Network Industries, Third Party Access and Competition Law in the European Union
Network Industries, Third Party Access and Competition Law
Network Industries, Thir d Party Access and Competition Law in the European Union
0 This Article is brought to you for free and open access by Northwestern University School of Law Scholarly Commons. It has been accepted for inclusion in Northwestern Journal of International Law & Business by an authorized administrator of Northwestern University School of Law Scholarly Commons
Part of the Energy Law Commons; International Law Commons; and the Transportation Law
In the last couple of decades, most of the world's major economies
have begun a fundamental shift toward a greater market orientation.
Governments have privatized state-owned enterprises, reduced the extent
and scope of regulation and sought to remove barriers to competition.
These cjianges have transformed the workings of major industrial sectors
including telecommunications, energy and transportation, replacing
stateowned or highly regulated monopolies with private firms competing in a
liberal market environment.
This article addresses a set of issues that arise in the context of market
liberalization for a special and important class of industries, the so-called
"network industries," which include electricity, natural gas, rail
transportation and telecommunications. Each of these industries combines
activities that are potentially competitive, such as generation of electricity,
with ones that are naturally monopolistic, such as transmission of
electricity. This combination produces a unique set of challenges to
competition law and policy in designing a market structure and regulatory
framework which maximize the benefits of liberalization while effectively
controlling any tendencies to monopolistic abuse.
These issues are of special relevance in Europe, where efforts are
currently underway to liberalize the natural gas and electricity industries
through the implementation of two recent European Commission
* The Brattle Group, 8-12 Brook St, London WIY 2BY, United Kingdom. We thank
Antje Gunther of Mayer, Brown & Platt for help in understanding European competition
law, and Charles Brealey, Matthew Curtis and Megan Talbott of The Brattle Group for
excellent research assistance.
directives. The directives allow each country to choose between "regulated
access" and "negotiated access." Under regulated access, the government
sets the prices and terms by which potential new competitors can obtain
access to essential networks. Under negotiated access, competitors must
negotiate the terms of access directly with incumbents. Negotiation must
be non-discriminatory and undertaken in good faith, with potential disputes
subject to an independent settlement procedure.
We analyze "Chicago School" theories that would support negotiated
access as an efficient option. According to the Chicago School,
incumbents naturally have incentives to offer access on terms that will
promote efficient entry. The predictions of the Chicago School, however,
are directly contradicted by the United Kingdom's experience with
voluntary negotiations in the case of British Gas. Over an extended period
of years, the British government relied on voluntary negotiations to provide
entrants with access to the British Gas pipeline network. However, British
Gas engaged in a pattern of behavior including obfuscation and
discrimination that made entry infeasible.
After exploring the apparent contradictions between Chicago School
predictions and British Gas's behavior, we analyze the theoretical
weaknesses in the Chicago School approach and identify several reasons
why voluntary negotiations should not form the basis of government
policy. We review the limited experience with negotiated access to
electricity transmission in Germany, which confirms our conclusion that
negotiated access would deter the development of competition in the
European gas and electricity industries. We propose an alternative frame
for regulation based on vertical separation of the network and regulated
third party access with cost-based pricing. Incumbents may merit
compensation for past investments or for continuing obligations such as
universal service. However, any such compensation should be provided by
transparent and competitively neutral finding mechanisms.
II. NETWORK INDUSTRIES
A number of common features distinguish the network industries.
First, provision of the final product to the consumer relies upon a
"network" such as the transmission grid, railroad system or local telephony
network. Second, this network is a "natural monopoly." Competition in
provision of network services would entail grossly inefficient duplication
of facilities. Third, the final services provided by the industry are
potentially competitive. Transmission of electricity and gas and the
provision of railroad track are all natural monopolies, but that in no way
rules out competition in generating electricity, supplying gas or running
Finally, these industries share a common historical pattern. Until
recently they have typically been dominated by monopolies which were
either state-owned, as in many European countries, or heavily regulated, as
in the United States. Consequently, the current market structure typically
consists of large, vertically integrated incumbent local or national
monopolies, competing with firms seeking entry. In France for example,
the electricity industry consists of a single, vertically integrated company,
the state-owned Electricit6 de France. In Germany eight utilities, each with
a historical monopoly in its area of operation, own and operate the
transmission grid, and together generate about 80% of all electricity
The peculiar characteristics of network industries lead to a
fundamental tension in competition policy. The central aim of
liberalization is to promote competition, to discipline inefficient firms and
to ensure that consumers benefit from the potential gains of deregulation.
In these industries competition can only come from new entrants who must
rely upon the incumbent, their main competitor, for access to the network
and therefore to the market. Competition policy first acknowledges the
natural monopoly of an essential network facility. It then faces the task of
protecting both entrants and consumers from the potential harmful effects
of the monopoly. While doing so it must also avoid the danger of
excessively harsh regulation of the incumbent. Over-regulation may lead
to inefficiency and, at its most extreme, excessive regulation may
constitute an effective "taking" or expropriation of the incumbent's
previous investment in the network.
The policy issues can be summarized as a sequence of fundamental
decisions. First is the question of market structure: should ownership of
the network be separated from participation in the final market? For
example, in the United Kingdom the national transmission grid (in England
and Wales) belongs to the National Grid Company, a regulated company
which plays no part in the generation, supply or retailing of electricity.'
This kind of "vertical separation" is however relatively unusual, and the
issue currently receives little attention in European policy debates. It arises
most naturally where liberalization occurs simultaneously with
privatization, providing a natural occasion to create a vertically separated
If the network owner is allowed to remain vertically integrated, then
the second key decision relates to market conduct. Specifically, is there a
need to regulate the terms under which competitor firms may access the
network, or should it be left open to negotiation subject only to general
provisions of competition law? Most European countries have chosen
formal regulation of so-called "Third Party Access" (TPA) but there are
exceptions, of which perhaps the most noteworthy is the German electricity
industry. Finally, if regulation is chosen, what form should it take?
Specifically, should the price of access be set on some basis related only to
the direct cost to the incumbent of providing access? Alternatively, should
IExcept through its control of certain specialized assets that provide "ancillary services"
related to the maintenance of frequency, avoidance of "blackouts," etc.
it include a component to compensate for any profits that the incumbent
may lose through the provision of access? As we shall see, this has been
the subject of considerable controversy.
I. BACKGROUND IN EUROPEAN LAW AND POLICY
Competition cases affecting more than one member state of the
European Union (EU) are subject to European law, which in those cases
overrides any provisions in domestic law. EU competition law is based on
Articles 85 and 86 of the Treaty of Rome and subsequent decisions and
directives of the European Court and Commission. Article 85 of the Treaty
relates to collusive behavior such as price-fixing agreements. Article 86
prohibits "[a]ny abuse by one or more undertakings of a dominant position
within the common market.. .in so far as it may affect trade between
Article 86 has been interpreted by the European Court and
Commission to cover "essential facilities," guaranteeing
nondiscriminatory access to networks and other "bottleneck" facilities.'
