Regional Innovation and Networks in Japan (International Perspectives in Geography, 16) 981162190X, 9789811621901

Citation preview

International Perspectives in Geography AJG Library 16

Yutaka Yokura

Regional Innovation and Networks in Japan

International Perspectives in Geography AJG Library Volume 16

Editor-in-Chief Yuji Murayama, The University of Tsukuba, Tsukuba, Japan Series Editors Yoshio Arai, Teikyo University, Utsunomiya, Japan Hitoshi Araki, Ritsumeikan University, Kusatsu, Japan Shigeko Haruyama, Mie University, Tsu, Japan Yukio Himiyama, Hokkaido University of Education, Sapporo, Japan Mizuki Kawabata, Keio University, Tokyo, Japan Taisaku Komeie, Kyoto University, Kyoto, Japan Jun Matsumoto, Tokyo Metropolitan University, Tokyo, Japan Takashi Oguchi, The University of Tokyo, Kashiwa, Japan Toshihiko Sugai, The University of Tokyo, Kashiwa, Japan Atsushi Suzuki, Rissho University, Kumagaya, Japan Teiji Watanabe, Hokkaido University, Sapporo, Japan Noritaka Yagasaki, Nihon University, Tokyo, Japan Satoshi Yokoyama, Nagoya University, Nagoya, Japan

Aim and Scope The AJG Library is published by Springer under the auspices of the Association of Japanese Geographers. This is a scholarly series of international standing. Given the multidisciplinary nature of geography, the objective of the series is to provide an invaluable source of information not only for geographers, but also for students, researchers, teachers, administrators, and professionals outside the discipline. Strong emphasis is placed on the theoretical and empirical understanding of the changing relationships between nature and human activities. The overall aim of the series is to provide readers throughout the world with stimulating and up-to-date scientific outcomes mainly by Japanese and other Asian geographers. Thus, an “Asian” flavor different from the Western way of thinking may be reflected in this series. The AJG Library will be available both in print and online via SpringerLink. About the AJG The Association of Japanese Geographers (AJG), founded in 1925, is one of the largest and leading organizations on geographical research in Asia and the Pacific Rim today, with around 3000 members. AJG is devoted to promoting research on various aspects of human and physical geography and contributing to academic development through exchanges of information and knowledge with relevant internal and external academic communities. Members are tackling contemporary issues such as global warming, air/water pollution, natural disasters, rapid urbanization, irregular land-use changes, and regional disparities through comprehensive investigation into the earth and its people. In addition, to make the next generation aware of these academic achievements, the members are engaged in teaching and outreach activities of spreading geographical awareness. With the recent developments and much improved international linkages, AJG launches the publication of the AJG Library series in 2012.

More information about this series at http://www.springer.com/series/10223

Yutaka Yokura

Regional Innovation and Networks in Japan

Yutaka Yokura Faculty of Economics Kyushu University Fukuoka, Japan

ISSN 2197-7798 ISSN 2197-7801 (electronic) International Perspectives in Geography ISBN 978-981-16-2190-1 ISBN 978-981-16-2191-8 (eBook) https://doi.org/10.1007/978-981-16-2191-8 © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2021 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Singapore Pte Ltd. The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721, Singapore

Preface

Networks and innovation both continue to be hot topics in the field of economic geography. Current research focusing on the process of regional innovation and knowledge creation highlights local linkages and extra-local networks. The main aim of this book is to empirically examine how various relationships affect the development and growth of industrial agglomeration. This book includes eight chapters that examine the various networks and the innovation process in Japan. Although most of the chapters draw on my own work previously published in Japanese academic journals, all of the content has been updated. Chapters 1 and 2 explore existing research on networks and innovation and present new perspectives on economic geography. Chapters 3 through 7 present empirical research on networks and innovation, which is conducted through case studies of industrial agglomerations in Japan. In this book, in addition to the interview survey, other novel tools for analyzing networks and innovation, such as social network analysis and covariance structure analysis, are applied for empirical analysis. Chapter 8 summarizes the future research prospects based on the results of the theoretical and empirical studies presented in this book. Parts of Chap. 1, 2, and 3 were written as my doctoral thesis for the University of Tokyo. I am indebted to my dissertation chair, Professor Hiroshi Matsubara, for his insight, useful comments, and great support. I am sincerely grateful to my dissertation co-chairs, Professor Yoshio Arai, Professor Junji Nagata, Professor Isao Mizuno, and Associate Professor Shin Kajita for their helpful advice that improved my work. I would also like to acknowledge the support of my colleagues and students at Kyushu University. Much of the research in this book was supported by JSPS KAKENHI Grant Numbers JP23720399, JP26770281, and JP17K03239. Finally, I want to thank my parents for their long-term support of my studies. Fukuoka, Japan

Yutaka Yokura

v

Contents

1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1 Regional Innovation and Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2 The Social Network Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.3 The Governance Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.4 The Spatial Scale of Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.4.1 Local Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.4.2 Non-Local and Global Networks . . . . . . . . . . . . . . . . . . . . . . . 1.5 New Research Trends Related to Networks . . . . . . . . . . . . . . . . . . . . . 1.5.1 Trust . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.5.2 Channels of Knowledge Flow . . . . . . . . . . . . . . . . . . . . . . . . . . 1.6 Towards a Novel Perspective on Networks and Innovation in Industrial Agglomerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1 1 2 3 3 4 4 5 5 6

2 Regional Innovation and Knowledge Creation . . . . . . . . . . . . . . . . . . . . 2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2 Spatial Dimensions of Innovation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.1 Innovation on a Local Scale . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.2 Innovation on a National Scale . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.3 Innovation on a Global Scale . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.4 Brief Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3 The Knowledge Base Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4 The Strength of Temporary Relationships . . . . . . . . . . . . . . . . . . . . . . 2.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

11 11 13 13 14 16 17 18 19 22 23

3 R&D Networks and Regional Innovation . . . . . . . . . . . . . . . . . . . . . . . . . 3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2 The Structure of R&D Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.1 Data and Social Network Analysis . . . . . . . . . . . . . . . . . . . . . . 3.2.2 Sectoral and Regional Structures of R&D Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3 The Spatial Dimension of Collaborative R&D Networks . . . . . . . . .

27 27 28 28

7 8

31 31 vii

viii

Contents

3.4 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Local Trade Fairs as Temporary Clusters: A Case Study of the Suwa Area Industrial Messe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2 Development of Japanese Trade Fairs . . . . . . . . . . . . . . . . . . . . . . . . . 4.3 Economic Context of the Suwa Area . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4 Building Relationships at the Suwa Area Industrial Messe . . . . . . . . 4.4.1 Goals of Attendance and State Support . . . . . . . . . . . . . . . . . . 4.4.2 Building Relationships With Non-Local Firms . . . . . . . . . . . 4.4.3 Processes of Relationship Building at Suwa Area Industrial Messe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Informal Networks and the Evolution of Industry: A Case Study of the Hamamatsu Area . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2 Economic Context of the Hamamatsu Area . . . . . . . . . . . . . . . . . . . . . 5.3 The Development of Informal Networks in the Hamamatsu Area . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3.1 Institutional Support System for Informal Networks . . . . . . . 5.3.2 Relational Structure of Informal Networks . . . . . . . . . . . . . . . 5.4 Relationships Between Informal and Formal Networks . . . . . . . . . . . 5.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Institutional Thickness in Regional Innovation Ecosystem: A Case Study of the Kyushu Semiconductor Industry . . . . . . . . . . . . . . 6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2 Kyushu’s Semiconductor-Related Industry . . . . . . . . . . . . . . . . . . . . . 6.3 Various Roles of International Semiconductor-Related Workshops . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.4 Formation of Transactional Relationships at Business-Matching Project . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.5 Toward Sustainable Support for Semiconductor Innovation Ecosystem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

42 42 45 45 47 50 52 54 55 56 58 60 61 61 63 66 66 67 72 76 77 79 79 81 82 87 90 91

7 Global Knowledge Flows and Corporate Values . . . . . . . . . . . . . . . . . . . 93 7.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93 7.2 Global Knowledge Flows and Inter-firm Relationship Structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95 7.2.1 Analytical Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95 7.2.2 Analytical Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95 7.2.3 Sectoral Comparison of Relationship Structures . . . . . . . . . . 100

Contents

ix

7.2.4 Using the Block Model to Contract the Relationship Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.2.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.3 Relationship Between Network Attributes and Corporate Value . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.3.1 Analytical Framework of Prior Studies Using Covariance Structure Analysis . . . . . . . . . . . . . . . . . . . . . . . . . 7.3.2 Simultaneous Analysis of Multiple Populations . . . . . . . . . . . 7.3.3 Comparison of Path Coefficients Among Sectors . . . . . . . . . 7.4 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

110 111 113 115 119

8 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.1 Networks and Innovation in Industrial Agglomeration . . . . . . . . . . . 8.2 Future Directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

121 121 123 124

105 107 109

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125

Chapter 1

Introduction

Abstract This book provides a novel perspective on networks and innovation in the field of economic geography. In this chapter, we present an overview of current research to establish the network concept as part of the economic geography field. Networks are a source of growth and innovation and are essential basic concepts in industrial agglomerations. Interdisciplinary research on networks is currently being conducted from various perspectives, including two perspectives in existing network theory. One examines where actors are situated using a social network perspective. The other perspective is centered on the process of governance, which reduces uncertainty. However, neither offers spatial implications; therefore, we distinguished network patterns in the three spatial dimensions occupied by networks. Additionally, we presented new research themes related to intangible networks, such as trust and knowledge flow. Keywords Industrial clusters · Social networks · Trust · Knowledge flow

1.1 Regional Innovation and Networks Recently, networks and innovation have become hot topics in the field of economic geography. Networks as a source of growth and innovation at the basis of industrial agglomerations are essential concepts, and interdisciplinary research is currently being conducted from a variety of perspectives. However, no consensus has yet been reached among researchers regarding the definition of networks because the concept itself includes a wide range of analytical fields. Conceptually, networks might be represented by physical networks that exist as part of infrastructure, such as those in the field of telecommunications or transportation. In addition, there are also more qualitative types of networks, such as personal networks and inter-organizational networks, that are not easily observable. In Japan’s economic geography field, the roles of networks and various actors1 in the knowledge creation and innovation process are attracting attention because of the influence of the “relational turn” (Bathelt and Glückler 2003) and “new agglomeration theory” (Scott 1988) in Western economic geography. Mizuno (2011, 2018) © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2021 Y. Yokura, Regional Innovation and Networks in Japan, International Perspectives in Geography 16, https://doi.org/10.1007/978-981-16-2191-8_1

1

2

1 Introduction

discussed the diversity and liquidity of actors in networks as catalysts for innovation. He criticized social network theory, which lacks a spatial dimension, and argued that introducing the concept of “proximity2 ” between actors was necessary to study networks from the perspective of economic geography. As observed in various “collaborations” such as open innovation (Chesbrough 2003), collective learning (Keeble and Wilkinson 1999), mergers and acquisitions (M&As), and strategic alliances, firm boundaries have become blurred in the contemporary economy. Therefore, it has likewise become important to establish a framework for analyzing these new economic phenomena from the perspective of economic geography. According to Grabher and Powell (2004), network research tends to proceed along two main perspectives—the social network perspective and the governance perspective. The social network approach investigates the performance of a collection of actors and their positions in a network. The governance approach examines “the institutional mechanisms by which networks are initiated, coordinated, monitored, recombined, and terminated” (Grabher and Powell 2004).

1.2 The Social Network Perspective Research on social networks has dramatically expanded due to the recent emergence of social network analysis as a distinct academic discipline. In 2007, a special issue on social network analysis appeared in the prestigious Japanese academic journal on business administration, Organizational Science. Yamada et al. (2007) explored a unique network of various actors, called “Kumi3 ,” in the Japanese film industry by analyzing data from 267 movies and the 3,170 people involved in their production. They verified that a cohesive network was formed in which tacit knowledge was easily shared through a dense set of relationships likely to give rise to innovation. Nakano (2007) criticized the fact that existing industrial agglomeration research focused only on descriptive, qualitative data but lacked a quantitative network structure approach. He argued that social network analysis is able to handle a large amount of complex data and, using transactional data from 5,111 businesses located in Ota-ku, Tokyo, created a network visualization of a supplier system. Some scholars (Cantner and Graf 2006; Graf 2006) have used social network analysis to illuminate knowledge flow4 as a horizontal relationship. They have examined networks that have created innovation as determined by the analysis of joint patent application relationships in Jena, Germany. Owen-Smith and Powell (2004) also visualized network structures by exploring information and knowledge transfer paths. There is a rich accumulation of research on social networks in economic geography (Yokura et al. 2013). Ter Wal and Boschma (2009) identified the challenges of using primary (i.e., survey) and secondary (i.e., patent) data for social network analysis and evaluated the discipline as having great potential for the research of industrial clusters and regional innovation systems. From data obtained from interviews, Giuliani and Bell (2005) applied social network analysis to knowledge flow

1.2 The Social Network Perspective

3

in a Chilean wine cluster. The originality of their research lies in their introduction of the spatial perspective into social network analysis.

1.3 The Governance Perspective A governance perspective in business administration “seeks to answer how to design, manage, and control networks in order to reduce uncertainties and improve competitive position” (Grabher and Powell 2004: xiii). In Japan, research from the governance perspective is being enthusiastically pursued in inter-organizational relations theory. Wakabayashi (2006) determined that the negative aspects of inter-organizational relationships—such as cartels, conspiracies, and the oligopolistic control of large firms—were emphasized in North American economic theory. However, in recent years, the positive aspects of long-term cooperation between firms, as seen in Japan, have been reconsidered and appreciated. Yamakura (1993) and Aldrich (1999) summarized the historical transition of perspectives in inter-organizational relations and evaluated the transaction cost perspective (Coase 1937; Williamson 1975) as providing an important paradigm for analysis. The transaction cost perspective has significantly influenced Japanese supplier systems research. Asanuma (1989) analyzed the competitive advantage conferred by long-term relationships between “core” firms and suppliers, using the Japanese automobile and electrical/electronic equipment industries as examples. He dubbed the supplier capability formed by repeated interactions with specific core firms as a “relation-specific skill” and determined it to be a source of competitive advantage. Although we have examined the research trends on networks from two perspectives, few of the research streams consider spatial implications, with a few exceptions from economic geography such as the work of Giuliani and Bell (2005). In the next section, we discuss the spatial scale of networks and refine the analytical framework for industrial agglomerations and inter-organizational relationships.

1.4 The Spatial Scale of Networks To avoid confusion in this analysis of networks, this chapter mainly discusses organizations5 as network components and excludes studies that have employed “cities” as a network node (i.e., synergy between cities). In the following discussion, representative research trends will be examined based on differences in spatial scales of networks.

4

1 Introduction

1.4.1 Local Networks Much of the previous research that has explored inter-organizational networks on the local scale6 has been heavily influenced by Marshallian industrial districts (Marshall 1890). Prime examples are studies of industrial clusters such as Third Italy (Piore and Sable 1984) and Silicon Valley (Saxenian 1994). Some economic geographers (Scott 1988; Scott and Storper 1987) have suggested “new industrial spaces” based on flexible specialization (Piore and Sable 1984) and have conducted empirical analyses of the interdependence between local small and medium-sized enterprises. Much research on the Japanese production system also exists, such as Toyota’s just-intime production system and local vertical relationships in which suppliers tend to be spatially concentrated around large firms. Other Japanese economic geographers (Matsuhashi 2002, 2005; Yamamoto and Matsuhashi 1999) have performed empirical analysis examining cooperation and collaboration between organizations in local agglomerations. In these studies, the “extra-firm network” concept (Yeung 1994) has been emphasized as key for explaining the competitive advantage of industrial clusters. Matsuhashi (2002, 2005) investigated the learning and innovation process in local industrial agglomerations and highlighted the social and institutional thickness supporting various actors, taking the electrical and electronic industry in the Tohoku region of Japan as an example. This research consists of empirical studies targeting local interorganizational networks as social capital. Many studies have revealed that networks that foster innovation feature non-local characteristics. In the following section, we will discuss research on trans-local relationships.

1.4.2 Non-Local and Global Networks The typology of industrial districts is helpful in understanding non-local networks among organizations. In the “satellite type” of industrial district (Markusen 1996; Park and Markusen 1995), relationships with the head office and factories outside of the cluster are more important than within the cluster. Park (2005) argued that local and non-local networks embedded in the region are a source of innovation, using as an example the case of Gangnam-gu and Sunchang-gun in South Korea. He posited that the combination of local human resources and non-local institutions (e.g., formal and informal conferences) was important for regional advancement. The innovative milieu approach developed by French and Italian scholars (Aydalot and Keeble 1988; Camagni 1991) also emphasizes non-local networks. Relationships with actors outside of the region are important as a source of development or innovation in the local milieu. In the context of uncertainty and rapid changes in technology and markets, non-local business networks are crucial to a local economy (Keeble et al. 1999). Some scholars (Gordon and McCann 2000; Markusen 2003) have criticized

1.4 The Spatial Scale of Networks

5

that the innovative milieu theory is supported by little empirical evidence but, in recent years, research that more fully conveys the actual situation is being conducted. Needless to say, non-local networks also appear on the global scale, such as the transnational community theory that examines the international transfer of knowledge. Saxenian and Hsu (2001) studied long-distance knowledge transfer between Hsinchu, Taiwan and Silicon Valley, in the United States (US), using the case of Taiwanese companies that have grown significantly based on their role as an original equipment manufacturer (OEM). They argued that Chinese entrepreneurs in Silicon Valley were the key actors that linked their native Taiwan to Silicon Valley, thereby creating a transnational community. Yeung and Olds eds. (2000) also emphasized the importance of transnational networks in addressing the problems of Chinese companies managing cross-border business. In network research with firms as the analytical unit, the research focus shifted from the local scale to the various relationships that arose from globalization, but qualitative research is still more common than quantitative, empirical analysis in the field. There is also little discussion about the content and quality of networks. In the next section, we will discuss new research trends related to the more intangible aspects of networks, such as trust and knowledge flow.

1.5 New Research Trends Related to Networks 1.5.1 Trust Organizationalists and economic sociologists have theoretically and empirically examined networks and trust. Murphy stated that “while there is widespread recognition of trust’s importance and role in facilitating regional development, technology transfer, and agglomeration economies, it remains rather undertheorized within economic geography and regional science” (2006: 428). The mutual trust created by long-term relationships has been identified as the source of Japanese companies’ international competitiveness. According to Dore (1983) and Sako (1992), the “goodwill trust” established between customers and suppliers promotes inter-organizational learning through cooperative relationships. This type of trust is reciprocal and is embedded in relationships rather than in arm’slength contracts. On the other hand, trust does not only exist in long-term relationships. Trust forged within a temporary organization such as a project or a trade fair also provides a new perspective for consideration. A project is a temporary system in which the end of the relationships between the participating actors is institutionally defined. Therefore, there is no mutual sharing of experience and familiarity before the project’s formation. According to Grabher, projects “seem to lack the normative structures and institutional safeguards that minimize the likelihood of failure” (2001a: 1329). Some

6

1 Introduction

authors (Meyerson et al. 1996) have pointed out that “swift trust” is built within the project to accomplish the assigned work. Grabher (2001b, 2002a) conducted empirical studies examining swift trust forged in projects, using advertising agencies in Soho, London as an example. By switching members for each project, novel knowledge was acquired and innovation was created. While trust based on long-term relationships is typical of collaboration in manufacturing industries like the automotive industry, the swift trust built in short-term projects is characteristic of cultural industries such as content creation industries like filmmaking, advertising, and publishing, which are more often agglomerated in urban areas. These research results provide a new analytical framework for the consideration of networks.

1.5.2 Channels of Knowledge Flow Since the mid-1990s, economic geographers have been focusing on the complex nexus of relationships among various actors causing dynamic changes in the spatial organization of economic activity. In economic geography, such shifts in research trends are called relational turns. The emphasis of this research has evolved from considering only economic linkages, such as transactional networks based on business relationships, to non-economic linkages, such as collective learning and buzz7 communities. According to Bathelt, a relational approach in economic geography “integrates economic and social, cultural, institutional and political aspects of human agency” (2006: 234). Bathelt et al. (2004) illuminated the utility of various relationships within and outside of industrial clusters as channels for the flow of knowledge. They focused on the role of “local buzz” and “extra-local or global pipelines” as catalysts for a cluster’s knowledge creation activity. The concept of buzz refers to rumors and casual conversations distributed locally (Storper and Venables 2004) that, in a broad sense, can be considered as tacit knowledge. Bathelt et al. (2004) argued that accidental buzz promotes good communication, collaborative problem-solving, trust, and mutual relationship development, which enables interactive learning among local actors. Therefore, similar to an “industrial atmosphere” (Marshall 1890), a buzz-distributing community tends to be connected to a locality. On the other hand, the global pipeline is a channel used in knowledge creation that is not completed locally. Some researchers (Owen-Smith and Powell 2004; Cooke et al. 2007) are conducting empirical studies on knowledge transfer between distant places through strategic alliances and joint research and development (R&D) activity using the biotechnology industry as an example. Recently, the accumulation of research on the interaction between local buzz and global pipelines can be seen in the case of trade shows (Bathelt and Zeng eds. 2015; Bathelt et al. 2014). Such studies provide a rich perspective for understanding the relationship between the growth of industrial clusters and extra-local networks.

1.6 Towards a Novel Perspective on Networks …

7

1.6 Towards a Novel Perspective on Networks and Innovation in Industrial Agglomerations There is much interdisciplinary research from various perspectives on networks that adopt complex and diverse network-related concepts. In this introduction, we tried to present an overview of current research to establish the network concept as part of the economic geography field. There are two perspectives in existing network theory. One examines where actors are situated using a social network perspective. The other perspective is centered on the process of governance, which reduces uncertainty. However, neither offers spatial implications, so we attempted to distinguish network patterns in the three spatial dimensions occupied by networks. In addition, we presented new research themes related to intangible networks, such as trust and knowledge flow. The aim of this book is to provide a novel perspective on networks and innovation in the field of economic geography. Chapter 2 considers the mechanism of the innovation process in which networks function as institutions and establishes this book’s theoretical foundation. In Chap. 3 and beyond, we conduct empirical research centered on statistical data analysis such as social network analysis and covariance structure analysis, as well as determine the real-world situation through interview surveys of related actors. We analyze various temporary systems in industrial agglomerations such as project-based joint R&D (Chaps. 3 and 5), trade fairs (Chap. 4), business workshops (Chap. 5), and international conferences (Chap. 6). Chapter 3 considers the structure and spatial patterns of R&D networks in Japan by making the structure visible and calculating network indices. In Chap. 4, we concentrate on local trade fairs held in industrial agglomerations and examine the development of various relationships between actors. Chapters 5 and 6 shed light on institutional thickness in industrial agglomerations. Chapter 7 spotlights quantitative and metrical examinations of inter-corporate relationships in terms of knowledge flows based on company-level data regarding technological alliances and ownership relationships between global corporations. The studies featured in Chaps. 6 and 7 serve to evaluate how Japan’s firms have adapted to radical changes under global competition. Chapter 8 summarizes the results of this book and presents future possible research topics. Notes 1. 2.

3. 4.

In this book, actors are defined as those who act with individual consciousness, such as human agencies and firms. The concept of proximity is used when discussing the distances between actors. Chapter 2 of this book focuses on the concept of organizational and cognitive proximity. Kumi is a term unique to the Japanese film industry, that means a production group centered on a film director. Knowledge flow means the spatial connection of knowledge that creates innovation (Matsubara 2007: 28).

8

5. 6.

7.

1 Introduction

In this book, the term “organization” means a body such as a firm, an association, or a public institution. Local scale means a geographically bounded spatial range based on certain characteristics (e.g. local government, local industry and cluster). Although the definition of a city varies from country to country, the term can be examined from the geographical extent of the local network. According to Bathelt et al., the definition of buzz “refers to the information and communication ecology created by face-to-face contacts, co-presence and co-location of people and firms within the same industry and place or region” (2014: 38).

References Aldrich HE (1999) Organizations evolving. Sage, London Asanuma B (1989) Manufacturer-supplier relationships in Japan and the concept of relation-specific skill. J Jpn Int Econ 3:1–30 Aydalot P, Keeble D (eds) (1988) High technology industry and innovative environments: the European experience. Routledge, London Bathelt H (2006) Geographies of production: growth regimes in spatial perspective 3—toward a relational view of economic action and policy. Prog Hum Geogr 30:223–236 Bathelt H, Glückler J (2003) Toward a relational economic geography. J Econ Geogr 3:117–144 Bathelt H, Golfetto F, Rinallo D (2014) Trade shows in the globalizing knowledge economy. Oxford University Press, New York Bathelt H, Malmberg A, Maskell P (2004) Clusters and knowledge: local buzz, global pipelines and the process of knowledge creation. Prog Hum Geogr 28:31–56 Bathelt H, Zeng G (eds) (2015) Temporary knowledge ecologies: the rise and evolution of trade fairs in Asia-Pacific. Edward Elgar, Cheltenham Camagni R (ed) (1991) Innovation networks: spatial perspectives. Belhaven Press, London Cantner U, Graf H (2006) The network of innovators in Jena: an application of social network analysis. Res Policy 35:463–480 Chesbrough H (2003) Open innovation: the new imperative for creating and profiting from technology. Harvard Business Review Press, Boston Coase HR (1937) The nature of the firm. Economica 4:386–405 Cooke P, Laurentis CD, Tödtling F, Trippl M (2007) Regional knowledge economies: markets, clusters and innovation. Edward Elgar, Cheltenham Dore R (1983) Goodwill and the spirit of market capitalism. Br J Sociol 34:459–482 Giuliani E, Bell M (2005) The micro-determinants of meso-level learning and innovation: evidence from a Chilean wine cluster. Res Policy 34:47–68 Gordon IR, McCann P (2000) Industrial clusters: complexes, agglomeration and/or social networks? Urban Stud 37:513–532 Grabher G (2001a) Commentaries. Environ Plan A 33:1329–1334 Grabher G (2001b) Ecologies of creativity: the village, the group, and the heterarchic organisation of the British advertising industry. Environ Plan A 33:351–374 Grabher G (2002) The project ecology of advertising: tasks, talents, and teams. Reg Stud 36:245–262 Grabher G, Powell WW (2004) Introduction. In: Grabher G, Powell WW (eds) Networks I. Edward Elgar, Cheltenham Graf H (2006) Networks in the innovation process: local and regional interactions. Edward Elgar, Cheltenham

References

9

Keeble D, Lawson C, Moore B, Wilkinson F (1999) Collective learning processes, networking and ‘institutional thickness’ in the Cambridge region. Reg Stud 33:319–332 Keeble D, Wilkinson F (1999) Collective learning and knowledge development in the evolution of regional clusters of high technology SMEs in Europe. Reg Stud 33:295–303 Markusen A (1996) Sticky places in slippery spaces: a typology of industrial districts. Econ Geogr 72:293–313 Markusen A (2003) Fuzzy concepts, scanty evidence, policy distance: the case for rigour and policy relevance in critical regional studies. Reg Stud 37:701–717 Marshall A (1890) Principles of economics. Macmillan, London Matsubara H (2007) Chishiki no kukanteki ryudo to chiikiteki innovation system (Spatial knowledge flow and regional innovation system). Tokyo Daigaku Jinbun Chirigaku Kenkyu (Komaba Stud Hum Geogr) 18:22–43 (in Japanese) Matsuhashi K (2002) Yonezawa-shi ni okeru denki/denshi kogyo wo meguru syakaiteki kankyo network (Development of extrafirm networks of the electronics industry in Yonezawa city). Sundai Shigaku (Sundai Hist Rev) 115:57–96 (in Japanese with English abstract) Matsuhashi K (2005) Hi-daitoshiken no sangyo shusekichiiki ni okeru chushokigyo no network tenkai no igi (Implications of SMEs’s networks proposing collective learning in agglomeration areas of non-metropolitan region: case studies on Yonezawa City in Yamagata prefecture, Kitakami city and Hanamaki city in Iwate prefecture). Keizaichirigaku Nempo (Ann Jpn Assoc Econ Geogr) 51:329–347 (in Japanese with English abstract) Meyerson D, Weick KE, Kramer RM (1996) Swift trust and temporary groups. In: Kramer RM, Tyler TR (eds) Trust in organizations: frontiers of theory and research. Sage Publications, Thousand Oaks Mizuno M (2011) Innovation no Keizai Kuukan (Economics spaces of innovation). Kyoto Daigaku Shuppan-kai, Kyoto (in Japanese) Mizuno M (2018) Sangyo shuseki no shinka to kinsetsusei no dynamics: chishiki gakushu to network no shiten kara (Cluster evolution and proximity dynamics: from a learning and network perspective). Shirin (J His) 101:261–292 (in Japanese with English abstract) Murphy JT (2006) Building trust in economic space. Prog Hum Geogr 30:427–450 Nakano T (2007) An integrating core of large-scale, regional cluster: assessment of some Japanese stylized facts from network analysis. Soshiki Kagaku (Organ Sci) 40(3):55–65 (in Japanese) Owen-Smith J, Powell WW (2004) Knowledge networks as channels and conduits: the effects of spillovers in the Boston biotechnology community. Organ Sci 19:549–583 Park SO (2005) Network, embeddedness, and cluster processes of new economic spaces in Korea. In: Le Heron R, Harrington JW (eds) New economic spaces: new economic geographies. Ashgate, Aldershot Park SO, Markusen A (1995) Generalizing new industrial districts: a theoretical agenda and an application from a non-Western economy. Environ Plan A 27:81–104 Piore MJ, Sabel CF (1984) The second industrial divide: possibilities for prosperity. Basic Books, New York Sako M (1992) Prices, quality, and trust: inter-firm relations in Britain and Japan. Cambridge University Press, Cambridge Saxenian A (1994) Regional advantage: culture and competition in Silicon Valley and Route 128. Harvard University Press, Cambridge Saxenian A, Hsu JY (2001) The Silicon Valley-Hsinchu connection: technical communities and industrial upgrading. Ind Corp Change 10:893–920 Scott AJ (1988) New industrial spaces: flexible production organization and regional development in North America and Western Europe. Pion, London Scott AJ, Storper M (1987) High technology industry and regional development: a theoretical critique and reconstruction. Int Soc Sci J 112:215–232 Storper M, Venables AJ (2004) Buzz: face-to-face contact and the urban economy. J Econ Geogr 4:351–370

