Sectoral Dynamics of EU Enlargement: Advocacy, Access and Alliances in a Composite Policy

Transcription

1 Sectoral Dynamics of EU Enlargement: Advocacy, Access and Alliances in a Composite Policy Ulrich Sedelmeier Central European University, Budapest ABSTRACT Theoretical studies of EU enlargement have mainly focused on the macro-level of enlargement, namely under which conditions the EU decides to enlarge. We still miss a conceptual framework to analyse the sectoral dynamics of enlargement that specifies under which conditions the preferences of the candidate countries are accommodated in EU policy. I argue that an accommodation of the candidates preferences depends crucially on policy advocates in the EU. The influence of these policy advocates is structured by the nature of enlargement as a composite policy, in which sectoral decisions are negotiated between the group of policy-makers with primary responsibility for enlargement as such, and various groups of sectoral policy makers. The policy advocates impact on sectoral policies thus depends (1) on their access to decision-making, which is shaped by the structure of the policy process, as well as (2) their ability to build alliances with sectoral policy-makers. Crucially, alliance-building does not only depend on the strength of sectoral interest groups, but also on the compatibility between the candidates preferences and the sectoral policy paradigms that shape the preferences of the sectoral policy-makers. I. INTRODUCTION 1 Over recent years, our conceptual understanding of EU enlargement has improved considerably. Especially the literature on eastern enlargement has started to address some crucial theoretical questions (Fierke and Wiener 1999; Friis 1998; Friis and Murphy 1999; Schimmelfennig 2001; Sedelmeier 1998, 2000). However, while there have been considerable advances with regard to what could be considered the macrolevel of enlargement, this is not matched with regard to the meso-level. Much of the theoretical debate with regard to eastern enlargement has focused on the macro dimension of enlargement - the broader dynamics that underpin the EU s enlargement policy and specifically on the EU s decision to enlarge. By contrast, the meso dimension - substantive policy outcomes in distinctive sectoral policy areas that are part of enlargement - have received very little attention. However, this dimension is no less central to EU enlargement. A key question with regard to these substantive policy outcomes is which factors facilitate or obstruct an

2 accommodation of the applicants preferences in sectoral policies such as trade liberalisation during the association period; regulatory alignment as part of the EU s accession (and pre-accession) conditionality, or the provisions accession treaties. Previous enlargement episodes suggest that a lack of flexibility by the EU can cause severe problems for the candidates and lead to disgruntled new members. In particular with regard to prospective members, an inability to accommodate their differences to a certain degree might jeopardise the sustainability of a broader European integration project. However, why the EU enlarges does not necessarily tell us much about these sectoral dynamics of enlargement. In particular with regard to eastern enlargement, the EU s general endorsement of enlargement is often at odds with substantive sectoral policy. Thus, the factors that facilitate an accommodation of the central and eastern European countries (CEECs ) demand for membership do not directly lead to an accommodation of their preferences in substantive policy. Some theoretical studies of the macro-level acknowledge the discrepancy between the general dynamics that lead to enlargement and restrictive policy practice on the substantive detail (Schimmelfennig 2001). Other theoretically-informed studies have attempted to problematise this discrepancy more explicitly (Friis 1997; Niblett 1995; Torreblanca 1997; Sedelmeier 1994). However, we still miss a conceptual framework that identifies in a more structured way the factors that determine the likelihood of an accommodation of the applicants preferences in substantive sectoral policies, or, in other words, which factors mediate between the broader dynamics of enlargement and specific sectoral outcomes. A key factor that might determine the likelihood of accommodation of the candidates preferences is the strength of sectoral interest groups that oppose such an accommodation. Given the unequal bargaining power between the EU and the applicants, liberal intergovernmentalist approaches (e.g. Moravcsik 1993) would suggest that the strength of domestic interest groups whose preferences diverge from those of the CEECs are the key variable constraining an accommodation of the CEECs preferences. Indeed, these assumptions underpin - at least implicitly - much of the early analyses of EU policy towards the CEECs which focus explicitly on substantive policy issues, mainly the extent of trade liberalisation under the Europe Agreements (EAs). These include contributions from trade economists (see e.g. Hindley 1993; Messerlin 1993; Rollo and Smith 2

