Technological Forecasting & Social Change 76 (2009) 608–619

Contents lists available at ScienceDirect

Technological Forecasting & Social Change

Spontaneous vs. policy-driven: The origin and evolution of thebiotechnology cluster☆

Yu-Shan Su a,⁎, Ling-Chun Hung b,1

a Chang Jung Christian University, Department of International Business, Tainan, Taiwanb Shih-Hsin University, Department of Public Policy and Management, Taipei, Taiwan

a r t i c l e i n f o

☆ The two authors contribute equally. We thank FreFulbright scholar at University of Texas at Dallas, whManagement (Anaheim, United States, 2008), Assoc(Cambridge University, United Kingdom, 2008). This rexpressed are those of the authors and do not represe⁎ Corresponding author. Tel.: +886 6 2785156; fax:

E-mail addresses: belle@mail.cjcu.edu.tw (Y.-S. Su)1 Tel.: +886 2 2236822x63465; fax: +886 2 22363

0040-1625/$ – see front matter © 2008 Elsevier Inc.doi:10.1016/j.techfore.2008.08.008

a b s t r a c t

Article history:Received 19 December 2007Revised 7 August 2008Accepted 20 August 2008

The biotechnology industry is at the heart of the fast-growing knowledge-based economy. Oneof the distinguishing characteristics of this industry is clustering. A cluster, like an organism,experiences origin, growth, and decline/reorientation. Our study constructs a framework toanalyze biotechnology clusters with different origins, “spontaneous” and “policy-driven”,through their life cycles. We use the Bay Area in the United States and Shanghai Zhangjiang Hi-Tech Park in China as two cases to represent spontaneous and policy-driven biotechnologyclusters. This study fills the gap in the literature by comparing these two types of biotechnologyclusters in an evolutionary perspective. The key success factors of both biotechnology clustersare their own human and financial capital, but they differ in their underlying processes forcreating and sharing these resources. The most fundamental differences arise from the impactof entrepreneurship, social capital and network patterns on the cluster's configuration.

© 2008 Elsevier Inc. All rights reserved.

Keywords:Biotechnology clusterSpontaneousPolicy-drivenOriginEvolution

1. Introduction

The biotechnology (biotech) industry is at the heart of the fast-growing knowledge-based economy, and it has become a focalpoint of many local, regional, and state economic development strategies. A distinguishing feature of the industry is clustering, forinstance in the Bay Area in the United States, Cambridge in the United Kingdom, Heidelberg in Germany, and Shanghai in China.

Clusters are a geographically proximate group of interconnected companies and associated institutions in a particular field,including product manufacturers, service providers, suppliers, universities, and trade associations [1]. A cluster, like an organism,experiences origin, growth, and decline/reorientation. Our study constructs a framework to analyze biotech clusters with differentorigins, “spontaneous” and “policy-driven”, through their life cycles. We address two research questions: What are the successfactors affecting the formation of each type of cluster? Which success factors shape the configuration of a cluster?

We conduct an in-depth longitudinal case analysis of two clusters, namely the Bay Area in the United States and ShanghaiZhangjiang Hi-Tech (ZJHT) Park in China, representing respectively spontaneous and policy-driven clusters. This study fills the gapin the literature by providing a side-by-side comparison of two types of biotech clusters in an evolutionary perspective.

Drawing on two case studies, we find that the success factors in both biotech clusters include their own human and financialcapital. Yet, they vary in their processes of creating and sharing these resources. The most fundamental differences arise from the

d Phillips and Klaus Meyer for helpful comments. This work was initiated at the time Dr. Yu-Shan Su was aile Dr. Ling-Chun Hung was a PhD student there in 2007. This paper was presented at the Academy oiation for Chinese Management Educators (Toronto, Canada, 2008) and Chinese Economic Associationesearch is supported in part by the Fulbright Association and the Taiwan Ministry of Education. All viewsnt those of the sponsoring organizations.+886 6 2785662., lchung@cc.shu.edu.tw (L.-C. Hung).325.

All rights reserved.

f

609Y.-S. Su, L.-C. Hung / Technological Forecasting & Social Change 76 (2009) 608–619

impact of entrepreneurship, social capital and network patterns on the cluster's configuration. The understanding of dynamics ofdifferent types of clusters is expected to benefit both policy makers and academic researchers.

Phillips and Su [2] questioned whether our understanding of evolutionary mechanisms is sufficiently deep and broad tosupport analogies to socio-technological change. Evolution is incorporated in Darwin's concept of natural selection, competitiveinteractions between species, and survival of the fittest. Clusters evolve aswell as interact with other clusters andwith the political,entrepreneurial, and other social environments. The present paper is aimed toward an evolutionary perspective on change indifferent types of biotech clusters.

This perspective will be fully realized when we are able to model the exchange of strategies, procedures and best-practices(“geneticmaterial”) that demonstrably occurs among cluster initiatives. It stands to reason that such exchange occurs differentiallyat different stages of each cluster's life cycle, and differentially according to whether the cluster is spontaneous or policy-driven.This paper's characterization of the two types of cluster life cycle is therefore an essential precursor to a full understanding of the“genetics” and “evolution” of industry clusters, and points out constructive directions for future research.

This paper includes five sections. Section 2 presents the theoretical background, followed by the analytical framework inSection 3. Section 4 applies the framework to analyze two cases of biotech clusters — the Bay Area in the United States and ZJHTPark in China. Section 5 concludes by discussing research findings, contributions, limitations and future direction.

2. Theoretical background

A cluster is defined as a geographical concentration of different actors such as interconnected companies, specialized suppliers,service providers, institutions, which compete and cooperate in the same industry [3]. The argument that clustering benefits aregion's economy can be traced back to Marshall's [4] external economies. Following Marshall, Arrow [5] analyzed the nature ofinvention and technological advance, andRomer [6] furtherdeveloped specialization. Together, their studies led to the developmentof a new theory regarding innovation and the growth of clusters, known as the “Marshall–Arrow–Romer” (M–A–R) theory.

