Introduction: The Organization of Science Collaboration

The internationalization or globalization of scientific collaboration is both a key feature of today’s science and a political mandate. Several mechanisms are contributing to the internationalization of science. First, some of the most important institutions and frameworks for conducting science have become increasingly global. An obvious one is the use of English as the global science language in publications, conferences, and communication (Ammon 2001). Closely related to this is the use of peer reviews for assessing the quality of publications, grant proposals, and concepts (Smith 2006). Second, international collaboration means accessing and combining the scientific resources, ideas, and competencies available worldwide. This can raise the quality not only of research, but also of education, and may accelerate discovery processes (Carlsson 2006; de Wit 2002; Kafouros et al. 2008). During recent decades, reductions in travel and communication costs as well as advances in communication tools, including the Internet and social media, have made internationalization possible and helped these forces to shape the organization of science. Politicians and managers in higher education and science, however, have responded to these changes, putting internationalization high on their agendas. One example is the demand for international collaboration in research and teaching as a means of quality control and a method of broadening the scope of activities (Knight 2004). An existing international collaboration is a performance measurement which is easy to monitor and use as part of evaluation and performance-based funding procedures (cf. Geuna and Martin 2003).

However, the counter-hypothesis of primarily national or even regional science organization is not far away (Ponds 2009). Powerful mechanisms such as national languages, policies, and funding agencies implicitly favor domestic collaboration. The advantages of close spatial proximity for frequent personal contacts and knowledge-sharing are obvious, and with respect to commercialization, local industry may seek scientific partners around the corner. These factors are at the core of some well-established theoretical concepts such as the National Innovation System (Lundvall 1992; Nelson 1993) and the Regional Innovation System (Cooke 1998). All these concepts provide the line of arguments for the hypothesis of a nationalism of technology systems (cf. Archibugi and Michie 1997), focusing on the complex interactions among the main groups within an innovation system, namely the industrial base, the governing state, and the science base, with scientific collaboration being one of the sub-dynamics within such complex innovation systems.

Network science allows us to analyze whether our images of how scientific collaboration is organized are accurate. For most scientific fields, network science shows that a national and regional collaboration bias is highly relevant and demonstrates features of a universally applicable law. While social networks have their own structural properties compared to other forms of networks (Newman and Park 2003), they represent a potentially large explaining factor for a universal function relating collaboration and space. A technical explanation for a decreasing probability of collaboration with distance is related to the fact that the actors and ties in networks are not randomly distributed. Large centers of knowledge creation are more likely to be found in urban agglomerations and their dominating organizations, that is, those which already have a large number of collaborations are more likely than small units to receive an additional one (cf. Yook, Jeong, and Barabási 2002). Still, this does not explain the spatial adjacency in which this linking activity is taking place.

Findings reported in the more recent literature on scientific collaboration in networks stem from three different perspectives: (1) looking at forces that increase internationalization; (2) regarding impediments to internationalization; (3) using network analysis to explain network structures and dynamics.

Many empirical studies have collected evidence for the internationalization of science and an increase in collaboration distances (Luukkonen, Persson and Sivertsen 1992: Archibugi and Iammarino 2002; Glänzel and Schubert 2005b; Leydesdorff and Wagner 2008; Tijssen, Waltman and van Eck 2011; Waltman, Tijssen and van Eck 2011; Ahlgren, Persson, and Tijssen 2013) that is sidelined by a general increase in research team sizes (cf. Hicks and Katz 1996). One important finding relates to an increase of the mean spatial distance between a paper’s co-authors from the 1980s to today. This gives quantitative proof for the assumption that internationalization is a structural feature of today’s science. However, Ahlgren et al. (2013: 782) also found that there is a quantitative counter-effect on the local scale, that is, an increase of local collaboration frequencies.

