Chillin'Competition

Relaxing whilst doing Competition Law is not an Oxymoron

The Economist Corner (2) – Patent Settlements in the Pharmaceutical Sector

with 3 comments

For this second edition of the Economist Corner, Benoît Durand from RBB Economics has sent us a post on patent settlements in the pharmaceutical sector. Benoît advocates against a per se approach to such agreements, and gives examples of pro-competitive settlements. The topic of Benoît’s post is timely. At a Brussels event last week, an EU official hinted that COMP would likely not treat such agreements under a per-se approach. He also said that the Commission would seek to provide as much legal certainty as possible on the issue. As some of you may know, the Commission dropped several settlement cases lately (GSK; AstraZeneca) but still continues to scrutinize other cases (J&J v. Novartis; Cephalon v. Teva; Servier (Perindropil); Lundbeck).

Following the conclusion of the pharmaceutical sector market inquiry in 2009, the European Commission has launched a number of investigations on patent-settlement agreements that include a payment between an originator and a generic company (also called “reverse payment” settlements[1]).  The Commission is worried that some of these payments may be used used by originators to reward generic manufacturers for postponing the launch of cheaper drugs on the market.  It is easy to see that the originator has a strong incentive to delay generic entry in order to continue earning a monopoly rent on the sale of its patented drug.  When the patent exclusivity expires generic drugs may begin challenging the originator’s monopoly position.  In this case, sharing part of the monopoly profit with a potential entrant is a better outcome than letting competition eat the rent away.[2]

However, as you might have guessed, reality is more complicated, and perhaps surprisingly, patent settlement agreements involving reverse payment need not necessarily be anti-competitive. The first thing to note is that the originator drug is protected by a patent, and it is only when the legal exclusivity expires that generic drugs may begin challenge the originator drug.  The second thing to note is that the validity of a patent is never a sure thing, and therefore its expiry date is uncertain.  Even though a pharmaceutical company has filed a patent, generic entrants may still challenge the incumbent before the formal expiry date.  Generic producers may consider that they have a good chance of challenging the patent in courts.  In the case of entry, the patent holder would seek an injunction to prevent entry, but judges may or may not grant the injunction, and they may or may not uphold a patent.

To save on litigation costs the originator and the entrant may decide to settle their dispute.  In this case a payment from the originator to the generic manufacturer might be required for the parties to agree.  The payment compensates the foregone profit opportunities for the generic producer.

Is this agreement anti-competitive?  The answer to this question involves an assessment of consumer welfare under the actual and counterfactual scenario of no settlement agreement.

On cursory analysis, entry of generic drug fosters consumer welfare in the short run, so it must be the case that delaying entry is anti-competitive.  But this view is oversimplistic.  Originators have the right to prevent entry (through legal actions) because of the patent they hold, which in turn, allows them to reap the returns from their investments in developing new drugs.  In the long run consumers also benefit from innovation.

In reality, if parties enter into a settlement agreement, it is because they are uncertain about the validity of the patent.  To take but one example of this, consider that the patent is “strong”—that is, there is a high chance that a court would find the patent valid.  Nevertheless, the originator may prefer to settle because the risk of an adverse outcome in case of litigation is simply too high—a patent holder may be very risk averse in particular when the patent dispute concerns a major source of revenue.  Even though the patent is strong, the originator would prefer a sure outcome—the settlement agreement—over a probabilistic win.  This means that when the patent is strong, the settlement agreement may allow generic entry before the formal patent expiry date, and this will benefit consumer welfare in the short-run .  This is because generic entry would take place earlier than under the counterfactual litigation scenario – i.e. no entry pre patent expiry – and price would be reduced as a result.  In the counterfactual case, there is a high probability that the courts would decide in favour of the originator, and entry would occur only at the end of the exclusivity period.

Under the circumstances described above, there is good chance that consumers are better off in the short run.  But there are other scenarios under which consumers would be made worse off, for instance, if the settlement agreement provides for entry post-patent expiry, or if the patent is weak.  All of that is to say that a formalistic approach is unable to determine whether this type of agreement is anti-competitive.  It is only through a thorough examination of these agreements, the size of the reverse-payment, the agreed timing for generic entry, the view of the parties at the time of the agreement (in particular their view about the patent strength and their aversion to risk), the litigation costs and the overall economic context in which the agreement took place that the authority would be able to assess whether such agreements are detrimental for consumers.

