European competition law and personalised price discrimination: a poor fit?
AbstractPersonalised price discrimination is a form of price discrimination which is enabled by the gathering and analysis of large amounts of data in order to offer personalised prices to consumers based on their data profiles. The profiles are based, for instance, on a consumer’s location, browse history, the way they accessed a website or a combination of such data points. With these profiles, firms are able to charge different prices to individual consumers for equivalent transactions, thereby approaching more and more a consumer’s true willingness-to-pay. In many instances, personalised price discrimination can be normatively desirable. For instance, when it improves consumer welfare on the aggregate level and low-end consumers can be reached. However, compared to uniform pricing based on aggregate demand, personalised price discrimination has an ambiguous effect on welfare, which typically differs on a case-by-case basis. There are situations in which this type of discrimination is undesirable. The issue is addressed to varying extent by various European legal regimes. Some academics have labelled competition law as a poor fit to address personalised price discrimination, whereas others see possibilities to enforce the activity as an abuse of dominance. This thesis evaluates whether price discrimination should be covered by competition law, both from a substantive and a procedural perspective, thereby also analysing the extent to which the practice is covered by other regimes. Next to analysing whether personalised price discrimination falls under Article 102 TFEU, the this paper examines whether recent proposals by the Benelux competition authorities for the introduction of new ex ante enforcement tools to tackle abuse of dominance in the data-driven economy can be part of the solution for online price discrimination practices that harm final consumers. The main takeaways of the research include that in the uncommon case that the practice is harmful, it is likely to be covered by at least one legal regime. For example, geographic price discrimination falls under laws such as the Geo-Blocking Regulation, even though its effectiveness remains to be seen. It is feasible, though not easy, to qualify harmful forms of personalised price discrimination as an (exploitative) abuse of dominance. However, competition law’s casuistic approach, potentially helped by the use of algorithmic, ex officio enforcement tools, makes it a candidate to deal with the practice. The Commission and Member State competition authorities should provide guidance documents in order to show under which specific circumstances the practice is abusive. In this way, the authorities can anticipate the proliferation of algorithmic pricing techniques which potentially harm consumers. The suggested use of ex ante remedies without the need to establish an infringement raises questions. If this far-reaching instrument were to be developed further, coming up with an appropriate legal benchmark in order to justify intervention is paramount.
Author: Mr. Drs. Y. (Yannick) van Weeren Supervisor: Dr. Mr. K.J. (Kati) Cseres E-Mail: yannickvanweeren@gmail.com Student ID: 12954306
Table of Content
1. Introduction ………..p. 3 2. What is personalised price discrimination? ………..p. 7 2.1. Welfare effects and normative desirability………..……….p. 10 3. Personalised price discrimination as an abuse of dominance ……….p. 14
3.1. Competition law compared to other legal regimes ………..p. 21 4. Ex ante guidance and remedies to address personalised price discrimination? ………..p. 27 4.1. Guidance papers and letters ……….p. 29 4.2. Ex ante remedies proposal: circumventing the infringement route? ………p. 32 5. Conclusion ………..p. 35 6. Bibliography ………...p. 37 7. Appendix ………p. 41
1. Introduction
Price discrimination is the act where different consumers are charged a different price for similar transactions.1 The availability of data and capabilities of algorithms allow firms to create ever more precise ‘profiles’ of their customers and their willingness-to-pay for a product.2 Their geographic location, online browse or purchase activities or a combination of characteristics are potentially used to offer different prices to different consumer profiles. The act where profiling is used to offer personalised prices for similar transactions will be referred to as Personalised Price Discrimination (“PPD”).
A market study3 conducted by the European Commission (“the Commission”) in 2018 researched four industries4 in order to find out whether systematic and consistent PPD took place. In 6% of the situations a price difference was observed between the personalised and non-personalised scenario. The differences were most prevalent in the markets for airline tickets and hotel rooms. Other studies showed that price discrimination between (groups of) consumers causes strong perceptions of unfairness.5 It is not always clear whether PPD has a positive or
negative influence on final consumers. For example, low-end consumers can often be reached when prices are personalised.6 However, compared to uniform pricing, the effect on aggregate consumer welfare can be negative, especially when output does not increase simultaneously with the implementation of the PPD strategy.7
Hence, it is questionable under which circumstances PPD should be regulated. Provisions from competition law, data protection law, consumer protection law and non-discrimination law all cover certain aspects of PPD.8 Nonetheless, all these different regimes serve different objectives
1 Botta, M., & Wiedemann, K. (2019). To discriminate or not to discriminate? Personalised pricing in online
markets as exploitative abuse of dominance. European Journal of Law and Economics, 1-24.
2 Ezrachi, A., & Stucke, M. E. (2016). Virtual competition. Cambridge, MA: Harvard University Press.
3 European Commission (2018). Consumer market study on online market segmentation through personalised
pricing/offers in the European Union. Published on July 19 2018. Retrieved from:
https://ec.europa.eu/info/publications/consumer-market-study-online-market-segmentation-through-personalised-pricing-offers-european-union_en
4 Involving the online industries for airline tickets, sport shoes, televisions and hotel rooms.
5 Cited by Townley, C., Morrison, E., & Yeung, K. (2017). Big data and personalized price discrimination in EU
competition law. Yearbook of European Law, 36, 690 (n. 6): Xia, L., Monroe, K. B., & Cox, J. L. (2004). The price is unfair! A conceptual framework of price fairness perceptions. Journal of marketing, 68(4), 1-15.; Spiekermann, S. (2006, April). Individual price discriminaton-an impossibility. In Proceedings of the Workshop
on Privacy and Personalization held with the International Conference for Human-Computer Interaction (CHI) (pp. 47-52).
6 OECD Report. (2018). Personalised Pricing in the Digital Era: Background Note by the Secretariat. Retrieved
from: https://one.oecd.org/document/DAF/COMP(2018)13/en/pdf
7 Chiang, R., & Spatt, C. S. (1982). Imperfect price discrimination and welfare. The Review of Economic
Studies, 49(2), 155-181.
and are linked to the concept in a slightly different manner. By some, competition law is regarded as a ‘poor fit’, since it will be difficult to establish an infringement under Article 102 TFEU. A reason for this is the requirement that dominance must be established and that it is difficult to prove which conduct repeatedly contributes to the abuse. What’s more, it is not easy to link the abuse to direct harm to final consumers.9 A decisive factor therein is the fact that the Commission has as of yet not prioritised exploitative abuses of dominance.10 In addition, the European Court of Justice (“ECJ”) stated in the cases T-mobile and GlaxoSmithKline11 that
harm to consumers through price-setting is not a prerequisite for an anticompetitive object and focused on the harm to competition instead.
