Internationalization and firm performance: the
importance of regional strategy
Master Thesis
Duco Rosier – S2522225
d.w.b.s.rosier@student.rug.nl
Supervisor: Dr. O. Lindahl
Co-assessor: Dr. C. Schlägel
MSc International Business and Management
Rijksuniversiteit Groningen, Faculty of Economics and Business
Date of submission: 18-01-2019
2
Abstract
The relationship between the degree of internationalization and firm performance is one of the most debated and researched subjects in the field of international business. However, to date, the empirical results remain inconclusive and show positive, negative, and even curvilinear relationships, which may be explained by the use of diverse conceptualizations throughout literature. The key objective of this study is to contribute to this debate by using a multifaceted measure of internationalization in combination with the recently introduced regionalization perspective. This is done by exploring the effects of regional intensity and regional scope on firm financial performance of the world’s largest 93 multinational corporations. Based on transaction cost economics, resource-based reasoning, industrial organization economics and consumer psychology, it is argued that a more regionally orientated firm will achieve higher financial performance. Additionally, moderating roles for marketing as well as R&D intensity are assessed in this context. By conducting a multiple regression analysis it is found that both regional intensity and regional scope have a positive impact on firm financial performance. However, against prior expectations, marketing and R&D intensity do not moderate the positive regional effects of the two dimensions of internationalization on firm financial performance. The findings of this study thereby add to the growing importance of regionalization literature and have several theoretical and managerial implications.
3
Table of content
Abstract ... 2 List of Figures ... 5 List of Tables ... 5 List of Abbreviations ... 5 1. Introduction ... 6 2. Literature review ... 8 2.1 Regionalization ... 92.2 Reasons for regionalization ... 11
2.2.1 Liability of foreignness and regionalization ... 11
2.2.2 Firm-specific advantages and regionalization ... 12
2.3 Internationalization and MNC financial performance ... 14
2.4 Hypotheses ... 15
2.4.1 Regional intensity and MNC financial performance ... 15
2.4.2 Regional scope and MNC financial performance... 17
2.4.3 The moderating effect of marketing intensity on the DOI-financial performance relationship ... 19
2.4.4 The moderating effects of R&D intensity on the DOI-financial performance relationship . 20 2.5 Conceptual model ... 22 3. Research methodology ... 23 3.1 Data collection ... 23 3.2 Sample ... 23 3.3 Measurement of variables ... 24 3.3.1 Independent variables ... 24 3.3.2 Dependent variable ... 25 3.3.3 Moderating variables ... 25 3.3.4 Control variables ... 26
3.4 Empirical data analysis ... 27
3.4.1 Robustness tests ... 27
4. Results ... 28
4.1 Test of basic assumptions ... 28
4.2 Descriptive statistics ... 29
4.3 Correlations ... 30
4 4.5 Robustness tests ... 34 5. Discussion ... 35 6. Conclusion ... 39 6.1 Theoretical implications ... 39 6.2 Practical implications ... 40
6.3 Limitations and future research ... 41
Acknowledgments ... 42
References ... 42
Appendices ... 52
Appendix A: Industry types ... 52
Appendix B: Preliminary analysis ... 53
Appendix C: Home-region and industry type frequencies ... 56
Appendix D: Correlation matrix 2016 ... 57
Appendix E: Regression results robustness test 1 (2016 data) ... 58
Appendix F: Regression results robustness test 2 (2017 ROS) ... 59
Appendix G: Regression results robustness test 3 (2016 ROS) ... 60
Appendix H: Regression results robustness test 4 (lagged ROA) ... 61
5
List of Figures
Figure 1: Conceptual Model
List of Tables
Table 1: Descriptive statistics Table 2: Correlation matrix Table 3: Regression analysis
List of Abbreviations
DOI Degree of internationalization FDI Foreign direct investment FSA Firm specific advantage GDP Gross domestic product IB International business LOF Liability of foreignness MNC Multinational corporation
NAFTA North American Free Trade Agreement RBV Resource-based view
R&D Research and development ROA Return on assets
ROS Return on sales SD Standard deviation
6
1. Introduction
Ever since its inception in the 1980s, the term globalization has become widely used and adopted by society. It is commonly believed that we live in an era in which national cultures, national economies, national borders, and national territories are dissolving; visualized as the world being a ‘global village’. Central to this perception is the notion of a rapid and recent process of increased economic interdependence among nations (Hirst, Thompson & Bromley,
2009). Multinational companies (MNCs) can be seen as the key actors of this globalization
process which is taking place at an ever more increasing pace. Globalization in itself, however, is still a topic that is poorly understood (Peng & Pleggenkuhle-Miles, 2009). From a strategic perspective, globalization requires firms to operate beyond their home-country borders to remain competitive. A significant body of literature has been published containing ideas and theories as to how companies can capitalize on the ever more interconnected world (e.g., Yip, 2003). Despite this common perspective that MNCs should pursue global strategies, there has been an emerging stream of studies that challenge the notion of global strategies, one even calling globalization ‘a myth’ (Rugman, 2001). Especially since the publication of their 2004 article, Rugman & Verbeke have made a radical departure from the prevalent understanding of international strategy. In this article, Rugman and Verbeke demonstrate that most of the world’s largest 500 companies are actually not ‘global’ but ‘regional’. The main idea of their regionalization approach and the research stream emerging from it is that few MNCs are truly global and most MNCs are regionally-orientated, thus calling for a regionalization strategy. This perspective has generated intense debate and testing on the subject of internationalization of MNCs, thereby leading to many unresolved conceptual and empirical issues.
7
introduced arguments from various theoretical disciplines in order to explain the performance effects of internationalization. Individually these studies offer different performance explanations, however, collectively they suggest that internationalization is a complicated phenomenon which should not be analyzed from a single theoretical perspective. Critics therefore argue that internationalization should be seen as a multidimensional construct (Goerzen & Beamish, 2003; Miller, Lavie, & Delios, 2016). Miller, Lavie, & Delios (2016) therefore suggest that a distinction should be made between three different dimensions of internationalization: (1) an intensity dimension, (2) a diversity dimension, and (3) a distance dimension. On top of this, the introduction of the regional component has opened a new door regarding the DOI-financial performance relationship (Verbeke & Asmussen, 2016). Combined, a multi-dimensional construct of DOI, as well as a regionalization perspective, might thus shine a better light on the long-debated relationship between DOI and financial performance.