3 1993; Baldwin 1994), as well as more theoretically informed studies (Haggard et al. 1993; Nicolaïdis 1993; Guggenbuhl 1995). Put simply, these analyses suggest that while political considerations and concerns about aggregate welfare gains might incline policy-makers to accommodate the CEECs preferences, this is undermined by vested economic interests in particular sectors. While these studies show convincingly that the pressure of domestic groups is an important factor, such pluralist and materialist approaches have limits. The strength of domestic groups alone is not always a reliable indicator for the extent to which EU policy accommodates the preferences of the CEECs. In the steel sector, for example, strong interest group opposition did not prevent an accommodation of the CEECs preferences. The EU liberalised trade considerably, despite strong countervailing pressures from EU producers (see section III below). This is not the only example in which the EU shifted towards a more accommodating policy, despite opposition from domestic groups (see e.g. Sedelmeier and Wallace 1996: 370, 376, 380). Thus, the existence of strong domestic veto groups alone does not always prevent an accommodation of the applicants preferences. This paper draws on evidence from EU policy towards the CEECs in order to develop a conceptual framework that specifies the conditions under which the EU is likely to accommodate the applicants preferences in sectoral policies. I argue that in order to understand these sectoral dynamics of enlargement we need to consider not only to the strength of societal interest groups, but two further factors, namely the structure of the policy process and sectoral policy paradigms. This argument is based on the following assumptions. First, given the asymmetrical bargaining power between the EU and the applicants, an accommodation of the preferences of the latter depends largely on policy advocates inside the EU. Second, the policy impact of these policy advocates is significantly affected by the nature of enlargement as a composite policy, i.e. a policy that consists of a macro policy and many sectoral meso policies and in which distinctive groups of policy-makers have primary responsibility for the macro policy and the various meso policies. The policy impact of such policy advocates then depends (1) on their access to decisions on sectoral policies, and (2) on their ability to form winning alliances with sectoral policy-makers. 3

4 The access of the policy advocates to decision-making depends significantly on the structure of the policy process. As in the case of eastern enlargement the policy advocates are among the macro policy makers, their access is greatest, the more centrally coordinated or hierarchical decision-making is. Conversely, the more fragmented the policy process, the more limited is their policy impact, and hence the likelihood of an accommodation of the CEECs preferences. While the structure of the policy process is a more formal constraint (or resource), materialist approaches do not easily capture a key factor that shapes alliance-building. The preferences of sectoral policy-makers are not simply a reflection of the interests of the strongest societal groups in particular sectors. Rather, they are crucially influenced by particular policy paradigms, i.e. the sets of ideas that respectively underpin policy in specific areas. These policy paradigms hence significantly shape the ability of policy advocates to forge alliances with sectoral policy-makers in favour of an accommodation of the CEECs interests. A mismatch between sectoral policy paradigms and the preferences of the CEECs is a key obstacle to successful alliance-building. Conversely, compatibility between paradigms and the applicants preferences increases the likelihood of an accommodation of the latter. The paper proceeds as follows. Section II sketches the analytical framework. Its first sub-section conceptualises enlargement as a composite policy, in which an accommodation of the candidates preferences depends largely on the influence of a group of policy advocates. The other two sub-sections elaborate respectively on the two factors that shape their influence and their ability to counterbalance interest group pressure - the structure of the policy process and sectoral policy paradigms. Section III presents evidence from two meso policy areas in the EU s eastern enlargement. These policy areas should be hard cases for an explanation focusing on interest groups, as in both cases sectoral interest groups where unified in strong opposition to accommodating the CEECs preferences. Yet in both cases the EU significantly accommodated the CEECs preferences. The case of steel trade liberalisation within the association policy illustrates that an emerging alternative policy paradigm enabled alliance-building that led to a significant degree of accommodation of the CEECs preferences, despite strong countervailing interest group pressure and a fragmented policy process. The EU s requirements for regulatory alignment during the pre-accession stage presents a more mixed picture. I 4

5 argue that variations in the degree of accommodation between social policy and environmental policy can be related to the role of the EU s internal market paradigm. Section IV draws the insights from the more inductive case studies together in order to sketch a framework that might be more generalisable for deductive research. II. POLICY ADVOCACY, ACCESS, AND ALLIANCES IN A COMPOSITE POLICY Policy Advocacy in a Composite Policy Given the asymmetrical bargaining power between the EU and the candidates, the likelihood of an accommodation of the candidates preferences depends to a large extent on the advocacy of these preferences by policy-makers inside the EU. In the case of eastern enlargement, such a group of policy advocates emerged around the Commission s unit for policy towards the CEECs and the cabinets of subsequent external relations Commissioners, Frans Andriessen and Leon Brittan (see e.g. Sedelmeier and Wallace 1996, 2000). The motivations behind their advocacy are analysed elsewhere. 2 Key for the analysis here is that the policy impact of these policy advocates is crucially shaped by the particular nature of the EU s enlargement policy. A key characteristic of EU policy towards the CEECs and enlargement more generally is that it is a composite policy (Sedelmeier 1998; Sedelmeier and Wallace 2000). It is a broad policy framework, or a macro policy, which is constituted by a range of sectoral meso policies. In other words, the EU s enlargement policy draws its substance from distinctive policy decisions across the range of policy areas which are part of the association policy or accession negotiations. This characteristic requires an analytical distinction between two dimensions of policy: (1) decisions about the macro level of policy, to determine the overall objectives and parameters of policy; and (2) decisions about the specific detail and substance of policy in particular sectors. Furthermore, the nature of a composite policy implies that different groups of policy-makers, or distinctive policy communities, have primary responsibility for specific parts of this policy. The distinction between macro- and meso policy makers would be a merely formal one, if macro- and meso policy makers from each member 5

