Cities and the Location Behavior of MNEs:
How
location choice at city level is influenced by FDI motives and
regionalization profiles
Master Thesis
MSc. Business Administration – International Management Supervisor: Dr. Johan Lindeque
Second reader: Dr. Alan Muller
Student: Nicky Caron
Student 1D: 10558128
Statement of originality
This document is written by Student Nicky Caron who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
Abstract
This study presents new insights on MNE location behavior at the city level of analysis and shows that both FDI motives and the regionalization profile of a company influence its location strategy, where the chosen city should present a good fit between the FDI motive and the city’s characteristics. This thesis finds that market seeking FDI is often directed at larger, cosmopolitan cities that have an abundance of highly skilled labor whereas resource seeking FDI is often linked to smaller cities or towns that possess raw materials and provide access to lower skilled manpower. Furthermore, there is a tendency for host regions investments to be located in Tier-1 cities, while home region investments take place relatively more often in Tier-3 cities.
Keywords: multinational enterprises, location choice, cities, investment motives, regionalization
Acknowledgements
First and foremost, I want to thank my supervisor Dr. Johan Lindeque for his guidance and supportive feedback, which helped me to push the development of my thesis further every time. I am also grateful for the fact that he could make time to provide me with feedback at the busiest times and even during his holiday. Furthermore, I want to thank my friends and family that supported me while writing my thesis.
Table of Contents
1. Introduction ... 7 2. Conceptual foundation ... 10 3. Methodology ... 23 4. Within-case analysis ... 35 5. Cross-case analysis ... 57 6. Conclusion ... 62 7. References ... 65 8. Appendix ... 72 "Index of Tables and Figures
Table 1. Overview case MNEs………...28
Table 2. Detailed description of cases and MNEs………..29
Table 3. Data sources of case MNE investments……..………..31
Table 4. Overview variables of city data and sources………32
Table 5. Nvivo codes………...………...33
Table 6. ADM: Entries per city category divided between home and host regions………...36
Table 7. ADM: Overview of investment decisions………37
Table 8. Bunge: Entries per city category divided between home and host regions………...38
Table 9. Bunge: Overview of investment decisions………...39
Table 10. Case description of food suppliers………..42
Table 11. Nestlé: Entries per city category divided between home and host regions………....43
Table 12. Nestlé: Overview of investment decisions……….44
Table 13. Kraft: Entries per city category divided between home and host regions………..46
Table 14. Kraft: Overview of investment decisions………...47
Table 15. Case description of food producers………50
Table 16. Carrefour: Entries per city category divided between home and host regions………...52
Table 17. Carrefour: Overview of investment decisions………53
Table 18. Tesco: Entries per city category divided between home and host regions………...54
Table 19. Tesco: Overview of investment decisions………..55
Table 20. Case description of food retailers………...57
Table 21. Overview intra- and inter-regional investments per company………...58
Table 22. Overview of investments divided between home and host regions………59
Table 23. Overview of first entries per company………...………60
Table 24. Overview propositions and results………...………...62
1. Introduction
Multinational enterprises (MNEs) locate their subsidiaries in multiple countries around the world for different investment reasons (Dunning, 1998). However, the world is not as globalized as some may assume (Ghemawat, 2003) and many MNEs do not pursue a truly global strategy, but tend to operate mainly in their home country or home region (Rugman & Verbeke, 2004). This has led researchers to emphasize processes of regionalization and it has been found that most firms indeed operate on a regional scale and often have the larger part of their sales in the home region (Rugman & Hodgetts, 2001; Rugman & Girod, 2003; Rugman & Verbeke, 2004, 2008). The regionalization phenomenon could arise from a liability of foreignness, which arises due to the company’s unfamiliarity with the foreign culture and way of doing business and is associated with costs that a local competitor does not incur (Zaheer, 1995; Hymer, 1972). Liability of foreignness is perceived as higher for firms investing in a host region compared to investments in their home region (Rugman & Verbeke, 2007). This study argues that liability of foreignness does not only influence MNE location behavior at a regional level, but also at the national, subnational and city level.
The regional and national characteristics of a firm’s location are important, but local characteristics have to be taken into account as well (Beugelsdijk & Mudambi, 2013). This thesis introduces cities as an important location choice determinant and argues that foreign direct investment (FDI) motives can be connected to a company’s preference for a certain type of city (Goerzen et al., 2013; Hymer, 1972), which is also influenced by the level of liability of foreignness the internationalizing company will incur. In the field of IB very little research has been done concerning the subnational location behavior of MNEs (Beugelsdijk & Mudambi, 2013). This thesis develops a conceptual framework to address this gap by integrating themes from the IB perspective, such as regionalization and liability of foreignness, and combining it with research on location choice from an economic geography
perspective. The conceptual framework will be the foundation for several working propositions, which form the basis to ultimately provide insights on the following research question:
How well do FDI motives and regionalization profiles explain city location choices?
This question will be answered by studying the location choices and entry decisions of six MNEs that operate in the food industry through a multiple-case study. These six MNEs are part of three case sectors: food suppliers, food producers and food retailers. Each case sector has two embedded units of analysis: Archer Daniels Midland (food supplier), Bunge (food supplier), Kraft (food producer), Nestlé (food producer), Carrefour (food retailer) and Tesco (food retailer). This selection of companies allows for a within-case analysis and cross-case analysis, improving the believability, generalizability and credibility of the findings.
This study finds that location behavior of MNEs at city level is indeed influenced by FDI motives and regionalization profiles. Firms internationalizing for market seeking reasons mainly invest in Tier-1 cities, cities that are cosmopolitan, have strong ties to the rest of the world and offer an abundance of highly skilled labor. Resource seeking investments are often linked to investments in Tier-2 cities, that have a large labor pool (middle to lower skilled) and have a strong infrastructure for information and communication systems, and Tier-3 cities, where raw materials are available and manpower is cheap. With regards to regionalization profiles, the findings show that home-region oriented companies are more prone to invest in their home region, while firms with a bi-regional or global strategy have most of their investments in one of their host regions. Furthermore, in the home region Tier-3 cities are relatively more favorable choice and in the host region Tier-1 cities are more
favored. When a firm enters a new country or region for the first time, the investment is likely to be in a Tier-1 city.
The thesis proceeds by reviewing the literature on international business strategy, regionalization and liability of foreignness, location choice, city categorization and FDI motives. This is followed by an explanation of the methodology and case sample. The findings of the analysis are discussed in the within-case analysis and cross-case analysis sections. The thesis ends with a concluding section that provides a summary of the results, elaborates on managerial implications of the study, its scientific relevance, limitations of the study and suggestions for future research.
