Multiscale analysis of interfirm networks in regional clusters $ Yuya Kajikawa a,b,n , Yoshiyuki Takeda a , Ichiro Sakata a,b , Katsumori Matsushima b a Institute of Engineering Innovation, School of Engineering, the University of Tokyo, 2-11-16 Yayoi, Bunkyo-ku, Tokyo 113-8656, Japan b Department of Technology Management for Innovation, Graduate School of Engineering, the University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-8656, Japan article info Keywords: Interfirm network Regional cluster Multiscale modeling Small world Module abstract Networks within an organization and also among organizations are expected to work as a conduit of resources and knowledge. In this paper, we construct a large database of interfirm networks in eight regional clusters in Japan, and investigate their multiscale structures combining existing organizational and network theories. We found that the small world properties among regional clusters show marked differences, which intensify as the network size evolves. We also found that connector firms bridging different modules are scarce and therefore we should focus on networking activity among different modules to support the development of well-connected networks. & 2009 Elsevier Ltd. All rights reserved. 1. Introduction In the last decade, there has been a widespread resurgence of interest in the economics of industrial locations, particularly in the issue of regional clusters (Porter, 1990; Gordon and McCann, 2000; Morosini, 2004; D¨ oring and Schnellenbach, 2006; John and Pouder, 2006). Innovative milieu (Aydalot and Keeble, 1988; Camagni, 1991, 1995), technology districts (Storper, 1992, 1993), and regional innovation systems (Cooke and Morgan, 1994; Cooke, 2001) are also used as regional innovation models although they tend to be used for different focuses, contexts, and applications. In these models, innovation is associated with places where relevant resources are easily accessed by firms in close proximity. Some regions have superior innovative capabil- ities, as evidenced by the localized production of patents (Jaffe et al., 1993; Acs et al., 2002; Frenken et al., 2005). Porter (1998) argued that enduring competitive advantages in the current global economy lie increasingly in local things – knowledge, relationships, motivation – that distant rivals cannot match, while companies in a global economy can source capital, goods, information and technology from around the world. Silicon Valley and the Route 128 zone of Boston (Saxenian, 1991, 1994), Cambridge (Segal Quince Wicksteed, 2000), Baden-W ¨ urttemberg (Cooke and Morgan, 1994) and‘‘Third Italy’’ (Brusco, 1982; Piore and Sabel, 1984) are typical example of such distinguished regions. Regional clusters can offer more opportunities for innovation than scattered locations, which is typically driven by reduced transaction cost (Fujita and Thisse, 1996), access to venture capitalists (Florida and Kenney, 1988; Powell et al., 2002; Stuart and Sorenson, 2003), local labor market pooling (Almeida and Kogut, 1999), entrepreneurial activity within the region (Sorenson and Audia, 2000; A ´ cs and Varga, 2005), enhancement of knowl- edge diffusion (Maskell, 2001; Tallman et al., 2004), and localized learning (Morgan, 1997; Maskell and Malmberg, 1999; Lorenzen, 2001). Already in 1920, Marshall (1920) developed the notion of ‘‘industrial districts’’ as agglomerations of firms operating in one industry sector within a well-defined and relatively small geographic area. On the other hand, regional clusters are defined as ‘‘geographic concentrations of interconnected companies and institutions in a particular field’’ (Porter, 1998). Regional clusters are distinguished from pure agglomerations by their intercon- nected nature, i.e. clusters are characterized as collaborative networks and concentrations of collaboration and competition, which offer significant opportunities and stimulate economic development (Porter, 1998). Another characteristic of regional clusters is the diversity of actors contained within. According to Porter (1990, 1998, 2000), an industrial cluster includes suppliers, consumers, peripheral industries, governments, and supporting institutions such as universities. In sum, the network among actors is the key to understanding the performance of regional clusters (Steinle and Schiele, 2002). Saxenian (1994) found that Hewlett Packard and other Silicon Valley firms improved their performance by developing long- term partnerships with physically proximate suppliers. She claims that proximity in high-technology industries facilitates the ARTICLE IN PRESS Contents lists available at ScienceDirect journal homepage: www.elsevier.com/locate/technovation Technovation 0166-4972/$ - see front matter & 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.technovation.2009.12.004 $ This study was performed through Special Coordination Funds for Promoting Science and Technology of the Ministry of Education, Culture, Sports, Science and Technology, the Japanese Government. n Corresponding author at: Institute of Engineering Innovation, School of Engineering, the University of Tokyo, 2-11-16 Yayoi, Bunkyo-ku, Tokyo 113- 8656, Japan. Tel./fax: + 81 3 5841 7672. E-mail address: kaji@biz-model.t.u-tokyo.ac.jp (Y. Kajikawa). Technovation 30 (2010) 168–180