The Interface Between Intellectual Property Rights and Competition in Developed Countries
Valentine Korah*
|
Table of Contents: 1. Introduction 1.1 Function of IPRs 1.2 Public goods 2. Should the competition rules protect consumers or competitors? 3. Should antitrust limit unduly broad IPRs? 3.1 United States Case Law 3.2 EC case law 3.2.1 Cases on refusal to supply goods 3.2.2 Refusal to Licence IPRs - Magill and Volvo 3.2.3 Back to goods and services Oscar Bronner 4. Licences again, IMS, Syfait and Microsoft 4.1 IMS 4.2 Syfait 4.3 Microsoft 4.3.1 Circumstances that were special 4.3.2 Commission’s remedies 4.3.3 Media player 4.4 Conclusion on refusal to deal or license 5. General conclusion |
||||||
Cite as: V Korah, "The Interface Between Intellectual
Property Rights and Competition in Developed Countries", (2005) 2:4 SCRIPT-ed
429 @: http://www.law.ed.ac.uk/ahrc/script-ed/vol2-4/korah.asp
|
||||||
|
© Valentine Korah 2005. This work is licensed through
SCRIPT-ed Open Licence (SOL).
Please click on the link to read the terms and conditions. |
||||||
1. Introduction
Intellectual property rights (“IPRs”) are rights to exclude others from exploiting a non-corporeal asset. They include patents, copyright, trade marks and other items. They may lead to significant market power when there are no substitutes on either the demand or supply side of the market. In other words, they may amount to entry barriers. Nevertheless, this is unusual. They cover only the inventions etc protected and, unless these are important, they are of little value. Most patents are never extended to other countries within the year of first application under the Paris Convention because the cost of making further applications, including translations, exceeds the expected value of the patent for further countries.
1.1 Function of IPRs
There are several reasons for granting IPRs : firstly, they may encourage various kinds of investment. As Professor Geroski says, patents are justified on the ground that without the prospect of an exclusive right, few firms would spend large sums on research and development. Secondly, patent applications and claims are published by national authorities and, even before the patent expires, the protected invention can be used for research and development, although not for commercial manufacture or sale. Consequently, the exclusive right can be perceived as the price of publication. Thirdly, patents facilitate the negotiation of licences. The licensee is unlikely to agree to pay a licence fee or accept other restrictions on its conduct unless it has been able to test the technology, but that would be difficult to arrange if the licensee were then free to exploit the invention without infringing the law. In this article I write mainly of patents and copyright, but many of the arguments apply to other kinds of IPR. Some have different justifications. Trade marks encourage the holder to invest in the reputation of the product, as consumers will be able to identify it by reference to the trade mark. In civil law systems, traditional artistic copyright gives effect to the natural law right to exploit the fruits of one’s artistic endeavour. The Anglo American tradition stresses the incentive to artistic endeavour. Most kinds of IPRs encourage investment thought to be useful and which would not be made, or not be so well or frequently made, without the prospect of an exclusive right.
Most IPRs are limited nationally. The holder of a UK patent cannot restrain anyone from exploiting the patented invention in the United States, unless it also holds a United States patent. The possibility of dividing the Common Market by licences limited to the patent in one member state, however, was early thwarted by the Community doctrine of exhaustion laid down by the European Court of Justice (“ECJ”) in the early 1970s. Once the product has been sold by or with the consent of the holder in one member state, the holder cannot rely on an IPR in another member state to restrain parallel trade.1 There was no need for this in the United States, which had been integrated long before the Sherman Act was adopted in 1890 and where IPRs have mostly been federal. The European Community does not recognise exhaustion of trade marks when the product was first put on the market outside the common market, unless there has been express consent by the holder to exploitation elsewhere. The lack of exhaustion when products first enter the common market has been criticised for dividing global markets.
Perceived ex post, the incentive to innovation may operate as a barrier to entry. There is a trade off between the incentive to innovate and the exclusive right, but competition in innovation is usually more important than competition from someone providing the same product in the same way. Joseph Schumpeter2 said that
the competition which counts [is] competition from the new commodity, the new technology, the new source of supply, the new type of organization . .. competition which commands a decisive cost or quality advantage and which strikes not at the margins of the profits and the outputs of the existing firms but at their foundations and their very lives.
He argued that
to survive in a capitalist competition, incumbents must withstand a perennial gale of competition in the form of the new consumer goods, the new methods of production or transportation, the new markets, the new forms of industrial organisation.
