NATIONAL INTERESTS AND EUROPEAN GAS

Transcription

1 KIRSTEN ZONDERVAN NATIONAL INTERESTS AND EUROPEAN GAS MARKET LIBERALISATION Explaining preferences for gas market policy in the Netherlands Msc Thesis Political Science Universiteit van Amsterdam Amsterdam, 3 July 2009

2 2 National Interests and European Gas Market Liberalisation Explaining preferences for gas market policy in the Netherlands Msc Thesis Research Project: The Political Economy of Energy International School for Humanities and Social Sciences University of Amsterdam Author: Kirsten Zondervan Student number: Supervisor: Dr. M.P. Amineh Second Reader: Dr. A.F. Correljé Cover Design:

3 3 Acknowledgements In the process of writing this thesis, I have enjoyed the help and support of several people whom I would like to express my gratitude to. First of all, I would like to thank my research supervisor Dr. Mehdi Parvizi Amineh for his invaluable support, encouragement and useful suggestions throughout the research process and for the interesting lectures he organised in the context of the research project The Political Economy of Energy, of which this thesis forms an integral part. I would further like to thank Dr. Aad Correljé, for taking up the role as second reader for this thesis and for providing me with some essential resources, which proved to be very helpful and considerably improved the quality of my research. I am deeply indebted to Dr. Daniela Obradovic, for her valuable suggestions on the theoretical part of this thesis. In addition, I would like to express my thankfulness to George Verberg and Jacques de Jong for taking the time to participate in interviews and for their valuable contributions to this report. Last but not least I would like to thank my parents for their continuous support through all my years of studying, my dear friend Jeanine for her everlasting patience and moral support during the writing process, my fellow students for their constructive feedback and my colleagues from APX for their moral support and genuine interest in my research topic.

4 4 Table of Contents Acknowledgements... 3 List of maps, tables and figures... 6 List of acronyms... 7 Chapter 1 Introduction Problem statement Theoretical and hypothetical framework Theoretical underpinnings of EU energy market reform Intergovernmental explanations of European integration: the logic of two level games Institutions as independent variables: the New Institutionalism Concluding remarks - Between intergovernmentalism and supranationalism An analytical model for endogenous preference change Methodology Organisation of the study Chapter 2 The need for liberalisation and regulatory reform Natural gas demand and supply in the EU The challenge of creating competition whilst sustaining the security of supply The rationale behind European gas market reform Extending the European single market to the energy sector The first steps towards the creation of an internal market for energy The first EU energy market directives: price transparency and transit rights for network operators The Hydrocarbons Licensing Directive Concluding remarks Chapter 3 Preferences for gas market policy in the Netherlands The structure of the Dutch gas market before EU liberalisation Uncertainties and status quo orientations of domestic actors Changing preferences for gas market organisation: the 1996 White Paper on Energy Policy Industrial realignments and increased support for reform... 63

5 Pressures for reform in the distribution sector The position of large industrial consumers The position of Dutch electricity producers Concluding remarks Chapter 4 Policy learning and leading norms of market integration Formal and informal norms of EU level negotiations Formal competences of the European Commission in the EU decision-making process The broad context of European integration Sector inefficiency and increased opportunities for transnational economic exchange Chapter 5 Conclusion References Appendix Appendix Appendix

6 6 List of maps, tables and figures Map 1. Member states of the European Union 9 Map 2. Map of the Netherlands 10 Table 1.1 Gross Inland Consumption of 1995 & 2005 in Mtoe 35 Figure 1.2 Gas demand forecast in EU25 36 Figure 1.3 Expected levels of natural gas consumption in the EU 37 Figure 1.4 Dependence according to energy product 37

7 7 List of acronyms BCM CCGT CEC DG TREN DSM DTe EBN ECB ECJ ECT EU Eurostat GDP HI IEA IEM LI MDW MEP MEZ MTOE NAM Billion Cubic Meters Combined Cycle Gas Turbine Commission of the European Communities Directorate-General for Transport and Energy De Nederlandse Staatsmijnen Dienst uitvoering en Toezicht Energie Energie Beheer Nederland European Central Bank European Court of Justice Energy Charter Treaty European Union Statistical Office of the European Communities Gross Domestic Product Historical Institutionalism International Energy Agency Internal Energy Market Liberal Intergovernmentalism Marktwerking, Deregulering en Wetgevingskwaliteit Member of the European Parliament Ministry of Economic Affairs Million Tons of Oil Equivalent Nederlandse Aardolie Maatschappij

8 8 NMa OPEC PSOs QMV RCI RES SEA SEP TPA TSO VEGIN VEMW WKK Nederlandse Mededingingsautoriteit Organisation of the Petroleum Exporting Countries Public Service Obligations Qualified Majority Voting Rational Choice Institutionalism Renewable Energy Sources Single European Act Samenwerkende Elektriciteits Produktiebedrijven Third Party Access Transmission System Operator Vereniging van Exploitanten van Gasdistributiebedrijven in Nederland Vereniging voor Energie Milieu en Water Warmtekrachtkoppeling

9 9 Map 1. Member states of the European Union in 2008 Source: European Commission. Audiovisual Service.

10 10 Map 2. Kingdom of the Netherlands Source: University of Texas Library.

11 11 Chapter 1 Introduction In the last decade European energy markets have gone through major restructuring processes due to reforms that were initiated at the European level. As part of the broader process of completing the European single market, the European Commission has continuously pushed for the integration of gas and electricity markets and the creation of a single European energy market. In the case of gas market liberalisation the first EU Gas Directive in 1998 marked a watershed in the development of the European gas sector, which until then was primarily characterised by a managed market structure and the absence of gas-to-gas competition (i.e. competition between gas suppliers). In traditional European gas markets, the strong natural monopoly characteristics of both gas production and transportation justified the highly monopolistic structure of the gas market, which was often dominated by a single national merchant transmission company. From the outset, the majority of member states as well as major players in the industry have strongly opposed the plans of the European Commission to liberalise the energy sectors. Member states largely feared that liberalisation would disrupt existing sectoral practices and therefore undermine the performance of public service obligations and other national policy objectives that were traditionally linked to the existing type of market organisation. Moreover, conflicting preferences of member states and divergent national gas market structures further complicated the decision-making process at the European level. However, despite initial opposition and conflicting interests, member states governments and the natural gas industry eventually came to embrace the Commission s proposals for the creation of an internal gas market, which resulted in the adoption of the first EU Gas Directive. Through the Gas Directive, competition and liberalisation were gradually introduced on the European gas market. And although the reforms were not as far-reaching as the Commission and the European Parliament had originally proposed - the directive only sets out very basic principles for reform and leaves it to member states to define the scope of their national regulatory framework - it paved the way for a profound restructuring of European gas markets and the creation of an internal market for natural gas. 1.2 Problem statement This study aims to assess how underlying national interests in the natural gas sector have changed in the course of EU level negotiations and how these interests interact with institutional forces at the European level. Gas market liberalisation in the European Union represents a puzzling case since the integration of national gas markets took place despite divergent and conflicting interests of member states. Furthermore, it resulted in deeper liberalisation in several member states, beyond the minimum requirements of the gas market directives. The unanimous adoption of the first Gas Directive, which

