European Union Case Law on Excessive Pricing: An Economic Assessment

Abstract

This paper examines the European Union’s case law on the prohibition of charging excessive pricing. Its cornerstone is the Court of Justice’s statement that a price is excessive if it bears no reasonable relation to its economic value under competitive conditions. This paper argues that economic value should correspond to the firm’s long-run average total costs, including incurred and opportunity costs, at the level of actual output. Accordingly, the paper concludes that the jurisprudence is based on sound economic principles, as it aims to protect consumers from the abuse of a dominant position, without significantly influencing investment decisions by allowing investors to earn a return on their investments commensurate with the risks they take. The main problem in sanctioning excessive pricing is practical: the difficulty of measuring economic costs. The case law considers a less cumbersome alternative: comparing the defendant’s price with the price of the same or comparable products in more competitive markets, but with obvious homologation complexities. Safeguards have therefore been put in place to minimize the risk of false convictions, including considering a price to be excessive only if it significantly exceeds costs and allowing defendants to rebut the competition authority’s findings with objective data.

I. INTRODUCTION

The purpose of this paper is to review, from an economic perspective, the application of the excessive pricing doctrine by the European Union’s competition authorities: the Court of Justice of the European Union (CJEU) and the European Commission (EC). Article 102 of the Treaty on the Functioning of the European Union (TFEU) prohibits abusive conduct by undertakings holding a dominant position on a given market in so far as it may affect trade between Member States. Abusive conduct includes the imposition of unfair purchase or selling prices. In the remainder of this paper, prices that are unfair because they are “unreasonably” high will be referred to as excessive prices, in contrast to the nomenclature of EU case law, which uses this term to refer to prices that are suspiciously high.¹

Between 1971 and 2021, the EU competition authorities ruled on 28 excessive pricing cases, of which eight emblematic decisions are analyzed in this paper². Three are CJEU judgements in response to questions referred for a preliminary ruling by national courts under Article 267(b) TFEU.³ These cases are Deutsche Grammophon in 1971,⁴ Lucazeau in 1989,⁵ and AKKA/LAA in 2017⁶. Four are decisions of either the CJEU or the EC, of which two are Prohibition Decisions (United Brands Company (UBC) in 1978⁷ and Deutsche Post AG DPAG in 2001⁸) under Article 7 of Regulation 1/2003⁹, and two are Commitment Decisions (Gazprom in 2018¹⁰ and Aspen in 2021¹¹) under Article 9(1) of Regulation 1/2003, which allows the EC to close an investigation into suspected infringements of competition law, other than cartels, if it considers that the commitments offered by the defendant meet its concerns. The eighth is the EC’s reply to a complaint (Scandlines in 2004¹²).

In only a few of these 28 cases is excessive pricing the only infringement (pure cases), as most of them are accompanied by ancillary exclusionary conduct. One possible reason for this is that Article 102 of the EC Treaty prohibits the abuse of a dominant position in the internal market in so far as it may affect trade between Member States. In the four decisions examined in this paper, the defendants’ price increase strategy was trade restrictive. Another explanation is the perceived greater difficulty in proving exploitative conduct than exclusionary conduct, discussed below.

According to EU jurisprudence, as articulated by the CJEU in the UBC case in 1978 and reaffirmed in subsequent CJEU and EC decisions, a price charged by a dominant firm is excessive if it bears no reasonable relation to its economic value under conditions of sufficient competition. This paper argues that such value should be equal to the firm’s long-run average total costs (LATC) at its output level.

Charging a price equal to the LATC satisfies the fundamental characteristic of perfectly competitive markets: zero economic profit, since the LATC includes both incurred (explicit) and opportunity (implicit) costs. The latter include, among others, taking into account the risk of the assets and the opportunity cost of land owned by the company,¹³ such as the land used by the port of Helsingborg to serve ferry operators in the Scandlines case.¹⁴ Moreover, the LATC should be that of an efficient version of the actual firm, since in perfectly competitive markets only efficient firms survive.

One method of calculating the LATC of a single-product company is to calculate the present value of all costs incurred over its lifetime (operating costs and investment costs less its residual value) plus opportunity costs, if any, and divide it by the present value of physical sales over its lifetime. Investment costs include, but are not limited to, research and development costs, equipment and construction costs, and start-up costs, most of which are incurred before operations commence. The discount rate used to calculate present values should be the company’s weighted average cost of capital,¹⁵ reflecting the risks of the investment, including the risk of failure and the risk of obsolescence of the assets.¹⁶

The economic value against which a company’s price should be compared therefore includes all of the company’s economic costs, but excludes any premium resulting from the exploitation of a dominant position.¹⁷ The EU’s excessive pricing jurisprudence is thus conceptually based on sound economic principles, as it aims to protect consumers from excessive prices without significantly distorting investment decisions by allowing investors to fully recover their economic costs, including opportunity costs.¹⁸ Moreover, LATC pricing would be fully “fair” in the sense that the consumer surplus would be 100 percent of the economic surplus, as the economic profits of the company would be zero,¹⁹ but it could still leave low-income consumers without access to the product.²⁰

Projecting costs and demand over the life of the project and estimating the cost of capital commensurate with the risks taken by the investor is a challenging task. One way around this difficulty used by EU competition authorities has been to compare the defendant’s price with the price of the same or comparable products in more competitive markets.²¹ The CJEU has held, in response to references for preliminary rulings, that it is possible, subject to certain conditions of objectivity, to compare the defendant’s price with the prices charged by other undertakings in other geographical areas for the same product/service (Deutsche Grammophon, Lucazeau and AKKA/LAA). This strategy, however, has its own homologation complexities, as it involves comparing prices from different time periods, products, or markets with different economic and legal environments.²²

In Aspen, the EC adds a third method: the “cost plus” method. In this approach, an estimate of the costs incurred by the company itself is increased by a ‘plus’ representing the profit margin. The latter is estimated as the median of the profit margins of a sample of similar companies in a competitive market, since any overpricing by the company would lead to an inflated profit margin. The difficulty of comparing prices from different markets explains the rationale for limiting the comparison to those components of economic value that are more difficult to estimate with the information available from the company.

