Localization of collaborations in knowledge creation


This study investigates the localization of collaboration in knowledge creation by using data on Japanese patent applications. Applying distance-based methods, we obtained the following results. First, collaborations are significantly localized at the 5% level with a localization range of approximately 100 km. Second, the localization of collaboration is observed in most technologies. Third, the extent of localization was stable from 1986 to 2005 despite extensive developments in information and communications technology that facilitate communication between remote organizations. Fourth, the extent of localization is substantially greater in inter-firm collaborations than in intra-firm collaborations. Furthermore, in inter-firm collaborations, the extent of localization is greater in collaborations with small firms. This result suggests that geographic proximity mitigates the firm-border effects on collaborations, especially for small firms.

This is a preview of subscription content, log in to check access.

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6


  1. 1.

    Duranton and Overman (2005) and Nakajima et al. (2012) show that nearly half of the manufacturing industries in the UK and Japan tend to be localized when using distance-based methods.

  2. 2.

    Crescenzi et al. (2016) is an exception. They use inventor-level data and estimate the impact of geographic and other proximities on collaboration in the EU by regression approach. They identify the roles of organizational proximity and of complementarity between several proximity measures and geographical proximities with regard to collaboration. While they focus on estimating the impact of geographical proximity on the probability of collaborations between these inventors, we adopt nonparametric approach to identify the strength and the distance range of localizations of collaboration. Furthermore, as is explained later, we use a finer measure of location of inventions in which are actually occurred.

  3. 3.

    This concept is the same as the “control patent approach” that is often used in the literature on the localization of patent citations (e.g., Jaffe et al. 1993; Thompson and Fox-Kean 2005).

  4. 4.

    We use Japanese patent data. Japanese patent system is much different from other major patent systems, except the convention of inventor’s address registered, which is explained later. Indeed, the Japanese patent system has been harmonized to the US and the EU through the European Patent Convention in 1977 and the Patent Cooperation Treaty in 1978. The details of the comparison of the patent systems across these countries are summarized in Japan Patent Office (2017). For a brief survey of the Japanese patent system, see the annual reports from the Japan Patent Office (Japan Patent Office 2014). In addition, Maskus and McDaniel (1999) provide a helpful overview of the Japanese patent system.

  5. 5.

    One may raise concerns about the transfer of workers. Consider the case in which an inventor transfers from an establishment during the process of collaboration, and she registers her new establishment on the patent application form. In this case, the observed distance between the two establishments that are registered on the patent does not necessarily represent the actual collaborating distance. To check this issue, we estimate the frequency of the inventor’s transfer for the average duration of invention. To identify the inventor’s transfer by patent information, addressing the “same name problem,” i.e., that different researchers have same name, is necessary. Saito and Yamauchi (2015) proposed focusing on the unique names observed in telephone directories. Based on the methodology, we restrict samples to researchers who have “rare names,” and we define movers as researchers who published patents in at least two establishments. We find that 31.1% of researchers transfer their establishments in the 20 years from 1986 to 2005. Suzuki (2011) found that it takes a median of 18 months from starting a project to applying for a patent in Japan. Thus, we roughly estimate that 2.75% (= 1–(1–0.311)18/12/20) of workers transferred establishments during the invention. On the other hand, share of researcher who registered at least one patent published with collaborators belonging in different establishments (which is “collaboration” in our definition in the analysis) in 18 months is 23.16%. Therefore, the magnitude of the pseudo-collaborations through worker transfer in our collaboration data is inconsequential.

  6. 6.

    A patent often has multiple IPCs. We use the primary IPC that is assigned to each patent.

  7. 7.

    For a robustness check, we conduct analyses using a more conservative definition for potential collaborating partners that restrict the establishments that have at least one collaboration experience. However, the results remained qualitatively unchanged. See Appendix 2.

  8. 8.