Clusters have been a popular research area for economists and geographers for decades. In 1990, Harvard Business Schoolprofessor Michael Porter examined industrial clusters from the perspective of business strategy and discussed the national andregional competitiveness. In consequence, cluster economies became a target for public policy [7]. Porter [8] argued that the role oflocations has been long overlooked in the age of open global markets, pointing out that “the enduring competitive advantages in aglobal economy lie increasingly in local things — knowledge, relationships, motivation — that distant rivals cannot match”.

A strand of studies has recognized the importance of the dynamics of clusters [3,8,9], while many earlier studies concentrated onthe participants and advantages of a cluster. Yet, there remains a lack of consensus over how clusters are started and to what extenttheir emergence can be set in motion by conscious design or policy interventions [10]. So, why does a cluster appear in a particularlocation? How does it start? How does it grow?What are the important ingredients in its growth? In this study, we construct the lifecycle of an industrial cluster as shown in Fig. 1. We map this life cycle from the origin, growth, to decline/reorientation. Althoughclusters decline or reorient as conditions change, this study only focuses on the stages of origin and growth. The Bay Area is a well-established cluster and China ZJHT Park is an emerging cluster. Both cases are in the growth stage and continue to evolve and expand.

Fig. 1. Life cycle of an industrial cluster.

610 Y.-S. Su, L.-C. Hung / Technological Forecasting & Social Change 76 (2009) 608–619

2.1. Origin

Clusters differ in their own origins. For example, Boston-based companies relied heavily on public research organizations, whilethe Bay Area biotech firms relied heavily on venture capital in [11]. According to Porter [8], the birth of a cluster might be rooted inhistorical circumstances, prior existence of supplier industries, or even chance. Clusters are created by a confluence of events:opportunity, existence of raw materials (including ideas, skilled human capital, etc.), emergence of an anchored firm, or someunexpected events, such as downsizing of the public sector inspiring local entrepreneurs [12].

Some other studies also provide evidence that serendipitous events or entrepreneurs are the forces to create clusters (e.g.[13,14]). Emphasizing on the role of individuals,2 Phillips [15] argues that outstanding individuals — change agents, visionaries,godfathers/godmothers, opinion leaders, connectors, mavens — shape attitudes and events, communicate across the rightnetworks, and think outside the box. For example of Northwest in the United States, the Portland Northwest Education Clusteroriginated from the idea of a professor idea at Oregon Graduate Institute and an investment banker in the region [16].

Chiaroni and Chiesa [3] suggest two major forms of cluster creation in biotech industry: (1) spontaneous clusters, that are theresult of the spontaneous co-presence of key factors; and (2) policy-driven clusters, that are triggered by the strong commitmentof governmental actors whose willingness was to set the conditions for the cluster creation, either as a response to an industrialcrisis or as a deliberate decision to foster the biotech sector. In a few cases, both forms of cluster creation coexist in a hybrid process[3]. While the literature focuses on spontaneous clusters, the number of clusters created by deliberate government policies hasbeen growing in recent years.

2.2. Growth: success factors

At the outset, a self-reinforcing cycle promotes the growth of a cluster, especially when local institutions are supportive andlocal competition is vigorous [8]. Studying the biotech clusters in five European countries (Denmark, Germany, France, Italy andUnited Kingdom), Chiesa and Chiaroni [3] describe the dynamics of cluster formation:

2 Phmany c

3 For

Once the process is started, a virtuous cycle often begins. The strong presence of new innovative biotechnology companiesincreases the area attractiveness, facilitating the establishment of new sites from large biotechnology or pharmaceuticalcompanies. The academic origin of some companies, moreover, facilitates the establishment of strong links and networksbetween industry and science. These two effects, in turn, reinforce the industrial and the scientific base of the area andtherefore provide the basis for the generation of new ventures and so on [3, p.167].

Five success factors in the evolutionary process of a biotech cluster are discussed in this study. The first success factor is strongscience and industry base, including strong networks between industry and science [3]. The cluster creation is driven by thecontinuous generation of new science-based companies, so-called Dedicated Biotech Firms, DBFs. A strong scientific and industrybase facilitates the growth of both academic and industrial spin-offs.

The second factor is finance supporting mechanisms, which are important for cluster formation because product developmentin biotech is typically long, risky and costly. Thus, funding availability is imperative for new biotech companies, including inparticular pre-seed capital, seed capital, venture capital and government funds [3].

The third success factor is entrepreneurship. Entrepreneurs, active agents who organize resources, are a critical element in theformation and viability of innovative industries and clusters [11,12]. Maskell and Malmberg [17] also propose that the key of clustersuccess is promotedmostly by local entrepreneurs, inwhich also facilitate the quality of themutual dependence amongorganizations.

The fourth factor is social capital, defined as the ability to secure resources by virtue of membership in social networks or largersocial structures [18]. This is becoming a key determinant of national and regional success in the global struggle for economicpredominance [19]. Phillips [15] argues that social capital is essential for a region to build trust among civic organizations; hencehigh-tech economic development should be based on trusting alliances and partnerships [19].3

The fifth factor is networking to create links between industrial participants at the core of a successful biotech cluster. Except forthe essential role of the geography, collaboration among organizations is a key component of success growth in the biotechindustry [11]. Thus, Owen-Smith and Powell [11] identify networking as one of the key factors supporting successful clusters.

The later three factors, however, are intertwined. In fact, entrepreneurs could not succeed without an atmosphere of trust.Social capital provides a mechanism to create trust and thus to help entrepreneurs to overcome uncertainties and to securetangible commitments from skeptical resource holders [17]. Furthermore, networks among organizations are established andpromoted by entrepreneurs based on rich social capital. These three factors contributing to the growth of cluster can hardly beprovided by governmental policies.

