by A Della Malva · 2010 · Cited by 26 — zwei Akteure darum wetteifern, eine Innovation als Erster zu vollenden, um sie dann zu patentieren, stellt defensives Publizieren eine attraktive Strategie

90 KB – 46 Pages

PAGE – 3 ============
Dis cus si on Paper No. Corporate Science in the Patent System: An Analysis of the Semiconductor TechnologyAntonio Della Malva and Katrin HussingerDie Dis cus si on Pape rs die nen einer mög lichst schnel len Ver brei tung von neue ren For schungs arbei ten des ZEW. Die Bei trä ge lie gen in allei ni ger Ver ant wor tung der Auto ren und stel len nicht not wen di ger wei se die Mei nung des ZEW dar. Dis cus si on Papers are inten ded to make results of ZEW research prompt ly avai la ble to other eco no mists in order to encou ra ge dis cus si on and sug gesti ons for revi si ons. The aut hors are sole ly respon si ble for the con tents which do not neces sa ri ly repre sent the opi ni on of the ZEW. Download this ZEW Discussion Paper from our ftp server:

PAGE – 4 ============
Non-technical Summary Corporate scientific publications are often present ed as a strategic means for firms to create prior a rt with the objective to prevent rivals from patenting related inventions. This defensive disclosure strategy has attracted the attention of academic sc holars in the field of law, economics and management science. The academic literature has dev eloped different answers to the question why firms engage in defensive publishing at the costs o f forgoing the right to exclude third parties from using the invention by applying for a patent. One s tream of the literature, mainly put forward by economic scholars, analyzes defensive publishing in the context of patent race models. In these models, the laggard in the race has incentives to p ublish preliminary results that are not patentable in order to prolong the patent race. Scholars in law a nd management science consider open disclosure as a complementary strategy to patenting. The latter p erspective finds partial confirmation in interviews with patent experts in the corporate sector, while the patent race model is largely rejected by practitioners. Both streams of the literature assum e that corporate publications enter the pool of pri or art which is relevant to judge the novelty of paten t applications at the patent office and that they h ave the power to block or hinder patent applications. T his proposed mechanism behind corporate publications as a means to preempt patents on relat ed technologies so far lacked empirical evidence. This paper provides a first (large-scale) analysis of the effectiveness of corporate scientific publications regarding their power to block rivals™ patents. With focus on the semiconductor technology we show that scientific publications by corporations challenge the novelty and the inventive step of patent applications at the Europe an Patent Office (EPO) significantly more than other pieces of prior art. Hence, we show that the mechanism behind the concept of defensive publishing works. Further, detailed information fro m the EPO patent examination procedure allows us to show that corporate publications threaten the no velty of patent applications in combination with other pieces of prior art, like patents, rather tha n as standalone documents. This supports the view o f practitioners and scholars in law and management sc ience who argue that corporate scientific publishing can be an effective means for firms to p rotect their freedom to operate if used a complementary part of a firms™ overall intellectual property protection strategy.

PAGE – 5 ============
Das Wichtigste in Kürze Wenn Unternehmen technische Informationen außerhalb des Patentsystems publizieren und damit darauf abzielen, Patente für verwandte Techno logien zu verhindern, wird das in der einschlägigen Literatur als fidefensives Publizieren fl bezeichnet. Die wissenschaftliche Literatur gibt verschiedene Antworten auf die Frage , warum sich Firmen entscheiden, Erfindungen zu publizieren statt diese zu patentier en und damit ein temporäres Monopolrecht auf die Erfindung erhalten. Eine Erklärung beruht a uf der Analyse des defensiven Publizierens im Kontext von Patentrennen. In solche n Modellen, in denen typischerweise zwei Akteure darum wetteifern, eine Innovation als Erster zu vollenden, um sie dann zu patentieren, stellt defensives Publizieren eine att raktive Strategie für den im Rennen zurückliegenden dar. Wenn dieser seine vorläufigen Fortschritte publiziert, kann der vorneliegende Wettbewerber seine Erfindung nicht al s Patent anmelden, da sie nicht mehr innovativ genug ist, und damit verlängert sich das Patentrennen. Eine alternative Erklärung sieht defensives Publizieren als komplementäre Stra tegie zum Patentieren. Diese Erklärung findet Bestätigung durch Patentexperten aus der Pri vatwirtschaft sowie durch Wissenschaftler der Rechtswissenschaften und der Betriebswirtschaft slehre. Beide Erklärungen für defensives Publizieren postulieren, dass Publikationen relevan t genug sind, um bei der Beurteilung eine Rolle spielen, und dass sie zudem die Macht haben, Patentanmeldungen zu blockieren. Diese Studie unternimmt die erste empirische Analys e, um den postulierten Mechanismus zu untersucht. Am Beispiel von Patentanmeldungen in de r Halbleiterindustrie zeigen wir, dass wissenschaftliche Publikationen von Unternehmen daz u in der Lage sind, Patentanmeldungen am Europäischen Patentamt (EPA) zu behindern. Anhan d detaillierter Informationen des EPAs zeigen wir weiterhin, dass Publikationen aus d er Privatwirtschaft insbesondere in Kombination mit anderen technischen Dokumenten, wie beispielsweise anderen Patenten, Patentanmeldungen blockieren. Dieses Ergebnis stütz t die Hypothese, dass das Publizieren von Erfindungen als komplementäre Strategie zum Pat entieren Teil einer effektiven Strategie zum Schutz geistigen Eigentums darstellen kann.

