Patent abuses under Article 82
This entry was added to http://www.reckon.co.uk/headlines-franckblog on 3 November 2005.
Updated 26 April 2006 upon publication of the Commission decision on AstraZeneca Losec.
Full blog table of contents available at Contents | viewpoint: Franck.
This article is mainly speculative. It considers the claims and counter-claims that have been reported in two current cases involving allegations of patent abuse that might be caught by the Article 82 prohibition on abuse of a dominant position:
Complaints against Qualcomm (3G phones)
This case emerged on Friday 28 October 2005. The best summary I have seen so far is in these two articles in the EE Times (and affiliated sites):
The complainants are mobile phone handset manufacturers. Qualcomm is a chip set manufacturer that also happens to own some patents over technology that is an essential part of 3G handsets.
The phone manufacturers are pursuing two allegations.
First, they claim that Qualcomm is unfairly trying to exclude other mobile phone chip set manufacturers from the market, by:
- refusing to licence essential patents on fair terms; and
- offering lower prices to handset makers who buy chip sets exclusively from Qualcomm.
Second, they claim that Qualcomm is charging excessive royalties for its essential patents.
Qualcomm's reported response to these allegations is that:
- The allegation of refusal to licence is disproved by the availability and wide take-up of licences for the relevant patents.
- Most of the complainants are amongst the licensees and are not therefore excluded from the market: all that they are doing is seeking to use Article 82 to reduce their payments to Qualcomm, and/or to bolster their positions as 2G incumbents by disrupting their competitors' access to 3G technology.
- The allegation about exclusionary rebates are "misleading" and Qualcomm's pricing merely reflects legitimate price competition.
- The allegation about excessive royalties is disproved by the downward trend of prices for 3G handsets.
There seems to be no dispute about Qualcomm's dominant position in the supply of licences to its essential patents: if they are essential then presumably there is no alternative supplier or alternative way of meeting the relevant needs of 3G handset manufacturers.
Refusal to licence
Under Article 82, an undue cessation of commercial relations, which might include a refusal to licence patents on reasonable terms, can be an abuse if it leads to the exclusion of all competition from a particular competitor (see for example Commercial Solvents). But there seems to be no allegation of undue cessation of commercial relationships by Qualcomm, or much likelihood of evidence that competitors are being entirely excluded from the downstream market because of Qualcomm's actions, at least if what Qualcomm says about widespread licensing is true.
In the absence of a cessation of an existing commercial relationship, a finding of abuse would probably have to draw on cases such as Magill and IMS Health, in which it was found that a dominant undertaking's "special responsibility" under Article 82 includes a duty to licence its technology or intellectual property to competitors only if:
- the refusal precludes all competition in a downstream market;
- the refusal prevents the emergence of a new product for which there is potential consumer demand; and
- the refusal is not objectively justified.
(See for example iTunes.)
These conditions seem unlikely to be satisfied in the present case, if it is true that there is already a competitive market of 3G handsets manufacturers who licence Qualcomm's patents, on terms that manifestly enable them to deliver 3G handsets to customers.
Given the above, it is not clear that there is anything of substance in this case under the heading of refusal to licence. Indeed, it is not even clear that the allegation has been made: the EE Times article only talks of "refusal to licence essential patents to potential chip set competitors on standard terms", which can be interpreted as pointing to the potential abuses discussed below rather than to a refusal to licence.
Exclusionary or discriminatory discounts
The law on the question of exclusionary or discriminatory discounts is fairly straightforward (although establishing the relevant facts might take some effort): it is an abuse for a dominant position to apply "dissimilar conditions to equivalent transactions with other trading parties, thereby placing them at a competitive disadvantage".
Here it seems clear that Qualcomm's customers are trading parties, that they compete with each other, and that the denial of a discount to one of them on the grounds that it did not wish to offer exclusivity to Qualcomm would unjustifiably place that customer at a competitive disadvantage in the downstream market — unless of course the discount was in fact justified by legitimate reasons such as different costs of supplying different volumes or otherwise different customer requirements, rather than linked to exclusivity.