Quantitative evidence alone may say little about the true importance of internationalization and be of no help assessing whether an internationalization of scientific collaboration is indeed desirable. The quality of scientific research, however, may also rise with collaboration internationalization. This is the essence of findings presented by Jones, Wuchty, and Uzzi (2008) and Narin, Stevens and Whitlow (1991), who show that multi-collaborator papers spanning organizational boundaries receive more citations and thus seem to be more successful. This finding relates to the diversity argument for complex processes in Kafouros et al. (2008), who argue that innovation activity (as a complex process) is enhanced when firms internationalize, that is, absorb ideas from diverse sources. Diversity as a driver of opportunities and performance is also a prominent topic in complex network research (cf. Eagle, Macy, and Claxton 2010). These two lines of argument (network diversity, geographical distance) are likely to be interrelated, that is, the network-based diversity increases with geographical distance, simply because the physically distant collaborators have different local embeddings (e.g., institutional framework, cultural idiosyncrasies). Network-based approaches lead to the second large body of literature, which is concerned with the structural impediments to internationalization of scientific collaboration, coming up with equally convincing results. Examples of empirical studies that address limits to internationalization and stress the importance of national border effects have been presented by Hennemann, Rybski, and Liefner (2012) and Hoekman, Frenken, and Tijssen (2010). Arguments against growing internationalization have also been expressed from the perspective of economic theory. For example, Beckmann (1993, 1994) suggested classical gravity models as a perspective from which to understand collaboration patterns. According to this view, scientific collaboration is mainly the outcome of (rational) choices of individual researchers, who wish to maximize or (being more modest) to satisfy their own research output (Beckmann 1993: 8). Due to cognitive limits, the individual does not know all potential collaborators (such search restrictions can be extended by transaction costs, e.g., for traveling, meeting, while other constraints include language barriers: Beckmann 1994: 242), and collaboration thus tends to be more local.

However, when considering the evolution of collaboration networks in science, Barabási et al. (2002) collected interesting empirical evidence that may help to save the idea of productive science systems with strong local structures by showing that it is possible to find decreasing average shortest path lengths among the collaborators in growing scientific networks. This finding partly contradicts findings from other research on (social) networks that are governed in a similar way by hub-and-spoke structures. It seems that the hubs (e.g., the top league of researchers in a field) are perfectly optimizing their immediate network neighborhoods and contribute to a greater system-wide knowledge access. Having said this, it could be hypothesized that knowledge produced and disseminated in such a global system may well be leading toward more globalized scientific knowledge over time, with diversified ideas improving the overall outcome quality. Jansen, von Görtz, and Heidler (2010) attempt to dissemble relevant structural properties of science networks by testing the influence of different types of sub-structures within larger networks on the knowledge production performance. Neither clustered areas of networks (e.g., intensively cross-connected groups of researchers) nor strategic positioning alone (e.g., researchers who attempt to utilize structural holes, that is, provide non-redundant connections in networks) are per se beneficial for enhancing system-wide knowledge production. One potentially hampering factor for quality enhancement can be derived from the deeper analysis of the results of Barabási and colleagues (2002), who conclude that in growing scientific networks, the actors already present dominate the community to the detriment of the new entrants, that is, the preference of collaborating with researchers already present is much higher than that with new entrants (2002: 611, cf. also Zitt, Bassecoulard, and Okubo 2000). One might wonder whether this poses a potential threat for scientific progress.

This brief review of the related literature has shown two things: first, there is evidence of a quantitative and qualitative trend toward the internationalization of scientific collaboration, while convincing counter-evidence is also present. Second, from analyzing the structures of collaboration, a number of arguments emerge that indicate which of the structures seem to be more desirable from the point of view of science, society, or politics. However, there is no common understanding of whether science is becoming more global, and whether this should indeed be the case.

Against this background, this chapter aims to provide a compressed piece of evidence that clarifies the respects in which scientific collaboration is international (or global) and the respects in which there is a persistent national and regional bias. It will be shown that both features can be present at the same time. This chapter further summarizes and explores several arguments as to what future development of scientific collaboration might be considered beneficial. There are two major equitable views of how to explain collaboration at the individual level. On the one hand, if maximizing the outcome is important, the best partner available will be chosen no matter where this partner is located geographically. On the other hand, if resource constraints are present, the selection of geographically adjacent partners will be favored.

In order to provide a consistent research framework for the purpose of evaluating the structures and trends in global science collaboration, it is necessary to define comparable research units, because similarities in research communities are considered to be important proxies for space-independent collaboration, thus leveling out the described network-based diversity effects, at least in terms of the core topics. Therefore, the concept of epistemic communities serves as a reference notion for the empirical section. Epistemic communities are mainly referred to as being informal collaboration networks (Ynalvez and Shrum 2011). Members of epistemic communities are commonly characterized as being professionally similar, that is, they share the same or similar notions or beliefs about subject-specific applications and techniques (cf. Haas 1992; Weisberg and Muldoon 2009). This should enable all members to understand one another easily. With increasing improvements in technical infrastructure, it is the cognitive proximity (cf. Boschma 2005; Nooteboom et al. 2007) that enables members of the epistemic community to compensate for a possible lack of spatial proximity, also transcending organizational boundaries (cf. Gertler 2003) and forming global science networks of researchers who specialize in similar fields (Moodysson 2008). Especially at the frontiers of science, new knowledge can be expected to be created through joint efforts in international teams of excellence (Power and Malmberg 2008).

