Introduction
The debate on the social dimension of the European Union is generating more political discussion and controversy than ever before. The call for expanding Social Europe mainly comes from the political left, which is preoccupied with strengthening and preserving the achievements of national welfare states, and considers the development of Social Europe as a necessity (Føllesdal et al., Reference Føllesdal, Giorgi and Heuberger2007). Politicians and scholars have presented a number of arguments in favour of strengthening Social Europe, including the expectation that it would increase the popular legitimacy of the European project as it could compensate for the negative consequences of economic integration by providing social protection at the EU level (Fernandes and Maslauskaite, Reference Fernandes and Maslauskaite2013).
Nevertheless, little scholarly attention has been paid to citizens’ attitudes toward the EU's social dimension and how these positions are related to their attitudes toward national welfare states. In this study, we approach Social Europe as referring to current or proposed EU governance that establishes supranational social policies and that affects social rights and policies in the member states (Martinsen and Vollaard, Reference Martinsen and Vollaard2014: 680). In some respects, Social Europe strengthens the national welfare state while, at the same time, challenging its foundations (Ferrera, Reference Ferrera, Vandenbroucke, Barnard and De Baere2017). For instance, social regulations can expand national legislation, whereas the coordination of social security rights prohibits social benefits and services being restricted to member states’ own citizens. These two dynamics of Social Europe may generate different expectations about how attitudes toward national welfare states relate to support for Social Europe. On the one hand, attitudes concerning the national welfare state can spill over to the European area, given that both Social Europe and the national welfare state aim to achieve similar goals. On the other hand, it can be expected that support for the welfare state hinders support for Social Europe, if the nation state and the European Union are considered as competing or substituting governance levels (Burgoon, Reference Burgoon2009).
Previous research shows that, in member states with high levels of income equality, citizens who support state intervention and income redistribution have more negative attitudes toward the EU (Garry and Tilley, Reference Garry and Tilley2015). Furthermore, citizens who are dissatisfied with national public services (Kumlin, Reference Kumlin2009), evaluate their coverage as insufficient, and have less confidence in the sustainability of their national welfare state (Beaudonnet, Reference Beaudonnet2015), are also less supportive of the EU. This may indicate that citizens either blame the EU for being the cause of their malfunctioning national system or consider it – at least in advanced welfare states – as a threat. Given these results, it remains unclear what citizens expect from the EU regarding social policy and, in particular, how their welfare attitudes are related to support for Social Europe. By investigating multiple welfare attitude dimensions, the current study aims to pinpoint how welfare state support is related to support for Social Europe and whether the relationship with welfare state attitudes varies across different dimensions of Social Europe. To fully understand this linkage, we distinguish between attitudes toward the basic principles of the welfare state and welfare state critique, since their relationship with support for Social Europe is based on a different underlying logic.
Explaining support for Social Europe
Welfare attitudes: Spillover or obstacle to support for Social Europe?
We propose two explanatory mechanisms for how welfare state attitudes may relate to support for Social Europe: namely, spillover or obstacle. On the one hand, Social Europe can be considered as an instrument that aims to achieve similar objectives to those of national welfare states. The social dimension of the EU has developed to counterbalance the negative impact of economic deregulation and to prevent a ‘race to the bottom’ in social standards (Fernandes and Maslauskaite, Reference Fernandes and Maslauskaite2013). In this respect, Vandenbroucke (Reference Vandenbroucke2013: 221) argues that a Social Union should guide and support the substantive development of national welfare states, based on general social standards and common objectives. From the perspective that Social Europe aims to defend the welfare state at the European level, it can be expected that advocates of the national welfare state will be more positive toward Social Europe. This spillover mechanism implies that citizens’ attitudes are congruent across policy levels (Muñoz et al., Reference Muñoz, Torcal and Bonet2011). A spillover of pro-welfare attitudes to the European area is consistent with cue-taking theory, which assumes that citizens use their attitudes about domestic politics as a proxy to evaluate European integration (Anderson, Reference Anderson1998). Given that many citizens have limited interest in and awareness of European politics, general attitudes toward the welfare state can be activated as a heuristic, and evoke attitudes toward the EU's social dimension.