Another difficulty in tackling PPD through competition law is caused by enforcement tools. PPD is a pricing strategy applied by online platforms in the data-driven economy, enabled by rapidly developing technological advances. The digital environment is dynamic, often requiring fast intervention. Especially when markets are characterised by entry barriers created through data collection, consumer lock-in effects, winner-takes-all tendencies and strong network effects, ex post enforcement can be too slow.12 The fact that first an infringement needs to be
established (which can be a lengthy procedure) before authorities can intervene, can be perceived as too little, too late in order to keep markets contestable and competitive.13 A remarkable recent proposal14 by a Dutch State Secretary contains three proposals to reform the competition rules. Among the proposals, more guidance and ex ante measures were presented as policy aims, which were also expressed15 by the Dutch competition authority, the Authority for Consumers and Markets (“ACM”). The views of the ACM were later again communicated
9 Botta, M., & Wiedemann, K. (2019). To discriminate or not to discriminate? Personalised pricing in online
markets as exploitative abuse of dominance. European Journal of Law and Economics, 1-24.
10 Guidance document on Article 102 TFEU (2009/C 45/02), para. 7
11 Case C-8/08 T-mobile Netherlands BV and Others [2009], paras. 38 and 39; Case C-501/06 P
GlaxoSmithKline [2009], paras. 63 and 64
12 Van Gorp, N., & Honnefelder, S. (2015). Challenges for competition policy in the digitalised
economy. Communications & Strategies, (99), 149.; Katz, M. L., & Shapiro, C. (1994). Systems competition and network effects. Journal of economic perspectives, 8(2), 93-115.; Crémer, J., de Montjoye, Y. A., & Schweitzer, H. (2019). Competition Policy for the Digital Era, final report presented to the European Commission. Retrieved from: https://ec.europa.eu/competition/publications/reports/kd0419345enn.pdf
13 See, for example, the proposal from the Benelux competition authorities: Joint memorandum of the Belgian,
Dutch and Luxembourg competition authorities on challenges faced by competition authorities in a digital world. Published in October 2019. Retrieved from:
https://www.acm.nl/sites/default/files/documents/2019-10/benelux-memorandum-over-toezicht-mededinging-in-digitale-economie.pdf
14 Keijzer, M.C.G. (2019, May 17). Toekomstbestendigheid van het mededingingsinstrumentarium in relatie tot
online platforms [Kamerbrief]. Retrieved from:
https://www.tweedekamer.nl/kamerstukken/brieven_regering/detail?id=2019Z09936&did=2019D20357
15 ACM. (2019, August 8). Extension of enforcement toolkit to increase effectiveness in dealing with
competition problems in the digital economy. Retrieved from:
via a joint memorandum16 by the Benelux competition authorities. The memorandum proposes two forms of ex ante guidance and the far-reaching introduction of an ex ante instrument providing for imposed remedies without the need to establish an infringement. With respect to guidance, the authorities suggest the introduction of guidance papers to tackle specific issues and guidance letters to address issues on a case-by-case level. Could these type of guidance documents prevent PPD practices that harm consumers?*
Both from a substantive and a procedural perspective, it is debatable whether (certain forms of) PPD should be addressed by competition law, considering that establishing an infringement can be a lengthy process. Due to the innovation-driven digital environment it occurs in, there is a chance that PPD will become more common in the future.17 Henceforth, it is relevant to find out whether competition law is the appropriate regime to deal with the conduct, taking into consideration proposals for new ex ante enforcement tools and the potential proliferation of PPD strategies in the future. This leads to the following question:
Taking into account the recent proposals regarding ex ante enforcement tools, is competition law the appropriate regime to deal with unfair and welfare reducing personalised price discrimination strategies as an abuse of dominance?
Subdividing the question leads to a number of sub-questions, which will be covered in different sections. What is PPD and how does it affect competition? Under which circumstances is it normatively (un)desirable? Should competition law cover PPD? Is PPD legal under competition law? Does competition law offer appropriate enforcement tools to deal with PPD? The thesis is structured as follows: Section 2 deals with the definition of PPD and the positive and negative effects it can have on competition. Section 3 evaluates the qualification of PPD as an abuse of dominance. In addition, it discusses briefly how other regimes, such as data and consumer protection laws, and anti-discrimination laws apply to form of PPD. Thereafter, in Section 4,
16 Supra n. 13
* During the writing process of this paper, the Digital Service Act package, an initiative by the Commission, has started to take shape. It is partly inspired by the proposals mentioned in this section. The initiative is currently under public review. The package includes an ex ante regulatory instrument to intervene in ‘gatekeeper’ scenarios and a new market investigation tool. For recent information, see https://ec.europa.eu/info/law/better- regulation/have-your-say/initiatives/12418-Digital-Services-Act-package-ex-ante-regulatory-instrument-of-very-large-online-platforms-acting-as-gatekeepers. A web-article that briefly summarizes the proposal: Bret, A. (2020, June 18). Gatekeepers and digital markets set for further EU regulation. Pinsent Masons. Retrieved from: https://www.pinsentmasons.com/out-law/analysis/gatekeepers-digital-markets-eu-regulation
17 See, for example: OECD Report. (2018). Personalised Pricing in the Digital Era: Background Note by the
the newly proposed ex ante enforcement tools are evaluated with respect to their suitability for the enforcement of abusive PPD.
2. What is personalised price discrimination?
Personalised price discrimination is a form of price discrimination that, enabled by the ability of firms to create data-based profiles of their customers, offers different prices to different (groups of) consumers. As mentioned in the Introduction, firms try to derive the willingness-to-pay of their customers based on various characteristics. In data-driven economies, those characteristics can include the characteristics of the buyers, the characteristics of the good or the (online) behaviour of the buyer. Essentially, these forms of price discrimination correspond to (a mixture of) second and third degree price discrimination as described by Pigou in 1920.18 Pigou distinguished three forms of price discrimination (based on a monopoly model).
First degree price discrimination, or ‘perfect price discrimination’ is the situation where the seller can derive the true willingness to pay of every individual consumer in the market. In reality, this would be an unattainable situation, since it requires perfect information.19 Second
degree price discrimination (or: versioning) is the situation where firms sell different versions of the same good at different prices.20 Second degree price discrimination also includes quantity
discounts and the imposition of a two-part tariff.21 Furthermore, there is third degree price discrimination, which entails pricing based on characteristics or behaviour of groups of customers, such as those referred to in Table 1 below. For instance, students or pensioners share certain characteristics, based on which retailers can personalise the price they charge each of them.22
The definition of PPD used in this paper is based on a definition from an OECD report,23 stating:
“… any [online] practice of price discriminating final consumers based on their personal characteristics and conduct [(which are retrieved through the collection and analysis of data)], resulting in prices being set as an increasing function of consumers’ willingness to pay.”
Slightly narrowing the definition (in brackets) is of importance since the data collected in digital markets enables firms to price-discriminate on a more granular level. In addition, as will be
18 Pigou, A. (1920), The Economics of Welfare, McMillan & Co.; also used in Botta & Wiedemann (2019),
supra n. 9 and Townley (2017), supra n. 5
19 Odlyzko, A. (2004). Privacy, economics, and price discrimination on the Internet. In Economics of information
security (pp. 187-211). Springer, Boston, MA.