2. Conceptual foundation
This chapter will focus on establishing the link between the regionalization of MNEs, location choice and investment motives to provide a theoretical foundation for cities as a location choice within the MNE regionalization strategies. The working propositions for the thesis are developed throughout the conceptual foundation. The chapter starts with an introduction to IB strategy on the internationalization motives of firms and continues with a discussion of the regionalization literature and the concept of liability of foreignness. Next, the location choice literature is reviewed by taking into account regional, national and local levels of analysis. This is followed by a discussion of the city as subnational location choice and this is linked to different investment motives of the MNE.
2.1 International business strategy and the regionalization of MNEs
Companies go abroad to obtain country specific advantages (CSAs) to combine with their firm specific advantages (FSAs) (Rugman & Verbeke, 1992; Rugman et al., 2011). CSAs are advantages specific to the host country where the MNE is present and FSAs are advantages that derive from the firm itself. FSAs can be location-bound (LB), which means that they can only be deployed within a limited geographic area, or they can be non-location-bound (NLB), which indicates that they can be easily transferred across countries or regions (Rugman & Verbeke, 1992). When an MNE internationalizes, non-location-bound FSAs are combined with host country CSAs to create a competitive advantage. However, distance or dissimilarities between home and host locations influence the recombination of non-location bound firm-specific advantages with host-country-specific advantages (Rugman et al., 2011), which is needed to create such a competitive advantage (Rugman et al., 2011). Many MNEs are mainly located in their home country and have the largest part of their sales in the MNEs
extended Triad regions, Asia Pacific (AP), the European Union (EU) or North America (NA), and compete within their home region or in their home region and only one other region of the Triad (Rugman & Verbeke, 2004). There are only a handful of truly global MNEs that have equally penetrated all three regions of the Triad in terms of sales (Rugman & Verbeke, 2004) and a global strategy is not suitable for most MNEs (Rugman & Hodgetts, 2001). Supply chains, which were always thought to be global, are regional in nature and a regional focus in terms of sales can contribute to improved performance (Rugman, Li & Oh, 2009). Regionalization can for a large part be attributed to the different degrees of liability of foreignness that MNEs experience within and between the extended Triad regions (Rugman & Verbeke, 2004).
2.2 Liability of foreignness
The difficulty of transferring, deploying, recombining and profitably exploiting FSAs abroad can be referred to as the liability of foreignness (LoF) of an internationalizing firm (Sethi & Judge, 2009; Zaheer, 1995). LoF has been broadly defined as all additional costs of operating in a foreign country that a domestic company would not incur (Zaheer, 1995) and is likely to influence ex ante location decisions and ex post subsidiary performance (Goerzen et al., 2013). The costs of doing business abroad originate from at least three sources: complexity, uncertainty and discrimination (Zaheer, 1995, Goerzen et al., 2013). Complexity originates from the difficulty of doing business at large distances with many coordination difficulties. Uncertainty comes from unfamiliarity with local culture and institutions. Discrimination costs originate from economic nationalism and the lack of legitimacy of foreign firms (Zaheer, 1995; Goerzen et al., 2013). Most MNEs are not able to exploit their FSAs across the world equally well and prefer to trade within their own region (intra-regional trade), because location-bound FSAs are more easily augmented to suit exploitation in home region countries,
which is due to higher levels of LoF for investments outside of the MNEs home region (Rugman & Verbeke, 2004, 2007). However, whether a MNE is able to exploit FSAs across regions depends on their regionalization profile (Rugman & Verbeke, 2004). Home-region oriented MNEs have at least 50 per cent of sales in their home region; bi-regional firms are defined as having at least 20 per cent of their sales in two regions, but less than 50 per cent in any one region; host-region oriented MNEs have more than 50 per cent of their sales in one of their host regions; and global firms have more than 20 per cent of their sales in each region of the Triad and not more than 50 per cent in one region (Rugman & Verbeke, 2004). Home-region oriented firms will experience higher LoF in host Home-regions, because the strategies of these firms are not adjusted to global penetration and do not have as much experience as bi-regional or global firms operating in other regions (Johanson & Vahlne, 1977; Rugman & Verbeke, 2004). Global or bi-regional firms pursue a strategy that is able to recombine FSAs and host-country CSAs across the world and have more experience doing business abroad. Therefore, the following propositions are formed:
Proposition 1a: MNEs with a home region orientation are likely to invest more often in their home region.
Proposition 1b: MNEs with a bi-regional or global strategy are likely to invest more often in host regions.
Internationalizing firms should take LoF into account before making any location or investment decisions, because levels of LoF are influenced by the region (Rugman & Verbeke, 2004), country (Zaheer, 1995) and subnational characteristics of a location (Goerzen et al., 2013). The next paragraph will take a closer look at location choice at three different
levels of analysis (country, subnational and city) and will integrate the concept of LoF with location choice at city level.
2.3 Location choice and country characteristics
Multiple international investments are the distinctive aspect of international business strategy (Ricart et al., 2004) and a location can be a source of competitive advantage for an MNE (Dunning, 1998). Makino et al. (2004) investigate location choice and the different factors that influence subsidiary performance. Their findings indicate that country characteristics have a significant impact on business performance. Nachum et al. (2008) confirm that country characteristics are important by showing that a country’s proximity to the world’s knowledge and markets are important drivers of location choice. However, while country characteristics are important, the specific characteristics of local environments also influence location choice (Piscitello, 2011). Subsidiary-specific advantages can be a source of competitive advantage and arise from local resources and knowledge spillovers from the local context, where the interaction between the firm and its location is very important (Piscitello, 2011). Local learning and mutual interaction with local actors can enhance competitive advantage (Piscitello, 2011). This can be linked to the FSA/CSA matrix (Rugman et al., 2011) where firm specific advantages are complemented by country specific advantages.
Mariotti et al. (2010) look at the spatial agglomeration of MNEs and show that geographical closeness to a location is necessary to reap benefits of local learning but is not enough to generate interaction. Furthermore, interaction does not necessarily lead to positive spillovers. Information externalities and observational learning influence the location choice of MNEs and leads MNEs to copy other MNEs that have already dealt with a particular problem in this location. There is also the potential of knowledge spillovers, which can be a net gain or a net loss for a particular firm. Which location is chosen and how spillovers work
out also depends on the subsidiary mandate or investment motive (Marrioti et al., 2010). Relatedly, Beugelsdijk et al. (2010) also urge companies to look at the effects of local characteristics and argue that too often a location is treated as synonymous with the country level of analysis without providing any detail on local features.
