Thursday, September 18, 2014

THE US PATENT SYSTEM

[When I was a professional economist in the 1970s and 1980s, I did some research on innovation. So a study of the US patent system by Adam Jaffe in 1999 caught my eye. This column was published in Business Standard of 5 October 1999.]


The Expanding US Patent System


For decades, the number of patent applications in the US remained around 60,000-70,000 a year; about a quarter of the applications were rejected, and the number of patents granted hovered between 30,000 and 50,000. The numbers began to decline in the 1970s – the applications gently, patent grants quite drastically. The decline of patenting activity was even more drastic when viewed in relation to population or R&D activity. The number of patents per thousand adults declined from almost 0.4 in the early 1960s to 0.2 in the late 1970s; the number of patents granted to US corporations per million 1992 dollars of research and development fell from over 0.6 to below 0.3. These sobering figures were grounds for widespread fear that the innovative spirit of the Americans was on the wane.
From the 1980s, however, as US growth picked up, so did patenting. The number of patent applications has risen over 100,000 in recent years, and the number of patent grants over 60,000. Though more gently than these absolute figures, the number of patents per adult and the patent productivity of corporate R&D have also risen. What is behind this revival of innovative activity in the US? Has the innovative spirit returned? The answer emerges from the recent work of Adam Jaffe at the National Bureau of Economic Research; apparently, the rise has as much to do with institutional changes as with changes in American creativity.
The US has had a lasting allergy towards monopoly, embodied in anti-trust laws going back to the 1880s. This allergy spread to patents as well in the 1960s; they were considered state-created monopolies, and the administration itself began to undermine the patent system. Thus, almost half the patent applications were rejected. Of those that were granted, a number were contested in courts. Once the district courts ruled on the validity of the patent, their verdict could be contested in appeal courts. Thus, the judicial process diluted the reliability of the patent.
This uncertainty was greatly reduced in the 1980s. In an effort to improve the quality of the judicial system, US Congress created a Court of Appeals for all federal district courts. All appeals on patent cases went to this Court, which greatly improved the consistency of the rulings; it was no longer possible for a litigant to find a court that was likely to be more friendly than another. At the same time, the US Supreme Court itself gave an important signal in its rulings that the monopoly granted by a patent overrode anti-trust laws. As a result, the judicial system today is more likely to uphold a patent.
Another major change is in the attitude of the government towards innovations made out of its funds. Originally, the government believed that its innovations were a common good which should be available free to everyone. But if an innovation required substantial investment to exploit, it was likely that making it available free might actually ensure that it would not be exploited. The patent system is supposed to stimulate the exploitation of innovations by giving a temporary monopoly to the innovator; temporary monopoly encourages the use of government-owned and government-financed innovations no less than of private ones. So in 1980, US Congress passed a law authorizing universities to patent their innovations even if these had been financed from federal funds; over the 1980s, the same right was extended to national laboratories.
Finally, both the US patent office and the Supreme Court have expanded the ambit of inventions that can be patented. Thus biological, genetic and software innovations that would have been unpatentable only a couple of decades ago are now patentable.
It is, of course, true that the proportion of applications in the new areas – biotechnology, genetics, software – has gone up considerably. Data processing and molecular biology, each of which in the late 1970s accounted for less than half a per cent of all patents granted, now account for over 2 per cent each. But it does not entirely account for the rise in the total number; there has been a surge in 70 per cent of the fields.
This suggests that there has been a rise in innovation activity. There are other indications: for instance, more new firms and small firms are now patenting. But there is also the phenomenon of competitive patenting. It has existed in pharmaceuticals for a long time. A drug can often be produced by a number of different processes; many products may show potential as drugs even though they may not be of obvious commercial value. Drug companies have long been in the habit of taking out a lot of related patents even though all are not commercially useful. The idea is to prevent raids in which other drug companies patent innovations which may later prove indispensable in the exploitation of a company’s own patents.
The result is a thicket of patents; one company cannot produce a product without straying into another company’s patents. As long as this dependence is mutual, the companies can come to cross-licensing agreements, or ignore the infringement of each other’s patents. They would not, however, be so tolerant of patent infringements by companies which did not hold extensive patents. Something similar has now happened in computer chip design. There too, companies cannot design chips without impinging on other companies’ patents; so similar cross-licensing agreements and tolerance of patent infringements arise. In this area, therefore, many innovations are patented whose value lies, not in their commercial potential, but in their importance as bastions of corporate defence.
One of the most remarkable increases in patenting is by US universities. They took out less than 200 patents a year in the early 1970s; now the number is touching 2500. This increase would not have taken place without a change in law. But following that change, universities have attracted research funding from industry, which expects the results to be patented or kept confidential. Before this change, university innovations used to be more important; they were more frequently cited in subsequent patents. Now the difference has disappeared. The number of patents granted to national laboratories is not so dramatic; but from being negligible, their number has risen to about 300 a year.

Thus the US has become less ambivalent about patents in the past two decades, and revamped its judicial system for better defence of patent rights. The result has been a considerable expansion of patenting activity, and its extension to new areas and to government-financed institutions that once did not patent at all. The change in domestic attitudes and activity was reflected in the Uruguay round, in which the US vigorously pushed for better protection of patent rights in developing countries, including India. This internal exposure is about to lead to changes in US patent practices: there are proposals to give priority to the date of filing, as in Europe, instead of the date of innovation, and to publish all patent applications 18 months after they are filed, thus making a challenge before grant of patents possible.