COMPETITION LAW AND POLICY IN THE EUROPEAN UNION

COMPETITION LAW AND POLICY IN THE EUROPEAN UNION

1. Foundations

The competition law of the EU responded to Europes mid-century economic conditions. Its development, driven by the imperative of market integration, profited from the symbiosis between the protection of competition and the promotion of open trade. Decisions of the European Court of Justice (ECJ), pursuing the goals of strengthening the community and eliminating trade barriers, established the legal framework underpinning an ambitious Community competition policy. The European Commissions Competition Directorate (DG Comp, formerly DGIV) is in a nearly unique position in the European Community system, because in the area of competition policy the Commission can apply direct enforcement power that is not dependent on national governments.

Community competition law is undergoing a profound transition, after moving beyond the initial goals of opening markets and establishing a competition culture to become a mature, comprehensive enforcement structure centred on the European Commission. The substantive principles that the Community institutions developed have now become a common legal framework shared with the national laws of the Member States. In the future, the law will evolve within the network of national and Community agencies that share responsibility for applying it. The principal focus of this study is the European Commission, as the administrative organ of the 25-country European Union. Most of the discussion would also apply in the context of the EEA, with its 3 additional countries and closely co-ordinated competition policy and enforcement.

1.1 Context and history

In post-war Europe, administered economies faced development needs and state monopolies. The institutions of the European Union were created in a context of state intervention through ownership and control over trade and prices, as Europe was rebuilding after depression and war. The designers of the new post-war political economy framework, seeking to expand and integrate markets and sustain development, concluded that competition policy would be a necessary element of the new structure, principally to curb abuses of national monopolies.

The concepts and institutions of EU competition law appeared first in the European Coal and Steel Community (ECSC). The need to stabilise Germanys post-war economy and integrate it into western Europe, while preventing German firms from dominating markets, was most evident in the key coal and steel sectors.

A new legal entity, the ECSC, was created by the 1951 Treaty of Paris to administer these sectors. The ECSC included most of the elements that were later incorporated into EU competition policy.

The competition rules in the Treaty of Rome build on those of the ECSC about agreements, dominance and subsidies, though not the ones about merger control. The rules for the Common Market add some precision to the prohibition of restrictive agreements, while strengthening rule against abuse of dominance into a prohibition. The Council gave the Commission broad powers to develop and apply the law. The Member States did not focus on competition policy very much during the 4 years it took to prepare the regulation to implement the Treaty rules, and thus the Commission ended up with more autonomy in this area than it might haveotherwise. (Goyder, 1998).

The 1962 enforcement regulation centralised responsibility in the Commission. Its approach to cartels emphasised the Treatys prohibition, because obtaining an exemption required a decision from the enforcer. Between 1957 and 1962, some national agencies had begun to apply the Treaty provisions. But the Council enforcement regulation marginalised the national agencies and courts by giving the Commission priority in investigations under Community law and exclusive competence over the key subject of exemptions. The Commission had to consult about enforcement actions with a committee of representatives from the Member States, but the committees views were advisory, not binding. The Council rejected a proposal to give this committee veto power.

When the system was implemented in 1962, the Commission received over 35,000 notifications requesting exemption or negative clearance. Case-by-case response proved impossible; general rules were obviously needed instead. The block exemption regulation of 1965 responded to the overload, while underscoring the Commissions autonomy in competition policy.

The Council delegated authority to the Commission to issue regulations setting generally applicable objective standards for exemption from the cartel prohibition. The Council has not delegated such authority to issue regulations to the Commission for any other substantive field. With encouragement from the judiciary, competition law framed an economic constitution. In the process of decision and appeal during the first decades of Community competition law enforcement, the dialogue between the Commission and the ECJ set the direction and scope of competition policy.

Increasing confidence in EU competition policy culminated in the 1989 merger regulation, completing the European competition-policy toolkit. After 20 years of laying the foundation, by the 1980s the Commission was taking stronger enforcement actions. The single-market program and the Single European Act of 1986 reinforced the market-integration objective and thus provided additional momentum for active competition policy. Merger control, which the drafters deliberately omitted from the 1957 Treaty, was finally adopted by a Council regulation after 17 years of effort. Business supported this move to establish a single point in Europe for regulation of large-scale mergers. Implementation involved a process of adjustment.

The first time the Commission prohibited a merger, in 1991, governments where the firms were located protested. It took some time to overcome the early impression that political factors could play a role in Commission merger actions. (Gerber, 1998) Wielding this significant power to affect key business decisions bolstered the visibility and prestige of Commission competition enforcement generally. State interventions and monopolies drew increasing attention.

Community competition policy is now being reshaped in terms of economic principle. In this project, which has been underway since the mid-1990s, the Commission is building on the increasing reliance on economic reasoning and analysis demanded by merger control and the liberalisation program. With the market integration goal largely accomplished concerning industry and trade, attention shifted both to the constraints on trade in services and to the standard competition policy fare of cartels and monopolies. The shift in enforcement focus entailed a shift in analysis away from formal categorisation. The foundation for the economic reconstruction was laid in the 1997 guideline about the definition of a relevant market. The first major substantive projects were the complete revisions of the rules about vertical and horizontal restraints, replacing long listings of specific requirements and prohibitions with general principles and market-share tests. The Commission is reformulating regulations and revising guidance to modernise Community competition law along these lines.

Closer judicial oversight led the Commission to improve its internal procedures. Addition of the new Court of First Instance (CFI) in 1989, which doubled the capacity of the Communitys judiciary, provided a more practical avenue for parties to appeal Commission decisions. The Community courts remain supportive of the Commissions competition policy initiatives. But the CFI has rejected Commission actions for procedural errors and for defects in the quality of its reasoning and its treatment of evidence. In 2002, the CFI rejected three Commission merger decisions within 4 months, in opinions that sharply criticised the Commissions economic analysis and its treatment of evidence.

