See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/330493259

How much of Raymond Vernon's product cycle thesis is still relevant today:
evidence from the integrated circuits industry

Article in International Journal of Technological Learning Innovation and Development · January 2019
DOI: 10.1504/IJTLID.2019.10018598

CITATIONS READS

3 3,851

2 authors, including:

Rajah Rasiah
University of Malaya
405 PUBLICATIONS 4,673 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Industrial Policy and Economic Development View project

Climate Change Mitigation View project

All content following this page was uploaded by Rajah Rasiah on 24 August 2020.

The user has requested enhancement of the downloaded file.

56 Int. J. Technological Learning, Innovation and Development, Vol. 11, No. 1, 2019

How much of Raymond Vernon’s product cycle thesis

is still relevant today: evidence from the integrated
circuits industry

Rajah Rasiah*
Asia-Europe Institute,
University of Malaya, Malaysia
Email: rajah@um.edu.my
*Corresponding author

Xiao-Shan Yap
Copernicus Institute of Sustainable Development,
Utrecht University, Netherlands
Email: x.s.yap@uu.nl

Abstract: Vernon’s product cycle thesis has increasingly been questioned as

industries experienced differentiation and dispersal of production stages, while
the need for proximity for interactions to take place between scientists and
engineers, and consumers ended following the introduction of computerised
inventory and production, and planning systems. This paper re-examines this
thesis using the integrated circuits (IC) industry. The results show that IC
multi-nationals have continued to retain frontier R&D and wafer fabrication
activities at locations endowed with strong human capital and research centers.
However, IC firms are attracted to relocate frontier R&D activities in distant
host-sites, such as USA, Japan, Spain, Germany, Russia and Israel that
are endowed with sophisticated science, technology and innovation
infrastructure, and scientists and engineers. Also, IC manufacturing is too
knowledge-intensive to be attracted to the least developed countries.

Keywords: product cycle; multi-national corporations; human capital; R&D;

integrated circuits; science, technology and innovation infrastructure.

Reference to this paper should be made as follows: Rasiah, R. and

Yap, X-S. (2019) ‘How much of Raymond Vernon’s product cycle thesis is still
relevant today: evidence from the integrated circuits industry’,
Int. J. Technological Learning, Innovation and Development, Vol. 11, No. 1,
pp.56–77.

Biographical notes: Rajah Rasiah is a Distinguished Professor of Economics

at the Asia-Europe Institute, University of Malaya. He is the 2014 recipient of
the Celso Furtado Prize awarded by the World Academy of Sciences for
advancing the frontiers of social science thought (Economics). He obtained his
Doctorate in Economics from Cambridge University in 1992 and was Rajawali
Fellow at Harvard University in 2014.

Xiao-Shan Yap is a Marie-Skłodowska-Curie Fellow at Utrecht University, the

Netherlands. She obtained her Doctorate in Economics from University of
Malaya.

Copyright © 2019 Inderscience Enterprises Ltd.

 How much of Raymond Vernon’s product cycle thesis is still relevant today 57

1 Introduction

Vernon (1966) had posited a pattern of relocation of the USA multi-national corporations
(MNCs) using the product cycle thesis associated with international trade. Being the
largest economy since the Second World War, Vernon felt that new products of the US
would be developed in the home country because of the close interaction needed between
researchers and engineers who possess the requisite skills and consumers who can afford
high-priced new products. As the production technology matures and becomes
standardised, and prices fall, its relocation to low wage locations is expected to occur,
which would enable the appropriation of profits in low income countries so long as prices
exceed marginal costs.
The product-cycle thesis came under challenge when Helleiner (1973) found evidence
to argue that MNCs relocated production stages in low cost sites simultaneously by
differentiating them by factor intensities and product costs. Light manufacturing offered
considerable evidence of labour-intensive stages relocated at labour-surplus economies:
clothing manufacturing and the assembly of semiconductors in the 1960s and 1970s
(Scibberas, 1977).
Since the 1990s, Cantwell (1995), Reddy (2000) and UNCTAD (2005) sought to
argue that MNCs had begun relocating R&D operations abroad, discrediting in the
process Vernon thesis altogether. However, Amsden and Tschang (2003) contested this
claim using evidence from Singapore of R&D operations abroad that have remained
confined to peripheral aspects that do not involve frontier research. While certain
propositions of the product-cycle rationale may no longer be applicable as pointed out by
Helleiner (1973) and Cantwell (1995), Vernon’s (1966) relocation argument was posited
on the basis of institutional development phases at production sites, which has yet to be
examined thoroughly by critics. Particularly, the institutional conditions of host-sites that
Vernon described may offer lessons for governments seeking to attract R&D operations.
Vernon’s (1979, pp.258–259) argument became clear in a later article in which he
conceded that the predictive power of his earlier model had declined because of the
evolution of MNC networks, and changes in the global environment as several economies
developed successfully.
Following the convincing argument of Nelson and Winter (1985) and Nelson (2008),1
we take the evolutionary position that location, timing and economic activity determine
the type institutional change and conduct of economic agents. Hence, using the integrated
circuits (IC) industry, this paper seeks to revisit Vernon’s (1966, 1979) product cycle
thesis with a focus on the location of new knowledge creating and using activities. The
rest of the paper is organised as follows. Section 2 explains the product cycle thesis.
Section 3 discusses aspects of the product cycle thesis that are still relevant today.
Section 4 evaluates aspects of the product cycle thesis that are no longer applicable.
Section 5 presents the methodology and data used in the paper. Section 6 analyses the
findings. Section finishes with the conclusion.

2 Vernon’s product cycle thesis

Vernon (1966) had argued that lead firms would retain operations in the developed
countries so as to enjoy product innovation rents from their high incomes, access to
58 R. Rasiah and X-S. Yap

superior human capital, and strong R&D infrastructure (Figure 1). While in this article
Vernon did not posit the support firms would enjoy from their national governments in
the form of incentives and grants, and protection from intellectual property rights (IPRs),
he had included it in his subsequent work (Vernon, 1971, 1981).
We use Figure 1 to illustrate Vernon’s (1966) product cycle thesis. In his original
article, firms would retain production in the US when products are at their introductory
(1) and growth (2) stages. In his subsequent works (see Vernon, 1979, 1981; Vernon and
Wells, 1991), it can be assumed that stages 1 and 2 can be started or located in the
developed and the rapidly industrialising countries. In addition to superior human capital
support and R&D infrastructure, stage 1 can also benefit from frequent interaction
between employees with sophisticated knowledge (R&D scientists and engineers) and
effluent consumers. Stage 2 benefits from strong demand in the developed countries,
including Western Europe and Japan. As product 1 reaches maturity (3) and prices start
to fall, falling innovation rents will result in the firm launching product 2 in the home
market to appropriate innovation rents again. At stage 3, production is relocated at less
developed locations endowed with low wages, while the cycle of each product will end at
stage 4. By stage 3, the production technology of product 1 becomes standardised and
simple enough to facilitate its relocation to developing economies by which time IPRs
lose their significance and prices exceed marginal costs. This process then continues with
subsequent product launches as production of mature products are relocated to the
developing economies.

Figure 1 Vernon’s product cycle thesis

Developed countries (stages 1 and 2)

1. Strong human capital support R&D

2. High income support high priced products
3. Strong suppliers
4. R&D incentives and grants
5. Strong IPR protection
6. Strong university and laboratories to support Product Product Product
R&D a b c

Developing countries
Stage 3: (stages 3 and 4)
Maturity a. Mature products
b. Standardised production
Stage 2: Stage 4: technology
Growth Decline c. Low product prices

Stage 1:
Introduction

Source: Drawn based on Vernon (1966)

Wells (1968, 1969) used Vernon’s (1966) product cycle and Hufbauer’s (1966) trade
theoretic frameworks to argue that firms in the US become exporters in the first phase as
their products become internationally competitive. Once the products mature and become
 How much of Raymond Vernon’s product cycle thesis is still relevant today 59

standardised, production of these products is relocated at the developing economies in the

second phase. In the third phase, subsidiaries ramp up operations to make production at
the foreign host-site competitive. In the fourth and final phase, low value-added products
from the foreign host-site are imported into the US. However, Wells (1968) extended
Vernon’s argument without differentiating production on the basis of novelty and
maturity by differentiating labour costs between the home (US) and host countries. Using
strong empirical evidence, he demonstrated how manufacturers abroad may seek to
introduce pricing differentials using lower labour costs (so long as prices are not less than
the marginal cost) to undercut prices so as to capture markets in the US.