The EU case law on excessive pricing does not distinguish between whether the dominant position was lawfully acquired and lawfully maintained. The infringement thus arises from the mere exploitation of customers by a dominant undertaking,²³ without any need for potential harm to the competitive process. The reason for this is that the impact of excessive pricing on consumer welfare does not depend on whether dominance was achieved by merit or by exclusionary conduct,²⁴ even though the exclusionary conduct may cause additional harm to consumers, whereas the process of achieving dominance by merit may benefit consumers.²⁵

Despite the clear doctrinal position of the EU competition authorities on excessive pricing, some authors have argued that the authorities have historically been reluctant to sanction this infringement, preferring to focus on exclusionary conduct.²⁶ That is, the authorities would have tended to protect consumers indirectly, by intervening against exclusionary conduct that may create market power, rather than directly, by prosecuting the abuse of market power itself,²⁷ which would leave purely exploitative cases unpunished. Some scholars argue that this predisposition of the authorities is due to the greater difficulty of proving excessive pricing compared to exclusionary conduct and not necessarily to doctrinal issues.²⁸ However, proving such theses would require a more detailed causal analysis, which is beyond the scope of this paper.

What is certain is the difficulty in determining the benchmark against which the actual price can be compared and the risk of false convictions. The recent upsurge in interest in prosecuting excessive pricing cases in the EU, as noted by some authors,²⁹ may reflect the confidence on the part of the competition authorities that the safeguards described below minimize, if not eliminate, the risk of false convictions and that the use of Commitment Decisions can lead to appropriate remedies that go beyond what the Commission could impose on its own, including structural remedies, such as the removal of relevant entry barriers and a stable supply obligation.³⁰

The remainder of this paper is structured as follows: Section II discusses the economic foundations of the excessive pricing doctrine and the practical obstacles to its application. Section III summarizes eight emblematic excessive pricing cases examined by the EU competition authorities, and the fourth section draws lessons from the analysis of these cases. The final section presents the conclusions.

II. CRITICISMS OF THE DOCTRINE

The doctrine of excessive pricing is intended to protect consumers,³¹ but it has nevertheless been highly controversial in the literature.³² The purpose of this section is to examine EU case law on excessive pricing in the light of the criticisms that the doctrine has received, both doctrinally and regarding its application.

The first doctrinal objection is that the control of allegedly high prices is incompatible with the functioning of a market economy, since prices are its primary method of organization. In particular, controlling prices would eliminate their function of signalling scarcity and thus prevent self-correcting market forces from operating.³³ Certainly, in perfectly competitive markets, where agents are price takers, a price that temporarily generates economic profits will encourage suppliers to expand and, more importantly, attract new competitors. In such circumstances, price control could discourage new entry and thereby potentially reduce the future intensity of competition.

In concentrated industries, however, the conditions may be different. First, the forces that led to concentration, such as economies of scale or the successful efforts of some firms, make it harder for new firms to enter. In addition, the possible strategic behaviour of incumbents increases the uncertainty faced by potential entrants by making it harder to predict the price that will prevail if they enter, thereby discouraging entry.³⁴ For example, an entrant who finds that the incumbent has lower costs is unlikely to enter, even if pre-entry prices are excessive.³⁵

Accordingly, in concentrated markets, where demand is relatively inelastic to price, it is possible for prices well above their economic costs to persist for a long time.³⁶ In such cases, the best remedy is to seek measures to promote competition rather than to control prices. Such measures may include lowering barriers to entry and introducing pro-competitive market rules that competition authorities can recommend to legislators and regulators.³⁷

However, when the industry is dominated by a single company protected by high and difficult-to-remove barriers to entry, abuses are more likely to occur, and, when they do, they tend to be protracted. Under these conditions, the possibility of self-correction in the absence of new entrants backed by significant technological innovation is quite low, and thus price control policies would be necessary to prevent welfare losses.

A second doctrinal argument against excessive pricing is that it could discourage investment. Several authors have pointed out that the ability to charge high prices provides the incentives to innovate and to invest in potentially highly profitable risky investments. Since investment benefits consumers by increasing the variety and quality of supply or reducing costs through innovation in production, it would be socially inefficient to try to correct high prices, even if they persist.³⁸

EU case law on excessive pricing should be immune to these criticisms because it is limited to companies in a dominant position and allows investors to earn a return on their investment commensurate with the risks they have taken. Moreover, the dominant position of a company that has achieved it by merit cannot be challenged by competition authorities and can earn a return on the additional investment commensurate with the scale and risk of its efforts.³⁹ These advantages should, in theory, be sufficient to attract investment.