6 state simply advance previously agreed national positions in inter-state bargaining. However, the intra-eu debate about substantive policy is often not so much characterised by national cleavages than by functional cleavages that relate to policymakers bureaucratic position in their administration. This is not to suggest that different member states attitudes to enlargement do not vary and that these national differences are insignificant for sectoral policies. However, the policy process is frequently characterised by a transgovernmental cleavage of cross-cutting coalitions that pitches foreign policy makers collectively against sectoral policy-makers from across the member states (see also Niblett 1995; Torreblanca 1997). This notion of a composite policy in which the sectoral dynamics of enlargement are shaped by the interactions between distinctive policy communities has affinities with a characterisation of EU policy towards the CEECs as overlapping games (Torreblanca 1998). There are also parallels to an analytical framework that emphasizes that different types of decisions are respectively dealt with by different groups of policy-makers, each guided by distinctive policy rationales (Peterson 1995; Peterson and Bomberg 1999). In case of (eastern) enlargement, the group of policy-makers which has primary responsibility for policy towards the CEECs includes officials from the member states foreign ministries and the Commission s DG for external relations (now DG Enlargement), as well as periodically the heads of state and government or the Commission president. These macro policy makers might be generally more inclined to accommodate the preferences of the CEECs. Again, for the purpose of the analysis, we can bracket the reasons for this bias: rationalist would emphasise that they are more likely to define enlargement in terms of longer-term political and security concerns and aggregate welfare gains, constructivists would point out that they are also more likely to be receptive to community norms that prescribe certain standards of appropriate behaviour towards the applicants. However, for the policy advocates to have an impact on policy, it is insufficient to influence the macro policy makers. The macro policy makers cannot take decisions on policy substance within the various meso-areas autonomously. Such decisions have to be negotiated, in a process of horizontal policy coordination, with those sectoral policy-makers who have primary responsibility for the respective mesopolicies. These meso policy makers are, however, more predisposed against policy 6

7 changes that would be necessary to accommodate the preferences of the CEECs. There is an institutional bias towards the status quo and policy change is likely to involve administrative costs and costs to sectoral interests groups. Thus, given the particular characteristics of a composite policy, the policy impact of the policy advocates depends (1) on their access to policy-making and decision-making on specific issues, as well as (2) on their ability to form alliances with sectoral policy-makers that are sufficiently strong to overcome countervailing interest group pressure. 3 The following two sections elaborate on the factors that respectively affect access and alliance-building, namely the structure of the policy process and the policy paradigms that underpin EU policy in particular sectors. Access in a Composite Policy: the Structure of the Policy Process In the case of EU policy towards the CEECs, the policy advocates that promote an accommodation of the CEECs preferences in intra-eu debate are among the macro policy makers. Their access to sectoral policy decisions is therefore strongly correlated to the structure of the policy process and the degree to which it respectively involves macro- and meso policy makers. The literature on intra-governmental policy coordination (see e.g. Scharpf 1993) and on veto points in the policy-process (Thelen and Steinmo 1992: 7) is particularly relevant if we want to understand how the structure of the policy process shapes the access of particular actors to policy decisions. We can broadly distinguish between different structures of the policy process according to their degree of fragmentation or centralisation, as well as between two different types of centralised policy processes hierarchical and centrally coordinated. The more centralised the policy-making process, the greater the influence of the macro policy makers, and of the policy advocates among them. Hierarchical decision-making either explicitly subordinates the concerns of the meso policy makers, or implicitly circumvents them. By reducing the access of the meso policy makers to decision-making, the outcomes of such a process are more likely to reflect the concerns and priorities of the macro policy makers. Of course the flip-side is that such decisions might insufficiently appreciate the problems that they might create for 7

8 coherent policy in the meso-areas. 4 Decisions that are imposed on a particular sector are therefore likely to remain contested and their sustainability might be doubtful. A centrally coordinated policy process usually takes the form of multilateral consultations between macro policy makers and various groups of sectoral policymakers. The ideal type is a form of positive coordination (Scharpf 1993). This might take the form of special task forces or working groups. A key characteristic is that while sectoral policy-makers would have to agree to policy change, none of them would unilaterally oppose changes in their particular policy area. The inherent pressure on all participants to compromise makes an accommodation of the candidates preferences likely. While such accommodation might be more limited than under hierarchical decision-making, it is more likely to be sustainable. At the other end of the spectrum, the more fragmented the policy process, the more likely the preferences of the meso policy makers will be reflected in the policy outcomes in the meso policy areas. The influence of the meso policy makers is greatest, if fragmentation means that they can decide policy autonomously, or if the meso policy makers have the power to veto policy changes in their policy area. In this case, there is a strong bias against an accommodation of the applicants preferences. In general, it is difficult to clearly identify one structure of the policy process for a given policy area. The structure of the policy process might vary within the same policy area over time, according to the stage of policy-making. Furthermore, precisely because of its importance for granting access, the structure of the policy process itself is subject to political strategies and manipulation by the political actors involved. The development of the EU s policy towards the CEECs reflects the relevance of the structure of the policy process in mediating the policy impact of the policy advocates on substantive sectoral policy outcomes. 5 One general feature of the evolving policy was the lack of continuous oversight by the foreign ministers in the Council (Sedelmeier and Wallace 1996), and the disengagement of the European Political Cooperation (EPC) framework from the association policy in early 1990 (Nuttall 1992). This greatly weakened the scope for more centralised policy coordination. The policy process largely took the form of bilateral consultations, between macro policy makers and individual groups of sectoral policy-makers, in which the latter had sufficient autonomy to block concessions in their respective 8

