Abstract
Introduction
An example of a truly transboundary crisis, the Covid-19 pandemic hit the world in early 2020. As of 7 December 2022, more than 646 million people across the globe have been infected and over 6.6 million have died (John Hopkins University, 2022). The COVID-19 pandemic is not only the largest health crisis for the European Union (EU) in recent times, but its socio-economic impact was also significant: GDP in the EU dropped by about 6% in 2020 (European Commission, 2022: 24). Employment in the Euro area contracted sharply, with 5.2 million fewer people employed in the second quarter of 2020 than at the end of 2019 (a 3.2% decline). Meanwhile, unemployment increased less drastically (by 1.2 percentage points between February and October 2020), reflecting the large-scale job retention schemes that reached unprecedented levels across the EU (see, e.g., Anderton et al., 2021).
The COVID-19 crisis has been a test for solidarity in the EU. Not only did the member states express solidarity by providing aid to other EU countries in need, but the EU has also taken extraordinary measures to tackle the crisis: it has agreed on a substantial recovery package that redistributes funds to the most severely affected countries while taking on common debts, thus breaking an important taboo that especially northern member states had staunchly defended during the European sovereign debt (Eurozone) crisis. Hence, the EU recovery package enacted during the COVID-19 pandemic also provides an opportunity to investigate determinants of citizens’ willingness to support cross-border redistribution among EU member states in the unique context of a transboundary health and economic crisis (see Bauhr and Charron, 2021; Baute and de Ruijter, 2021; Bobzien and Kalleitner, 2021; Goldberg et al., 2021).
We focus on attitudes towards redistribution drawing on the case of the EU rescue package during the COVID-19 crisis. As independent variables, we concentrate on three explanatory concepts, which are justice attitudes, Euroscepticism and perceived self-interest. Our results are relevant for gauging the degree to which the EU is slowly becoming more like a ‘normal’ political system where citizens evaluate policy at least partially based on long-standing distributive ideologies and perceived self-interest – like in the national political arena – or whether citizens’ view of the EU is predominantly determined by their image of the EU. This is important for assessing the democratic quality of the EU and could also provide hints at the future direction of European integration, especially with respect to the idea of a Social Europe.
We study the determinants of support for the EU rescue package through an original survey conducted in Austria, Germany and Italy in December 2020. Our three dependent variables are the degree to which citizen support: (a) a common EU package to alleviate the crisis; (b) the EU taking on common debt; and (c) EU aid being directed to the countries most affected. Through descriptive statistics and regression analyses, we find that pro-EU attitudes and, contrary to previous findings (Bauhr and Charron, 2021; Bremer et al., 2021), instrumental calculations are the strongest determinants. Justice attitudes, i.e., the adherence of individuals to ideologies such as egalitarianism or individualism, have a smaller influence.
The EU's response to the COVID-19 pandemic: ‘Next generation EU’
When the new European Commission (henceforth, Commission) took office in December 2019, nobody could foresee the need to manage the most severe global health crisis in a century, which would supersede its ambitious working programme. As dramatic pictures of overcrowded hospitals especially in Northern Italy reached global audiences, member states initially resorted to national measures. These included the uncoordinated reintroduction of border controls, as well as intra-EU export bans on medical equipment (Pirker, 2020; Wolff et al., 2020). Nevertheless, the EU also proved capable of launching several first aid measures in response to the crisis – whether it was through the ECB's Pandemic Emergency Purchase Programme (a temporary financial asset purchase programme launched in March 2020 and gradually increased to a total volume of €1.85 trillion), an additional credit line within the European Stability Mechanism, or support schemes for small and medium-sized companies (Ladi and Tsarouhas, 2020).