3 What still holds

Vernon (1966, 1979, 1981) had argued that firms would locate new product development
facilities in developed countries because of their superior infrastructure – e.g., human
capital, R&D institutes, universities and IPRs protection. Herein lies some confusion on
what Vernon meant. Unlike in Vernon (1971, 1974), the original Vernon (1966)
argument only referred to developed countries’ superior institutional support for R&D
activities, and its importance in the interactions between researchers and engineers with
their special skills and consumers. Hence, one should assume that it is within Vernon’s
(1966, 1979) argument to anticipate at some point for firms to seek sophisticated R&D
support from other developed locations. Cantwell et al. (2000) provided some evidence of
this, though they consequently dismissed Vernon’s (1966) product cycle argument.
While UNCTAD (2005) had quoted extensive evidence of R&D by MNCs in the
developing countries, this study did not differentiate between frontier and peripheral
activities. Similarly, Goshal and Bartlett (1988), Archibugi and Michie (1995), Cantwell
(1995), Patel and Pavitt (2000) and Reddy (2000) discussed strong evidence of MNC
subsidiaries undertaking R&D in the developing world but did not distinguish them
between frontier and peripheral activities. Such a differentiation is important as Vernon
(1966, 1979) emphasised the availability of research and engineering support for frontier
R&D activities. In the IC industry, the US enjoyed a huge technological lead until the
1980s in all its sub-sectors (UNCTC, 1987).
Hence, Amsden and Tschang (2003) were right in questioning the premature rejection
of Vernon’s (1966) argument on why R&D will remain in the developed countries as the
rationale was the availability of human capital, research universities and other supporting
organisations to undertake such activities, which are unlikely to exist in the developing
economies. Indeed, the whole rationale of how latecomer countries strategise to catch up
technologically is based on learning from foreign technology, including creatively
transforming older technologies (Gerschenkron, 1952; Freeman, 1985; Lall, 1980, 1987;
Ernst, 1990; Amsden, 1991; Kim, 1997; Amsden and Chu, 2003).
In addition, at the time that Vernon (1966) originally made his argument, apart from
the USA economy being the most sophisticated and largest, there was little heterogeneity
in income or infrastructure among the developing economies. The contrast was simply
like differentiating two sets of countries with an enormous development gap between
them. While Vernon’s (1966) focus on the pattern of MNC investment was associated
with international trade, the logic behind his argument – that the most developed country
will be endowed with the most sophisticated technology – is consistent with Amsden’s
(1989, 1991), Ernst’s (1990) and Kim’s (1997) emphasis on ‘learning from the
60 R. Rasiah and X-S. Yap

acquisition of foreign technology’. Vernon (1979, p.266) had acknowledged since his
original article that several developing countries have industrialised rapidly to enjoy the
infrastructure facilities to support R&D, which is why the predictability of his product
cycle thesis has decreased. MNCs have since the 1980s started to undertake R&D
through their subsidiaries in a wide range of newly industrialised countries. MNCs
motivations for internationalising R&D operations to the developing countries can be
classified into the following types.
1 To appropriate user-producer relations synergy in large economies that provide
sufficient demand (Vernon, 1979; Lundvall, 1992).
2 To appropriate innovation synergies from specialisation in processes and production
technology development (Rasiah, 1989). Sturgeon (2002) viewed this from the
perspective of moving up in modular value chains.
3 To support product proliferation and adaptation (Hobday, 1995).
4 To take advantage of natural resource endowments (Narula and Dunning, 2000).
5 To seek bio-prospecting of materials (UNCTAD, 2005).
As acknowledged by Vernon (1979) the topography of development has changed
considerably since the 1970s because of a convergence in the capabilities and incomes of
among countries. Also, MNC networks in the IC industry has experienced major changes
since the 1970s as the proliferation of cutting edge coordination software systems linking
production with supply and demand has reduced the significance of distance. However,
although the relationship between producers and users has changed, Vernon’s (1966)
general argument has remained strong in the IC industry as most of the R&D activities
carried out by MNC subsidiaries in the developing economies are – as noted by Amsden
and Tschang (2003) – in peripheral or activities that support assembly and processing.
Indeed, such activities are not related at all to the frontier IC activities of miniaturisation
of the minimum line width and the extension of the wafer diameter.
Vernon’s (1966) argument on this point becomes clear in Gruber et al. (1967) and
Vernon (1979). Gruber et al. (1967) used R&D intensities through the deployment of
R&D expenditure in sales and R&D scientists and engineers in the workforce to
distinguish MNCs R&D intensities and their relationship with exports comparing in the
process the USA with Western Europe and Japan using 1962 data. Also, the statistical
evidence used not only showed the US to enjoy the highest trade balance in industries
where R&D expenditure in sales and R&D personnel in workforces were the highest, but
also the US led the other developed countries on R&D-intensive exports. Vernon (1979,
p.257) even recognised that US MNCs had started to introduce new products in the
developing countries. In some cases the relocation of innovation abroad was a
consequence of US regulations as in the pharmaceutical industry (Gabrowski and
Vernon, 1977; Allen, 1978).2
Hence, to take on Vernon’s argument on the locational importance of R&D support it
is pertinent to examine frontier R&D activities rather than just to focus on all types of
R&D activities. Although Gruber et al. (1967) and Vernon (1979, pp.258–259)
underlined them within the rubric of R&D intensity in production, we use frontier R&D
activities that are targeted at radical innovations. Schumpeter (1943) referred to it as new
stocks of knowledge that require research in large R&D centres. Subsequent
developments have shown that even small firms can participate in R&D activities through
 How much of Raymond Vernon’s product cycle thesis is still relevant today 61

networks established with research universities and public R&D laboratories (Pavitt
et al., 1987; Audretsch, 1995).3 Nelson and Winter (1982), Cantwell (2002) and Cantwell
and Fai (1999) had distinguished these activities to establish the link between institutions
and radical innovation activities. Schumpeter (1961, p.66) distinguished such radical
innovation activities from incremental innovation with the latter referred to as minor
innovations achieved through a wide range of improvements to what already exists.

4 Declining predictability

As Vernon (1979, 1981) recognised later, the predictability of his model declined as the
world had changed dramatically since his original product cycle theory was posited,
which is a consequence of expansion in the network of MNC subsidiaries and rapid
industrialisation of several countries (Vernon and Wells, 1991). Vernon had argued that
firms expecting high innovation rents would launch new products in high income
countries so that researchers and engineers with the special skills and knowledge engaged
in R&D could get the feedback needed from consumers, and where exists the requisite
facilities to develop products for effluent consumers. His theory was based on the
assumptions that production is not divisible, transport costs would matter and that
proximity to facilitate interaction between researchers and engineers, and consumers
would be critical in the development of new products. All three of these assumptions
became less significant since the 1970s as technological change facilitated the
differentiation of production of the newest products by stages, while the costs of flying
light manufactured goods dropped with instantaneous exchange of knowledge that was
enabled by computerised information flows. MNCs began differentiating their production
and started to relocate the labour-intensive stages in the developing world to take
advantage of low wages and in some cases financial incentives, such as tax and tariff
exemptions (Scibberas, 1977; Lim, 1978). Indeed, by the late 1970s the assembly and test
of the latest ICs of US firms, such as Intel, AMD, Texas Instruments, Motorola, National
Semiconductor, Monolithic Memories and Hewlett Packard were undertaken abroad in
developing countries, such as, Malaysia, Philippines and Singapore (Rasiah, 1989).
Vernon’s assumption that production will not be differentiated was first demolished
by Helleiner (1973) who showed that MNC value chains were since the late 1960s
differentiated by stages, with the site of location of each stage was determined by relative
factor endowments. Rasiah (1989) went further to argue that political stability, security,
basic infrastructure and incentives were also important (Rasiah, 1989). Wells (1993,
p.178) subsequently distinguished three rounds of investors to the Third World. In the
third round he found mobile exporters from countries, such as Korea, Hong Kong,
Taiwan and Singapore relocating in countries, such as Indonesia, to produce products that
were not subject to the so-called orderly marketing arrangements or other forms of export
quotas at home. Exports were generated from third countries through both internalised
MNC networks, as well as from external contractors. Contract manufacturers became
important exporters of clothing and computers since the late 1980s and 1990s (Rasiah,
1995).
Also, proximity to major markets became insignificant in light manufacturing as the
trans-nationalisation of production by IC MNCs saw R&D and wafer fabrication retained
at home sites, while assembly and test operations relocated at politically stable low wage
sites until the end of the 1970s (Chaponniere, 1984). Indeed, even in locations where
62 R. Rasiah and X-S. Yap

