Journal of Business Venturing 20 (2005) 217 – 239

An analytical framework for science parks and technology

districts with an application to Singapore
Francis C.C. Koha,1, Winston T.H. Kohb,*, Feichin Ted Tschanga,2
a
School of Business, Singapore Management University, Singapore, Singapore
b
School of Economics and Social Sciences, Singapore Management University, 469 Bukit Timah Road,
02-08 Federal Building, Singapore 259756, Singapore

Abstract

This paper analyzes the question: What does it take for science parks and technology districts to
evolve and grow? We propose an analytical framework to examine the gestation, evolution, and
sustainability of science parks and related but broader regional phenomena such as technology
districts. The framework comprises three aspects of a science park’s development: growth
mechanisms, level of technological capabilities, and nature of its integration with national or global
markets. The main growth mechanisms we identify are government-led infrastructure provision,
agglomeration effects, and continual self-renewal through the creation of new businesses. We apply
this framework to analyze Singapore’s science park strategy and the recent One-North initiative.
D 2003 Elsevier Inc. All rights reserved.

JEL classification: O20; O32; O38

Keywords: Science parks; Silicon Valley; Hsinchu; Cambridge; Singapore

1. Executive summary

The current literature on science parks sits within two broad areas of study: (a) the
‘‘institutional’’ perspective focuses on whether science parks confer competitive advantages

* Corresponding author. Tel.: +65-68220853; fax: +65-6822-0833.

E-mail addresses: [email protected] (F.C.C. Koh), [email protected] (W.T.H. Koh),
[email protected] (F.T. Tschang).
1
Tel.: +65-68220719.
2
Tel.: +65-68220251.

0883-9026/$ – see front matter D 2003 Elsevier Inc. All rights reserved.
doi:10.1016/j.jbusvent.2003.12.002
218 F.C.C. Koh et al. / Journal of Business Venturing 20 (2005) 217–239

to the tenant firms, as well as positive spillover effects to firms located in its vicinity and the
regional economy; (b) the economic geography perspective considers a science park and its
surrounding region as an entity consisting of specialized firms with an evolving structure of
interfirm linkages and agglomerative effects. While both perspectives have highlighted key
challenges facing science parks, our focus in this paper is on the broader forces at work that
influence their growth paths. We present a framework that traces the development of science
parks and situate the different types of science parks in terms of their gestation and
subsequent evolution.
The three aspects that are critical to an understanding of how a science park operates
and grows over time are the following: (a) growth mechanisms—the factors and
capabilities that sustain a science park and enable it to grow over time; (b) technological
capabilities—the development and strengthening of R&D capabilities and the creation of
competitive advantages in specific technology sectors; (c) global role and market
integration—the linkages between the region and the global or national economies, the
degree of integration with regional or global markets, and the creation of the region’s niche
in the global system.
Our framework is used to situate three successful science parks/technology districts,
namely, Silicon Valley, Cambridge Science Park, and Hsinchu Science Park. Although these
three exemplars differ in their initial growth impetus, they all possess strong self-renewal
capabilities in new firm formation and continue to develop R&D competencies that propelled
their initial growth. Another common success factor is their access to talent. These factors
underline the ability to generate new sources of growth in the form of new technologies and
products for global markets. All three exemplars have matured into self-sustaining technology
‘‘districts’’ or regions and have gone beyond their boundaries to develop regional and global
linkages, as well as substantial technological capabilities. While Silicon Valley has become a
global hub for R&D with considerable success in creating world-class companies, Hsinchu
has plugged itself into Silicon Valley’s extended network, and Cambridge has acted as a
magnet for technology startups keen to take advantage of the facilities it offers.
The analytical framework is also applied to a study of Singapore’s earlier and recent
science park strategies to assess its development and to identify the challenges ahead. Like
many Asian science parks, the Singapore Science Park (SSP) was developed to provide and
upgrade the infrastructure for multinational corporations (MNCs) and local industries and to
form a focal point for R&D, particularly applied R&D to meet the needs of the manufacturing
sector. While there were substantial efforts over the years to define a global role for the SSP, it
has only achieved modest levels of success at plugging itself into the global network of high-
technology clusters. There was also little interaction between the actors within and with
others outside of the science park. The Singapore government, in a new push to create
domestic engines of growth, has committed US$8.6 billion to develop a new science district,
called the One-North Science Habitat, which will integrate the existing science park facilities
and other research centers into a mammoth science and technology district. There is a
renewed bid to attract companies to locate their R&D activities in Singapore, and to define a
role for itself as a regional R&D or R&D commercializing center, either on par with or in
connection to global technological clusters.
 F.C.C. Koh et al. / Journal of Business Venturing 20 (2005) 217–239 219

The One-North strategy represents a fresh attempt to chart a more aggressive growth path
for Singapore’s science parks. If the One-North Science Habitat succeeds in strengthening its
technological capabilities, developing its growth mechanisms, and linking itself to global
markets—in short, becoming a self-sustaining hub for R&D and innovation—then Singapore
stands a good chance to achieve greater success in defining a regional or global role for itself.

2. Introduction

Since the 1970s, many countries have established science parks and technology districts as
part of a strategy to develop new engines of growth. According to Felsenstein (1994), science
parks were generally established with two primary objectives in mind. The first objective of a
science park is to be a seedbed and an enclave for technology, and ‘‘to play an incubator role,
nurturing the development and growth of new, small, high-tech firms, facilitating the transfer
of university know-how to tenant companies, encouraging the development of faculty-based
spin-offs and stimulating the development of innovative products and processes.’’ The second
objective is to act as a catalyst for regional economic development or revitalization and to
promote economic growth.
Many Asian countries, including Singapore and Malaysia, have invested in and developed
science parks with these objectives in mind. Some governments had also hoped that the
science parks will also help to (a) raise the level of technological sophistication of local
industries, through the promotion of industrial R&D; (b) promote foreign investments,
especially in higher value-added activities; and (c) accelerate the transition from a labor-
intensive to a knowledge-intensive economy. In the late 1990s, many Asian governments
were particularly keen to invest in new science parks in an attempt to enhance economic
competitiveness and to replicate the success of Silicon Valley.
These developments lead to a number of research issues, which provide the motivation for
this paper. Specifically, we seek to develop a framework that provides a theoretical basis for
tracing the development of science parks, and to situate the different types of science parks in
terms of their gestation and subsequent evolution.3 We address the following questions in this
paper:

(a) Are there common lessons to be learnt from the development of different science parks around the
world?
(b) What does it take for a science park to evolve, grow and renew itself?
(c) Is there a framework that traces the evolutionary path of science parks?

Our objective is to develop a framework that will help situate evolving science parks and to
understand the future development of these entities.

