Encompassing Institutions and CO2 Emissions Institutional and political solutions to collective action problems in comparative perspective

Transcription

1 Encompassing Institutions and CO2 Emissions Institutional and political solutions to collective action problems in comparative perspective Klaus Armingeon University of Berne Department of Political Science March 2008 FIRST DRAFT Paper for ECPR Joint Sessions, Rennes, April 2008 Workshop 6: The Politics of Climate Change Abstract In this paper I discuss politico-institutional explanations for the variation of the change in levels of CO2 emissions in 38 countries for which the UN supplies data for carbon dioxide emissions in 1990 and The underlying theory is the power resources / parties do matter-theory and the theories about the role of encompassing organizations in the production of collective goods. I show that neither party differences, nor corporporatism, nor consociational democracies are statistically powerful explanatory variables. However, EU membership has a significant effect. This is due to the fact that environmental policies are only marginally related to the competition between large parties in democracies. Likewise they are not connected to traditional corporatist and to traditional consociational institutions. In contrast environmental policy has a much more central role in the policy agenda of the European Union. 1

2 I Introduction Carbon dioxide (CO2) is the most important greenhouse. CO2 emissions are a major determinant of climate change. By 2005 most post-communist EU member states have reduced their level of CO2 emissions in 1990 by 24 to 64%; Germany, Sweden, Denmark and UK recorded significant decreases of their former emissions. In contrast countries such as USA, Norway, New Zealand or Spain showed an increase of 20 to 60% of their level in To what extent can these variations be explained by politico-institutional variables? In particular I am interested in two types of variables. Does political competition matter in climate policy? Are governments in countries with strong green and left parties more successful to curb the increase of CO2 emission? Or is democratic competition mostly irrelevant for the outcomes of climate policy? Are democracies responsive to the preferences of their citizens as indicated by electoral choices? In particular this is important for left and green parties. They claim that they care more about environmental challenges than other parties. In addition they claim to make a difference. A second group of variables are encompassing organizations. Arguably climate policy is mainly about solving a collection action problem. Reduced CO2 emission is a collective good from which nobody can be excluded. Hence there are strong incentives to free-ride or to defect from common standards in order to avoid the risk to be exploited by others. These collective action problems occur on the international and the national level. On the international level, countries may be tempted to increase their CO2 emissions while benefiting from the reduction of emissions in other countries. Likewise a national government will not undertake significant steps to reduce CO2 emissions if it fears that other nations will not show efforts at improving the environment. On the level of nation state, this collective action problem may be visible between economic sectors or between groups of citizens. Workers and employers in one sector of the economy will see no reason to make sacrifices in order to contribute to the national goal of reduction of CO2 emissions if they have the feeling that they carry all costs while other national actors do not share the burdens. Political parties will not be willing to fight for reduction of CO2 emissions, if their socio-economic or 2

3 regional clientele suffers much more than others. In addition such strategies are risky in electoral terms, if political competitors can make the point that these parties are naïve and betraying the national interest in a world in which other nations do not sufficiently contribute to the common good. There are institutional solutions to these collective action problems (Olson, M., 1965). On the level beyond the nation state one solution are international organizations or regimes that oblige nations to common standards and that may sanction actors that defect from these standards. On the level of the nation state, these institutions are organizations, which are encompassing and which make it highly likely that other actors contribute their share in the production of the collective good. This applies to consociational democracy and to corporatism. Consociational democracy is a type of democracy, which avoids majority decisions as dominant decision technique. Rather elites try to negotiate agreements across a board range of political groups. Hence in consociational democracies an agreed climate policy is not subject of major political competition between political parties since all important political actors support it. Corporatism denotes the cooperation of interest groups and state in policy formation and implementation. Typically corporatism is not confined to sectors, but covers the whole national economy. These institutions may contribute to sharing the economic burden of national climate policy in a fair way. In this paper I test these hypotheses. I compare the change of CO2 emissions in those countries for which the UN (UNFCCC, 2007) publishes data on C02 emissions for 1990 and By implication I analyse the outcomes of various climate policies and not a set of selected policies (such a CO2 taxes or subsidies for renovation of houses or regulations with regard to car technology etc.) or the extent to which a particular country commits itself to international standards such as those of the Kyoto protocol. Hence the focus is on the combined effects of various policies. I analyse changes rather than levels since levels are determined largely by structures and constraints, which are not amenable to short-term political intervention. I argue that EU membership makes a difference. In contrasts to other international organizations the EU is the most integrated organization which can give its members the feeling that there is some pressure on all members to do something about climate change and that there is also more sanctions at least naming and shaming in a group of 27 actors than in any other international organisation or regime. In addition climate policy is high on the agenda of EU policies and it is 3

