Ole Gunnar Austvik:

Conflict and cooperation with respect to European natural gas regulations
 Applied Energy Elsevier Science, London, May-June 2003 Vol 75/1-2 pp 23-32 
ISSN: 0306-2619 

Download as pdf-file.

Download the article from publisher here.


Significant economies of scale and scope in the European gas industry make many transmission and local distribution companies natural monopolies in the markets in which they operate. Often, this gives them a strong market power and they experience little competitive pressure. Hence public interventions into the functioning of the market, as seen under the initiatives taken by the European Commission, such as the "Gas Directive" occur. This paper discusses a game between the public authority and the transporters, where various levels of conflict and cooperation will influence how far regulations will go and how they will be designed.

Author Keywords: European gas; EU; Natural monopolies; Regulation; Transmission


1. Introduction
2. Conflict with the regulator
3. Cooperation with the regulator
4. Pay-off-matrixes for transporter and the regulator
5. Efficiency improvements through partnership schemes
6. Conflict or cooperation?



Markets for natural gas transportation are generally highly concentrated with few actors involved. Transportation companies have a great potential for high profits as well as an important role for public services. Often, benefits of large-scale operations and scope economics in natural gas transportation effectively make barrier to entry for newcomers prohibitive in the European market. Generally, operations in the industry are either taken hand of by firms owned by, or they are private firms facing (potential for) strong regulations from, governments.

The argument behind various forms for public intervention in the operation of private natural monopoly transport utilities is that if they are allowed to behave as profit maximizers, without constraints, consumers and overall economic efficiency will suffer. Inefficient operation and possible opportunistic behavior among monopolistic firms, together with externalities in the use of gas as an important source of energy, the environment, concerns over economic activity, rent distribution, reduced dependency on Middle East oil and lack of information throughout the gas chain, have justified government intervention. It is, however, difficult to establish a fully competitive market in this sector. Rather a visible, regulatory hand, is the only option for policy makers to reach social goals. By intervening into firms’ functioning, governments wish to repair for market failures created by imperfect market structure and dominating enterprises.

The idea of regulating transporters’ terms of operations is that if the market itself does not produce optimal outcomes (when competition is introduced), then it can be mimicked to do so through regulatory and other public instruments. The social first-best solution would be a subsidized (publicly owned) enterprise that set tariffs according to marginal costs. This has been the tradition in many European countries in the aftermath of WW2. Even if nationalization and subsidization is the first-best solution in theory, second- or third-best outcomes may be its ultimate result. Due to lack of innovative pressure on and x-inefficiency in these companies, nationalized companies are today viewed as inferior to a system that regulates independent (privately owned) firms . When the European gas market becomes liberalized, part of the process in many countries is to (partially) privatize the transport utilities, and then start to regulate their terms of operation.

There are many techniques for interventions into the behavior of natural monopolies. Most well known is rate-of-return regulation, various forms of price discrimination, the use of subsides and multipart tariffs (Sharkey 1989, Berg & Tschirhart 1989, Train 1991, Tirole 1992, Laffont & Tirole 1993, Austvik 2000). Each of these techniques provides only second-best results from a social point of view. But they are better than leaving the firm unregulated. As long as regulators shall 'repair' misallocation of resources caused be imperfect markets, the system of regulated (private) enterprises may easily end up with situations that are either overdetermined or have too many degrees of freedom to yield the desired results.  Thus, in a market for a strategic and non-renewable commodity as for European natural gas, regulatory authorities will easily remain an arena of politically oriented interest groups, within and across borders.

Many EU member countries have already established regulatory procedures and authorities for their gas industries. In a few years we may even face a regulatory body on the European level.  With the help of a simple game set up, questions to be discussed in this paper are; Will gas transporters be better off by going into conflict with the regulator and try to halt or stop processes that aim at regulating terms for transportation? Or is it better to cooperate and try to “trap” the relevant authority in order to make him/her do the regulations in a way that they want? Is it better for society to find partnership schemes with the industry, or should it aim at maximum governmentally controlled arrangements?

Conflict With the Regulator

Generally, transmission companies and LDCs will receive lower margins when regulated as compared to an unregulated situation. The drop in profit will be distributed to producers, customers, and final consumers or to producing or consuming countries’ treasuries through taxation depending on how the system is liberalized (Austvik 1997). Even though transmission companies’ and LDCs' margins are rather stable both under the “old” system and in a liberalized one, their economic profit will be lost or, at least, reduced. Competition between transporters may to some degree be established, at least on some distances, which could make more variations in throughput, as well. Thus, from the outset, transport utilities have every reason to oppose almost any type of liberalization.

