Information Technology in a Global Economy

The Information Society, 28: 110–120, 2012 Copyright c© Michigan State University Board of Trustees ISSN: 0197-2243 print / 1087-6537 online DOI: 10.1080/01972243.2012.651004

Small and Medium-Sized Enterprises in Rural Business Clusters: The Relation Between ICT Adoption and Benefits Derived From Cluster Membership

Charles Steinfield, Robert LaRose, and Han Ei Chew Department of Telecommunication, Information Studies, and Media, Michigan State University, East Lansing, Michigan, USA

Stephanie Tom Tong Department of Communication Studies, University of Puget Sound, Tacoma, Washington, USA

This study examines the relationships between information and communication technologies (ICT) usage, the benefits a company derives from membership in a rural business cluster, and the suc- cess of rural companies. Analysis of 333 rural businesses located in northern lower Michigan showed a strong relationship between (a) ICT adoption and benefits derived from the membership in business clusters, (b) ICT adoption and self-reported business suc- cess, and (c) benefits derived from business clusters and business success. Although analysis indicates that these relationships may be industry specific, results suggest that ICT adoption by rural enterprises may have advantages for the region’s social capital and business success and may help reduce the digital divide experienced in rural communities.

Keywords information and communication technologies, business clusters, small and medium-sized enterprises, social capital

One strategy proposed to increase the adoption of in- formation and communications technologies (ICT) in ru- ral communities has been to ensure that rural small and medium-sized enterprises (SMEs) have adequate ICT ac- cess, as that would increase the chances of the technology diffusing into the rest of the community (Hollifield and Donnermeyer 2003; LaRose, Gregg, Strover, Straubhaar, and Carpenter 2007). In effect, it is hoped that exposure

Received 20 November 2009; accepted 26 April 2011. Address correspondence to Han Ei Chew, 12 Communication Arts

& Sciences Building, Michigan State University, East Lansing, MI 48825, USA. E-mail: [email protected]

to the Internet in the workplace would trigger a positive growth cycle, with the widespread diffusion of ICT leading to increased business formation and that in turn enhanc- ing technology diffusion into the rest of the community. The resulting increased in demand would in turn attract infrastructure investment in rural communities and the dig- ital divide would correspondingly be reduced. Using var- ious economic models, Katz and Suter (2009) estimated that approximately 128,000 jobs (or 32,000 jobs per year) could be generated from the deployment of broadband infrastructure.

Within this context, regional business clusters are of special relevance as an economic development strategy for rural areas (Porter 1990; Rosenfeld 2001). There is a vast literature exploring the impact of regional busi- ness clusters on economic development derived from Mar- shall’s (1920) groundbreaking work on industrial districts (Breschi and Malerba 2001; Porter 1990; 2000; Pratt 2000). Regional business clusters are typically defined as groups of companies in a common industry located in the same geographic area, often including a range of supporting players such as local trade associations and education and research institutional linkages (Porter 2000). Regional business clusters improve business per- formance by endowing certain localities with resource advantages and open up opportunities for e-business in- frastructure development, while simultaneously sparking innovation through competition among geographically proximate members (Breschi and Malerba 2001; Porter 2000; Pratt 2000). Regional business clusters are also as- sociated with regional social capital, defined broadly as the relational and informational benefits that accrue to the region arising from the connections among people in

110

SOCIAL CAPITAL AND RURAL BUSINESSES 111

area companies (Steinfield, Scupola, and Lopez-Nicolas 2010).

Clustering around common industries may be even more important for rural regions, given that rural areas can lack many of the transaction economies associated with proximity that are found in urban areas where eco- nomic activity is concentrated (Leamer and Storper 2001). Moreover, electronic linkages among firms may stimulate the formation of business clusters in rural areas that can improve the performance of firms (Porter 2004) and boost rural economies. Examples of rural clusters span the na- tion, from an aquaculture cluster on the coast of Maine, to a wood products cluster in Oregon (RTS 2003).

The use of broadband within a business cluster context has not been extensively studied, however, especially in a rural context. There is some indication, however, that firms in clusters derive greater benefit from their use of Internet-related services made possible by broadband con- nections than firms located outside any sort of business cluster (Steinfield and Scupola 2008). Among the possi- ble explanations for benefits of cluster membership, one plausible one is that diffusion of ICTs may be faster in a cluster due to imitation and learning effects. ICT use in turn facilitates interactions between cluster members and that helps generate regional social capital, which re- inforces the positive growth cycle (Steinfield and Scupola 2008). Moreover, when clusters enjoy a strong reputation, this may enhance business success as firms are better able to use ICTs to support transactions with distant clients (Steinfield and Scupola 2008).