4 closely linked to the core of the EU, i.e. economic liberalization within a common market. In contrast, corporatism and consociational democracy do not contribute significantly to reduction of CO2 emissions. Corporatism is about social and economic policies, about distribution and redistribution and about privat-public cooperation in the economy. Climate policy is very low on its agenda and there are no coporatist deals about climate policy. Likewise, consociational democracy is about governability in segmented societies. The institutions are geared towards accommodation between social, regional and cultural groups. But climate policy is not connected to any of these groups. There is no reason to assume that competitive democracies ( Westminster democracies ) are much better in climate policy than consociational democracies, but there are also no reasons to assume that they are much worse. I start with a discussion of the relevant literature and the underlying theories of party competition and solutions to collective action problems. Then I describe my research design and present my findings. Before concluding I discuss why EU makes a difference, but corporatism, consociational democracy and left/green parties fail to do so. Theory The dependent variable. Probably most of the political science literature about climate change is about policy decisions, such as the decision to sign the Kyoto protocol, or the decision for or against various measures to improve on the reduction of greenhouse gas emission. Valuable and insightful as they are, these studies tell us little how successful a political system is in changing the national level of CO2 emission. One brilliant policy may be counteracted by another policy, even outside the field of pure climate policy. Hence if we want to learn about the efficacy of a set of national policies we have to consider their complementarities. There are many examples to construct such composed indicators of climate policy. However, additive indices are vulnerable to the critique that they rest on mostly unproven assumptions of the weights of each component of the indicator. 4

5 Even more important, policies are frequently modified in the course of their implementation. The efficacy of implementation depends on the co-operation between public and private actors (for example firms). In an excellent study, Bättig and Bernauer (Bättig, M. and Bernauer, T., 2008) analysed the impact of democracy on climate policy. They found significant effects of democracy on outputs, but none on outcomes. Hence democracies are better in signing environmental protocols and in producing legal standards about emissions; but they are not better than autocracies in controlling the end-effect of all climate-relevant policies and actions. Or in other words, once the standards are going to be implemented at the interface between national administrations, private actors and economic and social structures, the end result of the process is only very remotely related to the original political decisions. A case in point are Canada, Spain, Portugal and Greece, which signed the Kyoto protocol. All these countries increased their CO2 emission by at least 25% during the last 15 years. Therefore the questions of the determinants of climate policy and of the outcomes of climate policies are substantively different, and one of the crucial links are the institutions which ensure the appropriate implementation of policy decisions. The independent variables However, a critical argument points to the importance of power resources and party differences. Climate policy has distributional consequences. It may be costly for employers; it may also endanger jobs and is a threat to workers in energy intensive industries. Political parties in democracies claim to offer a choice to voters. It depends on constellations of parties and the mobilization of socio-economic electorates what type of policies are implemented. This is the basic assumption of the theory that power resources and parties are decisive for policy development (Korpi, W., 1983; Castles, F.G., 1982; Huber, E. and Stephens, J.D., 2001; Schmidt, M.G., 1996). There are hardly political parties and elites that denounce the challenge of climate change or plead for passivity such as Czech president Vaclac Klaus, the Heartland Institutions or some prominent members of the Bush administration (Neue Zürcher Zeitung ). But there are differences in emphasis and determination. Green ( ecological ) parties put environmental concerns on top of their agenda. Parties left of the traditional social democratic parties have an argument by combining anti-capitalism with ecological policy by arguing that capitalist market 5

