Amazon’s E-Books antitrust saga - War now Peace?

Amazon has offered commitments to the European Commission to end the antitrust investigation into its use of ‘most favoured nation’ (MFN or parity) clauses in its e-books contracts with publishers, launched in 2015. The Commission is now inviting comments on these proposed commitments from customers and rivals. 

The Commission’s concern is that the clauses may breach EU antitrust rules and result in reduced competition among e-book distributors and less consumer choice.

Amazon’s MFN clauses require publishers to inform Amazon about more favourable terms or conditions offered to Amazon's competitors and to offer Amazon similar terms and conditions. This includes requiring publishers to offer Amazon any new or different distribution methods or release dates, any better wholesale prices or agency commissions, or to make available a particular catalogue of e-books.

The Commission considers that the cumulative effect of these clauses is to make it harder for other e-book retailers to compete with Amazon by developing new and innovative products and services. It also takes the view that imposing these clauses on publishers may amount to an abuse of a dominant market position.

In parallel, Audible, Amazon’s audio-books subsidiary, has announced the end of its exclusivity provisions in its distribution agreement with Apple following a joint antitrust investigation by the Commission and the German competition authority, the Bundeskartellamt. 

Amazon’s proposed commitments

Amazon disputes the competition law basis for the Commission’s investigation.  Nevertheless, in order to bring the investigation to a close (and to avoid the risk of a costly infringement decision), it has offered commitments:
  • Not to enforce:
    1. any clause requiring publishers to offer Amazon similar terms and conditions as those offered to Amazon's competitors; or 
    2. any clause requiring publishers to inform Amazon about such terms and conditions. 
  • To allow publishers to terminate e-book contracts that contain a clause linking discount possibilities for e-books to the retail price of a given e-book on a competing platform. Publishers would be allowed to terminate the contracts upon 120 days' advance written notice.
  • Finally, not to include, in any new e-book agreement with publishers, any of these clauses.
The commitments would apply for five years and (as is usual for behavioural commitments) be subject to oversight by a monitoring trustee.

E-Books - déjà vu? 

This is not the first time the Commission has investigated the e-books sector. In 2011 it opened antitrust proceedings against Apple and five international publishing houses (Penguin Random House, Hachette Livres, Simon & Schuster, HarperCollins and Georg von Holtzbrinck Verlagsgruppe) on the basis that it considered that they had colluded to limit retail price competition for e-books. In that case the companies also offered commitments to address the Commission's concerns (see our previous comment).

Where does this leave MFNs?

The Commission and national competition authorities have conducted investigations into MFN clauses in a number of other sectors, including online motor insurance and online sports goods retail, on which we have previously commented.  

While MFNs are not per se unlawful, and in some circumstances may even be pro-competitive, companies should carefully consider their possible anti-competitive effects before including them in new contracts. 

FTC settles abusive acquisition of pharma licensing rights

On 18 January, the FTC announced that Mallinckrodt ARD Inc. (formerly Questcor Pharmaceuticals, Inc.) and its parent company have agreed to pay $100 million to settle FTC charges that they violated antitrust laws when Questcor acquired the rights to a drug that threatened its monopoly in the U.S. market for adrenocorticotropic hormone (ACTH) drugs.  The announcement was made concurrently with the release of the FTC's complaint.

Antitrust (as opposed to merger) cases about acquisitions of competing technology are not an everyday occurrence.  However, this complaint has something of the flavour of the EU Commission’s Tetra Pak 1 decision.  In that case, the EU Commission objected to Tetra Pak’s acquisition (through a merger) of exclusive rights to what was at the time the only viable competing technology to Tetra Pak’s dominant aseptic packaging system.  The Commission (and subsequently the EU courts) held that this would prevent competitors from entering the market and therefore amounted to an abuse of a dominant position.  

The FTC’s Mallinckrodt complaint alleges that while benefitting from an existing monopoly over the only U.S. ACTH drug, Acthar, Questcor illegally acquired the U.S. rights to develop a competing drug, Synacthen (a synthetic ACTH drug which is pharmacologically very similar to Acthar).  This acquisition stifled competition by preventing any other company from using the Synacthen assets to develop a synthetic ACTH drug, preserving Questcor’s monopoly and allowing it to maintain extremely high prices for Acthar.  

To judge by the FTC’s complaint, the case appears to contain some pretty stark facts which may have contributed to the immediate settlement of the proceedings by Mallinckrodt.  Those facts also bring the case squarely into line with the US and EU competition regulators’ current concern over excessive pricing in pharma.