Unless we believe that officials can judge in advance what kinds of research and development are likely to be fruitful, and can award research grants efficiently, we have to leave the market to encourage investment in innovation. The only effective way so far used is the grant of exclusive rights for a limited time. After an important technological break through, such as the internal combustion engine, however, patent examiners may not understand how wide the exclusive right they are granting will be. In the United States many wide patents, such as the “one click” on the Amazon website, have been granted and many dubious ones have been upheld.3 In Europe the requirements of novelty and inventive step are probably stricter than in the United States, but we extended copyright from life plus 50 years to life plus 70 to avoid a German constitutional crisis. It might have been better to have allowed longer protection in Germany than elsewhere. At least the rest of us could have enjoyed the copyright works free of royalty sooner. Difficulties have also arisen from law reform by analogy, which I dislike. This has been favoured by European legislators, because they do not have to think, for instance, about what is the proper scope of copyright in software. They can just take over the existing rules for literary copyright and the onus is on the objectors to disagree. The result has been wide IPRs.
1.2 Public goods
As Professor Geroski said, it is
difficult to know how to reward those investing in public goods –
those which were expensive to create but can be exploited at
virtually no cost once they have been created. They would be used
most efficiently if access was granted at average variable cost –
often nil - but then there would be no incentive to make the original
investment. The option of leaving government to fund research and
development is not satisfactory, as civil servants have not been good
at spotting winners. The decision has therefore been taken in most
developed countries to let the market decide the worth of the
investment. This is done by giving an exclusive right to the first to
reach the patent office.4
It can then obtain the value of the invention by assigning or
licensing its IPR.
2. Should the competition rules
protect consumers or competitors?
Until about 1990, most officials in the European Community’s DG Comp, the EC Commission’s Competition Departments, following the ordo liberal tradition from Germany, thought that the competition rules should protect competitors, especially those that were of small or medium size. Many officials did not understand the difference between that and protecting consumers. If a competitor were ousted from the market consumers would be harmed. Few officials looked at the market as a whole to see whether enough competitors remained, or thought ex ante and at the margin. It was widely thought that exclusion was anticompetitive, and justifications based on free rider arguments and avoiding hold up problems were not widely accepted. In such a climate of thought, it is hardly surprising that officials and courts rather easily thought that IPRs should be limited by antitrust.
The philosophy of DG Comp has changed
radically in the last decade or so. Recent Commission documents5
frequently refer to the goal of consumer welfare. Many of the legal
secretaries at the Community Courts in Luxembourg have been
introduced to economic concepts at least during their courses for an
LL.M., and often by practice in competition law for five or six
years. There is now far more sympathy for the need to induce
investment.
3. Should antitrust limit unduly
broad IPRs?
Should antitrust limit IPRs that are wider than necessary to induce sufficient innovation? Even with hindsight, it is hard to determine when a right is palpably too broad. If the extent of protection is determined case by case, the resulting uncertainty may fail to induce sufficient investment. Have the judges got better training or facilities to decide the amount of protection than the legislature or patent examiners? Alternatively, if a natural monopoly is in the private sector, a regulator may be appointed to control excessive exploitation, although cost plus is a rotten criterion that encourages escalation in cost. Other criteria are little better. The advantage of IPRs is that the market is left to determine the value of innovation.
3.1 United States Case Law
The antitrust doctrine of essential facilities was developed in the United States to limit property rights and was first applied to physical assets rather than IPRs in circumstances where it was not so difficult to determine the compensation for the investors. In US v. Terminal Railroad of St. Louis Association,6 the bridge over the Mississippi had been built by a joint venture between the railroads that had started to go West, but found themselves barred by the Mississippi at St. Louis. There was nowhere to North or South for a hundred miles where a rival bridge could be built. A second bridge near St Louis would be a very expensive project as, not only is the river broad, its course shifts. Access was vital to the railroads that had arrived after the bridge was built. The court ordered that new railroads be given access to the bridge on terms similar to those agreed between the original railroads. Some adjustments should be made to provide for the risk incurred by the original joint venturers and leave them with a reward for their investment that would act as an incentive to further investments by anyone. That would have been more manageable than saying the level of remuneration should be somewhere between the cost of permitting entry (building lines to connect the railroad’s lines to those on the bridge) and the opportunity cost, the loss of the monopoly profit.
After a period when lower courts extended the doctrine to more dubious cases, the Courts were constrained by the classic article by Philip Areeda.7 Recently, the doctrine was further narrowed by the Supreme Court in Verizon Communications Inc .v. Trinko, LLP8 and by the Court of Appeals for the Federal Circuit in Covad Communication Company et al v. Bell Atlantic Corp et al. 9 In Europe, too, after the opinion of AG Jacobs and the judgment of the ECJ in Oscar Bronner,10 the doctrine has been greatly narrowed. There is concern that if the European courts intervene with property rights more than those in the United States, much of the research and development and production of highly technical products will be performed in the United States,11 with the loss of good quality jobs in the EU. In Oscar Bronner, the drawbacks of requiring supply to an essential facility were well set out by AG Jacobs: compulsory access reduces the incentive to invest in the original investment, or its duplication where that is practicable, and leads to detailed regulation to determine the amount of compensation payable.
