Concerns about inequality across domains

Literature on subjective economic inequality has mainly focused on perceptions and concerns about income gaps or wealth distributions (^{Castillo et al., 2022}; ^{Willis et al., 2022}). This approach, however, involves several theoretical and methodological limitations (^{García-Castro et al., 2019}; ^{García-Sánchez et al., 2022}). For instance, people may find it difficult to understand numeric and abstract representations of monetary resources, they could lack information about economic issues, and cognitive biases might influence their estimations (^{Castillo et al., 2022}; ^{Pedersen & Mutz, 2019}). Furthermore, recent theoretical proposals and empirical research have shown that perceived inequality is a multidimensional phenomenon that encompasses various domains of people’s everyday lives (^{García-Castro et al., 2021}; ^{García-Sánchez et al., 2022}; ^{London School of Economics and Political Science (LSE), 2018}). Therefore, the view of economic inequality solely based on the distribution of wealth and income excludes other information relevant to the way people perceive their society. For instance, it does not consider differences in access to health services or education opportunities.

Importantly, from a multidimensional angle, concerns about inequality may vary across domains because different distributive justice principles may apply for each of them (^{Jasso et al., 2016}; ^{Starmans et al., 2017}). For instance, people can justify income inequality by thinking that their socioeconomic position is the result of hard work (^{García-Sánchez et al., 2020}). However, this belief could be less relevant to a person suffering from cancer who needs an unaffordable treatment. In this regard, while merit is a relevant aspect for distributing economic resources, health and education can be considered human rights (^{United Nations, 1948}, art. 25 and 26), and therefore, might be less dependent on deservingness.

To our knowledge, very few studies to date have compared people’s concerns about economic inequality in different domains. As an exception, ^{Macchia and Ariely (2021)} asked participants to imagine that they were moving to a different country and that their place within the income distribution would be randomly assigned. Next, they had to indicate how they would distribute wealth and good educational and good health resources, across income quintiles. Results showed that people accepted more inequality with respect to wealth than in the domains of health or education, in which they desired an almost egalitarian distribution. Consistently, ^{Howarth et al. (2019)} showed that the number of people preferring egalitarian sharing was much lower in the wealth domain than in the health sphere (e.g., 5% vs. 46%). Although this evidence is preliminary, it points out that some domains of economic inequality might be less accepted than others. Furthermore, researchers must investigate whether concerns about inequality, beyond monetary or financial resources, can be associated with participation in collective actions.

Concerns About Inequality and Collective Actions

Whether economic inequality can foster collective actions is still an open question. From a theoretical perspective, social grievances-such as the experience of large economic gaps-are at the heart of protests (^{Jetten et al., 2021}; ^{van Stekelenburg & Klandermans, 2013}; ^{van Zomeren et al., 2008}). In fact, in an analysis of worldwide protests, ^{Ortiz et al. (2020)} observed that the lack of economic justice was the main reason that motivated people to protest. However, other empirical research has found that indicators of economic inequality at a macrolevel (e.g., Gini index) are marginally associated with collective actions to reduce it (^{Jo & Choi, 2019}), or they may even be negatively associated (^{Dubrow et al., 2008}; ^{Solt, 2015}).

Concerns about economic inequality, rather than objective indicators, might be a better predictor of people’s responses to inequality (^{Jo & Choi, 2019}; ^{Willis et al., 2022}). Although research on the relationship between concerns about economic inequality and support of collective actions is surprisingly scarce, some studies show a positive association (^{Jo & Choi, 2019}). For instance, the belief that the government should reduce income inequality between the rich and the poor was related to a greater participation in collective actions in Chile (^{Castillo et al., 2012}) and Latin America (^{Justino & Martorano, 2016}). Uniformly, using cross-national data from 45 countries, ^{Jo and Choi (2019)} showed that perceived income inequality and preferences for redistribution were positively associated with involvement in collective actions. Nevertheless, none of these studies have investigated whether concerns about economic inequality across various domains (e.g., health or education) could play a differential role in support of collective actions to reduce it.