    One may argue that establishments can choose their locations by considering the expectations of future collaborations, and if this is true, these establishments choose their locations where many potential collaborators are located. In this case, the counterfactual collaborations will be more localized than the collaborations of the establishments that choose their location regardless of the existence of future collaborators. Counterfactual distribution may include the establishments proximity by the above motivation, and it is interpreted that our estimation strategy identifies the degree of localization of collaborations given in the location distribution of establishments made by the above location choice strategy.

  9. 9.

    We use great-circle distance, which is the shortest distance along the surface of the globe.

  10. 10.

    Silverman’s (1986) optimal bandwidth and Gaussian kernel function are used as default in the K-density methodology. To maintain comparability with the previous literature (e.g., Duranton and Overman 2005; Nakajima et al. 2012, Murata et al. 2014), we use the default setting.

  11. 11.

    Following Duranton and Overman (2005) and Nakajima et al. (2012), we set 180 km as the upper bound of our focus distance. The choice of the upper bound does not qualitatively change our results.

  12. 12.

    One may raise the concern that a significant number of firms might only have two establishments. In this case, potential collaborations are unique and there are no variations. If there is a collaboration between the two establishments, the actual collaboration always coincides with the potential one. However, there are a significant number of firms that have only two establishments and there is no collaboration within those firms. If distance impedes collaborations, two distant establishments tend not to collaborate within a firm; on the other hand, two close establishments tend to collaborate with each other. Thus, in firms that have only two establishments, we still have enough variations to identify the relationship between distance and collaborations. We find that within firms that have more than one establishment, 27% of firms have more than two establishments. Thus, the identification strategy explained above plays a substantial role in our estimation results.

  13. 13.

    To confirm the role of geographic proximity, we conducted interviews with firms that have experience with collaborative works. The interviews were conducted in March 2016 in a provincial city in Japan. Managers of these firms mentioned that geographical proximity promotes trust between firms through monitoring and spillovers of each firm’s information and facilitates collaboration between firms.

  14. 14.

    Using a firm database (Tokyo Shoko Research database), we confirm that this categorization has a strong positive correlation with firm employments (0.40) and sales (0.41). However, one may raise the concern that the “single-firm” includes large-sized foreign-owned companies that have only one research establishment in Japan. From the Survey of Trends in Business Activities of Foreign Affiliates by the Ministry of Economy, Trade, and Industry which is the complete survey of foreign-owned companies (companies for which 1/3 of their stock is owned by foreign companies or investors), we confirm there are just 3185 foreign-owned companies included in the survey in 1998. On the other hand, the number of “single firms” in our data is 46,904. In this sense, the effect of foreign-owned companies on our results is mostly negligible.


  1. Agarwal S, Hauswald R (2010) Distance and private information in lending. Rev Financ Stud 23(7):2757–2788

    Article  Google Scholar 

  2. Akcigit U, Caicedo S, Miguelez E, Santcheva S, Sterzi V (2018) Dancing with the stars: innovation through interactions. NBER Working paper, No. 24466

  3. Arbia G, Espa G, Quah D (2008) A class of spatial econometric methods in the empirical analysis of clusters of firms in the space. Empir Econ 34(1):81–103

    Article  Google Scholar 

  4. Arzaghi M, Henderson V (2008) Networking off Madison Avenue. Rev Econ Stud 75(4):1011–1038

    Article  Google Scholar 

  5. Berliant M, Fujita M (2008) Knowledge creation as a square dance on the Hilbert Cube. Inter Econ Rev 49(4):1251–1295

    Article  Google Scholar 

  6. Boschma R (2005) Proximity and innovation: a critical assessment. Reg Stud 39(1):61–74

    Article  Google Scholar 

  7. Buzard K, Carlino G, Hunt R, Carr J, Smith T (2017) The agglomeration of American R&D Labs. J Urban Econ 101:14–26

    Article  Google Scholar 

  8. Cairncross F (2001) The death of distance: how the communications revolution is changing our lives. Harvard Business Press, Cambridge

    Google Scholar 

  9. Coe D, Helpman E (1995) International R&D spillovers. Eur Econ Rev 39(5):859–887

    Article  Google Scholar 

  10. Crescenzi R, Nathan M, Rodoriguez-Pose A (2016) Do inventors talk to strangers? On proximity and collaborative knowledge creation. Res Policy 45:177–194