First up is the finding that Questcor had a 100% share of the U.S. ACTH market and that it took advantage of that monopoly to repeatedly raise the prices of Acthar from $40 a vial in 2001 to more than $34,000 per vial today – an 85,000% increase.  The complaint details that in August 2007 Questcor increased the price of Acthar more than 1,300% overnight from $1,650 to $23,269 per vial and that it has taken significant and profitable increases on eight occasions since 2011 pushing the price up another 46% to its current $34,034 per vial.  Acthar is a speciality drug used to treat infantile spasms, a rare seizure disorder affecting infants, as well as being a drug of last resort (owing to its cost) for a variety of other serious medical conditions.  According to the FTC, Acthar treatment for an infant with infantile spasms can cost more than $100,000.  In Europe, Canada and other parts of the world doctors treat these conditions with Synacthen which is available at a fraction of the price of Acthar in the U.S. (Synacthen is not available in the U.S. as it does not have FDA approval.)  The FTC relies on the supra-competitive prices charged in the U.S. for Acthar as evidence of Questcor’s monopoly power as well as its 100% market share and the existence of substantial barriers to entry.

It is also part of the FTC’s case that Questcor disrupted the bidding process for Synacthen when the rights came up for acquisition.  According to the complaint, Questcor first sought to acquire Synacthen in 2009, and continued to monitor the competitive threat posed by Synacthen thereafter.  When the U.S. rights to Synacthen were eventually marketed in 2011, dozens of companies expressed an interest in acquiring them with three firms proceeding through several rounds of detailed negotiations.  All three firms planned to commercialise Synacthen and to use it to compete directly with Acthar including by pricing Synacthen well below Acthar.  In October 2012, Questcor submitted an offer for Synacthen and subsequently acquired the rights to Synacthen for the U.S. and thirty-five other countries and did not subsequently bring the product to market in the US.

In addition to the $100 million payout, the proposed court order requires that Questcor grant a licence to develop Synacthen to treat infantile spasms and nephrotic syndrome to a licensee approved by the FTC, a pretty far-reaching remedy.  

This case is the latest in a string of cases on both sides of the Atlantic relating to escalating pharma prices (as discussed in our previous blog posts here and here).  While companies retain significant scope to price products as they see fit, it reaffirms that pharma companies should be wary of implementing very significant price increases in the absence of good objective reasons for doing so.  This is particularly so where the increase is facilitated by commercial strategies such as acquiring IP rights to existing/potentially competitive products.  In the EU, it is also worth remembering that – as established by Tetra Pak I (on appeal to the General Court) – an agreement which falls within a block exemption can at the same time constitute an infringement of Article 102.  So companies and their advisors should remember to wear Article 101 and 102 hats when reviewing agreements.

Helen Hopson

Patent licensing and antitrust – Qualcomm in the firing line

Ten years after the European Commission opened an investigation into Qualcomm's royalties for 3G essential patents, and 8 years after it closed that investigation without any finding of infringement, the US FTC last week week (January 2017) brought a new complaint against the chipset manufacturer's patent licensing practices.

In the meantime, Qualcomm has not been long out of the competition authorities' sights, with statements of objections in the EU on exclusivity payments and predatory pricing, an infringement decision and $853m fine by the Korean FTC (the second imposed on the company by that agency) just after Christmas last year, and a settlement worth over $900m with the Chinese antitrust authorities in 2015.  It has also attracted private actions, as we reported here (Icera’s 2016 claim in the English courts) and as has been further reported in recent days (Apple’s new lawsuit in California, alleging overcharging for chips and failure to pay rebates due).

The FTC complaint focusses on three main areas*:

  • Requiring customers for its baseband processors (in which Qualcomm is said to have a dominant position for both CDMA (3G) and LTE (4G)) to take a patent licence, thus achieving elevated royalties;
  • Refusing to license its essential patents to competitors (such as Intel), contrary to Qualcomm’s FRAND commitments;
  • Exclusive dealing, notably with Apple, to cement its position in the 4G smartphone market

In combination, this conduct is said to impose a “tax” on mobile phone manufacturers, even when using non-Qualcomm processors.   The complaint also notes that licensees are passing up the opportunity to challenge whether Qualcomm’s rates are FRAND.  The FTC lists a number of reasons why Qualcomm’s licences might not be FRAND, such as the maintenance of Qualcomm’s royalty rates at a significant level despite its reduced patent share; the charging of royalties on the selling price of the whole handset, even though this includes many features not subject to Qualcomm’s patent claims; and the extraction of onerous cross-licensing terms.