Let's first consider the interest of the regulator (for example represented by the EU Commission) in a liberalization process simplified to a desire to unconditionally take away the transporters’ economic profit and give it to consumers. The interest of the transporter is assumed unconditionally to maintain as much profit as possible. Thus, the interests of the regulator and the consumers are assumed identical and conflicting. Under the assumption set up, the game is not zero-sum for society, as regulation is assumed to yield a greater surplus for consumers than the loss incurred on transporters. This is the general redistributional and net effect when markets are liberalized.

This binary situation (the choice between regulation and no regulation) is illustrated in figure 1. Both the regulator and the (potentially) regulated can chose between favoring a process that introduce regulation and a process where no regulation takes place. The outcome for the transporter is depicted in the upper right corner in each cell, and the outcome for the regulator is depicted in the lower left. Best possible outcome for the regulator (consumers) is 4 and for the transporter is 3. The different values correspond to the assumption that the gain for society under regulation (liberalization) is greater than the gain for transporters when the market is left unregulated (not liberalized). Worst possible outcome for both is value 0 (zero). All utility is considered ordinal, which means that each party may rank the outcomes, but do not know how much better or worse it is compared to another outcome.

If the regulator does not regulate, consumers get no extra surplus, which represents their and the regulator’s worst possible outcome, equal to the value 0 (zero). At the same time, no regulatory initiative is the best possible outcome for the transporter, achieving maximum profit, with the value of 3, as depicted in cell I.  On the other extreme, if the market should be perfectly liberalized, and the transporter fully accepts the regulator’s terms for operations on a normal profit basis, consumers' surplus is maximized. This outcome would be the worst possible for transporters, value 0 (zero), but the best possible for the regulator, value 4. The outcome when both parties favor regulation is depicted in cell III.

If the transporter opposes regulation and the regulator nevertheless chose to regulate, the outcome for the regulator (and consumers) must be assumed to be less than if the transporters just accept new terms for operation. Now, transporters fight against intervention, making as much difficulties as possible for the regulator, and tries to postpone and destroy regulator's initiatives. In spite of this resistance, the regulatory efforts can be expected to yield a better outcome for consumers than no regulation at all, but less than if the transporter adheres. This outcome for the regulator is depicted with the value 2 in cell IV. At the same time, transporters will gain compared to a strategy just following regulator's desires, but less than if no regulation was introduced, depicted with the value 1. Cell II represents a situation where transporters want to be regulated and the regulator don't and are, under our assumptions, considered an impossible combination of strategies. The greatest social surplus is given in cell III.

 Even if the outcome for each depends on the choice of the other, both the transporter and the regulator have dominant strategies independent of the other’s choice. The transporter will gain 0 (nothing) if regulation is supported, and 3 or 1 if regulation is opposed. Thus, opposing regulation will be a dominant strategy for the transporter. The regulator will gain 0 (nothing) if it does not regulate and 2 if it does. Thus, favoring regulation will be a dominant strategy for the regulator. Outcome from cell I (0,3 ), or status quo, will result if the regulator does not have the ability to force regulation on transporters without their acceptance. Outcome from cell III (4,0) will result if it freely can do so. he relative political strength between the regulator and the transporter will be the main variable in determining what is possible, and the final outcome. In the uncertainty of this strength, this will be a situation of direct confrontation between the parties, and we will end up with the results from cell IV (2,1).

Cooperation with the Regulator

 Let's now assume that the transporter knows that it cannot prevent regulation to be introduced. Now, the option “not regulate” does actually not exist anymore. Then, the question arises for the transporter whether it is best served by continuing making a maximum amount of difficulties for the regulator or if it is better to make an interplay with the authorities in order to design a regulatory regime that is favorable. This is known as a principal/agent problem, in which the agent tries to take control of his/hers principal and trap him/her to act according to it's desires (Binmore 1992: 526-530).

 In this situation, when the transporter continues to resist and the regulator nevertheless intervene, the outcome is the same as in the previous game, as depicted in cell IV (2,1) in figure 2. The transporter knows that the best result he can expect by opposing a new system is of value 1 (cell IV), because the regulator certainly will now introduce regulation (cell I will not be possible). However, by participating in the regulatory process, in stead of only opposing it, the transporter might succeed in achieving a value at least as high as when opposing regulation, even though it will still be lower than if no regulation is introduced, set to value 2 in cell III. By doing this, the outcome for the regulator (consumers) may simultaneously be reduced to less than if the transporter only adheres to regulator initiatives set to value 1. At the same time, when transporters participate in the regulatory process, better solutions can be found than if the regulator shall figure out all details. Thus, outcome for consumers may not necessarily be reduced compared to cell IV, an we could end up with a value closer to for example 2.