there existed little domestic demand, such as the Philippines, Malaysia, Singapore, South
Korea and Taiwan – output from distant subsidiaries, (which was originally flown back to
parent sites before final sales were made), production had evolved to generate direct
exports from host-sites to final consumers since the 1980s (Rasiah, 1989). The emergence
of successful catching up and leapfrogging experience of developing country firms, such
as Samsung Semiconductor in memory chips and Taiwan Semiconductor Manufacturing
Corporation (TSMC) in logic chips transformed the division of labour in the IC value
chain. Whereas Samsung created its own brand name, TSMC specialised on wafer
fabrication without its own brand. Consequently, South Korea and Taiwan have become
the most novel exporters of memory chips and logic chips respectively.
Vernon (1974, 1979, 1981) and Vernon and Wells (1991) acknowledged that the
rapid industrialisation of several economies had reduced the predictive power of his
product cycle thesis. He had anticipated this to happen in the other developed countries
and in rapidly industrialising countries, such as South Korea and India because of the
growth in the capabilities of researchers and scientists, and in domestic consumer demand
(e.g., China and South Korea). However, he did not foresee that these developments were
also evolving in subsidiaries where the host-location’s market were irrelevant, such as
Taiwan and Malaysia. The introduction of just-in-time (JIT), integrated materials
resource planning (MRPII) and cellular manufacturing since the 1980s had already
enabled computerised coordination between producers and consumers so that proximity
had become less important (Rasiah, 1989).4 Consequently, the appropriation of
monopolistic advantages through the internalisation of in-house capabilities at distant
host-locations became the innovation rent.5 Indeed, US IC firms quickly expanded their
R&D operations from just raising the sophistication of their inventions and
labour-saving characteristics to include the dimensions of material and efficiency in their
R&D focus since the 1980s to compete effectively with European and Japanese IC firms
(Rasiah, 1989; see also Figure 2). Consequently, US firms began to absorb JIT, MRP11
and kaizen6 practices to raise material and space utilisation efficiency. Hence, the
hypothetical ‘global scanner’ example that Vernon (1979, pp.261–262) had used has
indeed been globalised since the 1980s in the IC industry.
In addition, there are some secondary evidence of R&D undertaken abroad and
among the developing countries (Cantwell, 1995; Reddy, 2000). Similarly, employment,
export and learning opportunities offered by MNCs at foreign host-sites is documented
extensively (Rasiah, 1989; Lall, 1996). However, none of these studies have
distinguished between frontier R&D activities and re-engineering of older products and
processes, which led to Amsden (1991) and Kim (2003) arguing that the R&D
undertaken at subsidiaries abroad are limited to peripheral activities. Acquisition has been
one of the routes that IC firms have used to undertake re-engineering R&D abroad
(Edquist and Jacobssen, 1987; Brown and Linden, 2009).

5 Methodology and data

The study focuses on only one industry to ensure that industry-level differences did not
affect the conclusions, which is consistent with the arguments of evolutionary economists
(Nelson and Winter, 1982; Nelson, 2008). While the study covers all IC firms to identify
the location of each of their stages of production, the case studies focused on interviews
to identify the reasons for their relocation used the top four (Intel, Samsung
 How much of Raymond Vernon’s product cycle thesis is still relevant today 63

Semiconductor, Texas Instruments, TSMC and Freescale) and the next six (Global
Foundries, ST Microelectronics, Infineon, Freescale, Renesas and UMC) from the next
10 firms based on revenues reported in 2010, and 2 (Alterra and Avago) from others with
at least production operations in two countries (Avago and Alterra). The list of firms was
obtained from Gartner (2012a, 2012b), including the location of their operations
internationally by stages of operations.
The case study methodology was chosen because of the in-depth insights it offers,
especially on particular decisions made for relocating high technology operations abroad
(Bryan, 1964; Hanbrick, 1983; Thomas, 1982; Dyer and Nobeoka; 2000; Langlois and
Steinmuller, 2000; Best, 2001; Karnani, 2007; Inkpen, 2008). While detailed case studies
lack statistical rigor and generalisability, they offer the best accounts of complex
strategies adopted by firms, organisations and individuals. As pointed out by Doyle
(2003, p.326), interviews with the firms involved were preferred over gathering large
datasets for statistical analysis because the purpose of this study was interpretive rather
than predictive. Indeed, interviews offer the best primary source of information gathering
to establish the relocation strategies pursued by IC firms. To reduce the problem of
exaggerations by the respondents, the information then was checked with secondary
sources from electronics journals and Gartner (2012a, 2012b).
The interviews were carried out over eight years because of the difficulty associated
with establishing contacts with the managers and managing directors. We interviewed
two retired vice presidents and five retired managers from Intel, five retired managers
from Advanced Micro Devices (renamed as Global foundries following the acquisition
by Mubadala Corporation),7 three managers from Freescale, one retired managing
director and two managers from Texas Instruments, two managers from Renesas, two
managers from ST Microelectronics, one chief executive officer and one director from
TSMC, two managers from UMC, one manager from Samsung, one manager from
Avago, one manager from Alterra, and one manager from Infineon. In total we carried
out interviews with 31 managing directors, directors and managers from 12 IC firms with
operations in two or more countries.
Table 1 Officials interviewed, 2005–2013
Firm Number of officials Locations
1 Avago 1 Penang
2 Alterra 1 Penang
3 Freescale 3 Petaling Jaya
4 Global Foundries/AMD 5 Penang
5 Infineon 1 Kulim
6 Intel 7 Penang and San Jose
7 Renesas 3 Penang
8 Samsung Semiconductor 1 Kuala Lumpur
9 ST Microelectronics 2 Johore
10 Taiwan Semiconductor Manufacturing 2 Hsinchu
Corporation (TSMC)
11 Texas Instruments 3 Kuala Lumpur and Penang
12 United Microelectronics Company 2 Hsinchu
(UMC)
Source: Authors’ interviews, 2005–2013
64 R. Rasiah and X-S. Yap

Figure 2 Technological upgrading in IC value chain since the 1980s

Knowledge Vertical upgrading
intensity

Independent R&D
Horizontal upgrading

Capacity implant and specifications

Wafer
fabrication R&D support
Marketing (OBM)

Chip design

Sales

Assembly and test

Stages in value chain

Source: Upgraded from UNCTC (1987, pp.87–101)

