Can Personalised Price Discrimination Be Prohibited as an Abuse of the Dominant Position under Article 102 TFEU?
At the time of the Fourth Industrial Revolution, traditional systems of law are facing new challenges. During this revolution, machine-learning and Big Data are being actively integrated into business models. The rapid growth of these notions raises many unanswered legal questions. Which field of law should regulate the misuses of the Big Data? Data protection laws, of course, seem to be the most obvious answer to this question. However, what if the economic aspect is affected by such a misuse? In that case, would consumer protection laws and competition law come into play? One of the most widely discussed practices requiring attention is ‘personalised price discrimination’. Access of firms to the personal information of consumers has brought the notion of Big Data to a new level. Firms now have access not only to the data that has been volunteered to them by the consumers (such as names, email addresses and addresses for delivery) but also to ‘observed data’ (such as IP addresses, past purchases, search history and user’s location) and ‘inferred data’ (such as income, political ideologies, hobbies). This allows them to analyse each individual consumer’s willingness to pay and offer a tailor-made price for each consumer. The first time such a practice received public attention was in 2000. The online retailer Amazon offered different prices to consumers who had previously ordered from the website and those who had not. This became apparent when a consumer cleared his cookies and the price for DVD reduced. Furthermore, in 2013 several websites were analysed. It was found that the number of instances of price difference for the same product was the highest for Amazon. According to this research, the location of the person was the decisive factor for the price variation. Another example of how personalised price discrimination is used in practice is the website Staples.com. If the stores of competitors were located less than 20 miles away from the consumer, then the price was lower than usual. 
The focus of the discussion under EU Competition law is Article 102 TFEU. This article prohibits abuse of a dominant position. Although there is no case law directly dealing with personalised price discrimination, this essay will investigate whether this practice can (and should) be prohibited by Article 102 TFEU. First, Article 102 (a) TFEU prohibits “imposing unfair purchase or selling prices or other unfair trading conditions”. According to the case law of CJEU, abuse has occurred when the price of the product does not relate to the value of the product. Second, Article 102(c) prohibits behaviour which applies “dissimilar conditions to equivalent transactions with other trading parties, thereby placing them at a competitive disadvantage…”. Third, the list of abuses under Article 102 TFEU is non-exhaustive and companies can invent new practices that are not mentioned in the list. This paper will thus investigate whether there is a possibility to include a new abuse into the scope of Article 102 TFEU.
Personalised price discrimination has already come into the attention of various scholars and organizations. Townley, Morrison, and Yeung have mainly investigated the economic impact of this practice on the EU Competition Law and the fairness and justice aspects. Maggiolino focused on the impact of personalised price discrimination on consumer welfare and rather on the question ‘should’ it be prohibited under Article 102 TEFU. As to the privacy angle, Borgesius and Poort addressed the attitude of the consumers towards personalised pricing and what should be done in order to ensure GDPR compliance. Furthermore, OECD addressed questions of both privacy and competition law in a more general framework, beyond that of the European Union. The European Commission has also expressed its concerns with a focus on consumer welfare and attitude. This paper will conduct more detailed research into emerging trends in the case law of the CJEU. Although there has not been a case brought before the court yet, the fast growth of the digital market suggests that this will soon be one of the most important problems to deal with. Furthermore, this paper aims to answer a more practical question of whether the practice of online personalised pricing can be prohibited under Article 102 TEFU?
The first section of the essay will focus on the impact of personalised price discrimination on consumer welfare. The economic meaning of personalised price discrimination and its effect on consumers will be analysed. Although the aim of this paper is to investigate the practical aspects of the problem, the policy questions always lie with the very foundation of the Court’s decisions and therefore they also have to be analysed. In the second part, it will be considered whether this practice falls under any of the exciting abuses and if not, the possibility of including the personalised price discrimination into the scope of the Article as a new abuse will be examined. This paper will also briefly touch upon the interrelation between the competition law and data protection laws when considering the factors for the establishment of new abuse.
Scholars and practitioners seem to have a divergence of views on the definition of ‘personalised pricing’. It has often been used interchangeably with ‘price discrimination’ or ‘perfect discrimination’. The reason for this might be the fact that personalised pricing is a novel model, which has only become possible at the age of the big data. However, it has its roots in the economic notion of price discrimination. If economic terms are to be followed, personalised pricing is a form of first-degree price discrimination. Therefore, it is first necessary to define ‘price discrimination’. In economics, there are two dominant definitions. Under the margin definition, price discrimination takes place when “price-cost margins of two varieties of the same good are not the same”. Second, mark-up definition, where “price-cost mark-ups of two varieties of the same goods are not the same”. The CJEU, however, chose a simpler definition. Price discrimination was defined as “charging different customers or different classes of customers different prices for goods or services whose costs are the same or, conversely, charging a single price to customers for whom supply costs differ”. However, when a definition of personalised pricing is concerned, it is important to take into account the way this form of discrimination occurs – by collecting data. Thus, what differentiates the general price discrimination and personalised pricing is that in the latter, the price is set in accordance with the personal characteristics of the consumer, which are established with the use of big data. OFT offered a definition of the personalised pricing being “(…) the practice where businesses may use information that is observed, volunteered, inferred, or collected about individuals’ conduct or characteristics, to set different prices to different consumers (whether on an individual or group basis), based on what the business thinks they are willing to pay.”. OECD defined it as “any practice of price discriminating final consumers based on their personal characteristics and conduct, resulting in prices being set as an increasing function of consumers’ willingness to pay”. This paper will follow the second definition for several reasons. First, as will be shown in the next section, personalised pricing is always a form of discrimination. Second, the definition of OECD offers a broader scope of data collection. Instead of specifying the ways in which the data that is used to discriminate is collected, it focuses on the fact that the price is based on personalised characteristics, what gives a larger leeway for newly developing technologies to fall under the definition.
