Peter Katzenstein (1987, pp. 168–92) portrayed the West German welfare state as a highly segmented polity governed by consensual politics and providing generous social benefits. In fact, at first sight, not much has changed during the past two decades: compulsory insurance for all wage earners (sozialversicherungspflichtige Beschäftigung) is still provided by separate funds for pensions, health, unemployment, occupational accidents and – since 1995 – nursing care for the elderly (Pflegeversicherung). The system is still highly fragmented, and is administered according to a person's region of residence as well as that person's occupation. Thus, several regional funds are in charge of pensions for blue-collar workers (Arbeiter), whereas pensions for white-collar employees (Angestellte) are funded on a national basis. Civil servants (Beamte) receive their old-age benefits from current state budgets, while other public-sector employees are covered by the national white-collar workers' insurance scheme, supplemented by a complementary insurance which puts them on a par with civil servants. There is an entirely separate pension scheme altogether for miners, and finally, self-employed persons (Selbstständige) are allowed to opt out of the system altogether, which they normally do, since private pension and medical providers normally offer better levels of cover at lower costs than the statutory schemes.

Health insurance too still rests on a multitude of local, regional and national institutions. Although employees have been free to choose their health insurance fund since 1996, which has inevitably weakened the linkage to occupational status, the federal government has introduced a portfolio balance system to support those funds with bad risks.

Introduction

It is interesting to note that Katzenstein's original work on the semisovereign state did not include environmental policy as one of its specific domains for analysis (Katzenstein 1987). To a certain extent this is a curious omission, given that it was precisely at that point in the late 1980s that Germany's reputation as an ‘environmental leader’ (Weale 1992a, 1992b) was at its highest and that, in this policy domain at least, the German semisovereign state's capability to manage change appeared the most pronounced. Indeed, compared with other large European states, Germany in the 1980s was highly responsive to both the real environmental challenges associated with complex industrial societies and the higher levels of public unease about the trajectory of such societies, as manifested by the discourse of the ‘new politics’ and the emergence of Green parties (Lees 2000a, 2000b). As noted in chapter 1, some of those states, such as the United Kingdom, were characterised by unitary structures. Thus, the fact that the semisovereign Federal Republic outperformed states that were assumed to possess superior steering capacities was in itself worthy of note.

In defence of the original decision to omit environmental policy, however, one might also point out that it is when one is in the midst of change that it is hardest to define its overall parameters, trajectory and long-term consequences. Almost two decades later we are better placed to enjoy the benefit of hindsight and to take the long view.

Introduction: Federalism in Katzenstein's Policy and Politics in West Germany

Peter Katzenstein's treatment of federalism in Policy and Politics in West Germany (1987) was one of the classic accounts of co-operative federalism. It built on, reconciled and took forward what had hitherto been the pioneering texts: Fritz Scharpf et al. (1976) on Politikverflechtung, the interlocking of tiers of government in the policy process; and Gerhard Lehmbruch's Parteienwettbewerb im Bundesstaat, which explored how the dynamics of multi-tiered party competition could produce an equally interlocked ‘all-party proportional government’ through the interaction of different party majorities in Bundestag and Bundesrat (Lehmbruch 1976, p. 168). The arguments in Scharpf and Lehmbruch ran along parallel and largely unconnected trajectories. Parties did not really figure in Scharpf; Lehmbruch did not immerse himself in the institutional detail of the policy-making process. Each provided a brilliant but partial account. Katzenstein provided the linkage.

This linkage came in the juxtaposition of a ‘decentralised state’ with a ‘centralised society’ (Katzenstein 1987, p. 15). The decentralised state was a multifaceted antidote to Hitler. It was manifested in a strong Constitutional Court, the sectorised and regionalised operation of the civil service, ministerial autonomy under the Ressortprinzip, but above all federalism. West German federalism evolved in a way which was unusual. The division of powers was primarily functional, separating responsibilities for legislation, which was mainly carried out at the federal level, and implementation, which was mainly the responsibility of the Länder governments.