Concerning the interplay of the science sector with the (private) business sector, it is becoming increasingly prevalent for the overall economic development that comprehensive collaboration strategies, both national and international, are increasing competitiveness. This is particularly relevant in science-based industries and with respect to the technology and creative side of globalization, beyond production chains (cf. Archibugi and Michie 1997; Florida and Mellander, this volume, Chapter 15). Catchphrases such as “triple helix” (Etzkowitz and Leydesdorff 2000) or “national systems of innovation” (cf. Lundvall 1992; Nelson 1993) seek to explain this complex interplay of industries, governments, and the public research sector. It is not only advanced economies that are addressed by these concepts, but also rapidly growing latecomer economies, such as China, who are aware of the merits of intensive interaction, and the science sector can be seen to a certain extent as an integrator of global knowledge and local sources, even in the industrial sector (cf. Hennemann 2011; Barnard et al., this volume, Chapter 18). Therefore, the way science is organized geographically, as discussed in this chapter, relates directly to the ongoing debate concerning the strategic development paths of nations (e.g., the “techno-nationalism vs. techno-globalism” debate: Archibugi and Michie 1997; Ponds 2009).

How Researchers Connect Given a Set of Potential Partners at Various Distances

The measurement of scientific activity in quantitative terms is the major undertaking in scientometrics, a science field that dates back to the 1950s (cf. Garfield 1955; de Solla Price 1965). Scientometric research, as the science of science in a narrow sense, is commonly restricted to the measurement of pure (journal) publication activity. However, with the ubiquitous availability of other data sources such as research databases or the Internet, with its social communication platforms, this narrow view was relaxed to account for such new sources. This wider view is often termed informetrics, the measurement of information and knowledge production and dissemination. Consequently, three different sources of data are commonly used for the quantitative analysis, including academic publications, patents, and Internet resources. In recent years, the collaboration measurement in epistemic communities has gained importance in evaluating the knowledge production and dissemination in science, technology, and innovation systems at the system level. The evaluation and analysis of effects of such developments in the science sector is investigated within the methodological framework of the “science of team science” (Börner et al. 2010) or within the framework of “spatial scientometrics” (Frenken, Hardeman, and Hoekman 2009; Hennemann et al. 2012). The complex systems under investigation can be represented by networks (or mathematical graphs). Compared to simplistic frequency counting of the number of publications or patents in a given science field, the network approach to collaboration in science has the advantage of capturing a great variety of structural effects of the system itself that are present in most social systems, and allows for a direct translation of the network-oriented theoretical concepts into empirical research. The following empirical part of this chapter uses this systemic perspective and seeks to reveal the collaboration patterns in different epistemic communities in geographical space. The main question in this endeavor is to find out whether the geographical distance between any two collaborators in a given (and narrowly defined) science field is relevant for their probability of conducting joint research and eventually co-authoring a paper. Other questions of interest include the differences among science fields, for example, finding out whether analytical sciences such as biotechnology-related research is differently organized compared to more synthetically driven fields such as electrical engineering. One hypothesis often presented in recent decades relates to the impact of a shrinking geographical space due to the increasing capabilities of interacting through technical systems (e.g., the Internet) or through improvements in infrastructures that sharply reduce travel time (e.g., availability of high-speed trains, dramatically increased flight schedules). The basic rationale behind this is that knowledge production is made easier on a global scale, hence gaining in importance compared to local or national collaboration, and subliminally suggesting that this would improve the quality of the knowledge produced.

The Geographical Patterns and Collaboration Probabilities in Six Scientific Communities

Figures 16.1 to 16.6show the world maps for the collaboration activity on the left panel. While for the first two communities, bluetooth (Figure 16.1) and image compression (Figure 16.2), the Asian region is comparatively present, research on heart valves (Figure 16.3) in this part of the world is very weak. While Asian research activity can be neglected in heart-valve research, it is again very pronounced in the case of the bird flu virus research. African research steps in, showing contributions from a continent that is usually hardly visible in the global research system. Asia is the assumed origin of the bird flu virus (Figure 16.4). It is thus a reasonable result that these areas are included in the global research system, adding competence to the field. Tissue engineering appears to be driven by European and especially North American research organizations (Figure 16.5), while research on nanotubes seems to be equally distributed among the three main research regions North America, Europe, and Asia (Figure 16.6).

