Spatial mediation and moderated effect on FDI performance: Empirical study of Taiwanese fi

Description
Taiwanese businesses exhibit a specific pattern when they enter into an uncertain territory, and tend to
start off with a small amount of investments and explore markets by improvising strategies and tactics.
This paper integrates theories concerning transaction costs and systems, and develops antecedent
causes, and intervening and interference factors, by referring to the perspectives of economic geography.
It is found that when faced with high uncertainties in emerging markets, multi-national corporations are
highly concerned about institutional environments. Labor cost of different regions is an implicit indicator
of the average wages that local companies are willing to pay. Companies that are willing to pay a high
labor cost are less likely to face labor shortage, thus ensuring normal operation and better performance.
© 2015 College of Management, National Cheng Kung University. Production and hosting

Spatial mediation and moderated effect on FDI performance:
Empirical study of Taiwanese ?rms in China (1999e2008)
Hsu-Wei Hsu
a, *
, Yi-Long Jaw
b
a
Department of Business Administration, Shih Chien University, Taipei, Taiwan
b
Department of International Business, National Taiwan University, Taipei, Taiwan
a r t i c l e i n f o
Article history:
Received 13 February 2014
Accepted 18 March 2015
Available online 9 July 2015
Keywords:
Labor cost
Human capital
Multi-level research
Spatial analysis
a b s t r a c t
Taiwanese businesses exhibit a speci?c pattern when they enter into an uncertain territory, and tend to
start off with a small amount of investments and explore markets by improvising strategies and tactics.
This paper integrates theories concerning transaction costs and systems, and develops antecedent
causes, and intervening and interference factors, by referring to the perspectives of economic geography.
It is found that when faced with high uncertainties in emerging markets, multi-national corporations are
highly concerned about institutional environments. Labor cost of different regions is an implicit indicator
of the average wages that local companies are willing to pay. Companies that are willing to pay a high
labor cost are less likely to face labor shortage, thus ensuring normal operation and better performance.
© 2015 College of Management, National Cheng Kung University. Production and hosting by Elsevier
Taiwan LLC. All rights reserved.
1. Introduction
Since China embarked on a series of reforms and open-door pol-
icies in the 1990s, a large number of foreign businesses entered into
the Chinese market. For Taiwanese businesses, the low labor cost in
China is the key factor in cost reduction. In addition, the growing
emerging markets in Asia, as well as the continuous development of
the domestic market in China, has made China one of the most
popular destinations for foreign direct investments. The eclectic
paradigmis a theoryineconomics andis alsoknownas the OLI-Model
or OLI-Framework. Dunning (1998) added three more factors to the
theory: (1) Ownership advantages, (2) Location advantages, and (3)
Internalization advantages. The discussion on howlocal labor costs in
China assist effective operations of multinational corporations
(MNCs) has hence become a key issue for academia and practitioners.
1.1. Research motive
Taiwanese businesses exhibit a speci?c pattern when they enter
into an uncertain territory, and tend to start off with a small amount
of investments and explore markets by improvising strategies and
tactics. Meanwhile, China is not an open society, and is known for
complications and opaqueness in information and regulations.
Therefore, it is understandable that companies adopt conservative
strategies (Boisot & Child, 1999). Based on Cooke (2009), this paper
discusses the changes in the quantity and quality of human re-
sources of MNCs in China, and observes the differences between
China and the West in studies on joint-ventures with MNCs,
regarding the integration of corporate strategies and human re-
sources management (Gong, Shenkar, Luo, & Nyaw, 2005).
Most studies on the convergence of regional economies tend to
ignore spatial factors and treat regions as isolated islands. This
approach eliminates the possibility of interactions between regions
(Mankiw, 1995; Quah, 1996). Highly-recognized studies, such as
Baumol (1986), Barro and Sala-i-Martin (1995) and Mankiw, Romer,
and Weil (1992), all overlook spatial factors. As a result, it becomes
impossible to gauge whether spatial effects have material impact in
the processes of regional economic growth. Hence, Rey and
Montouri (1999) questioned the robustness of the empirical ?nd-
ings of those studies. Infact, it is necessarytoincorporategeographic
factors. Buckley and Ghauri (2004) suggested that international
business (IB) is de?nitely a major issue concerning the effects of
multinational enterprise (MNE) in the global economy, due to the
results of strategic shifts. Spatial transaction costs and knowledge
assets play a pivotal role in the process (Mudambi, 2008).
1.2. Research purpose
Past research on entry mode (or strategy) often discusses part-
ner selection, but in fact, local economic growth and labor costs are
* Corresponding author.
E-mail address: [email protected] (H.-W. Hsu).
Peer review under responsibility of College of Management, National Cheng
Kung University.
HOSTED BY
Contents lists available at ScienceDirect
Asia Paci?c Management Review
j ournal homepage: www. el sevi er. com/ l ocat e/ apmrvhttp://dx.doi.org/10.1016/j.apmrv.2015.03.002
1029-3132/© 2015 College of Management, National Cheng Kung University. Production and hosting by Elsevier Taiwan LLC. All rights reserved.
Asia Paci?c Management Review 20 (2015) 252e264
also critical factors. For example, Cheng and Kwan (2000) studied
the impact of market size, infrastructure construction, preferential
tax rates, and salaries costs on direct overseas investment in China.
This study selected Organizational Slack and Human Capital Rela-
tive Risk (HCRR) to discuss the performance of Taiwanese busi-
nesses after entering the mainland Chinese market. This can more
accurately re?ect the motivation of the Taiwanese businesses to
enter the mainland Chinese market. Moreover, the entry mode may
be affected by institutional factors; however, Cheng and Stough
(2006) and Cheng and Kwan (2000) did not discuss the institu-
tional factors affecting the entry mode.
In order to bridge this gap in literature, this research has three
purposes: (1) to investigate the effect of HCRR in the spatial cluster
and entry mode on the performance of joint ventures; (2) based on
the Local Institutional Theory, to discuss what role infrastructure
and labor costs have in organizational performance; (3) as China
has a relatively closed environment and political instability;
therefore, what is the effect of Organizational Slack when Taiwa-
nese business enter the mainland Chinese market? To make up for
de?ciencies of the research tools of previous studies, this study
especially considers the cross-level effect, and attempts to discuss
the difference between this study and the international enterprise
theory, as seen fromthe point viewof spatiography, and develops a
detailed framework for Taiwanese businesses to invest the main-
land Chinese market.
2. Literature review
Most studies on regional economic convergence did not
consider spatial elements, and regarded regions as isolated islands,
while excluding the possibility of regional interaction (Mankiw,
1995; Quah, 1996). The classic works of Baumol (1986), Barro and
Sala-i-Martin (1995), and Mankiw et al. (1992) did not consider
spatial elements, thus, not providing an understanding on whether
the spatial effect has signi?cant in?uence in regional economic
growth process. Rey and Montouri (1999) considered the robust-
ness of empirical research conclusions as questionable.
However, past studies have suggested that location choices
have a signi?cant impact on the entry strategies of transnational
corporations (Chen & Chen, 1998). From the cities’ perspective, the
urbanization of mainland China has developed for more than 10
years, and the economic momentum has shifted from the ?rst-tier
coastal cities to third-tier and fourth-tier inland cities as produc-
tion element and labor costs become higher and higher, thus,
many Taiwanese businesses have been shifted to inland cities. The
International Institute for Urban Development, Beijing, estimated
that China will continue to accelerate urbanization in the next 30
years (Nian, 2006).
Beugelsdijk, McCann, and Mudambi (2010) suggested that the
major three schools focus on the issues surrounding companies and
countries. In fact, traditional IB approaches do not delve into spatial
issues associated with geographic locations. Hence, this paper
concludes the following on the country level:
1) International trade theories: These theories consider companies
as the synonymof market structures and to a large extent ignore
many important organizational and spatial features and differ-
ences related to MNEs.
2) Economic geography and location study: One of the obvious
shortcomings of traditional economic geography and location
studies is ignorance in the complex multi-plant productions of
different countries and spatial behavior of MNEs. Few discus-
sions or analyses venture behind the description of key theories
(McCann & Mudambi, 2004, 2005).
