Clusters, Networks, And Firm Innovativeness

Strategic Management Journal Strat. Mgmt. J., 26: 287–295 (2005) Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/smj.448

RESEARCH NOTES AND COMMENTARIES CLUSTERS, NETWORKS, AND FIRM INNOVATIVENESS GEOFFREY G. BELL* Labovitz School of Business and Economics, University of Minnesota Duluth Campus, Duluth, Minnesota, U.S.A

This paper extends current knowledge of industry clusters by disentangling the effects of networks from cluster (i.e., distinctly geographic) mechanisms on firm performance as well as by studying the influence of these different mechanisms on firms located inside and outside the industry cluster. It also highlights the importance of simultaneously modeling multiple networks which may differentially influence important firm outcomes. In the paper, I model the innovativeness of Canadian mutual fund companies as a function of their geographic location—inside or outside the industry cluster of Toronto—and of their centrality in networks of managerial and institutional ties. I find that locating in the industry cluster as well as centrality in the managerial tie network enhances firm innovation, while centrality in the institutional tie network does not. Copyright  2005 John Wiley & Sons, Ltd.

INTRODUCTION Industry clusters—groups of geographically proximate firms in the same industry—are a striking feature of the geography of economic activity (Krugman, 1991) examined by industrial geographers at least since Marshall (1920). Strategy scholars are now beginning to study how clusters influence firm performance, and have yet to distinguish between benefits associated with enhanced social interaction effects (Harrison, 1992) and Keywords: innovation; networks; multiple networks; industry clusters; mutual funds

∗ Correspondence to: Geoffrey G. Bell, Labovitz School of Business and Economics, University of Minnesota Duluth Campus, 110 SBE, 412 Library Drive, Duluth, MN 55812, U.S.A. E-mail: [email protected]

Copyright  2005 John Wiley & Sons, Ltd.

agglomeration economies (DeBresson and Amesse, 1991; Harrison, 1992; Harrison, Kelley, and Gant, 1996; Pascal and McCall, 1980; Shaver and Flyer, 2000). Many existing studies model only firms in the cluster, ignoring both ties between cluster members and other (‘remote’) firms (Harrison, 1994, and Saxenian, 1994b, are notable exceptions) and the overall industry network structure (Storper and Harrison, 1991). This paper overcomes some of these shortcomings by asking, ‘What is the relationship among industry clusters, network centrality, and firm innovativeness?’ I examine these relationships by examining the influence of clusters and network structure on the innovativeness of Canadian mutual fund companies. My study advances the strategy literature by untangling cluster mechanisms from

Received 22 November 1999 Final revision received 14 September 2004

288

G. G. Bell

network mechanisms and showing how they differentially influence important firm outcomes. It extends network theory by comparing partially overlapping networks of managerial and institutional ties and showing that centrality in different networks distinctly influences innovativeness. The paper proceeds as follows. The next section of the paper develops theory and hypotheses. I then outline my study setting and methodology, and present results. Finally, I present conclusions, limitations, and implications for scholars and managers.

THEORY AND HYPOTHESES A cluster is a group of firms from the same or related industries located geographically near to each other (Becattini, 1990; Brusco, 1990; Harrison et al., 1996; Storper and Harrison, 1991). Scholars such as Harrison (1994) and Porter (1990) predict that firms in the cluster should be more innovative than others for at least two reasons. First, firms in the cluster benefit from agglomeration economies such as nearby suppliers attaining efficient scale (Scott, 1992), direct observation of competitors (Burt, 1987; Harrison et al., 1996), and ability to exploit collective knowledge (Dosi, 1988; Marshall, 1920). Second, firms in clusters benefit from network-based effects, especially enhanced social interaction (Harrison, 1992). I begin my examination of these influences by briefly defining key terms. Key terms: Innovation and centrality Innovation is the development and implementation of new ideas to solve problems (Dosi, 1988; Van de Ven, 1986). According to Van de Ven, ‘An innovation is a new idea, which may be a recombination of old ideas, a schema that challenges the present order, a formula, or a unique approach which is perceived as new by the individuals involved’ (Van de Ven, 1986: 591). Innovation includes a pattern of informal cooperative R&D (von Hippel, 1987), resulting from informal information exchange among firms, so firms better positioned to access information should be more innovative (Rogers, 1995). Innovation also countenances the possibility that an actor may try to imitate others, but in the process inadvertently generate new ideas (March, 1994). It entails the Copyright  2005 John Wiley & Sons, Ltd.