This chapter will first explain the three economic models of price discrimination and examine which model is better applicable for personalised pricing. Second, analysis of a perfectly competitive market and imperfectly competitive one will be made in order to determine the most suitable environment for personalised price discrimination. Third, the inquiry into the effect of personalised price discrimination on consumer welfare and general welfare will be conducted. This has its relevance to the EU Competition law because the economic welfare stands at the core of Article 102 TFEU, and no practice can be regarded as an ‘abuse’ if there is no effect on the economic welfare.
1.1. Models of Price Discrimination
According to Varian, there are three types of price discrimination. Although Varian and many other scholars  focus their research on the effect of three types of discrimination in a monopolistic marker, they are equally applied to the other markets.
In first-degree price discrimination, the supplier has perfect information about consumer and thus is able to charge the maximum price that the consumer would be willing to pay. According to many scholars, this type of discrimination is possible in theory, but not in any real-life setting. However, the assumption that such perfect price discrimination is not possible is only based on the presumption that companies cannot gather all the information about their consumers. Nowadays, big data is changing this paradigm. The abilities of companies to collect and analyse data can lead to first-degree price discrimination. As Cross and Dixit argue, the policies of companies have started to change from being ‘product-centric’ to ‘customer-centric’. This means that now, if companies able to study each individual’s preferences (and most companies have such an ability), they can create a custom-made price for each of the customers and discriminate on a first-degree basis.
Second-degree price discrimination occurs in instances where the supplier offers a menu of options that the buyer can choose from and the price will be changed in accordance with the chosen option. The most common example is selling a different number of goods at different prices. However, in the context of online price discrimination, Townley adjusted the term by explaining that the second-degree price discrimination occurs when the monopolist has no ex ante information about the willingness to pay but can learn it through consumer’s purchase decisions. However, Townley might have misinterpreted the example which was given by Varian. In the classic example of second-degree price discrimination, Varian referred to discrimination that is exercised by the way of self-selection. This means that the company offers two different types of packages, and the consumer with a higher willingness to pay would choose, for example, an unrestricted fare for an airplane ticket. In the example given by Townley, however, the goal of offering two different packages is to gather information about willingness to pay. As it can be deduced from Varian’s example, the essence of the second-degree price discrimination is self-selection, and although consumers might reveal some information, that comes as a by-product of this discrimination rather than as its end-goal. The same conclusion was reached by Areeda, who stated that although the second-degree price discrimination indeed allows the monopolist to observe the willingness to pay, the consumers are usually able to make a choice in accordance with their willingness to pay. It was also stated that for the second-degree price discrimination there is no need for the seller to acquire the information about the buyer due to the self-selection. This also goes contrary to Townley’s statement that in the classic models, the consumers are passive. In reality, in this model, consumers have full freedom to choose the price in accordance with the amenities required by them. Although Varian in his other article has himself stated that providing different prices in accordance with the purchase history is a form of second-degree price discrimination, he has also stated that it always involves the strategic response of consumers to those personalised prices.
Third-degree price discrimination means that a monopolist differentiates the prices by a group of people. Thus, different groups would get different prices, but each person within a group would get the same price. The classical examples are student and senior citizen discounts. In the context of online price discrimination, Townley pointed out that in this case, the monopolist has imperfect information and the prices vary in accordance with the observed characteristics. One of the examples of modern third-degree price discrimination can also be geo-blocking. According to the European Commission, in the context of the geo-blocking, geographical areas (most often countries) indicate the consumer’s willingness to pay. 
Some scholars argue that third-degree price discrimination is the most relevant for the algorithmic online personalised pricing. The reason for this is that second-degree price discrimination is indirect discrimination while for first-degree discrimination, the supplier has to be able to observe all the relevant factors, which is almost impossible. However, this might have only been true for the past decades. At the age of the big data and fast developing technologies, it seems like first-degree price discrimination is becoming more of the reality, than speculation. For example, Amazon and other online suppliers tailor their webpages to the preference of each individual consumer. If companies are able to do this, there is no reason to believe that they cannot collect enough information to also customize their prices for each individual. Therefore, this paper will proceed with assessing personalised price discrimination as first-degree price discrimination.
Three types of markets are relevant for the question of personalised pricing: monopoly, perfectly competitive markets, and imperfectly competitive markets. 
In the monopolistic market, all three models can work perfectly. When the company is a monopolist, it can acquire all the information about their consumers and can actually afford to charge them the maximum of what they are willing to pay. It is less clear, however, which form will personalised pricing take in the competitive market.
Most scholars agree that price discrimination cannot occur in perfectly competitive markets as the firm needs to have some degree of market power in order to be able to discriminate.This is because a perfectly competitive market is defined as one where the number of buyers and sellers does not allow the single firm to affect the market price. In a perfectly competitive market, the law of one price applies because no single firm can have long term market power. Therefore, it is not a suitable environment for personalised price discrimination. In economics, this phenomenon is explained by the fact that in the perfectly competitive market the price equals the marginal cost, while price discrimination only occurs when the price differs from the marginal cost. However, Geradin and Petit stated that perfectly competitive markets are particularly rare and that nowadays price discrimination exists even in highly competitive markets as most of the firms enjoy at least some degree of market power.