Next to this, the relationship between DOI and performance is also argued to be contingent on other factors (Bausch and Krist, 2007; Lu & Beamish, 2004). Especially firm-specific factors are found to have a moderating effect on the relationship of DOI and financial performance because these individual firm resources and capabilities are required to effectively maximize the advantages of international expansion (Hsu & Pereira, 2008; Lu & Beamish, 2004). Two of such firm-specific factors that positively influence financial performance are research and development (R&D) (e.g., Kotabe, 19990b) as well as marketing intensity (e.g., Morgan, Vorhies, & Mason, 2009), as they collectively allow firms to enhance their performance through premium pricing and superior products. It can therefore be argued that the impact of DOI on firm performance depends on a number of firm-specific factors, where R&D and marketing intensity have an especially important role within internationalizing firms as they enable firms to realize the inherent benefits of internationalization. However, only little research attention has been given on the possible interaction effects of R&D and marketing intensity on the relationship between DOI and the financial performance of MNCs.
8
since this has only received little research attention. This thesis thereby aims to bridge these research gaps by using the following main research question:
What is the effect of the regional degree of internationalization on financial performance of MNCs and how is this relationship moderated by their R&D and marketing intensity?
By examining the unclear DOI-financial relationship from a regional perspective combined with a multi-dimensional construct of DOI as well as the studying of two moderators, this thesis contributes to existing internationalization performance literature as well as to the emerging regional theory for MNCs. Also, important practical contributions are made by this thesis. As financial performance can be seen as the main goal of firms, new knowledge about how financial performance can be improved or influenced is highly useful for managers. This thesis contributes to this knowledge by showing how adopting a regional strategy affects financial performance.
In order to test the main research question, this paper will use multiple regression analysis on a sample of MNCs encompassing different industries. In the first part, the existing literature on internationalization and MNC financial performance is reviewed, as well as research on the influence of R&D and marketing intensity. Based on these key underlying concepts, the hypotheses and the core conceptual constructs of this research will be elaborated on. In the second part, the research design and methodology will be explained, including a description of the data. Subsequently, the data is analyzed in-depth and discussed. Last of all, the limitations and contributions of the study, as well as directions for future research, will be examined.
2. Literature review
9
over (local) competitors. Non-location-bound FSAs can be embodied in final products (usually leading to exporting), or shared across the MNC network (usually leading to the establishment of foreign subsidiaries), or exploited through contracting third parties (e.g. licensing). Regardless of how non-location-bound FSAs are transferred, they can lead to three types of ‘benefits of integration’ (Rugman & Verbeke, 1992): economies of scale (from concentrating specific activities in one location), economies of scope (from sharing resources across borders), and economies of exploiting national differences (from distributing the value chain across borders and coordinating these geographically dispersed activities) (Buckley & Casson, 1976; Caves 1971).
Internationalization, the degree to which a firm’s sales revenue or operations are gained from outside its home country, would thereby thus offer significant benefits for a firm. Partly due to Levitt (1983), the term global becomes mainstream in the 1980s, and firms are encouraged by both media and academics to ‘go global’. This globalization perspective focuses primarily on macro-level growth patterns in foreign direct investment (FDI) and trade and compares these data with national gross domestic product (GDP) growth rates. Next to the aforementioned benefits of integration, additional benefits of going global would, for example, include new market opportunities (Buhner, 1987), flexibility (Kogut, 1985a,b), avoidance of intense home-market competition (Elango, 2000), and learning (Ghoshal, 1987). The underlying assumption here is that globalization strategies should result in higher financial performance. However, next to the benefits, internationalization also brings costs and risks for MNCs. These costs and risks of operating internationally are also known as the liability of foreignness (Zaheer, 1995). One way for a firm to minimize this negative impact of internationalization is to seek markets that are closer to its home country. Returning to the surprising empirical observation of Rugman and Verbeke (2004), it becomes clear that the empirical evidence on internationalization of MNCs does not fully support the globalization perspective. Actually, their evidence points to the fact that most of the world’s largest MNCs follow a strategy that focusses on their close by home region; also called a regionalization strategy.
2.1 Regionalization
Semi-10
globalization implies that there is neither extreme geographical fragmentation of the world in national markets, nor that there is complete integration. Rugman and Verbeke (2004) study this phenomenon by analyzing micro-level growth data for the actors actually responsible for the FDI and trade flows: the MNCs; instead of analyzing the macro-level growth patterns for FDI and trade used by the mainstream globalization perspective. By examining the geographic sales data of the largest Fortune Global 500 MNCs in the three triad regions (NAFTA, European Union, and Asia), they find that most MNCs are actually not global, but regional. Rugman and Verbeke thereby classify MNCs according to 4 categories:
1. Home region orientated: at least 50% of sales are in the home region of the triad. This will systemically both shape and constrain the most important decisions and actions taken by the MNC.
2. Bi-regional: at least 20% of sales are in each of the two regions, but less than 50% in any one region.
3. Host region orientated: more than 50% of sales are in a triad market other than the home region of the MNC.
4. Global: more than 20% of sales in every three parts of the triad, but less than 50% in any one region of the triad.
Their sales data find only nine MNCs to be global, and only 25 to be bi-regional. The majority of the remaining MNCs in the Fortune Global 500 were found to be regional, thereby indicating the mislabeling of global associated with these firms. The empirical evidence for regionalization thereby suggests two important implications. First, much of the international activity of MNCs is conducted at the intra-regional level rather than the inter-regional (global) level (Rugman & Verbeke 2008). Second, many MNC operations are organized at the regional level instead of the global level (Rugman & Verbeke 2007).
Evidence confirming the regionalization hypothesis of Rugman and Verbeke has come from a number of studies. Whereas Dunning, Fujita, & Yakova (2006) confirmed the MNC home-region focus by looking at macro-level data on FDI stocks and flows, Rugman and Verbeke (2008a) and Rugman and Oh (2013) confirmed the prior results by including geographic distribution of assets into the analysis. Additionally, empirical studies (e.g., Arregle, Beamish, & Hebert, 2009; Asmussen & Goerzen, 2013; Qian, Khoury, Peng, & Qian, 2010) have shown that intra-regional expansion is much easier to accomplish than inter-regional expansion, at least in terms of final product market penetration.
11
(2007) support the findings of Rugman and Verbeke (2004), Dunning et al. find that regional concentration of MNC activity is more reflective of GDP and trade rather than of a particular MNC strategy. Moreover, Osegowitsch and Sammartino (2008) dispute Rugman and Verbeke’s (2004) classification criteria, and demonstrate in an empirical study that by using different schema, a significant proportion of firms gain global status. Additionally, Osegowitsch and Sammartino (2008) call for more longitudinal studies in the regionalization theory, as the data supporting the regional view only captures a snapshot in time. Overall, the main criticism thus seems to be one of measurement and taxonomy.
Notwithstanding its drawbacks, the study of Rugman and Verbeke (2004) has been the starting point of much academic attention concerning the globalization vs regionalization debate, even receiving a number of special issues in managerial journals (e.g., Management International Review, 45/1, 2005; Multinational Business Review, 18/1, 2010). Additionally, the new regionalization perspective has led to important implications for various strands of mainstream IB research. Part of the implications concerns the long-lasting IB debate on the relationship between the degree of internationalization and firm financial performance, wherein regionalization has opened a new way of assessing this relationship.
