3.2 Background: Earlier e-Books Case
The Commission had already intervened in the e-book sector in 2012 and 2013 in an earlier antirust case. As will become clear, the earlier e-books case also affected the present case.
In December 2011, the Commission opened proceedings against Apple and five major international publishers: Hachette Livre, Harper Collins, Simon & Schuster, Penguin, and Verlagsgruppe Georg von Holtzbrinck (collectively the “Five Publishers”). The Commission took the preliminary view that the Five Publishers and Apple had engaged in contracts that were aimed at inter alia raising the retail prices of e-books by jointly switching the sale of e-books from a wholesale model (where the e-book retailer determines retail prices) to an agency model (where the publisher determines retail prices and the e-book retailer acts merely as its agent) on a global basis, including on Amazon.
In order to eliminate the Commission’s preliminary concerns, the Five Publishers and Apple offered commitments, which were made binding by Commission Decisions in 2012 and 2013.
54 The commitments included the termination of all relevant agreements and a five-year ban on retail price, wholesale price, and commission parity clauses, as well as a two-year “cooling-off” period whereby the Five Publishers had to allow all e-book retailers that were on agency terms to discount the retail price of e-books up to a level of their respective aggregated annual commissions.
Following the switch by the Five Publishers to the agency model in 2010, an increasing number of publishers followed suit and entered into agency agreements with Amazon.
3.3 Description of the Relevant Clauses in the Amazon MFN Case
For brevity, this article focuses only on a subset of the clauses that were covered by the investigation and the commitments. The following list provides a brief and stylized description of the parity and notification provisions that are relevant for this article.
3.3.1 Retail Price Parity Clauses
The Agency Price Parity Clause contractually obliges the e-book supplier (usually a publisher) to set for a given e-book an agency price on Amazon that is not higher than the retail price (including promotions and discounts) at any competing e-book retailer.
55
The Discount Pool Provision is a term in agency agreements that relates to a “pool” of credits that Amazon may use at its discretion to discount agency prices on Amazon. This pool is filled if the agency price of a given e-book on Amazon exceeds the price of this e-book at competing e-book retailers.
3.3.2 Non-price Parity Clauses
The Business Model Parity Clauses contractually oblige the e-book supplier to notify and offer to Amazon the terms for the distribution of e-books under a given business model if that e-book supplier distributes e-books under that business model (for example, reseller, subscription, rental, bundling with physical books or book clubs, by download, partial downloads, streaming or any other form of digital distribution) through any e-book retailer other than Amazon.
The Selection Parity Clauses encompass a number of provisions that impose equal treatment for Amazon with regard to: (1) the catalogue available; (2) the specific format, including its features and functionalities, in which the titles are offered; and (3) the release date of the titles.
3.3.3 Retail Price Notification Provisions
These are provisions that contractually oblige the e-book supplier to notify Amazon if the agency price (including promotions) that is set by the e-book supplier on Amazon is higher than the retail price at a competing e-book retailer.
3.4 Preliminary Assessment
The Commission came to the preliminary view that Amazon had abused its dominant position by among others including the Agency Price Parity Clause, the Discount Pool Provision, the Retail Price Notification Provisions (together with requesting price parity), the Selection Parity Clauses and the Business Model Parity Clauses into its agency contracts for the following reasons.
56
3.4.1 Retail Price Parity Clauses
The Commission considered that the Agency Price Parity Clause and the Discount Pool Provision have very similar effects because each of these clauses allows Amazon to ensure that the effective retail price on Amazon does not exceed the retail price of competing e-book retailers.
57 Therefore, the effect of these Retail Price Parity Clauses is assessed together.
In line with Boik and Corts (
2016), the Commission found that the Retail Price Parity Clauses are likely to: (1) deter the expansion or entry of competing e-book retailers, which would strengthen Amazon’s dominant position; and (2) soften competition at the e-books retail distribution level. These two effects could inter alia lead to higher retail e-book prices to the detriment of consumers.
With regard to the first effect, a potential entrant or a competing e-book retailer could in the absence of the Retail Price Parity Clauses attempt to increase its market share by charging a lower commission to e-book suppliers, so as to induce them to set lower retail prices and, hence, attract buyers.
However, if Amazon had agreed on a Retail Price Parity Clause with a given e-book supplier, that e-book supplier could not set lower retail prices (compared to those on Amazon) at any competing e-book retailer. This automatically limited the ability of a competing e-book retailer to attract buyers by offering lower retail prices than those on Amazon. This may have discouraged competing e-book retailers from entering in the first place.
By restricting the scope for offering e-books at competing e-book retailers that were cheaper than on Amazon, price discounts could not be used to compensate customers for switching costs and to induce them to switch away from Amazon. This stifles the entry and expansion of competing e-book retailers in the first place.