12 12 paved the way for a structural reorganisation of the European natural gas sector, suggests that factors other than pure economic interests account for the changes in national preferences and regulatory reforms at the national level. A glance at the contemporary literature on the topic of gas market liberalisation reveals that many studies lack extensive research on the motives of European member states to liberalise the natural gas sector. Most of the studies that have been conducted on the topic of gas market liberalisation in the European Union focus on the domestic impact of the Gas Directives or the convergence or Europeanisation of national regulatory regimes as a result of liberalisation measures at EU level 1. Whereas many studies have been conducted with regard to the liberalisation of other network-based industries such as electricity (Eising 2002; Eising and Jabko 2001; Schmidt 1996, Andersen and Sitter 2007; McGowan 1993; Padgett 1992; Matláry, 1997) and telecommunications (Schmidt 1996; Sandholtz 1993; Thatcher 2001; Levi-Faur 2004; Hulsink 1999), empirical evidence for the factors underlying gas market reform in the European union seems to be lacking. The reason why so few scholars have engaged in empirical studies on the driving forces behind EU gas market liberalisation is presumably that the creation of an internal market for natural gas in the European Union is by many regarded as a spectacular failure. In recent years many scholars, policy makers and the European Commission to a certain extent as well, have argued that competition on the European gas market is lacking and the natural gas sector continues to reflect traditional market structures. Nevertheless, the fact remains that the natural gas sector in the European Union has undergone a profound restructuring process, a development which can largely be attributed to the impact of reform initiatives that were present at the European level. The ultimate aim of this research is to generate evidence on causal mechanisms that explain under which conditions the preferences of member states change and to what extent the institutional context of EU level decision-making shapes or influences the process of national preference formation. An indepth analysis of the gas market liberalisation process in the Netherlands is conducted with the aim to gain insight into the institutional mechanisms that foster change at the national level, and to explain political realignments of both government and the industry. In the Netherlands, both government and major players in the gas industry expressed strong opposition to the liberalisation plans of the European Commission. Since the existing structure of the natural gas market was in the interests of both parties, the main focus of the Netherlands in early liberalisation debates was on maintaining the status quo. However, a complete turnaround took place in the Dutch preferences for gas market reform in the course of EU negotiations in the mid-1990s. By the time the first EU Gas Directive was adopted, the Netherlands was at the forefront of radical change in this sector. 1 On the convergence of national regulatory regimes see especially Haase, 2008.

13 13 In this study gas market reform refers to the process in which traditional sector organisations are partly or entirely replaced by a market model of organisation, through new regulatory practices and the introduction of competition. It should be noted that liberalisation is not the same as privatisation; liberalisation refers to the removal of barriers to free competition between a large number of market players, whereas privatisation means the selling of public undertakings to private parties. Liberalisation of the European natural gas sector means that market players of other member states can compete freely with national gas undertakings, that import- and export restrictions are removed and that gas consumers have free choice of gas supplier. A precondition for liberalisation is that essential network facilities such as transmission networks are open for third parties, so that market players can compete with each other on a fair and non-discriminatory basis. In order to analyse how national preferences for gas market reform have shifted in the course of EU negotiations, this study focuses on a specific time period. The period under study starts with the Commission s first proposal for an Internal Energy Market (IEM) in 1988 and ends with the domestic implementation of the first EU Gas Directive in the Netherlands in The leading question of this research is: to what extent can changing national preferences towards gas market reform be explained by the institutional dynamics of EU negotiations? In order to provide a comprehensive account of how national preferences for gas market reform have changed in the course of EU level interactions, the following sub-questions will guide the analysis: 1) What stance did the Netherlands take in the EU liberalisation debate/negotiations and how did this position evolve over time? 2) Which policy actors, market players and other stakeholders were involved in the liberalisation process in the Netherlands and how did these actors respond to the plans of the European Commission to liberalise the European gas market? 3) How did the government legitimise its decision to liberalise the domestic gas market and depart from the status quo? 4) To what extent did the Netherlands move beyond the scope of EU level reforms? 5) To what extent do national reforms reflect European norms of market integration? The assessment of changing national preferences and underlying interests in the natural gas sector faces two important limitations. First, isolating preference changes from strategic reorientations or tactical adjustments of national governments is a very complex task. Empirical data such as official policy documents may not reveal the real preferences of member states, due to strategic motivations of governments. In order to uncover whether changes in preferences reflect changes in national interests or strategic realignments (for example in the form of side payments or package deals), some face-toface interviews were conducted to cross-check the findings of this research. In addition, secondary sources were used, including datasets and interviews that were conducted in the period under study, in

14 14 order to find out how state and non-state actors in the Netherlands have repositioned themselves in the EU liberalisation process. For the purpose of this research it is assumed that when member states move beyond the scope of EU level reforms, this provides sufficient evidence that a change in national policy preferences took place rather than strategic realignments. Secondly, due to limited time and resources only one national situation will be studied in detail. It is not to say that if the empirical findings of this study confirm our theoretical model, this model will also be valid in other national situations 2. Contrary to many existing studies on gas market liberalisation in the EU, this research involves neither an extensive policy analysis in the conventional sense, nor an in-depth assessment of the legal provisions of the Gas Directives. It rather seeks to connect contemporary developments in the European gas market to European integration theory by applying key theoretical concepts of the New Institutionalism to a specific policy sector, i.e. natural gas. Many mainstream analyses of European policy assume that the policy preferences of member states reflect domestic economic interests, which are fixed and stable over time and are developed independently of EU level interactions. However, in the case of gas market liberalisation the explanatory power of such approaches remains limited. When analysing policy changes in domestic sectoral arrangements, one needs to take into account the EU institutional context as well in order to explain why national governments preferred certain policy options to alternative or existing policies. Like the overall process of European integration, attempts to establish a single European market for natural gas went through periods of stagnation and conflict. In the European integration literature this stop-and-go nature of integration has led to some fundamental problems of asymmetrical explanation; whereas many studies have focused upon the factors that prevent common policies from emerging 3, only a few studies have attempted to explain how a common denominator emerges among various policy outcomes. When looking at the process of gas market integration in the EU, there is no agreement about how market integration in this sector has come about and what the driving forces and impediments are for further integration in this specific policy area. However, at a time when the political significance of the European Union (EU) is increasing both internally, due to EU decisionmaking impacting more and more areas of domestic life, and externally, due to the Union s enhanced scope for external action (Eilstrup-Sangiovanni, 2006, xiii), understanding complex political processes like gas market integration seems now more important than ever. 2 According to Aspinwall and Schneider (2002) one-at-a-time case studies do not permit valid generalizations as long as no causal model exists to assist in the inevitable task of counterfactual reasoning (p.24). It should be noted however, that extensive generalisation is not the main objective of the research. This study aims to explain certain preconditions for changes in member states preferences on the basis of a theoretical model and does not claim that the causal model will hold in all instances of sector-specific decision-making in the EU. 3 Examples include studies on the (non-emergence of the) EU s Common Foreign and Security Policy (CFSP).