On the other hand, it is argued that Social Europe challenges the foundations of national welfare states. Ferrera (Reference Ferrera2005) points out that national welfare states are based on ‘closure’, as compulsory social insurance draws clear boundaries between those who are entitled to social benefits and those who are not eligible to join the social security system. European integration, on the contrary, operates on the basis of ‘opening’, since it attempts to redraw the boundaries of welfare at the supranational level. The EU constrains the scope and content of national bounding decisions, as well as the right to impose boundaries as such (Ferrera, Reference Ferrera2005: 3). Others have expressed concerns that European integration will not be able to restructure, at the supranational level, the type of solidarity that exists in national welfare states (Hemerijck, Reference Hemerijck2012; Scharpf, Reference Scharpf2010). Integration not only affects the boundaries of social sharing systems, but also the internal design of welfare states: for instance, in the areas of pensions, health care and social assistance. The threat that Social Europe poses to the boundaries of welfare states can nourish competition for scarce resources, whereas the impact on the internal design of the welfare state can raise concerns – especially in the most advanced welfare states – about the maintenance of social protection levels. From this viewpoint, pro-welfare attitudes may obstruct support for Social Europe. First, citizens’ preferences regarding strong government responsibility for welfare and for income redistribution might be restricted to national boundaries. In this respect, pro-welfare state positions might coincide with welfare chauvinism. Second, citizens who are satisfied with their welfare regime may perceive Social Europe as a threat to institutionalised solidarity at the national level, whereas those who think that national institutions perform poorly may see it as an opportunity (Beaudonnet, Reference Beaudonnet2015; Sanchez-Cuenca, Reference Sanchez-Cuenca2000).
Dimensions of support for Social Europe
The notion of ‘Social Europe’ can refer to different basic principles and policy options, which vary in the extent to which they are intrusive to national welfare states. By intrusiveness, we understand the degree of interference in the boundaries of solidarity on which the welfare state is based. Accordingly, empirical research shows that citizens differentiate substantially in their attitudes concerning various aspects of Social Europe (Baute et al., Reference Baute, Meuleman and Abts2018). Whether citizens’ welfare attitudes facilitate or obstruct support for Social Europe might depend on how a particular policy instrument intervenes in the national welfare state.
The first, and least intrusive, aspect of current European social policy relates to the harmonisation of national social policies. This is mainly addressed through binding and non-binding social regulations: for instance, in the fields of health and safety at work, working conditions, and equality (Falkner, Reference Falkner, Castles, Leibfried and Lewis2010). The regulatory angle of Social Europe actually supports the welfare state, as it reinforces social protection, while leaving the boundaries of national welfare states untouched. More recently, the Open Method of Coordination was introduced to facilitate an upward convergence of social standards through mutual learning and peer pressure (de la Porte, Reference de la Porte and Greve2013). Given the large diversity of national welfare states, complete harmonisation is not aimed at. Nevertheless, the EU's instruments for harmonisation are a direct pressure toward more ‘bounded varieties of welfare’ (Falkner, Reference Falkner, Castles, Leibfried and Lewis2010).
Second, Social Europe also requires financial solidarity that goes beyond the national welfare state (Sangiovanni, Reference Sangiovanni2013). Member state solidarity is implemented through various structural funds, which aim to reduce regional disparities in income, employment, investment and growth within the EU (Allen, Reference Allen, Wallace, Pollack and Young2010). The fiscal aid to Eurozone countries, which overturned the ‘no bailout clause’ during the European sovereign debt crisis, is also considered as an instrument of international redistribution. In essence, member state solidarity provides financial assistance from more affluent regions to poorer ones, on top of existing forms of institutionalised solidarity within member states. Therefore, these transfers do not erode the autonomy of member states to conduct their own social policies.
Third, the development of European social citizenship (Faist, Reference Faist2001) is a cornerstone of Social Europe that operates according to an opposing dynamic to that of national welfare states. It implies that EU citizens acquire access to other member states’ social security schemes and that already-earned social security rights are transferrable between member states. The creation of an EU social citizenship space – which matches the EU's territorial borders – strongly infringes on the boundaries of national welfare states (Ferrera, Reference Ferrera2005, Reference Ferrera, Vandenbroucke, Barnard and De Baere2017). Currently, EU citizens receive equal social rights as nationals, without a European standard, as the amount, type and duration of benefits depend on the country of residence.
Turning to the most intrusive dimension, Social Europe can also be implemented by policy instruments that are based on interpersonal solidarity, defining rights and obligations among EU citizens (Sangiovanni, Reference Sangiovanni2013). Such policy instruments have been proposed in the form of a European unemployment insurance scheme (Dullien, Reference Dullien2013), a European minimum income benefit (Peña-Casas and Bouget, Reference Peña-Casas, Bouget and Natali2014), and a European child benefit (Levy et al., Reference Levy, Matsaganis and Sutherland2013). European social protection schemes would introduce new redistributive mechanisms and would enforce financial solidarity between EU citizens. Their level of contributions and benefits would be based on a relatively low common denominator between member states and could be topped up by the member states with national payments. Such measures would for instance be financed by a flat tax on all household income and limited to a maximum proportion of GDP. This is clearly the most intrusive component of Social Europe to the welfare state.