The third argument against the excessive pricing doctrine arises from its practical application.⁴⁰ Given the complexity of estimating a benchmark against which to compare the allegedly excessive price charged by a dominant firm, which is analyzed in detail in Section IV, critics claim that the application of the doctrine could lead to false convictions. This legal uncertainty could, in turn, discourage investment.⁴¹ It would also undermine the rule of law, as companies would not know when they were engaging in anticompetitive behaviour, given the leeway in determining whether a price is excessive.⁴²

In this respect, some authors have drawn an analogy between excessive pricing and other antitrust infringements, such as predatory pricing and price squeeze, which also require courts to rely on complex economic analysis followed by a judgement on a vague standard.⁴³ However, there is no empirical evidence that competition authorities have been any less reluctant to pursue these less controversial but equally difficult to prosecute infringements.⁴⁴

In any event, the practical application of the excessive pricing doctrine in the EU is subject to a number of safeguards that reduce the risk of erroneous convictions. First, for a price to be considered excessive, it must exceed the benchmark by a significant margin,⁴⁵ so that there is little doubt that is abusive; second, the burden of proof is on the EC,⁴⁶ which must also prove negligence or intent on the part of the defendant;⁴⁷ third, the defendant may rebut the Commission’s preliminary report with objective data, but it shifts the burden of proof to itself. Alternatively, the defendant may offer commitments under Article 9(1) to meet the authority’s concerns and close the case without a finding of guilt.⁴⁸ The latter provision allows defendants who may not be aware that they have committed an abuse of a dominant position to avoid sanctions.

Although there is no definition of when the margin between the defendant’s price and the benchmark is significant, the determining factors should be the accuracy of the defendant’s cost data or the similarity between the defendant’s market and the markets used for comparison, as appropriate. The EU case law appropriately leaves it to the competition authorities to determine when a price/benchmark margin is excessive, rather than setting a threshold. This task requires competition authorities to make a judgement on the vague concept of excessiveness, but, as mentioned above, it is no different from the rule of reason analysis used to decide other antitrust cases.⁴⁹

However, there is one situation where it may be very difficult to establish a benchmark. If the defendant owns a unique, scarce resource that it uses to provide the good or service, it would be virtually impossible to determine its opportunity cost and thus the defendant’s LATC. A prime example of such a unique factor is entrepreneurial talent in start-up companies that develop inventions and breakthrough innovations. Such talent is rare and heterogeneous, making it difficult to determine its opportunity cost.

Competition authorities would therefore likely have to refrain from pursuing cases where the dominant undertaking uses a unique resource, unless the excessive pricing is accompanied by other anticompetitive behaviour, a comparable resource is traded on the market, or the competition authority has direct evidence of wrongdoing, such as internal correspondence.

Competition authorities could use their intrusive investigative powers, such as unannounced inspections and examination of the books and other records of undertakings, as allowed by Articles 20(2)(b) and 20(2)(c) of Regulation 1/2003, but this issue is rarely mentioned in the debate. In any case, the impact of excluding these cases from excessive price control is likely to be small, as they are likely to occur in industries with rapid technological progress where new innovative entrants can challenge the incumbents.⁵⁰

Another criticism of the doctrine is that it would turn competition authorities into de facto regulators, a task for which they are ill-prepared.⁵¹ However, their task is to decide whether the dominant undertaking has charged an excessive price in the period under review and not to determine the ‘fair’ price or to monitor the undertaking’s future behaviour beyond their normal market surveillance role to deter all types of antitrust infringements, in particular those committed by dominant undertakings such as exclusivity agreements, loyalty rebates, and price squeezes.

An exception would be where the authority imposes behavioural remedies that need to be monitored. To this end, the authority could impose obligations on the undertakings to facilitate compliance, such as an obligation to report regularly on their implementation. In addition, where the competition authority imposes remedies that involve complex technicalities that it is not well-equipped to monitor, it could appoint an independent trustee to carry out this task and report on noncompliance.⁵² Commitment Decisions under Article 9(1), which, unlike Prohibition Decisions, are not primarily concerned with sanctioning past conduct, facilitate the implementation of such solutions.

In the Gazprom case, one of the commitments was a price review clause based on competitive benchmarks.⁵³ The Commission argued that this was an effective way to ensure that prices remained in line with competitive benchmarks without the need for the Commission to impose specific gas prices for individual gas contracts, as the clause allows Gazprom’s customers to react if the gas price starts to deviate from the competitive benchmarks, and the clause is a well-established market-based mechanism for long-term gas contracts.

An advantage of Commitment Decisions is that the companies propose the solutions that they consider most appropriate to address the Commission’s concerns, thus allowing for a more rapid implementation of more effective remedies that go beyond what the Commission could impose.⁵⁴ In addition, third parties can provide the EC with feedback on the proposed commitments to improve them. The Commission’s decision makes the commitments binding on the parties and closes the case without sanctions. However, the Commission can reopen the case, with the possibility of imposing a fine, if the company fails to comply with the commitments (Article 9(2) of Regulation 1/2003), which may be easier than proving unfair pricing.

Finally, in relation to the previous criticism, some scholars argue that, in the presence of sector-specific regulators with the power to sanction anticompetitive conduct, competition authorities should leave the control of excessive prices to the regulators, who, they argue, are better placed to perform this task.⁵⁵ However, intervention by competition authorities should not be ruled out in cases where regulators are not diligent in controlling market abuse,⁵⁶ in particular where their proximity to the industry makes them vulnerable to capture.⁵⁷ In any event, before intervening, the competition authorities should be able to prove that the sectoral regulator’s decision was manifestly wrong.