A striking element of all maps is the strong local emphasis in the collaboration activity, which can be derived from the way the connecting lines between collaborating organizations were drawn. The closer the partners are located, the higher the “spikes” are, whereas long-range connections can be identified by the smooth-curve character. This geographical collaboration pattern is also reflected in the probability plots on the right panels in Figures 16.1 to 16.6. All six plots show very similar curve shapes for all indicators presented. The empirical curve (circle line in the plots) shows the overall distance relation for all collaborations. This relation is subdivided into national (triangle-down) and international (triangle-up) collaboration. The collaboration according to the randomized baseline model is shown in the squared lines, with the dashed lines representing +/– 2 standard deviations from this estimated mean. As the plots are scaled in logarithm for both axes, straight lines would indicate a power-law relation. It is obvious that there is a very high propensity for local collaboration in all six communities. This finding is significantly different from the random model. In turn, international collaboration is significantly lower than expected. The mainly horizontal line for international collaboration shows that once a decision is made to select international partners, the location of that partner becomes irrelevant. Overall, the community structures as represented by the probability plots (right panels) show that the relation between distance and the chance of collaborating with a particular partner is indeed very similar in all fields analyzed.

In summary, the slope of the probability function can be interpreted as an indicator for the relevance of geographical proximity for collaboration in epistemic communities. The more negative the exponent is, the more important geographical proximity is between two potential collaborators. For exponents of –0.75, as typically found here, the probability of collaborating is 2.8 times higher if the potential collaborators are within the same country at a 50 km distance than if they are at a 200 km distance. If the collaborators are located in two different countries, the probabilities are different by two orders of magnitude (around 50 times) in favor of an intra-country collaboration.

While the distance plays a role in national collaboration activity (supporting the restricted resources hypothesis), it does not matter where the international partner is located once international partners are incorporated into the research team (supporting the outcome maximization hypothesis). Although not explicitly tested here, other authors found the national level to be especially important for academic-industry collaboration (cf. Ponds 2009).

Collaboration Over Time

The collaboration patterns presented so far represent the recent activity in the respective communities. In order to evaluate changes in collaboration over time, the case of the tissue engineering community was selected. The first paper on that topic was written in 1984. However, it was a single-author paper and therefore did not meet the requirements of this study. The fraction of co-authored papers increases over time. While two-thirds of the papers in tissue engineering were written by a single author in the 1980s and early 1990s in the initial phase of this science field, nowadays more than 80% are the joint work of two or more researchers. However, this trend is not exclusive to tissue engineering, but present in most disciplines (cf. Ponds 2009: 83, who found the same shares for a variety of disciplines). Reasons include increasing division of labor due to the increasing complexity in the scientific work process and the shortening of the knowledge life-cycle, which increases the pressure to publish, especially for younger researchers (cf. Bozeman and Corley 2004). The latter point is strongly linked to increasing evaluation and performance-based funding (cf. Liefner 2003). However, what is striking in the dynamics of a science field is the remarkable stability of the geographical influence (Figure 16.7). The exponent of the collaboration probability function fluctuates at around –0.9 to –0.7. Although it appears that the role of space is even stronger at the beginning of the evolution of the science of tissue engineering, this should be interpreted with caution due to the large confidence intervals around the median in the first years when the epistemic community was small. We can conclude that given the dramatic technological (e.g., Internet) as well as organizational developments (e.g., political incentives to internationalize) during the period from the 1980s until today, there is almost no sign of a change concerning the collaboration activity in geographical space. This is supported by the findings of Ponds (2009), who also found a stable proportion of international collaboration compared to national collaboration over this period of time.

Explanatory Factors: Institutions, Network Dynamics and Attributes of Knowledge

As Figure 16.7 illustrates, science collaboration is both international (involving distant international collaboration partners) and shaped by a strong distance decay (thus a high propensity of choosing partners nearby). The following section aims to answer the question of why the collaboration pattern in science has this dual feature and which factors contribute to its emergence and permanence.

Of greater importance are several bundles of factors briefly touched upon in the introductory section, namely the institutions that shape the organization of science, the mechanisms of network evolution, and the content of collaboration regarding the types of knowledge involved.

Science Collaboration: Truly Global and Better?