Indirect evidence suggests that exploring the effect of concerns about economic inequality in domains such as health and education (beyond income disparities) on collective actions is a promising direction. For instance, ^{Ortiz et al. (2020)} showed that 12% of protests worldwide denounced inequalities in income and wealth, but another 17% of protests were driven by reforms that threaten the quality and quantity of public services, such as education and health. From this perspective, recent outstanding social movements have had education inequalities at their cores, such as in the cases of Chile (^{Huenupi, 2021}), Colombia, and Brazil (^{Nava & Grigera, 2022}). In addition, protests against health disparities have increased, becoming much more visible after the onset of the COVID-19 pandemic (^{Daniels, 2021}; ^{Kim et al., 2020}; ^{Rodríguez-Bailón, 2020}). Thus, it seems that considering concerns across various domains -instead of just considering wealth or income inequality- could better explain participation in collective actions.

The present research

In this research, we addressed several limitations of previous literature. Although economic inequality permeates several social spheres, such as health or education, most of the studies in this field have focused exclusively on how people perceive and react to income or wealth disparities (e.g., salary gaps; ^{Castillo et al., 2022}). Therefore, concerns about economic inequality in other domains have been largely unexplored. Another important gap in the literature is that evidence regarding the relationship between concerns about economic inequality and support of collective actions is surprisingly scarce (^{Jo & Choi, 2019}). Specifically, to our knowledge, no one had tested whether concerns about economic inequality in health and education would predict support of collective actions.

We established two main objectives. First, we analysed concerns about economic inequality in health, education, and income. We predicted that people would be more concerned about education opportunities or health access than income inequality (H1). Moreover, we identified people’s profiles based on their concerns about health access, education opportunities, and income inequality. Second, we examined the relationship between concerns about inequality and support of collective actions to reduce inequality. We expected that concerns about health access, education opportunities, and income inequality would be associated with greater support of collective actions to reduce inequality (H2). In addition, we explored whether people’s profiles based on their concerns about inequality could predict support of collective actions to reduce inequality.

Furthermore, we used data from Latinobarometer 2020, which provides a unique perspective in the study of inequality. Latin America has one of the highest levels of income inequality in the world (^{UNU-WIDER, 2022}), despite the reduction of inequality that the region has experienced in recent decades (^{Justino & Martorano, 2016}). This pattern, however, may have been reversed by the COVID-19 pandemic, which has hit the most vulnerable social groups harder (^{Blofield et al., 2020}). Moreover, Latin America was one of the regions with the largest incidence of protests between 2006-2020 (^{Ortiz et al. 2020}), and this tendency continues to date (^{Díaz Pabón & Palacio Ludeña, 2020}). Therefore, the social unrest in Latin America might not solely be driven by changes in objective economic indicators, but could rather be associated with a combination of subjective mechanisms, such as concerns about inequality.

Method

Results

Concerns about economic inequality across domains

In line with H1, analyzing differences between proportions revealed that, in general, people worried more about education opportunities (χ² = 2264.7, p < .001, CI 95% (.22-.23)) and health access (χ² = 3009.6, p < .001, CI 95% (.26-.28)) than income inequality. A significant share of people mentioned education opportunities (43.1%; CI 95% (42.4-43.8%)) and health access (47.0%; CI 95% (46.3-47.7%)) as the worst expressions of inequality in their country. In contrast, only 20.2% (CI 95% (19.7-20.8%)) of people referred to income inequality. That is, the prevalence of inequality concerns in the domains of education and health was at least two times that of the domain of income inequality. At the country level, we observed the same tendency.

Despite finding some differences between countries², in general, concerns about education opportunities and health access were higher than concerns about income inequality (Figure 1). Likewise, the average percentages (between countries) of people who mentioned education and health domains were greater than the mean proportion of people who referred to income (Figure 2). Consistently, supplementary analyses showed that people were more willing to protest for better health and education than for higher wages and better working conditions (See Section 3.2. of Supplementary Materials).

Figure 1 Percentage of people that mentioned each domain of economic inequality in each country 

Figure 2 Violin plot with average percentage (between countries) of people that mentioned each domain of economic inequality 

We also found various profiles of people regarding inequality concerns through LCA. First, we estimated a one-class model, and then added classes until we identified the model with the best fit. We examined model fit based on our theoretical understanding of inequality concerns and the following statistical criteria (^{Weller et al., 2020}): the Bayesian information criterion and the Akaike information criterion, with lower values of these statistics indicating better model fit (^{Nylund et al., 2007}; ^{Weller et al., 2020}). Thus, we decided to retain a two-class model. Table 1 presents LCA results for various class models.

