Location choice and cities in the automotive industry

Location choice and Cities in the automotive

industry

Master Thesis

MSc. Business Studies – International Management

Supervisor:

Dr. Johan Lindeque

Second reader:

Dr. Lori DiVito

Student:

Dan Danes

Student ID:

10603115

Date:

27.06.2014

A

This research studies the suitability of the city as a level of analysis for evaluating Multinational Enterprises’ (MNE) location choices within their regional strategies. In comparison to the regional and country levels of analysis, the city level has been underexplored so far. Using a qualitative multiple case study design, the internationalizations of six MNEs are examined, focusing on their selection of cities as location choices. The sample consists of six MNEs that are selected according to their home country, home region and regional orientation. The cities in which the MNEs internationalized are divided into three categories of cities: tier I, tier II and tier III cities. Each city type has a specific role in the regional strategies of MNEs. The analysis of the data revealed that tier I cities, which are the global cities, are selected for strategic asset seeking motives and that they are able to decrease Liability of Foreignness (LoF). Tier I and II cities are selected for market seeking reasons while tier III cities are selected to locate the lowest level of operations and these cities are associated with efficiency seeking and resource seeking motives. Therefore, cities emerge as a highly relevant level of analysis that contributes to a better understanding of MNEs’ location choices.

K

:

A

First of all, I would like to thank my supervisor Dr. Johan Lindeque. His feedback, guidance and constructive criticism have helped me successfully finalize this paper that I am really proud of. He inspired me to work very hard on this thesis because of his enthusiasm and encouragement during all the discussion we had. His constant motivation made me push my limits.

Second, I wish to thank my family for giving me the opportunity to come and study in Amsterdam. It has been a great year in which I have learned many things. I will never forget the experience I had here.

Last but not least, I would like thank Anca for her love and constant support, during both good and bad times. Although the physical distance was huge, I never felt alone.

C

ONTENTS

Abstract ... 1

Keywords ... 1

Acknowledgements ... 2

Index of Tables and Figures ... 5

1. Introduction ... 6

2. Literature review ... 9

2.1. International Strategy and Regionalization ... 9

2.1.1. Liability of Foreignness ... 10

2.1.2. FSAs and CSAs ... 11

2.1.3. Industry Specificity ... 11

2.1.4. Asset Specificity ... 12

2.1.5. Entry modes ... 12

2.2 Location Choice ... 13

2.2.1. FDI Motives ... 13

2.2.2. Earlier conceptualization of Location Choice ... 14

2.3. Cities ... 16

2.3.1. Tier I Cities ... 17

2.3.2. Tier II Cities ... 18

2.3.3. Tier III Cities ... 19

3. Methodology ... 20

3.1. Philosophies behind the research ... 20

3.2. Qualitative research: multiple case studies ... 20

3.2.1. Quality criteria ... 21

3.2.2. Case Selection – the automotive Industry ... 22

3.4. Data Collection ... 24

3.5. Data Analysis ... 26

4. Within-case analysis ... 28

4.1. Japanese automotive companies: Mitsubishi and Toyota ... 28

4.1.1. Cities and Location choice with respect to fdi motives for Mitsubishi ... 28

4.1.2. Cities and Location choice with respect to fdi motives for Toyota ... 30

4.1.3. Japanese MNEs and Location choice: Cities ... 33

4.2. US automotive companies: Ford and General Motors (GM) ... 34

4.2.1. Cities and Location choice with respect to fdi motives for General Motors ... 34

4.2.2. Cities and Location choice with respect to fdi motives for Ford ... 37

4.2.3. US MNEs and Location choice: Cities ... 40

4.2. German automotive companies: Daimler and Volkswagen ... 41

4.3.1. Cities and Location choice with respect to fdi motives for Daimler ... 41

4.3.2. Cities and Location choice with respect to fdi motives for Volkswagen (VW) ... 44

4.3.3. German MNEs and Location choice: Cities ... 48

5. Cross-case analysis ... 49

6. Discussion ... 53

7. Conclusion ... 56

7.1. Limitations and suggestions for future research ... 58

7.2. Managerial implications and Scientific relevance ... 59

8. References ... 60

9. Appendix ... 66

I

T

Table 1. The three categories of cities ... 17

Table 2. The sample ... 23

Table 3. The sample and the internationalizations ... 25

Table 4. The sample and the collection of newspaper articles ... 26

Table 5. The codebook ... 27

Table 6. Categorization of cities ... 28

Table 7. Mitsubishi’s internationalizations ... 29

Table 8. Toyota’s internationalizations ... 31

Table 9. Japanese MNEs’ FDI Motives ... 34

Table 10. GM’s internationalizations ... 35

Table 11. Ford’s internationalizations ... 37

Table 12. USA MNEs’ FDI motives ... 40

Table 13. Daimler’s internationalizations ... 41

Table 14. Volkswagen’s internationalizations ... 44

Table 15. German MNEs’ FDI Motives ... 47

Table 16. Cross-case analysis ... 50

Table 17. Results of the working propositions ... 53

1.

I

NTRODUCTION

The location choices of Multinational Enterprises (MNE) have been studied intensely, as researchers agree that the decision of choosing a foreign country in which to invest is critical for the success of an MNE (Dunning, 1998). Studies have shown that the impact of the country upon the subsidiary is as strong as the industry in which the company operates, therefore making the location decision at least as important as the selection of the industry in which the MNE operates (Chan, Isobe, & Makino, 2008). The emphasis has moved from understanding how subsidiaries take advantage of different markets by using capabilities developed in their home country and home region to the advantages that host locations can generate, enhancing the capabilities of the MNE (Piscitello, 2011). The focus is now on selecting a location where the combination of existing Firm-Specific Advantages (FSAs) and Country-Specific Advantages (CSAs) will generate more value for the whole network, positively affecting all subsidiaries.

Nevertheless, as the environment switches towards a global economy, the proportions of global FDI in different countries have changed, depending not only on local economic and political activities but also on other country-specific factors that characterize the country of origin and the target country (Dunning, 1998). Capital, labour, natural resources, market size, knowledge and skills are all influencing the attractiveness of a host country (Cantwell, 2009). However, the location choices of MNEs are driven by the FDI motives. There have been identified four FDI motives: three traditional motives (natural resource seeking, market seeking, efficiency seeking) and a modern motive (strategic asset seeking) (Dunning, 1998).