In this article we examine the relationships between the adoption and use of broadband-related ICTs in a rural re- gion, the benefits that companies derive from rural cluster membership, and the overall success of rural enterprises. These relationships are investigated in the context of a sur- vey of rural enterprises within three industries clustered in the northern part of Michigan’s lower peninsula.

ICTs AND BENEFITS OF CLUSTER MEMBERSHIP

Clusters possess a stockpile of knowledge built over time based on experience of their members (Barkley, Kim, and Henry 2001). Cluster members can take advantage of this knowledge stock through what Bernat (1999) calls “knowledge spillovers.” However, all clusters are not sim- ilarly endowed. Therefore, the benefit a company is likely to derive from cluster membership depends on the endow- ments of the particular cluster to which it belongs. In other words, companies located in well-developed clusters are likely to benefit more from cluster membership than those in less-developed clusters. There are a number of reasons why this should be so. Stronger clusters are more likely to attract a more qualified labor force, enhancing the qual- ity of the information that flows in knowledge spillovers.

Moreover, greater reputational advantages may accrue to cluster members in strong clusters; for example, a tech- nology in firm in Silicon Valley may be perceived to be more competent than one in another part of the country, simply because it is situated in the region. The following hypothesis summarizes this expected relationship:

H1: The benefits derived from cluster membership are positively correlated with the strength of the business cluster.

Continuing with this line of reasoning, the benefits a company derives from a business cluster are likely to be directly related to its integration within the cluster. For example, companies that participate in trade meetings or- ganized by cluster members are likely to have greater op- portunities for learning from other cluster members. We therefore hypothesize:

H2: The benefits derived from the cluster are positively correlated with the strength of cluster membership.

One key difference between rural and urban business clusters is the greater physical distance that can separate the cluster members. These distances limit the opportuni- ties for social interactions in person. Within this context, greater use of electronic communication can facilitate the development of relationships among rural SMEs, which otherwise might remain quite isolated. Although such technologies also facilitate greater interaction with distant partners, and so might weaken clusters, prior research on clusters suggests that ICT use strengthens rather than weakens cluster interaction (Steinfield and Scupola 2008). Hence, the following hypothesis is proposed:

H3: Adoption of ICTs among rural SMEs is positively correlated with the benefits derived from business clusters.

Decades of research on ICT use in business suggest that greater adoption and use should be associated with improved company performance, despite the many ob- stacles to measuring productivity impacts of information technologies (Barua, Kriebel, and Mukhopadhyay 1995; Brynjolffson and Hitt 2000; Parker and Benson 1988; Strassman 1985; Zhu 2004). We therefore hypothesize that ICT use among rural SMEs should be positively as- sociated with business success.

H4: Adoption of ICTs among rural SMEs is positively correlated with business success.

Porter (2000) has provided significant evidence that the set of companies in well-developed business clusters outperform their counterparts in other regions where the same specific industry clusters are not present. We thus expect that companies that report stronger informational and relational benefits from their cluster should experience stronger performance, leading to our fifth hypothesis.

H5: Benefits of business cluster membership are posi- tively correlated to business success.

112 C. STEINFIELD ET AL.

FIG. 1. Conceptual model of ICT use, business cluster factors, and business success.

Social capital and increased ICT use are necessary in order for rural SMEs to develop. Together, they are hy- pothesized to have both direct and indirect effects on the success of rural SMEs. These hypothesized relationships (H1-H5) and theoretical model are presented in Figure 1.

Although it is predicted that an overall increase in pro- ductivity is associated with ICT adoption, it is likely that success will vary by the type of industry in which the clus- ter operates. Service industries and technology-intensive manufacturing firms are more likely than others to bene- fit from network connections (Windrum and de Berranger 2002). We therefore propose:

H6: The effect of ICT adoption on productivity will vary by industry type.