development of new products or services as well as new administrative systems (Damanpour, 1991; Nohria and Gulati, 1996). Centrality measures the involvement in the network (Knoke and Burt, 1983): the extent to which an actor is deeply involved in network relations (Burt, 1980; Wasserman and Faust, 1994). It considers access to and control over resources (Knoke and Burt, 1983; Wasserman and Faust, 1994), and thus is likely to be highly associated with innovation, as access to and control over information and resources are associated with innovation (Becker, 1970; Powell, Koput, and Smith-Doerr, 1996; Rogers, 1995; von Hippel, 1988). Clusters and firm innovativeness Firms in clusters have better access to information than do other firms (Bianchi and Bellini, 1991; Porter, 1990; Pouder and St. John, 1996), resulting from both direct cluster effects as well as network processes underlying the cluster (Becattini, 1990; Brusco, 1990; Harrison, 1994). Thus, the total effect of clusters on innovation may be mostly indirect, partially influenced by network position. In this section, I focus on the direct effect of clusters on innovation that operate independently of network effects. Such cluster effects will arise partially because there is common knowledge available to members of the cluster (Geroski, 1995) that is not consciously transmitted among them (Marshall, 1920), or is transmitted via chance meetings between executives that are fostered by geographic proximity (Saxenian, 1994b). Common knowledge is augmented and reinforced by public information sources, such as the local media or universities (Porter, 1998; Saxenian, 1994b). Over time, the common knowledge forms a cluster level of absorptive capacity (Cohen and Levinthal, 1990). The ability to understand and exploit this clusterlevel absorptive capacity is enhanced by the common lineage and heritage of the firms in the cluster and their executives. Specifically, firms in clusters often share lineage to a common parent firm, such as the many firms in Silicon Valley directly or indirectly related to Fairchild (Saxenian, 1994a). More broadly, executives in geographically proximate firms share a common background and understanding (Paniccia, 1998). This common lineage and heritage will enable executives to understand information they may share when they ‘run across each Strat. Mgmt. J., 26: 287–295 (2005)

Research Notes and Commentaries other’ in chance settings (Saxenian, 1994b). Also, because information is sticky and place-specific and the ability to transfer information decays with distance (Ormrod, 1990; Saxenian, 1994a; von Hippel, 1994), firms in the cluster will have better access to common knowledge than geographically remote firms (‘remotes’). Thus, they tend to search locally for information used in innovation (Almeida and Kogut, 1997; Jaffe, Trajtenberg, and Henderson, 1993). Additionally, the geographic proximity of firms in the cluster enhances direct observation of competitors (Burt, 1987; Pascal and McCall, 1980; Rogers, 1995). A firm that observes others may try to mimic them and inadvertently generate innovation (March, 1994). Such inadvertent innovation may operate even in the absence of direct network ties, when the imitator cannot simply contact the other firm to learn more about an innovation, but must rely on cues from observing the other, increasing the likelihood of mutation and innovation. Firms outside the cluster (remotes) would have access to neither the cluster common knowledge nor the ability to directly observe their rivals, so would not be able to use these conduits for innovation (Powell et al., 1996). Thus: Hypothesis 1: In a given industry, after controlling for network effects, cluster-member firms will be more innovative than remotes.

Modeling multiple networks Scholars have long recognized that organizations are embedded in multiple, only partially overlapping, networks (Powell, 1985; Powell and SmithDoerr, 1994). Powell (1985) found no boundary between the work and personal life of his subjects, suggesting that modeling multiple networks is needed to understand how different networks influence outcomes. Consequently, I model a managerial network (the network of informal ties among managers) and an institutional tie network (the network of formal ties between their firms) to capture both informal and formal ties.

