By Jason Rantanen
One of the main pieces of empirical data being drawn upon by folks arguing for and against various patent law reform proposals is the rate of new patent lawsuit filings. As the intellectual property scholars’ letter to Congress recently observed, while there were fewer patent lawsuits filed last year than in 2013, that year was itself a record-setting year for patent lawsuit filings. Professor Matthew Sag’s new study, IP Litigation in the United States District Courts: 1994 to 2014, illustrates the point of rising patent litigation dramatically: in the sixteen years from 1994 until 2010, the annual number of patent lawsuit filings doubled; it doubled again in the three years from 2010 to 2013.
While Professor Sag’s study provides a comprehensive empirical look at intellectual property litigation rates across all areas, over at Lex Machina legal data scientist Brian Howard has dug deeply into the patent lawsuit filing rate numbers to examine whether the America Invents Act led to a reduction in patent lawsuit filings once the anti-joinder provision of the AIA is accounted for. Even when controlling for the disaggregation of multi-defendant suits, he finds a rise in patent lawsuit filings during the two years after the enactment of the AIA. While the more recent period has seen filings turn downward, his findings about the short term effects of the AIA are surprising and worth thinking about further.
Finally, as I’ve noted in the past, I find it useful to think not just about new patent lawsuit filings, which can provide a barometer of future patent litigation activity, but also about currently pending patent litigation and lawsuit termination rates. In my view, these provide a better (although still imperfect) gauge of what kind of patent activity courts are presently seeing. Figures 1 and 2 below are updated versions of the pending patent lawsuit data that I’ve previously written about.
Figure 1
Figure 2
The bottom line for this data is that while pending patent litigation remains below the all-time highs of the late 2013’s and early 2014, this activity has plateaued over the last few months and even rose slightly in February. Digging a bit under the hood of the data, this rise appears due an increase in filings over that month as opposed to a drop in terminations, with February 2015 seeing about 500 new patent cases filed—a rise of 69 over January despite the shortness of the month. Given the close attention that is paid to patent lawsuit filing rates, I suspect that this is a point that will be jumped on by many. The below graph shows the number of patent actions filed, using LexMachina data, on a monthly basis from August 2014 – February 2015.
Figure 3
In the words of the awesome David Bowie: “ohhh, wham bam thank you ma’am!“
Jason / Dennis – I’d like to see any discussion of lawsuit quantity normalized against the quantity of patents currently alive in the U.S. Seems like common sense?
You can make a guestimate, at least, can’t you Mike? Just state your assumptions and do the math. The fundamental raw numbers are out there.
There’s more patents pending right now than at any point in the history of the world. Most of those patents are US patents. There’s been a huge explosion in the number of granted patents because — primarily — the bar for partenting was lowered to the very bottom.
Who doesn’t know that?
If you’re fascinated by “normalizing”, maybe you can normalize the number of patents to the US population and rates of population over the past 50 years.
Or add up the number of people who have been approached to license a patent and/or were threatened with a patent lawsuit and/or were sued and divide that by the US population. That might be interesting too.
The point, Mike, is that there isn’t some mandated minimum frequency of patent assertion that the public has assented to. Nor is there some deep “tradition” of some minimum frequency of patent assertion that True Americans Who Support Innovation are obligated to honor. If the PTO, for whatever reason, began granting 1,000,000 patents/yr or 5,000,000 patents/yr (not unthinkinkable in our lifetimes, given the unprecedent expansion we saw in the past decade or two) there’s no way that we (you know, the people who aren’t invested the game) would put up with the “corresponding” increase in patent assertion. That would be a far sadder joke even than the sad joke of a situation that than we’re in now — especially if we end up in the Information Is Everything world that the hardcore software patent proponents often predict we’ll be in.
Or maybe he has a really good point that innovation has an exponential growth so maybe the litigation is merely tracking that.
Here is some “innovation” people got filed on for this week….
When we all have to pay more for video games and groceries, it will be WELL worth it for all the wonderful things these patents bring us…
US 8370199
US5618045
Mike –
The video game patent wast grated in 1997.
Sounds like xbox live to me. $60 bucks/year (less if the ebay listings are to be believed) sounds reasonable to me.
not sure what you mean by “this week”….
Sorry – I guess that should be address to martin… I swear it said Mike when I first read it.
I mean it was asserted this week against Leapfrog.
OIC …
And your point is… that you think the disclosed/claimed system was known or obvious in 1995 and that the patent should not have been granted?
The claims in 8370199 were ineligible in 1995, for sure.
Anyhow… Back to my question. Let’s look at facts before overstating a particular position.
I don’t see how it makes any sense to normalize against patents. Like MM said, there have been a lot of patents granted in the last 10+ years. Why not normalize against population, GDP, or some other metric that isn’t directly tied to the patent system?
Ok – that’s a good idea as well. We should look at all of the above.
MM just wants to kick up dust because you have an excellent point.
I would suggest normalizing based on the size of the industry. So, information processing has grown from 5% to 20%+ of our economy over the last 30 years.
That’s another possible metric.
As long as there is some reason for the norming factor, the norming may tell you something reasonable. But watch for the dust-kicking when some random unrelated factor is proposed.
Dennis, that is an excellent point you made about patent litigation pendency being more relevant than the mere number of complaint filings, since the main problem being complained about is the discovery and other large costs and uncertainties of patent litigation. A suit that is settled rapidly before such costs occur, OR is promptly stayed for an IPR or CBM, or is immediately shot down on a 101 basis, will not have most such costs. Likewise suits in which judges make prompt Markman claim interpretations can lead to prompt lower cost S.J. non-infringement decisions.
Oops, credit to Jason, of course.
Got to love Markman hearing’s! And a big kiss to Octane Fitness.
The increase in patent lawsuits is due, at least in part, to the recent anti-patent rulings by the Supreme Court.
The Supreme Court has created uncertainty as to the validity of patents. Therefore, infringers are more willing to challenge patents in court rather than pay a fair and reasonable royalty.
That’s an interesting theory, subjectively hard to disagree with, especially re the Alice and Mayo 101 decisions.
But, to consider it further, how does one test for the number of private subjective decisions to refuse to pay for licenses for patent assertions that are not yet patent suits?
How many patent claims that appear readily 101-vulnerable are still now being sued on?
Also, note that this theory seems inconsistent with other assertions made on this blog that more and more patent suits are filed without any attempted prior license negotiations, and that more and more companies simply ignore licensing offers until they are sued? To what extent are these true?
[Personally I think it is too soon to draw conclusions from short term complaint filing numbers while the dust is still settling from these recent Sup. Ct. decisions and the impact of IPRs, but it is an interesting exercise.]
Paul, given Octane Fitness, would not it be more likely for an accused infringer to file a DJ action where he believes the patent particularly weak, or the infringement theories far-fetched?
Ned, as you know, there are a few companies that play hardball defense on a regular basis. But most companies would not logically file a D.J. action in response to a pre-suit licensing offer [if any] until at least finding out if the licensing offer is cheaper than their likely litigation costs to win. The latter is a considerable cost exposure, since a D.J. suit automatically triggers an infringement suit counterclaim, and many judges do not hold early-on Markmans and grant S.J.s before large discovery and other costs. Also, remember that about 97% of patent suits settle pre-trial. With no final decision there is no prevailing party to award attorney fee sanctions to! Thus, as others have also pointed out, the big hot hair arguments over attorney “fee shifting” are overblown. [Pun intended] It may affect less than a hundred suits out of the few hundred patent suits of any kind that go all the way through trial to a final decision at the Fed. Cir. each year. Furthermore, financial backing for PAE suits on vulnerable business method patents has already shrunk. The current flurry of 101 D.C. decisions is a one time blip for highly 101 vulnerable suits started before the Alice decision. [Not that that this legal confusion will go away, but it should reduce.]
Paul,
You’ve done this before: they are not in fact overblown, and your notes above merely apply pre-change conditions and ig nore the possible (and even hoped for) post change calculus.
Does Brian’s work differentiate between 830 cases and DJ actions (190)? Because last time I saw their research, they lumped the two together, which confounds cause and effect if you are going to pin the blame for increased patent litigation on NPEs/patent owners.
This D.J. significance question keeps coming up in other contexts, such as arguments for patent owners suing without any prior notice or license negotiation to get venue in E.D.TX rather than a defendant venue choice. Si it is high time someone made an actual determination of the percentage of patent suits started by D.J. actions rather than infringement suits. My impression is that it is so much smaller as to not significantly change who is starting patent litigations, but logically we ought to see actual numbers before buying into arguments made without them.
Paul
Given the adverse effects of sitting back and waiting to be sued now, I would expect DJs to be in fact “smaller” NOW % wise compared to before Medimmune because any smart patent owner preempts a defensive filing. But that doesn’t mean they are negligible.
Besides still objecting to speculative theories not supported by measurable facts on patent D.J. relative frequency, I would add that these days a smart defendant [not overly influenced by an O.C. partner too eager for litigation billing opportunities] would seriously consider either paying for a license that is less than realistic litigation costs or filing an IPR or CBM for vastly less costs and far better odds than litigation.
As previously noted, without norming the results appropriately by the number of possible claims (good proxy is number of active patents), the “rise” may in fact signify a drop in rate.
Comments are closed.