In sum, we define five success factors in the biotech clustering growth: strong scientific and industrial base, availability of funds,entrepreneurship, social capital, and networking between scientific and industrial bases.

illips [15] uses the term technopolis instead of cluster in his book. Although the precise definitions of these two terms are slightly different, they shareommon characteristics.the detailed discussion about the roles of social capital and culture in technopolis, please see [15] p.5.

611Y.-S. Su, L.-C. Hung / Technological Forecasting & Social Change 76 (2009) 608–619

2.3. Decline/reorientation

Although most successful clusters prosper for decades, they can and do lose their competitive edge and decline due to bothexternal and internal threats [8]. External threats include technology discontinuities, a shift in buyer's need, etc., while internalthreats include internal rigidities and the loss of innovative abilities, etc. By contrast, some scholars argue that clusters do notdecline, but the firms within it will reorient to a new development [12] or pursue other sustaining activity after the cluster'seffective growth. The sustainability implies coasting, or a relaxation of vigilant scanning for crises and opportunities [19]. Asmentioned earlier, the two cases of this study — the Bay Area and ZJHT Park — are both in the growth stage; the decline/reorientation stage is thus not the focus in this study.

3. The analytical framework

The life cycle perspective of an industrial cluster illustrates that different origins and success factors shape a cluster. Asmentioned earlier, spontaneous and policy-driven clusters are twomajor forms of the cluster creation. An interesting phenomenonis that most spontaneous clusters exist in the western countries, such as the Bay Area in the United States, Cambridge in UnitedKingdom andMarseilles in France; but the policy-driven clusters appear largely in the Asian countries, such as Shanghai ZJHT Parkin China, Tokyo-Kanto in Japan, and HsinChu Science Park in Taiwan. Even in late developing Vietnam, industrial parks played a keyrole in initiating clusters and attracting foreign investment [20].

In this study, we use a case study approach to analyze these two types of clusters (spontaneous and policy-driven) in theirorigin and growth stages. We try to answer the following questions by comparing and contrasting the two cases:

(1) What are the success factors driving the formation of the two types of the clusters?(2) Which success factor(s) shape the configuration a cluster?

Case study of an individual cluster is the most common approach in qualitative studies, [10] because this approach can capturethe dynamics of a cluster (e.g. [3,21,22]). We use the Bay Area in the United States and Shanghai ZJHT Park in China as the two casesto represent spontaneous and policy-driven clusters respectively.

Drawing on literature reviews, we construct an analytical framework of cluster formation, as shown in Fig. 2. In each case, wefirst introduce its origin and describe the brief history of its growth. Next, we examine how these success factors promote thegrowth of each cluster.

4. Two cases of biotech clusters

4.1. Spontaneous cluster: Bay Area in the United States

4.1.1. Formation

4.1.1.1. Origin: the cooperation between scientists and venture capitalists. The Bay Area is a biotech cluster created primarily by theinteraction between scientists and venture capitalists. The leading research institutes — Stanford University, University ofCalifornia, Berkeley (UC Berkeley), and University of California, San Francisco (UCSF) — created the technology, tools, and

Fig. 2. Analytic framework of cluster formulation.

Table 1Selected 2005 U.S. biotech public company financial highlights (by geographic area, ($m), percent change over 2004).

Source: [24].

612 Y.-S. Su, L.-C. Hung / Technological Forecasting & Social Change 76 (2009) 608–619

intellectual climate necessary to build a new industry. Following the pioneering work of Syntex Corporation and Cetus Corporation(founded in Berkeley in 1971), a university researcher (Herb Boyer, a UCSF biochemist) and a venture capitalist (Bob Swanson),founded Genentech, the world's first biotech company in 1976. This started the world's first biotech cluster [23].

4.1.1.2. Rapid growth: academic spin-offs. The biotech industry in the Bay Area experienced a tremendous growth in its firstdecade because of the efforts of scientists and the financial investments by venture capitalists. The decision of Supreme Court thatlife forms could be patented in 1980 also speeded this growth. Most of the biotech companies in the area were academic spin-offs.Over thirty years, the biotech industry in the Bay Area has grown to over 800 companies, employing 85,000 people and generatingover $2 billion in exports annually.4 Table 1 highlights the financial status of biotech companies in 2005 by geographic area in theUnited States. It shows that the Bay Area retained the leading place in this industry by several indicators: the number of publiccompanies, market capital, revenue, and R&D investment.

4.1.2. Spontaneous success factors in the evolutionary processMany of the success factors emerged in the Bay Area simultaneously. The formula of the effective growth process of biotech

industry in the Bay Area can be described as: the government funding goes through universities and research institutions to keepthe local “innovative engines” running, while and venture capital helps putting the insights of scientific research to practical uses.The shared prosperity5 is supported by the entrepreneurial spirit and social capital that has established networks amonguniversities, spin-offs and venture capital firms.

4.1.2.1. Scientific base. The Bay Area is home to numerous research universities and institutions, anchored by the three majoruniversities: Stanford University; UC Berkeley and UCSF. These world leading universities created strong scientific base in the BayArea. In 1973, Herb Boyer and Stanley Cohen, two Stanford geneticists collaborated to perfect the “recombinant DNA technology”.This provided a basis for much of the scientific progress that biotech has made in cloning cells and drug production [26], and hasdriven the development of modern biotech and the legendary company, Genentech. Table 2 shows the number of biotech-relatedpatents in the nine biotech clusters in the United States. Because of the strong science base, many scientists founded biotech start-ups on the basis of their own research results. Fig. 3 shows that more than 170 academic spin-offs were created from localuniversities or research institutions in the Bay Area.

4.1.2.2. Public and private financial funding. The Bay Area enjoys tremendous financial support from both public and privatesources. On the federal government level, the National Institute of Health and the National Science Foundation (NSF) are mainfinancial supporters for basic research in biotech. The local leading universities receive immense amounts of R&D funding fromthese two government departments. For instance, according to the NSF, the top universities located in the North California received

4 See: http://www.baybio.org/wt/home/Industry_Statistics.5 Although this term was explained by many persons with different meanings, we borrowed the term from Phillips [25]. He argued that “shared prosperity

involves the building of industry clusters and social capital, and action-oriented communities that can embrace change” [25].