According to the Treaty:
An undertaking which occupies a dominant position in the provision of
an essential facility and itself uses that facility (i.e. a facility or infrastructure,
without access to which competitors cannot provide services to their
customers), and which refuses other companies access to that facility without
objective justification or grants access to competitors only on terms less
favourable than those which it gives its own services, infringes Article 86 if
the other conditions of that Article are met. An undertaking in a dominant
position may not discriminate in favour of its own activities in a related
market. The owner of an essential facility which uses its power in one market
in order to protect or strengthen its position in another related market, in
particular, by refusing to grant access to a competitor, or by granting access on
less favourable terms than those of its own services, and thus imposing a
competitive disadvantage on its competitor, infringes Article 86.'
We discuss below whether a general "essential facilities" provision in
competition law is a suitable and sufficient instrument for the regulation of
In accordance with the principle of subsidiarity, cases which affect
only a single member state are subject to the provisions of domestic law.
Member states are, however, increasingly keen to ensure consistency
between domestic and European competition law. Germany's 1998 Act
Against Restraints on Competition and the United Kingdom's 1998
Competition Act are both explicitly designed to align domestic competition
law with European. Over time, domestic and European law can be expected
to form an increasingly seamless web, ensuring consistent treatment of
competition issues across Europe.
While the United Kingdom has more than a decade of experience with
privatized and liberalized energy markets, other European nations are only
now beginning down the same road. For many, the impetus toward
deregulation comes from the European Union. Competition law and policy
at the European level is in its infancy, but has already had powerful effects.
Most recently, the European Commission has promulgated two directives
aimed at liberalizing the key energy markets of gas and electricity. Both
directives have now received the approval of the Council and Parliament,
and passed into law.4 For many member states, implementation of these
directives will entail massive and wholesale restructuring of the industries
in question, leading to a new era in European energy.
The European Gas Directive entails the phased opening of European
gas markets to competition.5 Under the directive large customers may
choose between competing suppliers when purchasing natural gas, and the
network operator is obliged to facilitate such transactions by providing
network services in a non-discriminatory fashion. The directive defines
two methods of organizing access, referred to as "regulated" and
Article 15 of the Directive states:
1. In the case of negotiated access, Member States shall take the
necessary measures for natural gas undertakings and eligible customers either
inside or outside the territory covered by the interconnected system to be able
to negotiate access to the system so as to conclude supply contracts with each
other on the basis of voluntary commercial agreements. The parties shall be
obliged to negotiate access to the system in good faith.
2. The contracts for access to the system shall be negotiated with the
relevant natural gas undertakings. Member States shall require natural gas
undertakings to publish their main commercial conditions for the use of the
system within the first year following implementation of this Directive and on
an annual basis every year thereafter.6
Article 16 provides in part that:
Member States opting for a procedure of regulated access shall take the
necessary measures to give natural gas undertakings and eligible customers
either inside or outside the territory covered by the interconnected system a
right of access to the system, on the basis of published tariffs and/or other
terms and obligations for use of that system. This right of access for eligible
customers may be given by enabling them to enter into supply contracts with
4 1997 O.J. (L27) 40 (concerning common rules of the internal market in electricity);
1998 O.J. (L204) 41 (concerning common rules for the internal market in natural gas).
5 1998 O.J. (L204) 18. The directive provides for market opening in three stages (on
December 10, 2000, December 10, 2003, and December 10, 2018). At each stage, member
states are required to extend freedom of choice of supplier to a larger proportion of
customers. The initial proportion includes all gas-fired power generators and other final
customers consuming more than 25 million cubic meters ofgas per year, and must constitute
at least 20 percent of total annual consumption. At the two subsequent stages this
proportion must increase to first 28 percent and then 33 percent. Id.
6 1998 O.J. (L204) 15.
competing natural gas undertakings other than the owner and/or operator of
the system or a related undertaking.7
Member states are free to choose either or both procedures.8 They are
also required to "ensure that the parties negotiate access to the system in
good faith," and to "designate a competent authority... [to] settle disputes
concerning negotiations and refusal of access." 9
The Gas Directive entered into force in August 1998 and member
states have until August 2000 to implement its requirements. Pending its
implementation, negotiated access is the effective policy in most of
The European Electricity Directive similarly entails the phased
opening of European electricity markets to competition.'0 Large customers
may choose between competing generators when purchasing electricity,
and the transmission network operator is obliged to facilitate such
transactions by providing network services in a non-discriminatory
fashion.1 As in the case of natural gas, member states are free to choose
between negotiated and regulated access.' 2 Under negotiated access, the
directive requires that: "producers and... eligible customers... [are] able to
negotiate access to the system so as to conclude supply contracts with each
other on the basis of voluntary commercial agreements." 3
Regulated access gives "eligible customers a right of access, on the
basis of published tariffs for the use of transmission and distribution
systems, that is at least equivalent, in terms of access to14 the system, to the
other procedures for access referred to in this Chapter."'
The directive entered into force in February 1997 and member states
had until February 1999 to implement it.'" At present implementation
varies widely across Europe. Germany is the only state to have chosen
7Id. Art. 16.
"1d. Art. 14.
9 Id. Art. 21.
to As with natural gas, the electricity directive provides for market opening in three
stages (on February 19, 1999, February 19, 2000 and February 19, 2003). At each stage,
member states are required to extend freedom of choice of supplier to a larger proportion of
customers (the initial proportion corresponds to all customers with annual consumption
exceeding 40GWh, reducing to 20GWh and 9GWh at subsequent stages). 1997 O.J. (L27)
13Id. Art. 17.1.
14Id. Art. 17.4. The directive also allows for a "single buyer" procedure, which seems to
be very similar to regulated access. The main difference is that under the single buyer
system, supply contracts between customers and independent generators are combined with
a centralized electricity purchasing system through the use of "contracts for difference." Id.
1IId. Art. 27. Three states have been allowed to delay implementation. Belgium has
been allowed a delay of one year. Greece and Ireland have been allowed a delay of two
negotiated access as the basis for implementing the Electricity Directive.
Later in this article we discuss the implications of this choice for the
development of competitive electricity markets in Germany.
IV. THE ECONOMIC EFFICIENCY ARGUMENTS FOR NEGOTIATED
The most common concern surrounding negotiated access is that a
vertically integrated network owner may seek to deny its competitors
access to the network, thus "foreclosing" the market from competition.
Proponents of negotiated access argue to the contrary. Existing arguments
within antitrust theory, attributable to the "Chicago School" of economic
and legal theorists, are used in support of negotiated access. They imply
that concerns about market foreclosure are unfounded. Rather, negotiated
access will lead to an efficient outcome: more efficient firms will always
be able to enter the market through negotiation, while inefficient firms will
be driven out. All the potential economic gains from liberalization can be
achieved through negotiation, without imposing the burdens entailed in
In this section we introduce a simple example to analyze the prices
that we would expect to arise under negotiated access. We use the example
to illustrate the logic of its advocates' claims that voluntary negotiations
will automatically produce an access price that ensures efficiency in the
provision of competitive services.
Suppose that firm A is the incumbent electric utility, owning both
generation assets and a transmission network connecting those assets to
consumers. It costs firm A 4p/kWh to generate electricity, and an
additional lp/kWh to transmit. 6 The firm sells electricity at 7p/kWh,
giving it a profit of 2p/kWh. Suppose also that there is an entrant, firm B,
which owns its own power plant but has no transmission network of its
own, so relies upon access to firm A's transmission network.