Despite these initiatives, calls soon emerged for a long-term approach to mitigate the socio-economic consequences of the pandemic. Spurred by concerns over the unequal impact of the crisis across the EU and its potential to unravel European integration (Ladi and Tsarouhas, 2020; Morris and Birnbaum, 2020), the European Council (2020a) requested in April 2020 that the Commission develops a proposal for a post-pandemic recovery plan linked to the next Multiannual Financial Framework (MFF). The following month, a Franco-German proposal as well as the Commission's own recovery plan titled ‘Next Generation EU’ marked the beginning of long and controversial negotiations. Crucially, both proposals suggested that the Commission administers some funds not as loans but as (non-repayable) grants. Furthermore, it should finance the recovery fund by borrowing on the financial markets, with national commitments to the EU budget as a guarantee. These proposals prompted vocal opposition from Austria, Denmark, the Netherlands and Sweden. The so-called ‘frugal four’ – joined by Finland in the budget negotiations – strongly opposed the idea of debt mutualisation (Heikkilä and von der Burchard, 2020). This split, in turn, incited the resentment of many southern EU members, thus replicating the north-south divide that had characterised the Eurozone crisis.
On 21 July 2020, EU leaders finally reached a compromise on the recovery fund. The total budget was set at €750b, of which €390b in grants and €360b in loans. This measure effectively amounted to the introduction of ‘Eurobonds’, breaking ‘a major taboo of the Euro area era’ (Ladi and Tsarouhas, 2020: 1049). To compensate them for their concessions, the ‘frugal four’ (plus Germany) received rebates on their budgetary contributions (European Council, 2020b: 8). This breakthrough was an important step with the potential to fuel not only economic recovery but also closer European integration (Bradford, 2020; Financial Times, 2020). After the removal of a remaining barrier – a rule of law mechanism that introduced a certain conditionality into the allocation of EU funds – the EU institutions adopted both the MFF and Next Generation EU by the end of 2020.
Explaining public support for redistributive EU measures: Self-interest, justice attitudes and public opinion on European integration
According to Steele and Brenznau (2019: 3), a ‘redistributive policy’ is a policy that ‘draws from the common tax pool and then provides transfers or services that are taken up proportionally more by those who contribute less’. A classic example of a redistributive policy in this sense is child support financed through progressive taxation (Jacques and Noël, 2016). We use a broader understanding by also including other ways of financing redistribution than through taxes. A policy is redistributive in a relevant sense when those who profit from transfers and services are on average less involved in its financing. A redistributive policy could also be financed through debt or income from financial assets (Korpi and Palme, 1998).
On the EU level, outside of the Common Agricultural Policy – which only affects small parts of the EU's population – and EU cohesion policy, the EU has traditionally engaged very little in such policies and only redistributed comparatively small parts of its budget between member states (Beramendi, 2012: 70; Schraff, 2019: 85). However, more recent developments have challenged this principle (Daniele and Geys, 2015), such as the bailouts following the Global Financial Crisis in 2008/2009 or the COVID-19 pandemic, which has made the EU a major redistributor, if only temporarily.
With this increased prevalence of redistribution, a growing literature emerged investigating citizens’ attitudes towards redistributive policies at the EU level. This literature has, to a degree, tested the same explanatory variables as national-level studies, including self-interest or political ideology (Baute et al., 2019; Bechtel et al., 2014; Kanthak and Spies, 2017; Steele and Breznau, 2019). Yet, several studies have also added EU-specific variables such as trust in EU institutions (Bauhr and Charron, 2021; Daniele and Geys, 2015; Goldberg et al., 2021).
In this article, we want to add to this literature by investigating citizens’ support for the EU's redistributive efforts. Our study explores whether attitudes towards the EU or beliefs with respect to aspects of policies, similar to the national arena, influence citizens’ evaluations of redistributive policies on the EU level. We look at three relevant dimensions of EU level redistribution and asked respondents: (a) whether they support the EU's stronger involvement; (b) whether they back a common aid package and common borrowing; and (c) whether EU aid should disproportionately go to the countries most affected by the crisis. We test self-interest, distributive justice attitudes, and more general attitudes towards the European project as potential explanatory factors of citizens’ support for EU-level redistribution in three countries: Germany, Austria and Italy.