15 Theoretical and hypothetical framework The following paragraph sketches the theoretical framework of this study and provides the theoretical underpinnings for the model which will guide the research design. In this study, a thick neoinstitutional approach is used, which draws upon concepts of bounded rationality, policy learning and a logic of socialisation. In the neo-institutionalist approach, institutions are broadly defined as a set of formal and informal norms and rules that structure organisations and guide the behaviour and expectations of actors (or agents) about policy outcomes. It is assumed that national interests are embedded in institutions which provide the constraints and opportunity structures for strategic action. In this study, the basic policy preferences of EU member states for the organisation of the domestic gas sector, i.e. competition or maintaining the status quo, provide the starting point for analysis. However, instead of merely asserting that actors with given properties form their preferences on the basis of domestic material interests, this study also takes into consideration the effects of institutions - as formal and informal social norms - on the interests of member states. It is assumed that the EU institutional setting alters the perceptions of member states over preferable outcomes of different policy options, which ultimately leads to changes in policy preferences. In order to strengthen the claims derived from our theoretical model, the guiding hypotheses of this research will be systematically tested against a liberal intergovernmentalist model of preference formation. This allows us to incorporate a control variable which should disclose whether changes in member states preferences for gas market reform derive from interaction at EU level or changes in the domestic setting. The next section will provide an explanatory account of the guiding theoretical concepts of the New Institutionalism and will describe the liberal intergovernmentalist thesis of European integration. Moreover, this section will also provides a short overview of the theoretical underpinnings of the proposed EU gas market reforms, which illustrates in which light the proposed reforms should be interpreted Theoretical underpinnings of EU energy market reform In order to understand the background of the regulatory and structural reforms of European gas markets, it is necessary to look at the underlying theoretical assumptions that guided the proposals of the European Commission for gas market reform. First and foremost, one of the main objectives of the European Commission was to complete the internal market, a project which already started with the signing of the Single European Act (SEA) in The SEA provided for an internal market without frontiers in which the free movement of goods, services, persons and capital is ensured. Whereas the European gas and electricity sectors were initially not included in the 1985 White Paper on completion

16 16 of the Single Market, the internal market programme provided the main rationale for the energy liberalisation agenda in the years following the process of trade liberalisation and economic reform in the EU. From the beginning of the European integration process, the removal of economic barriers between EU member states has been one of the core integration objectives of the European Commission. According to Andersen and Sitter (2007) the Commission s effort to establish and extend the Single Market into areas formerly protected by national member states, fits into a broader global pattern of market- and institution-building that started in the late 1980s. After the collapse of communism in the Soviet bloc, general market-based rules were the order of the day as optimistic prophesies about the end of history and triumph of globalisation abound (p. 6). In the European Union, the completion of the Single Market took place against the backdrop of these developments in the global political economy and the revival of liberal ideology. The Energy Charter, which aimed to establish a free-trade area for energy beyond EU borders and promote energy cooperation between the states of Eurasia, should be interpreted in the light of these broader geopolitical developments. Moreover, efforts to liberalise the energy sectors in the European Union coincided with the liberalisation of network-based energy sectors elsewhere, for example in the United States where the results of liberalisation were very positive. In addition to the completion of the Single Market, the development of a competitive market for natural gas formed one of the core reform objectives of the European Commission. Through the gradual introduction of gas-to-gas competition, the European Commission aimed to increase efficiency, lower prices and enhance the security of supply (European Commission, 2006). According to the 2006 Green Paper of the European Commission the consolidation of the energy sector should be market driven if Europe is to respond successfully to the many challenges it faces and to invest properly for the future (ibid., p. 3). The liberalisation approach of the European Commission is founded on a set of neoliberal assumptions that guide the structural reforms in member states and provided for a justification of policy initiatives at the European level. The existing structure of European gas markets prior to liberalisation was first and foremost based upon principles of neoclassical economic theory. In the neoclassical view, rational utility-maximizing behaviour of individuals and firms is expected to lead to a Pareto optimal situation where market regulation is not necessary. Once a situation of perfect competition has been reached, supply and demand should be regulated by market forces rather than state intervention. However, in the neoclassical approach state intervention is not entirely absent; politics comes back into the field of economics in the theory of market failure.

17 17 In general, three broad types of market failure can arise, that is: market externalities, public goods and monopolies. Externalities refer to unintended consequences of exchange transactions for third parties that are not reflected in the market price. Public goods are products which cannot be produced for the market because they exhibit non-rivalry and non-exclusivity characteristics which may cause problems for the production of these goods. Non-rivalry refers to a situation where the consumption of a particular good by one person does not reduce the amount available for other persons. And nonexclusivity means that nobody can be excluded from enjoying the benefits of a good, even when they do not pay for it. Furthermore, monopolies and all other forms of market concentration which impede the free market entrance of competitors should be prevented through a market external actor such as the state or an independent regulator. In the neoclassical political economy, the role of the state is restricted to that of a market facilitator which should intervene only when the market fails to reach a situation of perfect competition (Spanjer, 2008, p. 12). However, the special characteristics of the natural gas sector invited extensive government intervention in order to remedy market imperfections and adjust market failures. In Continental Europe, many national energy market policy objectives included Public Service Obligations (PSOs), which were aimed to serve the public interest. PSOs are compulsory services which are allocated to public or quasi-public companies by the state in order to secure economic and social cohesion objectives, security of supply, quality and price of gas supply, and environmental protection (Spanjer, 2008). In the neoclassical perspective, goods and services which serve the general public interest should not be left to the market. Since public goods such as the security of supply and environmental protection exhibit market failures, government intervention is justified under this paradigm. Thus, in traditional European gas markets public interests were secured through direct public ownership or PSOs. In the EU Gas Directives, gas supply is still recognised as a matter of public interest for which member states may impose public service obligations. In addition to public service obligations, the natural monopolies present in almost all segments of the natural gas market (e.g. infrastructure, storage, production and supply to end users), provided another ground for state intervention under the neoclassical paradigm. Because substantial investments are required for developing new transport, production and exploration facilities, a monopolistic market structure, which spreads price and volume risks along the gas supply chain, is likely to emerge. Moreover, considerations with regard to environmental planning often prevented the establishment of multiple networks for gas transport, through which competition in this sector was largely ruled out. In continental European gas markets, integrated state-owned monopolies have traditionally assumed the responsibility for establishing new infrastructure and production facilities. As a result, the European natural gas sector managed to develop quickly.

18 18 Neoclassical economic theory thus invites government intervention in the market in the case of market failures 4. From the outset, the organisation of the natural gas sector in Europe has been justified on these principles. However, in the 1970s a paradigm shift occurred, in which the traditional neoclassical view was gradually replaced by a more liberal view on the functioning of markets. Increased efficiency and economic reform in the public sectors was propagated by a philosophy of rolling back the state and bringing in competition into formerly protected sectors (Correljé, 2005). The creation of the single European market through which formerly protected sectors are increasingly subject to European-wide competition should also be interpreted in this light. In order to enhance competition and facilitate the transition from a highly regulated monopolistic market towards a competitive market, regulation plays a central role in the energy market liberalisation process. Whereas the in the neoliberal approach state intervention is reduced to a minimum, neoliberals recognise the need for regulatory intervention to adjust market failures and create favourable conditions for competition. According to Majone (1994), regulatory intervention is based upon a single normative justification While nationalisations and other traditional forms of direct state intervention were [ ] justified by appealing to a variety of often conflicting goals, regulation has a single normative justification, improving the efficiency of the economy by correcting specific forms of market failure such as monopoly, imperfect information, and negative externalities (Majone, 1994, p. 378). In the liberalised European gas market, regulatory measures were introduced in order to make markets work. Instead of a mere retreat of state intervention in the gas market (letting markets work), regulation creates the framework through which a competitive European gas market can emerge Intergovernmental explanations of European integration: the logic of two level games In order to provide insight into the main theoretical propositions of liberal intergovernentalism, which represents the control variable in this research, an overview of intergovernmental explanations of European integration is provided in this section. After this, the basic propositions of the New Institutionalism will be discussed, which will provide the basis for the theoretical model that will guide the research. Traditionally, theories of European integration emphasize either the role of supranational institutions (supranationalism), or the enduring autonomy of nation states in the European integration process (intergovernmentalism). Intergovernmentalists historically attribute little influence to supranational actors. They explain processes of European integration largely in terms of coalition building activity of rational national governments on the basis of common and overlapping interests. Supranationalists 4 for a detailed overview and typology of market failures on the European gas market see Spanjer, 2008.