METHOD

A list of 4,482 small business firms was obtained from USADATA for a 29-county region at the northern tip of the lower Michigan peninsula bounded by Mason, Lake, Osceola, Clare, Gladwin, Bay, and Huron counties to the south. Businesses were initially screened by SIC code, and firms engaged exclusively in local service businesses that were unlikely to benefit from cluster membership (e.g., beauty salons) were excluded. Following Gibbs and Bernat (1997), wood products, metalworking, and machin- ery and computer-related goods and services were initially considered. Since USADATA provided SIC codes rather than NAICS codes, it was necessary to search across codes to identify related firms. This selection process resulted in the selection of 388 firms in the wood products industry, 469 in metalworking, and 222 in information technology, for a total of 1079 firms in the sample.

The firms in the final sample averaged 13 employees and had been in business for 19 years. On average they

derived 56 percent of their revenue from within the local region, 20 percent from elsewhere in Michigan, 19 percent from other states, and 2 percent from outside the United States. According to data compiled by USADATA, two- thirds reported annual revenues of $500,000 or less.

The tailored design method mail survey methodology (Dillman 2000) was followed, included a prenotification letter, a second mailing with the questionnaire booklet, cover letter, and a $1 cash incentive, and a follow-up post- card. After a period of one week, a duplicate of the first mailing, minus the incentive payment, was sent. Surveys were addressed to the chief executive officer (CEO) of each firm. Of the 1079 surveys, mailed 421 were com- pleted and returned. There were 62 bad addresses and 24 firms were disqualified on being informed that they were out of business, their owner was deceased, and other such reasons. Two surveys had high instances of missing data and were dropped from final analysis. The final response rate was therefore 42 percent or 333 surveys.

Items for the business surveys were selected from prior surveys of Internet adoption (e.g., LaRose and Hoag 1996) and information technology utilization (e.g., Pflughoeft et al. 2003; Riemschneider et al. 2003). The survey included indicators of the financial performance of the firm and factors known to be related to successful adoptions of Internet-based e-business applications (after Windrum and de Berranger 2002). Multi-item additive indices were constructed, with the (Cronbach alpha) internal consistency coefficients indicated in Appendix A. The dependent variable—business success—was based on a six-item additive scale for self-reported business characteristics over a two-year period. Items on the scale measured increase in the variety of products, improvement of sales outside of the region, addition of new positions to the payroll, increased profitability, improvement in the relationships with customers, and increased visibility

SOCIAL CAPITAL AND RURAL BUSINESSES 113

outside the region. The independent measures were grouped as firm characteristics, ICT scales, and two cluster scales. The strength of a company’s membership in a business cluster was measured by summing the number of trade associations to which the firm belonged.

The firm characteristics consisted of the number of employees, the revenue received from outside the state (in percentages), and sources of financing. ICT adoption was measured using three scales (see Appendix A): The ICT infrastructure of a company comprised items such as the ease of accessing the Internet, the type of Internet connection, the presence of an internal network and website, and the number of ICT staff. The ICT reliance scale measured the extent to which the company used ICT for various operations. The ICT critical scale measured the importance the company placed on effective use of ICT and whether it had computer literate employees and up-to-date computer systems. The strength of the business cluster was derived from the number of firms in the same industry, their reputation for excellence, and the degree of cooperation among them. The benefits measured the access to business relevant information, networking with business partners and customers, and other such benefits derived from cluster membership.

RESULTS

Table 1 shows the Pearson product–moment correlations among the dependent and independent variables. Data were consistent with H1: The benefits derived from clus- ter membership are positively correlated with the strength

of the business cluster, r(331) = .24, p < .01. Data were also consistent with H2 with results showing the benefits derived from the business cluster are positively correlated with the strength of a firm’s membership in trade associa- tions, r(331) = .12, p < .05.

Hypothesis 3 predicted a positive correlation between the adoption of ICTs among rural SMEs and benefits derived from business clusters. Analysis showed that all three ICT measures were positively correlated with benefits (rICT infrastructure = .18, p < .01; rICT reliance = .19, p < .01; r ICT critical = .53, p < .01). Hypothesis 4 predicted a positive correlation between ICT adoption and business success. Data were consistent with H4. Analysis revealed that all three indicators of ICT use were significantly correlated with business success (rICT infrastructure = .28, p < .01; rICT reliance = .20, p < .01; rICT critical = .38, p < .01). Lastly, H5 predicted a positive relationship between cluster benefits and perceived business success. Results indicated a strong positive correlation, r(331) = .34, p < .01.

