The Impact of Global Cities on Location Choices: Evidence
from German and French Automotive MNEs Offshoring R&D
and Production Facilities Towards China
Author: Cas Boudewijn Gaaf | 10682937 Date of submission: 15 August, 2014 First supervisor: Mr. S. von Delft PhD Second supervisor: Mr. dr. N. Pisani
Business Studies, International Management Faculty of Economics and Business
ACKNOWLEDGEMENTS
This study was inspired by the courses of “Theories of International Management” and “International Strategy” instructed by Mr. dr. Lindeque and Mr. dr. Pisani. Through the acquaintance with the concepts brought forward in these courses, this study saw the light. Due to the help and useful insights given by my thesis supervisor Mr. von Delft PhD, I managed to give this study face.
Cas Gaaf
Amsterdam, August 2014
ABSTRACT
This study examines the influence of attributes of global cities on location choices of MNEs offshoring parts of their R&D and production facilities towards China. It does so by integrating the concept of global cities with the concept of location choice in a conceptual model in which also the moderating roles of institutional effects and spatial agglomeration are included. The study broadens the concepts of global cities. More specifically, it makes a new combination of concepts together in a model and attempts to understand the underlying relationships. It mainly contributes to the on-going debate of increasing economic and geographic “embeddedness” related to globalization. Furthermore it uses the concepts of institutionalism and spatial agglomeration as moderating factors and broadens the existing literature on these concepts, as it challenges the understanding of the moderating role of these concepts in the relation between global cities and location choices made by MNEs. The study concludes by asserting that the underlying dimensions of global cities vary in their influence on location choice for production facilities and R&D facilities. The influence of global cities on locating R&D facilities gives the impression to be substantial, mostly in terms of the availability of a cosmopolitan environment. The institutional effects are inconclusive, mostly due to the timing of entrance of our automotive cases, but the agglomeration effects turn out to play a significant role.
Keywords: global cities; location choice; institutional effects; spatial agglomeration;
advanced producer services; international connectedness; cosmopolitan environment; MNEs; R&D; Production
TABLE OF CONTENTS
CHAPTER 1. INTRODUCTION ... 6
1.1 Research Problem And Relevance Of The Study ... 6
1.2 Problem Statement And Objectives ... 9
1.3 Outline Of The Study ... 11
CHAPTER 2. LITERATURE REVIEW ... 12
2.1 Location Choice ... 12
2.1.1 Foreign Direct Investment ... 12
2.1.2 International R&D ... 15 2.1.3 International Production ... 17 2.2 Global Cities ... 19 2.3 Institutional Effects ... 20 2.4 Spatial Agglomeration ... 23 2.5 Conclusion ... 26
CHAPTER 3. CONCEPTUAL FRAMEWORK ... 27
3.1 Introduction ... 27
3.2 Advanced Producer Services And Location Choice ... 28
3.3 Degree Of Interconnectedness And Location Choice ... 29
3.4 Cosmopolitan Environment And Location Choice ... 30
3.5 Institutional Effects, Global Cities And Location Choice ... 31
3.6 Spatial Agglomeration, Global Cities And Location Choice ... 32
3.7 Conclusion ... 33 CHAPTER 4. METHODOLOGY ... 34 4.1 Introduction ... 34 4.2 Research Method ... 34 4.3 Research Design ... 36 4.3.1 Data Collection ... 36 4.3.2 Data Analysis ... 38 4.4 Variables ... 39
4.5 Validity And Reliability ... 40
4.6 Conclusion ... 40
CHAPTER 5. THE CHINESE AUTOMOTVIE INDUSTRY ... 41
CHAPTER 6. ANALYSIS ... 43
6.1 Introduction ... 43
6.2 Within-Case Analysis... 43
6.2.1 The Focal Global Cities ... 43
6.2.2 The Focal Automotive Mnes... 46
6.3 Cross-Case Analysis ... 49
6.3.1 Global Cities And Location Choice ... 49
6.3.2 Institutional And Agglomeration Effects ... 52
CHAPTER 7. DISCUSSION AND CONCLUSION ... 55
7.1 Introduction ... 55
7.2 Implications ... 55
7.3 Limitations And Future Research Directions ... 60
7.4 Conclusion ... 61
REFERENCES ... 62
APPENDICES... 72
Appendix A: Map Of China ... 72
Appendix B: Timeline Of Important Events China ... 73
LIST OF FIGURES Figure 1.1-1 Foreign direct investments, net inflows ... 7
Figure 3.1-1 Conceptual framework ... 28
Figure 6.2-1 The automotive industrial zones of China […] ... 45
Figure 6.2-2 The focal automotive MNE cases and the geographically […] ... 48
LIST OF TABLES
Table 2.1-1 Theories explain location specific advantages of countries ________ 14 Table 2.1-2 Types of R&D internationalization ___________________________ 16 Table 2.1-3 Types of R&D facilities and how these are run by MNEs _________ 17 Table 2.2-1 Typology of global cities and attributes _______________________ 20 Table 2.3-1 MNEs engaging in the institutional environment ________________ 23 Table 3.1-1 The propositions with associated relationships _________________ 33 Table 4.3-1 Stages of data analysis ___________________________________ 37 Table 6.2-1 The focal global cities cases ranked and analysed […] ___________ 44 Table 6.2-2 Analysis of the focal automotive MNEs cases […] _______________ 47 Table 6.2-3 Comprehensive table containing all the cases […] ______________ 51 Table 6.3-1 Example of pioneer case VW in China […] ____________________ 54 Table 7.2-1 Outcomes concerning the propositions […] ____________________ 55 Table 7.2-2 Outcomes and relating back to table 3.1-1 ____________________ 56
CHAPTER 1. INTRODUCTION
1.1 RESEARCH PROBLEM AND RELEVANCE OF THE STUDY
Scholars in both international business (Goerzen et al., 2013; Alcácer and Chung, 2007; Nachum and Wymbs, 2005; Porter, 2001) and economic geography (Krugman, 1991; Lorenzen and Mudambi, 2013; Markusen, 1996) contemplate the concept of location in answering the question of where and why firms place specific activities in particular areas. As mentioned by Goerzen et al. (2013), although the significance of studying the concept of location has raised the last decennia, the combination of location and the geographic aspect of MNE behaviour remain weak ( see Beugelsdijk et al., 2010 for an exception).
International business scholars have examined the process of locating R&D and production facilities across borders extensively. Firms, especially MNEs, are constantly trying for an optimal outline of their global supply chain (Dunning, 2009) in order to deal with the context of economic globalisation and regional integration. Recent studies show that the process of (re)-locating R&D facilities overseas has changed from pattern. It has shifted from a focus primary on the USA, the EU and Japan towards emerging economies like China, India and Russia (Huggins et al., 2007). For instance, the shifting focus is according to Sun et al., (2006) the establishment of 700 independent R&D facilities in China after the 2000s and the associated amount of foreign direct investments in China. In figure 1.1-1, the FDI of China opposed to the US is shown as well as the growing trend for China as of 2009.