Information on technological capabilities of the firms was obtained from company
websites and information compiled by Gartner (2012a, 2012b). We then interviewed
managers and managing directors from 12 firms to identify the reasons that determined
their relocation patterns. Owing to the confidentiality of the study, no interviewee is
revealed in this paper.
Tremendous technological changes have taken place in the IC industry with
upgrading occurring both vertically and horizontally (see Figure 2). Hence, in addition to
specific questions on the dynamics of relocation, the 31 IC company officials were asked
to fill up a questionnaire using Likert scale scores ranging from 1 to 5, on the importance
of low wages, basic infrastructure, political stability, pool of R&D engineers and
scientists, supply of uniform-band power, proximity to universities that carried out
profound research on ICs, of home-country locations, tariff-free operations, tax
exemptions and R&D grants. Since the observations were small, the mean scores were
used to analyse the importance of these variables on the relocation of IC production
activities of frontier R&D, supportive R&D, wafer fabrication, chip design, and assembly
and test operations. We believe the detailed case study approach used in this paper is
superior to extracting regression results from large data sets owing the inductive approach
required for such a study.
 How much of Raymond Vernon’s product cycle thesis is still relevant today 65

6 Findings and discussion

Table 2 shows the distribution of plants of 186 IC firms across the world (compiled from
Gartner, 2012a, 2012b).8 Out of a total of 76 IC R&D centres,9 the US, South Korea (9),
Japan (7), Taiwan (6), and Germany (5) and UK (5) dominated R&D activity in the IC
industry. All of these countries enjoy strong presence of researchers and scientists related
to IC technology and research support facilities. Most R&D centres (71.2%) were also
located in the home countries. Only 4 (5.5%) of such centres were located outside the
developed countries. Among these countries, the centre in China is nationally owned
while foreign firms own the ones in Brazil, Israel and Russia. Among these centres only
the ones in Israel and Russia are engaged in frontier R&D activities, which have
benefited from both the presence of high quality researchers and scientists and their
interaction with the military consumers. The centres in the rapidly industrialising
countries of Brazil and China are not engaged in frontier R&D activities.
In addition, the distribution of knowledge-intensive wafer fabrication plants across
the world departs from the original product cycle argument of Vernon (1966) but
supports Vernon’s (1979, 1981) subsequent argument. All frontier wafer fabrication
activities targeted at the fabrication of 12-inch wafers that produce leading edge
microprocessors, memory and logic chips were situated close to cutting-edge R&D
centres in 2010 that are endowed with a strong pool of related researchers and engineers
(see also Figure 2).
The developed countries had 491 (86.9%) wafer fabrication plants in 2010 with the
largest number was located in Japan (201) followed by the USA (133), Taiwan (53),
South Korea (28), and Germany (24). It is interesting to note that the leading edge wafer
fabrication plants were only located in places endowed with quality researchers and
scientists and research support. Even Singapore, which has a per capita income higher
than the US, did not have a leading edge wafer fabrication plant because of the lack of a
sufficient pool or scientists and engineers and related research support. Also, only Intel
and Samsung had frontier wafer fabrication centres outside their home countries in 2010:
the former in Israel, Germany, Russia and Spain, and the latter in the US and Japan.
Interviews with key officials from AMD, Avago, Alterra, Freescale, Global Foundries,
Infineon, Intel, Renesas, Samsung, ST Microelectronics, Texas Instruments, TSMC and
UMC confirm these findings.10 These developments support Vernon’s (1979, 1981)
subsequent argument that locations endowed with researchers and engineers and support
from state-of-the-art research laboratories and universities matter for the relocation of
frontier R&D activities.
Like Singapore, a number of rapidly industrialising and developed countries have
managed to attract wafer fabrication plants but none of them are engaged in frontier
operations. The provision of grants and a strong science, technology and innovation
infrastructure (including researchers and scientists) but without frontier research
laboratories and universities have induced several firms to relocate their older fabrication
plants in Brazil, China, Singapore, and Malaysia to undertake re-engineering activities
targeted at product and process proliferation activities. Frontier research support is
66 R. Rasiah and X-S. Yap

essential to attract latest wafer fabrication activities (see Figure 2). Hence, the rapidly
industrialising and Eastern European countries had 60 (10.6%) and 14 (2.5%) wafer
fabrication plants respectively in 2010 (Table 1). Whereas all such plants in Eastern
Europe are nationally-owned, the 11 foreign owned plants (11.3%) in the rapidly
industrialising countries were engaged in older technologies.
Whereas wafer fabrication is highly capital-intensive, (which requires investment
outlays exceeding US$3 billion to break even), chip design is knowledge-intensive.
Slightly over half (58.2%) of chip design centres were located in foreign locations (see
Table 1). The presence of large pools of design and software engineers – either supplied
by domestic universities or from abroad – is the prime inducement for attracting such
activities. Hence, although the developed countries had the largest number of chip design
centres (72.9%), the rapidly industrialising countries had 24.7% of them in 2010. China
and India had 20 and 13 centres each, which was only exceeded by the US (40). Strong
supply of engineers – both domestically and the returning diaspora – have assisted China
and India to support large numbers of chip design centres (Saxenian, 2006).
In addition, US firms began to relocate sophisticated R&D operations abroad. For
example, Intel started frontier R&D activities in Israel and Russia. While the military
consumers of Israel and Russia have been important, the prime reason for the relocation
of frontier R&D activities in these countries were the availability of R&D scientists and
engineers, and research facilities. Intel has also started frontier software development
programmes in Bangalore (India). The development of strong pools of human capital in-
house in several international locations have helped strengthen coordination between
researchers and engineers, and consumers so much so that R&D activities have become a
globally determined strategy.
The acquisition of the wafer fabrication plants of AMD (formerly US owned) by
Global Foundries in 2009 following the global financial crisis, (which is owned by
Mubadala Development Company of Abu Dhabi), is an example where the wafer
fabrication and R&D is completely undertaken abroad in Dresden, New York and
Singapore. As with AMD before the acquisition, Global Foundries enjoys financial grants
and a strong pool of researchers and engineers at these sites.11 Vernon (1979, p.256,
1981) has acknowledged the possibility of these departures happening from his original
product cycle thesis as he recognised that US MNCs in a number of industries have
launched new products abroad in the rapidly industrialising countries. An example is in
Latin America because of restrictive government regulations at home in the
pharmaceuticals industry (Gabrowski and Vernon, 1977). However, Vernon’s (1979,
p.263) contention that location of R&D at prime research centres remains important point
as Mubadala Development Company established an agreement in 2014 to acquire the
R&D facilities of IBM located in Armonk and Santa Clara.12
The miniaturisation process has also raised the knowledge content of assembly and
test activities in the IC industry since the 1980s (Rasiah, 1989) so that IC firms have
increasingly introduced R&D to support these activities (see Figure 2). Hence, there were
115 supportive R&D facilities in the IC industry in 2010. With 90 of the facilities the
developed countries had 78.3% of supportive R&D facilities while the rapidly
industrialising countries had 19.1% (22).
Also, changes in production technology, especially computerisation in the use of
integrated MRPII, total preventive maintenance (TPM), quality control circles (QCC),
statistical process control (SPC), small group activities (SGA), and JIT, as well as,
supportive R&D have made cheap low-skilled labour reserves no longer important in the
 How much of Raymond Vernon’s product cycle thesis is still relevant today 67

IC industry since the mid-1980s (Rasiah, 1989). Nevertheless, these developments fit
well the changed world that Vernon (1979, 1981) had described later as the developing
countries still had 41.3% (222) of the assembly and test plants in 2010.
However, the rising knowledge content of IC assembly has meant that there were no
IC plants in any of the least developed countries (LDCs) in 2010. The proliferation of
supportive R&D has resulted in the return of significant numbers of assembly and test
operations to parent locations so that most of them were in 2010 located in Japan (125)
and the US (53). These countries still dominate foreign assembly and test plants with
75.7% (156) of the world’s plants. Also, the Taiwanese strategy of stage specialisation
has driven firms in assembly and test operations, such as, ASE, to invest in supportive
R&D activities (Yap and Rasiah, 2017).13
Table 2 IC firms by stages, ownership and country groups, 2010