To examine the relationships between business success for rural SMEs and their use of ICT as well as the impact of business cluster membership, a hierarchical regression was conducted. Firm characteristics were entered in the first block, ICT characteristics in the subsequent block, and cluster characteristics in the last block (see Table 2). The results of the regression indicated the predictors ex- plained 31.2% of the variance, F(3,318) = 17.46, p < .01, f 2 = .31. ICT measures had a moderate effect size, δR2 = .081, f 2 = .088, and business cluster measures had a smaller effect size, δR2 = .042, f 2 = .044.

TABLE 1 Pearson product–moment correlations among dependent and independent variables

Firm characteristics ICT scales Cluster scales

Variable 1 2 3 4 5 6 7 8 9 10 Mean SD

1. Business success 1.000 3.20 0.74 2. Number of

employees .288∗∗ 1.000 12.94 31.33

3. Revenue outside Michigan

.351∗∗ .203∗∗ 1.000 21.11 31.91

4. Sources of financing .222∗∗ .126∗ .040 1.000 3.52 0.97 5. ICT infrastructure .277∗∗ .236∗∗ .201∗∗ .083 1.000 7.55 3.60 6. ICT reliance .197∗∗ .123∗ .239∗∗ −.008 .624∗∗ 1.000 4.88 2.24 7. ICT critical .376∗∗ .179∗∗ .183∗∗ .135∗∗ .389∗∗ .440∗∗ 1.000 3.86 0.70 8. Strength of cluster

membership .205∗∗ .300∗∗ .086 .134∗∗ .181∗∗ .135∗∗ .179∗∗ 1.000 0.24 0.43

9. Cluster strength .163∗∗ .060 −.022 .269∗∗ −.010 −.150∗∗ .070 .131∗∗ 1.000 2.80 0.70 10. Cluster benefits .343∗∗ .131∗ .056 .103∗ .178∗∗ .186∗∗ .531∗∗ .123∗ .238∗∗ 1.000 3.55 0.62

∗Correlation is significant at the 0.05 level (two-tailed). ∗∗Correlation is significant at the 0.01 level (two-tailed).

114 C. STEINFIELD ET AL.

TABLE 2 Summary of hierarchical regression analysis for variables predicting business success (n = 328)

Unstandardized coefficients

Model B SE Standardized

coefficients, beta t Significance

1 (Constant) 2.510 .144 17.418 .000 Number of employees .005 .001 .209 4.090 .000 Revenue outside Michigan .007 .001 .301 5.951 .000 Sources of financing .146 .039 .186 3.729 .000

2 (Constant) 1.431 .233 6.153 .000 Number of employees .003 .001 .159 3.190 .002 Revenue outside Michigan .006 .001 .252 5.092 .000 Sources of financing .118 .038 .150 3.130 .002 ICT infrastructure .020 .013 .099 1.596 .112 ICT reliance –.007 .021 –.021 –.334 .739 ICT critical .282 .060 .256 4.706 .000

3 (Constant) .921 .335 2.749 .006 Number of employees .003 .001 .137 2.721 .007 Revenue outside Michigan .006 .001 .261 5.420 .000 Sources of financing .102 .037 .130 2.735 .007 ICT infrastructure .018 .012 .089 1.476 .141 ICT reliance .004 .021 .011 .169 .866 ICT critical .158 .066 .143 2.395 .017 Strength of cluster membership –.054 .084 –.031 –.639 .523 Cluster strength .087 .053 .082 1.635 .103 Cluster benefits .243 .069 .192 3.505 .001

Change statistics

R Adjusted R Standard error R squared F Significance Model R squared squared of the estimate change Change df1 df2 F change

1 .456 .208 .201 .67344 .208 28.338 3 324 .000 2 .537 .289 .275 .64121 .081 12.129 3 321 .000 3 .575 .331 .312 .62481 .042 6.689 3 318 .000

To test hypothesis 6, which posited that the effect of ICT adoption on productivity will vary by industry type, multiple regressions were calculated for each of the three industries. Results are reported in Tables 3 to 5 in Appendix B. The multiple regressions do, in fact, suggest variation in the relationship between ICT use and business success across industry type. For the woodworking industry, the importance ascribed to using ICT effectively and that of cluster benefits were significant predictors of business success in this industry. For the metalworking industry, the importance ascribed to using ICT effectively was a significant predictor. For the ICT industry, both cluster strength and cluster benefits were significant predictors of business outcomes.