Thus, Both US firms as well as European and Japanese firms are offshoring parts of their businesses to emerging economies, especially to China (Huggins et al., 2007). More specifically in terms of the automotive sector in China, statistics show a
sales growth in China from almost 6 million vehicles in 2005, up to more than nineteen million in the year of 2012 (OICA, 2014).
Figure 1.1-1 Foreign direct investments, net inflows BoP, current US$ -in billions
Source: Compiled from World Development Indicators of The World Bank.
The variation in economic development, innovation capacity, knowledge productivity and local innovative milieu are, according to Liu and Chen (2012) significantly dependent on which region in China one takes into account. The existence of global cities could be argued on to play a significant role in this variation among regions, since global cities possess the attributes mainly associated and necessary for the existence and further development of these above-mentioned factors.
According to Goerzen et al. (2013), the concept of global cities should be studied more extensively in future research to improve the insight into the MNE within its geographic space. Global cities have a thorough impact on the location strategies of MNEs (Sassen, 1991, 1994, 2001) and thus, as a result, the concept is associated with location choice as an outcome of these particular strategies. The different attributes the global cities concept possess, create forces that push or discourage the co-location of people, companies, and institutions, and therefore, are related to the concept of foreign direct investment and the associated location choices of MNEs. Also, the focus on global cities as one of the main concepts of this
0 100 200 300 400 2004 2005 2006 2007 2008 2009 2010 2011 2012 China United States Linear (China) Linear (United States)
study, contributes, in combination with location choice, to the on-going debate of increasing economic and geographic “embeddedness” related to globalization. As mentioned for future research in the study of Goerzen et al. (2013), global cities can vary among each other in their industrial traditions. Subsequently, this may lead to differences in the attractiveness to firms from different sectors and to preferable location choices depending on these industrial traditions. Moreover, Goerzen et al. (2013) comment on the concept of global cities that it is of exceptional importance for public policymakers, referring to the institutional environment in that this influences the process of globalization via cities (e.g. via stimulation or discouraging of FDI into cities or agglomerations).
Thus, the institutional environment has a thorough influence on the process of globalization and therefore may partly determine the location choices made by MNEs. This is especially true when it comes to FDI towards less developed countries. In these cases of offshoring (i.e. FDI) the institutional environment is seen as a highly important factor and as such is described in the literature in various ways (e.g. administrative distance by Chemawat (2001); policy-induced effects (Mariotti et al., 2010) and simply the taxation level in a particular country (Laamanen et al., 2011)).
Lastly, the spatial agglomeration can be argued on to play an important role in the location choice as it may for instance function as a pull factor for MNEs location choices in terms of supporting industries (Laamanen et al., 2011), or it may instigate knowledge spillovers between MNEs and other centripetal forces (Alcacer and Chung, 2007). On the other hand, as probed by Mariotti et al. (2010), it may cause centrifugal forces in terms of negative knowledge spillovers (e.g. the loss of intellectual property) and negative mimetic behaviour, as a result of information
externalities when MNEs enter a foreign country without the appropriate information (i.e. information asymmetries). In that case MNEs tend not to agglomerate but instead locate separate from each other. So, as in line with the categorization of Dunning and Lundan (2008), both institutional effects as well as agglomeration effects are important factors (endowment effects are a third relevant factor) in the location choice of MNEs and therefore highly relevant in the investigation of the relationship between global cities and location choices of MNEs.
1.2 PROBLEM STATEMENT AND OBJECTIVES
This study departs from the need to increase the understanding of which factors influence the location choice in offshoring R&D and production facilities. More specifically, the purpose of this study is to develop and enrich the conceptually and empirically validated understanding of how the different attributes of global cities influence location choices in offshoring production and R&D facilities by MNEs. Furthermore, the role of institutional effects (e.g. government policies and legislation) and agglomeration effects (e.g. mimetic isomorphism) are included as moderating factors, influencing the relationship between global cities and location choices. This leads to the central research question of this study:
Here, global cities are understood out of the functional tradition (i.e. rather than the demographic one) (Beaverstock et al., 1999), and refer to: (1) a high degree of
interconnectedness to local and global markets; (2) a cosmopolitan environment;
and (3) high levels of advanced producer services. These three factors represent “How do the different attributes of global cities affect the location choice in the process of offshoring R&D and production facilities?”
the distinction between global cities and other levels of analyses (e.g. megacities or industrial clusters) and entail several implications for MNE location choice (Goerzen et al., 2013). Also, global cities are pre-eminently applicable in functioning as cultural, political, economic, and social centers (Derudder et al., 2003; Short et al., 2000; Taylor et al., 2002), with the required characteristics to command and control globally. As a result, global cities function as a production space, in its broadest sense, for key inputs that globally operating MNEs need for their complex businesses to keep going (Sassen, 2012).
In order to provide an answer to the aforementioned problem statement, the study formulates three sub-research questions, which foresee in the conceptual and empirical research of this study and specify the correct focus that is needed to answer the problem statement properly.
To provide an insight into the moderating factors of the role of institutional effects and spatial agglomeration, the study formulates two sub-research questions, which additionally contribute to the conceptual and empirical research of this study.
(1) How does a high degree of interconnectedness to local and global markets affect the location choice in the process of offshoring (1a) R&D facilities, and (1b) production facilities?
(2) How does a cosmopolitan environment affect the location choice in the process of offshoring (2a) R&D facilities, and (2b) production facilities?
(3) How do high levels of advanced producer services affect the location choice in the process of offshoring (3a) R&D facilities, and (3b) production facilities?
The objective of this study is to develop more insight into the relation of global cities and location choice. It does so by separating three main concepts underlying global cities, which represent the distinction between global cities and other kind of cities (see section 1.2: problem statement and objectives). The study also keeps a managerial aim of providing automotive executive managers with more insight into the locational choices made in the past and for making future location choices, based upon the model as set out in chapter 3.
1.3 OUTLINE OF THE STUDY
The study consists of seven chapters. Chapter one provides a short introduction to the study and stresses the problem statement and objectives. After the introduction in chapter one, the study moves to the theoretical background in chapter two where a brief literature review is set out. In chapter three, the conceptual framework with the associated propositions are brought forward and supported by the literature. The methodology section describes the appropriate research design for the investigation of the central research questions and the underlying conceptual framework. Thereafter, chapter five entails the automotive industry of China with a short paragraph on the research aim and approach. Then, the study moves to the analysis and results. Finally, the last chapter will intricate on the theoretical and managerial implications, limitations and an overall conclusion involving future research directions.
(4) How do Institutional Effects influence the relation between global cities and location choice in the process of offshoring R&D and production facilities?
(5) How does Spatial Agglomeration influence the relation between global cities and location choice in the process of offshoring R&D and production facilities?
CHAPTER 2. LITERATURE REVIEW
2.1 LOCATION CHOICE
Location is the central question in defining international business research (Goerzen et al., 2013). The literature on location choice is largely build upon theories explaining specific advantages of location. Hence, the concept is closely related to the eclectic paradigm developed by Dunning (1977, 1993), which is used as a fundament in numerous studies to location choice and the MNE. Scholars frequently refer to the eclectic paradigm as the ownership, location and internalization tripod (OLI).