289

of others’ innovative efforts (Becker, 1970). They are less likely than others to miss valuable information (Becker, 1970), and have quick access to promising new ventures (Powell et al., 1996) that may generate innovation. Central firms may be better positioned to access the veracity of the information they receive as well as the information sources themselves (Burt, 1987). An information source may limit the information it provides, either for strategic reasons (to misrepresent the information) or to be ‘helpful’ (limiting information to what it believes the other party needs). The more a firm is involved in its network, the more it can compare information across sources and assess its veracity. Moreover, firms with multiple information sources are less likely to miss vital information as multiple information sources provide multiple channels to discover new information, and can combine information in novel ways to generate innovation (Van de Ven, 1986). Firms and their executives are involved in two distinct networks: a managerial network of informal ties among firm executives and an institutional tie network of formal ties between firms. The managerial network enhances information flow, especially the flow of tacit information among firms (Uzzi, 1996). It provides relatively high-trust context in which to communicate (Argyle and Henderson, 1985). Being central in this network may expose managers and their firms to a rich flow of tacit knowledge useful for innovation. Conversely, the institutional network provides opportunities to hear industry news. For example, if a trade association approves members’ new products, then serving on the association’s boards and committees provides early warning about competitor actions. Thus, centrality in each network should enhance innovativeness: Hypothesis 2: Centrality in the managerial network enhances firm performance. Hypothesis 3: Centrality in the institutional network enhances firm performance.

METHODS

Centrality and firm innovativeness

Research setting

Central actors are extensively involved in their networks (Burt, 1980; Freeman, 1979; Wasserman and Faust, 1994), so have highlighted knowledge

I collected the data on and from mutual fund companies listed in the January 1998 Membership Directory of the Investment Funds Institute

Copyright  2005 John Wiley & Sons, Ltd.

Strat. Mgmt. J., 26: 287–295 (2005)

290

G. G. Bell

of Canada (Investment Funds Institute of Canada, 1998b). One of the 78 firms was listed twice in the Directory, but included only once in the study. Thus, my final sample consisted of 77 Investment Funds Institute of Canada (IFIC) members, representing over 97 percent of assets under management in Canada (Investment Funds Institute of Canada, 1997). Forty-six of the 77 IFIC member firms maintain head offices in Toronto (10.6 firms per million people in the Toronto area); the other 31 maintain head offices in 10 cities across Canada, including nine in Vancouver (4.8 firms per million people) and seven in Montreal (2.1 firms per million people). Given the high concentration of mutual fund companies in Toronto, both in absolute number and relative to underlying population base, Toronto dominates the geography of this industry and firms in Toronto comprise the industry cluster.

(strongly disagree). Innovation is the summed and averaged responses received from all respondents for each firm. Cronbach’s alpha assessing the reliability of the three indicators of Innovation was 0.93, indicating a high level of reliability among the measures. Because Innovation was reverse scored (i.e., high values of Innovation indicate a lack of innovation), I reverse-coded the results as Innovativeness.

Clusters I coded the Cluster variable ‘1’ if a firm’s head office is in the cluster (Toronto), and ‘0’ otherwise. Data for this measure came from the Investment Funds Institute of Canada (1998a), company data (annual reports, etc.), BellCharts (1997), and Pape (1997).

Data and measures Mutual fund company innovativeness I collected information on firm innovativeness using a survey of industry experts: financial columnists in Canada’s business press and executives of mutual fund data services. I identified potential experts from lists of Southam News’ business columnists (including the Financial Post), investment and mutual fund reporters for the Globe and Mail, financial newsletter editors, and Canadian mutual fund data sources. My initial sample consisted of 23 persons, of whom 14 were general business writers unable to provide the detailed information I needed. The remaining nine persons had primary or sole duty analyzing the Canadian mutual fund industry. I received responses from seven of the nine experts, six of whom provided enough information to be usable. Five of the seven respondents were located in Toronto, two outside. Cronbach’s alpha score testing inter-rater reliability across the six respondents was 0.80, indicating broad consensus about firm innovativeness. I measured innovativeness along three dimensions by asking the industry experts to respond to three statements about each fund company: ‘This firm often leads the industry at (1) introducing new products/. . . (2) introducing new services/. . . (3) adopting new technologies.’ Respondents scored their responses to each statement on a 5point scale ranging from 1 (strongly agree) to 5 Copyright  2005 John Wiley & Sons, Ltd.