DISCUSSION

The analysis shows that firm characteristics were corre- lated with business success and were significant predic- tors of business success regardless of the SMEs’ indus- try. Firms that are more successful also tended to have more employees, have more business revenue outside of their locale, and have more sources of financing. All three ICT measures were initially found to be correlated to business success and in the subsequent multiple re- gression; only the importance placed on effective use of ICT was a significant predictor. This suggests that pre- dictors such as the availability of ICT infrastructure and the reliance of ICT at the workplace could be related to

SOCIAL CAPITAL AND RURAL BUSINESSES 115

business success via the importance placed on effective use of ICT. Additionally, cluster membership was found to have significant benefits to the overall success of rural SMEs.

The analysis showed the strength of cluster member- ship was correlated with business success, and benefits derived from business clusters were significant predictors of business success. This direct relation between benefits derived from membership in business clusters and busi- ness success supports the premise that rural clusters can be beneficial, corroborating findings from studies in other settings that have investigated value chain clusters that are geographically proximate (Porter 1990; 2000). Being in a business cluster increases the chances that a firm will do well economically.

We also found positive associations between ICT use and cluster benefits. In the case of rural firms that are geographically distant from their business clusters, this could mean that online connections can potentially com- pensate for the lack of physical proximity. For now, we do not think that ICTs will totally replace the role that ge- ographical proximity plays. Geographical proximity re- mains important in rural business networks because of potential benefits such as knowledge spillover, a com- mon labor pool, and value chain clusters (Scorsone 2002). What online connections can offer is a compensatory ef- fect when physical proximity is lacking. The strength of this effect would need to be investigated by further research.

Industry-Specific Analyses

The results of the analysis by industry imply that business success is contingent upon industry-specific factors. For example, in the metalworking industry, one reason ICT and cluster measures do not predict business success may be that the firms in this industry are in economic decline in Michigan. With the auto industry in Michigan facing major financial challenges in recent years, small rural firms that used to conduct business with the large automakers may be dying out.

Wood product firms may benefit from the region’s reputation for premier products, and our analysis did reveal that the benefits that wood product firms de- rived from cluster membership predicted their business success.

Finally, success of ICT firms is predicted by both cluster strength and the benefits that the company derives from the cluster. Understandably, ICT companies are more likely to be connected to networks, given the nature of their business. What the results suggest is that the degree to which ICT companies are connected to business networks could influence how successful they are.

Limitations

The generalizability of this study is limited by the use of only three sectors of the economy. The findings may vary for other industries since their products and inter- connectivity between firms would be different from those studied here. For instance, with the state government’s recent efforts to develop tourism in the region, firms in this industry could be changing their business processes and patterns of ICT use. Further, this study only surveyed firms in northern lower Michigan, and contextual differ- ences may be reflected in other rural regions. Lastly, the cluster measures are not as reliable as the others used in the study. Although reliabilities for cluster strength (α = .68) and cluster benefits (α = .62) were both acceptable for a new construct, these measures should be verified by future research.

The study is Internet-centric. We believe that broad- band was sufficiently important to investigate on its own. Nevertheless, we recognize that ICTs encompass a larger range of technologies. Therefore the implications of our study are limited to broadband.

This study found a significant relationship between ICT and cluster benefits; however, future research should ex- amine the nature of this relationship. It is possible that ICT use would enhance the benefits derived from a cluster membership but the converse is just as likely. It may be that the firms that are already clustered are more likely to use ICT to connect to one another. As such, future re- search should focus on establishing the causality in this relationship. Time-series studies or experiments to test these relationships are recommended.

CONCLUSIONS

This study contributes to existing understanding of pre- dictors of business success for rural SMEs by examining ICT use and benefits derived from membership in business clusters.

In particular, three of the findings stand out for their potential to inform policymaking.

One, cluster benefits were found to predict business suc- cess. For policymakers and community leaders seeking to develop businesses in rural communities, the received wis- dom about the benefits of membership in business clusters should be a key consideration. Persuading rural businesses to band together and creating the opportunities for cluster- ing are strategies that policymakers can fruitfully pursue.

Two, the importance that rural companies place on ICT use in their business processes was a significant predictor of business success. Business owners who have up-to- date computer systems and computer-literate employees also tended to enjoy greater business success. It should be noted that the other measures such as ICT infrastructure

116 C. STEINFIELD ET AL.

and ICT reliance did not predict business success signifi- cantly. The mixed findings with regard to the role of ICT in business success suggest that this relationship needs to be studied further. Mere infrastructure expansion is probably insufficient to improve the economic health of enterprises in the absence of information about the specific roles that ICT can play. However, what our research suggests is that the intersection of ICTs and business clusters could be a meaningful area to explore. Online tools that connect rural businesses may be able to generate the cluster benefits that were found in our study and also prior research.