Under negotiated access, firm A has a clear opportunity to foreclose
the market by refusing firm B access to its network. By doing so it will be
able to continue selling its own electricity, instead of being undercut by its
competitor. Suppose however that firm B's cost of generation is lower
than that of firm A, at say 3p/kWh. Then in this very simple example it is
clearly not in firm A's interest to refuse access, since it can make at least as
large a profit from transmitting firm B's electricity as from selling its own.
Firm A is currently making a profit of 2p from every kWh of electricity
consumed. Suppose that it allows firm B to make use of its transmission
network, at a charge of 3p/kWh. Since firm A's cost of transmission is just
lp/kWh, A will make the same 2p profit from every kWh sold by its
competitor. This transmission charge therefore leaves it indifferent as to
,6The abbreviation "p/kWh" stands for pence per kilowatt-hour. There are one hundred
pence in one British pound.
whether consumers buy electricity from itself or from its competitor, since
its profit is the same in either case. 7
Notice that such an arrangement would induce entry by firm B. Firm
B's total cost of supplying electricity will be 6p/kWh (3p for generation
plus another 3p in transmission charges), while it can sell electricity for up
to 7p/kWh, the current price being charged by A. Even after paying A's
access charges, therefore, it will still find it profitable to enter this market.
In fact B would be willing to enter at any access price up to 4p/kWh.
However, any price above 4p/kWh will effectively prevent entry, since it
will push B's total cost of supply above 7p/kWh, which is the highest price
it can hope to get for its electricity."
The effect of negotiated access in this example is therefore easy to
predict. Firm A must charge at least 3p/kWh for transmission to avoid a
reduction in profits as a consequence of providing access to firm B.
Furthermore, it can charge up to 4p/kWh for transmission and still induce
entry. Negotiations will therefore lead to an agreement giving firm B
access to the network at a price somewhere between 3p/kWh and 4p/kWh,
the exact figure depending on the bargaining skills of each side. At a price
between these two extremes, the effect will be to increase the profits of
firm A, and allow a positive profit to firm B.
A similar argument establishes the converse proposition: no inefficient
firm can survive under negotiated access. As we have seen, firm A will be
happy to provide access to its network provided it receives at least 3p/kWh
in transmission charges. Suppose now that firm B is in fact an inefficient
producer, that is, its generation costs are higher than those of firm A (say it
has a generation cost of 5p/kWh). Its total cost of supplying electricity to
the customer will now be at least 8p/kWh (5p for generation plus 3p for
transmission). Since the current price of electricity is only 7p/kWh, it is
impossible for firm B profitably to serve the market. Consequently, B will
not enter, or if it does it will operate at a loss and be forced to exit from the
" William J. Baumol, Some Subtle PricingIssues in ReinforcedRegulation, 10 INTL. J.
TRANSPORT ECON. 341 (1983). In this article, Baumol sets this price according to a
formulation he labels the "Efficient Component Pricing Rule" (ECPR). The ECPR specifies
that the access price should be set so as exactly to compensate the network owner for any
profit lost as a result of providing network access to a competitor. Almost everything we say
about negotiated access (both positive and negative) applies mutatis mutandis to the ECPR.
See also Robert D. Willig, The Theory of Network Access Pricing,in ISSUES INPUBLIC
UTILITY REGULATION (H. M. Trebing ed., 1979).
" It is convenient, and conventional in the economic literature, to assume that consumers
can be persuaded to switch suppliers (from A to B) if both firms charge the same price.
Similarly, we assume that B can be persuaded to enter if its expected profits are zero. These
assumptions have no substantive content however. We could easily suppose for example
that B must reduce its price to 6.9p/kWh to gain customers, and thereby enhance the realism
of our example at the cost of an increase in arithmetic complexity.
Negotiated access therefore appears to have the attractive property of
guaranteeing that the market will be served by the most efficient producer.
Far from foreclosing the market, the incumbent firm will positively
encourage new entrants provided they are more efficient than itself in
generation. Regulation of access is not needed to promote entry, or to
ensure that production is carried out by the most efficient firm.
This approach to the issue of market foreclosure in fact forms part of a
larger body of work in legal and economic scholarship concerning
"vertical" agreements, such as exclusive contracts between an auto
manufacturer and the dealers who sell the manufacturer's autos. 9 One of
the chief complaints made against vertical agreements is that they foreclose
competition. In the case of the auto manufacturer and the dealer, the claim
would be that an exclusive contract between the two will foreclose other
manufacturers from access to the dealer's retail facilities. Supporters of the
Chicago School however would argue that consumers will only suffer from
this arrangement if there are no other dealerships in reasonable proximity.
If there are, then other manufacturers can sell to consumers through the
other dealers and no problem arises. If not, then there is indeed a problem,
which is that the local market in car dealerships is insufficiently
competitive. However, this problem is a "horizontal one" that would exist
independently of the vertical arrangement with the manufacturer. It is not a
problem with the vertical arrangement per se. Generally speaking, the
Chicago School argues that vertical agreements exist to increase efficiency
rather than to diminish competition."
In the context of network industries, this would imply that the network
owner has no incentive to exclude competitor firms from the market: "a
monopolist has no inventive to gain a second monopoly that is vertically
related to the first, because there is no additional monopoly profit to be
taken."'" The whole monopoly profit can be earned by setting a high
enough access price, without interfering in the workings of the related
markets. If the incumbent is less efficient than a competitor in the
provision of the competitive service then it will negotiate a mutually
advantageous arrangement to have the competitor firm replace it in that
"9Other kinds of vertical contracting arrangements include such practices as "tying" and
Res2a0le Price Maintenance.
See, e.g., ROBERT BORK, THE ANTITRUST PARADOX: A POLICY AT WAR WITH ITSELF
225-245 (1993); RICHARD POSNER, ANTrrRUST LAW: AN ECONOMIC PERSPECTIVE 196-201
(1976). One well-known summary of this view is that "there is no such thing as a vertical
'problem'....The only possible adverse competitive consequences of vertical arrangements
inhere in their horizontal effects. Only where vertical arrangements facilitate restricted
output and raised prices-horizontal impacts-should they be inhibited." William Baxter,
then new assistant attorney general for the antitrust division of the U.S. Department of
Justice, quoted in JEANNE TIROLE, THE THEORY OF INDUSTRIAL ORGANIZATION 185 (1993).
2' BoRK, supra note 20, at 229. Bork notes that the reasoning relies on assumptions
("fixed proportions") which in this context would be equivalent to ignoring the possibility
of inefficient bypass of the network.
activity, not discriminate in favor of itself. Negotiated access therefore
creates no "new" anti-competitive issues. Any problems that arise are a
consequence of horizontal market structure, specifically the network
owner's monopoly status, and should be addressed directly rather than
through interference with the terms of access.22
V. BRITISH GAS: THE FAILURE OF NEGOTIATED ACCESS
An early European experiment with negotiated access can be found in
the United Kingdom's natural gas industry. The results of that experiment
provide unambiguous evidence for the weakness of negotiated access as a
means of facilitating entry and competition.
Before 1986, the natural gas industry in the United Kingdom consisted
essentially of a single, state-owned firm, the British Gas Corporation,
which held an effective monopoly over the sale of natural gas, and was sole
owner of the national transmission and distribution networks. 2 The Gas
Act of 1986 enabled privatization of British Gas. It deregulated supply in
the large users market,24 and established a new regulatory framework for
From our perspective, two aspects of the Gas Act are of special
interest. First, the legislation recognized the natural monopoly status of the
pipeline network, by prohibiting the construction of competing pipeline
networks.' Second, it sought to promote competition by a series of
provisions relating to TPA (referred to in the Act as "common carriage").