One of the recent debates in International Business (IB) literature highlights the contrast between two views: on one hand, the continuous globalization of the world that stimulates the unification of markets into a global one (Garrett, 2000), while on the other hand the fragmentation of the world into regions (Rugman & Verbeke, 2004). The main stimulus of globalization is represented by MNEs because they are responsible for a large proportion of global trade and FDI. The Fortune Global 500 list of MNEs represents the largest and most important MNEs constantly seeking opportunities for growth and development. However, most MNEs are not as global as their representatives affirm with pride. Rugman and Verbeke (2004) have demonstrated that few of the world’s 500 largest MNEs are global: in 2004, only 9 companies were found to be truly global. It is true that these giants are spread across the world, mostly in the three regions of the extended Triad (North America, Europe and Asia); however, their focus is not uniform. Each company tends to have a focus on a certain region and, intuitively, it usually is their home region, as it has been discovered by a number of studies (Filippaios & Rama, 2008; Rugman & Girod, 2003; Rugman & Verbeke, 2004; Rugman & Collinson, 2004). The globalization process has not managed to completely erase the differences between cultures and

countries and the borders between regions. This situation, also called semi-globalization, is characterized by an incomplete integration of cross-border markets (Ghemawat, 2003), which has led the companies to replace their global strategies with strategies at regional level (Arregle, Beamish, & Hébert, 2008). Rugman and Verbeke’s (2004) regionalization argument represents a firm level expression of this phenomenon, as captured by MNEs’ strategies.

The strategy of an MNE, whether it is regional or global, directly affects the firm activities at an international level, impacting the entry mode choice, their degree of local responsiveness or global integration but most importantly, their location choice. A large number of studies have evaluated the implications of location choice decisions of firms from numerous perspectives, such as analysing the distance between two countries (Ghemawat, 2001; Hofstede, 1984), the trade agreements (Copelovitch & Pevehouse, 2013) and other institutional barriers (Mansfield & Reinhardt, 2008).

The focus of analysis in IB has gradually shifted from the country level of analysis, where FDI in countries is measured, to the MNE level, where investment decisions are analysed, moving on to the subsidiary level (Rugman, Verbeke, & Nguyen, 2011). At the subsidiary level, knowledge flows between units has been the predominant topic: the isolation of some subsidiaries through geographic distance has been studied (Mudambi & Navarra, 2004), but the subnational level was never included. Another study conducted at the subsidiary level studied the development of FSAs through the subsidiaries (Birkinshaw, Hood, & Jonsson, 1998). It was discovered that the local environment was clearly an influence, underlining the local competition and the strategic importance of the country (Birkinshaw et al., 1998); however, the study did not seek to explain the influence of the subnational level where the subsidiaries are located. The first studies that focused on micro-locations were preoccupied with clusters, re-examining their role in the strategic choices of MNEs (Porter, 2000). They were followed recently by papers studying agglomeration effects such as knowledge spillovers and information externalities, discovering that the location choices of different MNEs are isomorphic not only at a country level, but also at a micro-locational level (Mariotti, Piscitello, & Elia, 2010). However, the common point of these studies is that they have gone through multiple levels of analysis but missed a critical level: the subnational level, represented by cities. International Business (IB) scholars have paid limited attention to understanding how cities influence the international strategies of MNEs. By contrast, sociology has seen a rise in the importance of micro-locations, specifically cities. Global cities have become a strong point of interest because of their peculiarity: these cities represent strong financial, economic, political and cultural centres (Sassen, 2004). Previous research has integrated these cities within the globalization discussion, concluding that these large global cities are found in large “globalization arenas” (Beaverstock, Smith, & Taylor, 1999). Hymer (1972) argues in his seminal work that the effects of globalization are not distributed evenly across countries; they are in fact centralized in cities. These global cities are now tied to their position in the global economy

and the way in which MNEs locate their operations there is directly affecting the economic development of the city, impacting employment, consumption and the spread of wealth (Hymer, 1972). These cities have a strong link with the companies that are located within them: the success of a company can represent the rise of a global city, while the downfall of a city can have a strong negative impact on local companies. Still, the reasons why MNEs chose to locate in a specific city as part of their regional strategies have yet to be found and there are still questions regarding the relationship of cities to earlier location choice studies. Therefore, the research question of this study is the following:

RQ: How do cities influence the location choices within the regional strategies of MNE?

This study attempts to explain the implications of cities within the regional strategies of companies. In order to answer this research question, the focus will be on the automotive industry. Six companies that operate in this sector will provide the object of study. These companies have been selected taking into consideration their regional orientation, home region and home country. Their regional orientation has been calculated using longitudinal data recorded between 2008 and 2012. Regarding their home region, each pair of MNEs comes from one of the three regions of the Triad: North America, Europe and Asia-Pacific. Moreover, they come from three countries that are known for their automotive heritage and that are now leaders in the production of motor vehicles in their region: United States, Germany and Japan.

The automotive sector is particularly interesting because the selected companies, which are featured on Fortune Global 500 list, are large organizations that internationalize heavily in order to obtain competitive advantage. They attempt to achieve a global network by efficiently spreading their facilities in different places in order to localize production and marketing and to extend the coverage of auto dealerships. The internationalizing firms also aim to take advantage of the growth opportunities and to increase profitability by tapping into new local markets, while also spreading production and thus decreasing risks. Previous studies have analysed the strategies of automotive companies at the global, regional or country level of analysis (Worrall, Donnelly, & Morris, 2003). For example, the success of Toyota has been analysed through the use of the double-diamond concept, underlining that the skilled workforce, well-established relationships with suppliers and sophisticated demand, which are all characteristic for Japan, have greatly influenced the company (Rugman, 2005). However, no researcher has ever attempted to narrow the focus down to the micro-locational level, the city level, to observe how the internationalization process is altered and how the regional strategies are influenced by the city as a location choice driver.

In order to answer the research question, this paper starts off with a review of the existing concepts in the regionalization and location choice literature. First, the main concepts in international strategy are presented, followed by Dunning’s FDI motives (1998). Finally, the focus is switched upon the existing

city literature. After the literature review, the methodology of this study is discussed: the research philosophies, the quality criteria and the multiple case study are presented. The next section contains the within-case and cross-case analyses, followed by the discussion of the working propositions and the inclusion of the city level of analysis within the location choice studies. The conclusion contains the key findings, as well as the limitations of the study, suggestions for future research and managerial implications.

2.

L

ITERATURE REVIEW

The literature review starts off by discussing the regionalization phenomenon and by presenting the main concepts in IB literature, such as Liability of Foreignness (LoF), FSAs and CSAs, asset specificity, industry specificity and entry modes. The focus is then moved on the nature of previous location choice studies and then FDI motives. The chapter ends with a brief overview of existing city literature, which mostly comes from streams of economic geography literature and finally, a three-tier categorization of cities is presented.

2.1.