3.2 EC case law
3.2.1 Cases on refusal to supply goods
The cases on refusals to supply goods to existing customers started with Commercial Solvents,12 but have been far from consistent over time. The ECJ held that where a dominant firm had a real stranglehold on a market downstream, and access to some product, goods or services, was indispensable to enter a connected market, a refusal to give access to a former customer amounted to an abuse of its dominant position contrary to article 82 (EC Treaty), if it eliminated all competition by the person requesting supply unless there was some objective justification.
Sometimes the ECJ went very far. In United Brands Company and United Brands Continental BV v Commission,13 the ECJ said:
182. . . . It is advisable to assert positively from the outset that an undertaking in a dominant position. . . – which cashes in on the reputation of a brand name known and valued by consumers – cannot stop supplying a long standing customer who abides by regular commercial practice, if the orders placed by the customer are in no way out of the ordinary.
This sounds more like regulation than competition. It represented French law of the time, which was based on ‘planification’ and fair shares for all dealers. In those days, however, as stated above competition law was seen by competition officials of the EC and by the judges in Luxembourg as protecting competitors; only in the last decade has consumer welfare been widely recognised as more important.
In Télémarketing14 the ECJ said:
26. [The ruling in Commercial Solvents15] also applies to the case of an undertaking holding a dominant position on the market in a service which is indispensable for the activities of another undertaking in another market. . . .
27. [An abuse] is committed where without any objective necessity an undertaking holding a dominant position on a particular market reserves for itself [or a subsidiary] an ancillary activity which might be carried on by another undertaking as part of its activities on a neighbouring market with the possibility of eliminating all competition from such undertaking.
In all these and other cases, the firm wanting access was a long standing former customer, the product was indispensable and without it a particular customer would be excluded from the neighbouring market. All the cases mentioned the possibility of justification, although property rights were held not to be a justification. What would amount to a justification was not much addressed.
The Commission adopted several decisions in this field, most of them relating to ship or air ports. It did not stress that the original investment was often made when the operators of the port were in the public sector or protected by special or exclusive rights. In that case, they were not forced by competition to consider carefully whether investment would be economic and incentives to invest would be less important. Moreover, as AG Jacobs observed in Oscar Bronner it is particularly difficult to compete with such a firm.
In the next series of cases, most of them involving IPRs, the basis of the duty to supply seems to have changed.
3.2.2 Refusal to Licence IPRs - Magill and Volvo
In Magill,16 a new entrant published a comprehensive guide to the television programmes to be transmitted by the three stations licensed to transmit in Ireland and Northern Ireland (separate Member States). All three obtained injunctions for breach of copyright. The ECJ confirmed the Commission’s decision that this amounted to the abuse of a dominant position. It confirmed its established case law that IPRs do not necessarily confer a dominant position (para 46). In Magill, however, each station had a de facto monopoly of its own list. In the absence of Community harmonisation, the ECJ held that it is for national law to define the scope of intellectual property. Nevertheless, the exercise of exclusive rights may amount to abuse in exceptional circumstances. The ECJ spelled out the circumstances that were special in that case (para 54):
The Appellants’ refusal to provide basic information by relying on national copyright provisions thus prevented the appearance of a new product, a comprehensive weekly guide to television programmes, which the appellants did not offer and for which there was potential consumer demand. Such refusal constitutes an abuse under heading (b) of the second paragraph of Article 82 of the Treaty.
The ECJ confirmed there was no justification for such refusal – copyright was not treated as a justification and the appellants (para 56):
reserved to themselves the secondary market of weekly television guides by excluding all competition on that market (citing Commercial Solvents) since they denied access to the basic information which is the raw material indispensable for the compilation of such a guide.
The question that has been much
discussed is whether the three circumstances listed as special in
paragraph 54 were cumulative, and whether they were exhaustive. I
believe they were cumulative, because of the conjunction and.
More important, are they exhaustive? Clearly novelty is a relevant
consideration, if the newcomer would produce something new, it would
help consumers. Nevertheless, there seems to me to be no reason why
there should be a duty to supply only if the newcomer is going
to supply something new. Parts that must fit cannot be new. As Sir
Jeremy Lever observed,17
the conclusion may have been right – there were only three
companies licensed to transmit TV in Ireland and Northern Ireland and
on that ground a requirement that they should not exclude others from
the complementary market in guides was sensible, but the reasoning
based on copyright was less satisfactory.
In Volvo,18 the ECJ had stated at para 8 that the right to restrain third parties ‘constitutes the very subject matter of that exclusive right (copyright in a front wing panel),’ but in the next paragraph gave three examples when a refusal to supply might be abusive:
the arbitrary refusal to supply spare parts to independent repairers, the fixing of prices for spare parts at an unfair level or a decision no longer to produce spare parts for a particular model even though many cars of that model are still in circulation, provided that such conduct is liable to affect trade between Member States.