2.1.1 FOREIGN DIRECT INVESTMENT
The OLI tripod, in which the specific motivations for foreign direct investment (FDI) are captured, is appropriate in explaining MNEs supply of foreign markets, rather than or succeeding of domestic ones (Dunning, 2001). The elaboration of a general theory of FDI is built upon the experience of industrialized country investors (Buckley et al., 2007). This general theory is according to Buckley et al. (2007) build upon two principles: (1) MNEs only undertake internalization of missing and imperfect external markets into intermediate products and knowledge from the point that costs associated with this internalization outweigh the benefits, and (2) locations are selected upon cost criteria (i.e. the minimization of a firms subsidiary or business unit costs of operations). The theory of the liability of foreignness (LoF) by Zaheer (1995) further identifies these costs associated with FDI by four key dependent sources (i.e. costs directly associated with spatial distance, firm-specific costs related to unfamiliarity with the host location, costs resulting from the host location and costs
from the host location). Plausibly, the LoF theory is used in accessing the challenges arising host country specific advantages.
As stated, the OLI tripod of Dunning (1977, 1993) consist of three variables, which each interact independently of each other. These variables determine (1) the extent (i.e. the ownership consisting of asset advantages and transaction cost minimizing advantages), (2) the geography (i.e. the location) and (3) the industrial composition (i.e. the internalization) of foreign production undertaken by MNEs (Dunning, 2000). Furthermore, these three variables each consist of sub-paradigms, which excavate the understanding of MNEs and their motive(s) to undertake FDI.
First, MNEs possess specific ownership advantages, also referred to as the competitive advantages of FDI-seeking firms (Dunning, 2000). According to Dunning (2000), the greater the ownership advantage of a particular MNE in a country compared to other MNEs investing in that country, the more likely its engagement and increase of FDI in that specific country. Second, the relative attractiveness of foreign countries, rather than domestic ones, for the investment of value-added activities (i.e. the more need for foreign presence, the more a firm will engage in FDI by exploiting their ownership specific advantages).
Third, the alternative ways of exploiting and creating core competencies MNEs possess given the presence of different countries and regions with each possessing certain locational attraction (e.g. buying and selling goods and services in the open market, so called inter-firm non-equity agreements, and internalization of intermediate product markets (Buckley and Casson, 1976, 2009; Hennart, 1982; Rugman, 1996).
The precise configuration of the eclectic paradigm by Dunning (1977) varies among MNEs due to its contextual dependence (Dunning, 2000). In Dunning (1998),
several motivations (i.e. resource-seeking, market-seeking, efficiency-seeking, strategic-asset-seeking) for undertaking FDI are stressed. As for the last mentioned, strategic-asset-seeking, terms as catch-up, diversification and R&D springboard (Demirbag and Glaister, 2010) are strongly associated concepts, predominantly extant in studies investigating outward direct investment (ODI) of developing countries MNEs (e.g. Buckley et al., 2006; Wong and Chan, 2003; Warner et al., 2004), which are strengthening their FDI flows globally.
Besides the eclectic paradigm there are other less influential theories underlying the location choice. Prior literature concerning the location (L) specific advantages out of the eclectic paradigm consists of numerous theories in which the primary focus lay on the advantage(s) of location(s) in certain countries (as summarized in table 2.1-1).
Table 2.1-1 Theories explain location specific advantages of countries
Theory Author, Year Focus (advantage of location)
I. Traditional location theories i.a. Lloyd and Dicken (1977)
Location choice is influenced by demand and supply oriented variables (e.g. availability, quality and price of natural resources and immobile assets)
Presence of competitors, exchange rates II. Theories of
internationalization
i.a. Johanson and Vahlne (1990)
Mainly based upon market-seeking and resource-seeking using traditional locational variables, but also using firm specific advantages
The role of physic distance, particularly in exploiting accumulated knowledge based ownership advantages.
III. Agglomeration theories i.a. Audretsch (1998)
Economies of scale and scope
Clustering of products for convenience of consumers and industrials
Supply related clusters based on asset augmenting activities, local accumulation of knowledge, and exchange of information and learning experience
Source: Compiled from Dunning (2000).
2.1.2 INTERNATIONAL R&D
The term R&D stands for Research and development and, “denotes a complex process of scientific and technological research, development, of new products and processes, manufacturing and marketing support, and provision of technical services (von Zedtwitz, 2004: 442).”
The process of locating R&D projects overseas has increased in conjunction with changes in the pattern and nature of internationalization of R&D (Atkinson, IV. Theories related to spatially
specific transaction costs
i.a. Scott (1996)
Spatially related transaction costs are hypothesized to lead to a clustering of related activities: (a) to reduce overall costs; (b) to maximize benefits of inter-related innovating and learning activities. V. Theories of complementary
assets
i.a. Chen and Chen (1999)
The presence of related activities which help lower transportation costs and promote joint economies in innovation, production and marketing
VI. Theories related to government induced incentives
i.a. Loree and Guisinger (1995)
Fiscal incentives leading to increase in demand for products of MNEs
Promotion of innovation-driven alliances Concessionary rights for exploiting
natural resources; IP-rights and tax advantages
VII. Theories related to
oligopolistic behaviour and product cycle
i.a. Graham (1998)
Follow my leader and other forms of oligopolistic behaviour (context specific)
VIII. Theories of risk diversification
i.a. Rugman (1979)
FDI determine location specific risks, but theory suggests that firms will diversify their portfolios to minimize risk exposure (exchange, political and economic risks) IX. Exchange rate theories i.a. Rangan
(1998)
Exchange rates or changes in these rates capture most of the differences in cross-border locational costs
Expectations of investors about future course of exchange rates also influence cross-border locational costs
X. Knowledge enhancing (dynamic) theories
i.a. Kuemmerle (1999)
Location choice influenced by level of opportunities accumulated for upgrading existing core competencies
Dynamic theories explain locational strategy and especially applies to R&D activity of all kinds of FDI
2007; Lewin et al., 2009; Manning et al., 2008). Certain patterns suggest MNEs are shifting their locational focus of value-added activities from the USA, EU and Japan towards developing countries including China, India and Russia (Huggins et al., 2007; Lewin et al., 2009). According to von Zedtwitz (2005) this particular pattern of R&D internationalization could be typed as ‘modern’ since it entails offshoring R&D from an advanced home country towards a developing country (e.g. US MNE offshoring R&D to China). In table 2.1-2 the other types of internationalization are emphasised.
Table 2.1-2 Types of R&D internationalization
Source: Compiled from von Zedtwitz (2005).