19 19 on the other hand, stress the relative autonomy of European institutions and focus on the impact of supranational actors in order to explain integration. In the theory of intergovernmentalism, a logic of two level games is introduced to explain the interaction between domestic politics and international relations. In the first game, states define their interests on the national level under pressure of domestic interests groups. In the second game which takes place at the level of interstate bargaining, national governments seek to maximize their own ability to satisfy domestic pressures and defend their national interests, while minimizing the adverse consequences of foreign developments (Putnam, 1988, p. 434). Although intergovernmentalism sees some potential for the cooperation between states, national sovereignty remains a key determinant of international politics (Eilstrup-Sangiovanni, p. 90). For this reason, intergovernmentalists assert that in the European integration process only technocratic and uncontroversial areas of low politics are susceptible to integration, whereas areas of high politics involving vital national interests are virtually immune from integration impulses (Hoffmann, 1965). This assertion, based on the propositions of classical realist thinking, served as the main source of criticism towards another grand theory of European integration, namely Neofunctionalism. Neofunctionalism, as developed by Ernst Haas (1958) and others in the late 1950s and 1960s, can be regarded as the first comprehensive theory of European integration (see Lindberg, 1963; Scheingold, 1970; Schmitter, 1969, 1971; Nye, 1965). Based on theoretical concepts of social pluralism, rational choice and spillover, Neofunctionalism regards European integration as an automatic process which is inherently expansive, irreversible and self-sustaining. Due to the nature of modern economies, economic integration in one sector is expected to automatically generate integration in other functionally linked sectors of the economy. In addition to functional spillover, which occurs when integration in one sector of the economy leads to economic distortions in other functionally linked policy sectors, neofunctionalists also distinguish a logic of political spillover. Political spillover occurs when supranational institutions, which were initially set up to manage technical issues of economic integration, gain a life of their own and provide an impetus for further integration and institutionbuilding in the EU. Furthermore, the activities of supranational institutions are expected to lead to convergence of national interests and expectations in the long term. As a result of supranational integration, domestic interests groups are expected to shift their loyalties, expectations and political activities towards a new centre since their interest are better served by supranational rather than national solutions (Haas, 1968). For this reason, neo-functionalism is also commonly referred to as a theory of supranationalism. The intergovernmentalist critique was essentially targeted at the neofunctionalist logic of spillover, which predicted a gradual spillover from economic cooperation to political integration and eventually

20 20 to a political union between member states (see especially Hoffman, 1965; 1966; 1982, Hansen, 1969, Moravcsik, 1998). Intergovernmentalists argued that this prediction was mistaken since states would never delegate control over sensitive areas that touch upon the realm of power and high politics. European integration in the intergovernmentalist perspective only occurs when the national interests of member states converge and when integration touches upon areas which are outside the scope of national security. Because of diverging interests in the area of high politics, integration is expected to take place only in the domain of low politics (Eilstrup-Sangiovanni, 2006). In addition, neofunctionalist theory came under the strain of empirical developments when the process of European integration came to a slowdown in the 1970s and the functional spillover from economic cooperation to political union failed to materialize (ibid.). The liberal intergovernmentalist (LI) model as developed by Moravcsik (1991, 1993, 1998) builds upon existing theories developed in the field of International Political Economy (IPE) as well as regional integration theory and international regime theory. Within the liberal intergovernmentalist model, the costs and benefits of increased economic interdependence between states are the primary determinants of national preferences The basic claim of LI is that the European Union can be conceived as an intergovernmental regime designed to reduce negative economic externalities between states through negotiated policy coordination (Moravcsik, 1993, p.474). Negative policy externalities occur where the policies of one nation imposes costs on the domestic nationals of another, thereby undermining the goals of the second government s policies. (ibid., p.485). These include for example protectionist policies pursued by foreign governments, currency devaluations and low standards for environmental protection. Furthermore, economic interdependence serves as an incentive for integration when unilateral policies dealing with the adaptation of the national economy to the changing structure of the global economy are either inadequate, ineffective or too expensive. The liberal intergovernmentalist framework for analysis is based upon a rational account of state behaviour, a liberal theory of how economic interdependence influences national interests and an intergovernmentalist theory of international negotiation (Moravscik, 1993, p.480). On the basis of these theoretical propositions, a model is developed which explains how national preferences are formed and under which conditions governments delegate or pool sovereignty to European institutions. rising interdependence domestic preference formation interstate bargaining delegation/ pooling of sovereignty zero sum policy outcomes

21 21 As discussed in the previous paragraph, rising economic interdependence between states provides the starting point for analysis in liberal intergovernmentalist model. In the following stage, national preferences are formed on the basis of the commercial interests of powerful domestic interest groups. According to Moravcsik (1993) groups articulate preferences, governments aggregate them (p.483). As a result, the ability of national executives to pursue certain policy goals depends on the degree of mobilisation of domestic interests groups. At the level of interstate negotiation political (zero sum) outcomes are determined by the relative bargaining power of states. The extent to which member state governments can pursue unilateral policy alternatives (threats of non-agreement), form alternative coalitions (threats of exclusion) or reach compromises, determines their relative bargaining power. Conflicts at the stage of interstate bargaining about the distribution of implementation and adaptation costs between states are resolved through coalition building, issue linkages and side payments. The fourth level explains institutional choice. In the theory of LI, governments decide to delegate or pool sovereignty to EU institutions in order to enhance the credibility of interstate commitments and to ensure compliance with interstate agreements. When applying the liberal intergovernmentalist model to gas market liberalisation in the EU, it becomes clear that this model does not grasp all the factors that contribute to political change at EU level. In the LI model, national preferences reflect the interests of domestic economic elites which provide the incentives for policy coordination at the European level. However, when the net expected costs of policy cooperation are considered too high, these elites will reject cooperation. This implies that European integration is only possible in cases where European governments can collectively overcome this opposition. Non-integration in the European gas sector could long be explained with reference to diverging national interests of European member states. Because national interests were too diverse and domestic interest groups (i.e. the national energy industry) showed no interest in energy market coordination at the European level, there was no impetus for integration in this sector. However, the organisation of natural gas sectors in Europe changed substantially as a result of reforms initiated at the European level in the 1990s. If the liberal intergovernmentalist model is correct, the willingness of European governments to cooperate at the European level should reflect changes in the domestic context. These include for example local performance and efficiency problems created by increased economic interdependence. Nevertheless, in studies on EU energy market integration it is often argued that this was not the case in the natural gas sector and that the liberal intergovernmentalist model of preference formation and interstate bargaining does not provide an adequate explanation for the empirical facts in this field (Andersen, 2000; Schmidt, 1996; Matláry, 1997, Eising, 2002). Because many policy choices in the area of energy market liberalisation could not be traced back to