I

S

R

Nowadays there are two conflicting views in the IB literature regarding the contrast between globalization and regionalization, also known as semi-globalization: the first view states that globalization is a strong force that drives cultures and countries, and consequently markets, to become similar and uniform while the second view affirms that local contexts still remain very different from each other and there are numerous benefits that can be reaped by MNEs (Meyer, Mudambi, & Narula, 2011). Globalization has not been able to completely erase the differences between cultures, countries and regions (Rugman, 2003). There are still numerous differences across countries and regions which result in MNEs operating in unique environments. This state, also called semi-globalization, is characterized by an incomplete integration of cross-border markets (Ghemawat, 2003), which has led the companies to replace their global strategies with strategies at regional level (Arregle et al., 2008). Globalization is driven by three processes which are cross-border trade, integration of financial markets and international cross-border investment (Garrett, 2000). The main driver of globalization is represented by MNEs. Their overall activities and progress have led the public to think that these MNEs are global companies that do business in all regions of the world. However, in their seminal work, Rugman and Verbeke (2004) prove that actually few companies in the Fortune Global 500 are truly global and instead operate mostly in their home region, having a small level of penetration in countries that are part of other regions of the Triad. This is due to the fact that the regional strategies of these MNEs co-evolve with the institutional environment characteristic for the region where they

operate (Cantwell, Dunning, & Lundan, 2010). MNEs are faced with the challenge of organizing their operations in heterogeneous contexts at two levels: firstly, they must balance the integration of subsidiaries within the whole MNE network and secondly, they must balance the MNE network across multiple host locations with different contexts (Meyer et al., 2011)

Rugman and Verbeke (2004) have studied the world’s largest 500 MNEs and have found that the dispersion of their sales is geographically imbalanced, leading to the identification four types of MNEs, according to their orientations: home region oriented, host region oriented, bi-regional and global. The majority of MNEs are home-region oriented, concentrating their downstream activities within their home region and around the home country. When the growth within the home region becomes exhausted, MNEs try to expand into another region, deploying their downstream activities in two regions and thus becoming bi-regional (Rugman, 2005). There is also the possibility of an MNE to have most of its sales in a different region that its home region, therefore becoming a host region oriented firm. Finally, only the MNEs that manage to spread their activities evenly across all regions are considered global (Rugman & Verbeke, 2004). The work of Rugman and Verbeke’s (2004) has been criticized because the regional orientation of MNEs is also influenced by the home country: there are numerous home region oriented MNEs, which is the result of having a large proportion of sales recorded in the home country.

2.1.1.

L

F

An important aspect of the FDI process is generated by the fact that MNEs investing in foreign countries have non-local status. Companies are burdened by costs such as spatial distance, costs generated by host or home location or general unfamiliarity costs (Sethi & Judge, 2009). These disadvantages of not being local are translated in the performance of firms, which is negatively impacted. The difficulties that arise for these multinationals in their internationalization process can be measured using the concept of distance, the cost of doing business abroad or the liability of foreignness (LoF) (Ghemawat, 2001; Sethi & Judge, 2009). For this study, the liability of foreignness is the most appropriate. LoF is one of the main reasons why MNEs are not becoming global and instead operate at a regional level (Rugman & Verbeke, 2007). Moreover, this concept has been adapted to the current debate about regional strategies. Rugman and Verbeke (2007) differentiate between two types of expansion: intra-regional, when FDI takes place within a single region and inter-regional, when FDI takes place across multiple regions. Therefore, it is required to make the distinction between two types of LoF: intra-regional LoF and inter-regional LoF (Rugman & Verbeke, 2007). The former is encountered when expanding intra-regionally, between countries, while the latter is seen when expanding between two regions. Inter-regional LoF has a direct relationship with inter-regional diversification, impacting the performance of the company (Qian, Li, & Rugman, 2013). The

two types of liabilities both influence inter-regional diversification, while inter-regional LoF has a bigger effect on firm performance than intra-regional LoF (Qian et al., 2013).

2.1.2.

FSA

CSA

The whole process undertaken by MNEs is a result of the combination of three factors that form the OLI framework: internalization advantages, location advantages and ownership advantages (Dunning, 2000). However, while the OLI framework has been extensively used in IB studies, the FSA/CSA concepts are better suited here. Firm-Specific Advantages (FSAs) are competences and capabilities that are unique characteristics of a firm, deriving competitive advantage out of them (Rugman et al., 2011). Besides giving the edge over competitors, FSAs of a firm are also aiding the firm in overcoming the Liability of Foreignness (LOF), reflecting Hymer’s (1976) original arguments. There are two types of FSAs: location bound (LB) FSAs and non-location bound (NLB) FSAs. Also, FSAs can be divided into tangible and intangible FSAs. The success of a company in its home country is the result of a combination of NLB FSAs with LB FSAs. The reason why many MNEs don’t become global is their inability to transfer their NLB FSAs across regions and to access local LB FSAs (Rugman & Verbeke, 2007).

As firms internationalize and transfer their FSAs across borders, these core capabilities interact with the Country-Specific Advantages (CSAs). CSAs are economic, political, institutional and cultural factors that are exogenous to the firm and are affecting its strategy (Rugman, 2005). These provide comparative advantage that can be leveraged by a firm using its FSAs, therefore resulting in competitive advantage. The whole process of internationalization can be seen as a combination of Firm-Specific Advantages (FSAs), which are the equivalent of ownership advantages, and Country-Specific Advantages (CSAs), representing the location advantages (Rugman, Oh, & Lim, 2012). From the combination of these two factors emerges a specific configuration which is highly contextual, depending on economic and institutional factors. Out of this configuration will emerge one of the four FDI motives that drive the investments of MNEs (Dunning, 1998). Therefore, strategies of MNEs are driven by the interactions between FSAs and CSAs.

2.1.3.

I

S

The regional orientation of MNEs is also influenced by industry characteristics. Researchers have studied the specificity of numerous sectors, such as the automotive industry (Rugman & Collinson, 2004), banking (Grosse, 2005), retail (Rugman & Girod, 2003), electric utilities (Kolk, Lindeque, & Buuse, 2013) and cosmetics (Oh & Rugman, 2006). Sectors that are “Location-bound”, such as banking, telecom or retail and other service sectors, are faced with numerous barriers imposed by governments, such as environmental, corporate, labor and other regulations. Consequently, their

regional sales volume is lower than manufacturing sectors (Rugman, 2005). On the other hand, some manufacturing sectors, such as computer and electronics, are global because the components are standardized, their supply chain is global and there are no government regulations. In case of the automotive sector, the orientation of car companies varies: there are home region oriented, host region oriented, global and bi-regional companies (Rugman & Collinson, 2004). All these companies are focused on the North-American market, since it is the most fragmented: domestic car producers hold only 63% of the market share, whereas in Japan imports represent only 4.5% of the total sales (Rugman & Collinson, 2004). Therefore, companies in the service sector do not expand their operations into foreign regions as much as the manufacturing companies do, affecting their regional strategies. In consequence, the regionalization of companies depends on the specific characteristics of the industry.

2.1.4.

A

S

In addition to industry specific factors, MNEs’ presence at a global level is also influenced by firm-specific factors. Following Rugman’s (2012) distinction between upstream and downstream activities, MNEs’ various FSAs can be separated for the different value chain activities. Upstream activities, such as production, R&D or logistics, are spread across the globe because firms want to take advantage of market imperfections in order to obtain competitive advantage; for this, they require strong upstream FSAs (Rugman & Verbeke, 2008). On the other hand, in order to expand downstream activities beyond the home region, strong downstream FSAs are required in order to capitalize on foreign markets. Still, these FSAs are location-bound and cannot be expanded globally because they contribute to the survival and profitability of the company within the home region and therefore are not easily transferrable across other regions (Rugman & Verbeke, 2008). Because of this, the penetration of foreign markets is influenced by these FSAs and therefore business units within a firm each have a different level of regional or global activities. As the degree of internationalization is measured by the spread of sales and assets across the globe, upstream activities, represented by assets, are less internationalized than downstream activities, represented by sales (Rugman et al., 2012). Moreover, MNEs also differ in their capabilities of taking advantage of different CSAs, therefore affecting their global expansion.