20 OECD. (2018). Personalised Pricing in the Digital Era: Background Note by the Secretariat. Retrieved from:
https://one.oecd.org/document/DAF/COMP(2018)13/en/pdf
21 Motta, M. (2004). Competition policy: theory and practice. Cambridge University Press (pp.494-495). 22Graef, I. (2017). Algorithms and fairness: What role for competition law in targeting price discrimination
towards ends consumers. Colum. J. Eur. L., 24, 541.
23 OECD (2018). Personalised Pricing in the Digital Era: Background Note by the Secretariat. Retrieved from:
shown in Section 2.1, this potentially brings along strong economic effects, both positive and negative.
PPD is not to be confused with dynamic pricing, targeted advertising, A/B testing or personalised ranking of offers.24 These activities are similar to PPD, but need to be distinguished, because they do not constitute price discrimination. Dynamic pricing includes price changes that are caused by exogenous shocks in demand and supply and only uniformly affect price levels. Targeted advertising shows personalised advertisements to consumers in order to increase the likelihood of a sale. A/B testing involves testing different prices on consumers in order to gauge responses. In addition, personalised ranking of offers (also referred to as search discrimination) involves differentiated rankings of offers to different consumes, based on their characteristics.25
Dwelling on the specifics of data analytics and machine learning would defeat the purpose of this research. Nonetheless it should be pointed out that the OECD defined three types of data that can be gathered in order to create consumer profiles on the basis of which a price discrimination strategy can be implemented. Especially the ‘inferred’ type should be noted, considering that it enhances the ability of firms to personalise without consumers having consented or being aware of the pricing strategy.
Table 1. Types of data gathered for PPD strategies. Source: directly imported from OECD report (p.11).26
Next to the ability of firms to assess a consumer’s willingness to pay, there are two other market conditions that increase the effectiveness of a PPD strategy. Firstly, the opportunity for
24 Ibid.
25 European Commission (2018). Consumer market study on online market segmentation through personalised
pricing/offers in the European Union. Published on July 19 2018. Retrieved from:
https://ec.europa.eu/info/publications/consumer-market-study-online-market-segmentation-through-personalised-pricing-offers-european-union_en
26 OECD (2018). Personalised Pricing in the Digital Era: Background Note by the Secretariat. Retrieved from:
consumers to arbitrage needs to be absent.27 Arbitrage in this context refers to the reselling of goods or services by consumers with a lower willingness to pay to consumers with a higher willingness to pay. This is likely to be the case for non-transferrable goods or services, such as hotel rooms, airline tickets, online courses or e-books. Secondly, sellers engaging in PPD activities need at least some degree of market power, at least to the extent that they can profitably raise prices above the price level of perfectly competitive markets.28 However, in reality it is highly unlikely that firms lack all market power.29 On top of that, product differentiation also provides some degree of market power. Nevertheless, firms with little market power are also very unlikely to impact market prices through a PPD strategy.30 Hence, markets that are characterised to some extent by economies of scale or scope, entry or switching costs, or direct or indirect network effects will constitute more fertile grounds for implementing a PPD strategy.
Traditional forms of price discrimination are common practice in many industries.31 In
competition jargon, examples include discriminatory input prices, discounts and rebates, tying and bundling, and selective price cuts.32 As will be described in Section 3, PPD can occur,
amongst others, as part of a predatory pricing strategy. In today’s high-paced, data- and innovation-driven markets, online retailers that have some degree of market power are likely candidates to engage in PPD practices.33 As mentioned in the Introduction, there are indications that PPD occurs on a large scale. However, not many reports that indicate PPD practices are cited yet in the literature.34
27 Motta, M. (2004). Competition policy: theory and practice. Cambridge University Press. p.492. 28 Ibid. citing Varian, H. R. (1989). Price discrimination. Handbook of industrial organization, 1, 597-654. 29 Ibid. Motta (2004), p.492.
30 Ibid.
31 Bishop, S. (2005). ‘Delivering Benefits to Consumers or Per Se Illegal? Assessing the Competitive Effects of
Loyalty Rebates’ in The Pros and Cons of Price Discrimination (Swedish Competition Authority, 2005).; Kolasky, W.J. (2004). What is Competition? A Comparison of US and European Perspectives. The Antitrust
Bulletin, 29, 34.
32 Geradin, D., & Petit, N. (2005). Price discrimination under EC competition law: The need for a case-by-case
approach. Collège d’Europe DP, 7(05).
33 OECD (2018). Personalised Pricing in the Digital Era: Background Note by the Secretariat. Retrieved from:
https://one.oecd.org/document/DAF/COMP(2018)13/en/pdf
34 Ibid. Citing: Hogan, K. (2018), Consumer Experience in the Retail Renaissance: How Leading Brands Build a
Bedrock with Data, https://www.deloittedigital.com/us/en/blog-list/2018/consumerexperience-in-the-retail-renaissance--how-leading-brand.html.; Clifford, S. (2012), “Shopper Alert: Price May Drop for You Alone”, The New York Times,
https://www.nytimes.com/2012/08/10/business/supermarkets-try-customizing-prices-forshoppers.html.; Valentino-DeVries, J., J. Singer-Vine and A. Soltani (2012), “Websites Vary Prices, Deals Based on Users’ Information”, The Wall Street Journal,
2.1. Welfare effects and normative desirability
Although economic effects of PPD are comparable to the traditional forms of price discrimination, they are likely to be stronger, assuming that data analytics and machine learning enable firms to compile more precise profiles of their customers. Nevertheless, PPD will probably be closer to imperfect (second or third degree) price discrimination than perfect (or: first degree) price discrimination, since the information sellers observe is merely correlated with their true willingness to pay and not a one-on-one representation.35
First of all, in an environment of imperfect competition (roughly somewhere between perfect competition and a monopolistic state of competition) PPD can improve static efficiency by improving total welfare (the sum of consumer and producer surplus), assuming that they expand output and serve more low-end consumers.36 This can be observed when comparing the
situation of uniform pricing (Graph 1, see Appendix) to imperfect discrimination (Graph 2, see Appendix), where the green surface represents the consumer surplus and the orange surface the producer surplus. Graph 2 clearly shows higher total welfare. However, one can imagine that it depends on the steepness of the price line37 and by how much (or whether at all) the seller is willing to expand output, how much total welfare will increase and what the distribution between consumer and producer surplus will be. Consequently, the effect of a PPD strategy on consumer surplus is ambiguous and a direct function of market power. It has to be noted that these types of economic models are subject to many assumptions. However, they tend to show the mechanisms and (welfare) effects of pricing strategies. In case of a monopoly, it is argued that third degree price discrimination will typically decrease consumer surplus.38
In reality, most markets that are suitable candidates for PPD strategies are characterised by imperfect competition. In these markets, the gathering of information about consumers is not an exogenous factor, since consumers potentially behave strategically by deceivingly showing or hiding information. There is always some degree of an incentive for consumers to switch sellers, depending on elasticities of demand between those sellers. This brings into the equation
35 Townley, C., Morrison, E., & Yeung, K. (2017). Big data and personalized price discrimination in EU
competition law. Yearbook of European Law, 36, 690
36 OECD (2018). Personalised Pricing in the Digital Era: Background Note by the Secretariat. Retrieved from:
https://one.oecd.org/document/DAF/COMP(2018)13/en/pdf
37 The more it rotates clockwise, the less consumer surplus is left, up until the point where the price line equals
demand, which would be the point where perfect (or first degree) price discrimination is reached.