In addition, if such discounts are successful in excluding Qualcomm's competitors from some relevant market by securing for Qualcomm the exclusive commitment of a critical mass of customers then such discounts, if not justified, would also be abusive because of their exclusionary effects.
If this analysis is correct, then the (difficult) questions of fact for the European Commission to consider are:
- What discounts were actually offered and given?
- Can they be objectively justified without reference to exclusivity?
Exploitative abuse
The final allegation reported by the EE Times is that Qualcomm's charges for licences are excessive, seemingly irrespective of their effects on anyone's competitive disadvantage or exclusion from the market.
I commented on exploitative abuse last week, and I still agree with myself that:
- If a firm has a dominant position in a market [...] it has a "special responsibility" not to charge excessive or unfair prices. And what characterises an unfairly exploitative price [is] the exploitation by the dominant firm of:
“opportunities arising out of its dominant position in such a way as to reap trading benefits which it would not have reaped if there had been normal and sufficiently effective competition” (United Brands, paragraph 249).
- Exploitative abuse is determined by comparing the prices charged with those that could have been charged under a counterfactual hypothesis of "normal and sufficiently effective competition".
- If a firm engages in an exclusionary abuse that has the effect of restricting competition in a market in which it has a dominant position, and if that restriction leads to higher prices than would otherwise had been achievable, then charging such higher prices is an exploitative abuse.
- The concept of a restriction on competition that is not "normal" extends wider than just illegal restrictions.
Here, there is no allegation that Qualcomm has acted illegally to obtain its patents or to bolster its position as an essential supplier to the 3G mobile phone industry.
However, there is an allegation of a "patent ambush". The complainants' allegations of unfairness also appear to draw on an allegation that Qualcomm is not complying with commitments it gave as part of the 3G standard-development process that it would licence its essential patents on fair, reasonable and non-discriminatory terms. Thus there may also be an allegation that Qualcomm misled the standard-making committees into adopting some of its technology on the basis of false promises.
Would misleading a standard-setting body about your licensing intentions, with the effect that the standard-setting body adopts a technology dependent on your patents, represent a restriction on competition that is incompatible with "normal and sufficiently effective competition"? I don't know, but I guess it would. If so, and if Qualcomm did mislead the process, and if Qualcomm has has a dominant position (i.e. if its licensing terms are not constrained by competition), then it would have (perhaps unwittingly) acquired a "special responsibility" under Article 82 not to charge more for patent licences than it would have been able to charge if it has not misled the standard-setting process.
On the assumption that all the above is true (and I have no evidence to suggest that it is), the special responsibility could be seen in two ways:
- Article 82 could be taken to require Qualcomm to comply with any promises it made about licensing during the standard-setting process (so that it would not have misled the process). If so, EC competition law would be an enforcement for certain kinds of promises by a dominant firm that have been relied upon by the market, even if these promises did not amount to contracts that are enforceable at common law.
- Alternatively, Qualcomm could be obliged to price its patent licences on the same basis as it would have done had there not been the restriction on competition arising from the standard-setting process. This line of reasoning could be seen to underpin the reported claim by the complainants that the benchmark for Qualcomm's licence fees should be the extent to which it contributed to the technology, as one might expect to happen in a competitive market in which there would be no ransom strip afforded to Qualcomm by the existence of a standard for which its patents are essential. According to the EE Times:
"The six [complainants] argue that Qualcomm has contributed less technology to the WCDMA 3G standard than it has to the CDMA2000 standard and, therefore, WCDMA royalties should be lower."
I seems to me that, in these (hypothetical) circumstances, Qualcomm would be entitled to choose which of the two methods above to use in order to comply with its special responsibility under Article 82. However, if it were "caught" in infringement of Article 82 then the Commission or a court might well seek to impose its own choice of remedy, perhaps even going further than what would have been required for compliance (as might be argued to have happened in Genzyme or Microsoft: the EC abuse of dominance case).
It should also be remembered that the special responsibility not to exploit its position in an abusive manner is without prejudice to other obligations that Qualcomm may have under Article 82, e.g. restrictions on undue discrimination between its customers if they would lead to a competitive disadvantage, as discussed above.