Managerial centrality I collected data on the managerial tie network of all 77 fund companies using a survey of executives (fund managers and top management) administered in February 1998. I developed the sample of fund company executives by contacting each mutual fund company and requesting a list of its executives, supplemented where necessary with alternate sources such as the firm’s web site. I sampled one to six executives from each firm. When the firm employed six or fewer executives, I surveyed all of them. When the firm employed more than six, I selected six randomly. Six firms were exceptions to this rule. Four cases consisted of two related companies from which I selected more than six respondents. In two cases, I sent more than six surveys because a respondent specifically suggested or requested that I send a survey to someone else in the firm. The procedures I followed closely parallel those advocated by Dillman (1978). I sent a cover letter and questionnaire to each executive in the sample asking them to respond by a certain date. Approximately 2 weeks later, I phoned the respondents and asked for their assistance, and 1 week later I sent non-respondents a second copy of the survey and an updated cover letter. I also called them and asked them to watch for the survey and assist me by completing it. Of the 305 surveys sent, 102 Strat. Mgmt. J., 26: 287–295 (2005)

Research Notes and Commentaries executives (33%) from 64 different firms (83%) responded.1 In the survey, I asked questions about the executives’ friendships, information, and advice networks which I used to develop the managerial tie network. For each of these variables, in cases where more than one executive from a given firm responded to the survey, I pooled their responses such that I record a friendship, information, or advice tie if any of the executives reported such a tie. I mapped each of the friendship, information, and advice networks in a 77 × 77 matrix. I coded cell ij ‘1’ if any executive at firm i was the recipient of a friendship, information, or advice tie, respectively, and ‘0’ otherwise. To assess agreement among the networks, I used the ‘simple matching’ QAP correlation routine (Borgatti, Everett, and Freeman, 1999), which calculates the extent to which there is the same entry in each cell in two matrices. The QAP correlation scores ranged from 0.90 to 0.93. I created the Managerial Centrality matrix by summing the three matrices and dichotomizing the result to record whether or not a tie existed from firm j to firm i. I symmetrized the results because centrality assesses involvement in the network (Knoke and Burt, 1983). I then measured in-degree (Wasserman and Faust, 1994) as Managerial Centrality for each firm by summing the columns to yield the number of firms tied to i. Institutional centrality I created the Institutional Tie matrix using data from the Investment Funds Institute of Canada (1998a), which listed the 1997–98 IFIC Board of

291

Directors, IFIC Management Committee, and IFIC Retail Council. I created three measures of institutional ties: IFIC Board of Director Ties, IFIC Management Committee Ties, and IFIC Retail Council Ties. Because of the reciprocal nature of these relationships, I measured these as symmetric ties. I recorded each of these relations in separate 77 × 77 matrices, wherein cell ij was coded ‘1’ if there was a tie between firms i and j , and ‘0’ otherwise. There was high agreement among these networks, with QAP correlations ranging from 0.95 to 0.97. I summed the matrices into the Institutional Tie matrix, whose cells are valued from ‘0’ (none of the three ties are present) to ‘3’ (all three of the ties are present). As with the Managerial Tie matrix, I created Institutional Centrality by dichotomizing the Institutional Tie matrix and summing the columns of the resultant matrix. Control variables Prior studies identified a significant positive relationship between firm size and innovativeness and a significant negative relationship between firm age and innovativeness. I control for these factors. Market Share is the firm’s assets under management as at December 31, 1997 divided by total industry assets under management at that date. Firm Age is 1998 less the earlier of either the date of founding of the fund company or (in two cases where that was unavailable) the date of inception of the company’s oldest fund.