Directives are one of several instruments that are used in EU social policy. This chapter puts them into perspective, outlining the wider context of the EU's social dimension over time and the important role played by Directives. The main finding is that the role of binding regulative action has not been diminished, despite the debates on the open method of co-ordination (OMC). Therefore, we argue, studying social Directives is crucial not only for understanding the past, but also the present and, very likely, the future of European social integration.

Competences and decision modes

The 1957 EEC Treaty basically left social policy in the hands of the member states. The Treaty did not provide for an outright Europeanisation of social policies since too many national delegations had been opposed to such a move. The dominant philosophy of this Treaty was that welfare would be provided by the economic growth stemming from the economics of a liberalised market, and not from the regulatory and distributive capacity of public policy (see, for example, Leibfried and Pierson 1995; Barnard 2000). It is indicative of the Treaty's pro-market bias that its only explicit legislative competence in the field of social policy related to the free movement of workers, which for the most part even allowed measures to be adopted on the basis of qualified majority voting (Articles 48–51 EEC Treaty).

However, the European Community's action capacity was incrementally increased in day-to-day politics.

Beyond the state of the art

Our study contributes to the existing literatureon European integration, policy implementation, public–private interaction in policy-making and policy analysis. While detailed arguments can be found in the above chapters, this chapter is where we highlight some of the most important findings.

1. The process of designing and implementing EU law is political. In fact, it isa prime example of multi-level politics in practice. This seemingly simple and unsurprising finding has a number of aspects.

• Since the founding of the European Economic Community in the late 1950s, and especially since the beginning of the 1990s, the stock of EU legislation in the area of social policy and labour law has acquired a considerable depth and breadth. Hence, this policy area is not entirely left to ‘courts and markets’, although these are important (Leibfried and Pierson 2000).

• EU labour law is not a sum of insignificant rules that fail to go beyond what already exists in the member states. First, qualified majority voting has allowed the adoption of Directives even in the face of explicit opposition from individual governments. Second, the negotiators in Brussels sometimes lack information on what a specific rule will change in their domestic system. And third, domestic governments in general seem to consider the EU level a ‘normal’ arena of policy-making that supplements the realm of domestic politics.

Aim and content of the Directive

The Pregnant Workers Directive is one of the ‘daughter Directives’ enacted as a follow-up to the 1989 Framework Directive on Health and Safety. While the Framework Directive introduced a general system of occupational safety and health, based on risk assessments, preventive measures, and the collaboration of employers, employee representatives and occupational physicians, the focus of the Pregnant Workers Directive is on new or expectant mothers, that is, on a particularly vulnerable group of workers who face specific risks at the workplace. The general aim of the Directive is ‘to encourage improvements in the safety and health at work of pregnant workers and workers who have recently given birth or who are breastfeeding’ (Article 1).

To fulfil this aim, the Directive includes a set of fourteen compulsory minimum standards. These can be divided into standards relating to occupational health and safety in a narrowsense, and into provisions belonging to the realm of employment rights more generally understood.

So far as health and safety issues are concerned, the Directive provides the following.

1. Employers have to evaluate the potential risks to new and expectant mothers working in their establishments, taking into account a list of agents, processes and working conditions specified in the first annex to the Directive.

2. Female workers and/or their representatives must be informed about the results of this assessment.

The preceding chapters have dealt with the compulsory reforms related to our six EU Directives. However, the domestic impact of EU policies is not necessarily confined to such obligatory adaptations. Member states may use the transposition process to push through voluntary reforms that go beyond the minimum level required by European Directives. On the one hand, this can be a reaction to European soft law and, in fact, recent social policy Directives have included a considerable number of concrete but legally non-binding recommendations. On the other hand, member states might also decide to surpass the level of the EU's minimum standards (say, by creating a parental leave scheme which offers six instead of the required three months of leave).