Three, in their efforts to improve the economic health of rural enterprises, decision makers need to be sensitive to the industry they are working with. While ICT use did not predict business success in ICT industry, this does not mean that ICT is not important. We think that the statis- tical nonsignificance of ICT use for this industry arises from the fact that there is little variation between firms within the industry. Naturally, firms in the ICT industry would be highly reliant on ICT for their business pro- cesses. For the wood products and metalworking indus- tries, the importance that business owners place on ICT use in their business processes predicted their economic success.

REFERENCES

Barkley, D. L., Y. Kim, and M. S. Henry. 2001. Do manufacturing plants cluster across rural areas? Evidence from a probabilistic modeling approach. Clemson University, Clemson, SC: Regional Economic Development Research Laboratory.

Barua, A., C. H. Kriebel, and T. Mukhopadhyay. 1995. Information technologies and business value: An analytic and empirical investi- gation. Information Systems Research 6(1):3–23.

Bernat, G. A. 1999. Industry clusters and rural labor markets. Southern Rural Sociology 15:170–87.

Breschi, S., and F. Malerba. 2001. The geography of innovation and economic clustering: Some introductory notes. Industrial and Cor- porate Change 10:817–34.

Brynjolfsson, E., and L. Hitt. 2000. Beyond computation: Informa- tion technology, organizational transformation, and business perfor- mance. Journal of Economic Perspectives 14(4):23–48.

Dillman, D. A. 2000. Mail and Internet surveys: The tailored design method. New York, NY: John Wiley.

Gibbs, R. M., and G. A. Bernat. 1997. Rural industry clusters raise local earnings. Rural Development Perspectives 12:18–25.

Hollifield, C. A., and J. F. Donnermeyer. 2003. Creating demand: in- fluencing information technology diffusion in rural communities. Government Information Quarterly 20:135–50.

Katz, R., and S. Suter. 2009. Estimating the economic impact of the broadband stimulus plan. Available at: http://www.elinoam.com/ raulkatz/Dr Raul Katz-BB Stimulus Working Paper.pdf (accessed September 4, 2009).

LaRose, R., J. L. Gregg, S. Strover, J. Straubhaar, and S. Carpenter. 2007. Closing the rural broadband gap: Promoting adoption of the In- ternet in rural America. Telecommunication Policy 31(6–7):359–73.

LaRose, R., and A. Hoag. 1996. Organizational adoptions of the In- ternet and the Clustering of innovations. Paper presented to the International Communication Association, Chicago, IL, May.

Leamer, E. E., and M. Storper. 2001. The economic geography of the In- ternet age. Journal of International Business Studies 32(4):641–65.

Marshall, A. 1920. Principles of economics, 8th ed. London: Macmil- lan.

Parker, M. M. and R. J. Benson. 1988. Information economics: Linking business performance to information technology. Englewood Cliffs, NJ: Prentice Hall.

Pflughoeft, K. A., K. Ramamurthy, E. S. Soofi, M. Yasai-Ardekani, and F. Zahedi. 2003. Multiple conceptualizations of small business web use and benefit, Decision Sciences 34(3):467–512.

Porter, M. 1990. The competitive advantage of nations. New York, NY: Free Press.

Porter, M. E. 2000. Location, competition, and economic development: Local clusters in a global economy. Economic Development Quar- terly 14:15–34.

Porter, M. E. 2004. Competitiveness in rural U.S. regions: Learning and research agenda. Cambridge, MA: Institute for Strategy and Com- petitiveness, Harvard University. Available at: http://www.eda.gov/ ImageCache/EDAPublic/documents/pdfdocs/eda 5frural 5fregions 2epdf/v1/eda 5frural 5fregions.pdf (accessed February 12, 2008).

Pratt, A. C. 2000. New media, the new economy and new spaces. Geoforum 31:425–36.

Riemenschneider, C. K., D. A. Harrison, and P. P. Mykytn, 2003. Understanding IT adoption decisions in small business: Inte- grating current theories. Information & Management 40:269– 85.

Rosenfeld, S. A. 2001. Networks and clusters: The yin and yang of rural development. Exploring Policy Options for a new Rural America. Kansas City, MO: Federal Reserve Bank of Kansas City.