3) IB and strategic literature: Dunning (1998, 2009) conducted a
precise analysis on multi-plant production and MNE activities,
and indicated that with few exceptions, most scholars in strat-
egy and IB often overlook location as a factor. In other words,
studies on IB and strategies are relatively unsophisticated in
spatial/geographic expressions (McCann & Mudambi, 2005).
However, a major change occurred about twenty years ago
(McCann & Mudambi, 2004, 2005), when three scholars started a
new era of possibilities from different domains: Scott (1988) in the
traditional economic geography, Porter (1998) in strategy, and
Krugman (1991) in trade theories. Their differing perspectives may
affect analysis of the other two domains. In particular, the research
scope of economic geography has rapidly expanded after long-
term studies by Krugman (1991). There has been mushrooming
perspectives developed from their literature, as scholars from
different traditions offered their cooperation with other disci-
plines, and International economics attempted to incorporate the
trade model of MNEs (Barba-Navaretti & Venables, 2004;
Markusen, 2002).
2.1. Place, space, and organization vs. IB
Location choices were dealt with by connecting IB with OLI
theories. Such connections examine the relationship between
companies and countries. Therefore, by lowering the focus fromthe
country level to the regional level, and incorporating the regional
effects on time and space, it is necessary to factor in the perspec-
tives of place, space, and organization (PSO).
From OLI to PSO, the research issues formed in the early 1990s
developed in two directions. The IB perspective is based on how
companies make their choices between exports and foreign direct
investment (FDI). This extends to various FDI strategies, including
horizontal options (the pursuit of markets) and vertical options (the
pursuit of ef?ciency) (Aw & Lee, 2008). Subsequently, the addition
of company-level heterogeneity into empirical models gained rapid
popularity. Ghironi and Melitz (2005) provided insight into the
change between trade goods and trade ?ows between countries.
They used micro-level statistics on exports to examine the rela-
tionship among trading cost reductions, companies, and geogra-
phies (Bernard et al., 2003; Tybout, 2003). In sum, IB mainly focuses
on the study of company heterogeneity and clustering.
IB and economic geography have interacted in PSO studies in
recent years (Beugelsdijk et al., 2010). The six major research ?elds
are clusters and entrepreneurship, spillovers, industry FDI and
regional developments, outsourcing and innovation, organiza-
tional geography, experience learning, company clusters, and
networks. This paper argues that in the context of spatial het-
erogeneity (rather than companies), the focus should be on (1)
spillover effects; (2) clusters, networks, and companies; and (3)
industry FDI and regional developments. These three domains can
be viewed in combination with country/geography characteristics
in the discussion of interactions between MNEs and spatial
environments.
2.2. Local Institutional Theory
Allen and Kendall (1992) proposed that the regional strategies of
MNEs can be viewed as alternatives to globalization. For example,
MNEs from the United States are known for empowerment;
whereas, MNEs from Japan centralize decisions and control at the
parent companies. When MNEs from the United States seek to
imitate the strategies of their Japanese peers, they often struggle
and become stuck in the middle. This is how the eclectic regional
strategy comes into play.
H.-W. Hsu, Y.-L. Jaw / Asia Paci?c Management Review 20 (2015) 252e264 253
The advantages of regional strategies are as follows:
1) It is not easy to achieve economies of scale on the global basis. It
is, however, relatively easy to attain economies of scale on the
regional basis.
2) It is easier to achieve time to delivery, customization, and low
inventory with regionalization.
3) Regional strategies can strike a balance between the requests
from the parent company and the need for subsidies to realize
their potential.
The system view holds that multinationals are simultaneously
embedded into the internal environment of the parent and the
external environment of the host country (Granovetter, 1985).
When a company enters a new overseas market, it is a given that
corporate strategies must be adjusted to the environmental spe-
ci?cs of the system in the host country (Martinez & Dacin, 1999;
Meyer, 2001; Peng, 2001). The systematic structure of the host
country, such as the restriction of foreign ownership percentage,
can sometimes be a hurdle for market entry (Delios & Beamish,
1999; Gomes-Casseres, 1989; North, 1990). However, it can also
work the other way to bene?t companies. Examples are the
exemption of import duties on raw materials, tax reductions, de-
ferrals and holidays, inexpensive labor, high-caliber personnel, and
government subsidies (Table 1).
The ?rst two topics in the above table emphasize the impor-
tance of systems and social environments, and promote the
achievements of knowledge production, via learning, to enhance
regional developments. The last two topics in the above table
explore spatial scales in regional studies. Beehive-like regional
borders and relationships form the theoretic differences in terms of
borderless regions. In sum, the above four approaches center on
industries gaining an understanding of a region as a concept.
In terms of local scale under the country scale, the density of
town clusters is far more important than distance. However, in the
context of country scale, distance outweighs density. The policy
challenge in the former situation is to achieve a reasonable density,
via market mechanisms, that encourage the concentration of eco-
nomic activities, and improve the living standards of villages, towns
and cities in sync. The policy challenge in the latter situation is to
assist ?rms and workers to avoid distance problems as a result of
concentration, which can be achieved via investments in in-
frastructures, to reduce transportation costs and enhance labor
liquidity.
The rapid rise of economic power in China has led to two trends.
First, MNEs have worked to establish their footprint, from Tier 1
and Tier 2 cities in the coastal area to Tier 3 to Tier 5 cities, and deal
with the challenge of localization. Secondly, the aggressive moves
from MNEs into the domestic market in China are forcing local
businesses in China toward globalization.
However, in the hierarchy of globe/countries to regions and
cities, these studies indicate varying mosaic features. In fact, in
terms of cities, it should be a vertical linkage of OLI to PSO and then
to cities, which places the social capitalism of cities into the spot-
light. There are different forms of capital: economic, cultural, and
social. Similar to economic capital, social capital can be accumu-
lated or exploited, as it is a type of wealth. Cities are in essence the
product of uneven geographic developments. Competition is
ubiquitous. In the era of global capitalism, global competition has
become a force never seen. During the past two decades, the roles
played by different localities in different competitions have led to
various forms and shapes of institutions, such as state, regional,
metropolitan, urban, and local governments, in the global eco-
nomic landscape. They resort to corporate-style governance stra-
tegies to develop local competitive advantages in order to enhance
structural competitiveness, attract foreign investments, and ensure
the accumulation of economic developments in their jurisdictions.
2.3. Hypothesis development
In recent years, the success of the clustering phenomenon raised
the concern of many scholars. The geological proximity also has
positive effect on manufacturer innovative performance (Audretsch
& Feldman, 2003; Feldman & Florida, 1994; Jaffe, Trajtenberg, &
Henderson, 1993). In general, earlier research literature focused
on consumers, sellers, and the space costs of natural distribution
(Porter, 1998; Saxenian, 1994). In recent years, many research in-
vestigations have found that, geography proximity and manufac-
turers in adjacent geological locations have advantages of lower
costs. Furthermore, the innovative activities of some areas in the
geological space are prominent and will promote regional indus-
trial growth (Cannon & Homburg, 2001).
2.3.1. Human Capital Relative Risk (HCRR) concept
De Long (1988) considered that human capital is the most
importance in?uential factor in regional analysis, thus, most
regional convergence research incorporates variable human capital.
Barro (1991) added the variable of human capital (enrollment rate
as a proxy) to the data from 98 countries from 1960 to 1985, and
signi?cant negative correlation exists between the level of the
primary areas and the economic growth rate. Mankiw et al. (1992)
followed the prediction of the neoclassic growth theory after con-
trolling the human capital, the investment ratio, and population
growth.
The empirical studies by Simon and Nardinelli (1996, 2002)
indicated that human capital has signi?cantly positive long-term
contribution to regional economic growth, and found that the
growth impetus of UK urban cities between 1861 and 1961 was not
from natural resources or large sized manufacturing, but bene?ted
from new knowledge gained by discussions between employees as
high human capital; cities and metropolitan districts with more
human capital in the early 20th century have higher economic
growth rate in the following century; in addition, employees
graduated from universities have higher positive in?uence on
Table 1
Evolution of regional studies.
Topic Content
Cluster study ? Focus on industry developments post Fordism. Highlight on how economic costs and non-trade interdependence encourage regional clustering
of companies and constitute networks of ?rms and suppliers into a new industry space.
Institutionalism ? Focus on non-economic infrastructure. Key concepts: learning regions, associational economy, and institutional thickness. Improvement
of regional competitiveness via upgrading systems and infrastructure.
