جهانی شدن تکنولوژی در بازارهای نوظهور: مدل جاذبه بر روی عوامل مؤثر بر همکاری های بین المللی ثبت اختراع

International technological collaborations (ITCs) and face-to-face interactions are an important vehicle of knowledge diffusion. This paper analyzes ITCs among USPTO patents’ inventors in eleven emerging economies and seven advanced countries (1990–2004) and a novel database on companies’ country of origin. Technological proximity and sharing a common language are key drivers of ITCs. When the applicant’s ownership is in the emerging country ITCs depend positively upon transport and communication costs (geographical distance and longitude) and negatively upon the strength of intellectual property rights (IPRs). Stronger IPRs positively affect ITCs from subsidiaries of multinational firms.

Endogenous growth models have shown that commercially-oriented innovation efforts by profit-seeking firms promote technological progress and productivity growth (Aghion and Howitt, 1992 and Romer, 1990) and international knowledge spillovers are key drivers of catching up and income convergence (Fagerberg, 1994 and Grossman and Helpman, 1991). Recent empirical literature on international knowledge flows has made important progress and identifies different channels of knowledge spillovers: import flows, cross-border investments, and a disembodied direct channel of codified information. Most of this literature focuses on developed or Organization for Economic Co-operation and Development (OECD) countries, however, the literature shows that imports are a significant channel of technology diffusion (e.g., Coe et al., 1997 and Keller, 2004). Some evidence suggests also that technical knowledge is transmitted through exports. Finally foreign direct investments (FDIs) from multinational corporations generate technology spillovers (in particular vertical spillovers), through the physical presence of the plant and labor turnover (Keller, 2010 and Keller and Yeaple, 2009a). In particular, as emphasized by Keller’s survey (2010, chap. 19), empirical evidence shows that geography and physical distance importantly shape the diffusion of technical knowledge. The idea that international knowledge spillovers affect productivity growth enhancing technological adoption and innovation in developing countries (Keller, 2010 and Montobbio and Sterzi, 2011) stimulates governments and international organizations to place the domestic dissemination of frontier knowledge high up on their policy agenda (e.g., World Bank, 2010). At the same time, recent empirical literature has also shown that knowledge spillovers tend to be localized1 and require absorptive capacity (Cohen and Levinthal, 1989 and Griffith et al., 2004). This is because technological knowledge includes not only materials and knowledge codified in blueprints, manuals, publications, and patents but also know-how, routines, and organizational capabilities, much of which is tacit in nature (Cimoli et al., 2009 and Dosi, 1988). Tacit knowledge (e.g., related to technical know-how or nonstandard production) is costly to transfer, and its transferability is limited by its embeddedness in individuals, teams, and organizations. As a consequence, knowledge diffuses more rapidly when interpersonal links in the form of joint research efforts and collaborations create opportunities for learning which go beyond the exchange of codified information. In particular, recent evidence underlines that research collaborations create social networks which can foster mutual learning and, as a result, individuals and companies that actively participate in a network of knowledge exchange (Breschi and Lissoni, 2009, Hoekman et al., 2009 and Singh, 2005) are more innovative. This paper therefore analyses international technological collaborations between patent inventors in a “North–South” gravity model looking at the interactions between emerging and advanced countries under the assumption that technological collaborations imply face-to-face interactions that are a key vehicle of knowledge spillovers. However, while scholars have been widely aware of the nature of globalization in terms of trade and financial openness, there is no clear consensus about the extent of globalization of technological activities. Academics and international organizations acknowledge that R&D activity is increasingly done at the international level (OECD, 2008). A number of communications technologies, such as fiber optics, social networks, and satellite communications, facilitate international technological activity and, in parallel with the decrease in communications and transport costs, geographical distance should have a declining impact on technological collaborations and research ventures. At the same time, some authors (Granstrand et al., 1992 and Patel and Vega, 1999) show that the technological activities of the world’s largest firms continue to be firmly embedded in their headquarters in the home countries. In parallel, Picci (2010), focusing on OECD countries, studies the degree of internationalization of innovative activities using patent data and finds a statistically significant impact of geographical distance. He shows that even if R&D internationalization is now more pronounced than it was 20 years ago there is a “lasting lack of globalization” that is surprising in the light of the abundant anecdotal evidence of both increased domestic R&D activities in emerging countries and offshoring R&D activities to countries such as China and India. Moreover, the scale and scope of international technological collaborations are affected by the legislation on intellectual property rights (IPRs) which has changed rapidly in recent years after approval of the Trade-related aspects of intellectual property rights (TRIPs) agreement signed in 1994 and adopted and implemented by different countries at different points in time. One of the main economic justifications of the TRIPs agreement is that IPR reinforcement in emerging countries facilitates knowledge transfer and dissemination from advanced countries.2 It is relevant then to control for the impact of IPR legislation on technology transfer and spillovers brought about by international technological collaborations between inventors. In addition, the impact of geographical distance and IPR legislation on international technological collaborations—and, in turn, on knowledge transmission—depends upon the typology of firms involved in the innovative project. It is therefore important to distinguish whether international technological collaborations occur with the joint contribution of different companies in different countries or within the laboratory of a multinational corporation (MNC) located in an advanced or emerging country or, finally, within the laboratory of a company from an emerging country. This paper contributes to the literature, building a novel database that takes into account not only the residential address of inventors and assignees but also the ownership of companies and their nationality. In parallel, the specific composition of the international team of inventors and the relative weight of the different countries in the team are also taken into account. For example, if the international team of inventors contains a large majority of inventors from an advanced country and the patent is applied for by a company with an address in the advanced country, we can expect that the international collaboration is the result of a movement of skilled labor from the emerging to the advanced country. This type of international collaboration (and its determinants) is clearly different from a collaboration occurring in a laboratory of a MNC subsidiary located in the emerging country. We use patent data from the US Patent and Trademark Office (USPTO) and we collect economic and institutional data from different sources. The sample covers 18 countries: a group of large emerging economies (Argentina, Brazil, India, Israel, China, South Korea, South Africa, Mexico, Malaysia, Singapore, and Turkey) and their relationship with seven advanced countries (USA, UK, Japan, Italy, Germany, France, and Canada). In order to model the impact of geographical distance and the impact of IPR reinforcement on technological collaborations between emerging and advanced countries, we use a modified version of a gravity equation and different empirical specifications, using panel data and Poisson pseudo-maximum likelihood (PPML) in order to tackle various econometric problems. Our main results are that geographical distance is not important per se and distance matters mostly through trade and cultural similarities. Results are slightly stronger for time zone differences. Technological proximity is a very important factor that favors collaboration. Fixed effects models show that countries experiencing an increase in IPRs protection tend to be more involved in international collaboration. This effect is greater for those countries that have stronger trade relationships, and is positive only in the emerging countries characterized by a very low level of IPR legislations before the TRIPs agreements. Importantly, for a subset of countries, we show that these determinants of international technological collaboration vary according to the type of collaboration considered and country of origin (emerging vs. advanced) of the companies involved. For example, for collaborations deriving from laboratories of multinational subsidiaries, we have no effects of geographical distance and a positive effect of IPR reinforcement. On the contrary, for collaborations that involve only a company from the emerging market, communication and transport costs—proxied by geographical distance—turn out to be important and the effect of the reinforcement of IPRs is negative. The paper is organized as follows. In Section 2 we present recent evidence on the geography of knowledge spillovers and discuss to what extent co-inventor relationships can be considered an indicator of knowledge flows. In Section 3 we present our model of weightless gravity used to study the determinants of international technological collaborations between emerging and advanced countries. In Section 4 we present data and the empirical model. Section 5 discusses the results of the econometric analysis. Finally, Section 6 concludes.