But why is this occurring? – why is it that licensees do not litigate Qualcomm’s royalty demands, as they do with other significant SEP holders?  The answer lies – according to the FTC – in the cost to licensees of litigation with Qualcomm.  It argues that rational licensees will be willing to litigate if the cost of doing so appears likely to be outweighed by the prospect of reduced royalties.  In this instance, however, it is said that the cost of litigation for prospective Qualcomm licensees is usually too great: as Qualcomm’s licensees are also customers of Qualcomm’s chipsets (for which there are few – if any – substitutes), they not only have to bear the cost of the litigation itself, but also, potentially of significantly impeded market access, arising from difficulties in obtaining Qualcomm chips. 

Patent licensing, once regarded as a largely benign or even pro-competitive business model, is now at the forefront of the antitrust authorities' attention around the world.  The fact that Qualcomm’s conduct is perceived to have effects on a market as valuable and crucial to the economy as the smartphone market have served to make it a particularly attractive target for the antitrust authorities.  This is not without controversy, however: US DOJ officials have previously expressed concerns about investigations seeking to  curb royalties would be liable to affect innovation, and – as Qualcomm has emphasised – the present FTC Complaint was launched on the basis of only 2-1 agreement by the FTC commissioners, in the face of opposition from Maureen Ohlhausen.  Whatever the divergence of views, it appears inevitable that Qualcomm will continue to appear in the antitrust headlines for some time to come.

* One further area has been identified by the FTC, but is subject to protective order.  

Compulsory Licencing: the Brave New World for (non-personal) data in Europe?

The Commission has published its Data Economy Package for non-personal data*, which is the final building block of its Digital Single Market (DSM) strategy – see our previous posts on the DSM here, here; and here.

With its new package, the Commission aims to: 

  • review the rules and regulations impeding the free flow of non-personal data and present options to remove unjustified or disproportionate data location restrictions; and
  • outline legal issues regarding access to and transfer of data, data portability and liability of non-personal, machine-generated digital data.
The package includes a Consultation on Building the European Data Economy, a Communication and Staff Working Paper.

Why is the Commission acting on data?

The economic rationale is that the EU data economy was worth €272 billion in 2015, and is experiencing close to 6% growth a year.  It is estimated that it could be worth up to €643 billion by 2020, if appropriate policy and legal measures are taken. Data also forms the basis for many new technologies, such as the Internet of Things and robotics.  The Commission’s ambition is for the EU to have a single market for non-personal data, which the EU is a long way from achieving.  The Commission refers to the issues in terms of – the “free movement of data”, suggesting something akin to a fifth EU fundamental freedom. 

What action is the Commission proposing to take? 

The Consultation sets out options for addressing the legal barriers to the free flow of non-personal data, in particular in relation to:

  • data access and transfer;
  • unjustified localisation of data centres;
  • liability related to data-based products and services; and
  • data portability.
Some of the more eye-catching (and interventionist) options set out by the Commission are the introduction of:

  • legislation to define a set of non-mandatory contract rules for B2B contracts when allocating rights to access, use and re-use data;
  • creation of a sui generis data producer right for non-personal machine-generated data, with the aim of enhancing tradability; an obligation to license data generated by machines, tools or devices on fair, reasonable and non-discriminatory (FRAND) terms; and
  • technical standards to facilitate the exchange of data between different platforms.
The Consultation is also seeking evidence on whether anti-competitive practices are restricting access to data.  In particular, the Consultation refers to: the use of unfair business practices; the exploitation of bargaining power when negotiating licences; and abuses of a dominant position.  Interestingly, it also asks whether current competition law and its enforcement mechanisms sufficiently address the potentially anti-competitive behaviour of companies holding or using data.

So where are we headed?

To date, competition law has mandated the compulsory licensing of IP rights only in exceptional circumstances, where the owner has a dominant position and there are no alternatives to the technology.  The Commission is now considering a range of regulatory options, of which the most interventionist could require access to be granted to non-personal data in a far wider range of contexts (albeit without any proposal to amend the existing database right and the new Trade Secrets Directive). These issues are likely to be of considerable concern for any company holding large amounts of non-personal data.  The Consultation runs until 26 April 2017. 


* Non-personal data includes personal data, where it has been anonymised

European Commission publishes two new studies on the interplay between patents and standards

In December 2016, the European Commission published two new studies on standard essential patents (SEPs).  As regular readers of this blog will know, SEPs protect technologies that are essential to standards such as 4G (LTE) and Wi-Fi, which rely on hundreds of patented technologies to function effectively.  For the same reason, SEPs will be crucial to 5G and the nascent “Internet of Things”.

The two studies form part of the Commission’s project to improve the existing IPR framework and to ensure easy and fair access to SEPs.  The specific aims of the Commission’s project were outlined in its April 2016 Communication “ICT Standardisation: Priorities for the Digital Single Market”, which we commented on here.

