The Pros and Cons of Price Discrimination

On November 18, The Swedish Competition Authority held a seminar on "The Pros and Cons of Price Discrimination". This was the fourth year we arranged a "Pros and Cons" seminar. The aim of the seminar is to allow researachers and practioners to meet and exchange ideas.

Introduction by Mats Bergman

Price discrimination is a pervasive phenomenon in many markets, such as consumer products markets, travel and transport, telecommunication, and many other services markets. It takes many forms and it is a phenomenon which can have both positive and negative effects, for consumers as well as for welfare. The analysis of price discrimination has deep roots in the economics discipline, where it has long been recognized that it can be for good and for bad – and sometimes even necessary.

Price discrimination has also featured prominently in many of the recent high-profile competition law cases. In the legal rhetoric, one sometimes gets the impression that price discrimination is all evil. At the same time, since most instances of price discrimination – even by dominant firms – goes unchallenged, it is clear that the legal practice is not as simple-minded as that. Consequently, discussing the pros and cons of price discrimination appears natural in a conference volume that brings together the views of academic economists and competition law experts.

According to Article 82 (c), if a dominant firm applies “dissimilar conditions to equivalent transactions with other trading parties, thereby placing them at a competitive disadvantage”, the firm abuses its dominant position. A possible interpretation is that all kinds of price discrimination are prohibited, if they are undertaken by a dominant firm. Note, however, the emphasis on the competitive position of the buyers that suffer from price discrimination. With this in mind, another interpretation is that price discrimination vis-à-vis final consumers is not illegal, while discrimination of downstream firms can constitute an abuse only if the firm that is discriminated against suffers a competitive disadvantage. In other words, it may be that for price discrimination to be abusive, the customers must be firms (and not consumers), the good that is sold on discriminatory terms must be an important (costly) input for the downstream industry – and the seller must be dominant.

In order to introduce the reader to the legal practice concerning price discrimination, the first paper in this volume is the only one not written by economists. In the paper, Damien Geradin and Nicolas Petit, who are both legal scholars and experts of competition law, analyse the scope of Article 82 (c). They distinguish between three main types of price discrimination that can be found in the EC competition law practice. First, primary-line price discrimination describes a situation in which a dominant firm price discriminates in order to exclude rivals. Typical examples are fidelity rebates and selective price cuts, both of which are employed in order to exclude a rival from the dominant firm’s market.

Next, secondary-line price discrimination occurs when one or more downstream firms are offered better terms than one or more other downstream firms, putting the latter in a position of competitive disadvantage. In one subtype of cases, the victims of secondary-line price discrimination have been foreign firms. For example, the owner of a port or an airport could offer better terms for domestic shippers or airlines, respectively. In another subtype of cases, involving essential facilities, the dominant firm favours its own downstream operations by providing access to the bottleneck facility on discriminatory conditions only.

Third, geographic price discrimination is used when a firm wishes to sell a product at different prices in different member states. Typically, this requires some measures in order to prevent trade between the member states from eliminating the price differentials.

Geradin and Petit argue that, for the most part, the legal practice on price discrimination has been misguided since it has failed to distinguish between these different types of price discrimination – and their different rationales. They argue that primary-line price discrimination should be challenged under Article 82 (b), i.e., that it should be abusive only to the extent that it is exclusionary, not because of the price discrimination as such. The same argument applies, according to the two authors, for secondary-line price discrimination employed by a vertically integrated firm. Furthermore, they argue that geographical price discrimination should not in itself be seen as an abuse. Instead, the competition authorities should focus on the measures that are used to prevent the arbitrage trade that geographical price discrimination stimulates. This leaves only secondary-line price discrimination undertaken by a non-integrated firm. In practice, such cases are few, since it would rarely or never be in the interest of the dominant firm. Mostly, this category of cases concerns undertakings that discriminate against foreign firms.

The following five papers are placed in alphabetical order. In the second contribution, Simon Bishop focuses on one specific form of price discrimination: loyalty rebates. Bishop begins by remind the reader of some of the pro-competitive reasons for using loyalty rebates: to give the retailer (or, more generally, the downstream firm) stronger incentives to provide complementary services, to reduce the problem of double marginalization and to allow efficient recovery of fixed costs. In favourable circumstances, this can be achieved by offering the retailers rebates on marginal quantities. Consequently, Bishop argues, a per-se prohibition of loyalty rebates (even if offered by a dominant firm) does not make sense. It follows that rebates should be analysed on a case-by-case basis.

