VoxEU Column COVID-19 EU policies Europe's nations and regions

After polarisation: Stable consensus on European mutual assistance is possible

Building a large and durable consensus for mutual assistance policies in the EU is challenging. Even in times of crisis, member states express different preferences, and policies must reckon with democratic politics. This column presents evidence from a randomised survey to assess support for various EU budgetary assistance packages across five member states. A majority of packages are supported in all countries, although individual design features have significant effects on public approval. Importantly, it is possible to design packages such that they obtain majority support across all sampled countries, a key condition for success with policies of this kind.

On 21 July 2020, after some 90 hours of negotiations spread over five days, the 27 leaders of the EU member states reached an agreement on a Recovery Fund of €750 billion to support countries to restore the devastation to their economies caused by the coronavirus pandemic. These negotiations followed on a political debate that was at times very polarising, with some bitter exchanges between ‘the frugal four’ (the Netherlands, Denmark, Sweden, and Austria) and member states severely hit in an early stage of the crisis, such as Italy and Spain. The program resulting from the agreement is temporary (albeit stretched out over a couple of years), even though new crises might emerge in the future. Hence, it is important to assess on what basis a durable and sufficiently large consensus across Europe’s national electorates can be built to sustain policies of mutual assistance. Buti and Carnot (2018), among others, make a strong case for such a capacity. The case for assistance, however, needs to reckon with democratic politics.

Democracy is about conflict: there is nothing surprising or wrong with that. The good news is that European leaders were able to compromise. However, a repetition of such polarising debates, which consume a lot of political capital and impair mutual trust between both leaders and electorates, would not be a good scenario, neither politically nor economically. Empirical research on citizens’ attitudes towards the establishment of an EU fiscal capacity shows that a basis for a sufficiently large pan-European consensus on the shape of mutual assistance policies exists, notwithstanding some significant differences in attitudes across the EU’s national electorates. Political leaders can now build on that basis.

That basis, it turns out, is conditional. Existing research on public attitudes shows, for instance, that public support for mutual assistance in Europe depends fundamentally on different factors. The nature of the crisis is important. For instance, a natural disaster triggers much more solidarity among respondents to surveys than a debt crisis. Furthermore, the policy domain that is supported is relevant. For instance, support targeted at medical countermeasures is more popular than financial assistance that is not earmarked. Also, conditions attached to the mutual support turn out to strongly undergird public support in a number of surveys. In general, attitude research shows that ‘European solidarity’ can carry significant support among citizens, but such support critically depends on the exact design of the policies at hand (Koos and Leuffen 2020, Genschel and Hemerijck 2018, Bremer et al. 2020, Gerhards et al. 2020, Vandenbroucke et al. 2018). 

While these and other studies have explored popular preferences on a range of EU-level policies, there is no analysis yet on detailed preferences towards alternative designs of an EU central fiscal capacity. In Beetsma et al. (2020), we address this gap by means of a randomised survey experiment fielded in March 2020 to 10,000 representative respondents from France, Germany, Italy, the Netherlands, and Spain, examining their support for EU budgetary assistance packages for countries in economic distress. We provide respondents with two possible frames, building upon a distinction between temporary and permanent shocks, which is conceptually important in the context of the optimum currency area theory (De Grauwe 2018). In one, the distress is temporary, which would typically be the result of a dip in the business cycle. The other framing describes a permanent negative shock, which would, for example, result from a permanent decline in an important economic sector. We choose these frames not only because of the empirical relevance of their distinction, but also because they may call for different policy responses that may command different degrees of popular support. In particular, offering fiscal room to address a permanent negative shock might be perceived as generating long-term redistribution – loved or loathed by publics.

We provide respondents with randomly selected proposals for assistance packages that differ with regard to (i) whether there need to be conditions on the support, (ii) how the resources are to be spent, (iii) how domestic taxation will be impacted, (iv) whether long-run redistribution among countries is tolerated or even an objective, (v) what role the European Commission should have, and (vi) whether and how countries should be punished in the case of non-compliance with the program’s conditions. Respondents have to choose between pairs of packages, and they have to indicate the extent to which they support each package. In total, 60,000 packages are rated. The most important advantage of such a conjoint setup is that it allows for causal inference of ‘treatment’ effects, resulting from varying policy packages along one or more of the aforementioned dimensions, on preferences.

Figure 1 shows by country, and overall, the fraction of packages that are ‘supported’, i.e. for which respondents indicate that they are ‘somewhat in favour’ or ‘strongly in favour’. The dark bars show the fraction when neutral judgments are counted as opposition to the package, leading to a very conservative assessment of support, while the light bars leave out the neutral judgment. In the latter case, a majority of the packages are supported in each country, in spite of the fact that these are randomly selected across all possible designs. Hence, a carefully selected design might earn substantially more public support. There are cross-country differences in support for an assistance program, which is on average highest for Spanish respondents and, depending on the measure of support, lowest for French or Dutch respondents.

Figure 1 Fraction of packages supported, pooled and by country

Suggestive leads to policymakers seeking a ‘suitable package’ are found in Figure 2, which shows the estimates of the so-called ‘average marginal component effects’ (AMCE). They indicate how individual design features of packages experimentally impact support. Concretely, compared to the baseline of ‘no budgetary conditions’, the introduction of ‘budgetary conditions’ on receiving financial assistance raises the likelihood that a package is supported by 7-8 percentage points, holding all its other features unchanged. The figure also shows the increase in the chance that in a pairwise comparison of packages one with budgetary conditions will be chosen. Compared to a package with ‘no budgetary conditions’, introducing ‘budgetary conditions’ makes a package on average about nine percentage points more likely to be chosen.

