Representation through parties? Environmental attitudes and party stances in Europe in 2013

Abstract

Public demand for environmental policies can only influence policies when political parties incorporate environmentalism into their platforms. The economic downturn that began in 2009 may have caused European parties to abandon their commitment to environmentalism as they focused on solutions to the economic crisis. Particularly in Central and Eastern Europe, where commitment to the environment is not as central to party competition, public environmental concerns may have been placed on the periphery of the policy agenda. Examining this issue, this study finds that political parties, especially mainstream parties, in Western Europe represented economic and environmental preferences of voters about equally as well in 2013 as they did prior to the economic crisis. However, in Central and Eastern Europe, parties tended to be less responsive to environmental preferences, presumably because economic differences dominate political divisions in newer democracies. It is concluded that environmental interests are well protected by mainstream actors in the West, though less so in Europe’s newer democracies.

Keywords:

• political parties
• environmental policy
• Eastern Europe
• public opinion

Notes

1. Complete information about all variables, data sources and question wording are available in Appendix A.

2. We excluded parties if the ISSP data identify fewer than 20 voters.

3. A detailed discussion of the advantages and shortcomings of expert surveys can be found in Rohrschneider and Whitefield (2012). Party expert positioning of the parties appears to be more accurate than using party manifestos, and highly correlates with where party officials and voters place their own party (Dalton et al. 2011, ch. 5; Rohrschneider and Whitefield, ch. 1).

4. We see a similar stability in absolute terms: in 2013, about 88% of parties stayed within one point of where they were in 2008.

5. See Dalton and Rohrschneider (2015) for a discussion on data sources.

6. We added both 5-point items into a single measure, and then we rescaled it to range from a minimum of 0 (favor environmental protection) to a maximum of 1 (oppose).

7. Unfortunately, we do not have information about public opinion about the environment from a later time point. However, this analysis still permits us to compare how well (and which) political parties represent citizens in the environmental and economic domain, since information about popular preferences is from the same surveys. Moreover, the crisis had already begun to dominate the news by 2010 so that citizens at least partially began to adjust their views by the time the survey was conducted. Finally, our party-level data is from 2013 – several years into the economic malaise – so that these no doubt began to adjust their policies.

8. We first calculated the mean scores of party positions on the environment and the economy (based on the expert data). We then subtracted from it the mean score of voters’ policy preferences for each policy domain (which we calculated on the basis of the 2010 ISSP). We use the absolute distances between parties and voters on the environment and the economy as our dependent variables.

9. See the ISSP codebook for country-specific coding decisions.

10. Because the ISSP does not measure voters’ party identification independently of their party preference in an election, we cannot examine whether the environmental attitudes of independent voters affect party choice differently than the environmental attitudes of partisan voters. See Rohrschneider and Whitefield (2012), chapter 2, for a discussion.

11. Another possibility might have been to model the influence of contextual variables on voters’ choice through a multi-level model. However, the relatively small number of countries leads us to limit the analyses to the individual-level relationships.

12. When we drop the post-materialism indicator, the environmental predictor becomes even stronger because the two indicators overlap.

13. Cronbach’s α = 0.68 (West) and 0.73 (CEE).

14. We extensively validated this indicator with Katz and Mair’s measure (1992) [KR not in reference list] data (Rohrschneider and Whitefield 2012). The summary indicator based on Katz/Mair and the Rohrschneider/Whitefield summary indicator are reasonably strongly linked (r = 0.63; N = 29), despite the fact that there is at least a 20-year time lag between both studies, and the very different modes of collecting information.