Self-interest and rational choice
The contention that self-interest is an important factor in citizens’ evaluations of policies stands in the tradition of rational choice theory. Rational choice theorists assume that individuals evaluate political proposals with respect to their effect on their utility (Downs, 1957). A ‘hard’ rational choice approach interprets utility mostly in terms of access to material goods (Opp, 1999: 174). With respect to support for European integration, there is some evidence that those who can be expected to profit from further integration tend to support it (Gabel, 1998: 337–338). However, these results are based on indirect measures of self-interest, such as proxies for income, education or skills, to assess whether someone is likely to gain from the specific opportunities generated by European integration rather than measuring directly whether these individuals perceive European integration to be in their self-interest. This approach runs the risk of falsely assuming that individuals correctly recognise whether European integration is in their material self-interest. Since individuals cannot usually expect to influence political events in a meaningful way, they have little material incentive to acquire political information (Downs, 1957). Therefore, we cannot infer that citizens’ subjective interests necessarily coincide with their objective material interests. Thus, instead of assuming the ‘hard’ assumption of fully informed rational agents, we use a ‘soft’ rational choice approach (Opp, 1999) where we focus on the individual
We look at the subjective perception of individual self-interest with respect to the EU rescue package. Furthermore, there is an individual and collective component of self-interest. First, we expect that if EU citizens think that EU policies benefit them personally, they are more likely to support such efforts. In addition, individuals are likely to expect that policies at the EU level benefitting their country are also of personal advantage. In contrast, if they think that either they personally or their country are net contributors, citizens might oppose redistributive policies at the EU level.
Justice ideologies
Worldviews or ideologies regarding what is just or unjust structure how individuals evaluate different kinds of facts. For example, the acceptance and legitimacy of economic systems and institutions depend fundamentally on whether citizens consider them to be just (Rodon and Sanjaume-Calvet, 2019: 384; Wegener, 1992). There is ample evidence that individuals’ evaluations of the welfare state and redistributive policies do not depend solely on their perceived self-interest but also on their value orientations (Dallinger, 2010; Steele and Breznau, 2019).
In political philosophy, we can differentiate between accounts of distributive justice which focus on individual merit (libertarianism), equality (socialism) and liberal equality (egalitarianism) (Cohen, 2009; Nozick, 1974; Rawls, 1999). Empirical research reflects this diversity: especially individual merit, equality and need are relevant values for people's perception of just socio-economic institutions and policies (Gerlitz et al., 2012: 264–265). Based on Wegener (1992), Gerlitz et al. (2012: 265) distinguish four justice ideologies, three of which are relevant here: individualism, egalitarianism, and ascriptivism. 1 ‘Each of them combines a certain worldview with complementary preferences for a distributive principle and a distributive agent. The more these preferences comply with the real distribution system, the higher its perceived justice should be’ (Gerlitz et al., 2012: 265).
‘Individualism’ is based on the belief that the market is the best distributive mechanism; its operation benefits everyone more than collective modes of distribution (Hayek, 2006). Individuals compete on the market and inequalities constitute important incentives. The fundamental distributive principle of this value orientation is individual effort, which should also be the main source of earned social inequalities. In contrast, ‘egalitarianism’ denies the justice of the market. Inequalities are the result of unfair allocations on the market, which significantly lacks ‘fair equality of opportunity’ (Rawls, 1999). Instead of the market, the state should be the agent of redistribution and correct for market allocations: accordingly, the state ‘represents the common good and holds the power to enforce redistribution’ (Gerlitz et al., 2012: 265). In contrast, ‘ascriptivism’ is an inherently conservative value orientation. Socio-economic inequalities are part of the just social order, which we should preserve.
In comparison to traditional two-dimensional measures in the welfare attitudes literature, like the left-right axis or redistribution preferences (see, e.g., Brady and Bostic, 2015), justice attitudes allow for the exploration of a multidimensional conceptual space that is theoretically connected to redistributive questions. Our measure of justice attitudes captures underlying ideological reasons agents might have when judging redistributive policies: it taps into the reasons for having attitudes towards different kinds of redistributive schemes.
In principle, supporting or rejecting each of these ideologies might influence how individuals evaluate the EU rescue package. First, we expect that those exhibiting an egalitarian ideology will support redistribution to correct for socio-economic inequalities and a common debt. In contrast, those who value individual merit will not support state interference in the market and oppose redistribution on the national or EU level because in the face of crisis individuals are expected to bear the risks themselves. Those who hold an ascriptivist ideology do not view inequality as problematic
Public opinion towards European integration
Social-psychological research has found that general ideological orientations – i.e., general belief-based predispositions and ideas, e.g., along the conservative/liberal or authoritarian/libertarian continuums – and internalised attitudes – i.e., psychological tendencies ‘expressed by evaluating a particular entity’, such as specific socio-economic issues (Eagly and Chaiken, 1993: 1) – influence how individuals process new information. Individuals tend to fit new information into their respective worldviews (Achen and Bartels, 2017). Individuals who strongly support or reject the EU thus tend to filter information according to their established viewpoint.