2.2 Reasons for regionalization
The regionalization perspective implicates that much of the international activity of MNCs is conducted at the intra-regional level rather than at the inter-regional (global) level. There are several reasons that can explain why this is the case.
2.2.1 Liability of foreignness and regionalization
12
several reasons why distance is lower when operating in the home region as compared to host regions. Taken together, these reasons are also known as the aforementioned liability of foreignness. The LOF is “(…) the costs of doing business abroad that result in a competitive disadvantage for an MNC subunit” (Zaheer, 1995: 342). In general, LOF can come from four sources: (1) costs directly associated with spatial distance, e.g. transportation costs; (2) firm-specific costs based on a firm’s unfamiliarity with and lack of roots in a local environment; (3) costs arising from the host country environment, e.g. economic nationalism; and (4) costs arising from the home country environment, e.g. restrictions on sales (Zaheer, 1995). The inter-relatedness among these four different dimensions causes the ‘spike’ in compounded distance at the regional border, mentioned above. On the one side, outsiders to the region face a compounded distance spike relative to the insiders. On the other side, insiders attempting to venture outside their home region face an inter-regional distance spike compared to the prevalent distances between countries in the home region (Aguilera et al., 2015). In regard to regionalization theory, Rugman & Verbeke (2007) argue that LOF is applicable on regions and thereby distinguish inter- and intra-regional LOF. In this sense, inter-regional LOF will be higher than intra-regional LOF, as well as the level of LOF, as, for example, spatial distance and the costs associated with it will be much smaller within a region than between regions. Due to the different levels of LOF, MNCs will thus prefer intra-regional expansion.
2.2.2 Firm-specific advantages and regionalization
13
commonality in some market characteristics, regionalization may serve as a good compromise between the two options of integration and responsiveness. By ‘upgrading’ their firm-specific advantages to ‘region-bound’ FSAs, firms could thus exploit some of the benefits of globalization, and at the same time, remain responsive to local market needs (Rugman & Verbeke, 2004).
This analysis of regional strategy can also be explained by the resource-based view (RBV) component of internationalization theory, as well as with the transaction cost economics dimension (TCE). First, the RBV component of internationalization theory points to the limits of international diversification (Narula & Verbeke, 2015). Firms’ resource bundles usually lead to more profitable growth opportunities in the home region compared to host regions, because there is a stronger similarity between the home country environment and that of other countries in the home regions (Rugman, Verbeke, & Nguyen, 2011a). Second, the TCE dimension of internationalization theory argues that (region-bound) FSAs must be protected through economizing mechanisms (Narula and Verbeke, 2015). Due to imperfect information and limited information processing capacity, seeking profitable growth inside the home region would seem consistent with the need to economize on bounded rationality (scarcity of mind). More intra-regional similarities both among country environments (e.g. regional commonalities in business taxation) and among the composition of firm-level activity in the relevant countries (e.g. semi-identical human resources management practices), reduces spatial transactions costs and thus encourages a home region strategy approach. As an outcome, bounded reliability challenges (scarcity of making good on open-ended promises) inside the home region can be substantially reduced, as compared to host regions (Verbeke & Greidanus, 2009).
14
compared to the home region, the effectiveness of these linked resources is therefore also much more uncertain.
2.3 Internationalization and MNC financial performance
Combined, the mechanisms outlined above have important implications for the financial performance of firms that choose to internationalize. Throughout the last four decades, this relationship between the degree of internationalization (DOI) and performance has been an important issue in IB literature, but one with inconclusive results (Annavarjula & Beldona, 2000; Thomas & Eden, 2004). Numerous studies have tested for this relationship, with different outcomes. Some have reported that the relationship between DOI and performance is
a positive linear one (Delios & Beamish, 1999; Kim, Hwang, & Burgers, 1989), while others
have found a negative linear one (Geringer et al., 2000; Majocchi & Zucchella, 2003). Also,
an U-shaped (Lu & Beamish, 2001), inverted U-shaped (Mauri & Sambharya, 2001; Caper & Kotabe, 2003), and even S-shaped relationship have been found (Lu & Beamish, 2004).
The mixed results for the relationship between DOI and financial performance may be explained by the use of different conceptualizations in existing literature (Miller, Lavie, & Delios, 2016). Individually, studies have offered a foundation for internationalization research by using mechanisms from distinct theoretical perspectives, such as learning theory (e.g., Lavie & Miller, 2008), industrial organization economics (e.g., Hennart, 2007), and resource-based reasoning (e.g., D’Aveni, Dagnino, & Smith, 2010). Collectively, however, these studies suggest that internationalization is a complex phenomenon, and thus a distinction should be made between different internationalization dimensions. According to Miller, Lavie, & Delios (2016), this distinction can be made between three different dimensions of internationalization: (1) an intensity dimension, which encompasses the firm’s commitment to serving customers in foreign markets; (2) a diversity dimension, which encompasses the breadth vs depth of internationalization; and (3) a distance dimension, which captures the differences between the characteristics of the firm’s home country and those of the host countries in which it is active. Most researchers have only focused on the first dimension, by using this as the one explanatory variable, thereby forgetting that the second and third dimension also might have an important influence in the analysis of internationalization.
15
Additionally, what also should be taken into account are the individual resources and capabilities that are actually required for MNCs to maximize the advantages of going abroad. Especially R&D and marketing intensity are inherent in realizing internationalization benefits
as they collectively allow firms to enhance their performance through premium pricing and
superior products.
2.4 Hypotheses
The review of the literature relating to internationalization reveals that results for the DOI-financial performance relationship are mixed. Important to note is, however, that these mixed results may be explained by the use of only one explanatory variable in most studies. Additionally, the introduction of the regional dimension in internationalization theory provides a new perspective on the DOI-financial performance relationship. Based on the benefits of the regionalization approach - lower LOF, more profitable growth opportunities from FSAs due to stronger similarity between the home country environment and that of other countries in the home region, and less bounded rationality and reliability problems - it appears that regional internationalization by firms, as opposed to global, would lead to a higher financial performance for MNCs. Combined, it can be argued that both a multi-dimensional construct of DOI as well as a regionalization perspective can reasonably explain the relationship between DOI and financial performance. Therefore, this study adopts a regional focus in capturing two different dimensions of DOI: regional intensity and regional scope. The regional intensity dimension thereby reflects the firm’s commitment to serving customers within, but not beyond, its home-region, whereas the regional scope dimension encompasses the breadth vs depth of regional internationalization by analyzing the spread of the firm’s operations within, but not beyond, its home-region.