6 economies are responsible for air pollution of the industrial and post-industrial society. Finally, even traditional social democratic parties tend to support ecological concerns. One of the underlying reasons is electoral competition and demographic change. The share of post-materialist or pro-ecology voters tends to grow in modern societies and if social democratic parties want to modernize their electoral basis (hitherto consisting mostly of workers) and if they do not want to lose too many votes to green parties, they have to include green issues in their strategies (Kitschelt, H., 1994). In a comparative analysis Neumayer found empirical evidence that green/leftliberal parties are associated with lower pollution levels, and he also found weak hints that the strength of traditional left-wing parties and the strength of left-wing parties in government leads to better environmental performance (Neumayer, E., 2004, Neumayer, E., 2003). Corporatist systems can solve the collective action problem by pursueing national rather than sectoral strategies and by developing effective schemes to compensate losers in economic adjustment, according to an analysis by Scruggs. In addition, corporatism works on the basis of governments, which often retain the threat to use direct regulation if voluntary cooperation between interest groups fails. This motivates for appropriate implementation. Further, corporatism is based on a history of government-producer trust and mutual learning allowing for long-term policy commitment (Scruggs, L.A., 1999, see also Crepaz, M.M.L., 1995; Jahn, D., 1998). Consociational democracies have to internalize externalities (Olson) since there is hardly any larger socio-economic, regional or cultural group that is excluded from the consociational regime. Developing and operationalizing the theory of consociational government, Arend Lijphard (Lijphart, A., 1999) argued that consensus-democracies tend to be kinder and gentler. Consensus democracies consist of two dimensions. The first dimension ( executive-parties ) is composed of structural pre-requisites and correlates of consociational democracies, such as grand coalitions, proportional electoral rules, or effective numbers of parties. This dimension is systematically correlated with positive environmental performance (energy efficiency and a composite index of concern for environment) according to Lijphart (Lijphart, A., 1999: ). 6

7 Research design I analyse all countries which are parties included in Annex I to the UN Convention on Climate Change and for which UN presents data on CO2 emissions for 1990 and (UNFCC, 2007: 19). Monaco and Liechtenstein are excluded from the following analysis since I lack the macro-economic data for the control variables. This leaves us with 38 countries. All these countries are democracies or nations that could be seen on the way to democracy (Russia, Turkey, Ukraine). This does not apply to Belarus. However I kept Belarus for the modelling of the base-line model and for the models for which I had data for at least one operationalization of the independent variables (consociational democracy, corporatism, EU membership). The models are robust with or without the inclusion of Belarus. The dependent variable is the change in CO2 emissions between 2005 and 1990, expressed in percentage of the level of These change rates are regressed on the theoretically important politico-institutional variables after controlling for socio-economic determinants and constraints. The political power distribution is measured by various variables. The first variable measures the cabinet seats of social democratic and other left-wing parties in percentage of total cabinet posts, weighted by days 2. The second measure is the share of parliamentary seats of green and left (non-communist) parties on the left of the traditional social democratic parties. The third measure is the share of parliamentary seats of traditional social democratic parties. Data sources are the Comparative Political Data Sets (Armingeon, K. and Careja, R., 2008, Armingeon, K. et al., 2008a, Armingeon, K. et al., 2008b). The data are calculated on an annual basis and these annual data were then averaged for the period Corporatism was coded as a dummy variable. The theoretical argument states that corporatism is effective and encompassing. Therefore the weak tripartisms of the post-communist world are not considered to be cases of strong corporatism. This applies not to Slovenia though. Based on the literature on established OECD 1 Total anthropogenic carbon dioxide emissions excluding emissions/removals from land use, land-use change and forestry 2 An alternative operationalization based on the share of parliamentary seats of governing left parties yielded nearly identical results in the regression analysis. 3 Missing values either because of presidential system or non-availability of data for Belarus, Croatia, Russian Federation, Turkey, and Ukraine. 7