Besides the typology of R&D internationalization, prior literature also offers typologies describing how R&D facilities are run by MNEs (Voelker and Stead, 1999). These are pooled in Table 3, where the typologies of Hood and Young (1982) and Bartlett and Goshal (1988) are combined, since both describe basically the same concept of decentralized R&D facilities (i.e. offshoring R&D facilities). Decentralization of R&D facilities is dependent upon the so-called ‘centre of gravity’ of a company (Chiesa, 1995), which entails the focus on either research or development. The process of ‘centre of gravity’ is usually kept at a central location
Author(s) Type of R&D
Internationalizat ion
Home Country & Host Country status
Example Haug et al. (1983) Traditional Advanced to advanced USA to EU Japan to USA Zedtwitz and Gassmann (2002) Modern Advanced to developing USA to China EU to India Lall (1987, 1990) Catch-up Developing to
advanced
China to US India to EU Kim (1980, 1997) Expansionary Developing to
developing
China to Brazil India to China
according to Chiesa (1995). Bartlett and Goshal type this pattern as a traditional one, in which the term ‘central-for-global’ (see: table 2.1-3) represent the central facility yielding all the R&D for a global purpose. Hence, the non-centre-of-gravity processes can be undertaken at distant locations (e.g. in developing countries) since these processes are less dynamic in terms of technology (e.g. firms from Europe opened facilities in the USA to take advantage of ‘centres of excellence’ (Voelker and Stead, 1999).
Table 2.1-3 Types of R&D facilities and how these are run by MNEs
Source: Compiled from Voelker and Stead (1999).
Research on R&D during the eighteens and nineteen’s mainly focussed on firms originated from advanced and economically developed countries (Zedtwitz, 2005), but as shown in table 2, the ‘catch-up’ and ‘expansionary’ R&D internationalizations by developing MNEs are increasing. Nevertheless, MNEs from advanced economies still clearly dominate domestic and international R&D in terms of patents and expenditures (UNCTAD, 2005).
2.1.3 INTERNATIONAL PRODUCTION
The concept of internationalization of production is of great public interest, particularly in countries with advanced economies. This is because of the fear of offshoring activities and the associated disappearance of manufacturing facilities,
Author(s) Types of R&D facilities run by MNEs
Hood and Young (1982)
Support of local use of production-technology and products which are designed or created outside the country
Foregoing Combined with a little design work intended for use by the company as a whole Develop products and/or processes for global production Bartlett and Goshal (1988)
Local-for-local Local-for-global Global-for-global
and further substituting of domestic investments, which may lead to low employment rates (Mayer et al., 2010). Not surprisingly, many related research is available on the wealth effects of locating production activities across borders (e.g. Feenstra, 1998; Glass and Saggi, 2001; Bernard et al., 2003).
Besides this public attention, scholars examined the internationalization of production extensively from other perspectives, mostly in terms of market-related motives of investment (Fish and Zschoche, 2012). Prior research firstly focused on the question whether MNEs establish cross border production facilities (e.g. for the sake of transaction and transportation costs, adapting to local tastes or for overcoming trading barriers (Dunning and Lundan (2008). In the field of economics, scholars examined the international location of production activities comprehensively. For instance the advantage that MNEs take of factor cost differentials between different countries by implementing a ‘cost-minimizing geographic organization’ of their production as a whole (Antras and Helpman, 2004; Blinder, 2006; Grossman and Rossi-Hansberg, 2008). Furthermore, equilibrium-based models are brought up as underlying explanation of the on-going optimization process of (re)locating production facilities, which denotes the permanent factor-cost discrepancies between countries (Fish and Zschoche, 2012). It is found difficult for firms to keep up with these differentials because synchronized investments and divestments are needed to achieve optimal dispersion of a firms production facilities across the globe. For MNEs, this process is found easier since they possess the advantage of having the flexibility to quickly transfer resources across borders (Kogut, 1983). Other studies, using stochastic programming, demonstrate that shifting production facilities across countries due to uncertainty in exchange rates is limited and maintained to two countries as MNEs benefit from this exclusively (Kogut
and Kulatilaka, 1994). Finally, the influence of labour costs changes (Belderbos and Zou, 2007) and economic crisis (Chung et al., 2010) as a quite recent theme in international business are studied and revealed that MNEs make use of their flexibility by adjusting their production configuration form time to time. As a result, this operational flexibility turned out to have a positive effect on the creation of firm value (e.g. Lee and Makhija, 2009a/b).
2.2 GLOBAL CITIES
Cities are studied by several disciplines over the past decades (Goerzen et al., 2013), which have led to several varieties of city types in the existing literature (e.g. great industrial cities, world cities, imperial cities, global capitalist cities, primate cities). Global cities, and the question how many there are, and what the hierarchy of these cities is (e.g. an international ranking), is still of an on-going debate (Beaverstock et al., 1999; Derudder et al., 2003; Short et al., 2000; Taylor et al., 2002) as mentioned by Goerzen et al. (2013) in their recent study of global cities and the location strategy of the MNE. In describing global cities, a separation have to be made between two traditions, which both are forthcoming out of different contexts (Beaverstock et al., 1999).
First, the demographic tradition, which has a national urban context, using demographic data and primarily established for the understanding of urban pre-eminence or hierarchy. The main focus in this tradition lay on human and ecological implications of large human populations in for instance ‘megacities’ (Gilbert, 1996).
Second, the functional tradition, which came forth out of the reinterpretation of the global economic role of cities (e.g. Friedmann, 1986; Sassen 1991), in which more focus is given to the interconnectedness and characteristics of certain cities (Brown et al., 2010). In the functional tradition, the particular functions cities play in Page | 19
the global economy as a result of the activities that go on within these cities, lead to the concept of the “Global City” (Goerzen et al., 2013).
According to Goerzen et al. (2013), prior literature suggests that the functional tradition has united around the concept of global cities due to three key attributes that are crucial in the characterization of global cities (i.e. advanced producer services, a cosmopolitan environment and a high degree of interconnectedness) and which separate global cities from the other abovementioned varieties of city types (e.g. the megacity Calcutta with over ten million citizens, without possessing any of the global city attributes, Beaverstock et al., 1999). Also, all three attributes have clearly different implications for MNEs their location strategies they implement around the globe. In table 2.2-1, the three attributes are shown and examples for each attribute are given.
Table 2.2-1 Typology of global cities and attributes
Source: Compiled from Goerzen et al. (2013).
2.3 INSTITUTIONAL EFFECTS
Institutional theory, also known as the institution-based view of business strategy plays a thorough role in all sorts of business contexts. As stated by Peng (2002), one of the fundamental founders of the institution-based view, all firms (especially MNEs) are exposed to the institutional framework in which they are embedded (i.e. “the set
Author(s) Required Attributes for
a Global City
Example
Sassen (1991, 1994) Advanced producer services
The existence of agglomerated and specialized service providers in finance, law, accounting, and advertising
Hall (1966) Cosmopolitan
environment
The existence of characteristics as diversity of national cultures, facilities for expatriates
Friedmann (1986) International
Connectedness
The existence of major financial centers, HQs of international institutions, rapid growth of business services sector, major transportation nodes
of fundamental, political, social, and legal ground rules that establishes the basis for production, exchange, and distribution (Davis and North (1971: 6)). Thus, it could be argued that there is barely any disagreement that institutions matter and that research on strategy in its widest form cannot just simply focus on industry conditions and firm resources as the only underlying drivers (Khanna and Palepu, 1997; Peng, 2000a). Institutions can be defined in numerous ways, but most research is based on the definition of North (1990: 3), stating that institutions are: “The rules of the game in a society or, more formally, the humanly devised constraints that shape human interaction.” Alternatively, a commonly used definition of institutions is provided by Scott (1995: 33). He states that institutions are: “Cognitive, normative, and regulative structures and activities that provide stability and meaning to social behaviour.” Besides the definition and concept of institutions, scholars developed the concept of “new institutionalism”, which entails a distinctive focus on both an economic (North, 1990; Williamson, 1985) as well as a sociological version (DiMaggio and Powell, 1991; Scott, 1995) of institutionalism. The economic focus entails the focus on efficiency, as the latter focuses on legitimacy. As in the article of Peng (2002), the underlying theoretical fundaments of these both versions of new institutionalism are beyond the scope of this study.