RESULTS Variable correlations

1

To examine potential problems associated with having responses from only 64 of 77 firms, I reran all analyses using matrices generated by data from the 64 responding firms. There were no significant differences from the results reported herein.

Table 1.

Table 1 presents means, standard deviations, and correlations among variables included in the analysis.

Means, standard deviations, and correlations

Variable

Mean

Standard deviation

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

Innovativeness Toronto Managerial centrality Institutional centrality Market share Firm age

2.70 0.62 15.17 4.60 1.30 18.81

1.01 0.49 11.47 7.44 2.31 13.63

∗∗∗

p < 0.001;

∗∗

1. – 0.39∗∗∗ 0.49∗∗∗ 0.43∗∗∗ 0.60∗∗∗ 0.12

2. – 0.26∗ 0.11 0.20† 0.13

3.

— 0.44∗∗∗ 0.43∗∗∗ 0.18

4.

— 0.61∗∗∗ 0.14

5.

— 0.36∗∗∗

p < 0.01; ∗ p < 0.05; †p < 0.10

Copyright  2005 John Wiley & Sons, Ltd.

Strat. Mgmt. J., 26: 287–295 (2005)

292

G. G. Bell

Table 2.

Predictors of firm innovativeness

Variable Toronto Managerial centrality Institutional centrality Market share Firm age Adjusted R 2

Innovativeness 0.50∗∗ (0.19) (0.01) 0.02∗ 0.00 (0.02) 0.21∗∗∗ (0.05) −0.01 (0.01) 0.47∗∗∗

∗∗∗ p < 0.005; ∗∗ p < 0.01; ∗ p < 0.05; †p < 0.10 Results for coefficients; standard errors in parentheses.

Hypotheses testing I tested the hypotheses using a regression equation model that modeled the influence of the cluster variable, the network variables, and the controls on firm innovativeness (see Table 2).

Factors influencing firm innovativeness Hypothesis 1 (locating in the Cluster enhances Innovativeness) is strongly supported (β = 0.50; p < 0.01). Firms in the cluster are significantly more innovative than remotes. Hypothesis 2 (Managerial Centrality enhances Innovativeness) was also supported (β = 0.02; p < 0.05), while Hypothesis 3 (Institutional Centrality enhances Innovativeness) was not supported (β = 0.004; n.s.). Market Share was positively significantly associated with Innovativeness (β = 0.21; p < 0.001) and Firm Age was not significant (β = −0.01; n.s.). Moreover, the possibility exists that innovation influences centrality as well as centrality influencing innovation (the dimension of causality I hypothesized here). I tested this by estimating the model with a two-stage least-squares (2SLS) technique using instrumental variables for the two centrality variables (results not reported). I applied a Hausman test, which tests whether the coefficients under the 2SLS estimate (which are efficient if innovation influences centrality) differs from the OLS estimates (which are efficient if innovation does not influence centrality). The Hausman test could not reject the null hypothesis that the two sets of estimates are the same (chi-square of χ 2 (5) = 6.34, Prob > χ 2 = 0.2745), meaning that we can rely on the OLS estimates. Copyright  2005 John Wiley & Sons, Ltd.

DISCUSSION Clusters, networks, and firm innovativeness This study set out to examine the relationship among clusters, networks, and firm innovativeness, and proposed that clusters and network centrality should enhance firm innovativeness. Indeed, locating in the cluster enhances firm innovativeness, even after separately accounting for the influence of network structure. Expanding on Marshall’s notion that in clusters information is ‘in the air’ forming a common good available to all (Marshall, 1920), common knowledge may reflect chance meetings that are likely to occur when executives are geographically proximate (Saxenian, 1994b) and when they share a common lineage (Saxenian, 1994a) and background (Paniccia, 1998). Alternatively, the positive influence of clusters may reflect ready access to geographically proximate supporting industries, such as commercial banks and the Toronto Stock Exchange. The study also shows the differential effects of managerial and institutional ties on firm innovativeness. As hypothesized, centrality in the managerial tie network enhanced firm innovativeness, suggesting that the informal friendship and communication network provides an important source of novel information useful in innovation (Uzzi, 1996). On the other hand, I found no significant relation between institutional ties and innovativeness. Perhaps, in this industry at least, institutional ties were used solely for the transmission of relatively well-known information, and failed to generate communication that was enough, deep enough, or novel enough (Uzzi, 1996) to enhance innovation.

