanding of how inequality influences political attitudes and behavior.

Individuals’ normative attitudes about inequality can provide both insights into the individual experience of inequality and, based on their origins in beliefs about justice/fairness, a potential mechanism for political participation. The results here suggest that their inclusion is part of the explanation for why current models performed poorly for models of both traditional and non-traditional political participation. More importantly, that perceptions of inequality partially drive political participation choices underscores the notion that democracy is salient to contest perceived market inequalities (Bollen and Jackman 1985). That is, there remains promise of democratic remedies to perceived disequilibrium in the balance of economic power. This corresponds with the wealth of evidence that individuals’ perceptions of inequality are linked to orientations to fairness and social balance (Wegener 2000; Verwiebe and Wegener 2000; Osberg and Smeeding 2006; Sen 1999; Rohrschneider 2005). Even in low inequality countries, the economic outcomes are judged more harshly if they are perceived to be the product of an unfair process (Bollen and Jackman 1985; Szelenyi and Kostello 1996; Reuveny and Li 2003; Whitefield and Loveless 2013). Given that democracy is maintained through consistent and revitalizing activity, a popular belief in the efficacy of democracy to right perceived wrongs is not unwelcome.

In addition, this analysis is a response to the literature’s call to pay closer attention to the role of individual perceptions of inequality on political choices (Gimpelson and Treisman 2018; Mols and Jetten 2017). Regardless of being substantially different than actual income inequality realities, normative attitudes toward inequality may be capturing what inequality means to individuals and thus shape their subsequent choices. In other words, perhaps the core theoretical/empirical trade-off is that we cannot know exactly how individuals experience and respond to inequality, but income is a poor proxy for that experience.

A second contribution has been to identify the strong and clear linkage between these normative attitudes and individuals’ non-traditional participation choices. This is evidence of the importance—and insight—of attempting to better individualize inequality. Given that normative democratic theory is underscored by appeals to direct and active popular participation and that the constituents of this system are considered politically equal, individuals are provided a means to address the effects of the market through policy (e.g., taxation and welfare). Ideally, when inequalities are perceived to be excessive, individuals can intervene politically to seek to remedy it. The evidence here suggests that is potentially the case but has important ramifications to our thinking about political participation, inequality, and democratic legitimacy. Inasmuch as citizens do not engage within democratic institutions in the form of traditional political participation or consistently go outside these to less traditional forms, normative opposition to inequality could be potentially corrosive to long-term democratic stability and viability.

Positively, we see normative attitudes toward inequality appear to provide an impetus to overcome the ‘participatory hurdle’ for such non-traditional participation choices; yet, they are also negatively correlated with traditional forms. Thus, the direction of the effect of normative attitudes toward inequality on political participation divides the concept in a possibly more meaningful way. First of all, for both political discussions and political interest, the impact is negative. That is, feeling that inequality should be more greatly addressed moves with less discussions and lower interest. For non-traditional political participation, normative attitudes toward inequality are significant for two of the four types of participations including the more ‘extreme’ joining in both boycotts and unofficial strikes. What is notable about the order of these participation choices is that it reflects the rank from the easiest to the most difficult to perform for individuals (Verba et al. 1979).

The ideological implication is that those who are most concerned with inequality are the most likely to develop a more wide spectrum approach to political participation beyond voting. This corresponds with work in which left-wing individuals' socio-tropic economic concerns over inequality—in that case social inequality—are more likely to be engaged in both traditional and non-traditional forms of political participation (Silagadze et al. 2023). In this sense, normative attitudes toward inequality increasingly serve as motivators to non-traditional political participation, a pattern we can see in Fig. 1 in which the strength of normative attitudes toward inequality become greater (and increasingly positive) as we move toward those more unconventional participation choices. In other words, there appears to be a tipping point between normative attitudes toward inequality reducing political participation (again, discussion and interest as well as signing a petition) and becoming part of the reason to choose increasingly difficult and unconventional political acts.

