AdobeStock_80648276.jpeg
VoxEU Column Productivity and Innovation

Innovative startup firms and the patent system

Patent protection is assumed to benefit entrepreneurs seeking investment, because patents signal quality and are an asset that can be resold if a startup fails. This column argues that the evidence for these benefits is inconclusive. Notably, patents acquired in a secondary market may be used for rent-seeking, rather than to incentivise innovation.

Is the patent system beneficial for new entrepreneurial firms? At first glance the answer to this question seems obvious, since ideas and inventions are almost the only things these firms have. Therefore, granting property rights to these intangibles would seem to secure the ability for them to grow, safe from the threat that someone else will copy their idea. Patents also create a securable asset in an environment where few assets are available, potentially unlocking external finance. Given the high probability of startup failure, however, the patent must have salvage value – i.e. when there is a secondary market for it. 

In spite of these apparent benefits, academic researchers and entrepreneurs have often criticised the patent system. They have argued that it creates high entry costs, and thus barriers to innovation, that exceed its benefits in some technological areas, notably information and communication technologies. 

In the words of Lemley (2012):

It is worrying that our fastestdeveloping, most innovative industries, the ones that are generating the most new innovations and the most money, by and large hate patents. They view patents as a tax and a cost of innovation, not as a benefit to innovation.” (p. 113)

For a contrasting view, a recent special issue of the Berkeley Journal of Law and Technology, edited by Leih and Teece (2018), argues that, at least in the smartphone sector, a large amount of innovation has gone on despite large amounts of patenting in the sector. 

In a recent survey (Hall 2018), I look more closely at the use of patents as a quality signal, or security, for obtaining finance when a firm has few ways of conveying information to potential investors or to protect them in case of failure. My conclusion is that although they function in this way in some cases, the need for a market in which to monetise the patent assets of failed firms has downsides. This is because those markets make rent extraction possible in a way that does not incentivise innovation. 

Patents as a signal and property right

In the first paper to obtain plausibly causal evidence of the positive impact on new firms of holding patents, Farre-Mensa et al. (2017) find that startups whose patents are granted averaged 55% higher employment growth, and 80% higher sales growth, after five years.  They also find that follow-on innovation by the same firms is of higher quality, and that venture capital funding is more easily obtained if they have been granted patents. 

Previously, Rosemarie Ziedonis and I found that semiconductor managers viewed the main benefit of the patent system as helping startups to obtain venture capital (VC) funding (Ziedonis and Hall 2001). Graham et al. (2009) surveyed entrepreneurs and investors in the biotechnology, medical device, and software sectors, and found that both groups rated financing and achieving a better exit valuation as 'moderately' to 'very' important motives for obtaining patents. This was more so in the fields of biotechnology and medical devices. This is consistent with surveys on the importance of patents in various industries (Cohen et al. 2000). 

Most other studies of patenting and access to finance for startups have found a positive relationship. Unlike Farre-Mensa et al. (2017), none fully differentiates between patents as a signal of a good idea and patents as a property right. 

Disadvantages of patents for new entrants

My finding that few startups patent, even when they are VC-backed, is therefore surprising. The most common reasons for not patenting reported by Graham et al. (2009) were cost and the desire to keep an invention secret. Not having a patentable invention was less frequently given as a reason.

In spite of the positive benefits of patents to exclude imitators and to secure of finance, many entrepreneurs and researchers criticise the way the patent system works for new entrants on two grounds:

  • Licensing cost. In the presence of cumulative innovation, the cost of licensing in the patented technologies on which the startup’s technology builds can be prohibitively expensive.
  • Sale of patents. The existence of a market for patents that so that patents can be sold on firm exit has a negative impact on surviving firms. 

There is anecdotal evidence from interviews with firms and VCs about the first problem (Ito 2005, US FTC 2003, US DOJ and FTC 2007), and some estimates of the cost. But it is difficult to measure the deterrent impact on firm entry of the need to license prior technologies, largely because we cannot observe firms that don't enter a market. In Hall et al. (2017), my co-authors and I try to measure this using comprehensive data on UK firms. We find that our measure of patent 'thickets' in a technology area leads to reduced entry by new firms. This is a suggestive result, but it would be better to have direct evidence.

A secondary market for patent and other IP assets is important to create salvage values for investors in failed startups. If investors might extract some value from patents after a failure, it reduces the cost of capital for the startup. So far, patent markets are in their infancy, or at best at the toddler stage. The main sellers are large operating companies rather than startups (Hall 2018). For patent markets to be a benefit, we must presume that the sale of patents occurs because they protect technology that is useful to other firms. In an ideal world, patent markets would allocate technologies to those best able to make use of them profitably and efficiently. This would enhance overall welfare. But the evidence suggests that these patents are being purchased for reasons that mostly do not involve actually using the protected technology.