Limitations of the study The study does not account for firm-specific factors influencing innovativeness, such as absorptive capacity. Because the study focused on the role of geographic clusters and networks, it did not measure absorptive capacity or similar firm-specific factors that may influence firm ability to translate information into innovation. On the other hand, size and age may provide reasonable controls for these differences. Strat. Mgmt. J., 26: 287–295 (2005)

Research Notes and Commentaries Implications for research This study challenges both strategy and network researchers. For strategy scholars, the study shows the importance of distinguishing cluster and network mechanisms. Simply aggregating cluster and network mechanisms into ‘cluster effects’ (Bianchi and Bellini, 1991) confounds benefits related to clustering per se with benefits arising from network ties. Thus, the study shows the importance of examining cluster effects in greater detail, clearly differentiating different mechanisms that drive performance. For network scholars, this study shows the importance of clearly choosing the network(s) modeled, and modeling multiple networks when necessary (Baum and Dutton, 1996; Powell and Smith-Doerr, 1994). Network scholars may choose the networks they model on the basis of convenience. However, if, as in this study, different networks distinctly influence important outcomes, then modeling only a single network on the basis of convenience risks attributing to one network effects actually generated by another, producing a cross-level fallacy (Rousseau, 1985). Thus, this study also amplifies the importance of studying multiple networks simultaneously (Galaskiewicz and Zaheer, 1999; Powell, 1985; Powell and Smith-Doerr, 1994). While Powell (1985) found that actors may use their networks for personal use, I find the opposite also holds: managers may use their personal networks for the benefit of their firms. Although not tested in this study, it is quite possible that different networks reflect different causal mechanisms. (For example, Galaskiewicz and Zaheer, 1999, assert that agents’ social networks are more likely to contain affective content than are interorganizational networks.) Thus, network scholars must seriously consider what mechanisms they expect to operate in the context they are studying and choose networks that capture those mechanisms. Implications for managers My study shows that, even accounting for network ties, firms in clusters are more innovative than are remotes. However, because the study did not examine specific mechanisms underlying cluster effects, we must be cautious in providing advice to managers. However, managers can take away several points. First, because cluster members are Copyright  2005 John Wiley & Sons, Ltd.

293

more innovative than remotes, managers who are concerned with innovation should consider locating some of their innovative-related operations to the cluster. I do not suggest that moving to the cluster will necessarily make a firm more innovative; however, doing so will likely increase access to firms that tend to be more innovative. Additionally, managers must recognize the importance of network ties, especially their managerial network ties, on the performance of their firms. While managers have recently focused their attention on structuring their firms’ network of formal ties, such as strategic alliances (Doz and Hamel, 1998), this study shows that managers need to manage their informal networks carefully as well (Burt, 1992).

CONCLUSION This paper responds to a call for researchers to examine innovation in service industries, especially in financial services (Drazin and Schoonhoven, 1996). It ‘untangles’ cluster effects by separately modeling cluster-based mechanisms and social interaction (network) mechanisms. In so doing, it finds that both cluster and network mechanisms influence Canadian mutual fund company innovativeness and that innovativeness, in turn, enhances firm prestige.

ACKNOWLEDGEMENTS I thank Milton Boyd, Phil Bromiley, Joe Galaskiewicz, Aks Zaheer, the University of Minnesota Strategy Research Group, and two anonymous SMJ reviewers for their input into this project. I also thank the University of Minnesota for providing financial assistance for this project.