Partly in response to problems that these decisions revealed, DG Comp created a new special economic unit and accelerated the recruitment of industrial economists generally in order to increase its capacity for economic analysis, and it introduced additional quality-control checks into its processes of investigation and case evaluation. After 40 years of experience, in 2004 the Community implemented a modernised enforcement process.

By removing the Commissions monopoly on deciding about exemptions, the new system makes it much more practical to apply the law through national institutions and processes. The founding treaty envisioned enforcement co-operation between the Commission and Member State national authorities, at least as a transition measure. But of the original six, only Germany had a similarly ambitious competition law system at that time. Now, all of the Member States have competition laws and enforcement agencies, and their national substantive laws have generally converged on the Community standards.

Some areas of divergence remain, and the new enforcement regulation deals with the questions of co-existence and supremacy. National law can be applied to conduct that meets the jurisdictional test concerning effect on trade between Member States, but the national authorities must also apply Community law at the same time, and national law cannot prohibit restrictive agreements that would not violate the Treaty. But national law can be stricter than Community law concerning unilateral behaviour.

1.2 Policy goals

The Treaty makes competition a principal goal, but it does not elaborate what the concept means. The activities prescribed for the Community institutions include several that directly implicate competition policy: to provide an internal market characterised by the abolition, as between Member States, of obstacles to the free movement of goods, persons, services and capital, and a system ensuring that competition in the internal market is not distorted. (Article 3) The Community and its Member States are to adopt a co-ordinated economic policy based on an open market economy with free competition. (Article 4) These parts of the Treaty thus set out the goal of free and undistorted competition for the Communitys internal market. The basic rules of Articles 81-87 do not limit the choice of policy goals. They do make clear that the competition rules address government measures as well as private conduct.

The Treatys opening statement implies that the most fundamental objective of the Community can be understood as promotion of economic welfare and progress, with competition policy being one of several instruments for that purpose. Thus the Treaty text also includes policy goals that might be construed as inconsistent with promoting competitionalthough they are typically phrased in a way that implies there will be no need to choose. For example, Community institutions are to support industry competitiveness, in part by encouraging an environment favourable to cooperation; however, this provision is not to be taken as a basis for any Community measure that could lead to adistortion of competition. (Article 157)

The goal of promoting market integration was important when the common market was still being established. Where industries were traditionally established within national markets, the challenge was to get them to transcend those boundaries. The market integration goal explains the emphasis on vertical relationships and intellectual property rights, which were seen as obstacles to crossborder trade. That goal was the link in the Commissions partnership with the ECJ.

The Treatys competition rules address conduct that may affect trade between the Member States. If an agreement as a whole is capable of affecting trade, Community law applies to all of it, including parts that individually do not affect trade, and to all of the parties, including ones whose individual contribution to that effect would be insignificant. The jurisdictional test can be met based on the expected effects of potential competition and on positive as well as negative changes in patterns of trade. The further requirement that the effect be appreciable limits this broad scope, though. The EU Guidelines define what is not considered appreciable based on cumulative thresholds of market share (5%) and aggregate turnover (EUR 40 million). In general, these levels define a rebuttable negative presumption. Below those levels, the enforcer has the burden of showing that there is nonetheless an appreciable effect on trade in order to establish Community law jurisdiction over an agreement or practice.

For agreements that by their very nature would affect trade, though, each threshold establishes a positive rebuttable presumption: above those levels, parties to such agreements have the burden of showing that there is nonetheless no appreciable effect on trade in order to avoid Community law jurisdiction. Virtually any agreement that controls imports or exports or that is implemented in more than one Member State would be covered by the positive presumption. There is also a clear presumption that an agreement or practice that covers all of a Member State meets the affecting trade requirement, because a cartel or abusive practice of such wide scope would necessarily affect the competitive prospects of potential competitors from outside. More judgment is called for in dealing with agreements or practices that apply only in a part of a single Member State.

Community law applies to entities that are undertakings, determined by function. This defined term is interpreted broadly, based on the nature of activities rather than formal structure. It excludes the sovereign functions of a state, but it includes the states commercial activities. It thus includes state bodies engaged in commerce, nationalised industries, municipalities, trade associations, private individuals, co-operatives and associations. Economic activity, not profit, is the important element. Some close cases have involved social insurance funds. Factors that may lead to finding that such a fund is not an undertaking include whether it is compulsory and motivated by solidarity or redistribution. On the other hand, it is more likely to be considered an undertaking if it is potentially in competition with similar private-business entities.

A standard analysis for defining markets is used for the tests based on market power. The Commissions 1997 Notice on the definition of the relevant market for the purposes of Community competition law lays out systematically the considerations used to identify product and geographic markets. Legal criteria come from the enforcement and merger regulations. A relevant product market comprises all those products or services which are regarded as interchangeable or substitutable by the consumer, by reason of the products characteristics, their prices and their intended use. The relevant geographic market comprises the area in which the undertakings concerned are involved in the supply and demand of products or services, in which the conditions of competition are sufficiently homogeneous and which can be distinguished from neighbouring areas because the conditions of competition are appreciably different in those areas. The method relies principally on demand substitutability, which is described as the most immediate and effective disciplinary force on the suppliers of a given product. This is tested by examining the markets likely reaction to a 5-10% permanent relative price increase. Supply substitutability may also be taken into account where its effects are equivalent to those of demand substitution.

Potential competition is not taken into account when defining markets, but may be considered in the competitive assessment. The guidance recognises that the market definition analysis might depend on the nature of the competition issue under examination. The focus in merger cases is prospective, anticipating market conditions in the future. By contrast, in dealing with restraints or abuse of dominance, the focus could be on present or past conditions, and it could include an examination of whether the conduct has affected conditions in the market and thus the evidence used to define it.