All the above is rather speculative: there is little EC case law about the application of Article 82 to patent ambushes in standard setting organisations. But a fairly clear statement of the starting point, which seems consistent with the logic set out above (and which highlights the differences between EC and US law in this area) was provided by the DG Competition representative to an American Bar Association International Roundtable in July 2002 as follows:
"As for Article 82, one must recall that unlike U.S. law, liability arises only for abuse of dominance, not anticompetitive creation thereof. Showing abuse may be problematic in a patent ambush context. The EC, moreover, has no equivalent to the Federal Trade Commission Act, which was the statutory basis for liability in Dell. To demonstrate this point: where a non-dominant [standard setting organisation] member intentionally conceals a patent that reads on the ultimate standard, and thereby becomes dominant as a result, is difficult to say liability arises under Article 82. Similarly, the subsequent assertion of IP rights against other members of the [standard setting organisation] may not constitute abuse of dominance, since the patent itself was properly granted in the first place. The only apparent area for Article 82 liability might arise if the IP holder applies unfair license terms, engages in excessive pricing or refuses to license in order to monopolize a downstream market."
A further matter that may concern the complainants if they are pursuing a claim for exploitative abuse is the lack of understanding of that aspect of EC competition law that seems to prevail, even amongst European regulators and professionals. This is illustrated by the focus on exclusionary and discriminatory abuses that appears to have been taken in the development of Article 82 guidelines, and in occasional attempts by officials to minimise differences between Article 82 and the US Sherman Act. Perhaps any complaint will need to go at some length over the authorities to ensure that the European Commission investigation team gives full consideration to the possibility of exploitative abuse.
On the other hand, by pursuing an exploitative complaint the handset manufacturers would avoid criticisms that they are only trying to re-negotiate the prices in their Qualcomm patent licences by the back door of Government intervention, because they would be doing it by the front door: they would be enforcing a claimed legal right to be charged less, irrespective of any complicated effects on competition.
Plug: I am happy to help, for a small fee.
Conclusion
I have already gone a long way into speculating about what claims might be advanced and what the facts might be, and I might have gone completely in the wrong direction already. So I have little to lose by going one step further and guessing how things might develop from here.
I predict that:
- Any refusal to licence complaint is likely to fail for the reasons advanced by Qualcomm: it does not look likely that there is a competitor from whom all competition was eliminated from the market as the result of Qualcomm's refusal to offer it its usual licensing terms.
- The complaint about discounts will succeed if, and only if, the European Commission finds evidence that Qualcomm did in fact offer a discounts for exclusivity that went beyond any justifiable discounts such as cost-reflective discounts for different distribution arrangements, volumes or timescales.
- The complaint about excessive royalties and any complaints about a patent ambush will succeed if, and only if, the following three conditions are met. First, the European Commission's investigation team will need to comprise people who understand the case law on exploitative abuse, or the complainants will need to explain it to them. Second, the European Commission will need to find evidence that Qualcomm misled the standard-setting process into specifying a technology over which it had essential patents, for example on the basis of false promises about future licensing. Third, the European Commission will need to show that royalty rates would have been lower if the standard-setting process had not been misled in this way.
The most interesting case from the point of view of an outside observer would be if the conditions for finding exploitative abuse are met. Besides providing a helpful application of a misunderstood area of EC law, a finding of exploitative abuse would also open up the door to private claims for damages under Article 82 to be pursued "on the coat-tails" of the Commission's decision: all that a licensee would need to do to establish his loss would be to compare the royalties paid with the royalties that would have been payable under the counterfactual hypothesis presented by the Commission as evidence that Qualcomm's charges were higher than the charges that would have prevailed under conditions of normal and sufficiently effective competition.
Update, 1 October 2007 (23 months later!): DG Competition has started an investigation into this, stating that it suspects exploitative abuse.