2.4.1 Regional intensity and MNC financial performance
16
1972). These costs are likely to be incurred regardless of the host countries chosen for operations and are therefore indifferent to diversity and distance. Additionally, the firm will face relatively high costs of tariffs, transportation, and communication with foreign stakeholders due to its limited commitment in international markets (Hitt, Hoskisson, & Kim,
1997). Also, costs coming from foreign consumers’ stereotyping will face the firm, due to
consumer ethnocentrism (Eden & Miller, 2004). This is a psychological tendency in which consumers attribute a negative image to products based on their country of origin, thereby discounting the products’ perceived value (Elliott & Cameron, 1994). As a firm with a low international intensity generally lacks recognition abroad, foreign consumers may react negatively to its products or services. In order to minimize all these costs, firms will only enter into nearby countries.
However, the firm can overcome the initial costs of internationalization by attaining a higher international intensity: intra-regional expansion. Essential here is that the firm should only expand further within its home-region, as opposed to entering particular countries located in host-region markets. This distinction is important because of the bigger differences in markets outside of the home-region, as compared to markets within the home-region. For example, trade barriers between the United States and Canada are much lower than those between the United States and China. By attaining a higher regional intensity the firm can leverage the reputation of an MNC and benefit from better familiarity of consumers with its products, which can lower the negative effects of consumer ethnocentrism (Eden & Miller, 2004). The firm can now also economize on its setup costs and at the same time gain additional revenue streams from its activities in foreign markets within its home-region. With a sufficient level of operations organized within its home-region, the firm benefits from regional economies of scale, due to the efficient use of its infrastructure in supporting international operations (Hennart, 2007). For example, by concentrating production of standardized electronic products into one location, an MNC can produce at lower costs per unit due to specialization of production and personnel.
17
(Gruca & Nath, 1994). Therefore, an MNC that expands its international intensity within its home-region can achieve a relatively higher financial performance as compared to an MNC expanding its international intensity beyond its home region. Hence, I hypothesize:
Hypothesis 1: Higher regional intensity will be positively related to an MNC’s financial performance
2.4.2 Regional scope and MNC financial performance
The second dimension of DOI, corresponding to the diversity dimension of Miller, Lavie, & Delios (2016), is regional scope. A firm can enhance its international scope by increasing the spread of its operations across different foreign countries, which the firm is likely to do in an incremental way. When entering foreign countries, the firm incurs LOF (Zaheer, 1995), resulting from spatial distance costs, costs arising from the managers’ unfamiliarity with local host-country conditions, and costs associated with the imposing of inconsistent and discriminatory requirements by the host and home country environment. Additionally, as the international scope increases, the firm spreads its resource investments across a rising number of countries. Some of these resources, such as key managing personnel, are typically required for supporting local responsiveness (Ghoshal & Bartlett, 1990), which makes them unavailable for use in other countries. International scope can thereby thus thin out the firm’s investments, which creates negative competitive consequences for the firm (D’Aveni, Dagnino, & Smith, 2010). This thinning out of resources is attributed to international scope instead of international intensity, as in the latter case resource investments can generate economies of scale. Next to this, according to TCE, dispersion of the firm’s international operations increases internal governance costs, associated with coordination, control, information exchange, flexibility and efficiency, which are required for managing activities across different countries (Buckley & Casson, 1976; Rugman & Verbeke, 1992). Social cohesion and communication frequency are undermined by increases in international scope, resulting in difficulties to monitor a particular number of foreign operations located across different host countries (Greve, 1998). Coordination costs also accumulate due to scheduling and capacity constraints (Ghoshal & Bartlett, 1990).
18
countries. By attaining a higher international scope within its home region, the firm benefits from access to an increasing pool of diverse resources, including low-cost labor, advanced technologies, and natural resources not available in its home country (Rugman & Verbeke, 2003). Also, the earlier coordination costs can now be capitalized on, as the firm has learned from its ongoing attempts to coordinate activities across different host countries (Bartlett & Ghoshal, 2003). Greater need for coordinating the firm’s foreign operations often leads to establishing an appropriate governance structure, which can now be used effectively within the home region. This need for coordination is directly associated with the number of countries in which the firm operates, rather than with the intensity of operations or the national characteristics of host countries. Additionally, with a sufficient number of countries hosting its foreign operations within its home region, the firm benefits from economies of scope, due to the internal sharing of firm-specific resources such as shared accounting practices, and technologies (Rumelt, 1974). By engaging in regional procurement, sharing production and marketing resources, regional servicing of customers, and combining knowledge across markets within the home region, the firm can improve its performance, and thus overcome the initial costs of international scope.
However, if the firm would increase its international scope beyond its home region to host regions, it would lose some of the regional performance benefits. The distance between countries forming a region is much smaller than the distance to countries outside of the region, and thus inter-regional LOF, as well as the level of LOF, will be much higher than intra-regional LOF (Rugman & Verbeke, 2007). The growing number of host region foreign operations will also bring additional different regulatory and trade laws, as well as currency fluctuations, which are difficult to deal with given greater (inter-regional) bounded rationality. Additionally, diseconomies of scope can occur because of an increase in mismatching across the firm’s assets and services (Bresnahan, Greenstein, & Henderson, 2011). Therefore, an MNC that expands its international scope within its home region can achieve a relatively higher financial performance as compared to an MNC expanding its international scope beyond its home region. Hence, I hypothesize:
19
2.4.3 The moderating effect of marketing intensity on the DOI-financial performance relationship
While the above reasoning provides insights into the (regional) performance impact of DOI, it does not incorporate the heterogeneity among the MNC’s ability to manage its respective DOI. In order to study this, the individual resources and capabilities that are actually required to effectively maximize the advantages of internationalization for the MNC should be incorporated as well (Lu & Beamish, 2004). The DOI-financial relationship is thus argued to be particularly contingent on firm-specific factors, as they enable the firm to fully exploit the advantages of internationalization. Two firm-specific factors that especially influence financial performance in a positive way are marketing (e.g., Morgan, Vorhies, & Mason,
2009) as well as R&D intensity (e.g., Kotabe, 19990b).
The marketing intensity of a firm is related to its ability to use its tangible and intangible resources to understand complicated consumer-specific needs, achieve product
differentiation relative to competitors, and build superior brands (Day, 1994; Dutta,
Narashiman, & Surendra, 1999).A firm can develop its marketing intensity when it combines
individual skills and knowledge of its employees with the available resources (Vorhies &
Morgan, 2005). By spending money on advertising and promoting its products the firm can increase its sales via both expanding the sales of the product category and by convincing customers to switch to its brands. With strong brand names, the firm can charge premium
prices, and thereby increase its profitability (Song, Nason, & Benedetto, 2008). This is also
the case in foreign markets, due to the presence of intermarket segments across countries for
many products. Important here is that the MNC makes a distinction between the types of
20
MNCs cannot use semi-standardized marketing programs in such an effective manner and additional costs for more tailored marketing will be necessary. Additionally, difficulties arise in processing large amounts of diverse ambiguous market and consumer information as internal coordination of marketing activities becomes more difficult.
