Sunday 10 October 2010

Patent Data and Innovation: Once Again, So Much and Yet So Little

Can there be a body of IP data that is more analyzed than patent filings and registations? The avalability of this data, and the large number of information fields that can be examined, seem to have made patents an irresistible source of attraction for quantitative analysis. More challenging, however, is the drawing of meaningful conclusions from this wealth of information.

A good example of this gap between the qualitative abundancy and qualitative
scarcity in analyzing patent data was underscored in the article that appeared in the October 2 issue of The Economist entitled "Trading Places: Innovation in Asia." Permit me to summarize points made in the article (more or less in the order of the article itself). The starting point is the statement that "[p]atents are a crude but useful measure of innovation". Based on the opaque disclaimer, of sorts, the article notes as follows:
1. Japan's dominance of Asian technology has been eroded by its neighbours. Consider, for example, the provenance of most of the major components of the Apple IPad (South Korea and Taiwan), but nearly none from Japan.
2. Between 2006-2009, patent filings in Japan have notably declined while those in China have soared. Based on the filing trends of these past years, filings in China may surpass those in Japan in 2010, "putting China in striking distance of America."
3. In 2008-2009, the Japanese filed 11% fewer PCT applications, while the Chinese filed 18% more. That said, the Japanese have a much better success rate in granted patents and enjoy a higher patent citation rate.
4. While Japanese companies are cutting back on R&D, Chinese companies are accelerating their R&D activities. Indeed, China may soon surpass Japan in domestic-spending R&D, "on purchasing-power terms."
5. China is not alone on this. Moving to South Korea, Samsung plans to double its research spending this year. Not by accident, perhaps, Samsung enjoyed profits last year that were greater than the largest nine Japanese electronic firms combined.
6. This push by China in the patent arena is attributed in large part to governmental policy. For example, China wants to wean itself off reliance on foreign patents (Chinese companies pay more than $10 billion to foreign in companies in licensing fees annually). Increasing Chinese patents will (i) avoid some of the need to pay royalties to foreigners: (ii) force foreign companies to take a license of Chinese technology and (iii) improve the position of Chinese companies in negotiating licences.
7. Japan is woefully behind other countries in patents that list a foreign co-inventor (4% for Japanese applications, compared with 40% for American filings).
8. Japan still has the most patents in force (1.9 million, compared to 1.4 million for America and 134,000 for China).
So what are we to make of the article?

First, the article does not really address the connection between patents and innovation. Is it really the case that increased patent filings in China and by Chinese points to increased innovation? Maybe yes, maybe no--the article does not really explain. In this regard, it is curious for The Economist to focus so much on the point given that the cover story of the same issue, entitled, "How India's Growth Will Outpace China's," seems to suggest that, over the long run, India will be a better bet than China with respect to innovation. The place of India in connection with Asian innovation, however, is not addressed in the article.

Second, why the scant attention to South Korea in the article? Given the so-called "lost decade" in Japan, data about patent doom and gloom is not very surprising. The more interesting comparison would be between China and South Korea. Alas, on this, there is nary a word.

Third, the observation that licensing fees are seemingly a major reason for the Chinese government push into patents is quite amazing, if true. Patents are filed for a variety of reasons, of which patent royalties (and the ability to extract cross-licences) is only one of them. If patent licensing has been identified as a major driver of patent activity in China, this seems to me to be a big deal. Sadly, the point is not pursued.

What stands out in the article is yet another effort to gather diverse bits and pieces of patent filing and registration information without either bringing them together into a coherent narrative or analyzing the points in any depth. That is too bad.


Unknown said...


As you and I know well, patent filings mean nothing unless they are coupled with an analysis of the attendant business strategy. Moreover, patent filings have little to do with innovation--innovation relates to things that get to the marketplace and make money, and need not be patented. If only 3% of patents make money (and it is probably about this number), then it can be said that patent filing information is only 3% directional in relation to innovation.

So, looking at patent filing data in the abstract is useless information. Unfortunately, few others recognize this, and there will be much groaning and gnashing of teeth over this "deep analysis" from the respected press outlet of the Economist. And, people will respond as they always do--by making erroneous conclusions about the state of innovation in Asia and elsewhere.

Maximiliano Marzetti, PhD said...


An excellent analysis and healthy exercise of critical thinking. Let's-jump-to-conclusions-that-may-attract-more-readers (or appease current ones) seems to be the editorial line of some respected sacred cows of economic journalism.

The lack of causal relationship between more patenting and more innovation is plainly bad argumentation (the old and still memorable "post hoc ergo propter hoc").

Serious research on IPR and innovation is badly needed. More than 50 years after Machlup's study we continue to "muddle through".


Kate Shore said...

The real issue may be that patent data is the best we can do... the data is relatively uniform across jusridictions, widely translate, easy to search, free, etc.

What other sources of data and information do you propose we use to measure innovation?