The rise and fall of voluntary agreements in German environmental policy.

Toller, Annette Elisabeth

1 Introduction

For decades, voluntary agreements have been regarded as "an element of a distinctly 'German' policy mix" (Lees 2007, p. 175). Indeed, in the 1980s and early 1990s, Germany had the largest number of such agreements (more than 100 altogether), even when compared to the Netherlands, which also made extensive use of such environmental agreements in Europe (e.g., Immerzeel-Brand 2002, p. 385; Glasbergen 1999). The reasons for the use of these instruments, however, are a matter of dispute: whereas some regard the use of such agreements as a response to the growing problems associated with the use of traditional, regulatory means of regulation (the "better regulation" hypothesis), others consider agreements to be the result of a bad deal made possible by growing business power during periods of globalisation (the power-centred hypothesis).

This article makes two arguments. First, although voluntary agreements were indeed a relevant element of the German environmental policy mix in the 1980s and early 1990s, their use was nearly completely discontinued in the new century. Second, when seeking out the causes of both the high use of voluntary agreements in the 1980s and 1990s and their more recent abrupt abandonment, neither the explanations based on functional arguments nor those based on the power of business can account for the developments observed. Instead, institutions (in particular European law and its effect on national policy decisions) and actors (in particular their party-political positions towards policy instruments) play a major role.

The article is based on an extensive theory-driven empirical study on the use of voluntary agreements in German environmental policy from the 1970s to the mid-2000s, covering both the macro-level and the meso-level, as reflected by 13 in-depth qualitative case studies (Toller 2012a). The article proceeds as follows: In chapter 2, I describe the general use patterns of voluntary environmental agreements in Germany and briefly summarise my case studies. In chapter 3, I first discuss the common explanations for the use of voluntary agreements in additional detail (3.1) and then present my findings, which are based on the approach of policy processes driven by momentum (3.2, see Bocher & Toller 2012b). Actors (driven by party politics) and institutions (European Union law) play a major role in explaining the use of voluntary agreements, whereas the roles of functionality (as expressed in the "better regulation" hypothesis) and business power (as expressed in the power-centred hypothesis) are moderate.

However, two caveats must be noted. First, for reasons of space, the article cannot account for all of the causal factors involved (see Toller 2012a, pp. 227-375). Second, the article only examines the use of voluntary agreements in Germany and its driving forces. It does not address the ecological effects, efficiency or merits or demerits of these instruments or what effect it has on the state to cooperate with societal actors.

2 Voluntary Agreements in Germany: Patterns of Use

2.1 Characteristics

To begin with a definition, voluntary agreements, (1) as used in Germany, are agreements between businesses (usually represented by business associations, sometimes by major companies) and the federal government (usually represented by the Ministry for the Environment, although other ministries or agencies and even the Chancellery may also be involved) to accomplish environmental objectives. (2) In contrast to the Netherlands, where such agreements can be private legal contracts, voluntary agreements in Germany are gentlemen's agreements only and are thus not legally binding. There used to be a strong tendency on the part of industry and the government to call such agreements industry self-obligations, (3) suggesting that they were the result of unilateral activity by business. However, in the majority of the cases, these self-obligations were the result of intense bilateral negotiations between business and government often came about under a more or less credible threat of hierarchical intervention--the famous "shadow of hierarchy" (Toller 2008c). Whereas these written agreements were formerly unilateral self-obligations on the part of business (while the government's pledge not to intervene so long as business complied with self-obligations remained implicit), more recent agreements entailed explicit obligations on the part of both parties and were signed by both (Toller 2012a, pp. 44-46). The bilateral character of the instrument is also an element of a definition based on a report by Mol and colleagues (2000): accordingly, voluntary agreements are characterised by a low degree of force (as opposed to statutory obligation) and a high degree of interaction between societal and governmental actors (as opposed to pure self-regulation).

In contrast to the Netherlands, where voluntary agreements are rooted in an overall governmental strategy to encourage business to become more proactively involved in environmental protection, voluntary agreements in Germany are adopted on a case-by-case basis (Toller 2008a), usually after the government has presented a proposal for a statutory regulation and a business association has reacted to this proposal by suggesting a voluntary regulation. This prelude is typically followed by negotiations over the regulatory substance of the agreement, which is finally either announced by the business and accepted by the government or signed by both parties.

Although voluntary agreements are often criticised for their tendency to bypass the appropriate legislative process and thus add to an overall tendency of deparliamentarisation (e.g., Rose-Ackerman 1995, p. 133; Porter & Ronit 2006, p. 49), in most of the cases under closer scrutiny here, the suggested regulatory alternative has been an executive ordinance in which Parliament would not have been formally involved (Toller 2012a, p. 68). However, the German Bundestag in most cases has discussed the issues at stake. In addition, voluntary regulations are not always an alternative to statutory regulation but rather can be combined with statutory regulation in various and interesting ways (Toller 2012a, pp. 69-72).

2.2 Macro-level Perspective: Patterns of Use

Although many publications label voluntary agreements as "new" policy instruments (e.g., Ingram 1999; Aggeri 1999; Grimeaud 2004; Jordan et al. 2005), at least in Germany, they are not new at all. Rather, this type of agreement dates back to the early days of environmental policy itself: the late 1960s and early 1970s.

If we consider the sheer number of voluntary agreements adopted each year, as displayed in figure 1, we can identify three phases: In a warm-up-phase from the late 1960s until roughly 1983, voluntary agreements were established and--with the exception of 1980--up to two agreements were adopted annually. The heyday of voluntary agreements was a long period from 1984 to 2001, when several agreements were adopted each year, with notable peaks in 1986 (14 agreements) and 1995 (ten agreements). From 2002 on, we observe a decline, with only four agreements adopted in total.

[FIGURE 1 OMITTED]

2.3 Meso-level Perspective: 13 Case Studies

To better understand why voluntary agreements were adopted (instead of statutory agreements), I analysed the adoption of such agreements using 13 cases studies. A case does not cover one agreement alone but rather a regulatory course of varying duration (the longest course spanned 28 years), in many cases including more than one agreement (see more on the research design in Toller 2012a, pp. 90-93). The case studies come from four fields: health protection, pollution regulation, waste management policy and climate protection. I selected the most interesting and complex cases addressing relevant ecological problems in each of the four fields.

For reasons of space, I can only present the core data on the case studies here (see table 1); all data on the policy processes can be found elsewhere (Toller 2012a, pp. 90-201). The results of these case studies are presented in chapter 3 (see, also, Toller 2012c).

Adopting a regulatory course perspective reveals aspects that cannot be observed in macro-data (figure 1) or case studies that focus on a particular case but neglect factors present before or after the adoption of the agreement. Whereas figure 1 only presents the number of newly adopted agreements each year (and the corresponding number of legislative acts), by studying regulatory courses, we can evaluate the "fate" of an agreement. In most cases, agreements are replaced by statutory regulation over time. However, whether this is the case (as is often assumed) because agreements were unsuccessful (rarely the case) or in spite of their success and for other reasons (more often the case) cannot be derived from the data presented here. This question is addressed in the following chapters.

3 Driving Forces

What is striking about the literature on voluntary agreements is that contributions tend either to be descriptive or to address questions of success or failure while remaining rather uninterested in why voluntary agreements are used at all (see Toller 2011, p. 503). If authors are interested, they generally employ implicit assumptions rather than systematically testing different hypotheses. Two broad lines of argumentation in the existing literature are presented below.

3.1 Popular Explanations: Better Regulation vs. Business Power

The first group of arguments, which I subsume under the better-regulation hypothesis, draws on the entire corpus of criticism concerning the numerous deficiencies of traditional regulatory policy instruments. Regulatory law is intrinsically inflexible and simple, and it requires complete information, which the government does not have, or in some cases, this information does not (yet) even exist (Black 2001, p. 107). The law is unable to adapt to all of the specific requirements of regulated firms, is based on an antagonistic relationship between business and government and is unable to positively motivate its subjects. Moreover, society is far too complex and idiosyncratic to be purposefully influenced by such simple instruments (Black 2001, p. 106). From this standpoint, new policy objectives, such as the pursuit of innovation or sustainability, especially require new instruments, which can be adopted more quickly; are more flexible, less antagonistic and costly; and even allow for collective learning (e.g., Kleger 1995, p. 107; Gunningham & Rees 1997, p. 366; Hofmann & Schrama 2005, p. 42; see Toller 2012a, pp. 206-215). In this context, many authors implicitly or explicitly argue in a functionalistic fashion that the advantages of voluntary agreements over statutory regulation explain their use (e.g., Ritter 1990, p. 58; Hansjurgens & Klock 2003, p. 10).

On the contrary, power-centred explanations begin from the Olsonian wisdom that specific interests are easier to integrate, organise and represent than diffuse interests (Olson 1965). Hence, the argument is that globalisation has intensified this inequality in power between business and collective interests. Business has become more powerful vis-a-vis the regulatory state because, under globalisation, it can credibly threaten to relocate its productive activities elsewhere (Greer & Bruno 1996, p. 21). "Whether real or not, the perception of mobility of firms and capital has limited the range of policy instruments used by governments" (Peters 2002, pp. 558-559.). From this perspective, voluntary agreements are the result of successful rent-seeking, particularly by transnational corporations. These agreements aim to avert serious statutory regulation and only give the appearance of regulation, whereas individual businesses can continue their activities, impairing the collective good. Thus, voluntary agreements serve business interests by " greenwashing" their activities, that is, by giving "an organization the appearance of ethicality and leadership when no such commitment exists" (Laufer 2003, p. 257) and without having to substantially modify their practices (Greer & Bruno 1996, p. 31; Gunningham & Rees 1997, p. 370; Beder 2002, p. 99; OECD 2003, p. 43).