A number of market characteristics may have exacerbated this effect. Those include: Amazon’s particularly strong market position; indications that e-book suppliers would support the expansion of competing e-book retailers in the absence of the Retail Price Parity Clauses; and that price appeared to be the most important parameter of competition especially for e-books without sophisticated illustrations or special functionalities.
As regards the second effect, the Commission considered that the Retail Price Parity Clauses are likely to soften competition between e-book retailers that work on the basis of an agency relationship by reducing their incentive to compete on commission fees.
In the absence of the Retail Price Parity Clauses, e-book retailers on agency terms had an incentive to compete on commission fees. Since commissions reduced the profits of e-book suppliers, the latter generally had an interest to steer sales to e-book retailers that charge lower commission fees: for example, by setting lower retail prices on their platforms. For this reason, e-book retailers anticipated that charging lower commission may increase the volumes of e-books that are sold on their platform and thereby increase their overall revenues and profits.
In contrast, where a Retail Price Parity Clause was in place, e-book retailers’ incentives to compete on commission were limited. When charging a lower rate of commission, a competing e-book retailer could no longer expect that e-book suppliers would respond by lowering selectively the retail prices of its e-books on that e-book retailer’s platform. This is because when bound by the Retail Price Parity Clauses, a lower retail price that was set by an e-book supplier on one e-book retailer’s platform would have needed to be matched by an equally low price on Amazon. Hence, a competing e-book retailer could no longer expect to attract customers from Amazon by offering lower rates of commission to e-book suppliers.
Conversely, with a Retail Price Parity Clause in place, as a consequence of a higher commission, e-book suppliers had to raise the retail price of their e-books uniformly on all e-book retailers’ platforms, instead of raising the retail price exclusively on Amazon. Therefore, Amazon may have benefitted from a higher commission and at the same time did not have to fear a substantial loss of consumers, as the latter may also have faced higher retail prices at competing e-book retailers.
In a recent working paper, Johansen and Vergé (
2016) show however that the Retail Price Parity Clauses can have pro-competitive effects if there is competition between e-book sellers and they can sell significant volumes directly to customers. However in the relevant e-book markets, the direct sales channel was of limited importance, and it appears to be unlikely that e-book sellers could credibly threaten to switch a large proportion of their sales to the direct channel.
3.4.2 Retail Price Notification Provisions
Because of the commitments that were made binding by the Commission in 2012 and 2013 on the Five Publishers, those Five Publishers could not introduce or maintain any retail price parity clauses for a five-year period in their e-book distribution agreements in the EEA. With those Five Publishers, Amazon replaced the Agency Price Parity Clauses with the Retail Price Notification Provisions.
The Commission found that once notified by an e-book supplier on the basis of those provisions, Amazon typically requested from that e-book supplier that Amazon should be offered the same low retail price (including promotions) that was charged at the competing e-book retailer. The Commission considered that Amazon had a number of measures at its disposal that constituted a credible threat to induce e-book suppliers not to set lower prices on competing e-book retailers’ platforms.
58 Such measures could be implemented swiftly, without any need of approval by the e-book suppliers and without terminating any agreement, which made them credible threats to induce e-book suppliers to modify their prices on Amazon.
Together with Amazon’s policy of requesting retail price parity, those clauses allowed Amazon to ensure that the retail prices of e-books that were offered on its platform did not exceed the lowest retail prices of the same e-books that were sold via competing e-book retailers with similar effects as already set out for the Retail Price Parity Clauses.
The Commission therefore had the preliminary view that Amazon abused its potentially dominant position by having these provisions included in its e-books distribution agreements.
3.4.3 Selection Parity Clauses
The Commission focused on two effects of Selection Parity Clauses: First, the Commission preliminarily found that these clauses are likely to reduce the incentives of e-book suppliers and e-book retailers to develop e-books with innovative features or functionalities such as highly illustrated or enhanced e-books. Second, these clauses are likely to deter their entry and expansion, thereby weakening competition between e-book retailers, to the detriment of consumers.
As regards the first effect, in the absence of Selection Party Clauses e-book suppliers would have had incentives to develop enhanced or highly illustrated e-books, mainly for two reasons. There appeared to be demand for enhanced or highly illustrated e-books of those e-book retailers that use e-book readers or electronic devices on which the illustrations, features, and functionalities of such e-books displayed well. In addition, developing enhanced or highly illustrated e-books with e-book retailers competing with Amazon for their e-book readers or electronic devices could have allowed e-book suppliers to increase the strength of those e-book retailers relative to Amazon and to promote entry and/or expansion.