A more robust solution would therefore be cooperation between sectoral regulators and competition authorities. This would help the latter to better understand how the sector operates, which is useful for assessing the defendant’s behaviour and designing effective remedies.⁵⁸

III. EUROPEAN UNION CASE LAW ON EXCESSIVE PRICING: EMBLEMATIC CASES

The following is an analysis of eight cases in which the EU competition authorities have ruled on excessive pricing, that is, prices that are unfair because they are excessive. The Court first ruled on excessive pricing in 1971 in the Deutsche Grammophon case in which it answered questions referred for a preliminary ruling under Article 267(b) (then Article 177).⁵⁹ In its judgement, the Court raised the possibility of using geographical comparisons to establish abuse of a dominant position, a view that has been confirmed and clarified in subsequent cases. The Court added that, to constitute an abuse, a price differential must not be justified by any objective criteria and must be particularly marked.

Six years later, the CJEU delivered what is considered to be the most important case law on excessive pricing in the EU. In the 1978 UBC judgement, the EC had found that UBC had infringed Article 102 (then Article 86) of the Treaty by (1) requiring its distributors/ripeners not to resell its bananas while still green; (2) charging different prices for equivalent transactions in Chiquita bananas; (3) refusing to supply Chiquita bananas to Th. Olesen A/S, Denmark; (4) and charging excessive prices for the sale of Chiquita bananas in the Belgo-Luxembourg Economic Union, Denmark, the Netherlands, and Germany.

The EC had ordered UBC to (1) pay a fine of one million units of account for infringement of Article 102 (then Article 86) TFEU, (2) cease the infringements, (3) inform its distributors/ripeners that the ban on the resale of green bananas had been lifted, and (4) inform the Commission twice a year for two years of the prices charged during the preceding six months. For each information obligation, a penalty payment of 1,000 units of account per day of delay from the dates set out in the decision was established.

The Commission based its finding of excessive pricing on an internal UBC document from 1974, which stated that the company “sold bananas to Irish ripeners at prices which allowed it a considerably smaller margin than in some other Member States”. From this document, the EC concluded that UBC had made high profits in those Member States where it charged much higher prices than in Ireland.⁶⁰ UBC appealed to the CJEU, arguing that it had made losses in Ireland during the period in question. The CJEU annulled only the excessive pricing decision, holding that it was for the EC to prove the infringement on the basis of its own analysis of the cost structure, rather than relying on the defendant’s documents, and therefore reduced the fine to 850,000 units of account.

In addition, paragraphs 248 to 253 of the UBC ruling set out a conceptual framework that has become the guiding principle for the assessment of excessive pricing cases in the EU. According to the judgement, charging a price that does not bear a reasonable relation to the economic value of the product and that allows the dominant undertaking to obtain benefits that it would not have obtained under normal and sufficiently effective competition constitutes an abuse under Article 102 of the EC Treaty.

The Court also proposed a two-part method, known as the UB test, for determining when a price is excessive. The first part of the test determines whether the difference between the actual costs incurred, and the actual price charged is suspiciously high and, if so, the second part determines whether the price charged is excessive either per se or in comparison with competing products. Thus, excessiveness may be determined either by direct comparison with an estimate of the economic cost of the product or, in the absence of sufficient cost data, by comparison with a competitive price benchmark, without excluding the use of other methods.

In 1989, in the Lucazeau case, the CJEU answered questions referred for a preliminary ruling on the conditions imposed on users by a collective society. The CJEU ruled that where the tariffs charged by a dominant undertaking are consistently and significantly higher than those charged for the same product/service by undertakings in other Member States; this difference must be regarded as an abuse of a dominant position. It added that in such cases, it is for the defendant to justify the price difference on the basis of objective differences between its situation and that of other Member States, shifting thereby the burden of proof to the defendant.

The CJEU also rejected the defendant’s argument that it charged significantly higher royalties to discotheques than collecting societies in other Member States because of its higher operating costs. The Court stated that if the proportion of the revenue absorbed by collection, administration, and distribution costs is significantly higher than that absorbed by those items in other Member States, it cannot be excluded that this is precisely due to the lack of competition on the market in question. Thus, according to the CJEU, if the costs are higher than those of collecting societies in neighbouring countries, the prices may still be excessive, even if the profits are not.⁶¹

The CJEU further clarified the case law in AKKA/LAA.⁶² In response to a request for a preliminary ruling from the Latvian Supreme Court, it stated that is appropriate to compare the defendant’s rate with rates in a sample of other Member States, provided that the selection of the sample is based on objective, reasonable, and verifiable criteria and that the comparisons are consistent.

The Court noted that if the rate differences found in such comparisons are significant, this is an indication of a possible abuse of a dominant position. However, it reiterated that the defendant should have the opportunity to defend itself by proving, based on objective factors, that its rates are fair, thereby shifting the burden of proof to the defendant.⁶³

It also noted that neither the Court nor the Commission had established what level of rate differential could be considered as excessive as that this depended very much on the product and the market in question. The Court did not question the fact that the Latvian Competition Authority compared the defendant’s tariff directly with tariffs in other markets, without first comparing them with the defendant’s cost estimates.

The EC, for its part, has consistently applied the CJEU’s excessive pricing jurisprudence. This includes the UB test, although not always the first part. In the Deutsche Post case of July 2001, the EC condemned DPAG for infringing Article 102 (then Article 82) of the EU Treaty by intercepting, surcharging and delaying incoming mail from the United Kingdom (UK) indicating a sender with a German address.

DPAG charged the same prices for the handling of incoming cross-border mail and for the delivery of domestic mail, without providing any justification. In the absence of reliable cost data for the relevant period, the Commission therefore used 80 percent of the domestic tariff as a benchmark for the costs of delivering incoming cross-border mail, a percentage which DPAG had provided to the Commission in a previous case.⁶⁴ The Commission therefore concluded that the price charged by DPAG for handling such incoming mail exceeded its costs by at least 25 percent.