Therefore, MNCs that spend more money on marketing (marketing intense MNCs) achieve bigger earnings from the different regional dimensions of DOI than MNCs with a lower level of marketing intensity. Specifically, for the first dimension of DOI, regional intensity, an MNC can achieve higher revenues due to a better fit and targeting of home region customers’ needs to its own products and services, as well as greater efficiency by developing semi-standardized marketing programs (marketing-related economies of scale) across home region markets and having better bargaining power with consumers (Takeuchi & Porter, 1986). However, this targeting and efficiency would be more costly and difficult to obtain for a globally focused marketing intensity. Hence, I hypothesize:
Hypothesis 3a: A higher marketing intensity will positively moderate the positive effect of regional intensity on an MNC’s financial performance
For the second dimension of DOI, regional scope, a higher marketing intensity would mean that an MNC can emphasize differentiation by heavy advertising marketing activities (marketing-related economies of scope), which makes it more likely to succeed in a larger
number of diverse home-region markets than a firm that does not do this (Helsen, Jedidi, &
DeSarbo, 1993). On the other hand, an MNC that does this globally will not be able to achieve this as differentiation becomes increasingly difficult. Hence, I hypothesize:
Hypothesis 3b: A higher marketing intensity will positively moderate the positive effect of regional scope on an MNC’s financial performance
2.4.4 The moderating effects of R&D intensity on the DOI-financial performance relationship
21
Jayachandran, 2008). A firm’s R&D intensity can be developed by combining individual
skills and knowledge of employees with available resources, in order to develop new technical knowledge that can be combined with existing technology. By spending money on R&D the firm can thus create superior products and services, which gives the firm an advantage over competitors. This becomes useful when the firm expands to foreign markets as it can then charge premium prices for its innovative products (Hitt, Hoskisson, & Ireland, 1994; Porter, 1986). Expanding to foreign markets, however, calls for an arising need to adapt and tailor the firm’s innovative products to the host market in order for them to appeal to local customers. Therefore, one of the main motivations of foreign direct R&D investment is to make this adaptation and tailoring of the MNC’s innovative products possible (Vernon 1977; Utterback, 1989). Next to this, another main motivation is to acquire local technology and talent as this can be used to maintain firm competitiveness (Kuemmerle, 1997). Managing and organizing increasing foreign R&D activity, however, becomes progressively difficult as more foreign R&D activities require more linkages to other corporate units to coordinate their activities (Bartlett and Ghoshal, 1989). Complicated reporting requirements and the impression of external control can have negative effects on both innovative performance and on the ability to acquire local technology and talent (Florida, 1997). MNCs that focus their
R&D on more far away markets will, therefore, have more costs and difficulties in achieving positive effects from this R&D on their different DOI dimensions as much R&D resources will be misspent. On the other hand, MNCs that focus their R&D on markets in their home region will find it easier to aim this R&D at the right activities, which will thereby lead to
positive effects of this R&D on the impact of the different regional dimensions of DOI on
MNC financial performance. Again, the main reason for this stems from a lower intra-regional LOF compared to a higher inter-intra-regional LOF (Rugman & Verbeke, 2007).
Therefore, MNCs that spend more money on R&D (R&D intense MNCs) achieve bigger earnings from the different regional dimensions of DOI than MNCs with a lower level of R&D intensity. Specifically, for the first dimension of DOI, regional intensity, an MNC can achieve higher revenues due to the point that it can provide home region customers with better, more innovative products, as well as greater efficiency due to the ability to spread the costs of research over a larger sales base (R&D-related economies of scale). However, this will be more costly and difficult to achieve on a global level. Hence, I hypothesize:
22
For the second dimension of DOI, regional scope, a higher R&D intensity would mean that an MNC can share R&D spillovers across its network and thereby use it for different lines of products and services; R&D-related economies of scope (Henderson & Cockburn, 1996). This makes the firm’s products and services more likely to succeed in a larger number of diverse home region markets than a firm that does not invest in R&D. On the other hand, an MNC that does this globally will not be able to achieve this as differentiation becomes increasingly difficult. Hence, I hypothesize:
Hypothesis 4b: A higher R&D intensity will positively moderate the positive effect of regional scope on an MNC’s financial performance
2.5 Conceptual model
From the provided theoretical framework and hypotheses a conceptual model can be derived, as shown in figure 1. This figure shows the proposed positive relationship between the independent variables (different dimensions of DOI) – regional intensity (H1) and regional scope (H2) – on the dependent variable financial performance. Additionally, the positive moderating effect of the variables R&D (H3a-b) and marketing intensity (H4a-b) on the positive DOI-financial performance relationship is incorporated in the model.
23
3. Research methodology
3.1 Data collection
In order to test the presented hypotheses, this section will outline the research methodology. The previously described DOI-financial performance relationship and the moderating effects of R&D and marketing intensity on this relationship will thereby be studied by a quantitative approach. The secondary quantitative data that is used for this approach comes from databases made available to students from the University of Groningen as well as publically available data. The data for the dependent variable financial performance, the independent variable regional intensity1, as well as the moderating variables marketing and R&D intensity is obtained via the Compustat North America database provided by Wharton Research Data Services. This database covers financial, statistical and market information from publicly traded companies in the U.S. and Canada. Compustat also offers a global database, however, this study uses the North America database because this database contains more firms from the Fortune Global 500, the intended sample from this study, than the Compustat Global database. For the independent variable regional scope and for the control variables MNC size, MNC age, and industry effect the Orbis database published by Bureau van Dijk is used. Some missing values are checked through corporate websites and annual report data.
3.2 Sample
As this study researches the different DOI dimensions of MNCs through a regional lens, the sample focusses on the world’s largest companies located in either one of the three broad triad regions of North America (NAFTA), European Union (28), and Asia-Pacific. The importance of these three regions is reflected in, for example, their high-tech production, urge for innovation, large financial and trade flows, advanced infrastructure, and the inclusion of the three largest markets for most new products in the world: the U.S., the EU and Japan
(Gunaratne, 2002; Rugman & Verbeke, 2004). Another useful indicator of the triad’s
24
products and services around the world, making them an interesting sample to test. The initial sample of this study is therefore selected on basis of the Global Fortune 500 list, which ranks the world’s largest corporations on basis of their revenue. The ranking published in 2018 is used, where after collecting the data 93 companies are included to form the final sample. This is sufficient for a good sample size given the rule of thumb that a sample should include at least 10 observations per variable (VanVoorhis, & Morgan, 2007).