Welfare state principles and welfare state critique
To gain detailed insight into the spillover and obstacle mechanisms, we incorporate two clusters of welfare attitudes in this study, namely support for basic welfare state principles and welfare state critique. The distinction between the two is relevant, because their relationship with support for Social Europe rests on different theoretical foundations.
Attitudes about welfare state principles refer to citizens’ preferences for redistribution and the government's responsibility for welfare. Welfare state principles refer to what Roosma and colleagues (Reference Roosma, Gelissen and van Oorschot2013) label the ‘welfare mix’: the goals of the welfare state and the preferred range of government responsibility for welfare. First, the welfare mix refers to the diversity of welfare-generating forces. The state, the market, civil society or the family can all operate as regulatory or redistributive institutions (Roosma et al., Reference Roosma, Gelissen and van Oorschot2013). Our primary interest is citizens’ preferences regarding the role of the government in providing welfare. A second welfare state principle relates to the redistributive goals of the welfare state (van Oorschot and Meuleman, Reference van Oorschot and Meuleman2012). Imposing equality of opportunities and/or outcomes is one of its main objectives (Esping-Andersen, Reference Esping-Andersen1990). Support for welfare state principles is thus strongly related to egalitarian views. The third aspect, the range of government responsibilities, refers to the specific areas of life in which the government should intervene (Roller, Reference Roller, Borre and Scarbrough1995). For instance, one might think that the government should ensure the provision of health care, pensions, unemployment benefits, child care, paid care leave and so on.
At the European level, a similar debate is ongoing regarding the role of government, the promotion of equality and the range of government responsibilities. Cue-taking theory (Anderson, Reference Anderson1998) would predict a positive relationship between support for basic welfare state principles and support for Social Europe. For instance, it assumes that citizens with a more liberal orientation are strongly in favour of the EU's internal market while being more opposed to European social policy, because they prefer less government intervention and fewer regulations. In line with this theory, empirical studies show that citizens with egalitarian views and those who prefer higher social spending are more in favour of member state solidarity (Beaudonnet, Reference Beaudonnet2014; Ciornei and Recchi, Reference Ciornei and Recchi2017). However, welfare state development has facilitated internal bonding between insiders by means of external bounding toward outsiders (Ferrera, Reference Ferrera2005). Citizens mainly think about state intervention and welfare redistribution within national boundaries. Literature on welfare chauvinism confirms that endorsing the basic principles of the welfare state is not necessarily consistent with support for welfare redistribution to non-nationals (van der Waal et al., Reference van der Waal, Achterberg and Houtman2010). As a result, support for welfare redistribution and government responsibility may translate into opposition to European interference in social policy, in particular regarding the Europeanisation of social rights.
In contrast to welfare state principles, welfare state critique does not refer to an ideological position as such, but to citizens’ assessment of the performance and consequences of their national welfare state. The performance of the welfare state generates a certain amount of institutional trust among citizens. We consider people's trust in the social security system as an overarching evaluation of their national welfare state. However, more specific components of welfare state critique can be found in literature. Perceptions of the mistargeting of welfare benefits, and especially the overuse of benefits, are a sensitive subject among the European public (Ervasti, Reference Ervasti, Ervasti, Andersen and Friberg2012). One might perceive that some beneficiaries are not deserving of, or not entitled to, receive social benefits, which in the latter case is considered as benefit abuse. Furthermore, perhaps the most criticised side effect of the welfare state is its economic consequences (van Oorschot et al., Reference van Oorschot, Reeskens and Meuleman2012). The critique is that the welfare state is a financial burden on the government budget, increases labour costs and tax levels, and makes labour markets inflexible. Together with (dis)trust in the social security system and perceptions of benefit overuse, we consider the perceived economic consequences as a major component of welfare state critique.