8 democracies (see Siaroff, A., 1999, Traxler, F. et al., 2001) and post-communist countries ( see also Armingeon, K., 2007, Meardi, G., 2007, Morley, J., 2007). I coded the following countries as corporatist. Australia, Austria, Belgium, Denmark, Finland, Iceland, Ireland, Japan, Luxembourg Netherlands, Norway, Slovenia, Sweden, and Switzerland. Germany and Italy are borderline cases, therefore I assigned a 0.5 to these two countries. This is a coarse classification. Therefore I repeated the analysis for the subsample of established democracies for which Allan Siaroff has gathered comparable metric data (Siaroff, A., 1999). Consociational democracy was coded as dummy variable on the basis of country reports and comparative descriptions (see e.g. Schmidt, M.G., 2000: ). Austria, Belgium, Luxembourg, Netherlands, and Switzerland are considered to be consociational democracies. Germany and Finland are borderline cases and were coded as.5. As an alternative to this crude classification I calculated the share of time a country has been governed by a grand coalition. Grand coalitions are arguable the major indicator for the politics of consensus. Finally, for the subsample of established democracies I could re-run the analyses with the Lijphart indicator (first dimension) of consensus democracy. EU membership has been coded as 1 for all EU-members in Hence Bulgaria and Romania are entered as EU members although they have not been yet members in 2005, at the end of the period under study. The other Western postcommunist countries have joined EU only in However, there was a long period of close EU surveillance before the official membership status and these countries have had to adopt to EU standards ( acquis communautaire ) during many years before acceptance. I checked findings for robustness by subdividing EU membership in those countries that have joined until 1995, and those that joined in 2004/2007. The development of CO2 emissions is influenced and constrained by socioeconomic variables, which are not amenable to short term intervention by governments. These variables have to be entered as controls. Probably the most important variable is whether a country is post-communist. Post-communist countries experienced a dramatic reduction of CO2 emissions simply due to economic crises, industrial restructuring and the abolition of ailing and energy intensive industries in the manufacturing sector during transition to capitalism and democracy. Hence I coded all countries whether they are postcommunist ( 1, elso 0 ). The change rate 8

9 may also depend on the level of CO2 emission. Therefore I entered the level in 1990, standardized by size of population. In theory there may be a Kuznet-like curve (inverted U-curve) describing the amount of CO2 emissions and economic development. In the early stages of development e.g. at the end of the agricultural period -- emissions are low. The same applies to very modern service societies, which do not have heavy industries to a significant extent. In contrast, industrial societies have strongest CO2 emission due to their industrial structures. From simple scatter plots of the countries under analysis (GDP pc * standardized CO2 emissions) there is no convincing hint that the countries are scatter over all areas to both sides of the inflection point. If there is a relationship among the countries of our sample we have to assume that our cases are on one side of the inflection point, which is in all likelihood the right side. Therefore in this country group we expect a negative correlation between emissions and GDP pc as indicator for economic development. Another indicator of stage of economic development is the share of the secondary sector in total GDP. Economic growth may be positively correlated with CO2 emissions since growing economies need more energy and mobility than stagnant economies. The direction of the coefficient of stage of economic development and economic growth may change depending on whether a country is in the special situation of transiting to capitalism. Therefore I entered interaction terms of post-communist countries * development and postcommunist countries * growth. Findings In a first step a base line model has been estimated in which all the socioeconomic context control variables have been entered. Successively, the least significant variables (p <.05 have been excluded). This led to a simple model that explains almost 80% of all international variation in change of CO2-levels between 1990 and 2005 (model 1). The explanatory variables are (a) whether a country is postcommunist (negative correlation), (b) the magnitude of economic growth (positive correlation), (c) the level of economic development (the more developed, the less increase), and (d) an interaction-term for level of economic development and postcommunist countries. In these countries the emission-reducing effect of economic development is almost halved compared to the other countries. 9