2.4 Within-country variation
Local contexts differ in their resource endowments as well as their institutional frameworks and are important because they have preserved distinctive differences despite the increased and more intense interactions attributed to globalization (Meyer et al., 2011). Resource endowments can be natural or created and form the foundation for the attraction of foreign MNEs and while one location’s resource endowments might fit well with advanced business activities, such as R&D, other resource endowments are better suited for standardized business activities, such as manufacturing (Meyer et al., 2011).
Subnational heterogeneity drives location decisions because firms locate in a particular agglomeration, region or city and not in a random location within a country (Beugelsdijk & Mudambi, 2013). When investigating motivations behind location choices the researcher should consider subnational distance effects and international border effects, because companies do not work with country averages but are affected by variation within countries (Beugelsdijk & Mudambi, 2013). Subnational location information is crucial and this should be integrated into international business research. Relatedly, Jensen and Pedersen (2011) show that the nature of the chosen location differs when a firm performs either more standardized or more advanced activities. Low cost manufacturing is moved to low cost destinations, while R&D activities are relocated to places with a more knowledge intensive industry. Many national subsidiaries focus on a very narrow and specific set of activities of the MNEs value chain (Jensen & Pedersen, 2011). These activities have to align with the local
resource endowment, knowledge base and other local features. The offshoring of manufacturing is primarily related to low-cost destinations and therefore follows the classic efficiency seeking strategy (Jensen & Pedersen, 2011). The link between investment motives and location choice at subnational level is further explained below through a focus on different city categories as a specific type of subnational location.
2.5 Cities as location choice
At the subnational level, cities are an important factor in location choice. City-regions are increasingly seen as the main driver of a country’s economy (McCann and Acs, 2011) and the business environment in cities directly affects national macroeconomic growth, because of the creation of output and the employment of people (Sridhar & Wan, 2010). The level of knowledge, creativity, innovation and connectivity are important drivers of a city’s productivity and cities as the center of knowledge can give rise to competitive advantage (McCann & Acs, 2011). Cities are fundamental territorial nodes of the modern global economy and their systematic and frequent exchanges among them help to shape the world system’s structure (Smith & Timberlake, 1995).
2.5.1 Cities and subsidiary activities
Cities can be linked to different types of subsidiary activities (Hymer, 1972), in the same way Jensen and Pedersen (2011) relate certain location choices to offshoring activities. This thesis integrates Hymer’s (1972) research on the link between subsidiary activities and cities with other streams of literature on cities to create a three-tier city framework to categorize cities.
Tier-3 cities are concerned with day-to-day operations and are spread around the world drawn by manpower, markets and raw materials (Hymer, 1972) and are often of a smaller size, because labor intensive firms do not locate in large or mid-size cities which can be
attributed to higher wage, training and attrition costs that is occurred in these cities (Sridhar & Wan, 2010). A stock market or top university is not found in a Tier-3 city.
Tier-2 cities feature subsidiary activities that include coordinating managers at the lower level and tend to be characterized by the need for white-collar workers and information and communication systems (Hymer, 1972). Regarding the size of the city, Tier-2 cities lie between Tier-1 and Tier-3, which means that Tier-2 cities could be considerably large, but lack the headquarter offices, cosmopolitan culture and advanced producer services that characterize Tier-1 cities. City size is only one of the factors taken into account and variables such as the type of labor available (high, medium or low skilled) is a more important determinant. Tier-2 cities might have a top university present, but will not feature a stock-market. One of the defining differences between Tier-2 and Tier-3 is that Tier-2 cities can feature a transportation-hub such as an airport or sea-port, which increases the importance of the city. Furthermore, while Tier-2 cities can be the capital of a country, province or state, Tier-3 cities will usually not have such an important role assigned to them.
Tier-1 cities are associated with goal determination and planning activities (Hymer, 1972), are typically somewhat larger, are close to the country’s capital market, media and government (Hymer, 1972) and have global-city features that include cosmopolitanism, global interconnectedness and an abundance of advanced producer services (Goerzen et al., 2013). These last three characteristics are more elaborately explained in the next paragraph. High connectivity to the rest of the world is necessary because firms that invest heavily in knowledge assets, must capture markets beyond country borders to generate the required returns on their knowledge investments (McCann & Acs, 2011) and while these returns were traditionally generated by exports, there is now an increasing premium associated with face-to-face contact time (McCann, 2007, 2008). Tier-1 cities are often capital cities of their
nation, province or state, feature a stock exchange, have an airport nearby and include one or more top universities.
It is expected that cities influence the location behavior of MNEs, because subsidiary activities have to form a good fit with city characteristics. However, Tier-1 or global cities have characteristics that help MNEs to do business abroad on another level as well by helping them to overcome liability of foreignness.
2.5.2. Global cities
Goerzen et al. (2013) find that MNEs have a strong propensity to locate in global cities, because specific features of global cities help MNEs to lower the costs that arise from doing business abroad (Goerzen et al., 2013). Global cities are defined by three characteristics: global interconnectedness, cosmopolitanism and abundance of advanced producer services. Global interconnectedness refers to the external global linkages that are constructed in cities and are defined by inflows and outflows of resources. Criteria that determine the connectedness of a city are the existence of financial centers, headquarters for international institutions, growth of the business services sector and transportation nodes (Goerzen et al., 2013). The criteria for interconnectedness are for a large part comparable to Level 1 activities of Hymer’s (1972) categorization and also describe the type of city that Hymer would link to Level 1 activities. A cosmopolitan environment refers to a high level of culture, politics, education and other social factors. Abundance of advanced producer services is the third characteristic of a global city and is important because MNEs need high quality services for their global operations.
The three the key features of Tier-1 or global cities help MNEs to reduce the costs of doing business abroad by lowering liability of foreignness. As stated earlier, these costs arise from complexity, uncertainty and discrimination when doing business abroad (Zaheer, 1995, Goerzen et al., 2013). Global cities can help MNEs to lower these costs. For example, global interconnectedness reduces complexity, because it makes the transfer of people, goods, capital and information easier. The cultural diversity of global cities makes complexity easier to manage because it facilitates the use of expatriates (Goerzen et al., 2013). The presence of global media and an elaborate IT infrastructure in global cities reduce uncertainty, because they make it easier for firms to find information and data about a country’s culture and institutions. Furthermore, a high level of cosmopolitanism in global cities reduces discrimination because its citizens are more open-minded towards other countries and cultures, have a preference for international consumption and an appreciation of diversity (Goerzen et al., 2013).