According to Peng (2002), an important phenomenon should be taken into account concerning the before mentioned definitions, concepts and focuses. This entails the fact that institutions can help reduce uncertainty for organizations as institutional frameworks interact with organizations by indicating which selections are tolerable and viable (Peng, 2002). In a studie of Dikova and van Witteloostuijn (2007), the aforementioned new institutionalism developed by Williamson (1985) is used to investigate the institutional effects in transition economies on MNEs dual
investment choices (i.e. both the mode of entry and the mode of establishment). They conclude by stating that the degree of a host country’s institutional advancement (i.e. the progress of changes in institutions over time in a particular country) moderates the effects of technological intensity and international strategy in relation to the dual investment choice of MNEs.
Another study on cross-national distance on the basis of an institutional approach by Berry et al. (2010) also mention the importance of institutional theories in terms of national business, governance and innovation systems. They do so by studying the foreign expansion choices of US companies to illustrate this. The most often-used approach, developed by Hofstede (1980), in which cultural dimensions explain differences among different countries, is not been used by Berry et al. (2010). Instead, they developed a conceptualization by making use of institutional theories, as in the study of Dikova and van Witteloostuijn (2007), and added recent institutional theory of Jackson and Deeg (2008) and Pajunen (2008). They distinguish nine dimensions (i.e. economic, financial, political, administrative, cultural, demographic, knowledge, global connectedness and geographic distance) in their study. In the case of most dimensions, institutions play the major role (i.e. economic, financial, political and connectedness dimensions of distance).
In recent studies towards institutions in combination with MNEs and the activities MNEs generate, the engagement of MNEs with institutions is stressed in terms of creativity of MNEs and institutional entrepreneurship. This is expected to lead to co-evolution of the institutional environment. According to Cantwell et al. (2010) this co-evolution is of relevance in understanding the interconnectedness between MNEs, its activities, and the institutional environment as a whole. In table 2.3-1 the
three different types of engagement are set out. Cantwell et al. (2010) start from institutional avoidance (i.e. the only chance MNEs have is choosing between Institutional environments through the use of certain entry modes) to adaptation (i.e. MNEs may try to influence, but only on a regional level) to ultimately co-evolution. In this final type of engagement, MNEs may even develop new institutions (e.g. International Accounting Standards) and are intertwined with the institutional environment through active agency and institutional entrepreneurship (e.g. Corporate Social Responsibility).
Table 2.3-1 MNEs engaging in the institutional environment
Source: Adapted from Cantwell et al., (2010).
2.4 SPATIAL AGGLOMERATION
Agglomeration is a traditional approach possessed by MNEs and entails the geographical proximity of MNEs locating near one another (Ellison and Glaeser, 1997). An agglomeration could be seen as a collection of MNEs which are somehow related to each other through various forms of formal and informal practices that binds these MNEs together in a specific spatial geographic nearness (Nachum and Keeble, 2003). According to Porter (1998), this spatial geographic nearness (i.e. cluster of firms) leads to competitive advantage and thus, a superior performance for
Type of engagement Main features
Institutional avoidance The institutional environment is totally exogenous to firms
Institutional adaptation Firms try to find the best fit within the institutional environment and try to influence these to a limited extend. Institutional is still exogenous.
Institutional co-evolution Firms doing co-evolution with the institutional
environment, as the institutional environment is partly endogenous.
these agglomerated MNEs. These firms may collaborate, compete or even be independent from each other and moreover, may be connected through market and non-market links (Nachum and Keeble, 2003).
One of the most important benefits of spatial agglomeration is the creation and transmission of knowledge (i.e. knowledge spillovers) between MNEs (Paci and Usai, 1999). For instance, it is requisite to promote social learning amongst MNEs and also keep joining with collaboration. Furthermore, according to Chung and Kalnins (2001), the spatial agglomeration of MNEs makes the road clear for production enhancement (i.e. MNEs get access to methods and proficiencies used by other MNEs) and a growing demand for products (i.e. the customer’ search costs are lower due to the agglomeration of firms). Besides production enhancement, the enhancement of innovation and R&D could be argued on to profit from spatial agglomeration. Not only can the agglomeration and the existence of external suppliers etcetera drive the inspiration and implementation for innovations (Hulsink et al., 2009), but also enhance productivity advantages (i.e. land, components and personnel), as described by Porter (2000).
Besides some of these important abovementioned benefits of spatial agglomeration, there are inherent disadvantages of it. More specifically, the interaction between MNEs and thus the creation and transmission of knowledge does not mean necessarily that these so-called spillovers are of a positive kind. In certain cases, MNEs do not choose to agglomerate because of negative knowledge spillovers. For example, the leakage of intellectual property (McCann and Mudambi, 2004) is regarded to be fateful for MNEs competitive advantage and is therefore considered in the literature as a possible negative outcome of agglomerations. Another example can be described in terms of innovativeness and firm
heterogeneity, firms that are far ahead of their competitors may choose not to agglomerate because this will give them not or only marginally boosts, instead of their opponents, which will gain viable strengths of the ahead firm (Shaver and Flyer, 2000).
Hence, knowledge spillovers are highly important in the extent to which MNEs locate near one another (i.e. co-locating) and to what extend they learn and speed the so-called “flow of ideas” (Ellison et al., 2007). Firstly, of importance is the extent to which the spatial proximity interaction implies interaction (i.e. as spatial proximity does not imply automatically interaction) and secondly, if the interaction creates positive spillovers or not (Mariotti et al., 2010). Furthermore, as Mariotti et al. (2010) states: “Firms may absorb knowledge, but they may also lose it” (Mariotti et al., 2010: 522), which make that MNEs have to make a trade-off between agglomerating or not, or balance the inflows and outflows of knowledge depending on the competitive position of the particular MNE towards its local competitors.
As for this competitive position, spillovers can be of two kinds: (1) intra-industry spillovers (Baptista and Swann, 1998), which are related to firms within the same industry (e.g. the automotive industry agglomerated in the area of Muenchen and Stuttgart), and (2) the inter-industry spillovers (Feldman and Audretsch, 1999), which are related to firms of a broader spectrum. This means that firms out of different industries agglomerate and enjoy the enhancement of diversity externalities (i.e. innovative activities across different sectors as for example the case in Brabant Stad).