From the perspective of national institutions, global science collaboration is tied to reputation and quality assessments, while the notions of network science and types of knowledge represent a view of global science collaboration as a means to ensure effective dissemination of mostly codified knowledge. So why should global science collaboration be worth the investment of time and money?

The core argument is centered on an increase of research productivity through international collaboration. For example, Ponomariov and Boardman (2010) find a positive relation between university research centers, collaboration, and improved productivity. Similarly, Barjak and Robinson (2008) claim that it is indeed beneficial to integrate researchers from another country in order to create well-balanced research teams. Scientometric analyses give support to these notions. At least for the level of inter-organizational collaboration, quality improvements could be empirically shown (cf. Jones et al. 2008). With its general early internationalization, and especially with the strong increase in international activity (as measured in internationally co-authored articles through the 1980s and 1990s), the academic world has somehow preempted trends in the corporate world (Archibugi and Iammarino 2002: 118). Referencing back, analogies from the business world help to rationalize the merits of global collaboration activities, since at least for companies, being present in many different markets is seen as a prerequisite for success (Kafouros et al. 2008, cf. also Carlsson 2006). Organizing science in such a way as to internalize the diversity of the world’s many national science systems can be thought of as equally promising. Overall, one cannot eliminate the possibility that the few international collaborations have the greatest impact in terms of scientific progress. However, direct proof for this is hard to find and the greater success for international science teams remains ambiguous (Abbasi and Jaafari 2013, cf. also the discussion in Persson 2010).

So there is little doubt that science can benefit from global collaboration. The main counter-arguments, however, have been explored above. First, international collaboration is more expensive than national or local cooperation, and second, keeping science national may secure some advantages in international competition. Competition in science may also spur innovation as it does in business.

Independently of these considerations, recent decades have witnessed some new entrants in global science, namely East Asian countries and research organizations, including China (see, e.g., Figure 16.4). Overall, the gap between countries traditionally leading in economic terms and those that follow at a distance has narrowed with respect to innovative capabilities such as patents and article publications (Castellacci and Archibugi 2008).

Consequently, one must ask whether this new trend is more an opportunity or a burden for developing country entrants (Archibugi and Pietrobelli 2003). Drawing on similarities with business again, international collaboration may benefit both parties involved, allowing learning on both sides (Thorsteinsdóttir et al. 2010). Upgrading the scientific competence of developing countries, for example with the help of foreign knowledge through science collaboration or by means of domestic investments, may thus contribute to technological convergence across countries (cf. Archibugi and Iammarino 1999; Castellacci and Archibugi 2008). However, the task of converting a newly acquired scientific competence into upgrading and competitiveness of domestic firms requires additional political support and rising absorptive capacity among firms (Ernst and Kim 2002).

Conclusion

Science is global, but at the same time also a national and local activity, with a much higher propensity for choosing short-distance collaboration. From a network perspective, this is not an unusual pattern, as it combines the benefits from long-distance networking with efficiency and cost-effectiveness of short-distance networking. There is enough evidence to assume that international collaboration can provide a different quality of knowledge input and can thus give unique stimulus for scientific activity. But there is no reason to demand that international collaboration should be promoted beyond its current level at the expense of national or local collaboration. Money should be spent wisely, and only science organizations with strong internal capacities or individuals with unique expertise can engage in meaningful international collaboration (cf. Barnard et al., this volume, Chapter 18).

The nation turns out to be still the most important entity regarding institutions and infrastructures that shape science collaboration and technology systems (cf. Archibugi and Michie 1997; Pavitt 1998; Archibugi and Iammarino 1999). In this context, it should also be acknowledged that global scientific progress may come not only through international collaboration, but also through competition among groups of local scientists in the arena of international science (e.g., Marginson 2006, cf. Dasgupta and David 1994).

Hence, the spatial pattern of global science collaboration can be expected to remain unchanged. Relatively few global connections will stretch between science hubs, with the majority of connections being established at short distance around the hubs. However, it seems desirable to include all nations in the world science system and not to forego ideas that may evolve under the particular circumstance found in countries that do not yet have a visible role in global collaboration.

