Strategic use of patents: The case of patent trolls Pénin Julien BETA Université de Strasbourg penin@unistra.fr DIMETIC Lecture March, 2010
Overview Patents as strategic instruments Much more than mere tools to exclude Patents, market for technologies and the raise of fabless firms The problem of patent trolls The dark side of market for technologies Elements of solution to limit patent trolling
Patents: practical details (1) A patent provides a monopoly over a technical invention on a given territory and for a maximum of 20 years (after the first application) (TRIPs agreement) A patent is a right to exclude (not to use) Prevent others from producing, selling and using the invention To be patentable, an invention must be new (not yet publicly available), non-obvious and must have an industrial application Application must be addressed to patent offices (EPO, USPTO, OMPI), which examine them. No European or world patent (only centralized procedures, EPO, PCT) Priority rule 18 months after the application, a description of the invention is disclosed (secrecy is broken)
USA Patents: practical details (2) Two main patent systems: USA / rest of the world - First to invent - 1 year grace period - Usefulness criteria Europe (ROW) - First to file - Invention must be new - Industrial application - Moral obligation to patent - Automatic disclosure after 18 - No publication obligation months - Disclose the «Best way» - Disclose one way to use tyhe - Reexamination invention - No first user right - Opposition - Geographic homogeneity - First user right - No geographic homogeneity
The traditional view of patents: Offensive strategy to protect a monopoly Traditional «arrovian» view: The incentives-diffusion dilemma To sum up, we expect a free enterprise economy to under invest in invention and research (as compared with an ideal) because it is risky, because the product can be appropriated only to a limited extent and because of increasing returns in use Further, to the extent that a firm succeeds in engrossing the economic value of its inventive activity, there will be an under utilization of that information as compared with an ideal allocation. Arrow (1962) Patent as a second best solution to this dilemma Enable appropriation and increase incentives (grants a monopoly) Favor diffusion of the underlying knowledge Monopoly deadweight loss The case of pharmaceuticals
Beyond the traditional view: A multiplicity of patenting strategies 1) Offensive strategy: to protect a monopoly (exclusion) 2) Market strategy: to trade technologies 3) Defensive strategy: cross-licensing 4) Reputation strategy: to certify and signal competences 5) Partnership strategy: to collaborate 6) Open strategy: to diffuse or free technologies Others: Blocking strategy, internal strategy (patent to manage employees), etc.
Patents: much more than tools of exclusion A wide spectrum of use Enable appropriation Prevent appropriation Strict appropriation Ex: pharmaceuticals Bargaining chips Ex: electronics Control the diffusion of the innovation Ex: network industries Grant licences more or less exclusive Ex: chemicals Reputation, image, signalling Free technologies Ex: biotech RT Many different possible role for a patent Sectoral specificities
Example: The market strategy Patents to sell technologies Inventions are not always intended to be used by the inventor Independent inventors Non key sectors Invention from the academia New market for the inventor (new country) Invention can be best valorized elsewhere The era of open innovation (Chesbrough, 2003) Rembrandts in the Attic (Rivette and Kline, 2000) A patent enables inventors to sell their inventions and therefore to directly make profits from it Chemical industry (licensing > R&D investments)
Patents and market for technologies The patent system eases the implementation of a market for technologies It enables to solve the Arrow paradox (1962) Ex: Kearns and the intermittent windshield wiper (Tirole, 2003) A patent has two properties: disclosure and exclusion Signal increases the visibility Protection sustains the signal Without patents it would therefore be very difficult to trade technologies (not impossible but difficult) The case of tacit knowledge (sticky vs. leaky) Patents and inter-firms collaborations
Patents, market for technologies and specialization
Patents, market for technologies and specialization Patents enables labour division and vertical specialization Specialization on core competences Example: Pharmaceuticals and the biotech paradigm In 2003, 40% of new market molecules were coming from biotech firms Random vs. rational drug screening NBF Patent License Big-Pharma Upstream research Development, clinical tests, marketing and selling
Emergence of new industrial organizations: The raise of «Fabless» firms Fabless = without fabrication (not virtual firms but firms specialized in knowledge production) Life science industries: start-ups and spin-offs Electronics: «Fabless» and «foundries» Those fabless firms strongly rely on intellectual capital Patents (and other IPR) play a central role since they give a legal existence to this capital
Market for technologies imply specialization and (generally) efficiency Yet, there is another side of market for technologies (the dark one): The emergence of «patent trolls»
Definition: Patent Troll A patent troll is a firm (or an individual) whose business model consists of generating revenue with its patent portfolio by relying on IPR litigation. Before threatening its victims (who most of the time are manufacturing firms), the troll operates in order to place them in position of Hold-up. Far from the usual industry business model, which is to use patent in order to prevent infringement, Trolls business model is therefore to be infringed (Reitzig et al., 2007) The Troll is looking for infringement