Against this background, we focus on the dimension of diffuse support for the EU, as opposed to specific support (Easton, 1975). Specific support refers to the outputs and performance of political authorities or policies based on a cost-benefit analysis. 2 Conversely, diffuse support is related to the object's properties and a general evaluation of what the object is. It often also relates to trust and legitimacy convictions vis-à-vis the political system. Citizens build up such diffuse support over time through experiences with the outputs of the system. For this study, we opt for a more general definition and refer to Euroscepticism as ‘opposition to some aspect of European integration’ (Leruth et al., 2018: 4). We see pro- or anti-European attitudes as a continuum. At the extreme of this continuum are ‘hard’ Eurosceptics characterised by their ‘outright rejection of the entire project of European political and economic integration and opposition to their country joining or remaining members of the EU’. On the positive side, there are qualified supporters of integration, who – while supporting European integration in general – see problems in the way the EU connects to its citizens.
The literature uses the notion of diffuse EU support as an explanatory variable for supporting certain EU/European policies. For example, there is evidence that an identification with the EU/Europe (defined and operationalised differently) correlates with support for further EU responsibility on social policy (Mau, 2005), backing for supranational fiscal governance (Kuhn and Stoeckel, 2014), support for European welfare solidarity (Gerhards et al., 2019), or with favouring a European minimum income benefit (Baute and Meuleman, 2020). We expect that somebody's stance on the pro-anti EU scale should also determine their opinion towards redistribution at the EU level.
The focus on self-interest, justice attitudes and attitudes towards the EU allows us to contribute to the literature on COVID-19-related EU-level redistribution, which is not unanimous, in several ways. To illustrate, through survey research in France, Italy and Romania, Bauhr and Charron (2021) find that self-interest has a relatively weak effect on citizen support for COVID-19-related EU redistribution. In their survey of France, Germany, Italy, the Netherlands and Spain, Bremer et al. (2021) also discover only a limited impact of self-interest. Conversely, Bobzien and Kalleitner (2021) find that self-interest is an important factor explaining citizens’ attitudes towards different components of the rescue package in Austria. The varying results might stem from differing operationalisations of the independent variable. For example, by focusing on perceived economic and health-related risks as the independent variable, Bauhr and Charron (2021) as well as Bremer et al. (2021) measure perceived collective self-interest indirectly. In contrast, Bobzien and Kalleitner (2021) look at how the expected long-term benefits for Austria influence citizens’ attitudes to supporting other countries. Unlike these approaches, we look at perceived individual and national self-interest with respect to the rescue package itself rather than generalised economic or health concerns. This way, we can better isolate the role of self-interest from other factors such as health related risks. With respect to normative beliefs, our focus on justice attitudes and Euroscepticism covers types of attitudes pertaining to either the EU or redistribution. We thus distinguish between socio-cultural elements of normative beliefs (Euroscepticism) and distributive beliefs (justice attitudes).
Case selection
We believe that contextual factors are potentially important for understanding individual attitudes towards EU policy and for excluding country-specific effects (and thus enhancing the generalisability of findings). We have therefore selected countries that differ with respect to being a net contributor or net beneficiary of the EU rescue package, the degree of affectedness by the pandemic and the government's position on the rescue package. These factors potentially influence individuals’ attitudes towards the rescue package. Citizens of net beneficiary countries or countries most affected by the pandemic could be more likely to expect the package to be in their interest. Their national government's position towards EU-wide redistribution could also signal to individuals whether the EU rescue package is in their interest. In this sense, the national government's narrative of the EU could be an interesting moderating variable: countries with a Eurosceptic government or strong Eurosceptic forces could be prone to an oppositional media narrative regarding EU policies. Thus, individuals might tend to view EU policies more negatively – independently of their individual degree of Euroscepticism.