Anticompetitive agreements are prohibited and void, unless exempted. Article 81(1) prohibits agreements that have the object or effect of preventing, restricting, or distorting competition. The prohibition extends to decisions by trade associations and to concerted practices. These terms are interpreted broadly, to include arrangements that are not legally enforceable contracts. The alternative characterisations, about the nature of the object or of the effect, also invite broad interpretation. But there are limits: for example, a suppliers arrangements with its customers may not be treated as a set of agreements if the customers had not acquiesced in the suppliers program.

Small-scale agreements are not usually considered to be likely to affect competition. In a series of Notices issued since 1970, the Commission has limited the scope of the Article 81 prohibition by describing transactions that are likely to be too small to have appreciable effects. The latest such de minimis Notice sets thresholds based on market share at 10% for agreements between competitors and 15% for agreements between non-competitors. Where there are parallel networks of similar agreements in a market, the threshold is lower, at 5%. The Notice also implies a collective threshold for that situation, stating that cumulative foreclosure is unlikely if less than 30% of the market is tied up in such parallel networks.

Regardless of market share, hard-core agreements cannot benefit from de minimis treatment. The Notices definitions of such hard-core agreements are the same ones used in the block exemption regulations and guidelines about vertical and horizontal agreements. The rationale for de minimis treatment is that competition concerns are unlikely if companies do not have a minimum degree of market power.

Economic benefits from an agreement can lead to exemption from the prohibition. Under Article 81(3), an agreement that would otherwise be prohibited may nonetheless be permitted, if it improves production or distribution or promotes technical or economic progress and allows consumers a fair share of the benefit, imposes only such restrictions as are indispensable to attaining the beneficial objectives, and does not permit the elimination of competition for a substantial part of the products in question. Improvements in productive efficiency obviously can be considered. Promoting progress could also include prospects for innovation that may be less immediately tangible. Neither efficiency nor progress implies a broad social balancing of economic advantages and disadvantages. Goals pursued by other Treaty provisions can be taken into account to the extent that they can be subsumed under the four conditions of Article 81(3). Guidelines from the Commission acknowledge that applying the 2 parts of Article 81 is a balancing process, seeking to identify the net economic effect of the restriction and the efficiencies. An increase in market power increases firms ability and incentive to raise price, but cost efficiencies may permit them to reduce price. In balancing these potentially opposite effects, the requirement that benefits be passed on to consumers results in a sliding scale.

2.2 Horizontal agreements

The Treaty specifies some of the horizontal agreements it prohibits. The listing, which is not exclusive, includes direct or indirect fixing of prices or trading conditions, limitation or control of production, markets, investment or technical development and sharing of markets or suppliers. Decisions have clarified what else Article 81 prohibits. All forms of agreements to divide markets and control prices, including profit pooling and mark-up agreements and private fair trade practice rules, are prohibited. Exclusionary devices such as aggregate rebate cartels are prohibited even if they make some allowance for dealings with third parties.

Joint purchasing and selling are permitted in some market conditions. Exchange of price information is permitted only after enough time has passed, and only if the exchange does not permit identification of particular enterprises. Something close to a per se rule can be used against hard-core conduct. It is not necessary to prove that price fixing, market division or output limits or quotas actually raised prices or reduced output. Decisions have made clear that those effects are presumed, and parties to the agreements cannot overcome the presumption by claiming they had no intention or capacity to achieve an anti-competitive effect. The decision to fix prices is enough to establish the infringement.

To set the fine, though, factors in addition to the nature of the infringement could also be relevant to showing the gravity of the offence. These factors could include the cartels geographic scope and its impact, if that can be measured. Showing implementation of the cartel agreement would not require evidence of marketplace effect; for example, it would be enough to show that the colluders announced agreed price increases or met to monitor compliance. For hard-core infringements, the ECJ has agreed that factors related to intent may be treated as more significant than those related to effects. Agreements that are tacit or undocumented may be prohibited as anticompetitive concerted practices. This term covers co-operative activity short of explicit agreement, which the ECJ has described as co-ordination between enterprises, that had not yet reached the point of true contract relationship but which had in practice substituted co-operation for the risks of competition. The term can include hard-core restraints. Cases about concerted practices typically look for evidence of arrangements for reaching and enforcing compliance with implied agreements to limit competition. The Commission has applied the concept to conduct that some other enforcers would treat as ordinary agreements, such as formal industry-wide market-sharing arrangements set up by trade associations. In this role, the term fills a conceptual gap in a legal tradition that privileges documentary formalism. But pure oligopoly interdependence would not be a prohibited concerted practice.

Intentional communication and awareness are needed, not just mutual awareness of the benefit of restraint. Because Article 81 explicitly applies to decisions of trade associations, Community law can deal with this common setting for cartel agreements in a straightforward way. There is no need to infer constructive agreements or resort to theories of collective dominance. The decisions that Article 81 prohibits can include an associations formal rules or its more informal actions or recommendations. Where the infringement is attributed to the association, the Commission may nevertheless take into account the sum of the members turnovers in calculating the associations fine. Under some conditions, the members may be liable for the payment of the fine by the association. The Commission will obviously go after the members for their own behaviour.

In its application to non-hard core agreements, Article 81 resembles a rule of reason. The Article 81(1) prohibition and the Article 81(3) exemption criteria require market analysis and balancing of positive and negative effects in cases of horizontal co-operation that does not amount to a hard-core cartel. The Commission has issued block exemption regulations about agreements for production specialisation and for research and development, accompanied by guidelines to show how Community law can allow competitor collaboration where it contributes to economic welfare without creating a risk for competition. The case law is not entirely consistent about whether the assessment of market conditions and consideration of potentially competing policies and effects are part of determining whether Article 81(1) prohibits an agreement or of determining whether Article 81(3) exempts it.