Some scholars argue that both national and European institutions are evaluated similarly (Anderson, Reference Anderson1998), assuming that citizens’ evaluations rest on institutional trust in general (Muñoz et al., Reference Muñoz, Torcal and Bonet2011). A spillover effect may also result from a direct involvement of national governments in EU decision-making (Kritzinger, Reference Kritzinger2003). In this regard, citizens’ support for Social Europe expresses support for incumbent authorities’ policies at the European level. The opposite argument – in favour of the obstacle mechanism – is also plausible: namely, that national and European systems are evaluated differently, since citizens make cost/benefit evaluations over transferring sovereignty to the European level (Sánchez-Cuenca, Reference Sanchez-Cuenca2000). Those who are satisfied with the functioning of the national welfare state may be more reserved about Social Europe, whereas those who perceive the welfare state as inefficient may embrace European social policy-making more enthusiastically (Beaudonnet, Reference Beaudonnet2015). This implies that strong criticism of the welfare state may boost support for Social Europe, whereas little criticism is accompanied by lower support for Social Europe.
Hypotheses
We have argued that both support for the basic welfare state principles and welfare state critique can either facilitate or hinder support for Social Europe. The contrasting mechanisms of spillover and obstacle lead us to formulate two alternative hypotheses:
Spillover hypothesis – H1: High levels of support for the basic principles of the welfare state (H1a) and low levels of welfare state critique (H1b) are positively related to support for Social Europe.
Obstacle hypothesis – H2: High levels of support for the basic principles of the welfare state (H2a) and low levels of welfare state critique (H2b) are negatively related to support for Social Europe.
Furthermore, the policy instruments of Social Europe do not intervene in national welfare states in the same way. EU social regulations and member state solidarity are the least intrusive to national welfare states, whereas European social citizenship and interpersonal solidarity challenge and undermine the sovereignty and boundaries of the welfare states to a larger extent. In line with the arguments set out in the previous section, we assume that the spillover mechanism of basic welfare state principles will be stronger for those aspects of Social Europe that are least intrusive because citizens will perceive them as supportive to their national welfare state. More intrusive instruments, such as European social citizenship and interpersonal solidarity, may be perceived as more threatening to national welfare states and raise concerns about the consequences of these policies. In sum, given that the various dimensions of Social Europe differ in the extent to which they are intrusive to the welfare state, we expect the following differentiation in the spillover and obstacle mechanism as referred to in the previous hypotheses:
H3: The spillover effect is stronger for less intrusive dimensions of Social Europe (H3a), whereas the obstacle effect is stronger for more intrusive dimensions (H3b).
Furthermore, we expect the relative importance of the welfare state principles and welfare state critique to depend on the intrusiveness of the different dimensions of Social Europe. For dimensions that intervene more strongly in national systems, citizens’ assessments of whether this will bring improvement or deterioration will gain importance. The opportunity costs of transferring social competences to the European level are much higher for citizens who are satisfied with the performance of their welfare state. Therefore, for the most intrusive components of Social Europe, the level of criticism aimed at the welfare state is likely to be more important than the level of support for the basic principles of the welfare state. When it comes to less intrusive policy instruments, citizens may reason more according to their ideological preferences regarding redistribution and regulations, which can be expected to have a stronger effect on support for Social Europe than their level of welfare state critique.
H4: For less intrusive dimensions of Social Europe, the level of support for the basic principles of the welfare state is more important in explaining citizens’ support for Social Europe (H4a) whereas, for more intrusive dimensions of Social Europe, the level of welfare state critique is more important (H4b).
Data and methodology
Data
We use data from the 2014 Belgian National Election Study (Abts et al., Reference Abts, Swyngedouw and Meuleman2015). This post-electoral survey was carried out among a register-based probability sample of Belgians entitled to vote in the 2014 national elections. On completion of a computer-assisted personal interview (response rate 47 per cent), respondents were asked to fill out a 20-page drop-off questionnaire, containing a specific module on Social Europe. Applying the principles of the Tailored Design Method (Dillman et al., Reference Dillman, Smyth and Christian2014), we were able to convince 74 per cent of the respondents to fill out and send back the questionnaire (N = 1403).
Variables
Support for Social Europe is measured by a second-order latent factor, consisting of the four constitutive dimensions distinguished above (see online Appendices 1–2 for question wording). Each of the dimensions is specified as a latent factor. First, attitudes toward harmonisation are measured by support for four types of social regulations covering health and safety at work, maximum weekly working hours, minimum terms for paid leave, and minimum terms for maternity leave. Second, attitudes concerning member state solidarity are measured by three items referring to solidarity between the richer and poorer member states. These concern support in times of economic difficulties, the amount of tax money being redistributed, and the continuation of member state solidarity in the future. Third, opinions about European social citizenship are operationalised by four items concerning citizens’ attitudes toward the access of EU citizens to social benefits and protection in Belgium. One item concerns equal social rights, two items relate to prioritising nationals, and one item refers to the conditionality of social protection. Lastly, interpersonal solidarity is measured by support for the implementation of a European social security system. One item concerns the implementation of an entire European social security system, whereas three items refer to European protection schemes for specific policy areas: child allowances, minimum income benefits, and unemployment benefits.
