WP 1 investigates how cross-border mobility has affected EU support during the COVID-19 pandemic. During the first wave of the pandemic and also partly during the second wave, governments have temporarily closed entire borders, they have imposed travel restrictions by declaring regions as high-risk zones and they enforced quarantines after traveling to other countries. This resulted in a dramatic decline of cross-border mobility.

Empirically, we test the effect of declining cross-border mobility in three steps. First, we conduct a cross-national analysis of the effect of cross-border mobility on EU support. Second, we rely on a difference-in-difference design that exploits variation in the closing of Germany’s borders with neighboring countries across regions and over time to estimate the causal effect of closed borders. We concentrate on border regions as they have been most heavily influenced by the closure of inner-European borders.

WP 2 examines how EU policy measures addressing the Corona crisis have affected EU support. First, WP 2 studies how the EU’s vaccination roll-out (and the first UK vaccine) has affected EU support.

Second, using data such as the COVID-19 economic stimulus index, or the COVID-19 Financial Response Tracker, we will create a dataset capturing funds funneled to regions in response to the pandemic, mainly through the NextGenerationEU Programme and rely on a differences-in-differences design to test the effect of these funds on EU support.

Finally, WP 2 fields a vignette experiment in order to identify policy measures which can help to overcome the economic crisis and at the same time also help to increase EU support.

WP 3 investigates how the political communication of governments has affected EU support. Throughout the pandemic, the EU is regularly used as a scapegoat for national failures.

Based on the cue-taking approach (Wessels 1995; Anderson 1998), we argue that the way governments communicate about the EU and other member states has an important effect on EU support. We will test our argument based on an innovative two step design consisting of a natural language processing analysis coupled with individual-level survey data and a survey experiment.

WP4 examines the role of Eurosceptic challengers in exploiting the pandemic to mobilize against the EU. Based on the cueing and the benchmarking approach (Anderson 1998; De Vries 2018), we expect that populist parties seek to shift the blame for the consequences of the pandemic to the EU. We furthermore argue that populist and Eurosceptic parties exploit the crisis and the unrest among citizens by mobilizing protests against the government and the EU which they blame as a corrupt elite thereby increasing Euroscepticism and reaping electoral gains.

By examining legislative debates in the five selected countries (Germany, the Netherlands, Poland, Spain, UK) through natural language processing techniques, we measure blame attribution during the crisis. We complement this analysis with a vignette experiment that randomly assigns different information about responsibility for the crisis, varying the actor responsible (national government/EU/none) and the source of the information (government/challenger/expert) and examine the effect on EU support.

WP 5 examines how media framing has affected EU support. First, building on the cue-taking approach (Wessels 1995; Anderson 1998; Foos & Bischof 2020), we argue that the way that media outlets have framed the role of the EU and other European countries, for example in containing the virus and in managing the vaccine roll-out, has an important impact on EU support and solidarity within the EU. We empirically test our argument based on a multi-method research design.

First, we compile a text corpus consisting of newspaper articles released by major newspapers in Germany, Spain, the Netherlands, the UK and Poland, and we use natural language processing to map the media discourse in each country during the crisis.

Second, to shed further light on the causal effect of media framing, we will afterwards conduct survey experiments in the selected countries in which we will randomly expose respondents to a) positive or negative media statements about the EU and other European countries and b) to statements that frame the crisis as a European vs. a national problem. To assure external validity of our experimental treatments, we will extract real media coverage from our media corpus.

WP 6 investigates the effect of fake news concerning COVID-19 on EU support. Building on the cuetaking approach (Wessels 1995; Anderson 1998; Foos & Bischof 2020), we argue that fake news that portray the EU negatively can fuel Euroscepticism. There is ample evidence that a large number of news pieces ranging from inaccurate to absurd gained high popularity during the crisis. These fake news may trigger negative emotions, notably anger, and affect political preferences and actions (Valentino et al. 2011).

To investigate the effect of fake news on EU support, we employ observational and experimental methods. We investigate the links between the contents of these news, their sources and their (social media) reach using amongst others natural language processing technologies. Second, we conduct a series of field experiments with Facebook users.