US PTO Mired, but Some Journalists Misinterpret

The Washington Post has a story about examiners at the US Patent and Trademark Office being driven away by production quotas, which haven't changed for decades even as patents have become increasingly complex and difficult to review, and the massive backlog that would take two years of the full time of all examiners to clear up, by which the incoming patents would have buried the office even more thoroughly. And for every two examiners hired, one leaves, according to this Information Week article on the report. But then, if you had to work through your vacations just to get caught up and avoid being fired, you might leave as well. A third of the examiners are apparently leaving after less than one year on the job, and 70% before five years - which, according to USPTO offices I've spoken with recently, is about the time that the examiners really come into their own and can work without supervision. In other words, most of the training the agency provides walks out the door.

Unfortunately, the Post reporters apparently spoke just to USPTO director John Dudas and reading a recent GAO report on the agency. Dudas seems to blame the backlog on poor information in the patent applications:

A quarter of applications arrive with no supporting materials and another quarter carry more than 25 references to supporting data, he said. Although an extreme example, Dudas said the agency once received an application that came in 28 boxes, with 2,600 pages per box.

What he doesn't say is that lawyers are afraid of putting themselves and their clients in a legal position of jeopardy. Include a moderate amount of supporting information, and a competing company could try to find so-called prior art that you missed and then claim you were essentially trying to pull one over on the USPTO, which could invalidate the patent. And so you have two schools of thought: provide nothing, so you don't exhibit any selectivity, or provide everything, so you can't be accused of skimping. Dudas's quote of needing "the best material. Not the kitchen sink. And not nothing," shows the problem. What constitutes the best?

By demanding the given levels of productivity, the USPTO is focusing on statistics that might make it look as though the situation is improving. That's understandable, when you need to lobby Congress for your operating budget (and Dudas is skilled at that), but it's really dealing with the symptoms and not the root causes. The European and Japanese patent systems take different fundamental approaches - allowing people a given amount of time to raise objections after the grant of a patent - than the US system.

And the focus that the USPTO makes on essentially blaming the users - and, frankly, some of that is warranted - also acts as a way to deflect attention from the agency's responsibility to find the appropriate prior art. And so the USPTO put into place the new continuation rules - a change in the basic approach to patents, and one that the office is apparently implementing even before their official start date, according to a number of people in the patent community that have found document ion showing this. If a company relies on the US patent system at all, then it will have to reevaluate its own approaches to patents and might possibly have to overhaul its strategic planning. That's how serious the situation now is.

Peer Patent Reviews Start at USPTO

The U.S. Patent and Trademark Office has just started on a year-long experiment in peer patent review. The way the process now works is that companies and people submit patents, often with broadly-phrased claims, in hopes of getting a license on a new way of doing something. PTO examiners will push back on many of the claims (fewer than half of applications reviewed last year were immediately granted a patent). But there's been question in the patent community with how well the process is working. For example, look at the case of a company suing the Blackberry's manufacturer for patent infringement. RIM lost in court and had to pay hundreds of millions to the plaintiff, yet now there are some challenges to the patent, which may not survive. It would be too late for Blackberry, which wouldn't get to recoup the money.

So the PTO is working with New York Law School's Institute for Information Law and Policy to try a peer review approach. Companies (including Microsoft and IBM) voluntarily provide software patent applications and anyone can comment and look for prior art that would show the invention isn't new or novel.

The idea isn't actually new; both the European and Japanese patent offices have used a different patent model from the US. Instead of doing all reviewing in-house, they grant applications and then providing a period of time in which others can challenge the basis of the patent. And software is a great place to try this - it's one of the biggest causes of patent application overload and one area where much of the community is already used to peer review in the form of open source software. But with backlogs that can run from a couple of years to five or six (particularly in biotech), it's worth a try. Almost anything is.