In this respect the Act was in fact an extension of already existing
Both the 1982 and 1986 Acts entailed relatively light-handed
regulation of TPA. Under the 1986 legislation, British Gas was required to
provide access to third parties wishing to supply gas, subject to conditions
of technical feasibility.27 The Act apparently envisaged that price and
non' It should be noted that in practice proponents of negotiated access typically represent
the incumbent firm, and consequently may be keener to emphasize the absence of vertical
problems than to point up the presence ofhorizontal ones.
I The industry was taken into state ownership in 1948, at which time it comprised over
1000 companies. For details of the history up to 1986 see MONOPOLIES AND MERGERS
COMMISSION, GAS - A REPORT ON THE MATTER OF THE EXISTENCE OF A MONOPOLY SITUATION
IN THE SUPPLY IN GREAT BRITAIN OF GAS THROUGH PIPES TO PERSONS OTHER THAN TARIFF
CUSTOMERS (1988) (hereinafter MMC).
24 Gas Act, 1986, ch. 44, § 8 (Eng.). The large users market is defined as those
customers purchasing over 25,000 therms per year.
' Id. § 7, ("Authorisation of public gas suppliers"). Subsection 9 of section 7 states in
part that "[a]n authorisation under this section shall not include in the designation any area
which is situated within 25 yards from a main of another gas supplier" (it goes on to give
certain rather special exceptions). Id. § 7-9.
26 Oil and Gas (Enterprise) Act, 1982, ch. 23 (Eng.). The Act contains provisions for
TPA, see id. §17, and makes a first step toward introducing competition in the large users
market, see id. § 12.
' Gas Act, supranote 20, §19(l)-(3).
price conditions for access in any particular case would be determined by
negotiation between British Gas and the firm seeking to supply gas.
However, in the event that the parties could not reach agreement on access
terms, the third party could ask the regulator to set terms, including
prices. 28 The Act provides guidance on the principles that should
determine access charges:
[T]he Director shall apply the principle that the public gas supplier
should be entitled to receive by way of charges for the conveyance of gas by
virtue of the right [of third parties to have gas conveyed by British Gas]-the
appropriate proportion of the costs incurred by the supplier in administering,
maintaining and operating his pipe-line system; and a return equal to the
appropriate proportion of the return received by the supplier (otherwise than
by virtue of the right) on the capital value of that system (including so much of
that return as is set aside to meet the need from time to time to renew that
This guidance was interpreted by British Gas as requiring prices
reflective of the average cost of transmission, including an appropriate
return on the capital embodied in the network." British Gas was further
required to publish information on common carriage, including examples
of the access prices it would charge.3
In 1988 the Monopolies and Mergers Commission investigated
allegations of market power abuse by British Gas. Their report provides a
detailed account of the failure of negotiated access. 3 2
British Gas had negotiated one common carriage agreement prior to
the 1982 Act. However, this agreement was of a rather special kind, since
it involved a gas producer that sold part of its output to British Gas and
retained some gas, carried by British Gas, for use in its own plant.33
Between 1982 and 1988, ten approaches were made to British Gas for
common carriage, five of them subsequent to the 1986 Act becoming law.
In 1988 none of these had led to common carriage, although the five
post1986 approaches were still under negotiation. 34 No access agreement was
actually signed until 1990.35
British Gas made use of a number of tactics in order to foreclose the
market. It provided only minimal information concerning access charges
and insisted on receiving commercially sensitive information concerning
the identities of the potential customer and supplier before it would
negotiate access. When it did negotiate, it appears to have done so in bad
29Id. § 19(4).
29Id. § 19(
30 See, e.g., MMC, supranote 23, 8.87 at 109.
1'Id. 3.37 at 24.
13Id. 3.34 at 24.
34 Id. 3.34-36 at 24.
35 M.A. ARMSTRONG, ET. AL., REGULATORY REFORM: ECONOMIC ANALYSIS AND BRITISH
EXPERIENCE 265 (1995).
faith. In addition, it used its monopsony position to attempt to deny
potential competitors access to sources of natural gas.
The general information provided by British Gas for firms wishing to
use its network was woefully inadequate. As a guide to what the charges
might actually be, it provided just two examples, for specified input and
destination points, load factors and quantities.36 However, British Gas
informed potential applicants that actual charges would "depend
principally upon the exact input point and destination of the gas[,]... the
load factor[,]. . .the quantity of gas conveyed and the duration of the
agreement" although additional charges might be involved to cover the
costs of any necessary extra expenses involved (including the cost of
The information content of these examples was close to nil. The
charges quoted did not include use of the local distribution system, which
would have been the largest component of the total charge.37 Nor did these
charges include any initial payment that British Gas might require.
Moreover, British Gas informed potential applicants that even in these
specific examples the prices quoted were not the prices that would actually
be charged. Instead the actual payments would take the form of an
undisclosed two-part tariff consisting of a capacity charge plus a use-based
charge, the latter with a minimum payment provision. Finally, the number
of examples was itself completely inadequate to provide any real guide to
the use of the pipeline system, since the United Kingdom has at least six
input points and dozens of possible destinations, with quantities and load
factors varying widely. It was impossible to extrapolate from the two
examples since British Gas provided no explanation of the underlying
methodology. When such an explanation was finally provided, the MMC
found that it contained significant biases:3 8
We have examined the method used by BG in calculating its two
published examples of common carriage charges. BG has employed a
different method of cost apportionment to that used for its own internal
purposes. It has also employed different principles of calculating costs, for
example the assumption of a 20-year asset life compared to up to 40 years in
its own accounts, the effect of which is to increase the calculated charge for
providing the common carriage system.
The MMC concluded that:
[w]e do not think that the information made available by BG provides
sufficient guidance to prospective users... We believe that BG's failure to
36 See MMC, supra note 23, Appendix 3.1 at 122. Appendix 3.1 of the Act reproduces
BRITISH GAS (BG), STATEMENT ON COMMON CARRIAGE (1988). For example, British Gas
"assum[ed] an average daily flow rate of 50mcfd.. .[g]as received by British Gas at
Bacton... and delivered to a point.. .near to Birmingham, at a load factor of 60
percent.. .4p/therm." Id.
17 id. 3.37 at 24. BG told the MMC that "use of the distribution system would add
about 7.5 pence per therm." Id.
3 'Id. 8.85 at 109.
provide adequate information is attributable to the monopoly situation, and
that this failure will make it difficult for third-party suppliers to estimate
transmission costs and negotiate contracts with customers. 9
more transparent and detailed information concerningMaMccCessthcahtarigtesp:r40ovide
British Gas resisted any suggestion from the
BG also argued that a [general] tariff [for common carriage] would be
inconsistent with the framework for common carriage set out in section 19 of
the Gas Act. In BG's view this section of the Gas Act presupposed the
possibility of negotiation between BG and third parties.. .a published tariff
would preclude this approach.