Figure 2 EU assistance program: Average marginal component effects (AMCE) plot, full sample of respondents

Note: Horizontal line pieces depict the 95% confidence intervals.

Respondents prefer to condition assistance on recipients reducing their debt in normal times. There is also substantial support for imposing conditions on how assistance should be spent. In particular, requiring that assistance be earmarked to healthcare spending raises the likelihood that a package is supported by 11-12 percentage points, while the corresponding number for requiring education earmarking is around six percentage points.1 In contrast, there is little support for spending on the banking system and deposit holders. There is support for an active role of the European Commission in terms of monitoring the implementation of the programs and providing guidance. However, the support for terminating a program and imposing a fine in the face of non-compliance is small. In other words, while conditionality and guidance towards effective policies are deemed important, respondents generally do not prefer a very punitive approach. Some long-run redistribution among countries resulting from the assistance program, and in particular to poor countries, is acceptable or even preferred. This is an important observation, because it is extremely difficult to design ‘pure risk sharing’ programs, i.e. programs that only share shocks but do not lead to ex-ante redistribution (this is difficult partly because of how hard it is to distinguish temporary and permanent economic shocks). Finally, financing the program through progressive taxes is preferred to financing through a flat tax increase for everyone. Somewhat surprisingly (not visible from Figure 2), our respondents’ preferences are essentially the same in both the temporary and permanent shocks scenarios. This may reflect the (overly) subtle difference in experimental framing, but it may also suggest that respondents are more concerned with the allocation of assistance than with its effects on stabilisation or redistribution.

In terms of the individual dimensions, the Dutch stand out (not shown in Figure 2). They are the only population against any cross-border long-run redistribution, and they are the only ones supporting termination and imposing fines in response to non-compliance with the program’s conditions. The Italian respondents are the only ones not strictly favouring budgetary conditions for financial assistance. These differences between the Italian and Dutch respondents in the survey appear to mirror similar differences in the positions by their respective governments in EU-level discussions on how to respond to the Covid-19 crisis. The preferences of the other countries’ respondents appear to reflect the more ‘middle-ground’ positions of their governments.

Still, there is overall substantial congruence among the preferences of the different populations. The key question is whether this opens the possibility of finding assistance packages that get majority support from all individual countries. A closer look at our results suggests that this is possible. Table 1 reports the most supported package of each individual country, as well as the support that these packages get in other countries. We observe that the package preferred by Germany and Spain can count on more than 50% support in each sample country. This package combines budgetary conditions, mandatory healthcare spending, monitoring and guidance by the Commission, redistribution to poor countries, progressive taxation and no termination, and fines following non-compliance. Replacing progressive with flat taxation or earmarking assistance for spending in other areas reduces support. Importantly, the support reported in Table 1 is measured in a very conservative manner assuming that neutral answers are counted as against. We find that if we drop neutral answers, even packages with flat taxation and spending in other areas can receive more than 50% support in each sample country (not reported in the table).

Table 1 Support for individual countries’ most-supported packages

Note: DE = Germany, ES = Spain, FR = France, IT = Italy, NL = Netherlands.

We conclude from these results that, among the European electorates, there is a durable basis for support for policies of mutual assistance such as instantiated by the Recovery Fund. Obviously, one should be cautious when inferring political traction from the results of a survey. After all, our survey is based on respondent views in a (partially) pre-political environment, i.e. before any concrete policy proposals are debated by political parties that seek the edges of polarisation. Hence, our survey gives respondents the opportunity to reason and form their own opinion about the assistance package, thereby providing the best possible guarantee of expressing their true views. The true challenge for European solidarity is a challenge with regard to how our democracies function: if political parties believe they can only survive by always seeking the edges of polarisation on European issues, solidarity policies will always have a hard time. However, our experimental results suggest that if governments demonstrate the willingness to express their support for further European economic integration, they can garner the support of their populations, making concrete progress possible in this direction.


Beetsma, R, B Burgoon, F Nicoli, A de Ruijter and F Vandenbroucke (2020), “What kind of EU fiscal capacity? Evidence from a randomized survey experiment in five European countries in times of corona”, CEPR Discussion Paper No. 15094, 24 July 2020.

Bremer, B, P Genschel and M Jachtenfuchs (2020), “Juncker's Curse? Identity, Interest, and Public Support for the Integration of Core State Powers”, JCMS: Journal of Common Market Studies 58(1): 56-75.

Buti, M and N Carnot (2018), “The case for a central fiscal capacity in EMU”,, 7 December.

De Grauwe, P (2018), The Economics of Monetary Union, 12th edition, Oxford: Oxford University Press.

Genschel, P and A Hemerijck (2018), “Solidarity in Europe”, Policy Brief, School of Transnational Governance (EUI), Issue 2018/01 (May).

Gerhards, J, H Lengfeld, Z S Ignácz, F K Kley and M Priem (2020), European Solidarity in Times of Crisis: Insights from a Thirteen-Country Survey, Routledge.

Koos, S and D Leuffen (2020), “Beds or bonds? Conditional solidarity in the coronavirus crisis”, Unversität Konstanz, Policy Paper No 1, 1 July.

Vandenbroucke, F, B Burgoon, T Kuhn, F Nicoli, S Sacchi, D van der Duin and S Hegewald (2018), “Risk sharing when unemployment hits: how policy design influences citizen support for European Unemployment Risk Sharing (EURS)”, AISSR Policy Report, No. 1.


1 The traction of healthcare spending may not be surprising in view of the moment, March 2020, the experiment was fielded. However, a test experiment on a Dutch sample from November 2019 corroborates this popularity of healthcare spending.

630 Reads