For non-hard core agreements, the ECJ has said that market conditions, market structure and the economic context determine whether Article 81(1) prohibits it. For a hard-core agreement, those considerations might come up, if at all, only to determine whether efficiency claims under Article 81(3) exempts it. The Commissions Guidelines state that the four conditions for exemption are exhaustive, so no other grounds can be invoked. (Guidelines on the application of Article 81(3), para. 42). Determining under Article 81(1) whether the agreement could restrict competition involves determination of the nature of the agreement, definition of markets, and evaluation of market structure and market power, including considerations such as the nature of the products, market concentration, barriers to entry, stability of shares and the countervailing power of buyers or suppliers. The horizontal guidelines presume that if parties have a low combined market share, co-operation is not likely to restrict competition. The guidelines do not prescribe a single rule, because conditions and effects can vary significantly, but they do suggest particular levels for some kinds of agreement.

For agreements about joint purchasing and commercialisation (that is, selling, distribution and promotion), the guidelines set a safe harbour at a combined market share of 15%. High market shares will not necessarily be a concern for standardisation agreements, and the assessment focuses more on whether the standards could raise barriers to entry. No market share test is needed for agreements that, because of their very nature, are unlikely to reduce competition. This could be the case, for example, if the parties could not carry out a project independently at all, or if their agreement is about an activity that does not affect any relevant parameter of competition. At the other extreme, no market share threshold applies to the hardcore restrictions of price-fixing, output limitation and allocation of markets or customers, which are generally prohibited irrespective of the parties market shares.

Cooperation horizontal agreements

The guidelines then explain the application of the cumulative Article 81(3) criteria about economic benefits and its caveats about sharing the benefit with consumers, indispensability of the restraint to achieving the benefit and not eliminating competition by dominating the market.

The block exemption regulation about specialisation treats production rationalisation in the same way that the guidelines treat similar practices which are not covered by the exemption. For agreements between competitors to specialise production, the regulation sets a market share threshold for exemption at 20% (for all parties combined, and subject to certain conditions, including the absence of hard-core restraints)

The block exemption regulation for research and development is generous, particularly to innovation that promises to create new markets. Agreements between competitors about research and development are exempt up to a combined market share of 25% (of the market for the product that is the subject of the joint research), subject to certain conditions and the absence of hard-core restraints. Here too, the guidelines about the application of the block exemption regulation use the same level to presume the lack of effect on competition and to explain why benefits would be presumed to outweigh harm to competition. If the collaboration is developing something for which there is not yet any market, the guidelines recognise that a successful first-mover effort should not necessarily be seen as an elimination of competition, even though it could lead to huge initial market shares once the product is developed. Thus, the block exemption permits such agreements to continue regardless of market share for the first 7 years after the product comes to market. At that point, the safe harbour of 25% applies.

Enforcement of Article 81 developed in an environment that had tolerated formal industry co-operation, sometimes amounting to self-regulation. Early cartel cases targeted national-scale industry associations and substantial international agreements about quinine, dyes, aluminium and chemicals. Some cartels had formal committees that kept minutes of their agreements. For other cases, obviously coordinated market actions could support an inference of agreement. Collection and exchange of information about price and output through trade associations has been a concern, as a means of tacit or even explicit co-ordination to confirm and police agreements. By the early 1990s, some cartel cases had resulted in total fines of over EUR 100 million. As the fines for price fixing mounted, Commission enforcement strategy began relying increasingly on insider evidence supplied by firms seeking clemency. This was formalised in the Commissions 1996 leniency notice, publicly announcing what had been an unofficial practice.

The level of enforcement against horizontal cartels has sharply increased since 2001. The Commission has issued an average of about 8 decisions per year, compared to fewer than 2 per year over the previous decades. The Commissions notice about setting fines treats hard-core cartels as very serious infringements, for which the fine, determined by gravity, would normally be at least EUR 20 million (before considering other factors). The fines imposed in these recent cases confirm that treatment. In 31 Commission cartel decisions since 2001, the fines totalled EUR 4 billion. The peak was in 2001, with EUR 1 836 million. The level then dipped, and in 2004 the total was about EUR 390 million. Despite the increased activity, the fines being imposed may not yet be enough to deter hard-core infringements.

2.3 Vertical agreements

Most restraints in agreements about supply and distribution are permitted, unless there is market power. The 1999 block exemption regulation for vertical cooperation agreements restated the Commissions analysis of these restraints. Recognising that parties typically enter agreements to manage the distribution chain in order to improve efficiency and that smaller-scale agreements are unlikely to affect competition either upstream or downstream, the regulation applies a market share screen. It exempts most agreements involving a supplier or buyer with a market share under 30%, considering that to be a level at below which vertical agreements would be expected to lead to an improvement in production or distribution and allow consumers a fair share of the resulting benefits (as long as they do not include certain particularly harmful restraints).

The buyers share of its downstream market is considered where the contract requires the supplier to sell exclusively to that buyer. Vertical agreements involving associations of retailers are also exempted, as long as the turnover of each is below EUR 50 million. Where parallel, similar networks of agreements account for more than 50% of a market, the Commission reserves the power to disapply the exemption, on 6 months notice; it has not yet done so, though. There is no presumption of violation where an agreement involves a supplier with a share over 30%. But with increasing market power come increasing concerns that agreements might impair competition, by foreclosing other suppliers, raising barriers to entry or reducing interbrand competition and facilitating collusion. The regulation also notes possible concerns about reducing intrabrand competition and about creating obstacles to market integration. The regulation applies to many types of agreements, replacing previous notices and regulations about topics such as exclusive distribution and franchising.