In this situation, regulator's dominant strategy will still be to regulate, as regulation would yield a better outcome for consumers no matter what the transporter does (1 or perhaps 2). The transporter, however, will change strategy towards collaboration, because it knows that regulation cannot be avoided. By participating in the formulation of regulatory mechanisms the situation can be improved for the utilities (value 2 in cell III) compared to opposing it (value 1 in cell IV). If the transporter considers that regulator will not get such authority, or it can be prevented by some means, it will still choose to oppose any intervention, as shown in cell I in figure 1. Interestingly, when the regulator ultimately achieves a level of competence that makes it able to intervene efficiently, it could be better served with a transporter that would not interplay with its action (rather been potentially captured by transporter).

Pay-off-matrixes for Transporter and the Regulator

 Transporters may have diverging views on the possibility of introducing a strong (enough) regulatory authority in Europe. However, the greater the number of transporters that think the regulator (will) get such an authority, the more of these transporters will start to influence regulatory design and, accordingly, increasingly set the premises for each transporter resisting. Thus, in the beginning, transporters would form coalitions in order to prevent that "too many" utilities accept the introduction of a regulatory process. In this multifirm dilemma, there may be a critical mass of firms (weighed with their quantity transported, sunk capital, strategic significance, political influence etc) that are needed to prevent the establishing of a regulator and/or to limit its competence as much as possible.

 If we, for simplicity reasons consider transporters acting as one firm towards the regulatory authority, the game-theoretic results from this regulatory process can be illustrated in a “Schelling-diagram" (Schelling, 1978), as shown in figure 3. On the vertical axis to the left, the utility for the transporter, U(T), is measured (by its profit) while on the vertical axis to the right utility for the regulator, U(R), is measured (by consumers' surplus). The horizontal axis between the two vertical axes measures the "level of liberalization". To the left, at point A, no liberalization is introduced; to the right at point B, the market is completely and perfectly liberalized. This is an unmeasurable continuum, but can for example be thought of as the number of regulatory initiatives; the more liberalized, the more interventions by government must take place such as increased competition and introduction of increasingly more regulatory details (and a regulatory authority).

 Maximum utility for the transporter is achieved if no regulation is introduced, as illustrated in point C.  In this situation, minimum utility for the consumer is attained, as illustrated in point A. If regulation is established, and the transporter just follows passively regulator's initiatives, maximum utility for consumers is achieved, illustrated in point D.  In this situation, minimum utility for the transporter is achieved, as illustrated in point B. Thus, the utility possibility curve goes from C to B for the transporters and from A to D for the regulator, as the market is liberalized. The curves' down- and upward directions illustrate that more (and efficient) regulation takes increasingly more profit from the transporters, which increases net social surplus. Maximum regulatory utility (point D) is drawn as greater than the maximum utility for transporters (point C). Point D is higher on the right axis than point C is on the left axis, because the gain for consumers should be greater than the loss for producers under regulation, and in this way it will be a positive net social surplus.

The outcomes in figure 3 can be traced back to the games illustrated in figures 1 and 2.  In figure 1, point C (value 3 for transporter) and point A (value 0 for regulator) represents cell I, where no regulation takes place. Cell III is represented by point D (value 4 for regulator) and point B (value 0 for transporter). Cell IV yields outcomes somewhere between C and B for transporters (value 1) and A and D for consumers (value 2). By opposing regulation, the transporter may succeed in either preventing it from being established, or to maintain some of it's profit (limit the competence of the regulator). This will simultaneously reduce the effect for consumers and is illustrated by the vertical line aa. Thus, under our assumptions, the line aa represent the worst outcome for transporters (value 1) when conflict with the regulator is chosen, and the best possible outcome for consumers (value 2). If the transporter adheres to regulations, it will end up in point D.

If the transporter knows that regulation will be established, it may start to interact with the regulator to design the system in a best possible manner for itself, as discussed under figure 2. By doing so, transporter's utility will at least measure value 1. If it really succeeds in capturing the regulator, real profit may be increased almost back to a monopoly level (point C). The vertical line bb illustrates a situation where the transporter has managed to regain most of its profit, but not all, through this interplay. Transporter's outcome is somewhere between 1 and 3, or value 2, while regulator's outcome simultaneously is reduced from value 2 to 1.