2.5 CONCLUSION
As has just been shown in this section, the choice of location should be discussed in more detail at the geographical level. The section moves from a country specific analysis (i.e. location choice, global cities and institutional effects) towards a
regional analyses (i.e. spatial agglomeration). From thereon, the concept of global cities comes into play. Due to the fact that global cities as a concept in the area of geographical studies are adolescent, the following section will elaborate on this topic.
CHAPTER 3. CONCEPTUAL FRAMEWORK 3.1 INTRODUCTION
In order to shape this study, this chapter provides a conceptual framework (see figure 3.1-1), wherein the theoretical foundations as discussed separately from each other in chapter two are connected to each other. It builds on the central problem statement and the associated sub research questions as presented in the introductory chapter. Hereby the various dimensions are linked together in an effort to explain the mutual relationships, which may be positive or negative. Firstly, the direct relationship which is advocated in this study is examined through the development of three propositions (i.e. global cities’ effect on MNEs location choice). Secondly, introducing a fourth proposition assesses the estimated moderating influence of institutional effects on this direct relationship. Thirdly and lastly, the predicted moderating effect of spatial agglomeration on this direct relationship is examined through setting up a fifth proposition. As shown in figure 3.1-1, the conceptual framework consists of four cubes where one of them is bold.
This manners the fact that this variable (i.e. location choice of MNEs) is an outcome variable in this research context. The other three cubes represent the direct (i.e. Global Cities) and moderating effects (i.e. Institutional Effects and Spatial Agglomeration) on this outcome variable. By including and combining these variables, the study avoids the “imbalance” as according to Cook (2006: 78): “Most authors focus on the economic dimensions of world cities/global cities to the exclusion of social, political and environmental criteria."
Figure 3.1-1 Conceptual framework
Through the formation of these propositions, which describe the relationships out of the conceptual framework, the study is able to proceed in the next chapter with describing the data collection procedure and case descriptions as the methodology used in this study.
3.2 ADVANCED PRODUCER SERVICES AND LOCATION CHOICE
The so-called New Economic Geography has led to the understanding that locational choices not only are the result of internal factors of MNEs, but also are highly dependent upon regional factors. In this spectrum the global city can be argued on to fulfil a regional factor in location choices made. As stressed in chapter 2, global cities can be divided into three aspects, of which advanced producer services is the first one to discuss here. The availability of foreign MNEs that are already based in a certain place (e.g. European Banking MNEs in Hong Kong) is described in the literature as a main determinant of location choices made by (other) MNEs
(Mucchielli, 1998). This is closely related to agglomeration, which in turn is included in the fifth proposition. In the study of Goerzen et al. (2013), the advanced producer services, which are said to be available in global cities, are described as an enabler for foreign MNEs to function abroad by servicing them in terms of consulting practices and other advices. Firms that are well established globally and locally do this (e.g. the big four accounting firms and American banks). The study therefore assumes that MNEs tend to locate especially their production facilities around these global cities with advanced producer services, as it is an important force for locating and it makes MNEs less vulnerable for FDI uncertainty as described in theories of the LoF (Mezias, 2002; Zaheer, 1995; Zaheer and Mosakowski, 1997) This presumption is set out in the following proposition:
P1: The availability of advanced producer services within a global city lead to a higher existence of production facilities
3.3 DEGREE OF INTERCONNECTEDNESS AND LOCATION CHOICE
The degree of interconnectedness should, according to Friedmann (1986), have severe influence on the complexity that MNEs face in their foreign operations. Therefore the location choice of MNEs should, among other criteria, be assessed in terms of the transfer of people, capital, goods and information between the MNE HQs in for instance Europe or the US and the focal subsidiaries in for instance China and India. The degrees of interconnectedness make these transfers possible and moreover determine at what speed and at what costs (Friedmann, 1986). For both production and R&D facilities, the study assumes this interconnectedness in its broadest sense is an indispensable factor for the attraction and eventual
establishment of these facilities. As for R&D the amount and supply of excellent brains and the availability of excellent IT communication systems are critical (Hatem 2007). For production facilities the supply of materials (e.g. import and export of semi-manufactured goods) is a critical factor, which again is contingent upon the degree of interconnectedness of a specific area. Moreover and lastly, the production facilities together with the R&D facilities are mutually beneficial (Defever, 2006), because the R&D activities strengthen the production activities in a certain area, since the link between both activities results in vertical links and cumulative effects (Krugman and Venables, 1996). Knowing all this, the second proposition of this study reads:
P2: The greater the international connectedness of a global city, the higher the existence of production facilities and R&D facilities.
3.4 COSMOPOLITAN ENVIRONMENT AND LOCATION CHOICE
Contrary to international connectedness and advanced producer services, a cosmopolitan environment could be seen as a less straightforward factor of the concept of global cities. It therefore is not clear to what extent a cosmopolitan environment could be argued on to effect the location choice concerning production facilities, so this study does not see ground for a relationship there. Instead, the assumption here is that a cosmopolitan environment affects the location choice of R&D facilities in that it attracts these facilities by the necessity of MNEs to attract research results engendered by foreign universities and competitors. Consistent with this is the highlight in the study of Colovic and Mayrhofer (2012), where on the basis of OECD data, it is stated that the knowledge spread around the world is quickly Page | 30
growing and MNEs need to attract this knowledge in order to stay competitive in their specific industry. This can be argued to be mostly done through the cooperation of MNEs with regional universities. This can also be deducted from the study of Colovic, (2006), in which it has been found that R&D location choices are for a great part determined by, inter alia, human factors (e.g. quality and profusion of scientific labour, international openness of researchers, scientific cooperation between companies and universities etc.). The conceptual framework of this study considers these factors as a part of a cosmopolitan environment, which is considered as highly present in global cities. Therefore, the following formulation is brought forward in this study:
P3: The more a cosmopolitan environment is available within a global city, the higher the existence of R&D facilities.
