The implications of the patent system for information and communication technologies (ICT) has been a focus of debate in recent years. One representative example is the smartphone dispute between Apple and Samsung, which led to over 40 international lawsuits over their design and utility patents between 2012 and 2014.
Patents are intended to provide incentives for innovation by granting inventors temporary monopoly rights to profit from their inventions. Therefore, the patent system has the potential to promote new technology development and economic growth. However, ICT products often consist of interlocking inventions. If most of them are protected by intellectual property rights (IPRs), patent thickets could exist: dense, overlapping webs of patents that any company trying to commercialize new technology must navigate. An analysis by Thomson Reuters concluded Apple held 1,298 patents related to its iPhone. This included 416 patents covering general smartphone technologies, 232 patents for user interface, and 149 patents for screen and image display.2
The presence of patent thickets may create challenges for ICT producers. When introducing a new product, a firm must identify patents its product may infringe upon. When these patents are dispersed among many owners, identifying relevant patents can be difficult and licensing negotiations with patent holders can be costly and time-consuming. Alternatively, if the firm does not obtain necessary licenses, it may face litigation that could lead to damages and potential injunction. Firms with an existing patent portfolio can navigate the patent thicket by cross-licensing their own patents with other patent holders. However, this may create incentives to engage in "defensive" or "strategic" patenting to defend against future lawsuits, which itself may increase the size of the patent thicket. Further, this cross-licensing strategy may not be feasible for small start-up firms who do not possess significant patent portfolios, and who cannot easily acquire them.
Some firms have attempted to proactively address the costs of identifying IPRs and obtaining licenses by opening their IPR portfolios to outside firms. IBM donated 500 patents to the "Patent Commons Project" (commonly referred to as "The Commons") in 2005, stating it would not enforce the patents against firms developing open source software (OSS). IBM, Sony, Bosch, and DuPont created the "Eco-Patent Commons" in 2008 to promote environmentally beneficial innovation: 100 patents have been pledged to this commons by 13 firms. Tesla Motors opened all of its electric car patents to other firms in 2014 and pledged not to launch patent lawsuits against anyone who wishes to use the technology protected by these patents "in good faith."
Firms might be willing to allow others to use patented technologies royalty-free to stimulate product growth in complementary markets. IBM's donation to the "The Commons" might spur innovation in OSS that promotes IBM server and service sales. Through opening up its patents to competitors, Tesla might be able to build a common platform to boost the demand for electric vehicles. If that is the case, one would expect the donated patents are valuable and could encourage innovation by outside firms. However, some have argued these types of donations are merely publicity campaigns. If this is the case, then their economic impact will be more limited.
We examined 2,054 start-up software firms to ascertain the economic impact of The Commons on new OSS products between 1999 and 2009.3 In January 2005, IBM pledged access to its 500 patents to any individual, community, or company working on or using OSS. Subsequent to IBM's action, several other firms pledged an additional 29 patents to The Commons. These patents cover technologies in a range of markets such as operating systems, databases, and software development tools. Standard metrics show the quality of the patents in The Commons was similar to those of other patents held by IBM and other firms in similar technology areas, suggesting these patents had real value to IBM. Patent contributions to The Commons stimulated new OSS product introductions by these start-ups for two possible reasons: by using patents in The Commons, start-ups faced lower invention and licensing costs; more importantly, the contribution behavior might signal patent contributors' commitment not to enforce IPRs against OSS firms, reducing litigation threats.
We found contributions to The Commons had a particularly large impact on stimulating new OSS product introductions in technology environments with denser patent thickets. In such environments, innovation builds heavily upon prior discoveries, and the boundaries of potential blocking patents are usually blurry. It is more difficult for start-ups to invent around existing patents and to determine what patents its needs to license. The promise of the patent commitments to The Commons seemed to be particularly valuable for start-up firms operating in these environments.