References

1. Abbasi, Alireza, and Ali Jaafari. 2013. “Research Impact and Scholars’ Geographical Diversity.” Journal of Informetrics 7(3): 683–692.
2. Ahlgren, Per, Olle Persson, and Robert J.W. Tijssen. 2013. “Geographical Distance in Bibliometric Relations within Epistemic Communities.” Scientometrics 95(2): 771–784.
3. Altbach, Philip G., and Jane Knight. 2007. “The Internationalization of Higher Education: Motivations and Realities.” Journal of Studies in International Education 11(3–4): 290–305.
4. Ammon, Ulrich (ed.). 2001. The Dominance of English as a Language of Science. Berlin: de Gruyter.
5. Archibugi, Daniele, and Simona Iammarino. 1999. “The Policy Implications of the Globalization of Innovation.” Research Policy 28(2–3): 317–336.
6. Archibugi, Daniele, and Simona Iammarino. 2002. “The Globalization of Technological Innovation: Definition and Evidence.” Review of International Political Economy 9(1): 98–122.
7. Archibugi, Daniele, and Jonathan Michie. 1997. “Technological Globalization or National Systems of Innovation?” Futures 29(2): 121–137.
8. Archibugi, Daniele, and Carlo Pietrobelli. 2003. “The Globalization of Technology and Its Implications for Developing Countries – Windows of Opportunity or Further Burden?” Technological Forecasting & Social Change 70(9): 861–883.
9. Barabási, Albert-Laszlo, Hawoong Jeong, Zoltan Néda, Erzsebet Ravasz, Andras Schubert, and Tamas Vicsek. 2002. “Evolution of the Social Network of Scientific Collaborations.” Physica A – Statistical Mechanics and Its Applications 311(3): 590–614.
10. Barjak, Franz, and S. Robinson. 2008. “International Collaboration, Mobility and Team Diversity in the Life Sciences: Impact on Research Performance.” Social Geography 3: 23–36.
11. Bathelt, Harald, Anders Malmberg, and Peter Maskell. 2004. “Clusters and Knowledge: Local Buzz, Global Pipelines and the Process of Knowledge Creation.” Progress in Human Geography 28(1): 31–56.
12. Beckmann, Martin J. 1993. “Knowledge Networks – The Case of Scientific Interaction at a Distance.” The Annals of Regional Science 27(1): 5–9.
13. Beckmann, Martin J. 1994. “On Knowledge Networks in Science: Collaboration Among Equals.” The Annals of Regional Science 28(3): 233–242.
14. Börner, Kati, Noshir Contractor, Holly J. Falk-Krzesinski, et al. 2010. “A Multi-Level Systems Perspective for the Science of Team Science.” Science Translational Medicine 2(49): 49cm24.
15. Bornmann, Lutz, and Loet Leydesdorff. 2011. “Which Cities Produce More Excellent Papers Than Can Be Expected? A New Mapping Approach, Using Google Maps, Based on Statistical Significance Testing.” Journal of the American Society for Information Science and Technology 62(10): 1954–1962.
16. Boschma, Ron. 2005. “Proximity and Innovation: A Critical Assessment.” Regional Studies 39(1): 61–74.
17. Bozeman, Barry, and Elizabeth Corley. 2004. “Scientists’ Collaboration Strategies: Implications for Scientific and Technical Human Capital.” Research Policy 33(4): 599–616.
18. Carlsson, Bo. 2006. “Internationalization of Innovation Systems: A Survey of the Literature.” Research Policy 35(1): 56–67.
19. Castellacci, Fulvio, and Daniele Archibugi. 2008. “The Technology Clubs: The Distribution of Knowledge Across Nations.” Research Policy 37(10): 1659–1673.
20. Cooke, Philip N. 1998. “Introduction: Origins of the Concept.” In Regional Innovation Systems: The Role of Governances in a Globalized World, ed. Philip N. Cooke, Hans-Joachim Braczyk, and Martin Heidenreich, 2–25. London: Routledge.
21. Dasgupta, Partha, and Paul A. David. 1994. “Toward a New Economics of Science.” Research Policy 23(5): 487–521.
22. de Solla Price, Derek J. 1965. “Networks of Scientific Papers.” Science 149(3683): 510–515.
23. de Wit, Hans. 2002. Internationalization of Higher Education in the United States of America and Europe: A Historical, Comparative, and Conceptual Analysis. Westport, CT: Greenwood Press.
24. Eagle, Nathan, Michael Macy, and Rob Claxton. 2010. “Network Diversity and Economic Development.” Science 328(5981): 1029–1031.