38 Aguirre, I., Cowan, S., & Vickers, J. (2010). Monopoly price discrimination and demand curvature. American
Economic Review, 100(4), 1601-15.; see also Bergemann, D., Brooks, B., & Morris, S. (2015). The limits of
price discrimination. American Economic Review, 105(3), 921-57: both articles cited by Townley et al. (2017),
search and switching costs that are experienced by consumers, taking into account the specific preferences that each (group of) consumers has for each alternative seller in the market and the likelihood that they are making an effort to search for the good (for whatever reason) and compare it to other goods.39 Both high switching costs and search costs lead to inelastic demand. The more inelastic demand from a certain (group of) consumers, the less sensitive they are to price changes, thereby making them more attractive targets for firms that contemplate a PPD strategy. Needless to say that switching and search costs are individual-specific and could lead to both reductions and improvements of consumer welfare. It needs to be taken into account though, that PPD amplifies the effect on consumer welfare. This makes intervention desirable in markets where PPD is likely to harm consumers. However, due to conditions of imperfect competition and the fact that one cannot generalise the effects of PPD on consumer welfare on the aggregate level, each case should be examined specifically.40
The desirability of PPD has been assessed from a normative perspective. Townley et al. (2017)41
evaluated PPD in the context of economic efficiency (with the goal to maximise consumer welfare) and fairness. Even though competition law pursues multiple goals42 (including fairness
and consumer welfare), consumer welfare is thought to be the more prevalent one by various academics (including Townley et al.).43 Whereas the ambiguous effect on welfare has been demonstrated above, fairness is much less easily defined. Since there is not one coherent definition, ‘fairness’ in relation to PPD can be interpreted in various ways.44 It can represent an overarching principle, embodying the social rationale behind competition law principles.45 The principle is repeatedly mentioned by Commissioner Vestager in speeches. On the 13th conference of the Global Competition Law Center she expressed how the fairness ‘standards’ are the moral fundaments of competition law that fairness is not about the application of competition law principles.46 Hence, fairness in her view is about creating a fair outcome for
39 Supra n. 35 40 Ibid. 41 Ibid.
42 Van den Bergh, R. J. (2017). Comparative competition law and economics. Edward Elgar Publishing. Ch. 3;
Van den Bergh, R., & Camesasca, P. D. (2001). European competition law and economics: a comparative
perspective. Intersentia NV. pp. 1-5; Gerber, D. J. (1998). Law and competition in twentieth century Europe: protecting Prometheus. Oxford University Press.
43 Cseres, K. (2006). The controversies of the consumer welfare standard. Competition Law Review, 3(2),
121-173.; Orbach, B. Y. (2011). The antitrust consumer welfare paradox. Journal of Competition Law and
Economics, 7(1), 133-164.
44 Akman, P. (2012). The concept of abuse in EU competition law: Law and economic approaches. Bloomsbury
Publishing, pp. 231-300
45 Gerard, D. (2018). Fairness in EU Competition Policy: Significance and Implications. Journal of European
Competition Law & Practice, 9, no. 4, 211-212
consumers and achieving “a fair share for consumers”. However, she also stresses that the mere notion that a conduct is unfair is not a sufficient condition for the competition rules to apply. DG COMP Director General Laitenberger mentioned during his speech at the same event that fairness is not an operational concept and that it should not replace analysis based on evidence.47 Perhaps this should lead to a more fairness-oriented approach to competition law when applied to PPD strategies.
Fairness can (inter alia) be regarded in terms of (un)fair dealing, as demonstrated by Townley et al. (2017). They identified two situations in which efficiency and fairness can be reconciled.48 Firstly, the circumstance where consumers wrongly believe that sellers apply a uniform pricing strategy may constitute both a form of unfair dealing and a reduction in consumer welfare. Secondly, there are groups of consumers who fail to switch between sellers where it would be economically rational to do so, due to a lack of sophistication or digital literacy. In case these sellers mistakenly interpret that choice as a manifestation of their preferences and willingness-to-pay, the consumer may be charged a higher price than better informed consumers. Whereas the first situation can be easily resolved by informing consumers of the PPD strategy, intervention in the second situation will be more complicated, since sellers cannot differentiate between high-value customers and unsophisticated or impulsive customers, or customers who fail for some reason to search for the best offer.49 Nevertheless, there are certainly situations in which PPD can be both unfair and welfare reducing. Both fairness and consumer welfare can be regarded as objectives of EU competition law.
Since both fairness and welfare are goals of competition law, unfair and (consumer) welfare reducing forms of PPD could be violating competition rules. By intervening in a situation where consumer welfare is reduced on the aggregate level, an unfair situation for final consumers can be remedied. However, is a PPD practice that is unfair and welfare reducing at the same time also against the competition rules? Under which circumstances would that be the case? Should it be a requirement that competition as such is harmed? This could be the case if a purely exploitative PPD strategy that directly harms consumers is combined with an exclusionary strategy such as predatory pricing or margin squeeze that demonstrates harm to competition. Should the Commission reconsider its priorities and also prioritise exploitative abuses? Already when the Guidance on the Commission’s enforcement priorities in applying Article 82 came
47 Ibid.
48 Townley, C., Morrison, E., & Yeung, K. (2017). Big data and personalized price discrimination in EU
competition law. Yearbook of European Law, 36, 690.
out in 2009, it was suggested in reaction to the prioritisation of exclusionary abuses that exploitative effects should be the mandatory test under the effects-based approach to establish anticompetitive effects. However, at the same time the detection of exploitative effects an sich would not be a sufficient condition to establish an abuse.50 In an OECD note, it is even suggested to define PPD as a sui generis abuse of dominance with a new theory of harm. For instance, if the amount of alternatives a consumer has is limited because of his demographic characteristics (such as postal code), it could be that PPD in that situation leads to the exploitation of vulnerable groups of consumers.51
50 Akman, P. (2009). Searching for the long-lost soul of Article 82EC. Oxford Journal of Legal Studies, 29(2),
267-303.