9 policy areas. Neither the Council nor the Commission established multilateral consultation networks or task forces for the association policy which might have led to positive coordination and agreement on collective burden-sharing across sectors and issues. These constraints were apparent in the process of establishing the Commission s negotiating directives for the EAs. DG I s original framework had been fairly accommodating of the CEECs preferences. The final negotiating directives turned out to be much more restrictive, after the details had been filled in by sectoral DGs in the Commission and agreed with sectoral policy makers in the Council. Likewise, the scope for the CEECs to press for a more far-reaching accommodation of their preferences during the EA negotiations was limited: the negotiations were characterised by a high degree of fragmentation and a hands-off approach of the foreign ministers in the General Affairs Council. Fragmentation of the policy process became most pronounced during the implementation of the EAs. For example, this fragmentation during the implementation phase made it very hard to prevent sectoral interest groups from using trade defence instruments to cancel out some of the concessions that the CEECs had obtained. However, there were also instances in which policy became insulated at the level of the macro policy makers, which facilitated an accommodation of the CEECs preferences. Such instances of more centralised coordination were apparent during the negotiations of the EAs during The context of the negotiations allowed the Commission negotiators on a number of occasions to force the General Affairs Council, rather than the Council s sectoral working groups, to become more directly involved in deciding hierarchically on revisions to the negotiations directives. This strategy also worked in some less publicised instances of circumventing opposition from sectoral policy-makers (see also, e.g. Torreblanca 1998). Similarly, in the lead-up to the Copenhagen European Council in June 1993, the endorsement of the CEECs membership perspective was decided by the foreign ministers in the Council, acting in considerable autonomy from sectoral policymakers. The package of trade concession announced at the European Council was agreed by a high-level group of foreign ministry officials. In the negotiations of the pre-accession strategy, presented at the Essen European Council in December 1994, the German presidency, in close cooperation with the Commission, considerably 9

10 restricted the access of sectoral policy-makers to the policy process. Discussions were kept at the level of the Permanent Representatives (Coreper), and the final decision was even taken in a Coreper restreint. More generally, the fact that the story of the association policy is so closely linked to decisions announced at specific European Council meetings also reflects a deliberate strategy by the Commission to use the anticipation of the political bias at the European Council to focus attention in the Council. In sum, a focus on the structure of the policy process suggest that the existence of veto groups might not directly or necessarily translate into restrictive policy outcomes. This overview suggests that while interest group pressure was an important factor that significantly restricted an accommodation of the CEECs preferences, there is a strong correlation between the effectiveness of this constraint and the structure of the policy process. The prevalent fragmentation of the association policy helped interest groups to oppose accommodation, while those instances in which policy was more centrally coordinated facilitated accommodation. Still, access to policy-making does not equal influence on policy outcomes. In order to wield influence, the policy advocates need to be able to build sufficiently strong alliances among the sectoral policy makers involved in this decision. The following section elaborates on the argument that policy paradigms are an important factor that affects the likelihood of successful alliance-building within a particular policy area. Alliance-Building in a Composite Policy: the Role of Policy Paradigms Predictions about the policy advocates ability to form winning alliances with sectoral policy makers depend crucially on our assumptions about the nature of the preferences of latter. In contrast to pluralist approaches in comparative politics, or a liberal intergovernmentalist approach to the EU, this paper argues that public policymakers do not simply act according to the interests of the strongest societal groups. Rather, policy-makers have preferences of their own which they might pursue even against societal pressure (see also e.g. Evans et al. 1985). These preferences are to an important degree shaped by policy paradigms (Hall 1993). Policy paradigms are an 10