There is a growing body of literature that underlines that face-to-face contacts and personal interactions are a crucial vehicle of knowledge transfer and spillovers. In emerging countries, access to foreign technologies, in particular, via collaborations with foreign counterparts, both in the domestic country and abroad is a hot political issue. As scientific research increasingly involves international teams and mobility of researchers is on the rise, it is possible to ask whether collaborative links with foreign laboratories rely more on relational and capability proximity than on geographical distance. Also, multinationals are increasingly delocalizing R&D activities in host countries, spurring a debate on what are the conditions under which the local community of researchers and firms can learn by tapping into foreign collaborative networks. Taking these issues as a point of departure, this paper analyzes international technological collaborations among inventors in emerging and advanced countries using USPTO patent applications and asks to what extent they are affected by a decrease in communication and transport costs (that are a function of geographical distance). In addition, it studies the effects of a set of economic and institutional variables like technological proximity, sharing a common language, or a colonial link, and, finally, the recent reinforcement of IPRs brought about by the TRIPs agreement. This paper uses a novel database that considers not only the residence of inventors and assignees but also the companies’ country of origin (in terms of ownership). In addition, the specific composition of the international team of inventors and the relative weight of the different countries in the team are taken into account. An important point of the paper is that results depend upon the type of collaboration considered and it makes a substantial difference whether the collaboration stems from a multinational company from an advanced country or a company from an emerging country. Overall, geographical distance seems to have a modest effect on international collaborations when controlling for trade relationships, technological, and cultural distances. However, differences arise when we consider the origin and the nature of such collaborations. In particular, we observe that distance affects international collaborations only when they originate in laboratories of companies from the emerging countries. On the contrary, geographical distance in itself is not important for those collaborations originating in MNC subsidiaries or via the international mobility of inventors from an emerging to an advanced country. If simple geographical distance has no strong (and negative) impact on international collaborations, time zone differences, to some extent, do. We also find that technological proximity is an important factor in explaining international technological collaborations. Sharing a common language is also always significant in the main models. This effect is mainly driven by the collaborations generated by the international mobility of highly-skilled workers from the emerging countries or from companies from the emerging countries. Sharing a common language has no significant impact for collaborations within MNC subsidiaries. Our paper contributes also to the policy debate on the effects of IPR reinforcement and our evidence suggests that there may be some positive effects on knowledge flows generated by the reinforcement of IPRs for those economies which started at the beginning of the 1990s with a low level of IPR protection. However, these results have to be taken with extreme care because the impact of the IPR regime is extremely complex and can vary from sector to sector and country to country. Importantly, we also show that the impact of IPR reinforcement varies according to the type of collaboration considered and country of origin (emerging vs. advanced) of the companies involved. For collaborations deriving from laboratories of multinational subsidiaries we have a positive effect of IPR reinforcement. On the contrary, for collaborations that involve only a company from the emerging market the effect of the reinforcement of IPRs is negative. Finally, our additional results show that a positive result may be confined to pairs of countries that are close trade partners and to those countries with a lower level of per capita GDP.