Thus, price discrimination mostly occurs in the imperfectly competitive setting (oligopolistic market). Although the imperfectly competitive market is a better environment for price discrimination as it gives companies leeway to impose such discriminatory measures, such a market has also its downfalls. If, for example, in the monopolistic market, a company has the ability to charge the maximum that consumer is willing to pay, on the imperfectly competitive market, on the other hand, it is hard to push consumers to their limits as they might choose another, cheaper brand. Thus, in such markets, discrimination can only take two forms. Firstly, the supplier can base its price on the additional value received by the consumer when choosing a particular supplier. Second, they can base the price on the search costs. Corts distinguishes these models as asymmetry and symmetry. In the asymmetric model, the consumers are distinguished into a ‘strong’ ones (consumers that prefer the brand of this supplier) and ‘weak’ ones (consumers that prefer rivals). While the prices can be raised for a strong market, they will also be decreased for a weak one in order to attract new customers. In the symmetric model, however, the ‘strong’ consumers are consumers that usually do not take a lot of time for shopping and make impulsive decisions. In this instance, the supplier can charge them more. ‘Weak’ customers would search for a lower price, which makes the suppliers discount them more.
It was important to initially determine the degree of price discrimination that personalised pricing represents because the effect on welfare is dependent on the degree of discrimination. This section will assess the factors that determine the consumer welfare and the effect that the personalised price discrimination has on this welfare. Furthermore, the opinion of the public on this practice will be analysed.
The first advantage of personalised pricing is that allocative efficiency might increase. Allocative efficiency means “an output level where the price equals the Marginal Cost (MC) of production.”. In the case of personalised pricing, the allocative efficiency will be increased by way of reducing the prices for consumers with lower (ability) willingness to pay while still making a profit on the consumers with higher (ability) willingness to pay. This can doubtlessly be achieved in first-degree price discrimination. The consumer will usually make a purchase if their willingness to pay is higher than the price. If the personalised price discrimination is exercised, such a scenario is always possible, as long as the price covers the marginal cost of the product. However, if there is no discrimination or the supplier does not have information about the consumer’s willingness to pay, they will set a general price, which will always be above costs as they need to increase their profits. In this case, only the consumer whose willingness to pay by default is higher than the price would make a purchase. The difference between these examples is that if there is personalisation, there is a high possibility that the supplier will set the price at the same level as the costs. However, in the second scenario, the price will always be higher than costs and in most of the cases will not be equal to the consumer’s willingness to pay. Thus, under the first scenario, there will be an increase in allocative efficiency.
The effect on welfare can also be ambiguous, depending on the standard of welfare chosen. As it has been shown before, the allocative efficiency will increase. However, general consumer welfare does not benefit from such measures as price discrimination means that the consumers with higher willingness to pay would pay way more than the marginal costs of the product. This creates a surplus from the consumers with lower willingness to pay to the ones with higher. Personalised pricing can also affect the distribution of surplus between consumers and producers.  In the classical models, the effect on the distribution of surplus is dependent upon the type of discrimination. If price discrimination is a form of third-degree discrimination and prices exceed the costs to the minimum extent, the consumers receive almost an entirety of the surplus. This is because the supplier is able to discount some consumers without charging the other consumers at a higher price. In such a case, the supplier’s benefit is that they can sell a higher amount of goods. In this model, the consumer’s with lower willingness to pay will benefit and the consumer surplus will increase. However, this paper argues that the essence of personalised pricing is first-degree price discrimination. In this case, the supplier is able to extract the maximum of the consumer, thereby capturing the surplus and reducing consumer welfare almost to zero.
The scholars that argue in favour of the personalised pricing also bring up the argument that it can foster competition and innovation between the firms. In the competitive market, competition becomes more intense if the personalised pricing is used because the firm has to fight for each individual consumer, instead of treating them as a mass. In this case, total welfare and dynamic efficiency will increase. The need in creating new mechanisms for personalization and increased competition will encourage the firm to make more innovative and creative solutions. In imperfect competition, for example, in the asymmetric model, the competition is increased because the ‘strong’ consumers of one supplier are the ‘weak’ consumers of the other one, and both of the companies are trying to attract the ‘weak’ customers of their rivals. In this model, the intensified competition can lead to all prices being lower than they would be if the uniform prices were to be applied. On the other hand, there is no impact on the total welfare as in the competitive market, companies make a lower profit with the asymmetric model. However, in the symmetric model, the ‘strong’ consumers would be the same for both sellers. Those consumers are characterised by their high search costs. This means that the consumers do not actually search for products with better price but just buy the first product they find. Therefore, no actions of the firms can help them to attract these customers.
Although public attitudes are not formally taken into account when Article 102 TFEU is concerned, it is necessary to consider whether those who are directly affected – consumers see personalised pricing as a problem. When the public opinion of the personalised pricing is examined, the first point to be considered is the percentage of the European population that is aware of ‘personalised pricing’. According to a survey conducted by the European Commission, only 44% of the people who conducted the survey in Europe have heard about personalised pricing, while for example, 67% heard about the targeted advertising and 62% heard about personalised offers. When the question about the benefits of the personalised pricing was asked, the consumers that have heard about it/ understand how it works, have mostly stated that (1) it allows online retailers to offer promotions and discounts, or (2) they do not see any benefits. The concerns are mostly connected to the storing of personal data. As to the overall opinion about this practice, 36% stated that they see both benefits and disadvantages, 33% said that they see primarily disadvantages and only 8% saw primarily benefits. In 2005, the survey conducted in the US illustrated that most of Americans object to any form of price discrimination as it is ‘ethically wrong’. The reasons for that included that: people agreed it would bother them to learn that other people pay less than they do for the same products and 66% dislike the fact that the supermarket where they shop at keeps detailed records of their buying behaviour. Another survey illustrated that even if the personalization can offer benefits, such as discounts, they did not want their prices on one website to be affected by their behaviour on the other websites. 