Liability of foreignness is particularly acute for market seeking horizontal MNEs, which is a multinational whose subunits are essentially replicas of each other that manufacture or distribute goods and services in different markets across the world (Caves, 1982) and whose operations are more on a local-for-local basis (Bartlett & Ghoshal, 1989). Resource seeking companies that can exploit economies of scale and scope may feel liability of foreignness less (Ghoshal & Nohria, 1989). Combining this with the finding that intra-regional liability of foreignness is lower than inter-intra-regional LoF, the following proposition is formed:
Proposition 2: Investments of a market seeking nature that are located outside of the MNEs home region are more likely to be located in a Tier-1 city.
When firms internationalize, they gradually gain more knowledge and experience and build relationships, which helps them to do business in the foreign country and ultimately overcome liability of foreignness (Johanson & Vahlne, 1977, 2009). Therefore, a company that enters a new country or region for the first time is likely to experience higher LoF than competitors who have been present for a longer time, because they have not had time to gain knowledge about the country’s customs and culture. Furthermore, MNEs experience higher liability of foreignness when internationalizing outside of their own region compared to intra-regional FDI (Rugman & Verbeke, 2004, 2007). Combining these two streams of research, the following proposition is formed:
Proposition 3: When a MNE enters a host region or country outside its home region, it is more likely to locate in a Tier-1 city.
2.5.4. Location choice and investment motives
Cities as the decisive factor in location choice, such as a company’s decision to locate in a global city or a small town, are not only influenced by the level of liability of foreignness that the company will incur. This thesis proposes that sub-national locational preferences also depend on the motives for the investment. Four main types of investment motives are identified (Dunning, 1998): resource seeking, market seeking, efficiency seeking and strategic asset seeking. When a company invests for resource seeking motives, it is in need of natural resources that can only be found in the host country. The availability, price and quality of these resources are important (Dunning, 1998). The location decision is also influenced by the presence of local opportunities for upgrading quality of the resources and the presence of local partners to promote knowledge or capital-intensive resource exploitation (Dunning, 1998). This study proposes that companies investing for resource seeking reasons will not be
likely to locate in a global, highly populated city, but look towards Tier-3 cities where an abundance of natural resources is available of where it is cheaper to manufacture products or Tier-2 cities that are somewhat larger, but are often located at a transportation hub, such as a large sea port and also have medium to low skilled labor available. Therefore, the following proposition is formed:
Proposition 4: MNEs with a resource seeking investment motive will locate in 2 or Tier-3 cities.
A market seeking motive indicates that a firm wants to expand its sales to a new foreign market and feels the need to be physically present in the new market (Dunning 1998). For example, a firm wants to start its own subsidiary to keep knowledge and other intangible assets in-house instead of risking that it might leak to export agencies or distributors of their product. The size of the market, the availability of skilled labor, the presence and competitiveness of competitors, the quality of the infrastructure, the need to be close to the end-user and the growth in importance of promotional activities are all important factors for market seeking firms (Dunning, 1998). Therefore this study proposes that market seeking firms will locate in Tier-1 cities with a large population that consists of highly skilled and sophisticated buyers or Tier-2 cities that also feature a large population and are signified by many white-collar workers. Therefore, the following proposition is formed:
Proposition 5: MNEs with a market seeking investment motive choose to locate in a Tier-1 or Tier-2 city.
The efficiency seeking motive is related to the possibility of lowering production costs or the presence of investment incentives, such as tax breaks. Efficiency seeking firms are attracted by the availability of cheap labor, the presence of competitors and the quality of the infrastructure (Dunning 1998). Efficiency seeking firms are mainly expected to locate in cities that are small to medium-sized, such as Tier-2 cities where white-collar workers are available and where lower level management is located, and Tier-3 cities, where raw materials are available and a large amount of cheap low-skilled labor is available. Therefore, the following proposition is formed:
Proposition 6: MNEs with an efficiency seeking investment motive choose to locate in a Tier-2 or Tier-3 city.
Strategic asset seeking involves firms that look for knowledge related assets or opportunities to exchange localized, tacit knowledge, ideas and interactive learning (Dunning 1998). Strategic asset seeking is influenced by access to different cultures and institutions, and different consumer demands and preferences. This investment motive is mainly linked to larger cities with highly skilled labor, sophisticated buyers or specialized clusters, such as Tier-1 cities. Therefore, the following proposition is formed:
Proposition 7: MNEs with a strategic asset seeking investment motive choose to locate in a Tier-1 city.
2.6 Conclusion
This study expects that MNE location behavior is influenced by multiple factors. The first is the firm’s regionalization profile, where a home region oriented company will mainly invest
in its own region, while a company with a bi-regional or global strategy will invest mainly in host regions. Market-seeking FDI and first-entries outside the host region are likely to be located in a Tier-1 city, because this type of city helps companies to overcome LoF (Goerzen et al., 2013). Finally, the FDI motive plays a role in location behavior as well, where different motives can be linked to the choice for a different city type. In the next chapter, the methods for analyzing the data with regards to these topics are discussed and the two chapters that follow will present the findings of the analysis.
3. Methodology
This chapter begins by introducing the ontological and epistemological assumptions, reflecting the postpositivist research paradigm adopted, which serve as the basis for the broader methodological choices. Afterwards, an explanation of the choice for qualitative research and in particular the choice for a multiple case study design are provided, including a discussion of quality criteria and of the theoretical case sampling. The chapter ends with an overview of how data was collected and analyzed.
3.1 Research philosophy
Research philosophy is concerned with the way researchers perceive the world and whether they see it objectively. Ontology deals with the question of how the researcher observes reality, where the researcher’s view of reality can range from extremely subjective in which reality is seen as a projection of human imagination to extremely objective in which reality is a concrete structure (Morgan & Smircich, 1980). In this thesis, the postpositivist view is adopted in which reality is seen as objective and can be understood through the mirror of science (Gephart, 2004). The postpositivist stance in this study requires methods of collecting and analyzing factual depictions that reveal truths or realities that can be used to evaluate or falsify propositions.
Epistemology deals with knowledge claims: what knowledge is, which knowledge humans can possess and how humans can obtain knowledge (Morgan & Smircich, 1980). A fundamental issue of epistemology is whether people can achieve any form of knowledge that is independent of their own subjective interpretation, since this knowledge is perceived or experienced by them (Morgan & Smircich, 1980). The epistemological stance of this thesis is one that flows naturally from the ontological point of view, as the two cannot be separated. The objective ontological stance adopted has a view of the world as an external and
independent reality of the researcher. However, the post-positivist view is one in which researchers cannot know what the real truth is and therefore there is a need to triangulate data from different sources in order to get a more complete picture of reality. From this point of view, the methods applied in this study should be coherent with this objective view and thus the post-positivist epistemology is appropriate.