2.1.5.

E

In the internationalization process of MNEs, the next step after choosing a location is to decide how to enter that specific market. Entry mode discussion has become a popular subject in the IB field and still continues to be debated by researchers (Beugelsdijk & Mudambi, 2013; Buckley & Casson, 1998; Dikova & Van Witteloostuijn, 2007) because it is a core element of the internationalization concept. There are two main streams of discussion in this field: the entry mode choice, referring to the degree

of ownership, and the establishment mode choice, pointing at the contrast between acquiring an existing firm or making a greenfield investment (Dikova & Van Witteloostuijn, 2007). Export, Joint Ventures (JV) and Wholly-owned subsidiaries (WOS) are considered to be on a continuum of increasing risk and investment (Brouthers & Hennart, 2007). A Joint-Venture is characterized by a lower degree of risk and commitment than a partial acquisition; finally, a full acquisition represents the largest degree of risk and commitment that a firm can be faced with (Pan & Tse, 2000). Joint-Ventures and partial acquisitions are employed when the firm is faced with a high degree of LoF (Goerzen, Asmussen, & Nielsen, 2013).

2.2

L

C

As many pioneers of the International Business field have recognized (Dunning, 1998; Iammarino & McCann, 2013; Vernon, 1966), one of the most important aspects of the internalization process of MNEs is the selection of a location to expand operations abroad. The reason why it is such an important strategic choice is the fact that it requires considerable investments: resources, expertise and time (Schotter & Beamish, 2013). According to Vernon’s (1966) early seminal work, there are many external forces that influence the location choice of an investment, such as labor costs, transportation costs but also communication costs and problems associated with host country’s economic and political environment. In concordance with the age of a product, firms have different needs: for new products, the MNE needs to have an excellent communication with suppliers and customers because the product has not reached the standardization stage and it can still be subject to change. On the other hand, an enterprise crafting an already mature product no longer needs flexibility, as the focus is on reaching economies of scale (Vernon, 1966). Therefore, there are shifts in the positioning of operation facilities in correlation with the needs of the firm.

Recently, as Dunning (1998) notes, the emphasis has been switched on the importance of intellectual capital, globalization and alliance capitalism. These elements, enforced by the integration of international markets and continuous technological progress, have completely changed the nature of FDI and location choices, positively altering the capabilities of MNE to adapt to new host environments. The balance of FDI among countries has been changed in favor of emerging countries, where companies can increase efficiency. Still, the geography of FDI is unique for each country, in close connection with economic, social and political factors (Dunning, 1998).

2.2.1.

FDI

M

MNEs establish operations abroad because of different advantages, such as CSAs, that characterize host locations. There have been identified four different motives: resource-seeking, market-seeking,

efficiency-seeking and strategic asset-seeking locations (Dunning, 1998). Each investment motive is related to different particularities of a location that, in combination with the FSAs of the MNE, generate competitive advantage. Resource seeking investments are targeting host locations that are endowed with tangible or intangible resources, such as natural resources and raw materials, labor force, investment incentives, or infrastructure; in case of automotive companies, the most sought-after resource is represented by cheap labor force. There are two types of resource seeking MNEs: general asset seekers, which are interested in physical resources and labor, and special asset seekers, who look for organizational or managerial capabilities and skills (Iammarino & McCann, 2013). Market seeking investments look for host country locations where there market and the demand are growing (Dunning, 1998). The efficiency seeking motive is related to foreign locations that enable them to rationalize their resources and investments. Clusters, supporting industries, and institutional support are all efficiency seeking motives. For example, automotive companies that seek to become more efficient will invest in cities that are located in clusters where suppliers, OEM manufacturers and distributors are also located. Finally, the strategic asset seeking motive drives MNEs to look for locations in which they can acquire valuable FSAs and capabilities that allow the MNE to reinforce its position and to gain competitive advantage (Iammarino & McCann, 2013). A typical strategic asset seeking investment is represented by acquisitions: often, large companies buy other smaller companies in order to extend their portfolio of brands or to obtain other intangible FSAs such as knowledge, patents or access to new technology. While the first motive has seen a decrease in recent times, most of the MNE’s investment is done to capture the advantages of local economies and institutional frameworks. The last motive has seen an important rise in the last two decades, signaled by the large number of mergers and acquisitions (Chapman, 2003).

2.2.2.

E

L

C

The location choice of MNEs has been studied from many points of view, using frameworks such as the OLI (Dunning, 1998), the Uppsala internationalization model (Johanson & Vahlne, 1977) or Ghemawat’s (2001) and Hofstede’s (1997) frameworks that measure distance between home and host location. The majority of location studies has taken the country as the unit of analysis (Chan et al., 2008; Makino, Isobe, & Chan, 2004; Meyer et al., 2011). As Beugelsdijk and Mudambi (2013, p.1) note, “location (L) in IB is almost always conceptualized and operationalized at the country level of analysis.” This central role of the country has been given by the role of national governments, their institutional context and international trade organizations but also by the historical aspect that national borders had in the past.

The literature has few studies that focus on the sub-national level of location and the majority of them focus on agglomeration effects, discovering that MNEs do not agglomerate with domestic firms in

order to avoid knowledge outflows unless these local companies own some kind of advantage but tend to agglomerate with other foreign MNEs because of the high chances of knowledge inflows (Mariotti et al., 2010). Also, there are papers that focus on clusters, underlining that companies co-evolve in clusters due to knowledge spillovers, mainly driven by foreign MNEs that locate there (Rugman & Verbeke, 2003). Clusters are strategically important for foreign MNEs in knowledge-intensive industries, stating that firms benefit from the intellectual property of the host economy (Mudambi & Swift, 2012). Most of the papers focus on flows of knowledge between the subsidiary and the environment (Iammarino & McCann, 2013; Mariotti et al., 2010; Mudambi & Navarra, 2004). Studies have also focused on agglomeration effects, clusters, supplier outsourcing and how MNEs influence each other’s location choices (Porter, 2000; Rose & Ito, 2008; Schmitt & Van Biesebroeck, 2013). It has resulted that the location choices of companies are similar because of uncertainties in the host market and investment constraints, generating similar FDI strategies (Rose & Ito, 2008) and that the co-location of firms and suppliers is not a necessary condition for successful sourcing (Schmitt & Van Biesebroeck, 2013).