Support for welfare state principles is measured by three latent factors. First, the role of the state versus the market is measured by the following agree-disagree statements recorded on a five-point scale: ‘Society would be better off if the government intervened less in the market’ and ‘Businesses should have more freedom; therefore, regulations for businesses should be reduced’. A higher factor score indicates that respondents support greater government intervention in the economy. Second, attitudes toward the principle of equality are measured by three items (answers on five-point scales): ‘The differences between classes ought to be smaller than they are at present’, ‘The differences between high and low incomes should stay as they are’, and ‘The government should reduce income differentials’. Higher scores indicate more egalitarian views. To measure attitudes concerning the range of government responsibilities, respondents were asked to what extent they think the government should be responsible for ‘a reasonable pension,’ ‘affordable health care’ and ‘a reasonable standard of living for the employed’. Responses range from ‘no responsibility’ (0) to ‘full responsibility’ (10).
Welfare state critique is also measured by three items. First, trust in the social security system is measured on a five-point scale ranging from ‘very little confidence’ (1) to ‘a great deal of confidence’ (5). Responses were recoded so that high scores indicate high levels of distrust. Second, perceptions of benefit overuse are measured by how often respondents think it occurs that people: ‘Use their health insurance although they are not sick’, ‘Receive unemployment benefits although they could have a job if they wanted’, and ‘Receive a minimum income although they are not actually poor’. Responses range from ‘very often’ (1) to ‘never’ (5). A latent variable was constructed with a higher score indicating greater perceptions of abuse. Lastly, attitudes to the economic outcomes of the welfare state are measured by a latent variable with three items: ‘The welfare state costs too much money compared with what it yields’, ‘The tasks of the welfare state are better left to the free market’, and ‘The welfare state costs companies too much and harms our economy’. Responses range from ‘completely disagree’ (1) to ‘completely agree’ (5).
We take into account basic social-structural variables such as age and gender (0 = male, 1 = female). Education level distinguishes between lower-secondary, higher-secondary, and tertiary education. Employment status consists of seven categories: white-collar worker, blue-collar worker, self-employed, pensioner, student, unemployed or disabled, and inactive. Sociotropic and egocentric benefits are included by citizens’ opinion about whether EU membership has brought advantages to Belgium and to people like them. Responses are recoded into ‘mainly advantages’ (1) and ‘mainly disadvantages or no opinion’ (0). European identity is measured by citizens’ attachment to Europe, ranging from ‘not at all attached’ (1) to ‘strongly attached’ (5). Since welfare attitudes are embedded in a larger political division of left-right ideology, we include left-right self-placement (0-10 scale). Responses are recoded into left (0–4), centre (5) and right (6–10).
An overview of the descriptive statistics is given in the online Appendices 3–4.
Methods
A second-order confirmatory factor analysis (CFA) is estimated that captures the common component of attitudes towards the four dimensions (see online Appendix 5).Footnote 1 The CFA model has a good fit; the RMSEA equals 0.032, the SRMR equals 0.043 and both the CFI (0.967) and TLI (0.960) are sufficiently close to 1. All first-order factor loadings have an absolute value larger than 0.48 and mostly above 0.60. This indicates that the items are sufficiently valid and are reliable indicators of the concepts they are intended to measure. The second-order factor loadings range from 0.329 (social regulations) to 1 (member state solidarity).Footnote 2
Hypotheses are tested by means of structural equation modeling, with welfare attitudes modeled as mediating variables between the social-structural control variables and support for Social Europe. The model captures the similarity of antecedents of the different dimensions of Social Europe by estimating general effects of the predictors on the second-order factor ‘Social Europe’. In addition, component-specific effects are included, by allowing significant direct effects of the predictors on the different dimensions of Social Europe (first-order factors). These direct effects indicate deviations from the common explanatory model, meaning that the impact of a certain predictor on a specific dimension of Social Europe is different from that on one's overall attitude towards Social Europe. Appendix 6 (online) shows a visual representation of the estimated model. All analyses are performed using Mplus software version 7.3 and weighted by age, gender and education level. Item non-response is addressed using full information maximum likelihood estimation.