10 --Table 1 about here In fact, little is left to explain after such structures and constraints are taken into account. Model 2 adds corporatism to this base line model. Model 3 uses the Siaroff index of integrated economies. It is restricted to the established democracies since the Siaroff index is not available for the post-communist countries. In both models the coefficient for corporatism is in the expected direction. Corporatist systems are associated with less increase in CO2 emissions. However, the coefficient is very far (all countries) or far away from the conventional level of significance (established democracies). Hence in line with Neumayer (Neumayer, E., 2003) I cannot replicate the most robust findings with regard to political institutions and climate policy found by authors such as Scruggs (1999), Jahn (1998), Matthews ( 2001), or Crepaz ( 1995). Does consociational democracy perform better in environmental policy than Westminister democracy? Model 4 is based on the coarse indicator of consociational democracy, model 5 uses grand coalitions as proxy for consociational democracy and model 6 is based on Lijpharts indicator of the first dimension (executive-parties) of consensus democracies. This latter analysis is restricted to the established democracies due to data availability. The result is similar to the result of the analysis of the impact of corporatism. Although all the coefficients are in the expected direction, they are not significant. Consociational democracy is associated with less increases in the level of CO2 emission, but this association is weak and not significant. Lijpharts (Lijphart, A., 1999) optimistic reading of his data cannot be replicated in this analysis. Is then party competition an important explanatory variable? Models 7, 8 and 9 include partisan variables. Model 7 looks at the average strength of left parties in government; model 8 at the parliamentary strength of left-green parties and model 9 at the parliamentary strength of traditional social democratic parties. All these estimations show that in fact left parties are associated with less CO2 emissions but in no case are these effects significant. Parties may make a difference, but this difference is small and statistically not important. Neumayer s (Neumayer, E., 2003, Neumayer, E., 2004) finding of a strong party effect cannot be reproduced in this sample and in regard to CO2 emissions. 10

11 Table 1: Determinants of the change in CO2 emissions between 1990 and 2005 Model 1 Model 2 Model 3 Model 4 Model 5 Constant ** ** ** ** ** Post-communist country(dummy) ** ** ** ** GDP Growth 6.03 * 6.04 * 8.42 * 5.84 * 5.09 * Interactive term (Post-communist*GDP p.c.) (/1000) 4.48 ** 4.50 * 4.36 * 4.90 * GDP p.c. (/1000) ** ** * ** * Corporatism (Dummy) Corporatism (Siaroff Index) Consociational democracy (Dummy) Consociational democracy (Grand coalition) Consensus democarcy (Lijphart Indicator first dimension) Share of cabinet seats of left parties Share of parliamentary seats of green and left-socialist parties (/100) Share of parliamentary seats of social democratic parties EU (yes=1, else=0) EU accession until 1995 (yes=1, else=0) EU accession in 2004/2007 (Yes=1, else=0) F ** ** 6.40 ** ** ** N R2(adj) Unstandardized regression coefficients p < 0.05: *; p < 0.01: ** 11

12 Table 1: Determinants of the change in CO2 emissions between 1990 and 2005 (continued) Model 6 Model 7 Model 8 Model 9 Model 10 Model 11 Constant * ** ** * ** ** Post-communist (Dummy) ** ** ** ** ** GDP Growth 9.14 * 7.98 ** 8.34 ** 8.47 ** 6.56 ** 7.10 ** Interactive term (Post-communist*GDP p.c.) (/1000) 5.64 ** 5.34 ** 5.46 ** 5.66 ** 6.36 ** GDP p.c. (/1000) * ** ** ** ** ** Corporatism (Dummy) Corporatism (Siaroff Index) Consociational democracy (Dummy) Consoc. democracy (Grand coalition) Consensus democracy (Lijphart Indicator 1st dimension) Share of cabinet seats of left parties Share of parl. seats of green and left-socialist parties (/100) Share of parl. seats of social democratic parties EU (yes=1, else=0) ** EU accession until 1995 (yes=1, else=0) * EU accession in 2004/2007 (yes=1, else=0) ** F 5.68 ** ** ** ** ** ** N R2(adj) Unstandardized regression coefficients p < 0.05: *; p < 0.01: ** 12

13 Models 10 and 11 add EU membership to the base line model. The findings are clear. EU membership is strongly correlated with less increases of the level of CO2 emissions. The findings are very significant and the effect can be found among the new and the old EU members. Why EU, why not corporatism, consociational democracy and parties? How can we explain that only one of the four variables is significantly associated with less increases in the level of CO2 emission over the period ? My basic argument is that actors have to be powerful and that they have to place environmental concerns in the upper part of their agenda. This applies both to party competition and to the mechanism of production of collective goods due to encompassing organizations. In case of party competition it are only the green parties (and at a maximum the other non-communist parties left of the social democracy) that put green issues in a dominant position. Social democratic parties are traditionally oriented towards questions of redistribution, state intervention and welfare states. Green issues are increasingly taken onto board but they are frequently balanced with other materialistic issues such as employment and growth. The concerns of the former German social democratic chancellor Schröder about VW s economic fate and EU s end-of-live vehicles draft directive (1999) are an example. On the other hand there is little evidence that liberal and conservative parties do not care about the environment -- at least as long as it is compatible with a thriving market economy. As argued by many authors (see e.g. the most recent OECD report on the economic costs of environmental policy, NZZ ) reduction of CO2 emission may not have serious economic costs even in the mid term perspective and not to speak of the economic costs of inactivity. Hence climate policy is not a major political cleavage that separates the main political parties in democracies. The argument about the effects of corporatism on CO2 emission is based on the assumption that actors make (nation-wide) agreements on industrial restructuring which include environmental concerns which are given an very important if not even 13