59 This could have reduced the e-book sellers’ dependence on Amazon and improve their bargaining position.
However, Amazon’s Selection Parity Clauses either oblige e-book sellers to produce a version of each enhanced or highly illustrated e-book with equivalent features that were compatible with Amazon’s e-book readers or notify Amazon that the e-book may not display well on Amazon’s e-book readers and provide Amazon with all assistance and materials that are reasonably required for Amazon to create an e-book of that title.
Depending on the additional resources that were required to develop equivalent features for Amazon’s e-book readers, this may have significantly increased the development costs in particular for enhanced or highly illustrated titles. The additional costs may have made the development process unprofitable even when taking into account that selling an e-book also on Amazon may give rise to additional revenues. There was evidence that indicated that even the obligation to assist Amazon in the creation of a version that was compatible with Amazon’s e-book readers could have been burdensome (and cause delays), in particular for certain enhanced or highly illustrated titles because it required cost and working time investments from e-book suppliers.
In line with these considerations, there were indications that e-book versions for a number of highly illustrated print books had not been produced in light of the Selection Parity Clauses.
Similarly, e-book suppliers were likely to be induced to keep functionalities of enhanced e-books simple and to avoid interactive and more advanced functions for all e-book retailers in order to simplify the adaption to Amazon’s e-book readers.
Also the ability and incentives of competing e-book retailers to develop innovative e-books and functionalities were likely to be compromised by the Selection Parity Clauses. Competing e-book retailers may have not been able to develop highly illustrated or enhanced e-books to the extent that e-book suppliers were hesitant to cooperate with them in light of the Selection Parity Clauses, even where the e-book reader and formats of those competing e-book retailers display features better than do Amazon’s e-book readers.
Competing e-book retailers may also have invested less in the development of highly illustrated or enhanced e-books as their willingness to invest in the development of such e-books depends on their expected benefits from such investments.
60
We now turn to the second potential effect of the Selection Parity Clauses: to weaken competition by deterring entry and expansion of e-book retailers. In the absence of Amazon’s Selection Parity Clauses, competing e-book retailers could have been able to offer e-books that were not (or only with less functionality) available on Amazon for the following reasons:
-
In light of technical difficulties to convert or to display certain highly illustrated or enhanced e-books on Amazon e-readers, such e-books would likely have been exclusively (or earlier) developed for e-book readers or electronic devices that were supported by competing e-book retailers.
-
In light of Amazon’s dominant position in the relevant e-book distribution markets, e-book suppliers may have had an interest in selectively supporting competing e-book retailers: for instance, by exclusive content; special editions (for example with specific features and functionalities); or early release dates.
The presence of Amazon’s Selection Parity Clauses may have limited competing e-book retailers’ scope to differentiate on the basis of content (including by offering titles or special editions not available on Amazon), or of particular features or functionalities of e-books or of earlier release dates.
As a result of Amazon’s Selection Parity Clauses, competing e-book retailers were likely to be hindered from developing and distributing e-books different to those available on Amazon or special editions (including editions with specific features and functionalities), or from choosing early release dates, as the Selection Parity Clauses hindered e-book suppliers from supporting this.
In particular, as set out above, the e-book retailers’ incentives to invest in enhanced or highly illustrated e-books or in non-price related promotional activities were likely to be reduced as a consequence of the Selection Parity Clauses. Lower investments by competing e-book Retailers (either in content or in non-price promotional activities) likely has deteriorated the attractiveness of competing e-book retailers from a customer perspective and hence may weaken competition at the e-book distribution level in the long run.
The risk of reduced competition at the e-book retail distribution level due to Amazon’s Selection Parity Clauses appeared to be more pronounced in markets where Amazon’s position in the retail distribution of e-books was particularly strong. Amazon’s high market share indicated that many e-book customers had a Kindle e-book reader or an Amazon account and that for them it is particularly convenient to buy e-books on Amazon. Buying e-books outside of Amazon’s closed Kindle ecosystem may impose a substantial burden (“switching costs”) on the users of Amazon’s Kindle e-reader. Therefore, to induce potential customers to switch away from Amazon, competing e-book retailers needed to provide additional value to consumers: for example, in the form of differentiated content or early releases of e-books.
Not being able to compete on the basis of a differentiated e-book catalogue or earlier release dates may have not only weakened competing e-book retailers; it could ultimately have forced competing e-book retailers to exit the market or have induced potential competing e-book retailers not to enter the market in the first place. Reduced competition at the e-book retail level may in itself result in higher prices.
3.4.4 Business Model Parity Clause
The Commission considered that Amazon’s Business Model Parity Clause is likely to hinder the emergence of alternative business models with regard to the distribution of e-books by: (1) reducing e-book suppliers’ incentives to support, and invest in, new and innovative business models for the e-book supplier’s own platform or for sales channels of e-book retailers; and (2) by reducing the ability and incentives of e-book retailers that competed with Amazon to develop alternative business models. Certain aspects of the assessment of Business Model Parity Clause and of the Selection Parity Clauses are hence conceptually similar.