Table 2US biotech-related patents (by geographic area).

Biotechnology centers 1975–79 1980–89 1990–99 1975–99

Boston 126 592 3007 3725San Francisco 414 1173 3991 5578San Diego 23 210 1632 1865Raleigh–Durham 27 204 796 1027Seattle 9 93 770 872New York 1420 3590 6800 11,810Philadelphia 679 1309 3214 5202Los Angeles 106 330 1399 1835Washington 121 470 2162 2753

Source: [27].

613Y.-S. Su, L.-C. Hung / Technological Forecasting & Social Change 76 (2009) 608–619

a record $2.8 billion in R&D funding in 2007.6 This generous research funding accelerated these universities' research progress, andindirectly facilitated technology transfer and the creation of academic spin-offs.

Venture capital plays a key role in the transition from innovation to commercialization in start-up firms. Biotech productdevelopment is very lengthy and costly. It takes on average 10 to 12 years for a biotech drug to pass from the initial research stage tothe commercialization stage. The cost of this process is estimated to be more than US$500 million [28]. Table 3 shows that the BayArea received the largest amount of venture capital investments in the biopharmaceutical industry from 1995 to 2001. Moreover,almost half of the highly active venture capital firms are located in the Bay Area. The large number of locally active venturecapitalists fuels biotech start-ups in this region. The Bay Area is now home to 34% of active US venture capital firms, a regionalconcentration that has existed since the 1980s.

4.1.2.3. Entrepreneurship. Entrepreneurship has been a driving force in the Bay Area for a long time. The Bay Area EconomicForum, a public–private partnership to support the economic vitality and competitiveness of the region, narrates on its website:7

“The success of the Bay Area economy is built on an unparalleled culture of entrepreneurship.” In addition, Ahmed Enany, the chiefexecutive officer of the Southern California Biomedical Council, identifies the entrepreneurial spirit as the key to the success ofbiotech industrial cluster in the Bay Area [29]. Stuart and Sorenson [30] explore regional differences by determinants of the localfounding rate, in which entrepreneurs assemble the resources to start new companies. They find relatively high rates of predictedIPOs and new venture creation in the Bay Area [30].

4.1.2.4. Social capital. The Bay Area shares valuable social capital which promotes innovation and competiveness. Cohen andFields [31] found that the social capital in the Silicon Valley created by a world of strangers, instead of a community of dense civicengagement (nobody knows anybody else's mother there). They argued that “Silicon Valley would be hard-pressed to present theimage of a close-knit civil society that, according to the social capital theorists, is the precondition for economic prosperity”. ButSilicon Valley owns a vastly different kind of social capital [31]. They describe it as:

6 ht7 Se

In Silicon Valley, social capital can be understood in terms of the collaborative partnerships that emerged in the regionowing to the pursuit by economic and institutional actors of objectives related specifically to innovation andcompetitiveness. It is the networks resulting from these collaborations that from the threads of social capital as it existsin Silicon Valley [31, p3].

Phillips [15] moreover notes that “In the U.S., where tech entrepreneurship has flourished, investors, entrepreneurs andemployees have learned to trust each other with eyes wide open.” On this trust base, the Bay Area develops its unique social capitalto encourage innovation and competitiveness. Thus Stuart and Sorenson's [30] conclude that the Bay Area affords moreopportunities to create new ventures than other areas because it holds rich social capital.

4.1.2.5. Tight networking within a spontaneous cluster. Without strong networks among research institutions, biotech start-ups,and venture capitalists, the Bay Area cluster could have never been as successful as it is today. Beginning with the story ofGenentech, the Bay Area has been developing strong ties between start-ups and venture capital firms. Moreover, the networkingactivities in the Bay Area are never static. Owen-Smith and Powell [11] find that direct links among Bay Area small DBFsoutweighed the links among DBFs with venture capital firms and DBFs with public research organizations by 1999.

In the initial stage, DBFs linked with venture capital firms and research institutions, because financial support and industrialincubators facilitated the commercialization of innovations. At more mature stage of the industry, the networks among DBFsbecame more intensive. The case of Bay Area thus provides evidence that tight, complex and delicate networking is developedspontaneously by the demand of the participants in the cluster. In addition, the continuous networking will evolve naturally overtime, as the description in Kauffman's [32] concept of “spontaneous sources of order”.

tp://www.baybio.org/pdf/IMPACT07_Policy.pdf.e http://www.bayeconfor.org/baefregion.html.

Fig. 3. Academic spin-offs in the Bay Area since the origin of the cluster. Source [3].

614 Y.-S. Su, L.-C. Hung / Technological Forecasting & Social Change 76 (2009) 608–619

4.2. Policy-driven cluster: Shanghai ZJHT Park in China

4.2.1. Formulation

4.2.1.1. Origin: a government planned area. ZJHT Park was established in 1992 as a national-level scientific park designed for high-technology development. However, the development of biotech industry did not start until 1996, when the agreement of NationalShanghai Biotechnology and Pharmaceutical Industry Base (NSBPIB) was singed to support and promote the development ofbiotech industry in the Park.

ZJHT Park is located in the middle part of Pudong New Area with a planned area of 25 km2, comprising Technical InnovationZone, Hi-Tech Industry Zone, Scientific Research and Education Zone, and Residential Zone. China's national science andtechnology policy targeted biotech as a key industry as early as 1986 (the National High-Tech Research & Development Plan, or the863 Plan), while the Torch program in 1988 also gave high priority to the biotech industry. In the national Ninth Five Year Plan(1995–2000), the Chinese government further stressed the importance of biotech R&D. In responding to the central governmentpolicy, in 1996, the NSBPIB established the biotech industry in ZJHT Park, and influenced key research institutes to resettle in thePark, including the Chinese Human Genome Center, the Institute of Materia Medica and the Shanghai Chinese Medical Universityand its affiliated hospital [33] (Fig. 4).