The Court ordered DPAG to immediately cease the infringement and imposed a fine of €1,000 after DPAG undertook to introduce a detailed procedure for the processing of incoming cross-border letter mail, which would avoid practical difficulties and facilitate the detection of future infringements, should they occur.⁶⁵

In the 2018 Gazprom case, the EC wrote that the first part of the UB test, which compares the actual costs incurred and the price charged, is a screening exercise to assess whether the price charged is suspiciously high, while the second part aims to determine whether the price is excessive, for which the Court suggested two alternatives: comparing the price with a competitive price benchmark or, where no appropriate price benchmark exists, by directly estimating the economic value of the product (‘unfair in itself’).

In line with the above reasoning, the EC first compared the average of the prices net of export taxes charged by Gazprom in Bulgaria, Estonia, Latvia, Lithuania, and Poland with the costs incurred and found that the margin was suspiciously high at 170 percent (first part of the UB test). In the second part of the test, the EC found that Gazprom’s front-month forward prices in all five countries were significantly and persistently higher than in the Dutch TTF gas hub and the German NCG gas hub⁶⁶ (for example, between 22 percent and 40 percent higher than the TTF hub price in 2009–2014). Given these margins and in the absence of any objective justification by Gazprom, the EC’s Preliminary Assessment was that Gazprom may have engaged in excessive pricing practices.

To address the EC’s concerns, Gazprom submitted a number of commitments to ensure the free flow of gas at competitive prices in the five countries, including the introduction of a price revision clause in its contracts. The commitment also included a non-exhaustive list of market segmentation clauses that Gazprom would refrain from using such as export bans or destination clauses. Following objections from third parties, Gazprom clarified and extended these commitments. The Commission decided that the revised commitments met its concerns and that there were no longer grounds for action under Article 9(1) of Regulation 1/2003.

In the 2004 Scandlines case, the Commission argued in its letter to Scandlines, the complainant, under Article 6 of Commission Regulation (EC) No 2842/98 of 22 December, 1998 that there was insufficient evidence to conclude that the prices charged by HHAB, the operator of the port of Helsingborg, to ferry operators for its services were excessive, even if they exceeded HHAB’s actual costs (incurred plus depreciation costs) in providing these services. In its letter, the Commission explained that it was difficult to determine the exact economic value of the services provided by HHAB, as some components were intangible and therefore not reflected in HHAB’s audited financial reports, and identified three such components:

First, the port of Helsingborg had very high sunk costs, which were not properly accounted for in the reports. The Commission argued that the investment costs to build a new ferry port in the same location to provide the same level of services would be much higher than the depreciation costs currently accounted for by HHAB.⁶⁷

Second, the benefit to the ferry operators of the location of the port, which was perfectly suited to their needs, was an intangible value that should be taken into account as part of the economic value of the services provided by HHAB.

Third, the land used by the port for ferry operations was very valuable in itself. Keeping ferry operations there instead of using the land for other purposes represented an opportunity cost for the City of Helsingborg, the sole shareholder of HHAB.

In its reply to Scandlines’ observations on the letter, the EC maintained its view that there was insufficient evidence to conclude that the prices charged by HHAB were excessive. The Commission argued that the economic value of the product/service should also take into account noncost factors, in particular with regard to the demand-side aspects of the product/service in question. In particular, it considered that consumers were willing to pay more for ferry services from Helsingborg because it was the shortest distance between Sweden and Denmark and the terminal was closer to the road and rail networks.

The Commission has since distanced itself from this letter. In an amicus brief in the Pfizer/Flynn case before the UK Court of Appeal,⁶⁸ the EC wrote that the Scandlines decision was not a final decision, but a mere rejection of a complaint and did not make any final findings. It also stated that it was the only example of demand-side factors being taken into account in the Commission’s practice and not a judgement of the Court of Justice or the General Court of the European Union.⁶⁹

In the 2021 Aspen case, the Commission’s Preliminary Assessment was that Aspen may have abused its dominant position by charging unfairly excessive prices for six oncology products it acquired from GlaxoSmithKline plc in 2009. The EC first compared the prices charged for the products with the costs of operation (incurred costs) plus a mark-up of 23 percent,⁷⁰ which corresponded to the Earnings Before Interest, Taxes, Depreciation, and Amortization (EBITDA)⁷¹ margin of a sample of companies with a profile like Aspen’s that operated in markets considered to be sufficiently competitive. The comparisons showed that, for the financial years 2013 to 2019, prices exceeded the cost-plus benchmark by between 280 and 300 percent on average, leading the Commission to conclude that Aspen’s prices were suspiciously high (first part of the UB test).

In the second part of the UB test, the EC looked for potential opportunity costs not included in the EBITDA margin. The Commission found that Aspen did not engage in commercial risk-taking, innovation, investment, or substantial improvement of the products, adding that the patents on Aspen’s products had expired 50 years ago and that it could be assumed that the R&D costs of the products had already been recouped. The EC added that since Aspen outsourced manufacturing and most of the marketing and distribution to third parties, Aspen’s profits could not reflect any production efficiency on its part.

Aspen claimed that the price increases were necessary to obtain a sufficient return on its investment in the products, given the purchase price Aspen had paid to GlaxoSmithKline for the products. The Commission rejected the purchase price as an appropriate proxy for the capital invested in the products, arguing that the purchase price paid by Aspen was likely to reflect mainly the capitalized value of future expected profits from the products and not the value of any tangible or intangible assets. The EC also noted that using the purchase price of the products to measure their profitability would create a circularity problem, as the purchase price itself is a function of the profits expected from the acquisition.