British Gas also claimed that it would be impractical to construct a
general tariff owing to the complexity of the United Kingdom network.4'
The MMC received testimony from companies that had expressed an
interest to British Gas in obtaining common carriage. A number of
sensitive information as a precondition for negotiatdioemn:a42nded commercially
respondents complained that British Gas had
A property company told us that BG adopted a negative attitude to the
use of its pipeline distribution system by third parties.. .BG had stated that a
prerequisite for such an arrangement was in effect the disclosure to BG of the
prospective alternative supplies. . . The company said that under these
circumstances new suppliers of gas.. .were unwilling to be identified.
In fact British Gas adopted a standard policy of requiring that both the
potential supplier and customer be identified. This would of course enable
them to approach the potential customer and undercut whatever price was
being offered by their competitor. They could also punish the supplier in
not aTlwhaeyres wbeaesnailnsogosoomdefaeitvhi:d43ence that when negotiation took place it had
MEUC [Major Energy Users' Council] said that the gas producers it had
approached had clearly been reluctant to be seen discussing direct supply with
industrialists, but MEUC had since met BG on behalf of a consortium to
discuss the transmission of gas through BG's pipeline system. MEUC told the
Commission [the MMC] that [it] had received a letter from BG quoting
between 13 and 20 pence per therm to carry third party gas. MEUC said that
its members regarded these terms as 'ludicrously high', and considered that
BG was clearly determined to inhibit competition in gas supply from
developing at all costs.
Finally, there was evidence that British Gas also used its dominance as
a purchaser of natural gas to restrict competition in supply:
AGAS [Associated Gas Supplies Ltd]... complained that all producers
approached by AGAS had stated that BG's purchasing policy was to demand
39 Id. 8.83 at 108.
exclusive rights to 100 percent of the gas reserves in any one field for the life
of that field. 'As a result of this practice, all current production is fully
committed to BG under contract and none is available for purchase by third
parties. In short, there is no free market in existing gas supplies.' 44
Not surprisingly, the MMC concluded that the operation of the natural
gas market in the United Kingdom was unsatisfactory in this regard, and
that British Gas was indeed seeking to exclude potential competitors from
the market for natural gas supply:
.. BG's failure to provide adequate information on the costs of common
carriage, its ability to use information obtained when negotiating common
carriage terms to identify potential customers of competing suppliers and the
potential source of gas, and its position as a dominant purchaser of gas, may all
be expe4c5ted to act against the public interest by deterring new entry into the
The United Kingdom has since retreated from its reliance on
negotiated access in the natural gas industry. In 1992, under threat of
proceedings before the MMC, British Gas agreed to targeted reductions in
its market share.46 In 1992, legislation was passed allowing for separate
regulation of the pipeline network.47 Whereas only the retail price of gas
was previously regulated, a separate set of regulations was introduced to
cover the prices for using the pipeline network. Shortly thereafter, British
Gas was forced to negotiate a comprehensive code concerning the terms of
access for all market participants. The negotiations were conducted under
the supervision of government officials, and the resulting Network Code
was approved in 1994. Another important step towards ensuring
nondiscriminatory 'access took place in 1997, when British Gas voluntarily
divested its trading operations by creating a new company, Centrica,which
was floated on the stock market as a separate entity. British Gas still
operates the pipeline business but no longer sells gas to retail customers,48
and owns no shares in Centrica. The United Kingdom government has
effectively abandoned negotiated access. Non-discriminatory access to the
British Gas network has been produced by a combination of regulated
access and vertical separation.
44Id. 6.88 at 62.
45Id. 1.4 at 1.
46See George Yarrow, Progress in Gas Competition, in REGULATING UTILITIES:
UND4E7RISdT.AN2D.3INaGt 7T0H.E ISSUES, 65, 68-72 (1998).
41Some British readers may be confused by the continued visibility of the British Gas
name in advertisements to retail customers. However, these advertisements simply reflect
Centrica's use of the British Gas name and logo in Great Britain. The associated rights were
transferred to Centrica upon its creation.
VI. NEGOTIATED ACCESS CANNOT BE RELIED UPON TO PROMOTE
The history of British Gas contradicts the claim that negotiated access
ensures efficient entry. In practice, the incumbent network owner may use
discriminatory provision of access to deny its competitors access to the
market. In this section we explain why, contrary to the arguments put
forward by Chicago School theorists, a vertically integrated incumbent
may seek to foreclose the market to its competitors. We also argue that
even if negotiated access actually promoted efficient entry, many other
regulatory schemes would have the same property. This is of particular
importance since negotiated access fails to address the central issue of
monopoly profits accruing to the network owner.
Recall the basic Chicago School argument that "monopoly profits can
only be earned once." The incumbent can set an access price that extracts
all profit from the market, and therefore, it has no need to foreclose or
otherwise discriminate against competitors. The elegance and simplicity of
this prescription to the would-be monopolist suggests one possible flaw in
its practical application. Using the access price in the manner suggested
would be a highly transparent way of extracting monopoly rents. To do so
would undoubtedly draw the attention of regulators, customers and
politicians to the existence of monopoly power and its potential and actual
abuse, thus inviting further regulation. It was safer for British Gas, because
it was more discreet to deny access on a case-by-case basis, thus engaging
in bad faith negotiation and relying on technical obfuscation rather than
publishing prices so high as to retain the full monopoly rent. In sum,
considerations of political economy contradict the simple economic
arguments we presented above.
We should also note a number of more technical objections to the
Chicago School arguments. Economists who have formalized these
arguments have observed that they rely on a number of implicit
The first of these points is correct in a famous special case. Suppose that
the two inputs are combined in fixed coefficients, that there are no economies
of scope between the vertically related activities, that there are no vertical
externalities, and that there are constant returns and perfect competition
between symmetric firms in the competitive sector (i.e., "generation" in our
example). Then it is indeed true that M [the network owner] can extract all the
monopoly profit that there is to be had while remaining in the transmission
activity alone. It will be apparent, however, that this is a very special case, and
the argument that there is no market power reason for vertical integration and
foreclosure may not be robust.49
Similarly, we note briefly that managerial risk-aversion and a taste for
empire-building would also tend to encourage foreclosure. Foreclosure
may be of significant benefit to the incumbent firm's management, even if
it is detrimental to the shareholders. By retaining an overall monopoly
position, managers avoid a number of risks. First, if they negotiate an
access contract with the aim of extracting full monopoly profits, they may
miscalculate or be outwitted in negotiation. Second, once other firms enter
the competitive activity, the entrants' managers may reveal themselves to
be more capable or efficient than the incumbent's. At its most extreme this
might even lead to the threat of a hostile takeover. Finally, by preventing
entry, managers can hope to ensure that they will retain the perquisites and
prestige traditionally associated with large national monopolies, especially
In some cases, negotiated access can create barriers to entry even if the
incumbent refrains from discriminatory behavior, as a consequence of the
well-known "hold-up" problem. 0 Returning to our example, suppose as
before that firms A and B have generation costs of 4p/kWh and 3p/kWh
respectively. For simplicity's sake, suppose that the total available market
is one billion kWh and that given access, firm B can capture the whole of
this market from firm A without having to lower its retail price." The
potential increase in industry profit, defined as the sum of the two firms'
profits, from using firm B's generation assets in place of firm A's is
therefore £10 million." When the two firms negotiate over the access
charge, they are effectively negotiating over the division of this £10
million. If the charge is set equal to 3p/kWh then firm A will make the
same profit as before, and firm B will get the £10 million. Conversely, if
the access charge is set at 4p/kWh, then firm A will increase its profit by
£10 million, while firm B will make zero profit. If we assume that each
firm has equal "bargaining power," then the access price arrived at will be
2.5p/kWh, and firm B will make £5 million.