51 OECD. (2018). Personalised Pricing in the Digital Era – Note by the Netherlands. Retrieved from:
3. Personalised price discrimination as an abuse of dominance
Generally, in order for a conduct to fall under Article 102 TFEU52 it has to be established that the ‘infringer’ has a dominant position in the relevant market. In addition, the conduct has to constitute an abuse and the abuse must have anti-competitive effects and bring harm to consumers which is not outweighed by efficiencies to the benefit of consumers (objective justifications).53 An OECD report also suggests that “the persistency of the effects” and “the source of discrimination” have to be assessed and identified. The former with the idea in mind to prevent intervention in situations where the market is likely to solve the issue itself and the effects are temporary in nature. The latter can be useful in designing the appropriate remedies, since various factors potentially facilitate an effective PPD strategy, such as price transparency, consumer inertia and data gathering.54
Under competition law and according to competition economics, when PPD directly harms final consumers, it is rather an exploitative than an exclusionary practice.55 PPD can mainly be
characterised as an exploitative abuse, even though there are until today no known cases in which price discrimination was recognised as such.56 Where PPD strategies could both exploitative and exclusionary elements, the Commission has taken a ‘hands-off’ approach and failed to make a clear distinction in legal treatment between them in its decision-making practice. The case law by the European Courts does not provide a coherent legal framework either.57 According to the (2009/C 45/02) Guidance on the Commission’s enforcement priorities in applying Article 82 of the EC Treaty to abusive exclusionary conduct by dominant undertakings (“the Guidance document”), behaviours such as PPD which are directly
52 Article 102
(ex Article 82 TEC)
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.
Such abuse may, in particular, consist in:
(a) directly or indirectly imposing unfair purchase or selling prices or other unfair trading conditions; (b) limiting production, markets or technical development to the prejudice of consumers;
(c) applying dissimilar conditions to equivalent transactions with other trading parties, thereby placing them at a competitive disadvantage;
(d) making the conclusion of contracts subject to acceptance by the other parties of supplementary obligations which, by their nature or according to commercial usage, have no connection with the subject of such contracts.
53 Sears, A. M. (2019). The Limits of Online Price Discrimination in Europe. Colum. Sci. & Tech. L. Rev., 21, 1. 54 OECD (2018). Personalised Pricing in the Digital Era: Background Note by the Secretariat. Retrieved from:
https://one.oecd.org/document/DAF/COMP(2018)13/en/pdf
55 Supra n. 8 and 9
56 Ritter, C. (2019). Price discrimination as an abuse of a dominant position under Article 102 TFEU:
MEO. Common Market Law Review, 56(1), 259-274.
57Graef, I. (2017). Algorithms and fairness: What role for competition law in targeting price discrimination
exploitative of consumers, are liable to infringe Article 102.58 However, although the Commission may decide to intervene if the protection of consumers cannot be otherwise properly be ensured, the Commission gives priority to exclusionary conduct. “For the purpose
of providing guidance on its enforcement priorities the Commission at this stage limits itself to exclusionary conduct and in, particular, certain specific types of exclusionary conduct which, based on its experience, appear to be the most common.” Only 7 percent of cases investigated
by the Commission between 2000 and 2017 were exploitative abuses.59 With that said, the possibility that PPD can be enforced as an exploitative abuse of dominance is certainly not excluded. Besides, Guidance papers are not binding on national courts, nor are they binding before the Courts of the European Union or NCAs.60 In addition, the Guidance paper to Article 102 stems from 2009, a year in which both the stage of development of technology and likelihood of price discrimination strategies applied by algorithms would not be representative of the possibilities today. The Commission does not seem to give a definitive answer to the question how it will deal with PPD.
Notably, there is a difference between the exploitative abuse where consumers are directly exploited as a consequence of PPD and the situation where a firm price-discriminates between downstream firms (without the incentive to exclude rivals) which results in a distortion of competition, which in turn leads to the exploitation of consumers. However, the situation in which the discrimination of downstream firms leads to the exploitation of consumers is very narrow and usually requires a test for exclusionary effects.61 When taking the consumer welfare approach, a reduction in aggregate consumer welfare leads to the establishment of the exploitative element. It proves harm to consumers. This would correspond to a more effects-based, economic approach, as was advocated by AG Wahl and later accepted by the ECJ in the
Intel case.62 Contrarily, a form-based approach with respect to PPD as an abuse of dominance would imply that after a firm is found to be dominant, it has a ‘special responsibility’63 not to harm competition and henceforth cannot engage in certain activities (such as price
58 Guidance document para. 7
59 Dethmers, F., & Blondeel, J. (2017). EU enforcement policy on abuse of dominance: Some statistics and
facts. European Competition Law Review, 38.; Cited in OECD (2018). Personalised Pricing in the Digital Era: Background Note by the Secretariat. Retrieved from:
https://one.oecd.org/document/DAF/COMP(2018)13/en/pdf
60 See, for example Case C-189/02 P Dansk Rørindustri [2008], paras. 209-211
61 OECD Report. (2016). Executive Summary of the Roundtable on Price Discrimination: Annex to the
Summary Record of the 126th meeting of the Competition Committee. Retrieved from: https://one.oecd.org/document/DAF/COMP/M(2016)2/ANN5/FINAL/en/pdf
62 Case C-413/14 P Intel [2017] and Opinion of Advocate General Wahl [2016]
63 Para. 1 of the Guidance on the Commission’s enforcement priorities in applying Article 82 of the EC Treaty to
discrimination, tying products or predatory pricing). Due to the ambiguous effect of PPD on welfare, as shown in the previous section, an effects-based approach would seem a logical approach, excluding a per se prohibition of the conduct.
It could very well be that an exploitative PPD strategy that harms consumer welfare is combined with an exclusionary strategy, such as predatory pricing, which would entail offering lower prices (below average avoidable cost and long run average incremental cost) to a competitors’ customers. Other possibilities include, amongst others, selective discounts and fidelity rebates.64 However, the technological developments and likely increase of PPD practices in the future, will make it relevant to reconsider the focus of the enforcement of Article 102 on exploitative abuses, where it has the potential to harm consumers directly. It would entail a change in enforcement choice and does not require amending the Treaty.65
Article 102(a) TFEU deals with directly or indirectly imposing unfair purchase or selling prices or other unfair trading conditions, including excessive pricing. Theoretically, in case a PPD strategy applied by a dominant firm leads to the exploitation of vulnerable groups of consumers, the conduct may be tackled by Article 102(a). This innovative interpretation was offered in an OECD note.66 However, it focusses on the unfairness of the price more generally and not explicitly on the price discrimination between consumers. On the other hand, it may refer more directly to the unfair treatment on final consumers, whereas it is suggested by some that Article 102(c) TFEU is targeted at price discrimination by a dominant firm against downstream customers.67 How to see this in light of the MEO case discussed further in this section? In that case, it was stated68 by the ECJ that discriminatory pricing (in relation to Article 102(c)) can only be abusive when it tends to distort competition. However, the use of Article 102(a) has as an advantage that it directly mentions the unfairness of the price and is not prone to the difficulties described below.
Geradin and Petit (2005) suggest that Article 102(c) TFEU should be used for forms of discrimination between customers of a non-vertically integrated firm, implying that the article
64 Motta, M. (2004). Competition policy: theory and practice. Cambridge University Press, pp.499-500 65 Botta, M., & Wiedemann, K. (2019). To discriminate or not to discriminate? Personalised pricing in online
markets as exploitative abuse of dominance. European Journal of Law and Economics, 1-24.