3
According to Porter (1991), a ‘‘framework . . . encompasses many variables and seeks to capture much of the
complexity of actual situations. Frameworks identify the relevant variables and the questions which the user must
answer in order to develop conclusions tailored to a particular industry and company.’’
220 F.C.C. Koh et al. / Journal of Business Venturing 20 (2005) 217–239

While there are many different models of science parks, a science park generally
encompasses business support and technology transfer mechanisms that encourage and
support the startup, incubation, and development of innovation-led, high-growth, knowl-
edge-based businesses. Most science parks also have formal and operational links with
institutions such as universities and research organizations. Although science parks are
supposed to provide focal points for R&D and innovation, the types of R&D conducted and
the sectors they focus on vary. Some science parks and technology districts are focused on
basic research (e.g., the Cambridge Science Park), while others are focused on applied
research (e.g., the SSP). There are also science parks and technology districts that possess
strong manufacturing capabilities, either within the park itself or in its vicinity (e.g., the
Hsinchu Technology District in Taiwan). By attracting new firms to locate within or in its
vicinity, science parks can create substantial agglomerative effects for the regional economy.
The plan for the rest of the paper is as follows. We briefly review recent studies on the
development and operation of science parks in Section 3. This review serves as the backdrop
for us to introduce, in Section 4, an analytical framework that identifies the key factors
influencing a science park’s development and growth. The elements of the analytical
framework are illustrated with the aid of three well-known exemplars, namely, Silicon
Valley, Cambridge Science Park, and Hsinchu Science Park. In Section 5, we apply the
analytical framework to an analysis of the SSP,4 which is an example of an infrastructure-led
model commonly seen in Asia. Section 6 concludes the paper.

3. A survey of recent literature

The literature on science parks sits within two broad areas of study, which we refer to as (a)
the ‘‘institutional perspective’’ and (b) the ‘‘technology district,’’ originating from the
economic geography or regional science field. The institutional perspective views a science
park as an institution providing assistance to its tenants in specific policy-based or
mechanism-based ways. This view emphasizes issues such as the functioning of incubators
and degree of spin-offs, and focuses on whether science parks confer competitive advantages
to the tenant firms, as well as positive spillover effects to firms located in its vicinity and the
regional economy. By contrast, the economic geography perspective views the science park
and its surrounding region as an entity consisting of specialized firms with an evolving
structure of interfirm linkages and agglomerative effects. This perspective examines the role
and contribution of science parks in the context of regional development (see Storper and
Harrison, 1991; Markusen, 1996, for examples of this perspective).
The recent literature covers both the institutional and the geographical perspectives. The
studies cover a range of geographic localities, such as Storey and Tether (1998), who

4
Although we speak of the Singapore Science Park, there are three separate, adjoining parks, separately
managed for administrative purposes, but under a common parent company, Ascendas Pte Ltd, which is a
government-related company.
 F.C.C. Koh et al. / Journal of Business Venturing 20 (2005) 217–239 221

provided an overview of science parks in Europe; Lofsten and Lindelof (2002, 2003) on
science parks in Sweden; Athreye (2002) on the agglomeration and growth of the
Cambridge science district; Saxenian (2001a,b) on the Hsinchu Technology District in
Taiwan; Bakouros et al. (2002) on science parks in Greece; Conceicao et al. (2002) on
Italian technology parks; Kihlgren (2003) on the St. Petersburg Technology Park in Russia;
Palmai (in press) on an innovation park in Hungary; and Phillimore (1999) on the Western
Australian Technology Park. A summary of the findings of several representative studies is
provided in Appendix A.
In the institutional perspective, the tendency to view and promote science parks as
specialized physical infrastructure to house technology companies has led to a particular
focus on the direct and tangible contribution of science parks and the institutions and
mechanisms within, e.g., job creation and quality of employment, contribution to R&D
investment and output, venture capital raised, as well as the roles of universities with the
science parks. Another focus is on the challenges of enterprise formation (see, for instance,
Lofsten and Lindelof, 2003; Bakouros et al., 2002). The incubation aspect has been broadly
addressed in the studies of science parks in the United Kingdom by Westhead and Storey
(1995) and Westhead and Batstone (1998). The issues of financing and the role of universities
were examined by Vedovello (1997) and Lofsten and Lindelof (2002).

4. An analytical framework for science parks

While both the institutional and geographical perspectives have highlighted the key
challenges facing enterprise formation within science parks, our aim in this paper is to
understand the broader forces that influence the growth paths of science parks. The central
question we ask is ‘‘What does it take for a science park, as a system, to evolve and grow?’’
To this end, we identify three primary aspects that are critical to an understanding of how a
science park operates and grows over time. These are:

(a) Growth mechanisms: These are the factors and capabilities that sustain a science park and enable it
to grow over time.
(b) Technological capabilities: This aspect is concerned with the development and strengthening of
capabilities in R&D and the creation of competitive advantages in specific technology sectors.
While a region’s technological capabilities are often indirectly addressed in discussions of a
region’s industrial sectors of competence, we shall examine how different regions specialize in the
chain of technology production.
(c) Global role and market integration: This aspect addresses the linkages between the region and the
global or national economies, the degree of integration with regional or global markets, and the
creation of the region’s niche in the global system.

The factors comprising our analytical framework are concurrently developed from and
used to situate the experience of our three exemplars of Silicon Valley, Cambridge Science
Park, and Hsinchu Technology District. These were all chosen based on their well-known
global successes, against which many science parks (including Singapore’s Science Park)
222 F.C.C. Koh et al. / Journal of Business Venturing 20 (2005) 217–239

were modeled after or benchmarked against. They also demonstrate diversity in the way they
were developed. More importantly, they all demonstrate critical attributes necessary for long-
term sustainability and growth.
In Table 1, we provide a comparison of the growth factors of the three exemplars. The
growth factors are broadly classified into subgroups: (a) gestation and takeoff factors, which
provided the initial impetus for the exemplar’s establishment and development; (b) growth-
sustaining factors, which are the capabilities that enable the science park to renew and sustain
itself.
According to Florida and Kenny (1990), ‘‘(f)or many, if not for most, Americans, Silicon
Valley and Route 128 stand out as symbols of economic and technological success.’’ Silicon
Valley has become the universal ‘‘role model’’ for science parks and technology districts
around the world. Dating back to World War II with the emergence of an electronics cluster in
the region, and supported by contributions from early semiconductor pioneers and Stanford
University, Silicon Valley has established itself as a high-tech R&D and industrial region.
In Europe, the Cambridge Science Park in the United Kingdom was initiated by Trinity
College of Cambridge University in 1970.5 The Cambridge Science Park took a long time to
take off. By 1978, 2 years after its officially opening, it had only seven tenants and only 20%
of the total area devoted to the science park was developed. This was partly because Trinity
College was very selective about the tenants that it allowed into the Cambridge Science Park.
However, since then, the broader region itself has attained a reputation as a high-tech district
(Green, 2002).
Established a decade later than the Cambridge Science Park, the Hsinchu Technology
District was initiated by the Taiwanese government in 1980, with the objective of attracting
Taiwanese technologists and entrepreneurs located in the Silicon Valley to return to Taiwan.
Presently, Hsinchu houses a thriving cluster of firms employing more than 50,000 people.
While the firms have started out as small and medium sized, many have become large global
multinationals in their own right (Amsden and Chu, 2003). While the Taiwanese government
was responsible for its establishment, it has subsequently maintained a hands-off policy,
leaving the private sector to drive the development of Hsinchu.
As we noted earlier, the exemplars we have chosen strongly illustrate the importance of
each aspect of our framework. We discuss them in turn.