The problem with these approaches is that they all appear plausible in some respects, but they consider the phenomenon from one side only and, in particular, do not reflect an adequately complex conceptualisation of politics. Politics is neither a matter of interests automatically producing rational choices and rational choices automatically producing policy outputs nor a purely rational process of benevolent government actors aiming at solving problems. In addition, the better-regulation explanation seems to exaggerate the shortcomings of traditional policy instruments and ignores that these shortcomings do not automatically produce new instruments. More generally, functional approaches have trouble explaining change (Schmidt 2007), although this is exactly what must be done.

3.2 Actors, Institutions and Momentum

To analyse my case studies, I applied an analytical approach that was first developed by Bocher and Toller for analysing the choice and change of instruments in environmental policy (Bocher & Toller 2007) and was later elaborated as a general approach for explaining environmental policy results (see Bocher & Toller 2012a; 2012b). Based on early works by Larry Kiser and Elinor Ostrom (Kiser & and Ostrom 1982) and not fundamentally unlike other approaches based on institutional theory (e.g., Mayntz & Scharpf 1995), we identify a number of factors that shape and influence the political process and its results: actors,

institutions (we see good reasons to place strong emphasis on institutions), problem structures, instrumental alternatives and process dynamics. However, the major difference is that our understanding of political processes is to some extent influenced by the work of John Kingdon (2003). Hence, we do not perceive political processes (as, for instance, the policy-cycle approach does) as a means-oriented, step-by-step problem-solving process but rather as a highly contingent process driven by coincidences in which solutions are in some instances searching for problems instead of the reverse (Bocher & Toller 2012b, pp. 461-470). E.g. actors or institutions (as explanatory factors) are not primarily oriented towards problem-solving but develop momentum.

3.2.1 Actors: On Party Politics and Intervening Variables

Apart from the truism that only actors can act, the concept of agency in policy analysis is as contingent as it is broad. One useful way to concretise our notion of agency is the theory of partisan difference (Castles 1982). Briefly, the theory assumes first, that political parties differ in their policy preferences (both because they have different ideological roots and they expect to gain votes from different societal groups); second, that once they hold governmental power, political parties aim to put their policy preferences into practice: and third, that they will be able to do so (see Hibbs 1977; Castles 1982). Currently, environmental policy in general--as opposed to welfare, education or budget policies --is not considered strongly determined by party politics (e.g., Seeger 2003), apart from the fact that the participation of the Green Party in government should have an effect.

However, the starting point of my considerations was that the decision of whether to adopt a voluntary agreement is not primarily a matter of environmental policy objectives but rather concerns the selection of one particular policy instrument over another. This is because policy instruments are not (as would be suggested by an approach that regards political processes as means-oriented problem-solving processes) technical, neutral means to achieve political aims. Rather, policy instruments are political entities of their own (see Majone 1976, p. 589; 1989, p. 116). They reflect understandings of the human being, the role of the state and democracy (Lascoumes & LeGales 2007; Immergut 2011) and are often either welcomed or damned for this reason rather than for their technical abilities to reach a politically defined objective (Bocher & Toller 2012a, pp. 74-83; pp. 178-199). Thus, the choice of policy instruments should be a matter of ideological dispute and thus also of partisan politics.

I began by examining party positions on the use of voluntary agreements and then analysed the de-facto use of voluntary agreements by different Minsters for the Environment. As voluntary agreements were not sufficiently important to be addressed in party programmes, I examined the numerous parliamentary debates on whether a specific environmental issue should be addressed by a voluntary agreement or an ordinance or legal act. The first interesting observation was that whereas environmental issues as such had already become the object of parliamentary activity (enquiries, in particular, as well as debates) in the late 1970s and early 1980s, the use of policy instruments only became an issue from 1987 on (see, e.g., Zittel 1996, pp. 130-142; Toller 2012a, p. 296). By examining these debates, it was possible to identify clear party positions on voluntary agreements.

The Greens have always been fundamentally opposed to the use of voluntary agreements. Based on the ecology movement, the Greens have consistently preferred strict statutory regulatory instruments, first because these instruments usually come with a substantial intervention in economic activity that threatens the public good, and second, on a symbolic level, because statutory instruments prohibit or restrict and thus morally discourage polluting activities (Toller 2012a, pp. 297-298). Thus, in parliamentary debates the Greens harshly criticised the use of voluntary agreements not only because, due to the lack of legally binding force, they expected the agreements to be ineffective but also because, on a symbolic level, the state negotiates ("horse trades") with business over the measures necessary to protect the environment (Toller 2012a, p. 298).

The Liberals can be placed on the other side of the spectrum. Liberals have consistently regarded voluntary agreements as a means of deregulation and eliminating red tape. Because agreements provide businesses with leeway to determine the measures to be pursued to protect the environment, the Liberals have favoured voluntary agreements irrespective of the aim (Toller 2012a, pp. 298-299). Social Democrats and Christian Democrats can be placed in the middle, albeit with differences in their specific attitudes.

Social Democrats have been moderately critical, stating that they are not generally against voluntary agreements but fear that these agreements would not be effective due to their lack of legally binding force and sanctions (Toller 2012a, pp. 299-300).

Christian Democrats have been moderately positive with respect to the use of voluntary agreements, but, as opposed to the Liberals, they demand that agreements be effective. (4)

When, in a second step, we consider the policies actually adopted under different party majorities, we can make two interesting observations (see table 2).

First, under a Green Minister for the Environment, far fewer voluntary agreements (1.7 on average) were adopted than under his two conservative predecessors. Although it took Jirgen Trittin three years in office to eliminate the unloved agreements (see figure 1), this pattern clearly confirms the assumptions made above based on partisan difference theory. Second, there is a relevant difference between the two conservative minsters that cannot be explained by party difference. Whereas Klaus Topfer adopted 4.4 agreements per year, under Angela Merkel, 6.8 agreements were adopted. How can we explain this difference? Here, the overall political context is important: Klaus Topfer as an environmental expert assumed office in 1987 and at least initially was expected to be a powerful minster. He had numerous plans to adopt modern but strict environmental policy instruments. Although he ultimately failed in numerous ambitious projects, he was not keen to adopt voluntary agreements (Toller 2012a, pp. 304-305). In Topfer's later days as a minister, and especially when Angela Merkel became the Minister for the Environment after the 1994 elections, Germany faced the strong pressure of the costs and consequences of German unification (Bocher & Toller 2012a, p. 32). That Angela Merkel, (then) a generally politically unknown woman from East Germany, was appointed to this position was an expression of the minor role that environmental policy was supposed to play in Chancellor Kohl's last term in office. In the coalition agreement, Liberals and Christian Democrats declared that in the much-disputed field of waste management policy, voluntary agreements should be given priority over authoritative measures. This policy was accompanied by a discourse strongly favouring deregulation to protect the "Standort Deutschland" (this emotionally charged term can only be loosely translated as "business-friendly Germany"). What is more, the German business association (BDI) placed strong pressure on the government to use voluntary agreements instead of statutory regulation. One can even argue that the use of voluntary agreements in environmental policy during this period became an aim in itself to demonstrate to business that the government was serious about deregulation (Toller 2012a, pp. 305-308); thus, the instrument itself gained momentum. This changing overall political context--as an intervening variable--can account for the difference between the two conservative ministers.

3.2.2 Institutions: On Legal Uncertainty Stemming from EU Law

It is the common wisdom of neo-institutional theories that institutions (defined as rule systems) influence what actors do, not only by entailing rules concerning allowable and unallowable actions but also by influencing how actors perceive their preferences (Hall & Taylor 2006; see Bocher & Toller 2012a, pp. 151-156). My argument here is that institutions do not fully determine what actors do, but they have a strong influence on agency.

Over the last 15 years, research on Europeanisation has addressed how European integration impacts, among other things, the public policies of EU member states (see Radaelli 2004; Exadactylos & Radaelli 2012; Toller 2012d). Briefly, the problem with this research is that it has primarily focused on the effects of positive integration (namely, harmonisation directives) while by and large neglecting the effects of negative integration. Moreover, this strand of research has heavily focused on the "desired" effects of European law while ignoring less desired, idiosyncratic effects.

The argument of this chapter is that the use of voluntary agreements in Germany, as presented in chapter 2, has been triggered to a substantial degree by European Law in general and the legal uncertainty stemming from negative integration in particular (see Toller 2012c). To elaborate my argument, I first briefly explain the substantial elements and second describe the procedural elements of negative integration. In a third step, I demonstrate how this integration brought about the "evasion" of voluntary agreements as an idiosyncratic mechanism of Europeanisation (Toller 2012c) and then present some--albeit brief--evidence from my cases. Finally, I examine the role of positive integration.

The Meaning and Substantial Arsenal of Negative Integration

The term negative integration means the abolition of all barriers to trade within the Community/Union and is based on Articles 34 and 36 of the EU Treaty (TFEU) and its interpretation by the European courts. By 1957, Article 34 (then Art. 30) had already stipulated the following:

"Quantitative restrictions on imports and all measures having equivalent effect shall be prohibited between Member States".