Results
Spillover or obstacle? Common patterns
Table 1 displays the standardised direct effects of the structural equation model. These include the common impact of the predictors on citizens’ overall support for Social Europe as well as the differential effects on the four specific attitudinal dimensions of Social Europe. The total effects of the predictors on attitudes toward Social Europe are reported in Appendix 7 (online). These are the sum of the direct effects shown in Table 1 and the indirect effects that run through the mediating variables. Information about the two types of effects (direct and total) is required to gain detailed insight into the explanatory model. Whereas the direct effects are useful to reveal differential impacts of predictors, the total effects provide insight into the general patterns in our data.
TABLE 1. Direct effects of social-structural variables and welfare attitudes on support for Social Europe and its different components (standardized estimates)
Note: n = 1327; χ² = 1371.12; df = 710; RMSEA = .026; CFI = .923; TLI = .905; SRMR = .036.
*** p ≤ .001; ** p ≤ .01; * p ≤ .05.
First, we look at the common effects of the social-structural variables on citizens’ overall support for Social Europe. Table 1 shows that, when welfare state attitudes are taken into account, neither age, nor gender, nor education has a significant direct effect on support for Social Europe. With regard to employment status, we observe that pensioners, students and the inactive are more in favour of Social Europe than blue-collar workers. The latter do not differ in their general attitude to Social Europe from white-collar workers, the self-employed, and welfare beneficiaries such as the unemployed and disabled. Furthermore, Table 1 shows that, if citizens perceive that their country has benefited from EU membership, they support Social Europe more strongly, whereas perceived egocentric benefits are irrelevant in this respect. People with a stronger European identity are also more supportive and right-wing voters are more opposed to Social Europe in general than those who position themselves at the centre politically.
The primary purpose of the analyses is to find out how support for the welfare state is related to support for Social Europe. To determine whether there is empirical evidence for a spillover or obstacle effect, we look at the common direct effects of the various welfare attitudes on the overarching Social Europe factor. With regard to welfare state principles, Table 1 shows that opinions concerning the welfare mix in society are not significantly related to overall support for Social Europe. Being in favour of government intervention in the economy neither increases nor decreases overall support for Social Europe significantly. By contrast, attitudes concerning social redistribution are an important predictor of citizens’ general support for Social Europe. People who are more strongly in favour of an egalitarian society are significantly more supportive of the EU's social dimension (β = .119; p = .031). Furthermore, Table 1 indicates that preferences for wide-ranging government responsibilities are unrelated to citizens’ overall support for Social Europe. In other words, people who think that the government should be responsible for providing reasonable pensions, affordable health care and a reasonable standard of living for the unemployed, do not differ in their overall attitude toward the EU's social pillar compared with those who believe that the government is not responsible for these matters at all. Turning to welfare state critique, lower levels of trust in the national social security system decrease support for European social policy (β = -.083; p = .041). Citizens who distrust their social security system do not put their trust in European social policy, as the obstacle hypothesis suggests. If people distrust their national social security, they also have more reservations about the EU's social policy. In addition, citizens who think that overuse of social benefits is common, hold much more negative attitudes toward Social Europe in general (β = -.217; p<.001). Those who believe that the effectiveness of social benefits is poor – in terms of whether they reach the ‘right’ people – are thus less willing to expand the circle of solidarity to the rest of Europe. Lastly, concerns about the negative economic consequences of the welfare state have no significant direct effect on overall support for Social Europe.
In summary, although only three out of six examined welfare state attitudes are significantly related to overall support for Social Europe, we notice a common spillover pattern. Endorsement of the basic welfare state principles is related to stronger support for Social Europe (H1a), which suggests that attitudes towards the national welfare state function as a proxy to form an opinion about Social Europe. Similarly, low levels of welfare state critique are accompanied by higher support for Social Europe (H1b), indicating that not only ideological preferences, but also institutional trust and evaluations are transferrable across policy levels.
Specific dimensions of support for Social Europe
The second purpose of this analysis is to figure out whether the relationship between welfare attitudes and support for Social Europe varies across different dimensions of Social Europe. Most of the welfare attitudes indeed have component-specific effects that deviate from the general pattern. This is indicated in Table 1 by the significant direct effects of the predictors on the specific attitudinal dimensions of Social Europe. From left to right, the four specific dimensions are ranked from less intrusive to more intrusive to the welfare state.