AstraZeneca v European Commission (Losec)
The second case considered in this article is the appeal brought by AstraZeneca, a pharmaceuticals manufacturer, against a finding by the European Commission that it had infringed Article 82 by:
- misleading patent offices into granting it "supplementary protection certificates" (patent extensions) that it was not entitled to; and
- de-registering some marketing authorisations so that they could not be relied on by competitors.
Both allegations relate to Omeprazole, the active substance in Losec, an ulcer treatment which has been one of AstraZeneca's "block-buster" drugs and went out of patent protection in the 1990s.
The Commission's infringement decision of 15 July 2005 is not yet published (I do wonder why it takes so long). It, and AstraZeneca's initial response, are outlined in this Finfacts article. AstraZeneca's notice of application to the Court of First Instance and an AstraZeneca briefing document (4 pages, PDF) on the case are also available.
Update 26 April 2006: decision published.
The Commission apparently alleges that both activities were undertaken as part of a plan to impede competition from generics and from parallel trade: the supplementary protection certificates do so directly, and the replacement of marketing authorisations for a capsule product with authorisations for tablets seems to have made life difficult for competitors who did not have access to the tablet product and could no longer rely on AstraZeneca's marketing authorisations as evidence for the safety of their capsules.
There are some interactions with EC regulations on pharmaceuticals, and a dispute about whether the misrepresentations that AstraZeneca is accused of were actually known to amount to misrepresentation at the time.
But these do not seem to go to the essence of the Article 82 case: presumably, not understanding the law is not a defence for abuse of a dominant position, although it might affect the appropriate level of any fine since there is no proportionate deterrent effect from a fine imposed for an abuse that was not negligent or intentional.
AstraZeneca's challenge to the decision is apparently based on the following points (my structure):
- The relevant market is broader than Omeprazole and AstraZeneca does not have a dominant position in it.
- Misleading patent offices into granting supplementary protection certificates is not an abuse unless and until the certificates are capable of being enforced. There is insufficient evidence that AstraZeneca misled patent offices anyway.
- There is no special responsibility under Article 82 to maintain marketing authorisations purely for the purpose of helping competitors.
- There was no strategy for excluding competitors by withdrawing marketing authorisations.
Taking these points in turn:
- The first point seems to be the usual big pharma bleat about market definition (bolstered with much bluster in the briefing document about how monopolies arising from innovation are not really monopolies that can be abused, honest...). It does not normally wash, and there is no reason to believe it will wash here: presumably doctors prescribe "Losec" or "Omeprazole" rather than "a pill that will cure your ulcer", and therefore as far as competition with generics and parallel importers is concerned AstraZeneca's supplies of Losec can only compete with other sources of Omeprazole.
- The point about abuse only arising when misappropriated rights are capable of being enforced does not seem right. Surely what matters is whether generic manufacturers and importers were, in fact, prevented from competing. Whether the impediment arose from a genuine possibility of enforcement against them or from a reasonable fear that AstraZeneca might be able to enforce a certificate and thereby destroy that part of their business in the future does not seem to matter under Article 82. In this case, abuse is, in a sense, in the eyes of the abused.
Update 26 April 2006: this appears to be the Commission's line. The Commission also presents the case as one of attempted exclusion, taking the view that it was abusive even if unsuccessful.
- The point about there being no duty to help competitors is surely right. But the question of fact that the court will probably need to consider evidence about is whether AstraZeneca effectively "sacrificed" a small amount of profit it could have made from continuing to sell capsules, or whether the incremental profits (net of cannibalisation) from doing so were actually outweighed by administrative costs of maintaining the authorisation. If the former, then the objective purpose of AstraZeneca's withdrawal will presumably be found to have been to exclude competitors, which looks like abuse. If the latter, then there is an objective justification for the withdrawal and no abuse.
Update 26 April 2006: the decision relies on (redacted) evidence that AstraZeneca's strategy was unusual to refute objective justification. AstraZeneca does not appear to have made a strong claim that there was a simple business justification for deregistration.