R&D Chip design Other Wafer Assembly and

Group centre centre R&D Fab test
N F N F N F N F N F
Developed* 52 21 66 58 46 44 427 64 266 49
Eastern 0 1 0 4 0 3 14 0 0 1
Europe
RICs 1 1 5 37 1 21 49 11 66 156
LDCs 0 0 0 0 0 0 0 0 0 0
Overall 53 23 71 99 47 68 490 75 332 206
Notes: * – classification of developed countries by World Bank (2010); N – national;
F – foreign; RICs – rapidly industrialising countries; LDCs – least developed
countries.
Source: Calculated from Gartner (2012a, 2012b)
Vernon’s (1966) original product cycle thesis did not specifically posit that MNCs will
not relocate R&D operations abroad. Instead, had argued that high technology activities
will be located at home-sites when first introduced because of the need for proximate
interactions between researchers and engineers with special skills, and consumers in the
initial stage. Hence, as Vernon (1979, pp.265–266) had discussed later, R&D activities
can be relocated abroad in developed or in rapidly industrialising countries once these
conditions are satisfied (Vernon, 1981; Vernon and Wells, 1991).14
It is only in his subsequent work that Vernon (1971) noted that R&D tends to be
located at parent nations because of nationalist sentiments behind the allocation of R&D
grants to their own national firms. While governments have been more favourable in
offering tax breaks to firms irrespective of ownership structures, they tend to finance
through grants their own national firms in activities, such as R&D. However, competition
between governments to attract frontier R&D operations resulted in foreign governments
providing such grants. For example, Singapore, Ireland, Germany and the USA have
provided grants to attract such activities, including in capital-intensive wafer fabrication.
Indeed, the New York state government offered a grant of USD1.4 billion to Global
Foundries to relocate its USD4.6 billion Fab 8 project at Saratoga (Timesunion, 2011).
However, while the national element no longer holds, this development is consistent with
Vernon’s (1966) earlier work in which he had argued that such activities will be situated
68 R. Rasiah and X-S. Yap

at locations enjoying strong supply of advanced scientists and engineers and research
facilities. Interviews with the 12 IC firms support this observation.
As Vernon (1979, p.266) had acknowledged later, the rapid development of the newly
industrialised countries of South Korea and Taiwan – all of which enjoy per capita
incomes exceeding some of the traditionally defined developed economies – have seen
them evolve their infrastructure to out-compete developed locations to support R&D
activities.15 The world’s lead firms in memory and logic chips, i.e., Samsung
Semiconductor and TSMC respectively, enjoy frontier research support their own
countries. These countries were underdeveloped and poor in the 1960s when Vernon
(1966) advanced the product cycle thesis but became developed by the time they were
able to support frontier R&D activities. Both countries have developed successfully their
human capital, and research universities and laboratories. Countries lacking such
endowments, including Malaysia, Singapore and Thailand have failed to attract frontier
IC R&D activities.
Another major development has taken place since the 1990s whereby knowledge
from R&D is globally shared within the IC MNCs networks. The fusion of knowledge
goes beyond interactions between R&D centres located in distant parts of the world.
Feedback linkages between all segments in the IC value chain enables flows of
knowledge back from wafer fabrication, chip design and assembly and test plants to R&D
centres. For example, Intel’s and AMD’s production engineers reported that the internet
has intensified interactions that have improved considerably the link between R&D and
the production processes.16
Also, evidence also shows that countries lacking basic infrastructure have failed to
attract any IC manufacturing stages. Examples include the Sub-Saharan economies,
Mongolia, Bangladesh, Cambodia, Myanmar, Laos, Indonesia, Papua New Guinea, Fiji
and Samoa. Interviews with the 12 IC firms confirmed that no part of IC assembly will be
relocated at cheap labour sites that do not have the capacity to engage in kaizen-type
incremental innovation activities, even in export processing zones that offer good basic
infrastructure, security and financial incentives.

6.1 Frontier R&D activities of lead firms

We undertake here a focused study of the three technologically top-ranked frontier firms
in the IC industry, i.e., Intel in microprocessors, Samsung Semiconductor in memories
and TSMC in logic chips. Intel undertakes frontier R&D activities in locations endowed
with requisite research laboratories, universities and human capital (see Table 3). Intel’s
most sophisticated R&D plants are in Hillsborough (USA) and Haifa (Israel).17 Its plants
in Spain, France and Russia are also important contributors to inventions that support
Intel’s activities. All these plants enjoy strong support from the presence of expert human
capital and research-based universities and laboratories. Intel has established R&D
collaboration at the frontier following its recognition of three key researchers and the
quality of research undertaken at laboratories at Universitat Politecnica de Catalunya
(UPC) in Barcelona, Spain.18
In addition to its own internalised operations in South Korea, Samsung has frontier
R&D operations in San Jose and Pasadena in the USA and Yokohama in Japan (see
Table 3). Following the pledge to provide grants by the New York State government, a
global consortium of firms was formed with Intel, IBM, Samsung, TSMC and Global
Foundries. The R&D collaboration carried out at Albany is targeted at developing the
 How much of Raymond Vernon’s product cycle thesis is still relevant today 69

world’s first 18 inch wafer that will reduce unit chip costs considerably. If this project
succeeds it will unprecedented as this is the first time IC firms from different countries
are attempting to jointly develop the next generation wafers. Albany is also the location
of the College of Nanoscale Science and Engineering (CNSE) and SEMATECH centres
of research. IBM and Global Foundries are the other participants in the Consortium
(Global Consortium, 2013).
Table 3 Foreign R&D centres, Intel, Samsung and TSMC, 2012

Firm Headquarters Foreign subsidiary Stage

Intel Santa Clara Duisburg R&D
Nuremburg R&D
Moscow R&D
Nuremburg R&D
Nizhny Novgorod R&D
Novosibirsk R&D
Sophia Antipolis R&D
Barcelona R&D
Albany (planned) R&D on wafer diameter
Samsung Yongin City San Jose R&D
Semiconductor Pasadena R&D
Yokohama R&D
Albany (planned) R&D on wafer diameter
TSMC Hsinchu Science Industrial Albany (planned) R&D on wafer diameter
Park
Source: Compiled by authors (2017)
The results of interviews with 31 top management firms of 12 ICs based on the stages
shown in Figure 2 are analysed below. The Likert-scale mean scores in table show that
R&D scientists and engineers (5.0), proximity of IC research universities (4.8), R&D
grants (4.6) and tax breaks (4.4) were the most important determinants of IC firms’
decision to locate frontier R&D operations. Political stability (4.3) and basic
infrastructure (4.0) were the next most important factors. Location and home sites (2.8)
and tariff-free benefits (2.8) were not as important. Uniform band power supply (1.3) and
low wages (1.9) were the least important. The respondents were not as un-scientists and
engineers.
Supportive R&D operations produced a slightly different set of results. The
respondents were not as unanimous as with frontier R&D operations on scientists and
engineers (4.0) and proximity of IC research universities (3.7), though these factors were
the most important in attracting such operations. In fact, tax breaks (4.2), political
stability (4.1), basic infrastructure (3.9) and R&D grants (3.9) ranked higher with
supportive R&D operations. Tariff-free operations (2.6) and home-sites (2–6) were of
average importance. Low wages (2.1) and uniform band power (1.6) were not important.
Scientists and engineers (5.0), uniform band power (4.9), R&D grants (4.5), basic
infrastructure (4.5), tax breaks (4.3), political stability (4.2) and IC research universities
(4.1) were the most important attractions for wafer fabrication. Tariff free operations
(3.9) were still important. Home site (2.5) and low wages (2.0) were not very important.
 70