4.2.1.2. Growth: governmental planned development. In 1999, Shanghai Municipal Committee and Municipal Governmentdeclared the strategy of “Focus on Zhangjiang” and identified the IC, software and biomedicine as the leading industries of the Park,and ZJHT Park began to develop rapidly ever since.

After 10 years of development, the Park has established a framework for biomedicine and information industry product chainand innovation. As a result, ZJHT Park is emerging as one of the major biopharmaceutical parks in China. By the end of 2004, thePark housed 141 biomedicine companies, of which 57 were foreign owned. The number of employees in the biotech industryexceeded 10,000 at the time, including 253 individuals returning from abroad and 413 people holding Ph.D. degrees (see Table 4).

4.2.2. Governmental role in the evolutionary processThemain force in ZJHT Park's growth process in the biotech sector is the aggressive intervention of the park administration and

the state and municipal governments by providing human resource and financial support.

Table 3US venture capital for biopharmaceuticals by region, 1995–2001.

Venture capital investments Highly active venturecapital firms

Initial publicofferings1995–2001

Biotechnology centers Number Amount Share (%) 1995–2001 1998–2001

Boston 211 1,915,654,300 19.7 10 3San Francisco 261 3,028,917,500 31.1 21 31San Diego 169 1,505,896,000 15.4 4 10Raleigh–Durham 54 379,687,000 3.9 2 1Seattle 44 419,954,000 4.3 1 8New York 63 639,099,000 6.6 5 5Philadelphia 51 457,550,000 4.7 3 2Los Angeles 26 180,761,000 1.9 1 1Washington 20 85,150,000 0.9 0 2

Source: [27].

Fig. 4. The location of Zhangjiang Hi-Tech Park. Source [34].

615Y.-S. Su, L.-C. Hung / Technological Forecasting & Social Change 76 (2009) 608–619

4.2.2.1. Human resource policies. Unlike many clusters that benefit from the proximity to universities and research institutions,the research force in ZJHT Park has been incubated and focused by the government. There were neither research institutionsnearby nor a good environment for industry development in the first place. However, more than thirty research institutions haveteamed up with about ten foreign-owned R&D centers (e.g. Dupont, Roche etc.) and about 210 small-to-mid size local biotechcompanies form a regional innovative system, with the guidance and support of the government, led by Shanghai Institute ofMateria Medica and Shanghai University of Traditional Chinese Medicine [35]. The 8580 biotech-related researchers provide aknowledge base for ZJHT Park's future development in the biomedicine sector. Outside the Park, Shanghai Jiao Tong University andFudan University provide ZJHT Park with high-quality manpower and on-the-job training [36].

The Park also aims to stimulate and benefit from the entrepreneurial spirit of Chinese overseas scientists. The policies plannedfor returnees include low rent, tax breaks and living needs [37]. In the end of 2004, there were 253 returnees in the Park. Partly aresult of these policies, the trend for Chinese overseas scientists to return and found a high-tech company has been accelerating.The number of companies founded by returnees in Shanghai has increased at the rate of one per day since 2002, reaching 3000 andamounting to a total investment of USD 660 million by the end of 2004.8 In 2003, there were 50 biotech companies founded byreturnees, among the approximately 500 newly-founded companies in ZJHT Park [38].

4.2.2.2. Government financial support. The government is the main financial supporter for biotech companies in the Park. Theventure capital system is immature and weak in China. According to Prevezer and Tang [33], venture capital companies have notemerged in China until around 1996 and the emergence process started with some local governments that set up investmentinstitutes which supplied funds to start-ups and incubators in high-tech parks. It appears that the Chinese government has tried toreplace the role of the venture capitalist in the beginning of cluster formation. In October 21, 2006, the Shanghai Pudong New AreaVenture Fund, a first policy-directed venture fund supported by the local government, was established with the amount of125 million USD. The fund aims to attract 2.5 billion USD venture capital.9 Table 5 presents the amount of government granted tostart-ups in 2004.

Some anecdotes help us to better understand the government role in providing financial funding. In interviews with 35entrepreneurs and industry experts, Sternberg and Müller [38] find that biotech entrepreneurs depend heavily on the state-runinvestors, especially the Shanghai Commission of Science and Technology, Shanghai Venture Capital and the Ministry of Scienceand Technology. In addition, the government uses tax incentives to encourage the development of biotech companies. For instance,the revenue of a R&D institution engaged in business of technology transfer, technology development, related technologicalconsultation, and technical service can exempt from business tax.

The following are success factors proposed by our framework but missing or not established in the ZJHT Park.

4.2.2.3. A missing ingredient: entrepreneurship. Although ZJHT Park aimed to attract Chinese overseas scientists partly due to theirentrepreneurship, entrepreneurship is still “the missing ingredient” in most Chinese Science and Technology Industry Parks(STIPs). After interviewing in several STIPs (including ZJHT Park), Watkins-Mathys and Foster [39] conclude that what appears tobe still missing inmany STIP-based firms are entrepreneurship andmanagement skills. In fact, entrepreneurshipwasmissing but isemerging in China when we look at the number of growing private companies like Lenovo China. This new phenomenon afterChina turns to more open economy can be learned from Peng's [40] book:

8 It w9 New

“In Mainland China, for nearly decades, there had been virtually no entrepreneurship, thanks to harsh communist policy. Inthe last two decades, as policies in China become more entrepreneur-friendly, the institutional transitions have opened thefloodgates of entrepreneurship, contributing to a booming economy. In a nutshell, it is not what is in people's “blood” thatmakes or breaks entrepreneurship; it is what institutions encourage or constrain that explains it. [40, p.200]”

as €330 million in the report. The exchange rate used is GBP: USD=2:1.s in www.people.com Oct. 22, 2006, retrieved on Sept. 22, 2007.

Table 4Statistics on biomedical manufacturer at the ZJHT Park in 2004.