These special circumstances are quite different from those that influenced the ECJ in Magill. So, one might assume that the circumstances found special in Magill are not exhaustive.19 Otherwise the judgments are inconsistent.
The judgments in Volvo and Magill differ from those relating to physical goods in that the complainants had never been supplied before and it seems that any number of people may be entitled to access – eg repairers of vehicles in Volvo. For a refusal to amount to an abuse, some conditions for a duty to supply seem to be the same – access must be indispensable for the trade of the new comer, and the possibility of justification is mentioned in all the judgments, but without any precision. Until Magill, the refusal had to eliminate all competition on the part of the complainant, but in Magill, it was to eliminate all competition in the market. The difference was not important on the facts, but if the change in formulation was deliberate, the duty to supply was substantially reduced. A dominant firm that feared new competitors might arrange to supply one firm that was not likely to compete very aggressively.
3.2.3 Back to goods and services Oscar Bronner
The next judgment of the ECJ, in Oscar Bronner,20 limited the requirement to supply. In one of his many fundamental opinions, AG Jacobs considered the United States case law on refusal to supply, and pointed out the three disadvantages of requiring access to essential facilities owned by a dominant firm: it reduces the incentive to invest in creating the facility initially, and to duplicate the facility where this is possible. Finally, he observed that since access is not granted voluntarily, someone will have to settle the amount of compensation and this may require detailed regulation.
The case concerned the only national home delivery service for newspapers in Austria and there were other ways of marketing papers, so the argument for compulsory access by a new paper was weak. Nevertheless, opinion and judgment were so much better reasoned than in most cases on essential facilities and they rolled back that concept considerably.
The ECJ in Oscar Bronner
returned to the old formula of eliminating all competition from a
particular trader, rather than all competition in the market.
Nevertheless, we still have rather hazy ideas about what amounts to
an essential facility. In judgment after judgment we are told that
access must be ‘indispensable’ and not merely convenient
to the newcomer. The Court recognised the change in the basis of the
case law in Volvo and Magill and said that even if the
IP cases applied to goods, there were sufficient substitutes for
marketing newspapers.
4. Licences again, IMS, Syfait and
Microsoft
4.1 IMS
In IMS,21 the ECJ delivered a formalistic judgment based mainly on the wording of the judgment in Magill. Neither AG Tizzano, nor the ECJ, gave any hint of being interested in policy. First the judgment treated the criteria in Magill as exhaustive: unless the newcomer intended to produce something new, there should be no duty to supply. No reason of policy was given for this statement. As I said in relation to Magill, I hope the duty to supply is narrow, but I do not see why novelty should be required. In the case of parts that must fit, novelty is inappropriate. So, the judgement in IMS is hard to reconcile with that in Volvo.
The Court reverted to the formula in Magill of excluding all competition from the market. Consequently, it is possible for a dominant firm to ameliorate its duty to give access by admitting a firm that is not likely to compete very aggressively. The ruling in Oscar Bronner was more sensible in this regard.
The IMS judgment and opinion included one helpful statement. The fact that no one was being supplied – that there were no transactions downstream – does not matter. There must be two markets, but the second market may be potential or even hypothetical. This seems to me to be a polite way of saying there need not be a second market. Provided that access is really indispensable, this seems desirable. It is probably worse for consumers if the dominant firm supplies no one than if it supplies only some.
Interestingly, in its decision, the Commission pointed out that the maps on the basis of which IMS presented data to its customers had become an industry standard, and suggested that that sufficed to raise a requirement that a dominant firm should license even if it enjoyed copyright in them
4.2 Syfait
Syfait22 was not an intellectual property case. The Greek competition authority asked the ECJ for a preliminary ruling whether it was per se illegal for a dominant firm to refuse supply of medicines with the intent of reducing parallel trade. Glaxo limited the amount of various medicines over which it held a dominant position to its Greek dealers in order to protect dealers in markets where higher prices might be charged. The common market in pharmaceutical products is unusual in that each member state controls the maximum price payable for medicines at a different level. Moreover, distribution was also controlled by national legislation, which prohibited the withdrawal of a medicine once introduced. The opinion of AG Jacobs was very interesting, but the ECJ declined jurisdiction under article 234 EC Treaty, since it thought that the competition authority was to some extent under the influence of the minister and , therefore, not a court or tribunal competent to make a reference. Consequently, it did not consider the questions on their merits.
The Advocate General had noted the differences between the early and the later case law. He added that a duty to supply does not arise easily or automatically (para 69), but that an intention to restrain parallel trade would probably be a reason to condemn a refusal to supply. Nevertheless, all the judgments admitted the possibility of justifications (para 71),and here, the national laws provided a justification. Parallel trade would make it impossible to ensure adequate supplies everywhere because they would be sourced from the country where the maximum price was lowest, in this instance Greece. The national regulations were segregated – it was not Glaxo that had divided the common market. If Glaxo were forced to supply, it might lead to some countries not being supplied with drugs, or at least to their introduction being delayed. Therefore a refusal to supply may be both reasonable and proportionate.