3.5 INSTITUTIONAL EFFECTS, GLOBAL CITIES AND LOCATION CHOICE
Scholars greatly investigated the effects of institutions on a wide range of areas as has been set out in the previous chapter. Also, it is very clear that institutions matter and that in order to formulate a successful strategy, firms have to adapt to the institutional environment, besides focussing on industry conditions and the resources the firm possesses in terms of firm specific advantages (FSA) as underlying drivers (Khanna and Palepu, 1997; Peng, 2000a). The moderating effect of institutions has been studied by Dikova and Witteloostuijn (2007). They mentioned institutions in terms of institutional advancement of a certain country and found evidence for differences between well developed and less or underdeveloped institutional advancement as being crucial in the process of selecting the appropriate entry-mode
and location choice. As they found evidence for in their study, an underdeveloped level of institutional advancement of a certain country does not facilitate the entry-mode of a greenfield (i.e. the set up of a subsidiary from scratch). This is the result of high establishing costs and the risk of losing intellectual property. This study uses these outcomes as a moderating factor in its conceptual framework by stating that institutional effects will have a negative influence in relation to global cities and location choice as it is expected that the particular case study will possess an underdeveloped slow changing institutional environment, which will not lead to great advancements. I will not specify the location choice of either production facilities or R&D facilities, as it is specified entirely on the direct relationship as shown in figure 3.1-1. The study therefore includes the following moderating proposition:
P4: Institutional effects have a negative influence on the relationship between global cities and location choice
3.6 SPATIAL AGGLOMERATION, GLOBAL CITIES AND LOCATION CHOICE
The final proposition of this study will elaborate on the role of spatial agglomeration and its influence relative to global cities and location choices. Spatial agglomeration is been described in the literature as generating both advantages and disadvantages for a broad scala of areas. One of the main is the number of foreign MNEs already established in a certain foreign country (Mucchielli, 1998), which, of course, indicates the importance of spatial agglomerations. In most research done so far, agglomerations are specified in three different variants, namely: economic activity, nationality and industry (Audretsch and Feldman, 1999; Head et al., 1995; Shaver and Flyer, 2000; Nachum and Wymbs, 2002; Belderbos and Carre, 2002). As for the Page | 32
latter, this study uses industry agglomeration (i.e. SIC and NAICS codes) in its conceptual model as the moderating factor. Industry agglomeration not only provides firms with a reduction of uncertainty but also stimulates the transfer of knowledge and as a result accelerates innovation among the firms that agglomerate (Audretsch and Feldman, 1996). The last proposition of this study therefore reads as follows:
P5: Spatial Agglomeration has a positive effect on the relationship between global cities and location choice
3.7 CONCLUSION
In this chapter, the conceptual framework has been specified through the declaration of the corresponding relations, direct or moderating, on the basis of propositions. These propositions serve as a guide for the analysis because they can answer the central question of this study. In table 3.1.1 below, the propositions are set out schematically.
Table 3.1-1 The propositions with associated relationships Independent variables Dependent variables Existence of Production facilities Existence of R&D
facilities Location choice P1:Advanced Producer
Services Higher (+) None (0) x
P2:International
Connectedness Higher (+) Higher (+) x
P3: Cosmopolitan
Environment None (0) Higher (+) x
Moderators
P4: Institutional effects x x Negative (-)
P5: Spatial
Agglomeration x x Positive (+)
Legend: Where “+” means a positive relationship; “-” means a negative relationship, “0” no relationship and “x” not applicable.
CHAPTER 4. METHODOLOGY
4.1 INTRODUCTION
This chapter discusses the methodology that is used in this study. Firstly, it starts with discussing the research method that is used for the purpose of moving towards a theory explaining the relation between global cities and location choices made by MNEs for both the establishment of R&D and production facilities overseas. Secondly, the research design is discussed, consisting of a paragraph on the data collection and data analysis procedure used for this study. Thirdly, the validity and reliability of the study are set out. This paragraph also involves the limitations and boundaries of this study. In accordance with all other chapters, this chapter finalizes with a short conclusion.
4.2 RESEARCH METHOD
This study adopts an inductive, theory-building approach rather than a deductive, theory-testing approach, since the existing literature lacks a predominant theory for these concepts in relation to each other. Consequently, this fits to the research (sub) questions, since these are all “how-questions”, and therefore best fit with a discovery-oriented approach (Voss et al., 2002), which is part of the inductive approach. Furthermore, the study investigates a real life context and therefore a qualitative research approach (i.e. describing rather than proving) is applicable (Yin, 2003). In addition to that, the study follows a multiple case study approach, entailing a parallel (nestled) approach in terms of the cases selected for the research (i.e. German automotive MNEs contrary to French automotive firms).
As stated in the study of McCutcheon and Meredith (1993), a case study stands for, according to Yin (1989): “An objective, in-depth examination of a contemporary phenomenon where the investigator has little control over events.” This definition covers some important facts (McCutcheon and Meredith, 1993). For instance, the data collecting may be through the use of primary sources (e.g. direct observations and interviews of people involved) or secondary sources (as shown in table 4.2-1). This study uses the latter for its data foundation. Table 4.2-1 shows an overview (Eisenhardt, 1989) of the selected cases, the research problem, data sources, investigators and output. The output of the case study entails a conceptual framework that is tested by the several cases of the study. The purpose is to generate general conclusions by investigating the cases in a real life context. Therefore the study uses propositions in order to form the conceptual framework and consequently to generate general conclusions through combining the current theory with the new insights (i.e. theory building). Important to note is that the data sources and associated research method is a qualitative one, which make it possible to create propositions. Future research could extend the study with a quantitative research method in order to really test the propositions or hypothesis forthcoming out of the generalisations made. Table 4.2-1 inductive case study research
Description of Cases Research Problem Data Sources Investigators Output
5 major German and French automotive firms (BMW AG, Daimler AG, VW AG, Renault S.A. and Peugeot S.A.)
Location choices of automotive MNEs in China, the influence of Chinese global cities hereon and the moderating role of institutional effects and spatial agglomeration
Documentary: online archives (e.g. website with statistical departments of each global city), annual reports, peer reviewed journals;
Single investigator Conceptual framework approaching the relation of global cities and location choices made by automotive MNEs in China, in conjunction with the moderating role of institutional effects and spatial agglomeration 6 Chinese global cities (Beijing,
Shanghai, Guangzhou, Shenzhen, Chongqing and Chengdu)
1 Hong Kong global city, Hong Kong
1 Taiwanese global city, Taipei.
Source: Compiled from Eisenhardt (1989).
4.3 RESEARCH DESIGN 4.3.1 DATA COLLECTION
The collection of the data (see also table 4.3-1) will be done case-by-case, meaning that all data sources containing information available for each selected automotive firm will be gathered and stored (only for the years after the crisis started in 2008). When enough information about R&D and production location is gathered for each of the firms, the information gathering process will point at the selected global cities as research object. Then, when also for these cases enough relevant data is selected, the interpretation of the data can begin.
The data selection for this study is done by documentary analysis of for instance annual reports of the selected MNEs and (governmental) online-archives (e.g. the World Investment Report (WIR) for the selected cities (see also table 1)). The specific form of documentary analysis is done by formal qualitative content analysis, which can be, according to Mayring (2000), all sort of recorded communication (e.g. transcripts of interviews done by third parties or other abovementioned types of documentary analysis). These are all available via the Internet and do not require additional fees to download (i.e. all annual reports of the automotive firms are free available to the public). To prevent poor data output, as a potential result of poor case selection, the selection of cases is done by undertaking a brief inspection of the available case companies beforehand (i.e. French and German automotive MNEs). By doing so, the selected cases in this study certainly provide enough (documentary) data for doing a full analyses.