REFERENCES Almeida P, Kogut B. 1997. The exploration of technological diversity and geographic localization in innovation: start-up firms in the semi-conductor industry. Small Business Economics 9(1): 21–31. Argyle M, Henderson M. 1985. The Anatomy of Relationships and the Rules and Skills Needed to Manage them Successfully. William Heinemann: London. Baum J, Dutton J. 1996. The embeddedness of strategy. In Advances in Strategic Management, Vol. 13, Shrivastava P, Huff AS, Dutton JE (eds). JAI Press: Greenwich, CT; 3–40. Strat. Mgmt. J., 26: 287–295 (2005)

294

G. G. Bell

Becattini G. 1990. The Marshallian industrial district as a socio-economic notion. In Industrial Districts and Inter-firm Co-operation in Italy, Pyke F, Becattini G, Sengenberger W (eds). International Institute for Labour Studies: Geneva; 37–51. Becker MH. 1970. Sociometric location and innovativeness: reformulation and extension of the diffusion model. American Sociological Review 35: 267–282. BellCharts. 1997. BellCharts for Windows. The Financial Post DataGroup: Toronto. Bianchi P, Bellini N. 1991. Public policies for local networks of innovators. Research Policy 20: 487–497. Borgatti S, Everett MG, Freeman LC. 1999. UCINET 5 for Windows: Software for Social Network Analysis (5th edn). Analytic Technologies: Natick, MA. Brusco S. 1990. The idea of the industrial district: its genesis. In Industrial Districts and Interfirm Co-operation in Italy, Pyke F, Becattini G, Sengenbeger W (eds). International Institute for Labour Studies: Geneva; 10–19. Burt RS. 1980. Models of network structure. Annual Review of Sociology 6: 79–141. Burt RS. 1987. Social contagion and innovation: cohesion versus structural equivalence. American Journal of Sociology 92(May): 1287–1335. Burt RS. 1992. Structural Holes: The Social Structure of Competition. Harvard University Press: Cambridge, MA. Cohen WM, Levinthal DA. 1990. Absorptive capacity: a new perspective on learning and innovation. Administrative Science Quarterly 35: 128–152 Damanpour F. 1991. Organizational innovation: a metaanalysis of effects of determinants and moderators. Academy of Management Journal 34(3): 555–590. DeBresson C, Amesse F. 1991. Networks of innovators: a review and introduction to the issue. Research Policy 20: 363–379. Dillman DA. 1978. Mail and Telephone Surveys: The Total Design Method . Wiley: New York. Dosi G. 1988. Sources, procedures, and microeconomic effects of innovation. Journal of Economic Literature 26(September): 1120–1171. Doz YL, Hamel G. 1998. Alliance Advantage: The Art of Creating Value through Partnering. Harvard Business School Press: Boston, MA. Drazin R, Schoonhoven C. 1996. Community, population, and organization effects on innovation: a multilevel perspective. Academy of Management Journal 39(5): 1065–1083. Freeman LC. 1979. Centrality in social networks: conceptual clarification. Social Networks 1: 215–239. Galaskiewicz J, Zaheer A. 1999. Networks of competitive advantage. In Research in the Sociology of Organizations, Andrews S, Knoke D (eds). JAI Press: Greenwich, CT; 237–261. Geroski P. 1995. Markets for technology: knowledge, innovation and appropriability. In Handbook of the Economies of Innovation and Technological Change, Stoneman P (ed). Basil Blackwell: Oxford; 90–131. Harrison B. 1992. Industrial districts: old wine in new bottles? Regional Studies 26(5): 469–483. Copyright  2005 John Wiley & Sons, Ltd.