Unlike previous regulations of similar topics, it does not contain a list of permitted clauses, an approach which had tended toward uniformity of practice out of a fear that whatever was not permitted would be prohibited. The vertical block exemption regulation takes the opposite position: below the 30% threshold, whatever is not prohibited is permitted. Fixing minimum resale prices and excessive territorial protection remain blacklisted, regardless of low market share. Resale price maintenance has always been treated as a per se infringement, at least with respect to minimum prices; however, recommending a resale price or requiring resellers to respect a maximum resale price are exempted, up to the 30% market share threshold, provided that the result is not a fixed or minimum sale price due to pressure from, or incentives offered by, the supplier. Territorial resale constraints are suspect, but some are permitted, to protect systems of exclusive dealership, preserve functional distinctions between wholesalers and retailers or prevent resale of components leading to competition with the supplier. The latest regulation followed a series of infringement actions that imposed fines totalling EUR 276 million. At the time (1998), the EUR 102 million fine against Volkswagen was the largest fine the Commission had ever imposed against an anticompetitive restraint. (The CFI later reduced this fine to EUR 90 million.) The experience persuaded the Commission to take a tougher line in this sector than the general vertical restraints regulation. The black list of forbidden clauses is unusually long and detailed. Manufacturers do not benefit from the block exemption if they do not allow their authorised dealers to sell competing brands or if they limit the dealers ability to open secondary outlets in other territories. The regulation applies the same market-share safe-harbour, of 30% (and 40% for a selective distribution system with a limited number of distributors), and it permits suppliers to use an exclusive system, but only if the system imposes no constraints on making passive sales to customers in other areas.

2.4 Abuse of dominance

Curbing abuses by firms that dominate markets and suppress competitors or harm consumers is the other main subject of Community antitrust law. Article 82 prohibits the abuse of a dominant position. Some acts that the Treaty lists as abuse are imposing unfair purchase or selling prices or trading conditions (either directly or indirectly), limiting production, markets, or technological development in ways that harm consumers, discrimination that places trading parties at a competitive disadvantage and imposing non-germane contract conditions. Other kinds of conduct by a dominant firm that disadvantage other parties in the market could also be abuses. Practices such as loyalty rebates that would not be objectionable when done by a firm without market power could be considered abuses when done by a firm in a dominant position.

To find infringement it is not necessary to show that an abusive practice produced an actual anticompetitive effect; it is enough that the conduct, when undertaken by a dominant firm, tends to restrict competition or is capable of having or likely to have that effect. There is no provision for exemption, although the case law has developed a doctrine that otherwise abusive conduct is not prohibited under Article 82 if it is objectively justified.

Dominance is a broader concept than economic market power over price. It is not the same as economic monopoly, although a monopoly would clearly be dominant. Dominance is often presumed at market shares over 50%, and it may be found at lower levels depending on other factors. The ECJs Hoffman-LaRoche (1979) and United Brands (1978) judgments explained the meaning of dominance under the Treaty, describing it as a position of economic strength enjoyed by an undertaking which enables it to prevent effective competition being maintained on

the relevant market by affording it the power to behave to an appreciable extent independently of its competitors, its customers and ultimately of the consumers. These still-authoritative judgments found dominance based on features such as vertical integration, because that enabled a firm to act independently of its suppliers of intermediate services, and at market shares of 40-45%. In current practice, there appears to be a safe harbour at a market share of about 25% and a rebuttable presumption of dominance at about 40-50%. Dominance depends on factors other than market share, such as the number and relative size of other firms and the conditions of entry.

A finding of dominance is more likely if entry is difficult or if there are no other firms of comparable size or with the capacity to counter the leaders strategies. Article 82 is not limited to single-firm misconduct. Under the theory of collective or joint dominance, several firms can share and abuse a dominant position. When the Commission first tried to apply Article 82 to oligopoly, the Court of First instance rejected the argument (although the court upheld the finding of infringement under Article 81 as a restrictive agreement). To find that several firms together hold a dominant position, the court demanded that the firms be united by economic links, such as a network of interdependent intellectual property licenses.

A formal cartel could possess collective dominance. Judgments applying the analogous language about dominance in the Merger Regulation, which have been cited as authority for the application of Article 82, imply that oligopoly interdependence might amount to collective dominance if the members could monitor each other effectively, if retaliation against defectors was credible enough to provide each member with an incentive to maintain co-ordination and if customer and consumer responses would not undermine co-ordination.

Exploitation of market power by charging high prices could be an abuse, although no final decision has actually condemned it. In the 1970s, reviewing Courts annulled 2 Commission decisions that had challenged high prices as abuses. The early Court judgments about abuse of dominance agreed with the principle, though, that unfair pricing could include setting prices to take advantage of market or monopoly power. Prices would be too high if they bore no reasonable relation to the economic value of the product. The ECJ has also endorsed the concept that Article 82 could prohibit prices for licensing intellectual property rights that are particularly high and not justified by the facts. Despite the theoretical support for the sweeping principle, in the end the decisions have declined to fiind infringements based upon the evidence in the cases.

For predation, anticompetitive intent can be inferred from prices too low to recover costs; the strength of the inference depends on the cost reference. Prices below average variable cost are presumed to be predatory, that is, intended to eliminate competitors. Prices above that level that do not recover total costs may also be treated as predatory, but that conclusion may depend on further evidence of the intent to eliminate or prevent competition.

Discrimination is included among the abuses listed in the Treaty text, which prohibits putting other parties at a competitive disadvantage through the application of dissimilar conditions to equivalent transactions. This language implies some common principles of competition laws about discrimination, notably that the discrimination must be between transactions involving the same product or service and that it must actually cause competitive harm. The harm could be to an individual competitor, though, and not necessarily to competitive conditions in the market. Price variations that correspond to different national markets have been treated as prohibited discriminations, in the absence of objective justification in terms of differences in costs or in the extent of the sellers exposure to different risks in different markets. Decisions have prohibited discriminations about inputs that impose a price squeeze on non-integrated competitors or that unfairly favour national-champion incumbents.