Mr. Jacobs went on at paras 89-91 to analyse the economics of the innovative pharmaceutical industry: its overheads were high and largely sunk, while its variable costs were very low. It was rational to sell wherever manufacturers could recover their variable cost, but it did not follow that they would be able to recover their total cost if all their sales were at the lowest price in the Community. This amounts, in all but name, to approving of ‘Ramsey Pricing,’23 something that neither the Commission nor ECJ had previously accepted.
It must be stressed that the situation was very unusual. The common market was divided not by Glaxo, but by the segregated controls imposed by Member States. There would be a real problem of ensuring that countries with the lowest maximum prices were supplied if a duty to supply was imposed. The benefits of parallel trade would go to the wholesalers in the countries were prices were limited at the lowest level and not to the patients nor to those paying for their treatment.
The narrow circumstances that might justify the refusal according to the Advocate General were important, as these arguments had been raised and dismissed in the early cases on exhaustion. He has not advocated a wholesale attack of parallel trade. Nevertheless, in all these cases, the Advocate General observed that there is a possibility of justification, and in this case, this may be possible where the common market is divided by member states and where prices and distribution are controlled by national law.
What will happen now that the ECJ has declined jurisdiction? It seems to me clear that the Greek authority should follow the Advocate General. His opinion on the substance has not been rejected by the judgment and retains some authority. The next step may be an appeal from the Authority by the dealers in the cheaper Member States, such as Greece. The relevant national Court will then probably make a further reference to the ECJ. Meanwhile, the Advocate General will have retired this year, and some of the judges may have done so. Indeed the terms of about half of them will have expired by the time the case gets back to the ECJ. The Chamber will consist of different judges.
Meanwhile, the question is whether
the Commission and national authorities should follow the Greek
Authority with the objective of achieving congruent decisions
throughout the common market.
4.3 Microsoft
The Commission’s decision in Microsoft,24 was adopted a month before the ECJ decided in IMS that the circumstances treated as special in Magill were exhaustive. The Commission has treated them otherwise.
The Commission found that Microsoft was overwhelmingly dominant on the market for operating systems for client personal computers (client PCs). Not only did it supply over 90% of them, its dominance was protected by high entry barriers: the ability to leverage from existing products; indirect network effects; and learning or switching costs.
Network markets are those where the more customers a supplier has, the more people want to buy from it. The usual example given is telephones. Having the only phone would be little use. The market may tip towards the first supplier to accumulate customers, even if that brand is less good or more expensive than another. Now that Microsoft’s personal computers are ubiquitous, everyone wants a PC that is compatible with Microsoft.
These direct network effects lead to what the Commission calls an ‘indirect network effect’ and some people call an ‘applications barrier to entry’. It is expensive to make software compatible with more than one standard. So, suppliers of applications that run on PCs target their efforts on Microsoft PCs, so as to have a larger range of possible clients. This reinforces Microsoft’s initial dominance over PCs, because many items of applications software run well on them.
Nevertheless, even in network markets, competition remains for the market, even if it is subdued in the market. There is a race to be the first firm to supply the next generation of technology. Network effects may last only for one generation of technology. Before Microsoft became the market leader for client PCs, IBM led the way. Moreover, if transportable ‘middleware’ were to develop, both network effects would be overcome. If someone develops a platform that can run on several operating systems for PCs, and which has sufficient applications programming interfaces to which applications can attach, all applications could be made compatible with the middleware, and ported with it from one operating system for PCs to another. Only if that market tips would network effects lead to the owner of middleware being dominant.
A third barrier to entry is the learning or switching costs. Many things are easier for buyers of the most popular system. It is easier to find support staff familiar with it, employees do not have to be taught to use the system etc. This barrier to entry may survive the next generation of technology.
There was found to be abuse of this dominant position by leveraging it in relation to operating systems for work servers and for media players.
Work group servers perform many mundane tasks, enabling people in an organisation to share information, imposing priorities to the printer and so on. Since the windows operating system for PCs is so popular, it is important that servers be compatible with it. The Commission found that when it first entered the server market, Microsoft gave the designers of operating systems for servers the exact protocols required for interconnection; but when it was gaining acceptance, the amount of information supplied was reduced (para 588). As a result, competing servers did not work so well with Windows (paras 592 & 613). Microsoft already had a dominant position over servers with some 70% of the market. The Commission did not allege that the market had yet tipped.