Table 4.3-1 Stages of data analysis
Stage Step Output
1a. The investigation of the
location choices of R&D and production facilities DAT A CO L L E CT IO N 1b. The investigation of annual reports of each of the automotive MNEs
1c. The investigation of other secondary data resources
DAT A ANAL Y S IS A ND R E S UL T S
1d. A detailed map with the facilities spread in China 1e. A table with an detailed identification of the focal companies with associated year of entrance, Chinese partner(s), joint companies and location(s) 1f. A detailed map of the industrial zones of the automotive sector in China, combined with the location of the focal global cities and locations as investigated under stage 2a.
2a. The investigation of the
global cities and the three
underlying factors (see also paragraph 4.4): International
connectedness; advanced producer services;
cosmopolitan environment
2b. The Gathering of data from miscellaneous sources for each of the focal global cities using the measurement criterions as described in paragraph 4.4
2c. An extension of the comprehensive table1of Chinese global cities performance for the each of the underlying factors. This table will finally serve as the joint analysis for the conceptual model, as it involves not only 2a. but also the output as under 1d., 1e. and 1f.
3a. The investigation of the
institutional effects
3b.The investigation of secondary data resources: peer-reviewed literature
3c. A historic timeline of the governmental influence in China, focussed on the automotive sector and several relevant quotes, which highlight certain phenomena.
4a. The investigation of the
spatial agglomeration
4b. The investigation of secondary data resources: peer-reviewed literature
4c. A detailed map of the industrial zones of the automotive sector in China, combined with the location of the focal global cities and locations as investigated under stage 1
DISCUSSION
5a. Combining the output as under 1d, 1e, 1f and 2c 5b. The consequence of the outcomes as found under 3c and 4c on the relationship as under 5a
5c. Finding certain patterns or inconclusiveness between the focal cases and apply these to the conceptual model and propositions
5d. Answers to the propositions
6a. Final integration of overall results
6b. Attempt to make a generalization for theory building purposes
6c. Answer to the main research question and sub questions; Implications, future research directions and limitations of the study
4.3.2 DATA ANALYSIS
The interpretation (see also table 4.3-1) will firstly focus on trying to find explicit patterns in the location choices of the automotive firms (or contradictions). It does so by first identifying the five automotive cases in terms of their partnerships with Chinese manufacturers. Then the joint companies are identified and lastly the step of looking where the production and R&D facilities are outsourced in China is made. This data will be listed in a table and then will be literally mapped to gain an overall oversight of the cases’ facilities throughout China. The next step is to combine this first map with data on the industrial zones of all automotive firms in China.
Secondly, the global cities come into the analysis process. The analyses proceed to look at the global cities and each of the configurations of these cities in terms of the attributes as described in the study. It does so by extending an existing scientific analysis on five of the cities that are selected for this study. The missing three cities are examined in the same manner, as done in the current scientific article and added to the table. The table consist of information for each of the three factors (i.e. International connectedness, advanced producer services and cosmopolitan environment), which make it possible to make generalisations after it is combined with the data found on the automotive cases.
Thirdly, the analysis will look into the institutional effects, which is done through combining several secondary data sources on institutional influences each of the automotive firms have dealt with. As for the spatial agglomeration, this moderating effect will be revealed through the overall combining of data, which is the last step in the data analysis process.
Thus, lastly, the analysis proceeds with combining the two maps (i.e. facilities map overview and industrial zones overview), the global cities configuration analysis,
and the institutional ‘story telling’ data. Again, it is important to search for certain patterns or contradictions, with the purpose to find out, making use of all the data gathered and the interpretation of this specific data (i.e. the eventual patterns and configurations found in the data), what the effect of global cities is on the location choice. It could also be that the effect is not verifiable at all, which can also be a viable conclusion and opportunity for future research. In the discussion contained in chapter 7, this all will be described in detail and conclusions will be attached.
4.4 VARIABLES
Advanced producer services: As part of the independent variable global cities, the advanced producer services is measured for each of the cities through the number of listed companies, domestic market capitalisation, value of share trading, yearly FDI and number of MNS Top-500 headquarters. These measurements fit with the description of advanced producer services by Sassen (1991, 1994).
International connectedness: Partly indicates how a certain city is connected to the rest of the world. Therefore, to measure international connectedness of the global city cases, the city airports, seaports and subways as well as the total value of import and export and the number of international trade fairs held are measured, which partly represents the examples given by Friedman (1986).
Cosmopolitan environment: The cosmopolitan environment is measured by analysing the foreign visitors arrivals, the permanent foreign population of the city, number of universities and GDP per capita. To enrich the assessment of the focal global cities, an example of a key mega-event will be given for each of the global cities. This is in line with the examples given by Hall (1966).
4.5 VALIDITY AND RELIABILITY
The strengths and limitations of the study’ research design are an important issue in terms of the quality of secondary sources (i.e. Authenticity, Credibility, Representativeness and Meaning (Scott, 1990), but also for the rigor (O’Sullivan, 2014) of the study (e.g. the degree of generalizability of the study as described below). One of the strengths of this study’ research design is the use of data from both the selected automotive companies andexternal sources like governmental databases and newspapers, which results in credible secondary data.. A more negative point in terms of rigor is that the study is not per se a generalization applicable for other sectors, as sectors can differ extensively from each other (i.e. the automotive compared to pharmaceutical and chemicals industry in terms of foreign R&D expenditure (UNCTAD, 2005). Nevertheless it can provide useful insights in the location choice of R&D and production facilities specifically made in the German and French Automotive sectors, and the influence of global cities in this choice.
4.6 CONCLUSION
This chapter discussed the methodology that is used in this study. It elaborated on the Case study approach as research method used for this study. After that, the research design respectively the validity and reliability were described. A schematically description of the stages, steps and outcomes of the data analysis process was provided. Lastly, this chapter stressed some limitations of the specific method used in this study, to be further elaborated in the discussion section of chapter 7.
CHAPTER 5. THE CHINESE AUTOMOTVIE INDUSTRY
This qualitative study is built on an empirical case of location choices of five European automotive firms in China (see appendix A for a map of China and Appendix B for a timeline of important events). As many other automotive MNEs, these firms have to consider whether or not to develop foreign R&D and production locations or offshore these facilities, due to growing globalized markets and competition. The automotive industry is a very suitable industry for analysing locational strategies as spatial dimension (i.e. spatial agglomeration) is of exceptional importance in this sector (Colovic and Mayrhofer, 2011). Especially European automotive MNEs, traditionally from the so-called Triad region (i.e. Western Europe, North America and Japan) felt the growing importance of a global automotive industry. Nowadays, these companies are still pursuing to locate production facilities in emerging markets, and to a lesser extent, R&D facilities as well. Since China has become the largest automotive market in the world in 2009 (CCFA, 2010), this study concentrates the analysis on this country. Also, the automotive industry in China is highly intertwined with the institutional environment in terms of obligations and requirements imposed by the (local) governmental institutions (i.e. institutional effects), which, in order to remain competitive, requests optimization of locations where activities of automotive firms in China take place (Schmid and Grosche, 2008). So this could be regarded as another motive, which makes China even more relevant as a context for research.
Additionally, in this study it is assumed that the existence of global cities in a country (China for this particularly study) does not automatically mean the presence of R&D and production facilities of automotive firms. Instead, the study