There are at least four problematic factors: 

  • Low-quality patents, the assertion of which will do nothing for innovation but which may be too costly to overturn (Lemley and Shapiro 2005).
  • Frequent parallel invention, which implies that the information in the patent was not used by a potential infringer.
  • Bargaining threat points, which lead to extraction of more than the value of the patented invention
  • Low returns to inventors or owners from the intermediaries in such markets, raises the question of how large an innovation incentive created by the secondary market really is. 

All seem to be features of patent markets, as well as of licensing more broadly (with the possible exception of point four). 

Who is buying patents? 

In their survey of patent sales offers by patent brokers and online platforms, Love et al. (2017) find that more than half are sold to patent assertion entities (PAEs) and defensive aggregators.1 In contrast, Serrano and Ziedonis (2017) find that 68% of the patents owned by their sample of failed VC-backed startups in the software, semiconductor, and medical device sectors had been resold five years after failure, mostly to operating companies. Because the sample used by Love et al. is based in Silicon Valley, it is likely to also be dominated by ICT firms. These results suggest that startups are more likely to sell to (potential) users of their technologies when they fail, whereas larger corporations are likely to be pursuing a more defensive strategy. 

The evidence on the overall impact of the patent system on startups, or indeed on innovation as a whole, is inconclusive. Patents seem to help some startups secure finance, but at a cost that discourages many firms from using them. The benefits to startups of the secondary market for patent assets are even more difficult to discern. First, the patent system is just that – a system – and evaluating its impact is a macro-economic question. We do not have alternative universes, other than past history, as a comparison. Second, and related to the weaknesses of the empirical evidence on the functioning of markets for technology, is the lack of transparency in these markets. The information we have tends to be selective, is often based on samples where the sampling methodology is not clear and for which a large number of licensing transactions go unobserved, except by the participants. 

References

Cohen, W M, R R Nelson, and J P Walsh (2000), "Protecting Their Intellectual Assets: Appropriability Conditions and Why Firms Patent or Not?", NBER working paper 7552.

Farre-Mensa, J, D Hegde, and A Ljungqvist (2017), "What Is a Patent Worth? Evidence from the U.S. Patent 'Lottery'", USPTO working paper 2015-05, revised version of "The Bright Side of Patents", published as CEPR discussion paper 11091. 

Graham, S J H, R P Merges, P Samuelson, and T Sichelman (2009), "High Technology Entrepreneurs and the Patent System: Results of the 2008 Berkeley Patent Survey", Berkeley Technology Law Journal 24(4): 1255-1328.

Hall, B H (2018), "Is there a role for patents in the financing of innovative firms?", Industrial and Corporate Change, forthcoming.

Hall, B H, C Helmers, and G von Graevenitz (2015, revised 2017), "Technology Entry in the Presence of Patent Thickets", NBER working paper 21455. 

Ito, J (2005), "One venture capitalist’s view on software patents", blog post, 8 July.

Leih, S and D J Teece (2018), "Introduction: Antitrust, Standard Essential Patents, and the Fallacy of the Anticommons Tragedy: Legal and Industrial Policy Concerns", Berkeley Journal of Law and Technology 32(4): 1301-1312.

Lemley, M A (2012), "The Regulatory Turn in IP", Harvard Journal of Law and Public Policy 36 (1): 109-115. 

Lemley, M A and C Shapiro (2005), "Probabilistic patents", Journal of Economic Perspectives 19(2): 75-98.

Love, B J, K Richardson, E Oliver, and M Costa (2018), "An Empirical Look at the “Brokered” Market for Patents", Missouri Law Review 83: 359.

Serrano, C J and R H Ziedonis (2018), "How Redeployable are Patent Assets? Evidence from Failed Startups", NBER working paper 24526.

Skillicorn, N (2016), How the current patent system actually hurts inventors, Inc.com.

US Federal Trade Commission (2003), To Promote Innovation: The Proper Balance of Competition and Patent Law and Policy.

US Department of Justice and Federal Trade Commission (2007), Antitrust Enforcement and Intellectual Property Rights: Promoting Innovation and Competition.

Ziedonis, R H and B H Hall (2001), "The Effects of Strengthening Patent Rights on Firms Engaged in Cumulative Innovation", in G Libecap (ed.), Entrepreneurial Inputs and Outcomes, Volume 13: New Studies of Entrepreneurship in the United States, Elsevier Science.

Endnotes

[1] A patent assertion entity is a firm or individual whose primary use of a patent is assertion against other firms and individuals who may be using the technology, in the hopes of forcing them to license the patent. A defensive aggregator is a firm or joint venture that obtains patents not for assertion, but to keep them out of the hands of those who might assert them. The owners of or subscribers to the defensive aggregator are typically large ICT firms who fear such threats.

1,470 Reads