The figure shows how we matched the information from the ESS and CHES to the different cabinets that were in place during the period of investigation. Coloured rectangles represent the duration of specific cabinets. All cabinets included in the analysis are labelled by the name of the prime minister. Different colours identify the specific CHES wave from which positions were attributed to the respective cabinet. Grey rectangles identify the duration of ESS field periods, that is, when the interviews were conducted.
Country | 2006 | 2008 | 2010 | 2012 | 2014 | 2016 |
Austria | 2405 | 1795 | 2010 | |||
Belgium | 1798 | 1760 | 1704 | 1869 | 518 | |
Bulgaria | 1400 | 2230 | 2434 | 2260 | ||
Denmark | 1504 | 1610 | 1576 | 1650 | 1502 | |
Finland | 1896 | 2195 | 1878 | 2197 | 2087 | 1925 |
France | 1986 | 2073 | 1728 | 1968 | 1917 | 2070 |
Germany | 2916 | 2751 | 3031 | 2958 | 3045 | 2852 |
Greece | 2072 | 2715 | ||||
Hungary | 1518 | 1544 | 1561 | 2014 | 1698 | 1614 |
Ireland | 1772 | 1764 | 2576 | 2628 | 2390 | 2757 |
Netherlands | 1889 | 1778 | 1829 | 1845 | 1919 | 1681 |
Poland | 1721 | 1619 | 1751 | 1898 | 1615 | 1694 |
Portugal | 2222 | 2367 | 2146 | 2151 | 1238 | |
Spain | 1876 | 2576 | 1885 | 1889 | 1922 | 1958 |
Sweden | 1926 | 1830 | 1497 | 1847 | 953 | |
United Kingdom | 2394 | 2352 | 2422 | 2286 | 2261 | 1958 |
The figure shows the mean incongruence across the different issues in all countries (with 95% confidence intervals). It can be seen that the means are signficantly different. It is also obvious that left/right exhibits the lowest degree of incongruence across all countries, while European Integration and - less often - redistribution is the issue with the highest mean incongruence.
The following figure shows the mean incongruence across all citizens for each individual issue. Obviously, incongruences on an left/right-axis are smallest while European integration exhibits the largest incongruences.
The figure replicates figure 3 from the manuscript in order to ease comparison between congruence measured on the basis of data from the European Social Survey and the European Election Study (see below).
The figure uses the data from the European Election Study and replicates the models based on ESS data (see figure 3 in the manuscript and above). Individual socioeconomic variables in the EES differ from the ESS. Education is measured as a categorical variable of the respondent’s age when he or she finished education: 15 or younger / No full-time education (= low), 16-19 (= medium), 20 or older / still studying (= high). There is no income variable in the EES. We therefore used a question with which respondent’s were asked to sort themselves into a high, medium or low social level.
The results corroborate the impression from the analyses using ESS data. Overall congruence and representational biases are somewhat lower on the left-right axis. Representational biases on European integration and redistribution tend to favor the better educated and particularly people from a higher social level.
In the following we present detailed results for each country studied in this paper. We first present information on the distribution of citizen preferences (from ESS and EES) and governments (Figure 1 in the manuscript). The plots also contain the mean preferences of the socioeconomic subgroups (people with high/low income/education) as measured by the ESS.
The regression tables and the coefficient plot present the results that underly Figure 3 in the text.