How to recognize a troll once you meet one? A troll is characterized by some specific attributes: (1) It has no tangible production (it is a fabless firm) (2) It does not invest in R&D and has a significant litigation department (lawyers) (3) It has a significant activity of patent buyout (4) It does not look for ex-ante licensing agreement but tries to enter into litigation as late as possible, once the victim has already made irreversible investments (Henkel and Reitzig, 2007; Lemley and Shapiro, 2007) Similarities and differences between technological firms and trolls / Both are fabless and use patents to generate direct revenues
Trolls business model: money extortion in 5 steps 1) Operate in sectors where technology is complex (like electronics) 2) Buy patents (to firms with financial difficulties, start-ups, universities, etc.) in order to gather a large patent portfolio 3) Identify infringers (which requires intelligence abilities) 4) Wait, before litigating, that the infringer has made sunk investments 5) Threat to stop the production and, at the end, compromise in order to obtain a compensation (which can be high or low according to the situation) (Remarks: defensive patents are inefficient with respect to trolls)
A famous case of trolling : RIM and the Blackberry In 2001, NTP accuses RIM, the firm which develops and commercializes Blackberry, of infringement In 2003 a US judge speaks in favor of NTP RIM must stop producing and selling Blackberry until 2012, when NTP patents expire (to much power to NTP?) The decision is confirmed in appeal In march 2006 RIM and NTP pass an agreement, RIM agreeing to pay NTP 612,5 millions of dollars to close the case (much more than the intrinsic value of NTP s technology) In between, the USPTO had released a report in which they considered NTP patent as invalid (it should not have been delivered) (Competences of USPTO and US judges? Too many low quality patents)
The issue of the quality of patents held by Trolls Patents held by trolls are often considered of bad quality (i.e. should not have been delivered / not new not novel) Trolls benefit from the imperfections of patents office (USPTO) It is right that low quality patents are compatible with trolls BM Yet new empirical insights (Fischer and Henkel, 2009) Complex technology Wide patents (high number of claims) High quality (measured by the number of backward citations) Conclusion: If compatible with their BM, low quality is not a constitutive characteristic of patent trolls Trolling is robust to improvements of patents attribution process
Patent trolls and social welfare In an ideal world (no uncertainty, perfect information, etc.), trolls are patent brokers They contribute to the good functioning of markets for technology They complement and push back technological firms (Credible threat of litigation for manufacturing firms) One step forward in the division of labor: technological firms invest in R&D and trolls specialize in technology transfer McDonough proposes therefore to replace the word patent troll by patent dealer The activity of patent dealers in their pure form benefits society (McDonough, 2006, p. 204)
Patent trolls and social welfare (2) In our world (far from ideal) Patent trolls are not mere brokers Uncertainty and information asymmetry decrease trolls desirability Furthermore, trolls are not technology brokers but patent brokers, which is much less valuable for society (they do not transfer knowledge but legal rights) At the end, patent trolls introduce a new tax to innovators (Jaffe and Lerner, 2004) They are only value destructor (Henkel and Reitzig, 2007).
Policy measures to limit the scope of patent trolling Trolling is linked to three main factors: Patents quality (and number) The structure of patent trials The possibility to move in order to place other firms in situation of hold-up Policy measures can, to a large extent, correct the first two factors. Yet, the third factor is largely robust to policy changes (Henkel and Reitzig, 2007)
Patents quality Current situation feeds patent trolling Too many bad quality patents (Jaffe and Lerner, 2004) Required changes: Increase patentability standards (or at least respect current ones: novelty and inventive step) Change the incentive structure of patent examiners: Harhoff (2009): The first mission of patent examiner should be to refuse patents not to grant them. They are the guardians of the temple Increase the cost of patent application? Open the examination process to the crowd? (Jaffe & Lerner, 2004) Limit: trolling is compatible with high quality standards (Fischer and Henkel, Pénin 2009) Julien Dimetic Lecture,
The structure of patent trials Currently, scenario is favorable to trolling Required changes: Probability to win the trial Cursor too far in favor of patent holder (CAFC in the US) But is it possible for policy makers to change this? Indemnification rules make it more profitable to litigate than to grant licenses ex-ante. Undue enrichment (Ger) / Royalties equal to the average industry rate / The payment of the overall costs of the trial (USA) Preliminary injunction Not justified in case of fabless firms Patent Reform Act in the US (2007)
Hold-up How to prevent firms to act in order to provoke hold-up (they are looking for litigation and delay the time of the attack)? Improve patent information Submarines patents (USA) The rule of 18 months before disclosure Compulsory licensing (very delicate topic) Yet, trolling is likely to remain viable (Reitzig et al., 2007) [Despite State intervention] it may still be more profitable to trap R&D manufacturers than to approach the latter for licensing fees in the first place (Henkel and Reitzig, 2007, p. 23).
Conclusion: How to limit patent trolling without damaging also technological firms? «Trolls» and technological firms are both «fabless» They both rely on a strong patent system Yet, limiting trolling activities implies somehow to weakening the patent system for fabless firms (ex: the US Patent Reform Act) But, technological firms are a critical features of a knowledge-based economy. Tradeoff: Find the right balance It may be necessary to tolerate a certain level of trolling in order to sustain the raise of technological firms! Work in progress