3.2 Multiple case study research design
There is no one optimal research strategy and each research strategy has its advantages and disadvantages (Benbassat et al., 1987). The nature of the research topic and the goals of the researcher determine which strategy is most suitable for a particular study. Qualitative research is very suitable to develop a deeper understanding of a phenomenon and is therefore the appropriate choice to study the complexities underlying location choice in relation to subnational locations within the regionalization strategies of MNEs. The research thus adopts a multiple-case study design in order to study location choices for entry decisions of MNEs, analyzing the choice for a particular city in relation to the investment motive of the entry decision.
The distinguishing characteristic of a case study is that it attempts to examine a contemporary phenomenon in its real-life context (Yin, 2009). A case study is an event or unit of analysis determined by the researcher and is usually examined as it changes over a period of time (Gephart, 2004). Case studies are often used to answer research questions that start with “how” and “why”, because they allow the in-depth analysis of a phenomenon, when there is no need to control the context and when the focus lies on contemporary events (Yin, 2009). Case study research is appropriate when theory about the phenomenon of the research is in an early, formative stage (Benbassat et al., 1987, Eisenhardt, 1989). Given that the topic of location choice in combination with cities and investment motives has rarely been studied
in-depth, a case study approach is very suitable to investigate this phenomenon. The aim of this research is to build theory, developed by recognizing patterns of relationships among constructs within and across cases and their underlying arguments (Eisenhardt & Graebner, 2007).
This research applies a multiple-case study design, because it enables comparisons across cases that clarify whether an emergent finding only applies to one case or is consistently replicated by several cases. Furthermore, it creates more robust theory than a single-case study, because the propositions are deeper grounded in varied empirical evidence (Eisenhardt & Grabner, 2007). There are three cases in this research that each contain two MNEs each as embedded units of analysis. The use of this multiple-case approach allows the thorough and detailed investigation of each MNE to establish a robust within-case analysis and allows a more credible comparison of different patterns across cases.
3.2.2 Quality criteria and the case study research design
The positivist approach requires the development of a theoretical framework before testing through empirical evidence and stresses the importance of reliability and validity (Brannick & Coghlan, 2007). Research actions to increase the rigor of a study can be grouped into four categories: construct validity, internal validity, external validity and reliability (Gibbert & Ruigrok, 2010, Yin, 1994).
Construct validity of a procedure refers to the extent to which a study investigates the phenomenon that it claims to investigate or to which extent the results are an accurate observation of reality (Yin, 2009). It mainly refers to the subjective judgments of the researcher while collecting data. To establish construct validity, data triangulation is applied by using different sources of data (Yin, 2009). Data triangulation is the combination of methodologies in the study of the same phenomenon and allows researchers to be more
confident in their results (Jick, 1979). Multiple data sources are used during the data collection phase and include qualitative data gathered through newspaper articles from the Financial Times London, company histories and company annual accounts, quantitative data from the Zephyr database on mergers and acquisitions and a mix of quantitative and qualitative data on cities (from multiple data sources). Another way to create construct validity is to establish a chain of evidence (Yin, 2009), which will be done by providing a detailed description of the data collection and analysis process in this chapter.
Internal validity refers to the presence of causal relationships between variables and results (Gibbert & Ruigrok, 2010) and mainly applies to the data analysis phase. The first concern is that the researcher incorrectly concludes that there is a causal relationship between two variables without knowing that some other variable may have actually caused it (Yin, 2009). A second concern involves inferences of the investigator about an event that cannot be directly observed, but is likely to have resulted from an earlier occurrence (Yin, 2009). To strengthen internal validity the researcher should establish a plausible causal argument that is strong enough to defend the research conclusions and therefore should be based on a conceptual framework. Ideally the conceptual framework should integrate multiple theoretical views to create theoretical propositions to base the case study analysis on. In the conceptual framework a number of working propositions have been developed to build the case study analysis on. Pattern matching also creates internal validity, which involves the comparison of different patterns across different contexts (Yin, 1994, Eisenhardt, 1989). The different contexts are represented by the different cases and case companies, using a multiple-case study design for this study. If the patterns from the cases coincide with the patterns expected from the theoretical framework, it will help to increase internal validity (Yin, 1994). By adopting a multiple-case study design with embedded units of analysis, it is possible for this
pattern matching to take place within as well as across cases, thus strengthening the overall credibility of the study.
External validity defines the domain to which the results of the study can be generalized and is related to the intuitive belief that theories have to account for phenomena outside the context in which they are studied (Gibbert & Ruigrok, 2010, Yin, 2009). Case studies do not allow for statistical generalization and therefore the goal of case study research should be to pursue analytical or theoretical generalization. To obtain this, findings have to be tested through replications of the findings across cases (Yin, 2009). This can be done by applying cross-case analysis, while also providing the rationale for a case selection and details on the case study context (Gibbert & Ruigrok, 2010).
Reliability refers to the absence of random error, enabling subsequent researchers to arrive at the same insight if they replicated the study (Denzin & Lincoln, 1994, Yin, 2009). Even though data collection and analysis cannot be free from assumptions made by the researcher, reliability can be achieved through detailed presentation of the data and careful documentation and clarification of the research procedures. To increase reliability, as many steps as possible in the data collection and analysis phase have to be made operational to ensure the ability to replicate the study (Yin, 2009).
3.3 Case selection and sample
This multiple case study includes three cases that each contains two MNEs from the food industry. Table 1 provides a quick overview and Table 2 provides a detailed description of each case and each MNE. The three cases are food suppliers, food producers and food retailers. The cases chosen are the following: Archer Daniels Midland (food supplier); Bunge (food supplier); Kraft (food producer); Nestlé (food consumer producer) Carrefour (food retailer); and Tesco (food retailer). The choice for each case to include two companies allows
for within-case analysis across different embedded units of analysis and the three cases together allow for analysis across cases. These six MNEs embedded in three cases form an interesting selection and fit with the research question, because they are positioned at different stages of the food industry value chain and will therefore pursue different FDI motives. The regionalization profiles of the companies are also taken into account to analyze its effect on the firm’s location behavior.
Bunge ADM Kraft Nestlé Tesco Carrefour North America Bi-regional Home-regional Bi-regional Europe Global Home-regional Home-regional
Food suppliers Food producers Food retailers
Upstream <---> Downstream
Table 1. Overview case MNEs Source: Author
This study is done in the context of the regionalization debate, which states that most MNEs are home-region oriented and do not operate on a global, but on a regional scale. The regional orientations of the case MNEs are taken into consideration as well. There are four home region MNEs (Carrefour, Archer Daniels Midland, Kraft and Tesco), one bi-regional MNE (Bunge) and one global MNE (Nestlé). Furthermore, three case MNEs come from North America (Archer Daniels Midland, Kraft Foods and Bunge) and three case MNEs have Europe as their home base (Carrefour, Tesco and Nestlé). The home regions of these firms are important, because of their influence on location choice. The diverse sampling of the cases across different home regions, regional orientations and locations at the industry value chain will increase the ability to generalize the results. The differences across these companies allow for a thorough evaluation of the influence of investment motives and regionalization patterns on the location choice at city level.