Standards-setting organizations might aid with patent thickets as well. In ICT industries, products are usually based on compatibility standards. A study suggests a laptop computer embodies 251 standards.1 Besides reducing technical uncertainty, the formal standardization process within standards-setting organizations often reveals the holders of essential patents to standardized technologies. Many standards-setting organizations also require these patent holders to license patents on reasonable and non-discriminatory terms. Therefore, formal standardization might decrease IPR uncertainty by identifying relevant patent holders and clarifying licensing terms.
Formal standardization can also increase legal risks. Incorporating technology into a compatibility standard makes it more difficult to use alternative technologies, so standards-essential patent holders can request large licensing fees. For example, Rambus did not disclose its patents in the JEDEC standard for memory chips until the standard was widely adopted. This behavior can increase the costs of using the standardized technologies.
We studied formal standards developed by the Internet Engineering Task Force (IETF) between 1994 and 2004, a period of rapid growth in the commercialization of the Internet.4 Standards-using firms were less likely to patent defensively in highly standardized areas. This suggests formal standards might reduce IPR uncertainty within this setting. However, IPR disclosure and licensing rules in use at standards-setting organizations vary widely, and changes in these rules can influence the outcome of standardization.
Although patents create incentives for firms to conduct innovation, patent thickets can create significant costs for ICT producers, particularly for start-up firms. Patent commons can reduce some of the costs of patent thickets by making certain technologies available to use royalty-free, and by signaling that patent holders will not enforce IPRs against certain types of uses. Standards-setting organizations may help to navigate patent thickets in a different way, by clarifying patent holders and reducing uncertainty surrounding licensing terms. The effects of both of these institutions will depend on their rules. For example, technology users will be able to navigate patent thickets more easily when standards-setting organizations set clear IPR disclosure and licensing rules that reduce frictions between standards-essential patent holders and users. ICT producers will need to understand these rules thoroughly to understand the implications of these institutions.
1. Biddle, B., White, A., and Woods, S. How Many Standards in a Laptop? (And Other Empirical Questions), 2010; http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1619440.
2. Thomson Reuters. Inside the iPhone Patent Portfolio: Data Offers Glimpse of Battery-Free Future, Insight into Smartphone 'Patent Wars', 2012; http://ip-science.thomsonreuters.com/m/pdfs/iphone-report.pdf.
4. Wen, W., Forman, C., and Jarvenpaa, S. Standards, intellectual property rights, and strategic patenting: Evidence from the IETF, 2015; http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2709645.
The Digital Library is published by the Association for Computing Machinery. Copyright © 2016 ACM, Inc.
I was very happy to see an article with this viewpoint published. This article, however, mentions some points I believe should be further stressed out as problematic and important. Namely, still at the introduction, after mentioning that patents are intended to provide incentives for innovation by granting to inventors temporary monopoly rights, the next paragraph continues, The presence of patent thickets may create challenges for ICT producers. When introducing a new product, a firm must identify patents its product may infringe upon.
The authors continue explaining the needed process But this simple statement should be enough to explain how the patent system is broken and needs repair.
A requisite for patenting an invention was originally the inventive and non-obvious characteristics. Anything worth being granted a patent should be inventive enough, it should be non-obvious to an expert in the field.
When we see huge bodies of awarded (and upheld) patents falling in the case the authors mention, it becomes clear that the patent applications were not thoroughly researched prior to their patent grant. Sadly, long gone are the days where the United States Patent and Trademarks Office employed minds such as Albert Einstein's; nowadays, the office is more a rubber-stamping bureaucracy where most patents are awarded, and this very important requisite is left open to litigation: If somebody is found in breach of a patent, they might choose to defend the issue that the patent was obvious to an expert. But, of course, that will probably cost more in legal fees than settling for an agreement with the patent holder.
The fact that in our line of work we must take care to search for patents before releasing any work speaks a lot about the process. Patents are too easily granted. They should be way stricter; the occurence of an independent developer mistakenly (and innocently!) breaching a patent should be most unlikely, as patents should only be awarded to truly non-obvious solutions.
Displaying 1 comment