Greener governments: partisan ideologies, executive institutions, and environmental policies

ABSTRACT

Why do some governments have more environmentally friendly policies than others? Part of the answer involves governing parties’ ideological positions on environmentalism and the constraints imposed by executive institutions. Here, this party-based explanation is elaborated and tested with uniquely comparable indicators of national environmental policies for governments in 27 countries in the European Union (EU). The findings show that governments with parties that emphasized environmental protection in their manifestos are more likely to propose pro-environment policies during EU-level negotiations. However, the effect of ideology is mediated by the centralization of the national executive branch. In centralized national executives, the environmental positions of prime ministers’ parties affect policies, while in decentralized national executives, the positions of environment ministers’ parties are relevant. The findings have implications for understanding the impact of parties’ environmental positions on government policies, as well as for policy making in coalitions more generally.

KEYWORDS:

• environmental policy
• ideology
• ministerial autonomy
• European Union
• political parties

Acknowledgments

The replication code (in R) and the data set used in this paper can be accessed on the authors’ websites (www.justinleinaweaver.com or www.robertthomson.info).

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. The party-family approach may be a better starting point for other research questions. For instance, if we wished to examine how ideology affects party members’ sense of identity, then the family approach may be preferable.

2. Table A1 gives a summary of the environmental issues raised in these legislative proposals. When a legislative proposal raised both environmental and non-environmental issues, we included the environmental issues and excluded the non-environmental ones.

3. Tests to compare the informants’ judgments with information from Council and EP documentation and with the estimates of different informants support the validity and reliability of these estimates (Thomson et al. 2006, pp. 329–347). In addition, König et al. (2007) compared 31 point estimates provided by key informants from the Council with estimates from informants in the EP and found that 30 match perfectly or almost perfectly.

4. Figure A1 illustrates the distribution of the original positional variable and the three-category ordinal variable we use.

5. Party identifications taken from the European Journal of Political Research’s Political Data Yearbooks.

6. In our sample, parties of the prime ministers commit between 0% and 13% of their manifestos to environmental protection, while parties of the environment ministers vary between 0% and 44%.

7. Our measures of the ideological environmentalism of the prime-ministerial and environment-ministerial parties are not highly correlated (r = 0.34). We also perform separate analysis on the cases in which the ideological environmentalism of the prime minister’s and environment minister’s parties differ.

8. Kassim (2013) also distinguishes among systems on the basis of whether their coordination ambition is comprehensive or selective, which indicates whether coordination is confined to nationally salient issues. Since all our cases refer to controversial issues, they are nationally salient to at least some extent; we therefore do not include this second dimension.

9. The Brant Test was applied to the ordered logit version of the second model in Table 2, and this indicated that the parallel regression assumption was violated (χ² for all coefficients together of 256.24, p = 0.0).

10. The Italian FdV Green Federation represents the most ‘ideologically environmental’ party in our observations. The Italian Greens produced party manifestos that emphasized environmental protection at a consistently high level across most of the manifestos reviewed by the CMP project (up to 44% in the mid to late 1990s). Re-estimating Model 2 from Table 2 while omitting the Italian FdV Green Federation does not significantly alter our coefficient estimates.

11. We did not add a second level, that of either proposals or states, to the multilevel models in the robustness test, since we have too few observations at the proposal level to justify doing so (<30) and given the lack of independence that exists at the higher levels.

12.. Given that concern, we also reran the models in Table 2 with standard errors clustered by issue and found similar results.

13.. While it is appropriate to test the robustness of the findings to measurement error, including these measures of uncertainty involves additional assumptions, notably that there is more certainty regarding measures derived from shorter election platforms.