Putting pressure on customers to enter into requirements contracts can be an abuse. Many cases have dealt with the foreclosure effects of loyalty rebates. Loyalty programs may be permitted if they are based on cost efficiencies. Only cost savings to the supplier can justify quantity-based rebates. Indeed, recent decisions imply that the only way to avoid liability is to base the loyalty scheme strictly on cost differences. Refusal to supply a customer can be an abuse, particularly if the customer is a long-time regular trading partner, but a reasonable justification, such as the customers poor credit, can overcome the prohibition. The prohibition of tying is implied in the Treaty text, which prohibits imposing non-germane conditions in contracts. The elements of tying, or refusing to supply a product unless the customer also takes another, allegedly unnecessary one, are dominance in the tying product, a separate tied product, coercion to take them together, an anti-competitive effect in the market for the tied product and the absence of an objective, proportionate justification for the tie. Bundling, which can make the price of the tied product effectively 0, can be construed as coercion. The most prominent application of Article 82 to tying tactics is the Commissions 2004 action against Microsoft.

Refusal to licence intellectual property, under some circumstances, can violate Article 82. The leading case reaching this result emphasised that the refusal prevented the production and marketing of a new product for which there was a potential consumer demand. The ECJ later refined the conditions for finding liability. The firm seeking the license must not be intending essentially to duplicate what the owner of the right already offers, there must be a potential consumer demand for the firms product, the refusal to license must not be justified by objective considerations; and the refusal must prevent, to the detriment of consumers, the development of a market in which the licence is an indispensable input.

A wide range of remedies is available to correct and deter abuse of dominance. The new enforcement regulation has authorised the Commission to order structural relief, which could include divestiture but might also include other dispositions of property rights. These measures must be proportionate. The Commission may only use structural measures to correct abuse of dominance where there is no equally effective behavioural remedy, or any behavioural remedy would be more burdensome to the entity that will be the object of the structural remedy. The new

enforcement regulation also confirmed that the Commission can order interim relief, which is particularly significant in cases about access. Behavioural orders and financial sanctions remain the principal tools. Substantial recent fines include EUR 13 million for exclusionary pricing in telecoms and EUR 24 million for loyalty rebates. The fine against Microsoft shows that Community sanctions against abuse of dominance can be vigorous; at EUR 497 million, it exceeds the total fines the Commission imposed against horizontal cartels in 2004.

2.5 Mergers

The inclusive legal standard for merger control can deal with all kinds of competitive effects. The Commission may prevent or correct a merger that would significantly impede effective competition in particular as a result of the creation or strengthening of a dominant position. This substantive standard is subsidiary to the Regulations fundamental criterion, whether the transaction is compatible with the common market. The 2004 revision of the Merger Regulation revised the original 1989 standard. The principal issue motivating the change was non-coordinated effects in oligopoly markets, where the merged firm might have market power without necessarily having an appreciably larger market share than the next competitor.

The guidelines presume that a merger does not impede effective competition if the new entitys market share would not exceed 25%; however, this presumption does not apply to coordinated effects, where the merged entity would be collectively dominant along with other third parties. Regardless of these levels, though, the guidelines warn that special attention will be paid if any party has a pre-merger share over 50%, or if there are obvious issues of potential or toe-hold entry, innovation, cross-shareholding, maverick market behaviour or indications of oligopoly behaviour in the industry. The guidelines discuss in detail the theories of non-coordinated and co-ordinated anti-competitive effect. Where non-coordinated effects are the concern, an important indicator where products are differentiated can be the closeness of substitution between the merging firms products, for homogeneous products, an important factor is the relative capacities of the merging firms and their rivals. A market share over 50% and a significant market share advantage over any rival may be a strong indication that the merger would create or strengthen a dominant position. Where co-ordinated effects are the concern, the guidelines describe the conditions for finding that a merger will create or strengthen a position of collective dominance.

Countervailing factors include buyer power and entry. Whether significant entry is likely is determined by inquiring whether an entrant would find it profitable to do so in post-acquisition market conditions. Entry must not only be likely but also sufficient and timely. The measure of timeliness could vary in different product markets, but the normal test is 2 years.

Efficiencies can also be a mitigating factor, if they are merger specific, timely, verifiable, and benefit consumers. Reductions in variable or marginal costs are more likely to lead to lower prices, and hence they would get more weight than savings in fixed costs. The guidelines disavow an efficiency offence, that is, that increases in productive efficiency, giving the merged firm a cost advantage over rivals, will be a reason to reject a merger.

If one of the parties is financially failing, the guidelines would permit an otherwise anti-competitive merger. The rationale is that the competitive structure would deteriorate equally absent the merger. The parties must show that the allegedly failing firm would in the near future be forced out of the market by financial difficulties if not taken over by another firm, that there is no less anticompetitive alternative acquirer, and that in the absence of a merger the assets of the failing firm would inevitably exit the market.

The merger regulation and the guidelines do not call for considering policies other than effects on competition. Efficiencies are taken into account as part of the competitive assessment. Member States could invoke these principles to block or regulate transactions that do not impede competition; however, they could not invoke them to authorise a transaction that the Commission has blocked.

A concentration cannot be put into effect before it is notified to the Commission and the Commission has cleared it. This merger control power applies only to transactions that are large enough to have a Community dimension. That status and the associated obligation to notify the Commission in advance are defined in terms of turnover, in total and within the Community. A transaction has a Community dimension when the combined aggregate worldwide annual turnover of all of the firms involved is more than EUR 5 billion, and the aggregate turnover within the EEA of each of (at least two) of them is more than EUR 250 million. An alternative definition captures certain transactions that have significant effects in several Member States. If the merging firms concentrate their Community business in a single Member State (with each of them having more than two-thirds of its Community turnover there), then the merger does not have Community dimension and the national competition authorities are responsible for it. There is also now a discretionary process for avoiding multiple national filings and reviews. If a merger

may have to be reviewed in 3 or more Member States, the merging firms can request that the Commission examine the merger, which it will do if none of the Member States objects. Merger control is more like a formal approval than a simple notification. The merger regulation, the Commissions implementing regulation and its best practice guidelines set out the process. It begins with informal contacts with DG Comp staff, including submission of a briefing memorandum and draft notification documents, before any formal filing is made. The formal process begins with submission of a detailed notification describing the transaction, its motivations, markets affected, market shares, and conditions of supply, entry and exit, and considerable other documentation. The Commission may use all of its other powers to get further information from the merging parties and from others. The Commissions decision process for mergers is similar to the process for other competition matters, except that merger decisions are subject to strict deadlines. The deadlines are now set in terms of working days. In the first phase, the issue is whether to clear the transaction or to open a second phase investigation; the deadline to finish the first phase is 25 days; if the matter continues to the second phase investigation and decision, the deadline is 90 days from the beginning of the second phase; if the parties offer modifications to deal with competition concerns, the first phase can be extended to 35 days, and the second phase, to 105 days.