Article 102 of the TFEU ensures that dominant firms do not abuse their market power in the internal market. This Article states that “Any abuse by one or more undertakings of a dominant position within the internal market or in a substantial part of it shall be prohibited as incompatible with the internal market in so far as it may affect trade between Member States.”. According to the Commission, one of the main goals of this Article is the protection of consumer welfare and effective distribution of the resources. The Commission has also stated that in applying this Article, it will focus to prevent the conducts that are the most harmful to consumers. The examples of the practices that are beneficial for the consumers are lower prices and a wider choice of new or improved goods. Thus, the end-goal is to increase economic efficiency while ensuring the well-being of consumers. It is also believed that the goal of the Article is to ensure that the competition is not distorted. This, in essence, means that the main aim of the Article is the preservation of the general objective of the Union. Article 102 TFEU does not contain an exhaustive list of abuses, but rather offers examples of the abuses. This chapter will first examine the possibility of personalised pricing falling under ‘excessive pricing’ in Article 102(a) TFEU. Second, it will be investigated whether it can be prohibited under Article 102(c) TFEU. Third, if none of the abuses listed in Article 102 TFEU would seem to encompass personalised pricing, the possibility of establishing a new abuse will be considered.
Preliminarily, it must be mentioned that there are two types of abuses: exploitative and exclusionary. Exploitative abuse is abuse that harms consumers directly, while exclusionary abuse harms them indirectly through distorting competition on the market. Exploitative abuses might include excessive pricing, while exclusionary abuses can take a form of rebates, predatory prices, tying. 
Article 102 (a) TFEU prohibits the imposition of “unfair purchase or selling prices”. It has been established that this Article includes both, excessive prices and predatory prices. This paper will, however, only focus on excessive pricing. The notion of excessive pricing was initially considered in the landmark case United Brands. The Court defined ‘excessive price’ as a price, which had “no reasonable relation to the economic value of the product”.  In order to determine whether the price has a reasonable relation to the economic value of the product, the Court proposed the following two-stage test:
1) determination of whether the difference between the costs and the price is excessive (whether the profit margin is excessive), and;
2) if the answer to the first point is affirmative, whether the price that has been imposed is unfair in itself or when compared to the competing product.
As to the first limb of the test, neither the Court nor the Commission has defined excessiveness of profit margin. It was only stated by the Commission that “the mere fact that revenues may exceed costs actually incurred is not sufficient to conclude that the difference is ‘excessive’ ”. In the case of Deutsche Post AG, the Commission found that for cross-border mailing the profit margin of 25% (from the amount of costs) was excessive, as the average profit margin on this market was 3%.  The Italian Competition authority, using the test established in the United Brands, found that a profit margin of 100-400% was excessive in the pharmaceutical market. Thus, it is evident that the excessiveness of the profit margin is dependent upon the market at stake and whether there is an appreciable alteration from the competitive levels. 
The second limb of the test was approached from the range of different angles. In order to determine whether the price was unfair, the Court used a benchmark of comparing the prices of the dominant undertaking in the past for the same product, dominant’s undertaking prices for other products in the same market, competitor’s prices on the same market, dominant’s undertaking prices on the same product market but on the different geographical market.
In the context of the personalised price discrimination, the second limb might not be efficient. If the company has a mechanism in place, where, for example, the website is tailor-made for a particular client, the similar products will also be offered at a higher price to this customer and the same website will also impose higher prices on the same customer in the different member state, and hence this practice would not be caught by this test.
The ineffectiveness of the test in the case of personalised pricing can be explained by the product-centric approach that EU Competition law takes. Methods mentioned above focus on a comparison of the prices on different products. However, in modern times, this approach might not be the best one to take. As was mentioned above, new technologies are changing the market from being a product-centric to being consumer-centric.A lot of business magazines and books are offering the firms solutions on how to become ‘consumer-centric’. The essence of the modern market can be inferred from the quote of Hernandez, partner at KPMG: “achieving customer centricity is no longer a differentiator. In the digital age it has become a matter of survival ”. KPMG in its brochure on customer-centric business provides an action plan for business to survive in the digital era, where the essential two steps are the collection of the data of each customer and on the basis of this data creation of a “new customer behaviour segments and ‘personas’ ”. Such a developing market strategy illustrates that the competition law has to also develop and also change from a product-centric approach to a customer-centric approach. In the context of online price discrimination, it is to be proposed not to compare prices of the similar products that the other firms charge/ the same firm charges for the similar products, but compare the prices offered to different consumers. This seems to be the only way that Article 102(a) can catch personalised price discrimination.
Therefore, the most suitable two-step test for the personalised price discrimination to be caught by ‘unfair pricing’ is (1) to find a price-cost difference and profit margin and (2) to analyse whether the profit margin is excessive by analysing the profit margin obtained from the different consumers. However, even such a test might be too complicated and puzzling as the price imposed on each consumer for this particular product will have to be analysed.
