شبکه های مخترعان و نقش دانشگاه: اکتشاف داده های ثبت اختراع ایتالیا
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|20038||2004||19 صفحه PDF||سفارش دهید||محاسبه نشده|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Research Policy, Volume 33, Issue 1, January 2004, Pages 127–145
This paper proposes a quantitative analysis of social distance between Open Science and Proprietary Technology. A few general properties of social networks within both realms are discussed, as they emerge from the new economics of science and recent applied work on “small worlds”. A new data-set on patent inventors is explored, in order to show that social networks within Proprietary Technology are much more fragmented than Open Science ones, except for science-based technologies. Two propositions are then put forward on the “open” behaviour expected from academic inventors, namely university scientists getting involved in Proprietary Technology networks by signing patents. Both propositions are confirmed by data, which show academic inventors to be more central and better connected than non-academic ones. The database and methodology produced for this paper are suggested to be relevant for the more general debate on the role of geographical and cognitive distance in university–industry technology transfer.
University–industry knowledge transfer is nowadays a key research subject both in economics and management studies, as well as a top entry in the science and technology policy agenda of a number of developed and developing countries. “Distance” between the two realms of academic and industrial research has been increasingly called in to explain whether the former may, or may not, benefit the latter. Two concepts have attracted most of the attention: geographical and cognitive distance. Within the geographical realm, it is usually suggested that both scientific and technical knowledge are largely “tacit” and “non-codifiable”, and require distance-sensitive transmission means such as frequent face-to-face clarifying discussions and on-site demonstrations (Feldman, 1999). As for cognitive distance, patent citations have been exploited to measure the impact of university patents and scientific publications for innovations in industry, and differences in the relevance of different research fields (Jaffe, 1989 and Tijssen, 2001). Data from innovation surveys have provided useful additional evidence on the impact of other academic activities, such as meetings and informal contacts with university researchers (Mansfield, 1995). These remarks suggest that both geographical and cognitive distance matter in so far as they contribute to reduce a more fundamental kind of distance between the academic and industrial realms, namely social distance. The exchange of tacit knowledge between university and corporate researchers requires the two social groups to share some acquaintances and/or a few codes of behaviour in terms of reciprocity and fairness (both in case of market transactions and in case of free sharing). Similarly, academic researchers’ mobility to and from industrial labs (either in the position of employees or entrepreneurs) requires a web of personal contacts for exchanging information on job and financing opportunities, and again some codes of behaviour that do not punish such mobility by portraying it as free-riding. While case studies on the theme of social distance abound, large-scale quantitative research on the same subject is more of a rare breed, limited as it is by highly demanding data requirements. The present paper summarizes the early results of a research program that aims at producing and exploiting a large data set for Italy, with information on individual inventors’ location, activity, and social ties. To date, the chief output of that program is the EPO-INV data-set on the social ties of Italian inventors, as measured by their participation to patents registered at the European Patent Office, from 1978 to 2000. A nested data-set, named EPO-INV-DOC, identifies those inventors who in 2000 were employed as full and associate professors or researchers by Italian universities. In Section 2, we discuss our choice of individuals as the key observational units. We first recall the theoretical debate on the role assigned by the ‘New Economics of Science’ to social networks as knowledge diffusion vehicles. Then we illustrate how the EPO-INV database can serve the purpose of exploring the expected general properties of those networks. In Section 3, we introduce the EPO-INV-DOC database and notion of “academic inventor”, which help moving the measurement of universities’ contribution to patenting away from the institutional to the individual level. The move may be of crucial importance for studying countries such as Italy, where universities are not organized to manage Intellectual Property Rights. In Section 4 we provide a few exploratory statistics on both the EPO-INV and the EPO-INV-DOC databases, which help identifying the structure of the Italian social network of inventors, and the role played by academic inventors. In Section 5 we conclude by sketching our future research plans.
نتیجه گیری انگلیسی
This paper reports the early steps of quite an ambitious research program, whose ultimate goal is assessing the role of geographical and knowledge proximity in technology transfer not just on the basis of a few assumptions on the nature of knowledge exchanges, but as a function of the social structure supporting them (see also Breschi and Lissoni, 2003). The steps undertaken here are limited to the exploration of the social structure of Italian inventors, as it emerges from data on patent applications and co-invention relationships. In particular, we have explored whether the social positioning of academic inventors is by any means outstanding, as the new economics of science seems to suggest. Two propositions have been put forward. Taken together, they suggest that academic inventors may retain some of their “openness” when moving from the realm of Open Science to that of Proprietary Technology. We also outlined the reasons why testing those propositions require large data sets on individual researchers, such as the EPO-INV and the EPO-INV-DOC databases. Our empirical enquiry has just started, so that we could produce little more than descriptive statistics. While preliminary, the conclusions we reach are nevertheless promising. • Networks of inventors, which we take as representative of social networks within Proprietary Technology, are highly fragmented, with the exception of technological fields wherein science plays an important role, such as chemical and, to a lesser extent, electronics. • Academic inventors that enter the network are, on average, more central than non-academic inventors: academics exchange information with more people and across more organizations. Therefore; they play a key role in connecting individuals and network components. However interesting, these conclusions are little more than a good start. First and foremost, we need to know more about non-academic inventors, whom by-and-large we have been dealing with as a sort of residual, thus ignoring their heterogeneity. At the very least, we ought to be able to distinguish between researchers from large corporate labs from other industrial researchers, something we can do only by checking the applicants’ names thoroughly. Second, in order to know more about academic inventors’ attitudes towards patenting we need to proceed with interviews and questionnaire enquiries. They will help us checking the identity of the co-inventors of academics: are they university technicians, PhD students or retired professors (who have escaped our EPO-INV-DOC database)? Or are they not academics, such as researchers from the patent applicants’ labs? More information on this point will help us refining the positive evaluation we have tentatively expressed when commenting the high centrality of academic inventors. Does centrality really signal a deep involvement in an applied research field, to which academic inventors contribute also by connecting people and introducing young researchers to the community of industrial researchers? Or does is merely reflect an opportunistic behaviour, by which university scientists sell some of their research ideas on an occasional basis, and rely either on colleagues or on the applicants’ employees to set up an improvised research team? Finally, we need to produce data-sets for countries other than Italy, comparable with EPO-INV and EPO-INV-DOC. It is the only way to judge whether the amount of patents produced by Italian academic inventors is high enough to force a revision of the usually harsh judgements on Italian universities’ contribution to innovation, or it is a mere reflection of the original way we found to calculate university patents. The same question holds for Europe as a whole, when compared to the US: could re-classifying patents by inventor, rather than by applicant, lead to a more positive evaluation of university–industry links in Europe, one which takes into account the relative inexperience of European universities in handling IPRs? Although demanding, these are research questions which we expect to answer in the near future, on the basis of the data and methodology presented here.