First gas directive (98/30/EC)a
Second gas directive (2003/55/EC)b
Third gas directivec (2009/72/EC)
Third party access
Commercial principles (Article 3)
Regulated by a national authority (Article 14)
Not changed (Article 18)
Negotiated (Article 15) or Regulated access (Article 16)
Member states shall ensure the regulation (Article 18)
Regulated (Article 32)
Complete (ownership unbundling) (Article 21)
Separate in its legal form, organisation and decision-making (Article 10)
Ownership unbundling (Article 7) or independent system operator (Article 9)
Separated accounts (Article 13)
Specification about the separation in its legal form and a clause that ownership unbundling is not an obligation (Article 9)
Ownership unbundling (Article 9, 14, 26) or Independent system operator (Article 13,17)
General dispute resolution mechanisms (Article 25)
National regulatory authority (Article 22)
National regulatory authority and european level coordination through an agency (Article 24a-24f)
Competent authority (Article 21)
Competent bodies with the function of regulatory authorities (Article 25)
National regulatory authorities and european level coordination (Article 35-40)
Operators shall not be controlled by a person or persons from third countries (Article 7a)
Certification in relation to third countries (Article 11)
31 December 1992
100 % by 1 January 2005 (Article 18)
18 months (Article 2)
20 % by 2000
100 % by 1 July 2007 (Article 33)
By March 2011 Article 11 by March 2013
28 % by 2003
33 % by 2018
For a systematic comparison, we had to limit the scope of issues, which we have derived from the broader liberalisation instruments introduced by the Commission even before the negotiations over the first gas directive have started. The European Commission’s definition of liberalisation can be summarised as breaking up of national monopolies in energy production, transmission and distribution (European Commission 1988). The instruments of achieving this goal are generally associated with “sector restructuring, introduction of competition in wholesale generation and retail supply, effective regulation of transmission and distribution networks, establishing an independent regulator, and privatisation” (Jamasb and Pollitt 2005, p. 2). Based on this definition, we observe the evolution of the liberalisation efforts and thus of the dynamics of agenda-setting and policy making by looking at the specifics across five issues: (1) Third Party Access (TPA) to transmission and distribution networks, (2) Form of Unbundling in transmission and distribution, (3) Form of Dispute Resolution, (4) Deadlines for transposition and (5) Regulations for external participants. A systematic comparison of how these issues were presented in the initial legislative proposal and what was in the end adopted might serve as a rough indicator to detect successful agenda-setting. This can later be used as a basis to trace the actors who framed the successful issues and the strategies of these actors. A number of important observations arise from the comparison based in Table 10.1. Certainly, one of the key turning points in the liberalisation process was the first directive (as pointed by Herweg in Chap. 4). If we think of the directives in terms of the degree of change (Hall 1993), then the first directive was a “third order” change (the most significant change) because it marked an establishment of new policy goals. It thus conditioned the subsequent directives regarding the main instruments of liberalisation. However, it also displays a very light form of liberalisation with no significant pressure to break up the monopolies in the distribution and transmission. Table 10.1 also makes it clear that across the three directives, the first directive presents the biggest mismatch between the initial proposal and the final act. This of course is mainly the result of the directive being the first and that in subsequent negotiations the Commission has probably learned to exclude the issues that were not able to find broader political support. Additionally, the mismatch between the initial proposal and the final act makes the tracing of successful decision agenda much more difficult.
The second directive then represents a calibration of the existing policy instruments, thus the “first degree” change (the smallest change). The adjustment and stricter regulation of the third party access and the form of unbundling in transmission and distribution networks were necessary because of the growing discontent with the lack of reciprocity and the growing “merger culture” of French and German companies. Significant impulses for a new directive were provided by the pro-liberalisation coalition led by the UK and supported by the European Parliament (Brutschin 2013; see Ciambra and Solorio in Chap. 8). Stricter regulation instruments were not included due to prevailing strong opposition from Germany, in particularly regarding the regulatory oversight (Pariente-David and Swanson 2003) and from France because of strong domestic resistance to unbundling (Brutschin 2013).
The third directive could be interpreted as “second degree” change since it not only adjusted existing policy instruments but also added some new ones. The biggest prominence was the new regulation of access for the third countries or the so-called “(Anti-) Gazprom clause”, which protects the European gas market from third countries and requires consultations with the Commission regarding any acquisitions from third countries. Additionally, new approaches to regional cooperation and solidarity were included in the directive as well. The calibration of the existing policy instruments went beyond general expectations (Eikeland 2011). Not only was there finally a clause on ownership unbundling (even though it was relativized by permitting the so-called independent transmission operator and independent system operator) the third directive also significantly upgraded the powers of the national regulators. Given these significant changes and the inclusion of new issues in the third directive, which were negotiated after the eastern European enlargement in 2004 and in 2007, the third gas directive represents an interesting case study to trace the possible effects of the changing number of participants and the role of the eastern European countries.
10.1.2 Identifying Successful Decision Agenda
Table 10.1 can also be used to trace the successful decision agenda from the third gas directive: the issues that were included in the initial proposal as well as in the final legislative act. Generally, it seems that there was still significant calibration within the issues from the first and second gas directive. The third party access was further regulated even though it was not on the initial proposal of the Commission. The main bone of contention throughout the whole liberalisation process, the issue of transmission and distribution unbundling, was further negotiated and the initial proposal was significantly weakened. In turn, the powers of the national regulatory authorities and the coordination mechanisms were much more specific and upgraded in the final act as compared to the initial proposal.