Based on these considerations, we have chosen Germany, Austria and Italy for our survey in order to see whether public opinion differs in countries that exhibit variation in key aspects related to the impact of the pandemic, attitudes towards the EU in general, or the government stance on the EU rescue package. With respect to the level of EU support, public opinion has been more in favour of (further) European integration in Germany than in Austria and Italy in the past few years (see ESS8, 2016; ESS9, 2018). By including two relatively Eurosceptic countries, we want to ensure that we have enough Eurosceptic individuals in our sample to test our hypothesis that they might view the EU rescue package negatively. It will also be interesting to see whether citizen perceptions in Italy – and perhaps also Germany 3 – differ from those in Austria due to the Austrian government being part of the coalition of the ‘frugal four’. Despite the fact that we have carefully selected our three countries, we do not claim that they are necessarily representative of the whole EU. Even if there are differences between them, Germany, Austria and Italy feature among the richer EU countries and are in geographical proximity. Hence, we do not assume that our results can be generalised for all EU countries. Yet, our cases should allow us to find general patterns of public support for EU redistribution that potentially travels beyond these three countries.
Data and variables
We use an original survey to test the explanatory power of self-interest, justice attitudes and attitudes towards European integration on EU citizens’ opinions towards increased redistribution at the EU level. Our survey, administered by the survey company CINT, went into the field in the first week of December 2020. CINT drew participants from three online samples in Germany, Austria and Italy, each comprising several hundred thousand people. The size for each country sample was slightly over 1000. Using these large samples, we asked CINT to implement quotas for age, gender and regional dispersion. Despite the quotas, our sample only roughly matches the general population. For example, our sample is a bit younger than the general population (i.e., the mean age in the sample is slightly over 38 years; this compares to approximately 45 years in the three countries) and more educated (i.e., exactly half of the population has at least a high school degree, which is about one third higher than in the general population). Summary statistics are reported in the Online appendix.
To capture different aspects of redistribution on the EU level, we opt for three dependent variables gauging: (a) whether people support the EU rescue package to alleviate the COVID-19 crisis; (b) whether the EU should allow common debt to alleviate the COVID-19 crisis; and (c) whether aid funds should go to those countries most affected by the pandemic. For each of these three variables, we use a 0 to 100 scale, with 0 indicating strong disapproval and 100 signifying strong approval. We operationalise our main explanatory theories as follows: for self-interest, we use two proxy variables that respectively measure citizens’ assessment of the benefit of the EU rescue package for their country and for themselves personally. To test the explanatory power of the three justice attitudes, we add a proxy for egalitarianism (i.e., whether the state should guarantee a minimum standard of living), individualism (i.e., whether it is right that businesses make a lot of money; this will benefit everyone) and ascriptivism (i.e., whether it is correct that parents pass on their money to their children). For measuring these justice attitudes, we employ the items developed by Stark et al. (2014). To gauge our third explanatory concept, Euroscepticism or EU support, we ask whether respondents think that EU integration has gone too far or should go further. We also measure these independent variables on a 0 to 100 scale (see the Online appendix for details). 4
We also add several control variables, even if the direction of the control variables might not always be clear beforehand. First, we add respondents’ education and financial resources. Even if they might not benefit directly, economically well-off and educated individuals could see opportunities and benefits of redistributive policies at the EU level; they might, therefore, be more likely to support such measures (for a similar argument, see Anderson and Reichert, 1995; Gabel, 1998). Second, individuals’ political ideology might influence their opinions on redistributive policies. For example, citizens on the right of the political spectrum should value national identities and nation state sovereignty. These individuals may disapprove of redistributive policies at the EU level (Carey, 2002; McLaren, 2002). Our proxy variable for political ideology is citizens’ self-placement on a left-right scale. Third, we control for citizens’ satisfaction with their national democracy. We posit that individuals may rely on domestic cues like satisfaction with democracy as information shortcuts for developing their views on policies at the European level (Anderson, 1998; Franklin et al., 1994); this might include policies about redistribution. Finally, we control for age (at the time of the survey), gender (a nominal three-value variable capturing men, women, and others) and place of residence (a nominal variable with four categories: (1) farm or small village in the countryside; (2) small town in the countryside; (3) suburbs of a big city; (4) big city). More details are available in the Online appendix.