25. Ernst, Dieter, and Linsu Kim. 2002. “Global Production Networks, Knowledge Diffusion, and Local Capability Formation.” Research Policy 31(8–9): 1417–1429.
26. Etzkowitz, Henry, and Loet Leydesdorff. 2000. “The Dynamics of Innovation: From National Systems and “Mode 2” to a Triple Helix of University–Industry–Government Relations.” Research Policy 29(2): 109–123.
27. Frenken, Koen, Sjoerd Hardeman, and Jarno Hoekman. 2009. “Spatial Scientometrics: Towards a Cumulative Research Program.” Journal of Informetrics 3(3): 222–232.
28. Galison, Peter, and Bruce William Hevly (eds.). 1992. Big Science: The Growth of Large-Scale Research. Stanford, CA: Stanford University Press.
29. Garfield, Eugene. 1955. “Citation Indexes for Science: A New Dimension in Documentation through Association of Ideas.” Science 122(3159): 108–111.
30. Gertler, Meric S. 2003. “Tacit Knowledge and the Economic Geography of Context, or the Undefinable Tacitness of Being (There).” Journal of Economic Geography 3(1): 75–99.
31. Geuna, Aldo, and Ben R. Martin. 2003. “University Research Evaluation and Funding: An International Comparison.” Minerva 41(4): 277–304.
32. Glänzel, Wolfgang, and András Schubert. 2005a. “Analysing Scientific Networks through Co-Authorship.” In Handbook of Quantitative Science and Technology Research, ed. Henk F. Moed, Wolfgang Glänzel, and Ulrich Schmoch, 257–276. Heidelberg: Springer.
33. Glänzel, Wolfgang, and András Schubert. 2005b. “Domesticity and Internationality in Co-authorship, References and Citations.” Scientometrics 65(3): 323–342.
34. Haas, Peter M. 1992. “Introduction: Epistemic Communities and International Policy Coordination.” International Organization 46(1): 1–35.
35. Hennemann, Stefan. 2011. “The Role of Transnational Corporations in the Chinese Science and Technology Network.” Erdkunde 65(1): 69–81.
36. Hennemann, Stefan. 2012. “Evaluating the Performance of Geographical Locations within Scientific Networks Using an Aggregation–Randomization–Re-Sampling Approach (ARR).” Journal of the American Society for Information Science and Technology 63(12): 2393–2404.
37. Hennemann, Stefan, Diego Rybski, and Ingo Liefner. 2012. “The Myth of Global Science Collaboration – Collaboration Patterns in Epistemic Communities.” Journal of Informetrics 6(2): 217–225.
38. Hicks, Diana M., and J. Sylvan Katz. 1996. “Science Policy for a Highly Collaborative Science System.” Science and Public Policy 23(1): 39–44.
39. Hoekman, Jarno, Koen Frenken, and Robert J.W. Tijssen. 2010. “Research Collaboration at a Distance: Changing Spatial Patterns of Scientific Collaboration within Europe.” Research Policy 39(5): 662–673.
40. Hoekman, Jarno, Koen Frenken, and Frank van Oort. 2009. “The Geography of Collaborative Knowledge Production in Europe.” The Annals of Regional Science 43(3): 721–738.
41. Jansen, Dorothea, Regina von Görtz, and Richard Heidler. 2010. “Knowledge Production and the Structure of Collaboration Networks in Two Scientific Fields.” Scientometrics 83(1): 219–241.
42. Jones, Benjamin F., Stefan Wuchty, and Brian Uzzi. 2008. “Multi-University Research Teams: Shifting Impact, Geography, and Stratification in Science.” Science 322(5905): 1259–1262.
43. Kafouros, Mario I., Peter J. Buckley, John A. Sharp, and Chengqi Wang. 2008. “The Role of Internationalization in Explaining Innovation Performance.” Technovation 28(1): 63–74.
44. Knight, Jane. 2004. “Internationalization Remodeled: Definition, Approaches, and Rationales.” Journal of Studies in International Education 8(1): 5–31.
45. Leydesdorff, Loet, and Caroline S. Wagner. 2008. “International Collaboration in Science and the Formation of a Core Group.” Journal of Informetrics 2(4): 317–325.
46. Liefner, Ingo. 2003. “Funding, Resource Allocation, and Performance in Higher Education Systems.” Higher Education 46(4): 469–490.
47. Lundvall, Bengt-Åke (ed.). 1992. “National Systems of Innovation: Towards a Theory of Innovation and Interactive Learning.” London: Pinter Publishers.