66 OECD. (2018). Personalised Pricing in the Digital Era – Note by the Netherlands. Retrieved from:
https://one.oecd.org/document/DAF/COMP/WD(2018)124/en/pdf
67Graef, I. (2017). Algorithms and fairness: What role for competition law in targeting price discrimination
towards ends consumers. Colum. J. Eur. L., 24, 541.
should be used secondary rather than primary line of injury.69 Article 102(c) TFEU70 provides that the condition for (personalised) price discrimination to actually constitute an abuse is that if a firm differentiates prices for ‘equivalent transactions’ and thereby placing certain customers at a ‘competitive disadvantage’ when compared to ‘other trading partners’. In the United Brands case the ECJ set out the conditions for ‘equivalent transactions’.71 In British Airways, these
conditions were put into practice where the sale of airline tickets, albeit to different destinations, were marked as equivalent transactions, since both the cost of supply and the nature of the service were equivalent.72
The ‘competitive disadvantage’ does not have to be quantified by competition authorities, but the conduct should be “capable of distorting competition”73 and it is argued that this implicates
“potential” rather than “purely hypothetical” harm as the standard of proof.74 In MEO the ECJ
mentioned75 that ‘all relevant circumstances’ should be taken into account, which is approximately (this is debatable)76 in line with the effects-based approach set out in the Intel
case.77 The mere fact that one firm is charged a higher price than another is no sufficient ground
for an infringement, since it still has to be tested whether competition is distorted.78 Placing
trading partners at a ‘competitive disadvantage’ as in Article 102(c) TFEU has been interpreted by the legal literature as a sign that this provision protects consumers, or at least downstream customers.79 Hence, it is argued that the article provides protection against secondary line injury. On the other side, notwithstanding, since the finding of an abuse is hereby linked to the presence of a downstream market in which the customer will be active, it is thought that the provision protects small purchasers against large sellers and hence provides more primary line protection.80 In the American Robinson-Patman Act, the requirement to establish harm to a competitor is defined as a ‘significant but unwarranted competitive disadvantage’, which is further defined as the absence of other sources of supply. With respect to Article 102(c), the requirement is less clearly defined.
69 Geradin, D., & Petit, N. (2005). Price discrimination under EC competition law: The need for a case-by-case
approach. Collège d’Europe DP, 7(05).
70 Supra n. 51
71 Case 27/76 United Brands v Commission [1978], par. 228-229
72 Case C-95/04 P British Airways plc v Commission [2007], paras. 136-141
73 Case C-525/16 MEO [2018], par. 37
74 Supra. n. 54
75 Case C-525/16 MEO [2018], paras. 27-31
76 Supra. n. 50; Supra n. 56 77 Case C-413/14 P Intel [2017]
78 Whish, R., & Bailey, D. (2018). Competition law. Oxford University Press, p. 782. 79 Supra n. 44: p.241 , including the literature cited there (notes 46-50)
Where the Article 102(c) speaks of ‘other trading parties’ and placing them at a ‘competitive disadvantage’, it remains unclear whether harm to consumers is also covered, since it appears to cover relations between undertakings only. It has been argued that discrimination between consumers by an undertaking or group of undertakings is caught by Article 102(c). The argument is that the conception of ‘trading partners’ can be interpreted broadly, meaning that consumers can also be included.81 Others believe that the prohibition of applying dissimilar conditions to equivalent transactions should only apply to situations in which an upstream firm sells at different prices to two competing downstream firms, excluding situations related directly to consumers, especially since consumers can only be put at a ‘competitive disadvantage’ in cases of true scarcity, which are rare.82 I think that consumers should be
included in the definition of ‘other trading partners’ and that putting them at a competitive disadvantage can easily be broadly interpreted, as explained by Akman. In case Article 102(c) does not catch the conduct, it can always be dealt with through the broader Article 102(a) or the general provision of Article 102.83
Even though EU courts have never ruled out the application of Article 102(c) TFEU to price discrimination as an exploitative abuse of dominance and the CJEU in the MEO case84
recognised that dominant undertakings have no interest in discriminating between trading partners in the downstream markets (theoretically also including consumers) - since it would negatively influence its reputation - no EU court ever explicitly specified that consumers can be regarded as ‘targets’.85 In fact, AG Wahl in MEO suggested that exploitative price
discrimination is extremely rare and that Article 102(c) TFEU should only cover discrimination in the primary line of injury.86 Only in case of, for example, very dominant online platforms or markets characterised by high entry barriers, should Article 102 be used, appears to be the view.87 In addition, it is debatable to what extent consumers can be put at a competitive disadvantage.
81 Akman, P. (2009). Searching for the long-lost soul of Article 82EC. Oxford Journal of Legal Studies, 29(2),
267-303.
82 Supra n. 44: p.48 supported by: O'Donoghue, R., & Padilla, J. (2014). The law and economics of Article 102
TFEU. A&C Black (pp.247-249)
83Graef, I. (2017). Algorithms and fairness: What role for competition law in targeting price discrimination
towards ends consumers. Colum. J. Eur. L., 24, 541.
84 Case C-525/16 MEO [2018], para. 35 in line with AG Wahl’s Opinion paras. 79-80
85 Botta, M., & Wiedemann, K. (2019). To discriminate or not to discriminate? Personalised pricing in online
markets as exploitative abuse of dominance. European Journal of Law and Economics, 1-24.
86 Ibid.
87 Botta, M., & Wiedemann, K. (2018). EU Competition Law Enforcement vis-à-vis Exploitative Conducts in the
Data Economy Exploring the Terra Incognita. Max Planck Institute for Innovation & Competition Research
Indeed, competition authorities have been reluctant to investigate exploitative abuses of dominance, arguably due to the (by some perceived as) high burden of proof and potential overlap with sector regulation.88 They tend to give priority to exclusionary abuses. In decisions
Football World Cup89 and Deutsche Post90 the European Commission sanctioned parties for
discrimination in the processes of buying tickets and delivering post, respectively. Remarkably, in paragraph 100 of Football World Cup, the Commission states that the protection of consumer interest can be achieved by:
“..prohibiting conduct by dominant undertakings which impairs free and undistorted competition or which is directly prejudicial to consumers.”
Even though some have called for a limited application of Article 102 to exploitative abuses, claiming that an effect on the market should be established first, others claim that it just needs a refined application, advocating a two-step test, where harm to consumers is established after the consideration of potential anticompetitive effects. Combatting exploitative abuses of dominance is both consistent with the case law, as well as the wording of the Treaty.91
According to Akman, it probably does not make sense to put a prohibition of discrimination in Article 102 TFEU, considered that economics suggests that effects of price discrimination on welfare are ambiguous and hence require a case-by-case analysis. She assessed the phenomenon in terms of ‘welfare’ and ‘fairness’.92 She holds the opinion that even though Article 102
provides a prohibition instead of an instruction to balance welfare and fairness requirements, this is what a judge should do in order to assess whether (personalised) price discrimination qualifies as an abuse.