    Article  Google Scholar 

  11. Duranton G, Overman H (2005) Testing for localization using micro-geographic data. Rev Econ Stud 72(4):1077–1106

    Article  Google Scholar 

  12. Ellison G, Glaeser E, Kerr W (2010) What causes industry agglomeration? Am Econ Rev 100(3):889–927

    Article  Google Scholar 

  13. Gomes-Casseres B, Hagedoorn J, Jaffe A (2006) Do alliances promote knowledge flows? J Financ Econ 80(1):5–33

    Article  Google Scholar 

  14. Goto A, Motohashi K (2007) Construction of a Japanese patent database and a first look at Japanese patenting activities. Res Policy 36(9):1431–1442

    Article  Google Scholar 

  15. Griffith R, Lee S, Van Reenen J (2011) Is distance dying at last? Falling home bias in fixed-effects models of patent citations. Quant Econ 2(2):211–249

    Article  Google Scholar 

  16. Häusler J, Horn HW, Lütz S (1994) Contingencies of innovative networks: a case study of successful interfirm R&D collaboration. Res Policy 23(1):47–66

    Article  Google Scholar 

  17. Hoekman J, Frenken K, van Oort F (2009) The geography of collaborative knowledge production in Europe. Ann Reg Sci 43:721–738

    Article  Google Scholar 

  18. Hoekman J, Frenken K, Tijssen R (2010) Research collaboration at a distance: changing spatial patterns of scientific collaboration within Europe. Res Policy 39:662–673

    Article  Google Scholar 

  19. Huo D, Motohashi K (2015) Understanding two types of technological diversity and their effects on the technological value of outcomes from bilateral inter-firm R&D alliances. RIETI Discussion Paper Series 15-E-06

  20. Inoue H, Nakajima K, Saito YU (2014) Localization of knowledge-creating establishments. Jpn World Econ 43:23–29

    Article  Google Scholar 

  21. Inoue H, Nakajima K, Saito YU (2015) Innovation and collaboration patterns between research establishments. RIETI Discussion Paper Series, 15-E-049

  22. Jaffe A, Trajtenberg M, Henderson R (1993) Geographic localization of knowledge spillovers as evidenced by patent citations. Q J Econ 108(3):577–598

    Article  Google Scholar 

  23. Japan Patent Office (2014) Japan Patent Office Annual Report 2014, Tokyo

  24. Japan Patent Office (2017) Comparative Research on the Patent Systems of Japan, the United States and Europe, Tokyo

  25. Katz S, Martin B (1997) What is research collaboration? Res Policy 26(1):1–18

    Article  Google Scholar 

  26. Keller W, Yeaple S (2013) The gravity of knowledge. Am Econ Rev 103(4):1414–1444

    Article  Google Scholar 

  27. Kerr W, Kominers S (2015) Agglomerative forces and cluster shapes. Rev Econ Stat 97(4):877–899

    Article  Google Scholar 

  28. Lerner J (1995) Venture capitalists and the oversight of private firms. J Finance 50(1):301–318

    Article  Google Scholar 

  29. Lychagin S, Pinkse J, Slade M, Van Reenen J (2016) Spillovers in space: does geography matter? J Ind Econ 64(2):295–335

    Article  Google Scholar 

  30. Madsen J (2007) Technology spillover through trade and TFP convergence: 135 years of evidence for the OECD countries. J Int Econ 72(2):464–480

    Article  Google Scholar 

  31. Marcon E, Puech F (2003) Evaluating the geographic concentration of industries using distance-based methods. J Econ Geogr 3(4):678–693

    Article  Google Scholar 

  32. Marcon E, Puech F (2010) Measures of the geographic concentration of industries: improving distance-based methods. J Econ Geogr 10(5):745–762

    Article  Google Scholar 

  33. Marshall A (1920) Principles of economics. Macmillan, London

    Google Scholar 

  34. Maskus KE, McDaniel C (1999) Impacts of the Japanese patent system on productivity growth. Jpn World Econ 11(4):557–574