Regarding the first effect, in relation to e-book suppliers’ incentives, in the absence of a Business Model Parity Clause they appeared to be willing to support alternative business models of competing e-book retailers mainly for two reasons:
-
E-book suppliers appeared to have an interest in experimenting with alternative business models—for example with smaller e-book retailers on a smaller scale (one country/region), or on a selection of their catalogue (children books, classics), or targeting only certain customers groups—without having to test such business models on a larger scale for the mass market.
-
Supporting alternative business models of competing e-book retailers would have allowed e-book suppliers to increase the strength of those e-book retailers relative to Amazon and to promote entry and/or expansion as set out above.
The presence of the Business Model Parity Clause may have reduced the incentives of e-book suppliers to support and invest in alternative business model in various ways including the following:
-
By effectively preventing e-book suppliers from launching alternative business models on their own or with a single or few e-book retailers, the Business Model Parity Clause denied e-book suppliers the opportunity of testing the effect of alternative business models on a small scale. This forced e-book suppliers into a position where they either introduced the alternative business models with Amazon on a larger scale or dropped such projects and did not develop them at all.
-
Alternative business models could not be used to strengthen competing e-book retailers relative to Amazon.
In addition, the Business Model Parity Clause may have also negatively affected the e-book retailers’ ability and incentives to develop alternative business models. Specifically, the Commission preliminarily considered that the Business Model Parity Clause diminished the e-book retailer’s incentives to invest in alternative business models in various ways, including the following:
-
E-book suppliers became reluctant to support alternative business models as discussed above. This, in turn, undermined the ability of competing e-book retailers to develop such alternative business models as competing e-book retailers need the e-book suppliers’ support, cooperation, and materials in order to implement alternative business models.
61
-
Moreover, even in instances where an alternative business model was launched, the e-book supplier’s obligation to grant the same business model to Amazon was likely to diminish the competing e-book retailer’s incentives to invest in alternative business models for the above-mentioned reasons.
Consequently, in its Preliminary Assessment, the Commission considered that the development of such alternative business models would be reduced or delayed, to the detriment of consumers. Consistent with these considerations, the evidence available to the Commission indicated that the Business Model Parity Clause had prevented the emergence and/or development of alternative models with competitors including: (1) print and e-book bundles; (2) pay-as-you-read and book club models (where readers do not necessarily have to acquire the e-book for an unlimited period of time, but are rather given a license to access only parts thereof); (3) subscription models; and (4) applications for smartphones that give access to e-book versions of classics.
As regards the second effect, Amazon’s Business Model Parity Clause was found likely to reduce the competitiveness of e-book retailers. This was likely to reduce the intensity of competition at the e-book distribution level and was in itself to the detriment of consumers since less competition may result in higher e-book prices and less choice.
A number of observations were important for that conclusion:
As already set out above, weaker competition likely in itself results in consumer harm.
3.6 Relation to Investigations in the Online Travel Agency Sector
Since 2010 a number of national competition authorities (“NCAs”) have investigated
inter alia retail price parity clauses that are used by online travel agencies (“OTA”). Whereas many NCAs expressed concerns about wide retail price MFNs, differing approaches have been implemented concerning narrow MFNs.
64
In this context, OTAs have argued that narrow parity clauses are indispensable to prevent hotels from free-riding on OTA investments. Their argument is that—absent such parity clauses—consumers will use OTAs to search for and compare hotels, but will then book more cheaply on the hotel’s website, thereby depriving the OTA of commission revenue.
In the OTA cases, the same theories of harm as in the Amazon case for the Retail Price Parity Clauses (see Sect.
3.4.1) were pursued although in those cases a potential infringement of Article 101 of the Treaty on the Functioning of the European Union was investigated. However, given that in many markets several large OTAs operated, the concern that MFNs would soften competition among OTAs appeared to be more pronounced than the concern that weaker OTAs could be excluded. Prima facie, it seems indeed plausible that potential exclusionary effects of retail price MFNs are more pronounced in environments where: (1) very strong firms are competing against relatively small/weak competitors; (2) the incumbent enjoys a quality advantage; or (3) the market position of the incumbents is protected by significant switching costs.
Second, the direct sales channel played a much more prominent role in the OTA investigations compared to the Amazon case. Indeed, in the relevant e-book markets the direct channel appeared to play a rather marginal role. This probably has to do with the benefits that are linked to using a given platform (for example, the Kindle devices work with e-books in the proprietary Amazon format). Consequently, also the claim that parity clauses are needed to restrict free-riding appears to be less convincing in the Amazon case.