Framing the Eurozone crisis in national parliaments: is the economic cleavage really declining?

ABSTRACT

While it is undisputed that the Eurozone crisis has contributed to the politicization of European integration, the longer-term impact of this politicization on the structure of political conflict in Europe remains unclear. This article engages with research findings which argue that the crisis has contributed to a shift of political cleavages in Europe, from an economic (left vs. right) to a transnational (pro- vs. anti-EU) divide. We examine whether there is any evidence of such a shift in parliamentary debates about the crisis in four Eurozone states (Germany, Austria, Spain and Ireland) between 2009 and 2014. We use a combination of content and cluster analysis to identify the discursive frames that parliamentarians employed to make sense of the crisis, and then assess which factors affected how these frames were used. Our findings show that the economic (left-right) cleavage remained highly influential in shaping the four parliaments’ crisis discourse.

KEYWORDS:

• Eurozone crisis
• politicization
• political cleavages
• national parliaments
• political discourse
• frame analysis

Acknowledgments

Previous versions of this article were presented at the European Community Studies Association-Canada (ECSA-C) Biennial Conference in Toronto in May 2018 and at the Université de Montréal in November 2018. We are grateful to Laurie Beaudonnet, Scott Edward Bennett, Thomas Lehner, Amy Verdun, and two anonymous reviewers for constructive comments.

Disclosure statement

No potential conflict of interest was reported by the authors.

Supplementary material

Supplemental data for this article can be accessed here A replication dataset is available at 10.5683/SP2/9JHYPJ.

Notes

1. Our approach differs from other recent studies of Eurozone-related debates in national parliaments, which focus on only one of the three frame elements. For instance, Wendler (2016, 115–129) and Wonka (2016) analyze evaluative standards used in parliamentary discourse, distinguishing various resource/output-oriented and normative arguments, while Maatsch (2014) concentrates on neoliberal and Keynesian solutions proposed for the crisis. Closa and Maatsch (2014) link the pragmatic, ethical and normative justifications used by parliamentary parties to their support (or lack thereof) for Eurozone rescue measures. Kinski (2018) examines the definition of the constituency – national or European – in whose name parliamentarians claim to speak.

2. The material was coded by the four authors of this article. An intercoder reliability test was conducted using a random selection of twenty contributions; it resulted in acceptable levels of reliability (Krippendorff’s alpha of 0.711 for diagnosis, 0.734 for evaluation, and 0.713 for prescription).

3. More detailed information on the aggregation process, including a list of all individual combinations of arguments within each frame element, is presented in Online Appendix A2.

4. Auto-clustering statistics are available in Online Appendix A3.

5. It is noteworthy that the number of outliers is higher in Germany and Austria than in Spain and Ireland. This finding proves consistent across the various cluster solutions we calculated, including ones with a higher number of clusters. We hence interpret it as evidence of more idiosyncratic discourse patterns in these countries, rather than as an indication of an undefined cluster that we have failed to identify.

6. The outlier cluster is an appropriate reference category for this analysis since it constitutes a ‘neutral’ category of all those speeches which cannot be allocated to any of the three frames.

7. We use a generalized estimating equation (GEE) approach – calculated through the complex samples module in SPSS – to address the fact that some parliamentarians are represented more than once in the sample. A diagnostic examination indicates an acceptable model. Both the model as a whole and each of our four independent variables are statistically significant at least at the 0.01 level. Collinearity diagnostics were satisfactory. The full list of parameter estimates, as well as information on sample design, variables and model effects, is available in Online Appendix A4.

8. As an alternative indicator for non-economic political cleavages, we also tested the GAL-TAN variable in the Chapel Hill Expert Survey; however, its effects on frame usage were not statistically significant.

9.. For ‘build a better Europe’, the predicted probabilities correlate more strongly with positions on European integration. A graph that displays this correlation is available in Online Appendix A5.

Additional information

Funding

Research for this article has been supported by an Insight Grant from the Social Sciences and Humanities Research Council of Canada [Project No. 435-2013-1813] and by a Mitacs Globalink Research Internship.