From an abstract point of view, three different logics of treating the binding and non-binding parts of a Directive can be specified: minimalism, maximalism and alogic of domestic politics (see Table 10.1). The most frequently discussed assumption in the competitiveness-oriented literature is that the reaction of member states to potentially costly EU policies is driven by a logic of minimalism. According to this view, only binding law has the potential to harmonise the different domestic working conditions. In the absence of obligation, no additional costs to the national enterprises or the administration would be accepted in response to European policies.

This chapter extends the study of Europeanisation from the sphere of policy content to policy-making patterns, specifically to public–private relations. Since the beginning of the 1990s, remarkable developments have taken place in EU social policy at this procedural level. Since this specific style of public–private co-operation is restricted to one policy area only, we prefer not to speak about ‘Euro-corporatism’ (Gorges 1996), but rather about a ‘corporatist policy community’ (Falkner 1998).

The EC Treaty's social provisions (see Articles 136–48) now contain three layers of social partner participation in the policy process. First, a member state may entrust management and labour, at their joint request, with the implementation of social policy Directives. Secondly, the European Commission now has a legal obligation to consult both sides of industry before submitting social policy proposals. And thirdly, but most importantly, management and labour may, on the occasion of such consultation, inform the Commission of their wish to conclude social partner agreements instead of proceeding with traditional EU legislation. Such agreements may, at the joint request of the signatory parties, be implemented by a Council decision based on a proposal from the European Commission. Thus, since the 1990s, the social partners have been formal co-actors in EU policy-making.

The member states of the European union are characterised by deeply rooted systems of public–private interaction that exemplify their respective processes of public policy-making.

Constructing a typology: methodological and practical background

It has been mentioned in the previous chapters that the considerations of domestic politics have been underrated in recent writing on compliance with EU law. The great importance of national preferences and ideology for the implementation performance of many countries is one major finding of this study. However, this is not a single overriding factor which determines the compliance performance of member states and could thus serve as a safe anchor for predicting the success or failure of future implementation cases in all of our fifteen countries. Therefore, it should not be read as a new over-generalised theory for explaining the implementation of EU law. In fact, an untidy overall picture emerges once the manifold hypotheses we derived from the different literatures have been discussed: no causal condition pre-supposed by existing theories is able to explain our empirical observations. This suggests, once again, that the search for law-like generalisations and for simple isolated causes that could explain complex empirical phenomena is futile (see, for example, Scharpf 2002b).

Classic factors from the comparative welfare state literature do not help either. Most importantly, there is no direct correlation between social expenses and compliance records. For example, the UK ranks fourth when it comes to delays in transposing our six Directives, but is on the lower end of the scale of social expenses in Europe.

The implementation of EU Directives is but one example of the broader phenomenon of ‘Europeanisation’. This term has become a catchword in recent political science literature, referring to a number of slightly different phenomena that are located on at least four different levels. First, the term is a times used to refer to the EU-level development of policies and/or policy networks (e.g. Risse et al. 2001). Second, it can mean the reactions in domestic systems to top-down influences from the EU level, be they directly induced by EU law or indirectly by European policies such as the Maastricht convergence criteria (e.g. Ladrech 1994; see also Radaelli 2000). Third, Europeanisation is used to point out changes at the national level induced by transnational influences (Kohler-Koch 2000a). Finally, some authors take a very broad view and include the sum of all of these notions/levels in their understanding of Europeanisation (e.g. Borzel 1999; Falkner 2000b). For the purpose of this book, we shall adopt the top-down perspective as referred to by Robert Ladrech, andwe will try to isolate, as far as possible, the effects stemming from EU politics and (social) policy from other aspects included in some of the concepts mentioned above.

Research on Europeanisation – even if understood in this comparatively more narrow sense – targets a broad and complex phenomenon since all parts of the domestic political system may be affected (i.e. policies, politics, and polities).

It is obvious and has been argued throughout the book that the effect of EU social policy Directives depends on timely and correct implementation in the member states. This chapter will briefly outline which instruments the EU Commission can use to make non-compliant member states fulfil their European duties. The aimis to assess the phenomenon empirically and to confront the occurrence of relevant failures with the Commission's enforcement policy.