With regard to support for EU social regulations, the least intrusive dimension of Social Europe, Table 1 reveals several differential effects. First, we look at social-structural variables. The difference between blue-collar workers and the self-employed is larger with regard to support for EU social regulations than support for other facets of Social Europe. This can be explained by the fact that the self-employed are unlikely to benefit from these regulations, which are targeted at employees. Moreover, employers have to bear part of the costs of these regulations. The difference between the reference category of blue-collar workers and the unemployed or disabled is much larger for EU social regulations, with the latter group being more in favour of these measures. This is not illogical, since the unemployed and disabled are in higher need of social protection and the EU social regulations are most reinforcing of national welfare states compared with the other aspects of Social Europe. Turning to our main area of interest, all three welfare state principles have a differential effect on opinions about EU social regulations. Controlling for general attitudes to Social Europe, preferring state intervention above market forces is strongly positively related to support for EU social regulations (β = .238; p<.001). Obviously, aversion to government intervention obstructs the approval of active social policymaking in a regulatory way, regardless of the policy level. Egalitarian attitudes seem to be disproportionally more important regarding support for EU social regulations than for other components of Social Europe. The additional effect of egalitarianism is positive (β = .123; p = .034), indicating that citizens who are in favour of an equal society approve of the EU's social regulations even more than we would expect given their general level of support for Social Europe. Furthermore, preferring wide-ranging government responsibilities for welfare has a direct, positive effect on support for EU social regulations (β = .191; p<.001). This means that even when we take into account citizens’ general score on the Social Europe factor, those who think that the state should provide reasonable pensions, unemployment benefits and affordable health care are significantly more in favour of the EU's regulatory actions in the social domain.
Second, several differential effects are found concerning attitudes toward European social citizenship. Education level has a direct positive effect on support for this facet of Social Europe. The fact that the higher educated have, on average, more transnational interactions might explain why they differ so much from the lower educated with regard to their support for this particular dimension of Social Europe. European social citizenship creates new social opportunities beyond the welfare state, which are mainly advantageous to the higher educated. Furthermore, the gap between left-oriented voters and those who position themselves at the centre politically is larger concerning support for European social citizenship than for other dimensions of Social Europe. With regard to welfare state principles, a negative differential effect is observed for egalitarianism (β = -.107; p = .009). This indicates that egalitarianism has a weaker impact on support for the Europeanisation of social rights compared with the other dimensions of Social Europe. In fact, the total effects shown in Appendix 7 (online) indicate that, overall, egalitarian views are not significantly related to support for European social citizenship (β = -.050; p>.05). The opening up of national social security to mobile EU citizens might be perceived as most threatening to citizens’ own social position, which could explain why egalitarian views do not result in stronger support for equal social rights for EU citizens. Second, two components of welfare state critique have direct effects. While perceptions of benefit overuse have a strong negative impact on support for Social Europe in general, support for European social citizenship is even more strongly related to perceptions of benefit overuse (β = -.273; p<.001). This means that voters who believe welfare overuse is very common are even more sceptical about granting social rights to EU citizens than one would expect given their overall score on the Social Europe factor. In other words, if the overuse of social benefits is perceived as more frequent, this hinders support for European social citizenship disproportionally more than citizens’ support for social regulations, member state solidarity and an EU social security system. Further, controlling for citizens’ general attitude toward Social Europe, Table 1 shows a direct negative effect of concerns about the economic consequences of the welfare state on support for European social citizenship (β = -.092; p = .035). People who believe that the welfare state is harmful to the economy are more negative with regard to granting EU citizens access to their welfare state. The logic behind this might be the belief that the number of beneficiaries of social protection schemes should not be expanded, in order to restrain the negative consequences for the welfare state.
No differential effects are found for member state solidarity. As the confirmatory factor analysis shows (see Appendix 5 online), attitudes toward member state solidarity coincide completely with citizens’ general disposition toward Social Europe. This explains why antecedents cannot have differential effects on member state solidarity. With regard to support for a European social security system, we find that attachment to Europe has a positive differential effect. Stronger identification with Europe increases support for this policy instrument of interpersonal solidarity even more than it influences citizens’ overall level of support for Social Europe. This finding indicates that identity is an essential element of public support for transferring more social competences to the European level. Contrary to our expectations, we find no direct effects of welfare attitudes on support for a European social security system, which is the most intrusive to national welfare states.
Our results indicate that the spillover effect of welfare attitudes from the national to the European level is not uniform for all facets of Social Europe. In addition to a set of common predictors, we find specific effects of welfare attitudes that deviate from the general pattern, confirming both hypotheses H3 and H4. More specifically, the spillover effect of welfare state principles is strongest for the dimension that is least intrusive to the national welfare state: namely, EU social regulations. For example, the impact of egalitarian views is much stronger on support for EU social regulations than on support for more intrusive dimensions of Social Europe. For European social citizenship, the spillover effect seems to have weakened in favour of the obstacle effect. By contrast, the spillover effects of different types of welfare state critique are strongest for European social citizenship. These findings indicate that distrust and concerns regarding the effectiveness and consequences of welfare policies easily spill over to the European policy level, whereas support for basic principles such as government intervention and redistribution tend to be more nationally demarcated. The spillover effect depends not only on the dimension of Social Europe, but also on the type of welfare attitudes.
