![Sample our Politics & International Relations journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5947/TOC-Subject-Sample-2021-Politics-International-Relations.jpg"})
Abstract
Is the public EU membership evaluation similar to the support for further integration or different? Starting from the assumption that the EU is – compared to stable nation states – still a ‘polity in the making’, the paper answers the question which factors can explain the desired speed of integration. I first hypothesise that the standard economic, political and cultural explanations have the same effect for the (prospective) desired speed of integration than they have for the (retrospective) membership evaluation analysed in previous studies. Additionally, I develop two temporal explanations specifically for the explanation of future support. The results on the basis of Eurobarometerdata from 2008 suggest that the canonical economic and political theories work only for the explanation of the membership evaluation and not for the desired speed.
Notes
1. This is in line with the finding from Hooghe and Marks Citation2004 who find a correlation of 0.409 in their analysis of the data from 2000.
2. This is due to the fact that this wave is the latest that contains all relevant indicators for this study, especially the speed indicator.
3. There are six notable researchers and research groups using the indicator as well: Sanchez-Cuenca (Citation2000) uses it as a proxy for general attitudes towards Ei, Ray (Citation2003) includes the speed dimension as a robustness check for the unification indicator, Hooghe and Marks (Citation2004) incorporate the Eurodynamometer in their index of overall public opinion on the EU, Brinegar et al. (Citation2004) use it as the dependent variable for a test of the ‘varieties of capitalism’ hypothesis and Vetik et al. use a somehow related indicator asking if candidate countries should join the EU ‘as soon as possible’. However, none of these studies is really interested in integration speed; they have rather technical reasons to use this indicator as a representation for the general evaluation of the EU. Recently, Boomgaarden et al. (Citation2011) include the dynamometer as one item of their more encompassing ‘strengthening’ dimension. Additionally, Hobolt and de Vries (Citation2016) take parts of the speed indicator in their overview article under the surprising label ‘policy support’.
4. Following the general distinction of political psychology, it can be said that the current perceived speed is the cognitive component, the desired speed is the affective component and the net desired speed is the evaluative component of the integration speed.
5. The category left out for comparison was ‘manual worker’. So the effects of all other occupations are seen in comparison to this category.
6. A note on the technical aspect: the speed indicator has the parameter values from −6 to 6, therefore the OLS conditions are fulfilled. The membership indicator has only three realisations, −1, 0, 1. Therefore, we applied an ordered logit model. Checking this against the OLS multi-level outcomes, it appeared that there were no significant differences. So, since the interest is not on the effect size but only on the direction of the effects, we finally use for both dependent variables OLS models since they are easier to compare. As a robustness check, I transformed the speed indicator into a variable with three realisations as the membership measure. Generally, the results stay robust and the model fits converge.
7. Table A1 of the appendix contains the average marginal effects for the interaction of both trust levels. Trust in national government in addition to trust in the EU lowers the membership evaluation significantly, whereas this value is insignificant for the desired speed.
8. Three control variables are included in every model; however, due to space constraints, they are not included in the results tables. They are: respondents’ age, gender and left–right placement.
9. To check the robustness of the results, a reduced model of individual-level factors is used. My selection criterion was to include all individual predictors that reached significance for the desired net speed.
10. Since the CPI is based on individual evaluation of the political performance of a country, we additionally test the robustness of my results in models 1 and 2 of Table A2. We can see that the variable governmental effectiveness as a concrete measure of political institutional performance from the World Bank indicators has the same effect.