10

1 Introduction

Ter Wal ALJ, Boschma RA (2009) Applying social network analysis in economic geography: framing some key analytic issues. Ann Reg Sci 43:739–756 Wakabayashi N (2006) Nihon kigyo no network to shinrai: kigyokan kankei no atarashii keizai shakaigakuteki bunseki (Network and trust in Japanese inter-organizational relationships). Yuhikaku, Tokyo (in Japanese) Williamson OE (1975) Markets and hierarchies: analysis and antitrust implications-a study in the economics of internal organization. Free Press, New York Yamada J, Yamashita M, Wakabayashi N, Kanki N (2007) Social capital for highly performing film projects: the case of Japanese filmmakers’ networks. Soshiki Kagaku (Organ Sci) 40(3):41–54 (in Japanese) Yamakura K (1993) Soshikikan kankei: kigyokan network no henkaku ni mukete (Interorganizational relationships: towards the reformation of interfirms networks). Yuhikaku, Tokyo (in Japanese) Yamamoto K, Matsuhashi K (1999) Chusho kigyo shuseki chiiki ni okeru network keisei: Suwa Okaya chiiki no jirei (Networking in an industrial area characterized by agglomeration of SMEs). Keizai Shirin (Hosei Univ Econ Rev) 66(34):85–182 Yeung HWC (1994) Critical reviews of geographical perspectives on business organizations and the organization of production: towards a network approach. Prog Hum Geogr 18:460–490 Yeung HWC, Olds K (eds) (2000) Globalization of chinese business firms. St. Martin’s Press, New York Yokura Y, Matsubara H, Sternberg R (2013) R&D networks and regional innovation: a social network analysis of joint research projects in Japan. Area 45:493–503

Chapter 2

Regional Innovation and Knowledge Creation

Abstract The traditional myth that codified knowledge can be transferred over long distances but tacit knowledge embodied in human beings that requires face-to-face contact can hinder the current understanding of more diversified innovations. In this chapter, we reviewed trends in regional innovation studies from three perspectives: the spatial scale of innovation, the knowledge base approach, and temporary organizations. To understand the spatiality of innovation processes, it is useful to divide the concept into three dimensions: global, national, and sub-national innovation. We categorize network structures, which are considered to be the foundation of innovation, and discuss the role of extra-regional relationships. This study also examines innovation in terms of the economic performance of inter-organizational networks. Additionally, this chapter considers the development of temporary organizations and their role in innovation. Keywords Spatiality · Innovation systems · Knowledge base · Temporary clusters

2.1 Introduction The importance of tacit knowledge has gained importance in the academic discussion of industrial agglomerations. The knowledge that is embodied in human beings and cannot be easily transferred has been considered as one of the primary factors giving rise to cluster establishment. In the field of economic geography, there is an abundance of research on the benefits of knowledge creation based on learning, such as untraded interdependencies (Storper 1997) and horizontal collaboration or competition (Malmberg 1996). Innovation research emphasizes the importance of non-local networks (Mizuno 2011; Yokura 2009). Boschma focused on geographical proximity’s impact on innovation, stating that because “networks are defined and demarcated in a non-territorial manner, it does not seem correct to assume that knowledge spillovers are spatially bounded” (Boschma 2005: 69). In this vein, Asheim and Herstad (2005) stressed sourcing non-local knowledge by developing inter-organizational connections such as project-based productions, strategic alliances, and user-customer relationships. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2021 Y. Yokura, Regional Innovation and Networks in Japan, International Perspectives in Geography 16, https://doi.org/10.1007/978-981-16-2191-8_2

11

12

2 Regional Innovation and Knowledge Creation

Today, in industries characterized by a rapid pace of technological progress such as the semiconductor and pharmaceutical industries, building networks with a variety of actors to reduce uncertainty and forging strategic alliances with both domestic and overseas firms has become necessary (Powell and Grodal 2005). Multinational enterprises innovate globally in response to various demands and market conditions and efficiently adopt different products and technologies using overseas research and development (R&D) (Narula and Zanfei 2005). In Japan’s economic geography research trends, the characteristics of knowledge essential for innovation are often emphasized. According to Mizuno (2005), to obtain novel knowledge for innovation, connectivity with geographically distant external actors is essential. For novel knowledge to circulate, industrial agglomerations require diversity and liquidity. The openness and structure of the network influence its ability to acquire novel knowledge (Mizuno 2007). The transfer and acquisition of tacit knowledge from remote areas are also important for innovation. According to Breschi and Lissoni (2001), “tacit messages can be sent over long distances by means of a variety of communication media” (989). Yamamoto (2005) argued that tacit knowledge does not have to be fixed in a specific place but can be transmitted between remote settings through communication and meetings. This implies that the spatiality of innovation requires further investigation. To understand the spatiality of innovation processes, it is useful to divide the concept into three spatial dimensions: global, national, and sub-national1 innovation (Matsubara 2007). Regarding innovation on the sub-national scale, various types of innovation are considered from those originating in local regional blocks to those created in urban areas. However, existing studies focus on local learning processes of actors within regions. According to Matsubara (2007), with the formation of overseas R&D by multinational enterprises, further analysis is required on innovation that accounts for the acquisition and utilization of knowledge on a global scale is required. Similarly, Bunnell and Coe (2001: 583) indicated that “one spatial scale will rarely be adequate for a full understanding of innovation processes” and that “a range of scales from the global to the regional/local” can provide an important lens for examining innovation. In this chapter, we categorize network structures, which are considered to be the foundation of innovation, and discuss the role of extra-regional relationships. This chapter presents the spatiality of innovation from the perspective of its local, national, and global dimensions. This study also examines innovation in terms of the economic performance of inter-organizational networks. Industry-specific knowledge bases (Asheim and Coenen 2005) have a great influence on the spatiality of innovation because the innovation process differs for each type of knowledge base and some industries can source key innovation knowledge locally while others need to search for knowledge on a national or global scale. In discussions on the knowledge creation that leads to innovation, the importance of mutual learning through temporary contact at MICE2 events such as conferences and trade fairs has recently drawn attention. This chapter considers the development of those new concepts and their role in innovation. Section 2.3 reviews existing

2.1 Introduction

13

studies on innovation from the perspective of three spatial dimensions (i.e., the local, national, and global scale). Section 3.3 describes the relationship between an industry-specific knowledge base and innovation. Section 3.4 clarifies the role of temporary communication between actors in innovation. In the last section, future research issues are considered.

2.2 Spatial Dimensions of Innovation 2.2.1 Innovation on a Local Scale Research has accumulated on local-scale innovation in regional innovation system theory (Cooke 1992) and local milieu theory by the GREMI3 group (Camagni [ed.] 1991). In this study, we define an innovation system as a network built by actors connected to each other according to various intellectual connections forged for the purpose of creating innovation. Innovation systems can be divided into three distinct types according to the network’s spatial dimension—regional, national, and global innovation systems. According to Tödtling and Trippl (2005), “innovation should be seen as an evolutionary, non-linear and interactive process, requiring intensive communication and collaboration between different actors” (1205) such as universities and public institutions. Innovation system theory emphasizes cooperation, trust, knowledge-sharing, and entrepreneurship between local actors and the defining regional characteristics known as the “innovation climate” (Matsubara 2007). The local milieu, the means for coordinating and governing an innovation system, is composed of physical elements (e.g., infrastructure), non-physical characteristics (e.g., knowledge), and institutional features (Matsubara 2006; Yamamoto 2005). Two types of knowledge transfer occur between the actors in a regional innovation system—formal, contract-based exchanges and informal, non-contract-based exchanges. Formal knowledge exchanges include market transactions, such as licensing and contract research (Tödtling et al. 2006), and joint research endeavors, such as industry-academia-government collaboration (Etzkowitz and Leydesdorff 2000; Etzkowitz 2008). Informal knowledge exchanges include observations made at conferences and trade fairs (Tödtling et al. 2006: 1048), industry “buzz” words circulating among people in the same region (Storper and Venables 2004; Bathelt et al. 2004), and the mobility of people who themselves embody knowledge in their entrepreneurship and job changes (Saxenian 1994, 2007). For an actor to efficiently perform informal knowledge exchange, it is necessary that they be situated in a local milieu. The role of the local milieu is to function as an external environment that promotes informal knowledge transfer.

14

2 Regional Innovation and Knowledge Creation

2.2.2 Innovation on a National Scale 2.2.2.1

Innovation Networks and Knowledge Pipelines

The national innovation system (NIS) theory emphasizes the role of the nation-state in fostering innovation (Lundvall 1992; Nelson [ed.] 1993). NISs consist of various components and relationships that are located in a certain country or that have a specific origin (Toda 2004). By interacting with these components, the output of radical innovation from other countries can be diffused throughout the nation-state. Institutions, organizations, governments, and policies all play an important role in advancing such innovation processes on a national scale. NISs are important because institutional design, organizational construction, and R&D investment can differ significantly between various NISs and policies affecting the innovation process are formulated and implemented on a national scale (Edquist 2005). Some scholars have established the role of the university as an important “institutional actor” in NISs and have argued that universities exhibit strong nationstate characteristics. Mowery and Sampat (2005) determined the economic returns from university-based R&D activities in Organization for Economic Cooperation and Development (OECD) countries and clarified that the policies regarding commercialization technologies, such as universities’ promotion of patents and licenses, are globally diffused. However, factors such as history, path dependence, and institutional embeddedness often make it difficult to imitate other countries’ policies. Based on GREMI’s innovation networks and knowledge pipelines (Owen-Smith and Powell 2004), understanding the role played by the local milieu in NISs emphasizes the nation-state framework. Camagni (ed.) (1991) attaches great importance to relationships with extra-regional actors as a growth factor for the local milieu. According to Keeble et al. (1999), the local milieu must be seen in conjunction with the parallel importance of wider inter-firm networks (326) to decrease the uncertainty of rapidly changing market opportunities. A trans-local innovation network includes strategic partnerships and licensing as well as customer-supplier relationships (Yamamoto 2005). Owen-Smith and Powell (2004) proposed the concept of the “knowledge pipeline,” which is a communication channel between actors that is utilized in non-local knowledge creation activities. They used the case of the Boston biotechnology industry to explain extra-regional knowledge-sourcing through strategic alliances between countries. Scholars like Bathelt et al. (2004) have argued that pipelines need to be built to reduce the cost of monitoring and controlling firms outside the local area. By building pipelines to facilitate extra-local interactions, it is possible to obtain information on new markets and technologies that is not available in an industrial cluster. Bathelt et al. noted that pipelines support industrial agglomeration’s “cohesion and strengthen the internal translation processes” (2004: 41). Innovation networks and knowledge pipelines both support interactions that function as the source of the extra-local knowledge necessary to promote agglomeration growth. Organizational proximity enables the establishment and development of

2.2 Spatial Dimensions of Innovation

15

such interactions by firms. Organizational proximity is defined as commonality and similarity of routines and practices between entities. Knowledge transfer can more easily be facilitated between closely related, organizationally proximate firms such as subsidiaries and affiliates (Mizuno 2007: 489).

2.2.2.2

Organizational Proximity and Cognitive Distance

The concept of cognitive distance can be used to effectively measure organizational proximity. According to Nooteboom, the “notion of cognitive distance derives from a social constructivist view of knowledge” (2006: 140), which means that mental proximity is rooted in shared values. Nooteboom et al. argued that “to the extent that people have developed along different life paths and in different environments, they interpret, understand and evaluate the world differently” (2007: 1017). Therefore, inter-organizational interpretations and value arrangements give rise to similarities between the routines and practices (i.e., organizational proximity) of firms. Although Nooteboom insisted that the “central task of organizations is to sufficiently reduce cognitive distance” (2006: 140), cognitive proximity does not immediately lead to innovation. Nooteboom posits that there is an optimal value for cognitive distance, which can be assessed as the “absorptive capacity” (Cohen and Levinthal 1990) of organizations. Absorptive capacity is defined as the required “organizational capabilities to assimilate information, internally distribute it and implement knowledge in design, development, production, and marketing” (Nooteboom 2006: 140). Absorptive capacity, which is based on mutual understanding between two organizations, improves as cognitive distance decreases. Conversely, the farther the cognitive distance, the greater the amount of knowledge that cannot be acquired on a daily basis and the higher its novelty value. Therefore, if the learning efficiency that enables innovation is expressed by the product of absorptive capacity and novelty value, cognitive distance has an inverted U-shaped effect, as illustrated in Fig. 2.1 (Nooteboom et al. 2007: 1018). Learning efficiency is highest at the intersection of absorptive capacity and novelty value, so this intersection represents the optimal cognitive distance. Investment in R&D increases absorptive capacity but does not affect novelty value. R&D investment shifts the intersection to the right, thereby increasing the optimal cognitive distance, illustrating the truism “the more one knows, the further away one has to look for novelty” (Nooteboom et al. 2007: 1031). Giuliani and Bell (2005) examined the knowledge system of a wine cluster in Chile, focusing on the firms’ absorptive capacities and cognitive distances. They categorized absorptive capacity into the following two types—meso- and microlevel. The absorptive capacity of a meso-level (cluster) is defined as “the capacity of a cluster to absorb, diffuse and exploit extra-cluster knowledge” and the absorptive capacity of a micro-level (firm) is defined as the capacity that “reflects the stock of knowledge accumulated within the firm, embodied in skilled human resources and accrued through in-house learning efforts” (Giuliani and Bell 2005: 49). Firms with a higher absorptive capacity tend to build intellectual relationships with extra-cluster

16

2 Regional Innovation and Knowledge Creation

Fig. 2.1 Optimal cognitive distance (Source Redrawn from Nooteboom et al. [2007: Fig. 2.1: 1018])

knowledge and firms with different absorptive capacities occupy various cognitive positions within an industrial cluster. The originality of this study is that they collected data illustrating the relationships between the firms using an interview survey and applied social network analysis to examine the structural positions of the actors.

2.2.3 Innovation on a Global Scale Innovation networks and knowledge pipelines can be extended to the global dimension. Knowledge transfer across borders can be hindered by country-specific historical factors, path-dependence, and institutional embeddedness. Conversely, institutional and cultural proximity (Gertler 2004), community of practice (Wenger et al. 2002), and network-based trust (Brown and Duguid 2001) are factors that promote the flow of knowledge globally. According to Bunnel and Coe (2001), the innovation system approach emphasizes dense local networks, while the extra-local networks that lead to innovation have been relatively neglected in the literature. They insist that the direction of research on global-scale innovation should investigate the social and cultural proximity of extra-local networks. Cultural proximity creates external economies of scale, reduces transaction costs, promotes collective learning processes, and enhances flexibility (Boschma et al. 2002). Gertler (2004) investigated the transfer of corporate

2.2 Spatial Dimensions of Innovation

17

best practices between countries with different institutions and cultures in a case study about a technology transfer to North America from a German machine tool manufacturer. He identified the problems faced by North American user companies such as labor market systems, labor relations, and corporate governance in Germany and North America. Malmberg (2003) criticized the industry cluster theory for overemphasizing the strength of local interactions and instead emphasized the significance of global connections. He empirically determined the following four phenomena related to industrial clusters. Firstly, global transactional relationships dominate local networks; secondly, formal collaborations follow global value chains; thirdly, the degree of local rivalries varies; and fourthly, many firms perceive labor mobility in the local milieu as a problem rather than an advantage (153). Even though people are globally dispersed, tacit knowledge can be shared by a community of practice, which fosters mutual recognition and trust, enabling mutual learning to proliferate globally (Malmberg 2003: 157–158). Fromhold-Eisebith (2007: 218) identified three distinct dimensions of innovation systems (i.e., international, national, and regional) and argued that mutually supportive interdependencies exist. In a discussion about innovations’ sectoral systems (Malerba 2002), she proposed establishing a national supersystem of innovation (NSSI) as a concept to functionally and comprehensively capture the three distinct spatial dimensions of innovation systems. NSSI is the extension of the discussion on innovation systems to a global scale, providing a new perspective. According to Fromhold-Eisebith (2007: 226), national authorities act as “masters of scales” in promoting innovation and possess various competencies required for coordination and institution-building. She also insisted that close interactions with local organizations are also essential for innovation and the need for an NSSI to provide opportunities to establish socially embedded collaborations and information exchanges. High-tech industries like the information and communications technology (ICT) industry are usually analyzed in NSSI research, so further empirical research is required.

2.2.4 Brief Summary Figure 2.2 illustrates the existing innovation research organized by spatial dimension. The relationships between the organizations depicted in the figure range from vertical relationships such as transactional relationships to horizontal relationships including joint R&D, licensing, and strategic alliances. In this chapter, the “milieu” concept, which is defined as an external environment that promotes informal knowledge transfer, is adopted as a spatial unit for understanding regional-scale innovation. In the discussion of innovation on a national scale, the framework of the nation-state, which is the agent of policy implementation, was emphasized. The concepts of organizational proximity and the knowledge pipeline are useful in understanding extra-local networks. In the discussion of innovation on

18

2 Regional Innovation and Knowledge Creation

Fig. 2.2 Spatial scale of innovation

the global scale, institutional frameworks, such as unique history and culture, play an important role in the knowledge creation process. An NIS can take various forms in each country. This also applies to innovation systems on both the local and global scale. However, the innovation system concept considers the global scale such as NSSI lacks the perspective of treating regions with such diverse institutions. Depending on the prominent industries in an industrial agglomeration, the geographical manifestation of its organizational relationships differs and innovation can be created in various spatial dimensions. In the next section, we will review innovation research centering on the differences between industrial knowledge bases.

2.3 The Knowledge Base Approach The knowledge creation process transverses networks in various spatial dimensions ranging from local to global. Therefore, the traditional myth that codified knowledge can be transferred over long distances but tacit knowledge embodied in human beings requires face-to-face contact can hinder the current understanding of more diversified innovations. Some scholars (e.g., Asheim and Coenen 2005; Asheim and Gertler 2005) have identified two distinct types of knowledge bases—synthetic and analytical—that are instrumental in the knowledge creation process. A synthetic knowledge base includes engineering knowledge based on inductive processes such as problem-solving, whereas an analytical knowledge base includes scientific knowledge based on deductive processes such as patents. According to Asheim and Gertler, firms’ innovation process is “shaped by their specific knowledge base, which tends to vary systematically by industrial sector” (2005: 295).

2.3 The Knowledge Base Approach

19

Asheim and Coenen (2006: 165) insisted that synthetic knowledge bases dominate plant engineering, advanced industrial machinery, and shipbuilding. The innovation process of such industries “tends to be oriented toward the efficiency and reliability of new solutions, or the practical utility and user-friendliness of products from the perspective of the customers” (Asheim and Gertler 2005: 295–296). Therefore, innovation tends to be incremental rather than sudden. Scientific knowledge is important for analytical knowledge bases. Innovation is based on codified, rational processes. Analytical knowledge bases are prevalent in biotechnology and information technology (Asheim and Gertler 2005: 296). Those industries have their own in-house R&D departments and rely heavily on the research output of universities and other institutions, emphasizing the necessity of cooperation between academia and industry. An analytical knowledge base can give rise to radical innovation. Asheim and Gertler (2005: 297–298) explained that the innovation process in analytically based industries tends to be spatially concentrated for the following three reasons. First, by sharing experience and mutual understanding, the circulation of knowledge is highly localized. Second, the labor market, which provides workers with a high level of education and lucrative employment opportunities, is unevenly distributed. Third, employers that provide a high quality of life and attract talented workers have become especially important in the creative industry. The innovation process depends heavily on the nature of the knowledge base on which the industry depends. Innovation activities tend to spatially concentrate depending on the “institutional thickness” of universities and industrial support organizations. Local inter-organizational relationships are emphasized in synthetically based industries, while non-local collaboration is required in the analytical knowledge-based industries in which innovation networks are more widespread. Asheim et al. (2007) proposed three types of knowledge bases—synthetic, analytical, and symbolic (see Table 2.1). Symbolic knowledge bases are “related to the aesthetic attributes of products, to the creation of designs and images, and to the economic use of various cultural artifacts” (664) and dominate cultural industries such as advertising, filmmaking, and publishing. Buzz plays an important role in the exchange of symbolic knowledge. The innovation process of industries that rely on symbolic knowledge bases tends to proceed based on project organizations. Project organizations are temporary systems in which the duration of the relationships between the constituent actors is determined systematically. In the next section, we examine in more depth the learning and innovation processes created by temporary inter-organizational relationships.

2.4 The Strength of Temporary Relationships Grabher (2001, 2002) conducted an empirical study on temporary project organizations in Soho, London’s advertising industry in which the concept of “learning by switching” was described. By changing each project’s participating actors, it is

20

2 Regional Innovation and Knowledge Creation

Table 2.1 Typology of knowledge bases Knowledge Base

Analytical

Synthetic

Symbolic

Innovation

Innovation by the creation of new knowledge

Innovation by the novel application or combination of existing knowledge

Innovation by recombining existing knowledge in novel ways

Input for Innovation

The importance of scientific knowledge often based on deductive processes and formal models

The importance of applied and problem-related knowledge (engineering), often acquired through inductive processes

The importance of reusing or challenging existing conventions

Interactions between Actors

Research collaboration Interactive learning between firms (e.g., with clients and R&D departments) and suppliers research organizations

Learning by interacting with the professional community, learning from youth/street culture or “fine” culture, and interacting with “border” professional communities

Knowledge and Technology

Dominance of codified knowledge due to documentation in patents and publications

Reliance on tacit knowledge, craft, and practical and search skills

Examples

Biotechnology and Industrial machinery information technology and shipbuilding

Dominance of tacit knowledge due to more definite know-how, craft, and practical skill

Advertising, filmmaking, and publishing

Source Based on Asheim et al. (2007: Table 2.1: 661) and Gertler (2008: 214)

possible to avoid developing the negative lock-in that often arises in fixed relationships. Projects organizations make “access to know who highly relevant in finding out which actor has the technical skills needed, who is most innovative, whom one can collaborate with” (Asheim et al. 2007: 660), and the utilization of buzz is the best way to disseminate such information. According to Alderman (2005), in project-based productions, worker mobility plays a more significant role in knowledge creation than local embeddedness. Sydow and Staber (2002) explored how project networks function in television content production in two of Germany’s media regions. They explored the difference in institutional thickness between the two regions and determined that the flow of resources, information, and knowledge was facilitated by embedding temporary collaborative networks in the region. Bathelt (2005) highlighted the case of the Leipzig media cluster and project works and found that a mix of local and non-local transactions was necessary to promote sustainable growth of the media cluster.

2.4 The Strength of Temporary Relationships

21

Temporary projects can commonly be found in the construction industry (Eccles 1981) and the software industry (Grabher 2004) as well as in emerging industries such as the content industry and the cultural industry. In addition, project-based consortiums have been formed in high-tech industries such as biotechnology and nanotechnology. Torre and Rallet (2005) identified the concept of proximity and examined the interaction between geographical and organizational proximity in industrial clusters. They observed that “the need for geographical proximity is generally not permanent” (53) and argued that temporary geographical proximity, exemplified by shortterm business trips, enables organizations to exchange information. Tödtling et al. (2006: 1048) also noted that temporary inter-organizations such as trade fairs and conferences function as channels for informal knowledge exchange. Some scholars (Bathelt and Schuldt 2008; Maskell et al. 2006) have investigated temporary contacts among actors in global knowledge exchange and have observed that they are characterized by a knowledge exchange mechanism similar to permanent clusters or industrial agglomerations. The place in which such temporary interaction occurs is called a temporary cluster (Maskell et al. 2006: 999). The strength of temporary clusters lies in their ability to deal with market uncertainty, as is the case with project organizations. Figure 2.3 shows a knowledge creation process based on local buzz and trans-local knowledge pipelines in temporary and permanent clusters (Bathelt and Schuldt 2008: 856). According to Maskell et al. (2006), contemporary firms see international trade fairs as an opportunity to access distant markets, knowledge pools, and future business partners. By participating in trade fairs, firms can build trust between actors and form continuous cooperation for joint R&D and trade. Therefore, “temporary clusters are

Fig. 2.3 Pipeline creation and the complementary relationship between temporary and permanent clusters (Source Bathelt and Schuldt [2008: Fig. 2.1: 856])

22

2 Regional Innovation and Knowledge Creation

complementary to, rather than substitutes for, permanent” clusters (Maskell et al. 2006: 1005). During trade fairs, the actors participate in various interactions; for example, suppliers and customers exchange information about recent market trends and future product requirements. These relationships are vertical interactions (as shown in Fig. 2.3B). Therefore, international trade fairs can be the “source of information for adjustments in strategies and innovations, as well as for the establishment of new and the maintenance of existing pipelines” (Bathelt and Schuldt 2008: 856). As indicated by Tödtling et al. (2006), it is possible to observe and compare the products and strategies of competitors at trade fairs. These horizontal interactions reinforce firms’ future investment behavior and improved practices. Temporary relationships forged by project-based organizations and trade fairs of an institutionally fixed duration enable actors located in existing industrial clusters to acquire information and knowledge that is not locally available. Temporary clusters also play an important role in responding to rapidly changing markets and tracking competitors’ trends. For firms that cannot enjoy the benefit of geographical proximity, temporary relationships are essential for knowledge creation and building business transactions.

2.5 Conclusion In this chapter, we reviewed trends in innovation studies from three perspectives: (1) the spatial scale of innovation, (2) the knowledge base approach, and (3) temporary organizations. Table 2.2 demonstrates innovation factors according to knowledge base and cluster duration. In the joint R&D and production systems of the cultural industry, the knowledge creation process begins based on projects. Regardless of differences in knowledge bases, information and knowledge are exchanged at temporary places like trade fairs. In industrial agglomerations based on analytical knowledge, cognitive proximity promotes the formation of close networks between firms, universities, and public Table 2.2 Typology of innovation factors by knowledge base and cluster duration Knowledge Base

Analytical

Synthetic

Symbolic

Temporary cluster

• Project-based R&D • Trade fairs

• Interactive learning between suppliers and customers • Trade fairs

• Swift trust by one-off project production • Trade fairs

Permanent cluster

• Triple helix of universityindustry-government interactions • Highly skilled labor

• Reciprocal trust between suppliers and customers

• Local embeddedness by sharing project experience

2.5 Conclusion

23

institutions. Having a pool of highly skilled labor increases the local absorptive capacity and contributes to the advancement of industrial agglomerations. In industrial agglomerations based on synthetic knowledge, long-term rather than arm’slength relationships between firms are key to the development of mutually reciprocal relationships. In industrial agglomerations based on symbolic knowledge, relationships between participating actors established through common projects are embedded in the locality and give rise to the development of institutional thickness. Linkages or pipelines between organizations that are located on multiple spatial scales (i.e., from local to global) enable the shift from temporary to permanent clusters. In addition to the strength of local networking, the trans-local knowledge linkages between actors contribute to the successful development of industrial agglomerations. Notes 1. 2. 3.

In this book, the sub-national scale refers to the wide-area economic block and includes the regional scale, which is geographically wider than the local scale. The Meeting, Incentive tour, Convention/Congress, and Exhibition/Event industry is generally known as the MICE industry. GREMI is an acronym for “Groupe de Recherche Européen sur les Milieux Innovateurs” (Maskell and Kebir 2006: 44).