Company Archer Daniels Midland Bunge Kraft Foods Nestlé Tesco Carrefour Regional focus Home region Bi-regional Bi-regional Global Home region Home region
Home Region North America North America North America Europe Europe Europe
Ranking Global Fortune 500 (2012)
92 160 170 71 59 39
Case Food supplier Food supplier Food producer Food producer Food retail Food retail
Case description Food suppliers are involved in raising crops, producing food, processing food (preparation of fresh products for the market or manufacturing prepared food products), wholesale and distribution of food and trading commodities. Both companies in this case operate all over the world and are among the biggest food suppliers of the world.
Food producers are operating in between food production companies and food retailers, using input from food suppliers to make products that are sold to food retailers. Both companies in this case are producers of top brands which are well known to consumers.
Food retailers operate many different types of stores where people can buy groceries and other products. It ranges from mini markets to hypermarkets and from urban to rural locations. Food retailers are located at the end of the food industry value chain.
Company description Archer Daniels Midland
is an American food-processing and commodities-trading corporation. At their 270 worldwide plants, they convert corn, oilseeds, wheat and cocoa into products for food, animal feed, industrial and energy uses. ADM has an extensive crop origination and transportation network, connecting crops to markets in more than 140 countries and 6 continents.
Bunge is a global agribusiness based in the United States. Its main activities are originating and crushing grains and oilseeds, crushing oilseeds to produce oils, margarines, mayonnaise and other food products, crushing sugar, milling wheat and corn, selling fertilizer to farmers and trading commodities. Bunge is present in more than 40 countries. Kraft is an American grocery manufacturing and processing conglomerate. They have a portfolio of brands in the beverage, cheese, refrigerated meals and grocery categories. After 2012, Kraft split into a North American part and an international part. Kraft sells its products in more than 80 countries worldwide.
Nestlé is a Swiss food and beverage company and is the largest food company of the world in terms of revenues. Their products include baby food, bottled water, breakfast cereals, coffee and tea, confectionery, dairy products, ice cream, frozen food, pet foods and snacks. Nestlé operates in 194 countries and employs around 333.000 people.
Tesco is a British multinational grocery and merchandise retailer and is the second-largest retailer of the world based on profits. Since the 1990s Tesco started diversifying their product offer to attract a wide range of customers and added new store formats, like Tesco Express. They employ 530.000 people in their stores across 12 countries.
Carrefour is a French format, multi-channel, multi-location retailer. Their product offer is based on a few principles, including a broad selection, the lowest prices, the highest quality, compliance with manufacturing conditions and a balanced diet for everyone. They operate more than 10.000 stores in 34 different countries. Table 2. Detailed description of cases and MNEs
3.4 Data collection
The study makes use of primary and secondary data. Primary data comes in the form of newspaper articles, company histories and company annual accounts. Newspaper articles are gathered from the Financial Times London through Lexis Nexis. Using the name of the case MNEs as key search term, articles were collected for the period 2000 to 2013. The articles were systematically reviewed to identify articles that provide significant information on company entry decisions. Company reports and histories were obtained from company websites and provide additional insights on entry decisions, location choices and investment motives and were also used to classify entry decisions as Greenfield. Secondary data was used to acquire information on the international investments of the MNEs from the Zephyr database on mergers and acquisitions. The regionalization profiles of MNEs were established by using data of Master Thesis students of the Amsterdam Business School of the University of Amsterdam, who gathered this data under the supervision of Dr. J. Lindeque (LMNERDB, 2013). The articles from the Financial Times and the company annual accounts were coded and analyzed using Nvivo software. The coding and analysis processes are described in more detail below. An overview of the data sources is provided in Table 3. It shows how many articles were found in the Financial Times and between brackets how many of them were analyzed, and this is repeated for the investments that came out of the Zephyr database. “Location choices identified” indicates the amount of locations that were connected to a country, whereas “Cities identified” shows how many locations could be connected to a city as well. “Investment motives identified” is the number of FDI motives that could be extracted from the FT articles, company history or annual reports.
Company name ADM Bunge Carrefour Kraft Nestlé Tesco Financial times 378 (21) 275 (9) 1875 (56) 2308 (30) 2997 (47) 7745 (103)
Zephyr database 63 (29) 55 (26) 227 (40) 97 (13) 221 (48) 225 (39)
Location choices identified 41 32 42 13 48 44
Cities identified 35 29 26 8 36 17
Investment motives identified 13 5 7 6 12 28 Table 3. Data sources of case MNE investments
Note: The numbers represent how many articles (Financial Times) or entries (Zephyr)
were found and the numbers in brackets represent the selected articles and entries, respectively. Source: Author
To divide the different types of cities into three categories, multiple variables were established and data was collected from a multitude of sources. The first variable was population, chosen because it is an important indicator of market size. The second selected variable indicated whether the city is a capital of its country or province/state, because it shows the importance of the city within its nation or state. The third variable signals whether the city has a stock market, because it indicates its economic importance within the world. The fourth variable states whether a top university is present, because this is an indicator of a high-skilled labor market and of the sophistication of the market. The fifth variable shows if there is an airport located in or near the city, indicating the level of connectivity of the city to the rest of the world. The sixth selected variable signals whether the city is situated at a coast, indicating the efficiency for transportation matters. The final variable shows if the city is a global city according to Beaverstock et al. (1999), which indicates whether the city is likely to be Tier-1 city. Table 4 provides an overview of the variables and the data sources. The cities were categorized using the discussed variables. Tier-1 cities were categorized as such when it has more than 500.000 inhabitants and a positive outcome to at least four out of six variables. Tier-2 cities had to be larger than 50.000 inhabitants and a positive answer to at least two of the six variables. Tier-3 cities were characterized by having less than 50.000 inhabitants and none or only one positive indication of the six variables. Appendix A shows a table with all cities and their categorization.
Variable Source
Size of the city population Population statistics at http://www.citypopulation.de Capital of country or province/state Various sources, including government websites Stock exchange presence
Global stock valuation database at
http://www.numeraire.com/exchanges/abc.htm
Top university presence Academic world ranking at http://www.shanghairanking.com/ Airport presence World airport database at http://www.world-airport-codes.com/ Location of city at the coast Google maps at http://www.maps.google.com
Global city according to Beaverstock (1999)
Goerzen, A., Asmussen, C. and Nielsen, B., 2013. Global cities and multinational
enterprise location strategy, Journal of International Business Studies, vol. 44, pp. 427 – 450.
