Why We Need Better Patent Quality: There Are a Lot of Invalid Patents Out There

Prof. Michael Risch from Villanova University School of Law recently put out a great study on patent quality and patent troll litigation. Prof. Risch’s study is possibly the most thorough one on patent quality to date. He followed patent lawsuits from the 2000’s, filed by both patent trolls and non-patent trolls. For every patent in each of those lawsuits, Prof. Risch found every lawsuit involving the patent. He looked at over 1300 non-troll cases and over 900 troll cases that were active from 2000 to 2009.

There are a number of findings in the paper, but I want to focus on one in particular.

Invalidity Decisions Usually Go Against Patent Owner

Risch found that, of the patents where there was a decision on invalidity, 91% of troll patents had at least one invalid claim (over half were completely invalidated), while non-troll patents had at least one claim invalidated a little over half the time. As Prof. Risch points out, not all cases have an invalidity challenge. But the fact that so many patents are invalidated at least partially is stunning, because the standard of proof is so high. (To invalidate a patent, the accused infringer has to prove invalidity by clear and convincing evidence, which is much harder to do than the typical burden in a civil case. Normally, you would just have to prove something is more likely true than not.)

Patents don’t do well when they’re challenged at the USPTO either. As of this summer, in 59 out of 78 cases the patent challenged in an inter partes review was completely invalidated, and another 19 cases invalidated at least some of the challenged claims.

A Simple Explanation

While it’s possible that we have a whole bunch of rogue juries and judges, there’s a simpler explanation: most issued patents (or at least most of the issued patents worth enforcing) are invalid.

The USPTO, at least in recent history, has done a poor job of weeding out applications that should not be awarded patents. This isn’t baseball where getting a hit a third of the time is great. Considering that the federal government is giving out a monopoly that’s difficult and expensive to challenge, it’s unacceptable to hand out patents like door prizes.

As I wrote recently, the culture at the USPTO focuses on pushing applications out the door, which results in more allowed patents. The net effect is, as we’re seeing, a lot of patents that should never have been issued.

Under current law, an issued patent is strongly presumed valid. Considering how badly patents do when challenged, this presumption makes no sense. Unfortunately, changing the presumption requires action by Congress, and political realities make that unlikely.

The USPTO Should Not Be Selling Patents

The place we can improve things is at the USPTO itself. The USPTO is going to have to change its focus from processing applications as quickly as possible to doing the job right. If that means that applications pile up from stubborn applicants who refuse to concede, then so be it. Patents should not be for sale. If you don’t deserve a patent, you shouldn’t get one. If you do, you should.

We cannot continue the excesses of the past.  Invalid patents don’t benefit innovation, they block innovation. And we have a patent system where a substantial portion of the issued patents, if not most, are invalid. And the patents in the software area are even worse.

The USPTO has got to shift its focus to patent quality. Now.


Matt Levy

Image of Matthew Levy

Matt Levy is Patent Counsel at the Computer and Communications Industry Association, where he handles legal, policy advocacy, and regulatory matters related to patents and is lead blogger for CCIA’s Patent Progress. He is recognized as an expert on patent litigation reform and is cited and quoted widely on the issue.

Matt joined the CCIA in 2013 from private practice, where he got first-hand experience in both patent prosecution and patent litigation, including defending clients against patent assertion entities.