The hold-up problem arises as follows: we add one more detail to the
hypothetical scenario above, assuming that before negotiations began firm
B had made an (irreversible) investment of £6 million to establish itself in
the market. For example, it might have spent that sum on building its
power station in the area covered by firm A's network.53 Then, overall it
will have lost money, since the £5 million profit will not compensate for
the £6 million investment. However, by the time negotiation begins the
investment is a "sunk cost" that adds little or nothing to firm B's
bargaining power. Although there might be an ethical case for firm A to
lower the access price in recognition of firm B's investment, that case has
no legal or practical force.
'0See, e.g., OLIVER E. WILLIAMSON, THE ECONOMIC INSTITUTIONS OF CAPITALIsM (1995).
S We ignore the time dimension to avoid tiresome calculations. On the assumption that
no price reduction is needed to induce a switch, see supranote 19.
2A saving, and hence an increase in profit, of lp/kWh from switching to firm B,
multiplied by one billion kWh.
11 For simplicity's sake, we assume that the power station will have no alternative use if
firm B cannot access firm A's network.
Unless some way around this problem can be found, firm B will never
build its power station. This will clearly be inefficient, since the £6 million
investment produces a benefit of £10 million. In this case a simple solution
presents itself: before the power station has been built the two firms should
sign a long-term contract, fixing the access price at a rate which makes the
investment an attractive one. Notice that before the investment is made,
firm B can rationally refuse to accept an access price that does not produce
a profit on the investment. In more complex situations, however, it may
not be possible to write such a long-term contract. For example, the
irreversible investment may be in developing a new technology. By its very
nature, the technology may be difficult to describe or define until it has
been developed. Alternatively, it may be that by entering into negotiations
firm B will be forced to reveal to firm A some commercially sensitive
information which will enable firm A to reduce its own costs to 3p/kWh
without making any deal.54 Whenever there are obstacles to the writing of
long-term contracts, the danger of this kind of hold-up will occur.
Consequently, negotiated access can deter firms with potential innovations
from ever entering the market, even though both entrant and incumbent
Finally, we turn to the issue of preventing inefficient entry. It is
plausible that negotiated access will have the desirable effect of preventing
inefficient firms from entering the market. However, it is also clear that
any non-discriminatory regulatory regime will have the same effect.
Natural competitive forces mean that inefficient firms can survive only
under a scheme which artificially subsidizes them in some way." In the
absence of such subsidies, negotiated access is not needed to deter
VII. NEGOTIATED ACCESS FAILS TO ADDRESS THE MONOPOLY
POWER OF THE NETWORK OWNER
Whatever its flaws, the "Chicago School" approach does perform a
valuable task in focusing attention on the fundamental problem of network
regulation: the monopoly power enjoyed by the network owner. In the
Chicago terminology, this is a "horizontal" rather -than a "vertical"
problem. However, since the network is a natural monopoly, the usual
kind of "horizontal" solutions such as forced demerger or the removal of
barriers to entry are by definition inappropriate. The Chicago approach
therefore begs the question of how the network monopoly problem is to be
resolved. In our example of the electricity market, although negotiated
access led to efficiency in production, it did little or nothing to reduce the
final price of electricity. From a purely economic point of view, the
efficient outcome would be a price equal to the (marginal) cost of
producing and transmitting electricity in the most efficient manner
possible. If firm B has generating costs of 3p/kWh, then this price would
be 4p/kWh 6 However, firm A uses its control of the network to extract
monopoly profits from the market, leading to a much higher price than is
Proponents of negotiated access and the ECPR do not claim that either
approach can solve the problem of monopoly power:
A key problem is that the bottleneck [i.e., the network] owner is a
monopolist, albeit a regulated one, and its final product price may therefore be
set at a level that yields monopoly profits. Such monopoly profits may be
among the profits foregone as a result of a lost sale of final product and,
consequently, constitute a part of opportunity cost, [and therefore the ECPR
price]... We have consequently always maintained that efficiency requires
both ECPR and some arrangement that prevents overpricing of both final
product and bottleneck input[.]57
In this section we discuss possible solutions to this problem and
conclude that there is no effective solution that is compatible with
negotiated access. Specifically, we first ask whether general provisions of
competition law might be sufficient to prevent monopolistic abuse by the
network owner, without the need for more burdensome forms of regulation.
Unfortunately, this appears not to be so. We then examine the only other
approach compatible with negotiated access, that of regulating the final
price charged to consumers.58 We argue that such an approach is
incompatible with some of the fundamental aims of liberalization.
Attempts to use general provisions of competition law against abuse of
the network monopoly have already failed in the well-known case of
Telecom Corporation of New Zealand, Ltd. v. Clear Communications,
Ltd. 9 This case took three years to adjudicate and involved several million
"6Efficiency requires that production take place at the least possible cost, hence that firm
B becomes the producer. This gives a production cost of 3p/kWh, to which is added the
transmission cost of lp/kWh to give the total of 4p/kWh.
" William J. Baumol, Janusz A. Ordover & Robert D. Willig, Parity Pricingand Its
Critics: A Necessary Conditionfor Efficiency in the Provision of Bottleneck Services to
Competitors, 14 YALE J. ON REG. 145, 150-51 (1997).
s We ignore a third "solution," which would be to accept the monopoly as a necessary
evil and take no steps to mitigate it, on the grounds that the costs imposed by regulation
would outweigh its benefits (in the economists' jargon, that market failure is less costly than
government failure). In practice, regulation typically does impose substantial costs, both
directly (the existence of a regulatory agency, and the demands it makes upon the time and
administrative resources of the regulated firm) and indirectly (through poor regulatory
decisions). However, in the European context a policy of laissez-faire towards the industries
in question is ruled out by the requirements of EU Competition Law and is therefore merely
of theoretical interest.
" Telecom Corp. of N.Z. Ltd. v. Clear Communications Ltd., [1992) 5 T.C.L.R. 166,
rev'd  T.C.L.R. 138 (P.C), rev'd 4 N.Z.B.L.C. 340 (C.A.), rev'd  1 N.Z.L.R.
dollars in litigation fees.6" It was characterized by disagreement over the
proper scope of the laws in question, and complex economic arguments
which at times clearly confused the courts. 6' Even a decision by the Privy
Council failed to settle the issue: Clear Communications continued to
lobby the New Zealand government for intervention until reaching a
settlement with Telecom in March 1996. Since then a new round of related
litigation has begun.62 The costs in time and resources, and the courts'
difficulties in evaluating economic arguments, illustrate the drawbacks of
competition law as a regulatory instrument. Moreover, this is by no means
an extreme example. The economic arguments in the case remained at the
general and theoretical level, thus avoiding the immense added
complications that would arise if a court actually engaged in setting prices.
