Euroscepticism in Italy from the Onset of the Crisis: Tired of Europe?

ABSTRACT

Attitudes towards the European Union (EU) have changed deeply in Italy: the level of support for EU membership has dramatically declined among Italian citizens and, especially after the 2018 elections, the Italian government has often been on a collision course with the EU. The article maps attitudes towards the EU in Italy, at the level of citizens, parties, and political elites. First, we provide a longitudinal analysis where we assess the scope of euroscepticism in this country. Second, we assess if recent changes in EU attitudes have been common to both elites and masses. Finally, we formulate some propositions with regard to the relationship between Italian democracy and the EU.

KEYWORDS:

• European integration
• EU
• citizens
• political parties
• elites
• public opinion
• attitudes
• survey

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1. See on this https://www.corriere.it/politica/18_gennaio_09/di-maio-non-piu-momento-uscire-dall-euro-f9941a32-f55a-11e7-b250-16cc66648122.shtml?refresh_ce-cp.

2. See on this https://www.corriere.it/politica/17_dicembre_21/elezioni-2018-salvini-sfida-maio-sull-euro-bbc3627e-e5c6-11e7-bb03-a8143f47e27e.shtml;see also http://tg.la7.it/politica/nel-centrodestra-ora-salvini-frena-uscire-dalleuro-potrebbe-costarci-caro-28-02-2018-125109.

3. In particular, the mass surveys were conducted between June and July 2016 (I wave) and between June and October 2017 (II wave). The Elite surveys were conducted between April and November (I wave) and between October 2017 and February 2018 (II wave). The mass surveys have a panel structure, with the II wave based on a sample of citizens from the first wave pool of respondents. The total number of observations (after having dropped incomplete interviews) was 74 (I wave) and 60 (II wave) for the elite sample and 2,207 (I wave) and 1,278 (II wave) for the mass sample.

4. We relied on the survey question ‘Which party do you feel closest to?’ to grasp individual party affiliations at mass level.

5. More specifically, we computed the average position of the mass sample on the same unification scale by their declared party closeness and we assigned this value to each MP elected with the same party (for this information, we rely on the 2016 mass survey, as the unification question has not been included in the 2017 mass questionnaire).

6. A 0–7 scale, which we have rescaled as 0–10, where 0 means that the party leadership strongly opposes European integration and 10 means that it is strongly favourable to European integration. The CHES 2014 data are the last available before our mass/elite survey was fielded starting from 2016.

7. We dropped the mass and the CHES positions from the model because of collinearity problems.

8. A 0–7 scale, which we have rescaled as 0–10, where 0 means that the party leadership strongly opposes the EU authority over member states’ economic and budgetary policies, and 10 means that it strongly favours it.

9. The logarithmic form was chosen under the assumption that each additional year spent in parliament has a decreasing marginal impact on the dependent variable.

10. It classifies MPs according to whether they have a primary, secondary, or university/postgraduate education.

11. A note of caution is required when interpreting these results. Our dataset contains data on MPs that belong to a set of different parties. This could lead to residuals that are not independent within the same party, something that would violate one assumption of OLS regression (Steenbergen & Jones 2002). A way to deal with this kind of data structure could be to cluster standard errors on parties, but in our case (by narrowing standard errors) this could be misleading given the limited number of clusters (i.e. parties) at our disposal. We have, nonetheless, estimated the same OLS models adjusting standard errors for (party) clustering (and using robust standard errors that provide a general test of significance that does not depend on normality and homoscedasticity). These tests did not change our main findings significantly.

Additional information

Funding

This work was funded by the EUENGAGE project of the European Union’s Horizon 2020 research and innovation programme under grant agreement No 649281 — EUENGAGE — H2020-EURO-2014-2015/H2020-EURO-SOCIETY-2014

Notes on contributors

Nicolò Conti

Nicolò Conti is Professor of Political Science at the Unitelma Sapienza University of Rome. His main research interests are in the fields of parties, political elites and public opinion. His work has appeared in journals including West European Politics, European Union Politics, Party Politics, South European Society and Politics, International Political Science Review, Acta Politica, Politics, Italian Political Science Review.

Francesco Marangoni

Francesco Marangoni is Associate Professor of Political Science at the University of Siena. His research interests include comparative government, legislative behaviour and comparative political elites. His more recent work has appeared in journals such as Parliamentary Affairs, Journal of Legislative Studies, Italian Political Science Review.

Luca Verzichelli

Luca Verzichelli is Professor of Political Science at the University of Siena. He has published extensively in the fields of comparative political institutions and political elites. His articles have appeared in journal such as European Journal of Political Research, West European Politics, South European Society and Politics, Journal of Legislative Studies, Swiss Political Science Review, Italian Political Science Review.