PAGE – 6 ============
Corporate Science in the Patent System: An Analysis of the Semiconductor Technology * Antonio Della Malva a and Katrin Hussinger a,b,c aMaastricht University, School of Business and Econo mics, Department of Organization & Strategy (The Netherl ands) b K.U. Leuven, Dept. of Managerial Economics, Strateg y and Innovation (Belgium) c Centre for European Economic Research (ZEW), Mannhe im (Germany) November 2010 Abstract Corporate scientific publications are often present ed as a strategic means for firms to create prior art with the objective to prevent others from patenting related inventions. This presumes that co rporate publications enter the pool of prior art which is relevant to ju dge the novelty of patent applications at the patent office and that corporat e science has the power to block patent applications. This paper analyses for the first time whether the presumed mechanism behind corporate publication s as a means to pre- empt patents works. With focus on the semiconductor technology our results show that scientific publications by corpor ations challenge the novelty of patent applications at the European Pate nt Office (EPO) significantly more than other pieces of prior art. Detailed information from the EPO patent examination procedure allows us to s how that corporate publications threaten the novelty of patent applica tions if combined with other pieces of prior art like patents (rather than as standalone documents). This supports the view that corporate scientific pu blishing can be an effective means for firms to protect their freedom to operate if used as a complementary part of a firms™ overall IP protectio n strategy as proposed by scholars in the field of law and management scie nce. Keywords: defensive corporate publishing; freedom to operate ; patents; semiconductor JEL-Classification: O34, O32 * We thank Michele Pezzoni granting us access to his text search algorithm and subsequent fitechnicalfl support, Fleur van Dullemen for her precious assist ance during the data gathering process and Thorsten Doherr for granting us access to his search engine as well as for providing us with an update of the patent application outcomes. For helpful comments w e thank Dirk Czarnitzki, Bronwyn Hall, Cedric Schneider, Andy Toole and Brian Wright as well as t he participants of the CISS summer school in Turkey, 2010, the Brown Bag Seminar at the KU Leuve n, 2010, the EPIP Conference, 2010, and the Innovation Seminar at the UC Berkeley, 2010.

PAGE – 8 ============
2 development. In these models, the loser typically r eceives no or little return to her investment, while the winner gets all (e.g. Reinganum, 1982). Defensive publishing is introduced into these models by allowing for the option to disclose knowledge outside the patent system. In such a scenario, the laggard has incentives to disc lose her findings before the leader applies for a patent in order to expand the state of the ar t and, therewith, increase the patentability requirements for the rival and prolong the patent r ace. The patent race explanation for defensive publishin g has raised criticisms on both legal and strategic grounds (Eisenberg, 2000). With regar d to the U.S. patent legislation, Eisenberg (2000) argues that the first-to-invent rule 2 and the statutory bars of the U.S. patent doctrine 3, which define prior art as of one year prior to the application date, limit the effectiveness of preemptive publications. From a strategic point of view, Eisenberg (2000) criticizes that the disclosure of relevant information outside the pate nt system might not be beneficial for the laggard since it can speed up the completion of the rival™s invention by means of knowledge spillovers (De Fraja, 1993) and invite additional c ompetitors, while at the same time, it increases the patenting threshold for the laggard f irm as well (Rinner, 2003). Scholars in the fields of law and management scienc e suggest defensive publishing as a complementary strategy to patenting (Colson, 2001, Buxbaum, 2001, Rinner, 2003, Johnson, 2004, Henkel and Pangerls, 2008). In this setting, companies ponder the costs and benefits of patent protection for their inventions. Johnson (20 04) develops a theoretical framework in which companies can choose between patenting, publi shing and keeping the invention secret depending on the respective costs and benefits. A m ajor difference between this model and the patent race models is that the latter models co nsider defensive publishing as a strategy for the laggard who is unable to patent, while Johnson (2004) models defensive publishing as an equivalent alternative to patenting and secrecy. H is model predicts that publishing is most attractive for inventions that are less technically challenging or that can be easily invented 2 The first-to-invent rule allows the patent applica nt to prove the invention was completed earlier tha n the conflicting prior art (U.S. patent doctrine, se ction 102(a)). This practice is referred as fisweari ng behind the referencefl or fiRule 131 affidavitfl (Eise nberg, 2000: 2360) 3 See U.S. patent doctrine, section 102(b).