Methods
First, we display a histogram of any of the three dependent variables. This gives us an idea of general support ratings for the EU's COVID-19 funds and the distributions in answers overall and in the three countries. Second, we present some bivariate scatterplots displaying the effect of our three independent variables – justice attitudes, general EU support, and self-interest – on the three dependent variables – support for the aid package, common EU debt, and EU aid allocation for those countries most in need. Third, we display the results of three sets of multiple regression models, featuring the three dependent variables on the left-hand side, respectively. On the right-hand side are the proxy variables for justice attitudes, self-interest, and EU support, as well as the control variables. We also add country dummies for Austria and Italy, respectively, with Germany serving as the reference category. For each set of models, we present four equations. The first model includes the proxy variables for socio-tropic instrumental calculations, egalitarianism, and latent EU support. The second model replaces the socio-tropic instrumental calculations dummy with the ego-tropic one. The third and fourth models replace egalitarianism with individualism and ascriptivism, respectively. 5 Finally, we add interaction terms of our proxy variables for rational choice, justice attitudes and latent EU support with the two country dummies for Austria and Italy. As we cannot directly interpret these interaction coefficients, we also show the predicted probabilities, graphically displaying the effect of all these interactions.
For presentation purposes, we display below all analyses for our first dependent variable (whether the EU should have an aid package to alleviate the COVID-19 crisis). For our second and third dependent variables, measuring whether the EU should allow common debt to alleviate the crisis and whether aid funds should go to those countries most affected by the pandemic, we show all analyses in the Online appendix, but discuss them briefly below.
As for any cross-sectional study, endogeneity might be an issue for the current analysis. However, we think that the associational chain should go from justice attitudes and general EU beliefs towards redistributive preferences at the EU level. Justice attitudes and EU attitudes are stable long-lasting beliefs that likely influence EU-level redistributive preferences. In contrast, it is theoretically unlikely that citizens form opinions about redistribution at the EU level first, and that these opinions then influence general beliefs about the EU or moral values.
Results
The first result that stands out is that our respondents appear to be rather supportive of redistribution at the EU level. The histograms in Figure 1 as well as in the Online appendix illustrate that the average ratings on our 0 to 100 scales are 64 points in support of an EU aid package, 59 points on whether EU aid should go disproportionately to countries most affected, and 52 points on the common debt question. The bivariate scatterplots in Figure 2 and the Online appendix seems to confirm that our three theories have explanatory power and account for variation in the three dependent variables. However, the various graphs also illustrate that some variables appear to have more explanatory power than others. More specifically we find that self-interest and latent EU support appear to have a larger influence than justice attitudes when it comes to citizens’ support of the EU aid package. For both socio-tropic instrumental calculations, as well as latent EU support, Figure 2 roughly portrays a 40-point difference in support for the rescue package if we go from the negative to the positive endpoint for the two independent variables. We also find greater influence of socio-tropic instrumental calculations compared to ego-tropic instrumental calculations (see the Online appendix). In fact, the former appears to influence attitudes towards the EU aid package twice as strongly as the latter one. For justice attitudes proxies, even if the average influence of any of the three proxies is rather limited, egalitarianism seems to have a stronger influence compared to individualism and ascriptivism.
Public support for an EU-funded COVID-19 relief package.
The bivariate relationship between instrumental calculations, justice attitudes and latent EU support on respondents’ support for an EU-funded COVID-19 relief package.
The multiple regression results appear to confirm the bivariate results (see Table 1 and the Online appendix). 6 All three factors – self-interest, justice attitudes and EU support – seem to have predictive power; their unstandardised regression coefficients are positive and generally point in the expected direction. If we compare the respective coefficients of our three 0 to 100 scales, we confirm that the substantive impact of latent EU support and socio-tropic instrumental calculations is highest. For example, Table 1 roughly predicts that for every point a person thinks that her country benefits more from the EU-funded COVID-19 package her support increases by .3 points. For latent EU support the predicted increase in EU COVID-19 aid is roughly .25 points for every point a person becomes more EU friendly. In contrast, a one-point increase on the egalitarianism scale, as the justice attitude with most predictive power, appears to only lead to .15 increase in support for an EU COVID-19 rescue package. The regression models also seem to confirm that socio-tropic instrumental calculations trump ego-tropic calculations. In fact, the coefficient for ego-tropic instrumental calculations is not statistically significant in Model 2 (Table 1). More broadly, the limited influence of ego-tropic instrumental calculations confirms results from other areas of public opinion research, in which socio-tropic considerations also trump ego-centric considerations. An example would be anti-immigration attitudes, the main drivers of which are socio-tropic sentiments of cultural threat, rather than ego-tropic cultural threat perceptions (see Hainmueller and Hopkins, 2014).