The main contribution of the paper is a discussion of a step-by-step economic methodology for assessing the anti-competitive effects of loyalty rebates. First, one needs to establish that the firm that offers the loyalty rebates has market power (a dominant position). Second, one must examine whether rivals have alternative routes to the market; i.e., whether it would make a difference or not that some downstream firms are locked up with the dominant provider. Third, one must evaluate if the rivals can match the rebates or nor. As a final step – if it has been found that a firm with market power locks up essential customers in a way that an efficient rival cannot match – the adverse consequences of this must be weighed against the possible efficiency-improving effects of the rebates.

Bishop is most specific in his discussion of the third step. He suggests that the competition authorities compare the incremental revenues (including the effects of the loyalty rebates) and the incremental costs for those quantities that are in practice open to competition. Under this analytical scheme, a loyalty rebate would not be prohibited because of its form, but because the effective rebates over the competitive range are so large that the dominant firm is in effect pricing below costs for these quantities.

In the third paper, Yongmin Chen focuses on price discrimination in a symmetric duopoly situation. In the first of the two models he presents, the products of the two firms are initially identical, but once a consumer has bought from one of the firms, the consumer will experience switching costs if he or she buys from the other firm in the next period. This creates a lock-in situation, but it also creates a temptation for the two firms to try to “poach” each others customers. If they can price discriminate by the customers’ purchase history, both firms will offer lower prices to the other firm’s customer. This intensifies competition, lowers prices and lowers profits – but the practice may or may not benefit consumers.

In the second model that Chen presents, the products of the two firms are assumed to be differentiated. This means that some consumers will prefer to buy from one firm, while other consumers will prefer the other firm. However, when the firms compete by offering better prices to their rivals’ customers than to their own, some consumers will be lured to buy from the “wrong” firm, i.e., from the firm that produces products that they like the least. The outcome of this type of competition will still benefit consumers, since the lower prices more than compensate for the worse consumer-producer fit. However, total welfare (and profit) will be lower than if price discrimination were not possible.

Which type of model is the best representation of real-world competition depends on the characteristics of the market that we want to analyse. However, a general conclusion is that price discrimination based on purchase history intensifies competition. This is in contrast with the view taken by competition authorities and courts in a number of cases under the competition laws. However, these cases typically concern asymmetric markets, i.e., markets where a dominant firm competes with one or more smaller firms, while the theoretical analyses focus on symmetric competition. Chen acknowledges that there is little economic theory that deals with price discrimination based on purchase history in asymmetric settings.

Thomas P. Gehrig and Rune Stenbacka, in the fourth contribution to the volume, take a step back and ask: What are the arguments in favour of – and against – price discrimination? They identify a number of arguments in both directions. First, price discrimination increases the flexibility of pricing, often with the effect of increasing efficiency. For example, higher prices during peak-demand periods allow for a more efficient use of capacity. Second, price discrimination improves fairness between consumers: consumers that would gain much from buying a product at an average price will have to pay a high price, while consumers that would gain little at the average price can now buy at a low price. Third, and finally on the positive list, in markets that are reasonably competitive, the use of price discrimination makes competition more intense.

However, there are also possible reasons to oppose price discrimination. First, it appears that many people prefer simple rules, for example one price only, over more complex rules (here, multiple prices). Second, one can argue that equal prices imply greater equality than unequal prices, hence disregarding that the latter may imply a fairer distribution of gains from trade. Third, if competition is not effective, price discrimination may not be good for consumers.

The conclusion of the authors is that a ban on price discrimination cannot be justified with reference to fairness considerations since, in their view, price discrimination will typically tend to increase fairness. Furthermore, in relatively competitive markets, price discrimination will typically tend to make the competition even keener. However, in markets without effective competition, price discrimination will tend to hurt consumers. This conclusion is consistent with the fact that the prohibition only applies for dominant firms. (On the other hand, in situations where two or more firms are dominant in their own home markets, while they have a small presence in each others’ markets, the prohibition of price discrimination will tend to reinforce the divided market structure.)

In the fifth contribution to this volume, Anne Perrot argues that competition authorities’ policies towards price discrimination should by governed by the effect of a particular type of price discrimination, not by its form. To set the stage for such an analysis, she provides an overview of what economic theory has to say about price discrimination.