These facts led the Commission to conclude that Aspen’s prices may have been excessive, an assessment reinforced both by the questionable means by which the price increases were implemented and enforced and by the potential harm to patients and healthcare systems. While indicating that it disagreed with the Commission’s preliminary assessment, Aspen offered commitments under Article 9(1) of Regulation 1/2003. It undertook to reduce the prices of the products in the European Economic Area by an average of approximately 73 percent for ten years, subject to a one-off review after five years in the event of a significant increase in the products’ direct costs.⁷²

Under the terms of the agreement, Aspen undertook to appoint a monitoring trustee, a natural or legal person, to ensure compliance with the commitments, subject to the approval and supervision of the Commission. In particular, the monitoring trustee will have to (1) propose to Aspen the measures necessary to ensure compliance with the commitments and (2) propose to the Commission the necessary measures if Aspen does not comply with the trustee’s proposal.

In February 2022, after receiving some clarifications from Aspen, the EC concluded that the commitments addressed its concerns, as prices after the reductions would be on average 10–20 percent above the cost-plus level, and therefore made Aspen’s commitments legally binding and closed the case.

IV. ANALYSIS OF THE EU CASE LAW ON EXCESSIVE PRICING

This section draws some conclusions on the main issues related to the EU case law on excessive pricing. Some of these are already in the literature; others may be new.

A. Economic Value

The CJEU ruling in the UBC case, which is the main source of EU doctrine on excessive pricing, states that charging a price that is not reasonable in relation to the economic value of the product and that allows the dominant undertaking to obtain benefits that it would not have obtained under normal and sufficiently effective competition constitutes an abuse under Article 102 of the EC Treaty. The first question to be addressed is therefore the meaning of economic value.

Under conditions of sufficiently effective competition, the economic profit, not the accounting profit, of any firm should be zero. This condition implies that the price it charges should be equal to its LATC for that level of output. Indeed, total economic costs must include incurred costs and opportunity costs; otherwise, the investor will prefer to allocate his resources to options where he can recover all his costs.

Thus, the benchmark against which a dominant firm’s price should be compared is, or at least should be based on, its LATC for its level of output.⁷³ Interpreted in this way, the EU’s excessive pricing jurisprudence is conceptually based on sound economic principles, as it aims to protect consumers from excessive prices without significantly distorting investment decisions, as it allows investors to fully recover all their economic costs, but excludes any premium resulting from the exploitation of a dominant position.

B. Willingness to Pay

The LATC against which the company’s price should be compared can be affected by demand. An increase in the consumer willingness to pay for a product/service, that is an outward shift in the demand curve, will increase consumption and thus the demand for the product, which, in turn, will increase the demand for factors of production. If one of these factors is scarce, its increased demand would raise its price and thus the cost of the final product/service. On the other hand, increased demand for the product could allow economies of scale to be exploited and costs to be reduced. Only if the company has constant economies of scale will the LATC not be affected by changes in demand. In short, the economic value of the dominant firm’s product should be its LATC at the actual output level, which depends on supply- and demand-side factors.⁷⁴

Furthermore, assuming that the firm is a monopoly and that the demand and LATC curves for the product are textbook, then the output level at which the inverse demand curve and the the LATC intersect is the equilibrium quantity if the firm charges the LATC. Since the consumer willingness to pay for an additional unit of the good decreases with quantity, and the inverse demand curve shows this relationship, in equilibrium only the willingness to pay for the last unit purchased is equal to the firm’s LATC. Note that the willingness to pay for the other units consumed does not affect this LATC, no matter how much consumers want them.⁷⁵

Since the dominant firm is likely to have the lowest LATC among competitors, in many cases, it will be equal to the willingness to pay for the last unit consumed. However, in some situations, the willingness to pay for each unit purchased may be higher than the company’s LATC. Such a situation would arise, for example, if all users of a life-saving drug had a willingness to pay higher than the LATC. However, this situation does not change the fact that the benchmark to determine whether the price charged by the firm is excessive should be its LATC. Conversely, if some people who also need the drug are unable to pay for it, their willingness to pay would be below the LATC. This painful prospect is not a question of competence policy, but of public health policy.⁷⁶

In conclusion, consumer willingness is not an additional factor to the LATC to be considered in determining economic value, nor is it the sole factor determining economic value, as some writings suggest.⁷⁷ On the contrary, it is a determinant of the LATC, together with supply-side factors, and therefore understanding its role shows that its consideration does not preclude a finding of abuse through excessive pricing.

C. Price Comparison with Other Markets

Estimating the ATC is difficult because it requires projecting costs and demand over the life of the project and estimating the cost of capital, which reflects the risks taken by the investor. In addition, if the company is multiproduct, it is necessary to allocate the fixed costs of production to the different products. EU jurisprudence, probably in response to these difficulties, considers the alternative of comparing the allegedly excessive price with the prices of the same or comparable products in other more competitive markets, or with the price in the same market at a time when it was under more competitive pressure.⁷⁸

In Aspen, the EC explains that under the UBC analytical framework, the conclusion as to whether a price is excessive can be reached in two alternative ways: either (1) the price is excessive in itself or (2) the price is excessive when compared with competing products.

There is an economic rationale for using the prices from more competitive markets as benchmarks: the theoretical long-run equilibrium price in competitive markets should be the firms’ ATC at their lowest level. Since the competitive price may be subject to large short-term fluctuations, the price benchmark should be determined by analyzing historical series of competitive prices. Nevertheless, it should be less cumbersome than estimating the defendant’s ATC.

