Study: Most Important Innovations Are Not Patented

by Stephan Kinsella on November 30, 2013

Patents are often used as indicators for economic and innovative progress.¹ The assumption is that many patents represent innovation, and also that many innovations are patented. Patent records thus correlate with innovation.

A fascinating new paper, “Reassessing patent propensity: evidence from a data-set of R&D awards 1977-2004,” by Roberto Fontana, Alessandro Nuvolari, Hiroshi Shimizu, and Andrea Vezzulli (Department of Economics at the School of Economics and Management (ISEG), Technical University of Lisbon, Working Papers series number 2013/09; Mar. 2013), carefully examines this issue. The authors studied the innovations from the “R&D 100 Awards” competition organized by the journal Research and Development, which, since 1963, “has been awarding this prize to the 100 most technologically significant new products available for sale or licensing in the year preceding the judgments.” I.e., this is a list of important and “technological breakthrough” innovations over almost 3 decades.

The authors searched the patent records for all of these innovations to determine which of them were patented or not, and conclude “a relative low number of important innovations are patented“. In particular, they found that only about 10% of “important” inventions are patented (this number varied a bit based on the industry). This implies that most important innovations are not patented. For most innovations, the innovating companies rely on trade secrecy, lead time/first to market advantages, or other strategies, instead of on the patent system.

One obvious conclusion to be drawn from this study is that patents are not a significant driver of most innovation, if 90% of important inventions are never patented in the first place. Proponents of the patent system often claim that patents are necessary to provide an incentive to innovate; some even (ridiculously) claim that without patents, all innovation would grind to a halt (the truth is the opposite: if patents were made universal and had a perpetual term, all human life would grind to a halt; no offense Galambosians). But even if the 10% of innovations that are patented would never have resulted without the incentives provided by a patent system (an absurd assumption), the great bulk of technological innovations and breakthroughs in modern times would still have come about.

(This paper, by the way, deals with “important” or technological breakthrough innovation. What about more trivial or mundane innovations? There are three possibilities: (a) an even lower percentage of trivial innovations are patented, (b) about the same percentage of trivial innovations are patented as for important innovations, or (c) a higher percentage of trivial innovations are patented. In the latter case, this simply means that a large fraction of the patents actually issued cover junk and trivial innovation.)

None of this is surprising, given the way the patent system is often used in corporate America.² This jibes with the evidence, too. For example, in ch. 9 of Against Intellectual Monopoly, economists Boldrin & Levine take a look at 15 fundamental medical and pharmaceutical discoveries and innovations (as selected by medical doctors and medical journals such as the British Medical Journal). They are:

Penicillin, x rays, tissue culture, ether (anaesthetic), chlorpromazine, public sanitation, germ theory, evidence based medicine, vaccines, the pill, computers, oral rehydration therapy, DNA structure, monoclonal antibody, technology, smoking health risk.

How many entries in this list were patented, or were due to some previous patent, or were obtained during a research project motivated by the desire to obtain a patent? Two: chlorpromazine and the pill. Is this a fluke? We do not think so. In the same issue (freely available on line) of the BMJ you can find references to other similar lists. A particularly interesting one was compiled since 1999-2000 by the US Centers for Disease Control and Prevention (CDC): a top 10 list of public health achievements of the 20th century in the United States. How do medical patents score on this one? Zero. The editor of the BMJ, recognizing the intrinsic arbitrariness of any top-N list, somewhere in the editorial presentation names her three beloved ones among the excluded, “Where are Aspirin, Helicobacter pylori, and Medline?” Good point, and we ask: do they owe anything to patents? Not at all.

So, let’s see. The patent system distorts, skews,³ and hampers innovation and competition, imposes hundreds of billions of dollars of net costs on the US and global economy every year,⁴ and is not necessary to stimulate innovation (in fact, it does not stimulate innovation at all).⁵ People say it is needed to incentivize innovation, though this is clearly false: there has been and would be innovation without and before patent law; people innovate for a variety of reasons; there is no proof that there is any net innovation stimulated by the modern patent system nor that any such net innovation is worth the cost of the patent system.⁶ What patents do is create barriers to competition, increased consumer prices, reduced innovation, and oligopolized industries.⁷

“But wait!” as the informercial carnies will tell you—”there’s more!” Patent law is not just about incentivizing innovation; it’s about inducing people to disclose innovation they would otherwise maintain as a trade secret.⁸ Except … the patent system does not really do this.⁹ Most innovations are disclosed when the inventor starts selling the product embodying the new idea; in many cases the new features are loudly trumpeted in advertising, packaging, and PR. The rare inventions that can be kept secret while the resulting product is sold (e.g. a mundane product like a chemical produced by a unique process) will not be patented and will be maintained as trade secrets, even with a patent system in place.

In other words, a patent system does not encourage innovation, and it does not result in disclosure of innovation. It is simply a dead weight drag on the economy, society, the free market, and human liberty, and a wealth transfer from consumers and small companies and ordinary innovators to the connected state-favored IP grantees.

Update: See Mike Masnick’s post Over 90% Of The Most Innovative Products From The Past Few Decades Were NOT Patented, who notes:

A stunning 91% of all of the technologies receiving the prize were not actually patented. That’s covering approximately 3,000 technologies winning this award as the most innovative advancement of the year over a period of about three decades. What’s interesting to me is that this actually matches very closely with one of my favorite studies on patents, from economist Petra Moser, who looked at historical patenting rates from the 19th century using data on products displayed at the Crystal Palace exhibition of 1851 and the Centennial exhibition in Philadelphia in 1876, which against showed very few of the “economically useful” inventions were patented. Over 80% were not patented. Of course, you might think that back in the 1800s there was less interest in patenting, but this new study suggests a rather similar rate to what Moser found from 150 years ago.

… All in all, this is a really interesting paper and a significant contribution to the discussion over whether or not patents are really a good judge of innovation. It would seem from the data available that the answer is a very loud “no.” In fact, it would appear that very few of the most significant and important innovations are being patented. That should, at the very least, raise considerable questions concerning those who argue that our patent policy is necessary to encourage innovation, or those who argue that numbers from the patent system are a good judge of innovation.