References Alderman N (2005) Mobility versus embeddedness: the role of proximity in major capital projects. In: Lagendijk A, Oinas P (eds) Proximity, distance and diversity, issues on economic interaction and local development. Ashgate, Aldershot Asheim BT, Coenen L (2005) Knowledge bases and regional innovation systems: comparing Nordic clusters. Res Policy 34:1173–1190 Asheim BT, Coenen L (2006) Contextualising regional innovation systems in a globalising learning economy: on knowledge bases and institutional frameworks. J Technol Transf 31:163–173 Asheim BT, Coenen L, Vang J (2007) Face-to-face, buzz, and knowledge bases: sociospatial implications for learning, innovation, and innovation policy. Environ Plan C: Gov Policy 25:655–670 Asheim BT, Gertler MS (2005) The geography of innovation: regional innovation systems. In: Fagerberg J, Mowery DC, Nelson RR (eds) The Oxford handbook of innovation. Oxford University Press, Oxford Asheim BT, Herstad SJ (2005) Regional innovation systems, varieties of capitalism and nonlocal relations: challenges from the globalising economy. In: Boschma RA, Kloosterman RC (eds) Learning from clusters: a critical assessment from an economic-geographical perspective. Springer, Dordrecht Bathelt H (2005) Cluster relations in the media industry: exploring the ‘distanced neighbour’ paradox in Leipzig. Reg Stud 39:105–127 Bathelt H, Malmberg A, Maskell P (2004) Clusters and knowledge: local buzz, global pipelines and the process of knowledge creation. Prog Hum Geogr 28:31–56 Bathelt H, Schuldt N (2008) Between luminaires and meat grinders: international trade fairs as temporary clusters. Reg Stud 42:853–868 Boschma R (2005) Proximity and innovation: a critical assessment. Reg Stud 39:61–74

24

2 Regional Innovation and Knowledge Creation

Boschma R, Lambooy J, Schutjens V (2002) Embeddedness and innovation. In: Taylor M, Leonard S (eds) Embedded enterprise and social capital: international perspectives. Ashgate, Aldershot Breschi S, Lissoni F (2001) Knowledge spillovers and local innovation systems: a critical survey. Ind Corp Change 10:975–1005 Brown JS, Duguid P (2001) Knowledge and organization: a social-practice perspective. Organ Sci 12:198–213 Bunnell TG, Coe NM (2001) Spaces and scales of innovation. Prog Hum Geogr 25:569–589 Camagni R (ed) (1991) Innovation networks: spatial perspectives. Belhaven Press, London Cohen WM, Levinthal DA (1990) Absorptive capacity: a new perspective on learning and innovation. Adm Sci Q 35:128–152 Cooke P (1992) Regional innovation systems: competitive regulation in the new Europe. Geoforum 23:365–382 Eccles RG (1981) The quasifirm in the construction industry. J Econ Behav Organ 2:335–357 Edquist C (2005) Systems of innovation: perspectives and challenges. In: Fagerberg J, Mowery DC, Nelson RR (eds) The Oxford Handbook of Innovation. Oxford University Press, Oxford Etzkowitz H (2008) The triple helix: university-industry-government innovation in action. Routledge, London Etzkowitz H, Leydesdorff L (2000) The dynamics of innovation: from national systems and “Mode 2” to a triple helix of university–industry–government relations. Res Policy 29:109–123 Fromhold-Eisebith M (2007) Bridging scales in innovation policies: how to link regional, national and international innovation systems. Eur Plan Stud 15:217–233 Gertler MS (2004) The institutional geography of industry practice. Oxford University Press, Oxford Gertler MS (2008) Buzz without being there? communities of practice in context. In: Amin A, Roberts J (eds) Community, Economic Creativity, and Organization. Oxford University Press, Oxford Giuliani E, Bell M (2005) The micro-determinants of meso-level learning and innovation: evidence from a Chilean wine cluster. Res Policy 34:47–68 Grabher G (2001) Ecologies of creativity: the village, the group, and the heterarchic organisation of the British advertising industry. Environ Plan A 33:351–374 Grabher G (2002) Cool projects, boring institutions: temporary collaboration in social context. Reg Stud 36:205–214 Grabher G (2004) Temporary architectures of learning: knowledge governance in project ecologies. Organ Stud 25:1491–1514 Keeble D, Lawson C, Moore B, Wilkinson F (1999) Collective learning processes, networking and ‘institutional thickness’ in the Cambridge region. Reg Stud 33:319–332 Lundvall B-Å (ed) (1992) National systems of innovation: towards a theory of innovation and interactive learning. Pinter Publishers, London Malerba F (2002) Sectoral systems of innovation and production. Res Policy 31:247–264 Malmberg A (1996) Industrial geography: Agglomeration and local mileu. Prog Hum Geogr 20:392–403 Malmberg A (2003) Beyond the cluster: Local milieus and global connections. In: Peck J, Yeung HWC (eds) Remaking the global economy: economic-geographical perspectives. Sage, London Maskell P, Bathelt H, Malmberg A (2006) Building global knowledge pipelines: the role of temporary clusters. Eur Plan Stud 14:997–1013 Maskell P, Kebir L (2006) What qualifies as a cluster theory? In: Asheim B, Cooke P, Martin R (eds) Clusters and regional development: critical reflections and explorations. Routledge, London and New York Matsubara H (2006) Keizai chirigaku: ritchi/chiiki/toshi no riron (Economic geography: theories relating to industrial location, regional economy, and urban structure). University of Tokyo Press, Tokyo (in Japanese) Matsubara H (2007) Chishiki no kukanteki ryudo to chiikiteki innovation system (Spatial knowledge flow and regional innovation system). Tokyo Daigaku Jinbun Chirigaku Kenkyu (Komaba Stud Hum Geogr) 18:22–43 (in Japanese)

References

25

Mizuno M (2005) Innovation no chirigaku no doukou to kadai: chishiki, network, kinsetsusei (The geography of innovation reconsidered: knowledge, network, and proximity). Keizaichirigaku Nempo (Ann Jpn Assoc Econ Geogr) 51:205–224 (in Japanese with English abstract) Mizuno M (2007) Keizai chirigaku ni okeru shakai network ron no igi to tenkai hoko (Social network approach and knowledge in economic geography). Chirigaku Hyoron (Geogr Rev Jpn) 80:481–498 (in Japanese with English abstract) Mizuno M (2011) Innovation no keizai kuukan (Economics spaces of innovation). Kyoto Daigaku Shuppan-kai, Kyoto (in Japanese) Mowery DC, Sampat BN (2005) Universities in national innovation systems. In: Fagerberg J, Mowery DC, Nelson RR (eds) The Oxford handbook of innovation. Oxford University Press, Oxford Narula R, Zanfei A (2005) Globalization of innovation: the role of multinational enterprises. In: Fagerberg J, Mowery DC, Nelson RR (eds) The Oxford handbook of innovation. Oxford University Press, Oxford Nelson RR (ed) (1993) National innovation systems: a comparative analysis. Oxford University Press, Oxford Nooteboom B (2006) Innovation, learning and cluster dynamics. In: Asheim B, Cooke P, Martin R (eds) Clusters and regional development: critical reflections and explorations. Routledge, London and New York Nooteboom B, Haverbeke WV, Duysters G, Gilsing V, van den Oord A (2007) Optimal cognitive distance and absorptive capacity. Res Policy 36:1016–1034 Owen-Smith J, Powell WW (2004) Knowledge networks as channels and conduits: the effects of spillovers in the Boston biotechnology community. Organ Sci 19:549–583 Powell WW, Grodal S (2005) Networks of innovators. In: Fagerberg J, Mowery DC, Nelson RR (eds) The Oxford handbook of innovation. Oxford University Press, Oxford Saxenian A (1994) Regional advantage: culture and competition in Silicon Valley and Route 128. Harvard University Press, Cambridge Saxenian A (2007) The new argonauts: regional advantage in a global economy. Harvard University Press, Cambridge Storper M (1997) The regional world: territorial development in a global economy. The Guilford Press, New York Storper M, Venables AJ (2004) Buzz: face-to-face contact and the urban economy. J Econ Geogr 4:351–370 Sydow J, Staber U (2002) The institutional embeddedness of project networks: The case of content production in German television. Reg Stud 36:215–227 Toda J (2004) Innovation system approach to innovation no kuukansei (Spatiality of innovation: from the perspectives of innovation systems approaches). Keizaigaku Kenkyu (J Political Econ) 70:45–62 (in Japanese) Tödtling F, Lehner P, Trippl M (2006) Innovation in knowledge intensive industries: the nature and geography of knowledge links. Eur Plan Stud 14:1035–1058 Tödtling F, Trippl M (2005) One size fits all? towards a differentiated regional innovation policy approach. Res Policy 34:1203–1219 Torre A, Rallet A (2005) Proximity and localization. Reg Stud 39:47–59 Wenger E, McDermott R, Snyder W (2002) Cultivating communities of practice: a guide to managing knowledge. Harvard Business School Press, Boston Yamamoto K (2005) Sangyo shuseki no keizai chirigaku (Economic geography of industrial agglomeration). Hosei Daigaku Shuppan-kai, Tokyo (in Japanese) Yokura Y (2009) Sangyo shuseki chiiki ni okeru innovation no ketteiyouin bunseki: chiikishinsei consortium kenkyu kaihatsu jigyo wo jireitoshite (An empirical analysis of innovation in industrial agglomerations: a case of consortium R&D project for regional revitalization). Keizaichirigaku Nempo (Ann Jpn Assoc Econ Geogr) 55:351–368 (in Japanese with English abstract)

Chapter 3

R&D Networks and Regional Innovation

Abstract Spatial dimension analysis is indispensable when studying knowledge flow from the perspective of economic geography. This chapter adapted the knowledge pipeline framework to investigate external networks and applies social network analysis to the structure and innovation of joint research projects. Social network analysis can be performed without incorporating the spatial dimension; however, the special aspects of R&D networks and the research aims of this chapter are best fulfilled by its use. We focused on R&D networks that exist between industry, academia, and the public sector and considered the case of the Consortium R&D Project for Regional Revitalization in Japan. This chapter disclosed differences between network structures and patterns of collaborative R&D with regard to local and non-local actors. Additionally, the spatial patterns of collaborative R&D depend on both the technical field of the joint research projects and the parties involved. Keywords Social network analysis · Research and development · Knowledge flow

3.1 Introduction Recent studies on innovation in industrial agglomeration have emphasized the role of non-local or external networks. Social network analysis is key to clarifying network structure. Existing studies have applied social network analysis to interorganizational transactional relationships in Japan. Using the electronic data of the main customers of approximately 3,700 companies in Japan, Sugiyama et al. (2006) proposed a structural superiority index for corporate transaction networks. A structural superiority index can be used to demonstrate a certain node’s advantageous position in comparison with the other nodes in a network. Sugiyama et al. (2006) examined the correlation between the structural superiority index and a variety of financial indices and posited that the aggregate market value of a company increases the higher its structural superiority index is. Wakabayashi (2006) examined mutual trust according to long-term transactional relationships. He presented a case study of the subcontract enterprise cooperation associations of two major electric companies in the Tohoku region of Japan and © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2021 Y. Yokura, Regional Innovation and Networks in Japan, International Perspectives in Geography 16, https://doi.org/10.1007/978-981-16-2191-8_3

27

28

3 R&D Networks and Regional Innovation

demonstrated the importance of trust in quality control by visualizing the interorganizational network. These studies proposed a new perspective by applying structural analysis, but a problem remains. Although they analyzed industrial agglomerations, they did not determine the relationship between “internal” and “external” networks and industrial and knowledge clusters. Giuliani and Bell (2005) and Cantner and Graf (2006), instead, focused on the spatial dimensions of innovation using social network analysis. When studying knowledge flow from an economic geographical perspective, the analysis of spatial dimensions is indispensable. This study adopts the knowledge pipeline framework of Bathelt et al. (2004) and Owen-Smith and Powell (2004) to investigate external networks and applies social network analysis to the relational structure and knowledge creation processes of joint research projects in Japan. The next section will present this study’s data and methodology and compare the structures of research and development (R&D) networks in regional blocks in Japan. Section 3.3 will illustrate the spatial patterns of knowledge flow in various technical fields. The last section will draw conclusions and present possible directions for future research.

3.2 The Structure of R&D Networks 3.2.1 Data and Social Network Analysis Social network analysis will be employed in this study to achieve the abovementioned research purposes. Social network analysis is useful from an economic geographical perspective, particularly when examining relational structures of networks (Yokura 2008). While social network analysis can be performed without incorporating the spatial dimension, the special aspects of R&D networks and the research aims of this chapter are best fulfilled by its use. Giuliani and Bell (2005) investigated the cognitive positions and absorptive capacity of network actors in the emerging wine cluster in Chile. Utilizing firm-level empirical evidence, they classified the cluster’s cognitive positions into four distinct types—knowledge sources, absorbers, mutual exchangers, and isolators. In addition, they measured the absorptive capacity and discovered a robust correlation between absorptive capacity, nonlocal knowledge linkages, and actor centrality. Graf (2006) used social network analysis to analyze knowledge flow by studying the patent data of innovators in Germany and determined that differences in network positions influenced the actors’ activities. A large collection of data regarding the attributes of actors is necessary to apply social network analysis to an industry-academia-public sector collaboration. We use data collections on the Consortium R&D Project by the Ministry of Economy, Trade and Industry (METI) of Japan. This program aims to develop R&D collaboration networks between industry, academia, and the public sector to promote start-ups resulting from the related collaborative research. Each project’s duration is up to two

3.2 The Structure of R&D Networks Fig. 3.1 Bipartite graph of research projects and actors

29

A

Project

a

Actor

B

b

c

C

d

e

f

years. METI provides each project a grant of up to a hundred million yen the first year and fifty million the next year for collaborative R&D support. This data enabled us to learn the names of the projects and the identities and locations of the actors who participated in these joint research projects. Projects accepted between 2001 and 2007 were analyzed in this study. The total number of projects undertaken was 911 and 4,547 actors participated in these projects. We constructed a bipartite graph of actors and research projects, illustrating, for example, that Actors a and b participated in a joint R&D project called Project A. Figure 3.1 is an illustration of a bipartite graph of research projects and actors. To build a representation of a network of actors, incidence matrix X is constructed in which the rows correspond to actors and the columns to projects. ⎛

1 ⎜1 ⎜ ⎜ ⎜0 X =⎜ ⎜0 ⎜ ⎝0 0

0 1 1 1 1 0

⎞ 0 0⎟ ⎟ ⎟ 0⎟ ⎟ 1⎟ ⎟ 1⎠ 1

Next, a symmetrical matrix is computed by multiplying the two matrices, X, and transposing X. The adjacent matrix shown below is an example of the network of actors. For example, the value (4, 5), the element, shows that Actors d and e participated in the same two R&D projects. ⎛

1 ⎜1 ⎜ ⎜ ⎜0  XX = ⎜ ⎜0 ⎜ ⎝0 0

1 2 1 1 1 0

0 1 1 1 1 0

0 1 1 2 2 1

0 1 1 2 2 1

⎞ 0 0⎟ ⎟ ⎟ 0⎟ ⎟ 1⎟ ⎟ 1⎠ 1

30

3 R&D Networks and Regional Innovation

The network drawing program NetDraw was used to graphically illustrate the network of actors (Fig. 3.2). We visualized the relational structure of collaborative R&D networks and compared the network structures of nine regional blocks (Fig. 3.3). Fig. 3.2 Network among actors through joint research projects

d

a

b

c

f

e

Fig. 3.3 Regional blocks according to METI’s regional bureaus

3.2 The Structure of R&D Networks

31

3.2.2 Sectoral and Regional Structures of R&D Networks The Consortium R&D Project was classified into six technical categories. Table 3.1 lists the number of R&D projects undertaken in each of the following technical fields: life sciences technology, information and communication technology, nanotechnology, manufacturing, environment and energy, and cross-cutting technology. Kanto had the greatest number of projects (207 projects) and Okinawa the least, with only 19 projects accepted. Table 3.1 shows that each of the nine regional blocks specialized in selected technological fields, although almost every technological field was represented by approved projects in all regions. Five regional blocks (viz., Hokkaido, Kanto, Kinki, Chugoku, and Kyushu) accounted for about three-quarters of the all of the projects. Certain actors were selected for an in-depth analysis of the R&D network’s structure. Figure 3.4 illustrates the relational structure of R&D networks in the five regional blocks. The National Research Institute, “S.S.,” which has many joint research partners, became the core actor for R&D in both Kanto and Kinki. The influence of engineering faculties from universities such as “Sa.U.,” “G.U.,” and “Sz.U.,” were often positioned near the National Research Institute in the relational structure of R&D networks, which is another indicator of the rather similar of relational structures in these two regional blocks. In peripheral regions such as Hokkaido and Kyushu, the local universities and National University (not only those with engineering faculties) played an essential role. In such regions, these two actors participated in many of the same R&D projects, forming a rigid dyadic connection. Table 3.2 shows the descriptive statistics of the collaborative R&D network of nine regional blocks. The degree centrality of the public sector is larger than that of industry or academy in the regional blocks with the exception of Hokkaido. Degree centrality means the number of direct ties that one node has to other nodes, i.e., a node’s exposure to what is flowing in the network. Although the degree centrality of industry is similar in all of the blocks, the degree centrality of academia and the public sector varies depending on the characteristics of specific regions. The number of R&D partners of academia is larger than that of the public sector in Hokkaido; therefore, academia has become a hub actor for circulating local knowledge. In Kinki and Chugoku, the degree centrality of academia and the public sector is larger than in the other regional blocks, indicating that the number of collaborative R&D projects each actor joins is large. Table 3.2 indicates that each block features one or two large components that connect each actor directly or indirectly to other actors.

3.3 The Spatial Dimension of Collaborative R&D Networks We now examine the spatial dimensions of the R&D networks of six technological fields. The collaborative R&D networks were mapped by Geographic Information Systems (GIS) data from the 2001–2007 period.1 Figure 3.5 shows that the spatial

-

129

3

17

6

8

17

9

45

12

12

90

159

358

607

12

85

28

41

76

40

187

70

68

-

163

0

23

10

17

38

21

31

13

10

Projects

146

200

475

821

0

113

53

105

193

104

145

56

52

Actors

Nanotechnology

-

154

3

26

11

13

21

21

35

17

7

Projects

159

213

554

926

8

146

69

80

138

135

201

114

35

Actors

Manufacturing

Source METI data for Consortium R&D Project for Regional Revitalization 2001–2007

118

435

Public

-

Industry

810

257

176

total

28

74

64

41

197

105

142

24

135

Academia

18

6

Okinawa

12

Shikoku

Kyusyu

41

22

Chubu

7

34

Kanto

Chugoku

6

Tohoku

Kinki

30

Actors

Information and Communication

Projects Actors Projects

Life science

Hokkaido

Blocks

Table 3.1 Number of accepted projects by technical field and region

178

3

28

13

20

31

26

37

10

10

144

200

553

897

21

136

67

121

149

113

179

52

59

Projects Actors

Environment and Energy

-

48

0

3

1

5

7

2

18

6

6

37

73

114

224

0

20

3

26

29

7

85

26

28

Projects Actors

Cross-Cutting Technology

-

63

4

10

9

9

6

12

7

5

1

56

67

139

262

15

44

37

47

25

45

26

18

5

Projects Actors

Others

32 3 R&D Networks and Regional Innovation

3.3 The Spatial Dimension of Collaborative R&D Networks (a) Hokkaido

H.U.(p)

N.Inc. S.Inc.

33

(b) Kanto

U.T.(e)

O.U. S.S.

Sa.U.(e)

H.U.(m) H.K.

Sn.U.(e)

R.K.

Yn.K.

S.S.

H.U.(e)

Yn.U.(e) Sz.U.(e) G.U.(e)

Mu.I.T.(e)

(c) Kinki

(d) Chugoku

O.S. K.U.(e)

M.C.

O.U.(e) S.S.

F.K.

H.U.(e)

F.U.(e)

S.S.

Sm.U.(s) K.I.T.

Tt.U.(e)

(e) Kyushu

O.K.

Abbreviations

Legend

Industry

Ky.I.T.(e) Ku.U.(s)

(a) agriculture (b) engineering

F.K.

Academia

N.U.(e) S.S.

(c) medicine

Ky.U.(a)

Ky.U.(e)

(d) pharmacy Public Sector

Ka.K. Ka.U.(e) M.U.(e)

(e) science U. University

Fig. 3.4 Structure of R&D networks in five regions

distance between the actors that joined the same life sciences projects was often within 100 km (km) of Hokkaido, Kanto, Kinki, and the northern region of Kyushu. Hokkaido has abundant agricultural and fisheries resources, and many collaborative R&D projects utilize those resources. Therefore, local circulation of specialized knowledge is critical for fostering innovation. In addition, Fig. 3.5 depicts the fact that long-distance collaborative R&D can be found in current state-of-the-art of biotechnology projects in Japan. Figure 3.6 demonstrates that there are many collaborative R&D projects within 100 km in provincial urban areas. In other words, the R&D networks of information and communication technology are decentralized in Japan. This is because it is easy

215

2

14.24

10.72

6.93

2098

261

Tohoku

547

2

14.88

8.67

5.85

4818

686

Kanto

336

1

14.50

9.43

5.89

2620

369

Chubu

Source METI data for Consortium R&D Project for Regional Revitalization 2001–2007

238

Public sector

Size of the largest component

12.47

9.44

Academy

1

5.92

Industry

–

No. of components with ten or more actors

1794

–

Degree centrality

245

No. of actors

No. of links

Hokkaido

Regional Blocks

Table 3.2 Descriptive statistics of the collaborative R&D networks

496

1

20.84

13.53

6.19

4192

528

Kinki

298

1

21.21

13.43

6.72

2594

302

Chugoku

203

1

17.56

11.13

6.02

1686

220

Shikoku

357

1

15.67

13.63

5.67

2888

381

Kyusyu

32

2

7.00

6.07

5.02

358

65

Okinawa

34 3 R&D Networks and Regional Innovation

3.3 The Spatial Dimension of Collaborative R&D Networks

Fig. 3.5 R&D networks in life science technology

Fig. 3.6 R&D networks in information and communication technology

35

36

3 R&D Networks and Regional Innovation

Fig. 3.7 R&D networks in nanotechnology

for researchers to participate in project over long distances in the telecommunications field, since software development often utilizes codified knowledge. However, we can see from Fig. 3.6 that the co-location of actors is also necessary for exchanging this knowledge. Nanotechnology makes use of nanometer (10−9 m) materials and devices. The leading scientists with advanced knowledge can be the key actors in emerging fields such as nanotechnology. Figure 3.7 shows that the universities with which these leading scientists in nanotechnology are affiliated participate in joint research projects not only with local companies and research institutes but also with other organizations located in distant places. We see from Fig. 3.8 that the amount of R&D that occurs within 100 km of manufacturing technology is greater than that of any other field. The processing technology of the metal molds and machine systems is a central task of joint research in manufacturing. Therefore, the level of skill demanded is not higher than in other fields. Furthermore, the control of distribution costs and the necessity of close ties existing among universities and local research institutes require the localization of R&D networks. Figure 3.9 illustrates the fact that R&D networks within 100 km of these locations are decentralized in the environment and energy field. These connections are remarkably strong in Kanto, Chubu, Kinki, and Kyushu and many actors located in those regions joined the same research projects. As Fig. 3.10 indicates, although the amount of cross-cutting technology is small, a few knowledge clusters can be found in Hokkaido, Kanto, Kinki, and Kyushu.

3.3 The Spatial Dimension of Collaborative R&D Networks

Fig. 3.8 R&D networks in manufacturing technology

Fig. 3.9 R&D networks in environment and energy technology

37

38

3 R&D Networks and Regional Innovation

Fig. 3.10 R&D networks in cross-cutting technology

Fig. 3.11 Spatial reach of collaborative R&D networks by technical field (Source METI data for Consortium R&D Project for Regional Revitalization 2001–2007)

3.3 The Spatial Dimension of Collaborative R&D Networks

39

Fig. 3.12 Spatial reach of collaborative R&D networks for pairs of actors (Source METI data for Consortium R&D Project for Regional Revitalization 2001–2007)

We can see from Fig. 3.11 that the ratio of R&D to distance differs in each technical field. The ratio of collaborative R&D within 100 km is over 50% in every technical field; the ratio over 500 km is clearly higher in analytical knowledge-based technical fields. By contrast, that of synthetic knowledge-based technical fields is less than 10%. This result indicates that in manufacturing collaborative R&D, there is relatively little need for the knowledge of non-local actors. Figure 3.12 indicates the length of distance between a pair of actors according to actor type—industry, academia, or the public sector. First, the ratio of “academiaacademia” with a distance of over 100 km is 50% or more and the ratio of collaborative R&D over 500 km is almost 20%. The ratio of “industry-academia” over 500 km is the second highest (10% or more) and collaborative R&D within 100 km is high. As for joint research in which the public sector is included, collaborative R&D networks are located within a short distance in general. Therefore, we may conclude that the public sector plays an important role in the local knowledge circulation. Conversely, universities and some companies are responsible for transmitting knowledge from non-local actors. Figure 3.13 summarizes the model of the collaborative R&D networks according to the type of relational structure of network (decentralized or centralized) and technical field (synthetic knowledge-based or analytical knowledge-based) in the Consortium R&D Project. There are limited hubs or core actors in centralized relational structures like those in Kanto and Kinki. On the other hand, there are several nodes that have many joint research partners in the decentralized relational structures like those in Tohoku and Kyushu. Local actors positioned in the same clusters constitute decentralized, synthetic knowledge-based networks. Some universities and the

40

3 R&D Networks and Regional Innovation

Fig. 3.13 Model of collaborative R&D networks in the Consortium R&D Project for Regional Revitalization in Japan

public sectors have many R&D collaborations and form a multicore relational structure. Conversely, the number of core actors with many joint research units is limited in centralized, synthetic knowledge-based networks. In this study, the innovation performance of R&D networks is measured by the number of start-ups. Table 3.3 presents the ratio of start-up projects to all R&D projects in the regional blocks. With regard to analytical knowledge-based projects, centralized regional blocks (i.e., Hokkaido and Okinawa) created more business startups than the decentralized regional blocks of Chugoku and Shikoku. Conversely, with regard to synthetic knowledge-based projects, decentralized blocks (i.e., Tohoku) performed better than centralized blocks (i.e., Chubu and Kyushu). Collaborative R&D were open to external actors in decentralized and analytical knowledge-based networks. Universities and companies located in distant regions joined collaborative R&D projects as “outsiders.” The role of the public sector was as a “coordinator” promoting the formation of collaborative R&D projects with local actors. A few universities acted as a hub in concentrated and analytical knowledgebased network structures. These actors were transmitters of external knowledge from universities and companies and played a critical role in regional innovation.

42

221

0.19

Business start-up (a)

Total (b)

Ratio(a/b)

–

DS

0.34

225

77

Tohoku

CS/CA

0.25

632

157

Kanto

CS

0.22

278

60

Chubu

CS/CA

0.16

492

81

Kinki

DA

0.11

290

31

Chugoku

DA

0.12

158

19

Shikoku

CS

0.18

376

69

Kyusyu

CA

0.24

46

11

Okinawa

Note DS: Decentralized and Synthetic knowledge-based, CS: Centralized and Synthetic knowledge-based, DA: Decentralized and Analytical knowledge-based, CA: Centralized and Analytical knowledge-based Source METI data for Consortium R&D Project for Regional Revitalization 2001–2004

CA

-

Number of actors

R&D type

Hokkaido

Regional Blocks

Table 3.3 Ratio of start-up projects and type of R&D network

3.3 The Spatial Dimension of Collaborative R&D Networks 41

42

3 R&D Networks and Regional Innovation

3.4 Conclusion This study focused on R&D networks that exist between industry, academia, and the public sector and considered the case of the Consortium R&D Project for Regional Revitalization in Japan. It applied social network analysis to examine the relational structure and regional innovation of collaborative R&D projects. The resulust of this study reveal differences between relational structures and patterns of collaborative R&D with regard to local and non-local actors in Japan. The regions can be classified into two types of relational structures. One is the “decentralized type,” i.e., a multicore structure with many R&D collaborations. The other is the “concentrated type,” with a limited number of core actors. The spatial patterns of the collaborative R&D networks generally depend on both the technical field of the joint research projects and the parties involved. While long-distance, decentralized networks dominated, collaborative R&D networks in information and communication technology or nanotechnology tended to be more spatially concentrated in a few locations that are intra-regionally embedded. Academia-academia cooperation has much greater spatial reach than that with companies The public sector plays a critical role for regional innovation processes, while up-to-date academic research is more important in some peripheral parts of the counties such as Hokkaido and Kyushu. This study considered the start-up business as a measure of innovation because the available data are limited. However, there is room for improvement in the definition of “innovation.” Patent applications, the publication of papers, and entrepreneurship are also innovative activities that could serve as a proxy for innovation performance. We can apply the framework of this study using data that are obtained from other Japanese cluster policies. A future direction of this study will be to examine not only formal project networks, but also informal networks, such as conferences and trade fairs. Note 1.

The maps of Fig. 3.5~3.10 are defined in the Universal Transverse Mercator (UTM) and cover between approximately 26–46 degrees north latitude and between approximately 127–147 degrees east longitude.

References Bathelt H, Malmberg A, Maskell P (2004) Clusters and knowledge: local buzz, global pipelines and the process of knowledge creation. Prog Hum Geogr 28:31–56 Cantner U, Graf H (2006) The network of innovators in Jena: an application of social network analysis. Res Policy 35:463–480 Giuliani E, Bell M (2005) The micro-determinants of meso-level learning and innovation: evidence from a Chilean wine cluster. Res Policy 34:47–68 Graf H (2006) Networks in the innovation process: local and regional interactions. Edward Elgar, Cheltenham

References

43

Owen-Smith J, Powell WW (2004) Knowledge networks as channels and conduits: the effects of spillovers in the Boston biotechnology community. Organ Sci 19:549–583 Sugiyama K, Honda O, Ohsaki H, Imase M (2006) Network bunseki syuho ni yoru nihon kigyokan no torihiki kankei network no kozo bunseki (Appliction of network analysis techniques for Japanese corporate transaction network). Shakai Johogaku kenkyu (J Socio-Inf Stud) 11(2):45–56 (in Japanese) Wakabayashi N (2006) Nihon kigyo no network to shinrai: kigyokan kankei no atarashii keizai shakaigakuteki bunseki (Network and trust in Japanese interorganizational relationships). Yuhikaku, Tokyo (in Japanese) Yokura Y (2008) Keizai chirigaku oyobi kanren syobunya ni okeru network wo meguru giron (A consideration of discussions in economic geography and related fields on “network”). Keizaichirigaku Nempo (Ann Jpn Assoc Econ Geogr) 55:351–368 (in Japanese with English abstract)

Chapter 4

Local Trade Fairs as Temporary Clusters: A Case Study of the Suwa Area Industrial Messe

Abstract This chapter investigates the ways in which various relationships are created among actors at local trade fairs held in industrial agglomerations in Japan. Trade fairs can be considered as temporary clusters in that they are economic phenomena where actors gather at a particular place for a limited duration with specific intentions. We focus on Japanese local trade fairs held in industrial agglomerations, specifically the Suwa Area Industrial Messe, and investigate how the various relationships and networks among actors, such as exhibitors and visitors at the trade fair, are developed. The study’s findings offer important implications for improving industrial agglomerations using local trade fairs as temporary clusters. Specifically, the following three findings are noteworthy. First, trade fairs function as a place to acquire information and relevant knowledge by observing the skills of other exhibitors. Second, trade fairs facilitate communication with non-local visitors who are potential customers, thus making it possible for exhibitors to acquire new orders during the fair. In particular, the trade fair strengthens the marketing power of small and medium-sized enterprises. Third, trade fairs offer exhibitors the opportunity to help build sustainable relationships and mutual trust by inviting existing customers to attend. Keywords Temporary clusters · Trade fairs · Vertical · Horizontal relationships

4.1 Introduction Trade fairs can be seen as temporary clusters, as they are economic phenomena where actors gather at a particular place for a limited duration, with specific intentions. Recent studies on knowledge transfer and innovation focus on the role of temporary geographical proximity (Torre 2008). These studies indicate that temporary Faceto-Face (F2F) communication, such as the communication that occurs at trade fairs and business conferences, enables actors to acquire information and new, relevant knowledge similar to what may occur in permanent clusters. As shown in Chapter 2, the temporary cluster’s characteristic of the extraordinary is an important element for economic performance. Hansen (2004: 3–4) suggested © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2021 Y. Yokura, Regional Innovation and Networks in Japan, International Perspectives in Geography 16, https://doi.org/10.1007/978-981-16-2191-8_4

45

46

4 Local Trade Fairs as Temporary Clusters: A Case Study …

that trade shows (or trade fairs) serve a number of important purposes such as improving employee and customer motivation, in addition to supporting corporate image building and marketing activities. Participants are open and responsive to stimulation and tend to generate novel ideas and relevant knowledge through various communication opportunities that arise during trade shows. From the perspective of knowledge-based conceptualization as developed by Asheim et al. (2007), trade fairs can be classified into two categories. The designintensive type requires symbolic knowledge and artistry or creativity. This would include trade fairs held by industries such as furniture and other culturally-based industries. The technology-intensive type is based on analytical or synthetic knowledge and attaches special importance to personal meetings as it is difficult to evaluate technical properties from exhibitions (Bathelt and Schuldt 2010: 1964). Focusing on the comic book industry as an example of a design-intensive type, Norcliffe and Rendace (2003) argued that networking among creators and publishers at major conventions is significant for comic book production. Yoon and Malecki (2010) drew attention to global production networks in the animation industry and argued that professional gatherings contribute to knowledge sharing and collaboration among geographically dispersed animation artisans. Other investigations focused on trade fairs held by the Swedish furniture industry (Ramírez-Pasillas 2008, 2010; Power and Jansson 2008). Using social network analysis, it was shown—albeit with limited evidence—that novel knowledge spills over into firms that did not participate at the trade fairs, through their local personal networking and partnerships with exhibitors (Ramírez-Pasillas, 2008, 2010). In studies related to technology-intensive trade fairs, Chen (2009) demonstrated the importance of informal networks in sourcing knowledge from overseas. International trade fairs can support those informal networks among exhibitors and visitors, allowing them to acquire information about different markets as well as technological know-how. Although these findings suggest trade fairs play a number of roles, there is a need to better understand the interactions and learning process in temporary clusters. Few empirical studies explicitly examine how temporary clusters affect actors’ activities in industrial agglomerations (or permanent clusters), and vice versa. This study focuses on Japanese local trade fairs held in industrial agglomerations, specifically the Suwa Area Industrial Messe, and investigates the development of diverse relationships and networks among actors that participate in the trade fair, such as exhibitors and visitors. The Suwa Area Industrial Messe, a trade fair of high-precision technology, has expanded considerably and is now recognized as an important business event for firms in the local industrial cluster of the Suwa area, Nagano prefecture (Yokura 2014). Exhibitors are limited to the firms in or near the Suwa area as the fair is intended to promote the local economy. Compared with other local trade fairs, the Suwa Area Industrial Messe has attracted a substantial number of domestic exhibitors and visitors. To gather information for this study, I conducted an extensive interview with the Executive Committee Chairperson of the Suwa Area Industrial Messe in November 2010, and collected data and lists of exhibitors and visitors from the previous

4.1 Introduction

47

fairs. Data used for the analysis were derived from survey questionnaires that were conducted by the executive committee and the Nagano Economic Research Institute in 2003 and 2009. The questionnaires were distributed to all exhibitors of the Suwa Area Industrial Messe, and used the same questions about the exhibitors’ reasons for exhibiting and the number of business transactions that occurred during and after the fairs. Of the 178 questionnaires sent to the exhibitors, 126 were returned in 2003, and 188 of 252 questionnaires were returned in 2009. In the next section, I describe the development of the Japanese trade fair. Next, I discuss the Suwa area and its manufacturing industry based on statistical data. Then, I describe the structure of Japanese trade fairs more broadly, and the role of the Suwa Area Industrial Messe is explained as an example of a local trade fair. Finally, I offer concluding remarks.

4.2 Development of Japanese Trade Fairs Major Japanese trade fairs are concentrated in metropolitan areas such as Tokyo, Osaka, Nagoya, and Fukuoka. The Tokyo Big Sight has the largest facility for exhibitions and holds the largest number of events in Japan (Table 4.1), and the Makuhari Messe and the INTEX Osaka both have exhibition areas of over 70,000 square meters. The seven largest exhibition facilities account for more than three-quarters of the total number of exhibitions held in Japan annually. Like trade fairs elsewhere, Japanese trade fairs can be classified into two types: (1) “technology-intensive” exhibitions that display production goods such as industrial machinery, measuring equipment and basic technology; and (2) “design-intensive” Table 4.1 Facilities of exhibitions held in Japan

Name of facility

No. of events

Total exhibition area (m2 )

Tokyo Big Sight

259

80,660

Pacifico Yokohama

50

20,000

INTEX Osaka

41

70,000

Makuhari Messe

36

75,000

Port Messe Nagoya

13

34,000

West Japan General Exhibition Center

9

16,400

Marine Messe Fukuoka

6

11,351

Others

134

-

Total

548

-

Source JETRO’s online trade fair database (J-messe) and Handbook of Events and Exhibitions 2011(Mihonnichi tenjikai sougou handbook 2011). Tokyo: POP Inc

48

4 Local Trade Fairs as Temporary Clusters: A Case Study …

exhibitions that display consumer goods related to the furniture, advertising, entertainment and cultural industries. Figure 4.1 shows the distribution of trade fairs in Japan based on the above classifications. The numbers of technology-intensive and design-intensive trade fairs are almost equal in Tokyo; however, there are more design-intensive than technologyintensive trade fairs in Osaka, Aichi and Fukuoka. In the Kanto region, designintensive fairs such as jewelry and housing also feature international firms aside from local firms. Hokkaido and Shizuoka, which are famous as major furniture producing and wood processing areas, are also leaders in design-intensive fairs. In contrast, technology-intensive fairs are prominent in the Hokuriku region and in

Fig. 4.1 Geographical distributions of trade fairs by category, 2010 (Note Since the number of trade fairs held in Okinawa Prefecture is zero, it is excluded from the map. Source JETRO’s online trade fair database [J-messe] and Handbook of Events and Exhibitions 2011 [Mihonnichi tenjikai sougou handbook 2011].Tokyo: POP Inc)

4.2 Development of Japanese Trade Fairs

49

Nagano prefecture, which are known for their importance in the precision machinery industry. Table 4.2 shows the major Japanese trade fairs by type of knowledge base (technology versus design-intensive) and location (metropolitan versus local). Beginning with the period of Japan’s high economic growth in the mid-1950s, large-scale exhibitions such as the Tokyo Motor Show have been held in Japan’s metropolitan areas Table 4.2 Major trade fairs in Japan (Year 2010) Trade Fair

Exhibitors foreign Visitors

Cloud Computing Expo 1,241 Japan

n.a

Foreign Location

124,056 n.a

Tokyo Big Sight

Type TM

Eco-Products

745

4

183,140 n.a

Tokyo Big Sight

TM

CEATEC JAPAN

616

196

181,417 n.a

Makuhari Messe

TM

Tokyo Pack

551

91

170,859 2,516

Tokyo Big Sight

TM

FOOMA JAPAN

406

26

140,576 1,968

Tokyo Big Sight

TM

Shiga Environmental Business Exhibition

313

5

36,580

n.a

Nagahama Dome

TL

Suwa Area Industrial Messe

255

n.a

24,180

n.a

Lake Suwa Event Hall

TL

BARI-SHIP IMABARI 179 MARITIME FAIR

14

13,985

n.a

Texport Imabari

TL

Techno Fair in Hokuriku

163

n.a

17,621

n.a

Fukui Industrial Hall

TL

MEX Kanazawa

103

n.a

50,068

n.a

Ishikawa Industrial TL Exhibition Hall

Tokyo International Gift Show

2,506

255

201,245 1,354

Tokyo Big Site

DM

International Jewellery Tokyo

1,257

402

35,763

n.a

Tokyo Big Site

DM

Osaka Automesse

251

n.a

210,118 n.a

INTEX Osaka

DM

Tokyo Game Show

194

91

207,647 n.a

Makuhari Messe

DM

Tokyo Motor Show

113

14

614,400 21,504

Makuhari Messe

DM

Shizuoka Hobby Show

79

n.a

80,000

n.a

Twin Messe Shizuoka

DL

Food Service Industry 75 Exhibition in Hokuriku

n.a

28,100

n.a

Ishikawa Industrial DL Exhibition Hall

Kofu Jewelry Fair

73

n.a

2,223

n.a

I Messe Yamanashi DL

SHIZUOKA KAGU MESSE

67

n.a

7,500

n.a

Twin Messe Shizuoka

DL

Tochigi Housing Fair

53

n.a

10,000

n.a

Marronnier Plaza

DL

Note n.a. = data not available; TM = Technology-intensive and Metropolitan; TL = Technologyintensive and Local; DM = Design-intensive and Metropolitan; DL: Design-intensive and Local. Source JETRO’s online trade fair database (J-messe)

50

4 Local Trade Fairs as Temporary Clusters: A Case Study …

for the purpose of industrial development. The main purpose of trade fairs in the early days was to promote the industry rather thanto encourage business interactions and negotiations. As shown in Table 4.2, the number of foreign visitors is not published for many trade fairs, and the number of foreign exhibitors is small. In general, trade fairs with a strongly international slate of participants are rare, reflecting cultural and other entry barriers for foreign firms. The main purpose of almost all Japanese trade fairs is to cultivate business in the domestic market. This is notably different from the goals in other Asian countries such as China, Korea and Singapore, where the states explicitly and strategically attempt to target overseas markets. There are several problems with Japan’s modern trade fairs. First, they do not draw upon superior organizing skills given that international trade fairs in Japan do not have a long history. Second, since the available data refers only to Japanese trade fairs and was collected locally, we do not have a basis for comparing Japanese trade fairs to those in other countries. However, the Japan Tourism Agency, which is under the Ministry of Land, Infrastructure, Transport and Tourism, is focused on international tourism and promoting Japanese trade fairs in order to develop the overseas markets. The Japan Tourism Agency defined 2010 as “Japan’s MICE Year” and attempted to increase the number of inbound tourists during meetings, incentive tours, conventions, and exhibitions (MICE). In 2012, the Japanese government approved the “Tourism Nation Promotion Basic Plan” which sought to increase the number of MICE (conventions) held in Japan, aiming to become the top host country in the MICE category in Asia by 2016. Based on this, the number of foreign exhibitors and foreign visitors as shown in Table 4.2 will likely increase in the future.

4.3 Economic Context of the Suwa Area The Suwa area in Nagano prefecture is located in central Japan. The area is well known as an industrial district for high-precision technology companies and is often called the “Switzerland of the East” in Japan. The headquarters and branch plants of Seiko Epson Corp. and Nidec Sankyo are located in the Suwa area, and leading Japanese semiconductor and optical device firms such as Kyocera and Olympus have subsidiaries nearby. Many small- and medium-sized enterprises populate the Suwa area industrial agglomeration. The central government of Japan has supported the Suwa area with regional innovation policies such as “Industrial Cluster” and “Knowledge Cluster” programs in the 2000s. Both programs focus on the development of industrial agglomerations by increasing the national competitiveness of innovative products and technologies. The primary goal of Japanese regional innovation policies is to develop R&D networks among industry, academia, and the public sector to support start-up processes based on collaborative research. Researchers in public institutes, universities and firms perform joint research for each R&D project. The Suwa area experienced a severe industrial restructuring after the 1970s. The establishment and enterprise census of Japan shows that the number of manufacturing firms and employees has decreased since the mids-1980s (Fig. 4.2). The Suwa area

4.3 Economic Context of the Suwa Area

51

l

l l

l

l

l

l l

l

l

Fig. 4.2 Number of industrial employees (a) and establishments (b) in Suwa (Source Establishment census of Japan)

52

4 Local Trade Fairs as Temporary Clusters: A Case Study …

has a significantly higher proportion of firms and employees in mechanical industries such as general machinery, electrical machinery and precision instruments, compared to the Japanese economy overall. Additionally, as production shifted overseas in the latter half of the 1980s, the precision instruments industry was replaced by the electrical machinery and electronic parts and devices industry as the dominant industrial sector in the Suwa area. In this socio-economic environment of restructuring, several methods of industrial revival were pursued that were designed to takes advantage of Suwa’s historical and institutional knowledge of high-precision technology. One such method was the local trade fair that was proposed by the Suwa Chamber of Commerce and Industry to create regional brands.

4.4 Building Relationships at the Suwa Area Industrial Messe The Suwa Area Industrial Messe began in 2002 with 174 exhibitors and over 12,000 visitors. It is held every year at an event hall that utilizes the former factory of Toyo Valve, which was a representative company in the Suwa area. Although the Suwa Area Industrial Messe does not (and cannot) attract as many visitors and exhibitors as the trade fairs held at much larger exhibition centers such as the Tokyo Big Sight and Makuhari Messe, it has become famous and compares favorably with other local trade fairs that are far removed from metropolitan areas as it is superior in generating various relationships (Table 4.2). The purpose of the Suwa Area Industrial Messe is to emphasize business negotiations, in addition to promoting the industry. Additionally, the Executive Committee tries to offer overseas exhibitors in Suwa support from the Japanese government. Therefore, the Suwa Area Industrial Messe has different features compared to other large trade fairs in Japan. The price for hosting the Suwa Area Industrial Messe was approximately 40 million yen (US$320,000) in 2002. Half of the total expense was covered by subsidies from local governments in the Suwa area, the Nagano prefecture, and the central government. The Suwa area chambers of commerce and industry groups also funded one-quarter of the expense. The balance of the expense was covered by exhibitor fees. According to an interview with the Executive Committee Chairperson of the Suwa Area Industrial Messe, the budget for hosting the fair had increased to 60 million yen (US$480,000) by 2010. Though the central government eliminated subsidies for the Suwa Area Industrial Messe, local governments and the Nagano prefecture continued to support the fair through contributions totaling about 20 million yen (US$160,000). The amount borne by corporate sponsors and exhibitors rose to account for more than half of the total budget, due to an increase in the number of exhibitors. The number of visitors increased rapidly until 2005. Then, it gradually increased until the 2009 global recession (Fig. 4.3). The number of exhibitors has remained

4.4 Building Relationships at the Suwa Area Industrial Messe

53

Fig. 4.3 Growth of the number of exhibitors and visitors of the Suwa Area Industrial Messe, 2002–2017 (Source Based on data from the executive committee of the the Suwa Area Industrial Messe)

unchanged since 2004 because the Japanese Fire Service Act limits the number of booths within the exhibition hall. In Japan’s provincial cities there are only a limited number of exhibition halls for large-scale events such as trade fairs, and it is usually difficult to find alternative halls. Therefore, the fair uses a screening system whereby the executive committee requires exhibitors to meet certain qualifications, and priority is given to exhibitors located in the Suwa area or nearby. Among all exhibitor applications, nearly 20 percent were excluded from the 2010 fair. Spaces for exhibitors’ booths are separated into four zones according to the following categories: (1) Processing and Engineering (cutting, pressing, optics, diecasting); (2) Machinery and Finished Products (jigs and tools, molds, machine tools); (3) Industrial-Academic Research (universities, research institutes); and (4) Solutions (software, telecommunications, finance). Figure 4.4 shows that the proportion of exhibitors from the Suwa area is the largest in all categories in comparison to exhibitors from outside the Suwa area, except Industrial-Academic Research. This result implies that institutional R&D support from outside of Suwa was required because the area had no science-based university until 2002. The executive committee of the Suwa Area Industrial Messe surveyed the characteristics of visitors by collecting visitors’ name cards at the entrance since the first

54

4 Local Trade Fairs as Temporary Clusters: A Case Study …

l

l

l

Fig. 4.4 Exhibitors by industry group and place of origin at the Suwa Area Industrial Messe, 2002 (Note Numbers in the graph are the number of exhibitors. Source Based on data from the executive committee of the Suwa Area Industrial Messe)

year in 2002 (the collection rate was over 40% in 2002, and 30% in 2009). The survey shows that the proportion of visitors from outside the Suwa area has steadily increased from 48.6% in 2002 to 62.7% in 2009).

4.4.1 Goals of Attendance and State Support The survey questionnaires used in this study were designed to gather information about the exhibition’s purpose and the practices of building relationships with other participants at the fair. The questionnaires were distributed to all exhibitors in 2004 and 2009. Of 178 questionnaires sent to exhibitors in 2004, 126 were returned, and in 2009, 188 of 252 were returned. Regarding the exhibition’s purpose, the 2009 survey revealed that acquiring new orders was the most important goal of exhibitors in attending the fair (rated important by 135 (72.2%) of 187 exhibitors; multiple answers were allowed). Achieving name recognition (49.7%) and exchanging information between exhibitors and visitors (32.6%) were also cited as important reasons for participating. While the average number of business conversations per exhibitor was 6.4, only 15 firms (12.4% of the 121 respondents) acquired new orders during the fair. However, a follow-up survey one year after the fair showed that the number of firms that subsequently acquired new orders increased (38 firms of 107 respondents).

4.4 Building Relationships at the Suwa Area Industrial Messe

55

In an interview, the Executive Committee Chairperson of the Suwa Area Industrial Messe stated that building transactional relationships during the fair was not recommended, as explained by the following: Business relationships that are immediately established during the fair are often not sustainable. It may be too unprofitable or too difficult for exhibitors to follow through on a deal. (The interview was conducted in Japanese on September 10, 2010)

The executive committee of the Suwa Area Industrial Messe stressed that the role of the trade fair was to offer an opportunity for exposure to potential customers who might be unfamiliar with the exhibitors. The Chairperson considered the first step of building long-term relationships to be the exchange of business cards1 and broad information during the fair. There is distinct institutional support for facilitating interactions between exhibitors and visitors at the Suwa Area Industrial Messe. First, an “on-site business meeting planning” invites and guides interested visitors to the exhibitors’ factories. In 2009, approximately 20 exhibitors used this plan to build trust by having visitors observe their facilities and see their equipment. It is virtually impossible to hold on-site business meetings at metropolitan trade fairs because of the distance between the exhibition halls and the production facilities. Second, “private consultation meetings for opening new markets” are held by the Suwa Area Monozukuri Promotion Organization, a non-profit organization (NPO). The NPO was established by Suwa area firms through the Suwa Chamber of Commerce and Industry in 2005 to provide administrative and clerical support to the Suwa Area Industrial Messe, to coordinate industry-academic collaboration, and to develop human resources. The NPO collects firms’ registration data and technical information in advance of the fair and supports business matching. In 2009, 30 firms registered for such support, and 17 of those firms acquired new orders.

4.4.2 Building Relationships With Non-Local Firms In 2002, the fair obtained full support from Toyota Motor Corporation. For example, a call for visitors to the fair was sent to related firms and subcontractors, resulting in many visitors from the Chubu region where Toyota is headquartered. Additionally, more visitors from the Kanto region have come to recent fairs. According to the 2009 questionnaire sent to visitors (577 respondents), approximately 28% of the visitors from outside Nagano prefecture were from Tokyo, 36% were from Kanto (outside Tokyo), and only 11% were from Chubu. Local financial institutions and the local government arrange chartered bus tours and invite non-local firms to the fair. One of trade fairs’ primary purposes is to provide opportunities for local exhibitors to build long-term business relationships and mutual trust with non-local existing customers by inviting them to attend the fair. According to the 2009 questionnaire, more than half of exhibitors advertised the exhibitions in advance and sent out invitations to existing customers. Exhibitors not only guide the firms’ factories tours

56

4 Local Trade Fairs as Temporary Clusters: A Case Study …

during the fair, they also take visitors sightseeing around the Suwa area. They also attempt to establish business contacts by holding private receptions at restaurants and hotels outside the fair. The executive committee vigorously pursues overseas markets in cooperation with national and local institutions. The JETRO, an independent agency of the Japanese government, invites foreign firms that are interested in marketing and conducting business meetings during the fair. In 2009, the JETRO invited automobile parts firms from Switzerland, France, the United States and Canada and covered the visitors’ travel expenses (airfare and accommodations). In advance of the fair, the local NPO recommended firms in Suwa as potential partners. In 2009, 50 firms participated in the JETRO’s business meetings and two local firms acquired overseas orders during the fair. In addition, another local support organization invited foreign firms related to the machine tool industry and conducted business meetings at the fair. Analyzing the participation of overseas firms during the fair, this study found that manufacturing wholesalers play a significant role in acquiring information about the needs of non-local markets. The executive committee had rejected manufacturing wholesalers as exhibitors since the fair’s introduction in 2002, because the purpose of the fair was supposed to be building relationships among local firms and potential customers based on F2F communication during the fair. However, the executive committee reconsidered the role of wholesalers as providers of information about overseas markets in 2005, and the number of manufacturing wholesaler exhibitors increased to ten in 2010. The wholesalers’ reliability in advertising Suwa firms’ high-precision products also supports wider recognition of regional brands.

4.4.3 Processes of Relationship Building at Suwa Area Industrial Messe Having analyzed the specific of relationship building and interfirm interactions during the Suwa Area Industrial Messe, we now attempt to identify broader patterns of relationship building by distinguishing three stages: preparing for the trade fair, activities during the fair and activities after the fair. Bathelt and Schuldt (2008) offered hypotheses regarding the processes of trans-local pipeline creation and effects of international trade fairs. As shown in Chap. 2, the actors participate in various interactions at the different stages of trade fairs. In this study, we explicitly differentiate among these various stages and clarify the relationships among exhibitors and visitors. The Suwa Area Industrial Messe not only offers opportunities for business meetings but also promotes R&D connections and matching during the fair. The region is well known for cross-industrial associations/networks that meet regularly through various events, such as study meetings and research workshops. These include university researchers, firms’ engineers and members of research institutions in Suwa for collaborative R&D efforts. Those R&D associations have participated in the fair

4.4 Building Relationships at the Suwa Area Industrial Messe

57

almost every year since its inception in 2002. Pior to exhibiting at the fair, they frequently gather in groups and collaborate in researching and developing new technologies and products. The Suwa Area Industrial Messe thus reinforces and strengthens collaborative R&D relationships as the partners demonstrate their technological capabilities. In addition, it is one of the most critical business activities for small and medium-sized local firms to promote the fairs in advance and send invitations to existing customers to attend (Fig. 4.5). During a trade fair, visitors and exhibitors exchange information about industrial trends, and observe other exhibitors’ levels of technology (Fig. 4.6). The Suwa Area Industrial Messe thus works as the place to acquire new information and relevant knowledge. It is extremely important for local firms to build relationships with nonlocal potential partners, which begins by exchanging name cards at the fair. Some firms attempt to create and deepen trust and establish relationships through private receptions at neighboring restaurants and hotels outside the Suwa Area Industrial Messe. Additionally, visitors can observe exhibitors’ skills directly through the “on site business meeting planning.”

Suwa Area

visitor

exhibitor

academy

Promotion to existing customers

Collaborative R&D

Fig. 4.5 Relationship building in preparation for the trade fair

Searching for business partners Private meetings

Onsite business meeting planning

Observation of participants’ skill

Fig. 4.6 Relationship building during the trade fair

58

4 Local Trade Fairs as Temporary Clusters: A Case Study …

Continuous business meetings

Technology spillovers through collaborative R&D

Strengthening relationships Fig. 4.7 Relationship building after the trade fair

Manufacturing wholesalers and exhibitors that have pursued business in overseas markets play an important role in sharing information concerning the needs of those non-local markets. The fair strengthens the marketing capability of small and medium-sized enterprises (SMEs) in the Suwa area. Though it is difficult to establish business contacts during the fair, many exhibitors are able to acquire new orders through continued business meetings and conversations after the fair (Fig. 4.7). The fair stimulates not only vertical relationships among exhibitors and visitors, it reinforces horizontal relationships such as research workshops in the industrial agglomeration that improve exhibitors’ technological skills. Through existing horizontal relationships, new information and technology introduced during the fair spill over to non-participating firms. These indirect spillover effects improve the industrial agglomerations and support the creation of a regional brand. The establishment of the regional brand for the Suwa region as a mecca of high precision technology firms also motivates firms to participate in the fair, creating a self-reinforcing growth cycle.

4.5 Conclusions This study investigates how various relationships are created among participants at local trade fairs held in industrial agglomerations in Japan. The findings of this study offer important implications for methods of upgrading industrial agglomerations through local trade fairs as temporary clusters. Since its first fair in 2002, the Suwa Area Industrial Messe became an important event for showcasing the area’s high precision technology and has expanded its scale significantly since its early years. It is now recognized as an important business event for firms in local indusrial cluster of the Suwa area. The reasons for participating at the trade fair are summarized as follows. First, the ability to observe other exhibitors’ skills and new products is important. Second, communication with non-local firms who are potential customers,

4.5 Conclusions

59 䞉Promotion to existing customers

Creating regional brands of local specific technology

䞉Technology spillovers through collaborative R&D

Preparation for the fair

Increasing the number of exhibitors and visitors 䞉Private meetings

Permanent cluster

䞉Business meetings and workshops

Positive growth & development cycle After the fair

Increasing technological capabilities of local firms

䞉Introduction of novel knowledge and non-local market information 䞉Strengthening existing relationships

䞉Searching for business During partners the fair 䞉Gathering information for marketing Development of (1) vertical relationships among firms and (2) horizontal relationships among firms and academia

Fig. 4.8 Upgrading industrial agglomerations through local trade fairs

and acquiring new orders is facilitated by the fair. In particular, the fair strengthens the marketing power of the SMEs in the Suwa area. Third, the fair offers exhibitors opportunities to build sustainable relationships and mutual trust by inviting existing customers to attend. Manufacturing wholesalers that have succeeded in overseas markets play an important role at the trade fair by providing information about the needs of nonlocal markets. This external information can spill over to local firms that do not participate at the trade fair via horizontal relationships such as research workshops. These indirect spillover effects improve the industrial agglomerations and support the building regional brand. The region’s reputation as a mecca of high precision technology firms also motivates firms to participate in the fair, creating a positive growth cycle (Fig. 4.8). Japanese trade fairs are at crucial turning points. Major cities in other Asian countries have been eager to attract international trade fairs. However, Japan is lagging in its trade fair business compared with other countries because there are not enough large facilities or specialized workers to support mega events and exhibitions in Japan. Expanding institutional support from the Japanese government will be key to addressing this competitive threat. Note 1.