Therefore, Art.102(a) TFEU is not appropriate for personalised pricing even if the consumer-centric test is used. There has always been a strong separation between excessive pricing and predatory pricing. When those two are intertwined, this does not become unfair pricing, but margin squeeze, which falls outside the scope of this paper. The essence of personalised pricing is making a profit by charging some consumers too much, while charging others too little. Therefore, it is clear that the abuse of excessive pricing (or predatory) alone would not be enough. The issue of personalised pricing has to be tacked from both sides and therefore, the next section will address discriminatory abuse under Article 102 TFEU.
Article 102(c) TFEU prohibits ‘applying dissimilar conditions to equivalent transactions with other trading parties, thereby placing them at a competitive disadvantage’. Therefore, there are three elements that must be assessed under this Article: (1) applying dissimilar conditions to equivalent transactions (2) what results in the other trading parties being at a competitive disadvantage.
According to the Court, discrimination is not only found when dissimilar conditions are applied to identical transactions but also when similar situations are treated differently, or different situations are treated identically. In general, this means that a dominant undertaking applies different prices for two transactions that have the same costs. However, ‘equivalent’ still lacks a definition. For example, in Clearstream Banking AG, the Court found that the two transactions were equivalent because the two companies that were compared each provided secondary clearing and settlement services and there was no clearing and settlement services that the dominant undertaking provided for one firm but did not provide for another one. Although the imposition of the different prices is one of the possible interpretations of the term ‘dissimilar conditions’, it alone will not allow establishing the abuse. The costs difference must reflect the difference in price. For example, if additional costs incurred by the dominant undertaking are insignificant or marginal, the dissimilar conditions would be found. However, if the dominant undertaking has actually borne more costs for one of the transactions and thus had to set a higher price, then the conditions would not be dissimilar for the purpose of Article 102(c) TFEU.
The most controversial part of the section is, however, ‘competitive disadvantage’ that has to be caused to ‘another trading parties’. Although it might seem as ‘another trading parties’ means ‘traders’ or ‘undertaking’ and thus the Article only applies to the downstream competitors, the rest of the provisions that actually apply to another business use the term ‘undertaking’. Thus, the term ‘another trading parties’ can, in theory, include consumers. For example, when a consumer buys something from the online website, it enters into a trade with the supplier. 
Most of the scholars argue that competitive disadvantage can only occur in the transactions between the firms because consumers do not compete with each other. However, the notion of ‘competitive disadvantage’ has been interpreted ambiguously by both, commission and the Court. There seemed to be a tendency moving towards inclusion of the B2C transactions into the scope of the section. In fact, according to the Commission, there is no need for the other trading parties to be negatively affected, but merely the fact that consumers are negatively affected is enough.  The Commission supported this view in the Football World Cup case. In this case, the dominant undertaking required consumers to provide their address in France for the sale of football tickets, thus not allowing people from non-French member states to purchase the tickets. The Respondent, in this case, argued that it did not derive any commercial advantage from this action.  Commission, on the other hand, stated that first of all, finding the competitive advantage acquired by the firm is not necessary in order to establish an abuse. Second, this conduct could not be deemed to fall outside the scope of the section, as the behaviour was aimed at discrimination against the customers on the grounds of nationality, what is against the fundamental principles of the European Union. Although this decision hints at the general discrimination under Article 102 TFEU and not subsection (c), in the latter case of Deutsche Post, Commission cited the Football World Cup case to refer to abuse under Article 102(c) TFEU. This clearly shows that the Commission was accepting the view that Article 102(c) applies without causing the competitive disadvantage, thereby only having a negative impact on the consumers as a result of the application of dissimilar conditions.
However, there is an inconsistency between the views of the Commission and the Court. In the recent case of MEO, the Advocate General stated that price discrimination by itself is not problematic under the competition law as it is not always harmful to competition. He argues that the practice of price discrimination can increase the efficiency and well-being of the consumers and that “the rules of competition law are designed to safeguard competition, not to protect competitors”. However, the case of MEO dealt with the downstream competition, where the dominant undertaking imposed discriminatory pricing on the other undertaking.  In this case, it is clear that there should be an anti-competitive effect. The Court has supported the view of the Advocate General and provided that the mere existence of competitive disadvantage does not mean that abuse is prohibited under Article 102(c) TFEU, but there is always a need to establish a competitive disadvantage. 
What can be inferred from the analysis of the Commission decision and case law of the Court is that while Commission was ready to accept that discrimination against consumers without having a competitive disadvantage under Article 102(c) TFEU, the Court seems to be reluctant to accept this point of view. In fact, the only case where the Commission found discrimination against the end consumers took place 20 years ago, and it was never appealed to the Court. If the recent case of MEO is to be applied by analogy (although it dealt with undertakings), the Court did not take into account the fact that Article 102 TFEU is non-exhaustive, but strictly applied the wording of section c) of the Article. Therefore, the conclusion can be drawn from a recent case of MEO that the Court would not accept the application of Article 102(c) TFEU to the personalised pricing, at least when consumers are concerned. It might be different, however, if there would be a case of downstream competition between undertakings, which falls outside the scope of this paper.