Article 82 lists some examples of abuse, including “(b) limiting production, markets of technical development to the prejudice of consumers”. In Microsoft, the Commission considered that making it difficult for its competitors to operate limited technical development and markets, contrary to article 82 (paras 693-701). It rejected the view that the IPRs held by Microsoft justified the failure to supply the interface information.
The Commission alleged inter alia that Microsoft had extended its dominant position (quasi monopoly) over operating systems for client PCs for many years, to the adjacent markets in operating systems for servers (paras 533, 779-781, 1065).25 The Commission decided that withholding the information necessary to design competing programmes for work group servers compatible with the Windows was an abuse (paras 780-784) and risked eliminating competition from the server market (paras 585-589 and 692), stifling innovation and reducing consumers’ choice by locking them in (para 781).26 The Commission decided that this foreclosed competitors from designing work group servers fully compatible with Windows.
At paragraphs 548-559 the Commission shortly considered several of the judgments on refusals to license in special circumstances. It decided that the special circumstances mentioned in Magill27 were not exhaustive (para 555) (contrary to the ruling of the ECJ in IMS a month later that they were exhaustive.)
4.3.1 Circumstances that were special
The Commission found several circumstances that in combination made the circumstances special: Microsoft’s refusal to supply the complainant, Sun Microsystems, was part of broader conduct of not supplying vendors of work group servers with information necessary to achieve interoperability (paras 573-577) It involved disruption of previous patterns of cooperation when full interface information had been made available (paras 578-584. This observation may be an attempt to bring the case within the early case law on a refusal to supply goods and services to an existing customer, but the relevance of the disruption was not spelled out.
The Commission added that Microsoft’s conduct created a significant risk of eliminating competition in the supply of work group servers (paras 585-692) and harming consumers (paras 693-701). If Sun and other providers of work group servers were unable to compete, innovation by them would cease. If they lack the interface protocols for Windows, there will be no point in their investing in innovations, which they will not be able to use themselves, but would be able only to sell to the dominant incumbent. In the past they had introduced new features that customers had bought. If they are finally squeezed out of the market, the only innovations will be those introduced by Microsoft, and its incentive to innovate will be reduced if it faces no competition in innovation (para 725).
Operating systems for servers were complements to operating systems for PCs and, for some purposes, substitutes. When Microsoft first entered the server market, it supplied full interface information. Its software for servers was inferior to that of its competitors. Making competitors’ servers work well with Windows operating systems for PCs enabled it to sell more licenses for Windows operating systems for PCs. The Commission observed , moreover, that if other servers worked well with Windows operating systems for PCs, the value of the latter would be higher (para 587). Microsoft was thus able to extract the full value of its dominant position over operating systems for PCs: it was indifferent whether it obtained royalties from other producers of servers or increased its sales of licences of operating systems for PCs.
The Commission noted further that once Microsoft’s workgroup server’s operating systems gained acceptance, Microsoft ceased to supply as much information as previously (para 588). The Commission pointed out that that was rational. Other suppliers of servers were competitors whose market shares were declining (paras 590-597).28 Although the Commission did not state that the market for servers had already tipped in favour of Microsoft, it said that rivals’ products were marginalized and, in turn, this buttressed Microsoft’s dominance over operating systems for PCs (para 769). This is not the traditional objection to leveraging, which is the extension of market power from the tying product to that tied, although it is now generally accepted by economists now-a-days. The Commission was mainly concerned that the increasing incompatibility between Windows operating systems and rival servers would buttress the original monopoly over operating systems for PCs, rather than that Microsoft was monopolising the tied market (para 769).
4.3.2 Commission’s remedies
The Commission’s decision required Microsoft to make available the necessary interface information to its competitors in servers on reasonable and non-discriminatory terms (paras 998-1009). Only the interface information had to be licensed, not the source code (para 1004). The Commission thought that this prevented Microsoft’s competitors from cloning its ideas (paras 713-722).
Forbidding Microsoft to discriminate in favour of its own server operations is not helpful in setting the price in the long term, as it could always increase the price to its own server or division or subsidiary, but it may help in deciding what information must be disclosed. The requirement that the price for competitors must be reasonable is controversial. This price could be anywhere between the cost of providing the information, virtually nil, and the opportunity cost – the cost of losing its exclusive right. The Commission has expressly excluded opportunity cost, but the range of possibilities is still immense. So, the Commission insists on the appointment of a monitor, to be paid by Microsoft, to ensure compliance with the order (para 1043-1048). The monitor will have to make many controversial decisions, most of them arbitrary.
4.3.3 Media player
The Commission also objected to
Microsoft tying its media player to the operating system for PCs, but
this does not relate to whether competition law should trump IPRs,
and will not be considered here. This part of the decision is weaker
than that relating to refusal to supply the interface information.
