Lies, damn lies and innovation

Nuala Moran

What do patent statistics tell you about innovation? It depends how you count them, says the OECD. Nuala Moran investigates.

At first glance the latest international league tables of patent numbers published by the Organisation for Economic Co-operation and Development (OECD) appear embarrassingly out of date. Although they claim to cover the period up to the end of 2005, the figures for the European Patent Office are for 2002, while those for the US Patent and Trademark Office refer to 2000.

The explanation is that the OECD considers the date an invention was made, rather than the date it was patented, as the key measure of innovative output. Obviously, the OECD is only interested in successful applications, and so it has to wait until patents are granted before it can assess the innovativeness of any year.

As the authors, Helene Dernis and Mosahid Khan, note in the Compendium of Statistics 2005, a major drawback with this approach is “the apparent deterioration of the timelines of patent indicators”.

The authors argue that whereas their statistics reflect actual inventive and creative activity, data from national patent offices merely reflect how hard they have worked. “Patent data reported in annual reports ... are designed to reflect their patenting activity and are primarily for administrative purposes, for example, budget planning.”

Look for the invention, not the patent

But patent statistics can also be an important measure of innovative output, as they reflect the inventive performance of countries, regions, and technologies. And they can be used to track the diffusion of idea across countries. To get at these measures it is more important to know when the invention was made, not the date the patent was granted.

For example, between 1994 and 2000 the number of patents granted by the European Patent Office fell year on year, implying that innovation was in decline. But in the same time frame the number of applications increased, and this is reflected in an increase in patent grants from 2000 onwards.

Similarly, the number of patents granted by the US Patent and Trademark Office increased gently from 1990 to 1997, a period when there was a dramatic increase in the number of patents filed.

The authors argue also that the figures from individual national offices are weak indicators of international technology performance. This is mainly due to the home advantage bias, with applicants are more likely to file in their own countries.

The trouble with counting

Another weakness of looking at single patent offices is that activity is influenced by factors other than technology, such as differing patenting procedures. Even more significantly, straight counting says nothing at all about the value of one patent against another.

To counter this, the OECD developed the concept of triadic patent families. A triadic patent is one granted in the Japan, Europe and the US to protect the same innovation. This eliminates home advantage and the influence of geography. And because it reflects the extra cost and effort taken to file in three jurisdictions, the authors argue it is a reflection that the patents are high value.

The OECD is developing an international statistical infrastructure for patents as the basis for improving the comparability and quality of patent indicators. The aim is to create better tools for policy makers and researchers and provide a more meaningful measure of innovation.