In Japan, there is a business practice of trying to communicate with the people whom one meets for the first time by exchanging business cards.

60

4 Local Trade Fairs as Temporary Clusters: A Case Study …

References Asheim B, Coenen L, Vang J (2007) Face-to-face, buzz, and knowledge bases: sociospatial implications for learning, innovation, and innovation policy. Environ Plan C Gov Policy 25:655–670 Bathelt H, Schuldt N (2008) Between luminaires and meat grinders: international trade fairs as temporary clusters. Reg Stud 42:853–868 Bathelt H, Schuldt N (2010) International trade fairs and global buzz, part I: ecology of global buzz. Eur Plan Stud 18:1957–1974 Chen LC (2009) Learning through informal local and global linkages: the case of Taiwan’s machine tool industry. Res Policy 38:527–535 Hansen K (2004) Measuring performance at trade shows: scale development and validation. J Bus Res 57:1–13 Norcliffe G, Rendace O (2003) New geographies of comic book production in North America: the new artisan, distancing, and the periodic social economy. Econ Geogr 79:241–263 Power D, Jansson J (2008) Cyclical clusters in global circuits: overlapping spaces in furniture trade fairs. Econ Geogr 84:423–448 Ramírez-Pasillas M (2008) Resituating proximity and knowledge cross-fertilization in clusters by means of international trade fairs. Eur Plan Stud 16:643–663 Ramírez-Pasillas M (2010) International trade fairs as amplifiers of permanent and temporary proximities in clusters. Entrepren Reg Dev 22:155–187 Torre A (2008) On the role played by temporary geographical proximity in knowledge transmission. Reg Stud 42:869–889 Yokura Y (2014) Performances and roles of local trade fairs in Japan: case study on the Suwa area industrial messe, Nagano prefecture. Komaba Stud Hum Geogr 21:85–100 Yoon H, Malecki EJ (2010) Cartoon planet: worlds of production and global production networks in the animation industry. Ind Corp Change 19:239–271

Chapter 5

Informal Networks and the Evolution of Industry: A Case Study of the Hamamatsu Area

Abstract The research presented in this chapter examines the business workshops held at a local industrial agglomeration in Japan. We focused on the case of the Hamamatsu area that is a well-known industrial district with a rich entrepreneurial culture. The relationships established through business workshops are characterized as informal networks. One role of the business workshop is providing an opportunity for participants to engage in collaborative R&D. The learning process is advanced when the actors forge trusting relationships that lead to innovation through collaborative R&D. This study revealed differences related to geographical distribution and network structure according to the type of actor (i.e., industry, academia, or public institution). The study also discovered that participating in multiple business workshops in a specific technological field plays an important role in innovation and knowledge creation by sharing heterogeneous knowledge among the various Hamamatsu area workshops. Keywords Informal networks · Business workshops · External networks · Social network analysis

5.1 Introduction The concept of networks is often invoked to suggest the strength of industrial agglomerations. Mizuno (2011) examined the issues of innovation studies in economic geography and suggested that the network perspective is effective in understanding the knowledge diffusion and learning processes that occur in industrial agglomeration. Mizuno (2019) focused on the dynamism of networks as an evolutionary process and argued that the effect of proximity should be considered. Extant research in relational economic geography focused on the relational structures of clusters and argued that their performance is improved by mutual connections between both internal and external actors (Bathelt et al. 2004; Bathelt and Schuldt 2008, 2010). These studies suggest that trade fairs play an important role in strengthening external networks. Trade fairs and research workshops held in industrial clusters are called temporary clusters and have been discussed as an important setting © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2021 Y. Yokura, Regional Innovation and Networks in Japan, International Perspectives in Geography 16, https://doi.org/10.1007/978-981-16-2191-8_5

61

62

5 Informal Networks and the Evolution of Industry …

for estrablishing trust between actors (Yokura 2014, 2015). Temporary clusters have institutional start and end points and provide “short-lived hotspots of intense knowledge exchange, network building, and idea generation” (Maskell et al. 2006: 997). Such temporary cluster phenomena can often be found in industrial agglomerations or permanent clusters. Business workshops and gatherings that are held several times a year in industrial agglomerations are also considered to be temporary clusters. No transactions occur, including the exchange of goods or money, in such business workshops except for payment of the participation fee. Although research and development (R&D) activities (Yokura et al. 2013) and content production systems (Hanzawa 2016) are based on projects and contracts, business workshops are not based on strict contracts. Therefore, the relationships between actors that are established through business workshops are characterized as informal networks. One of the roles of the business workshop is to provide an opportunity for participants to engage in collaborative R&D. Business workshops provide a meeting place for various actors to interact and participants can build trust by repeatedly participating in such temporary and informal networks. The learning process progresses when the actors forge trusting relationships, which lead to innovation creation through collaborative R&D. Some studies have shed light on business workshops and gatherings in Japan, arguing that they deepen trusting relationships in industrial agglomerations. Yamamoto (2002) examined the historical development of business workshops governed by public research institutes in Nagano Prefecture and indicated that business workshops promote untraded interdependencies (Storper 1997) between firms located in industrial agglomerations. Yokura (2016) examined international workshops on microelectronics assembly and packaging held in Kyushu and investigated how to strengthen personnel networks to promote the industrial advancement of Kyushu semiconductor industry. It is important to quantitatively evaluate the potential of informal networks to assess the impact of informal networks on innovation creation. This study conducts social network analysis of informal networks in industrial agglomerations. Social network analysis is helpful from an economic geographical perspective, particularly in determining network structures (Yokura et al. 2013: 495). In this paper, the actors who transfer heterogeneous knowledge by participating in several business workshops are identified and the potential of informal networks is quantitatively examined. The impact of informal networks on formal networks characterized by rigid contracts, such as collaborative R&D, are also explored. The research presented in this paper examines the business workshops held at the Hamamatsu area in Shizuoka Prefecture. The Hamamatsu area is well-known as an industrial district with a rich entrepreneurial culture, known as the “yaramaika” spirit in Japan. The area is situated between Tokyo and Osaka, so transportation is very convenient. Hamamatsu’s manufacturing industry was financially supported by the Technopolis project (Sternberg 1995) of the former Ministry of International Trade and Industry (MITI, now METI) and current innovation programs (Yokura 2012; Yokura et al. 2013) of METI and the Ministry of Education, Culture, Sports, Science

5.1 Introduction

63

and Technology (MEXT). These organizations led to the development of “institutional thickness” (Amin and Thrift 1992), consisting of various industry support organizations that coordinate and govern business workshops in the Hamamatsu area. Interviews for this research were conducted in November 2011 and August 2016 at the Hamamatsu city office and at the Hamamatsu Chamber of Commerce and Industry. Some of the participant lists for business workshops and collaborative R&D meetings were collected at that time. The remaining lists were obtained from the websites of industry support organizations in the Hamamatsu area. The next section describes Hamamatsu’s manufacturing industry on the basis of statistical data. Then, utilizing the data, this study conducts social network analysis to determine the relational structure of informal networks and measure their potential. Section 5.4 considers the relationship between informal networks based on business workshops and formal networks based on collaborative R&D. Finally, concluding remarks are presented.

5.2 Economic Context of the Hamamatsu Area The Hamamatsu area consists of Hamamatsu City and Kosai City, which are located around Lake Hamana. Japan’s central government has supported the Hamamatsu area with regional policies such as the “Technopolis,” “Industrial Cluster,” and “Knowledge Cluster” programs during the 2000s (Yokura 2014). The area can be considered a homogeneous region characterized by economic and social cohesion. Figure 5.1 shows the historical evolution of the manufacturing industry in the Hamamatsu area, which was once one of Japan’s leading industrial regions best known for its cotton textile industry. Then, the machine tool and the motorcycle industries emerged based on foundational technology such as the weaving machine. The Hamamatsu area was previously a timber distribution center, which gave rise to the evolution of the lumber industry. Woodworking machinery and musical instrument manufacturing later developed in the area. Textiles, musical instruments, and transportation equipment are called the “three major industries” in the Hamamatsu area. In recent years, the Hamamatsu area has also become well-known for the optical industry as an emerging technological field. The Hamamatsu Optronics1 Cluster program has been promoted under MEXT’s Knowledge Cluster initiative since 2002. With financial support from METI and MEXT, this area has tried to establish several hubs of industry-academia-government collaborations for the creation of innovations related to opto-electronic technology since 2009. The software companies that support the manufacturing industry are agglomerated in this area. The beginning of the software industry’s clustering process started with sales competition between motorcycle manufacturers Honda and Yamaha in the early 1980s. As a result of intense competition, Yamaha recruited voluntary retirees

64

5 Informal Networks and the Evolution of Industry …

past

1900

1950

1980

present

Woven coƩon fabric Weaving machine

Machine tools Marine engines SoŌware MuniƟon industry

Industrial Robot

Motor cycle, motor vehicle Aerospace

Opt-electronics Life science

Agriculture

Electronics Wood working machinery

General machinery

Lumber Plywood technology

Propeller technology

Musical instrument

Fig. 5.1 Historical evolution of the manufacturing industry in the Hamamatsu area (Source Based on Otsuka [1986] and data from the Hamamatsu Chamber of Commerce and Industry)

and many software-related spin-off companies were started (Nagayama 2009). The diffusion of specialized technology by transportation equipment manufacturers led to the continuous creation of spin-off companies in the Hamamatsu area. Figure 5.2 indicates trends in the number of manufacturing establishments and employees in the Hamamatsu area from 1981 to 2012. Similar to national trends, the number of both establishments and employees has been decreasing in recent years. In 1981, there were over 2,000 textile establishments, which accounted for the largest share of the manufacturing industry. However, by 2012, this number had dropped significantly—only 400 textile companies remained in the Hamamatsu area. Conversely, in terms of trends in the number of employees, the share of transportation equipment manufacturing was the largest in 1981. The shares of the textile industry and of “other manufacturing industries,” including musical instruments manufacturing, in terms of total employment, were about half the size of transportation equipment manufacturing in 1981. The number of employees in the transportation equipment and electrical machinery industries increased significantly in 1986 and, since then, these industries have created a lot of employment opportunities. Table 5.1 presents the top 10 trends in the value of shipments of four-digit industrial sub-classifications in the Hamamatsu area from 1980 to 2010 based on the Census of Manufacture.2 We can see that “Motor vehicles parts and accessories” and “Motor vehicles, including motorcycles” were ranked first and second for all years; they were the leading industries in the Hamamatsu area. “Miscellaneous musical instruments,

5.2 Economic Context of the Hamamatsu Area Fig. 5.2 Number of establishments (a) and employees (b) (2-digit) in the Hamamatsu area, 1981–2012 (Source Establishment census of Japan)

65

66

5 Informal Networks and the Evolution of Industry …

Table 5.1 Rank of top 10 trends of 4-digit industrial subclassification in the value of shipments from 1980 to 2010 Year

1980

1990

2000

2010

Motor vehicles parts and accessories

1

1

2

1

Motor vehicles, including motorcycles

2

2

1

2

Miscellaneous musical instruments, parts and materials

3

3

3

18

Wood products, n.e.c., including bamboo and rattan

4

–

–

–

Miscellaneous manufacturing industries, n.e.c

5

–

–

–

General sawing and planing wood

6

19

35

–

Pianos

7

44

60

–

Electric audio equipment

8

11

9

4

Auxiliary equipment for internal combustion engines

9

23

21

6

Wooden furniture, except japanned

10

14

34

38

Note –: Not included in the top 60 Source Census of Manufacture

parts and materials” remained in third place until 2000 but fell to 18th place in 2010. A similar declining trend applies to “Wood products” and “Wooden furniture,” which were significantly lower in the rankings. Conversely, growth industries include electrical machinery and marine engine manufacturing such as “Auxiliary equipment for internal combustion engine” and “Miscellaneous electronic parts.” The role of R&D is very significant in such emerging industries, so a series of business workshops were established in the Hamamatsu area with various actors participating across all of these fields.

5.3 The Development of Informal Networks in the Hamamatsu Area 5.3.1 Institutional Support System for Informal Networks Since the 1980s, the Hamamatsu area has been supported by various regional science and technology promotional policies such as the Technopolis project. Therefore, this area features more industrial support organizations than other areas. The “Organization for Hamamatsu Technopolis”3 established many business workshops related to industries that are expected to grow and are considered important institutions. According to the 2010 business report of this organization, there were six business workshops that were held five to six times a year with various actors participating in each workshop. In addition, the Hamamatsu Chamber of Commerce and Industry hosted multiple business workshops and actively tried to foster innovation. After the start of the industrial cluster plan by METI in 2005, this organization established

5.3 The Development of Informal Networks in the Hamamatsu Area

67

five business workshops for the transportation industry and aerospace technology in addition to the opto-electronic industry. According to an interview with the Hamamatsu City Office, business workshops sponsored by private companies were rare, so the workshops run by industrial support organizations were very prominent in the Hamamatsu area. Table 5.2 lists the business workshops that were targeted by this study for analysis. These data were obtained directly from the industry support organizations that managed the workshops and by collecting lists published on websites. The business workshops in this study are categorized according to their establishment time—into those held between 1982 and 1998 (the first term) and those held starting in 2005 (the second term). The first term of business workshops was governed by the Organization for Hamamatsu Technopolis and the second term by the Hamamatsu Chamber of Commerce. Figure 5.3 depicts the distribution of participants in the Hamamatsu area business workshops. Many firms, universities, and public institutions were located in densely inhabited districts (DID) that extended from the southern part of the Hamamatsu area. Most large enterprises with over 300 employees participated in the business workshops in both the first and second terms. Many enterprises located in the two industrial parks in the central part of the Hamamatsu area also participated in the business workshops during both terms. In addition, some small and medium-sized enterprises (SMEs) located in the south did not participate in the business workshops in the first term but started to participate in the second term. Figure 5.4 shows the geographical distribution of the participants. Almost 8% of the participants were from the Hamamatsu area in both the first and second terms. There was a big difference between the two terms in the proportion of participants from outside the Hamamatsu area. Ten percent of the participants were from distant municipalities, except for Shizuoka and Aichi prefectures, in the first term. In the second term, there were only three participants from afar. The second term was characterized by many participants from Shizuoka prefecture, excluding the Hamamatsu area.

5.3.2 Relational Structure of Informal Networks Social network analysis was employed to examine the relational structure of the informal networks in the Hamamatsu area. Large collections of data are necessary to apply social network analysis to informal networks.4 In this study, in applying social network analysis, it was assumed that a link existed between the actors participating in the same business workshops. Table 5.3 shows the betweenness centrality5 of the top 10 actors in the first and second terms. Many actors who participated in multiple study groups in the first term and played a central role in building informal Hamamatsu area networks participated only in specific study groups in the second term. Although many actors who participated in multiple business workshops in the first term played a significant role in

Organization For Hamamatsu Technopolis Organization For Hamamatsu Technopolis Organization For Hamamatsu Technopolis Nakano Special Steel (Private company)

Software

All industries

All industries

Software

Electronics

All industries

Opt-electronics

Technoland Hosoe

TM plaza Hamamatsu

Hamamatsu technology All industries interchange plaza

All industries

Software industry promotion research

Miyakoda techno park

Hamamatsu system development cooperative

Research association for organic devices

Miyakoda associate

Semiconductor laser application research

1984

1983

1982

Year of foundation

Hamamatsu Chamber of Commerce & Industry

Organization For Hamamatsu Technopolis

Shizuoka University

Denkosha (Private company)

1998

1993

1991

1990

1990

1987

1987

Technical Support (Private company) 1986

Organization For Hamamatsu Technopolis

Venture business & All industries venture capital research

Organizer

Organization For Hamamatsu Technopolis

Technical field

Life science association Life science

Name of business workshops in the Hamamatsu area

1st term

Table 5.2 List of business workshops and gatherings in the Hamamatsu area

60

64

31

4

11

25

65

13

42

24

5

(continued)

No. of actors

68 5 Informal Networks and the Evolution of Industry …

General machinery

Precision technology research

Aerospace Life science Agriculture Opt-electronics All industries Motor cycle Motor cycle Aerospace

Space airplane technology

Hamamatsu medical-industrial collaboration

Hamamatsu agriculture-commerce-industry collaboration

Hamamatsu optical technology application

Next-generation design & manufactruing structrures

Transportation equipment industry strategy

Environmental cars pilot program conference

Hamamatsu aerospace project

Source Based on data from organizers and interviews

Technical field

1998

Year of foundation

Orion Tool (Private company located in Hamamatsu area)

Hamamatsu City Hall

Hamamatsu Chamber of Commerce & Industry

Organization For Hamamatsu Technopolis

Hamamatsu Chamber of Commerce & Industry

Hamamatsu Chamber of Commerce & Industry

Hamamatsu Chamber of Commerce & Industry

Hamamatsu Chamber of Commerce & Industry

Organizer

Organization For Hamamatsu Technopolis

Organizer

Name of business workshops in the Hamamatsu area

2nd term

Technical field

Name of business workshops in the Hamamatsu area

1st term

Table 5.2 (continued)

2010

2010

2009

2007

2006

2005

2005

2005

Year of foundation

65

9

15

49

15

23

48

101

51

No. of actors

No. of actors

5.3 The Development of Informal Networks in the Hamamatsu Area 69

70

5 Informal Networks and the Evolution of Industry …

Fig. 5.3 Distribution of actors in informal networks (Source Based on the list of business workshops and gatherings in the Hamamatsu area)

building informal networks, they participated only in specific business workshops in the second term. There are some companies, such as Sugiyama Media Support Corporation and Nihon Sekkei Kogyo Corporation, that did not participate in second-term business workshops. Examples of actors with a high value of betweenness centrality for both the first and second terms are NST, Denkosha, and Enomoto Kogyo Corporation. These are SMEs focused on R&D who had a significant incentive to participate in the business workshops. In addition, firms such as Technical Support Co. and Pulstec Industrial, which had relatively low betweenness centrality in the first term, played a significant role in linking the business workshops in the second term. As Table 5.2 indicates, Technical Support governed “Technoland Hosoe,” but in the second term, it participated in several business workshops such as “Space airplane technology” and “Hamamatsu medical-industrial collaboration” and existed as an important node in informal

5.3 The Development of Informal Networks in the Hamamatsu Area

71

Fig. 5.4 Geographical distribution of actors in informal networks (Source Based on the list of business workshops and gatherings in the Hamamatsu area)

networks. Pulstec Industrial had a high evaluation technology related to optics, so by participating in business workshops such as “Hamamatsu optical technology application” and the “Hamamatsu agriculture-commerce-industry collaboration,” it played the role of mediating between multiple workshops. As described above, the main actors with high betweenness centrality in both the first and second terms are the firms that have the strongest links with other industries. If the participants in each business workshop are fixed to a specific industry, a “structural hole” (Burt 1992) will occur between the various workshops. These structural holes hinder the flow of knowledge between actors in different industries. Firms with highly applicable technologies play a role in bridging such structural holes and efficiently conveying specialized information by participating in multiple workshops. Novel knowledge is circulated throughout fixed and closed networks by the functioning of actors characterized by high “liquidity” (Mizuno 2011).

72

5 Informal Networks and the Evolution of Industry …

Table 5.3 Top 10 actors by betweenness centrality 1st term Rank

Name of actors

Betweenness centrality

1

Hamamatsu Photonics

Industry

10.05

0

2

Sugiyama Media Support

Industry

7.85

–

3

Shizuoka University

Academia

7.15

0

4

Nihon Sekkei Kogyo

Industry

5.79

–

5

NST

Industry

5.07

6

NEC

Industry

4.22

7

Denkosha

Industry

4.1

1.89

8

Muramatsu Seiki

Industry

3.55

–

9

Kunimoto Industry Industry

2.49

0

10

Hamamatsu City Hall

2.42

–

Public

Value in 2nd term

11.19 –

2nd term Rank

Name of actors

1

NST

Betweenness centrality

Value in 1st term

Industry

11.19

5.07

2 3

Enomoto Kogyo

Industry

7.56

2.2

Pulstec Industrial

Industry

4.11

0.01

4

Yamaha Moror

Industry

2.7

0

5

Technical Support

Industry

2.12

0.13

6

Daiichi Kogyo

Industry

2.12

–

7

Tenryu Saw Mfg

Industry

2.12

–

8

Shinba Iron Works

Industry

1.92

1.3

9

PaPaLaB

Industry

1.89

–

10

Denkosha

Industry

1.89

4.1

Note –: Not participated in workshops Source Own calculations, based on data from organizers and interviews

5.4 Relationships Between Informal and Formal Networks Japan’s science and technology promotional policies have provided much needed support for industry-academia-public sector collaborations6 in the Hamamatsu area. This study examined the influence of informal networks on formal networks based on collaborative R&D. Table 5.4 shows eight science promotional policies that contributed to the establishment of formal networks in this study. A feature that these policies had in common was that the joint research projects were promoted

5.4 Relationships Between Informal and Formal Networks

73

Table 5.4 List of scientific support policies in the Hamamatsu area Ministries or institutions

Science and technology policies

Accepted year

No. of projects

No. of actors

JST

Joint-Research Project for Regional Intensive

2000

2

21

METI

Consortium R&D Project for Regional Revitalization

2001–2007

14

63

MEXT

Knowledge Cluster Initiative

2002, 2007

6

71

METI

Cross-Field-Partnership New Business Development Plans

2005–2015

30

94

JST

Comprehensive Support Programs for Creation of Regional Innovation

2006–2008

4

13

METI

Supporting Industry Project

2006–2014

28

102

CAO

Special Zones for Innovative Technology

2008

1

6

METI

Regional Innovation R&D Program

2008–2010

6

27

METI

Subsidiary program for experimental study on the creation of regional innovations

2012

1

3

METI

Measures to Support Global Technical Collaboration

2012

1

3

MEXT

Center of Innovation Science 2013 and Technology based Radical Innovation and Entrepreneurship Program

1

11

Source Based on data from ministries and institutions

under a specific technological theme to foster the creation of innovation and support start-ups. This study assumes that innovation is created through the following process. First, informal networks are built by participation in business workshops, fostering trust among the actors. Trust promotes the mutual learning of specialized knowledge and technology. Collaborative R&D then becomes possible by advancing the participants’ technical levels and their observation by others. Although not all collaborative R&D leads to innovation, formal networks based on trust relationships can create innovation. The innovations created through formal networks are abundant in the Hamamatsu area. For example, according to the ex-post evaluation report of the “Joint-Research Project for Regional Intensive,” 38 patent applications were achieved by 2005. This project also contributed to the establishment of universities related to optics and provided a pool of professional engineers in the Hamamatsu area. In the Knowledge Cluster Initiative, they also supported 20 advanced technology start-ups in fields such

74

5 Informal Networks and the Evolution of Industry …

as endoscopic surgery and digital camera systems, leading to the achievement of 254 patent applications. Figure 5.5 depicts the formal networks for eight science promotional policies using the network drawing program NetDraw. If two actors participated in two or more of the same collaborative R&D projects, the dyadic linkage is illustrated by a heavy line and the size of each node is proportional to the betweenness centrality. As Fig. 5.5 suggests, Shizuoka University functioned as a hub, directly or indirectly linking each actor with others. In addition, Hamamatsu University of Medical Science, Hamamatsu Industrial Research Institute of Shizuoka Prefecture, and Hamamatsu Photonics collaborated on multiple R&D projects to establish strong ties. There are 15 small-scale components composed of two to four actors, but they are composed only of private firms. In this study, we examined the relationship between the structural positions of actors in informal networks based on business workshops and whether the actors participated in collaborative R&D. Table 5.5 presents the results of the Mann– Whitney U test for the degree centrality and betweenness centrality of actors in informal networks. In the first term, the degree centrality and betweenness centrality of the actors that participated in collaborative R&D were higher than those of actors that did not participate. This result clearly demonstrates that actors with high network centrality tended to build formal networks through collaborative R&D. The possibility of collaboration is increased by participating in various business workshops with many participants.

Shizuoka University

Hamamatsu University School of Medicine

Hamamatsu Photonics

Industry Academia Public sector

Fig. 5.5 Relational structure of formal networks (Source Based on data from ministries and institutions)

5.4 Relationships Between Informal and Formal Networks

75

Table 5.5 Mann–Whitney U-test for the network centrality of actors in formal networks Term

1st term

Centrality

Degree centrality

2nd term Betweeness centrality

Degree centrality

Betweeness centrality

Participated in project No. of actors

33

33

30

30

Average rank

183.91

182.85

123.95

137.47

232

198

198

Non-participated in project No. of actors

232

Average rank

125.76

125.91

113.07

111.02

Mann–Whitney U

2148

2183

2686.5

2281

Two sided p-value

0.000

0.000

0.387

0.004

Source Own calculations, based on data from ministries and institutions

In the second term, the betweenness centrality of actors that participated in collaborative R&D was statistically high, however, the degree centrality was not statistically different. This suggests that it was more important to participate in many business workshops and deepen interactions, rather than to participate in only one workshop in the second term. To start collaborative R&D projects, it was necessary to search for appropriate partners by participating in multiple business workshops. If a business workshop is held for many years, a closed or fixed network will be formed. In such network, the number of homogeneous actors increases and knowledge diffusion becomes limited. As Fig. 5.4 indicates, in the first term, relatively speaking, there were many participants from outside the Shizuoka or Aichi Prefectures, so it was possible for novel knowledge to flow into the Hamamatsu area. Business workshops involving actors inside and outside the Hamamatsu area supported access to a diverse pool of knowledge and contributed to the formation of a formal innovation network. Through this process, the Hamamatsu area seems to have avoided “cognitive lock-in” (Grabher 1993), which restricts entrepreneurship and innovation. As Table 5.2 indicates, many of the business workshops held in the second term were focused on cutting-edge technologies such as next-generation automobiles and the aerospace industry. Even if the number of participants was small, a formal network would be formed with actors who shared a common purpose by participating in the business workshops, which were highly specialized and expected to have a high market value in the future. By participating in many business workshops, actors active in collaborative R&D will be able to collect leading-edge knowledge and market information that cannot be obtained by participating in only one workshop. The progress of these innovation processes has contributed to the advancement of agglomerations in the Hamamatsu area.

76

5 Informal Networks and the Evolution of Industry …

5.5 Conclusion The industrial structure has significantly changed in the Hamamatsu area. To respond to changes in the socioeconomic environment, various business workshops have been held with the aim of promoting specific industries such as the automobile and the opto-electronic industries. The support system for informal networks in the Hamamatsu area differed greatly between the first term, which was governed by the Organization for Hamamatsu Technopolis, and the second term, which was governed by the Hamamatsu Chamber of Commerce. The findings of this study reveal differences related to geographical distribution and network structure according to the type of actor (i.e., industry, academia, or public institution). It is also apparent that participating in multiple business workshops in a specific technological field plays an important role in innovation and knowledge creation by transferring heterogeneous knowledge among the various Hamamatsu area workshops. In the first term, actors who successfully formed formal networks participated in large workshops. In the second term, the high level of “liquidity” of the workshop participants was important at the start of collaborative R&D. This suggests that it is important that workshop participants are not fixed and that they can connect with a variety of actors. Participation in various business workshops by actors with advanced knowledge and market information contributes to the formation of formal networks. We suggest that informal cooperation among the industrial support institutions in the Hamamatsu area maintained high liquidity in the workshops. When many workshops are held in an agglomeration such as in the Hamamatsu area, a high degree of redundancy7 in the workshops can become a problem. According to an interview with the Hamamatsu Chamber of Commerce and Industry, in the Hamamatsu area, there was frequent coordination among public support institutions so that the workshop themes for the first and second terms would not be redundant. This institutional thickness contributed to the efficient transfer of novel knowledge and information in the Hamamatsu area. The implication of this study is that industry support institutions in the Hamamatsu area should continue to design business workshops so that informal networks can function and avoid cognitive lock-in. The commitment of large firms inside and outside of the Hamamatsu area will be essential for effective learning success. According to interviews with the Hamamatsu Chamber of Commerce and Industry and the Hamamatsu City Office, business workshops in the Hamamatsu area are focused on improving SMEs’ technological capabilities and supporting the search for seeds of innovation; large firms with novel knowledge and technology did not participate in the workshops often. Most of the highly centralized actors were not large firms but SMEs in the Hamamatsu area, as demonstrated in Table 5.3. The Hamamatsu Chamber of Commerce and the Hamamatsu City Hall argued that there was a need to develop new markets and circulate knowledge using the unique networks of large firms. In the future, it is expected that open innovation will be promoted

5.5 Conclusion

77

by large firms by donating contributions and human resources to various science projects in the Hamamatsu area. Notes 1. 2. 3. 4. 5.

6.

7.

The term of “Optronics” comes from “optics” and “electronics”. The Census of Manufacture is implemented by METI and lists the top 60 fourdigit industries in the value of shipments by industrial district or regional block. This organization was renamed “Hamamatsu Agency for Innovation” when it was integrated with related organizations in 2012. Yokura (2016) and Yokura et al. (2013) conducted social network analysis on relational structures wherein various actors participate. Betweenness centrality is defined as “the ratio of the number of the shortest paths between two nodes passing a node to the number of all possible such shortest paths in a graph” (Hung and Wang 2010: 125). Yokura et al. (2013) used collections of data of private firms, universities/colleges, and public research institutions on the Consortium R&D Project for Regional Revitalization by METI. There are two different perspectives on whether redundancy is a strength or a weakness. Hanzawa and Yamamoto (2017) focus on the strength of redundancy in the innovation of highly uncertain industries.

