The EU member states’ swift agreement to set up a far-reaching, albeit temporary, stabilisation capacity and to issue Eurobonds as part of the Recovery and Resilience Facility (RRF) in July 2020 took many political observers by surprise (e.g. Corsetti et al. 2020, Garicano 2020, Giavazzi and Tabellini 2020). After all, in the aftermath of the euro crisis, policymakers were unable to make significant progress with fiscal integration, which is seen by many (e.g. Jones et al. 2016) as a necessary condition for ensuring the long-term stability of the euro area.
One important obstacle to fiscal integration has been the concern of national governments about a Eurosceptic and populist backlash of their voters. Are these concerns valid? Existing research suggests that policymakers have overestimated public scepticism towards fiscal integration of the EU (Daniele and Geys 2015). Substantial support for EU-level unemployment stabilisation existed well before the pandemic (Vandenbroucke et al. 2018); fiscal integration is well received by European citizens and ‘spaces of agreement’ are indeed possible (Beetsma et al. 2020); and Germans and Italians prefer some form of joint debt emission over a breakup of the euro area (Baccaro et al. 2020). These studies show that some form of fiscal integration is met with sufficient support from European citizens, under the right circumstances and when appropriately designed.
However, the political context of the pandemic as well as the design of the RFF fund will likely affect citizens’ preferences for fiscal integration and risk-sharing. As such, existing research is not well placed to explain citizens’ preferences towards the specific forms of the pandemic recovery fund. It remains to be seen what Europeans think about the pandemic recovery fund and which policy characteristics drive support and opposition.
Public support for a European pandemic recovery fund
To address these questions, we conducted a new experimental survey on public support for different designs of a European pandemic recovery fund as they were discussed by policymakers in July 2020. We fielded the survey in five European countries (France, Germany, Italy, the Netherlands, and Spain) immediately before and after the decisive European Council summit. The sample in each country is nationally representative in terms of age, gender, education, regional distribution, and – to a slightly lesser extent – income and profession.
In our conjoint survey experiment, respondents were asked to compare and evaluate possible alternative programme scenarios for a European pandemic recovery fund. Respondents were presented with two such policy alternatives side by side, and asked to indicate their preferred scenario.
The content of these policy packages was randomised across a number of dimensions: (i) purpose of the programme, (ii) sources of financing, (iii) solidarity in repayment, (iv) distribution of the funds, (v) duration of the programme, and (vi) governance of the programme. While some of these features may sound technical and abstract, we put great care in ensuring that they were presented in the most simplified and concise way possible, adopting a language that is easy to understand.
To assess the average effect of each option on a given policy dimension on support for a programme alternative, we compute the average marginal component effect (AMCE) of each option on decreasing or increasing support (Figure 1). In other words, everything else being equal, the values in Figure 1 tell us to what extent a certain option increases or decreases support. We show the variation across countries in Figure 2, which displays the predicted probability to support a pandemic recovery fund scenario if it contains the policy option in question.
Figure 1 Estimated AMCEs of different dimensions of the pandemic recovery fund
Note: The figure shows the average component-specific marginal effect (ACME) of a change in the value of one of the six dimensions on the probability that respondents choose a package. Responses from the full sample are included and 95% confidence intervals are shown. Respondents saw three iterations of the experiment. N = 45,006.
Figure 2 Estimated marginal means of different dimensions of the pandemic recovery fund
Note: The figure shows the conditional marginal means (MMs) and 95% confidence intervals for all levels of the attributes by country. MMs denote the predicted probability of support for package if it contains the level in question, irrespective of all other levels. As such, MMs display the level of favourability toward policy packages that share a particular policy level. N = 45,006.
Figures 1 and 2 show that in terms of the purpose of the RRF, citizens have a strong preference for accompanying healthcare measures with an additional goal: If an RRF policy package also aims at economic recovery, public support increases by 11 percentage points compared to the baseline. The ‘green deal’ version is also appreciated: if a pandemic recovery fund also aims at tackling climate change, support increases by seven percentage points with respect to a programme directed at supporting health care systems only. When it comes to financing, respondents are somewhat sceptical about the idea of joint European borrowing: on average, it decreases support by about four percentage points with respect to the alternative of repurposing existing resources. Importantly, this effect is negative in all countries, including Spain and Italy, but the magnitude of this effect is small. Similarly, the average effect of joint repayment (through grants) is also slightly negative and decreases support by three percentage points. On this dimension, however, differences between countries are starker, with Spaniards and Italians strongly supporting grants, while citizens of other countries favour loans (see Figure 2). On all other dimensions, public opinion in all countries is far more similar and there is a clear pattern: citizens prefer distribution on the basis of financial need rather than on size, permanent duration rather than temporary, and Council decision-making rather than Commission or autonomous national decisions.