Multiple regression models measuring the effect of justice attitudes, rational choice and EU support on public support for an EU-funded COVID-19 relief package.
If we hone into justice attitudes more closely, egalitarianism appears to have a stronger influence than ascriptivism; and individualism seems to have the weakest influence. These patterns appear to be consistent regardless of the dependent variables (i.e., support for an EU aid package, support for common debt or support for unequal redistribution). In Table 1, as well as the Online appendix, the three proxy variables for individualism and ascriptivism appear to moderately influence our three dependent variables, albeit with minor variations (e.g., individualism seems to have no impact on general support for an EU aid package, but a weak and unexpectedly positive effect on the other proxies measuring support for EU-level redistribution).
Finally, our results seem to indicate that individual characteristics shape citizens opinions about redistribution more than the country context an individual lives in (see Figures 3 to 6, as well as Table 2). In fact, the context has only some, albeit rather limited, moderating influence (further details can be found in the Online appendix). For justice attitudes, these differences are rather nuanced and might differ based on our dependent variable. For example, egalitarian attitudes appear to have a stronger bearing on support for EU-directed redistribution in Germany compared to Austria or Italy. In contrast, individualism tends to have no impact on the three EU-distribution proxies in Germany and Austria, but an unexpected slightly positive impact in Italy. When it comes to rational choice, socio-tropic instrumental calculations seem to have a large influence in Germany and Austria as compared to Italy. Finally, concerning EU support, the link between latent EU support and support for more EU-wide redistribution appears weakest in Italy and stronger in Austria and Germany (see also the results presented in the Online appendix).
The predicted effect of socio-tropic instrumental calculations, justice attitudes and latent EU-support on respondents’ support for an EU-funded COVID-19 relief package.
The predicted effect of ego-tropic instrumental calculations, egalitarianism and latent EU support on respondents’ support for an EU-funded COVID-19 relief package in Germany, Austria and Italy.
The predicted effect of socio-tropic instrumental calculations, individualism and latent EU support on respondents’ support for an EU-funded COVID-19 relief package in Germany, Austria and Italy.
The predicted effect of socio-tropic instrumental calculations, ascriptivism and latent EU support on respondents’ support for an EU-funded COVID-19 relief package in Germany, Austria and Italy.
Multiple regression models measuring country differences in the effect of justice attitudes, rational choice and EU support on public support for an EU-funded COVID-19 relief package in Germany, Austria and Italy.
Conclusion
Our article makes several contributions to the literature. First, we find that above 50 percent of our respondents support the EU rescue package. In addition, all independent variables – self-interest, general EU support, and justice attitudes – have explanatory power in predicting citizens’ opinions towards EU-level redistributive policies. However, we observe that latent EU support and instrumental calculations trump justice attitudes in their explanatory power – in contrast to several studies in the COVID-19 context that only find a weak effect for self-interest (cf. Baute and de Ruijter, 2021; Bauhr and Charron, 2021; Bremer et al., 2021). Furthermore, we find that citizens judge policy initiatives such as a COVID-19 support fund more with regard to how much such funds benefit their home country rather than themselves. And for EU support, our results indicate that positive attitudes towards more integration also spill over to support for EU-level redistribution. This result nicely complements Papageorgiou and Immonen (2023), who find that individual satisfaction with the solidaristic actions of member states towards each other lead to a more positive assessment of the EU overall. Taken together, the results of Papageorgiou and Immonen (2023) coupled with ours point to a positive feedback loop, in which citizens, who support the EU want to see widespread cooperation among member states including social and financial matters. Once they perceive such cooperation, they might become even more positive about the EU.