14 a dominant role. This is a heroic assumption though. The reports on corporatist deals in the past relate mainly to issues of wages, prices, social and tax policy (Compston, H., 1994, Compston, H., 1995a, Compston, H., 1995b; Armingeon, K., 1983). In the reports about most recent social pacts we also do not find environmental policy in a prominent position (Hassel, A., 2006). It might be very well the case that governments, trade unions and employers also consider environmental questions when they conclude social pacts. However there is very little empirical evidence that this is really the case and these potentially existing but hidden contents of corporatist deals are not reflected in our data. Finally the beneficial effects of consociational democracy for climate policy are based on the assumptions that elites in these non-competitive democracy extent the decision technique of negotiation to all political conflicts. This is a heroic assumption, too. Originally the consociational bargains were related to the conflicts between segmented socio-cultural and regional groups. None of these historical conflicts is connected to the environment, though (Lehmbruch, G., 1974). Even if this technique is extended to new policy issues such as environment, we can expect significant effects only if this issue ranks very high on the political agenda and if it is a controversial issue in domestic political competition between the major players. The strong and highly significant effects of EU come as a surprise. The EU has a reputation for watering down goals in lengthy intergovernmental bargains, for failing to implement its policies, for little compliance of many member states in complying with directives and standards and for being very slow in coping with external challenges. This applies in particular to positive integration such as social policy (Falkner, G. et al., 2005; Börzel, T.A. and Risse, T., 2000, Börzel, T.A., 2006) or foreign policy (Wallace, W., 2005). It also applies to environmental policy (Knill, C., 2003, Knill, C. and Lenschow, A., 2005, Lenschow, A., 2005). However, there are also findings that point to stronger convergence towards stricter environmental regulation in the EU (Holzinger, K., 2008). These findings correspond to the regression coefficients of this analysis. The underlying causal reason may be that environmental policy has become a dominant issue in EU politics over the last ten to 15 years (Knill, C., 2003, Lenschow, A., 2005). An example is the leading role of the Commission during the Bali negotiation. Environmental policy became important for the EU since it could be framed as policy of negative integration for a long time. Similar environmental standards have been considered to be crucial for the creation of 14

15 a level playing field within a common market. Supported by its market-enabling role environmental policy could move up towards the top of the EU agenda. For the powerful national European governments and for the Commission there are good reason to be active in environmental policy. As a union of many highly developed economies it cannot neglect the externalities of its economies and societies. As national governments within the EU there are at least two reasons to support strict and efficient regulations within the nation state: First, nations that do not cooperate will suffer from naming and shaming and increasingly from ECJ and Commission action if they violate EU standards. Probably more importantly, governments see EU environmental policy as a creditable commitment of all members. Hence the fear of a national administration may be reduced that they contribute to the collective good while other exploit them and free-ride. Compared to other encompassing institutions such as corporatism and consociational democracy the EU is arguably the encompassing institution that has powerful actors that put environmental concerns on the top of the political agenda. This may explain, why EU matters for CO2 reduction while corporatism and consociational democracy do not. Conclusion In this paper I made the simple argument the EU can be considered to be an encompassing institution that is concerned about environmental policy. This explains why EU membership matters for the reduction of increases in the level of CO2 emissions. In contrast, party competition, corporatism and consociational democracy are not powerful explanatory variables for the variation in change of levels of CO2 emissions between 1990 and This is in stark contrast to the findings by Neumayer, Crepaz and Scruggs (Neumayer, E., 2003, Crepaz, M.M.L., 1995; Scruggs, L.A., 1999). The explanation of this difference can be found in the research designs. All these works are based on cross sectional pooled time series (panel data), they explain levels rather than changes and they include institutional variables on the right hand side of the equitation that are hardly time-variant. This creates serious problems of autocorrelation, heteroscedasticity, and unit root processes. The latter denotes highly persistent time series. Frequently national levels of CO2 emissions or 15