While Figures 1 and 2 are informative about how individual design choices impact public opinion, they only tell part of the story. In particular, they do not reveal to what extent specific combinations of policy options find public support. Using respondents’ ratings of each RRF scenario as an additional dependent variable, we present the aggregate levels of support for four alternative packages in Figure 3: the least-liked package; the most-liked package; the RRF as it was agreed in the Next Generation EU programme; and a permanent version of the RRF, which we label ‘Permanent Next Generation EU’ (see Table 1).
Table 1 Scenario composition of three selected European pandemic recovery fund scenarios
Note: The Next Generation EU scenario is operationalised here as being administered by the European Commission. In practice, the Council can request a ‘brake’ on the Commission’s assessment, which the Commission is unlikely to ignore.
Here, we provide two estimates: a conservative estimate that counts neutral views of a recovery fund scenario as opposition, and the standard view in which we only include those who have negative or positive views of a package (excluding those who had expressed neutral views). Support for the least-liked package varies from a minimum of 33% of respondents (in the conservative estimate) to a maximum of 56%. Support for the most liked package varies from a minimum of 57% in the conservative estimate to well over 80%. The picture for the actual RRF agreement (measured as a scenario with healthcare and economy focus, joint debt, grants rather than loans, distribution dependent on need, temporary duration, and Commission governance) is less clear-cut, with support varying from 47% in the conservative estimate to about 69%. However, if policy-makers were to move towards a permanent RRF, this would render the agreement certainly politically sustainable, with support varying from a minimum of 50% to about 72%.
Figure 3 Estimated levels of support for different versions of the pandemic recovery fund
Note: The figure shows the predicted level of support for four selected packages; 95% confidence levels are shown. Respondents were asked to rate packages on a scale from one to five. The ‘conservative estimate’ counts people who chose three as opposed to the package; the other model drops all respondents who chose three.
The road from the euro crisis to the EU’s first attempt at creating a fiscal capacity has been a long one. Importantly, we find that support for a joint European fiscal instrument is high and that the RRF agreed under ‘Next Generation EU’ is a well-tailored instrument for the purpose, at least according to citizens in the five countries studied. While citizens are not enthusiastic about common debt, this issue is not important enough for citizens to negatively evaluate the programme as a whole. Transforming the RRF into a permanent fund to deal with future crises would provide a clear pathway to even further political sustainability – ensuring widespread political support among European citizens.
Baccaro, L, B Bremer and E Neimanns (2020), “Everyone Thinks that Germans Oppose "Coronabonds": Our Research Shows How They’re Wrong”, Washington Post Blog Monkey Cage, 20 April.
Beetsma, R, B Burgoon, F Nicoli, A de Ruijter and F Vandenbroucke (2020), “After polarisation: Stable consensus on European mutual assistance is possible”, VoxEU.org, 21 August.
Daniele, G and B Geys (2015), “Public support for European fiscal integration in times of crisis”, Journal of European Public Policy, 22(5): 650–670.
Corsetti, G, A Erce and A Garcia Pascual (2020), “Perpetual bonds are not the best way to finance the European Recovery Fund”, VoxEU.org, 14 May.
Garicano, L (2020), “Towards a European Reconstruction Fund”, VoxEU.org, 5 May.
Giavazzi, F and G Tabellini (2020), “Covid Perpetual Eurobonds: Jointly guaranteed and supported by the ECB”, VoxEU.org, 24 March.
Jones, E, R D Kelemen and S Meunier (2016), “Failing Forward? The Euro Crisis and the Incomplete Nature of European Integration”, Comparative Political Studies, 49(7): 1010–1034.
Vandenbroucke, F, B Burgoon, T Kuhn, F Nicoli, S Sacchi, D Van Der Duin and S Hegelwald (2018), “Risk Sharing when Unemployment Hits: How Policy Design Influences Public support For European Unemployment Insurance”, University of Amsterdam: AISSR Policy Report Nr 1.