4.4 Conclusion on refusal to deal or license
If Advocate General Jacobs is right in Syfait and the Commission in IMS and Microsoft, a dominant firm may have be a duty to supply
1) a former customer when a refusal would eliminate competition from the complainant,
2) third parties who have never been supplied before where the circumstances are special (Magill and IMS), whatever this may mean, or
3) if the refusal would reduce parallel trade.
4) The Commission suggested in IMS,
that a dominant firm may be required to grant access to a de facto
industry standard. This is repeated in its decision in Microsoft.
5. General conclusion
Competition and property fundamentally pursue the same aims – consumer welfare, together with the increase of useful investment. The law of property grants an exclusive right in the hope that this will induce people to make investments in things that people want to use. The investor will be able to use the fruits of the investment itself, rent out or sell it to others. This should lead to the optimal amount of investment being made. This is not the same as a fair return, but enables the market to decide the compensation due to the investor.
Problems arise in the case of IPRs because it is particularly difficult to know where the investment should be made. The law favours successful investors who may have obtained important rights with little effort, but the possible alternatives, awarding research funds to likely winners, or the state paying for successful innovations, are thought not to be practical.
The US Agencies conducted hearings in 2003 and we expect a report at the end of 2005 on the extent to which antitrust should override IPRs. The hearings disclosed less conflict than had been expected.29
The EC Commission is preparing a paper on the reform if Article 82 and it is hoped that it will not extend the doctrine of refusal to supply far. One of my concerns is that it adopts very narrow market definition on the basis of a relative change of price of 5 or 10%. This may work fairly well for commodities where the market is competitive, but not where a firm has significant market power, and there is no evidence of what a competitive price would be; nor when the market power is based on investments in developing an essential facility in a network market. Fortunately, the EC Commission is reconsidering the way it defines markets. Pressure is building up for such an enquiry.
The ECJ is stressing that some dominant positions are very strong.30 If the doctrine of refusal to license were confined to such cases, refusal to supply would be a less frequent disincentive to investment, but this may not be true in the pharmaceutical industry, where markets are usually confined to medicines for a particular problem and markets are considered concentrated.
The cost of awarding IPRs is that, perceived ex post, they operate as barriers to entry, sometimes to very important assets. Antitrust in the US has not attempted to provide access through refusals to license being treated as monopolisation. In the EC, however, the Commission and courts were relatively willing to require access, but only where the incumbent has a strong monopoly, a stranglehold downstream and access is indispensable to the complainant’s business. Both EC Commission and EC courts have had great difficulty in deciding what other conditions are required for refusal of access to constitute an abuse of a dominant position and the case law has been far from consistent.
The tension between increasing supply in the short term and sometimes in the longer term, on the one hand and retaining the incentive to investment provided by IPRs remains, and there is no simple answer as to where the balance lies.
* Professor Emeritus of Competition Law at University College London, Honorary Professor at the College of Europe at Bruges, author of many books and articles, including
- EC Competition Law and Practice, 8th. ed., Hart Publishing, Oxford, 2004,
- Cases and Materials on EC Competition Law, 2nd ed. September 2001, Hart Publishing, 3rd edition in preparation for 2006.
- Intellectual Property, being chapter 9 of European Competition Law, Second Edition, Lexis Nexis, General Editor Valentine Korah, forthcoming release 2005.
- Intellectual Property and the Competition Rules, forthcoming, Hart Publishing.
1 Centrafarm BV and de Peijper v Sterling Drug Inc (15/74), 31 Oct 1974, [1974] ECR 1147, [1974] 2 CMLR 480, CMR 8246 and many other cases affecting not only patents but other IPRs apart from performing rights.
2 J.A. Schumpeter (1943) “Capitalism, Socialism and Democracy” George, Allen & Unwin Ltd, London, 84.
3 By the Federal Circuit for the District of Columbia, which hears all patent appeals in the United States
4 In the United States, it used to be to the first to invent, but that had the drawback of not being transparent, as a firm might spend resources on inventing something of which only one firm knew, only to find that the latter was entitled to the patent (although the second inventor might be allowed to use its own invention).
5 E.g., Commission notice, Guidelines on the Application of Article 81(3) of the Treaty, OJ 2004, C101/97:
13. The objective of Article 81 is to protect competition on the market as a means of enhancing consumer welfare and of ensuring an efficient allocation of resources. Competition and market integration serve these ends since the creation and preservation of an open single market promotes an efficient allocation of resources though out the Community for the benefit of consumers.
6 224 US 383, 32 S.Ct.507, 56 L Ed 810 (1912).
7 ‘Essential Facilities: an Epithet in Need of Limiting Principles,’ (1989) 58 Antitrust Law Journal, 841.
8 540 US 398 (2004).
9 March 1, 2005, 540 US 398 (2004)
10 Oscar Bronner GmbH & Co. KG v Mediaprint Seitungs- und Zeitschriftenverlag GmbH & Co. KG and others (C-7/97), 26 November 1998, [1998] ECR I-7817, [1999] 4 CMLR 112, [1999] CEC 53 (“Oscar Bronner”). See also text to note 21.