Previous streams of IB literature have used the country as a unit of analysis because of the cross-border activities that MNEs undertake. But the significance of state cross-borders is constantly decreasing because of globalization (Iammarino & McCann, 2013). It has been emphasized that the focus should move from country level to regional level. The home country should be replaced by the home region, since this is the place where the majority of MNEs do most of their business activities (Rugman & Oh, 2012). In accordance, firms should enhance their regional experience instead of their country experience, which is more relevant for their progress. However, the utilization of countries and regions as the unit of analysis of the location choice has become questionable since these units are too vast to capture all the details of a location. Moreover, globalization, with all of its implications, has lowered the significance of these units of analysis and has brought into attention the sub-national level, the cities (Iammarino & McCann, 2013). The process of selecting a location for a future investment has a hierarchical structure. The first level of decision is concerned with the selection of a country and the second level of decision implies the selection of a subnational geographic location (Iammarino & McCann, 2013).

This subnational level has been overlooked in IB research streams because it narrows down the characteristics of a country and determines more specific locational features. Global cities represent important centers of influence while being part of the worldwide network that leads the global economy (Goerzen et al., 2013). These cities are distinct from the previous units of analysis because they are more focused and narrow; therefore, the utilization of cities generates significant insights in the strategies of MNEs. The activity and overall performance of MNEs is affected firstly by firm-level and subsidiary-level factors, but also by micro-locational factors that can be seen at the global city

level, such as connectivity, advanced producer services and cosmopolitan environment (Goerzen et al., 2013).

2.3.

C

Global cities have been studied in connection with globalization, both in sociology (Sassen, 2004), economy (Goerzen et al., 2013; Hymer, 1972), and economic geography (Beaverstock et al., 1999; Beaverstock, Smith, Taylor, Walker, & Lorimer, 2000; Beaverstock, Smith, & Taylor, 2000). Large cities provide MNEs with numerous opportunities to increase the efficiency of their operations and their productivity. World cities drive the growth of economies in developed regions while the lack of any large cities in poor regions hinders the development of these areas (Iammarino & McCann, 2013). Because of globalization, state borders have become more and more insignificant, resulting in a borderless world that consists of a network of cities connected through material and immaterial flows, such as information flows (Beaverstock et al., 2000). As a consequence, global cities have become valuable because of the flows that go through them and not because of what is fixed within them (Sassen, 2004). Because the global city literature is approached mostly from a sociological perspective, it is important to note that the value of a city is measured through the degree of global “connectivity”, which is established by measuring the flows of knowledge and information that pass through it; also, global connectivity can be measured through corporate functions, decision-making links, research collaborations or trade exchanges (Iammarino & McCann, 2013).

The first study that attempts to classify world cities belongs to Hymer (1972). Based on the corporate structure identified by Chandler and Redlich (1961), there are three levels of cities corresponding to each level of business administration: the lowest level, where day-to-day operations are managed, the second level where managers that coordinate the day-to-day operations are located and the first level, where top management is located (Hymer, 1972). According to location theory, all these activities are spread over the globe in different cities (Hymer, 1972). Other researchers (Asmussen, 2009; Beaverstock et al., 1999) each use their own categorization system for separating cities into different tiers. However, the problem is that there currently is no unanimously accepted system of measurement of world city functions and relations between them, since there is a lack of necessary data; therefore, a final methodology and classification for world cities has not been created yet (Beaverstock et al., 2000; Taylor, 1997).

Country-Specific Advantages (CSAs) represent an important element in the regional strategies of MNEs. However, since the focus of analysis is switched from the country level, it is necessary to adapt the CSA concept for the city level. Each specific micro-location is endowed with its specific advantages, which may be represented by local resources, geographical advantages and other

level advantages. such as international connectedness, cosmopolitan environment or advanced producer services (Goerzen et al., 2013). City-Specific Advantages (City-SAs), such as proximity and accessibility are becoming increasingly valuable. However, these advantages are insignificant in comparison with the role cities have in fostering innovation, knowledge and creativity, generated by their human capital (Iammarino & McCann, 2013).

For this study, a three-tier classification of cities is suitable. Three-tier classifications have been used before (Beaverstock et al., 1999; Beaverstock et al., 2000; Hymer, 1972). Following Hymer’s original denomination, this study considers tier I, tier II and tier III cities (Hymer, 1972). Each type of city is endowed with different City-SAs that make it at an attractive location for MNEs

2.3.1.

T

I

C

Tier I cities, the global cities, are characterized by high knowledge flows, global connectedness and advanced producers services (Goerzen et al., 2013). These cities are usually located in the world’s most productive economies (Iammarino & McCann, 2013). The workforce of these cities is highly skilled and specialized; also, the presence of universities generates a large pool of intangible knowledge, facilitating the use of specialized assets and transforming these cities into large centres of innovation and information technology, making them attractive for knowledge-based MNEs to develop intangible FSAs here (Iammarino & McCann, 2013). MNEs are attracted by these cities, as the flows of knowledge facilitate highly specific investments there, such as R&D centres. Moreover, the high degree of connectedness enhances the transferability of NLB FSAs in host locations. Tangible LB FSAs such as plants and factories are highly specific investments and are spread across all three categories of cities and therefore they are not related to any category.

All global cities are part of a global city network, having intangible and tangible flows that connect them (Sassen, 2004). The effect of this network of global cities has been studied, discovering that

Table 1. The three categories of cities. (Source: Author)

Tier I City Tier II City Tier III City

Ci

ty

-S

As

- advanced producer services - part of global city network - high information & material flows

- highly skilled workforce - sophisticated market - presence of expatriates - links with other global cities

- large market - strategic location - specialized economy

- presence of industry partners - less concentration

- sometimes these cities are capitals of less-developed nations - cheap labor - resources - local embeddedness - strategic positioning (proximity) - infrastructure - available land space

global cities reduce MNEs’ LoF: global cities helps reduce the discrimination of the MNE’s products in a country while also facilitating the communication with other subsidiaries of the MNE (Goerzen et al., 2013). Also, it is easier to obtain information about the local environment while being in a global city because of the presence of numerous local experts that are used to dealing with foreign MNEs (Goerzen et al., 2013). The large pool of expatriates enhances the coordination of MNEs’ activities (Belderbos & Heijltjes, 2005), while also managing the cultural diversity (Goerzen et al., 2013). Tier I cities host a large proportion of foreign MNEs, having numerous links with other global cities. In fact, the development of tier I cities is led by the global economy more than the local economy (Nachum & Wymbs, 2005). According to Hymer (1972), MNEs place top management in tier I cities. These cities are large political and cultural centres (Beaverstock et al., 1999), but also financial centres because of presence of international capital (Friedmann, 1986).

WP1a: Strategic asset seeking internationalization focused on intangible FSAs will take place in a tier I city.

WP1b: Strategic asset seeking internationalization focused on tangible FSAs will not prefer one city type over another.

WP1c: Efficiency seeking internationalization that aims to decrease LoF will take place in a tier I city.

2.3.2.

