Measuring Innovation: A New Index Of Patent Quality

01 Apr 2004

Economic researchers have developed a new way to measure innovation based on the quality of patents. Jean Lanjouw and Mark Schankerman have used data on more than 100,000 US patents produced in seven technology areas to construct their index. Their study, published in the April Economic Journal, shows that patent quality is strongly associated with variations in the stock market value of firms, especially in pharmaceuticals and other health sectors. This suggests that the patent quality index may be useful for understanding how the stock market values patents, and for evaluating bundles of patents for cross-licensing and patent pooling arrangements.

The index also reveals that average patent quality has increased over time and that this accounts for a significant part of the apparent decline over time in research productivity. Research productivity, as measured by the number of patents per dollar of R&D, has declined sharply over the last 40 years in the United States and elsewhere. By 1990, the number of patents produced per US scientist or engineer had fallen to just 55% of its 1970 level, with even steeper declines in Europe. These statistics have attracted attention because of concern about the apparent slowdown in total factor productivity.

This fall in research productivity could be due to expanding markets. Market growth increases the profitability of R&D. In response, more investment is directed toward R&D. Since the extra dollars are invested in more ''marginal'' projects, this would naturally lead to falling productivity. But earlier studies have found that markets have not grown fast enough to explain the observed fall in the ratio of patents to R&D inputs.

The evidence of declining research productivity raises the spectre of ''technological exhaustion'' – getting less inventive output for a given level of R&D investment. This would lower innovative output directly and, by reducing the private payoff to R&D, it would also reduce private R&D investment and undermine productivity growth.
Is this threat real? Are we really getting less innovation for our R&D investments? This research uses extensive data on US firms and detailed information about the patents they hold to show that there is no evidence of technological exhaustion.

Changes in the number of patents per R&D reflect two very different things: the number of patents per invention, and the number of inventions per R&D. A fall in research productivity, as it is typically measured, may be real – a declining invention/R&D ratio – or only apparent – a declining patent/invention ratio. Since we do not have information on the number of inventions, what appears to be technological exhaustion may be a consequence of bad measurement. If inventors make less use of the patent system, or if the average value of a patent is increasing, growth in the number of patents understates true growth in the amount of innovation. To identify technological exhaustion, these measurement issues need to be addressed.

To do so, the researchers develop a composite index of patent quality using four different variables: the number of claims specified in the patent, the number of subsequent citations to the patent, the number of citations the patent makes to previous patents and the number of countries in which the patent is applied for.

Using data on more than 100,000 US patents during 1975-93 in seven technology areas – drugs, biotechnology, other health, chemicals, electronics, computers and communications, and mechanical – they construct an estimate of quality for each patent and show that this composite index greatly improves our ability to identify patent quality.

The index shows that rising patent quality over time accounts for a significant part of the apparent decline over time in research productivity. But it varies by technology area. For example, in electronics and medical instruments, the patents/R&D ratio first fell sharply until the late 1980s and then recovered, ending 13% and 20% higher, respectively. But adjusting for changes in quality, the increase in productivity over the period was more than twice as large. In chemicals, the patents/R&D ratio fell by 20%, but adjusting for patent quality, the decline is only 7%. In pharmaceuticals, there was a very sharp rise in R&D spending and a precipitous decline in R&D productivity, by about 50%, and adjusting for quality makes almost no difference in this sector.

In addition, as predicted by economic theory, R&D productivity at the firm level is negatively related to the level of market demand and patent quality. Importantly, however, there is no evidence of technological exhaustion at the firm level: there is no evidence of negative time trends in research productivity, given the level of R&D.

Contrary to expectations, R&D productivity over time is not strongly (negatively) related to the quality of patents held by a firm. But there is a strong association between the stock market valuation of firms and the average quality of the patents they hold. This relationship is particularly strong in pharmaceuticals and medical instruments.

The researchers estimate that increasing the average patent quality of a firm from the 50th percentile (median) to the 95% percentile would raise market value by 20% for firms in pharmaceuticals and medical instruments, 7.3% in electronics and 3.9% in mechanical sectors.

''Patent Quality and Research Productivity: Measuring Innovation with Multiple Indicators'' by Jean Lanjouw and Mark Schankerman is published in the April 2004 issue of the Economic Journal. Lanjouw is in the ARE Department at the University of California, Berkeley and the Brookings Institution; Schankerman is at the London School of Economics.

Mark Schankerman

020-7955-7518 | M.Schankerman@lse.ac.uk