References Amin A, Thrift N (1992) Neo-Marshallian nodes in global networks. Int J Urban Reg Res 16:571– 587 Bathelt H, Malmberg A, Maskell P (2004) Clusters and knowledge: local buzz, global pipelines and the process of knowledge creation. Prog Hum Geogr 28:31–56 Bathelt H, Schuldt N (2008) Between luminaires and meat grinders: international trade fairs as temporary clusters. Reg Stud 42:853–868 Bathelt H, Schuldt N (2010) International trade fairs and global buzz, part I: ecology of global buzz. Eur Plan Stud 18:1957–1974 Burt RS (1992) Structural holes: the social structure of competition. Harvard University Press, Cambridge Grabher G (1993) The weakness of strong ties: the lock-in of regional development in the Ruhr area. In: Grabher G (ed) The embedded firm: on the socioeconomics of industrial networks. Routledge, London and New York Hanzawa S (2016) Contents sangyo to innovation: terebi anime game sangyo no shusheki (Content industries and innovation: The agglomerations of the television production, animation and console videogame industries). Keisoshobo: Tokyo (in Japanese) Hanzawa S, Yamamoto D (2017) Recasting the agglomeration benefits for innovation in a hits-based cultural industry: evidence from the Japanese console videogame industry. Geogr Ann B Hum Geogr 99:59–78 Hung S-W, Wang A-P (2010) Examining the small world phenomenon in the patent citation network: a case study of the radio frequency identification (RFID) network. Scientometrics 82:121–134 Maskell P, Bathelt H, Malmberg A (2006) Building global knowledge pipelines: the role of temporary clusters. Eur Plan Stud 14:997–1013

78

5 Informal Networks and the Evolution of Industry …

Mizuno M (2011) Innovation no keizai kuukan (Economics spaces of innovation). Kyoto Daigaku Shuppan-kai, Kyoto (in Japanese) Mizuno M (2019) Sangyo shuseki to network heno shinkateki approach: Utrecht gakuha no jissyo kenkyu wo chushinni (Evolutionary approaches to regional clusters and networks: a review of empirical studies by ‘the Utrecht school’ of evolutionary economic geography). Keizaichirigaku Nempo (Ann Jpn Assoc Econ Geogr) 65:239–259 (in Japanese with English abstract) Nagayama M (2009) Atarashii sangyo shuseki no keisei mechanism: Hamamatsu chiiki to Sapporo chiiki no software shuseki keisei ni okeru spin-off rensa (The mechanism of new industrial cluster creation: chain reaction of spin-off in creation process of Sapporo and Hamamatsu software clusters). Mita Gakkai Zasshi (Keio J Econ) 101:741–768 (in Japanese) Otsuka M (1986) Chiho toshi kogyo no chiiki kouzo: Hamamatsu Technopolis no keisei to tembo (The regional structure of provincial urban industry: formation and prospect of the Hamamatsu Technopolis). Kokonshoin, Tokyo (in Japanese) Sternberg R (1995) Supporting peripheral economies or industrial policy in favour of national growth? An empirically based analysis of goal achievement of the Japanese ‘Technopolis’ program. Environ Plan C: Polit Space 13:425–439 Storper M (1997) The regional world: territorial development in a global economy. The Guilford Press, New York Yamamoto K (2002) Gakusyuu suru chiiki toshiteno naganoken suwa okaya chiiki: kikai kinzoku kogyo gijyutu no gakusyu to kakushin (The Suwa-Okaya district of Japan as a learning region: learning and innovation in technology and machine and metal manufacturing skills). Keiz Shirin (Hosei Univ Econ Rev) 69(4):271–302 (in Japanese with English abstract) Yokura Y (2012) Kyodo kenkyu kaihatsu no kankei kouzou to kuukanteki pattern: Chiiki kesyuu gata kyodo kenkyu kaihatsu jigyo wo jirei toshite. Tokyo Daigaku Jinbun Chirigaku Kenkyu (Komaba Stud Hum Geogr) 20:39–56 (in Japanese) Yokura Y (2014) Performances and roles of local trade fairs in Japan: case study on the Suwa area industrial messe, Nagano prefecture. Komaba Stud Hum Geogr 21:85–100 Yokura Y (2015) Building relationships at local trade fairs in Japan: a case study of the Suwa Area Industrial Messe. In: Bathelt H, Zeng G (eds) Temporary knowledge ecologies: the rise of trade fairs in the Asia-Pacific region. Edward Elgar, Cheltenham Yokura Y (2016) Temporary space and business-matching networks of the semiconductor industry in Kyushu, Japan. J Int Econ Stud 30:3–12 Yokura Y, Matsubara H, Sternberg R (2013) R&D networks and regional innovation: a social network analysis of joint research projects in Japan. Area 45:493–503

Chapter 6

Institutional Thickness in Regional Innovation Ecosystem: A Case Study of the Kyushu Semiconductor Industry

Abstract Japan’s industrial cluster policies have reinforced Kyushu’s semiconductor industry by attempting to gain a competitive advantage in system large-scale integration (LSI) and advanced three-dimensional semiconductor packaging. In this chapter, we discuss two promotional semiconductor-related industry projects that were established by a regional think-tank called the Kyushu Economic Research Center. The first project is an international workshop on semiconductor assembly and packaging technologies for building global networks. The workshop has become an important event for both domestic and foreign firms and has strengthened existing networks. However, the lack of a coordinator has hindered the establishment and maintenance of inter-organizational relationships. The second project is support for business-matching for semiconductor-related firms, which helped build mutual trust between firms with low mutual recognition and established new business relationships with firms outside Kyushu. Keywords Innovation ecosystem · Transactional relationships · Semiconductor industry · Social network analysis

6.1 Introduction The theory of innovation ecosystems—in which an analogy is drawn between technological ecosystems and biological ecosystems—reflects the interrelationships between the various actors such as firms, governments, universities, researchers, consumers, and workers whose aim is to create and foster innovation (Fransman 2018; Frenkel and Maital 2014: 12). Based on this ecosystem concept, Japanese policymakers have been promoting the regional innovation ecosystems project since 2016. The origin of innovation policy focused on technological development in Japan can be traced back to the “Technopolis” project (Sternberg 1995) by the Japanese Ministry of International Trade and Industry (METI). After the establishment of the Technopolis project in 1980, Kyushu’s semiconductor industry was regarded as a “sunrise industry” by local governments and small- and medium-sized enterprises

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2021 Y. Yokura, Regional Innovation and Networks in Japan, International Perspectives in Geography 16, https://doi.org/10.1007/978-981-16-2191-8_6

79

80

6 Institutional Thickness in Regional Innovation …

(SMEs) in Kyushu. Since then, the Japanese government has reinforced Kyushu’s semiconductor industry with the aim of strategically fostering industries. Since 2001, METI’s Industrial Cluster Program has promoted the Kyushu Silicone Cluster Plan and supported the establishment of semiconductor-related SMEs located in Fukuoka, Kumamoto, Oita prefectures in Kyushu. Especially in Fukuoka Prefecture, many public R&D institutions of the semiconductor industry, such as the Research Center for Three-Dimensional Semiconductors and the Experimental Center for Social System Technologies, were constructed under the leadership of the governor. The Ministry of Education, Culture, Sports, Science, and Technology (MEXT) has also aimed to promote system LSI technology in Kyushu. Since 2002, the Knowledge Cluster Initiative, launched by MEXT, has supported semiconductor cluster in the northern Kyushu area. Yamasaki (2003) suggested that Kyushu Silicone Cluster Plan by METI was important for the following reasons: (1) Kyushu’s regional cluster of semiconductor-related firms; (2) the location of universities and colleges that can become collaborative R&D partners; (3) international conferences on semiconductors; (4) the expansion of venture companies from outside of Kyushu; and (5) as the site of Sony’s semiconductor division’s head office. Many mass production factories of major semiconductor companies are located in Kyushu that Kyushu is often called “Silicon Island” (Kondo 2017). However, under the influence of the rise of newly industrialized economies, Kyushu has shifted its focus from mass-production facilities to becoming a research hub of semiconductorrelated industry. By establishing a nexus of diverse actors located both inside and outside of Kyushu, the semiconductor-related industry has aimed to advance the industry. This chapter focuses on institutional thickness in Kyushu. It discusses two promotional semiconductor-related industry projects that were established by a regional think-tank called the Kyushu Economic Research Center (KERC). The first is to hold an international workshop on three-dimensional semiconductor packaging technologies for building global networks. The second is a support project of business-matching for semiconductor-related firms. Temporary clusters offer critical spaces wherein trust between various actors is established through individual events such as international conferences (Maskell et al. 2004, 2006; Bathelt and Schuldt 2008, 2010; Yokura 2014). This chapter conducted social network analysis to explore the structure of temporary relationships. In June 2011, interviews were conducted with the research staff of the KERC to collect data regarding previous business-matching support. An earlier version of this chapter was previously published in Yokura (2016). In June 2020, another interview with the KERC regarding the development of the two projects was conducted. In the next section, we survey the historical circumstances of Kyushu’s semiconductor industry using statistical materials. In Sect. 6.3, we examine the diverse functions of international workshop on Kyushu’s semiconductor industry. In Sect. 6.4 we explain the influence of business-matching support projects on relationships and reports the results of social network analysis designed to understand the evolution of inter-firm networks. Section 6.5 concludes.

6.2 Kyushu’s Semiconductor-Related Industry

81

6.2 Kyushu’s Semiconductor-Related Industry The evolution of Kyushu as the center of semiconductor production in Japan was significantly influenced by the major electrical equipment manufacturers. In 1967, Mitsubishi Electric constructed a mass production factory specializing in the preprocess of semiconductor manufacturing in Kumamoto. In addition, Kyushu NEC (founded in 1969) and Sony Kokubu (founded in 1973) also began production of semiconductor in Kyushu. The barriers to entry to the semiconductor production were comparatively low when major semiconductor device makers started operations in Kyushu. Therefore, new entrants from outside the industry such as local wholesalers and shipping companies established semiconductor subcontracting firms that formed industrial clusters in Kyushu. When the demand for integrated circuits (ICs) drastically increased in the mid-1970s, Kyushu’s semiconductor-related industry, which specialized in dynamic random access memory (DRAM), rapidly expanded. By 1985, Kyushu accounted for 30 percent of the country’s semiconductor production. However, many major semiconductor-related manufacturers shifted their laborintensive processes and products overseas because of the Japanese yen’s rapid appreciation after the Plaza Accord in 1985. Furthermore, low-cost competition intensified as Asian countries competed for semiconductor manufacturing. As a result, Kyushu’s semiconductor industry shifted its DRAM production to other Asian countries such as Taiwan and began specializing in the high value-added production of LSI circuit manufacturing. Figure 6.1 illustrates trend of the production value in IC commodities in Kyushu between 1990 and 2015. Although the production value of metal oxide semiconductor (MOS) logic ICs—used as a logic division of system LSI—was increasing at the time, the production value of MOS memory IC began to drastically decrease in the 2000s. As Fig. 6.2 shows, there were many employees in the prefectures in which the production bases of major semiconductor companies were located. In the Tohoku region, many electronic component manufacturing companies were located near the highway; the specialization coefficient of each prefecture is therefore high. The semiconductor industrial clustering along the Tohoku Expressway is called “Silicon Road.” In addition, the specialization coefficient of each prefecture is over three in the Kyushu region. Figure 6.2 also indicates a high degree of specialization in the semiconductor industry, especially in Kumamoto, Kagoshima and Oita Prefectures. Figure 6.3 presents the trend in the semiconductor-related industry’s exports to countries and regions in the Kyushu economic area. Because the semiconductor industry in Kyushu mainly targets the Asian market, exports to the United States and the European Union are small. Although exports to ASEAN countries are declining today, the presence of South Korea and China in the semiconductor market is still noteworthy. Kyushu’s semiconductor industry is the most competitive in the high-density assembly and packaging technologies. In the process of the miniaturization of mobile phones and personal digital assistants, the technology to mount LSI chips into the modules was critical. Kyushu’s semiconductor industry also has a competitive

82

6 Institutional Thickness in Regional Innovation …

Fig. 6.1 Trend of the production value according to the integrated-circuit commodity item in Kyushu (Source Kyushu Bureau of Economic, Trade, and Industry Report)

advantage in high-density assembly and packaging technologies such as Systemin-Package (SiP) and three-dimensional packaging. Several local governments and public industrial institutions devote remarkable effort to forming global nexus to advance Kyushu’s semiconductor industry.

6.3 Various Roles of International Semiconductor-Related Workshops The Japan Science and Technology Agency has supported the growth of Kyushu’s semiconductor industry by several grants in the Joint-Research Project for Regional Intensive from 1997 to 2002. This project contributed to building relationships between Kyushu’s semiconductor-related firms and researchers. Many SMEs in Kyushu called for the establishment of an international workshop with the aim of promoting the advanced semiconductor technology and building personnel networks both at home and abroad. In 2001, METI Kyushu and others hosted the first International Workshop on Microelectronics Assembling and Packaging (MAP) to support

6.3 Various Roles of International Semiconductor-Related Workshops

83

Fig. 6.2 Distribution of semiconductor-related employees (Source Census of Manufacture)

SMEs in Kyushu build relationships with foreign firms. The aim of MAP was as follows: To facilitate the creation of a value network by linking Kyushu’s over 700 IC-affiliated companies with companies in Asia through open discussion and the sharing of state-ofthe-art ideas, techniques, and perspectives on microelectronics assembling and packaging among leading industrial companies and academia. (Excerpt from the MAP brochure)

The important role of MAP is to provide participants with an opportunity to attend poster exhibitions or oral presentation in English. Most of the participants were engineers from semiconductor-related firms, but there were also presentations by university researchers. Figure 6.4 shows that both the number of poster exhibitions

84

6 Institutional Thickness in Regional Innovation …

Fig. 6.3 Trend of the exports to primary countries and regions of the semiconductor-related industry in the Kyushu economic area (Source Trade Statistics of Japan [Moji Customs])

and oral presentations increased in 2003 when business meetings for buyers and suppliers began during MAP. There were a significant number of poster exhibitions from China in 2003 and 2004 and from South Korea in 2007. Furthermore, there were many poster exhibitions from India in 2010 because MAP’s executive committee specifically invited Indian buyers to MAP to expand Kyushu’s overseas market. As shown in Fig. 6.4(b), the ratio of oral presentations from overseas has fluctuated between 25 and 45 percent. In recent years, the number of participants from the U.S. and the EU has also increased. Thus, the nationalities of the presenters have diversified with each MAP event. The number of oral presentations by foreigners peaked in 2008 with India having a prominent presence at the event. This is because the India Semiconductor Association signed a memorandum of understanding with the Asia Semiconductor Trading Support Association (located in Kyushu) in 2008. As Fig. 6.5 shows, there is a strong link between the number of MAP participants and the new-order ratio (defined as the number of new orders divided by the total number of business negotiations), which indicates that MAP has been effective at expanding the semiconductor’s overseas market. The new-order ratio was high in both 2003 and 2004 (e.g., there were 70 new orders in 2003 and 182 in 2004), which led to the rapid increase in the number of participants at MAP in 2005. However, the new-order ratio dropped significantly in 2007 (there were only 29 new orders). Thus, in 2008, the number of both domestic and foreign participants decreased. MAP’s registration brochures list the number of successful deals in the past and give

6.3 Various Roles of International Semiconductor-Related Workshops Fig. 6.4 Trend of the number of poster exhibitions (a) and oral presentations (b) at MAP by country and region (Source Based on date from Kyushu Economic Research Center)

85

86

6 Institutional Thickness in Regional Innovation …

Fig. 6.5 Trend of the number of participants at MAP and the new-order ratio (Source Based on date from Kyushu Economic Research Center)

participants information on the business opportunities. The executive committee also makes a list of semiconductor-related firms located in Kyushu and East Asia and provides them to participants. Therefore, MAP not only provides semiconductorrelated firms with the opportunities to publish their research results but also offers a place for business negotiations. MAP also plays a critical role in strengthening existing personnel relationships. According to interviews with KERC’s research staff, it was important for MAP participants to openly exchange information. Almost half of the participants were “repeat participants” and semiconductor engineers, who contributed to reinforcing MAP’s personnel networks. In addition, engineers who participated in MAP sometimes established semiconductor-related entrepreneurial ventures together. Thus, MAP has facilitated the mobility of specialized human resources in Kyushu.

6.4 Formation of Transactional Relationships …

87

6.4 Formation of Transactional Relationships at Business-Matching Project The Kyushu Semiconductor & Electronics Technology Innovation Association (known as SIIQ) was established to support the Kyushu’s semiconductor industries as the organization operating METI’s Industrial Cluster Program. Since July 2008, SIIQ has promoted business-matching between Kyushu’s semiconductor firms through the project called “SIIQ DIRECT.” The KERC was in charge of SIIQ DIRECT, which was designed to exploit Kyushu’s network of semiconductor firms to help SMEs finding new customers. SIIQ DIRECT also had management capabilities—introducing appropriate suppliers to customers according to each customer’s requests. The SIIQ DIRECT coordinators, who were familiar with the semiconductor industry, played an important role in building relationships between customers and suppliers in Kyushu. Most customers were private firms since it was infrequent for universities and public research institutions to become customers. In fact, there were 247 customers that utilized SIIQ DIRECT in 2010, but only 15 of them were non-private firms. In the early stages of SIIQ DIRECT in 2008, the formation of transactional relationships within Kyushu was the most important goal. In other words, it was assumed that both suppliers and customers were in Kyushu. However, the person in charge of SIIQ DIRECT believed that it was important to build close relationships throughout East Asia along with Kyushu’s local network for the growth of the Kyushu’s semiconductor industry. This enabled firms located outside Kyushu to utilize SIIQ DIRECT. We illustrate SIIQ DIRECT’s customer–supplier networks from 2008 to 2010 by using the NetDraw software (Fig. 6.6). The transactional relationships established in the SIIQ DIRECT are represented by the thick lines. Figure 6.6 also shows SIIQ DIRECT’s network flows by adding an arrow from the customer to the supplier. The size of each node is proportional to the degree centrality, which is the sum of the indegree and outdegree, and each node’s color is dependent on its location (i.e., within or outside Kyushu). As shown in Fig. 6.6, although it was a sparsely linked network in 2008, it later developed into a dense network as specific nodes grew into hubs. In other words, specific nodes were suppliers as well as customers, and they were able to many build business relationships. Table 6.1 shows the descriptive statistics of the top six firms by degree centrality in 2010. Five of the six firms, except for Firm F, had been using SIIQ DIRECT project since 2008. Firm F was located outside Kyushu and participated in the project for the first time in 2010. Although six firms succeeded in establishing business relationships in 2010, the ratio of having transactional relationships by Firm B was relatively low. Firm A received many orders in 2008 and was able to build many business relationships in 2010 because of its competitive advantage in semiconductor assembly and packaging technology. Although Firm B used SIIQ DIRECT as a customer in 2008, but in 2010 it become a supplier and participated in the project. Firms C, D, and E received orders as suppliers in 2008 and 2009 and also used SIIQ

88

6 Institutional Thickness in Regional Innovation …

2008

within Kyushu outside Kyushu

2009

A

2010

Customer

Supplier

Fig. 6.6 Inter-firm network in the SIIQ DIRECT project (Source Based on date from SIIQ DIRECT)

6.4 Formation of Transactional Relationships …

89

Table 6.1 Top six firms by degree centrality Firms

Year

Indegree

Outdegree

Degree centrality (a)

No. of transactional relationships (b)

Ratio(b/a)

A

2008

17

4

21

10

0.48

2009

28

8

36

18

0.50

2010

78

58

136

86

0.63

2008

2

9

11

2

0.18

2009

17

22

39

6

0.15

2010

25

33

58

14

0.24

2008

3

1

4

0

0.00

2009

8

1

9

0

0.00

2010

11

41

52

26

0.50

2008

2

2

4

2

0.50

2009

4

3

7

3

0.43

2010

4

34

38

34

0.89

2008

1

0

1

0

0.00

2009

1

1

2

1

0.50

2010

3

30

33

16

0.48

2008

0

0

0

0

–

2009

0

0

0

0

–

2010

33

0

33

26

0.79

B

C

D

E

F

Source Own calculations, based on date from SIIQ DIRECT

DIRECT project as customers in 2010. On the other hand, Firm F was specialized in semiconductor substrate design and product evaluation technology, but it had not received orders from firms located in Kyushu. With the participation in SIIQ DIRECT project, the number of orders had increased in 2010. Figure 6.7 shows the network between the top six firms. Firm B made an inquiry to Firm A regarding the development and trial manufacturing of semiconductor device in 2008, but no transactional relationships was built at that time. Although the network between the top six firms did not change in 2009, but the relationships diversified in 2010. Firms A and B established a transactional relationship in semiconductor assembly and packaging in 2010. This was a complementary relationship as the two firms demanded advanced technology for semiconductor mounting. They had continuously exchanged information on semiconductor assembly and packaging since 2008, and mutual trust has been established. Firm E used SIIQ DIRECT project less until 2009, but the business relationship between Firm A and Firm E in semiconductor mounting was built in 2010. In addition, Firms A, C, and E were seeking new business partners, and in 2010 formed a relationship with Firm F, which had cutting-edge semiconductor technology. As a result, a close relationship was built between the top six firms.

90

6 Institutional Thickness in Regional Innovation …

E

B

B

E A

A

F

F C

D

C

D

2008 and 2009

2010 Customer

Supplier

Fig. 6.7 Customer–supplier networks among the top six firms by degree centrality (Source Based on date from SIIQ DIRECT)

6.5 Toward Sustainable Support for Semiconductor Innovation Ecosystem Japan’s cluster policies by METI and MEXT in 2000s have supported the semiconductor innovation ecosystem in Kyushu by having a competitive advantage in system LSI and three-Dimensional packaging technologies. Since then, various actors in the Kyushu’s semiconductor industry had adapted to changes in the external environment, such as the reorganization of large firms and’ technological “catching-up” with Japan in East Asian countries. According to an interview with the KERC, two projects were successful, but after that, continuing support for Kyushu’s semiconductor industrial agglomeration was discontinued because a valuable human resource who coordinated international workshops died in 2016, and the scale of the event decreased significantly at that time. In addition, the SIIQ DIRECT project was reduced in scale since a semiconductorrelated budget was no longer available from the government. Although the MAP has become an important event for domestic and foreign firms, strengthening existing networks, but the lack of coordinators hindered the establishment and maintenance of inter-organizational relationships. SIIQ DIRECT helped

6.5 Toward Sustainable Support for Semiconductor Innovation Ecosystem

91

build mutual trust between firms with low mutual recognition and establish new business relationships with firms outside Kyushu. The relational structure of networks in SIIQ DIRECT project was sparsely connected in 2008 and 2009 because it was difficult for firms with little degree centrality to exchange information. However, with the employment of knowledgeable coordinators in the semiconductor industry in 2010, the inter-firm network has evolved to link various actors. Finally, in order to promote the competitive advantage of the semiconductor industry in Kyushu both domestically and internationally, it is important to facilitate the business-matching project. In this case, strong ties with offshore firms would be useful. Furthermore, when various countries collaborate in the specialized area of semiconductor technology, they will help industrial advancement of the semiconductor industry in Kyushu.

References Bathelt H, Schuldt N (2008) Temporary face-to-face contact and the ecologies of global and virtual buzz. SPACES online 6(2008–04):1–23, Toronto and Heidelberg.www.spaces-online.com Bathelt H, Schuldt N (2010) International trade fairs and global buzz, part I: ecology of global buzz. Eur Plan Stud 18:1957–1974 Fransman M (2018) Innovation ecosystems: increasing competitiveness. Cambridge University Press, Cambridge Frenkel A, Maital S (2014) Mapping national innovation ecosystems: foundations for policy consensus. Edward Elgar, Cheltenham Kondo A (2017) Sangyo ritchi kara mita nihon no handoutai sangyo no kyousou retsui heno ichi kousatsu (A study on the development and international competitiveness of the Japanese semiconductor industry: from the viewpoints industrial location). Keizai Ronshu (J Econ, Kumamoto Gakuen Univ) 23:247–262 (in Japanese with English abstract) Maskell P, Bathelt H, Malmberg A (2004) Temporary clusters and knowledge creation: the effects of international trade fairs, conventions and other professional gatherings. SPACES online (2004– 04):1–34, Toronto and Heidelberg. www.spaces-online.com Maskell P, Bathelt H, Malmberg A (2006) Building global knowledge pipelines: the role of temporary clusters. Eur Plan Stud 14:997–1013 Sternberg R (1995) Supporting peripheral economies or industrial policy in favour of national growth? an empirically based analysis of goal achievement of the Japanese ‘Technopolis’ program. Environ Plan c Politics Space 13:425–439 Yamasaki A (2003) Chiiki sangyo seisaku toshiteno cluster keikaku (Cluster program as a regional industrial policy). In: Ishikura Y, Fujita M, Maeda N, Kanai K, Yamasaki A (eds) Nihon no sangyo cluster senryaku: chiiki ni okeru kyosou yuui no kakuritsu (Strategy for cluster initiatives in Japan: establishment of the competitive advantage in regions). Yuhikaku, Tokyo (in Japanese) Yokura Y (2014) Performances and roles of local trade fairs in Japan: case study on the Suwa area industrial messe, Nagano prefecture. Komaba Stud Hum Geogr 21:85–100 Yokura Y (2016) Temporary space and business-matching networks of the semiconductor industry in Kyushu, Japan. J Int Econ Stud 30:3–12

Chapter 7

Global Knowledge Flows and Corporate Values

Abstract This chapter designed a database of actual cases of foreign investments and visualized relationship structures by using social network analysis to identify knowledge flows among global firms. A model was built to examine which actors in what sort of structural positions are showing good economic performance. The following findings were obtained. First, network statistics were used to classify 18 sectors of manufacturing industry. The statistics were compared through a visualization of related structures. From this, groups were extracted where certain Japanese and foreign textile, pharmaceutical, and electric appliances firms have relationships with nodes positioned on the periphery of their networks, contributing to the formation of huge components. Second, groups were identified such as food firms that have small-scale components with limited technology as compared to other sectors. Further, sectors such as machinery have many actors whose networks, however, are split so that they have no hubs that work as a key to facilitate the circulation of knowledge flows. Finally, comparisons were made among sectors through a simultaneous analysis of multiple populations. The influence of a network advantage worked strongly in specific industry sectors that required scientific and analytical knowledge. Keywords Knowledge flow · Corporate value · Social network analysis · Covariance structure analysis

7.1 Introduction Various linkages with actors outside of industrial clusters are thought to spur inflows of new knowledge and information, help overcome the lock-in problems that are caused by limited and rigid relations, and also promote innovation within industrial clusters. In Chap. 3, the argument focuses on the process of innovation that differs significantly depending on the attributes of the knowledge bases that industries rely on, pointing out that the existence of a highly skilled local labor force and the combining of knowledge across organizations beyond the local level are factors that lead to innovation within industrial agglomerations.

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2021 Y. Yokura, Regional Innovation and Networks in Japan, International Perspectives in Geography 16, https://doi.org/10.1007/978-981-16-2191-8_7

93

94

7 Global Knowledge Flows and Corporate Values

Although networks are a key concept in the research trend described above, the importance of adopting an approach that focuses on the structures of relationships among actors may be mentioned as an empirical issue. Network research in the field of economic geography has conventionally consisted primarily of qualitative descriptions focusing on the social contexts and systems of industrial agglomerations. As analytical tools developed, more focus has come to be placed on approaches that analyze relationships among actors quantitatively and consider individual actors’ activities and performance in the context of relationship structure. In other words, as evidenced in the works of Wakabayashi (2006) and Graf (2006), research that applied social network analysis in inter-organizational business relations and knowledge flows and focused on the structural positions of actors within networks was being accumulated. Yokura et al. (2013) focused on knowledge flows in joint research and development (R&D). Using exhaustive data on participating actors in R&D project consortiums for regional revitalization, Yokura et al. (2013) employed social network analysis to examine the relationship structure of networks based on shared R&D topics. They then classified the research areas dealing with life sciences, information and communication, and manufacturing technology into “manufacturing (monozukuri)” and “science” types based on engineering knowledge and practical technology and scientific knowledge, respectively. Yokura et al. (2013) found that while local actors play central roles in the manufacturing and science types of research areas, there is still a requirement for actors from a broad range of geographical areas. However, such joint development projects that encompass broad geographical areas are limited to Japan, and the need for introducing a global perspective is discussed as a future challenge of research on knowledge flow. In a discussion on the spatiality of innovation, Matsubara (2007) observed that the situational analysis of the actual state of the acquisition and combination of global knowledge by multinational corporations was accumulating primarily in the field of international business management. However, research attempting to quantify global knowledge flows is just getting started. Although studies such as an analysis of coauthored articles in the biosciences field by Cooke (2009) and analyses by the OECD (2009a, b) of joint international patent applications have been undertaken, the number of such studies remains limited. Therefore, our study begins with a quantitative and metrical examination of inter-firm relationships in terms of knowledge flows based on company-level data regarding technological alliances and ownership relationships among global corporations. In Sect. 7.2, we describe our analytical data and methodologies and then use social network analysis to visualize the relationship structure of networks according to manufacturing sectors and perform a comparative analysis. After that, we consider the types of economic actors that play central roles in global knowledge flows. In Sect. 7.3, we present a model to examine the types of structural positions in which economic actors are performing well in the global inter-firm network. Using this model, we examine the differences in degrees of influences of network characteristics on corporate performances among different sectors. Finally, in Sect. 7.4, we summarize our research findings and discuss topics for future research.