4.1. Growth mechanisms

The first aspect of our framework is the set of growth mechanisms that drive a science
park’s development. In the case of Silicon Valley, what distinguishes it from other technology
districts is the way it continually renews itself (Seely Brown and Duguid, 2002). Silicon

5
The University had commissioned in 1969, a report by the Nobel Laureate Professor Sir Neville Mott, on the
role of the University in the local economy. The report concluded that it would be in the interest of Britain, the city
of Cambridge, and the University to encourage the growth of existing and new science-based industry and other
applied research organizations in Cambridge.
 F.C.C. Koh et al. / Journal of Business Venturing 20 (2005) 217–239 223

Table 1
A comparison of exemplars of science parks and technology districts
Silicon Valley Cambridge Hsinchu
No. of firms (year) More than 10 000 About 1000 (2002) 272 (2000)
Sectors R&D, manufacturing Scientific R&D Manufacturing (various)
(limited)

Gestation factors
Gestation and early Initially government led, University-led effort Government-led effort to
advantage followed by strong in its creation, followed build infrastructure and
industry and university by private firms choosing institutions, which provided
linkages to locate within the park environment for spin-offs
from research institutions
Evolution Self-directed. Take-off Self-directed. Established Government-led with self-
occurred in the 1950s in 1970 direction after spin-offs
matured. Established in 1980
Constraints Overcrowding has Small size of firms. No Taiwan is still largely a
forced some smaller major industries developed manufacturing base.
firms to leave. The around Cambridge Urgent need to move
quality of life has into innovation
declined somewhat

Growth-sustaining factors
Tenants Local firms either Local firms interested Domestic small and
spun off from Stanford to locate near Cambridge medium enterprises
University, established University, some recent spun off from government
firms, or relocating from high-tech spin-offs institutes (e.g., ITRI) or
outside of the region (e.g., biotech) founded by returnees
from the USA
R&D stages Mainly known for basic Mainly known for pure Applied research related
contained within research, but also science and basic research to design, manufacturing
park or region includes pure science,
and applied research
Additional Enabling A global hub for Strong intellectual capital Connected to and supported
factors venture capital financing, base attracts continual by a global production
technology markets inflow of entrepreneurial network (emanating from
initiatives Silicon Valley)
Source: Silicon Valley: Saxenian (1994), Lee et al. (2000). Cambridge: http://www.cambridgesciencepark.co.uk,
Athreye (2002), Green (2002). Hsinchu: http://www.sipa.gov.tw, Lee and Yang (2000), Saxenian (2001a,b).
Singapore: http://www.sciencepark.com.sg, Amsden and Tschang (2003).

Valley was the epicenter of several waves of technological innovation (semiconductors,

computers, and more recently IT and e-commerce) and is often the region that first emerges
with innovative new technological sectors (Kenny, 2000). Agglomeration effects are also
considered to be significant in Silicon Valley, as the concentrations of talent and venture
capital financing continually pull in other talent and firms, thereby creating a virtuous cycle.
The success of Silicon Valley can be attributed to a number of competitive advantages that
support the creation of new firms: (a) a large pool of technical talent, (b) availability of
224 F.C.C. Koh et al. / Journal of Business Venturing 20 (2005) 217–239

preexisting infrastructure and large network of suppliers, (c) access to venture capital, (d)
access to excellent educational facilities and research institutions, and (e) well-developed
information networks (see Amirahmadi and Saff, 1993; Lee et al., 2000).
Whereas many of Silicon Valley’s firms have gone on to become global companies, many
of the tenants in the Cambridge Science Park have remained relatively small. This partly
reflects the orientation of the Cambridge Science Park towards basic scientific research.
While Trinity College limited the tenant base to technology-focused companies, many firms
that did not locate within the Cambridge Science Park still benefited from ‘‘spillover effects’’
by locating in its vicinity, with the emergence of a support ecosystem of ancillary services
(Green, 2002). The Cambridge Science Park and the surrounding region is estimated to
account for 60% of all high-tech establishments and over 70% of all high-tech employment in
the Cambridgeshire Country (Athreye, 2002).
Finally, the Hsinchu Science Park has had considerable success in the creation of
technology startups (see Hou and Gee, 1993; Saxenian, 2001a,b). Although many of the
startups were initially spun off from the government-linked Industrial Technology Research
Institute (ITRI), an abundance of venture capital and technical and managerial talent has
emerged over the years to support early-stage and later-stage entrepreneurial activities
within Hsinchu.
We can discern three types of growth mechanisms: (i) government-directed mechanisms, in
the form of infrastructure provision with perhaps a signaling motive; (ii) agglomerative effects;
(iii) new-firm creation and self-renewal. We consider these growth mechanisms in turn.

4.1.1. Government-directed mechanisms

The primary government-directed mechanisms present in many science parks involves the
funding of institutions of research, such as Hsinchu’s ITRI, Cambridge University, and
Silicon Valley’s early defense firms. On the other hand, the primary motivation of many
Asian governments is the provision of infrastructure. The development of high-quality
infrastructure was also part of a concerted strategy to attract MNCs to invest and locate
manufacturing as well as R&D operations in these Asian economies. This was the case in
Malaysia, Thailand, and Singapore, where the development of science parks was an important
part of the efforts to attract foreign investment. The presence of a science park—even one
with moderate physical infrastructure and tenuous linkages to universities and research
institutions—signals a country’s commitment to a high-tech economic growth strategy. As
foreign investments flow in and more MNCs locate in the science parks (or neighboring
districts), the advantages of the science parks are strengthened.
Unlike Asian economies, the signaling motive was not as strong a factor in the establish-
ment of European science parks, although those parks do serve to help attract investments to
the region by the presence of the existing firms. This is explained further as a process of
agglomeration next.

4.1.2. Agglomerative effects

The presence of a science park and a cluster of high-growth technology firms can stimulate
job creation and regional growth as new firms, talented labor, and investments are drawn to
 F.C.C. Koh et al. / Journal of Business Venturing 20 (2005) 217–239 225

the region. In turn, firms can have access to a greater number and variety of suppliers,
technical expertise, and potential business partners, all located within close proximity. The
agglomerative forces also include knowledge spillovers, resulting from the informal transfers
of knowledge and exchange of ideas.
Silicon Valley exemplifies this process, and provides the benchmark against which
other innovative regions are assessed (Kenny, 2000). For European science parks,
agglomerative effects were important considerations for their establishment and long-
term sustainability. In Asia, though, agglomerative effects were not often cited as the
primary reasons for the establishment of science parks, as noted in Xue (1995), for
example.