11 interpretative framework, based on a coherent set of ideas that underpin policymakers understanding of policy in their respective areas of responsibility: policymakers customarily work within a framework of ideas and standards that specifies not only the goals of policy and the kind of instruments that can be used to attain them, but also the very nature of the problems that need to be addressed. Like a Gestalt, this framework is embedded in the very terminology through which policymakers communicate about their work, and it is influential precisely because so much of it is taken for granted and unamenable to scrutiny as a whole (Hall 1993: 279). Policy paradigms can best be seen as one feature of the overall terms of political discourse. They suggest that the policymaking process can be structured by a particular set of ideas, just as it can be structured by a set of institutions (Hall 1993: 290). Policy paradigms are specific to particular policy areas, as a distinctive feature within the normative institutional structure of these policy areas. Policy paradigms are not simply causal ideas about means-ends relations for an effective policy. They also incorporate principled ideas about what constitutes a legitimate policy, for example with regard to policy goals, appropriate policy tools, or about priorities between different groups that the policy affects. Paradigms become embedded by acquiring a dominant position in the policy discourse that characterises a policy area. This view of the role of policy paradigms has strong affinities with the concept of policy frames (Rein and Schön 1991) which has found recently wider application in analyses of EU policy-making (see Jachtenfuchs 1996; Lenschow and Zito 1998; Dudley and Richardson 1999). There are also certain parallels to an analysis of individual EU policy areas as distinctive governance regimes (Armstrong and Bulmer 1998). In this paper, I use the term policy paradigm for the set of ideas that underpins policy in a particular meso policy area. Policy paradigms shape the preference formation of sectoral policy-makers, albeit not exclusively, but independently from the material structure of the interests in this sector. The ability of policy advocates to build successful alliances in the meso policy areas is thus not simply a function of the configuration of societal interest. The degree of compatibility between the preferences that they promote and sectoral policy paradigms is a key factor shaping alliance-building. A mismatch between the applicants preferences and sectoral paradigms might lead sectoral policy-makers to oppose an accommodation of these preferences, even independently of interest group 11

12 pressure. Conversely, compatibility facilitates accommodation. Such compatibility increases the likelihood that the policy advocates might forge a strategic alliance among sectoral policy makers that is sufficiently strong to bring about a policy change, even against countervailing interest group pressure. However, the very concept of policy paradigms would lead us to expect that such cases of compatibility are rare. An accommodation of the applicants preferences would usually require a change in the status quo of a sector, since accommodation is otherwise highly unlikely to provoke opposition from sectoral interests. Yet the institutional bias in favour of the status quo is precisely so strong because it is usually underpinned by policy paradigms. Policy paradigms are therefore most likely to form significant obstacles to an accommodation of the applicants preferences, in addition to interest group pressure. The most likely cases where paradigms favour accommodation despite strong opposition from interest groups, are thus cases in which paradigms shifts. By definition, we would expect such cases of paradigm shifts to be rare. The notion of policy paradigms is based precisely on the assumptions of a high degree of stability and taken-for-grantedness of the underlying ideas. However, policy failure and changing circumstances might discredit the ideas which underlie a policy paradigm and the policy paradigm itself. In this case accommodation might be achieved, if such a policy change is compatible with alternative policy paradigms that challenge and successfully replace the dominant policy paradigm. An important question is therefore whether the policy advocates can form a strategic alliance with sectoral policy-makers who promote an alternative policy paradigm that is more compatible with their preferences. I argue in section III that this is precisely what happened in the case of trade liberalisation in the steel sector. A special case of paradigm compatibility, which is however particularly relevant for enlargement is the question of a temporary accommodation of the applicants preferences. Applicants might ask for such temporary accommodation during the pre-accession period or in post-accession transition periods. Such accommodation might be possible if at least it is not considered to challenge a dominant sectoral policy paradigms. In section III, I argue that this was the case with certain aspects of the internal market. 12

13 Methodologically, policy paradigms can be identified within the policy discourse of a particular meso policy. Public statements and documents in which policy makers justify their action with reference to certain problem definitions or normative ideas reflect such policy paradigms. Sometimes these might be more implicit. The identification of policy paradigms requires an inductive analysis of the history of policy development in a particular meso policy. This analysis has to reveal whether the emergence of a particular policy regime is indeed based on coherent sets of ideas about cause/effect relations and legitimacy claims. An assessment of the political impact of policy paradigms requires an analysis of how consensual or contested these understandings are, and to what extent they are challenged by alternative ideas. A key methodological implication is thus that it is difficult to make predictions about the way in which policy paradigms affect a political process prior to the analysis. Table 1 summarises the initial propositions about how the factors that affect the policy advocates access and alliance-building structure the influence that they can bring to bear on sectoral policies. If an accommodation of the candidates preferences is opposed by sectoral interest groups, their advocacy is unlikely to be successful, unless the policy the conditions for both access and alliance-building are favourable, i.e. the policy process is centrally coordinated and policy paradigms are compatible with the candidates preferences (box 2). If policy paradigms are incompatible, then a centralised policy process might facilitate temporary accommodation, which is, however, unlikely to be sustainable in the longer term (box 4). Even if sectoral policy paradigms are compatible with the candidates preferences, advocacy is likely to remain unsuccessful, as long as the policy process remains fragmented (box 1), but as soon as access is achieved, a sustainable accommodation is possible. The following section draws on empirical evidence of two episodes of the eastern enlargement policy in which the CEECs preferences were accommodated despite interest group opposition, in order to explore the effect of the interplay between policy paradigms, policy process and interest group opposition on the likelihood of accommodation. 13