Scalarities ? Multilevel scale analysis of international politics and economic relationships, from international organizations (i.e., WTO), multinational
organizations (i.e., European Union), countries to cities, to establish an analytical framework for global politics and economics.
Relational
geography
? Regions as a complex and endless network extended from social relationships, rather physical borders. Regions de?ned as the location in
a spatial network consisting of goods, technologies, knowledge, personnel, ?nance, and information ?ows and cycles.
Source: MacLeod and Jones (2007).
H.-W. Hsu, Y.-L. Jaw / Asia Paci?c Management Review 20 (2015) 252e264 254
employees graduated from high schools (Simon, 1998; Simon &
Nardinelli, 1996, 2002).
Shapiro (2006) further found that, 0.8 of employment growth
rate is resulted from every increase of 10% of the population with a
bachelor degree in US metropolitan areas from 1940 to 1990.
However, Shapiro (2006) found no evidence to indicate a growth
effect for high school graduates. Rauch (1993) found that areas with
more highly educated workforces would result in more positive
externalities of more human capital, which can improve regional
productivity and attract more talents. Florida (2002) investigated
50 metropolitan areas, and found that a high percentage of bach-
elor degrees is helpful for the concentration of high-tech enter-
prises and growth of income per head.
2.3.2. Relationship of infrastructure, labor cost, HCRR and
organizational performance
Markusen (1995) showed that multinational companies (MNCs)
preferred to FDI than export when the tariffs or transportation costs
of large foreign markets are higher. This study suggests this implies
that costs are relatively lower when the size of the market is larger.
MNCs may have the cluster effect, which may form hotspots. From
the perspective of economic geography, the driver of globalization
is both technologies and system mechanism arrangements, which
result in free mobility of funds, ?nances and freight services, but
does not fall into the technological determinism trap (Castells,
1996; Harvey, 1989). Rivoli and Salorio (1996) further indicated
that FDI is inevitable if manufacturers have high-level special lo-
cations—internalized advantages of transactions. However, in-
vestments are affected by uncertainty, investment delay, and
withdrawal. As investment is not necessarily completed the ?rst
time, we can select and wait for more suitable investment times.
Chen (1996) and Balasubramanyam, Salisu, and Sapsford (1996)
indicated that local economic development may be impaired if
the education, infrastructure, and other preconditions of the host
country are not adequate. Larger markets have higher pro?ts and
economies of scale in production. In addition, the ?xed costs of
investing in a foreign country mean adapting to environmental
factors, such as: laws, culture, and language, and such costs and
sales offset each other.
Manufacturers rely heavily on supporting infrastructures, such
as: roads, airports, and telecommunications, in order to maintain
trade with suppliers and customers, and such business activities
(?rms' activities) form interactions. Transportation infrastructure
services can be regarded as the proxy of a business support
infrastructure. In economic development, common comparisons
are made between public expenditures for highways and the
expenditures of other infrastructure (Fisher, 1997). A well-
developed transportation infrastructure allows businesses to
adopt economic and ef?cient production and marketing
strategies.
Cheng and Stough (2006) investigated the factors that attract
Japanese investors to invest in mainland China from 1997 to 2002,
and found that market size, infrastructure, land, labor costs, labor
quality, etc., are the main factors attracting Japanese businesses to
direct foreign investment in mainland China. Buckley (2004)
considered that, for many multinational companies, China is a
country with low wages and high productivity, and thus, it can
attract market-orientated, cost-orientated, and input-orientated
FDI. Areas with lower labor wages can shorten the time of wait-
ing for foreign investors. Namely, investment by manufacturers
has higher probability if the host country has lower labor wages
(Rivoli & Salorio, 1996). During the early reform era of China, the
manufacturing areas had unbalanced distribution, and the eastern
coastal areas had more advantages than other areas. Between
1980 and 2001, the eastern coastal manufacturing areas in the
three northeastern provinces and western areas shrunk relatively,
especially the Yangtze River Delta, Pearl River Delta, and Bohai
Gulf. Due to the industrial cluster of the eastern coastal areas, the
total output may increase. The cluster effect may increase FDI, as
well as the immigration work force in the middle and western
villages, thus, the expansion of regional difference in China is
closely related to its industrial clusters. Jenkins and Tallman
(2010) indicated geographical proximity, and implied assump-
tions of knowledge ?ow between remote clusters. Therefore, it is
inferred that this region or city cluster may come from social
capital.
H1a. HCRR intermediates relation between Infrastructure and
Organizational Performance
H1b. HCRR intermediates relation between Labor Cost and
Organizational Performance
2.3.3. Relationship of infrastructure, labor cost, entry mode and
organizational performance
There are many previous studies on the entry mode. The
empirical study by Chowdhury (1992) showed sole proprietor-
ships have a higher percentage of export sales and better business
performance than joint ventures. However, for the market exit
rate, joint ventures have better performance than sole pro-
prietorships. Wilson (1980) and Shaver (1998) suggested that the
survival rate of sole proprietorships is higher than mergers and
acquisitions (M&A) in selecting entry mode. Woodcock, Beamish,
and Makino (1994), and Nitsch, Beamish, and Makino (1996),
used three entry modes: sole proprietorships, joint ventures, and
M&A to discuss the investment of Japanese manufacturers in the
US and Western Europe, and found that sole proprietorship has
the best performance, while M&A has the worst performance.
Anderson and Gatignon (1986) discussed entry mode and business
performance using an entry model with high, medium, and low
degrees of control. The results showed that business performance
is better in the entry mode with a high degree of control. Minor,
Wu, and Choi (1991) divided business performance into long-
term and short-term business performance, and found that the
entry model with a low degree of control has better short-term
performance, while the entry model with a high degree of con-
trol has better long-term performance.
Regarding the material level, manufacturers can form front
and back industrial correlation through intermediate goods;
regarding the non-material level, as knowledge is public, shared,
and non-excludable, when manufacturers create new knowl-
edge, they cannot retain the new knowledge within itself.
Meaning knowledge will spill out, and other manufacturers can
bene?t (Capello, 1994), thus, spillover effects are generated. From
the above, frequent liaison necessarily exists between manufac-
turers. As manufacturers are not always in the same geological
space, regional interdependence exists, and manufacturers rely
heavily on supporting facilities, such as roads, airports, and
telecommunications to maintain trade with suppliers and cus-
tomers. These business activities interact with each other. The
transport infrastructure can be regarded as a proxy of their
business infrastructure. In economic development, common
comparisons are made between the public expenditures of
highways and the expenditures of other infrastructure (Fisher,
1997). A well-developed transportation infrastructure allows
businesses to adopt economic and ef?cient production and
marketing strategies. Therefore, this paper proposes the hy-
pothesis as follows:
H2. Entry Mode intermediate relation between Infrastructure and
Organizational Performance
H.-W. Hsu, Y.-L. Jaw / Asia Paci?c Management Review 20 (2015) 252e264 255
2.3.4. Relationship of HCRR, Organizational Slack and
Organizational Performance
The Resource-Based View (RBV) emphasizes resource hetero-
geneity. However, Foss and Foss (2005) observed that empirical
papers prefer to use the production function to establish relation-
ships between enterprise resources and performance, and adopt an
input base to measure resource reserves. Thus, the greater the input
amount, the higher the book value, meaning resource reserves are
high. Due to continuous accumulation of resources, misbalance
occurs between enterprise resources and activities, and in such
misbalance, organizational slack resources (surplus resources) can
emerge at the right moment. As slack resources exist in an
incomplete market, organizations must continue to pursue growth
in order to develop economic ef?ciency of resources. Therefore,
growth decision making of an enterprise depends on its owned
resources (Mishina, Pollock, & Porac, 2004; Penrose, 1959; Rugman
& Verbeke, 2002).
Penrose (1959) suggested that enterprise growth is an evolu-
tionary process of exploring and using resources. When enterprise
resources are fully and effectively utilized, the enterprise can ach-
ieve growth, and thus, enter different growth phases. A newgrowth
phase can produce newslack resources, which allows enterprises to
continue to pursue growth. In order to use slack resources, enter-
prises often adopt a growth strategy that pursues economies of
scale or scope. Moreover, slack resources are a safety mechanism
against environmental uncertainty. This mechanism can reduce
failure probability and increase the ?exibility of enterprises in
response to environment changes when enterprises adopt new
strategies and launch in new markets (Fuentelsaz, G omez, & Polo,
2002). Slack resources can be used as backup support for the
growth strategy, allowing enterprises to pursue growth.