4.1.3. New firm creation and self-renewal

The single most critical factor for a science park’s long-term sustainability is its ability to
foster the creation of new firms, in both existing and emerging sectors, in order to
continually renew itself. This was the key factor in Silicon Valley’s success (Lee et al.,
2000). Many science parks have sought to enhance their capabilities in this process with the
establishment of incubators and colocation of venture capital firms to accelerate the
entrepreneurial process.
A region’s existing capability to generate new research and other knowledge is an
important factor in the new-firm creation process. In the case of Cambridge and Silicon
Valley, the presence of world-class universities and other research institutions helped
stimulate the entrepreneurial process, partly as new firms were spun off from the research
programs of these institutions, and partly as these institutions served to draw new firms
desiring some spillovers from those institutions. Other important factors include the
proximity to centers of commerce, the willingness of established businesses to try new
technology products (i.e., as early adopters and customers), and a high-quality workforce.
As a whole, these factors will draw in new firms, and ultimately strengthen the
agglomerative forces. Clearly, if a science park or technology district is unable to renew
itself through new sources of growth, decline will set in. Some researchers have commented
that this was perhaps a reason underlying the relative lack of vitality in Boston’s Route 128,
since the late 1970s, relative to Silicon Valley.6

4.2. Technological capabilities

The second aspect of our analytical framework is the level of research or technology
development capabilities within a science park or technology district. This aspect is
important for assessing the status of technological sophistication and the level that is
aspired to. The R&D framework presented in Amsden and Tschang (2003) is applied to

6
Saxenian’s (1994) earlier study contrasted the different ‘‘cultures’’ of Silicon Valley and Route 128, as well
as the latter’s dependence on government funding. Silicon Valley’s ability to find new sources of growth and
renew itself has been extensively studied. A recent study is that of Bresnahan et al. (2001).
226 F.C.C. Koh et al. / Journal of Business Venturing 20 (2005) 217–239

this purpose. In the simplest possible terms, this typology consists of five categories, as
described in Table 2. The R&D stages range from pure science and basic research
(involving fundamentally similar techniques and results, but different motivations) to
applied research (involving research oriented to more practical, product-related consider-
ations), to exploratory development, (involving prototyping) and to advanced development
(involving manufacturing considerations).
We use this typology to classify the technological capabilities of the three exemplars.
Silicon Valley has established itself as a global hub for both basic and applied research
(Saxenian, 1994), although a sizable amount of R&D is also conducted in applied research
and product development. Its phenomenal growth is oriented around specific technology
sectors and concentrations of firms, which took ideas out of laboratories and into the
commercial world. While government defense spending in the 1950s catalyzed the develop-
ment of Silicon Valley, it was the technology companies spun off from both the universities
and established companies such as Fairchild Semiconductor that propelled growth in the
1960s. Other successful waves of companies included those in computers, e.g., Silicon
Graphics, Apple Computer, and Sun Microsystems, and more recently, those oriented around

Table 2
A classification of the stages of research and development
Stage of Description of activity Entity performing it
research and output and skill requirements
Pure science Research for the pursuit Universities (scientific
of knowledge, with outputs expertise needed)
usually in scientific journals
and possibly patents
Basic research Similar outputs as pure science Corporate laboratory
approach, but conducted with (scientific expertise
specific long-term corporate needed)
objectives (such as future
products) in mind
Applied research Medium-term research Corporate laboratory
on known technologies; (less scientific, more
typically involves transforming engineering expertise
or localizing existing product needed)
knowledge, or reapplying known
research results to other areas
Exploratory development Development and prototyping Corporate product
of design and other systems development departments
(product development
expertise)
Advanced development Addressing of manufacturing Manufacturers
considerations for products (manufacturing and
product development
expertise)
Source: Adapted from Amsden and Tschang (2003).
 F.C.C. Koh et al. / Journal of Business Venturing 20 (2005) 217–239 227

networks and other new computing technologies. These companies played a role in shaping
the global production networks that emanate from Silicon Valley (Florida and Kenny, 1988).
In the case of the Cambridge Science Park, the main type of R&D appears to be the pure
science research output of Cambridge University and other nationally funded research
institutes in the region. In the 1970s and 1980s, many of the companies that originated from
the Cambridge Science Park were led by research scientists who were first-time entrepre-
neurs. The initial focus was in electronics and software, which broadened to other technology
sectors later. Increasingly, major companies with strong science and technology resources
have also based research outposts in Cambridge. As a result of all this, in the 1990s,
biotechnology, R&D consultancy, and software were the key sectors in the Cambridge
Science Park.
The specialization of the Hsinchu Science District is primarily in production-oriented
technology (Lee and Yang, 2000)—what we term exploratory development and advanced
development (see Table 2). Hsinchu’s development has benefited from the proximity to
institutions such as the National Chiao Tung University, the National Tsing Hua University,
and the Industrial Technology Research Institute, as well as the three national laboratories—
National Center for High Performance Computing, the Synchronous Radiation Research
Center, and the National Space Program Office.

4.3. Global or national role definition and integration of markets

The third and final aspect of our framework consists of the linkages that a given science
park has with national and global economies, and as a result, the role or niche that it plays.
At the national level, this may concern the ability of the economy to commercialize the
output of the science park’s research, and to offer a viable market for its products. At the
global level, this may concern the ability of a region to connect with different regions and
the global value chain. Science parks that cannot match the sophistication of the
technologically more advanced players must find and develop an alternate role, such as
the production of products and services that can serve the local or national economy (Seely
Brown and Duguid, 2002).
The three exemplars exhibit different levels of interaction with the national and global
economies. Silicon Valley is home to a large number of enterprises that have globalized their
operations while retaining their R&D core in the Silicon Valley region (Lee et al., 2000). It is
considered a global R&D leader and ‘‘the market to be in’’ for many technology sectors. As a
global hub for both emerging technologies and venture capital, it has strong linkages with
other technology districts in other countries.
By contrast, Cambridge Science Park has acted mainly as a magnet within the United
Kingdom for global and national enterprises to locate in its vicinity (e.g., Microsoft’s first
non-U.S. laboratories were established in Cambridge, UK). With Cambridge University
playing a key role in fostering university–industry linkages, the broader Cambridge science
district has provided a vibrant environment for science-based industries. Athreye (2002)
argues that Cambridge is unique among all other IT clusters that followed in the wake of
Silicon Valley’s success in that no other European region has shown the same scale of
228 F.C.C. Koh et al. / Journal of Business Venturing 20 (2005) 217–239

entrepreneurial activity as Cambridge or developed a similar set of institutions without state

intervention. However, in marked contrast to Silicon Valley, Cambridge’s global impact has
been limited. The number of large technology firms in the Cambridge area remains low, and
few have made a significant contribution at the national or global level (Athreye, 2002;
Green, 2002).
Although it lags behind in the capabilities to develop cutting-edge technology, Hsinchu is
generally regarded as a key player in the manufacturing of new semiconductor and other
electronic technologies. Firms located in Hsinchu are among the top manufacturers of
personal computer-related systems and components. Hinschu has developed strong links
with Silicon Valley, effectively becoming part of the latter’s global network (Saxenian,
2001a). These linkages extend to cross-border investment flows, key manufacturing partner-
ships between Taiwanese and U.S. firms, and the transfer of entrepreneurial talent and
management practices from Silicon Valley into Taiwan. In 1998, 109 of the 272 companies
established in the Hsinchu district were founded by Taiwanese entrepreneurs who have had
managerial, business, and entrepreneurial experience in the United States (Lee and Yang,
2000).

4.4. Summary

From the above discussion, we can discern different models of successful science park
development. While Silicon Valley became a global hub for R&D, possessing considerable
success in creating world-class companies, Hsinchu plugged itself into Silicon Valley’s
extended global network, and Cambridge has acted as a magnet for technology startups keen
to take advantage of the facilities it offers. In each of the models, the most important trait for
sustaining growth is that of an ability to attract or create new firms. A common factor that
underlies the success of the exemplars is the access to talent. All these factors underline the
ability to generate new sources of growth in the form of new technologies and products for
global markets.
Hsinchu provides an interesting contrast to Cambridge and Silicon Valley. A primary
motivation in its establishment was to provide high-quality infrastructure to support
technology startups and help develop national capabilities in high-tech manufacturing.
Mechanisms were put in place to facilitate the transfer of know-how from research
institutions to the private sector. Agglomerative effects were equally important in the
subsequent development of Hsinchu, and although the government provided the initial
stimulus, it was the private sector that drove its subsequent development (Saxenian
2001a,b).
By contrast, neither Cambridge nor Silicon Valley was preoccupied with infrastructural
concerns. Instead, the desire to commercialize new technologies and products was the
principal motivation in the early development of both Cambridge and Silicon Valley.
Their evolutionary paths, however, have been different since. Silicon Valley is a global
hub in both R&D innovation and in production. Although Cambridge has a limited role
in global production, it is also one of the global clusters of intellectual development and
research.
 F.C.C. Koh et al. / Journal of Business Venturing 20 (2005) 217–239 229

5. An application of the analytical framework to Singapore

We turn our focus next to an analysis of the SSP. The framework we have developed is
used to evaluate the SSP, and to examine whether the SSP exhibits the hallmark character-
istics of a global self-sustaining science park.
The SSP has been a core component of Singapore’s development strategy for the past two
decades. As an island economy with little natural resources, an urgent task in the early years
of Singapore’s independence was to attract MNCs to invest in and set up manufacturing
operations in the country. While the strategy had been successful in accelerating economic
growth in the 1960s and 1970s, it also led to Singaporean companies playing a largely
supporting role to the MNCs. As the cost of doing business escalated and other developing
economies offered cheaper locations for manufacturing, this prompted the government to
undertake a review of the competitiveness of the Singapore economy. This led to the
development of the SSP in 1980.

5.1. The Singapore Science Park strategy: 1980–2000

The origin and development of the SSP7 must be viewed against the set of coordinated
government policies on science and technology policy (including research and human capital
formation), IT infrastructure, and more recently, promotion of entrepreneurship (see Ministry
of Trade and Industry, Singapore, 1986, 1991). The SSP has been an integral part of the
technology policy that underpins Singapore’s economic growth strategy. Like many Asian
science parks, one of the initial motivations of the SSP was to provide and upgrade local
infrastructure to house MNCs as well as new industries that require proximity to institutions
of higher learning. Additionally, the SSP was established to provide a focal point for R&D in
Singapore, with an emphasis on industrial R&D. A secondary objective of the SSP was to
signal Singapore’s readiness to develop high-tech industries.
The provision of infrastructure went beyond just physical facilities, and included the
creation—with government encouragement in the form of tax breaks and other incentives—of
a supporting infrastructure for the MNCs. This supporting infrastructure consisted of
domestic suppliers, service providers, and potential business partners. The multiplier effects
of these investments and the development of the supporting infrastructure had significant
stimulative effects on the domestic economy.
Singapore’s science park strategy has until recently been driven largely by the government.
Private sector participation was limited, unlike in the case of Silicon Valley or Hsinchu. While
there were substantial efforts over the years to define a global role for the SSP, the strategy
has only managed to achieve modest levels of success at plugging Singapore into the global
network of high-technology clusters. There was also little interaction between the actors
within the SSP and between them and other regions, although it has improved in recent years.

7
Information on the SSP can be found at http://www.sciencepark.com.sg. A recent paper that evaluates the
SSP from a geography perspective is Phillips and Yeung (2003).
230 F.C.C. Koh et al. / Journal of Business Venturing 20 (2005) 217–239

5.2. The Singapore Science Park strategy: into the future

To extend and complement the SSP, the government is building a new One-North Science
Habitat, a 200-hectare development that will integrate the existing science park facilities and
other research centers into a mammoth science and technology district.8 Conceptualized in
2000, the project is estimated to cost S$15 billion, or about US$8.6 billion, over 15 years. It is
envisaged to create the ambience of a multifaceted research community, with schools, public
transport, and other amenities. It will provide a wider focal point for R&D and entrepreneurial
activities in the biosciences and information technology sectors. In the planning of the
project, detailed studies were made on a number of science parks and incubators, including
Hsinchu, Cambridge, Silicon Valley, Sophia Antipolis, and the MIT Medialab.
There are several areas that distinguishes the One-North master plan from the original SSP.
Firstly, it aims to provide infrastructure that offers seamless connectivity at both the
individual and business level. In contrast, the sprawling nature of the existing Science Parks
I and II, modeled after the low-density environment in Silicon Valley, does not facilitate close
interaction. Secondly, ‘‘dynamic planning’’ will be emphasized in One-North to encourage
the vertical and horizontal integration of tenants and their different uses of the space. The aim
is to enable different companies to work closely together and facilitate cross-fertilization of
research ideas. Greater private sector participation will be encouraged. For instance, although
the biomedical sciences is a key sector in the first phase of One-North’s development—as
shown by the large investment in the Biopolis, a new biomedical research facility9—the
current thinking is to create self-evolving industrial structures that will allow One-North to
catch the next wave of emerging technologies.
The One-North strategy represents an attempt to address some past deficiencies. While the
development of state-of-the-art infrastructure remains a key attraction for overseas companies
to locate in One-North, there are also renewed efforts to project Singapore as a regional centre
of R&D. The Singapore government is encouraging companies from Australia, New Zealand,
and elsewhere to locate their R&D activities in Singapore, and position the island to be a
gateway to the markets in China, India, Southeast Asia, and Indochina.
While the One-North project represents a new approach in Singapore’s science park
strategy, a key constraint will be the limited pool of technical talent available in Singapore.
The Singapore government has recognized this limitation, and has introduced policies aimed
at strengthening its technological capabilities and expanding the pool of expertise.10 For

8
Based on interviews with officials involved in the One-North development project, we understand that the
planning and conceptualization of the One-North project took more than a year. Details of the One-North project
can be found at http://www.one-north.com.
9
This facility is being developed a short distance from the SSP and the National University Hospital. The
establishment of Biopolis is part of the government’s current plan to develop Singapore into a biomedical hub and
to create new engines of growth.
10
With the benefit of the Hsinchu experience, another objective of the One-North plan is to attract overseas
Singaporeans to return home to contribute their talents and entrepreneurial skills, just like in the case of the
returnee Taiwanese entrepreneurs from Silicon Valley.
 F.C.C. Koh et al. / Journal of Business Venturing 20 (2005) 217–239 231

instance, more postgraduate scholarships are being awarded for PhD studies in the sciences
and there is a renewed emphasis on basic research (as opposed to the former emphasis on
applied research) in the government-funded research institutes and in the local universities. In
the short term, the government is making concerted efforts to attract foreign talent to
Singapore to supplement the local talent pool (see Koh and Koh, 2002).