3.3 Measurement of variables
3.3.1 Independent variables
The measurement of the two dimensions of internationalization, regional intensity and regional scope are each measured at the firm-level. First, with regard to the intensity dimension, previous studies have indicated several measures. The one that is most widely practiced is the percentage of foreign sales to total sales (the foreign sales ratio) (e.g., Geringer et al., 2000; Kim et al., 1989; Caper & Kotabe, 2003). However, this measurement can be problematic because it neglects the differences between overseas markets, as argued in this study. Therefore, to accurately measure the regional intensity of an MNC, differentiation is needed between the various countries or markets an MNC has entered. Hence, following Rugman & Verbeke (2004), I operationalize regional intensity as sales within the home geographic region (including sales in the MNC country of origin) divided by the total sales of an MNC in the year 2017 (see formula 1). In this light, a region is defined according to the three broad triad regions of North America, Europe, and Asia. As the outcome of this variable is a percentage it can range between zero and one hundred and is therefore measured as a continuous variable.
Formula (1): Regional sales ratio = Sales within the home region
Total sales
25
2008). Yet again, this measurement does not take into account the differences between
overseas markets, which will be less present within a region. Hence, I operationalize regional scope as the number of subsidiaries an MNC has within its home region (including subsidiaries in the MNC country of origin) divided by the total subsidiaries of an MNC in the year 2017 (see formula 2). Again, as the outcome of this variable is a percentage it can range between zero and one hundred and is therefore measured as a continuous variable.
Formula (2): Regional country ratio = Number of subsidiaries in the home region
Total subsidiaries
3.3.2 Dependent variable
As is common in most internationalization studies, return on assets (ROA) will be used to measure the dependent variable financial performance (e.g., Delios & Beamish, 1999; Geringer et al., 2000; Hitt et al., 1997). An accounting-based measure is highly appropriate to use, contrary to market-based measures such as Tobin’s Q because market-based measures are not only affected by internationalization, but also by future prospects and earnings surprises (Ball & Kothari, 1991). Additionally, ROA is also more appropriate to use compared to return on equity, as this measure can be relatively easily influenced by a firm’s capital structure (Hitt et al., 1997). Return on sales is also used in internationalization literature to measure financial performance, however, it is argued that ROA and return on sales correlate strongly, thereby making no difference in using either one (Chao & Kumar, 2010). Hence, I operationalize an MNC’s financial performance as return on assets in the year 2017. ROA is calculated by dividing the MNC’s net income by the total value of the MNC’s assets at the end of the period (see formula 3) and indicates how efficient a company is in using its assets to generate earnings. The outcome is a percentage which can take any value, both positive and negative, making it a continuous variable.
Formula (3): ROA = Net income
End of period assets
3.3.3 Moderating variables
26
understanding of customer needs. In order to achieve this goal, an MNC needs to increase expenditures in marketing related activities. Therefore, consistent with earlier studies (e.g., Dutta et al., 1999; Slotegraff, Moorman, & Inman, 2003), I operationalize marketing intensity as the total advertising expenses divided by total sales in the year 2017 (see formula 4). As the outcome is a percentage which can take any value, both positive and negative, this makes it a continuous variable.
Formula (4): Marketing intensity = Total advertising expenses
Total sales
This logic is similar to an MNC’s R&D intensity, as the R&D goal of an MNC is to create effective new products and services, as well as improving old ones. This goal is, again, also partly reflected in an increase of sales through superior products and services. Achieving this goal can thereby be done by increasing R&D expenditures. Hence, following previous studies (e.g., Kotabe, 1990b; Mansfield, 1981), I operationalize R&D intensity as the total R&D expenses divided by total sales (see formula 5). Additionally, according to Langowitz & Graves (1992), it takes time for R&D spending to have an effect on performance; therefore a 3-year lag is chosen. As the outcome is a percentage which can take any value, both positive and negative, this makes it a continuous variable.
Formula (5): R&D intensity = Total R&D expenses
Total sales
3.3.4 Control variables
27
(e.g., Jung, Noh & Chung, 2016; Riahi-Belkaoui, 1996), MNC age is measured in this study as the difference in years between the MNC’s year of inception and 2017. This variable is measured as a discrete variable because it can take any integer value bigger than zero. Third, industry effect is also included as a control variable in this study as prior research showed that the type of industry the firm operates in influences firm financial performance (e.g., Schmalensee, 1985). Following Cuervo-Cazurra et al. (2018), four different industry sectors are used (see Appendix A): (1) natural resource-based industries; (2) manufacturing industries; (3) service industries; (4) others. Hence, industry sectors are measured as categories indicating the industry type an MNC operates in. For example, this implies that an MNC that produces cars is classified into category two. As this variable is measured with different categories, it is a categorical variable.
3.4 Empirical data analysis
In order to test the suggested hypotheses, IBM SPSS version 24 is used. As the study examines multiple independent variables to explain one continuous dependent variable, a multiple linear regression analysis is appropriate to use. Since there are moderation effects in this study, moderated multiple regression is used. This requires the independent variables and the continuous moderators marketing and R&D intensity to be mean centered for the analysis, as well as the computing of interaction variables (Hayes, 2013). Additionally, because regressions can only be run with ratio measured variables as well as dummy variables, the control variable industry effect required the creation of dummy variables. Therefore, the three categories are dummy coded into either a value of one if the MNC operates in the corresponding industry or zero if this is not the case.
In the case of missing data for one of the variables, the MNC is still included in the analysis of other hypotheses as this method (pairwise deletion) maintains valuable data (Enders, 2010).
3.4.1 Robustness tests
28
1999; Geringer et al., 2000; Hitt et al., 1997), and can be calculated by dividing the MNC’s net income by net sales (see formula 6). The outcome is a percentage which can take any value, both positive and negative, making it a continuous variable.
Formula (6): ROS = Net income
Net sales
Finally, in order to control for possible effects of endogeneity and reverse causality, a one-year lagged measure for the dependent variable MNC financial performance (both ROA and ROS) is tested. By analyzing this, the possibility that firm performance could affect the dimensions of DOI (Chen, 2007) and the moderators is checked for. The analysis uses the data for the independent variables, moderators, and control variables for the year 2016 and the data for the dependent variable ROA for 2017, and for robustness purposes ROS for 2017, creating a one-year lag.
4. Results
4.1 Test of basic assumptions
29
making a scatterplot with standardized residuals. This scatterplot shows randomly distributed dots, thereby proving that there is no pattern and thus meeting the assumption of linearity and homoscedasticity (Hayes, 2013). For details see Appendix B.
Additionally, several outliers occurred during the data analysis. As these outliers might influence the dataset, they are removed from the analysis.
4.2 Descriptive statistics
The descriptive statistics are presented in table 1. From the 93 sampled MNCs 48 operate in North America, 27 in the EU, and 18 in Asia-Pacific (see Appendix C, table A.1). Regarding industry type, 10 MNCs operate in natural resource-based industries, 48 in manufacturing industries, 31 in service industries, and 4 in other industries (see Appendix C, table A.2).