Table 4

Low Basic Political Uniform band Scientists and IC Research Home Tariff Tax R&D
R. Rasiah and X-S. Yap

wage infrastructure* stability power engineers University site free break grants
Frontier R&D 1.9 4.0 4.3 1.3 5.0 4.8 2.8 2.8 4.4 4.6
Supportive R&D 2.1 3.9 4.1 1.6 4.0 3.7 2.6 2.6 4.2 3.9
Wafer fabrication 2.0 4.5 4.2 4.9 5.0 4.1 2.5 3.9 4.3 4.5
Chip design 3.0 4.0 4.0 2.0 4.5 3.0 2.0 4.0 4.0 3.0
Assembly and test 3.1 4.2 4.1 1.8 3.4 2.9 2.1 4.1 4.4 3.3
N 31 31 31 31 31 31 31 31 31 31
Notes: Basic infrastructure refers to roads, airports and airline services, water supply, security, basic education, healthcare and mean Likert-scale
scores.
Source: Firm study
Important factors in IC firms’ relocation decisions by stages, 2005–2013
 How much of Raymond Vernon’s product cycle thesis is still relevant today 71

Scientists and engineers (4.5) were the most important attraction for chip design. In fact,
two former managing directors from Intel reported that the firm solicits chip design from
specialists in Bangalore despite having no semiconductor manufacturing in India. Good
basic infrastructure (4.0), political stability (4.0) and tax breaks (4.0) were the next most
important factors. R&D grant (3.0), low wages and IC research universities (3.0) were
only slightly above average in importance. Uniform band power (2.0) and home site (2.0)
were not important.
Tax breaks (4.4), and basic infrastructure (4.2), political stability (4.1) and tariff free
operations (4.1) were the most important pull factors for IC assembly and test operations.
Scientists and engineers (3.4) were fairly important. The significance of scientists and
engineers complement the direct interviews to show that chip assembly and test is fairly
knowledge-intensive, and hence, is unlikely to be considered for the LDCs even if good
basic infrastructure could be offered at export processing zones. Low wages (3.1) and IC
research universities (2.9) were only slightly above average in terms of importance.
Uniform band power (1.8) and home sites (2.1) were not important.
The results show that home sites do not figure as important in attracting any of the IC
stages and activities. However, the availability of related scientists and engineers, R&D
grants and IC research universities are very important in attracting the most
knowledge-intensive and investment-intensive activities of frontier R&D and wafer
fabrication. Tax breaks are important in attracting all IC activity but the mean Likert
scale scores range from 4.0 to 4.4. Chip design is knowledge-intensive, and hence,
scientists and engineers are the most important pull factor. Uniform band power is only
important for wafer fabrication because the processes require such support in processes,
such as lithography.
Overall, as acknowledged later by Vernon (1979, 1981, 1994), our findings show a
significant decline in the predictive power of the product cycle thesis. Also, in the IC
industry, Vernon’s (1979) hypothetical case of MNCs acquiring global scanning capacity
to coordinate demand no longer holds as computerised control systems have diffused
through all subsidiaries since the 1980s. However, Vernon’s (1966) argument over the
location of R&D in developed countries in general, and where there is the requisite
scientists and engineers, and frontier research infrastructure still holds. Furthermore, as
acknowledged later by Vernon (1979) the rapidly industrialising countries have managed
to attract R&D operations albeit only in the peripheral activities. The major implication
from this evidence is that governments will have to offer such an infrastructure that is
rich in related scientists and engineers, research universities and laboratories, and grants
if they are to succeed in attracting frontier R&D activities. In addition, rapidly
industrialising countries endowed with strong STI infrastructure but without sophisticated
scientists and engineers and frontier research facilities can attract supportive R&D, chip
design and wafer fabrication activities using grants but not in frontier R&D activities.

7 Conclusions

As acknowledged by Vernon (1979, 1981, 1994) the predictive power of his product
cycle thesis has indeed declined over the years. Among the limitations that Vernon
(1966) did not anticipate include the divisibility of production into different stages for its
international relocation based on international endowments. Subsequently IC MNCs
72 R. Rasiah and X-S. Yap

evolved the capacity to coordinate production taking account of supply and demand
factors globally through the use of highly computerised materials resource planning
systems and human capital developed in-house or imported from abroad so that proximity
between researchers and engineers and consumers is no longer necessary. Vernon’s
(1979) hypothetical example of the global scanner has indeed moved beyond the firm and
proximate customer locations to the globe in IC production.
Vernon (199, 1981, 1994) had acknowledged that several developing economies had
since experienced rapid economic progress to offer sophisticated infrastructure to attract
R&D activities. Intel has relocated frontier R&D operations in Germany, Israel, Russia
and Spain to tap into their research support facilities and human capital, and their military
markets. Countries, such as, South Korea and Taiwan have deepened their pool of
researchers and scientists, and research laboratories. The countries offer grants to support
frontier R&D but only to national firms. Countries, such as, Singapore, and Brazil, China,
India and Malaysia have evolved their STI infrastructure to attract at least the
developmental aspects of R&D activities. Lacking in cutting-edge research support from
laboratories and universities, Singapore, China, Malaysia and Brazil have only managed
to attract second-tier wafer fabrication plants by offering grants and strong STI
infrastructure.
Chip design, which is knowledge-intensive but not capital-intensive, has spread to
countries enjoying large endowments of related design and software engineers. Hence,
while the US is still the leader, the rapidly industrialising economy of India has also
figured prominently in chip design operations by foreign MNCs. The lack of a sufficient
pool of software engineers have reduced the attractiveness of Southeast Asian nations for
chip design operations.
Since the entire IC value chain has become knowledge-intensive since the 1980s
supportive R&D has increasingly become important even in assembly and test activities,
which explains why slightly over half of such activities have been relocated in the
developed countries by 2010. While the emergence of supportive R&D activities has
spread to the rapidly industrialising countries, it has also reduced further the likelihood of
IC assembly and test being relocated in the LDCs.
While the predictive power of Vernon’s (1966) product cycle thesis has fallen owing
to changes in the world economy, it still holds in two key stages in the IC value chain,
i.e., frontier R&D activity and frontier wafer fabrication activities as they are only located
at sites endowed strongly with human capital, research universities and laboratories and
R&D grants. Although several governments, such as, Singapore and Malaysia, have
offered grants to attract frontier R&D and frontier wafer fabrication activities they have
not been successful because of the lack of strong research facilities and human capital.
These findings have wide implications for policy. Governments seeking to attract
frontier wafer fabrication and R&D activities must first develop such infrastructure in
their own countries. The evidence shows that it is possible to attract R&D and wafer
fabrication related to non-frontier technologies and chip design with grants and good
basic infrastructure, which is what Singapore, China, and Malaysia have managed to do.
India has the advantage with a massive pool of software engineers to attract chip design
in particular locations. However, host-sites will have to have strong human capital and IC
research support to attract frontier R&D and frontier wafer fabrication operations. South
Korea and Taiwan developed those facilities in their countries through the acquisition of
foreign high technology firms, establishment of strong research centres at home and
 How much of Raymond Vernon’s product cycle thesis is still relevant today 73

inviting back their diaspora endowed with frontier knowledge. Finally, assembly and test
activities have also become too knowledge-intensive for its relocation at LDC sites.

Acknowledgements

We are grateful to Loius T. Wells, Richard Nelson and John Cantwell for their
constructive comments. We are thankful to Dwight Perkins, Anthony Saich and
Kaori Urayama for offering support at the John Ash Center for Democratic Governance
and Innovation, Harvard. University. We wish to also acknowledge University of
Malaya’s High Impact Research (HIR) Grant (Ref No. UM.C/625/1/HIR/ASH/019),
which we used to finance the fieldwork trips abroad, and in Malaysia since 2012. The
usual disclaimer applies.