14 Table 1: Influence of policy advocates on sectoral policies in a composite policy Policy Process Fragmented Centralised Policy Paradigms Compatible Incompatible +/- (1) + + (2) - - (3) -/+ (4) III. EMPIRICAL EVIDENCE FROM ASSOCIATION AND PRE-ACCESSION POLICIES TOWARDS THE CEECS Trade Liberalisation in the Steel Sector The case of trade liberalisation towards the CEECs in the steel sector illustrates how the emergence of an alternative policy paradigm which is more compatible with the preferences of the CEECs facilitated accommodation, despite opposition from interest groups. 6 The ability of the policy advocates to form a strategic alliance with those sectoral policy makers who favoured a change in steel policy more generally played an important role in overcoming sectoral opposition to an accommodation of the CEECs preferences. The EU s steel sector is considered one of its most import-sensitive sectors and has traditionally enjoyed a high degree of protection. Before the negotiations of the EAs, steel trade was regulated by a highly restrictive regime of Voluntary Restrain Agreements (VRAs) that imposed both quantitative and price restrictions on CEEC exports (Wang and Winters 1993). In addition, EU producers in the steel sector were particularly strongly opposed to liberalising trade with the CEECs and there was no countervailing pressure from consuming industries on the EU side. This status quo and the balance of interest group pressure makes the steel sector a hard case for which interest group-based 14

15 explanations would suggest that a policy change in favour of greater market access for the CEECs was highly unlikely. However, the EU accommodated the CEECs preferences for free and unconditional access to the EU market to a considerable extent (see also Rollo and Smith 1993). The EAs ended the VRAs and immediately eliminated all quantitative restrictions. Tariffs (which in any case did not constitute a major trade barrier) were phased out. Moreover, the centralisation of the policy process in the framework of the EA negotiations enabled the policy advocates to prevent the insertion of provisions that would have maintained the possibility of reintroducing VRAs. Crucially, they also thwarted subsequent attempts to undermine some of the benefits granted to the CEECs. The fragmentation of the policy process in the implementation phase meant that on a few occasions EU producers were able to use trade defence instruments against CEEC companies. However, the policy advocates successfully resisted interest group pressure for the wholesale re-introduction of a restrictive regime as part of the 1993/94 EU plan for the restructuring of the steel industry. While this outcome appears to contradict interest-based expectations, a focus on the changes in the structure of the policy process the temporary shifts from a predominantly fragmented process to greater central coordination partly explains the accommodation of the CEECs preferences. Crucially however, the role of the policy paradigm underpinning EU steel policy provides a plausible explanation for the sustainability of accommodation. The policy paradigm which dominated EU steel policy since the inception of the European Coal and Steel Community (ECSC) could be described as a managed restructuring paradigm. This paradigm is based on the assumption that the specific production and market structure of the steel industry (see e.g. Glais 1995) makes the nature of competition in this market crucially different from other sectors (see e.g. Mény and Wright 1986). During periods of depression, the free play of market forces would thus not induce firms to undertake the necessary adjustment measures. It would merely lead to cut-throat competition, in which not necessarily the most economically viable firms would survive. Firms would only restructure and cut over-capacity if this was organised as a collective effort, under the guidance of public policy-makers, and in a stable external environment i.e. 15

16 without the disruption of cheap imports. The CEECs preferences for the liberalisation of market access were thus incompatible with this dominant sectoral policy paradigm. However, in the early 1990s, an alternative policy paradigm which I call the non-intervention paradigm started to emerge (see also Dudley and Richardson 1999). Different factors opened a window of opportunity for a paradigm change. Technological changes, such as the emergence of mini-mills, made it easier for producers to adjust to changing demand patterns. Changes in the market structure, such as an accelerating trend towards mergers and privatisation in the European steel industry facilitated rationalisation and increased competitiveness. Finally, the anticipation of the expiration of the ECSC treaty in 2002 gave rise to a debate whether the EU should continue to consider the steel sector as a special case. The non-intervention paradigm precisely rejects the idea of the steel sector as a special case. In contrast to the dominant paradigm, it regards external competition as a positive contributing factor to restructuring. This paradigm focuses no longer on the protection of the EU market, but on the need for open markets abroad. It is thus much more compatible with the CEECs preferences. However, while EU producers increasingly supported a less regulated international steel trade regime, they were united in strong opposition to greater market access for the CEECs. The ascendancy of the non-intervention paradigm enabled the policy advocates to find support for trade liberalisation among some policy-makers in the steel sector. Although this alternative policy paradigm emerged largely independently from eastern enlargement, the interaction between the policy advocates and supporters of the non-intervention paradigm for EU steel policy played an important role. The sectoral policy-makers who supported the nonintervention paradigm provided arguments and a supportive alliance for the policy advocates push for greater CEEC access to the EU steel market. This alliance was sufficiently strong to bring about, and sustain, a significant liberalisation of the EU steel market for CEEC exporters, despite the strong opposition from EU steel producers. 16