In the context of globalization, in addition to low production
costs and preferential measures, what method can be used for local
development? Some scholars believe in the local clusters of the
global production network, where local internal characteristics,
especially speci?c institutional conditions, such as local in-
stitutions, sharing rules, habits, and knowledge intensity, can pro-
vide different methods for local development. Localization and
globalization are interwoven during institutionalization (Amin &
Thrift, 1994), meaning local structure is initiated as the response
to globalization. Local economic development and corporate
strategy re?ect entrepreneurship and market power, as well as the
economic development track created by the ?exible intuitional
strategies deployed by local participants, such as institutional in-
frastructures based on suitable organizational support structure
and local mosaic social capital. Knowledge creates networks, trust,
and connection, which supplement human resources and hardware
construction investment. Such “relational assets” are the core of
institutional thickness, and are regarded as key to local economic
development. In other words, the dimension of “institutional turn”
understands that institution is the factor to explain local economic
development in the real world.
Therefore, this “?eld domain” research collects detailed empir-
ical evidence in order to recognize the regional change process, the
nature of spatial differences, and the causes and results. Regarding
economic geography, the institutional turn can be further extended
to the institutional research method of explaining different local
economic life. The institution, as the intermediate of shaping
different local economy development tracks and products, can be
understood as a method to produce and exhibit spatial differences
in economic development. From another point of view, geograph-
ical dimension may become an important factor for explaining
institutional change. Therefore, this paper proposes the hypothesis
as follows:
H3. HCRR interferes relation between Organizational Slack and
Organizational Performance
3. Methodology
To discuss the issues of social capitalism in geographic space,
sociologists Merton and Lazarsfeld (1950) highlighted the impor-
tance of multilevel analyses, and argued that macro systems affect
micro processes. Woodman, Sawyer, and Grif?n (1993) suggested
an interactive model of organizational innovations and believed
that individual innovations are subject to the in?uence of individ-
ual and system factors. All such theoretic discourses indicate the
multilevel characteristics of organizations. However, varying fo-
cuses or domains result in different issues under different spatial
scales in different localities.
This study sampled the outward investments of Taiwan to China.
Data suggest a total of 3752 investments from listed companies in
Taiwan. The entries with unspeci?ed addresses and outliers (before
1979) are eliminated. According to the data of the Ministry of
Transportation & Communications and the Ministry of Finance,
there are a total of 35,823 outward investments from individuals
and small and medium enterprises (SMEs) from Taiwan to China in
2000~2009. In this study, a total of 18,067 entries are selected after
the removal of those with missing values.
According to Filatotchev, Strange, Piesse, and Lien (2007),
Taiwanese businesses refer to the data of the prior year for their
strategic selections of investments in China. When manufacturers
evaluate investment proposals, they visit the destination sites in
advance for due diligence. Therefore, this paper refers to the
assumption of Filatotchev et al. (2007) and samples the regional
data in 1999~2008 of municipalities and prefectural-level cities in
China. Based on the literature of the last section, the development
research framework is shown in the following Fig. 1.
3.1. Spatial level variable
3.1.1. Human Capital Relative Risk measure
This study used SaTScan software(SSS) to measure clusters and
relative risks. SSS is de?ned and developed by Kulldorff and
Nagarwalla (1995) to scan and measure spatial and time statistics.
Its applications include: (1) locality monitoring to detect whether
there is signi?cant clustering in time or space; (2) veri?cation of
whether speci?c distributions in time, space, or both are random;
(3) evaluation of whether clusters are statistically signi?cant; (4)
regular monitoring to identify clusters as early as possible.
SSS scans moving windows (which represent potential clusters)
over time or in space, and maintains a detailed record of the actual
values and expected values of all the windows. Researchers can
de?ne the maximum range of time and space windows and refer to
likelihood ratios or relative risks (RR) to examine whether the
noti?cation ratio within moving windows is higher than that
outside moving windows. If the ratio within the moving windows is
higher than the outside, it is de?ned as clustering. The location of
the cluster is then used to identify the reasons for such clustering.
The regional data of concern is calculated with statistical methods,
where p values are the basis for the identi?cation of the most
signi?cant range of clustering (Lawson & Waller, 1996).
Bernoulli and Poisson model are different random samples. The
null hypothesis (H0) is that a survival time distribution exceeds the
boundary of space. The actual distribution may be unclear or even
varied. To solve this problem, it is possible to randomly observe
conditional survival times and infer the replacement geographic
coordinates and survival times. The sign 0/1 is converted into
H.-W. Hsu, Y.-L. Jaw / Asia Paci?c Management Review 20 (2015) 252e264 256
corresponding time. This random process ensures the accuracy of p
values.
The height of purely spatial scan statistics re?ects potential
clustering over a period of time. In every possible time and space, it
is necessary to examine the possible geographic location and size.
In fact, regions of varying sizes and shapes can be obtained to cover
possible clustering throughout the entire research area. The same
method can be applied to other variables such as gender, race, or
education, as well as adjusted variables on the regional level
(Kulldorff & Feuer, 1997).
Many studies refer to education as a variable to measure human
capital. Hence, this paper uses investments in education in mu-
nicipalities and prefectural-level cities as the proxy variable to
measure whether HCRR (Human Capital Relative Risk) in a given
city exhibits spillover, and examine how such spillover in?uence
FDIs fromTaiwanese businesses. This paper de?nes RR as the risk of
over investment in education in a city.
Fig. 2 shows that relative risks in the region tend to cluster into a
common RR. This spatial heterogeneity causes co-movement ef-
fects among neighboring cities (Fig. 3).
3.1.2. Other spatial variables
Transportation can be deemed as the most important business
support infrastructure. This paper refers to public buses for pas-
sengers and cargo, as well as the volume of postal services, as the
proxy variable. Principal components analysis is performed to
extract the overall indicators of the infrastructure (Table 2, Fig. 4).
The majority of operation premises set up by Taiwanese busi-
nesses are in the proximity of highways. Taylor (2004) argued that
most cities are between ordinary cities and world cities, and thus,
they are in different competitive positions. Other scholars have
mentioned that the rapid decline of transportation and communi-
cation costs has resulted in an unprecedented imbalance between
the regions and cities in the developing world. Meanwhile, the gap
continues to widen between regions in the industrialized world.
The mobility of capital and labor, contrary to conventional wisdom,
contributes to the concentration of productions rather than an even
distribution of activities (Sunley, 2003). Therefore, this paper refers
to the adjusted overall indicator of the infrastructure as a proxy
operational variable.
During the past decade, major cities in China have witnessed a
shortage of labour. Cheng and Kwan (2000) suggested that labor
costs are the most important consideration for outward investments
by MNEs. Labor costs are mostly measured with average wages
(Coughlin, Terza, & Arromdee, 1991; Sun, Tong, & Yu, 2002;
Lansbury, Pain, & Smidkova, 1996a, 1996b). In the spatial context,
high wages in a city mean relative better access to a high-quality
labor force. Therefore, this paper uses the average monetary wages
in a city in China as the proxy variable. Average monetary wages
refer to the average monetary compensations that an employee re-
ceives in companies, enterprises, or government agencies. They
represent the level of labor incomes during a certain period.
3.2. Firms level measure
Voss, Sirdeshmukh, and Voss (2008) argued that physical re-
sources and human resources are the operational resources of
companies. Any excess of operational resources indicate excess
capacity. To enhance resource utilization ef?ciency, managers
actively seek growth. Meanwhile, if excess capacities emerge due to
market saturation, demand declines, competitive pressure in-
creases, and product lines become stale. Hence, companies must
expand product portfolios and market shares, or diversify opera-
tions, in order to enhance the probability of survival (Seth, Song, &
Pettit, 2002). The more excess capacity in the form of tangible re-
sources, the higher the pressure to pursue growth. Therefore, this
paper refers to the internal capital of Taiwanese businesses, i.e., an
organizational slack for investment projects, as a proxy variable.
Finally, this paper uses book values at the end of investment as
the performance measurement. The stake percentage of Chinese
subsidiaries held by parent companies is the proxy model for the
entry model.