The use of competing prices is not a perfect solution as it involves comparing prices from different time periods, products, or markets with different economic and legal environments. In this sense, the comparator most likely to provide meaningful comparisons is the price charged by the defendant before it acquired market power, if such information is available.⁷⁹ There are also questions as to whether the benchmark represents a sufficiently competitive market.

In Aspen, the EC follows a middle way approach, which it calls “cost-plus”, where the term “cost” corresponds to the costs incurred by the company and the term “plus” is a proxy for a reasonable profit margin under competitive conditions. The latter term is based on the median measure of profits of companies with a profile like Aspen’s, operating in markets considered to be sufficiently competitive. The profit margin is an accounting measure, so it does not necessarily include all the firm’s opportunity costs. For this reason, the EC in Aspen complements the cost-plus method with a search for opportunity costs not included in the cost-plus approach.

D. Efficiency

According to EU case law, the benchmark against which the defendant’s price must be compared is the efficient economic cost. Indeed, in the UBC case, the CJEU stated that the undertaking’s profits should not be higher than they would have been had normal and sufficiently effective competition prevailed (paragraph 249). Since only efficient firms survive in a competitive market, it follows that the benchmark should reflect efficient costs. The Court in Lucazeau confirmed this position by rejecting the company’s defence that it charged higher royalties because it had higher costs and pointing out that this could be due to the company’s inefficiency.

Although competitive market discipline “punishes” inefficiency, it seems odd to impose a legal penalty on a firm for being inefficient. Moreover, calculating the efficient costs would require the design of an efficient firm, a task more appropriate for a regulator than for a competition authority. In practice, competition authorities are more likely to seek to prohibit the use of unjustified costs, particularly where the misrepresentation benefits the owners or managers of the firm. For example, in the Aspen case, the EC refused to consider the purchase price paid by Aspen for the products as a proxy for the capital invested in the products because the Court considered it to be an inflated price. Therefore, when comparing a firm’s sales price with its own costs (excessiveness in itself), a thorough analysis of the defendant’s actual costs should be carried out, correcting, inter alia, overstated costs.

The efficiency requirement is implicit in benchmarks constructed using data from firms operating in competitive markets. The EC in Aspen recognizes the heterogeneity of firms and notes that not every deviation from the average or median profitability of a sector is directly indicative of excessive pricing, as market participants may earn margins above the average or median. Although perfectly competitive markets are a chimera, using prices from highly competitive markets as comparators may overburden the defendant. The way out should be to increase the accepted margin of tolerance between the defendant’s price and the benchmark.

E. The UB Test

In the Gazprom, Scandlines (Article 6 letter) and Aspen cases, the EC compares the “costs incurred-plus” with the prices charged by the defendant to decide whether they are suspiciously high under the first part of the UB test. The EC finds that the price charged by Gazprom exceeds the costs incurred by 170 percent. This difference is partly explained by the fact that incurred costs do not include all opportunity costs, in particular the economic rent of a scarce resource: Gazprom’s gas reserves. For the second part of the test, the EC compares the prices charged by Gazprom with gas prices at hubs that are considered to be sufficiently competitive and that include the cost of natural gas reserves.

In Aspen, the Commission compares the prices charged by the defendant with its incurred costs plus the average EBITDA margin of similar companies operating in more competitive markets. The comparisons show that, for the financial years 2013 to 2019, the prices exceeded the “cost-plus” benchmark by an average of between 280 percent and 300 percent. In the second part of the test, the EC looks in a nonformalized way for possible unaccounted opportunity costs that could explain this margin, but does not find any and therefore concludes that the prices charged are excessive.

In the Scandlines case, in the Article 6 letter, the Commission compares the price charged by HHAB with the costs incurred plus depreciation, which corresponds to the first part of the UB test. As regards the second part of the test, the Commission argues that the economic value should take into account the opportunity costs (intangible assets) that are not reflected in HHAB’s audited accounts and identifies three of them: the opportunity cost of the investment, which may be quite different from the depreciation cost, the opportunity cost of the location of the port of Helsingborg and the opportunity cost of the land used by the port. The EC added that it was very difficult to determine these costs.

In short, in the cases examined in this paper, the EC uses incurred accounting costs plus some opportunity costs to determine whether a price is suspiciously high (the first part of the test). Second, the Commission uses an estimate of economic value, which includes all opportunity costs, to determine whether a price is excessive (the second part of the test). In Scandlines, the Commission lists the main opportunity costs but does not calculate them and therefore does not conclude whether the price is excessive or not. In Aspen, the Commission examines an exhaustive list of possible opportunity costs, which are not necessarily included in the profit margins, and finds none. In Gazprom, the second part of the test is a comparison with prices in more competitive markets, which leads to the preliminary conclusion that Gazprom’s prices are excessive.

F. Scarce and Unique Resources

Firms’ inputs may include scarce resources. If they are purchased on the market, they should be treated like any other input. If the resource is owned by the firm, its cost is not included in its incurred costs, but is an opportunity cost that should be taken into account. This is the case with Gazprom’s natural gas reserves. If the scarce resource is not traded in the market estimating its opportunity cost may be problematic because it is difficult to separate its opportunity cost from the company’s profits.

An example of a nontradable scarce factor is the location of the port land in Helsingborg, as correctly identified by the EC in its Article 6 letter. The letter also suggests a way to determine this opportunity cost when the port is not congested: the best value the city could find for other uses of the land. If the port is congested, calculating of the opportunity cost of the land is more complex. For simplicity, assume that the land does not allow the port capacity to be expanded in which case land is a unique and scarce factor. The opportunity cost of land must now incorporate the valuation of scarcity. Indeed, the opportunity cost of land to be included in the economic cost should be such that, for a port fee equal to the ATC, the demand is equal to the capacity of the port.