22 22 pure national economic interests of member states, some of these studies have pointed to the effects of European institutions on the preference formation of member states. The idea that national economic interests do not always account for policy choices at the EU institutional level, fits into a broader criticism on intergovernmental explanations of European policy. Whereas the theoretical framework of Liberal Intergovernmentalism may well apply to major treatymaking bargains, it does not take into account the EU s administrative policies and day-to-day decision-making within the framework of the EU (Eilstrup-Sangiovanni, 2006). Furthermore, intergovernmentalists have a strong tendency to explain every political outcome at the European level in terms of near total member-state control, even when this does not fit the empirical facts. Where policy outcomes do not conform to the expected preferences of member states, they may be explained as part of a nested game or as an instance of side payments. (Pierson, 1996, p.305). Intergovernmentalists often rely on a post hoc examination of institutional choice to explain what the preferences of state must have been prior to the stage of interstate bargaining level. However, in order to be able to predict future outcomes and to provide a coherent account of the overall integration process, intergovernmentalists should explain what motivated states in the first place to achieve these functional outcomes. In the next section, a neo-institutional model is proposed in order to explain what remains exogenous to rationalist intergovernmentalist explanations of European integration, i.e. the endogenous sources of change in the policy preferences of member states Institutions as independent variables: the New Institutionalism Contrary to intergovernmentalist accounts of European integration where institutions are denied any meaningful role and are merely perceived as outcomes (i.e. to ensure credible commitments of member states), institutionalist theories of European integration envisage a more important role for institutions. According to Eilstrup-Sangiovanni (2006) the most important feature of new institutionalist analyses is that they focus on EU institutions not only as outcome variables whose origin and design needs explanation, but also as independent and/or intervening variables which crucially affect actors strategies and goals in the area of integration (p.194). In the EU literature, the debate between intergovernmentalists and supranationalists on the role of institutions often resulted in a deadlock. The introduction of new institutionalist approaches to European integration thus offered a welcome alternative to overcome this impasse. According to Pollack (1997) the primary virtue of the New Institutionalism [...] is that it allows us to transcend the intergovernmentalist-neofunctionalist debate by acknowledging the initial primacy of the member states and, proceeding from this point, to generate a series of hypotheses about

23 23 supranational autonomy and influence more precise than those generated by either neofunctionalist or intergovernmentalist theory (Pollack, 1997, p.101). The New Institutionalism was largely developed in reaction to the behaviouralist turn in the social sciences in the 1970s. As many other theoretical approaches in the EU integration literature, the basic theoretical propositions of the New Institutionalism were largely derived from the international relations literature. However, despite the fact that the New Institutionalism did not originate in the field of European studies the institutional structure of the European Union represented an ideal test case for the study of institutions and political outcomes. This allowed for the further refinement of neo-institutionalist premises and the development of a set of testable hypotheses in the field of European politics. The new institutionalism differs from traditional institutionalism in the sense that the old institutionalism focused more on the formal aspects of decision-making, whereas the New Institutionalism also stresses the importance of informal norms and rules within political decisionmaking structures. According to March and Olsen (1984) the New Institutionalism argues that preferences and meanings develop in politics, as in the rest of life, through a combination of education, indoctrination, and experience. They are neither stable nor exogenous. (p.739). Since political preference formation, as all other human behaviour, is shaped through historical experiences, it does not make sense to assume that preferences are exogenous to the political process (ibid., p. 739). Neo-institutionalism presents a theory that is capable of endogenising the underlying rational beliefs and desires of actors preferences and can explain the variation in political outcomes and policy preferences over time (Kerremans, 1996). According to Hall and Taylor (1996), three different new institutionalisms can be distinguished with distinctive ontological approaches towards social and political problems. These are historical institutionalism, Rational Choice Institutionalism and sociological institutionalism. According to Hall and Taylor (1996), the New Institutionalism does not constitute a unified body of thought (p. 936). Instead, every type within the New Institutionalism paints quite a different picture of political processes, emphasizing other levels of analysis and relying on other outcome variables. In the next section, a brief outline of all three approaches will be provided with the aim to provide some background information on the selection of the explanatory and outcome variables in this study. Historical Insitutionalism Historical institutionalism (HI) focuses on the development of institutions over time and the unintended consequences of established institutions for its creators, i.e. the EU member states. Historical institutionalists define institutions as formal or informal procedures, routines, norms and conventions embedded in the organisational structure of the polity or political economy. (Hall and

24 24 Taylor, 1996, p. 938). Institutions are thus viewed as the rules and norms that are derivative of formal organisations. One of the distinguishing features of historical institutionalism is that it accepts the intergovernmental premise that national governments are the primary actors in the creation of multilateral institutions. Namely that institutions are created as functional means to serve the collective interests of national governments. (Eilstrup-Sangiovanni, 2006). However due to the dynamics of path dependence, institutions over time become more difficult to control by national governments. According to Pierson (2000) the high costs associated with the creation of new intuitions makes institutional development vulnerable to increasing returns. Once institutions are established and a country or region has started down a track, the costs of exit or change are rising with every step in that given direction, which reinforces the strength of institutions and determines further development (Levi as cited in Pierson, 2000, p. 252). Because of the high costs of change, initial policy choices will determine subsequent political developments, hence the notion of history in historical institutionalism. In addition, institutional decision-making rules such as voting procedures and learning mechanisms are also expected to lead to path-dependent outcomes. In the case of policy learning, the costs of policy change increase when governments adapt national institutional arrangements to the new policy and make investments based on expectations about this new policy. According to Pierson (2000) these investments increase the attractiveness of existing policies relative to hypothetical alternatives. As social actors make commitments based on existing institutions and policies, their cost of exit from established arrangements generally rises dramatically. (Pierson, 2000, p. 259). In addition to the logic of increasing returns, historical institutionalism also stresses the relative autonomy of institutional actors in the European integration process. Institutions, once they are established are expected to use their autonomy to advance their own interests, which results in a loss of control over the integration process by national governments. Furthermore, the policy preferences of national governments with restricted time horizons may change over time, leading to a divergence between prior institutional choices and national preferences. Where these gaps occur, member states find it more difficult to regain control over the policy-making process and institutional actors are more likely to pursue policy in a direction which suits their own institutional preferences. Moreover, a combination of national governments focusing on electoral gains instead of long-term effects of policy choices and the technical complexity of EU policies implies a significant potential for unintended consequences. Historical institutionalism can be described as a mid-level theory that treats institutions as intervening variables in the EU integration process. Nonetheless, the theory fails to provide a coherent account of the overall integration process, since the ultimates cause of events remain exogenous to the

25 25 institutional setting. HI for instance does not explain the motivations behind institutional choice and merely focuses on institutional stability and change, which renders the theory far better equipped to explain the persistence and stability of institutions than to account for change (Eilstrup-Sangiovanni, 2006, p.200). Rational Choice Institutionalism Rational Choice Institutionalism (RCI) is the second approach within the neo-institutionalist framework. This version of the New Institutionalism does not represent a coherent body of thought, but rather constitutes a broad approach towards the social sciences and the study of institutions in general. Rational Choice Institutionalism is associated with a high degree of empiricism and does not attempt to provide a grand theory of European integration. Within the study of EU politics, rational choice theories have been applied to EU decision-making processes in order to explain under which circumstances member states delegate power to supranational agents and to what extent national governments can control the effects of authority delegation to institutional actors (see Scharpf, 1988; Majone, 2000; Tsebelis,1994; Pollack, 1996). Rational Choice Institutionalism (RCI) resembles the liberal intergovernmentalist approach to the study of European decision-making to a certain extent, since both RCI and LI are based on the precept of methodological individualism, which means that they both focus on the behaviour and actions of individuals in order to explain social phenomena. Furthermore, both Liberal Intergovernmentalism and Rational Choice Institutionalism perceive of institutions as a functional means to reduce transaction costs and to solve problems of incomplete contracting. Cooperation, therefore, is instrumental, and is not necessarily a socially-ingrained and habitual practice (Aspinwall & Schneider, 2002, p.11). Another important feature of Rational Choice Institutionalism, which distinguishes RCI from the other approaches within the neo-institutionalist paradigm, is that it regards preference formation as exogenous to the institutional setting. According to Aspinwall and Schneider (2002), there is a methodological explanation for the fact that rational choice theorists do not endogenise preference formation in their model. The scientific method used by these theorists namely stipulates that at least one of the explanatory variables should be held constant for the theory to be testable. If a model allows every variable to change, nothing can be explained [ ] preference change can thus only be modelled if other factors are held constant. (ibid., p.11). In the EU integration literature, rational choice analyses have become dominant in the study of executive bodies within the EU, such as the European Commission and the European Central Bank (ECB). In these studies the main research questions centre upon the conditions under which member states delegate power and authority to supranational agents. However, another very important question that is posed within rational principal-agent analyses relates to the unintended effects of delegating