3. Institutional issues: enforcement structure and practices

The administrative process for applying the law is adapting to strengthen investigative powers and better incorporate economic concepts and evidence in decision-making, in order to convince the courts while maintaining policy consistency in a system of decentralised enforcement.

3.1 Competition policy institutions

The European Commission, as the Communitys executive body, implements its competition policy. By Treaty, the Commissioners shall, in the general interest of the Community, be completely independent in the performance of their duties, neither seeking nor taking instructions from a government or anyone else, and refraining from any action incompatible with their duties. Member governments undertake to respect the principle and not to seek to influence members.

In the Commission staff, the Directorate-General for Competition is principally responsible for competition policy and enforcement. Headed by a Director General who is a career Community manager, DG Comp is organised into 10 directorates, for management, antitrust and merger policy, cartel enforcement, sectoral expertise (4 directorates) and state aid (3 directorates, including one for policy). DG Comps complement has been stable for several years at just over 600 permanent staff. In addition, DG Comp relies on contract and auxiliary personnel and experts seconded from Member State competition agencies.

3.2 Enforcement processes and powers

The Commission uses broadly similar basic procedures and investigative tools for dealing ex post with infringements of Articles 81 and 82 and for decisions about notified mergers. A complaint or merger notification, or a decision to start a procedure at the Commissions own initiative, is followed by an investigation by DG Comp, managed by a case handler. The evidence and proposed remedy are presented to the respondent in a statement of objections. The respondent has a right of access to the investigative file and an opportunity to reply, in writing and at an oral hearing. The decision is taken by the Commission, on a recommendation of the Competition Commissioner. The Commission no longer processes applications for individual exemption or negative clearance, because it no longer has the power to issue them. The enforcement regulation provides for the possibility of a guidance letter, but it sets conditions to discourage routine requests. Now that the historically important system of notification and exemption has been eliminated, the legal criteria describing what is prohibited and what qualifies for exemption apply

directly. In effect, this approach puts a burden on companies to evaluate their agreements accurately, as they are at risk for making a mistake.

A complaint is submitted on a prescribed form. Comprehensive background information and documentation are required. The complainant is asked to detail the factual basis of the claim, supply information about markets and market shares, submit documents and statistics relating to the complaint, give names of persons who could testify about it, explain the complainants legitimate interest in the matter and specify the relief sought. Investigative tools and powers deal principally with documentary information, since Commission procedures rely heavily on documentary evidence. A request for information may be a simple request or a request by decision. Each will state the basis and purpose of the request, specify the information requested and the deadline, and indicate the consequences of incorrect or misleading response. There is no penalty for failing to respond to a simple request, although a company can be fined if its responses are false or misleading; by contrast, a company risks substantial fines for failure to respond to a request by decision. The Commission now has a limited additional power to interview persons during investigations.

Providing false or misleading information in an investigation, or failing to provide complete and accurate responses within the time set by a request by decision, may result in fines of to 1% of turnover. Daily periodic penalties of up to 5% of average daily turnover may also be imposed to compel complete and accurate responses to a request by decision.

Advisory Committees composed of officials from Member State competition agencies play a role in both antitrust and merger matters. These committees meet regularly and submit opinions about proposed decisions. Their opinions about mergers are appended to the decision and published. Their opinions about other cases are appended to the decisions and may be published if the Committee recommends it. Advisory Committee views do not control DG Comps recommendation to the Commission. But the consultation process is a valuable avenue for achieving general consensus, and DG Comp takes the views seriously.

The draft decision is prepared by the DG Comp case-handler (or team of casehandlers) and reviewed by DG Comp management. It is also reviewed by the decision scrutiny unit. The Competition Commissioner consults with the Legal Service and the Advisory Committee from the Member States before proposing a decision to the Commission. There are several other means for providing a measure of quality control over the staffs recommendation, in addition to the possibility of n internal scrutiny panel.

Application of fines and sanctions

The principal sanction for substantive infringements, set out in the Council regulation on enforcement, is an administrative fine against the infringing undertaking. This fine can be as high as 10% of the undertakings global annual turnover. The regulation does not authorise fines against individuals. The process begins with a basic amount, which is then adjusted upwards for aggravating circumstances and downwards for attenuating circumstances. The basic amount depends on the gravity of the infringement (which includes its nature and impact (if measurable) and the size of the market) and its duration. The guideline recognises three classifications in terms of the gravity of the infringement: minor, calling for a fine between EUR 1 000 and EUR 1 million; serious, up to EUR 20 million; and very serious, over EUR 20 million. These approximate levels are not fixed. That amount can then be increased to account for duration, up to 50% for medium duration infringements and up to 10% per year for longer ones. The aggravating circumstances could include repeat offences, refusal to co-operate, leadership in the violation, retaliation against other firms and the need to set the penalty greater than the gain. The attenuating circumstances could include with a passive role in the violation, non-implementation, termination upon Commission intervention, reasonable doubt about whether the conduct was an

infringement, negligence and co-operation with the Commission (outside of the leniency programme). The enforcement regulation now makes clear that the Commission can accept binding commitments to correct infringements and that the Commission can impose a fine for failure to comply with these commitments; however, commitments are not to be accepted in cases where the Commission intended to impose a fine in the first place.