    Article  Google Scholar 

  35. Murata Y, Nakajima R, Okamoto R, Tamura R (2014) Localized knowledge spillovers and patent citations: a distance-based approach. Rev Econ Stat 96(5):967–985

    Article  Google Scholar 

  36. Nakajima K, Saito YU, Uesugi I (2012) Measuring economic localization: evidence from Japanese firm-level data. J Jpn Int Econ 26(2):201–220

    Article  Google Scholar 

  37. OECD (2008) Compendium of Patent Statistics. OECD, Paris

    Google Scholar 

  38. Østergaard C, Timmermans B, Kristinsson K (2011) Does a different view create something new? The effect of employee diversity on innovation. Res Policy 40(3):500–509

    Article  Google Scholar 

  39. Pittaway L, Robertson M, Munir K, Denyer D, Neely A (2004) Networking and innovation: a systematic review of the evidence. Int J Manag Rev 5/6(3&4):137–168

    Article  Google Scholar 

  40. Ponds R, van Oort F, Frenken K (2007) The geographical and institutional proximity of research collaboration. Pap Reg Sci 86(3):423–444

    Article  Google Scholar 

  41. Rosenthal S, Strange W (2001) The determinants of agglomeration. J Urban Econ 50(2):191–229

    Article  Google Scholar 

  42. Rosenthal S, Strange W (2003) Geography, industrial organization, and agglomeration. Rev Econ Stat 85(2):377–393

    Article  Google Scholar 

  43. Saito YU, Yamauchi I (2015) Inventors’ mobility and organizations’ productivity: evidence from Japanese rare name inventors. RIETI Discussion Paper Series, 15-E-128

  44. Saxenian A (1994) Regional advantage: culture and competition in Silicon Valley and Route 128. Harvard University Press, Cambridge

    Google Scholar 

  45. Silverman B (1986) Density estimation for statistics and data analysis. Chapman and Hall, New York

    Google Scholar 

  46. Suzuki J (2011) The structural characteristics of research and development by Japanese companies, and issues for the future. RIETI Discussion Paper Series, 11-J-002

  47. Thompson P, Fox-Kean M (2005) Patent citations and the geography of knowledge spillovers: a reassessment. Am Econ Rev 95(1):450–460

    Article  Google Scholar 

  48. Wuchty S, Jones B, Uzzi B (2007) The increasing dominance of teams in production of knowledge. Science 316(5827):1036–1039

    Article  Google Scholar 

Download references


This study is conducted as a part of the Project “Inter-organizational and Inter-inventors Geographical Proximity and Networks” undertaken at Research Institute of Economy, Trade and Industry (RIETI). We thank two anonymous referees for their useful suggestions that significantly improved the paper. We thank Victor Couture, Jonathan Dingel, Gilles Duranton, Tatsuaki Kuroda, Yasusada Murata, Sadao Nagaoka, Ryosuke Okamoto, Yukako Ono, and the participants at the UEA in Ottawa, the ARSC at Toyama University, and at seminars at Fukuoka University, Hitotsubashi University, Kobe University, Nihon University, RIETI, Temple University, and the University of Tokyo for their helpful comments. We also thank the Center for Spatial Information Science, University of Tokyo, for providing us with geocoding services. We gratefully acknowledge financial support from the Japan Society for the Promotion of Science (Nos. 15K01217, 16H02018, 16K13367, 17H02508, 17H02514, 17H02517, 17H02518, 18H00859, and 18K04615). We gratefully acknowledge NIRA for providing funding and setting up interviews with patent-publishing companies.

Author information



Corresponding author

Correspondence to Kentaro Nakajima.


Appendix 1: Methodology of extracting establishments

We identify the establishments and their collaborations from the Japanese patent database. Here, we describe the methodology in detail.

