Drivers of national climate policy

Abstract

Patterns of national climate policy performance and their implications for the geopolitics of climate change are examined. An overview of levels of emissions performance across countries is first provided. Substantial changes in emissions trends over time are documented, notably with GHG emissions trajectories, which are shaped less and less by the developed/developing country divide. Various patterns of policy convergence and divergence in the types of policies states implement are then surveyed. Four broad types of explanation that may account for these trends are then explored: (1) variation in the institutional form of country-level governance regimes, (2) patterns of dependence on fossil fuel energy, (3) broad systemic differences among states (specifically in their population densities, carbon intensity, and per capita incomes, and (4) variations in the traditions of economic intervention by states. The article contributes to the growing body of work on comparative climate policy, and provides a first attempt at exploring the comparative politics of instrument choice. The analysis challenges the continued importance of a North–South divide for the future of climate policy, thus reinforcing a sense of the ‘new geopolitics’ of climate change. Some of the implications of the analysis for debates about the form of future international agreement on mitigation policy are also explored.

Policy relevance

The article contributes to the understanding of the variety of institutional conditions under which policy makers develop policy and thus the constraints and opportunities for the design of international agreements under these conditions.

Keywords:

• climate change mitigation
• climate change policies
• institutional frameworks
• international negotiations
• policy instruments
• political processes

Acknowledgements

We thank Philippe Descheneau and Grégory Kudish for research assistance in preparing this article, Niklas Höhne for helpful guidance in finding documentation for some of the primary data, Charlotte Streck and Martial Foucaul for extremely helpful responses on an earlier draft, and three anonymous reviewers for their helpful comments and constructive feedback.

Notes

Other quantitative indicators of climate policy performance – such as the targets established at Kyoto – are left to one side because they correlate poorly with trends in actual emissions, and are thus not the most valid indicator of an individual country's commitment to reducing GHG emissions.

When calculating emissions change, in cases where the 1990 value was missing, data were input from the earliest available year as emissions in 1990 (e.g. Belgium, Russia, Saudi Arabia).

It is worth emphasizing that these figures exclude land use, land-use change and forestry (LULUCF) emissions. Brazil has managed to radically reduce these emissions in recent years (Hochstetler & Viola, 2012), so incorporating this would make Brazil's performance even more impressive.

Estimates in regressions were based on case-wise deletion, so sample sizes vary.

As a robustness check, the average annual rate of growth in emissions over the same time periods was regressed, and nearly identical results (not reported here) were obtained.

All models were estimated with robust standard errors.

These results were cross-checked in additional models (not reported here due to space limitations) and were robust when the dependent variable was operationalized as the average annual growth rate in emissions, and when the richest 10% of countries were excluded from the sample.

We follow convention here. Countries garnering a score of 6 or greater on the Polity IV scale are generally considered democratic (Chang, Kayser, Linzer, & Rogowskiet, 2010). The Polity IV scale ranges from −10 to +10.

More recently, Bernauer & Böhmelt (2013) found a stronger explanatory effect for democratic regimes, but this was in relation to an index that combines both emissions trends and policy outputs in a single measure of climate policy performance. The earlier article by Bättig & Bernauer (2009), like ours, disaggregated these two aspects of climate policy performance.

Specifically, all regressions were re-run using an alternative operationalization of the dependent variable (average annual emissions growth) and removing the 10 richest countries from the sample. The results were nearly identical. The only substantive difference was that the negative coefficient on parliamentary regime lost significance in model 3 (restricted, post-Kyoto).

For liberal and coordinated market economies (LMEs and CMEs, respectively), data were drawn from Hall and Soskice (2001). The developmental states were coded by the authors.

This typology, now well established in the climate policy literature, notably excludes direct action by states in terms of state investment in new energy technologies or low-carbon infrastructure, which may be primarily framed as something other than climate policy (e.g. economic development). Insofar as such spending initiatives may be the predominant form of climate policy in some (e.g. developing) states, this typology may under-represent their climate policy performance.

The aim was to generate a distinction in the general category of ‘economic instruments’ by considering carbon pricing (e.g. emissions trading or carbon taxes) as distinct from other economic instruments (e.g. subsidies for renewable energy), which probably follow a different political dynamic.

A difficulty in coding policy in such systems as China stems from the fact that local governments in the country are often responsible for implementing policy set at the national level (Harrison & Kostka, forthcoming). Implementation thus varies widely from region to region (see e.g. Rommeney, 2008, p. 52), as does the reliability of the data, which makes it difficult to assess actual policy developments across the country. The typology here is therefore limited in that it misses action at lower government levels in China, and elsewhere (e.g. the Regional Greenhouse Gas Initiative and Western Climate Initiative in Canada and the US).

GLOBE refers to the ‘Global Legislators for a Balanced Environment’, an association of parliamentarians working on environmental questions. The survey was conducted by a research team based at the London School of Economics, UK.

Specifically, the following sources were used: Islamic Republic of Iran (2011), Kingdom of Saudi Arabia (2011), Government of Spain (2009), Government of Australia (2010), Ecofys/Climate Analytics (2011), and DB Climate Change Advisors (2012). These were used to provide data on the policy types for Australia, Iran, Saudi Arabia and Spain, which were not covered by the GLOBE report.

The category of BRICS is taken here to include South Africa and Mexico, in addition to the original BRIC members of Brazil, Russia, India, and China.

 The case of the EU perhaps complicates the picture here. All four CMEs with carbon pricing (Germany, France, Spain, Italy), as well as one of the two LMEs (the UK) get this through participation in the EU ETS.

Note that participation in the CDM was excluded here. Although participation by developing countries in the CDM creates a certain type of carbon price domestically, at least for industries that could envisage developing CDM projects, it is not the result of the domestic policy processes that are the focus of this article.

There are two possible interpretations here. Differences between LMEs/CMEs and developmental states may be taken as evidence that little has in fact changed. Behind the diplomatic shifts in apparent power and negotiation, as well as interesting shifts in emissions trends, there is still the reality that only Annex B countries have in fact introduced any significant policies designed to affect emissions. However, this seems implausible given that the variation in emissions trends highlighted previously does not follow such a pattern. So it may more plausibly represent the weakness in the typology of policies hinted at above, notably in ignoring direct action by states through investment in renewable energy deployment, low-carbon infrastructure, and so on. These policies may be more prevalent in developmental states, especially those that still have significant elements of command economies.

Although this is not explored systematically here, it might be expected that this logic will extend to a distinction between federal and unitary states – unitary states should be more capable of developing climate policy than federal ones – a conclusion that is borne out by comparisons of Canada and the US to the UK and France (the latter a federal state but significantly more centralized than either the US or Canada).

To illustrate this, the category of presidential system includes not only democracies like the US, South Africa, and Mexico, but also autocratic regimes like China, Saudi Arabia, and Indonesia.

That is, in all countries in the sample with Polity IV scores of 6 and above (therefore excluding China, Indonesia, and Iran).

Although it is not the focus here, this argument would be strengthened if combined with recognition of the growing role and diversity of transnational climate change governance (Bulkeley et al., 2012).