How and why is the US patent system so broken? News stories about dubious patents generate grumbling, annoyance, frustration, and perplexity.
An exceptionally good book, Innovation and Its Discontents explains what’s wrong with the patent system and how to fix it. Written by two economics professors at Harvard, Adam Jaffe and Josh Lerner, the book is short, clear, well-argued, and wears its erudition lightly.
Things haven’t always been this bad. But in the 1980s and 1990s, two separate reforms — of the patent courts and the patent office — combined for a pernicious result. Bad patents became much easier to get, and harder to overturn.
In 1982, the patent appeals court system was consolidated from 12 regional courts, which had vastly uneven standards, to one centralized court. The reform halted the practice of “forum-shopping”, whereby patent-owners rushed to accuse infringers in patent-friendly courts, while challengers rushed to seek hearing in patent-friendly courts.
The practices of this centralized court made it much easier to sue for patent infringement and win. The percentage of patents upheld increased from 62% to 90% in the few years after the central court started.
A few years later, in the mid-90s, the Patent Office changed from a tax-supported agency, whose mission was to ensure that patents are valid, to a fee-for-service agency, whose mission was to quickly issue patents to those who apply. The fees from the Patent office are siphoned into the general federal budget, while the office can’t keep qualified staff. 55% of patent examiners have less than two years of experience.
The result is that bad patents sneak through without good scrutiny. The average patent claim is reviewed for only 16-20 hours, which is half the time spent in the European Union. In the time available, patent examiners look for information that is easiest for them to find — other patents in patent databases. They don’t have the time or experience to look for other sources — like existing software and academic research — that prove that the “invention” is obvious, or not new.
Meanwhile, the patent review process is mostly closed — there isn’t a good way for third parties to share relevant information about prior art until after the patent is granted. Once the patent is granted, the legal system presumes that a patent is valid, and stacks the deck against attempts to overturn a patent.
A reform in 1999 was intended to create a “reexamination process”, but the process was watered down so badly that it is almost never used. The only kind of evidence that a challenger can present is other patents (not pre-existing software, evidence of historical business practices, or academic papers). The challenger doesn’t have the opportunity to explain the evidence. If a challenger applies for a patent re-examination and loses, they lose the right to sue later.
As a result, a lot of bad patents get issued, and they are very hard to protest or overturn. Technology companies use patents to gain license fees from competitors, who will settle rather than go to court, even if the patent is bad, because an infringement allegation is too costly and risky to defend. Large competitors create cross-license patent libraries that maintain the advantage of the leaders, and freeze out smaller players.
So how can the system be improved? Jaffe and Lerner recommend a tiered approval and review process, where patents can be issued quickly, but there are several stages where challengers and third parties can submit prior art and try to prove that the patent is obvious or not new. They also recommend reduced use of juries, who lack expertise to evaluate the information.
The book has interesting observations about the failure of patent reform efforts in the 90s. Talk show celebrities including Oliver North and G. Gordon Liddy used the issue to grandstand against Japan, who were competing against US manufacturers. Patent lawyers, who gain from the current system, were well-organized. At the time, the technology industry was not well-organized, and there was little public interest in patent reform.
Thanks to Doug Barnes for recommending the book, which joins my short list of favorite non-fiction. It takes a puzzling and potentially abstruse subject, and explains it clearly. It uses stories and well-chosen research data to make its points. And it shows a potential exit for the tangled mess of the US patent system.
Patent reform is in the works again in Congress. The book is very helpful context for the debate.