16 institutional variables may be highly persistent. One consequence of such persistent time series on both sides of the equation is the danger to create many cases by copying and pasting the few cases which reality provides. The major effect of this procedure, which does increase technically the number of cases without increasing the variance on both sides of the equation, is to increase the likelihood of statistically significant results. In the simple cross sectional design of this paper with change rates as the dependent variable, however, we avoid these problems and increased the number of cases by added much more nations to the sample than in the other works. I find the same direction of coefficients for parties, corporatism and consocational democracy but without meeting the criterium of statistical significance. Since all the studies are no random samples, statistical significance is only a heuristic tools to decide about the importance of findings, though. And using these statistics in an appropriate manner I find that corporatism, consociational democracies and parties are not important explanatory variable for changes of CO2 emissions. In contrast, EU membership is important for change rates of levels of CO2 emissions. The work shop director asked us for political consequences which could be drawn from our preliminary findings. I have to offer four such consequences: (1) Most variance in change of CO2 levels is explained by economic and structural variables. These are hardly amenable to political steering. In particular we know that governments have little effective choices in growth policies (Obinger, H., 2004). (2) For many authors the finding that corporatism does not matter comes as a normative disappointment. On the other hand this finding is also devastating for another hypotheses, namely that corporatist systems are a burden for democracies since they slow down political reforms and benefit the egoistic interests of certain groups. (3) The same applies to consociational democracies. It is not worse than Westminster democracies in coping with climate challenges. Westminster democracies are not superior in dealing with major challenges quite the contrary what many proponents of the model of competitive democracy claim. (4) EU integration may contribute to improving the environmental performance of democratic nations. 16

17 References ARMINGEON, K. (1983) Neo-korporatistische Einkommenspolitik. Eine vergleichende Untersuchung von Einkommenspolitiken in westeuropäischen Ländern, Frankfurt a. Main: Haag & Herchen. ARMINGEON, K. (2007) 'Kleinstaaten in Weltmärkten. Drei Ergänzungen der Katzenstein-These', Zeitschrift für Sozialreform, 53(3): ARMINGEON, K. & CAREJA, R. (2008) Comparative Data Set for 28 Post- Communist Countries, , Berne: Institute of Political Science ( ARMINGEON, K., CAREJA, R., GERBER, M. & LEIMGRUBER, P. (2008a) Comparative Political Data Set III, , Berne: University of Berne, Institute of Political Science. ARMINGEON, K., GERBER, M., LEIMGRUBER, P., BEYELER, M. & MENEGALE, S. (2008b) Comparative Political Data Set I , Berne: Institut of Political Science ( BÄTTIG, M. & BERNAUER, T. (2008) 'National Institutions and Global Public Goods. Are democracies more cooperative in climate change policy?' ETH Zürich. BÖRZEL, T. A. (2006) 'Participation Through Law Enforcement. The Case of the European Union', Comparative Political Studies, 39(1): BÖRZEL, T. A. & RISSE, T. (2000) 'When Europe Hits Home: Europeanization and Domestic Change', European Integration online Papers, 4( CASTLES, F. G. (ed.) (1982) The Impact of Parties. Politics and Policies in Democratic Capitalist States, London/Beverly Hills: Sage. COMPSTON, H. (1994) 'Union Participation in Economic Policy-Making in Austria, Switzerland, The Netherlands, Belgium and Ireland, ', West European Politics, 17(1): COMPSTON, H. (1995a) 'Union Participation in Economic Policy Making in France, Italy, Germany and Britain, ', West European Politics, 18(2): COMPSTON, H. (1995b) 'Union Participation in Economic Policy Making in Scandinavia, ', West European Politics, 18(1): CREPAZ, M. M. L. (1995) 'Explaining National Variations of Air Pollution Levels: Political Institutions and Their Impact on Environmental Policy-Making', Environmental Politics, 4(3): FALKNER, G., TREIB, O., HARTLAPP, M. & LEIBER, S. (2005) Complying with Europe. EU Harmonisation and Soft Law in the Member States, Cambridge: Cambride University Press. HASSEL, A. (2006) Wage Setting, Social Pacts and the Euro. A New Role for the State, Amsterdam: Amsterdam University Press. HOLZINGER, K. (2008) 'Environmental Policy Convergence: The impact of international harmonization, transnational communication and regulatory competition', International Organizations, forthcoming HUBER, E. & STEPHENS, J. D. (2001) Development and Crisis of the Welfare State. Parties and Policies in Global Markets, Chicago and London: The University of Chicago Press. 17