How should competition law respond to a clash of welfare and fairness considerations? It is argued that personalised price discrimination should be lawful, i.e. not violating Article 102, in case it enhances both consumer welfare and fairness. As a corollary, personalised price discrimination that decreases both fairness and consumer welfare, can be regarded as violating Article 102. However, there are also situations where consumer welfare increases as a consequence, but fairness decreases or vice versa. What do we make of that, assuming that all requirements of Article 102(c) have been fulfilled? For instance, when an e-commerce firm
88 Ibid.
89 Case COMP/36.888 Football World Cup 1998 [2001]
90 Case COMP/38.745 Deutsche Post [2004]
91 Supra n. 76
(online platform) operates a discriminating pricing algorithm which leads to an increase in aggregate consumer surplus (welfare), but at the cost of some minority group (e.g. elderly) that fails to switch to a supplier that matches their preferences at a lower price due to lack of digital literacy (‘Situation I’).93 The converse situation could look like a retailer who sells a product to
both men and women94 at the same price (and where a pricing algorithm could increase consumer surplus), but at the cost of gender equality (‘Situation II’).
Townley et al. (2017) developed the approach that consumer welfare and the establishment of the single market should be considered the most prominent goal of Article 102 at the stage of establishing an ‘abuse’, since it is regarded by many as the quickest way to provide high levels of output, lower prices and greater variety of choice for consumers and it fosters the allocation of resources to the users who value them the most.95 It is more effective, since fairness “does not render itself to rational and operational rules that also provide the undertakings with sufficient legal certainty.”96 When the notion of fairness applies in relation to the article, it
should be clear how fairness relates to the abuse, in order to prevent that for different situations different notions of fairness are used. This would potentially lead to arbitrary outcomes. Achieving ‘a fair share for consumers’ may be a way in which fairness applies to the article. For example, if a strategy leads to a decrease in aggregate consumer welfare, this could be perceived as not allowing a fair share for consumers. The ECJ focuses at the same time on goals such as freedom of competition and protecting a market’s structure. They also propose that fairness considerations should be dealt with at the ‘defences’ stage and should be protected by
“legal regimes that are directly concerned with the goals of ensuring collective fairness and distributive justice.”97 Hence, from that perspective fairness considerations can only be dealt
with in case an abuse is established.
That approach implies that Situation I would not constitute an abuse according to Article 102, since it enhances consumer welfare. In the approach suggested by Townley and associates, this would imply that the part of the situation that undermines fairness should be dealt with by EU rules that are directly oriented towards promoting fairness, such as non-discrimination (or other regimes described in Section 3.1). In Situation II, where welfare is reduced for the benefit of
93 Supra n. 76
94 Who have on the aggregate level different demand curves, with a different willingness-to-pay at the same
price. Hence, charging them the same price would result in a suboptimal consumer surplus.
95 Supra. n. Townley p.40; Motta, M. (2004). Competition policy: theory and practice. Cambridge University
Press (pp. 17-22).
96 Supra n. 44 p.262 97 Supra n. 44
gender equality, an abuse could be established. Where the burden of proof rests on the competition authority, the dominant firm can justify the behaviour at a later ‘defences stage’ by putting forward ‘objective justifications’ (which are rarely accepted in practice).
To conclude, there are some obstacles in qualifying certain forms of PPD as an abuse of dominance. In addition, the ECJ and the Commission do not provide a coherent legal framework where it concerns abuses that directly harm final consumers and not indirectly through an exclusionary abuse. Technically, it is feasible to apply Article 102 TFEU to forms of PPD that harm final consumers. However, the use of competition law for PPD should also be regarded in relation to the extent other regimes cover certain aspects of the behaviour where final consumers are harmed. Of course, competition law should add to the remedies already provided by other regimes. In the next section, the application of other regimes to certain aspects of PPD will be discussed.
3.1. Competition law compared to other legal regimes
Consumer laws, data and privacy protection laws, non-discrimination laws and competition laws98 all cover certain aspects of PPD.99 All areas try to protect consumers or their welfare.
Although every regime serves separate objectives, has different enforcement mechanisms and vary in scope of application, it is worth mentioning that consumer law, data protection law and competition law are connected through a number of “family ties”, which becomes especially apparent in a data-driven economy.100 It has, for instance, been suggested to connect these regimes through the notion of ‘fairness’ as a common underlying normative principle.101 Other suggestions include the use of data protection and consumer law principles for the enforcement of competition law.102 Whereas consumers tend to dislike price discrimination, regulators are
98 It may be the case that there are other provisions or laws that also influence the legality of PPD. However, the
author will in that case not be aware of any.
99 Sears, A. M. (2019). The Limits of Online Price Discrimination in Europe. Colum. Sci. & Tech. L. Rev., 21, 1.;
See also OECD. (2018). Personalised Pricing in the Digital Era: Background Note by the Secretariat. Retrieved from: https://one.oecd.org/document/DAF/COMP(2018)13/en/pdf
100 Costa-Cabral, F., & Lynskey, O. (2017). Family ties: the intersection between data protection and competition
in EU Law. Common Market Law Review, 54(1), 11-50.; Graef, I., Clifford, D., & Valcke, P. (2018). Fairness and enforcement: bridging competition, data protection and consumer law. International Data Privacy
Law, 8(3), 200-223.
101 Graef, I., Clifford, D., & Valcke, P. (2018). Fairness and enforcement: bridging competition, data protection
and consumer law. International Data Privacy Law, 8(3), 200-223.
102 Botta, M., & Wiedemann, K. (2019). To discriminate or not to discriminate? Personalised pricing in online
markets as exploitative abuse of dominance. European Journal of Law and Economics, 1-24.; Costa-Cabral, F., & Lynskey, O. (2017). Family ties: the intersection between data protection and competition in EU
hesitant to address PPD directly.103 This section covers briefly the application of data protection, consumer and non-discrimination laws (in that order) to PPD.
Regarding data protection laws, the General Data Protection Regulation (“GDPR”) and the ePricacy Directive are part of the legal framework that is deemed relevant for PPD.104 Firms use (mostly) cookies to gather ‘personal data’ about their potential customers.105 The Breyer case106 has determined that even when information that allows the identification of an individual is stalled at a third party, an IP address can still constitute personal data. Personal data can be used to fuel PPD strategies. Data and privacy protection laws are to a high degree consent-based and require firms to inform their customers about automated decision-making and data processing and have to ask permission to use cookies.107 The goal is to create transparency for consumers. However, there is one significant problem with asking the consumer for consent or informing them about a certain practice (such as PPD). Who reads these kind of notifications in detail, let alone thinks of its implications? With respect to cookies, is it not true that the ‘accept all’ button is a more attractive option than customising the range of optional cookies that you opt-in or opt-out for? If a consumer explicitly consents with the PPD strategy, firms can use it. In addition, it will be difficult to exactly describe to consumers the working of the machine learning and data analytics techniques.108 Apparently, as one study suggests, 87 percent of consumers decided to ignore cookie notification banners, whereas only 0,73 percent decided to opt out.109 The newly announced ePrivacy Regulation allows consumers to set their preferences and choose which type of cookies they consent to. Many browsers, including Google Chrome, already announced to block out third party cookies entirely.110 However, they do not exclude first party cookies (cookies that are stored by the website that consumers visit directly) and activities such as fingerprinting, cache inspection, navigation tracking and
103 Supra n. 90, citing European Commission (2018). Consumer market study on online market segmentation
through personalised pricing/offers in the European Union. Published on July 19 2018. Retrieved from:
https://ec.europa.eu/info/publications/consumer-market-study-online-market-segmentation-through-personalised-pricing-offers-european-union_en
104 General Data Protection Regulation (EU) 2016/679, ePrivacy Directive 2002/58/EC (amended in Directive
2009/136/EC); see also Sears (2019)
105 Zuiderveen Borgesius, F. J. (2015). Personal data processing for behavioural targeting: which legal
basis?. International Data Privacy Law, 5(3), 163-176.