Certainly, there has been much debate about what types of measures have such equivalent effects. In its Dassonville decision in 1974, the ECJ ruled that "all trading rules enacted by Member States which are capable of hindering, directly or indirectly, actually or potentially, intra-Community trade are to be considered as measures having an effect equivalent to quantitative restrictions" (C 8/74; Geradin 1997, p. 10). Thus, long before the internal market was realised based on the SEA (1987), the EC Treaty restricted the national regulation of harmful substances, and as such, regulation could be considered a measure restricting intra-community trade. This restriction only became potentially relevant with the development of environmental regulation in the EC member states beginning in the early 1970s.

However, at that point, Art. 36 (now Art. 36 TFEU) already allowed for "prohibitions or restrictions on imports, exports or goods in transit justified on grounds of public security; the protection of health and life of humans, animals or plants". Such prohibitions or restrictions shall not, however, constitute a means of arbitrary discrimination or a disguised restriction on trade between Member States (Shapiro 1992, p. 130; Geradin 1993, p. 179; 1999, p. 13). In several judgments, the ECJ emphasised that barriers to trade erected for compelling reasons (such as the protection of consumers, etc.) could be acceptable as long as they were not discriminating and disproportional (Cassis de Dijon, C 120/78, for environmental policy: old oil case, C 240/83; Geradin 1993; Kramer 1995, p. 113). In Germany, the so-called Danish bottle case in 1988 attracted substantial attention, as this was the first time that the court had to decide whether a national environmental regulation was proportional and thus permissible under the EC Treaty (C 302/86; Geradin 1993, p. 183). (5) In this judgment, environmental protection was established as a compelling reason for adopting barriers to trade, as long as they are proportional.

However, if the Community adopted a harmonising measure (positive integration), national leeway would shrink to a greater extent (Palme 1992, p. 53), despite the establishment of a number of exceptions since the Single European Act came into force in 1987 (Art. 100a, now Art. 114 TFEU, and Art. 130s-130t, now art. 191-193 TFEU): under certain conditions, stricter national measures could be upheld, or new ones could be adopted (Kramer 1995, pp. 99-106). (6)

As noted above, negative integration has not thus far been the focus of the literature on Europeanisation. Other studies, however, have addressed possible effects at the national level. Fritz Scharpf has argued since the early 1990s that the internal market would bring about a regulatory race to the bottom, as negative integration (as privileged by its establishment in the Treaty itself) would abolish old, and prohibit new, national regulation, whereas positive integration would depend on unanimity in the Council that, for structural reasons, would be unlikely (Scharpf 1996; Esty & Geradin 2000, p. 238). Others have argued that a race to the top could also emerge (Holzinger & Knill 2004, p. 26; Vogel & Kagan 2004, p. 17; Knill & Lenschow 2005, p. 123). Susanne Schmidt, however, argued that more often than not, the effect of negative integration at the national level is legal uncertainty, which means that the extent of the remaining national scope of action is ambiguous (Schmidt 2004, p. 353; 2008). Thus, whereas between the early 1980s and the mid-1990s, national policy makers felt somehow restricted in their scope of action by European law (Hartkopf & Bohne 1983, p. 168; Miiller 1986, pp. 11, 136, 266; Pehle 1998, p. 248), the precise nature of this restriction and, in turn, what the remaining options would be, were not at all clear.

This uncertainty had a substantial dimension as characterised above. Notification procedures that allowed for a switch from negative to positive integration added a procedural dimension to this uncertainty, as will be argued below.

The Procedural Arsenal: Notification Procedures

Provided that no European measure has been proposed or adopted in a particular field, only negative integration is at work, i.e., all (new) measures are analysed (politically and legally) with regard to whether they could be a discriminating and disproportional barrier to trade according to the Treaty (then Art. 30, now Art. 34). Yet, in all fields where there is a Community competence, the Commission, with its monopoly of initiative, always has the option to "switch" from negative to positive integration. The procedural links between negative and positive integration are notification procedures, which entail relevant standstill duties that have a pivotal effect on what member states can and cannot do. Whereas negative integration has attracted substantial scholarly attention, notification procedures definitively have not (however, see von Bogdandy 2003). Although there are three different notification procedures that apply in environmental policy, the most relevant is the so-called information procedure in the field of norms and technical regulations that was introduced in 1983 and is now well established (O.J. 1983 No, L 109/8, see O.J. 1998 No. L 204/37). Accordingly, member states are required to inform the Commission of all proposals for all types of technical regulation. If the Commission or another member state expresses, within three months, the opinion that this measure could be a barrier to trade, the measure must not be implemented for six months. This standstill period is extended to 12 months if the Commission informs the member state within three months that it intends to present a proposal for a European measure.

The objective of this procedure is to prevent a national measure from endangering the common market and challenging European harmonisation (Meyring 2003, p. 945). Furthermore, this process provides the procedural framework for switching from negative to positive integration. Although I have presented only the most important notification procedure here (for a full account, see Toller 2012a, pp. 264-266 and 2012c, pp. 6-7), all three procedures have four aspects in common. First, they allow for control if planned national measures could be a barrier to trade. Second, they oblige the member state planning such a measure to "stand still", i.e., not to implement the measure so the Commission and other member states can examine the proposed measure before the measure becomes effective. Third, they allow the Commission to identify areas of interest in which it might wish to "switch" from negative to positive integration. Fourth, in this case, member states are further obliged not to act.

Although the two procedures not presented here have been of limited relevance, the information procedure has gained paramount importance, especially after the ECJ decided in the mid1990s that failure to report a national measure would lead to the non-applicability of this national measure (ECJ C-194/94; von Bogdandy 2003, p. 38; Meyring 2003, p. 957). Hundreds of measures are reported by member states each year, and roughly half of all product-related regulations are sent to Brussels (Toller 2012a, p. 266).

My argument thus far has been that negative integration produces substantial legal uncertainty at the national level because member states are unaware of the extent to which their scope for autonomous regulation has been restricted and what options remain possible. Moreover, the notification procedures generate procedural uncertainty, as member states must not implement national measures during the standstill periods. Furthermore, there is always the possibility that the Commission will consider the political salience of an issue as a reason to propose a European measure, in which case the autonomous national options would finally come to an end. This situation, as I will show in the next paragraph, has been a strong reason why the German government has resorted to non-legislative regulation.

Recourse to Voluntary Agreements

The specific German type of voluntary agreements has a number of characteristics that have made these agreements interesting options in a pressing situation in which the government wishes to adopt an environmental, in particular a product-related, regulation. First, the agreements--albeit always the result of bilateral negotiations--take the form of unilateral voluntary self-obligations by business (usually business associations). Second, in Germany (as opposed to the Dutch case), voluntary agreements are gentlemen's agreements and thus have no legal quality. As such, these agreements can neither be assigned to the government as an actor nor challenged as legally relevant acts. Thus, they cannot violate the EC treaty and are hence the perfect instruments with which to bypass these supranational norms. Oldiges made this observation in his 1973 article on voluntary agreements (Oldiges 1973). In 1992, Palme discussed the problem that national environmental regulations could come to an end because they are seen as measures having equivalent effect, whereas

"private self obligations do not fall in this category, they cannot constrain the free trade of goods because enterprises are free to comply with them. [...] Such options are left to national environmental policy" (Palme 1992, translation from German A.E.T.).

Third, as voluntary agreements have no legal quality, they do not have to be reported and cannot bring national decision making to a standstill. Thus, these agreements allow for the evasion of not only substantive but also procedural uncertainty. Fourth, when the Commission is planning to propose a directive, member states wishing to adopt a national measure feel that they have to hurry and present their measure before the Commission officially proposes one of its own.7 In this situation, voluntary agreements can typically be adopted more rapidly. (8) Fifth, voluntary agreements allow the government to not only "do something" about a publically perceived environmental or health problem, thus demonstrating to its citizens that it is able to act, but also set standards (in terms of substantive standards and organisational models) that (politically speaking) cannot be ignored by a European proposal. If a government chooses a voluntary agreement instead of a legislative regulation for these reasons, we refer to this action as evasion.

Relevance of European Law in the Cases Examined

European law did not play a relevant role in only three of the 13 in-depth case studies noted above. In all other cases, the threat coming from European law was a relevant argument for the adoption of a voluntary regulation instead of a legislative measure. In most cases, both substantive and procedural uncertainty played a role: the government was uncertain as to whether a national product regulation would be considered compatible with the Treaty, and irrespective of the outcome, once the notification procedure would begin, the national ability to act would vanish. In a number of cases, the agreement was adopted shortly before a European measure was proposed. In all cases, the federal government did not report the agreement (even if it was, in some cases, combined with an ordinance), whereas in some cases, the Commission found that at least the ordinance had to be reported. Finally, in a number of cases, the European proposal--in terms of regulatory substance or strategy--bore some similarity to the German voluntary agreement, which supports the argument made above.

Below, I support my argument by presenting a selection of three cases in the field of product regulation (from the fields of health and climate protection). The cases involving waste management, as noted briefly, display very similar patterns.