3.3. Control variables
This paper controls one variable on each of the two levels. The
variable “investment horizon” is controlled on Level 1. This oper-
ational variable is based on the year 2000 (coding ¼ 0), and
calculation is conducted accordingly for subsequent years. Some
studies treat time as the explanatory variable in order to establish
an understanding of the effect of time horizon on joint ventures.
However, this paper does not explore time as a factor; rather, time
as a variable is controlled in order to eliminate the possible prob-
lems associated with autocorrelation of time series. The variable
“actual amount of FDI during the year” is controlled on Level 2. It is
found that foreign investments promote local economies. Whether
the government can properly utilize foreign investments is
considered part of governmental effectiveness. The following offers
an explanation of why FDI is a control variable on Level 2, in the
context of spatial autocorrelation (Fig. 5).
Fig. 1. Research framework.
H.-W. Hsu, Y.-L. Jaw / Asia Paci?c Management Review 20 (2015) 252e264 257
Fig. 2. RR cluster.
Fig. 3. RR Cluster of Taiwanese businesses in China.
H.-W. Hsu, Y.-L. Jaw / Asia Paci?c Management Review 20 (2015) 252e264 258
Table 2
Principle component analysis result.
Proxy Communalities Extraction sums of squared loadings
Before adjustments Volume of passenger transportation 71.3% 49.82%
Volume of cargo transportation 76.7%
Volume of postal services 63.2%
After adjustments Volume of passenger transportation 81.11% 65.788%
Volume of cargo transportation 81.11%
Fig. 4. FDI points in the proximity of highways.
Fig. 5. FDI/Census of China.
H.-W. Hsu, Y.-L. Jaw / Asia Paci?c Management Review 20 (2015) 252e264 259
“Spatial autocorrelation” and “spatial randomness” are two
contrasting concepts. Without human or natural intervention, the
spatial distribution of a given variable should be random. However,
regarding social phenomena, spatial distributions are rarely
random due to signi?cant differences in geographic and environ-
mental resources.
Global Moran, as a general phenomenon, refers to the existence
of correlation of one performance and a neighbor's performance. In
fact, this general trend is an aggregation of individual relationships.
In other words, an individual location has an interactive relation-
ship with surrounding locations. Local Moran I is often used as an
indicator to measure this relationship, and is de?ned as follows:
I
i
¼
Y
i
ÀY
P
n
i¼1
À
Y
i
ÀY
Á
2
X
n
j¼1
W
ij
À
Y
i
ÀY
Á
where n is the number of locations, Wij is the element value of the
ieth row and the j-th volume in neighboring matrix W, W is the
standardized row, Y is the variable, and Y-bar is the mean of Y. If the
value is positive, it indicates positive correlation, while a negative
value indicates negative correlation (Fig. 6).
4. Results
Hierarchical linear modeling (HLM) is conducted to validate the
research hypotheses. According to Bryk and Raudenbush (1992),
grand mean centering is performed on the variable of Level 1.
This paper refers to performances as the outcome variable. In-
vestment performances are factored into the null model for the
calculation of interclass correlation coef?cients (ICC
(1)
), in order to
con?rm the necessity for multilevel analysis (Bryk & Raudenbush,
1992). The ICC of the investment performances in this paper is
0.074 (P < .01). Cohen (1988) suggested that if ICC is smaller than
0.059, indicating a low level of intra-class correlation. If it is be-
tween 0.059 and 0.138, it indicates a medium level of correlation. If
the value is higher than 0.138, it indicates a high level of intra-class
correlation. He supposed that a medium level of intra-class corre-
lation means the possible presence of similarities. Therefore, if ICC
is greater than 0.059, it is necessary to consider multilevel analysis;
in other words, HLM should be used rather than GLM (generalized
linear modeling). This paper continues to test the research hy-
potheses with HLM.
4.1. Main effect and mediation effect
To test the research hypotheses, Baron and Kenny (1986) argued
that it is necessary to ?rst con?rm whether there is a signi?cant
correlation between independent variables (X; Mo; Me) and
dependent variables in the examination of mediating and inter-
vening effects. Table 3 shows that after the time factor is controlled,
both variables Organization Slack (b ¼ 0.063; P < .01) and Entry
Mode (b ¼ 3.211; P < .01) are statistically signi?cant. Follow-up
steps are then performed to test the hypotheses.
In terms of main effects on Level 2, this paper controls the time
effect and FDI effect. The results suggest that Infrastructure as a
variable is not statistically signi?cant (b ¼ 0.7677; P > .05). How-
ever, Labor Cost (b ¼50.713; P < .05) and HCRR (b ¼À10.113; P < .1)
are both statistically signi?cant (Table 4).
Baron and Kenny (1986) suggested that to explore mediating
relationships, it is necessary to clarify the relationship among X,
Mo, and Me, in Step 2. However, it is not possible to de?ne a Level 2
variable as a dependent variable in the HLMprocess. Therefore, this
paper refers to the results summarized in Table 5 to explain the X-
> Me relationship. M2-1 shows a signi?cant and negative correla-
tion between HCRR and Infrastructure. M2-3 indicates a signi?cant
and positive correlation between Entry Mode and Labor-cost.
4.2. Cross-level mediation and moderated effect
The two sets of mediating effects in this paper are multi-level in
nature. Models 3-1, 3-2, and 3-3 test multi-level mediating effects,
and the results show that HCRR is not statistically signi?cant
(b ¼À7.938; P > .05). The value for Labor Cost declines from50.713
to 47.380872, which remains signi?cant (P < .05). This paper uses
the mediating model developed by Baron and Kenny (1986). HCRR
mediates the relationship among Infrastructure, Labor Cost, and
Performance. Prior studies suggest that if X to Y is not statistically
signi?cant, it is still possible to continue the procedures (Kenny
et al., 1998; Shrout & Bolger, 2002). Therefore, it is possible to
evaluate the relationship with correlation coef?cients. Meanwhile,
HCRR as a variable does not mediate the relationship between
Infrastructure and Labor Cost to Performance. However, Infra-
structure (g¼À0.197**), Labor Cost (g¼À0.290**) is the antecedent
of HCRR, while the correlation between HCRR and Performance is
negative (b ¼ À10.113173*). In theory, it is still possible to infer a
mediating relationship. Thus, H1a and H1b are supported.
Model 3-4 shows the relationship among Infrastructure, Entry
Mode, and Performance. Entry Mode is statistically signi?cant
(b ¼ 3.283; P < .01). The coef?cient of Infrastructure drops from
0.768 to À2.836, which is not statistically signi?cant (P > .05).
Model 1-4 suggests that Infrastructure as an independent variable
Fig. 6. LISA for Moran's I.
Table 3
Level-1 main effect.
M1-1 M1-2
Intercept 369.056*** 381.068***
Control
Time 38.403*** 39.432***
FDI
Level-1
Organization slack 0.063***
Entry mode 3.211***
Deviance 303244.777 303111.063
H.-W. Hsu, Y.-L. Jaw / Asia Paci?c Management Review 20 (2015) 252e264 260
becomes insigni?cant, which is not consistent with the testing
procedures designed by Baron and Kenny (1986). Thus, H2 is
partially supported.
According to the testing procedures designed by Hofmann
(1997) for intervening effects in HLM, the ?rst step is to validate
the signi?cance of the Null Model. In the Random Coef?cient
Regression Model, Organization Slack (b ¼ 0.034; P < .05) is sta-
tistically signi?cant. In the Fixed Effects Model, Moderated Effect is
statistically signi?cant (b ¼ 0.030; P < .01). Thus, H3 is signi?cantly
supported (Fig. 7).
As seen above, HCRR causes intervening effects to Organization
Slack and Performance. In other words, the HCRR of a given city
undermines Organization Slack and brings about Performance
(Table 6).
5. Conclusions
Dunning (1980, 1981) developed the Eclectic Theory, and
argued that location advantages refer to politics, cultures, pro-
duction costs, and wage differences, of host countries. Companies
pro?t from overseas operations as a result of location advantages.
However, location advantage is a static concept in the Eclectic
Theory. Thus, this study used SSS to explore the dynamic
dimension.