Number of companies 141State owned 84Foreign owned 57

Employees 10,424Female 5095Oversee returnees 253Ph.D. 413Master 750

Number of patent applied 730Asset 875Capitalization (end of year) 1453Liabilities (end of year) 622Revenue 467R&D investment 56Profits −3Subsidies 2

Source: http://www.zjpark.com/; USD Million (RMB: USD=8:1).

616 Y.-S. Su, L.-C. Hung / Technological Forecasting & Social Change 76 (2009) 608–619

4.2.2.4. Different kind of social capital: Guanxi. Some scholars use the term “Guanxi” to describe social capital in China.10 It isimpossible to ignore Guanxi when doing business in China. In Watkins-Mathys and Foster's [39] study, all parties related to STIPsrecognize the importance of good Guanxi for developing their business activities. Generally, Guanxi refers to relationship amongspecific groups or individuals based onmutual benefit exchange. Often, this includes networking with government departments inthe business world. But too much Guanxi might create an exclusive and close network for some specific groups that do not benefitinnovation and competiveness. Watkins-Mathys and Foster's [39] study provides some evidence for this concern. They report thatinterviewees perceive the main challenge of doing business in China to develop good Guanxi with the departments of localgovernment in order to gain access to larger customers [39]. In China, it is crucial to build up social capital that facilitatesinnovation and competitiveness.

4.2.2.5. A loose networking in the policy-driven cluster. There is currently no tight networking among biotech companies in ZJHTPark. The aggressive governmental interventions attract high-quality human and financial capital to the Park, but governmentpolicies cannot create tight networking among research institutions, biotech companies and venture capital firms. In ZJHT Park,links are developing between the local government and biotech companies. Prevezer and Tang [33] found a symbiotic relationshipbetween the local government and biotech companies, which has formed the backbone for the positive feedback between localgovernment initiatives and entrepreneurial responses. In other words, biotech companies in the Park make more efforts tonetwork with the local government other thanwith other companies or venture capital firms in both initial and growth stage. Thereasons are: first, the financial support comes mainly from the government funding or from quasi-government institutions;second, a good relationship with the government helps biotech companies to communicate their needs with the government, suchas regulation change or the needs of basic infrastructure.

5. Discussion and conclusion

5.1. Research findings

We summarize our comparison between Bay Area and ZJHT Park from the origin to the growth stages in Table 6. Severalresearch findings arise from these two cases of clusters of different origins and following different evolutionary paths.

First, the success factors of both biotech clusters are human and financial capital, but the underlying forces for creating andsharing these resources are very different. Table 6 illustrates that the Chinese government has been the main provider of bothhuman capital and financial support for ZJHT Park. In contrast, academic spin-offs and venture capital are the twomajor sources forhuman and financial capital in the Bay Area.

Second, we show that the key differences between the two cases are grounded in their entrepreneurship, social capital andnetwork patterns,whichdevelop interdependently in the evolution of an industrial cluster. In the BayArea, strongentrepreneurshipand social capital promote tight networking. The dynamic networking has shaped the cluster configuration and continues to evolve.In contrast, ZJHT Park has been planned by the government. Both human and financial capital have been important in initiating thebiotech cluster in ZJHT Park. However, the still emerging entrepreneurship plus lacking of social capital result on loose networks,although Guanxi, the Chinese version social capital, might facilitate the networking among biotech companies in the long run. ZJHTis still a young biotech cluster that needs time to demonstrate its ability to adapt to the environment in its further growth.

10 Knight and Yueh [41] point out that an important aspect of Chinese society — whether in traditional China (–1949), in the period of central planning (1949–1978), or in the period of economic reform (1978–) — is the Chinese variant of social capital known as Guanxi.

Table 5Government grants to start-ups in ZJHT Park, 2004.

Industry Number ofincubated firms

Grants from the NationalTechnology Innovation Fund forsmall-mid size business

Grants from Shanghai TechnologyInnovation Fund for small-mid sizebusiness

Grants from Shanghai PudongNew Area Technology Fund

Projects Amount (1000 USD) Projects Amount (1000 USD) Projects Amount (1000 USD)

IC-Industry 291 15 1138 4 104 24 563Biomedicine 87 6 563 0 0 10 319Others 35 2 175 4 143 10 179Total 413 23 1875 8 246 44 1060

Source: Zhangjiang Hi-Tech Park Administrative Office. http://www.zjpark.com/. (RMB: USD=8: 1).

617Y.-S. Su, L.-C. Hung / Technological Forecasting & Social Change 76 (2009) 608–619

Finally, we demonstrate the different configuration of clustering and networking in the evolutionary processes. In the Bay Area,biotech companies linked to research institutions and venture capital firms in the initial stage. In themature stage, these links haveshifted towardmore tight connections between biotech companies themselves. In contrast, ZJHT Park has better links between thebiotech companies and the government from the outset, and in the early growth. In contrast, the Bay Area cluster owes its capacityto evolve and adapt to the environment to its rich culture, history and social capital for innovation and competiveness. The organic,active interaction between factors shapes a more checkered but ginger cluster. Since ZJHT Park has been planned by governmentwithin a definite boundary, it is in a relatively artificial order. Nevertheless, it is too early to make inferences about ZJHI Park'scapacity to adapt to the environment.

5.2. Research contributions

This study offers three contributions. First, we compare two types of biotech clusters, “spontaneous” and “policy-driven”, whichhas been lacking in the literature. Many previous studies compared clusters with the same origins or in a nearby geographic area.We provide essential insights into the dynamics of different types of clusters through their industrial cycle. Second, our cross-country comparative analysis sheds light on how culture, history, and social capital shape the configuration of different clusters.The two cases thus illustrate that country-specific institutional context is very valuable in shaping national innovation system [42].

Finally, the detailed exploration of the origins and growth processes of two cases provides valuable insights. Many clusterstudies have difficulty in identifying the precise origins and the critical founding events. The critical issue is how to draw policylessons on the formation of clusters when their precise origins are so difficult to ascertain [10]. Our study overcomes this challengeand thus adds to the literature. Moreover, we demonstrate the intertwined and dynamic relationship among success factors in theevolutionary process which is valuable in cluster developing.