Beginning with the regulation of substances harmful to human health or the environment, asbestos has been known for its harmful health impact since the 1930s. In the 1970s, a number of minor measures, e.g., for the protection of workers, were adopted in Germany (Lautenbach et al. 1992, p. 47). Towards the end of the 1970s, asbestos was banned in Denmark, the Netherlands and Sweden. A report issued by the German Environmental Agency (UBA) in 1980 was initially intended to recommend a ban on specific asbestos products but ultimately did not. A major reason for this failure (in addition to strong pressure by industry and the fear of job losses) was that the UBA was uncertain whether a national asbestos ban (by an ordinance) would be considered a barrier to trade by the European Commission (UBA 1980, pp. 391392; Lautenbach et al. 1992, pp. 53; Wicke & Braeseke 1998, p. 143). In addition, the parliamentary secretary of state argued before the Bundestag in 1980 that

"the European Commission is preparing a directive containing particular protective measures and bans of asbestos" (BT P1Pr 9/9. Session 255),

which might have meant that Germany would no longer be able to adopt a national measure. Indeed, when the Commission presented a proposal in the spring of 1980, a standstill was imposed on national regulation. However, at that time, the understanding of the precise nature of the European restriction by the relevant actors appears to have been vague, to say the least. Moreover, the press cited the 10,000 persons dying of cancer each year in Germany as a consequence of not banning asbestos in construction. Thus, the government came under strong pressure. As a result, a voluntary agreement was negotiated in 1982 (see table 1, case nr. 1), obliging industry to reduce the share of asbestos in construction material by between 30 and 50 per cent and to prefabricate 95 per cent of asbestos products to reduce the risk of unnecessary exposure (Toller 2012a, p. 99). A further agreement adopted in 1984 required that no asbestos be used in construction after 1990 (Lautenbach et al. 1992, p. 48). Industry complied fully with these agreements. In 1991 and 1994, asbestos was banned by an ordinance in line with the European directive. By then, however, asbestos was no longer used in construction in Germany.

Another case in point in the context of climate protection is the regulation of CFCs (see table 1, case nr. 10). In 1974, Rowland and Molina argued for a connection between CFCs and the depletion of the ozone layer, which, however, could not be proven scientifically before 1987/88 (Lautenbach et al. 1992, p. 63). In 1977, CFCs became a highly disputed issue in the Bundestag (Toller 2012a, p. 169). After the hole on the ozone layer was "discovered" in 1985, the CFC issue became even more prominent in the Parliament, which, by the end of 1987, established an Inquiry Commission on the protection of the atmosphere. In 1977 and 1987, the federal government negotiated agreements with the aerosol industry on the reduction of the use of CFCs in sprays (Toller 2012a, pp. 167-182). Against the backdrop of the reduction mandates in the Vienna Agreement for the Protection of the Ozone Layer (1985) and the Montreal Protocol (1987), adopted to implement the Vienna agreement, it became clear towards the late 1980s that the absolute amount of CFCs produced in Germany had not decreased, although industry had complied with the reduction objectives. Thus, in 1989, the members of the Bundestag (9) called upon the federal government to adopt more stringent

restrictions on the use and production of CFCs (BT-Drs. 11/4133, p. 6; Toller 2012a, p. 172). However, MPs were well aware that EC law restricted the national options. As MP Muller suggested, "The problem is that with such strategies we are increasingly dependent on legislation and conditions set by the European Commission. It is indeed not at all clear if we can achieve a national ban [on CFCs]" (BT P1Pr 11/946439).

One year later, the Minister for the Environment, Klaus Topfer, made it very clear:

"We did [thus far] speak of voluntary agreements, not because we naively believe that we can expect industry to achieve everything they promise but because voluntary agreements do not entail problems with regard to European law. This is one reason why we have chosen this option. An authoritative ban, which I adopt today, has to be reported to the European Commission. The time needed for notification is much more than the time we would need to negotiate a voluntary agreement. We chose this option not for a lack of backbone in facing industry but because we are aware of the options that European law leaves open to us" (Klaus Topfer, Minister of the Environment, 1989. BT P1Pr 11/164, p. 12407). (10)

In 1988, the EC adopted a regulation on CFCs, which, however, did not establish particularly demanding reductions. Yet, no further voluntary agreement was concluded in Germany, as neither the government nor Parliament was content with the further reductions offered by industry. The German government evaluated whether a more demanding national regulation would be viable in terms of European law. Ultimately, the government decided that because the European regulation was based on Art. 130t ECT, member states should have the option to adopt stricter regulation. In late 1989, the Environment Ministry presented a proposal for an ordinance restricting the use of CFCs in sprays (BT-Drs. 11/6203, p. 35). The proposal banning CFCs in sprays, based on the act on the regulation of chemical substances (1990), was reported to the Commission and adopted in 1991 (BGB1. 1991 I, p. 1090; Osorio-Peters 1999, p. 237, BT-Drs. 11/8166, p. 15f.).

Against the backdrop of the cases presented above, the regulation of PCP (see table 1, case nr. 2) is, from a certain time on, a deviant case. Here, the government switched from evasion to confrontation. Thus, it appears plausible that evasion was indeed a rational choice under the given conditions.

PCP is a substance that had been used in wood protection since the end of the 19th century (SRU 2004, p. 443) and was used in Germany from the late 1960s on for indoor use (Luhmann 2001, p. 194). Towards the end of the 1970s, an increasing number of cases were observed in which humans displayed serious toxic reactions to PCP, such as fatigue, headache and concentration deficits (Jacob 1999, p. 156; Luhmann 2001, p. 104). However, it remains impossible to demonstrate a causal effect of PCP on human health. In the 1970s and early 1980s, a number of safety recommendations regarding PCP were adopted. In 1983, victims of PCP founded an association that engaged in public awareness efforts and sought an accounting of the economic damage to both buildings and human health. In Frankfurt in 1984, persons allegedly harmed by PCP went to court and initiated what was later known as one of the largest environmental criminal cases in Germany.

Beginning in 1983, political parties in the Bundestag began calling upon the federal government to adopt a ban of PCP. The government, however, faced the problem that there was no scientific proof of the harmfulness of PCP (Jacob 1999, p. 158). As industry realised it had to react to public concerns, in 1984, the association of manufacturers of wood-protecting substances declared that they would stop using PCP by April 1985, which they did. In 1986, after a major industrial accident at a plant owned by Sandoz (which had nothing to do with PCP), the government (and the newly established Ministry for the Environment in particular) was strongly criticised for the slow implementation of the chemical substances act. Thus, the newly appointed Minister for the Environment, Walter Wallmann, proposed an ordinance banning PCP in wood-protecting substances (11) that was adopted by the cabinet of his successor, Klaus Topfer, in 1987 and was proudly presented as the first ordinance based on [section]17 of the chemical substances act that totally banned a substance to protect human health or the environment (Presse- und Informationsamt der Bundesregierung, Bulletin: May 27, 1987).

In the literature, the PCP ban was regarded as an example of the types of regulation for the sake of human health and the environment that EC member states can adopt (e.g., Pehle 1998, p. 253; Jacob 1999, p. 157; Kramer 1995, p. 117). In 1987, the ordinance was reported to Brussels (ECJ C-41/93, p. 1844). The Commission received an objection raised by the Belgian government that extended the standstill period to six months (Jacob 1999, p. 157). The Commission then requested that the federal government cease implementation of the ordinance for 12 months, as it planned to propose a European directive on the regulation of PCP (ECJ C-41/93, p. 1844). In 1988, the Commission presented its proposed directive. In 1989, the federal government decided to implement the ordinance adopted in 1987 (BGB1. I 1989, p. 2235). The European directive that was adopted in 1991 (against the votes of Germany, Denmark and the Netherlands in the Council) stipulated far-reaching restrictions on the use of PCP but no general ban (O.J. 1991 No. L 85/34; Jacob & Janicke 1998, p. 526).

The German government, supported by the Bundestag's environmental committee, insisted on implementing the stricter German provision and called upon the safeguard clause envisaged in Art. 100a ECT (now Art. 114 TFEU), according to which stricter national measures are allowable as long as they are not arbitrary barriers to trade (see above). The European Commission confirmed the information presented on the first of August 1991 by the German government that Germany intended to apply the German ordinance instead of the European directive (ECJ C41/93, p. 1846; O.J. 1992 No. C 334/8; Kramer 1995, p. 117). At that point, France took action before the ECJ, arguing that the Commission was wrong to confirm the German safeguard and thus was violating Art. 100a ECT. (12)

This case was the first on the new Art. 100a, introduced by the SEA in 1987 (ECJ C-41/93, p. 1831). In its decision taken in 1994, the court rejected the Commission's decision because it violated elementary formal requirements but left the German ordinance as such untouched (ECJ C-41/93, p. 1831). Therefore, although the German ban had been in force since 1989, legal certainty was reached only ten years after the voluntary agreement had been adopted. Thus, this deviant case demonstrates the difficulties the government potentially avoided in the cases in which it restricted itself to regulating via voluntary agreements.

Similar patterns can be identified in the field of product-related waste management regulation (Toller 2012a, pp. 272-276). The regulation of used batteries and scrap cars (see table 1, cases 7 and 8) essentially followed the evasion pattern outlined above (similar to the asbestos and CFC cases). Here, as in the case of packaging waste, the government seemed almost haunted by the problem of legal uncertainty and the threatening standstill for national options once the Commission had presented a proposal for a European measure. Moreover, in adopting voluntary agreements, the government was quite successful in influencing the regulatory substance and organisational structure of the European measures adopted (particularly in the case of batteries but also in the case of scrap cars). The case of the regulation of packaging waste (see table 1, case 6) is, however, a deviant case similar to the PCP regulation: beginning in 1988, the government opted for confrontation. The case was settled by an ECJ decision in 2004 (Haverland 1999, p. 259; ECJ C 463/01, p. 4). Over 16 years, the core of the conflict was whether an environmentally driven refund duty for drink containers was a disproportional barrier to trade and whether there were sound ecological reasons for stipulating specific quotas for multi-use systems (as the German ordinance on package waste adopted in 1991 did; see Haverland 1999, pp. 264-270; ECJ C-463/01).