T

II

C

Tier II cities are large cities that are not as concentrated as tier I cities (Hymer, 1972). These cities have a large population, their economy is significant but the material and immaterial flows are significantly smaller than tier I cities (Beaverstock et al., 1999). These cities are typically selected by MNEs to place their middle-management functions (Hymer, 1972). Because of their large population, they are attractive for MNEs that internationalize because of the ability to reach out to a larger customer base. While these cities are not global cities, they are still important cities within a country or a region, from where MNEs can serve an entire nation or region (Beaverstock et al., 1999). Because of the urbanization wave, a large number of tier II cities will be faced with a high growth that represents untapped market potential for MNEs (Dobbs, Remes, & Schaer, 2012).

Also, these cities are far more specialized than global cities and some of these cities are actual capitals of less-developed countries; their development is highly dependent on the country’s economy (Friedmann, 1986). Because of their specialized character, there are many suppliers and business partners for different industries that are agglomerated in tier II cities. In addition, these cities contain highly specific assets such as plants and facilities, which represent tangible LB FSAs that are attractive for MNEs. These industrial tier II cities can be specialized in one industry or they can contain multiple

industries. Therefore, it is expected that companies that activate in a sector will internationalize in cities where MNEs from the same industry are located because of presence of suppliers, business partners and existing infrastructure, that enable the company to decrease its transport and communication costs:

WP2a: Market seeking internationalization will take place in a tier II or tier I city.

WP2b: Efficiency seeking internationalization that is industry specific will take place in a tier II or tier III city.

2.3.3.

T

III

C

Tier III cities are small cities where MNEs place their lowest levels of operation, the daily operations such as production (Hymer, 1972). These cities are particularly attractive because of different endowments, such as cheap labor or presence of natural resources. The presence of infrastructure is a precondition for MNEs to internationalize there. These City-SAs are particularly attractive for MNEs that tend to spread their operations across the globe in order to spread production and to ‘diffuse industrialization to the less developed countries’ (Hymer, 1972). The diffusion of operations happens because of MNEs’ ability to coordinate large distribution networks by taking advantage of the advancement of telecommunications (Smith & Timberlake, 1995). In addition to cheap labor and resources, tier III cities can also have a strategic geographic position within a region or a country that allows the internationalizing MNE have better and cheaper access to other areas. Regardless of their regional orientation, MNEs looking to make their operations more efficient will target tier III cities, where labor is cheap and where investments, such as land, are cheaper than in tier II and I cities. While tier I cities are characterized by presence of intangible FSAs, tier III cities are expected to contain tangible FSAs, such as plants, factories and manufacturing facilities. Tier III cities are attractive especially for manufacturing MNEs that have the chance to reach economies of scale while having to deal with low transport costs (Boddy, 1999; Cantwell & Piscitello, 2002); service sector MNEs are expected to have an interest only in upper-tier cities.

WP3a: Resource seeking internationalization aiming to obtain tangible City-SA will likely take place in a tier III city.

WP3b: Efficiency seeking internationalization aiming to take advantage of cheap labor will take place in a tier III city

3.

M

ETHODOLOGY

The following chapter presents the main methodological concepts of this study. The section debuts with a presentation of the ontological and epistemological foundations that support the research methodology. Next, the basic principles of qualitative research are presented, followed by the presentation of the sample. Finally, the data collection process is briefly described and the chapter ends with explanations regarding the methods of data analysis.

3.1.

P

Ontology is a concept that represents the researcher’s philosophy about the reality and the world. There are two ontological views: subjective and objective (Brannick & Coghlan, 2007). Subjective ontology argues that the reality is in fact a product of the human mind; therefore, each individual has its own projection of reality. On the other hand, objective ontology argues that an independent researcher examines the reality as it is, without altering it (Brannick & Coghlan, 2007). Objectivism pictures the reality that is external to the observer (Saunders & Lewis, 2012). Regarding epistemology, post-positivism focuses on the creation of new knowledge and extension of established theories (A. Ryan, 2006). This research principle is based on the premise that theory and facts are two sides of the same coin and therefore research is done by considering both of them during the whole process. Assuming the existence of an objective world, post-positivist epistemology assumes researchers do not influence the study with their values and personal beliefs; moreover, the reality is mirrored by theory (Gephart, 2004). Objectivism is the most applied paradigm in management studies while the post-positivist epistemology reinforces the idea that the reality cannot be known in its totality, but only probabilistically (Gephart, 2004). Because this research philosophy is adopted according to the studied phenomena, an objective post-positivist approach is adopted for this research.

3.2.

Q

:

Qualitative research in IB literature has been overlooked by scholars, who still rely on quantitative research methods that are built on statistical analysis observer (Saunders & Lewis, 2012). Quantitative analysis serves for discovering causal relationships among variables, while qualitative research can provide rich descriptions of real actions in real life contexts that generate insights which cannot be provided by statistical analyses (Gephart, 2004). However, the combination of quantitative and qualitative data can provide an in-depth analysis, generating unique insights that are hard to observe otherwise (Eisenhardt, 1989). Unlike quantitative analysis, the explanatory nature of qualitative case studies allows for a better understanding of the realities and complexities that surround the object of study (Bryman, 2012). Because the nature of qualitative research has not been precisely defined, a

clear set of steps for conducting this type of research does not exist yet, reflecting the rich diversity of the qualitative studies; however, broad approaches are emerging around the design of credible qualitative case studies (Bryman, 2012).

This research will be conducted using a multiple case study research design similar to Kolk, Lindeque and van den Buuse (2013), framed around Eisenhardt’s (1989) and Yin’s (2009) approach to multiple case study research design. According to Yin (2009), a case study is a method of research in which real phenomena are studied within the real-life context that surrounds them in order to obtain an in-depth understanding of the selected cases. According to Eisenhardt (1989), case studies allow the researcher to discover patterns and to generate new theory. However, while this thesis does not adopt the purely inductive approach advocated by Eisenhardt (1989), numerous of the techniques and design elements she suggested are still adopted to strengthen the more deductive approach of Yin (2009). Yin (2009) and Eisenhardt (1989) both adopt an approach that Shepherd and Sutcliffe (2011) refer to as combining deductive approaches with bottom-up theorizing, providing sound grounds for combining their approaches.

Since the object of this study is to explain how cities can be used as a level of analysis within the regional strategies of MNEs, case studies are suitable because they facilitate the findings of patterns regarding the cities that are selected as location choices. The study will consist of three cases, each with two embedded units of analysis: the cases will consist of two Japanese MNEs, two German MNEs and two from USA. In the within-case analysis, both firms from each case are first studied individually and then patterns are identified by comparing them. In the cross-case analysis, the patterns emerge by contrasting and comparing the internationalizations of all six sampled MNEs.

3.2.1.

Q

Case study research has been criticized because of its lack of generalizability, rigor, and lack of methodic-ness (Yin, 2010). However, the quality of a case study research can be increased if quality criteria are respected. Construct validity, external validity, internal validity and reliability are principles that certify the quality of a study (Yin, 2009).

Construct validity refers to the accurate measurement of the objects of study. This is achieved through the use of multiple sources and through the use of proven measures (Yin, 2009). This study is characterized by data triangulation, since there are multiple sources of data: annual reports of the studied MNEs, newspaper articles, M&A databases for qualitative data and company annual accounts for dispersion of sales for quantitative data. Moreover, the establishment of the regional orientation of companies has been done according to highly cited previous studies (Rugman & Verbeke, 2004; Rugman & Collinson, 2004).