For each patent, we conduct the following procedure. First, we identify whether the patent is applied for by firms. This identification is performed using an applicant’s name. If the applicant’s name includes the term “company limited” (in Japanese, “kabushikigaisha”), the applicant is defined as a firm. This definition simultaneously excludes all relatively small firms, such as private limited companies. Second, we identify the establishment information. We check whether the inventor’s address includes the firm’s name. If the inventor’s address includes the firm name, we consider the inventor’s address to be the address of the firm’s establishment. In practice, there are many different written forms in addresses. Thus, we use the geocoding service provided by the Center for Spatial Information Science of the University of Tokyo to convert addresses to longitudes and latitudes. This definition of establishments by latitude and longitude simultaneously aggregates two different establishments in the same building.

In some patents, the firm name is not included in the inventor’s address, and there are three possible causes for this. First, the inventor registers their establishment address but does not include the firm name. In this case, we try to match the address to the establishment address that includes a firm name. Second, the inventor’s residential address is registered, although the inventor works in an establishment. Third, the inventor does not belong to the firm. Because there is no solution for these last two cases, we ignore them. We check the fractions of inventors whose establishments can be identified. We define a unique inventor by the pair of inventor’s name and address information. In total, we find 1,216,718 inventors registered as inventors on patents whose applicants are firms. Of them, we ultimately identify 873,633 inventors’ establishments to which the inventors belong. This implies that we can identify and use the 71.8% of inventors’ establishments information attached to the patents for which the firms applied. Considering that researchers in public research institutes and universities are also involved in patents applied for by firms (these would be included in the third case above), we can identify the information of a substantial number of inventors belonging to firms. Out of the 873,633 inventors, we identify 97,374 inventors whose addresses do not include firm names but can be matched to establishment addresses collected by other inventors’ addresses, which corresponds to the first case of the above three cases. The remaining inventors (28.2%) correspond to the second and third cases.

Appendix 2: A more conservative definition of potential partners

The key idea of our analysis is to control for potentially collaborative establishment locations. Therefore, our results may depend on the definition of potential collaborators. In this appendix, we assess the robustness of our results by adopting a more conservative definition of potential partners.

In our main analysis, we define a potential collaborating partner as an establishment that has applied for at least one patent in technology class \( A \). It does not matter whether the establishment has experience in collaborative work. However, our dataset includes many establishments that have never applied for collaborative patents with other establishments. These establishments may not be potential collaborators. To control for this possibility, we now adopt another definition of potential collaborators. We restrict the establishments to those that have applied for at least one collaboration patent in technology class \( A \in {\mathfrak{A}} \). Below, we describe the results by using the definition of potential collaborating partners and counterfactual collaborations.

Figure 7 shows the baseline result. Similar to Fig. 1, the collaborative relations are statistically localized with a localization range of approximately 85 km. Our baseline results remain unchanged, although a more conservative definition of potential collaborators is applied.

Fig. 7

All collaborations

Panels (a) and (b) of Fig. 8 show the results. Similar to the baseline results shown in panels (a) and (b) of Fig. 4, the localizations of both collaborations are statistically significant, and the range of localization is shorter in inter-firm collaborations.

Fig. 8

K-densities of collaborating relations. a Intra-firm collaborations. b Inter-firm collaborations

Finally, we check the results for firm size. Figure 9 shows the results. Similar to the baseline results that are shown in Fig. 6, in every collaboration pattern (single–single, multiple–multiple, and single–multiple), stronger collaboration localizations are found for small firms, and all collaboration localizations are statistically significant at the 5% level. Furthermore, the range of localization is shorter for small firms.

Fig. 9

K-densities of inter-firm collaborations. a Single–single collaborations. b Multiple–multiple collaborations. c Single–multiple collaborations

Our baseline results remain completely unchanged, although a more conservative definition of potential collaborators than the originally adopted definition is applied. Localization is statistically significant for establishments that collaborate to conduct innovative activities, and we observe firm-border effects, particularly with respect to small firms.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Inoue, H., Nakajima, K. & Saito, Y.U. Localization of collaborations in knowledge creation. Ann Reg Sci 62, 119–140 (2019). https://doi.org/10.1007/s00168-018-0889-y

Download citation

JEL Classification

  • R12
  • O31