Positive Integration

In environmental policy as much as elsewhere, the transition from negative to positive integration has occurred in stepwise fashion at different paces and times and still remains incomplete in some fields, although today, most sub-sectors of environmental policy have been the subject of complete or almost complete European harmonisation. Even once harmonisation has occurred, negative integration can still be at work, as the ECJ emphasised (ECJ C-463/01; Stumpf 2005, p. 54).

The major point related to my argument here is that European directives usually require legislative acts to be transposed into European law. Whereas voluntary agreements that have (as the Dutch agreements do) the character of a private contract can under certain conditions be used (or upheld) to transpose a European directive, the German version of voluntary agreements is not accepted as a suitable measure for transposing directives by the European Commission (O.J. 1996 No. L 333/59).

Thus, as much as negative integration (by giving government actors a strong motivation for evasion) fostered the use of voluntary agreements, positive integration was a strong reason for the government to rescind existing voluntary agreements (as occurred in the field of waste management policy--not only in Germany but also in Austria and even in the UK; Okoburo & Ogut 2004, p. 71; Baggott 1986, p. 65). Moreover, as contemporary German environmental policy is strongly determined by Brussels (Toller 2008b), there are very few areas left in which there are no European directives and in which the government can autonomously conclude voluntary agreements.

Without neglecting other factors, as discussed above and below, we can affirm that the development in European law as described above was a major cause both of the extensive use of non-legislative, voluntary agreements between the early 1980s and the late 1990s and of the end of their use more recently.

3.2.2 What of Functionality and Business Power?

When analysing my case studies, I also considered both the 'better regulation' hypothesis and the power-centred hypothesis.

If my study came to the conclusion that business power did not play any role in averting statutory regulation and triggering the adoption of voluntary agreements, something would be missing. (13) The assumption that business prefers voluntary regulation to statutory regulation and no regulation to voluntary regulation (14) (Wicke & Braeseke 1998, p. 36; Croci 2005, p. 13) proves correct in most cases. (15) In most of the cases analysed, business associations (and individual companies) not only expressed their clear preference for voluntary instead of statutory regulation but also used the strategy of threatening major job losses to support their position. This was the case, for example, when the harmful effects of asbestos (Lautenbach et al. 1992, p. 53), the waste management of used batteries or drink packages, early measures for C[O.sub.2] reduction, the termination of nuclear energy and many other environmental regulations were at stake (Toller 2012a. pp. 346-377).

However, whereas in the 1980s, as in the asbestos case, threatening the loss of thousands of jobs appeared quite convincing, more recently, even threatening job reductions has become more complex. For example, when the quota for combined heat and power generation was at stake, there were 20,000 jobs in public utility companies versus 400,000 jobs in coal mining. Thus, although constellations have become more complex, contemporary business is more successful not in confronting common good interests with specific economic interests but rather in re-framing the economic interests of different branches as competing common good interests.

However, the threat of closing factories and eliminating jobs does not always have the same weight. The effect is stronger if the overall economic situation is tense, the government is known to have sympathies for business (which can be, but is not necessarily, dependent on party politics), and the Ministry for the Environment is particularly weak in relation to the Ministry of Economics. (16) Argumentative support from Brussels also helps.

Thus, business power does have a role to play in explaining the use of voluntary agreements, but it is far from being deterministic as the popular power-centred explanations would suggest.

What about the better-regulation hypothesis? What role do the much-debated deficiencies of statutory regulation, in particular for addressing specific types of problems, play in explaining the use of voluntary agreements?

In my 13 case studies, there is only one case in which a statutory alternative seemed unable to achieve the defined aims (quite apart from the question of whether the voluntary agreement was expected to do so). When in 1994/1995, after eight years of debate, the proposed heat utilisation decree (Warmenutzungsverordnung) was finally dropped in favour of the climate declaration (table 1, case 12), it had indeed proven difficult to regulate all technical details associated with heat reutilisation for all types of plants (Kohlhaas & Praetorius 1994, p. 46; Toller 2012a, p. 338). The case of EDTA regulation is the only case in which--given the lack of suitable substitutes--the production of knowledge was expected from the voluntary agreement and its implementation via expert networks. However, there were a number of cases (such as the regulation of PCP, CFCs and NTA) in which, when regulation was deemed necessary, the scientific proof for the problematic nature of a substance was not sufficient to satisfy the demands required by law. This situation, however, was less a problem of regulatory law as such than one of the stipulations regarding the burdens of proof in the specific law (which in the case of the chemicals act were relaxed in 1990; Toller 2012a, p. 334). Similarly, a number of regulations (ordinances) could not be adopted easily because the legislative act on which the ordinance was intended to be based did not entail the necessary authorisation (e.g., in waste management until the revision in 1986). This again was not a problem of regulatory law as such but a deficiency of a specific law that, given the necessary political majorities, could be remedied (Toller 2012a, pp. 335-338).

Thus, the literature clearly exaggerates the structural shortcomings of traditional policy instruments while not adequately realising that the deficiencies of such regulatory instruments are often the result of a lack of political will and not legal-technical capabilities.

What is the Role of the Success or Failure of the Agreements?

Although the question addressed in this article is why voluntary agreements are or are not adopted, a related question (which was raised following the examination of the regulatory courses in chapter 2) is why, once agreements are adopted, they continue to be implemented or are terminated. The consensus in the literature is that agreements are terminated when they do not succeed, whereas they continue to be implemented when they do succeed (e.g., UMK 1997, p. 3; Hansen 2005, p. 85). Although the success of the agreements is beyond the scope of this article, it is worth noting that most agreements were implemented and the defined objectives were achieved (see, also, Wicke & Braeseke 1998). Whereas agreements regarding the termination of the use or production of certain substances were terminated because they were successful (the adoption of a statutory regulation simply occurred to secure the legal situation), in the case of other agreements (e.g., in the field of waste management policy), success or failure had no effect on the continuation or termination of the agreements. Although some clearly failed agreements continued for a long period (as in the field of package waste regulation, when the necessary political majority to adopt a statutory regulation was not reached in the Bundesrat), successful agreements (such as, e.g., the scrap car agreement) were terminated because European law required it.

4. Conclusion

This article began with one surprising finding and one assumption. The finding was that whereas environmental voluntary agreements have been considered "an element of a distinctly 'German' policy mix" for decades (Lees 2007, p. 175), they were relevant in the 1980s and early 1990s, but their use almost completely ceased in the new century. Second, when seeking the causes of both the widespread use of voluntary agreements in the 1980s and 1990s and the more recent abrupt termination of their use, neither explanations based on functional arguments nor those based on the power of business can account for this development. Instead, beginning from an analytical approach developed by Bocher and Toller, institutions (in particular, European law and its effect on national policy decisions) and actors (in particular, governmental actors determined by party politics) were assumed to play a major role, whereas political processes were regarded as less means-oriented and characterised by step-by-step problem-solving than is usually the case in policy analysis.

What can account for both the rise and the fall of voluntary agreements as a relevant instrument in German environmental policy?

If we consider the driving forces for the rise of voluntary agreements, in particular from the mid-1980s to the late 1990s, there are four relevant aspects elaborated in this article. The potentially most important aspect was the uncertainty stemming from European law, in particular, when only negative integration was at work. This uncertainty was a strong motivation for governmental actors to resort to informal forms of regulation, namely, voluntary agreements. Almost as important were problems with the legal basis or burden of proof of an ordinance, which, however, were not due to the general problems of regulatory law but simply to a lack of opportunity or political will to adopt legal acts with a lower burden of proof or a more explicit authorisation for ordinances regulating business.

Business power (i.e., its preference for voluntary regulation instead of statutory regulation and its threats of site relocation and job losses) played a role in most cases, but its impact was highly dependent on other factors, such as party politics, the overall economic situation and the actual power of the Ministry of the Environment vis-a-vis the Ministry of Economics (see below). In particular, the peaks in the mid-1990s are also an effect of the debate on the "Standort Deutschland", in which--in Kingdons's sense--the Ministry under Angela Merkel was searching for problems to which the instrument was suited rather than the reverse. In this phase, voluntary agreements became an element of deregulation, which they had not been before. A further aspect that has not been described thus far is the policy of the German Trust agency (voluntary agreements are agreements among competitors and thus cartels), which had long tolerated the agreements (Toller 2012a, pp. 244-252); however, this policy began to change in the mid-1990s.

As suggested above, the factors identified did not operate in isolation. In particular, the institutional argument (European law) is never deterministic but only shapes how actors perceive their options. Thus, for political actors who were in favour of voluntary agreements (e.g., Angela Merkel; see Toller 2012a, p. 308), arguments emphasising the problems stemming from European law (or that the burden of proof mandated in the law was excessive or the authorisation would not suffice for an ordinance) were more welcome than for others. However, whether in a single case the government opted for evasion or for confrontation also depended on the "felt" relevance of a policy issue to the broader audience (as in the case of package waste) and on what we called process dynamics (as in the case of PCP, when the government felt political pressure to do "something").