Table 1 Fit statistics for latent class model solutions of concerns about inequality of education, health, and income 

AIC = Akaike Information Criterion; BIC = Bayesian Information Criterion; df = degrees of freedom; G ² = likelihood ratio/deviance statistic; MLL = maximum log-likelihood; χ² = chi-square goodness of fit.

Regarding profiles in concerns about inequality, Class 1 was the most prevalent (69.36%, n = 11,419), while Class 2 was less frequent (30.64%, n = 5,044). Participants in Class 1 (unconcerned about inequality) had a low probability of mentioning education opportunities (23.78%) and health access (28.15%), and an even lower probability of considering income inequality (14.12%) one of the worst expressions of inequality in their country. In contrast, people in Class 2 (concerned for health and education inequalities) had a very high probability of mentioning education opportunities (97.34%) and health access (97.49%), and still a low probability of referring to income inequality (37.95%). Importantly, whereas people unconcerned about inequality were unlikely to mention any of the domains, people concerned about health and education were still less concerned about income inequality (Figure 3). See Supplementary materials (Table S3) to know about various determinants of class membership (e.g., age, gender, political ideology).

Figure 3 Probability of mentioning each domain of economic inequality as a function of latent class membership 

Concerns about economic inequality across domains and support of collective actions to reduce inequality

Consistent with H2, multilevel regression analyses revealed that concerns about inequality across domains might lead to greater support of collective actions to reduce inequality. Specifically, concerns about inequality in the domains of health and education were positive and significantly associated with support of collective actions across all three measures, even after controlling for income inequality concern (Table 2; M1a, M2a, M3a). We also controlled for covariates³. That is, the greater concern about health access and education opportunities, the greater willingness to participate and demonstrate for higher wages, better working conditions, better health and education, and a more egalitarian society.

Table 2 Multilevel regression models of collective actions to reduce inequality 

Note. * p < 0.05 ** p < 0.01 *** p < 0.001

Moreover, class membership obtained in the LCA was also a significant predictor of collective actions (Table 2; M1b, M2b, M3b). More specifically, people highly concerned (vs. unconcerned) about education and health were more willing to demonstrate and protest for higher wages, better working conditions, better health and education, and a more egalitarian society. All models confirmed our second hypothesis.

Discussion

The first aim of our research was to explore concerns about economic inequality in three domains: health, education, and income. Confirming H1, in general and in almost every country, people were more concerned about health and education disparities than about income inequality. Notwithstanding, we observed variations between countries that may reflect differences in their political systems and should be further explored. Furthermore, we found two profiles of people (or classes) by combining their concerns about economic inequality across domains. The first class was people not concerned about economic inequality in any domain, and the second class was people concerned about health access and education opportunities. Crucially, both classes had a low probability of mentioning income inequality as one of the worst expressions of inequality in their country.

These results support findings of prior research showing that people might desire more egalitarian distribution in the domains of health and education compared to income or wealth (e.g., ^{Howarth et al., 2019}; ^{Macchia & Ariely, 2021}). More broadly, these results speak in favour of recent empirical evidence (^{García-Sánchez et al., 2018}, ²⁰²²) and theoretical proposals (, 2018) that point to economic inequality as a multidimensional phenomenon that goes beyond income or wealth domains. Importantly, our findings might signal that, compared to health and education, income could be a less relevant domain in people’s concerns about economic inequality. This does not mean that income disparities are less important. Instead, we argue that taking into account other types of inequality, rather than only income disparities, could be critical to better understand people’s concerns about economic inequality (^{Jachimowicz et al., 2020}). As such, it may be important to increase awareness of inequality by encompassing all the domains that impact people’s everyday lives.

Our second aim was to analyse the relationship between concerns about economic inequality and support of collective actions to reduce inequality. In line with the H2, we found that concerns about health access and education opportunities (besides income inequality) were positively associated with support of collective actions to reduce inequality. In the same line, results showed that people concerned about health and education (Class 2) were more prone to engage in collective actions than those unconcerned about inequality (Class 1).

