Tuesday, March 10, 2015

Dueling Patent Reform Letters to Congress

We now have dueling patent reform letters to Congress by economists and legal scholars.

The pro reform side. The anti-major-reform side. My work - including the very same article - is cited in the bibliography added to both letters. I take professional pride in this fact, but it also means that there are arguments to be made by both sides with the same data.

I signed on to the second letter, which was a little bit uncomfortable, because that letter (rightfully) criticizes the generalizability of my data. This has long been a concern about this particular string of articles (there are now three) because I focus on the most litigious patent plaintiffs. In the articles, I explain the methodology and what I think can and cannot be drawn from the data, but readers will have their own takeaways.

In the end, though, I signed the second letter because I think the first letter overstates the case for patent reform. It was intended, I believe, to counter the argument that there's no evidence of harm from patent trolls. But I think this argument is a straw person; I'm not sure anyone says that there is no harm from bad actors in the system. The dispute, it seems to me, is about how much harm there is, and whether particular patent reforms are solutions to that harm. On this, I remain unconvinced.

Regardless of how you fall on the spectrum, these two letters together form a large bibliography on the topic, though they each leave out salient points from the other side, which is nothing new. After the jump, I'll summarize what I think the bulk of these studies tells us, and what remains unanswered. (Note that the only copy of the full pro-reform letter I can find online is on SSRN.)

Here's the accumulation of what I think these studies (and some other studies not cited by either) tell us:

  1. The population is growing
  2. There have been significant technological advances in particular areas in the last 25 years
  3. The number of patents grows as population and technological advances grow
  4. The number of  patents has been growing
  5. The amount of patent litigation grows as patents grow
  6. The amount of patent litigation grows as a shock when there are technological advances
  7. The amount of patent litigation has been growing
  8. The rate of litigation is at the high end of historical norms in terms of number of patents, but seem to have spiked in the last 5 years
  9. It's unclear how much of the spike is due to the AIA requirement to file separate suits against every defendant, because the number of defendants has grown, but not as rapidly as the number of suits
  10. The percentage of companies making patent demands or filing litigation, but that do not sell products (NPEs) has been growing
  11. People cannot agree about how much the percentage has grown, in part because people cannot agree on a classification for who NPEs are or which ones are studied in any given study
  12. Some NPEs are run by the inventors, who get the lion's share of the proceeds, some are research firms that spend money on research and development, and some buy their patents
  13. Different people feel differently about the value to society of each of these firms, but most agree that it is better if more money is returned to inventors
  14. Most studies do not measure dynamic effects; that is, how much money would be returned to inventors in a system where they could not sell their patents for someone else to enforce (and thus might eliminate the ability to obtain payment for patents or technology ex ante). We don't know how much effort inventors put into exploiting their patents (and transferring technology) before working with a firm to assert the patent. There are case studies, anecdotes, and court testimony that imply many do, but inventors rarely share such information after they have begun asserting their patents and we do not know about aggregate efforts
  15. Companies that make and sell stuff seem to do better in litigation than NPEs, though how much better depends on the type of NPE and how aggressive the parties are, because the more people you sue, the more likely it is that your patent will be invalidated
  16. Though some NPEs tend to do worse, they do not lose cases at a rate that implies frivolousness as compared to other plaintiffs. Their patents survive summary judgment challenges at rates similar to other patent plaintiffs
  17. Product companies appear more willing to try cases, while many NPE cases are of shorter duration and settle more often
  18. It is on average cheaper to litigate a case against an NPE
  19. While the exact magnitude is highly disputed, companies pay a lot of money in the aggregate to litigate and settle cases with NPEs. No one has studied how this compares to amounts paid to litigate and settle cases with other plaintiffs. This number could be more or less than the amount paid to NPEs. There is no evidence that NPEs extract higher license fees
  20. Small companies that are sued by patent owners can struggle. Surveys on this do not show that the struggle is higher or lower when it is an NPE that files suit.
  21. Small companies have sold their patents to NPEs to raise cash for survival.
  22. Large companies have used patents as a barrier to entry against small companies, especially in software
  23. Patent litigation might reduce money spent on research and development by defendant firms. There is no evidence that the type of patent owner makes a difference
  24. It is better to have ex ante technology transfer than ex post assertion licensing. This is true no matter who the plaintiff is.
  25. One concern in patent litigation is a nuisance settlement, because patent litigation is so costly. People differ on how high a settlement must be to be considered a nuisance. There is no evidence about whether nuisance demands are for better or worse patents. There is no cheaper procedure for resolving disputes to reduce nuisance demands
These are  my 25 takeaways from the studies. Those who favor strong patent reform would probably says I've shaded these against the findings that support their position a bit. Perhaps, but that's probably why I didn't sign that letter. I just don't think the studies say more than this.

It may well be that too much patenting and patent litigation is harming innovation so much that we should make it harder to assert valid patents; but I haven't seen good evidence of that particular fact, nor have I seen a strong empirical case that the patent owner type should make a difference when we mete out policy. Yes, getting sued is costly, and might limit innovation. But what we don't know is what would happen in the alternative world.

That said, the key takeaway for me is, and for some time has been, that patent litigation is too expensive. It's not like I think it's great for startups (or anyone else) to spend so much money defending against lawsuits, especially baseless ones. There needs to be a cheaper way to get to the bottom of whether a patent owner has a patent right to vindicate. Post grant review proceedings hold some promise, but they seem to be proving that only large companies can bring such actions, and they are a tool for delaying, rather than hastening, determinations.

I favor patent small claims courts, but I don't know whether they would work in practice (though I wrote some PTO comments suggesting ways). Other methods may be arbitration. I also favor phased or otherwise limited discovery that allows courts to get at the key disputes early in the case, rather than allowing cases to languish in costly electronic discovery when the whole case hinges on construction of a couple claim terms. But I don't think that broader reforms will be helpful to achieving the end of reducing litigation costs; like the PTAB reviews, I agree with others that fee shifting provisions are likely to cause patenting to become even more consolidated among the large firms that can afford it, and I don't see that as the type of outcome pro-reformers are looking for.

No comments:

Post a Comment