Benoît Durand , 14 March 2012

Ps: I have received interesting and challenging comments on the issue related to state aid in the banking sector.  I will come back on this topic, which is still very much in the news with the recent court decision regarding ING.

Ps 2: I have also noted the beginning of a debate between Alfonso and Nicolas about the role of economics in competition policy, legal certainty and whether we should let economists rule.  I will provide some view on that debate in a future post.


[1] Typically, apatent settlement involves a payment from the infringer to the patent holder.  After all, the patent holder has been deprived of its right to exclude, which results in lost profit.  In some cases, the settlement is reversed, and the payment remunerates a generic entrant that has not yet launched its product on the market.  The payment serves to compensate the generic’s foregone profit.

[2] In the US, under the Hatch-Waxman, the first generic entrant is granted a marketing exclusivity of 180-days.  In this context, reaching a settlement agreement with the first entrant has the benefit to delay entry by all other generics as well.  The law was changed in 2003. The marketing exclusivity was dropped if a settlement took place.  In Europe, the originator has to agree with all potential entrants.

About these ads

Written by Nicolas Petit

21 March 2012 at 11:03 am

3 Responses

Subscribe to comments with RSS.

  1. Benoit, I agree with your nice post, but there is an important twist which you put aside. It seems clear that most patent settlement cases center around the uncertainty of the validity of patents. Both originators and generics can save transaction costs and reduce risk by settling on terms that reflect the expected strength of the patents. But monetary exchanges are prima facie not necessary for this. Firms could simply negotiate over the timing of entry alone.

    Prima facie, a settlement on “delay without pay” should be expected to reflect the expected strength of patents. If the firms believe that the originator would be likely to prevail in patent litigation, they will settle on a relatively late entry date (not much before the expiry of the originator’s patents). If, on the other hand, firms believe that the originator’s patents are weak, they will settle for an earlier entry date (i.e., only moderate delay). Such settlement on “delay without pay” can generally be expected to benefit consumers: it overcomes uncertainty and transaction costs, but at the same time it creates no risk of collusion. The generic would have no reason to agree to delay entry more than is appropriate given the estimated strength of the originator’s patents. In that way, the generic has incentives to negotiate in the interest of consumers and concede delayed entry only to the extent that the originator’s patents are in fact perceived to be strong.

    In other words, in order to explain why a monetary payment is necessary to induce an efficient market outcome, more is needed than the recognition that patents are probabilistic. Now, I think there can be a number of good reasons why payments are necessary. But in the simple world sketched in this reply and your original post, a payment on top of delay would simply amount to a bribe to delay entry further than is necessary to reflect the expected strength of patents. The discussion should, in my view, therefore mainly focus on the specific reasons why “pay without delay” is not good enough to generate beneficial outcomes for consumers.

    hzenger

    21 March 2012 at 1:01 pm

    • Hans, I agree with you that a payment is not always necessary for the originator and the generic producer to settle their dispute. However, if they negotiate only on the entry date, this may not be enough to reach an agreement. In particular when the parties’ respective assessment of the probability that the patent is valid is very different, in that case, it might be helpful to have a payment to bridge the gap.
      I recognise that this might not have been very clear from my inital post.

      Benoit Durand

      22 March 2012 at 7:50 pm

  2. Hans, Benoît,

    Even if the parties’ respective assessment of the probability rhat the patent is valid is the same, it may be the case that joint-profits are higher when parties go to court rather than settling without delay. This could be the case with small court costs even in very simple models.
    More over, in the presence of asymmetric information (extreme case of what was suggested by Benoit in his comment), delay might also play a crucial role.
    It thus makes it difficult (although this should not prevent competition authorities from trying) to identify unjustified delay.

    Thibaud

    26 March 2012 at 1:40 pm


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Follow

Get every new post delivered to your Inbox.

Join 1,019 other followers

%d bloggers like this: