The arts of prosecuting, litigating, monetizing and teaching patents can progress only so far without benefit of empirical, quantitative data. To that end, we’re gathering up the most useful patent statistics available today. Most of them are little known, because they come from a wide variety of scattered sources where they are typically presented in raw form or in some other user unfriendly manner.
We’ll be expanding and updating our collection of patent statistics on a weekly basis. The section below, Featured Statistics, is a preview of things to come. As the collection grows, we will organize the statistics by topic and subtopic.
From 2005 to 2013, U.S. patents litigated to final decision were invalidated under the following grounds at the following rates:
- 1.7% under § 102(a) earlier public knowledge or publication
- 3.9% under § 102(b) time-bar publication or patent
- 1.4% under § 102(b) public use
- 1.7% under § 102(b) on sale
- 1.0% under § 102(e) earlier-filed US patent of another
- 0.5% under § 102(f) derivation or incorrect inventorship
- 0.7% under § 102(g) prior invention by another (novelty)
- 8.7% under § 103 obviousness
- 2.6% under § 112(1) inadequate written description
- 1.5% under § 112(1) lack of enablement
- 0.5% under § 112(1) best mode (no longer a ground for invalidity)
- 3.9% under § 112(2) claim indefiniteness
- 0.2% for double patenting
- 0.4% for improper broadening of reissue
Some comments about these numbers are in order.
First, the dataset on which they are based consists of 3316 case decisions from 2005 to 2013, which were coded as to rulings on 40 legal issues by the University of Houston Law Center. The dataset includes all Federal Circuit rulings (including summary affirmances under Rule 36) and all reported lower rulings from district courts and the ITC. See U.S. Patent Litigation Statistics, Institute for Intellectual Property and Information Law, University of Houston Law Center, at patstats.org (visited Jan. 2017)
Second, note that, under all parts of 102(b) combined, 7% of patents are invalidated under 102(b), which makes it four times more likely than 102(a) to invalidate a litigated patent.
Third, more patents are invalidated under 102 (10.9%, all 102 sections combined) than under 103 (8.7%), which belies the notion that non-obviousness is the primary gatekeeper of patent law.
Fourth, indefiniteness under 112(2) has become more important. At an invalidation rate of 3.9%, it now accounts for almost as many invalidations as written description and enablement combined (4.1% combined)! From 2005 to 2009, claims were invalidated under 112(2) in only about 9.4 cases per year; from 2010 to 2013 that number doubled to about 20.8 cases per year. In light of recent academic work and Supreme Court decisions expressing concern about indefiniteness, we suspect the importance of 112(2) has only grown since 2013.
Fifth, for cases from 2010-2013, the dataset identifies each case for which each ground of invalidity was decided. This data shows that, when enablement and written description are both adjudicated, they rise and fall together. That is, in the 23 cases from 2010-13 in which both enablement and written description were decided, the outcome for both was the same: either both requirements were found satisfied or both were found unsatisfied. This tends to corroborate Ariad’s position in Ariad Pharmaceuticals v. Eli Lilly, 598 F.3d 1336 (Fed. Cir. 2010) (en banc), in which Ariad argued there is no substantive difference between enablement and written description.
The numbers below are based on the University of Houston’s dataset of 3316 cases decided 2005-2013, discussed above in the section on Invalidity.
Inequitable Conduct. In 3.6% of the cases, the patent was held unenforceable due to inequitable conduct. When the issue was litigated to decision, the accused infringer won only about 26.5% of the time.
Note that the rate at which the issue was litigated to decision fell over the years. From 2005 to 2011, it was litigated to decision about 53 times per year on average. From 2012 to 2013, it was litigated to decision only about 35 times per year on average. We suspect the rate has fallen a bit since 2013.
Patent Misuse. Patent misuse was decided in 33 (or 1.0%) of the cases. Of the 33 cases, the accused infringer won only 1 (or 3.0%) on this ground.
Laches. Laches was decided in 94 (or 2.8%) of the cases. Of the 94 cases, the accused infringer won 22 (or 23%) on this ground.
Equitable Estoppel. Equitable estoppel was decided in 61 (or 1.8%) of the cases. Of the 61 cases, the accused infringer won 17 (or 28%) on this ground.
Failure to Mark. Failure to mark was decided in 44 (or 1.3%) of the cases. Of the 44 cases, the accused infringer won 29 (or 66%) on this ground.
Preexisting License. Whether the accused infringer had licensed the patent in question was decided in 54 (or 1.6%) of the cases. Of the 54 cases, the accused infringer won 28 (or 52%) on this ground.
Exhaustion. Whether first sales had exhausted the patentee rights was decided in 81 (or 2.4%) of the cases. Of the 54 cases, the accused infringer won 48 (or 59%) on this ground.
Exemption under 271(e). Whether the accused infringer was exempt under the FDA safe harbor provision of 271(e) was decided in 15 (0.45%) of the cases. Of the 15 cases, the accused infringer won 7 (or 47%) on this ground.
Exemption for Experimental Use. The common law defense of experimental use was not decided in any of the cases. It may therefore behoove patent law professors to remove experimental use from their syllabi.
Doctrine of Equivalents
Of every 100 findings of direct infringement by U.S. courts, only about 10 are findings of direct infringement under the doctrine of equivalents.
This number sheds light on the importance of DOE scope relative to literal scope. Some applicants may spend a lot of time and energy on the DOE and the nuances of the Festo case and its progeny, or go to lengths to avoid amending claims in ways that could curb DOE protection. When that translates into giving short shrift to the literal language of claims, it’s the tail wagging the dog. Applicants should bear in mind that when a patent is found infringed, nine out of ten times it’s found literally infringed.
Note also that findings of direct infringement likewise dominate findings of indirect infringement. For every 100 findings of infringement, 81 are findings of direct infringement (literal or DOE), whereas 19 are findings of indirect infringement (primarily inducement under 271b or contributory infringement under 271c). Whether the accused infringer induced infringement under 271(b) was litigated to decision in 222 (or 6.7%) of 3316 cases from 2005-2013. Of the 222 cases, the patentee won 95 (or 42.8%). Whether the accused infringer contributed to infringement under 271(c) was litigated to decision in 140 (or 4.2%) of the cases. Of the 140 cases, the patentee won 58 (or 41.4%).
The numbers above are based on the University of Houston dataset discussed above in the section on Invalidity.
Patent Litigation Win Rate by Venue
From 2000-2010, the national average rate at which patent cases settled was 75.5%. 15% of patent cases went to final judgment. The overall win rate for patentees was 32.5%. Win rate, however, varies dramatically by venue. For example, the patentee win rate in the Northern District of Texas (which decided 405 cases) was almost five times higher than it was in the Northern District of Georgia (which decided 457 cases)! Listed below are the venues with the six highest and six lowest patentee win rates.
Top Six Patentee Win Rates
- N. District of Texas 55.1%
- M. District of Florida 46.3%
- District of Nevada 46.2%
- District of Delaware 45.3%
- District of Oregon 45.2%
- E. District of Texas 40.3%
Bottom Six Patentee Win Rates
- W. District of Wisconsin 24.0%
- District of New Jersey 21.0%
- W. District of Washington 20.0%
- E. District of New York 17.6%
- E. District of Wisconsin 17.1%
- N. District of Georgia 11.5%
These numbers come from a study by Professor Mark Lemley, in which he examined a dataset of 21,667 patent cases decided between 2000 and 2010 and reported the win rates for all venues that decided 25 or more cases over that 10-year period. To see the complete list of win rates, See Mark A. Lemley, Where to File Your Patent Case, 38 AIPLA Quarterly Journal 1 (Fall 2010).
Lemley also reports the average times to trial. The two quickest venues were the Western District of Wisconsin (0.67 years) and the Eastern District of Virginia (0.96 years). The two slowest venues were the Eastern District of Michigan (3.41 years) and the Eastern District of Wisconsin (3.51 years).
The national average rate at which patent cases went to trial was 2.8%. The rate was highest in the District of Delaware (11.8%) and the Eastern District of Texas (8.0%). It was lowest in the Southern District of Ohio (0.5%) and the District of Columbia (0.0%).
Jury Trial vs. Bench Trial
Over the last 5 years, patentees won 73% of jury trials and 53% of bench trials. The median jury award was 16X higher than the median bench award! See the PWC study entitled 2016 Patent Litigation Study: Are we at an inflection point? (May 2016).
Given the numbers reported in this study and also in the study discussed below under Declaratory Judgment, we predict that the percentage of trials heard by juries (75% over the last 5 years) will grow even higher.
If a patent suit will be decided by a jury, the patentee greatly increases its odds of winning if it files suit first, i.e., before the accused infringer files for declaratory judgment. According to a study by then Professor Moore, in cases decided by juries, when patentees file suit they win 65% of the time, but when accused infringers file declaratory judgment actions patentees win only 42% of the time. See Kimberly A. Moore, Judges, Juries, and Patent Cases-An Empirical Peek Inside the Black Box, 99 Mich. L. Rev. 365 (2000).
The first thought that arises in response to this is that suits initiated by accused infringers must tend to differ on the merits from suits filed by patentees. But here’s the rub: in bench trials, win rates do not vary with who files first. According to the study, the win rate in bench trials (about 50% for patentees) remained the same regardless of who filed first. The reason for this surprising finding is unclear. Perhaps jurors tend to believe (somewhat incorrigibly) that whoever filed the suit is the deserving party.
Injunctions & Enhanced Damages
From 2005 to 2013, patentees obtained preliminary injunctions in about 4.1% of patent cases. This figure is not much changed from years past. One study of preliminary injunctions issued in 1990-91 showed that patentees obtain them in about 5% of cases.
Patentee requests for permanent injunction were ruled on in 253 cases from the date of eBay v. MercExchange 547 U.S. 388 (2006) through 2013. Patentees won 74.7% of those requests.
From 2005 to 2013, damages were enhanced for willfulness in only about 3.2% of cases. Patentees won attorneys fees in about 3.3% of cases; accused infringers won them in about 3.6% of cases.
The 5% figure for older preliminary injunction cases comes from Jean O. Lanjouw and Joshua Lerner, Preliminary Injunctive Relief: theory and evidence from patent litigation, NBER Working Paper No. 5689 (NBER 1996). The figure for permanent injunctions is based on a supplementary dataset reported by the University of Houston. The other figures are based on the University of Houston’s dataset of 3316 cases decided 2005-2013 , discussed above in the section on Invalidity.
Foreign Owners of U.S. Patents
Foreign owners of U.S. patents are about five times less likely than domestic owners of U.S. patents to sue for infringement of their U.S. patents. Original source: Jean O. Lanjouw and Mark Schankerman, Stylized Facts of Patent Litigation: value, scope and ownership, NBER Working Paper No. 6297 (NBER 1997); Chapter: Patent Litigation – Enforcement of Patent Rights in the United States, Patents in the Knowledge-Based Economy (Nat’l Academy of Sciences 2003).
These studies are based on old patent cases, namely, those from 1975 to 1991. But assuming it’s still true that foreign owners of U.S. patents are less likely to sue, the question remains as to why. The study authors suggest that it might be harder for foreign patentees to detect infringement and also that their costs of litigation might be higher. I would add that many foreign patentees face greater uncertainty about both the costs and the benefits of U.S. litigation. Also, they may simply be less inclined to litigate for cultural reasons – Japan comes to mind.