References
Allen, T.J. (1978) Managing the Flow of Technology, MIT Press, Cambridge.
Amsden, A.H. (1989) Asia’s Next Giant: South Korea and Late Industrialization, Oxford
University Press, New York.
Amsden, A.H. (1991) ‘Diffusion of development: the late-industrializing model and greater
East Asia’, American Economic Review, Vol. 81, No. 2, pp.282–286.
Amsden, A.H. and Chu, W.W. (2003) Beyond Late Development: Taiwan’s Upgrading Policies,
MIT Press, Cambridge.
Amsden, A.H. and Tschang, F.T. (2003) ‘A new approach to assessing the technological
complexity of different categories of R&D (with examples from Singapore)’, Res Policy,
Vol. 32, No. 4, pp.553–572.
Archibugi, D. and Michie, J. (1995) ‘The globalization of technology: a new taxonomy’,
Cambridge Journal of Economics, Vol. 19, No. 1, pp.121–140.
Audretsch, D. (1995) Innovation and Industry Evolution, MIT Press, Cambridge.
Best, M. (2001) The New Competitive Advantage, Oxford University Press, Oxford.
Brown, C. and Linden, G. (2009) Chips and Change: How Crisis Reshapes the Semiconductor
Industry, MIT Press, Cambridge.
Bryan, S.E. (1964) ‘TFX – a case in policy kevel decision making’, Academy of Management
Journal, Vol. 7, No. 1, pp.54–70.
Cantwell, J.A. (1995) ‘The globalisation of technology: what remains of the product cycle model?’,
Cambridge Journal of Economics, Vol. 19, No. 1, pp.155–174.
Cantwell, J.A. (2002) ‘Innovation, profits and growth: Penrose and Schumpeter’, in Pitelis, C.N.
(Ed.): The Theory of the Growth of the Firm: The Legacy of Edith Penrose, pp.215–248,
Oxford University Press, Oxford.
Cantwell, J.A. and Fai, F.M. (1999) ‘Firms as the source of innovation and growth: the evolution of
technological competence’, Journal of Evolutionary Economics, Vol. 9, No. 3, pp.331–366.
Cantwell, J.A., Iammarino, S. and Noonan, C.A. (2000) ‘Sticky places in slippery space – the
location of innovation by MNCs in the European regions’, in Pain, N. (Ed.): Inward
Investment, Technological Change and Growth, Macmillan, London.
Chaponniere, R. (1984) ‘The ASEAN integrated circuit: the electronics industry in
ASEAN – issues and perspectives’, ASEAN Economic Bulletin, Vol. 1, No. 2, pp.136–151.
Cook, G. and Johns, J. (2013) The Changing Geography of International Business, Pelgrave
Macmillan, London.
74 R. Rasiah and X-S. Yap

Doyle, L.H. (2003) ‘Synthesis through meta-ethnography: paradoxes, enhancements, and

possibilities’, Qualitative Research, Vol. 3, No. 3, pp.321–344.
Dunning, J.H. (1988) ‘The eclectic paradigm of international production: a restatement and some
possible extensions’, Journal of International Business Studies, Vol. 19, No. 1, pp.1–31.
Dyer, J.H. and Nobeoka, K. (2000) ‘Creating and managing a high-performance knowledge-sharing
network: the Toyota case’, Strategic Management Journal, Vol. 21, No. 3, pp.345–367.
Edquist, C. and Jacobbsen, S. (1987) ‘The integrated circuit industries in India and South Korea in
an international techno-economic context’, Industry and Development, Vol. 21, No. 1,
pp.1–62.
Ernst, D. (1990) Technology and Global Competition: The Challenge Ahead for Newly
Industrializing Countries, Springer-Verlag, Berlin.
Freeman, C. (1985) Technology Policy and Economic Performance: Lessons from Japan, Pinter
Publishers, London.
Gabrowski, H.G. and Vernon, R.M. (1977) ‘Consumer protection in ethical drugs’, American
Economic Review, Vol. 67, No. 1, pp.359–364.
Gartner (2012a) Fab Database: Worldwide, 2012 Update, Gartner Database.
Gartner (2012b) Semiconductor Packaging, Assembly and Test Facilities: Worldwide, 2012
Update, Gartner Database.
Gerschenkron, A. (1952) ‘Economic backwardness in historical perspective’, in Granovetter, M.
and Swedberg, R. (Eds.): The Sociology of Economic Life, Westview Press, Boulder.
Global Consortium (2013) Global Consortium To Make The Move To 450 mm [online]
http://www.siliconsemiconductor.net/article/75801-Global-Consortium-To-Make-The-Move-
to-450mm.php (accessed 22 September 2018).
Goshal, S. and Bartlett, C. (1988) ‘Creation, adoption and diffusion of innovations by subsidiaries
of multinational corporations’, Journal of International Business Studies, Vol. 19, No. 3,
pp.365–388.
Gruber, W., Mehta, D. and Vernon, R. (1967) ‘The R&D factor in international trade and
international investment of United States industries’, Journal of Political Economy, Vol. 75,
No. 1, pp.20–37.
Hanbrick, D.C. (1983) ‘An empirical typology of mature industrial-product environments’,
Academy of Management Journal, Vol. 26, No. 2, pp.213–230.
Helleiner, G. (1973) ‘Manufactured exports from less-developed countries and multinational
firms’, Economic Journal, Vol. 83, No. 329, pp.21–47.
Hobday, M. (1995) Innovation in East Asia: The Challenge to Japan, Edward Elgar, Cheltenham.
Hufbauer, G.C. (1966) Synthetic Materials and the Theory of International Trade, Harvard
University Press, Cambridge.
Hymer, S. (1960) International Operations of National Firms, Doctoral thesis approved by MIT
(published in 1976), MIT Press, Cambridge.
Imai, M. (1986) Kaizen: The Key to Japan’s competitive Success, McGraw-Hill International
Edition, Singapore.
Inkpen, A.C. (2008) ‘Knowledge transfer and international joint ventures: the case of NUMMI and
General Motors’, Strategic Management Journal, Vol. 29, No. 4, pp.447–453.
Karnani, A. (2007) ‘Doing well by doing good – case study: ‘fair & lovely’ whitening cream’,
Strategic Management Journal, Vol. 28, No. 13, pp.1351–1357.
Kim, L. (1997) Imitation to Innovation: The Dynamics of Korea’s Technological Learning,
Harvard University Press, Cambridge.
Kim, L.S. (2003) ‘The dynamics of technology development: lessons from the Korea experience’,
in Lall, S. and Urata, S. (Eds.): Competitiveness, FDI and Technological Activity in East Asia,
Edward Elgar, Cheltenham.
 How much of Raymond Vernon’s product cycle thesis is still relevant today 75