106 Case C‑582/14 Breyer [2018]
107 Articles 12-14 and 22 GDPR, Article 5(3) ePrivacy Directive. 108 Supra n. 99
109 Howlett, D. (2014). Is the EU cookie law providing useful? Retrieved from:
https://diginomica.com/eu-cookie-law-proving-useful
110 Slefo, D. (2020, February 4). Google Chrome begins third party cookie evolution. Retrieved from:
network level tracking, which constitute comprehensive cookie workarounds.111 All these activities still allow PPD strategies to be implemented. It also depends on the effectiveness of data protection authorities, which leaves a lot to be desired in many cases, especially considering the fact that many authorities suffer from insufficient capacity.112
In addition, there is consumer protection law that may apply to PPD strategies. Consumer protection law supports consumers as the weaker party in a transaction, thereby eliminating barriers to the internal market.113 An advantage is that it directly applies to the B2C relation. Consumer law includes legislative acts, such as the Unfair Contract Terms Directive (“UCTD”)114, Unfair Commercial Practices Directive (“UCPD”)115 and the Consumer Rights Directive (“CRD”)116, which have recently been amended as part of the ‘New Deal for Consumers’ by Directive (EU) 2019/2161 (which has to be transposed before 28 November 2021).117 The provisions of the CRD will not be very useful, since it requires the consumer to be completely aware of the PPD strategy and at the same time of an offer considerably cheaper elsewhere or for other consumers on the same site.118 Next, there is the UCPD. Its provisions
are ‘fairness’-based and broadly speaking there are two types of unfair behaviour. Either ‘aggressive’ or ‘misleading’ behaviours.119 However, firms are allowed to implement PPD
strategies, as long as they inform the consumer.120 Again, who is going to check how the algorithm exactly works? In addition, it is difficult to explain the working of such an algorithm to consumers. Even if it is possible to do so, it will be very unlikely that consumers will bother
111 https://www.chromium.org/Home/chromium-privacy/privacy-sandbox
112 See, for instance, this article in the Dutch Financial Times, where specifically the consent-based approach and
the authorities’ lack of clout were criticised: Piersma, J., Pols, M. (2020, May 25). Privacywet mist na twee jaar nog steeds tanden. Het Financieel Dagblad. Retrieved from: https://fd.nl/ondernemen/1345538/privacywet-mist-na-twee-jaar-nog-steeds-tanden
113Graef, I. (2017). Algorithms and fairness: What role for competition law in targeting price discrimination
towards ends consumers. Colum. J. Eur. L., 24, 541; citing K.J. Cseres, The Controversies of the Consumer
Welfare Standard, COMPETITION L. REV., 30.
114 Council Directive 93/13/EEC of 5 April 1993 on unfair terms in consumer contracts
115 Directive 2005/29/EC of the European Parliament and of the Council of 11 May 2005 concerning unfair
business-to-consumer commercial practices in the internal market and amending Council Directive 84/450/EEC, Directives 97/7/EC, 98/27/EC and 2002/65/EC of the European Parliament and of the Council and Regulation (EC) No 2006/2004 of the European Parliament and of the Council
116 Directive 2011/83/EU of the European Parliament and of the Council of 25 October 2011 on consumer rights,
amending Council Directive 93/13/EEC and Directive 1999/44/EC of the European Parliament and of the Council and repealing Council Directive 85/577/EEC and Directive 97/7/EC of the European Parliament and of the Council
117 Directive (EU) 2019/2161 of the European Parliament and of the Council of 27 november 2019 - amending
Council Directive 93/13/EEC and Directives 98/6/EC, 2005/29/EC and 2011/83/EU of the European Parliament and of the Council as regards the better enforcement and modernisation of Union consumer protection rules
118 Supra n. 99 (Sears), p.16
119 Article 5(4) UCPD; Supra n. 99, p. 16
120 Ibid. citing: European Commission, Commission Staff Working Document. (2016). Guidance on the
reading it. What’s more, Article 5(2) UCPD requires that the PPD strategy “distorts or is likely
to distort the economic behaviour with regard to the product of the average consumer whom it reaches or to whom it is addressed, or of the average member of the group when a commercial practice is directed to a particular group of consumers.” According to Sears (2019), it will be
unlikely that consumer behaviour is distorted, since consumers will still be able to compare to prices on other websites. Besides, it will prove difficult to define the ‘average consumer’ and the impact of prices on that consumer.
With respect to the UCTD, there might be a chance that PPD can be enforced, since Article 33 CRD has amended the UCTD, by stating that Member States are required to notify the Commission, if they were to include the adequacy of a price in assessment of unfair contract terms.121 However, only Finland, Sweden and Portugal decided to do so. Does this indicate that other Member States hold the view that other fields of law are better equipped to deal with a potential negative outcome of PPD? There is more to the consumer protection Directives then described in this section. However, due to limited space, covering them in more detail would go beyond the scope of this paper.
Yet another law regime covers PPD strategies, namely a set of non-discrimination laws. Some of these laws may apply where the strategy applies to disadvantaged groups within a society. It may be difficult to detect the group of victims of the strategy where its constituents are dispersed over and not concentrated to a sufficient extent within one geographic area.122 Many of these non-discrimination laws have their roots in international and European human rights laws, which are not likely to be effective in relation to PPD strategies.123 However, they inspired EU legislative acts. Some of these acts see to discrimination in employment settings. Others extend it to goods and services. In particular, gender, race and nationality or place of residence are covered by these provisions. Claims can be made before national courts and is not available as an EU law procedure. Nonetheless, the courts of the Member States can refer questions for a preliminary ruling. Whereas under the Directives for Racial Equality124 and Gender Goods and Services125 it will be difficult to establish a prima facie infringement, the Services in the
121 Supra n. 99 p.18 122 Supra n. 99 p.27 123 Supra n. 99 p.29
124 Council Directive 2000/43/EC of 29 June 2000 implementing the principle of equal treatment between
persons irrespective of racial or ethnic origin
125 Council Directive 2004/113/EC of 13 December 2004 implementing the principle of equal treatment between