Harrison B. 1994. Lean and Mean: The Changing Landscape of Corporate Power in the Age of Flexibility. Basic Books: New York. Harrison B, Kelley MR, Gant J. 1996. Innovative firm behavior and local milieu: exploring the intersection of agglomeration, firm effects, and technological change. Economic Geography 72(3): 233–258. Investment Funds Institute of Canada. 1997. December Statistics. The Investment Funds Institute of Canada: Toronto. Investment Funds Institute of Canada. 1998a. Listing of Board and Committee Members. The Investment Funds Institute of Canada: Toronto. Investment Funds Institute of Canada. 1998b. Membership Directory. The Investment Funds Institute of Canada: Toronto. Jaffe AB, Trajtenberg M, Henderson R. 1993. Geographic localization of knowledge spillovers as evidenced by patent citations. Quarterly Journal of Economics 108(3): 577–598. Knoke D, Burt RS. 1983. Prominence. In Applied Network Analysis: A Methodological Introduction, Burt RS, Minor M (eds). Beverly Hills: CA; 195–222. Krugman P. 1991. Geography and Trade. MIT Press: Cambridge, MA. March JG. 1994. A Primer on Decision Making: How Decisions Happen. Free Press: New York. Marshall A. 1920. Principles of Economics. Macmillan: London. Nohria N, Gulati R. 1996. Is slack good or bad for innovation? Academy of Management Journal 39(5): 1245–1264. Ormrod RK. 1990. Local context and innovation diffusion in a well-connected world. Economic Geography 66(2): 109–122. Paniccia I. 1998. One, a hundred, thousands of industrial districts: organizational variety in local networks of small and medium-sized enterprises. Organization Studies 19(4): 667–699. Pape G. 1997. Gordon Pape’s 1998 Buyer’s Guide to Mutual Funds. Prentice-Hall Canada: Scarborough, Ontario. Pascal AH, McCall JJ. 1980. Agglomeration economies, search costs, and industrial location. Journal of Urban Economics 8: 383–388. Porter ME. 1990. The Competitive Advantage of Nations. Free Press: New York. Porter ME. 1998. Clusters and the new economics of competition. Harvard Business Review 76(6): 77–90. Pouder R, St. John CH. 1996. Hot spots and blind spots: geographical clusters of firms and innovation. Academy of Management Review 21(4): 1192–1225. Powell WW. 1985. Getting Into Print: The DecisionMaking Process in Scholarly Publishing. University of Chicago Press: Chicago, IL. Powell WW, Koput KW, Smith-Doerr L. 1996. Interorganizational collaboration and the locus of innovation: networks of learning in biotechnology. Administrative Science Quarterly 41: 116–145. Powell WW, Smith-Doerr L. 1994. Networks and economic life. In The Handbook of Economic Strat. Mgmt. J., 26: 287–295 (2005)

Research Notes and Commentaries Sociology, Smelser NJ, Swedberg R (eds). Princeton University Press: Princeton, NJ; 368–402. Rogers EM. 1995. Diffusion of Innovations (4th edn). Free Press: New York. Rousseau DM. 1985. Issues of level in organizational research. In Research in Organizational Behavior, Vol. 7, Cummings LL, Staw BM (eds). JAI Press: Greenwich, CT; 1–37. Saxenian A. 1994a. Lessons from Silicon Valley. Technology Review 97(5): 42–51. Saxenian A. 1994b. Regional Advantage: Culture and Competition in Silicon Valley and Route 128 . Harvard University Press: Cambridge, MA. Scott AJ. 1992. The role of large producers in industrial districts: a case study of high technology systems houses in Southern California. Regional Studies 26(3): 265–275. Shaver JM, Flyer F. 2000. Agglomeration economies, firm heterogeneity, and foreign direct investment in the United States. Strategic Management Journal 21(12): 1175–1193. Storper M, Harrison B. 1991. Flexibility, hierarchy and regional development: the changing structure of

Copyright  2005 John Wiley & Sons, Ltd.

295

industrial production systems and their forms of governance in the 1990s. Research Policy 20: 407–422. Uzzi B. 1996. The sources and consequences of embeddedness for the economic performance of organizations: the network effect. American Sociological Review 61: 674–698. Van de Ven AH. 1986. Central problems in the management of innovation. Management Science 32(5): 590–607. von Hippel E. 1987. Cooperation between rivals: Informal know-how trading. Research Policy 16: 291–302. von Hippel E. 1988. The Sources of Innovation. Oxford University Press: New York. von Hippel E. 1994. ‘Sticky information’ and the locus of problem solving: implications for Innovation. Management Science 40(4): 429–439. Wasserman S, Faust K. 1994. Social Network Analysis: Methods and Applications. Cambridge University Press: Cambridge, U.K.

Strat. Mgmt. J., 26: 287–295 (2005)