PAGE – 9 ============
3 around. Publishing can be a defense against fipatent piratesfl that strive for patent rights with the aim to exploit these patents in litigations. Th ese predictions received partial support from law scholars and intellectual property (IP) experts who strongly advise to rely on defensive publishing only for minor improvements on existing core technologies which are already well protected (Buxbaum, 2001, Colson, 2001, Rinner, 200 3). Henkel and Pangerls (2008) confirm the use of defensive publishing as a complementary strategy to patenting based on interviews with 44 patent experts from 37 publicly listed comp anies in the German business sector. They strongly dismiss the patent race model: companies w ould barely know if they are in a race and they would, in particular, not be able to determine their position in the race (whether they are the laggard or the leader). Previous literature leaves two questions unanswered : First, it is unclear whether the proposed mechanism of corporate publications as a m eans to preempt patents on related technologies works. In other words, do corporate pu blications enter the pool of prior art that is used by the patent office to evaluate new patent ap plications; and if so, can they block the grant of patent applications? Second, the question arises whether corporate publications can hinder patent grant decisions as standalone documen ts – as proposed by the patent race models – or whether they have to be combined with other IP strategies like patenting, for instance, in order to be effective – as proposed by the manageme nt science and law literature. We provide an empirical analysis for these research questions based on a sample of patent applications at the European Patent Office (EPO). W e answer the first research question by investigating the impact of corporate scientific pu blications 4 cited in patent applications on the patent examination outcome. We employ detailed information about the reason for which the specific references are relevant for the patent application, which is available from the EPO (Webb et al., 2005, Harhoff et al., 2005). This all ows us to distinguish prior art cited as non- 4 We focus on scientific corporate publications only because we want to ensure that the corporate disclosure has technical significance as evaluated by the peers. Other forms of disclosure like corpor ate postings on disclosure platforms or presentations a t trade fairs might be vague regarding its technica l specificities. Corporate scientific publications ha ve been identified as the second most effective mea ns for defensive publishing (Henkel and Pangerl, 2008) .

PAGE – 10 ============
4 infringing state of the art from documents that thr eaten the novelty of the patent application in question. Focusing on the novelty threatening refer ences to corporate science allows us to determine the blocking power of corporate publicati ons in the patent system as opposed to other pieces of prior art, like patents and publica tions taken out by the public sector. The second question can be answered by distinguishing b etween references in patent applications that are novelty threatening if taken alone and suc h that are blocking if combined with other pieces of prior art, like for instance with a paten t. The managerial view on the effect of corporate science in the patent system would be sup ported if we would find that corporate science is not blocking itself but only in combinat ion with other pieces of prior art. This would imply that firms cannot rely on publications to protect their IP but should use complementary IP protection mechanisms like patents . Besides the access to detailed information about th e patent examination procedure at the EPO, focusing on the European patent system has the advantage that Eisenberg™s (2000) critique against the effectiveness of corporate sci ence as an IP protection strategy based on legal grounds does not apply. For one, the EPO has a first-to-file rule in place rather than a first-to-invent regulation, and second, the EPO def ines prior art as all inventions being known before the first date of filing. This renders defen sive publishing a potentially more effective strategy in the European context than in the U.S. Our analysis focuses on the semiconductor technolog y in the period 1995-2000. At least two characteristics render this technology attracti ve for our analysis. First, it is characterized by a strong reliance upon scientific discoveries (B reschi and Catalini, 2010). Second, the semiconductor technology is a complex technology wh ere a single product relies on several patents. Firms in complex technologies own fragment ed patent portfolios and pursue fiportfolio maximizingfl IP protection strategies in order to avoid being held up by other firms (Hall and Ziedonis, 2001, Ziedonis, 2004). Semicond uctor firms, accordingly, engage in patent portfolio races rather than in traditional p atent races (Hall and Ziedonis, 2004). In this context, the dissemination of scientific results by firms can represent an effective

PAGE – 11 ============
5 complementary means (e.g. to patenting) for securin g their freedom to operate technologies as suggested by the management science and law literat ure. This study makes two contributions to the literatur e on corporate disclosure. First, we provide the first systematic large-scale analysis o f the effectiveness of corporate scientific publications in the patent system. We show that cor porate publications can help secure the freedom to operate a technology by blocking the gra nting of related patent applications as pieces of prior art. This shows that the mechanism imposed by both economics as well as scholars in management science and law, i.e. that c orporate science enters the pool of prior art and by doing so increases patenting thresholds, wor ks. Prior empirical evidence on the effectiveness of defensive publishing is limited to case illustrations within the field of genomics (Eisenberg, 2000, Maurer, 2002, Merges, 20 04) and interviews with IP experts in the private sector (Henkel and Pangerl, 2008). Seco nd, we show that corporate publications are especially effective against patent grant decis ions in combination with other pieces of prior art rather than as standalone documents. In l ine with the law and management science literature, this suggests that corporate publicatio ns when used for defensive publishing purpose are most effective as a complementary part of a firm™s overall IP protection strategy. The remainder of the paper is organized as follows: section 2 describes the patentability criteria at the EPO and details how scientific publ ications can impact patent grant decisions. Section 3 describes the construction of the data se t and presents descriptive statistics. The fourth section shows the empirical results. Section 5 provides additional tests to strengthen our interpretation of the main findings. The last s ection concludes. 2. The European Patent System This section outlines the patentability requirement s at the EPO and details how scientific publications can impact patent grant decisions.

90 KB – 46 Pages