Economic perceptions and attitudes towards the European Union: A survey experiment

Motivation

Do negative perceptions about the economy erode opinions about the European Union (EU)? Does this relation vary depending on each country’s experience during an economic crisis? The recent Great Recession revived the interest about the impact of economic assessments on support for the EU. This crisis has been an important milestone, not only because of its magnitude but also because it required the assistance of the EU, alongside other international institutions, in providing bailouts to some countries. The Great Recession can be considered the first EU-wide crisis that affected -with different intensities- all its members and put the stability and integrity of the European Monetary Union at risk. Building on the classical utilitarian explanations, recent studies have found that economic conditions and perceptions are still relevant to explain EU attitudes (Braun and Tausendpfund, 2014; Foster and Frieden, 2021, 2017). These studies, however, draw in almost all cases from observational evidence.

Limited research has tried to remedy this. Maier, Adam and Maier (2012) and Yordanova et al. (2020) test with experimental methods the effect of negative economic messages on EU support. While the former find that economically negative campaign messages affect citizens’ attitudes towards the EU, the latter finds no effect of negative economic messages on a hypothetical EU membership referendum. Beyond the mixed empirical results, these studies focus on experimental treatments about the general economic benefits of EU membership and not on the effects of economic crises. It is reasonable to expect that manipulating perceptions about how beneficial it is to be an EU member state affects its support. However, if we want to understand the impact of economic crises on citizens’ opinions on the EU, the key issue is whether citizens connect their general perceptions about economic performance with the EU, shaping their attitudes towards it accordingly.

Using a quasi-experimental framework, Colantone and Stanig (2018) exploits exogenous economic shocks to explain how worsened local economic conditions in the United Kingdom’s counties fueled Brexit voting in the 2016 referendum. This is a very valuable contribution that does allow us to establish a causal connection between economic conditions and support for the EU. However, this relies on political choices (voting) and has a country-specific focus. It does not account for the attitudinal mechanisms that underlie those political choices or any cross-county differences in these mechanisms. Citizens’ experiences vary considerably across countries and there might be different mechanisms by which an economic crisis can have an impact on EU attitudes.

In this article, I overcome these limitations by exploring: (a) the impact of economic perceptions on opinions about the EU with a survey experiment that manipulates economic perceptions, and (b) addressing the differences between a creditor country (Germany) and a debtor country (Spain). I find that having worse perceptions about the impact of the financial crisis erodes attitudes towards the EU. However, while in Germany worse perceptions about the economy are linked with believing that the EU is a less economically beneficial project, in Spain these perceptions are connected to worse opinions about democratic representation in the EU.

Research design

I conducted a vignette survey experiment in Spain and Germany in August 2018 for a sample of 1166 Spanish and 1147 German respondents. ¹ The rationale of the country selection is twofold. First, by selecting two of the biggest and most influential countries within the EU, we can delve into the circumstances that can place the EU project at risk. The worsening of EU opinions in countries like Germany or Spain will be more consequential for the future of the EU. Secondly, each country represents very different experiences during the Great Recession. In Germany, the economic impact of the financial crisis was much softer than in Spain. However, the debt crisis did raise a public debate on whether the EU was a worthy project. In countries like Spain, on the contrary, the economic impact of the crisis was substantially higher, and the political debate around the crisis focused on the constraints over policy-making and the limits on democratic representation that the creditor countries and EU institutions imposed, which raised large contestation. This allows to explore whether different experiences matter to explain how citizens rationalize the crisis and their EU attitudes.

In the survey, a treatment group – formed by half of the sample of each country – read a paragraph about the intensity of the Great Recession and the debt crisis in Europe and then answered to a series of questions, covering their opinions about the EU and their perception about (a) the economic benefits of membership and (b) democratic representation within it. The control group answered the same questions without having read the vignette. The treatment reads as follows:

Vignette: From 2008 on, the European Union has suffered the deepest economic and debt crisis since its creation. As a consequence of the economic crisis, there has been a sharp increase in unemployment in European countries, where unemployment rose from 6.8% in 2008 to 11% in 2013 in the whole European Union. Public debt rocketed, rising from 57% of GDP in 2007 to 86% in 2014 to finance the public deficits of European countries. As of today, the levels of public debt and unemployment have not returned to pre-crisis levels.

In the Online appendix, I show that control and treatment groups are correctly randomized across relevant covariates. In addition, to make sure that the treatment worked, a battery of manipulation check questions was run. This ensures that citizens in the treatment group are more likely to consider that the European financial crisis has been a strong negative economic shock both at the European and national level. The manipulation checks can also be found in the Online appendix. Once the treatment group has been manipulated to have worse economic perceptions, the main dependent variables are two standard questions on EU attitudes. First, individuals express their opinion about the EU on a 1 to 5 Likert scale. To the question ‘Overall, how is your opinion about the European Union?’, respondents could answer: 1 ‘Very negative’, 2 ‘Somewhat negative’, 3 ‘Neither negative nor positive’, 4 ‘Somewhat positive’ or 5 ‘Very positive’. I employ this variable as a continuous indicator. The second question asks respondents to assess the EU membership of their own country. The question wording is ‘And specifically for [Germany/Spain], do you think that being a Member State of the European Union has been…?’, and the answers follow a similar Likert scale as with the previous variable: 1 ‘Very negative for [Germany/Spain]’, 2 ‘Somewhat negative for [Germany/Spain]’, 3 ‘Neither negative nor positive for [Germany/Spain]’, 4 ‘Somewhat positive for [Germany/Spain]’ or 5 ‘Very positive for [Germany/Spain]’. ²

After accounting for the impact of crisis perceptions on EU attitudes, I assess whether the treatment modifies the variables that measure perceptions over two dimensions of EU membership. This allows us to investigate the underlying mechanisms in the relation between citizens’ perceptions of the economy and their opinions about the EU.