If the recent developments are to be analysed, the conclusion can be drawn, that the Commission and the Court do not have a uniform view on the issue. From the wording of Art.102(c) TFEU and case law, it can be deduced that the notion of the abuse in this article means discrimination towards several competing downstream undertakings which have a potentially exclusionary effect. In this case, discrimination towards end consumers which has an exploitative effect falls outside of the scope of the section. The recent case law developments in the sphere point out that in accordance with Article 102 TFEU and understanding of Commission and the Court, Article 102 as a whole is aimed at protection of competitors, instead of consumers. This is also supported by the fact that the Commission has decided to adjust its approach to Article 102 TFEU but only with relation to exclusionary abuses, leaving the exploitative outside of the development. According to the Commission, the aim of the communication on the exclusionary conduct under Article 102 TFEU is “to ensure that dominant undertakings do not impair effective competition by foreclosing their competitors in an anti-competitive way, thus having an adverse impact on consumer welfare”. This is inconsistent with a general distinction between exploitative/exclusionary abuses. If the Commission claims that protection of the consumer welfare is one of the main goals, it has to directly deal with the exploitative abuses, which have a wider effect on the welfare of the end consumers. It seems that the aim of protecting consumers comes in only as secondary, effective competition being the primary one. This is, also inconsistent with the goals set by the commission itself, which is to “enhance consumer welfare”.
2.3. Can Personalised Price Discrimination Be Included into the Scope of Article 102 TFEU as a Separate Abuse?
This section will attempt to illustrate that there is still a possibility to include personalised price discrimination into the scope of Article 102 TFEU. Although, as it has been shown above, the exploitative conduct of the undertakings, such as excessive pricing and discrimination, have always been of little interest to the Commission. The end-goal of Article 102 TFEU and the underlying rationale of the prohibition of exclusionary conduct is also the protection of consumers. Furthermore, in the early case law, there was a tendency to conflate exclusionary practices and exploitative ones. Excessive pricing and price discrimination have sometimes been used in order to support exclusionary considerations. Therefore, although the Commission has never given any clear guidance on the exploitative abuses, the general tendency can still be found if the analysis of the exclusionary cases is to be conducted. This section will make an analysis of the notion of ‘abuse’ under Article 102 TFEU and will seek to detect a criterion for the establishment of abuse and whether personalised price discrimination can be deemed as an ‘abuse’. This will be done by conducting an analysis of the case law where the abuse with similar characteristics as personalised pricing was found.
It has to be preliminarily stated that although there is no direct evidence that personalised pricing occurs in the internal market on the large scale, the European Commission and European Parliament both acknowledge the threat of this practice. Commission has prepared a report together with OECD on personalised pricing, where it supported the fact that although Article 102(c) requires discrimination between undertakings, Article 102 TFEU is non-exhaustive and personalised pricing might fall under the broad notion of ‘abuse’. At the same time, the European Parliament made an amendment to a proposal of the Commission for New Deal for Consumers directive. In this amendment, recital 45 together with Article 6 demand the consumers to be informed about any personalised pricing on the basis of the automated decision making. What is peculiar about this amendment is that the European Parliament does not prohibit the personalised pricing but merely imposes an obligation to inform.
The Court has repeatedly held that Article 102 TFEU prohibits both, practices that are harmful to the consumers directly, and practices that have a negative effect on the consumers through their impact on the competition. The concept of the ‘abuse’ relates to the conduct of the dominant undertaking, which hinders the competition on the market by its mere presence and that by the particular conduct weakens the existing competition. When considering whether a conduct falls under the notion of ‘abuse’ “it is necessary to consider all the circumstances and to investigate whether the practice tends to remove or restrict the buyer’s freedom to choose his sources of supply, to bar competitors from access to the market, to apply dissimilar conditions to equivalent transactions with other trading parties, or to strengthen the dominant position by distorting competition”. 
One of the most important aspects of the ‘abuse’ within the meaning of Article 102 TFEU is that the dominant undertaking, due to its powerful position on the market has a special responsibility not to let its conduct impair the competition. Therefore, when proposing a new abuse, the principles of legal certainty and foreseeability have to be preserved: the dominant undertaking when applying particular conditions has to be able to assess whether its conduct will be deemed lawful or not. In essence, when such a broad, non-exhaustive provision as Article 102 TFEU is concerned, the undertaking should assess the legality of its conduct on the basis of reference to relevant rules and interpretation of them by the Court.
One of the characteristics of personalised price discrimination is the imposition of the higher prices to ‘strong consumers’ (loyal consumers) and lower to ‘weak’ consumers (consumers that are loyal to the competitors). In the case of AKZO, the Commission fined the dominant undertaking for aggressive pricing strategy with an aim of elimination of the competitors. AKZO asked its customers to obtain quotes from its competitors, implying that whatever the price competitors have, it would offer them a lower one. The Commission considered, that both, applying selective prices to certain customers and collecting the information about competitor’s prices (which was not freely available on the market) contribute to the establishment of the abuse. The decision of the Commission was later supported by the Court which stated that “with regard to the acquisition of information, it must be observed that where such a practice forms part of a plan to eliminate a competitor, as in the present case, it cannot be regarded as a normal means of competition.”. 
This can have a direct relation to the personalised price discrimination. Collecting of data, otherwise not being available and charging a lower price with a view of weakening competition are characteristics of the personalised price discrimination and the Court has accepted that those factors can be taken into account when establishing an abuse.
Another factor that was taken into account by the Court when establishing an abuse, was whether there is any economic justification for the practice of offering different conditions to different consumers. Although the case of Hoffmann-La Roche dealt with fidelity rebates, it is relevant for the personalised price discrimination, because the effect of those rebates is ‘to apply dissimilar conditions to equivalent transactions with other trading parties in that two purchasers pay a different price for the same quantity of the same product depending on whether they obtain their supplies exclusively from the undertaking in a dominant position or have several sources of supply’. Thus, in order to be able to establish discrimination, it has also to be established whether there is any economic justification for the existence of the practice. Such an economic justification has to show that the application of dissimilar prices is dependent upon saving of the costs. One of the most essential characteristics of price discrimination, however, is that prices do not reflect the costs, but only reflect the willingness to pay of each individual consumer. 