26 26 powers to institutional agents and the conditions under which state principals cannot control these effects. What if an agent such as the Commission, the Court of Justice, or the ECB behaves in ways that diverge from the preferences of the principals? (Pollack, 2006, p. 39). In order to answer the second question, rational choice institutionalists point to the administrative procedures and institutional monitoring procedures that were set up by member state principles in advance in order to control the behaviour of institutional actors. Contrary to the liberal intergovernmentalist approach to European integration, institutions in the rational choice approach are not mere reflections of underlying national preferences; they are also crucial in distributing decision-making powers among actors. Variation in political outcomes is explained in terms of different institutional procedures and varying power between different policy actors at the institutional level. One of the main strengths of Rational Choice Institutionalism is that it specifies under which conditions member states can or cannot control the unintended effects of delegation of authority to institutional actors. However, its main weakness lies in its ontological blindness to the issue of endogenous preference formation and change (Pollack, 2006). RCI fails to account for changes in member states preferences because it specifies functions in advance and assumes that preference formation takes places in isolation from the institutional venue. This makes prediction impossible. Sociological institutionalism Sociological institutionalism offers a viable alternative to overcome the problem of endogenous preference change. In the sociological institutionalist account, institutions are broadly defined as collective norms, rules and procedures that guide actors behaviour and shape identities and interests. Sociological institutionalism has much in common with the broader constructivist approach in International Relations and European Union studies, which also emphasises the role of informal rules, norms and ideas in explaining political outcomes at the European level. Contrary to rational choice models of preference formation where actors are driven by a logic of utility maximisation, constructivists argue that member states behaviour is largely determined by a so-called logic of appropriateness which includes socially constructed norms, ideas and values. As a result of social interaction in political decision-making processes, actors are exposed to a number of socialisation mechanisms which may alter their beliefs and expectations about certain policy options and ultimately lead to a change in policy preferences. In the rational choice approach social interaction may lead to changes in strategic action of member states, but not to a redefinition of their preferences. The process through which agent properties and preferences change as a result of social interaction is only assumed by non-rationalist (constructivist) scholars. Because rational choice theorists regard preference change as exogenous to social interaction, this hypothesis is not valid in rationalist analyses.

27 27 According to some constructivist scholars in the field of European integration, interaction within European institutions can be characterised as a process of social learning (Hall, 1993; Checkel, 1999, 2003). Through processes of argumentation, deliberation and persuasion 5, member states acquire new insights, which may affect their understanding of what is in their own interest. As a result, member states may change their behaviour and policy preferences in ways that cannot be explained by material incentives (Eilstrup-Sangiovanni, 2006). Within the sociological account of institutionalism, socialisation takes place when member states internalise certain patterns of behaviour, norms or rules of the institutional setting in which they interact. A member state has internalised shared norms of the institutional level when no external sanctions are needed to ensure compliance with new social norms. Because of the fact that the perception of rational and legitimate policies is socially constructed, the policy preferences of member states will reflect the rules for appropriate behaviour within the institutional setting. Rather than act in accordance with their calculated self-interest, actors will do what they feel is most appropriate given their socially defined role. (Eilstrup-Sangiovanni, 2006, p.396). The constructivist logic of appropriateness implies that strategic action of states is embedded in the institutional setting, which influences actor s perceptions about the desirability, feasibility and legitimacy of certain policy options and challenges existing policy paradigms. A policy paradigm refers to a framework of ideas and standards that not only specifies the means and ends of policy, but also the very nature of the problem that should be addressed (Hall, 1993). According to Hall (1993) 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 (p. 279). In the process where an existing policy paradigm is replaced by a new paradigm, argumentation and persuasion represent two important sources of change. Persuasion is a cognitive process which changes the attitudes of actors about cause and effect of certain policies in the absence of coercion (Checkel, 2003). In other words it represents a social learning mechanism through which underlying interests of actors may be redefined and policy preferences may change. In the EU institutional setting, well-positioned actors such as the European Commission, often have considerable leeway to convert their beliefs into broader, shared understandings among member states. Here, the redefinition of policy issues in a highly universalistic and abstract manner allows for the accommodation of many different interests policy preferences. Through strategic political entrepreneurship and its right of initiation, the European Commission can be very successful in turning its own beliefs and preferences into broader normative ideas at the European level. However, the extent to which the European Commission can influence the policy preferences of member states is 5 For a comprehensive theoretical overview of the role of persuasion and argumentation in preference change see Checkel, 2003

28 28 dependent on a number of preconditions. Checkel (2003) argues that influencing preferences through political entrepreneurship is more likely to succeed when certain policy windows are open. This is the case when member states face a policy problem which is difficult to approach or is new and unknown. In this situation social interaction often leads to the breakdown of initial policy preferences. Through its formal agenda-setting powers, the European Commission can largely determine the scope of new EU policies and the options available to member states. In a situation where member states face uncertainty with regard to the implications of different policy options, the Commission is likely to influence the preferences of national governments. (Eising, 2002, p.88). The research in this study will mainly draw on the theoretical concepts derived from sociological and historical institutionalism. RCI falls outside the scope of this model, since it regards preference formation as exogenous to the institutional setting and it does not represent a suitable model for the object under study, namely the influence of the EU institutional setting on member states policy preferences. Moreover, rational choice and constructivist approaches are based upon contradicting social ontologies. The fundamental sources of human behaviour from the rationalist perspective come from outside the agent and her context [...] from the sociological and, to an extent, the historical perspective, the agent and her context are co-determining; they evolve simultaneously, and do not exist independently. (Aspinwall & Schneider, 2002, p.2). Another important underlying difference between RCI on the one hand and historical and sociological approaches on the other, concerns the time frame. Whereas historical and sociological institutionalists focus on the long-term effects of institutional decision-making, rational choice theorists first and foremost concentrate on the dynamics of short-term decision-making Concluding remarks - Between intergovernmentalism and supranationalism From the above sketched overview it should be clear that the New Institutionalism provides a middleground between intergovernmental theories of European integration which stress the enduring control of national governments over the integration process and supranational approaches which tend to overemphasize the role of supranational European institutions. The various approaches within the neoinstitutionalist paradigm acknowledge the initial primacy of member states, but depart from this assumption when arriving at the stage of interstate bargaining by arguing that institutions are not mere neutral arenas for the enhancement of exogenous preferences but also affect actors strategies and goals in the process of social interaction. Instead of merely perceiving EU institutions as outcome variables, as is the case in Liberal Intergovernmentalism, the New Institutionalism regards institutions as independent and intervening variables, which allows for the analysis of endogenous preference change.