5.3. Analyzing Singapore’s science park strategy

The evolution of Singapore’s science park strategy can be analyzed with our framework.
We discuss, in turn, the three aspects of growth mechanisms, technological capabilities, and
global role definition.

5.3.1. Growth mechanisms

The original SSP was conceived as part of a national policy to promote R&D and to
signal the country’s readiness to promote high-tech industries. As such, the desired
agglomerative effects were less emphasized in the early stages of its development.
Although the MNCs located in Science Parks I and II conducted more R&D than
Singaporean organizations, the impact on knowledge creation and spillover was limited
(Hu and Shin, 2002). Due to Singapore’s small size, the agglomerative effects could be due
as much to national-level advantages than to the SSP. In fact, it can be argued that the
whole of Singapore should be considered an ‘‘industrial district,’’ since travel time is only
two hours from one end of the island to the other, and often less than an hour between key
points of business and industry.
As discussed earlier, the ability of a science park to foster the creation of new firms and
renew itself is intimately linked to competitive advantages such as the availability of technical
talent and R&D capabilities. In Singapore’s case, these competitive advantages were
relatively weaker than larger, more advanced countries. Although there were attempts in
the 1990s to foster entrepreneurial capability by locating the Kent Ridge Digital Laboratory (a
large national IT Laboratory) and two venture capital firms in the SSP, this did not translate
into successful high-tech spin-offs for the most part.
In many science parks, knowledge spillovers are sometimes more important than physical
infrastructure as a source of attraction for potential tenants. Thus, providing better quality
infrastructure alone is unlikely to attract firms to new science parks such as One-North, at a
time when every country in Asia has a surfeit of infrastructure. Already, Malaysia has
announced that it will be embarking on a project similar to Biopolis, and is aiming to attract
companies in a similar manner to that of Singapore’s Bipolis.
Recognizing the challenges ahead and to differentiate itself from other competing new
science parks, One-North intends to attract new firms by creating an environment that
facilitates spillovers and agglomerative effects through the development of a dense
technological metropolis. Specifically, it intends to create an environment that will allow
organizations in different sectors and technological activities to combine their different
capabilities into new ones, in the process transforming themselves, as required, to adapt
to new situations.
232 F.C.C. Koh et al. / Journal of Business Venturing 20 (2005) 217–239

5.3.2. Technological capabilities

One of the biggest difficulties facing Singapore is the expansion and deepening of its
technological capabilities. While Singapore’s government-funded research institutions and
local universities managed to spin off a number of companies in the late 1990s, the
combined impact of these efforts within the SSP had been relatively modest. Access to
larger markets and the ability to sustain a critical mass of entrepreneurs are frequently cited
challenges.
Most of these technology spin-offs observed in and around the SSP occurred in 1999
and 2000, just before the burst of the Internet bubble. Many of these startups were
launched after reaching the applied research or development stages, i.e., after a relatively
short 2- to 3-year period of R&D. Consequently, there was little innovative technology
that was embodied in their products or services. Many of these enterprises also had
difficulties scaling up their operations to serve a bigger market before other imitators
enter the market and copy their strategies. These problems were further compounded by
Singapore’s small domestic market, which made it necessary for these companies to
expand into other Asian markets in order to survive. However, most of these startup
companies lacked the necessary networks and resources to establish a foothold overseas.
The outcome is that many of these startups declined rapidly and closed down after a
couple of years.
Beyond the need to rely on the other parts of the science and technology system, One-
North has also expressly addressed the need to increase Singapore’s technological
capabilities. It plans to attract research-performing firms from overseas (such as firms
from Australia and New Zealand) by offering them a superior intellectual property regime
and by providing better marketing channels to the rest of the Asian region. This presents
another challenge, as MNCs generally prefer to locate their overseas R&D operations in
markets with a ready supply of highly skilled labor (e.g., scientists with PhDs) and low
costs of operation.

5.3.3. Global role definition and market integration

The success of the Hsinchu Science District has shown that strong linkages to
technology centers such as Silicon Valley, via the transnational community of Taiwanese
entrepreneurs located in the United States, can have a significant impact on the success
of a science park. Like Hsinchu, Singapore sought earlier to create a global role for
itself. However, apart from local contract manufacturers that rose to service MNCs
located in Singapore, there was no significant transnational community like Hsinchu’s to
link up startups in Singapore and help them create inroads into the U.S. and European
markets.11

11
Singapore’s few contract manufacturers are more likely to locate their production and R&D facilities outside
of Singapore! Singapore’s most globally known product maker—Creative Technologies—is still limited to niches
within the global PC peripherals market.
 F.C.C. Koh et al. / Journal of Business Venturing 20 (2005) 217–239 233

If One-North’s strategy to attract companies to locate their R&D activities in

Singapore succeeds, Singapore could then have the basis for defining a role for itself
as a regional R&D hub or R&D commercializing center, either on par with or in
connection to new global technological clusters. It remains to be seen how this role
evolves: whether Singapore can attract firms and sustain the growth of firms with
what is still a foreign investment-led (i.e., infrastructure- and government-policy-led)
policy, and whether it will then be able to plug itself effectively into a regional
production chain, or into the global production chain (as Hsinchu did). Either strategy
will not be easy. The former requires developing Singapore as a gateway to other
Asian markets and establishing a regional role for itself in high-tech R&D, perhaps
with a constellation of intellectual property protection provider and other resources. On
the other hand, a global role requires developing a stellar capability in R&D, so as to
serve a more global function as an R&D hub, or to become a manufacturing base,

Fig. 1. The growth path of Singapore’s science park strategy.

234 F.C.C. Koh et al. / Journal of Business Venturing 20 (2005) 217–239

both of which are harder to do when increasing numbers of countries have higher-
skilled but lower wage workers in abundance. The alternative strategy is for One
North to develop a role in the national economy, but since Singapore is a city–state,
and has a smaller market, this may be a less relevant strategy. Developing a relevant
global or regional role effectively sends a strong signal that it is indeed a significant
base (for R&D or whatever else) in the region, further helping to consolidate its
position.