Table 4. Overview variables of city data and sources Source: Author
3.5 Data analysis
Data analysis consists of examining, categorizing and tabulating the evidence to address the initial propositions of a study (Yin, 2009). The goal of the data analysis is to treat the evidence fairly, produce compelling analytic conclusions and rule out alternative interpretations (Yin, 2009). During the data analysis, information and data from different sources was constantly crosschecked between the newspaper articles, the Zephyr database, company histories and company annual accounts to increase reliability and accuracy of the findings. Qualitative sources such as newspaper articles from the Financial Times will only be cited when used in the text or tables of this thesis document.
To increase internal validity this study makes use of pattern matching and explanation building. Pattern matching refers to the comparison of predicted patterns based on theory and patterns that emerge during the data analysis (Yin, 2009). Explanation building is used to stipulate a set of causal links about a phenomenon and includes comparing case findings to theoretical propositions (Yin, 2009). Both pattern matching and explanation building are based on the theoretical propositions that were formed in the conceptual framework of this study. These theoretical propositions also form the basis for thematic codes that are used to guide the analysis and thus enable systematic generalization to the relevant theory (Gibbert & Ruigrok, 2010, Yin, 2009).
Thematic analysis is useful in capturing the complexities of meaning within a textual data set (Guest et al., 2012). In this study the textual dataset consists of the selected articles from the Financial Times and annual accounts from the case companies. The use of pre-established themes from theory will help to connect the data to theory (Ryan & Bernard, 2003). The thematic codes that are established (see Table 5) will guide the analysis by allowing the researcher to code the data while analyzing it in Nvivo. When all relevant data is coded it allows the researcher to search for codes in the program to compare data and recognize patterns within and across cases.
Code Sub-codes Definition
Working proposition Entry mode Greenfield
FDI where the MNE sets up a subsidiary
independently
Joint venture
FDI whereby a subsidiary is set up together with
another company
Acquisition FDI whereby a subsidiary is bought by the MNE
First entry First entry When a company enters a new country Proposition 5 Investment
motive Market seeking The reason for FDI is market seeking Proposition 1 Resource seeking The reason for FDI is resource seeking Proposition 4 Efficiency seeking The reason for FDI is efficiency seeking Proposition 3
Strategic asset
seeking The reason for FDI is strategic asset seeking Proposition 2 City
characteristics Tier 1 The city is categorized as Tier 1 or global
Tier 2 The city is categorized as Tier 2
Tier 3 The city is categorized as Tier 3
Non-city The investment is not related to a city MNEs Archer Daniels Midland Bunge Carrefour Kraft Tesco
Region Asia Pacific
Central America
European Union
Latin America
North America
Table 5. Nvivo codes Source: Author
The analysis consists of within-case analysis and cross-case analysis. During the within-case analysis, narrative accounts of each case are constructed that will describe unique patterns of each case. The goal of within-case analysis is for the researcher to become intimately familiar with each case as a stand-alone entity, which allows unique patterns of each case to emerge before researchers push to generalize patterns across cases (Eisenhardt, 1989). During the cross-case analysis, emerged patterns from each of the cases are compared and contrasted across cases and are also matched with the theoretical propositions (pattern-matching). During the analysis there is a high risk of investigators reaching premature and even false conclusions as a result of information-processing biases (Eisenhardt, 1989). Therefore, to increase construct validity, the data was looked at in multiple ways, comparing and contrasting data from annual accounts, Zephyr, company history and newspaper articles about the same entry and location decision. The emerged patterns across cases will be discussed in the cross-case analysis where each proposition will be answered separately. The research concludes with a section answering the research question in general and discussing the limitations of the research, its relevance for managers and implications for future research.
4. Within-case analysis
The following section will present the results of the within-case analysis, which will provide an overview of the expansion process of the six embedded case companies between 2000 and 2013, providing a unique case description for each MNE. This expansion process is examined along the lines of the different themes of the conceptual framework in order to take a closer look at location behavior of MNEs at the city level of analysis. The patterns that emerge from each embedded case unit are compared and contrasted with the other MNE in the case to provide a complete picture of the case.
4.1 Food production companies
Both Bunge and Archer Daniels Midland (ADM) are food suppliers and are characterized by a variety of activities that include extraction, crushing, processing and transportation of different food products, oils and fats. Bunge and ADM are seen as two of the four biggest food-suppliers of the world. They are located at the beginning of the food industry value chain, because they source and process raw products. Their main reasons for investing abroad are resource seeking and market seeking.
4.1.1. Archer Daniels Midland
ADM is one of the world’s largest suppliers and processors of oilseeds, corn, wheat and cocoa. Their customers include global food manufacturers, food processors, oil companies and industrial firms. They are present in 60 countries, employ over 26.000 people and operate more than seven hundred processing plants. Even though ADM has a presence in 60 countries, with 52% of their sales taking place in North America their strategy is home-region oriented according to the Rugman and Verbeke (2004) classification (LMNERDB, 2013).
An overview of the entry decisions per city category for both the home region and host regions is shown in Table 6. Based on the regionalization literature, this study would expect ADM to invest more in its home region than outside of it due to its home-regional orientation. However, the table shows that a slightly higher percentage of investments takes place outside of the home-region. There is no evidence that investments outside of the home region are more likely to be in a Tier-1 city, supporting the idea that resource seeking companies do not feel liability of foreignness as strongly as market seeking companies (Ghoshal & Nohria, 1989) and supports the proposition that companies with a resource seeking investment have the propensity to locate mainly in Tier-2 and Tier-3 cities. Between 2000 and 2013 ADM does not perform any first entries and therefore there is no data to analyze the influence of first entries on location choice for this firm.
Home or host
region Tier-1 Tier-1 % Tier-2 Tier-2 % Tier-3 Tier-3 % Total investments Total percentage Host regions 5 25% 10 50% 5 25% 20 57% Home region 4 27% 4 27% 7 47% 15 43%
Table 6. ADM:Entries per city category divided between home region and host regions. Source: Author
Table 7 displays an overview of all investments that were linked to an investment motive and shows that ADM engages in FDI mainly for resource seeking reasons. These investments are often linked to a Tier-3 city. For example, ADM’s purchase of six flour-milling plants in the UK, four of which are located in a Tier-3 city. A Tier-3 city could be attractive for this type of investment, because Tier-3 cities are characterized by low labor costs and the availability of raw material (Hymer, 1972). Most headquarters of companies acquired by ADM are located in a Tier-1 city, such as Wilmar Holdings in Singapore. A tier-1 city is a suitable location for a headquarters office, because of the city’s high connectivity to the rest of the world (Goerzen et al., 2013) and a Tier-1 city is a typical location for goal determination and planning (Hymer, 1972).