11 The United States Guidelines permit more restraints on conduct than the EC. In particular, they place the burden of proving lack of efficiency is on the person alleging illegality. Antitrust Guidelines for the Licensing of Intellectual Property, 6 April 1995, (1995) 68 ATRR 462, CCH TRR 13,132, [1995] 7 EIPR 3, 3.4.
12 Istituto Chemioterapico Italiano SpA and Commercial Solvents Corp v Commission (6 & 7/73), 6 March 1974, [1974] ECR 223, [1974] 1 CMLR 309, CMR 8209.
13 (27/76), 14 February 1978, [1978] ECR 207, [1978] 1 CMLR 429, CMR 8429.
14 Centre Belge d’Etudes du Marché Télémarketing SA (CBEM) v Compagnie Luxembourgeoise de Télédiffusion (311/84), 3 October 1985, [1985] ECR 3261, [1986] 2 CMLR 558, CMR 14246.
15 Istituto Chemioterapico Italiano SpA and Commercial Solvents Corp v Commission (6 & 7/73), 6 March 1974, [1974] ECR 223, [1974] 1 CMLR 309, CMR 8209.
16 Radio Telefis Eireann and Others v Commission (C 241 & 242/91 P), 6 April 1995, [1995] ECR I-743, [1995] 4 CMLR 718, [1995] 1 CEC 400 (ECJ) (“Magill”).
17 At the International Intellectual Property Law and Policy conference at Fordham in April 2005.
18 Volvo AB v Erik Veng (UK) Ltd (238/87), 5 October 1988, [1988] ECR 6211, [1989] 4 CMLR 122, CMR 14498 (“Volvo”).
19 Contrast IMS below.
20 Oscar Bronner GmbH & Co. KG v Mediaprint Seitungs- und Zeitschriftenverlag GmbH & Co. KG and others (C-7/97), 26 November 1998, [1998] ECR I-7817, [1999] 4 CMLR 112, [1999] CEC 53. Also considered above.
21 IMS Health GmbH & Co KG v NDC Health (C-418/01), 29 April 2004, [2004] 4 CMLR 1543. Note that the appeal from the Commission’s interim decision requiring supply has been struck out for lack of substance now that the Commission’s interim decision has been withdrawn. My comments relate to the preliminary ruling of the ECJ in response to the German court that granted an injunction for breach of copyright.
22 (C-53/03), Opinion Jacobs, 28 October 2004, judgment, May 31, 2005 (“Syfait”).
23 F.P. Ramsey, ‘A Contribution to the Theory of Taxation,’ (1929) 37 Economic Journal 47-61. Provided that no price is below variable cost, no one is worse off if most of the sunk overhead cost is recovered from those willing to pay more and most customers are better off. In the low priced market, supplies will be available for those able and willing to pay the variable cost, and this may even provide some contribution to the overhead, which would benefit those have to pay more. Economists, all of them, generalise the theory and argue that supply will be most efficient if the overhead is recovered from different markets in inverse proportion to the elasticity of demand.
24 Press Release IP (04)382, [2004] 4 CMLR 1231 (“Microsoft”).
25 Earlier, the United States Department of Justice had attacked Microsoft for extending its market power over Windows PC operating systems to the complementary market for internet browsers. This prevented Netscape from attracting enough applications software to become satisfactory middleware that might have removed the applications barrier to entry into the market for client PCs.
26 See F Leveque, ‘Innovation, Leveraging and Essential Facilities: Interoperability Licensing in the EU Microsoft Case,’ [2005] 28 World Competition Law and Economics Review 71.
27 The ECJ’s preliminary ruling in IMS, was decided a month after the Commission’s decision and did not mention the Commission’s view of what circumstances are special.
28See again F Leveque, ‘Innovation, Leveraging and Essential Facilities: Interoperability Licensing in the EU Microsoft Case,’ [2005] 1World Competition Law and Economics Review 71, at pp.82-85.
29 Federal Trade Commission Hearings on Enforcement Policy, 69 BNA ATRR 487, 670-682. See also FTC report “Anticipating the 21st Century – Competition Policy in the new High-Tech Global Market Place”, May 1996. Available at http://www.ftc.gov/opp/global/report/gc_v1.pdf .
30 See AG Fennelly in Compagnie Maritime Belge Transports SA v Commission (C-395/96 P), 16 March 2000, [2000] ECR I-1365, [2000] 4 CMLR 1076 and AG Jacobs in Oscar Bronner, quoting AG Warner in Commercial Solvents. In Compagnie MaritimeBelge, the ECJ stressed at para. 119 that Compagnie MaritimeBelge had a market share exceeding 90%. See R Whish (2005, 5th edn) “Competition Law” Butterworths, London 189 – 191.