5.4. The evolution of Singapore’s science park strategy

In our analytical framework, we postulated three aspects in the gestation, evolution, and
sustainability of science parks, namely, (a) growth mechanisms, (b) level of technological
capabilities, and (c) global role definition and integration of markets. The dynamic nature of
growth mechanisms and global role definition can be illustrated in Fig. 1, which shows the
evolution of Silicon Valley, Cambridge, and Hsinchu.
All the three exemplars in our study have matured into self-sustaining technology
‘‘districts’’ or regions, in part because of local firms, and have gone beyond their
boundaries to develop global linkages and substantial technological capabilities. If
Singapore’s One-North is successful in developing its growth mechanisms, strengthening
its technological capabilities, and linking itself to global markets—in short, becoming a
self-sustaining hub for research and innovation—Singapore stands a good chance to
achieve greater success in defining a global or regional role for itself. In this case, the
future growth path for Singapore’s science parks could shift course towards the other
exemplars, as shown in Fig. 1. The task will not be easy, and requires a substantial
amount of investment and efforts to develop the supporting institutions in the years
ahead.

6. Conclusion

Our aim in this paper was to present an analytical framework to study the development of
science parks and technology regions, and the factors contributing to their successes. The
proposed analytical framework focuses on a science park’s growth mechanisms, the
sophistication of its research capabilities, and the role that it plays in the national and global
economy. The framework was built on the experiences of Silicon Valley, Cambridge Science
Park, and the Hsinchu Science District. Although the three exemplars differ in their initial
growth impetus, they all possess strong self-renewal capabilities in new firm formation.
Furthermore, these exemplars have continued to sustain the R&D competencies that propelled
their initial growth.
Finally, we applied the framework to Singapore’s earlier and recent science park strategies
to assess its development and to identify the challenges ahead. While the One-North strategy
represents a fresh attempt to chart a more aggressive growth path for Singapore’s science
parks, it is still primarily an instance of the infrastructure-led variety of growth mechanism.
 F.C.C. Koh et al. / Journal of Business Venturing 20 (2005) 217–239 235

Whether the efforts would succeed would depend on how successful One-North can acquire
the other requisite growth mechanisms. We believe that greater participation by the private
sector, as well as the development of linkages with Silicon Valley and other science parks and
technology hubs based on Singapore’s own development of strong native capabilities, could
augur well for the future of Singapore’s science parks.

Acknowledgements

We would like thank Kok-Huat Goh and Danny Phuan for briefing us on the
conception and strategy of the One-North project; Thomas Menkhoff, Marshall Meyer,
and Sock Yong Phang for discussions on the subject; Don Siegel, Philip Phan, the
Editors, and two anonymous referees for their many helpful comments and advice in
the revision of the paper. Financial support from the Wharton-SMU Research Centre,
Singapore Management University is also gratefully acknowledged. The usual
disclaimer applies here.

Appendix A. Recent studies of science parks and technology districts

Paper Subject Key findings Remarks

Storey Various science1. Science parks in Europe 1. This was a paper from the
and Tether parks in the were significantly smaller ‘‘institutional’’ perspective.
(1998) Europe than those in the USA. 2. The authors compared scale of
2. Science parks in Europe parks, contributions to economic
made modest contributions growth and inability to spawn fast-
to employment and growth science-based enterprise.
technology transfer. 3. The authors also argued for more
3. European science parks government support to assist
also did not seem to spawn technology-based firms to grow.
rapidly growing science-
based enterprises.
Lofsten Various science 1. This study compared 1. This paper the institutional
and Lindelof parks in Sweden on-park and off-park firms perspective of science park
(2003) to analyze the added value development focusing on their
that science parks offer to ‘‘added value.’’
new technology-based 2. The authors argued for initiatives
firms (NTBFs). to promote new technology-based
2. Firms located within firms within science parks.
science parks had more
links with local universities.
3. NTBFs within parks
yielded a higher rate of
job creation.
(continued on next page)
236 F.C.C. Koh et al. / Journal of Business Venturing 20 (2005) 217–239

Appendix A (continued)
Paper Subject Key findings Remarks
Athreye Cambridge 1. Cambridge had 1. This paper projected the ‘‘economic-
(2002) Hi-Tech Cluster, developed an array of geography’’ perspective of science park
United Kingdom institutions, university – development.
industry links and local 2. It highlighted Cambridge’s premier
technology venture capital position as a centre for science-based
that have nurtured entrepreneurial activity in sectors in
entrepreneurship in science- Europe.
based industries from its
humble beginnings.
2. Compared to Silicon
Valley, there was an absence
of large firms based on
product market successes.
3. Growth of employment
came mainly from the growth
in the number (not the size)
of new establishments.
4. Unlike Silicon Valley,
Cambridge was not highly
specialized in high-technology
production in the UK.
5. However, there was a
significant amount of science-
based entrepreneurship and
some local network effects
among the scientists.
Florida and Silicon Valley 1. Silicon Valley (‘‘SV’’) 1. This study critically appraised the
Kenny and Route 128, and Route 128 projected negative ‘‘institutional’’ aspects of
(1990) USA the image of freewheeling, success.
high-technology 2. SV and Route 128 might have stood
entrepreneurship and out as symbols of economic and
quick-shooting venture technological success to the Americans
capital activities. (and the emerging countries of Asia),
2. They had created new but they did not ‘‘reinvigorate and
and highly innovative renew traditional industries.’’
companies at breakneck 3. SV’s phenomenal growth has bred
speed. They ‘‘can catalyze innovative companies that may not be
the world’s most advanced equipped to follow on with large-scale
breakthrough innovations.’’ product commercialization.
3. These centers also
generated a high degree
of internal competition and
a serious problem of
industrial fragmentation.
4. The unfortunate reality
of SV and Route 128 was
one of severe, at times,
devastating competition.
 F.C.C. Koh et al. / Journal of Business Venturing 20 (2005) 217–239 237

Appendix A (continued)
Paper Subject Key findings Remarks
Saxenian Hsinchu 1. The Hsinchu region of 1. This is a paper from both the
(2001a) Science Taiwan grew from a ‘‘institutional’’ and ‘‘economic
District, government effort in 1980 geography’’ perspectives.
Taiwan with the development of 2. The author argued that there is an
infrastructure. It started agglomeration effect as firms benefited
with clusters of small firms. from proximity and interfirm mobility.
It drew in FDI from the United 4. The author also observed that the
States and Japan, domestic connections between clusters are critical
and foreign talents, and for their success.
Taiwan returnees from SV. 5. From our perspective, the domestic
It had grown to be a world alumni network and external
centre for integrated circuits, connections with Silicon Valley together
computers, and peripherals. provided Hsinchu the added advantage
2. The Hsinchu region, like for its sustainable growth into the future.
them SV, was an example
of Marshallian external
economies, in which the
localization of skill,
specialized materials and
inputs, and technological
know how generates cost
reductions for individual
firms and increasing returns
to the region as a whole. It
has created an
‘‘agglomeration’’ effect.
3. Interfirm mobility
insured the diffusion of
tacit knowledge and
facilitates the process of
new firm formation.
4. The flows of people,
information, and know-
how linking SV and Hsinchu
were so great that the latter
is like an extension of SV.