5.1. Discussions
In the context of globalization, what are the answers to local
developments, other than cheap production costs and in-
centives? Some studies have argued that, in terms of local
clusters in global production networks, institutional qualities
(e.g., local systems for knowledge sharing) provide alternatives
to the drive of local development. Localization and globalization
are intertwined in the process of systemization. Local structures
respond to globalization and become part of globalization in the
process (Amin & Thrift, 1994). Local economic developments
and corporate strategies not only re?ect entrepreneurship and
market powers, but also echo local tracks and take the form of
?exible systems and strategies to push economic development.
For example, appropriately organized and supporting structures,
which are locally embedded and constituent social capitalism.
Institutional infrastructures, such as knowledge creation net-
works, trust, and mutual communications, supplement human
resources and investments in facilities and infrastructures.
These relational assets are the core of system intensity and the
key to local economic development. In the real world, institu-
tional shifts can be interpreted as how systems drive local
economies.
This study integrated regional institutional theories and TCE.
However, as the TCE theory can only be tested in a qualitative
manner, the testing of actions taken by companies is based on
theoretic descriptions as the correlation between entry models and
performances (Brouthers, Brouthers, & Werner, 2003). In fact, as it
is costly and time-consuming to build networks in local markets,
joint ventures are effective means to enter new markets with
limited investment. MNEs mostly intend to set up overseas sub-
sidiaries and work with local companies in order to reduce the
political risks of host countries. This means that joint ventures
remain the second best choice, next to sole ownership (Hennart,
1988). When it is dif?cult to gauge expected asset values in
different industries, joint ventures help to collect information and
mitigate the problems associated with an insuf?cient knowledge of
assets (Balakrishnan & Koza, 1993).
The pro?tability of outward investments to new markets is
generally lower than that of existing players in the same markets;
however, the learning effects narrow this gap over time. Mean-
while, most investments cannot be completely pulled back;
hence, companies do not make immediate investments. (If in-
vestments are totally irrevocably, it will not be necessary to wait
Table 4
Level-2 main effect.
M1-3 M1-4 M1-5 M1-6
Intercept 501.496*** 500.906*** 283.987*** 519.430***
Control
Time 38.789* 38.774*** 40.658*** 37.785***
FDI À28.292*** À28.209** À37.3099** À28.195*
Level-1
Organization Slack
Entry Mode
Level-2
Infrastructure 0.7677
Labor Cost 50.7129**
HCRR À10.113*
Deviance 303233.9854 303223.595 303221.218 303224.565
Table 5
Relationship among X, Mo, and Me.
Dependent variable
M2-1 M2-2 M2-3
HCRR Organization slack Entry mode
Level-2
Infrastructure À0.197**
Labor Cost 0.035 0.144
*
Fig. 7. Moderated effect.
H.-W. Hsu, Y.-L. Jaw / Asia Paci?c Management Review 20 (2015) 252e264 261
for the optimal timing for investments). If the pull-back costs (i.e.,
invested costs are high) or a pull-back is impossible or will be
detrimental to reputations, companies are more likely to wait for
the right time. The uncertainty and a lack of complete irrevoca-
bility are often present for investment projects; hence, in-
vestments are often delayed; or else, the investments are carried
out immediately (such as in the case of expiration of the approved
period or patent rights). Therefore, Organization Slack is one of
the key considerations for Taiwanese companies
(b ¼ 0.063361***). HCRR intervenes in Organization Slack and
Performance. The ?gure on intervening effects shows that HCRR
enhances Organization Slack Strategy and enhances Performance.
A high local HCRR indicates longer investment cycles, and better
operational performances.
5.2. Management implications
The theoretic implications of this paper are identical to the
speci?c assets mentioned in the transaction cost theory. Local
infrastructures and human resources in China affect the choices
made by the control mechanisms of joint ventures set up by MNEs
(David & Han, 2004). With faced with high uncertainties in
emerging markets, MNEs emphasize the importance of institu-
tional environments. DiMaggio and Powell (1983) stressed the
importance of enforceability, regulations, and imitability. This
study focused on the multi-level analysis of relative risks to
Taiwanese businesses in the context of local social capital, and
explored the behavior differences as a result of entry models and
performances.
This study treated average wages as a proxy for Labor Cost.
Rising wages are inevitable for labor-intensive tech manufacturers
in China. Wages are implicitly re?ective of local conditions. The
lower the environmental uncertainties (compared with cities of
superior economic/political setups and infrastructure), the more
comfortable Taiwanese companies will be to increase investments
for better performances. However, rising wages will pose a chal-
lenge to managerial capability. If revenues do not increase
accordingly, or costs cannot be reduced to offset rising wages via
relocations or automation, pro?ts will shrink.
In practice, the shortage of labour is a prevalent problem in
China. Entry-level workers are not loyal to employers and easily
switch to jobs that offer higher pay. Therefore, Labour Cost in
different cities is also an implicit indicator of the average wages
local companies are willing to pay. Companies that are willing to
afford a higher labor cost are less likely to meet labor shortage, thus
ensuring better operational performances.
Regarding city-level factors, the wide gap between major cities
and remote towns is unique to China. Better-educated talents in
remote locations often seek job opportunities by moving to major
cities. This study used the educational expenses of local cities as a
proxy variable; however, these numbers do not re?ect the number
of employees from other locations.
5.3. Contributions
Regarding increased risks, strategic forces and IB lengthen the
distance between the host country and the home country, ac-
cording to the Uppsala School of Internationalization. However,
simply tackling strategic forces and IB may be insuf?cient regarding
studies on geographic regions and connecting economics, due to a
lack of speci?cs. Fewstudies have examined the spatial interactions
between companies in the investigation of inner company factors
and country/geographic characteristics (Beugelsdijk, 2007).
This study thus performed a spatial analysis in the empirical
studies of Taiwanese companies in mainland China, and achieved
the following three academic contributions:
1) Research methodology: This study visualized geographic infor-
mation and made it easy to understand the geographic distri-
bution investments of Taiwanese businesses. However, spatial
auto-correlation violates the assumptions that samples are in-
dependent in the context of quantitative geography, which
renders the traditional approach of econometrics invalid.
Therefore, HLM was used to resolve erroneous assumptions in
the conventional OLS approach.
2) Theoretic contributions: The concept of relative risk stems from
the domain of public health. Traditionally, IB or strategic studies
do not explore the issues associated with relative risks. How-
ever, the clustering of relative risks is a fact. Hence, this study
incorporated this concept in the explanation of the effects on
the performance of Taiwanese joint-ventures.
3) Practical contributions: Taiwanese businesses exhibit a speci?c
pattern in their outward investments. Their FDIs tend to be
conservative in highly uncertain countries or regions. Initial
investments can be viewed as sunk costs and may not be
recoverable for joint ventures. Therefore, this study included the
concept of organizational slack in the examination of
companies.
Table 6
Cross-level mediation and moderated effect.
M3-1 M3-2 M3-3 M3-4 M4-1 M4-2 M4-3
Intercept 500.906*** 283.987*** 312.5192*** 528.786*** 366.137*** 519.430*** 520.697***
Control
Time 38.774*** 40.658*** 39.728*** 40.470*** 38.186*** 37.785*** 37.4701***
FDI À28.209** À37.3099** À36.792** À30.433** À28.195* À27.521*
Level-1
Organization slack 0.0343** 0.041***
Entry mode 3.2830***
Level-2
Infrastructure 0.7677 1.864 À2.836
Labor cost 50.71286*** 47.381**
HCRR À7.938 À10.113* À13.248**
Organization Slack _(mean) 0.0243
Cross level
HCRR*Organization Slack 0.0295***
Deviance 303223.5953 303221.218 303205.418 303082.3180 303254.387 303224.565 303239.578
H.-W. Hsu, Y.-L. Jaw / Asia Paci?c Management Review 20 (2015) 252e264 262
This study integrated theories concerning transaction costs and
systems, and developed antecedent causes, and intervening and
interference factors, by referring to the perspectives of economic
geography. Despite a lack of qualitative explorations due to the use
of secondary data in the analysis, this study achieved major
breakthroughs in research methodology and theoretic
contributions.
Con?ict of interest
All contributing authors declare no con?icts of interest.
References
Allen, D. L., & Kendall, C. (1992). Explaining differences between qualitative and
quantitative data: a study of chemoprophylaxis during pregnancy. Health Ed-
ucation Quarterly, 19, 41e54.