What can account for the fall of voluntary agreements, namely, the almost complete cessation of their use? First, again, European law must be considered: negative integration producing legal uncertainty has not fully ceased but has been considerably reduced in its scope, as throughout the 1990s most fields of environmental policy were harmonised by European directives (see, e.g., Toller 2012a, pp. 284-285 for the field of waste management policy). This harmonisation meant not only that legal uncertainty declined as a driving force of adopting voluntary agreements but also that voluntary agreements could no longer be in place (nor could new ones be adopted). At present, 80 per cent of German environmental legislation is influenced in one way or another by European law (Toller 2008b). There is simply little room left for purely national voluntary regulation.

Second, party politics clearly played a role. The decrease in agreements adopted after 2001 is to a large extent due to the Green Minister for the Environment Jurgen Trittin's disapproval of this voluntary instrument. One case in point is that a further voluntary agreement on package waste was suggested in 2001 by business associations and various Lander (Federal States), which was rejected by Trittin (Toller 2012a, p. 143). Moreover, Jurgen Trittin considered EU law to be a good reason to terminate existing agreements (Toller 2012a, p. 311).

Finally, in the early 1990s, a number of problematic constellations emerged in the field of waste policy (problematic at least from the point of view of competition policy). Thus, in the mid-1990s, the German Trust Agency changed its tolerant policy towards voluntary agreements. In 1996, as Angela Merkel was planning to adopt a further agreement in the field of package waste, the Trust Agency intervened (Toller 2012a, p. 138).

If the use of voluntary agreements in German environmental policy has practically come to an end, what is the benefit of this article except for as an examination of history?

First, although the use of voluntary agreements in environmental policy has ceased and will most likely not resume, voluntary agreements remain a relevant instrument in many other policy fields. Recent examples are the Corporate Governance Codex, the self-obligation by the pharmaceutical industry and the current heated debate over how to regulate the proportion of women in the management of DAX companies. Second, although this type of evasion of EU law has not been observed in other EU member states, a similar mechanism has been identified beyond the European Union: the emergence of transnational certification schemes was, among other things, driven by the need to evade the restrictions deriving from WTO free trade law (see Bartley 2003 and Toller 2012c). Thus, when explaining public policies, we should more generally have a greater interest in institutions and their idiosyncratic effects.

Third, in my 13 in-depth case studies on environmental policy decision making from the late 1970s to 2005, I found substantial evidence for our analytical approach that considers actors, institutions, problem structures, instrumental alternatives and process dynamics as explanatory factors for environmental policy results. I found support not only for emphasising the pivotal but idiosyncratic role of institutions but also for our understanding of the political process. The political processes analysed in this study were not step-by-step problem-solving processes but rather often highly contingent processes driven by, among other factors, actors seeking to prove their capability to act. Such actors are motivated by the desire to implement their pet solutions, are un certain with regard to the latitude European law leaves them to act and do not wish to make fools of themselves. Indeed, in some instances, solutions have been in search of problems instead of the reverse (Bocher & Toller 2012b, pp. 461-470).

References

Aggeri, F. (1999). Environmental policies and innovation. A knowledge-based perspective on cooperative approaches. In: Research Policy, 28, 699-717.

Baggott, R. (1986). By Voluntary Agreement: The Politics of Instrument Selection. In: Public Administration, 64, 51-67.

Bartley, T. (2003). Certifying Forests and Factories: States, Social Movements, and the Rise of Private Regulation in the Apparel and Forest Products Fields. In: Politics and Society, 31, 433-464.

Beder, S. (2002). Global Spin. The Corporate Assault on Environmentalism, revised edition, Vermont: Green Chelsea.

Black, J. (2001). Decentring Regulation: Understanding the role of Regulation and Self-Regulation in a ,Post-Regulatory World'. In: Current Legal problems, 54, 103-146.

Blauberger, M., Toller, A. E. (2011). Competition Policy. In: Michele Knodt; Hubert Heinelt (Eds.), Policyfields in the European Multi-Level System. Instruments and Strategies of European Governance (pp 123-152), Baden-Baden: Nomos.

Bocher, M., Toller, A.E. (2007). Instrumentenwahl und Instrumentenwandel in der Umweltpolitik. Ein theoretischer Erklarungsrahmen. In: K. Jacob, K. Biermann, P.O. Busch, P. Feindt.(Eds), PVS-Sonderheft Politik und Umwelt (pp. 299-322). Wiesbaden: Westdeutscher Verlag.

Bocher, M., Toller, A.E. (2012a). Umweltpolitik in Deutschland. Eine politikfeldanalytische Einfuhrung. Wiesbaden: VS/Springer.

Bocher, M., Toller, A.E. (2012b). Eigendynamik und Zufall als Triebkrafte der Umweltpolitik? Ein Ansatz zum Erklaren umweltpolitischer Entscheidungen. In: Zeitschrift fur Um weltpolitik, 4, 450-479.

Bogdandy, A.v. (2003). Links Between National and Supranational Institutions: A Legal View of a New Communicative Universe. In: Beate Kohler-Koch (Ed.), Linking EU and National Governance (pp. 24-52), Oxford: Oxford University Press.

Buchanan, J. M., Tullock, G. (1975). Polluters' Profits and Political Response: Direct Control Versus Taxes. In: The American Economic Review, 65(1), 139-147.

Castles, F. G. (1982). The impact of parties: Politics and policies in democratic capitalist states, London: Sage

Croci, E. (2005). The Economics of Environmental Voluntary Agreements. In: Edoardo Croci (Ed.), The Handbook of Environmental Voluntary Agreements, Design, Implementation and Evaluation Issues (pp. 3-30), Dordrecht: Springer.

Esty, D. C., Geradin, D. (2000). Regulatory Co-Opetition. In: Journal of international economic law, Oxford, 3 (2), 235-255.

Geradin, D. (1993). Free trade and environment protection. In: Journal of transnational law & policy, 2, 144-197.

Geradin, D. (1997). Trade and the enviroment. A comperative study of EC and US law. Cambridge: University Press.

Glasbergen, P. (1999). Tailor-Made Enviromental Governance: On the Relevance of the Covenanting Process. In: European Enviroment, Journal of European Enviromental Policy, 9(2), 49-58.

Greer, J., Bruno, K. (1996). Greenwash. The Reality Behind Corporate Environmentalism, Penang: Third World Network/ New York: The Apex Press.

Grimeaud, D. J. E. (2004). Convergence or Divergence in the Use of >> Negotiated Environmental Agreements << in European and US Environmental Policy: An Overview. In: Vig. N.J., Faure, M.G. (Eds.), Green Giants? Environmental Policies of the United States and the European Union (pp. 159-182), Cambridge MA.

Gunningham, N., Rees, J. (1997). Industry Self-Regulation: An Institutional Perspective. In: Law & Policy, 19 (4), 363-413.

Hall, P.A., Taylor, Rosemary C. R. (2006). Political Science and the Three New Institutionalisms. Political Studies, 44 (5), 936-957.

Hansen, L.G. (2005). Aspects of the Political Economy of Environmental Voluntary Agreements. In: Edoardo Croci (Ed.), The Handbook of Environmental Voluntary Agreements, Design, Implementation and Evaluation Issues (pp. 67-89), Dordrecht: Springer.

Hansjurgens, B., Kock, W (2003). Kooperative Umweltpolitik--Eine Einfuhrung. In: Hansjurgens, B., Kock, W., Kneer, W. (Eds.) Kooperative Umweltpolitik (pp 9-15), Baden-Baden: Nomos.

Hartkopf, G., Bohne, E. (1983). Umweltpolitik, Part1: Grundlagen, Analysen und Perspektiven, Opladen: Westdeutscher Verlag.

Haverland, M. (1999). National Autonomy, European Integration and the Politics of Packaging Waste, Amsterdam: Purdue University Press.

Hibbs, D. A. (1977). Political Parties and Macroeconomic Policy. In: American Political Science Review. 71(4), 1467-1487.

Hofman, P. S., Schrama, G. J.I. (2005). Dutch Target Group Policy. In: T. de Bruijn, V. Norberg-Bohm (Eds.), Industrial Transformation. Environmental Policy Innovation in the United States and Europe (, pp. 39-63), Cambridge, London: The MIT Press.

Holzinger, K., Knill, C. (2004). Competition and Cooperation in Environmental Policy: Individual and Interaction Effects. In: Journal of Public Policy, 24(1), 25-47.

Immergut, E.M. (2011). Demokratietheorie und Staatstatigkeit: Vier Modelle von 'Policy, Politics and Choice'. In: Der Moderne Staat, 4(1), 69-86.

Immerzeel-Brand, E. (2002). Assessing the performance of negotiated environmental agreements in the Netherlands. In: P. ten Brink (Ed.), Voluntary Environmental Agreements. Process, Practice and Future Use (pp. 384-389), Sheffield: Greenleaf Publishing.

Ingram, K. (1999). From Sparring Partners to Bedfellows: Join Approaches to Enviromental Policy- Making. In: European Enviroment, 9 (3), 41-48

Jacob, K. (1999). Innovationsorientierte Chemikalienpolitik. Politische, soziale und okonomische Faktoren des verminderten Gebrauchs gefahrlicher Stoffe, Munchen: Herbert UtzVerlag.

Jacob, K., Janicke, M. (1998). Okologische Innovationen in der chemischen Industrie: Umweltentlastung ohne Staat? Eine Untersuchung und Kommentierung zu 182 Gefahrstoffen. In: Zeitschrift fur Umweltpolitik und Umweltrecht, 21(4), 519-547.