It was furthermore mentioned in the report by the European Union, that whatever measure is taken against price discrimination, it has to be assessed on the basis of the reference to consumer welfare.  When competition law is concerned, consumer welfare takes precedence over the general welfare.  As it was mentioned above, the impact on consumer welfare will mostly be negative as almost the whole surplus will be captured by the supplier. Although it can be argued that the effect on the consumers with lower willingness to pay is positive as they do not have to pay more than they can afford when the total consumer welfare is concerned the outcome will not be positive. This is because the consumers with a higher willingness to pay will have to pay the maximum or their limit and thus the effective distribution will be impaired. Furthermore, although the welfare of consumers with lower willingness to pay may be increased, charging them low price can also characterise as conduct that aims at excluding other undertakings from the market.
Furthermore, discrimination under Article 102 TFEU without any reference to subsection c is not a novel concept for the Court. For example, the Court found an abuse of dominant position when a dominant undertaking imposed different delivery terms for so-called loyal consumers and consumers that used a product of competitors as well. It was held that undertaking has a right of imposing a certain criterion for the treatment of consumers as long as the criterion does not pursue a discriminatory purpose and the exclusionary effect does not result from it.
Another aspect that can be taken into account when establishing an abuse is data protection laws and especially whether there has been a violation of GDPR by a dominant undertaking. The Court has confronted the interplay between competition law and data protection only once, in the 2006 case of Asnef. There, it was stated that ‘any possible issues relating to the sensitivity of personal data are not, as such, a matter for competition law, they may be resolved on the basis of the relevant provisions governing data protection.’ However, the case was decided 13 years ago, prior to the GDPR’s entry into force. As to the recent case, German federal competition authority as recently fined Facebook for exploitative business terms in the light of inadequate data processing, basing its decision on the GDPR. In this case, Facebook used the data collected from the other companies of the Facebook group (Instagram, WhatsApp, etc.) to use it for its own advertising.  It was held that competition law can intervene in cases that are essential for data protection if the conduct involving data protection is applied in the manifestation of market power.  It was also stated that competition law involved balancing the constitutional rights of individuals against the behaviour of the dominant undertaking. The dominant position of Facebook, in this case, was directly linked to the exploitative conduct, as it had the power to impose certain conditions on the end-users due to its market position. 
This section will, therefore, examine whether data protection and competition law can intervene together in cases of personalised pricing.
It should be stated that the protection of personal data is one of the fundamental rights in the European Union enshrined into Art.16 TFEU, which states “Everyone has the right to the protection of [their] personal data”. GDPR applies at the moment when personal data is processed.  Personal data is ‘any information relating to an identified or identifiable natural person’. Although personalised price discrimination mainly deals with instances of ‘cookies’, Article 29 Working Group stated that ‘cookies’ fall under the definition of personal data because they usually contain a unique user ID.  Furthermore, the Court also gave a broad definition to the notion of personal data, stating that a dynamic IP address constitutes ‘personal data’. 
Does processing with the purpose of the price customization have a legal basis?
Does Article 22 GDPR allow automated decision with the purpose of the price customization?
Even more relevant for personalised pricing is Article 22 GDPR. This Article states that individuals have a right not to be subject to processing done by solely automatic means, including profiling which produces a legal effect or similarly significant effect. Although Article 22 GDPR is in essence interpreted as giving a right to individuals, it implies in principle a prohibition on decisions based on solely automatic means. For the practice to be prohibited under this Article, 4 conditions have to be met: (1) there should be a decision (2) based on solely automatic means, including profiling (3) which produce a legal or similarly significant effect on an individual. When the personalised pricing is concerned, usually the first two conditions are fulfilled.  Profiling is ‘any form of automated processing of personal data consisting of the use of personal data to evaluate certain personal aspects relating to a natural person, in particular to analyse or predict aspects concerning that natural person’s performance at work, economic situation, health, personal preferences, interests, reliability, behaviour, location or movements.’. Personalised pricing includes profiling because an automated mechanism analyses the data collected by an individual in order to determine their willingness to pay. As to the third condition, ‘significantly affects’ has never been defined in the GDPR. However, in Recital 71, the given examples are ‘automatic refusal of an online credit application’ and ‘e-recruiting practices without any human intervention’. Article 29 working group stated that in order to fall under the notion of ‘significantly affects’ the practice needs to tend to ‘(1) significantly affect the circumstances, behaviour or choices of the individuals concerned, (2) have a prolonged or permanent impact on the data subject; or (3) at its most extreme, lead to the exclusion or discrimination of individuals.’. According to this, personalised pricing would fall under the notion of ‘significantly affects’ as it automatically leads to the discrimination between consumers.  Article 22(2) provides for three exceptions (1) if it is necessary for the performance of the contract between the data subject and a data controller; (2) it is authorized by Union or member states; (3) it is based on data subject’s explicit consent.
Therefore, it can be concluded that for the practice to be lawful under GDPR the only option for data controllers will be to obtain consent from the data subjects. However, it is believed by scholars that such consent will be almost impossible to obtain; even if a particularly dominant undertaking is able to obtain the consent, it will most probably take the form of terms and conditions. However, the terms and conditions that provide for discrimination based on the collected data can be considered unfair by the competition authorities and thus prohibited.