Amin, A., & Thrift, N. (1994). Living in the global. In A. Amin, & N. Thrift (Eds.),
Globalization, institutions and regional development in Europe (pp. 1e22). New
York, NY: Oxford University Press.
Anderson, E., & Gatignon, H. (1986). Modes of foreign entry: a transaction cost
analysis and proposition. Journal of International Business Studies, 17, 1e26.
Audretsch, D. B., & Feldman, M. P. (2003). Small-Firm strategic research partner-
ships: the case of biotechnology. Technology Analysis & Strategic Management,
15, 273e288.
Aw, B. Y., & Lee, Y. (2008). Firm heterogeneity and location choice of Taiwanese
multinationals. Journal of International Economics, 75(1), 167e179.
Balakrishnan, S., & Koza, M. P. (1993). Information asymmetry adverse selection and
joint-ventures. Journal of Economic Behavior and Organization, 20, 99e117.
Balasubramanyam, V. N., Salisu, M., & Sapsford, D. (1996). Foreign direct investment
and growth in EP and IS countries. Economic Journal Royal Economic Society, 106,
92e105.
Barba-Navaretti, G., & Venables, A. J. (2004). Multinational ?rms in the world econ-
omy. Princeton, NJ: Princeton University Press.
Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in
social psychological research: conceptual, strategic, and statistical consider-
ations. Journal of Personality and Social Psychology, 51, 1173e1182.
Barro, R. J. (1991). Economic growth in a cross section of countries. Quarterly Journal
of Economics, 106(2), 407e443.
Barro, R. J., & Sala-i-Martin, X. (1995). Economic growth. New York, NY: McGraw-Hill
Press.
Baumol, W. J. (1986). Productivity growth, convergence, and welfare: what the
long-run data show. American Economic Review, 76(5), 1072e1085.
Bernard, L., Einspanier, U., Haubrock, S., Hubner, S., Kuhn, W., Lessing, R., …Visser, U.
(2003). Ontologies for intelligent search and semantic translation in spatial data
infrastructures. Photogrammetric-Fernerkundung-GeoInformation, 6, 451e462.
Beugelsdijk, S. (2007). The regional environment and a ?rm's innovative perfor-
mance: a plea for a multilevel interaction approach. Economic Geography, 83(2),
181e199.
Beugelsdijk, S., McCann, P., & Mudambi, R. (2010). Introduction: place, space and
organization-economic geography and the multinational enterprise. Journal of
Economic Geography, 10(4), 485e493.
Boisot, M., & Child, J. (1999). Organizations as adaptive systems in complex envi-
ronments: the case of China. Organization Science, 10, 237e252.
Brouthers, K. D., Brouthers, L. E., & Werner, S. (2003). Transaction cost enhanced
entry mode choices and ?rm performance. Strategic Management Journal, 24,
1239e1248.
Bryk, A. S., & Raudenbush, S. W. (1992). Hierarchical linear models: Applications and
data analysis methods. Newbury Park, CA: Sage.
Buckley, P. J. (2004). The role of China in the global strategy of multinational en-
terprises. Journal of Chinese Economic and Business Studies, 2(1), 1e25.
Buckley, P. J., & Ghauri, P. N. (2004). Globalisation, economic geography and the
strategy of multinational enterprises. Journal of International Business Studies,
35(2), 81e98.
Cannon, J. P., & Homburg, C. (2001). Buyer-supplier relationships and customer ?rm
costs. Journal of Marketing, 65(1), 29e43.
Capello, R. (1994). Spatial economic analysis of telecommunications network exter-
nalities. Aldershot, UK: Avebury.
Castells, M. (1996). The rise of the network society. Oxford, UK: Blackwell.
Chen, C. (1996). Regional determinants of foreign direct investment in Mainland
China. Journal of Economic Studies, 23, 18e30.
Chen, H., & Chen, T. J. (1998). Network linkages and location choice in foreign direct
investment. Journal of International Business Studies, 29, 445e468.
Cheng, L. K., & Kwan, Y. K. (2000). What are the determinants of the location of
foreign direct investment? The Chinese experience. Journal of International
Economics, 51, 379e400.
Cheng, S. M., & Stough, R. R. (2006). Location decisions of Japanese new
manufacturing plants in China: a discrete-choice analysis. Annals of Regional
Science, 40, 369e387.
Chowdhury, J. (1992). Performance of international joint venture and wholly owned
foreign subsidiaries: a comparative perspective. Management International Re-
view, 2, 60e70.
Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.).
Hillsdale, NJ: Eribaum.
Cooke, F. L. (2009). A decade of transformation of HRM in China: a review of
literature and suggestions for future studies. Asia Paci?c Journal of Human Re-
sources, 47, 6e40.
Coughlin, C. C., Terza, J. V., & Arromdee, V. (1991). State characteristics and the
location of foreign direct investment within the United States. The Review of
Economics and Statstics, 73(4), 675e683.
David, R. J., & Han, S. K. (2004). A systematic assessment of the empirical support for
transaction cost economics. Strategic Management Journal, 25(1), 39e58.
De Long, J. B. (1988). Productivity growth, convergence, and welfare: comment. The
American Economic Review, 78(5), 1138e1154.
Delios, A., & Beamish, P. W. (1999). Geographic scope, product diversi?cation and
the corporate performance of Japanese ?rms. Strategic Management Journal, 20,
711e727.
DiMaggio, P. H., & Powell, W. W. (1983). The iron cage revisited: isomorphism and
collective rationality in organizational ?elds. American Sociological Review,
48(2), 147e160.
Dunning, J. H. (1980). Toward an eclectic theory of international production: some
empirical tests. Journal of International Business Studies, 11, 9e31.
Dunning, J. H. (1981). International production and the multinational enterprise.
London, UK: George Allen & Unwin.
Dunning, J. H. (1998). Location and the multinational enterprise: a neglected factor?
Journal of International Business Studies, 29, 45e66.
Dunning, J. H. (2009). Location and the multinational enterprise: a neglected factor?
Journal of International Business Studies, 40(1), 5e19.
Feldman, M. P., & Florida, R. (1994). The geographic sources of innovation: tech-
nological infrastructure and product innovation in the United States. Annals of
the Association of American Geographers, 84, 210e229.
Filatotchev, I., Strange, R., Piesse, J., & Lien, Y. C. (2007). FDI by ?rms from newly
industrialized economies in emerging markets: corporate governance, entry
mode and location. Journal of International Business Studies, 38, 556e572.
Fisher, M. (1997). What is the right supply chain for your product? Harvard Business
Review, 75(2), 105e117.
Florida, R. (2002). The economic geography of talent. Annals of the Association of
American Geographers, 92(4), 743e755.
Foss, K., & Foss, N. J. (2005). Resources and transaction costs: how property rights
economics furthers the resource based view. Strategic Management Journal,
26(6), 541e553.
Fuentelsaz, L., G omez, J., & Polo, Y. (2002). Followers entry timing: evidence from
the Spanish banking sector after deregulation. Strategic Management Journal,
23(3), 245e264.
Ghironi, F., & Melitz, M. (2005). International trade and macroeconomic
dynamics with heterogeneous ?rms. The Quarterly Journal of Economics, 120,
865e915.
Gomes-Casseres, B. (1989). Ownership structures of foreign subsidiaries, theory and
evidence. Journal of Economic Behavior and Organization, 11, 1e25.
Gong, Y., Shenkar, O., Luo, Y., & Nyaw, M. K. (2005). Human resources and inter-
national joint venture performance: a system perspective. Journal of Interna-
tional Business Studies, 36, 489e491.
Granovetter, M. (1985). Economic action and social structure: the problem of
embeddedness. American Journal of Sociology, 91, 481e510.
Harvey, O. J. (1989). Muzafer Sherif (1906e1988). American Psychologist, 44,
1325e1326.
Hennart, J. F. (1988). A transaction cost theory of equity joint venture. Strategic
Management Journal, 9, 361e374.
Hofmann, D. A. (1997). An overview of the logic and rationale of hierarchical linear
models. Journal of Management, 23, 723e744.
Jaffe, A. B., Trajtenberg, M., & Henderson, R. (1993). Geographic localization of
knowledge spillovers as evidenced by patent citations. Quarterly Journal of
Economics, 63, 577e598.
