Don't attack patent law, but improve it

There is a school of thought among some economists and legal academics who espouse the hypothesis that patent rights, although intended to stimulate innovation, actually threaten it. The hypothesis goes something like this: First, that patents expire 20 years from the date of application. From the time of grant, the patentee has an exclusive right over the subject-matter of the patent. Second, that an exclusive right means what it says, the right to exclude. The patentee can exclude others from any commercial activity over that subject-matter. This exclusive right does not, though, stop others from merely doing research about the invention that is the subject of the main patent. Later inventors can patent any improvements provided they are new and inventive. But because the main patent excludes any commercialisation of the improvement, the reality is that people will not try to make improvements whilst the main patent remains in force.

Until it expires, the main patent is therefore seen as a disincentive to any innovation that comes within its scope. People will be less willing to try to make improvements, particularly if the scope of the main patent is wide, in other words, a big invention. And a further problem arises if there are lots of overlapping patents for minor inventions in the same field, known as “patent thickets.” In a report to the UK government, Prof. Ian Hargreaves asserted that patent thickets ‘obstruct entry to some markets and so impede innovation.’

This hypothesis may sound plausible enough, but is it true? Let’s consider it as a natural scientist would when developing a theory from a hypothesis. Only when a hypothesis is found to accord with observation and experiment can it be treated as valid – as a theory. Economists’ hypotheses can sometimes be tested by observation, but seldom by experiment. So the test for converting an economist’s hypothesis to theory ought to be largely observation.

Does the hold-up hypothesis I have described stand up to observation? Nearly 200 years have passed since the beginning of the industrial age, so there should be some solid supporting evidence by now if it is correct. As there is none, the answer must be a clear ‘no’; there is no credible observation, or evidence that patents threaten innovation or have ever done so.

First, observation of the real commercial world shows that patents provide a major incentive to innovate. Why spend billions on research and development, much of which will be fruitless, if the results of success can be copied by someone without sharing those costs? Ask innovative companies how they fund their research, and the answer is always the same – with the profits from other successful research efforts. The telecom innovator Qualcomm, for instance, invests 20% of its profits in further research. Consider what happened where there was no patent system: Neither the Soviet Union nor Communist China produced a single new class of pharmaceutical compound. Yet since introducing market reforms, China has now built a patent system with amazing success in such a short time.

Second, consider the history of patent abolitionism. Arguments for abolishing patents were loud in the mid-19^th century. The Netherlands, then a backward non-innovative economy, even abolished patents for 35 years. That enabled a Mr. Philips to set up a copycat electric lamp company free from the Edison and Swan patents – but Philips, once it became an innovator in the 20^th century, then emerged as one of the great innovation and patenting companies. Switzerland also had weak patent laws, while the German patent laws of that time did not protect products manufactured abroad, but when they strengthened those laws, the great and ever-innovative Swiss patent-based pharmaceutical industry came into being. As such, I believe that just as those old abolitionists lost, the new abolitionists that exist and the quasi-abolitionists – those trying to water down the patent system, such as competition authorities – should now lose.

Does the evidence of history bear out the hypothesis? I cannot think of a single example to support it. There are several supposed examples, some of almost mythical status. The first is the proposal that the Boulton and Watt patent (for condensing the steam of a steam engine in a separate condenser) held up the development of “strong steam” – using much higher pressure cylinders and not bothering to condense steam at all. Detailed research has shown the story isn’t true. Studies of the sewing machine wars of the mid-1900s likewise show anything but hold-up as rival inventors fought in the courts and in the market: sewing machine technology burst on consumers much as mobile phone technology has done today. Other examples recent scholarship has shown of supposed patent hold-up that have no basis in reality are the supposed hold-up of electric light technology by Edison’s patents, the supposed hold-up of Ford after he lost a patent action in the 1900s and the supposed hold-up of aeroplane development by the Wright brothers’ patents in the First World War. Examination of the facts, including descending into the technology and examining the scope of the patents concerned and finding out what really was happening, has exploded all these myths.

What then of the thicket part of the hypothesis? Again, the evidence doesn’t support it. The best supposed modern example of this is that of mobile phone technology, where in reality innovation is extremely rapid: New models with new features appear so fast that one can hardly keep up. And prices are falling – especially considering the increasing number of features with every new model from competing companies. If you look without any preconception at actual technology, you will find that “thickets” are a sign of fast-moving innovation, quite the reverse of what the hypothesis predicts.

As well as this, the hypothesis ignores the reality of evidence from companies whose business depends on innovation. These companies know that their patents are time-limited, so unless they can innovate their businesses will fail. There is competition in innovation itself – a competition which depends on the patent system to make it possible. You need only ask pharmaceutical companies or telecom companies such as Qualcomm how much their business models and their investment in R&D depend on the patent system to explode the hypothesis.

A final reality check is the fact that patent law provides limited protection for patents developing new uses of old and known medicines. It is inherently unlikely that the first use of a new medicine developed at a huge cost of time and money is the only, or even the best, use of that medicine. Yet because of a lack of patent incentive, there isn’t nearly enough research into new uses of old medicines. Again, this can be examined by asking pharmaceutical companies what they are doing – and in the case of new uses for old medicines, there is a huge reluctance to invest in research. A striking and seriously alarming example is that of the bisphosphonates. Developed for treating loss of bone density, there are good indications that they may have valuable uses in rheumatoid arthritis, certain cancers and heart disease. But no one is following up because they cannot get a useful patent if they do the research.

This is not to say the patent system is perfect; far from it. There are real problems with the proportion of invalid patents granted, with expensive litigation systems (especially that of the U.S. with its jury trial, preposterously over-elaborate discovery system and over-compensatory awards of damages partly fuelled by lawyers on contingency fees taking large proportions of any damages). But only a man blinded to the evidence by his own hypothesis would say that patents threaten innovation. All attention should move away from attacks on the principle behind patent law, and concentrate on improving its machinery.