On the basis of the aforementioned, it can be concluded that in order for the personalised practice to be prohibited under the EU Competition law as an abuse it has to fulfil certain criteria.
Thus, this paper proposes a test.
1) intends to weaken competition or/and exclude competitors from the market
2) by applying different prices to different customers on exactly the same transactions
3) where prices for each customer are based on the collected data and consumer’s willingness to pay
4) and the undertaking either has not obtained the valid consent for the collection of this data under GDPR or this consent has taken place in the form of unfair terms and conditions
5) in the absence of any economic justification
If the conduct of the undertaking satisfies this, it will automatically mean that this conduct causes a negative effect on consumer welfare as discussed in Sections 1.3 and 2.3.4 of the present paper. Therefore, the abuse of the dominant position can be established, provided that principles of legal certainty and foreseeability are preserved.
This paper aimed at answering the question of whether personalised price discrimination can be prohibited under Article 102 TFEU. The answer to this question seems to be dubious. On the one hand, the practice of personalised pricing is outrageous discrimination because it causes different treatment to people of different background, income, location, habits. Furthermore, it increases the welfare of suppliers while decreasing the welfare of consumers. On the other hand, there is no conclusive evidence that personalised price discrimination is actually taking place and it cannot be certainly said which form will take it in the future. This paper took an extreme approach, looking into personalised pricing as first-degree price discrimination. The reason for taking this approach is simple: nowadays it indeed became possible for companies to know almost every single detail about individuals and their preferences. The lack of evidence at the moment does not mean that the practice is not taking place or that it is impossible. The explanation for this can be the fact that it is not taking place on a large scale yet or that the technologies are not advanced enough yet. However, even if the practice would be proved to take place, this paper illustrated that EU Competition law might not be prepared to tackle it for several reasons. First, practices that directly affect the consumers have never been given enough attention to by the Commission. It is true, however, that most of the cases where both Commission and the Court were involved dealt with the practices having an exclusionary effect. This can be caused by the fact that up to now, companies could not impose the practices that would be making more profit for them by directly affecting the consumers. Instead, the best way to optimize profit used to be the exclusion of the rivals. Second, the EU Competition law has always taken an approach of focusing on the product. However, at the moment, markets are changing to focus on the customers. This is because Big Data brings the possibilities that were not even imaginable 10 years ago.
This paper, in its first section, explained the economic side of the personalised price discrimination. The analysis illustrated that even economic opinion of the scholars developed in the past years. When the personalised price discrimination just came into the play, it was mainly argued that it takes the form of third-degree price discrimination as it was allegedly impossible for companies to be able to discriminate on a first-degree basis. However, now a lot of scholars accepted the fact that first-degree price discrimination can indeed take place on the imperfectly competitive markets. The degree of price discrimination changes the effect on welfare. This paper concluded that if personalised price discrimination takes a form of first-degree price discrimination, then the effect on the consumer welfare is negative, as the distributive outcomes are uneven and most of the surplus is captured by the supplier, leaving consumers with nothing.
The second section firstly explored the possibility of the personalised price discrimination being included into excising abuses of the dominant position under Article 102(a) TFEU and 102(c) TFEU. Under Article 102(a) TFEU it was examined whether personalised price discrimination can fall under ‘excessive pricing’. The outcome of the analysis illustrated that neither existing product-centric test, nor a consumer-centric test would be appropriate for the personalised price discrimination. This is because the essence of personalised pricing is charging some consumers too much while charging others too little. However, Article 102(c) TFEU, which deals with discrimination, also has shown not to be suitable for personalised price discrimination. The main reason for that is that in order for the practice to be prohibited under Article 102(c) TFEU, there should be a ‘competitive disadvantage’. As the case law shows that this criterion is applied consistently if not particularly strictly. Personalised price discrimination, on the other hand, mostly deals with consumers, who are not in competition with each other.
In its last section, this paper explored the possibility of the personalised price discrimination being included in the broad scope of Article 102 TFEU independently of the list included in this Article. In order to establish a test for abuse, this paper explored case law that would fit the characteristics of the personalised pricing. It was derived that in order to establish abuse, first and foremost, the principles of legal certainty and foreseeability should be adhered to. This means that the dominant undertaking has to be able to foresee that its conduct will be seen as illegal by EU Competition authorities. It was revealed in this paper that in order for the personalised pricing to be prohibited as an abuse of the dominant position, the following requirements have to be fulfilled: the conduct of the undertaking should (1) be intended to weaken competition or/and exclude competitors from the market , (2) by applying different prices to different customers on exactly the same transactions (3) where prices for each customer are based on the collected data and consumer’s willingness to pay and (4) and the undertaking either has not obtained the valid consent for the collection of this data under GDPR or this consent has taken place in the form of unfair terms and conditions (5) in the absence of any economic justification. This paper has combined different decisions of the Commission and the case law of the Court in order to establish this test in the context of personalised pricing.
Thus, this paper has concluded that personalised pricing can be prohibited under Article 102 TFEU as a separate abuse of a dominant position as the Court has already assessed all the conditions of the proposed test, but separately. Furthermore, it was concluded that personalised pricing should be prohibited by Article 102 TFEU as it has a negative impact on the consumer welfare, an interest which stands at the core of Article 102 TFEU and European Competition law as a whole.
Our law services are included but limited to:
– Law Essay Writing Services
– Law Dissertation Writing Services
– Law Assignment Writing Services
– Law Coursework Writing Services
– Law Report Writing Service
– BVC/BPTC Writing Service Online