Efficiency improvements through partnership schemes

The discussion above concentrated on power and counter-power forces between public authorities and the transport companies. If however the transporters could influence regulation in a way that also improves efficiency, and not only their own profit, as compared to a situation with no interplay with the regulator, there may be Pareto improvements in the process. This may happen because regulator's insight into the industry's complexity may be limited and partly be depending on transporter's information. Such examples can be found in the U.S. regulatory history, where regulator has made inadequate decisions for the industry with huge losses in efficiency and resulting stop-and-go-policies (Austvik, 2002). In this case, the pay-off matrixes may not necessarily contain straight (linear) utility possibility curves for either party. In figure 4, U(T) is dropping when some regulation is introduced. After a while, the transporter starts to interact with the regulator in the formulation of new interventions, and manages to maintain its profit without reducing the benefit for the regulator/consumers.

This is due to the fact that it can suggest arrangements that are more efficient than the regulator could do itself. Overall surplus in the market is increased compared to the more static first strategy. In fact, in this situation, an increasingly number of steps towards a more liberal market is taken to the net benefit of the society and only moderate loss to the transporter. At some level of liberalization, however, illustrated by the line cc, transporters may start to loose more substantially. At his point regulatory interventions are so comprehensive that transporter's utility curve drops more steeply and ultimately down to point B when the market is completely liberalized. The transporter would loose so much by passing cc, that it starts to oppose regulation again. In this situation, it is possible that the best point for the regulator could never be reached, because he lacks the ability to liberalize the market perfectly in an efficient manner and, thus, actually needs the collaboration from the transporter. By trying to move the transporter all the way to point B, the outcome for consumers may be worse than if stopped at cc. Thus, net social surplus may drop if more regulation beyond this point is attempted forced through.

Conflict or cooperation?

Of course, the two ways the utility curves are drawn are just examples on their many possible natures. They may be bowed in various ways or even be discrete. However, independent of the shape of the utility curves, we observe that transporters’ strategy depend heavily on whether a regulatory authority gets the power and ability to liberalize the market or not. The transporters would probably adopt a dual strategy opposing any initiatives taken by authorities on market intervention and simultaneously prepare for interplay in designing optimal regulatory regimes, if or when they come.

Transporters will be best served if they succeed in delaying or destroying political decisions giving such power to regulatory authorities, pointing out the complexity of regulations, security issues, risk or any other arguments that work. If possible, they will form coalitions, possibly even fusion wit each other. But when or if a decision about actual regulation is made, nevertheless, transporters should shift to a partly collaborative strategy. The regulator should, on its side, try to penetrate a possible collaboration between transporters by starting to design regulatory regimes with only one or a few of them. If a critical mass of transporters interact, the rest must follow, as well.

 In the dynamics of this decision making process, the strategies may shift from conflict to elements of cooperation, and back. When and how the parties should or would collaborate and when they confront each other, depends on the shape of the curves. The shape depends on market complexity, competence among each party, ability to intervene etc. If one accepts that it is difficult to reach a fully and perfectly liberalized market, one should rather discuss what would be the optimal degree and form for regulation, not only in the sense of economic efficiency, but also in terms of political feasibility (cc).

Austvik, Ole Gunnar, 1997: "Gas pricing in a Liberalized European market; Will the rent be taxed away?", EnergyPolicy vol 20/no.12 pp. 997-1012, London, Elsevier Science.

---, 2000: Economics of Natural Gas Transportation.  Research Report no. 53 Lillehammer College.

---, 2002: Norway as a Major Gas Exporter. Europaprogrammet
Berg S.V. & Tschirhart J., 1989: Natural monopoly regulation. Principles and practice, Cambridge University Press, Cambridge

Binmore, Kent, 1992: Fun and Games. A Text on Game Theory, Heath & Company

European Union (EU), 1998: The Single Market for Natural Gas, IGM Directive 98/30 (the ”Gas Directive").

---, 2001:  “The Internal market for Gas and Electricity: Completing the internal energy market”, Package
of Commission documents adopted on 13 March 2001.

Gerwig, Robert W., 1962: "Natural Gas Production: A Study on Costs of Regulation", Journal of Law and Economics, October, vol page 69-92

Laffont J.J. & Tirole J., 1993: A Theory of Incentives in Procurement and Regulation, MIT Press, Cambridge

Schelling, Thomas, 1978: Micromotives and Macrobehavior. W.W. Norton.

Sharkey, W.W., 1989: The Theory of Natural Monopoly, Cambridge University Press, Cambridge

Tirole, J., 1992: The Theory of Industrial Organization, MIT Press, Cambridge

Train, K., 1991: Optimal Regulation. The Economic Theory of Natural Monopoly, MIT Press, Cambridge