29 An analytical model for endogenous preference change In this section a theoretical model is proposed which guides the research design of this study and specifies which explanatory variables will be used in order to explain the changing position of the Netherlands in the process of European gas market liberalisation in the 1990s. While most sociological and historical institutionalist analyses are based upon inductive research, aiming to uncover the mechanisms behind institutional change, this study adopts a more deductive approach towards the research. A set of hypotheses derived from the New Institutionalism will be tested against a control variable from the liberal intergovernmentalist model to see if neo-institutional theory can explain the variation in political outcomes and national preferences in the case of gas market reform in the Netherlands. By developing a coherent framework for causal analysis which consists of a limited number of explanatory variables and by incorporating a control mechanism in the form of a counter variable derived from LI, this study attempts to respond to the main source of critique towards institutionalist analyses which is centred upon a systematic lack of analytical rigour and an ad hoc selection of independent variables. In this research it is assumed that member states preferences change as a result of social interaction at EU level. Because member states internalised new norms that were introduced at the European level in the form of a new policy paradigm, they adjusted their preferences in favour of liberalisation of the European gas market. If this theory is correct, we should observe that national governments and other domestic policy actors changed their preferences in accordance with the emergence of new norms at EU level. Hence the first hypothesis: if member state governments change their preferences in accordance with new European norms, national debates and national policies should reflect these (internalised) new norms. The second hypothesis represents a counterhypothesis that should demonstrate that observable implications generated by the liberal intergovernmentalist model cannot account for the changes in member states preferences for the reorganisation of domestic gas sectors. For this reason, the second hypothesis consists of the following: If member state preferences reflect mere domestic economic interests, we should detect a significant change in material circumstances in the natural gas sector prior to the shift in national preferences for gas market reform. The explanatory variables of our theoretical model include 1) formal and informal norms of EU level negotiations, 2) the formal (agenda-setting) competences of the European Commission in the EU decision-making process and 3) the broader context of European integration. A fourth control variable has been added to test the main hypothesis of this research against competing claims of the liberal intergovernmentalist approach. This variable includes (4) the inefficient allocation of domestic resources and rising opportunities for domestic actors to profit from transnational economic exchange.

30 30 Liberal theories of European integration claim that increased economic interdependence creates negative policy externalities for member states, which in turn creates incentives for policy coordination and market liberalisation at the European level. It is assumed that trade restrictions and protectionism are not in the interest of domestic interest groups which explains their motivations to press for market integration. In addition, the inefficient allocation of domestic resources and increased competition in international markets gives rise to pressures from domestic interest groups for liberalisation. On the contrary, neo-institutionalist analyses of European integration emphasize the importance of norms in providing justifications for certain policy choices at the national and the European level. In our analysis, the liberal model of market integration on the basis of which the European Commission justified the expansion of the single market into the energy sector, represents the leading norm in the process of European gas market liberalisation. In addition, norms of consensus that are strongly present within Council negotiations represent another important factor in the explanation of preference change in the course of EU level negotiations. Since strong social consensus norms restrain member states from rigidly pursuing their own national interests, negotiations within the Council of Ministers may lead to a reassessment of domestic situations and even to a change in fundamental policy preferences Although obstructionist or purely self-serving negotiating tactics remain possible, there is a risk of backlash if a majority of Council members is antagonized. Those member states willing to defend their national settings are challenged to come up with compromise proposals that expand the areas of agreement. This in turn limits the range of legitimate arguments available to them unless they are willing to face the responsibility of permanent deadlock or failure of the negotiations. (Eising & Jabko, 2001). One of the main explanatory variables in respect to preference change with regard to gas market reorganisation is the introduction of a new EU level policy frame or paradigm which consists of a number of policy concepts that structure member states views and expectations regarding the appropriateness of certain policy choices and the preferred type of market organisation. The emerging neoliberal discourse of market integration and internal-market logic in which the debate over gas market reform was largely embedded, made it increasingly difficult for both national governments and the gas industry to defend traditional organisations of the gas industry on the basis of arguments against market integration. Because market integration provided a leading norm in the overall EU integration project, governments could not challenge the principles of market integration as such. They could only argue that the natural gas sector was not suitable to be subsumed under such principles, otherwise they would undercut the validity of the whole EU integration project.

31 31 According to Andersen (2000), the emergence of a new policy paradigm at the European level involves three interlinked stages. Argumentation introduction of the new policy frame Negotiation formulation of EU policy paradigm Implementation internalisation of new policy paradigm In the first stage, the new policy frame is introduced through a logic of argumentation. In this stage arguments about the appropriateness of certain organisational principles and type of market organisation play an important role. In the second stage, the content of the new policy frame is specified through debate and negotiation. At this stage, the implications of alternative organisational forms are widely debated in order to implement new policies at the third level. In the implementation stage the new policy paradigm is internalized through national implementation of EU level policies which gives further substance to the new EU level paradigm (Andersen, 2000). The second explanatory variable includes the formal competences of the European Commission in the EU decision-making process. The Commission s right of initiative grants it the authority to determine the scope of the EU policy agenda to a great extent. Furthermore, Commission proposals and progress reports provide national governments with new information about the consequences of EU policy and arrangements in other member states, which may lead them to subject their domestic situation to detailed scrutiny and re-evaluate initial policy preferences. As Eising (2002) maintains, the adoption of an EU directive by national governments greatly depends on the perceptions of member states about the impact of European policy on their domestic setting. By changing member states perceptions about policy outcomes, the European Commission can greatly influence the way in which member states perceive their own interests, which can ultimately lead to a redefinition of national preferences. Moreover, as long as the Council of Ministers (the Council of the European Union) has not decided on a Commission proposal, the Commission has the right to withdraw its proposals. By adjusting its proposals to the preferences of supportive coalitions, the Commission can put pressure on other member states to adjust their preferences as well. The broad context of European integration represents the third explanatory variable. This variable mainly refers to the ideological commitment of member states to the European integration project and the extent to which national policymakers refer to Brussels in order to legitimise domestic policies.

32 Methodology The following paragraph presents the methods and data which were used to test the hypotheses as described in the previous section. In this research a range of qualitative methods was used to operationalise the research questions. As stated earlier, the aim of the research is to generate evidence on causal mechanisms which explain how the preferences of government and industry for gas market reform have changed and how underlying interests in the natural gas sector interact with institutional forces at the European level. In the empirical part of this study the national liberalisation debate in the Netherlands is scrutinised in order to gather evidence for the claim that a shift in government and industrial preferences for gas market reform has occurred in the Netherlands. The second step was to demonstrate that EU institutional factors, instead of changing economic interests of domestic interest groups, account for the variation in national preferences with regard to liberalisation. In order to determine the influence of the EU institutional setting on national preferences, a technique of process tracing was used to reconstruct the motivations of the main policy actors and the industry at different points in time in the EU decision-making process. The empirical reconstruction of the preferences and motivations of the actors involved occurred on the basis of official policy documents, newspaper articles and in-depth interviews with people that were involved in the liberalisation process in the Netherlands. Furthermore, some secondary sources were consulted, which include for example the journalistic work of Noud Köper Tegenpolen (2003) about the liberalisation of the Dutch energy market and the dataset of Dertig Jaar Nederlands Energiebeleid (2005) by Jacques de Jong, which presents an overview of Dutch energy market policy from 1973 onwards and provides insight into the policy process and the positions of the main actors involved in the energy market liberalisation process in the Netherlands. In this study, process tracing serves two specific research goals. First, it should provide evidence that the relevant actors changed their policy preferences rather than strategic positions. Moreover, it should demonstrate that preference changes were induced by negotiations at the European level and not by changes in material circumstances at the domestic level. In order to establish whether and to what extent policy learning has occurred as a result of EU level negotiations, domestic adjustments of gas market policy will be examined in relation to EU level reforms. It should be noted however that a change in national gas market policy does not always imply that policy learning has occurred at the European level. Nevertheless, when member states exceed minimum EU standards in the national implementation of the directives, this should provide sufficient evidence that national governments were persuaded of the need for policy change rather than forced into a bargain. In order to examine to