Jordan, A., Wurzel, R.K.W., Zito, A. (2005). The Rise of 'New' Policy Instruments. In: Comparative Perspective: Has Governance Eclipsed Government? Political Studies, 53 (3), 477-496.

Kingdon, J.W. (2003). Agendas, alternatives, and public policies, 2. Edition, New York: Addison-Wesley Educational Publishers inc..

Kiser, L., Ostrom, E. (1982). The Three Worlds of Action: A Metatheoretical Synthesis of Institutional Approaches. In: Ostrom, E. (Ed.), Strategies of Political Inquiry (p. 179222), Beverly Hills.

Kleger, H. (1995). Verhandlungsdemokratie. Zur alten und neuen Theorie des kooperativen Staates. In: Voigt, R. (Ed.), Der kooperative Staat. Krisenbewaltigung durch Verhandlung? (p. 93-117), Baden-Baden: Nomos.

Knill, C., Lenschow, A. (2005). Compliance, Communication and Competition: Patterns of EU Environmental Policy Making and Their Impact on Policy Convergence. In: European Environment, 15, 114-128.

Kohlhaas, M., Praetorius, B. (1994). Selbstverpflichtungen der Industrie zur C[O.sub.2]-Reduktion. Moglichkeiten der wettbewerbskonformen Ausgestaltung unter Berucksichtigung der geplanten C[O.sub.2]-/Energiesteuer und Warmenutzungsverordnung, Berlin: Duncker & Humblot.

Kramer, L. (1995). E.C. Treaty and Environmental Law, London: Sweet& Maxwell. 2. Edition.

Lascoumes, P., Le Gales, P. (2007). Introduction: Understanding Public Policy through Its Instruments--From the Nature of Instruments to the Sociology of Public Policy Instrumentation. In: Governance, 20(1), 1-21.

Laufer, W. S. (2003). Social accountability and corporate green-washing. Journal of Business Ethics, 43, 253-267.

Lautenbach, S., Steger, U., Weihrauch, P. (1992). Evaluierung freiwilliger Branchenvereinbarungen im Umweltschutz, Koln: Bundesverband der Deutschen Industrie.

Lees, C. (2007). Enviromental Policy in the United Kingdom and Germany. In: German Politics, 16 (1), 164-183.

Liefferink, D., Arts, B., Kamastra, J. and Ooijevaar, J. (2009). Leaders and laggards in environmental policy: a quantitative analysis of domestic policy outputs. In: Journal of European Public Policy 16. 677-700.

Luhmann, H.-J. (2001). Die Blindheit der Gesellschaft. Filter der Risikowahrnehmung, Munchen: Gerling Akademieverlag.

Majone, G. (1976). Choice among Policy Instruments for Pollution Control. In: Policy Analysis 2, 589-613.

Majone, G. (1989). Evidence, Argument and Persuasion in the Policy Process, New Haven.

Mayntz, R., Scharpf, F.W. (1995). Der Ansatz des akteurzentrierten Institutionalismus. In: R. Mayntz, F. W. Scharpf (Eds.), Gesellschaftliche Selbstregelung und politische Steuerung (pp 39-72), Frankfurt: Campus.

Merkel, A. (1997). Umweltbezogene Selbstverpflichtungen--ein wichtiger Baustein moderner Umweltpolitik. In: L. Wicke, J. Knebel, G. Braeseke (Eds.), Umweltbezogene Selbstverpflichtungen der Wirtschaft--umweltpolitischer Erfolgsgarant oder Irrweg? (pp. 73-86), Bonn: Economica Verlag.

Meyring, B. (2003). Europarechtliche Stillhalteverpflichtungen bei der nationalen Gesetzgebung. In: Europarecht, 38(5), 949-959.

Mol, A., Liefferink D., Lauber, V. (2000). Introduction. In: A. Mol, A., V. Lauber, D. Liefferink (Eds.), The Voluntary Approach to Environmental Policy. Joint Environmental Policy-making in Europe (pp. 1-9), Oxford: Oxford University Press..

Muller, E. (1986). Innenwelt der Umweltpolitik. Sozialliberale Umweltpolitik--(Ohn-) Macht durch Organisation? Opladen: Westdeutscher Verlag.

OECD, 2003: Voluntary Approaches for Environmental Policy. Effectiveness Efficiency and Usage in Policy Mixes. Paris: OECD Publication Service.

Okoburo/Ogut, 2004: Environmental Governance und Umweltvereinbarungen, Wien. http://doku.cac.at/Governance_und_Umweltvereinbarungen.pdf, last view: Oct. 26 2011.

Oldiges, M. (1973). Staatlich inspirierte Selbstbeschrankungsabkommen der Privatwirtschaft. In: Wirtschaft im Recht, 2(1). 1-29.

Olson, M. Jr. (1965). The Logic of Collective Action, Cambridge: Harvard Univ. Press.

Osorio-Peters, S. (1999). FCKW-Ausstieg in Deutschland. In: Klaus Rennings (Ed.), Innovation durch Umweltpolitik, ZEW Wirtschaftsanalysen. Schriftenreihe des ZEW, Band 36, 227-251.

Palme, C. E. (1992). Nationale Umweltpolitik in der EG. Zur Rolle des Art., 100a IV im Rahmen einer Europaischen Umweltgemeinschaft, Berlin: Duncker& Humblot.

Pehle, H. (1998). Das Bundesministerium fur Umwelt, Naturschutz und Reaktorsicherheit, ausgegrenzt statt integriert? Das institutionelle Fundament der deutschen Umweltpolitik, Wiesbaden: Westdeutscher Verlag.

Peters, B. G. (2002). The Politics of Tool Choice. In: Lester M. Salamon, L.M. (Ed.), The Tools of Government. A Guide to the New Governance (pp 552-564), Oxford: Oxford University Press,

Porter, T., Ronit, K. (2006). Self-regulation as policy process: The multiple and criss-crossing stages of private rule making. In: Policy Sciences, 39, 41-72.

Prittwitz, V. v. (1990). Das Katastrophenparadox. Elemente einer Theorie der Umweltpolitik, Opladen.

Radaelli, C. (2004). Europeanization: Solution or Problem? EloP, 8 (16). Available online at: http://www.eiop.or.at/ei op/pdf/2004-016.pdf (last view: November 26 2012)

Radaelli, C., Exadactylos, T. (Eds.) (2012). Establishing Causality in Europeanization Research. London: Palgrave Macmillan.

Ritter, E.-H. (1990). Das Recht als Steuerungsmedium im kooperativen Staat. In: Staatswissenschaften und Staatspraxis, 1, 50-88.

Rose-Ackerman, S. (1995). Umweltrecht und -politik in den Vereinigten Staaten und der Bundesrepublik Deutschland, Baden Baden.

Scharpf, F. W. (1996). Politische Optionen im vollendeten Binnenmarkt. In: M. Jachtenfuchs, B. Kohler-Koch (Eds.), Europaische Integration (pp. 109-140), Opladen: Leske + Budrich.

Schmidt, S. K. (2004). Rechtsunsicherheit statt Regulierungswettbewerb: Die nationalen Folgen des europaischen Binnenmarktes fur Dienstleistungen, FernUniversitat Hagen.

Schmidt, S. K. (2008). Beyond Compliance: The Europeanization of Member States through Negative Integration and Legal Uncertainty. Journal of Comparative Policy Analysis, 10(3), 297-306.

Schmidt, V. A. (2007). Bringing the State Back Into the Varieties of Capitalism Discourse and back Into the Explanation of Change. Center for European Studies Program for the Study of Germany and Europe Working Paper Series 07.3, http://www.ces.fas.harvard.edu/publications/docs/pdfs/Sch midt.pdf, last view March 11 2008.

Seeger, B. (2003). Umweltpolitik in den 16 Landern: Wahlprogramme und Regierungshandeln. Ein Bundesvergleich, Dissertation, Heidelberg, archiv.ub.uni heidelberg.de/volltextserver/3667/1/Seeger_Promotion_05_07_2003.pdf, last view November 26 2012.

Shapiro, M. (1992). The European Court of Justice. In: Alberta Sbragia (Ed.), Euro-Politics. Institutions and Policymaking in the New European Community. Washington: Brookings Institutions Press.

SRU (Sachverstandigenrat fur Umweltfragen) (2004). Umweltgutachten des Rates von Sachverstandigen fur Umweltfragen: Umweltpolitische Handlungsfahigkeit sichern, BT Drs. 15/3600.

Stumpf, C.A. (2005). Die deutsche Verpackungsverordnung als Beschrankung der europaischen Warenverkehrsfreiheit. In: Zeitschrift fur das gesamte Lebensmittelrecht. 32(1), 47-64.

Toller, A.E. (2008a). Kooperation als Trend? Verwendungsmuster und Ursachen kooperativer Politikformen in den Niederlanden, Deutschland und den USA. In: Zeitschrift fur Vergleichende Politikwissenschaft, 2(2), 315-346.

Toller, A.E. (2008b). Mythen und Methoden. Zur Messung der Europaisierung der Gesetzgebung des Deutschen Bundestages jenseits des 80%-Mythos. In: Zeitschrift fur Parlamentsfragen, 39 (1), 3-17.

