Patent Reform and the Progress of Innovation
Senate Passes America Invents Act 95-5
Patent Reform 101
To quote the Supreme Court, a patent may be granted for “anything under the sun that is made by the hand of man,” provided that the invention fulfills the three basic requirements of being novel, useful, and non-obvious. Among the immense variety of inventions that have been patented are new materials or devices; methods for making those materials or devices; novel life forms created through the manipulation DNA; mathematical algorithms used to analyze data, organize information, or predict human behavior.
Read more about patent reform at Science Progress.
The U.S. Constitution empowers Congress to design a patent system that will promote the “progress of … the Useful Arts”—that is, the progress of innovation. A system intended to spur innovation will, from time to time, itself benefit from change. But achieving legislative consensus on what change is necessary can be fiendishly difficult. Even if, upon dispassionate parsing, the relevant economic data clearly support a particular reform, such reform will generally create losers who are unlikely to give up without a fight.
Indeed, for the greater part of the last decade, legislative proposals for patent reform have languished as competing industry and trade group interests fought to a standstill. Meanwhile, the Supreme Court and the executive branch have taken up the mantle of reform, producing significant improvement in such areas as standards for evaluating patentability, remedies for patent infringement, and numerous administrative reforms to improve both the timeliness and quality of review.
Yet there is important patent reform that only the legislature can implement. Thus, last week’s 95-5 Senate vote in favor of the America Invents Act represents something of a watershed. Indeed, given the background history of fierce clashes between interest groups, and the everyday difficulties of achieving bipartisan agreement, the Senate action deserves commendation irrespective of what ultimately happens in the House.
There is no question that our patent system needs further improvement. It currently takes at least 34 months on average for the Patent and Trademark Office, or PTO, to finalize a patent application. The average wait is even greater (more than 40 months) if one counts the large number of applications that are resubmitted one or more times. As a consequence, a backlog of more than 700,000 patent applications currently awaits review. These long waits create uncertainty for inventors, investors, and entrepreneurs, dampening innovation. What’s more, although the PTO is entirely supported by applicant fees, it does not even possess fee-setting authority. No self-supporting institution that lacks the ability to charge for the costs it incurs can possibly perform efficiently.
The America Invents Act, if taken up by the House of Representatives and signed by the president, would address these problems in several ways. First and most significantly, it would confer fee-setting authority upon the Patent and Trademark Office. As I have discussed at length elsewhere, with fee-setting authority of the sort given in the Senate bill (and a related guarantee that fees won’t be diverted by Congress to other uses, as has happened repeatedly in the past), the PTO could much more readily work to create processing times that were suitable both for applicants and for potential competitors. The PTO could also, if it so desired, use this authority to create incentives for applicants to file well-drafted applications that could be processed in one cycle without having to be resubmitted, thereby further reducing administrative delays.
Another significant (though contested) feature of the legislation is its enactment of a first-to file system. Currently, unlike every other major global economy, the United States operates under a “first-to-invent” system. At least in theory, the U.S. approach therefore accords patent rights to whomever can prove they first invented a product or technology. The operative word is theory, however, as the actual impact of the first-to-invent system is far from clear. So-called interference disputes—that is, the length and expensive administrative proceedings in which first and second patent filers settle who was first to invent—account for only about .01 percent of cases. Meanwhile, the asymmetry between the U.S. system and that of other major economies creates problems both for applicants who file globally and for PTO efforts to reuse the work product of other patent offices.
In response to data regarding the negligible role of interferences, proponents of the status quo (typically small inventors) raise two reasonable points. First, ordinary patent prosecutions, not interferences, are the arena where a first-to-invent system matters. In the prosecution context, the U.S. system allows an applicant to argue that third-party disclosures of invention that arise after the date of invention, but before the date of filing, should not be used to reject the patent application. As a recent study by Dennis Crouch documents, however, attempts by applicants to assert a pre-filing invention date arise in only 0.1 percent of cases. Moreover, small inventors appear to assert pre-filing invention dates “less often and less successfully than large, publicly traded companies.” Thus, at least in the prosecution context, the first-to-invent system appears to disadvantage precisely those small inventors who are lobbying to keep it.
Second, some critics assert that empirical studies of current interference proceedings cannot necessarily predict what would happen in the counterfactual situation of a shift to a first-to-file system. Perhaps in that context, critics of reform argue, large firms that no longer feared an interference challenge would race to the patent office and file applications, including applications based on information appropriated from small inventors. That scenario seems unlikely, however. The current interference system would hardly seem to present much of a barrier to racing activity on the part of large firms. Additionally, the Senate bill bars those who appropriate information from others from using that information to file patents or even to create prior art.
In addition to fee-setting authority and first-to-file language, the Senate bill also contains improved administrative alternatives to costly and inefficient litigation over the validity of issued patents. Although these improvements may not go as far as one might like, they are certainly a step forward.
The incremental improvements driven by the judicial and executive branches in the last few years must be accompanied by legislation that addresses what only the legislature can do. The Senate bill is thus a worthy and necessary endeavor.
Arti K. Rai is the Elvin R. Latty Professor of Law at Duke Law School. From 2009–2010, she served as the administrator for external affairs at the USPTO. This article represents her views only.
Comments on this article