Third, we find that the effect of justice attitudes is smaller on people's opinions on EU redistribution. Interestingly, egalitarian justice attitudes seem to have a larger influence than either individualistic or ascriptivist justice attitudes. Also, in some models, individualistic justice attitudes have a positive impact on support for the rescue package. This finding may be understood as a sign that some individuals do not interpret the recovery package in terms of their ideological predisposition but as an extraordinary situation that requires extraordinary measures. Finally, all these results are relatively stable across countries with only moderate differences between them.
We believe that our findings have broader implications for how citizens view the EU and EU policies. First, we find evidence that individuals who support EU integration seem to support it holistically; this also includes redistribution. This implies that if general support for further integration wanes/increases, support for more redistribution at the EU level will also likely diminish/become stronger. Second, the relevance of socio-tropic instrumental calculations suggests that individuals must have the feeling that their country benefits from the EU in some way. Thus, support for EU-level redistribution also depends on whether policies are designed in such a way that individuals have reason to believe they somehow benefit materially from these policies. This finding points to some similarities with mechanisms proposed in theories of welfare state expansion that suggest an important role of interest-based coalitions in support for redistributive policies (cf. Esping-Andersen, 1990; Korpi and Palme, 1998). Third, we need to be careful in interpreting the (smaller) effect of justice attitudes. The literature (see Eagly and Chaiken, 1995; Toikko and Rantanen, 2017) agrees that abstract attitudes tend to have a smaller effect than more concrete attitudes. The fact that we have only asked abstract questions about justice attitudes, but concrete questions about citizens’ instrumental calculations, might explain at least part of this larger influence of self-interest. Thus, it is reasonable to think that egalitarian beliefs among European citizens could be a source of support for the establishment of more redistributive policies on the European level. This speculation could be further strengthened by the fact that contrary to our findings, studies on the question of public borrowing (Eurobonds) during the Eurozone crisis have demonstrated that public support may (strongly) correlate with (egalitarian) justice attitudes (Daniele and Geys, 2015).
A major caveat of our study's results is that they are based on an online sample. Even though our sample is diverse and despite the fact that we implemented quotas, there are some discrepancies in the demographic and socio-economic characteristics between our sample and the general population in Germany, Austria and Italy. For instance, the fact that our sample is younger and more educated could bias our results towards higher support ratings of the EU, and potentially also more support for redistribution. Participants in online panels are also ‘professional’ survey takers and they receive some compensation for their work. While we do not believe that such considerations will tremendously skew our results, we still believe that it would be ideal to replicate our study with a fully representative non-online sample to corroborate our findings.
Supplemental Material
sj-pdf-1-eup-10.1177_14651165231166284 - Supplemental material for What guides citizen support for redistributive EU measures as a response to COVID-19: Justice attitudes, self-interest or support for European integration?
Supplemental material, sj-pdf-1-eup-10.1177_14651165231166284 for What guides citizen support for redistributive EU measures as a response to COVID-19: Justice attitudes, self-interest or support for European integration? by Doris Unger, Jürgen Sirsch, Daniel Stockemer and Arne Niemann in European Union Politics
Supplemental Material
sj-do-2-eup-10.1177_14651165231166284 - Supplemental material for What guides citizen support for redistributive EU measures as a response to COVID-19: Justice attitudes, self-interest or support for European integration?
Supplemental material, sj-do-2-eup-10.1177_14651165231166284 for What guides citizen support for redistributive EU measures as a response to COVID-19: Justice attitudes, self-interest or support for European integration? by Doris Unger, Jürgen Sirsch, Daniel Stockemer and Arne Niemann in European Union Politics
Supplemental Material
sj-dta-3-eup-10.1177_14651165231166284 - Supplemental material for What guides citizen support for redistributive EU measures as a response to COVID-19: Justice attitudes, self-interest or support for European integration?
Supplemental material, sj-dta-3-eup-10.1177_14651165231166284 for What guides citizen support for redistributive EU measures as a response to COVID-19: Justice attitudes, self-interest or support for European integration? by Doris Unger, Jürgen Sirsch, Daniel Stockemer and Arne Niemann in European Union Politics
Footnotes
Acknowledgements
Author contributions
Funding
ORCID iDs
Supplemental material
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