Table 1: Descriptive Statistics
2017 2016
Min Max Mean SD Min Max Mean SD
ROA (%) .12 10.18 5.07 1.89 -1.37 8.88 4.12 2.10 ROS (%) .28 12.39 6.06 2.63 -.62 10.24 5.61 2.51 Regional intensity (%) 10.23 99.45 57.97 20.62 9.39 99.69 56.98 22.13 Regional scope (%) 5.00 96.39 61.03 21.15 5.00 96.39 59.98 21.23 Marketing intensity (raw data) .09 5.88 1.65 1.27 .05 5.39 1.64 1.27 Marketing intensity (SQRT) .29 2.43 1.19 .49 .22 2.32 1.18 .50 R&D intensity (raw data) .04 13.87 3.54 3.58 .04 13.65 3.66 3.43 R&D intensity (SQRT) .21 3.72 1.62 .96 .21 3.69 1.67 .94 MNC size 10.10 11.97 10.91 .39 9.90 11.73 10.87 .40 MNC age (raw data) 3 171 61.42 42.61 2 170 59.77 42.83 MNC age (SQRT) 1.73 13.08 7.30 2.87 1.41 13.04 7.14 2.98
30
The data for the 2017 sample shows that the average ROA for an MNC is 5.02% (SD=1.93), while the average ROS is 6.06% (SD=2.63). From the variables regional intensity and regional scope, it becomes clear that MNCs mainly sell and operate in their home region, as intra-region sales are on average 57.97% of total MNC sales and MNCs have on average 61.03% of their subsidiaries in the MNC home region. The variables marketing and R&D intensity (raw data) show that the data is on average skewed toward the lower-value range, with marketing intensity (raw data) ranging from 0.09% to 5.88% (mean=1.65%) R&D intensity (raw data) ranging from 0.04% to 13.87% (mean=3.54%). The new and transformed marketing and R&D intensity (SQRT) variables shift the value range for the data and create a decreased standard deviation (marketing SD=0.49; R&D SD=0.96). This is also the case for the variable MNC age.
The 2016 sample used for robustness purposes, shows that an MNC’s ROA is, on average, 4.12% (SD=2.10), which is a bit lower compared to 2017. This is also the case for the average ROS, which is 5.61% (SD=2.51). Regional intensity (mean=56.98%) and scope (mean=59.98%) remain relatively unchanged, although a bit lower for 2016. The same goes for the moderator variable marketing intensity (mean=1.64%). Interesting to note is that R&D intensity is somewhat higher in 2016 (mean=3.66%) compared to 2017, although only very little. All in all, the variables remain almost unchanged.
4.3 Correlations
31
Table 2: Correlation matrix
ROA Regional intensity Regional scope
Marketing intensity
R&D
intensity MNC size MNC age
32
4.4 Regression results
The results of the regression analysis can be found in table 3. As can be seen in table 3, eight different models are used for this regression analysis. In the first model, the effects of the control variables on MNC financial performance are tested. MNC size is the only control variable that can explain the dependent variable on a significant level (p≤0.1).
In model 2 the independent variable regional intensity is added in a separate manner to the control variables, in order to test hypothesis 1. The results show that this variable has a significant positive effect (B=.513; p=.000), also when looking at the general regression model 4 (B=.291; p=.077). As both models support hypothesis 1, it can be accepted.
After this, the effect of the independent variable regional scope on MNC financial performance (hypothesis 2) is tested in model 3. From table 3 it becomes clear that regional scope also has a positive significant effect (B=.496; p=.000), which is again confirmed in model 4 (B=.277) at a significant level (p≤.1). Hence, both models support hypothesis 2, which can therefore be accepted.
Model 5 shows that the moderating variable marketing intensity does not really have a direct effect on the dependent variable (B=.093; p=.445). By adding the interaction terms in model 6, it becomes evident that the variable itself does not only rarely explain the dependent variable (B=.093; p=.451), it also rarely and insignificantly interacts with the two independent variables, which is against the prior indication of the correlation results. Interesting to note is that for both the interaction effects, regional intensity (B=-.016) as well as regional scope (B=-.001), the interaction is negative and small. Hence, marketing intensity does not seem to positively moderate the DOI effects on financial performance and hypotheses 3a-b need to be rejected.
33
Table 3: Regression analysis
Financial performance (ROA)
Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7 Model 8
Independent variables Regional intensity .513 (.010)*** .291 (.015)* .299 (.016)* .298 (.016)* .261 (.016) .377 (.016)** Regional scope .496(.010)*** .277 (.014)* .295 (.015)* .296 (.016)* .279 (.015)* .299 (.015)* Moderating variables Marketing intensity .093 (.469) .093 (.477) R&D intensity .261 (.232)** .224 (.227)* Interactions
Regional intensity X Marketing intensity -.016 (.039) Regional scope X Marketing intensity -.001(.037)
Regional intensity X R&D intensity .142 (.016)
Regional scope X R&D intensity .158 (.014)
Control variables
MNC size -.235 (.590)* -.169 (.519) -.274 (.516)** -.219 (.491)** -.232 (.541)** -.231 (.554)* -.278 (.537)** -.283 (.529)**
MNC age -.139 (.081) -.140 (.071) -.156 (.071) -.149 (.065) -.160 (.071) -.161 (.075) -.141 (.069) -.143 (.067) Natural resource-based industries (dummy) .216 (1.311) .166 (1.143) .191 (1.143) .174 (1.039) .214 (1.176) .211 (1.202) .253 (1.125) .170 (1.120) Manufacturing industries (dummy) .238 (1.147) .321 (1.001) .280 (1.000) .309 (.910) .319 (.993) .320 (1.015) .226 (.976) .103 (.984) Service industries (dummy) .180 (1.163) .094 (1.015) .103 (1.016) .088 (.923) .115 (1.016) .119 (1.041) .103 (.980) .004 (.982) R square .084 .316 .315 .344 .351 .351 .394 .449 Adjusted R square .011 .250 .249 .269 .264 .239 .312 .352 F-statistic 1.153 4.777** 4.753*** 5.405*** 4.054*** 3.138*** 4.802*** 4.642***
Note: N=93; The other industries is the baseline industry for the dummy effect; Standard errors in parentheses
34
4.5 Robustness tests
The results for the first robustness test, the 2016 analysis, can be found in Appendix E. This analysis appears to be consistent with the 2017 analysis, with some small differences. Models 2 (B=.570; p=.000) and 4 (B=.236; p=.070) support the significant positive effect of regional intensity, thereby confirming hypothesis 1. This is also the case for the significant positive effect of regional scope, which is supported in models 3 (B=.604; p=.000) and 4 (B=.446; p=.000), thereby also confirming hypothesis 2. The interactions from regional intensity (B=.208; p=.123) and regional scope (B=-.161; p=.228) with marketing intensity are stronger in the 2016 analysis compared to 2017 but remain insignificant, thus rejecting hypotheses 3a-b. Different to 2017, the insignificant interaction from marketing intensity with regional intensity is no longer negative in the 2016 analysis but positive, although it does remain negative for the interaction with regional scope. Finally, model 7 shows that there is no direct significant effect of R&D intensity (B=.150; p=.168), which is also the case in model 8 (B=.157; p=.158). This is different from the 2017 analysis, where there is a direct significant effect. Also, the interaction terms in model 8 with both the independent variables regional intensity (B=-.181; p=.211) and regional scope (B=-.028; p=.843) show no significance, which therefore rejects hypotheses 4a-b. Interesting to note is that these insignificant interactions are negative in the 2016 analysis, where they are positive in the 2017 analysis. All in all, the 2016 analysis supports the results of 2017.