These findings are consistent with theoretical approaches of distributive justice signaling the key role of justice evaluations about the actual distribution (e.g., concerns about inequality) as a driving force to restore justice (^{Jasso et al., 2016}). In our study, concerns about inequality were associated with greater willingness to participate in collective actions to reduce economic inequality. Importantly, people were especially concerned about health and education, which may signal that redistributive justice principles could be different for each domain. We did not explore the different mechanisms that may operate in justice evaluations of inequality in each domain as it was beyond the scope of this research. Nevertheless, we hypothesise that while income inequality can be seen as fair by reasons of effort and deservingness (^{Barr & Miller, 2020}, ^{García-Sánchez et al., 2020}), it may be harder to justify inequality in health and education with such meritocratic beliefs.

Results also speak in favour of theoretical models of collective actions describing that concerns about inequality may trigger social mobilisation (^{Jetten et al., 2021}; ^{van Stekelenburg & Klandermans, 2013}; ^{van Zomeren et al., 2008}). In the same line, previous empirical evidence has shown that concerns about income inequality may promote collective actions (e.g., ^{Jo & Choi, 2019}; ^{Justino & Martorano, 2016}). However, our research might add that concerns about economic inequality across different domains (e.g., health or education) could play a differential role in participation in collective actions to reduce it. That is, concerns about each domain may have independent effects that can contribute, all together, to collective actions aimed at reducing economic inequality. In this sense, we argue that taking into account that people understand and react to economic inequality in different dimensions might be relevant for pursuing social change.

Another important contribution of our research is that it may help to understand the case of Latin America, where economic inequality has decreased but protests have risen in last decades (^{Díaz Pabón & Palacio Ludeña, 2020}; ^{Justino & Martorano, 2016}; ^{Ortiz et al., 2020}). Some have argued that despite reductions of inequality, people remained dissatisfied with the quality of public services, such as education or health (^{Justino & Martorano, 2016}). For instance, Chile’s recent social movement in favour of public education illustrates this reality (^{Huenupi, 2021}). Our findings highlight the importance of concerns about economic inequality in these domains and their clear connection with participation in collective actions to reduce inequality. This research might shed some light on a path for the emergence of social movements to reduce economic inequality in Latin America, still one of the most unequal regions in the world (^{UNU-WIDER, 2021}).

At this point, we acknowledge some limitations of this research and outline some directions for future investigations. First, this study design does not allow probing relationships of causality. Future studies should implement experimental designs to test whether concerns about economic inequality in different domains could actually promote collective actions. Second, we analysed responses to single-item measures, which have limited psychometric properties. Nevertheless, the series of results presented were based on various collective-actions indicators used in survey research. In the future, we encourage other researchers to use other measures, such as scales with several indicators, registering participation in collective actions, or designing behavioural tasks. Third, the characteristics of the sample may jeopardise the generalisation of our findings in other regions of the world, where public services of health and education might be covered more efficiently (such as Europe). Future studies should explore this question in different regions of the world, so we can examine potential differences between political systems, cultures, economic models, and so forth.

Conclusions

One key idea that we can take away from this research is that people might be more concerned about other domains of economic inequality (e.g., health access or education opportunities) rather than income differences. Nevertheless, the message is not that income inequality does not matter, as income disparities are at the base of economic and social inequalities. In this sense, our findings may denote a lack of connection between income differences and their effect on related domains such as health or education. We must reduce income disparities to reduce inequality in the other domains. Working only on education or health inequalities would have a minor impact on economic inequality (^{Breen & Chung, 2015}) and its several negative effects for individuals and our society.

The other important message of our research is that concerns about other domains of inequality besides income disparities (e.g., health, education, or other unexplored domains) may contribute to social mobilisation. In this regard, as concerns about health and education could be higher than concern about income inequality, they might serve as a common ground for initiating collective actions to reduce economic inequality. This preliminary evidence may have important implications for policy implementation. Specifically, media and political discourses could emphasise the effect of economic inequality in health and education as a way of increasing concerns about economic inequality and mobilising the public to reduce it ^{4 5}.