Traditionally, economists have distinguished between three types of price discrimination. Under first-degree price discrimination, the seller is able to extract all consumer surplus by setting each consumer’s individual price exactly equal to the maximum price that consumer would be willing to pay. (Of course, this is a theoretical concept, but sometimes it is very useful in theoretical analyses.) Under second-degree price discrimination, the seller offers different prices according to the terms of the sale – e.g., quantity discounts and high (low) peak (off-peak) prices – so that the buyers can self-select according to their willingness-to-pay. Under third-degree price discrimination, buyers are asked for different prices depending on their characteristics. An example would be lower prices for children, students or the elderly.

However, when assessing the competitive effects of a price-discrimination scheme, Perrot argues that it is useful also to make a distinction between the strategic and the non-strategic effects of price discrimination. Simply put, non-strategic price discrimination refers to practices employed by monopoly sellers that aim to increase their revenues. At first glance, one would think that higher revenues, while good for the seller, are always bad for the collective of consumers. However, price discrimination will tend to benefit low-valuation consumers. If the total volume increases because of price discrimination, the practice may actually benefit consumers collectively. In some situations, e.g., when fixed costs are high, price discrimination may actually be necessary for any production at all to take place.

Strategic price discrimination, in contrast, occurs in oligopoly situations and in vertical relations. Sometimes, strategic price discrimination is efficiency increasing, for example when it is used to create economically correct incentives for down-stream customers. In an oligopoly setting, the possibility to price discriminate will sometimes tend to intensify competition (along the lines argued by Chen and Gehrig and Stenbacka in this volume). In other instances of strategic competition, however, price discrimination can be used as a component of a predatory or exclusionary strategy. Perrot argues that, in order to make a proper effects-based analysis of price discrimination, the competition authorities must lay bare the mechanisms through which the practice tends to reduce competition.

In the final contribution, David Spector focuses on the strategic use of price discrimination. He argues that price discrimination can be an essential element of a predatory strategy, because it makes predation less costly for the dominant. It is often argued that predation is a costly strategy for a dominant firm because, being dominant, it will experience the greatest profit loss due to lower prices of all firms. However, price discrimination allows the predatory activities to be targeted directly at the rival, making the dominant’s immediate profit loss – and the consumers’ immediate gain – smaller. Because of this, price discrimination can be an aggravating circumstance in exclusionary-abuse cases.

A similar mechanism is at work in bundling and exclusive-dealing practices: price discrimination makes these practices less costly – and therefore more effective – for the dominant. It follows that analysing the effect of price discrimination can often be a key step when analysing exclusionary behaviour.

An isolated ban on price discrimination, however, can be counter-productive, according to Spector. This is so, since price discrimination has many positive effects. As argued in some of the other contributions to this volume, benefits can arise from both non-strategic price discrimination and from strategic price discrimination; in the latter case because price discrimination sometimes intensifies competition. Spector identifies one possible exception to the principle of not seeing price discrimination in isolation as abusive. In the context of a vertically integrated firm, price discrimination between the dominant’s downstream activities and the downstream rivals can make sense, from the point of view of the dominant. However, this brings us into the realm of the essential-facilities doctrine.

An interesting point raised by Spector is that banning what Geradin and Petit (and others) call secondary-line price discrimination can hurt the competitive process and that a ban can actually be in the interest of the dominant. The reason is that a manufacturer selling to competing retailers will be tempted to increase its profit by selling to many retailers, in effect undercutting itself. This process will drive down both wholesale and retail prices, even if the manufacturer holds a monopoly. In order to escape this paradoxical result, the manufacturer needs to commit to a high price; one way to achieve this is to commit not to price discriminate between the different retailers. In situations like this, a legal prohibition of price discrimination will be an ideal commitment device.

The overall conclusion of this volume appears to be that price discrimination is a complex phenomenon and that competition authorities would be ill-advised to see all instances of price discriminations as violations of Article 82, in particular since price discrimination can increase efficiency and intensify competition. Indeed, the authors appear to agree that price discrimination in itself, i.e., not associated with other abusive practices, should with very few exceptions not be seen as a violation of the prohibition of abuse of dominance. On the other hand, price discrimination can often be an essential element of a successful – from the point-of-view of the perpetrator – exclusionary strategy.

This suggests that a thorough analysis of the effects of the price discrimination is required. Hopefully, this volume contributes towards a better understanding of the mechanisms through which price discrimination has an impact on markets – and towards a more effective enforcement of the competition rules.

For more information, please contact:
Mats Bergman, Chief Economist, tel +46-8-700 15 57
Niklas Strand, Head of the organizing committee, tel +46-8-700 16 64
Jimmy Dominius, Press Officer, tel +46 8 700 15 80 or +46 73 773 15 80