5.3. Future research directions

As all studies, this research has limitations. The first limitation arises with the data used in this study. Statistics about US patentand venture capital in Tables 2 and 3 which are the best datawe could find, since we cannot access other recent statistics regardingthe metropolitan areas in the United States. However, the trend of statistics provides insights of the R&D and financial capacity inthe Bay Area and other clusters in the United States. The second limitation is that we focus on biotech clusters only. The biotechindustry has unique characteristics such as highly-educated scientists and large financial capital, etc. It might be not appropriate to

Table 6Comparison for two types of clusters.

Cluster type Spontaneous cluster Policy-driven cluster

Example Bay Area in the United States Zhangjiang Hi-Tech Park in China

Origin Birth with the founding of Genentech: the cooperationbetween a scientist and a venture capitalist in 1976.

Birth in a government planned area in 1996.

Growth: success factors(1) Human capital (science base) Strong scientific capacity supplied by leading universities. Policy-planned manpower: government actively

attracts research talents locally and oversea.(2) Financial capital Abundant governmental and venture capital funding. Most funding comes from the government.

(3) Entrepreneurship Splendid entrepreneurship Emerging entrepreneurship

(4) Social capital Valuable social capital for innovation and competitiveness. Different kind of social capital: Quanxi.

(5) Networking Tight networking among biotech companies, venture capitaland research institutions at first. Now the direct links amongbiotech companies became the main networking.

Loose networking among biotech companies, venturecapital and research institutions. Biotech companiesmake more efforts to network with the local government

.

618 Y.-S. Su, L.-C. Hung / Technological Forecasting & Social Change 76 (2009) 608–619

generalize the findings to industrial clusters in other industries. Finally, this study is limited by its focus on two cases. The Bay Areaand the ZJHT Park may not be the only or the best representatives for this type of cluster.

Three directions of the future research arise from this study. First, both interview and questionnaire survey can be conducted todevelop more in-depth and comprehensive view of the origin and evolution of clusters from different origins. Second, the issue ofevolution/adaptation and decline/reorientation of an industrial cluster can be explored as conditions change. Other related issuesthat may be interesting to explore in future research include: Does the function of the spontaneous cluster or the policy-drivencluster change, or they just keep what they are doing until changing conditions lead to decline stage of their life cycle? Is it easierfor the spontaneous cluster change than the policy-driven cluster? And does the government role drive the policy-driven clusterchange to respond the environmental change? Finally, another interesting issue is whether the spontaneous clusters or the policy-driven clusters are emulated and evolve by other clusters and by other area. That is, do they share their “genetic code” with otherclusters in other regions? For example, the Bay Area cluster is emulated and evolves in the San Diego and Seattle; and some clustersare planned in Shanghai, Beijing, and Shenzhen in China. Whether the clusters' networks (their connections with other regions, ortheir intra-regional company memberships) change over time can be explored further in the future.

5.4. Concluding remarks

Feldman et al. [12] argue that it is dangerous to jump to easy conclusions regarding the essential ingredients of a successful cluster.However,wewish to offer two concluding remarksbasedon this study. First, spontaneous clusters have the capacity to evolve. Kauffman[32] suggests that spontaneous sources of order provide inherent order that evolution has to work with ab initio and always. Our studyshows that the Bay Area, a spontaneous cluster, originates from academic spin-offs and has been evolving by the powers of venturecapital and social capital organized by entrepreneurs. Continuous evolution leads to the tight networking in spontaneous clusters.

Second, this study demonstrates the presence of social power in the cluster formation. Phillips [15] proposes that sociology andhistory trump pure economics in shaping the regional development. It is interesting to observe how different kinds of social capitaland the accompanying networks have contributed to the different configurations of the two clusters.

References

[1] Cluster Mapping Project, Institute for Strategy and Competitiveness, Harvard Business School.[2] F.Y. Phillips, Y.S. Su, Advances in evolution and genetics: implications for technology strategy, Technol. Forecast. Soc. Change 76 (4) (2009) (In Print).[3] V. Chiesa, D. Chiaroni, Industrial Clusters in Biotechnology—Driving Forces, Development Processes and Management Practices, Imperial College Press,

London, 2005.[4] A. Marshall, Principal/Principles? Of Economics, Reprint. Originally Published: 8th ed. London: Macmillan, 1920. Published by Prometheus books. Oxford

University Press, 1997.[5] K.J. Arrow, The economic implications of learning by doing, Rev. Econ. Stud. 29 (3) (1962) 155–173.[6] P.M. Romer, Increasing returns and long-run growth, J. Polit. Econ. 94 (5) (1986) 1002–1037.[7] S. Rosenfeld, Industry clusters: business choice, policy outcome, or branding strategy? J. New Bus. Ideas Trends 3 (2) (2005) 4–13.[8] M.E. Porter, Clusters and the new economics of competition, Harvard Bus. Rev. 76 (6) (1998) 77–90.[9] J. Tan, Growth of industry clusters and innovation: lessons from Beijing Zhongguancun Science Park, J. Bus. Venturing 21 (2006) 827–850.[10] D.A. Wolfe, M.S. Gertler, Clusters from the inside and out: local dynamics and global linkages, Urban Stud. 41 (5/6) (2004) 1071–1093.[11] J. Owen-Smith, W.W. Powell, Accounting for emergence and novelty, in: P. Braunerhjelm, M. Feldman (Eds.), Boston and Bay Area Biotechnology in Cluster

Genesis: Technology-Based Industrial Development, Harvard Business School Press, Boston, MA, 2006, pp. 61–86.[12] M.P. Feldman, J. Francis, J. Bercovitz, Creating a cluster while building a firm: entrepreneurs and the formation of industrial clusters, Reg. Stud. 39 (1) (2005)

129–141.[13] M.P. Feldman, Y. Schreuder, Initial advantage: the origins of the geographic concentration of the pharmaceutical industry in themid-Atlantic Region, Ind. Corp.