RTS (Regional Technology Strategies). 2003. Snapshots of rural inno- vation: A compendium of rural industry cluster vignettes. Minneapo- lis, MN: Author. Available at: http://www.hhh.umn.edu/img/assets/ 9140/rosenfeld1.pdf (accessed February 4, 2008).

Scorsone, E. A. 2002. Industrial clusters: Enhancing rural economies through business linkages. Mississippi State, Mississippi, South- ern Rural Development Center. Available at: http://srdc.msstate.edu/ publications/scorsone.pdf (accessed February 13, 2008).

Steinfield, C., and A. Scupola. 2008. Understanding the role of ICT networks in a biotechnology cluster: An exploratory study of the Medicon Valley. The Information Society 24(5):319– 33.

Steinfield, C., A. Scupola, and C. López-Nicolás. 2010. Social capital, ICT use and company performance: Findings from the Medicon Val- ley Biotech Cluster. Technological Forecasting and Social Change 77(7):1156–66.

Strassman, P. 1985. Information payoff: The transformation of work in the electronic age. New York, NY: Free Press.

Windrum, P., and P. de Berranger. 2002. The adoption of e-business technology by SMEs. Maastricht, Netherlands: Maastricht Eco- nomic Research Institute on Innovation and Technology. Available at: http://edocs.ub.unimaas.nl/loader/file.asp?id = 643 (accessed February 4, 2008).

Zhu, K. 2004. The complementarity of information technology infras- tructure and e-commerce capability: A resource-based assessment of their business value. Journal of Management Information Systems 21(1):167–202.

SOCIAL CAPITAL AND RURAL BUSINESSES 117

APPENDIX A: SURVEY ITEMS AND RELIABILITIES

Index Questionnaire items Scoring Alpha

Business success (6-item scale) Over the past two years we . . . 5 = strongly agree .80 Increased variety of our products; improved our sales

outside of the region; added new positions to our payroll; increased profitability; improved our relationships with customers; increased our visibility outside the region.

ICTs scales ICT infrastructure (5-item scale) Your ICT infrastructure. .. 5 = very easy

Access to broadband Internet service in northern lower Michigan; type of connection; internal network; company website; number of employees whose primary responsibility is the management of ICT.

None/Dial-up/ Broadband

.72

Yes/ No ICT reliance (9-item scale) The extent to which our company relies on . . . 5 = a great deal .88

Email; leads generated through your website; online payment system; online research; online recruiting for job openings; online training; group collaboration software; making long-distance calls over the Internet; percentage of employees that use computers as part of their daily duties.

ICT critical (3-item scale) Our success as a company depends on more effective use of information technology; having more computer literate employees; we have up to date computer systems.

5 = strongly agree .71

Cluster scales Cluster strength (4-item scale) Our industry has a high profile in the region; our regional

industry has a national reputation for excellence; there is a lot of cooperation within the region among the firms in our industry; we benefit from having a critical mass of firms in our region.

5 = strongly agree .68

Cluster benefits (4-item scale) Our success as a company depends on: access to business-relevant information; maintaining good relations with partners outside the region; exchanging knowledge with our business partners; exchanging information with our customers.

5 = strongly agree .62

118 C. STEINFIELD ET AL.

APPENDIX B Table 3. Regression coefficients of firm characteristics, ICT measures, and cluster measures on success of the wood

products industry, with dependent variable “success” (n = 114) Unstandardized

coefficients

Model B SE Standardized

coefficients, beta t Significance

1 (Constant) 2.406 .206 11.694 .000 Number of employees .005 .003 .160 1.810 .073 Revenue outside Michigan .007 .002 .300 3.476 .001 Sources of financing .134 .058 .204 2.308 .023

2 (Constant) 1.453 .328 4.429 .000 Number of employees .003 .003 .094 1.060 .292 Revenue outside Michigan .006 .002 .256 2.911 .004 Sources of financing .124 .055 .188 2.231 .028 ICT infrastructure .020 .026 .080 .765 .446 ICT reliance −.036 .042 −.091 −.853 .395 ICT critical .284 .080 .316 3.560 .001

3 (Constant) .913 .537 1.701 .092 Number of employees .001 .003 .042 .439 .662 Revenue outside Michigan .006 .002 .263 2.990 .003 Sources of financing .132 .056 .200 2.361 .020 ICT infrastructure .024 .026 .095 .906 .367 ICT reliance −.028 .043 −.073 −.664 .508 ICT critical .174 .097 .193 1.787 .077 Strength of cluster membership −.008 .139 −.005 −.055 .956 Cluster strength .063 .095 .058 .662 .509 Cluster benefits .205 .104 .204 1.965 .052