17 Conditionality for the Regulatory Alignment of the CEECs with the Internal Market The EU s policy for the regulatory alignment of the CEECs with the internal market illustrates a different way in which sectoral policy paradigms affected the accommodation of the CEECs preferences. The general orientation of the Commission s White Paper on which the EU s regulatory alignment policy is based, accommodates the CEECs preference for selective alignment with product regulations during the pre-accession period. However, while this was particularly the case for environmental policy, this is in contrast to the area of social policy. The process of regulatory alignment is similar to that of regulatory harmonisation to remove non-tariff barriers (NTBs) in the single market program and to the EEA regime which extended the internal market to most of the EFTA countries. Regulatory alignment of the CEECs is based on a Commission White Paper (WP) (Commission 1995) that the Cannes European Council endorsed in June The WP has the status of non-binding recommendations for the CEECs, while final, legally binding decisions are only taken in actual accession negotiations. 7 Essentially, the CEECs prefer selective regulatory alignment, which would give them greater flexibility to set their own priorities (see Smith et al. 1996; McGowan and Wallace 1996; Young and Wallace 2000). Such flexibility would allow them to opt for early adjustment in areas where they consider it beneficial for their economic restructuring or to prepare their companies for internationalisation. Conversely, selective alignment would enable the CEECs to delay alignment in those areas where this involves high costs (either public spending or for private companies) and where the only rationale for doing is directly related to EU accession. Especially if in such areas EU regulations do not concern product standards but the production process, they might want to implement alignment only during post-accession transition periods. While product standards might constitute NTBs that would undermine the functioning of the internal market, process regulations only affect production costs. Such process regulations that impose considerable costs on CEEC firms, as well as to some extent, governments, primarily concern measures in the area of EU social policy, 17

18 such as health and safety at the workplace, and certain elements of EU environmental policy, such as pollution from stationary sources. However, EU producers strongly oppose the CEECs preference for selective alignment (UNICE 1997; see also Chiris 1998). Rather, they argue that the EU should demand from CEEC companies strict and early compliance with EU process regulations. Early alignment would reduce competitive advantages that the latter might have from producing according to less stringent regulations. The balance of interest group pressure would therefore suggest that an accommodation of the CEECs preferences is highly unlikely. Furthermore, two central elements of the policy paradigm underpinning the EU s internal market militate against selective alignment. The first is a distinctive understanding of the nature of economic competition. It is based on a level playing-field argument, which suggests that the purpose of internal market regulations is not only to remove NTBs in order to guarantee the free circulation of products. Single market legislation also has to ensure fair competition, which requires a harmonisation of regulations that concern the production process. This harmonisation is necessary, as certain environmental standards, e.g., affect firms production costs. The second element is the broad acceptance inside the EU that the internal market is not merely based on a rationale of competition and market integration, but that it also serves the protection of broader public policy objectives, notably high levels of social and environmental protection (Wallace and Young 1996). In sum, both these elements of the EU s internal market paradigm constrain the extent to which EU policy-makers view the CEECs demands for selective alignment as feasible and legitimate. However, while these arguments dominate the EU s discourse on the internal market, the assumptions on which the EU s internal market paradigm is based, are far from generally accepted among economists. The argument can be made that while the harmonisation of product standards is essential for the functioning of the internal market, the harmonisation of process standards is not. The latter affect the production costs of firms, and thus indirectly affect competition in the internal market. However, they do not constitute NTBs and therefore do not impede the free circulation of products. Furthermore, it can be argued that there is nothing unfair about competition in the absence of 18

19 harmonised environmental or social standards (see, for example, Smith et al. 1996). From such a perspective of the internal market, there would be no reasons to refuse the CEECs the possibility of selective alignment and for implementing process regulations in the course of long transition periods after accession, rather than at the pre-accession stage. This might grant CEEC firms a certain a temporary competitive advantage, but it would not endanger the integrity of the internal market. Despite the fact that strong groups inside the EU, as well as the EU s internal market paradigm, are opposed to selective alignment, the WP accommodates the CEECs' preferences to a significant extent. The EU s regulatory alignment policy was essentially driven by the policy advocates inside the Commission, which included officials from the CEEC unit in DG I and from the cabinet of Leon Brittan, the then external relations Commissioner (Sedelmeier 1998: ). These policy-makers understanding of the specific situation in the CEECs led them to question the viability and desirability of early alignment according to the EU s internal market paradigm. They successfully convinced the policy-makers of DG XV, responsible for the internal market and the drafting of the WP, to take a selective approach to the CEECs regulatory alignment. The agreement of DG XV to a more selective approach was on the one hand a pragmatic choice, as the sheer amount of process regulations in agriculture alone would have contradicted the purpose of the WP as a useful guide. On the other hand, however, the widespread acceptance of a the narrower version of the internal market paradigm facilitated alliance-building with policy makers of DG XV. The prevalent view in DG XV was that the core of the internal market consisted of the elimination of NTBs and the free circulation of products, which was compatible with a temporary accommodation of the CEECs preferences for selective alignment. The drafting of the WP was carried out in a process of central coordination between DG XV and sectoral DGs. In this process, the former was able to persuade most of the sectoral DGs to concentrate in their particular sections of the WP on product regulations and to exclude process regulations. This selective approach is most clearly reflected in the area of environmental policy, where such process regulations as the pollution from stationary sources are excluded from the 19