7.2 Global Knowledge Flows and Inter-firm Relationship Structures

95

7.2 Global Knowledge Flows and Inter-firm Relationship Structures 7.2.1 Analytical Data The analytical data used in this study come from Kigyo betsu gaishi donyu soran (Jojo kigyo hen) [Guide to foreign capital investment by company (Listed firms)], published by the Keizai Chosa Kyokai.1 This document contains summaries on foreign capital investment in 1,083 firms listed on the first and second sections of the Tokyo Stock Exchange and that have introduced foreign investment. Our analytical sample was narrowed down to 760 manufacturing firms; we built a database containing the names of the foreign firms, their nationalities, and summaries of investments using 10,593 instances of foreign investment. Because the Kigyo betsu gaishi donyu soran classifies the manufacturing industry into 18 sectors,2 our study also uses the same classification. In this document, foreign investment is classified into (1) technology introduction (2) direct investment of capital in the form equity and the like and the establishment of joint-venture firms, and (3) imports and sales. Figure 7.1 shows the percentage of each area of foreign investment according to industrial sectors. For all the sectors, technology introduction represents the area with the largest share of foreign investment. In a number of sectors, it comprises 90% or more of all foreign investment. Sectors with a relatively low percentage of technology introduction are food, oil and coal products, and non-ferrous metals, among others. Among the industrial sectors, the foods sector has the highest percentage of imports and sales; oil and coal products and non-ferrous metals have a notably higher percentage of equity and joint-venture investment than in other sectors. Figure 7.2 presents the shares of partner countries for technology introduction by sector. The United States has the highest shares for all the sectors, with shares in excess of 70% for electric appliances and precision instruments sectors. Following the United States, European countries like Germany, France, and the U.K. stand out in terms of their shares. Some sectors like oil and coal products show specific features, where the Netherlands has a notably high share along with the U.K.; Germany, France, and Sweden each accounted for more than 10% in the railroad vehicles sector.

7.2.2 Analytical Methodology This study simultaneously depicts the relationship of various foreign investments by sector to form a multiple network. In our analysis, we assigned varying degrees of strength to these relationships. With regard to the relationships between technology introduction, equity investment, and imports and sales, we assigned the strength of 1 if only one relationship existed among these three, 2 if two relationships existed, and 3 if all three relationships existed. Table 7.1 shows the results of computing descriptive

96

7 Global Knowledge Flows and Corporate Values

Precision Instruments

n=440

Motor Vehicles

n=422 n=33

Railroad Vehicles

n=734

Ship Building Electric Appliances

n=3,103

Machinery

n=1,308

Metal

n=102

Electric Wires

n=180

Non-ferrous Metals

n=116

Iron and Steel

n=597

Glass and Ceramics

n=240

Rubber

n=177 n=61

Oil and Coal

n=704

PharmaceuƟcals

n=1,225

Chemicals

n=21

Pulp and Paper

n=372

TexƟles

n=306

Foods 0%

20%

40%

60%

80%

100%

Technology introducƟon Direct investment of capital & joint-venture companies Imports and sales

Fig. 7.1 Foreign investment according to industrial sectors (Source Based on Kigyo betsu gaishi donyu soran (Jojo kigyo hen) [Guide to foreign capital investment by company (Listed firms)], published by the Keizai Chosa Kyokai)

7.2 Global Knowledge Flows and Inter-firm Relationship Structures

97

Precision Instruments

n=411

Motor Vehicles

n=359 n=30

Railroad Vehicles

n=712

Ship Building Electric Appliances

n=2,869

Machinery

n=1,162 n=91

Metal

n=161

Electric Wires

n=87

Non-ferrous Metals Iron and Steel

n=549

Glass and Ceramics

n=206

Rubber

n=157 n=45

Oil and Coal

n=595

PharmaceuƟcals

n=1,016

Chemicals

n=21

Pulp and Paper

n=322

TexƟles

n=207

Foods 0%

20%

40%

60%

U.S.A.

Germany

U.K.

Others

80%

100%

France

Fig. 7.2 Ratio of partner countries for technology introduction by sector (Source Based on Kigyo betsu gaishi donyu soran (Jojo kigyo hen) [Guide to foreign capital investment by company (Listed firms)], published by the Keizai Chosa Kyokai)

3.58

Mean range

Total number 5 of components

4.00

Average degree

0.23

Node ratio included in the largest component

230

40

Total number of components

115

3.22

Mean range

No. of links

3.66

Average degree

No. of actors

618

Total degree

Non-ferrous metals

338

No. of actors

Sector

Foods

Sector

2

3.53

4.43

354

160

Electric wires

0.91

16

4.88

4.38

744

340

Textiles and apparels

0.25

7

1.69

3.00

42

28

22

2.20

3.32

204

123

Metal products

Pulp and paper

56

5.89

4.13

2594

1256

Machinery

0.91

24

4.60

4.87

2434

999

Chemicals

Table 7.1 Descriptive statistics of networks by industry sector

29

3.97

6.74

5958

1767

Electric appliances

0.96

7

4.22

5.12

1402

548

Pharmaceuticals

1

3.47

4.62

1434

621

Ship building

0.77

5

3.49

4.07

122

60

Oil and coal products

3

2.02

3.67

66

36

Railroad vehicles

0.60

9

3.92

4.07

354

174

Rubber products

15

5.15

4.56

820

360

Motor vehicles

0.52

17

4.04

3.90

480

246

Glass and ceramics products

16 (continued)

4.34

4.44

872

393

Precision instruments

0.95

7

3.94

4.29

1184

552

Iron and steel

98 7 Global Knowledge Flows and Corporate Values

Non-ferrous metals

0.98

Electric wires 0.18

Metal products 0.80

Machinery 0.95

Electric appliances 1.00

Ship building 0.61

Railroad vehicles 0.88

Motor vehicles 0.84

Precision instruments

Source Own Calculations, based on Kigyo betsu gaishi donyu soran (Jojo kigyo hen) [Guide to foreign capital investment by company (Listed firms)], published by the Keizai Chosa Kyokai

Node ratio 0.75 included in the largest component

Sector

Table 7.1 (continued)

7.2 Global Knowledge Flows and Inter-firm Relationship Structures 99

100

7 Global Knowledge Flows and Corporate Values

statistics for each sector in our multiple network using UCINET, a software for social network analysis. The number of actors signifies the scale of each sector’s multiple network. “Total degree” refers to the total number of links between actors, while “average degree” denotes the number of links per entity. “Mean range” refers to the average number of necessary links joining one node with another node. “Component” refers to a completed subgraph in which each node is directly or indirectly linked within the network. We attempted to categorize the 18 sectors by using Ward’s method of cluster analysis on the network’s descriptive statistics, such as average degree, mean range, the total number of components, and the node ratio included in the largest component. In the analysis, we standardized each index and used the standard Euclidean distance as the distance between indices. From the results of the analysis, we were able to extract the following six groups. Group 1 consists of the four sectors of foods, pulp and paper, metal products, and railroad vehicles that contain several small-scale components and are characterized by their small average degree. Group 2 consists of the four sectors of oil and coal products, rubber products, glass and ceramics products, and non-ferrous metals whose network’s descriptive statistics have average values. Group 3 consists of the four sectors of textiles, chemicals, motor vehicles, and precision instruments and is characterized by the large size of its largest component and by a large mean range. Group 4 consists of the four sectors of pharmaceuticals, steel, electrical wire and cable, and shipbuilding; it is also characterized by the large size of its largest component, but its mean range is smaller than that of Group 3. Group 5 consists of the machinery sector alone and is distinguished from other groups by its large mean range and a large number of total components. Finally, Group 6 consists of the electric appliances sector and is characterized by a high average degree with a huge component structure in which more than 95% of the nodes are included in one component. Below, we choose six sectors, that is, foods, rubber products, textiles, pharmaceuticals, machinery, and electric appliances as typical examples that best represent the characteristics of these groups, and examine their relationship structures.

7.2.3 Sectoral Comparison of Relationship Structures Figure 7.3 is a visualization of the multiple networks of the six sectors in our analysis using NetDraw, a drawing software. The shape of the nodes varies depending on their nationality. Further, the size of each node is in proportion to the value of the standardized betweenness centrality, which measures the degree to which they mediate relationships. In the case of foods in Fig. 7.3a, we see components for such firms as Kirin Beer, Sapporo Beer, and Meiji Seika and relatively large components for firms such as Takara and Ajinomoto. However, we found that they have few nodes in their relationship structure that function as hubs connecting multiple actors but are characterized by a number of small-scale components with low average degrees.

7.2 Global Knowledge Flows and Inter-firm Relationship Structures a.Foods

101

Japan

U.S.A.

France

U.K.

Germany

Others

Sapporo Beer

Takara

Salins Procter & Gamble

Ajinomoto

Kirin Beer

Novo Nordisk

Meiji Seika

b. Rubber Products

Yokohama Rubber

Toyo Rubber

Continental Goodyear Tire & Rubber

Bridgestone

AT&T

Nitta

Fig. 7.3 Multiple networks in the introduction of foreign investment by industry sector

102

7 Global Knowledge Flows and Corporate Values c. Textiles and Apparels

Optima

Cluett Peabody & Co. Toray

Toyobo

Du Pont

Teijin Asahi Kasei

d. Pharmaceutical

Chugai

Takeda

Fujisawa Pharmacia

Merck Taisho

Yamanouchi

Sankyo

Fig. 7.3 (continued)

7.2 Global Knowledge Flows and Inter-firm Relationship Structures

e. Machinery

Kubota

Ebara

Sumitomo Heavy

General Electric

Sumitomo Precision

Litton Industrial Products

Komatsu IBM

f. Electric Appliances

Panasonic Hitachi

Fujitsu Mitsubishi Electric

Toshiba

Fig. 7.3 (continued)

General Electric

Motorola

NEC

103

104

7 Global Knowledge Flows and Corporate Values

In the case of rubber products, we found large manufacturers like Bridgestone and Yokohama Rubber to be prominent nodes. Further, while they have several relationships with global giants like Continental (Germany) and Goodyear (United States), for the provision of reinforced rubber technology that is used in run-flat tires, they constitute one component as a whole (Fig. 7.3b). At the upper left side of Fig. 7.3b, we can see the presence of several small components. One of the characteristics of textiles, which represent a group with a large component scale and mean range, is the markedly high betweenness centrality of certain overseas firms within the network (Fig. 7.3c). In other words, the U.S. chemical company DuPont along with major Japanese textile manufacturers like Toyobo and Teijin has developed multiple relationships in the provision of synthetic textile technology for interlacing yarn and in the establishment of joint-venture firms. Further, in the case of Cluett Peabody & Company (United States), it provided shrink-proof process technology for various types of textiles. As a result, we can see that it has contributed to the formation of the textile industry’s huge network Next, looking at the relationship structure in pharmaceuticals, which represents an example of a group with a large scale of the largest component and a small mean range (Fig. 7.3d), the Japanese drug-makers Takeda Pharmaceutical, Fujisawa Pharmaceutical, Taisho Pharmaceutical, and Yamanouchi Pharmaceutical are notable for their large nodes. Among foreign firms, major American drug-makers such as Pharmacia and Merck have relationships with Japanese manufacturers in the development of hepatitis B vaccine technology and drugs for treating osteomyelitis, demonstrating a marked degree of betweenness centrality. Figure 7.3e shows a relationship structure in which only the largest component is extracted in the machinery network that belongs to the group with the large mean range and a large number of total components. Among Japanese firms, the size of the nodes for Komatsu, Ebara, and Kubota stands out. Among foreign firms, U.S. corporations like General Electric, Litton Industrial Products, and IBM have several relationships in such areas as control technologies for machine tools and equipment manufacture, and their betweenness centrality is relatively notable. Figure 7.3f shows the largest component of the inter-firm network for electric appliances firms, which is characterized by a high level of average degree and huge components. At the center of this network, Toshiba, Hitachi, and the U.S. firms General Electric and Motorola stand out for their large betweenness centrality nodes. These firms are part of many relationships through the manufacture of semiconductor equipment and the establishment of joint-venture firms. Thus far, we could confirm the nodes of high betweenness centrality and the differences in the scale of their components by visualizing the relationship structures of networks. However, in the networks of the sectors like pharmaceuticals, machinery, and electric appliances, the number of actors contained is large. These nodes have intricate relationships, so we need to identify the nodes that play a key role in global knowledge flows by expressing the relationship structure more simply. The next section thus attempts to simplify the relationship structure.

7.2 Global Knowledge Flows and Inter-firm Relationship Structures

105

7.2.4 Using the Block Model to Contract the Relationship Structure In this study, we use a contraction method of networks called “block model.” The block model makes it possible to “classify the nodes into several blocks based on their position in the network and find the relationship structure within and between the blocks” (Yasuda 2001). Using the CONCOR algorithm (White et al. 1976), a network analysis tool, it is possible to contract a complex network by extracting those blocks that have similar relationship structures. The block model classifies nodes into several blocks (groups); when this happens, the nodes that belong to a single block are considered to be structurally equivalent3 to one another. Table 7.2 is an 8 × 8 density matrix constructed with the CONCOR algorithm. A density matrix presents the interrelationships among blocks. In this study, we extracted only those values that are greater than the mean density and deemed those instances as indications of relationships between blocks.4 Consequently, the 558 nodes for pharmaceuticals were divided into eight groups, and the inter-block relationship was depicted as a contracted graph, as shown in Fig. 7.4. Table 7.3 shows Table 7.2 Density matrix of the pharmaceutical sector Block

1

2

3

4

5

6

7

8

1

0

0

0

0

0.008

0

0

0.001

2

0

0

0

0

0.009

0

0

0.001

3

0

0

0

0.211

0.005

0.001

0

0.001

4

0

0

0.211

0.200

0.025

0.006

0

0.005

5

0.008

0.009

0.005

0.025

0.011

0.003

0.007

0.004

6

0

0

0.001

0.006

0.003

0.017

0

0

7

0

0

0

0

0.007

0

0

0

8

0.001

0.001

0.001

0.005

0.004

0

0

0.043

Source Own Calculations, based on Kigyo betsu gaishi donyu soran (Jojo kigyo hen) [Guide to foreign capital investment by company (Listed firms)], published by the Keizai Chosa Kyokai

Fig. 7.4 Reduced graph of relational structure in the pharmaceutical sector

8

1

4

5

2

7

3

6

88

Sanofi-Synthelabo

No. of actors

Name of firms

73

2

Pharmacia Glaxo Smith Klein Astra Zeneca

57

3 Takeda F. Hoffman Merck

5

4 Fujisawa Yamanouchi Sankyo Boehringer Mannheim

161

5

Chugai Nikken Chemicals

72

6

51

7

Taisho Pharmaceutical Company

41

8

Source Based on Kigyo betsu gaishi donyu soran (Jojo kigyo hen) [Guide to foreign capital investment by company (Listed firms)], published by the Keizai Chosa Kyokai

1

Block

Table 7.3 Number of actors and lists of firms belongings to each block in the pharmaceutical sector

106 7 Global Knowledge Flows and Corporate Values

7.2 Global Knowledge Flows and Inter-firm Relationship Structures

107

the number of actors in each block and the names of the firms that have the highest betweenness centrality for each block. Further, in Fig. 7.4, the numbers above the nodes correspond to the names of the blocks in Tables 7.2 and 7.3. Figure 7.4 has links that double back on themselves. This means that there are relationships among nodes in the same block, that is, among nodes in a structurally equivalent position. This contracted graph shows that firms in blocks 4 and 5, such as Takeda, Merck, and Yamanouchi, contribute to the structure of the huge component in the pharmaceutical network. Table 7.4 and Fig. 7.5 show the results of contraction for machinery and electric appliances. In Fig. 7.5a, blocks 7 and 8 are connected, while connections also exist among blocks 4, 3, 6, and 5. However, we see an absence of blocks that would serve as hubs linking nodes belonging to different blocks. The formation of this type of dispersed structure would be a factor contributing to the high mean distance in the machinery sector. Conversely, in Table 7.4b and Fig. 7.5b, in the electric appliances sector, Japanese firms like Toshiba, Hitachi, NEC, and Fujitsu in block 8 form a typical center– periphery structure (Kanemitsu 2003), in that they have relationships with nodes that belong to other blocks. Therefore, having a structure with a global hub for information flows is considered to reduce the mean distance of the electric appliances sector’s network.

7.2.5 Summary This chapter takes the data on the introduction of foreign investment that is aggregated for each company and classifies sectors based on network statistics to examine the relationship structures among both Japanese and global actors. Further, using the block model generated by the CONCOR algorithm to simplify the networks made it possible to identify the nodes that constitute hubs and contribute to the formation of a dense network; that is, the groups that are crucial to growing and developing the networks. Our analysis found that in cases of a dispersed structure like that of the machinery sector, relationships are limited to those between nodes that belong to certain blocks. At the same time, we found that the electric appliances sector’s network has a center–periphery structure and that certain Japanese firms are crucial to the smooth circulation of global knowledge flows. In light of the above findings, let us describe the issues that remain. The characteristics of the relationship structures that we have seen so far need to be considered in view of how they affect the firms participating in the network or the network’s profitability and performance. Therefore, the next section will consider the relationship between network attributes and corporate performance using the networks’ descriptive statistics for each sector and each company’s financial statements as the analytical data and employing covariance structure analysis as the analytical tool.

122

Ebara IBM

No. of actors

Name of firms

Hewlett-Packard RCA

Name of firms

306

2

5

6

Sumitomo Heavy Industries

3

SHINKO ELECTRIC

176

5

General Electric MacNeal-Schwendler Air Products & Chemicals

108

General Electric Philips Electronics Motorola Lucent Technologies

203

4

Sumitomo Precision Products Litton Industrial Products Teijin Seiki

151

4

Microsoft IBM Texas Instruments

199

3

Komatsu Takuma DAIKIN INDUSTRIES

607

3

Toyo Engineering V/O Licencintorg Arther D. Little Enterprise

TOKAI RIKA Dolby Laboratories Licensing

492

6

7 69

8

54

7

Toshiba Hitachi NEC Fujitsu Panasonic

103

8

CHIYODA Exxon Research & Engineering Ford Motor

70

Source Based on Kigyo betsu gaishi donyu soran (Jojo kigyo hen) [Guide to foreign capital investment by company (Listed firms)], published by the Keizai Chosa Kyokai

1

234

No. of actors

Kubota George Fischer

126

2

Block

(b) electric equipment

1

Block

(a) machinery

Table 7.4 Number of actors and lists of firms belonging to each block in the (a) machinery and (b) electric equipment sectors

108 7 Global Knowledge Flows and Corporate Values

7.3 Relationship Between Network Attributes and Corporate Value Fig. 7.5 Reduced graphs of relational structure in the (a) machinery and (b) electric appliances sector

109

(a)Machinery 1

3

6

7

2

4

5

8

(b)Electric Appliances 7

6

1 8 5 2 3

4

7.3 Relationship Between Network Attributes and Corporate Value Yukawa (2004) and Sugiyama et al. (2006) both use social network analysis to examine the correlation between relationship structures between actors and economic performance. Yukawa (2004) takes an analytical sample of internet-related firms located in Tokyo and analyzes the cooperative relationships through their executive teams, shareholders, and banking institutions, as well as the networks comprised of firms in competitive relationships, meaning that they use the same suppliers and client firms. Upon considering the interrelations between network statistics and such corporate performance metrics as sales and profits, Yukawa (2004) found that networks

110

7 Global Knowledge Flows and Corporate Values

with cooperative relations had no correlation or a negative correlation with performance. Therefore, it was thought that inter-firm networks might not be functioning properly. At the same time, Yukawa (2004) asserts that in networks with competitive relations, a company’s performance is affected by the relationship that it has with analytical firms. Using a massive database on the business relationships of Japan’s listed firms and aided by social network analysis, Sugiyama et al. (2006) analyzes the correlation between various financial statements and networks’ relationship structures. They concluded that there is a high correlation with market capitalizations of firms. Thus, Yukawa (2004) and Sugiyama et al. (2006) examine the correlation between networks’ descriptive statistics and the sole financial metrics for corporate performance, that is, sales and market capitalization. However, as Inori (2001) points out, although financial statements show one important aspect of corporate management, ascertaining an abstract concept like corporate performance requires comprehensive criteria that consider several metrics. Our study, therefore, uses covariance structure analysis to formulate comprehensive metrics for examining the relationship between networks’ attributes and firms’ financial statements.

7.3.1 Analytical Framework of Prior Studies Using Covariance Structure Analysis Although the application of covariance structure analysis in economic geography has begun very recently, using covariance structure analysis makes it possible to establish several factors based on data derived from direct observation (factors underlying the data) and to quantitatively identify the relationships between these factors. Let us examine some prior studies that have employed covariance structure analysis. Focusing on the importance of theme-sharing in an artificial satellite project in the Higashi Osaka region, Kanai et al. (2006) used a questionnaire survey of 145 manufacturing firms located in that region. They conducted a covariance structure analysis of the major factors in conventional research on industry–academia–government linkages and inter-firm networks and the role of regional revitalization for increasing a region’s level of business with other regions and improving its namerecognition level. Based on their findings, Kanai et al. (2006) argues that similar to industry–academia–government linkages and inter-firm networks, firms’ sharing of a common theme is established as a metric for explaining the elements of learning areas. Inori (2001, 2002) uses firms’ financial statements as analytical data to construct latent variables (factors) such as “corporate value,” as measured by market capitalization, and cash flow and “profitability,” measured by return on shareholders’ equity. Further, covariance structure analysis was used to identify the cause-and-effect relationship among these latent variables. As Inori (2001) points out, covariance structure analysis has an advantage in that it uses the same metrics that are ordinarily

7.3 Relationship Between Network Attributes and Corporate Value

111

employed in financial analysis to measure corporate value and profitability and that are thought to “exist latently even though they cannot be measured directly.” Inori (2001) builds a model in which profitability has a positive effect on growth and the rate of growth increases corporate value. Inori (2002) shows that higher profitability and the improvement of profitability changes lead to a sound financial structure and higher corporate value. However, this study lacks explainability because its overall analytical findings have a low level of applicability, so the model has considerable room for improvement. Asakawa and Nakamura (2005) take researchers from the corporate research labs of Japanese pharmaceutical firms as their analytical sample to investigate whether the acquisition of knowledge through exchanges with external R&D sectors is linked with individual researchers’ R&D achievements. They studied 130 researchers from four Japanese and three non-Japanese firms; covariance structure analysis was employed to investigate the factors leading to a sense of accomplishment that research results were achieved. The study found that exchanges with external “universities and academic societies” were effective only when internal information was available through exchanges with “internal sectors.” Matsumoto (2006) examines “organizational trust” and used covariance structure analysis to identify structures that create various characteristics such as trust in the area of civil society (e.g., non-profit organizations and religious bodies) and trust in systems (e.g., the National Diet and the police). Matsumoto (2006) analyzes the case of Japan, assuming the two latent variables of system trust and civil trust and presents a model of organizational trust that has a stable structure. Matsumoto (2006) also focuses on the structure of organizational trust using anonymized data from eight countries and regions: Japan, Beijing, Shanghai, Hangzhou, Kunming, South Korea, Taiwan, and Hong Kong. The outcome of the simultaneous analysis of multiple populations suggests that different structural models apply to the four regions of mainland China and the other four countries/regions, that is, Japan, South Korea, Taiwan, and Hong Kong.

7.3.2 Simultaneous Analysis of Multiple Populations An advantage of covariance structure analysis is that it can depict complex relationships, as shown in the path diagram in Fig. 7.6. (Kano and Miura 2003; Toyoda ed. 2007). The rectangle in the figure is an observation variable whose data is obtained directly from reference materials and questionnaire surveys. The ovals denote latent variables that cannot be observed directly. The one-way arrows (paths) stand for the cause and effect from the beginning toward the end, and bidirectional arrows signify correlations among variables (covariances). The symbols in circles—e1 and e2—are error variables, which denote the portions that cannot be forecast or explained by the causal relationships among latent or observation variables. We use AMOS 5.0.1, a specialized software, to carry out the following covariance structure analysis.

112

7 Global Knowledge Flows and Corporate Values

Fig. 7.6 Path graph in covariance structure analysis

The model in Fig. 7.6, considered in this paper, posits two latent variables: “structural advantage” and “corporate value.” Structural advantage refers to an advantage in a company’s structural position in a global inter-firm network. It is assumed to influence three observation variables that represent network attributes: degree centrality,5 betweenness centrality, and equivalence.6 Corporate value, a comprehensive metric for corporate performance, is an abstract evaluation axis that influences the soundness of firms’ financial conditions and their growth. Our study considers that corporate value determines the three observation variables, equity capital, market capitalization, and net assets per share.7 Further, positing that structural advantage has an impact on corporate value allows us to consider the relationship between a network’s relationship structure and corporate value. To compare the sizes of the model’s path coefficients between sectors (or the effects between the latent variables, or between the latent variables and the observation variables), we performed a multi-population simultaneous analysis that assumes multiple groups.8 In addition, because we cannot obtain a proper estimate if the number of samples for a sector is too small, we need to exclude such sectors from our analytical sample. Therefore, our study includes 12 sectors: foods, textiles, pharmaceuticals, chemicals, rubber, glass and ceramics, steel, metals, machinery, electric appliances, motor vehicles, and precision instruments. In a covariance structure analysis, when we perform a multi-population simultaneous analysis, we need to use the following procedure (Kano and Miura 2003). First,

7.3 Relationship Between Network Attributes and Corporate Value

113

Table 7.5 Fit indices of each model in multi-population simultaneous analysis GFI

CFI

RMSEA

Model 0 (configural invariance)

0.930

0.982

0.034

AIC 479.610

Model 1 (measurement invariance)

0.632

0.665

0.115

1601.784

Model 2

0.615

0.644

0.114

1668.944

Model 3

0.605

0.643

0.110

1664.826

Model 4

0.483

0.440

0.116

2349.083

Source Own calculations. based on Kigyo betsu gaishi donyu soran (Jojo kigyo hen) [Guide to foreign capital investment by company (Listed firms)], published by the Keizai Chosa Kyokai

we consider the Model 0 (configural invariance9 ), for which there is no constraint that all estimated parameters differ from population to population. Then, we must consider Model 1 (measurement invariance), in which only the path coefficients are equivalent; Model 2, in which the variance-covariance of the path coefficients and the latent coefficients is equivalent; Model 3, in which the variance-covariance of the path coefficients and error variables is equivalent; and finally, Model 4, which assumes that all parameters are equivalent. Table 7.5 shows four goodness-of-fit (GFI) indices10 resulting from the analysis of each model under the different constraints. In the table, we see that in Model 0, the GFI is good, and the comparative fit index and root mean square error of approximation have very good values. However, from Model 1 onward, the values of at least one of the goodness-of-fit indices are bad and do not support the models. Therefore, in a simultaneous analysis of multiple populations comparing sectors, it was demonstrated that configural invariance is established but that measurement invariance is not established. In other words, it was suggested that while the models in Fig. 7.6 are applicable to all sectors, the impact of the paths showing the cause and effect between variables differs from sector to sector. Because configural invariance has been established, we will now compare the sizes of the path coefficients from sector to sector.

7.3.3 Comparison of Path Coefficients Among Sectors Figure 7.7 shows the estimation results for the coefficients in the foods sector using the maximum likelihood method and based on Model 0. We see here that the path coefficients from structural advantage to betweenness centrality and degree centrality are high with values of more than 0.9. Further, the path coefficient to equivalence is a large negative value.11 Therefore, it seems that these three observation variables are appropriate as indices representing structural advantage. Furthermore, although the path coefficient coming from corporate value is low at 0.21 for net assets per share,12 the path coefficients going toward equity capital and market capitalization are around

114

7 Global Knowledge Flows and Corporate Values

Fig. 7.7 Estimation of path coefficients in the foods sector

0.9, and the signs do not contradict either. Therefore, these three observation variables can all be considered appropriate indices for corporate value. Now, let us look at the value of the path coefficient running from structural advantage to corporate value and examine the differences in impact from sector to sector. Table 7.6 shows the standardized estimates and non-standardized estimates of the path coefficients from structural advantage to corporate value. For standardized estimates, machinery is the lowest at 0.49, motor vehicles is the highest at 0.84, and the impact is positive for both sectors. This shows that the more structurally advantageous the position, the higher the corporate value. Table 7.7 gives the test statistics (z statistics) for the differences in the path coefficients of non-standardized values for the 12 sectors. The test statistics of these differences approximately follow normal distributions. From Table 7.7, we can see that the values of the path coefficients from “structural advantage” to “corporate value” for foods and metals are significantly lower than those for the electric appliances, motor vehicles, and precision instruments sectors. For machinery, the difference with pharmaceuticals is also significant in addition to these three sectors. For steel and chemicals, significant difference in the path coefficients was observed only with the motor vehicles sector. Meanwhile, in the textile, rubber, and glass and ceramics sectors, the path coefficient values were in the middle range for all 12 sectors, and no significant differences were observed among the sectors. From the above, we can surmise that among the 12 sectors, structural advantage has a relatively low impact on corporate value for foods, metals, and machinery.