The alternative instrument usually proposed by Chicago-style theorists
as a means of preventing the network owner from obtaining monopoly
rents is to regulate the final price of the service in question. In our
example, this would involve a cap of 5p/kWh on the price charged to
consumers for electricity. Firm A would earn zero profit (i.e., no
monopoly rent -- recall that "costs" already include a reasonable return on
capital) since 5p/kWh is by design exactly equal to its average costs. If
negotiation allows firm B to serve consumers, also at 5p/kWh, then this
will create a surplus of 1p/kWh. Of course the exact division of this
surplus between the two firms will depend on the details of the access
contract negotiated. However, the surplus is a return on firm B's
technological superiority, which is generally regarded as a legitimate
source of supernormal profit, in contrast to any surplus that would be
earned by firm A (absent the price cap).6"
Unfortunately, this idea contains an irremediable flaw. Recall that one
of the characteristics of network industries is that provision of the final
service is potentially competitive, with only the network itself having the
characteristics of a natural monopoly. Indeed perhaps the key purpose of
liberalization is to enable competition to occur in activities such as
electricity generation. The suggestion that final prices be regulated is
therefore highly paradoxical. It would entail regulation of activities which
are inherently competitive and hence have no need for regulation,
combined with a laissez-faire approach to a natural monopoly. Such an
approach would remove many of the major benefits of competition, for a
purpose that can better be achieved simply by directly regulating the
pricing of access to the monopoly facilities.
VIII. THE TAKINGS ARGUMENT FOR NEGOTIATED ACCESS
In this section, we examine an entirely separate argument for the use
of negotiated access. It involves the assertion that prior to deregulation
there existed an implicit "regulatory contract" between the government and
the incumbent. Under this contract the incumbent made investments, at the
behest of government, which would not have been economic in a
deregulated environment. Examples of these so-called "stranded costs"
include investments in nuclear power, and the construction of infrastructure
to serve remote areas. According to the theory, regulated prices contained
an element of compensation for these investments, as the government's
side of the regulatory contract.'
Deregulation represents a potential breach of the regulatory contract
and the government should compensate the incumbent for any damages
incurred. This obligation is, of course, in line with the prohibitions found
in many legal systems against uncompensated expropriation of private
property by the government." The advantage of negotiated access is that it
enables the incumbent to carry on earning the same profits as it would have
absent the breach, and thus indemnifies the incumbent against its effects. It
can also be used to compensate for the cost of ongoing "incumbent
burdens" such as a "Universal Service Obligation" (USO), for which
similar arguments can be made.
This argument has recently been developed in great detail by two
scholars, Gregory Sidak and Daniel Spulber.66 Their discussion takes place
mostly in the context of the U.S. telecommunications industry and draws
heavily on U.S. legal history and scholarship. However, the underlying
I Proponents of this theory view monopoly rents as a gratuity rather than part of a valid
regulatory contract and therefore do not propose compensation for the potential loss of
monopoly rents by incumbents. For example: "The regulatory contract is a bargained-for
exchange between the state and individual firms that is intended to benefit consumers. That
relationship between the private firm and state differs fundamentally from the relationship
that frequently exists when.. .the state or federal government confers a statutory gratuity on
a firm[.]" J. GREGORY SIDAK & DANIEL F. SPULBER, DEREGULATORY TAKINGS AND THE
REGULATORY CONTRACT: THE COMPETITIVE TRANSFORMATION OF NETvORK INDUSTRIES IN
THE UNITED STATES 455-56 (1998) (hereinafter Sidak).
65This legal prohibition can be justified on economic grounds. The requirement to pay
compensation for breach of contract, whether applied to private parties or governmental
bodies, has the same efficiency properties. Efficiency means that the gains to the party that
breaches outweigh the costs to the injured party, which equal the compensation that has to
be p6a6id. Hence breach is profitable ifand only if it is efficient.
See Sidak, supranote 64.
logic could apply to any of the network industries.67 As Sidak and Spulber
point out," the prohibition on confiscation of property embodied in the
Takings Clause of the Constitution69 is not unique to the United States. It
can be found in other English-speaking nations and is enshrined in
European treaties including the European Convention on Human Rights:
"Every natural or legal person is entitled to the peaceful enjoyment of his
possessions. No one shall be deprived of his possessions except in the
public interest and subject to the conditions provided for by law and by the
general provisions of international law."70
However, while we approve of the goal, we are skeptical that
negotiated access is an effective or appropriate means of- providing
compensation for incumbent burdens. First, we suspect that the
compensation provided prior to deregulation was of a highly inexact kind.
We doubt that regulators prior to deregulation set prices that exactly
compensated the regulated firm for its uneconomic investments.
Consequently, even if the profits earned under a system of negotiated
access were close in magnitude to those earned prior to deregulation, they
are unlikely to provide the correct level of compensation.
Second, we note that deregulation is likely to produce radical changes
in the whole market environment, and therefore have a very significant
effect on the level of profits enjoyed by the incumbent. Under negotiated
access they are likely to be much higher than the profits earned under
regulation, since negotiated access allows the incumbent to earn full
monopoly profits. Consequently, there is even less reason to think that the
profits earned by a deregulated incumbent subject to negotiated access will
bear any relation to the cost of uneconomic investments. Negotiated access
is an illogical and opaque mechanism of compensation for those costs.
IX. AN ALTERNATIVE APPROACH: VERTICAL SEPARATION,
COSTBASED PRICING AND COMPETITIVELY NEUTRAL STRANDED-COST
We are now in a position to enumerate the three fundamental goals for
competition law and regulatory policy toward network industries: first, to
ensure that competition occurs in the potentially competitive sector of the
industry; second, to prevent the network owner from earning monopoly
profits from the network; and third, to compensate properly the network
owner for stranded costs and ongoing regulatory burdens. We believe that
it is possible to achieve all three through well-designed regulation.
67 The argument as a whole is however, clearly not relevant to the case where
deregulation occurs simultaneously with privatization.
68 See Sidak, supra note 64, at 3.
69 "[N]or shall private property be taken for public use, without just compensation." U.S.
CON70ST. amend. V.
See Article 1of Protocol No. I, European Convention on Human Rights.
The correct regulatory approach can itself be divided into three parts,
corresponding to the goals listed above. Vertical separation of the network
will prevent the incumbent from stifling competition in the potentially
competitive sector. Clearly this advantage would come at a high cost if
there are significant economies of scope between the provision of network
and complementary services, but we see little evidence for that claim.
Whether or not vertical separation occurs, access prices should be regulated
directly through "cost-based pricing" to remove the element of monopoly
profits. Finally, explicit funding of stranded costs and ongoing burdens
through dedicated user fees or surcharges will provide proper
compensation in an efficient, non-discriminatory and transparent manner.
Taken together these elements provide the best possible basis for effective
regulation ofnetwork industries.
We have already discussed the problems inherent in allowing the
network owner to compete in downstream markets. We believe that the
problems in the United Kingdom natural gas industry were to some extent
unavoidable while British Gas remained vertically integrated." It is
probably not possible for the regulator to prevent a vertically integrated
incumbent from discriminating against its competitors. Moreover, if it is
possible, it requires such frequent intervention in the company's affairs, on
so burdensome a scale, that the company itself might prefer vertical
separation. As noted above, this was the path eventually chosen by British
In some countries, full vertical separation is politically infeasible. If
so, we recommend an alternative involving an "independent system
operator" or ISO for the network. The concept of the ISO has developed in
the United States to ensure non-discriminatory access to the transmission
networks owned by vertically integrated incumbents.72 The incumbent
continues to own the transmission network, but relinquishes management
of network operations and the provision of transmission services to a
neutral third party, the ISO. The ISO has no incentive to discriminate in
providing access to competitors.
