Monday 22 April 2013

Rich relation, poor relation, correlation

Having not-quite-so-recently received my copy of the March/April issue of The Patent Lawyer, I rediscovered a piece which I had written for it on patent statistics and what, if anything, they mean, having been slightly inspired by a playful piece of writing by Swiss IP lawyer Mark Schweizer that was hosted on the IPKat, here.  Looking over it, I though I may as well share it with readers of the IP Finance weblog too, since it will surely interest, amuse or irritate some of them.  Here it is:

Rich relation, poor relation, correlation 
Perhaps because he had nothing better to do, or possibly because he suspected something, multi-talented Swiss lawyer and judge Mark Schweizer played around with some official statistics to see what conclusions they might yield when he matched them against each other.  First he took the per-capita rate at which patent applications were filed in 19 developed countries (18 European Patent Organisation members plus the United States). He then plotted the patent filing rate against the ratio of public debt to gross domestic product (GDP ) in each of those countries.  The result? While there were a couple of outliers, of which the most significant was patent-rich, low-debt Switzerland, the general pattern was clear: countries in which public debt represented a lower proportion of GDP were also those in which the number of patents filed per capita was higher. 
Now, one might seek to explain these figures in some rational manner, as some commentators have done so; one might deny the validity of this exercise by sticking to the point that these figures represent correlation, not causation: one may as well argue that a low level of public debt is caused by a high level of patent filing activity as the other way round.  In any event, no-one disagrees that Schweizer’s figures are a snapshot, a view which is based on two sets of available figures, and that we cannot without further ado establish that we are staring at a universal truth or even a short-term trend, rather than a one-off coincidence. 
Schweizer’s exercise, which was never intended as anything other than a good-natured and playful prod at the patent community, is however important.  Within other areas of intellectual property law, notably copyright, arguments in favour of reform have increasingly been subjected to the requirement that they be “evidence-based” .  The standard of proof is not that of the courts but that of a cross-disciplinary mix of economics, social psychology, anthropology and business studies (but not moral philosophy).  However, the burden of proof lies with he who would reform the existing system. 
In contrast, while there are studies a-plenty on the operation of the patent system and its impact in different industrial sectors, economies and time-frames, the momentum towards reform in the US, Europe and even China appears to have been the result of a political push.  There may be little or no harm in this, but every patent system ultimately stands or falls on the basis of its performance.  This need to appraise a patent system’s performance is something that we are often poor at recognising, since for so many of us the existence of a patent system is axiomatic and its form is so closely prescribed by international law. We are also ill-equipped to carry out the sort of appraisal the patent system needs since, with differing economic, social, and cultural goals, we cannot easily agree the criteria by which we must assess it. 
This is where the Schweizers of this world come in. They encourage us through their example to take publicly available data and use it as a way of reflecting how the patent system performs relative to other known data. Pegging patent applications to public debt may be no more meaningful than correlating them to the speed at which patents are granted, the cost of litigation, the availability of investment funds or credit, the navigability of corporate red-tape, access to technical data and assistance, sightings of unidentified flying objects  or the national consumption of beer. However, patent applications, grants, renewals and expiries are all more meaningful when pegged to something, preferably repeating the exercise at regular intervals in order to build up a set of impressions that have some trajectory through time as well as space.
Once we have our data, our correlations, our causations and an image of the patent system which is historical, current and, to the extent that it can be projected, future-facing, we can either use it in order to make better decisions as to how the patent system works or, if the mood takes us, we can continue to run the system by taking decisions that are essentially political in nature but have a better idea as to what might happen if we do.

No comments: