Clusters and Innovation in Biotechnology

This paper focuses on biotechnology innovation in a peripheral region of Canada. It views the industry sector from a cluster standpoint. The paper presents descriptive data on nine companies and 13 support organizations that have been visited as part of a continuing study. At this point in time, it is not clear whether the collection of organizations in Halifax, Nova Scotia represents a biotechnology “cluster” as defined in the literature. Presented at the 19. Industrial Marketing and Purchasing Conference Lugano, Switzerland September 4-6, 2003 Clusters & Innovation: Biotechnology / 2 Introduction In an interesting development, three papers at the 18. IMP Conference (Johnston & Araujo 2002, Hakanson, Tunisini & Waluszewski 2002, Waluszewski 2002a) explicitly examined the spatial elements of business processes, including marketing and purchasing activities. Continuing in the same vein, the present paper examines the geographic embeddedness of companies and their interactions. It originates from a study of innovation in an emerging biotechnology industry sector. The setting is an agglomeration of companies and support organizations that exist within a 100-kilometre radius of Halifax, in Nova Scotia, Canada. The paper reports on research that is in progress. The structure of the paper is as follows. First, a brief literature review is provided. Second, some contextual material on biotechnology is presented, along with information on the industry in Canada. Third, the study is described and findings from preliminary interviewing revealed. The data are mostly drawn from interviews with nine companies but are supplemented with information from 13 support organizations. The 22 completed interviews represent about one-half of those planned. Several key variables have been selected for analysis and presentation in the paper; this provides a sampling of what will follow. The paper concludes with a discussion of the preliminary findings. Background Literature In this section, the theoretical basis for the present study is outlined. This is organized around three terms: business networks, industry clusters, and innovation. IMP researchers have pioneered the use of a business networks approach to the study of industrial markets. The focus of early IMP work was on interaction in buyer-seller dyads but the emphasis shifted over time to the analysis of networks and interorganizational relationships. The change in approach was prompted by a desire more fully to understand the complexity of business markets and to model these more accurately. A substantial literature has resulted from the work of IMP researchers, who have written numerous books and journal articles on a wide range of related topics. The changing nature of IMP enquiry is, perhaps, best reflected in the collections edited by Ford, now in a third edition (2002). Although IMP work is wide ranging, geographic aspects of networks have seldom been explicitly considered. This leads Johnston and Araujo to comment that “research in interorganizational networks has stressed functional links in industrial networks but has neglected the role of space in fostering particular forms of relationships” (2002, p. 1). Similarly, Håkansson, Tunisini and Waluszewski (2002) argue that the location of interacting companies (or “place”) has received limited attention in work by IMP researchers. The relevant body of knowledge on industry clusters is large and spans several disciplines, including economic geography, evolutionary economics, regional innovation systems, economic development, as well as organizational studies. Alfred Marshall wrote about the concentration of industry in local areas in Britain more than one hundred years ago but it was the success of craft-based industrial districts in Italy, as well as in other technologies and locations, that gave rise to numerous studies of clusters over the past decade. The recent interest in clusters also results from Porter’s influential work on the competitive advantage of nations, regions and cities (1990, 1998). Although there is no generally accepted definition of clusters, Porter’s is illustrative. Specifically, a cluster is "a geographically proximate group of interconnected companies and associated institutions in a particular field, linked by commonalities and complementarities.“ Following this line of reasoning, clusters promise a "collective efficiency“ that produce higher returns for firms from a combination of external economies and joint actions (Rosenfeld 2001). More specifically, Malmberg & Maskell (2002) posit that the benefits of clusters stem from shared costs for infrastructure, the development of a skilled work force, transaction efficiency, and knowledge spillovers that produce learning and innovation. Proximity is central to the realization of these benefits, 1 Mota and de Castro (2001) provide an exception to this general observation. 2 See Martin and Sunley (2003) for a recent critique of clusters and Porter’s work. Clusters & Innovation: Biotechnology / 3 especially those that depend on fact-to-face contacts and the transmission of tacit knowledge. The benefits of clusters and the allure of enhanced business performance has led many government agencies to support clusters as a tool of economic development. Whether this faith will be rewarded is unclear. Much of the research on clusters is anecdotal and descriptive (Staber 2001), leading some writers to state that the performance of clusters has still to be scientifically demonstrated (Martin & Sunley 2003). Clearly, networks and clusters are related concepts. A simple view of their relationship might be that clusters include those members of networks lying within a given geographic range. An alternative view is presented in Table 1. Innovation is a central activity in business networks and industry clusters. Regardless of the focus or setting, organizations collaborate with one another so as to maintain or enhance competitiveness. Thus, process or product improvements provide much of the impetus for companies to work together (Lorenzoni & Lipparini 1999) and collaboration is just as much a part of the craft industry landscape as it is in sectors labelled as “high technology” (Perry 1999). From the standpoint of this study, a number of contributions to the literature on innovation deserve mention. Innovation is increasingly studied at the regional (as opposed to national) systems level since local infrastructure, resources and conditions very much influence activity and success. An illustration is provided by Feldman and Florida, who argue that “geographic proximity of these inputs promotes information transfer and spill-overs that lower the costs and reduce the risks associated with innovation” (1994, p.214). There is a strong IMP tradition of examining innovation processes at the firm, dyad, network and industry level. Håkansson (1987) proposed that product development in any firm is likely to involve a process of interaction with many others and that it should, therefore, be looked at from a network perspective. This approach is employed in Lundgren’s (1993) study of innovation in digital image technology in Sweden. Rather than studying innovation in a single company, Lundgren looks at the evolution of a new technology and the development of a network of innovating and support organizations. A final issue concerns innovation in peripheral regions. Rosenfeld (2002) identifies the barriers that exist to innovation (and cluster development) in such regions as well as actions that can be taken. The barriers include: “a weak infrastructure; lack of access to technology, innovation and capital; regional insularity and isolation; low educational levels and low skilled work force; absence of talent; and an overly mature or hierarchical industry structure” (p. 9). Geography is an important element in business activities. However, spatial factors are considered differently in the bodies of literature described above. The clusters and innovation literatures treat questions of location and place more explicitly than does the networks literature. However, this apparent oversight is being addressed in current IMP work. Building on Dicken and Malmberg’s (2001) notion that firms are embedded both territorially and functionally, Johnston & Araujo propose “an agenda for research into the role of spatial proximity in understanding industrial systems” (2002, p. 1) and outline a study that is underway. Håkansson, Tunisini and Waluszewski (2002) also note the lack of attention paid to place in academic research. The authors present two case studies illustrating how location has been used (rather than simply being accepted) by the companies in question, and go on to propose a schema for the strategic use of place by companies. This is a promising development. Finally, Waluszewski (2002a) employs a range of methods, including the schema mentioned above, to examine the biotechnology cluster in Uppsala. Her early findings are that technological and commercial development have evolved over many years, includes many actors, and does not follow a simple linear pattern of development. The industrial and geographic setting of the present study are described in the next sections. 3 Rosenfeld (2002) identifies three types of less developed region: (1) older industrialised regions that have lost their cost advantage to other nations, (2) semi-industrialised regions with small, craft-based firms with low levels of technology, and (3) peripheral or less populated regions where productivity and out-migration concerns mean that resource based industries must be supplemented by those with growth opportunities. Clusters & Innovation: Biotechnology / 4 Table 1: Networks and Clusters Compared