48. Luukkonen, Terttu, Olle Persson, and Gunnar Sivertsen. 1992. “Understanding Patterns of International Scientific Collaboration.” Science, Technology & Human Values 17(1): 101–126.
49. Marginson, Simon. 2006. “Dynamics of National and Global Competition in Higher Education.” Higher Education 52(1): 1–39.
50. Maslov, Sergei, and Kim Sneppen. 2002. “Specificity and Stability in Topology of Protein Networks.” Science 296(5569): 910–913.
51. Moodysson, Jerker. 2008. “Principles and Practices of Knowledge Creation: On the Organization of ‘Buzz’ and ‘Pipelines’ in Life Science Communities.” Economic Geography 84(4): 449–469.
52. Narin, F., K. Stevens, and Edith S. Whitlow. 1991. “Scientific Co-operation in Europe and the Citation of Multinationally Authored Papers.” Scientometrics 21(3): 313–323.
53. Nelson, Richard (ed.). 1993. National Innovation Systems: A Comparative Analysis. Oxford: Oxford University Press.
54. Newman, Mark E.J., and Juyong Park. 2003. “Why Social Networks Are Different from Other Types of Networks.” Physical Review E 68: 036122-8.
55. North, Douglass C. 1990. Institutions, Institutional Change, and Economic Performance. Cambridge: Cambridge University Press.
56. Nooteboom, Bart, Wim Vanhaverbeke, Geert Duysters, Victor Gilsing, and A.J. van den Oord. 2007. “Optimal Cognitive Distance and Absorptive Capacity.” Research Policy 36(7): 1016–1034.
57. Pavitt, Keith. 1998. “The Social Shaping of the National Science Base.” Research Policy 27(8): 793–805.
58. Persson, Olle. 2010. “Are Highly Cited Papers More International?” Scientometrics 83(2): 397–401.
59. Ponds, Roderik. 2009. “The Limits to Internationalization of Scientific Research Collaboration.” Journal of Technology Transfer 34(1): 76–94.
60. Ponomariov, Branco L., and P. Craig Boardman. 2010. “Influencing Scientists’ Collaboration and Productivity Patterns Through New Institutions: University Research Centers and Scientific and Technical Human Capital.” Research Policy 39(5): 613–624.
61. Powell, Walter W., and Paul J. DiMaggio. 1991. The New Institutionalism in Organizational Analysis. Chicago, IL: University of Chicago Press.
62. Power, Dominik, and Anders Malmberg. 2008. “The Contribution of Universities to Innovation and Economic Development: In What Sense a Regional Problem?” Cambridge Journal of Regions, Economy and Society 1(2): 233–245.
63. Smith, Richard. 2006. “Peer Review: A Flawed Process at the Heart of Science and Journals.” Journal of the Royal Society of Medicine 99(4): 178–182.
64. Suresh, Subra. 2011. “Moving Toward Global Science.” Science 333(6044): 802.
65. Thorsteinsdóttir, Hella, Christina C. Melon, Monali Ray, et al. 2010. “South-South Entrepreneurial Collaboration in Health Biotech.” Nature Biotechnology 28(5): 407–416.
66. Tijssen, Robert J.W., Ludo Waltman, and Nees Jan van Eck. 2011. “Collaborations Span 1,553 Kilometres.” Nature 473(7346): 154–154.
67. van Rijnsoever, Frank J., Laurens K. Hessels, and Rens L.J. Vandeberg. 2008. “A Resource-Based View on the Interactions of University Researchers.” Research Policy 37(8): 1255–1266.
68. Waltman, Ludo, Robert J.W. Tijssen, and Nees Jan van Eck. 2011. “Globalisation of Science in Kilometres.” Journal of Informetrics 5(4): 574–582.
69. Watts, Duncan J., and Steven H. Strogatz. 1998. “Collective Dynamics of ‘Small-World’ Networks.” Nature 393: 440–442.
70. Weisberg, Michael, and Ryan Muldoon. 2009. “Epistemic Landscape and the Division of Cognitive Labor.” Philosophy of Science 76(2): 225–252.
71. Ynalvez, Marcus Antonius, and Wesley M. Shrum. 2011. “Professional Networks, Scientific Collaboration, and Publication Productivity in Resource-Constrained Research Institutions in a Developing Country.” Research Policy 40(2): 204–216.
72. Yook, Soon-Hyung, Hawoong Jeong, and Albert-Lászlo Barabási. 2002. “Modeling the Internet’s Large-Scale Topology.” Proceedings of the National Academy of the USA 99(21): 13382–13386.
73. Zitt, Michael, Elise Bassecoulard, and Yoshiko Okubo. 2000. “Shadows of the Past in International Cooperation: Collaboration Profiles of the Top Five Producers of Science.” Scientometrics 47(3): 627–657.