18 JAHN, D. (1998) 'Environmental performance and policy regimes: Explaining variations in 18 OECD-countries', Policy Sciences, 31( KITSCHELT, H. (1994) The Transformation of European Social Democracy, Cambridge, Mass.: Cambridge University Press. KNILL, C. (2003) Europäische Umweltpolitik, Opladen: Leske&Budrich. KNILL, C. & LENSCHOW, A. (2005) 'Compliance, Communication and Competition: Patterns of EU environmental policy making and their impact on policy convergences', European Environment, 15( KORPI, W. (1983) The Democratic Class Struggle, London: Routledge & Kegan Paul. LEHMBRUCH, G. (1974) 'A Non-Competitive Pattern of Conflict Management in Liberal Democracies: the Case of Switzerland, Austria and Lebanon', in MCRAE, K. D. (ed.) Consociational Democracy: Political Accommodation in Segmented Societies, Toronto: McClelland and Stewart, LENSCHOW, A. (2005) 'Environmental Policy', in WALLACE, H., WALLACE, W. & POLLACK, M. A. (eds.) Policy-Making in the European Union, Oxford: Oxford University Press, LIJPHART, A. (1999) Patterns of Democracy: Government Form and Performance in Thirty-Six Countries, New Haven: Yale University Press. MATTHEWS, M. M. (2001) 'Cleaning Up Their Acts: Shifts of environment and energy policies in pluralist and corporatist states', Policy Studies Journal, 29(3): MEARDI, G. (2007) 'More voice after more exit? Unstable industrial relations in Central Eastern Europe', Industrial Relations Journal, 38(6): MORLEY, J. (2007) 'Labour market developments in the new EU Member States', Industrial Relations Journal, 38(6): NEUMAYER, E. (2003) 'Are left-wing party strength and corporatism good for the environment? Evidence from panel analysis of air pollution in OECD countries', Ecological Economics, 45(2): NEUMAYER, E. (2004) 'The environment, left-wing political orientation and ecological economics', Ecological Economics, 51(3-4): OBINGER, H. (2004) Politik und Wirtschaftswachstum. Ein internationaler Vergleich, Wiesbaden: Verlag für Sozialwissenschaften. OLSON, M. (1965) The Logic of Collective Action. Public Goods and the Theory of Groups, Cambridge/London: Harvard University Press. SCHMIDT, M. G. (1996) 'When Parties Matter: A Review of the Possibilities and Limits of Partisan Influence on Public Policy', European Journal of Political Research, 30(2): SCHMIDT, M. G. (2000) Demokratietheorien 3. Aufl., Opladen: Leske und Budrich. SCRUGGS, L. A. (1999) 'Institutions and Environmental Performance in Seventeen Western Democracies', British Journal of Political Science, 29(1): SIAROFF, A. (1999) 'Corporatism in 24 industrial democracies: Meaning and Measurement', European Journal of Political Research, 36(2): TRAXLER, F., BLASCHKE, S. & KITTEL, B. (2001) National Labour Relations in Internationalized Markets, Oxford: Oxford University Press. UNFCCC (2007) 'National greenhouse gas inventory data for the period ', Subsidiary Body For Implementation, Twenty-Seventh Session. Bali, United Nations. 18

19 WALLACE, W. (2005) 'Foreign and Security Policy', in WALLACE, H., WALLACE, W. & POLLACK, M. A. (eds.) Policy-Making in the European Union, Oxford: Oxford University Press,