7.3 Relationship Between Network Attributes and Corporate Value Table 7.6 Path coefficient from “structural advantage” to “corporate value”

115

Sector

Standardized path coefficient

Non-standardized path coefficient

Foods

0.56

0.68

Textiles and Apparels

0.80

1.09

Chemicals

0.68

0.96

Pharmaceuticals

0.80

1.22

Rubber Products

0.79

1.18

Glass and Ceramics Products

0.70

1.22

Iron and Steel

0.78

0.80

Metal Products

0.71

0.67

Machinery

0.49

0.70

Electric Appliances

0.70

1.20

Motor vehicles

0.84

1.57

Precision Instruments

0.79

1.41

Source Own calculations. based on Kigyo betsu gaishi donyu soran (Jojo kigyo hen) [Guide to foreign capital investment by company (Listed firms)], published by the Keizai Chosa Kyokai

For sectors like pharmaceuticals, electric appliances, motor vehicles, and precision instruments, the impact is great. In addition, for textiles, steel, chemicals, rubber, and glass and ceramics, the impact is in the middle range, suggesting that we can classify the 12 sectors into three groups according to the degree of impact from structural advantage. In sectors like foods, metals, and machinery, processing technology and synthetic knowledge13 are required. In sectors like pharmaceuticals, electric appliances, motor vehicles, and precision instruments, a great deal of scientific and analytical knowledge is required and M&A is proceeding rapidly. In this scenario, firms are required to uniquely position themselves in their relationship structures, distinguishing themselves from their competitors. They must actively seek technology introduction arrangements with foreign firms, which we consider to be of importance to improve the corporate value as measured by equity capital or market capitalization.

7.4 Conclusion Research that quantitatively extracts and describes the global networks used in knowledge creation is at an early stage. This paper constructed a database of actual cases of foreign investments and visualized relationship structures by using social network analysis to identify the knowledge flows between global firms. We examined network

0.65 0.43

3.25 1.34 −

2.08 2.56 −

0.35 0.13 −

1.48 0.40 −

0.85 1.06 −

0.01 0.08 −

0.08 −

−

0.34

1.50

0.37

0.06

0.07

2.22

2.99

2.14

Glass and Ceramics

Iron and steel

Metal

Machinery

Electric appliances

Motor vehicles

Precision instruments

1.39

2.21

1.15

1.36

1.16

0.57

0.54

1.11

0.07

2.07

1.83

1.28

0.01

0.08

0.47

0.86

0.04

1.21

1.19

0.85

0.08

0.44

0.89

0.07

1.55

1.48

1.06

1.61

2.26

1.39

0.35

0.40

1.28

2.09

2.87

2.08

0.13

1.19

1.83

2.25

3.25

2.56

1.55

1.21

2.07

0.65

1.34

1.39

0.07

0.04

0.07

0.43

2.87

2.26

0.89

0.86

1.11

2.21

−

2.25

2.09

1.61

0.44

0.47

0.54

1.39

Source Own calculations. based on Kigyo betsu gaishi donyu soran (Jojo kigyo hen) [Guide to foreign capital investment by company (Listed firms)], published by the Keizai Chosa Kyokai

0.88

1.50

0.41

1.59

1.43

0.90

0.76

0.54

0.99

0.21

1.15

0.88

2.14

0.41

1.36

1.50

2.99

1.90

1.16

0.41

2.22

Rubber products 1.19

0.57

1.59

0.07

Pharmaceuticals

0.76

1.43

0.06

0.54

0.90

0.37

0.99

0.34

1.50

−

0.51

0.21

1.19

1.20

1.90

Chemicals

Precision instruments

0.41

Motor vehicles

1.20

Electric appliances

0.51

Machinery

1.47

Metal

−

Iron and steel

1.47

Glass and ceramics

−

Rubber

Textiles

Pharmaceutical

Foods

Chemicals

Textiles

Foods

Table 7.7 Test statistics for the differences in the path coefficients of non-standardized values for the 12 sectors

116 7 Global Knowledge Flows and Corporate Values

7.4 Conclusion

117

attributes and the effects of networks on knowledge creation, as well as their causeand-effect relationships with economic performance by adopting various statistics representing network attributes that were calculated by social network analysis as variables of covariance structure analysis. We then built a model to examine which actors in what sort of structural positions are showing good economic performance. We review the findings of this paper and propose some research topics for the future. First, we used network statistics to classify the 18 sectors of manufacturing industry and compared them through a visualization of their relationship structures. From this, we extracted the groups where certain Japanese and foreign textile, pharmaceutical, and electric appliances firms have relationships with nodes positioned on the periphery of their networks, contributing to the formation of huge components. At the same time, we identified groups like foods firms that have small-scale components and have not introduced as much technology as other sectors. Further, sectors like machinery have a large number of actors whose networks, however, are split so that they have no hubs that work as a key to facilitate the circulation of knowledge flows. Then, we made comparisons among the sectors through the simultaneous analysis of multiple populations with respect to the relationship between the network attribute indices derived from our social network analysis and firms’ economic performances based on their financial statements. Our findings show that for all sectors, the better the position in the relationship structure, the more positive the impact on corporate performance as measured by such metrics as market capitalization. However, the degree of impact varies by sector, suggesting that in sectors such as pharmaceuticals and electric appliances, where more analytical knowledge is required, being in a structurally advantageous position is essential to increasing corporate value. This is consistent with the fact that M&A is more prevalent in sectors requiring analytical knowledge than in other sectors. Finally, there are two remaining issues. First, individual firms when they engage in economic activities, do not necessarily include the overall network that occupies multi-tiered spatial dimensions ranging from the local to the global in their perspective. Future research initiatives should synthesize the findings of analyses of both local economic activities and global activities and identify where each organization’s economic activities are positioned in the overall network structure. This should enable researchers to quantitatively understand the economic effects that network relationships occupying various spatial dimensions—from local to global—have on firms and geographical areas. The second issue is the need for informal network analysis. In the project-type joint-R&D networks, and the global inter-firm networks based on incoming foreign investment, discussed in the present study, relationships between actors are contractbased. Therefore, they have a formal configuration. In contrast, recent discussions about innovation have been focused on participation in industry conferences, trade fairs, study groups, and research groups as the key channels for knowledge creation (as shown in Chaps. 3, 4, and 5). In the past, the mainstream studies that have addressed this type of informal means of inter-organizational knowledge creation

118

7 Global Knowledge Flows and Corporate Values

have been qualitative empirical studies from the standpoint of social capital. The existing empirical studies need to be complemented by applying the analytical framework of our study and by systematically identifying overall network structures focusing on structures among participating actors. Notes 1. 2.

3.

4. 5. 6.

As the Keizai Chosa Kyokai is no longer in existence, data from the Kigyo betsu gaishi donyu soran are available only from March 2002. The 18 sectors are foods, textiles, pulp and paper, chemicals, pharmaceuticals, oil and coal products, rubber products, glass and ceramics products, steel, non-ferrous metals, electrical wire and cable, metal products, machinery, electric appliances, shipbuilding, railroad vehicles, automobiles, and precision instruments. Nodes that are structural equivalents have the characteristic of being substitutable (their relationship with each other does not change even if they are substituted for each other). There are multiple criteria for determining whether there are relationships among blocks. See Kanemitsu (2003) for details on other criteria. Degree centrality is measured by the total number of links that each actor possesses. For each company in the groups formed by the block model, the value is obtained by dividing the number of actors that are structurally equivalent (e) by the value of the network’s scale (n) used as the metric (equivalence) for measuring a company’s position in the relationship structure. However, as this metric is a ratio, the logit-converted value is used as the observation variable. This observation variable is depicted as follows:

log(e / n 7.

8.

9.

1 e / n

)

There are various candidates identified as metrics for measuring corporate value such as total assets, ordinary income margin, and return on equity. In this study, metrics were used that measure profitability and efficiency, it was deduced that the three variables used here showed the highest fit for the model. Incidentally, for the analysis, all three observation variables were used by logarithmically converting them. Furthermore, financial statement data were obtained from each company’s annual securities reports for the fiscal year ending March 31, 2002 (or the most recent accounting period). Analyzing each sector separately leads to problems such as loss of stability of the estimation value owing to fewer samples and the inability to assess differences among groups (sectors) in the model in general (Toyoda ed. 2007). Configural invariance denotes the hypothesis that the estimation values differ even though the same path diagrams apply among groups (Toyoda ed. 2007).

7.4 Conclusion

10.

11.

12. 13.

119

GFI and CF are indices depicting how well a model fits in a covariance structure analysis, and the closer they are to 1, the better the model’s explainability. A value of 0.9 or more is good, and a value of 0.95 or more is very good. The RMSEA is considered good if it is 0.05 or lower. AIC stands for Akaike’s information criterion, which is used when performing relative comparisons of several models. The smaller the value, the better the model. In the analysis, the square of the value of the standardized path coefficient is equivalent to the dispersion of the variable at the destination end of the path. As a result, a path variable of more than 0.9 means that at least 80% can be explained by the variable at the originating end of the path. Further, the fact that the sign of the path variable running from structural advantage to equivalence is negative means that it possesses few nodes positioned similar to itself in the relationship structure and has an advantage to the degree of having uniqueness. The path coefficient of 0.21 means that only 4% can be explained by corporate value. Synthetic knowledge relates to engineering-type knowledge that is grounded in inductive processes such as experience in business decision making. Analytical knowledge refers to scientific knowledge grounded in deductive processes such as academic articles and patents.

References Asakawa K, Nakamura H (2005) Seiyaku kigyo no kenkyusha level ni okeru kenkyu seika tassei no joken: naigai collaboration wo tujita knowledge kakutoku no koka (External and internal collaboration as the sources of R&D performance: an empirical investigation of R&D researchers in the pharmaceutical firms). Keiei Kodo Kagaku (Jpn J Adm Sci) 18:223–234 (in Japanese with English abstract) Cooke P (2009) The economic geography of knowledge flow hierarchies among internationally networked medical bioclusters: a scientometric analysis. Tijdschr Econ Soc Geogr 100:332–347 Graf H (2006) Networks in the innovation process: local and regional interactions. Edward Elgar, Cheltenham Inori M (2001) Kigyo zaimu data to kyobunsan kozo bunseki ni yoru shuekisei to kigyo kachi no bunseki (Profitability and corporate value analysis using corporate financial data and covariance structure analysis). Nempo Keiei Bunseki Kenkyu (J Bus Anal) 17:62–69 (in Japanese) Inori M (2002) Shinyo risk no hyoka to kyobunsan kozo bunseki ni yoru kigyo model (Credit risks and corporate models using covariance structure analysis). Kyoto Manag Rev 1:185–206 (in Japanese with English abstract) Kanai A, Matsubara H, Niwa K (2006) Gakushu chiiki ni okeru theme kyoyu no jyuuyousei (The importance of shared themes in learning regions). Kenkyu Gijyutsu Keikaku (J Sci Policy Res Manag) 21:294–306 (in Japanese) Kanemitsu J (2003) Shakai network bunseki no kiso: Shakaiteki kankei sihon ni mukete (The basics of social network analysis: towards social capital). Keiso Shobo, Tokyo (in Japanese) Kano Y, Miura A (2003) Graphical multivariate analysis by AMOS. Gendai Sugakusha, Kyoto (in Japanese)

120

7 Global Knowledge Flows and Corporate Values

Matsubara H (2007) Chishiki no kuukanteki ryudo to chiikiteki innovation system (Spatial knowledge flow and regional innovation system). Tokyo Daigaku Jinbun Chirigaku Kenkyu (Komaba Stud Hum Geogr) 18:22–43 (in Japanese) Matsumoto W (2006) Higashi Asia ni okeru soshiki ni taisuru shinraikan: kokusai hikaku no tameno shinraikan no bunseki (Sense of trust on organizations in East Asia: analysis for a cross-national comparative study). Kodo Keiryo Kagaku (Jpn J Behaviormetrics) 33:25–40 (in Japanese with English abstract) OECD (2009a) OECD patent statistics manual. OECD Publishing, Paris OECD (2009b) OECD regions at a glance 2009. OECD Publishing, Paris Sugiyama K, Honda O, Ohsaki H, Imase M (2006) Network bunseki syuho ni yoru nihon kigyokan no torihiki kankei network no kozo bunseki (Application of network analysis techniques for Japanese corporate transaction network). Shakai Johogaku kenkyu (J Socio-Inf Stud) 11(2):45–56 (in Japanese) Toyoda H (ed) (2007) Kyobunsan kozo bunseki [AMOS hen]: kozo hoteishiki modeling (Covariance structure analysis [Amos]: structural equation modeling). Tokyo Tosho, Tokyo (in Japanese) Wakabayashi N (2006) Nihon kigyo no network to shinrai: kigyokan kankei no atarashii keizai shakaigakuteki bunseki (Network and trust in Japanese inter-organizational relationships). Yuhikaku, Tokyo (in Japanese) White HC, Boorman SA, Breiger RL (1976) Social structure from multiple networks. I. blockmodels of roles and positions. Am J Sociology 81:730–780 Yasuda Y (2001) Jissen network bunseki: kankei wo toku riron to gihou (Practical network analysis: theory and techniques to unravel relations). Shinyosha, Tokyo (in Japanese) Yokura Y, Matsubara H, Sternberg R (2013) R&D networks and regional innovation: a social network analysis of joint research projects in Japan. Area 45:493–503 Yukawa K (2004) Kigyokan network kara mita netto kigyou no cluster to kigyou senryaku: netto kigyou ni okeru kyocho to kyosyo no kankei kozo (Clusters of internet firms and corporate strategy from the perspective of inter-firms networks: relational structure of cooperation and competition in internet firms). FRI Kenkyu Rep 214:1–35 (in Japanese)

Chapter 8

Conclusions

Abstract Various relationships have been established in industrial agglomeration to respond to radical changes in the economic environment. In this book, we analyzed the relationships among local organizations and extra-local networks and focused on their spatial dimensions to describe the roles that various networks play in the knowledge creation process. This book focuses on hot topics such as “networks” and “innovation” in economic geography and presents new findings through theoretical foundations and empirical analysis. This book’s primary research aim is to clarify the evolution of the development mechanism of industrial agglomerations by understanding the inter-organizational network from a structural perspective. In this chapter, we discuss possible directions for future research. Keywords Industrial agglomerations · Temporary systems · Dynamic perspective

8.1 Networks and Innovation in Industrial Agglomeration So far, in the discussion of industrial agglomeration, we have considered the existence of tacit knowledge, which is embodied by humans and lacks mobility, as one of the factors that forms industrial agglomeration. Current research examining the process of regional innovation and knowledge creation focuses on the various linkages between actors both inside and outside of industrial agglomerations. The influx of novel knowledge and information from outside will overcome the negative rigidity resulting from industrial agglomerations’ limited and fixed relationships to foster innovation. The study of networks and innovation has broadened its research field from vertical relationships, such as business transactions, to horizontal, inter-organizational relationships, such as joint research and development (R&D), in which intangible information and knowledge exchange is carried out. In early economic geography studies on networks and innovation, qualitative descriptions of the social contexts and institutions of industrial agglomeration were the main focus. In recent years, with the development of additional analytical tools, the approach of quantitatively analyzing the relationships among the relevant actors has come to the forefront. While new © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2021 Y. Yokura, Regional Innovation and Networks in Japan, International Perspectives in Geography 16, https://doi.org/10.1007/978-981-16-2191-8_8

121

122

8 Conclusions

analytical tools such as social network analysis provide a very rich perspective for network and innovation research, they lack the required spatial perspective. In this book, we analyzed the relationships among local organizations and extralocal networks. We focused on networks for knowledge creation that evolved in various spatial dimensions, from local to global scale. In Chaps. 1 and 2, we provided overviews of discussions on networks and innovation in economic geography and related fields. While there are two perspectives on networks—i.e., the social network approach and the governance approach—neither offers spatial implications. Therefore, we reviewed existing research from the different perspectives of three spatial dimensions (i.e., local, non-local, global) and examined the forms of networks and innovation. When studying intangible relationships such as trust and knowledge flow, this study provides new perspectives by introducing spatiality and the knowledge base into quantitative analytical tools such as the social network approach. Chapter 3 considered the structures and spatial patterns of R&D networks in Japan by visualizing their structure and calculating network indices. Spatiality differs considerably depending on whether the technical field is synthetic knowledge-based or analytical knowledge-based. In addition, some universities and technical colleges play an important role in long-distance collaborative R&D. In Chap. 4, we discussed the local trade fairs commonly held in industrial agglomerations and examined the development of the various relationships among their key actors, such as exhibitors and visitors, as temporary clusters. One purpose of participation in a trade fair is the observation of the other exhibitors’ skills and new products. Business talks with non-local firms who are potential customers and the acquisition of new orders are also important for local actors. Specifically, such fairs enhance the generally weak marketing power of small- and medium-sized firms. The role of trade fairs is to offer exhibitors an opportunity to construct long-term business relationships and establish mutual trust by inviting existing customers to attend. Various relationships have been established in industrial agglomeration to respond to radical changes in the economic environment. As we suggested in Chap. 5, informal networks established through business workshops could improve resilience and tolerance capabilities in industrial agglomerations. Informal networks expand when specific actors participate in multiple workshops. Those actors may have novel knowledge and market-valuable information, so they play an important role in innovation creating processes and developing formal networks such as joint research projects. The case studies featured in Chaps. 6 and 7 evaluated how Japan’s firms adapted to radical changes under global competition by using new analytical tools such as social network analysis and covariance structure analysis. In Chap. 6, the focus was on the case of Kyushu semiconductor industry, which adapted to external environmental changes such as the reorganization of major companies and East Asian countries catching up technologically with Japan. In this case, a coordinator with knowledge of excellence in the semiconductor industry played a significant role in facilitating the formation of inter-firm networks. However, after the coordinator died, the international conference based on these inter-firm networks was reduced in size, making continuous information acquisition difficult. This suggests that even in relatively

8.1 Networks and Innovation in Industrial Agglomeration

123

successful industrial support projects, the lack of key actors complicates sustainable support systems. Chapter 7 constructed a database of actual cases of foreign investments in Japan and visualized the relationship structures by using social network analysis to identify the knowledge flows between global companies. Furthermore, we examined the effect of structural advantage on corporate behavior using covariance structure analysis. The impacts of structural advantage work strongly in specific industry sectors that require scientific and analytical knowledge.

8.2 Future Directions When evaluating current industrial agglomerations, it can be seen that some regions have problems with negative lock-in due to a lack of economic flexibility (Boschma 2005) and others feature agglomeration economies that promote entrepreneurship and collective learning (Saxenian 1994; Keeble et al. 1999). Based on this reality, this book states that, for industrial agglomerations to continue fostering innovation, not only the institutional thickness within the industrial agglomeration but also a widearea network outside the industrial agglomerations is crucial. This book specifically analyzes networks and innovation based on the project-type joint R&D, trade fairs, business workshops, and loose coupling established through international conferences. Such economic activity performed in industrial agglomerations functions as a temporary system. As we discussed in Chaps. 1 and 3, the geographical manifestation of interorganizational networks varies greatly depending on the industries that are prominent within industrial agglomerations. In other words, geographical proximity to diverse local organizations plays a critical role in industrial agglomerations of “monozukuri” industries, in which synthetic knowledge is important. On the other hand, relational proximity to extra-local actors is essential for creating innovation in “science-based” industrial agglomerations in which analytical knowledge is significant. Temporary systems such as the project-based joint R&D and trade fairs featured in this book create permanent agglomeration economies by forging relationships with non-local organizations. This book also suggests that the institutional thickness of industrial clusters with universities and public research institutes is also a source for agglomeration economies. This book focused on hot topics such as “networks” and “innovation” in economic geography and presented new findings through theoretical foundations and empirical analysis. Most of the research aim of this book is to clarify the evolution of the development mechanism of industrial agglomerations by understanding the interorganizational network from a structural perspective. However, this book does not answer every research issue. Later in the book, I mention possible directions for future research. First, it is necessary to refine the definition of innovation. As mentioned in Chap. 3, “innovation” in this book is narrowly defined and, due to the limitation of analytical

124

8 Conclusions

data, innovation is replaced by the achievement of commercialization through joint R&D. Therefore, this research was unable to include different types of innovation, such as incremental innovation or radical innovation, in its analytical framework. Process innovations such as the standardization of production technology are also not included in this analysis. In the future, performing research that includes various innovation forms into the scope of analysis is necessary. Second, applying the dynamic perspective to empirical research is required. In the structural analysis of networks in this book, comparison between regions on a single point is the main focus, so it is crucial to introduce historical dynamism. Finally, it is necessary to elaborate and deepen the policy implications of these findings. In this book, we clarified the kind of structure that is beneficial to the performance of industrial agglomeration by visualizing the structure of the network constructed by its various actors. However, this does not mean that there is a “best” network structure that can be adapted in every region. The relationships established in industrial agglomerations have embedded within them local characteristics related to their history, culture, and institutions, and they are influenced by radical changes in the economic environment and various other networks that occupy diverse and multi-layered spatial dimensions. In further research, exploring a network structure suitable for the actual conditions of each individual region will be necessary.

References Boschma RA (2005) Editorial role of proximity in interaction and performance: conceptual and empirical challenges. Reg Stud 39:41–45 Keeble D, Lawson C, Moore B, Wilkinson F (1999) Collective learning processes, networking and ‘institutional thickness’ in the Cambridge region. Reg Stud 33: 319–332 Saxenian A (1994) Regional advantage: culture and competition in Silicon Valley and Route 128. Harvard University Press, Cambridge

Index

A Absorptive capacity, 15, 16, 23, 28 Adjacent matrix, 29 Agglomeration economies, 5, 123 AMOS, 111 Analytical knowledge, 18, 19, 22, 39–41, 115, 117, 119, 122, 123 Arms-length relationships, 23 Average degree, 98, 100, 104

B Best practice, 17 Betweenness centrality, 67, 70–72, 74, 75, 77, 100, 104, 107, 112, 113 Block model, 105, 107, 118 Branch plant, 50 Buzz, 6, 8, 13, 19–21

C Capability, 3, 58 organizational, 3 technological, 57, 76 Codified knowledge, 18, 20, 36 Cognitive lock-in, 75, 76 Collective learning, 2, 6, 16, 123 Co-location, 8, 36 Community of practice, 16, 17 Competitive advantage, 3, 4, 82, 87, 90, 91 Component, 3, 14, 31, 34, 74, 81, 98–100, 104, 107, 117 Content industry, 21 Creativity, 46 Cultural, 6, 16, 19, 21, 22, 48, 50

industry, 6, 19, 21, 22, 48 proximity, 16

D Degree centrality, 31, 34, 74, 75, 87, 89–91, 112, 113, 118 Design-intensive fairs, 48 Diffusion, 61, 64, 75 Distance cognitive, 15, 16 Euclidean, 100 Diversity, 2, 12

E Embeddedness, 14, 16, 20, 22 Entrepreneur, 5 Entrepreneurship, 13, 42, 73, 75, 123 Equivalence, 112, 113, 118, 119 Evolutionary, 13, 61 External network(s), 27, 28, 61

F Face-to-face (F2F), 8, 18, 45 Fixed network, 75 Flexibility, 16, 123 Flexible specialization, 4

G Global firm, 115 Globalization, 5 Global production networks, 46

© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2021 Y. Yokura, Regional Innovation and Networks in Japan, International Perspectives in Geography 16, https://doi.org/10.1007/978-981-16-2191-8

125

126 Governance, 2, 3, 7, 17, 122

H Heterogeneous knowledge, 62, 76 High-tech industries, 17, 21 Homogeneous region, 63 Honda, O., 63 Horizontal, 2, 11, 17, 22, 58, 59, 121 collaboration, 11 relationship(s), 2, 11, 17, 22, 58, 59, 121 Human resources, 4, 15, 55, 77, 86, 90

I Incremental innovation, 124 Industrial, 1–4, 6, 7, 11, 12, 14, 16–23, 27, 28, 46, 47, 50–52, 56–59, 61–64, 66, 67, 70–72, 76, 81–83, 90, 91, 93–96, 108, 121–124 atmosphere, 6 cluster(s), 2, 4, 6, 14, 16, 17, 21, 22, 46, 50, 61, 63, 66, 81, 93, 123 Industrial Cluster Program, 80, 87 Industrial districts, 4, 50, 62, 77 Innovation network, 14, 16, 19, 75 Innovation policy, 50, 79 Innovation system, 2, 11, 13, 16–18 Institutional thickness, 4, 7, 19, 20, 23, 63, 76, 123 Interactive learning, 6, 20, 22 Inter-firm network, 14, 80, 88, 91, 94, 104, 110, 112, 117, 122

J Joint venture, 95, 104 Just-in-time, 4

K Knowledge base, 12, 13, 18–20, 22, 49, 93, 122 linkage, 23, 28 pools, 21 sharing, 46 transfer, 2, 5, 6, 13, 15–17, 45 Knowledge cluster initiative, 63, 73, 80

L Labor, 17, 19, 81, 93 market, 17, 19 mobility, 17

Index Learning process, 12, 46, 61, 62 Local trade fair, 7, 46, 47, 52, 58, 59, 122 Lock-in, 20, 93, 123 Loose coupling, 123

M Metropolitan areas, 47, 49, 52 trade fairs, 55 MICE, 12, 23, 50 Milieu, 4, 5, 13, 14, 17 Mobility, 13, 17, 20, 86, 121 Monitoring, 14 Monozukuri, 55, 94, 123 Multiple populations, 111, 113, 117 Mutual learning, 12, 17, 73 recognition, 17, 91 trust, 5, 27, 55, 59, 89, 91, 122

N National innovation system (NIS), 14, 18 Nation-state, 14, 17 NEC, 72, 81, 107, 108 NetDraw, 30, 74, 87, 100 Non-profit organization (NPO), 55, 56, 111 Novelty, 15

O Offshore company, 91 Open innovation, 2, 76 Optimal cognitive distance, 15 Organizational learning, 5 proximity, 14, 15, 17, 21 relations, 3 relationships, 3, 18, 19, 90, 121 trust, 111

P Patent, 2, 14, 18, 20, 28, 42, 73, 74, 94, 119 Path dependence, 14 Process innovation, 124 Project-based consortiums, 21 organizations, 22 production, 11, 20 Proximity cognitive, 7, 15, 22 geographical, 11, 21, 22, 45, 123

Index mental, 15 organizational, 14, 15, 17, 21 relational, 123 Public research institutes, 62, 123

R Radical innovation, 14, 19, 73, 124 Reciprocal relationships, 23 Redundancy, 76, 77 Relational turn, 1, 6 Routines, 15

S Science-based projects, 30–33, 66, 73, 74, 77, 94 technical fields, 31, 32, 68, 69 university, 53 Shareholders, 109, 110 Silicon Valley, 4, 5 Skilled labor, 22, 23 Small and medium-sized enterprises (SMEs), 58, 59, 67, 70, 76, 79, 80, 82, 83, 87 Social capital, 4, 118 Social network analysis, 2, 3, 7, 16, 27, 28, 42, 46, 62, 63, 67, 77, 80, 94, 100, 109, 110, 115, 117, 122, 123 Sony, 80, 81 Spillover, 11, 58, 59 Spin-off, 64 Start-up, 28, 40–42, 50, 73 Strategic alliances, 2, 6, 11, 12, 14, 17 Structural holes, 71

127 Subsidiaries, 15, 50, 73 Swift trust, 6, 22 Symbolic knowledge, 19, 23, 46 Synthetic knowledge, 18, 19, 23, 39–41, 46, 115, 119, 122, 123

T Tacit knowledge, 2, 6, 11, 12, 17, 18, 20, 121 Technological alliances, 7, 94 capabilities, 57, 76 catch-up, 90, 122 Technopolis, 62, 63, 66–69, 76, 79 Temporary cluster(s), 21, 22, 45, 46, 58, 61, 62, 80, 122 Third Italy, 4 Toyota, 4, 55 Transaction costs, 3, 16

U UCINET, 100 Uncertainty, 3, 4, 7, 12, 14, 21 Untraded-interdependencies, 11, 62

V Vertical, 4, 17, 22, 58 interactions, 22 relationships, 4, 17, 58, 121

Y Yamaha, 63, 72