- The question of whether there was an overall "strategy" usually arises in the context of predatory abuses. In these cases, if there is no objective evidence of conduct that could only be commercially justified by its exclusionary effect, then an abuse case can be built on the basis of evidence that there was intent to exclude unfairly. Presumably AstraZeneca comments on both routes here because the Commission alleged that they had objectively sacrificed profits in particular instances, and that even if they had not it was all part of an exclusionary plan. That seems another difficult question of fact for the CFI to get its teeth into.
Update 26 April 2006: the decision uses (redacted) information about AstraZeneca's strategy to rebut the hypothesis that the misrepresentations etc. were mistakes by low-level employees (i.e. not negligent or intentional abuse by the company), as well as to indicate the lack of a proper business justification for deregistration.
There is one further obvious issue with the Commission's apparent view that AstraZeneca had infringed Article 82 by abusing some patent office processes through misrepresentation. It is simply this: Article 82 prohibits the abuse of a dominant position, not the abuse of a capacity to lie to a patent office official.
Maybe this is what AstraZeneca is actually getting at with the point about abuse only resulting from the ability to enforce the certificates. If so, I think that this is a strong point (whereas it does not seem to amount to much if interpreted as a rebuttal of the claim of impediments to competitors' businesses, as I argued above).
It seems to me that, if the Commission is trying to extend Article 82 to catch all anti-competitive conduct in the world, irrespective of whether the conduct actually uses a dominant position, then I think that it is bound to fail, as this is simply not what Article 82 says or has been interpreted as saying.
There is a risk that someone in the Commission might have taken some US Sherman Act or FTC Act precedent about (attempted) monopolisation or the unfair acquisition of market power and assumed that things worked in the same way under Article 82. But I don't think they do: Article 82 does not prohibit the acquisition of market power, by whatever means, except when the acquisition itself involves the abuse of a dominant position (as in a price squeeze case).
Instead of prohibiting its acquisition, Article 82 restricts the use of market power; and it does so to a considerably greater extent to US federal anti-trust law appears to do.
Update 26 April 2006: at a first glance, the decision seems clear enough that the abuse was the possibility of enforcement of the ill-gotten SPCs, not the misrepresentations that led to the grant of the SPCs. So the risk noted above does not appear to have materialised.
Despite this, it seems natural to think that there is something wrong in a pharmaceutical company extending a Government-granted patent protection by lying, if this is indeed what happened, over and above the allegations of impediments to parallel trade (where a special responsibility might arise from AstraZeneca's likely dominant position in supplying the products in the exporting country) or abuse of a dominant position in relation to marketing authorisations (where there might be some sort of dominant position associated with the ability to influence approvals of medicines).
I am afraid that this brings me back to exploitative abuse, as above and last week. Again. Sorry. But what else do you call the exploitation of restrictions on competition arising from a fear of enforcement unfairly obtained supplementary protection certificates by generic manufacturers and parallel traders? Indeed Neelie Kroes is reported to have said:
"By preventing generic competition AstraZeneca kept Losec prices artificially high."
If there is beef behind that particular bit of bluster, i.e. if the Commission can demonstrate that prices were higher than they would have been in the absence of any proven exclusionary withdrawal of marketing authorisations and any proven unfair acquisition of supplementary protection certificates, then it seems clear to me that there would be evidence that AstraZeneca (assuming a relevant dominant position) was charging prices higher than they would be under normal and sufficiently effective competition.
If so, an Article 82 infringement, together with an opportunity for health services and insurance companies across Europe to consider whether they have might have a claim for financial compensation, would seem to have been established. That could be interesting (for spectators).
Update 26 April 2006: at a first glance, there is no such beef in the decision. The case is of exclusionary abuse only, and the Commission relies on case law that there is no need to show an actual effect on prices or consumers under Article 82. But the claims of exclusionary abuse appear (at a first glance) to have been well constructed by the Commission.
Key links
- Viewpoint: Franck
My regulation and competition economics blog.
Entry added by Franck Latrémolière on 3 November 2005, updated on 26 April 2006
There may be comments and notes on this page. You can contribute by using the "Add a comment" button above. Advice on making contributions (including trackback) is in the Guidelines.
Last changed by Franck at 11:26 AM on Monday 1 October 2007.