Economic dimension: This variable measures the degree to which citizens believe that the EU provides economic benefits with the two following statements: ‘The EU is a project that is beneficial from an economic perspective.’ and ‘From an economic perspective, it is worth for [Germans/Spaniards] to be part of the EU.’ Answers to both questions were measured in a 1–5 scale, where the responses where 1 ‘Completely disagree’, 2 ‘Somewhat disagree’, 3 ‘Neither disagree nor agree’, 4 ‘Somewhat agree’ or 5 ‘Completely agree’. I add up both variables to obtain a 0 to 10 ‘Economic dimension’ variable.

Representation dimension: This variable measures the degree to which citizens believe that the EU restricts democratic and political representation, based on the two following statements:‘The EU has restricted too much the set of policies that governments can implement’ and ‘The EU does not allow national governments to implement the policies that citizens demand’. Again, answers ranged on a 1 to 5 scale, from 1 ‘Completely disagree’ to 5 ‘Completely agree’. The ‘Representation Dimension’ variable is the sum of the responses to both statements.

Results

The main results are contained in Figures 1 and 2. I display the average treatment effects for both dependent variables for the total sample and by country. I present results with and without covariates. ³ Both figures show that exposing respondents to a vignette about the economic crisis worsens opinions about the EU. The finding is not only evident in the overall sample, but also separately in the German and Spanish samples. In both countries, we find that worse economic perceptions deteriorate attitudes towards the EU. ⁴ This provides experimental support for the claim that perceptions about the economic situation can worsen opinions about the EU. ⁵ These results have relevant implications as negative opinions about the diminishing EU support. ⁶

OPEN IN VIEWER

Figure 1.

ATEs on EU opinion.

OPEN IN VIEWER

Figure 2.

ATEs on perception of the EU good for the country.

Following the analysis of the treatment effects on attitudes towards the EU, I study the underlying motivations for the worsening of EU opinions in Germany and Spain. The Great Recession and the subsequent debt crisis had different consequences in the two countries under investigation. While in Germany the main focus was whether debtor countries would pay their debts and, therefore, whether, the benefits of being part of the EU could be hampered in Spain the main focus was on the constraints over policy-making and the limits on the democratic representation that were imposed on the country. Hence, the impact of the economic crisis could produce an erosion in positive EU opinions through different mechanisms.

I test this claim in Figure 3. I show the treatment effect on perceptions about the economic benefits of the EU in Figure 3(a). Those exposed to the information about the Great Recession at the European level worsen their perceptions about how beneficial the EU is as an economic project. This effect, however, is only significant for the German sample, and not for the Spanish one. Thus, German citizens are more likely to rationalize that the financial crisis implies that the European project is not economically beneficial. In Spain, although the sign points to the expected direction, there is no significant evidence of an erosion of the EU’s economic dimension.

OPEN IN VIEWER

Figure 3.

ATEs on two dimensions of the crisis.

These findings are in contrast to those reported in Figure 3(b), which displays the treatment of effects on perceptions of erosion of political representation for both national samples. Spanish respondents connect the economic crisis with a belief that the EU provides limits on political representation. Being a country whose financial sector was bailed out by the EU and where consecutive left-wing and right-wing governments imposed austerity against the popular will, the Great Recession is perceived as a constraint to democratic representation that reduces support for the EU. Conversely, the effect, is virtually zero for the German respondents, where the EU did not enforce policies that limited democratic representation.

Concluding remarks

From this article, we can draw the following concluding remarks. First, more negative perceptions about the economic consequences of the Great Recession erode attitudes towards the EU. Second, the implications of the crisis are different across countries. In a creditor country like Germany, the Great Recession deteriorates the perceptions of the economic benefits of EU membership. In contrast, in a debtor country like Spain, the crisis is a phenomenon that aggravates EU opinions through a different mechanism: an erosion of perceptions that the ideal of democratic representation can be fulfilled within the EU.

Although the scope of the results is limited by the external validity of survey experiments, the findings confirm previous observational research that output legitimacy, which has been one of the traditional pillars of the EU, can decrease as a result of economic perceptions. The public image of the EU worsens as a result of bad economic assessments.

This study analyses general attitudes towards the EU, whereas it cannot identify how the worsening of attitudes translates into a demand for EU reform or its direction. Therefore, it would be relevant to investigate how lower output legitimacy can eventually lead to more willingness to change the country’s status quo with regard to the EU. In addition, recent experimental research – see, among others, Bechtel, Hainmueller and Margalit (2017) or Beetsma et al. (2020) – has shown that attitudes towards the EU are multidimensional and that different individual- and country-level traits explain support for different areas of integration. In future research, it would be useful to decompose attitudes towards the EU and explore which specific dimensions of European integration (fiscal, political, etc.) are affected by a worsening of economic perceptions.

Supplemental Material

Supplemental Material

Supplemental Material