Toller, A.E. (2008c). Kooperation im Schatten der Hierarchie: Dilemmata des Verhandelns zwischen Staat und Wirtschaft. In: Schuppert, G.F., Zurn, M. (Eds.), Governance in einer sich wandelnden Welt, PVS-Sonderheft (pp 282-312), Wiesbaden: Westdeutscher Verlag.

Toller, A.E. (2009). Freiwillige Regulierung zwischen Staat und Markt: Der Deutsche Corporate Governance-Kodex (DCGK). In: Der Moderne Staat, 2 (2009), 301-322.

Toller, A.E. (2011). Voluntary approaches to regulation--patterns, causes and effects. In: D. Levi-Faur (Ed.), Handbook on the politics of regulation (pp. 499-510), Cheltenham: Edward Elgar..

Toller, A.E. (2012a). Warum kooperiert der Staat? Kooperative Umweltpolitik im Schatten der Hierarchie. Baden-Baden: Nomos.

Toller, A.E. (2012b). Voluntary Regulation under which Shadow? The Case of the German Pharmaceutical Regulation Scheme (FSA). Paper for the ECPR-Regulation Conference, Exeter University, June 26-29, 2012.

Toller, A.E. (2012c). Evasion to non-legislative regulation as a neglected mechanism of Europeanization. Paper for the ISA- Panel on "Resisting international norms", Paper for the ISA 2012 San Diego Conference, Panel on "Resisting International Norms.

Toller, A.E: (2012d): Concepts of Causality in Quantitative Approaches to Europeanization. In: Claudio Radaelli, Theofanis Exadactylos (Ed.): Establishing Causality in Europeanization Research, London: Palgrave McMillan, 44-63

Vogel, D., Kagan, R. A. (2004). Introduction. Dynamics of Regulatory Change. How Globalization Affects National Regulatory Policies. In: D. Vogel, R. A. Kagan (Eds.), Dynamics of Regulatory Change. How Globalization Affects National Regulatory Policies (pp. 1-41), Berkeley, Los Angeles, London.

Wicke, L., Braeseke, G. (1998). Selbstverpflichtungen als Instrumente des Umweltschutzes--Okonomischer T eil (Schlussbericht), Teilbericht. In: Jurgen Knebel/Lutz Wicke/Michael Gerhard: Selbstverpflichtungen und normersetzende Vertrage als Instrumente des Umweltschutzes, Institut fur UmweltManagement (IfUM), Berlin.

Wilks, S. (2005). Competition Policy. In: Helen Wallace/William Wallace/Mark A. Pollack (Eds.), Policy-making in the European Union (pp. 113-139), Oxford: Oxford Universitx Press..

Zittel, T. (1996). Marktwirtschaftliche Instrumente in der Umweltpolitik. Zur Auswahl politischer Losungsstrategien in der Bundesrepublik. Opladen, 130-142.

Annette Elisabeth Toller, Fern Universitat in Hagen

(1) On a broader variety of voluntary instruments, see Toller 2011.

(2) Although most voluntary agreements can be found in environmental policy, they are by no means restricted to this field. Early voluntary agreements concerned advertising for cigarettes and pharmaceuticals (Toller 2012a, pp. 83-84). Recently, the share of women in management (2001), the rules of corporate governance more generally (2002, Toller 2009) and the practices of the pharmaceutical industry (2004) have been addressed by voluntary regulation (see Toller 2012b).

(3) On the multitude of labels, which are very revealing of their authors' attitude, see Toller 2012a: 48.

(4) The Bavarian branch of the Christian Democrats, the CSU, however, occasionally displayed a more interventionist stance than its sister party, the CDU.

(5) In 1981, Denmark adopted a waste management regulation that only allowed beer and soft drinks to be sold in returnable containers that had to be licensed by the national environmental agency. Thus, e.g., aluminum cans and other one-way containers were banned. The European Commission attacked this regulation before the ECJ as an unjustifiable barrier to trade. The ECJ determined that the stipulation that drinks could only be sold in returnable containers was proportional, whereas the need to have containers licensed was not because it disadvantaged non-Danish producers and traders (C302/86; Geradin 1993: 183; Kramer 1995: 114).

(6) Whereas the free trade norm could be a challenge for product regulation, since the late 1980s, European state aid policy as part of a policy to invigorate European competition has also formed part of the arsenal of negative integration (Wilks 2005; Blauberger & Toller 2011). State aid policy has the particular potential to challenge national regulation of processes if this regulation, e.g., favours particular processes (e.g., of energy production) vis-a-vis others.

(7) I chose the formulation member states "feel" obliged because, given the lack of legal quality of the agreements discussed above, it does not really matter if such an agreement is adopted before or after the Commission proposes a directive.

(8) The assumption that voluntary agreements take less time to adopt than parliamentary legislation is considered a major reason for their use in the literature (Merkel 1997, p. 79; Liefferink 1997, p. 231; Okoburo & Ogut 2004, p. 12). In most of the cases analysed, the negotiation process took between 12 and 18 months. However, in some cases, the negotiations over agreements took years (Toller 2012a, p. 292).

(9) This is the only case in which we can observe a conflict between the Parliament and the government, whereas all other cases follow the logic of

the parliamentary system in that the parliamentary majority supports the position of the government while the opposition complains.

(10) Furthermore, at that time, the German act on the regulation of chemical substances did not provide a stipulation that allowed for the regulation of CFCs (Lautenbach et al. 1992, p. 67).

(11) As the voluntary agreement concluded in 1984 had been complied with and PCP was no longer used, the relevant regulatory result of this regulation was 1) to create legal certainty and 2) to avoid PCP-containing imports. Primarily, the regulation served a symbolic function to demonstrate the government's determination to totally ban a dangerous substance.

(12) The French government argued that the data submitted by the German authorities did not prove that the ban was justified by the particular situation in Germany. In particular, there was no proof that the environment was under such a threat that a regulation stricter than the European directive was required. Furthermore, the ban was considered disproportional given the trade restrictions that it caused (ECJ C-41/93, p. 1836). In terms of procedure, France argued that the Commission had not adequately ensured that the national safeguard clause was not a measure of arbitrary discrimination (ECJ C-41/93, pp. 1836, 1839).

(13) I conceptualised business actors (companies and associations) as one category of actors, but their behaviour also played a role regarding the structure of the problem (see Toller 2012a, pp. 312-330 and 344-348).

(14) Exceptions to this rule occur when particular branches are interested in protecting their present markets against future competitors (Buchanan & Tullock 1975) or when the branch of environmental technology benefits from the regulation of other branches (von Prittwitz 1990, p. 116).

(15) For a detailed discussion of assumed reasons, see Toller 2012a, pp. 315-318.

(16) The relationship between the Ministry of the Environment and other ministries, the Ministry of Economics in particular, has always been characterised by an inferior position of the former, yet with some variation over time (Toller 2012a, p. 243).

Table 1: Thirteen regulatory courses from 1977-2005 (Toller
2012a, pp. 90-201)

Health Protection

1. Regulation of Asbestos

1982/84 Voluntary Agreement I (building above ground)
1988 Voluntary Agreement II (building underground)
1991/1994 Statutory Regulation

2. Regulation of PCP (pentachlorophenol)
1984 Voluntary Agreement
1987 Statutory Regulation

3. Regulation of Nuclear Energy ("Atomausstieg")
2000 Voluntary Agreement

Pollution Regulation

4. Regulation of NTA (nitrilotriacetic acid)

1984 Voluntary Agreement I
1986 Voluntary Agreement II

5. Regulation of EDTA (ethylene diamine tetra acetate)

1991 Voluntary Agreement I
1998 Voluntary Agreement with the photography industry
2000 Voluntary Agreement II

Pollution Regulation

4. Regulation of NTA (nitrilotriacetic acid)

1984 Voluntary Agreement I
1986 Voluntary Agreement II

5. Regulation of EDTA (ethylene diamine tetra acetate)

1991 Voluntary Agreement I
1998 Voluntary Agreement with the photography industry
2000 Voluntary Agreement II

Waste Management Regulation

6. Package Waste Regulation (drink containers)

1977 Voluntary Agreement I
1987 Voluntary Agreement II
1988 "lex Coca-Cola"
1991 Waste Ordinance and DSD (dual system Germany)
1996 Voluntary Agreement III (failed)
2003 Failure of DSD in the field of drink containers
2005 Revision of the Waste Ordinance

7. Used Batteries Regulation

1980 Voluntary Agreement I
1988 Voluntary Agreement II
1995 Voluntary Agreement III (failed)
1998 Ordinance

8. Scrap Cars Regulation

1996 Voluntary Agreement
2002 Ordinance

9. Waste Electronic Equipment Regulation

1995/1996 Voluntary Agreement (failed)
2002 Ordinance

Climate Protection

10. Reduction of CFC Use

1977 Voluntary Agreement I
1987 Voluntary Agreement II
1991 Ordinance

11. Termination of CFC Production

1995 Voluntary Agreement

12. Reduction of C[O.sub.2] Emissions
1995/96 Voluntary Agreement ("climate declaration")

13. Promotion of Power-heat Cogeneration

2001 Voluntary Agreement

Table 2: The adoption of voluntary agreements
by different Environment Ministers based on data
by Toller (2012a, p. 311).

Environment Minister Average Number of Voluntary
 Agreements Adopted Each Year

Klaus Topfer (CDU) 1987-1994 4.4
Angela Merkel (CDU) 1994-1998 6.8
Jiirgen Trittin (GREENS) 1998-2005 1.7