State of the art

Within the field of research on the EU, a number of authors have dealt with the question of how well member states follow their commitment to implement EU law. Generally speaking, these studies can be divided along two lines, i.e. the approach taken and the data used. Thus most of the literature could easily be sorted into a four-box matrix, where the x-axis distinguishes between quantitative and qualitative studies, and the y-axis differentiates between cases where the initiative lies with an individual complainant or with the European Commission.

Aim and content of the Directive

The general aim of the Working Time Directive is to improve the health and safety of workers by laying down minimum standards for the organisation of working time (Article 2 of the Directive). The Directive is based on a wide interpretation of occupational health and safety which assumes that working long hours is harmful to workers’ health and thus has to be limited.

The Working Time Directive applies in principle to both public and private sectors. It includes twelve compulsory minimum standards.

1. As a general rule, workers may not work longer than forty-eight hours per week, averaged out over a period of four months.

2. Every worker has to be granted a consecutive daily rest period of eleven hours.

3. Every worker has to be granted a consecutive weekly rest period of thirty-five hours, averaged out over a period of two weeks.

4. Every worker has to be granted a break if the working day is longer than six hours.

5. Moreover, employees are entitled to at least four weeks paid annual leave.

6. The four weeks' paid annual leave may not be replaced by an allowance.

7. Night workers may not work more than eight hours per day (averaged out over aperiod to be defined by national law or collective agreement), while night workers whose job involves ‘special hazards or heavy physical or mental strain’(Article 8) must work no more than an absolute limit of eight hours per day.

Policy-making within the multi-state polity of the European Union is an intricate affair. Given heterogeneous policy legacies in the member states as well asthe diverse preferences of national governments and other domestic actors, one-size-fits-all solutions are often neither politically feasible nor normatively desirable. A certain amount of flexibility and variation is thus needed in order to find solutions that are applicable to all member states. In this context, recent EU policies follow a new regulatory method based oncompulsory minimum standards, possibilities for derogations and non-binding recommendations, which tries to combine both ‘community and autonomy’ (Scharpf 1994). However, is a flexible governance style necessarily a good thing? Could it even be a dangerous development? Focusing on the field of EU social policy, we offer the first in-depth survey of voluntarist EU policies as set forth in different EU Directives, on the basis of empirical research into the impact of such steering efforts in the EU's multi-level system.

EU social policy: a successful combination of community and autonomy, or potentially dangerous voluntarism?

EU social policy is confronted with a ‘regulatory conundrum’ (Rhodes 1995). The disparate social systems and standards of the member states do not allow (at least in any practical way) for detailed harmonisation in the sense of simply equalising social standards via EU law ‘from above’. At the same time, liberalisation of the economy in the internal market has increased competitive pressures on the national systems of social and labour law.

Aim and content of the Directive

The Parental Leave Directive was the first EU-level social policy measure tobe based on a framework agreement by the major European federations of management and labour (UNICE, CEEP, and ETUC). The Parental Leave Directive did no more than give general legal force to the social partners pact. None of the latter's substantive provisions was modified, which is best illustrated by the fact that the agreement was attached, unchanged, to the Directive.

The general aim of the Directive is, according to the preamble preceding the main text of the social partners agreement, ‘to set out minimum requirements on parental leave and time off from work on grounds of force majeure, as an important means of reconciling work and family life and promoting equal opportunities and treatment between men and women’. The purpose of the agreement is therefore to enable working parents to take a certain amount of time off from work to take care of their children. In this context, particular emphasis is put on enabling and encouraging men to take on a greater share of childcare responsibilities.

The compulsory minimum standards of the Directive thus encompass seven provisions:

1. workers must be granted the right to at least three months' parental leave;

2. this entitlement is to be an individual right of both male and female workers;

3. parental leave has to be provided not only for parents with children by birth, but also for those who have adopted a child;