Type of entry
Date Acquired company First entry
Activity Investment motive Country City Category Quotes
Acquisition 2001 Intrade NV No Marketing, supply and trading
of fuels and oils Resource seeking Netherlands Amsterdam 1
Acquisition 2001 Kao No North water treatment plant Resource seeking USA Decatur, 1llinois 3 “ADM will build a pilot plant to produce oil at its Decatur, Illinois, base - and use this to develop sales in the US and globally.” (Tait, 2001 p.38)
Acquisition 2001 Doysan Yag Sanayii No Vegetable oil production Resource seeking Turkey HQ: 1stanbul 1 Acquisition 2001 Sociedad Aceitera
del Oriente
No Vegetable oil production Resource seeking Bolivia Santa Cruz de la Sierra 2 Acquisition 2002 Associated British Foods No Flour mills Resource seeking UK Edinburgh 1 Liverpool 2 Castleford 3 Knottingley 3 Corby 3 Tewkesbury 3 Joint venture 2005 Wilmar Holdings No Production of ethanol and
bio fuel
Resource seeking Singapore Singapore 1 “Archer Daniels Midland plans to invest an extra Dollars 900m over the next two years to boost production of ethanol and bio-diesel fuels as the world's largest grain processor bets that the increasing use of renewable fuels will accelerate amid high oil prices.” (Grant, 2005 p.27)
Joint venture 2008 Grupo Cabrera No Sugarcane agriculture and production
Resource seeking Brazil Limeira Do Oeste 3 Acquisition 2009 Schokinag Schokolade
Industrie Herrmann No Chocolate manufacturer Market seeking Germany HQ: Mannheim 1
Table 7. ADM: Overview of investment decisions. Source: Author
4.1.2. Bunge
Bunge mainly engages in the processing of oilseeds, grains, sugarcane, wheat and corn, and the production of different types of food products. They are present in more than 40 countries globally. Bunge can be classified as a bi-regionally oriented company according to the Rugman and Verbeke (2004) classification, because they have 29% of their sales in North America, 32% in Europe and 18% in Asia (LMNERDB, 2013).
Table 8 shows that Bunge has 85 per cent of its investments outside of its home region, which confirms the expectation of this thesis that a company with a bi-regional strategy will perform most of its investments in host regions. Although Bunge’s investments are spread across the different city types, inside the home region it has a preference for Tier-3 cities. This reinforces the idea that companies locate more often in Tier-3 cities inside their home region because intra-regional liability of foreignness is lower. With respect to first entries, Bunge invests in China for the first time in 2005 by purchasing a plant located in Rizhao (Tier-2 city). This is a logical choice for the location of a plant since Rizhao is located at the coast and is a port-city, which makes transportation of goods very efficient. It also supports the idea that liability of foreignness is less acute for resource seeking companies and might therefore be less bothered by LoF (Ghoshal & Nohria, 1989). On the other hand, the percentage of investments in Tier-1 cities outside of the home region is higher than inside the home region, which does support the idea that Tier-1 cities help companies overcome liability of foreignness, even for resource seeking firms.
Home or host region
Tier-1 Tier-1 % Tier-2 Tier-2 % Tier-3 Tier-3 % Total investments Total percentage Host regions 8 35% 7 30% 8 35% 23 85% Home region 1 25% 1 25% 2 50% 4 15%
Table 8. Bunge: Entries per city category divided between home region and host regions. Source: Author
Type of entry Date Acquired company First entry
Activity FDI motive Country City Category Quotes
Greenfield 2000 - No Trading Market seeking China Shanghai 1 Acquisition 2001 Geepee Ceval No Oilseed processing Resource seeking India Kota 2
Acquisition 2002 Cereol No Soybean processing Efficiency seeking France Paris 1 “In agribusiness it is so important to be efficient. It is highly competitive - few businesses are as transparent as commodities . . . the acquisition is very important to address scale and efficiency, and enable us to reduce logistics expenses.” (Daniel, 2002 p.28) Acquisition 2005 Sanwei Group Yes Soybean processing Market seeking China Rizhao 2 “The move underscores the interest of western
commodity processing companies in providing grain and oilseed processing capacity to China as the country's appetite for protein-based commodities grows.” (Grant, 2005 p. 26)
Acquisition 2008 Corn Products
International No Food processing Resource seeking USA Westchester, 1llinois 3 “Bunge, the world's largest oilseed processor, has agreed to buy Corn Products International, a producer of high-fructose corn syrup and food additives, for $4.4bn, as the boom in commodity prices prompts US agribusinesses to seek increased global scale.” (Weitzman, 2008 p. 15)
Table 9. Bunge: Overview of investment decisions. Source: Author
Bunge internationalizes for different reasons and an overview of Bunge’s entry decisions that were linked to an investment motive can be found in Table 9. Investments for resource seeking reasons are connected to the choice for Tier-2 or Tier-3 cities. Bunge agrees to buy Corn Products International, which is headquartered in Westchester, Illinois, USA (Tier-3), a suburb of Chicago (Tier-1) and therefore has access to the city’s high skilled labor and can make use of Chicago’s high connectivity to the rest of the world. Market seeking investments are connected to the choice for Tier-1 and Tier-2 cities. Locating their Chinese trading office (market seeking) in Shanghai is a logical choice, because it will have access to the stock market and many advanced producer services are present (Goerzen et al., 2013).
4.1.3. Case synthesis
An overview of the food supplier case is provided in Table 10. A pattern that emerges from both embedded units is that Tier-3 cities are chosen more often when the investment takes place inside the home region. Host region FDI doesn’t seem related to Tier-1 cities for these companies, strengthening the idea that liability of foreignness is less serious for resource seeking investments (Ghoshal & Nohria, 1989). With respect to FDI motives, it seems that resource seeking can often be related to Tier-2 or Tier-3 cities when the investment involves a plant or production facility, but headquarters are likely to be located in a Tier-1 city, following the expected outcome that plants and production facilities are attracted to locations that are abundant in low skilled labor and raw materials such as Tier-3 cities, while headquarters need the highly skilled employees, connectedness and advanced producer services that can be found in Tier-1 city. Another pattern that emerges is that market seeking FDI is connected to Tier-1 and Tier-2 cities, reinforcing the proposition that expects this association.