Internal validity is a construct that refers to the causality of the identified relationships (Yin, 2009). In this study, internal validity is secured through detailed explanations of the studied relationships, including rival explanations; all within-case analyses are conducted following the theoretical framework and the results are analyzed in the cross-case analysis (Yin, 2009). In addition, methodological triangulation, the combination of quantitative and qualitative analysis, is used because it enables the researcher to gain multiple perspectives regarding the object of study.

External validity refers to the generalizability of findings (Yin, 2009). This is achieved through the use of multiple cases and through the use of deductive and inductive principles (Eisenhardt, 1989). Even if this study starts off as a deductive study, from theory-generated hypotheses, it is in fact a combination of inductive and deductive elements, going back and forth from theory to data (Shepherd & Sutcliffe, 2011). Unlike statistical generalization, where the findings from the analysis of the sample can be generalized to the whole population, theoretical generalizability is achieved by comparing the empirical results of the case study with the theoretical framework (Yin, 2009). Through replication logic, the theory is tested by comparing the results of each case (Yin, 2009). In addition, the external validity is also strengthened by the use of multiple cases with two embedded units of analysis, which allows for a better generalizability (Eisenhardt, 1989).

Reliability is achieved when a study is rigorously conducted so that later investigators can repeat the same steps (Yin, 2009). In case of this research, reliability is achieved by using multiple sources of data and by following a rigorous data collection process in which all cases are stored in a case study database and all steps are recorded and described. Moreover, all the steps of this research are mentioned in the data collection and data analysis sections of this chapter.

3.2.2.

C

S

–

I

While Rugman and Verbeke (2004) demonstrate that most Fortune Global 500 firms’ sales take place in their home region and that globalization is a myth, their work has been criticized by several authors. In their analysis of regionalization, Osegowitsch and Sammartino (2008) posit that Rugman’s benchmarks regarding the home and host region sales percentage are too rigid. Another contrasting view is brought in the IB field by Dunning, Fujita and Yakova (2006), who state that the degree of concentration of FDI and sales in a specific area give the regional orientation of an MNE rather than the actual distribution of sales. But even if some methodological aspects of the paper have been criticized, it has been recognized as a highly relevant study in the IB literature, as it has brought new insights about international strategies of MNEs. Therefore, the replication of this study in order to set the regional orientation of the selected firms is the most appropriate.

Table 2. The sample. (Source: Author)

1 _{– More than 50% of sales are recorded in home region.} 2 _{- Between 20% and 50% of sales in all three regions of the triad. For the other two orientations, host region} oriented firms record more than 50% of sales in a region other than the home region. Bi-regional MNEs record between 20% and 50% of sales in two of the three regions of the triad

.

Region Orientation Description

Case 1 Japan

Mitsubishi Motors Corporation

118 Japan Asia-Pacific

Home Region Oriented1

Mitsubishi Motors Corporation designs, develops, manufactures, assembles, markets, sells, and imports automobiles and automobile parts. It primarily offers passenger cars and commercial vehicles; and

automobile engines, transmissions.

Toyota Motor Corporation

8 Japan Asia-Pacific

Home region oriented1

Toyota, the largest car manufacturer in the world, engages in the design, manufacture, assembly, and sale of passenger cars, minivans, commercial vehicles, and related parts and accessories primarily in Japan, North America, Europe, and Asia. It also owns the brands Lexus

and Scion.

Case 2 USA

General Motors Company

22 USA North-America

Home Region Oriented1

General Motors Company (GM) designs, manufactures, and markets cars, crossovers, trucks, and automobile parts worldwide. The company

markets its vehicles primarily under the Buick, Cadillac, Chevrolet, GMC, Opel, Holden, and Vauxhall brand names, as well as under the

Alpheon, Jiefang, Baojun, and Wuling brand names.

Ford Motor Company

28 USA North-America

Home Region Oriented1

Ford Motor Company develops, manufactures, distributes, and services vehicles, parts, and accessories worldwide. It also owns the Lincoln Brand. It markets cars, utilities, trucks, service parts, and accessories

through distributors and dealers in North America, South America, Europe, Turkey, Russia, and the Asia Pacific region.

Case 3 Ger-many

Daimler A.G.

23 Germany Europe Global2

Daimler AG, through its subsidiaries, develops, produces, distributes, and sells passenger cars and off-road vehicles, trucks, vans, buses, and

related spare parts and accessories worldwide. The company operates through Mercedes-Benz Cars, Daimler Trucks, Mercedes-Benz Vans,

Daimler Buses, and Daimler Financial Services segments.

Volkswagen A.G.

9 Germany Europe

Home Region Oriented1

Volkswagen AG, together with its subsidiaries, manufactures and sells automobiles primarily in Europe, North America, South America, and

the Asia-Pacific.Volkswagen AG provides its products under the Volkswagen Passenger Cars, Audi, ŠKODA, SEAT, Bentley, Bugatti,

Lamborghini, Porsche, Ducati, Volkswagen Commercial Vehicles, Scania, and MAN.

23

The automotive industry represents the focal industry that this study is based upon. In a previous study about the automotive industry, Rugman and Collison (2004) find out that out of 29 companies in this sector, none of them are global. Instead, 23 are home region based, two are bi-regional and two are host region based (Rugman & Collinson, 2004). The automotive industry cannot become a global industry at the moment because companies face numerous barriers in their pursuit of a global strategy. First of all, this industry operates in large clusters: Original Equipment Manufacturers (OEMs) assemble vehicles using a large network of suppliers, distributers and other manufacturers that are all spread around the world and bound to a specific location (Rugman, 2005). Another barrier encountered by these companies is a cultural barrier: in Europe, Diesel cars are very popular while in North America, Gasoline is still prevalent. Also, in Europe, consumers are attracted by performance and efficiency, while the North-American market is still focused on large cars, with a tendency for SUVs. Lack of uniformity of car demands across the globe forces automotive manufacturers to adjust their product portfolios according to regions (Rugman, 2005). Consequently, the industry has not managed to produce a global car that is suitable for all markets. This study addresses the way in which cities affect the regional strategies of automotive MNEs.

Following Rugman and Verbeke (2004) and Rugman and Collinson (2004), six automotive companies are sampled: Toyota, Mitsubishi, Volkswagen, Daimler, Ford and General Motors. The only difference between Rugman and Collinson’s (2004) sample and this research’s sample is the exclusion of Honda and the inclusion of Mitsubishi. Moreover, the six sampled companies come from three countries that are leaders in the automotive industry in their region: Toyota and Mitsubishi from Japan, Daimler and Volkswagen from Germany and Ford and GM represent USA.

3.4.