Discussion
Although the role of the EU and the scope of its competences in social policy are contested, the attitudes of citizens regarding Social Europe have received little previous scholarly attention. Given the strong historical link between welfare policies and the national state, we analyse whether public support for the national welfare state spills over to or obstructs support for Social Europe. Our results provide evidence for the spillover effect, as citizens holding more positive welfare attitudes are more in favour of Social Europe than those generally less supportive of the welfare state. This indicates that citizens recognise that Social Europe ultimately has very similar objectives to those of national welfare states. It should be noted that Belgians already enjoy a relatively high level of social protection and thus that their opportunity cost of transferring competences to the European level is higher compared to citizens in less-advanced welfare states. This suggests that the spillover effect might be even stronger in other EU member states. The fact that support for the welfare state facilitates support for Social Europe instead of being an obstacle – even in a country with an advanced welfare state – points to a window of opportunity for European policymakers to implement a more ambitious social agenda. In established Western-European welfare states, European policymakers proposing a deepening of the EU's social dimension should not fear resistance from welfare state advocates. Instead, they will find a social basis for it among those who prefer strong welfare states and who praise their national social security system to a larger extent. Citizens who support social protection and welfare redistribution in general also see a role for the EU in these areas. Opposition to the growth of Social Europe will mainly be articulated by citizens who hold more negative views about the welfare state.
Our analysis also shows, however, that creating public support for European social policy is more challenging in the case of policy instruments that are more intrusive to the national welfare state. The spillover effect of support for welfare state principles is strongest for European social policy instruments that leave the boundaries of the national welfare state untouched, and merely extend social protection along existing lines. Advocates of the welfare state support European regulations on health and safety at work very strongly, whereas their enthusiasm for instruments that imply redistribution between member states or EU citizens is somewhat more moderate. This indicates that citizens in favour of welfare redistribution especially welcome policies that support national welfare states in their key functions, without reshuffling the boundaries of welfare too profoundly. In addition, our analysis indicates that critical evaluations of the performance and consequences of the national welfare system have a stronger detrimental impact on support for European social policy that blurs the boundaries of national welfare states, namely equal social rights for mobile EU citizens. However, our expectations regarding support for a European social security system – the most intrusive to national welfare states – are not confirmed. Future studies should analyse whether this is due to measurement error or whether citizens perceive it as less intrusive than a European social citizenship. Nevertheless, our findings indicate that more-intrusive policy instruments of Social Europe might not be primarily rejected on the basis of Europe's social engagement as such, but seem to be related to reservations about the practical feasibility, negative side effects or potential abuse. For instance, if citizens perceive that social benefits are mistargeted, they are less willing to open up the boundaries of solidarity by granting EU citizens access to their welfare system. Although the less-intrusive policy instruments of Social Europe might seem a safer option to avoid public contestation, they do not tend to contribute much to the visibility of Social Europe in advanced welfare states. An important task for both European and national policymakers is therefore to inform citizens about how the EU engages in social policy and how it could protect ordinary citizens. Furthermore, this study illustrates that individual variation in support for Social Europe is not only explained by welfare attitudes but also by European identity and citizens’ perceptions of their country's benefits of EU membership. Social Europe thus evokes a very diverse set of attitudes that cannot be reduced to welfare issues alone.
It should be noted that the scope of welfare attitudes included in this study is not exhaustive. For instance, we did not include accurate measurements of satisfaction with the coverage or quality of national welfare provisions. If citizens think that their national welfare system does not provide enough protection, they might be more supportive of EU social policy. Furthermore, the positive relationship between support for the principles of the welfare state and support for Social Europe might even be stronger in less-advanced national welfare states. In these countries, citizens are protected to a lesser extent by their welfare policies, which can trigger high expectations about Social Europe among left-wing voters. Future research is needed to examine whether the strength of the spillover effect varies across EU member states.
Acknowledgements
This study was made possible by grants from KU Leuven Research Council (OT/13/30), the Research Foundation FWO – Flanders (Grant number G068816 N) and the Belgian National Lottery.
Supplementary material
To view supplementary material for this article, please visit https://doi.org/10.1017/S0047279418000314