Change 5 (1996) 839–862.[14] S. Klepper, The capabilities of new firms and the evolution of the US automobile industry, Ind. Corp. Change 11 (2002) 645–666.[15] F. Phillips, The Technopolis Columns: Social Culture and High Tech Economic Development, Palgrave Macmillan, London, UK, 2006.[16] The New York Times Small Companies That Try to Bring Innovative Technology to Teaching By James Flanigan Published: June 21, 2007.[17] P. Maskell, A. Malmberg, Localized learning and industrial competitiveness, Camb. J. Econ. 23 (1999) 167–185.[18] Portes, M. Mooney, Social capital and community development, in: M.F. Guillen, R. Collins, P. England, M. Meyer (Eds.), The New Economic Sociology—

Developments in an Emerging Field, Russell Sage Foundation, New York, 2003.[19] F. Phillips, Sustainability of regional initiatives for technology entrepreneurship, EFMD, 35th EISB Conference on “Sustaining the Entrepreneurial Spirit over

Time,” Barcelona, Spain, September 12–14, 2005.[20] K.E. Meyer, H.V. Nguyen, Foreign investment strategies and sub-national institutions in emerging markets: evidence from Vietnam, J. Manag. Stud. 42 (1)

(2005) 63–93.[21] R. Pouder, C.H. St. John, Hot spots and blind spots: geographical clusters of firms and innovation, Acad. Manage. Rev. 21 (4) (1996) 1192–1225.[22] A. Saxenian, Regional Advantage: Culture and Competition in Silicon Valley and Route 128, Harvard University Press, Cambridge, MA, 1994.[23] Office of the Governor of the State of California, Bay Area Council Monitor Group: Taking Action for Tomorrow—Bay Area Life Sciences Strategic Action Plan,

2003.[24] Ernst, Young, Beyond Borders: Global Biotechnology Report, 20 anniversary edition, Ernst and Young Global Limited, 2006, p. 28.[25] F. Phillips, Toward an intellectual and theoretical foundation for ‘shared prosperity’, Syst. Pract. Action Res. 18 (6) (2005) 547–568.[26] C. Robbins-Roth, From Alchemy to IPO, Perseus Publishing, Cambridge, MA, 2001.[27] J. Cortright, H. Mayer, Signs of Life — the Growth of Biotechnology Centers in the U.S., the Brookings Institution Center on Urban and Metropolitan Policy,

Washington, DC, 2002.[28] R. Ng, Drugs-from Discovery to Approval, John Wiley& Sons, New Jersey, 2004.[29] A.M. Efendioglu, Bay Area (California, USA) and Hsinchu (Taiwan) biotech clusters: a comparative analysis, J. Asia-Pac. Bus. 6 (4) (2006) 45–60.[30] T.E. Stuart, O. Sorenson, The geography of opportunity: spatial heterogeneity in founding rates and the performance of biotechnology firms, Res. Policy 32

(2003) 229–253.[31] S.S. Cohen, G. Fields, Social Capital and Capital Gains, or Virtual Bowling in Silicon Valley, Working Paper 132, Berkeley Roundtable on the International

Economy, BRIE Working Papers Series, 1998, pp. 1–31.[32] S.A. Kauffman, Origins of Order: Self-Organization and Selection in Evolution, Oxford University Press, Oxford, UK, 1993.

619Y.-S. Su, L.-C. Hung / Technological Forecasting & Social Change 76 (2009) 608–619

[33] M. Prevezer, H. Tang, Policy-induced clusters: the genesis of biotechnology clustering on the East Coast of China, in: P. Braunerhjelm, M. Feldman (Eds.),Cluster Genesis: Technology-Based Industrial Development, Harvard Business School Press, Boston, MA, 2006, pp. 113–132.

[34] S.M. Walcott, Chinese industrial and science parks: bridging the gap, Prof. Geogr. 54 (3) (2002) 349–364.[35] G. Zeng, C. Xiao, The study of Shanghai Pudong biomedicine cluster, E-J. China Urban Stud. 1 (1) (2006) 9–17 (in Chinese).[36] H.C. Lai, J.Z. Shyu, a comparison of innovation capacity at science parks across the Taiwan Strait: the case of Zhangjiang High-Tech Park and Hsinchu Science-

based Industrial Park, Technovation 25 (2005) 805–813.[37] D. Zweig, Learning to Compete: China's Strategies to Create a 'Reverse Brain Drain,Working Paper No. 2. Center on China's Transnational Relations, Hong Kong,

China, 2005.[38] R. Sternberg, C. Müller, Entrepreneurship in Regional Innovation Systems— a case study of the biotechnology industry in Shanghai, DRUID, Tenth Anniversary

Summer Conference, Copenhagen, Denmark, June 27–29, 2005.[39] L. Watkins-Mathys, M.J. Foster, Entrepreneurship: the missing ingredient in China's STIPs? Entrep. Reg. Dev. 18 (MAY 2006) 249–274.[40] M.W. Peng, Global Strategy, Thomson South-Western, Cincinnati, OH, 2006.[41] J. Knight, L.Y. Yueh, The Role of Social Capital in the Labour Market in China, Discussion Paper Series No. 121, Department of Economics, Oxford University,

2002.[42] S. Bartholomew, National systems of biotechnology innovation: complex interdependence in the global system, J. Int. Bus. Stud. 28 (2) (1997) 241–266.

Yu-Shan Su (PhD, National Taiwan University) is Assistant Professor of International Business at Chang Jung Christian University, Tainan, Taiwan. She was aFulbright scholar at the University of Texas at Dallas in 2006–2007. Her research interests are innovation and R&Dmanagement in the biotechnology industry. Shehas published her works in the Asia Pacific Journal of Management and others.

Ling-Chun Hung (PhD, University of Texas at Dallas) is Assistant Professor of Public Policy and Management at Shih-Hsin University, Taipei, Taiwan. Her researchinterests are innovation evaluation and public policy analysis.