Lall, S. (1980) ‘Monopolistic advantages and foreign involvement by U.S. manufacturing industry’,
Oxford Economic Papers, Vol. 32, No. 1, pp.102–122.
Lall, S. (1987) Learning to Industrialize: The Acquisition of Technological Capability by India,
Macmillan, London.
Lall, S. (1996) Learning from the Asian Tigers Studies in Technology and Industrial Policy,
Pelgrave Macmillan, London.
Langlois, M. and Steinmuller, W.F. (2000) ‘Strategy and circumstance: the response of American
firms to Japanese competition and semiconductors’, Strategic Management Journal, Vol. 21,
Nos. 10–11, pp.1163–1173.
Lim, L.Y.C. (1978) Multinational Firms and Manufacturing for Export in Less-Developed
Countries: The Case of the Electronics Industry in Malaysia and Singapore, Unpublished
Doctoral thesis approved by Michigan University, Ann Abhor.
Lundvall, B.A. (1992) ‘User-producer relations, national systems of innovation and
internalization’, in Lundvall, B.A. (Ed.): National Systems of Innovation: Toward a Theory of
Innovation and Interactive Learning, Pinter, London.
Narula, R. and Dunning, J. (2000) ‘Industrial development, globalization and multinational
enterprises: new realities for developing countries’, Oxford Development Studies, Vol. 28,
No. 2, pp.141–167
Nelson, R.R. (2008) ‘What enables rapid economic progress: what are the needed institutions’, Res.
Policy, Vol. 37, No. 1, pp.1–11.
Nelson, R.R. and Winter, S. (1985) An Evolutionary Theory of Economic Change, Harvard
University Press, Cambridge.
Oviatt, B.M. and McDougall, P.P. (1994) ‘Toward a theory of international new ventures’, Journal
of International Business Studies, Vol. 25, No. 1, pp.45–64.
Oviatt, B.M. and McDougall, P.P. (2004) ‘The internationalization of entrepreneurship’, Journal of
International Business Studies, Vol. 36, No. 1, pp.2–8.
Patel, P. and Pavitt, K. (2000) ‘National systems of innovation under strain: the internationalisation
of corporate R&D’, in Barrell, R., Mason, G. and O’Mahony, M. (Eds.): Productivity,
Innovation and Economic Performance, pp.217–235, Cambridge University Press,
Cambridge.
Pavitt, K., Robson, M. and Townsend, J. (1987) ‘The size distribution of innovating firms in the
UK: 1945–1983’, Journal of Industrial Economics, Vol. 35, No. 3, pp.291–316.
Rasiah, R. (1989) ‘Competition and restructuring in the semiconductor industry: implications for
technology transfer and absorption in Penang’, Southeast Asian Journal of Social Sciences,
Vol. 17, No. 2, pp.41–57.
Rasiah, R. (1995) Foreign Capital and Industrialization in Malaysia, Macmillan, Basingstoke.
Reddy, P. (2000) Globalization of Corporate R&D: Implications for Innovation Systems in Host
Nations, Routledge, London.
Saxenian, A.L. (2006) The New Argonauts: Regional Advantage in a Global Economy, Harvard
University Press, Cambridge.
Schumpeter, J.A. (1943) Capitalism, Socialism and Democracy, Harvard University Press,
Cambridge.
Schumpeter, J.A. (1961) The Theory of Economic Development: An Inquiry into Profits, Capital,
Credit, Interest and the Business Cycle, translated from German by R. Opie, Harvard
University Press, Cambridge.
Scibberas, E. (1977) Multinational Electronic Companies and National Economic Policies, JAI
Press, Connecticut.
Sturgeon, T. (2002) ‘Modular production systems: a new American model of industrial
organization’, Industrial and Corporate Change, Vol. 11, No. 3, pp.451–496.
76 R. Rasiah and X-S. Yap

Thomas, H. (1982) ‘Screening policy options – an approach and a case-study example’, Strategic
Management Journal, Vol. 3, No. 3, pp.227–244.
Timesunion (2011) Global Foundries Cash Grant Largest Ever Awarded in U.S.,
9 October [online] https://www.timesunion.com/business/article/GlobalFoundries-cash-grant-
largest-ever-awarded-2209888.php (accessed 13 September 2018).
Tsai, T. and Cheng, B.S. (2006) The Silicon Dragon: High-Tech Industry in Taiwan, Edward Elgar,
Cheltenham.
UNCTAD (2005) ‘World investment report: transnational corporations and the internationalization
of R&D’, United Nations Conference for Trade and Development, Geneva.
UNCTC (1987) Transnational Corporations in the International Semiconductor Industry, United
Nations Center for Transnational Corporations, New York.
Vernon, R. (1966) ‘International investment and international trade in the product cycle’, Quarterly
Journal of Economics, Vol. 80, No. 2, pp.190–207.
Vernon, R. (1970) ‘Location of economic activity’, in Dunning, J.H. (Ed.): Economic Analysis and
the Multinational Enterprise, pp.83–114, George Allen and Unwin, London.
Vernon, R. (1971) Sovereignty at Bay: The Multinational Spread of U.S. Enterprises, Macmillan,
London.
Vernon, R. (1974) ‘Competition policy toward multinational corporations’, American Economic
Review, Vol. 64, No. 2, pp.276–282.
Vernon, R. (1979) ‘The product cycle hypothesis in a new international environment’, Oxford
Bulletin of Economics and Statistics, Vol. 44, No. 2, pp.255–267.
Vernon, R. (1981) ‘Sovereignty at bay ten years later’, International Organization, Vol. 35, No. 3,
pp.517–529.
Vernon, R. (1994) ‘Contributing to an international business curriculum: an approach from the
flank’, Journal of International Business Studies, Vol. 25, No. 2, pp.215–227.
Vernon, R. and Wells, T.L. (1991) Economic Environment of International Business, Prentice Hall,
Englewood Cliffs.
Wells, L.T. (1968) ‘A product life cycle for international trade?’, Journal of Marketing, Vol. 32,
No. 3, pp.1–6.
Wells, L.T. (1969) ‘Test of a product cycle model of international trade: U.S. export of consumer
durables’, Quarterly Journal of Economics, Vol. 83, No. 1, pp.152–162.
Wells, L.T. (1993) ‘Mobile investors: new foreign investors in Asia’, in Froot, K.A. (Ed.): Foreign
Direct Investment, Chicago University Press, Chicago.
Yap, X.S. and Rasiah, R. (2017) Catching Up and Leapfrogging: The New Latecomers in the
Integrated Circuits Industry, Routledge, London.

Notes
1 The pattern of entrepreneurship and spatial distribution involving new ventures can vary
significantly between industries (see for example Oviatt and McDougal, 1994, 2004; Cook and
Johns, 2013).
2 Dunning (1988) captured a number of this in his eclectic paradigm.
3 Just as Vernon (1966) did not anticipate the differentiation of production stages, Schumpeter
(1943) did not anticipate the linkages that could evolve between small firms and R&D
laboratories, especially university-initiated start-ups that appropriate institutional support from
incubator facilities at science parks. The small firm, Phison, which invented the pen-drive was
started by five persons with the findings of the founder’s master thesis as the building blocks.
The firm was incubated at the Hsinchu Science Industrial Park and in 2010 was engaged in
designing activities with 200 employees and a share value exceeding US$1 billion (author
interview with Phua Kian Seng on 12 September 2010 at Hsinchu).
 How much of Raymond Vernon’s product cycle thesis is still relevant today 77

4 Savings from space, labour and material enabled Intel to close down its assembly and test
plants in Puerto Rico and Barbados in the mid-1980s and integrate them in Penang (Rasiah,
1989).
5 This development is consistent with the argument of Hymer (1960) and Vernon (1970),
though, the particular route through which the monopolistic advantage was gained was not
anticipated then.
6 Refers to continuous improvement (Imai, 1986).
7 Internship of nine months at AMD in 1986 (firm acquired from Monolithic Memories
Incorporated) and in 1990 over 6 months exposed Rajah Rasiah to the detailed operations of
the firm in Penang, including the division of labour, inventory and quality control systems,
and product and process technologies. Similar firm visits over a full day was then carried out
in Intel, Freescale, Hitachi and Fairchild.
8 Because the subsidiaries and parent plants are registered separately the number is significantly
larger than the total number of IC firms in the world. Nevertheless, it does not affect the
analysis as the focus of the paper is on plants located in different countries.
9 R&D centres in IC firms are primarily focused in the miniaturisation (reducing the minimum
line width), increasing IC yield, the extension of the wafer diameter and increasing the
functions of IC implants.
10 Author interviews carried out from 2005 until 2013 (see Table 1).
11 [online] http://www.mubadala.com/en/what-we-do/semiconductors/globalfoundries (accessed
21 November 2014).
12 Mubadala Development Corporation had on 14 October 2014 achieved a definitive agreement
to acquire IBM’s microelectronics division – subject to regulatory reviews – to strengthen its
R&D capabilities [online] http://globalfoundries.com/newsroom/press-releases/2014/10/20/
globalfoundries-to-acquire-ibm’s-microelectronics-business (accessed 21 November 2014).
13 Taiwanese IC firms, such as, TSMC, UMC, ASE, Vanguard and Winbond specialise in
particular stages of the IC value chain (see Tsai and Cheng, 2006; Yap and Rasiah, 2017).
14 See also Vernon’s (1994) recollections later.
15 Based on per capita incomes of countries in 2000.
16 Interview conducted by the authors on 15 September 1993 in Penang.
17 Interview with managing director and vice president of Intel in Penang on 19 September 2008.
18 [online] http://www.broadcastnewsroom.com/article/Intel-Expands-European-RD-Operations-
10430.

View publication stats