Thus, the EC’s rejoinder to Scandlines’ comments on the Article 6 letter obscures the EC’s original position. The opportunity cost of the port’s land, due to its privileged location, should be included in the economic costs of the service provided by HHAB, but is not a “non-cost factor related to the demand-side aspects of the service” as stated by the EC. However, it must be recognized that this cost would increase with demand if the port were operating at full capacity.

It is even more difficult to determine the value of factors that are both unique and heterogeneous. This is the case of copyright collecting societies where there is no way to value a catalogue of heterogeneous works. These societies are set up to collectively manage copyright to reduce transaction costs and are usually de jure or de facto monopolies. The only benchmarks for comparison are other societies that are also monopolies (Lucazeau and AKKA/LAA), so there is an obvious problem of circularity.⁸⁰ Entrepreneurial talent for start-ups that develop inventions and breakthrough innovations, whose value is not directly observable because their results also depend on luck, is another case of similar complexity.

G. Safeguards Against Miscarriages of Justice

EU case law reflects the concern of competition authorities about the risk of false convictions. Since the Deutsche Grammophon ruling in 1971, the CJEU has repeatedly stated that the difference between the benchmark and the price of the product must be “particularly marked” to be decisive for a finding of abuse, and that what constitutes a marked difference depends very much on the product and the market in question (AKKA/LAA). For its part, the EC noted in Aspen that data availability, cost allocation assumptions or methodological choices affect the accuracy of profit calculations and that therefore only profits significantly above the median should be considered excessive (paragraphs 135 and 136).

A second safeguard introduced by the EU competition authorities against the risk of a false conviction is the possibility for defendants to justify their prices on the basis of objective factors after the competition authority has made its preliminary assessment (Lucazeau and AKKA/LAA).⁸¹ First, the EC determines whether the price charged by the defendant is excessive, then the defendant can rebut the EC’s preliminary assessment with new information.

Alternatively, following the EC’s Preliminary Assessment, the defendant may request the closure of the case on the basis of Article 9(1) of Regulation 1/2003 (see footnote 5). This provision, used by the EC in the Gazprom and the Aspen cases, allows the defendant to offer commitments without admitting guilt after the EC has issued its Preliminary Assessment. In this way, the EC can avoid penalizing dominant undertakings for anticompetitive behaviour which they did not foresee as such.

Finally, the sanctioning of an anticompetitive practice requires proof of negligence or intent, which provides an additional safeguard against a miscarriage of justice. In the UBC case, the CJEU found that the defendant knew or should have known that it was infringing Article 102 of the Treaty by adopting measures which, taken as a whole, had the effect of strictly partitioning national markets and concluded that the Commission’s finding that UBC’s infringements were at least negligent was justified.

In the DAPG case, the EC jointly analyzed the defendant’s actions as a pattern of behaviour, from which the EC concluded that DPAG must have been aware that it was impeding the free flow of mail from the UK and adversely affecting competition, and that such behaviour was therefore at least negligent. In the cases leading to the Commitment Decisions, the Commission’s preliminary assessment was that the respondents had engaged in price-increasing strategies, suggesting a series of deliberate actions aimed at raising prices. The law thus establishes a standard of proof that makes it more difficult to wrongly sanction excessive pricing.

V. CONCLUSION

According to EU case law, an abuse of a dominant position occurs when a dominant firm negligently or intentionally charges an excessive price, that is, a price that is disproportionate to its economic costs. The abuse is satisfied by the mere abuse of a dominant position that it enjoys. As economic costs include both incurred costs and opportunity costs, the EU case law on excessive pricing has a sound economic basis as it seeks to avoid welfare losses resulting from the exercise of market power without discouraging investment.

In addition, EU competition authorities, aware of the obstacles to construct an appropriate benchmark against which to compare an allegedly excessive price, have adopted a number of safeguards when prosecuting excessive pricing cases to avoid false convictions. These safeguards include considering a price to be excessive only if it significantly exceeds the competitive benchmark and allowing defendants to rebut the EC’s Preliminary Assessment with objective data. In addition, EU law allows defendants to offer commitments to address the concerns raised in the Commission’s Preliminary Assessment to close cases without a finding of guilt.

Of the cases examined in Section IV which have resulted in sanction or Commitment Decisions, only Aspen has the imposition of excessive prices as its sole alleged infringement. However, even in this case, the EC found that Aspen’s pan-European price increase strategy included the allocation of quotas and, where necessary, the withholding of product supplies in some Member States.

At first sight, these references to trade restrictions could be explained by the fact that Article 102 TFEU prohibits the abuse of a dominant position that is capable of affecting trade between Member States. However, this requirement does not seem to be the main reason, as excessive pricing in an internal market can in itself affect trade if the product is or can be exported to other States. A possible explanation for alleging exclusionary infringements together with excessive pricing is the perception that their inclusion increases the likelihood of a conviction because they are easier to prove. In the UBC case, the Court upheld the EC’s finding that UBC had infringed Article 102 on three counts, but not on excessive pricing. Although the inclusion of ancillary infringements may make it easier to prove negligence or intent, there are cases of pure overcharging where it should not be difficult to prove intent or at least negligence. One example is an unjustified price increase by a dominant undertaking.

The most difficult cases to prosecute are those where a unique scarce resource is embedded in the good/service, as in this case there would be no guidelines to separate the payment of these resources from the profits of the company, as in the Scandlines case. However, this should not be an excuse for not applying the doctrine in situations where it is possible to do so, which can lead to significant welfare gains.

Footnotes

References

Author notes