The second lesson from British Gas is that the price and non-price
terms of access must be formally regulated. The rules should be
nondiscriminatory, which will be of particular concern in the absence of
vertical separation. Prices should be regulated via some form of cost-based
pricing. By cost-based pricing we mean any approach that aims to ensure
that the network owner's total access revenues are no higher than what is
"! The MMC came to the same conclusion in 1993. See MMC, British Gas plc, Aug.
1993, Cnnd. 2315-2317, at 32.
7 For example, see Order No. 888, Final Rule, "Promoting Wholesale Competition
Through Open Access Non-discriminatory Transmission Services by Public Utilities," 75
FERC 61,080 (Apr. 24, 1996); Order No. 592, "Merger Policy Statement," FERC 31,049
(Dec. 18, 1996); Notice of Proposed Rulemaking, "Regional Transmission Organizations,"
87 FERC 61,173 (May 13, 1999).
needed to cover costs. Here the notion of costs naturally includes an
appropriate return on the capital employed.73 Many possible pricing
schemes satisfy this requirement. Choice among them is complex and will
depend on the particular industry and the availability of information.74
Even without vertical separation, there is every reason to choose cost-based
The recovery of stranded costs and compensation for ongoing
regulatory burdens are, from an economic point of view, essentially
questions about optimal taxation. In principle, such compensation could be
provided from general tax revenues, which might well be the most efficient
approach theoretically." In practice, the government usually requires that
the necessary revenues come directly from the industry in question. In that
case, it should be raised in a "competitively neutral" fashion, minimizing
the associated distortions. This is best achieved by applying
nonbypassable fees to all customers who use the network. Such an approach
has the added advantage of transparency, in contrast to the negotiated
access approach where consumers have no direct way of seeing how much
of the cost of service they pay is applied to stranded cost recovery.
X. IMPLICATIONS FOR THE EUROPEAN GAS AND ELECTRICITY
In most of Europe, no attempt at vertical separation has been made
during the implementation of the gas and electricity directives. Several
countries have chosen regulated access, but Germany has chosen
negotiated access, and negotiated access remains the effective regime
pending full implementation of the directives throughout Europe.
Consequently, in the next few years we anticipate continued competitive
problems concerning access to natural gas pipelines and electricity
Our concerns find confirmation in the experience to date with
negotiated access in Germany, where claims of undue denial of
7 Our definition includes UK-type "RPI-X" schemes, under which the regulated firm can
increase its return on capital by achieving greater than expected improvements in efficiency.
74 Here we avoid discussions of non-linear pricing, the multi-product case, "Ramsey
pricing" and so forth. These technical topics are covered exhaustively in the economics
literature: see J.J. Laffont & J. Tirole, Creating Competition Through Interconnection:
Theory and Practice,10 JOURNAL OF REGULATORY ECONOMICS 227 (1996). Note that their
use of the term "cost-based" differs from ours. An approach such as Ramsey pricing, where
the access charge depends on demand elasticities as well as cost factors, is "cost-based"
according to our definition, but not according to theirs.
71 In the United Kingdom, the central government pays direct subsidies to railroad
companies in return for the provision of uneconomic services considered to be socially
desirable. The level of subsidy is decided by a public tender, where competing firms made
bids stipulating the level of subsidy they would require in order to provide the services in
question. The scheme is run by the Office of Passenger Rail Franchising
transmission access have already been brought before the federal antitrust
authority, the Bundeskartellamt.76 Access to transmission in electricity in
Germany is now governed by a "network code" that lays out a
methodology for determining transmission prices and principles, the
Verbdndevereinbarung ("VV"), negotiated by associations of incumbent
utilities and large customers. We have analyzed the charges under the
German VV and found that, for comparable transactions, they significantly
exceed those in the United Kingdom, Norway and the United States." The
VV's charges are most unreasonable for transactions of particular interest
to entrants, such as transmission over longer distances, over short-term
periods and particularly at off-peak periods.78 The level of service offered
under the VV is extremely inflexible relative to other transmission markets
and imposes scheduling and approval requirements that are
disproportionately burdensome for entrants.79 Finally, the VV is devoid of
specific rules that would ensure transparent and non-discriminatory access
to transmission networks, and leaves significant aspects of transmission
access to the complete discretion of incumbents.
We conclude that negotiated access to electricity transmission in
Germany is likely to impede the development of effective competition
under the directive. Similar problems can be anticipated in other countries
that choose negotiated access. Even for countries choosing regulated
access, competitive issues are likely to arise pending development of a
comprehensive set of regulations to ensure transparency and prevent
The shift toward a competitive, market-oriented environment for the
network industries has the potential to produce enormous increases in
efficiency and consumer welfare. However, inadequate or misguided
regulation can limit the realization of these gains. The experiences we
have outlined above suggest the importance of proper regulation of the
network monopoly, in order to prevent monopolistic abuse and remove
barriers to competition in the potentially competitive sectors. As major
states in Europe begin upon this path, we expect to see a common pattern
of experience as the problems posed by a vertically integrated network
owner become apparent, exacerbated in places by the initial choice of
excessively light-handed regulation. These problems can be fully resolved
76See Power in Europe, 288 Financial Times Energy, Dec. 7, 1998, at 5; Power in
Europe, 296 Financial Times Energy, Mar. 29, 1999, at 13-14.
" Johannes Pfeifenberger et al., Transmission Access in Germany Compared to Other
Transmission Markets (Dec. 1998, updated Feb. 10, 1999). (unpublished article, on file
with the NorthwesternJournalofInternationalLaw &Business).
"'Id. at 6-10, 29-44.
71 Id. at 5-6, 24-25.
80Id. at 6, 25-29.
only by the approach we have outlined in this article: vertical separation of
the network, cost-based access charges, and transparent, competitively
neutral mechanisms for the recovery of stranded costs.
14 A* discussion of "incomplete contracts" can be found in OLIVER HART, FIRMs, CONTRACTS AND FINANCIAL STRUCTURE ( 1995 ).
5 See William B. Tye & Carlos Lapuerta, The Economics of Pricing Network Interconnection: Theory and Application to the Market for Telecommunications in New Zealand, 13 YALE J . ON REG . 419 ( 1996 ) (hereinafter Tye) .
60 Fees cited in "Better Rules Needed," New ZealandHerald (Oct . 15 , 1993 ), length of case cited in Telecom Corp. of NZ Ltd. vs . ClearCommunications Ltd ., [ 1993 ] 6 T.C.L.R. 138 , 148 (P.C. ), ( W. EMMONS & M. CALLES , CLEAR CoMMuNICATIONs LTD . VS. TELECOM CORPORATION OF NEw ZEALAND LTD . (B), (Harvard Business School Case Study No. N9- 798-091 , 1998 ).
6 See Tye , supranote 55 .
62 W. EMMONS & M. CALLES , CLEAR COMMUNICATIONs LTD . VS. TELECOM CORPORATION OF NEW ZEALAND LTD. (B) , (Harvard Business School Case Study No. N9-798-091 , 1998 ).
6' The distinction exists because any rent earned by A would be a consequence of the artificial scarcity (there is only one network) created by regulation. In contrast, firm B's technological superiority is genuinely scarce - - if other firms can also produce at the same price as B, then competition will lower the price of electricity to 4p/kWh, eroding B's extra profit .