20 WP. Of course, this accommodation of the preferences of the CEECs was facilitated by the fact that the WP did not mobilise much interest group pressure, in part because of its non-binding character, in part because of the speed at which it was drafted. Significantly however, the outcomes of accession negotiations in the area of environmental policy appear to confirm that the EU is prepared to grant transition periods in areas where substantial adaptation of infrastructure is necessary and where substantial investments need to be spread over time. By contrast however, the area of social policy constitutes an exception to the general rule. Officials from DG V insisted that the WP should include process regulations, such as health and safety at the workplace. Significantly, their opposition to selective alignment was not the result of external pressures. Rather than defending the interests of particular interest groups inside the EU, these policy-makers defended the dominant internal market paradigm that emphasises a level playing field and high levels of social and environmental protection. Even a temporary accommodation was incompatible with their specific understanding of, and preference for, the operation of the EU s internal market. In turn, the presentation of their arguments with reference to assumptions underpinning the broadly accepted internal market paradigm made it hard for DG XV to deny the legitimacy of their resistance to a more flexible approach. In sum, regulatory alignment with the internal market shows two contrasting examples of how policy paradigms structured the interactions of macro- and meso policy. The policy advocates argument that a temporary accommodation was compatible with a narrower view of the internal market allowed them to build an alliance with policy-makers in DG XV in favour of selective alignment. Environmental policy-makers did not oppose such a temporal accommodation, as they were assured that it did not challenge the dominant internal market paradigm. In the area of social policy, the defence of the dominant policy paradigm by public policy-makers was the main obstacle to accommodating the preferences of the CEECs. This was the case although, at least at that stage, there was no interest group pressure for doing so. 20

21 IV. INTEREST GROUP PRESSURE, POLICY PROCESS, POLICY PARADIGMS AND THE SECTORAL DYNAMICS OF ENLARGEMENT Table 2: Likelihood of an accommodation of applicant preferences in sectoral policies if strongly opposed by sectoral interest groups Policy Process Fragmented Centralised Accommodation Accommodation Compatible possible, likely to be sustainable likely [general orientation Policy Paradigms Incompatible [steel] Accommodation unlikely of WP] (Temporary) accommodation possible, but difficult to sustain [environment, social policy] Table 2 summarises the propositions that emerge from the case studies about how the interplay between the structure of the policy process and sectoral policy paradigms affect the likelihood of an accommodation of the candidates preferences if such an accommodation is opposed by the dominant sectoral interest groups. It suggests that interest group opposition is an effective constraint on an accommodation of the applicants preferences only in those cases in which the policy process is fragmented and sectoral policy paradigms are incompatible with the applicants preferences. Sectoral policy paradigms are a key variable, concerning both the likelihood and sustainability of accommodation. Incompatibility between sectoral paradigms and preferences are a key obstacle to accommodation, even in a centralised policy process. In such cases, centralisation might still make accommodation possible, but only temporary accommodation. In the case of social policy, an incompatible policy paradigm constrained an accommodation even in the absence of sectoral opposition. 21

22 By contrast, compatibility between paradigms and preferences can lead to accommodation, despite interest group opposition. This was the case of the general orientation of the WP on regulatory alignment. Another example is the case of environmental policy in the EFTA enlargement negotiations, which allowed the new members to maintain higher standards of protection. Furthermore, compatible paradigms facilitate accommodation even in a predominantly fragmented policy process. The case of steel trade liberalisation illustrates that fragmentation makes it difficult for the policy advocates to gain access to decision-making. However, once access has been achieved, the policy paradigm facilitates alliance-building, which makes accommodation possible and likely to be sustainable. However, as suggested earlier, cases in which paradigms are compatible with an accommodation of the applicants preferences and interest groups are opposed to their accommodation are likely to be rare (and conversely, cases in which paradigms are incompatible and interest groups favour an accommodation of the applicants preferences). Such cases might be limited to periods of paradigm shifts, which makes it difficult to systematically test the relative importance of paradigm compatibility and interest group pressure. A key question to which the case studies draw attention is the likelihood of temporary accommodation (such as post-accession transition periods) when the candidates preferences are incompatible with sectoral policy paradigms. In this regard the contrast between temporary accommodation in the area of environmental policy and resistance to it in social policy is striking. This variation seems to present a puzzle as in both cases there was no interest group pressure at this stage of the policy process, the policy process was centrally coordinated and an accommodation of the candidates preferences is incompatible with the broader internal market paradigm that underpins both sectors. A tentative starting point might be to investigate more closely the status of the sectoral paradigms. Arguably, while in the EU there is a strong acceptance of high levels of environmental protection, acceptance of a further expansion of social policy is much more problematic. In other words, while the respective sectoral policy paradigms are strongly consensual within both sectors, their external acceptance in the broader context of the EU differs. While the acceptance of environmental policy outside the sector is high, it is more contested in the case of social policy. 22