D

C

The process of collecting data has a number of phases: first, data is collected to establish the regional orientation of sampled MNEs; second, a list of internationalization deals is established; third, newspaper articles about these deals are gathered; the last step is the collection of data regarding each city in which an MNE internationalized. All three streams of data are interdependent and offer a holistic view regarding the internationalizations of MNEs and the city level of location choice: while the quantitative data is used to determine the regional orientation, which affects the regional strategies of MNEs, the newspaper articles provide the qualitative data that is necessary to analyze the location choices, which have been identified using the Zephyr database and official company history.

The data necessary for this study is documentary data and it comes from multiple sources. The regional orientation of the selected companies is established through the replication of Rugman and Verbeke’s (2004) study to find out the regional orientations of the sampled MNEs. First, longitudinal

24

data over the last 5 years (2008-2012) regarding the spread of assets, revenue and employees across regions is collected from each company’s annual reports. Then, the data is centralized in a database where the calculations of Rugman and Verbeke (2004) are applied, including their initial benchmarks. According to their set of criteria, the regional orientations of the sampled firms are established. However, lack of data regarding the proportions of assets and employees across regions meant that the primary variable in establishing the regional profile was the sales value.

The next step is the collection of data regarding the sampled firms’ internationalizations. The Zephyr database is used to find all completed mergers and acquisitions (M&As) that the company conducted in the sampled period. Because Zephyr does not have greenfield investments, official company websites will be used to identify the greenfield investments. Using these sources, a list of all internationalizations in foreign countries of these companies between 1993-2013 is established. The country, the city, the date, the status of the deal and the entry mode are recorded in a database. Therefore, out of this database, a list of cities will emerge. However, this list does not contain all the internationalizations, since the Zephyr database and the company websites are incomplete: in some cases, Zephyr does not contain the name of the city, while the company websites do not have a clear list of all internationalizations and their respective cities.

The next phase is to download newspaper articles from Financial Times London for each FDI: using the company name as a keyword, all articles published from 1993-2013 will be downloaded and will serve as the basis of analysis. There are however some special cases. In case of General Motors, the terms “General Motors Company” and “General Motors Corporation” are both used, since the latter is the official name of the company before its bankruptcy in 2009 and the former is the current name. Also, in case of Daimler AG, the keyword “Mercedes-Benz” is also used to find articles since the “Daimler AG” term is not indexed in all related articles in LexisNexis database.

Company Zephyr investments Company website investments Studied investments

Total Studied Total Studied

Mitsubishi 12 4 4 3 7 Toyota 58 5 17 10 15 General Motors 24 4 7 4 8 Ford 36 9 13 9 18 Daimler 65 5 7 7 13 Volkswagen 38 8 21 9 17

Table 3. The sample and the internationalizations. (Source: Author)

3.5.

D

A

The aim of the analysis of the qualitative data is to understand the working propositions, which is achieved using thematic coding. Thematic coding starts with the establishment of themes, ranking the themes according to each theme’s importance and building a hierarchy; then, the themes are linked to the theoretical framework (G. W. Ryan & Bernard, 2003). Analyzing data through thematic coding allows for the identification of patterns because coding’s main function is categorization and identification of patterns. According to the inductive and deductive approaches that are used in this study, themes arise from the theoretical framework as well as during the data collection process. Thematic coding through the use of CAQDAS (Computer Assisted Qualitative Data Software) enables the creation of categories of themes as codes. The themes are identified by noticing repetition and through the use of theory-related material (G. W. Ryan & Bernard, 2003).

In order to understand how cities impact the location choices of MNEs, newspaper articles from 1993-2013 from specialized newspapers such as Financial Times London were collected. For each analyzed internationalization, at least one relevant newspaper article was found. Along with official company websites, these materials are analyzed using NVivo. Through thematic coding, themes are divided into categories which are then associated to codes; the software is used to associate the data with the codes.

In order to study the city level of analysis, each internationalization is associated with a city. However, due to methodological reasons, all cities must be divided into three categories. However, the city literature does not have a generally accepted methodology of categorizing cities and positioning them into a hierarchy yet (Beaverstock et al., 2000). Some authors classify the cities by calculating the number of MNE subsidiaries (Beaverstock et al., 2000; Beaverstock et al., 2000), different types of flows such as air traffic or material flows, or even the number of times a specific city was mention in a publication (Beaverstock et al., 2000). Therefore, there will be three types of cities, according to Hymer’s (1972) original classification: Tier I cities, Tier II cities, and Tier III cities. Each city in which one of the sampled MNEs internationalizes will be part of a tier. The categorization will take

Case Company Total FT Articles

Found Used FT articles

Case 1 Japan Mitsubishi 8956 13 Toyota 8907 32 Case 2 USA General Motors 10170 14 Ford 7862 30 Case 3 Germany Daimler 5792 17 Volkswagen 8765 31

Table 4. The sample and the collection of newspaper articles. (Source: Author)

into account the population, the degree of connectivity, number of top universities, average wage, presence of airports and seaports and whether it has the status of capital of the country. This procedure will be done especially for tier II and III cities; tier I cities have been studied before and their belongingness to this category is unanimously accepted by a number of authors (Beaverstock et al., 2000; Beaverstock et al., 2000; Friedmann, 1986; Goerzen et al., 2013).

The categorization of the cities takes into account the variables that have been previously collected. First of all, the population is a good all-around indicator: it signals the presence of potential workforce and customer base. The change in population over time offers an idea about the direction where the city is going: whether it is developing or not. A tier I city will have a large population, indicating the presence of an large existing customer base but also signaling the existence of skilled labor. On the other hand, tier II and tier III cities will have a smaller population and usually the value of the population change is negative, due to migration to tier I cities. Airports and seaports are key strategic elements that can be typically found in tier II and tier I cities. Also, a capital city signals the presence of institutions and other MNE headquarters, making it a potential tier I city, but also a tier II city if the country is less developed. The GDP per capita is an indicator of the standard of living in a city: the bigger the value, the richer the population is, making it a potential market for high-end products and it is another indicator for the wage and income of the inhabitants. The last three indicators are representing the degree of world connectivity of the cities (Taylor, 2001): a city with a high degree of connectivity is usually a tier I city, a global city, while tier III cities have a small degree of

Code Sub code WP Description

Strategic Asset Seeking

Intangible FSA – tier

I WP1a

All internationalizations that target intangible FSAs take place in a tier I city

Tangible assets – no

tier WP1b

All internationalizations that target tangible assets regardless of the type of city

Market Seeking Market Seeking –

Tier II WP2a

All internationalizations that seek a customer base and take place in a tier I or II city

Efficiency Seeking

Cheap Labor – Tier

III WP3b

All internationalizations that seek efficiency by aiming to take advantage of cheap labor and take place in tier III cities

Industry specific –

Tier II and III WP2b

All internationalizations that are industry specific, seek efficiency and take place in tier II and III cities LoF – Tier I WP1c All internationalizations that aim to reduce LoF Resource

Seeking

Tangible CitySAs –

Tier III WP3a

All internationalizations that aim to obtain tangible City SAs and take place in a tier III city.

Table 5. The codebook (Source: Author)