Change statistics

R Adjusted R Std. error R squared F Significant Model R squared squared of the estimate change Change df1 df2 F change

1 .438 .192 .170 .60826 .192 8.695 3 110 .000 2 .543 .295 .255 .57602 .103 5.220 3 107 .002 3 .572 .327 .269 .57076 .032 1.660 3 104 .180

SOCIAL CAPITAL AND RURAL BUSINESSES 119

Table 4. Regression coefficients of firm characteristics, ICT measures, and cluster measures on business success for the metalworking industry, with dependent variable “success” (n = 141)

Unstandardized coefficients

Model B SE Standardized

coefficients, beta t Significance

1 (Constant) 2.687 .283 9.510 .000 Number of employees .004 .001 .227 2.821 .005 Revenue outside Michigan .006 .002 .271 3.384 .001 Sources of financing .137 .073 .146 1.882 .062

2 (Constant) 1.552 .417 3.717 .000 Number of employees .002 .001 .140 1.751 .082 Revenue outside Michigan .004 .002 .188 2.288 .024 Sources of financing .045 .074 .048 .611 .542 ICT infrastructure .036 .026 .142 1.417 .159 ICT reliance .003 .036 .007 .073 .942 ICT critical .324 .115 .256 2.824 .005

3 (Constant) 1.279 .582 2.197 .030 Number of employees .002 .001 .143 1.741 .084 Revenue outside Michigan .005 .002 .214 2.540 .012 Sources of financing .055 .075 .059 .738 .462 ICT infrastructure .031 .027 .121 1.161 .248 ICT reliance −.005 .037 −.013 −.143 .887 ICT critical .232 .130 .183 1.781 .077 Strength of cluster membership −.044 .141 −.026 −.311 .757 Cluster strength −.015 .086 −.014 −.173 .863 Cluster benefits .213 .140 .144 1.523 .130

R Adjusted R Std. error R squared F Significant Model R squared squared of the estimate change Change df1 df2 F change

1 .422 .178 .160 .71283 .178 9.909 3 137 .000 2 .525 .275 .243 .67693 .097 5.973 3 134 .001 3 .537 .288 .240 .67835 .013 .813 3 131 .489

120 C. STEINFIELD ET AL.

Table 5. Regression coefficients of firm characteristics, ICT measures, and cluster measures on success for the ICT industry, with dependent variable “success” (n = 73)

Unstandardized coefficients

Model B SE Standardized

coefficients, beta t Significance

1 (Constant) 2.517 .260 9.693 .000 Number of employees .026 .008 .331 3.121 .003 Revenue outside Michigan .007 .003 .282 2.660 .010 Sources of financing .121 .073 .175 1.658 .102

2 (Constant) 1.100 .666 1.650 .104 Number of employees .023 .009 .294 2.458 .017 Revenue outside Michigan .006 .003 .238 2.227 .029 Sources of financing .122 .074 .178 1.652 .103 ICT infrastructure −.002 .020 −.010 −.078 .938 ICT reliance .059 .059 .129 .986 .328 ICT critical .242 .155 .179 1.569 .121

3 (Constant) .798 .792 1.007 .318 Number of employees .021 .009 .269 2.420 .018 Revenue outside Michigan .005 .002 .195 1.952 .055 Sources of financing .032 .073 .046 .440 .661 ICT infrastructure −.009 .019 −.056 −.483 .631 ICT reliance .056 .055 .123 1.009 .317 ICT critical .131 .149 .097 .883 .380 Strength of cluster membership −.187 .190 −.098 −.980 .331 Cluster strength .268 .107 .272 2.510 .015 Cluster benefits .244 .132 .198 1.844 .070

R Adjusted R Std. error R squared F Significant Model R squared squared of the estimate change Change df1 df2 F change

1 .495 .245 .212 .65069 .245 7.470 3 69 .000 2 .551 .304 .241 .63883 .059 1.862 3 66 .145 3 .661 .437 .357 .58788 .133 4.978 3 63 .004

Copyright of Information Society is the property of Routledge and its content may not be copied or emailed to

multiple sites or posted to a listserv without the copyright holder's express written permission. However, users

may print, download, or email articles for individual use.