A major limitation of this analysis is that we must acknowledge that citizens' perceptions about inequality—and its fairness or utility—do not emerge in a social vacuum. Individual-level studies of political participation and national inequality would benefit from greater attention to the ‘supply side’ of participation, that is, the role of parties, interest groups, and even social networks, to mitigate the relationships identified here and elsewhere (see Campante 2011). Here, the theory focuses on the individual level not for the explicit exclusion of other processes or levels but for the meaningful sake of identifying the essential relationship posited. Undoubtedly, there are a number of related processes, in particular, party/actor mass appeals, programs, campaigns, and policies as well as the ebb and flow of media attention (to name only two). Our inability to confront and control for these can be seen as an avenue to expand on this and related research.

Finally, we note that this investigation is similar to work on individuals’ views about the preferences for redistribution (see Kenworthy and McCall 2008; Corneo and Gruner 2002; Finseraas 2012; Rehm 2009). The study of such views focuses on the preferences for strong state action or allowing markets to function freely to re-balance (economic) inequities in society. Here, there is no claim about how citizens view the role of government (i.e., “government is the solution/problem”) but simply that political participation represents citizens’ need to express their concerns about this issue. The importance of individuals’ participatory responses to inequality should not limited to whether or not they are interested or advocate for redistribution. This is not even the most crucial question. Instead, we can more clearly understand the relationship between inequality and democracy. If individual forgo participation altogether, their policy preferences are moot on the basis of their non-participation. It is the demobilizing effect of inequality—or as here, its push toward unconventional participatory choices—that is more salient to our understanding of inequality’s impact on individuals’ political participation choices.

Conclusion

“We suggest that most theories about political effects of inequality need to be reframed as theories about effects of perceived inequality” (Gimpelson and Treisman 2018, p. 27). Here, we have proposed a shift from a strict socio-economic explanation to a socio-psychological approach to reflect the growing body of literature that has demonstrated that (1) individuals cannot accurately and consistently identify national levels of inequality; (2) individuals’ socio-economic location weakly conditions individuals’ perceptions of income inequality; (3) citizens see democracy as the most effective means to address inequality; (4) changes in national-level inequality have little direct effect on political participation choices; and (5) political choices are better understood by taking seriously how individuals feel (about inequality). The crucial implication is that individuals’ normative attitudes about inequality shape political participation choices because citizens see democracy as the means to address inequality. If perceived excessive inequality can best be remedied by democratic action, it makes sense to engage in politics in traditional—and non-traditional—ways.

Is this relevant to Europe today? Yes, particularly as there is a greater proportion of Europeans who both are—and see themselves as—less well off than previous generations (Bukodi et al. 2019; Buscha and Sturgis 2018). Further, recent scholarship has argued that these misalignments of perceptions and reality are not merely stimulants for inter-group hostility but also within group contests, even among the well-off (Mols and Jetten 2017). They show that perceived prosperity and success of one’s self—and of others—can create prejudicial and negative orientations not merely to ‘others’ but to members of one’s own socio-economic cohort. While again, objective economic realities appear less directly relevant to individuals’ political attitudes and choices, negative changes in economic conditions can provoke an increase in those who are—or can conceive of themselves—in increasingly precarious positions.

Yet, the implication of this study is not simply the narrow conclusion that individual-level beliefs matter, as we already know that. Instead, it allows us also to question the role of context in comparative political behavior. That is to say, in addition to sophisticated methods such as hierarchical modeling, comparativists are encouraged not to lose sight of the fact that the theoretical linkage between macro-phenomena and micro-level processes, regardless its intuitiveness, is not always direct, obvious, or uncomplicated. That is, the implication is that the mismatch between perceptions and actual levels of national-level inequality challenges core assumptions of multi-level behavioral models in which changing contexts are considered sufficient mechanisms for change in individuals’ behavior. If contexts—such as national-level income inequality—are perceived differently than what they actually are, this assumption is substantially weakened. And we need to know to what effect. The findings here are congruent with a growing body of work suggesting that individuals’ perceptions of and attitudes toward macro-economic and political phenomena are more important to their choices than reality itself.