33 33 what extent national gas market reforms have exceeded minimum EU standards, national gas market legislation 6 was compared with the provisions of the Gas Directive. To assess how the policy preferences of domestic actors have shifted in response to EU level developments, the positions of the main actors were assessed at different points in time. The actors involved in the liberalisation process in the Netherlands include both state and non-state actors, i.e. the Ministry of Economic Affairs, which is responsible for energy market policy in the Netherlands, the national gas industry as represented by the monopolist gas supply and transmission company Gasunie and gas producers, large industrial consumers and local and regional distribution companies. Because of interrelated developments in power production sector, the position of large electricity producers will also be taken into account. In recent years, natural gas in the Netherlands has become the preferred fuel for (combined heat and) power production, which has led to a large increase of gas consumption in the electricity production sector. 1.4 Organisation of the study In chapter 2 an overview will be provided of the natural gas supply and demand situation in the European Union. In this chapter, the problem of ensuring security of supply in the liberalised energy market will be highlighted. Furthermore, the underlying factors for a structural reorganisation of the European gas market will discussed, together with the historical efforts of the European Commission to create an internal market for natural gas. After this, a systematic overview will be presented of the negotiations leading up to the adoption of the first EU Gas Directive and the adoption of two earlier directives on price transparency and transit rights for European network operators. Chapter 3 describes the structure of the natural gas sector in the Netherlands and provides an overview of Dutch gas market policy since the discovery of the large Groningen gas field in Furthermore, this chapter includes a reconstruction of the motivations of the actors involved in the liberalisation process in the Netherlands and describes how the policy preferences of both state and industrial actors have changed throughout the process. Chapter 4 seeks to connect the empirical findings of this study with the theoretical framework. On the basis of the theoretical model as presented in section 1.2.5, developments in the liberalisation process in the Netherlands are tested against two competing claims, in order to find out whether changing preferences for gas market policy reflect the internalisation of European norms or material changes in the domestic context. Chapter 5 presents a brief summary of the findings and the conclusions resulting from this study. Moreover, it discusses the theoretical implications of the research outcomes. 6 In the Netherlands, the Gas Act (Wet van 22 juni 2000, houdende regels omtrent het transport en de levering van gas) was designed to implement Directive 98/30/EC. The Gas Act entered into force on 22 June 2000.

34 34 Chapter 2 The need for liberalisation and regulatory reform In this chapter, an overview of the natural gas demand and supply situation in the European Union will be provided, in which the import dependency of European countries on natural gas will be highlighted. The risks associated with Europe s high degree of import dependency will be discussed, which provided the main rationale for a structural reorganisation of the European gas market. Moreover, this chapter presents a historical overview of the Commission s efforts to create an internal market for natural gas and discusses the challenge of ensuring effective competition whilst sustaining the security of supply. Security of supply in this context refers to the degree to which, in the short and the longterm, the prospect of uninterrupted supply of gas can be assured (International Energy Agency [IEA], p. 48). It concerns in particular the capacity to maintain supplies in periods of peak demand, the marginalization of supply failure risks and the mobilization of sufficient volumes of gas. However, security of supply is also related to reasonable prices. Gas must be sufficiently available at reasonable prices. 2.1 Natural gas demand and supply in the EU Until the 1960s, natural gas did not play a significant role in the European domestic energy consumption. The natural gas sector in the European Union largely developed in reaction to the discovery of major gas fields in the Netherlands in 1959 and the United Kingdom in the 1960s. Since then, natural gas production and consumption increased rapidly in Europe. In many European countries, natural gas began to form part of the national energy mix because it allowed for a diversification of energy resources in the national energy mix and moreover reduced countries high dependence on oil. Demand for natural gas largely increased in response to higher oil prices, caused by OPEC (Organisation of the Petroleum Exporting Countries) price increases and the growing insecurity of Middle East oil supply. The two oil crises in the 1970s caused European countries to diversify imported energy sources away from the Middle East and focus more on indigenous energy resources, such as natural gas. In addition to rising oil prices and growing concerns about the security of supply, gas represented the cleanest energy source compared to other fossil fuels such as oil and coal. Natural gas emits fewer pollutants into the atmosphere and contributes to a healthy energy consumption balance. Whereas natural gas was long considered a second-rate energy product linked to the exploitation of oil, it soon found its way into all segments of the European energy consumption. In the 1970s, Norwegian gas provided for a further diversification of energy resources for countries that were heavily dependent on imports for their energy supplies. However, while countries such as Germany, Italy and France also discovered some indigenous gas resources, this was not sufficient to

35 35 meet growing European demand for gas. Eventually, Britain s indigenous gas reserves could not even meet domestic demand without additional imports (Stern, 1998). Today, gas consumption in the European Union is highly dependent on imports from outside the EU. Due to the growth in use of gas on several markets such as electricity generation, heat production and in households, the enormous increase in demand exceeds domestic production. When looking at the gross inland consumption of energy in the EU-27 in the period (see table 1.1), we can observe that natural gas recorded the highest absolute increase from 333Mtoe (Million Tons of Oil Equivalent) in 1995 up to 445Mtoe in Table 1.1 Gross Inland Consumption of 1995 & 2005 in Mtoe Change Total % Oil % Gas % Nuclear % Hard Coal % RES % Lignite % Note: RES represents the share of renewable energy sources in the overall consumption rate. Source: Eurostat (2007) Energy, Transport and Environment Indicators. Office for Official Publications of the European Communities: Luxembourg. The causes for the large increase in gas consumption differ per region. However, according to statistical information of the International Energy Agency (IEA), the growing importance of natural gas in the power generation sector represents the main cause for the overall increase in demand. Increased efficiency, lower greenhouse gas emissions, low capital costs and quick returns associated with combined cycle gas turbine (CCGT) power plants, make natural gas increasingly attractive compared to other fuels for power generation such as coal. In IEA projections of EU gas demand in 2004, the power generation sector represents the main driver for the projected increase in gas consumption over the next 20 years, as shown in figure 1.2. In the coming years, around 70 percent of the increase in demand will derive from the power generation sector.

36 36 Figure 1.2 Gas demand forecast in EU25. Source: International Energy Agency, World Energy Outlook In the European Union, the overall gas demand is expected to increase by 40 percent by In 2030 gas demand is expected to reach a level of 770 billion cubic metres, while European production is expected to stagnate in the coming years due to the levelling-off of maturing gas fields (see figure 1.3) As a result, the EU will increasingly have to rely on imported gas from suppliers outside the EU. The EU s gas import dependence on external supplier countries like Russia, Norway and Algeria is expected rise to 70 percent by 2030 (figure 1.4). Some EU member states are already completely dependent on imports for their gas supplies, other countries will see their import dependence rise up to 100% in the coming years.

37 37 Figure 1.3 Expected levels of natural gas consumption, production and imports in the EU in 2000 (in Mtoe). Source: European Commission: Green paper on energy security (2000) The overall dependency rate of the EU s reliance on imported energy products has been set at 52,3 percent in However, this rate is expected to increase to two thirds of the overall energy demand over the next 20 years. At the moment, the EU relies on imports for 82.2% of its total EU oil consumption and 58% of natural gas consumption (compared to 74.4% and 43.6% respectively in 1995) (Eurostat, 2007). Figure 1.4 Dependence according to energy product Source: European Commission: Green paper on energy security (2000) EU-30