The Commission has had a formal leniency program since 1996. The first undertaking that comes forward can receive full immunity from fines. If immunity has already been granted, or if the Commission already has enough evidence to find an infringement, reductions of fines remain possible for companies that provide significant added value to the Commissions case. For the second firm to come in, the fine could be reduced by 30-50%; for the third, 20-30%; for others, no more than 20%. There are conditions: the applicant must co-operate throughout the Commissions proceeding and end its involvement in the cartel, and it may not have coerced others to participate in the violation. Applications continue to increase. There were 16 in 2003 and 29 in 2004.

3.3 Judicial review

Commission decisions are subject to oversight by the two European courts. The ECJ, which was established in the original ECSC, ensures enforcement against Member States, decides disputes between the Community and Member States (and between Community institutions) and ensures uniform interpretation of Community law by deciding questions referred to it by national courts. The CFI was created to reduce the ECJ workload and backlog by dealing with the cases with no political or constitutional importance and those involving complex facts. The ECJ can review CFI judgments, but only on matters of law.

3.4 Other means of applying EU competition law

Private parties can sue in national courts, under national procedures, for relief from infringements of Community competition law. The usual Commission infringement proceeding has elements of privately-initiated litigation. An advantage of taking the trouble to meet the demands of a formal complaint at the Commission is that complainant gains party standing, including the right to participate to some extent in the proceedings at the Commission and to seek judicial review if the action is unfavourable.

The Commission has long tried to encourage resort to private suits in national courts and issued a

notice on co-operation with national courts to call attention to the option and provide guidance. Notably, national judiciaries must make available to claimants under Community competition law all remedies and procedures, such as injunctions and compensation, that are available to claimants under analogous or related national laws. DG Comp is again trying to promote more private enforcement, to empower those who are the object of infringements of the law..

Member State competition agencies and courts can apply Community substantive law. All Member States have taken the necessary national legislative steps to be sure that national institutions have the power to apply Community law.

Indeed, national authorities now have an obligation to apply Community law if the jurisdictional requirement of Community effect is met. National authorities can, under Community law, order cessation of an infringement, order interim measures, accept commitments, impose fines and make negative determinations that there are no grounds for action.

The informal European Competition Network (ECN) is the medium for facilitating inter-agency co-ordination. The ECN is conceptually and functionally distinct from the Advisory Committee that must be consulted about Commission enforcement proposals. It is also distinct from the association of European national competition agencies, which has been in existence for several years (and which includes agencies in countries that are not Member States of the EU). The basis for the ECN is in the enforcement regulation requirements for consultation and provisions about information exchange and case allocation. The network is mentioned in the recitals of the enforcement regulation, but the ECN has no legal status. The notice on co-operation explains how it will work, to allocate cases, handle information and ensure consistency. A joint declaration in the modernisation package describes the agencies non-binding expectations about co-operation.

Normally, a case would be handled by the authority of the Member State that is most affected, and others would stay their own proceedings. The Commission would normally deal with matters that substantially affect more than 3 Members or where its intervention is appropriate in order to ensure efficient enforcement or to set policy. Where a national authority is already acting on a case, the Commission would only exercise its power to take over if it looks like it will be in conflict with others or with established Community case law, if it is taking too long, or if similar

problems are appearing widely and a Community decision is needed to clarify Community policy. The Commission would not normally adopt a decision in conflict with a national authoritys decision if it has been kept fully informed about the case. The ECN is also a forum for informal co-operation. The members hold occasional plenary meetings at the policy level. There are also working groups on topics such as leniency, transitional issues, sanctions and procedures, and Article 82, and sector subgroups about railways, electricity and insurance.

As agencies across the Community share enforcement responsibilities, complications will arise about parallel investigations and leniency applications in several jurisdictions. Exchange of experiences has led to the introduction of generally consistent leniency programs in the Member States. Programs are in place in 17 Members now, and programs are planned in several more. Enforcement could suffer it multiplicity and inconsistency deter firms from coming forward.

Differences among the programs are few, mostly about the obligation to stop the infringing conduct.

4. Limits of competition policy: exclusions and sectoral regimes

There is no general principle in the Treaties or in Community jurisprudence about how to deal with a conflict between the demands of competition law and those of other Community-level laws or official actions. Conflicts with Member States laws and official actions have been a much more important topic. The Member States capacity to use national law to prevent or restrain competition is limited. A Member State may constrain the freedom to set prices, if this is manifestly an exercise of government authority and not rubberstamping of a private agreement and if the power has not been removed by Community legislation. But business firms cannot usually defend their infringements of Article 81 or Article 82 by claiming that national regulation eliminated their freedom of competitive action.

Member States may adopt rules that constrain competition in order to promote some sectoral or other policy, as long as there was no contrary Community policy on the issue and other Treaty obligations are respected. Since those other Treaty obligations include market openness and non-discrimination, national measures that might affect trade would not be permitted. But the Treaty texts combine to disfavour a broad state-action exemption. Article 10 of the Treaty obliges Member States to take all appropriate measures to fulfil their obligations and to facilitate the Commissions task, and it requires that they abstain from any measures that could jeopardise attainment of Treaty objectives. One of those objectives is a system ensuring that competition in the internal market is not distorted. (Article 3(1)(g))

Labour is not subject to Community competition law. The Treaty covers labour separately, in Article 39. Employees are not undertakings; rather, they work for undertakings. Trade unions would be considered undertakings to the extent that they enter into commercial activities, but not when they are dealing with labour market issues. In general, agreements concluded in good faith on core subjects of collective bargaining, such as wages and working conditions, which do not directly affect third markets and third parties are beyond the scope of Community competition law.