The regulation of access to transmission as well as distribution networks from third countries was a completely new issue that could not be traced in the first or second directive. This clause clearly represents a case of successful decision agenda as it is included in the initial proposal and in the final directive. Additionally, this so-called “Gazprom” clause or “Anti-Gazprom clause”—an important mechanism to coordinate common energy policy—was apparently heavily lobbied for by Poland (Eikeland 2008). This leads again to the question whether the significant change observed during the third gas directive and the Gazprom clause can be explained through the changing number of the participants and the eastern European enlargement. If this is the case, then what are the exact mechanism behind it?
10.2 Theoretical Framework: Economic Interests or Framing?
Essentially, the main mechanism that will allow actors to push the policy into their preferred direction is the polysemic nature of liberalisation that allows many different interpretations. The concept of liberalisation offers multiple framing directions. Framing (definition of a problem) that will mobilise significant number of participants within a policy process is considered to be one of the main strategies for a successful agenda-setting (Princen 2011). However, in order to show that framing or a new definition of a problem was one of the main explanatory factors, it is necessary to show that a pure shift in the economic interests is not a better explanation.
Especially given the enlargement, one could argue that new member states have strengthened the pro-liberalisation coalition and this is why we could observe changes and new agendas. At the same time, if we follow the logic of the veto player theory (Tsebelis 2002), which deduces that an increasing number of veto players will increase policy stability by reducing the winset—the political space that includes points that can change status quo (SQ) and by increasing the core—the political space that includes points that cannot change the SQ, we would expect that an increasing number of actors might lead to a decreasing number of successful agendas. If it however can be shown that the winset was not significantly changed through the enlargements, then we can suspect that framing strategies deliver better explanations for observed change. In order to show whether the winset was changed, we need to find a way to assess member states interests. This is addressed in the following section.
10.3 Methods and Sources
In Chap. 8, Andrea Ciambra and Israel Solorio discuss the methodological concerns within agenda-setting studies that arise when scholars attempt to infer from language. This is especially problematic when we want to infer actors’ interests regarding certain issues or later when the framing strategies and their success are analysed. I thus suggest first utilising an alternative measurement of actors’ interests based on the so-called material interests (Brutschin 2013) and then to apply process-tracing techniques to analyse the framing strategies.
10.3.1 Positions Based on Material Interests
Scholars have utilised expert interviews (Benoit and Laver 2007), hand coding content analysis (Veen 2011), roll calls (Hagemann 2007) or opinion surveys (Hug and König 2002) to measure interests, preferences and positions of actors on certain issues. More recently, researchers started also to rely on quantitative text analyses and scaling of interest group positions from their position papers (Klüver 2009). Unfortunately, none of these techniques could be applied to gas market liberalisation because there are no systematic sources of information.
Nonetheless, general interests could also be inferred based on the so-called material interests. The main assumption of material interest based studies is that certain variables that are specific for each actor can significantly shape the actor’s positions on a certain issue. The interests are then inferred based on the assumption that actors will not be in favour of an issue that is too costly for them. This approach thus operates within the Rational Choice framework.
The applications of material variables can be found in research on environmental policies. So for instance, Sprinz and Vaahtoranta (1994) could show that preferences of states in international environmental negotiations are largely shaped by the countries’ ecological vulnerability and abatement costs. In a more recent study, looking at the US debates on climate change Fisher (2006) could trace that senators from coal dependent states were more likely to oppose the Climate Stewardship Act of 2003. Both examples suggest that certain material conditions can largely determine the starting negotiation positions or even explain why actors change their positions. Gas market, given its relative stable market structures, lends itself to this sort of analyses. Before we can proceed with the analysis, it is necessary to determine, which dimensions to include in the inference of material interests.
Starting from the basic assumption that material variables should reflect the costs of possible policy changes, the first dimension of a high importance is the share of gas as a source of energy for the economy of a country. This can serve as a proxy to assess the salience of the issue. It is plausible to assume that countries which heavily rely on natural gas as a source of energy will be much more interested in the developments of the gas markets (The Netherlands, UK, Austria) than the countries that barely include gas in their energy mix (Cyprus, Greece, Portugal). This, however, does not allow predictions regarding the possible position towards the direction of change. Here an additional dimension that can reflect additional costs of change is needed. Therefore, the market concentration could be taken as the main indicator to assess the possible costs of completely liberalising the market. By assuming that the further away from the complete liberalisation a member state is, the more costly it will be to change the market; it can be argued that member states that have a highly concentrated market will oppose liberalisation, while member states that have a liberalised market will support liberalisation to share the benefits.
Decisive for the assessment of possible material interests is the combination of the discussed dimensions. It can further be deduced that if the domestic market of a member state is not very concentrated and the consumption is very low, a member state will probably support the liberalisation but not to the same degree as states with the higher level of consumption. In a similar manner, the member states that have high concentration but low consumption should also oppose liberalisation but not to the same degree as the member states that have high levels of consumption. It was decided to use simple average to distinguish between low and high levels across both dimensions. Table 10.2 summarises the expectations based on the level of consumption and the level of concentration.
Deriving material interests
High concentration (above average)
Low concentration (below average)
High consumption (above average)
Low consumption (below average)
The data was then collected through Eurostat, the official source of the European statistics. The data for gas market concentration was not available and therefore the data for the concentration of the electricity market was used as it can be assumed that for a given member state, there will be minimal differences between the electricity and gas market concentration. We will look at the year 2007, the year when the proposal for the third gas directive was initiated.