Jenkins, M., & Tallman, S. (2010). The shifting geography of competitive advan-
tage: clusters, networks and ?rms. Journal of Economic Geography, 10(4),
599e618.
Kenny, D. A., Kashy, D. A., & Bolger, N. (1998). Data analysis in social psychology. In
D. Gilbert, S. Fiske, & G. Lindzey (Eds.), Handbook of social psychology (4th ed.,
pp. 233e265). New York, NY: McGraw-Hill.
Krugman, P. (1991). Increasing returns and economic geography. Journal of Political
Economy, 99(3), 483e499.
Kulldorff, M., & Feuer, E. J. (1997). Breast cancer clusters in the northeast United
States: a geographic analysis. American Journal of Epidemiology, 146(2),
161e170.
Kulldorff, M., & Nagarwalla, N. (1995). Spatial disease clusters: detection and
inference. Statistics in Medicine, 14, 799e810.
Lansbury, M., Pain, N., & Smidkova, K. (1996a). Foreign direct investment in Central
Europe since 1990: an econometric study. National Institute Economic Review,
156, 104e113.
Lansbury, M., Pain, N., & Smidkova, K. (1996b). The determinants of foreign direct
investment in Central Europe by OECD countries: an econometric analysis. In
G. Cs aki, G. F oti, & D. Mayes (Eds.), Foreign direct investment and transition: The
case of the Visegr ad Countries (pp. 137e159). Budapest: Institute For World
Economics of the Hungarian Academy of Sciences.
H.-W. Hsu, Y.-L. Jaw / Asia Paci?c Management Review 20 (2015) 252e264 263
Lawson, A. B., & Waller, L. (1996). A review of point pattern methods for spatial
modelling of events around sources of pollution. Environmetrics, 7, 471e488.
MacLeod, G., & Jones, M. (2007). Territorial, scalar, networked, connected: in what
sense a regional world? Regional Studies, 41(9), 1177e1191.
Mankiw, N. G. (1995). The growth of nations. Brookings Papers on Economic Activity,
1, 275e326.
Mankiw, N. G., Romer, D., & Weil, D. N. (1992). A contribution to the empirics of
economic growth. The Quarterly Journal of Economics, 107(2), 407e437.
Markusen, J. R. (1995). Incorporating the multinational enterprise into the theory of
international trade. Journal of Economic Perspectives, 9, 169e189.
Markusen, J. R. (2002). Multinational ?rms and the theory of international trade.
Cambridge, MA: The MIT Press.
Martinez, R. J., & Dacin, M. T. (1999). Ef?ciency motives and normative forces:
combining transactions costs and institutional logic. Journal of Management, 25,
75e96.
McCann, P., & Mudambi, R. (2004). The location behavior of the multinational en-
terprise: some analytical issues. Growth and Change, 35(4), 491e524.
McCann, P., & Mudambi, R. (2005). Analytical differences in the economics of ge-
ography: the case of the multinational ?rm. Environment and Planning A, 37(10),
1857e1876.
Merton, R. K., & Lazarsfeld, P. L. (1950). Continuities in social research: Studies in the
scope and method of “The American Soldier”. Glencoe, IL: Free Press.
Meyer, K. E. (2001). Institutions, transaction costs, and entry mode choice in Eastern
Europe. Journal of International Business Studies, 32, 357e367.
Minor, M., Wu, W.-Y., & Choi, M.-K. (1991). A proposition-based approach to in-
ternational entry strategy contingencies. Journal of Global Marketing, 4(3),
69e87.
Mishina, Y., Pollock, T. G., & Porac, J. F. (2004). Are more resources always better for
growth? Resource stickiness in market and product expansion. Strategic Man-
agement Journal, 25(12), 1179e1197.
Mudambi, R. (2008). Location, control and innovation in knowledge-intensive in-
dustries. Journal of Economic Geography, 8(5), 699e725.
Nian, J. (2006). Report on the life of quality in Chinese cities in 2006. Beijing, China:
Economics Press.
Nitsch, D., Beamish, P., & Makino, S. (1996). Entry mode performance of Japanese
FDI in Western Europe. Management International Review, 36, 27e43.
North, D. C. (1990). Institution, institutional change and economic performance. New
York, NY: Cambridge University Press.
Peng, M. W. (2001). The resource based view and international business. Journal of
Management, 64, 165e187.
Penrose, E. T. (1959). The theory of the growth of the ?rm. New York, NY: Oxford
University Press.
Porter, M. E. (1998). On competition. Boston, MA: Harvard Business School.
Quah, D. T. (1996). Regional convergence clusters across Europe. European Economic
Review, 40, 951e958.
Rauch, J. E. (1993). Productivity gains from geographic concentration of human
capital: evidence from the cities. Journal of Urban Economics, 34(3), 380e400.
Rey, S. J., & Montouri, B. D. (1999). US regional income convergence: a spatial
econometric perspective. Regional Studies, 33(2), 143e156.
Rivoli, P., & Salorio, E. (1996). Foreign direct investment and investment under
uncertainty. Journal of International Business Studies, 27(2), 335e357.
Rugman, A. M., &Verbeke, A. (2002). EdithPenrose's contributiontotheresource-based
view of strategic management. Strategic Management Journal, 23(8), 769e780.
Saxenian, A. (1994). Regional advantage: Culture and competition in silicon valley and
route 128. Cambridge, MA: Harvard University Press.
Scott, A. J. (1988). New industrial spaces. London, UK: Pion.
Seth, A., Song, K. P., & Pettit, R. R. (2002). Value creation and destruction in cross-
border acquisitions: an empirical analysis of foreign acquisitions of U.S. ?rms.
Strategic Management Journal, 23(10), 921e940.
Shapiro, J. M. (2006). Smart cites: quality of life, productivity, and the growth effects
of human capital. The Review of Economics and Statistics, 88(2), 324e335.
Shaver, J. M. (1998). Do foreign-owned and U.S.-owned establishments exhibit the
same location pattern in U.S. manufacturing industries. Journal of International
Business Studies, 29, 469e492.
Shrout, P. E., & Bolger, N. (2002). Mediation in experimental and nonexperimental
studies: newproceduresandrecommendations. Psychological Methods, 7, 422e445.
Simon, C. J. (1998). Human capital and metropolitan employment growth. Journal of
Urban Economics, 43, 223e243.
Simon, C. J., & Nardinelli, C. (1996). The talk of the town: human capital, informa-
tion, and the growth of English cities, 1861 to 1961. Explorations in Economic
History, 33, 384e413.
Simon, C. J., & Nardinelli, C. (2002). Human capital and the rise of American cities,
1900e1990. Regional Science and Urban Economics, 32, 59e96.
Sunley, P. (2003). Urban and regional growth. In E. Sheppard, & T. J. Barnes (Eds.),
A companion to economic geography (pp. 187e201). Oxford, UK: Blackwell
Publishing.
Sun, Q., Tong, W., & Yu, Q. (2002). Determinants of foreign direct investment across
China. Journal of International Money and Finance, 21, 79e113.
Taylor, P. J. (2004). Is there a Europe of cities? In E. Sheppard, & R. B. McMaster
(Eds.), Scale and geographic inquiry (pp. 213e235). Oxford, UK: Blackwell
Publishing.
Tybout, J. R. (2003). Plant- and ?rm- level evidence on “new” trade theories. In
E. K. Choi, & J. Harrigan (Eds.), Handbook of international trade (pp. 388e415).
Oxford, UK: Blackwell.
Voss, G. B., Sirdeshmukh, D., & Voss, Z. G. (2008). The effect of slack resources and
environmental threat on product exploration and exploitation. Academy of
Management Journal, 51(1), 147e164.
Wilson, B. D. (1980). The propensity of multinational corporation to expand
through acquisitions. Journal of International Business Studies, 2, 59e65.
Woodcock, C. P., Beamish, P. W., & Makino, S. (1994). Ownership-based entry mode
strategies and international performance. Journal of International Business
Studies, 25, 253e273.
Woodman, R. W., Sawyer, J. E., & Grif?n, R. W. (1993). Toward a theory of organi-
zational creativity. Academy of Management Review, 18(2), 293e321.
H.-W. Hsu, Y.-L. Jaw / Asia Paci?c Management Review 20 (2015) 252e264 264

doc_519589170.pdf