The first new study, titled “Transparency, Predictability and Efficiency of SSO-based Standardization and SEP Licensing”, and prepared by economics consultancy Charles River Associates (CRA), examines a number of issues relating to the standardisation process and SEP licensing.  Building on a previous 2014 report on patents and standards, and on the responses to a 2015 public consultation, the authors outline what they see as the main “problems which have real significance and impact ‘on the ground’”.  They then go on to consider a number of specific policy options which might help alleviate those problems.  Particular focus is placed on “practical and readily implementable solutions” which would, according to the authors, enhance the transparency of the standardisation process and reduce the transaction costs of SEP licensing.

One of the CRA study’s most notable – and doubtless controversial – proposals is the imposition of a ceiling on the aggregate royalty for a given standard.  The authors suggest that a commitment by SEP holders to observe a maximum total royalty burden would go a long way to tackling the problems of patent hold-up and royalty-stacking*.  While the study recognises that there would be a number of difficulties in implementing such an approach, it arguably underestimates the challenges.  The first problem would be determination of the aggregate royalty level.  Assuming that can be overcome, allocation of total royalties between SEP holders would be a formidable challenge, even for a ‘static’ standard.  The landscape here is far from static, however.  Not only do SEPs change hands regularly (as the second report by IPlytics emphasises), but telecoms standards themselves evolve, through the addition of new releases which improve on or supplement existing technologies.  When you throw into the mix the lack of public information about licence fees charged across the industry (something which the authors also have in their sights**), and the multiplicity of methods for comparing the relative values of SEP portfolios, it is difficult to see how such a system would work in practice – except, perhaps, as very general guidance.

The CRA study goes on to emphasise the importance of preserving flexibility on issues such as the appropriate royalty base and the level of the value chain at which SEP licensing should occur.  In the authors’ opinion, economic analysis of these issues suggests there is no appropriate one-size-fits-all solution.  This stands in contrast to the conclusion on royalty-stacking, where greater control is advocated.

The second new report, prepared by the Berlin-based data analytics company IPlytics, uses a dataset of over 200,000 SEPs to paint a more quantitative portrait of the SEP landscape.  It provides detailed empirical evidence on a number of issues, including:

  • Technology trends – The report shows that most declared SEPs relate to communication technologies, followed by audio-visual and computer technologies.  More than 70% of all SEPs are declared as essential to ETSI.
  • Regional trends – The proportion of SEPs filed at the Korean and Chinese patent offices has increased in recent years (particularly in the telecommunications sector), reflecting the growing importance of Asian markets in the global economy.
  • SEP transfers – More than 12% of all SEPs have been transferred at least once.  The study reveals that the top sellers of SEPs are Motorola, Nokia, Ericsson, InterDigital and Panasonic.  The most active buyers include Qualcomm, Intel and – perhaps surprisingly, given its recent suit alleging that Nokia evaded FRAND by transferring patents to two PAEs – Apple.  
  • Comparison with non-SEPs – A comparison with a control group of patents which have not been declared as standard essential suggests that SEPs are more frequently transferred, litigated, renewed and cited as prior art than non-SEPs.  This implies that SEPs are generally more valuable than non-SEPs, but the study refrains from considering whether the technology protected by SEPs is intrinsically more valuable than that protected by non-SEPs, or whether the higher value of SEPs is merely a product of their incorporation into a standard.

One striking feature of both studies is their attempt to grapple with the thorny issue of ‘over-declaration’.  The authors of the CRA study point to research showing that, when tested rigorously, only between 10 and 50 per cent of declared SEPs turn out to be actually essential.  Both studies propose some form of independent essentiality testing to address the problem.  The CRA study claims that random testing of a sample of each SEP holder’s portfolio would provide useful information about how royalty payments should be allocated between SEP holders; and that the benefits of such testing would be especially pronounced when combined with the imposition of an appropriate ceiling on the total royalty stack.  According to the IPlytics report, patent offices have the requisite technical competence and industry recognition to perform essentiality testing at a reasonable cost.

To conclude, the two studies provide a reminder – if any were needed – that issues relating to the standardisation process and SEP licensing remain high on the Commission’s agenda.  The Commission says it intends to draw fully on the studies’ findings when assessing the interplay between patents and standards in the EU Single Market.  However, whether the Commission will embrace any of the practical solutions proposed by the studies remains to be seen.

* As mentioned in this December 2015 blog post, royalty-stacking refers to the situation where the royalties independently demanded by multiple SEP holders do not account for the presence of other SEPs, potentially resulting in excessively high total royalty burdens for implementers.

** See pages 71 and 85 of the CRA study.