35
hypotheses 4a-b. The interaction with regional intensity in the 2017 analysis shows a positive result, while the interaction with regional scope shows a negative result. In the 2016 analysis, both interactions are negative. Contrary to the 2016 ROA analysis, but consistent with the main 2017 ROA analysis, both ROS analyses deliver a direct positive significant effect for R&D intensity.
Lastly, the fourth and fifth robustness test check for endogeneity and reverse causality (see Appendix G for lagged ROA; Appendix I for lagged ROS). It appears that these two analyses also mostly confirm previous tests, although one important difference can be noted. Whereas regional scope is positive significant again in both model 3 and 4 for both tests, confirming hypothesis 2, regional intensity is only positive significant in model 2 in both tests. In the general regression model 4, regional intensity becomes insignificant in both years. So, while model 2 in both tests partially supports hypothesis 1, model 4 rejects it. Moreover, in the lagged ROA test, regional intensity has a(n) (insignificant) negative effect (B=-.094; p=.528), but this effect is too small to add much meaning to the results. In the lagged ROS test this effect still remains positive (B=.126; p=.396) in model 4. Next to this, the interaction effects of the moderators marketing and R&D intensity do not specifically change in comparison to all the other analyses. Although the effects are somewhat changed, all interactions remain insignificant for both the lagged ROA and the lagged ROS test, thereby rejecting hypotheses 3a-b and 4a-b.
Overall, most results are confirmed in the robustness checks, although there are some small differences. As most of these differences are insignificant effects, the differences do not impact the general outcome of this study. The differences in significance that do occur in the robustness tests are evaluated in the subsequent discussion part.
5. Discussion
36
This result is also generally consistent with existing regionalization research (e.g., Chen & Tan, 2012; Qian et al., 2008; Rugman & Oh, 2010; Sukpanich & Rugman, 2007).
Additionally, although only sporadically tested in existing internationalization research, interactions with marketing and R&D intensity for DOI dimensions are also tested within this study. However, these do not show explanatory power.
Being one of the most researched topics in international business literature, the impact of DOI on financial performance remains a topic of very mixed results. Contractor, Kundu, & Hsu (2003) and Lu & Beamish (2004) have independently argued that all these contradictory results might be reconciled in a three-stage theory of internationalization. Initially, in stage 1, the DOI-financial performance relationship has a negative slope as costs and barriers to internationalization need to be tackled. In stage 2, the relationship becomes positive because the initial costs are overcome by increased internationalization. In stage 3, the relationship becomes negative again due to the point that the growth of coordination and governance costs exceed the benefits of increased expansion. Taken into consideration this three-stage model of internationalization, the regionalization approach may offer an interesting explanation for the second stage. It is in this stage that the benefits of international expansion arise and that MNCs can reap from higher financial performance. As Oh & Contractor (2014) show, the ‘where’ factor of MNC internationalization matters significantly. MNCs that expand in their home region increase firm performance, while foreign-region expansion lowers firm performance.
37
explanatory power of the multidimensional construct of internationalization might also be of added value. Lavie & Miller (2016), for example, also include distance in their construct, which adds other explanatory mechanisms to the DOI-performance relationship, such as cultural distance and economic distance.
Focusing especially on regionalization studies, the significant positive explanatory effect found in this thesis for regional intensity also seems to be consistent with existing studies
(Chen & Tan, 2012; Sukpanich & Rugman, 2007). Moreover, the result holds true
independent of what year’s dataset is used or what the measurement method is for financial performance. Even when testing for endogeneity and reverse causality the results are partially supported, although not in all models. This finding could validate the argumentation that an MNC that commits to serving more foreign customers within, but not beyond, its home-region can benefit in general from higher financial performance, mostly due to obtained economies of scale. The results thereby show that it is not just foreign sales that matter but more importantly, where foreign sales are generated. It is therefore important in internationalization studies to take a more fine-grained approach by considering host country/region effects (Li, 2005; Rugman & Oh, 2007). Important to note is, however, that how a region is defined in internationalization studies can make a significant difference to the results (Asmussen, 2009; Osegowitsch & Sammartino, 2008). In this and much other regionalization studies Rugman and Verbeke’s (2004) original definition of the triad (NAFTA, EU, and Asia-Pacific) has been used but this is not without certain caveats, such as the geographic nature of this definition. The question of how to conceptually define a region is still unresolved as there are multiple ways to do this. Next to the geographic approach of Rugman and Verbeke, regions can also be defined by cultural or institutional similarities or by the World Value Survey (Peng& Pleggenkuhle-Miles, 2009). Furthermore, regions are not static and may change over time due to ongoing economic political and financial integrations, as well due to increased labor mobility. Then to make things even more challenging, there is also the fact that many MNCs adopt their own version of what constitutes a region, in terms of organizational structure as well as in managerial practices (Verbeke, Kano, & Yuan, 2016). The above implies that studying the internal working of regional MNCs requires a comprehension of their particularities, as to what constitutes a region for them.
The significant positive findings in all regression analyses for the second DOI dimension used in this thesis, regional scope, is also consistent with prior research (Qian, Khoury, Peng,
& Qian, 2010;Asmussen, 2009;Qian et al., 2008). The best explanation for this result may be
38
Additionally, the importance of economies of scope obtained at a regional level may also have an important positive effect on firm performance. However, as discussed above, the definition of a region used in internationalization studies can make a significant difference to the results, for all the dimensions of internationalization analyzed.
The hypothesized positive moderating effect of marketing intensity on the two dimensions of DOI on MNC financial performance is not supported in this study as the different conducted tests lead to different results. Although the moderating effect of this variable is only sporadically tested in internationalization research, the results of this study are contradictory to prior research (Kotabe, Srinivasan, & Aulakh, 2002). Where prior research
shows a positive moderating effect which was also hypothesized in this study, almost about