References

Amirahmadi, H., Saff, G., 1993. Science parks: a critical assessment. J. Plan. Lit. 8 (2), 107 – 123.
Amsden, A.H., Chu, W., 2003. Beyond Late Development: Taiwan’s Upgrading Policies. MIT Press, Cambridge,
MA.
Amsden, A.H., Tschang, T., 2003. A new approach to assessing the technological complexity of different
categories of R&D (with examples from Singapore). Res. Policy 32, 553 – 572.
Athreye, S.S., 2002. Agglomeration and growth: a study of the Cambridge Hi-Tech Cluster. Stanford Institute for
Economic Policy Research Discussion Paper 00-42. Stanford University.
238 F.C.C. Koh et al. / Journal of Business Venturing 20 (2005) 217–239

Bakouros, Y.K., Mardas, D.C., Varsakelis, N.C., 2002. Science park, a high tech fantasy?: an analysis of the
science parks of Greece. Technovation 22, 123 – 128.
Bresnahan, T., Gambardella, A., Saxenian, A., Wallsten, S., 2001. ‘‘Old economy’’ inputs for ‘‘new economy’’
outcomes: cluster formation in the new Silicon Valley. Stanford Institute for Economic Policy Research
Discussion Paper 00-43. Stanford University.
Conceicao, P., Heitor, M.V., Piperno, W., Rubini, D., 2002. Perspectives for the observation of Italian technology
parks (Paper presented at the). 6th International Conference on Technology and Innovation, August 12 – 15,
Kansai, Japan.
Felsenstein, D., 1994. University-related science parks—‘seedbeds’ or enclaves of innovation? Technovation 14
(2), 93 – 110.
Florida, R.L., Kenny, M., 1988. Venture capital-financed innovation and technological change in the USA. Res.
Policy 17, 119 – 137.
Florida, R.L., Kenny, M., 1990. Silicon Valley and Route 128 won’t save us. Calif. Manage. Rev. 33 (1), 68 – 88.
Green, C., 2002. Learning by comparing technopoles of the world—the Cambridge phenomenon (Paper
presented at the) . 6th International Conference on Technology and Innovation, August 12 – 15, Kansai,
Japan.
Hou, C.M., Gee, S., 1993. National systems supporting technical advance in industry: the case of Taiwan. In:
Nelson, R. (Ed.), National Innovation Systems. Oxford University Press, Oxford, pp. 384 – 413.
Hu, A.G., Shin, J.-S., 2002. Climbing the technology ladder. In: Koh, A.T., Lim, K.L., Hui, W.T., Rao, B., Chng,
M.K. (Eds.), Singapore Economy in the 21st Century. McGraw-Hill, Singapore, pp. 229 – 320.
Kenny, M., 2000. Understanding Silicon Valley: Anatomy of an Entrepreneurial Region. Stanford University
Press, Stanford, CA.
Kihlgren, A., 2003. Promotion of innovation activity in Russia through the creation of science parks: the case of
St. Petersburg (1992 – 1998). Technovation 23 (1), 65 – 73.
Koh, F.C.C., Koh, W.T.H., 2002. Venture capital and economic growth: an industry overview and Singapore’s
experience. Singap. Econ. Rev. 47 (2), 243 – 267.
Lee, W.H., Yang, W.T., 2000. The cradle of Taiwan high technology industry development—Hsinchu Science
Park. Technovation 20, 55 – 59.
Lee, C.M., Miller, W.F., Hancock, M.G., Rowen, H.S., 2000. The Silicon Valley Edge: A Habitat for Innovation
and Entrepreneurship. Stanford University Press, Palo Alta, CA, USA.
Lofsten, H., Lindelof, P., 2002. Science parks and the growth of new technology-based firms—academic – industry
links, innovation and markets. Res. Policy 31, 859 – 876.
Lofsten, H., Lindelof, P., 2003. Determinants for an entrepreneurial milieu: science parks and business policy in
growing firms. Technovation 23 (1), 51 – 64.
Markusen, A., 1996. Sticky places in slippery places: a typology of industrial districts. Econ. Geogr. 72 (3),
283 – 313.
Ministry of Trade and Industry, 1986. The Singapore Economy: New Direction. SNP Publishers, Singapore.
Ministry of Trade and Industry, 1991. The Strategic Economic Plan: Towards a Developed Nation. SNP Publish-
ers, Singapore.
Palmai, Z., in press. An innovation park in Hungary: INNOTECH of the Budapest University of Technology and
Economics. Technovation.
Phillimore, J., 1999. Beyond the linear view of innovation in science park evaluation: an analysis of Western
Australian technology park. Technovation 19, 673 – 680.
Phillips, S.M., Yeung, H.W., 2003. A place for R&D? The Singapore Science Park. Urban Stud. 40, 707 – 732.
Porter, M.E., 1991. Towards a Dynamic Theory of Strategy. Strategic Management Journal 12, 95 – 117 (Special
Issue: Fundamental Research Issues in Strategy and Economics).
Saxenian, A.L., 1994. Regional Advantage: Culture and Competition in Silicon Valley and Route 128. Harvard
University Press, Cambridge, MA, USA.
Saxenian, A.L., 2001a. The Silicon Valley – Hsinchu connection: technical communities and industrial upgrading.
Berkeley Plan. J. 15, 3 – 31.
 F.C.C. Koh et al. / Journal of Business Venturing 20 (2005) 217–239 239

Saxenian, A.L., 2001b. Taiwan’s Hsinchu region: imitator and partner for Silicon Valley (Paper presented at the).
Conference on Silicon Valley and Its Imitators, July 28, 2000.
Seely Brown, J., Duguid, P., 2002. Local knowledge: innovation in the networked age. Manag. Learn. 33 (4),
427 – 438.
Storey, D.J., Tether, B.S., 1998. Public policy measures to support new technology-based firms in the European
Union. Res. Policy 26, 1037 – 1057.
Storper, M., Harrison, B., 1991. Flexibility, hierarchy and regional development: the changing structure of
industrial production systems and their forms of governance in the 1980s. Res. Policy 20, 407 – 422.
Vedovello, C., 1997. Science parks and university – industry interaction: geographical proximity between the
agents as a driving force. Technovation 17, 491 – 502.
Westhead, P., Batstone, S., 1998. Independent technology-based firms: the perceived benefits of a science park
location. Urban Stud. 35 (12), 2197 – 2219.
Westhead, P., Storey, D.J., 1995. Links between higher education institutions and high technology firms. Omega
Int. J. Manag. Sci. 23 (4), 345 – 360.
Xue, L., 1995. Developing science parks in Asia Paper presented at the. 4th World Conference on Science and
Research Parks, September 26 – 28, Beijing, PR China.