Multinational Enterprises and the Extent of their
Cross-border Integration: The Relationship Between
International Strategies and the Geographic Scale and
Scope of Internationalization
Faculty of Economics and Business
MSc. Business Studies – International Management
Master’s Thesis
By: Lysbet Dekker Student Number: 5941229 Date: 27-06-2014 First Supervisor: Dr. Niccolò Pisani Second Supervisor: Dr. Carsten Gelhard
2
Abstract
Multinational enterprises (MNEs) need to decide on international strategies for expanding global business activities. Different international strategies can have various effects on
MNEs’ scale and scope of internationalization and require extensive analysis. While it is hard to adopt multiple international strategies at the same time, MNEs are expected to have their main focus on a particular international strategy. Three widely adopted international strategies analyzed in this research are aggregation, adaptation, and arbitrage as presented in the AAA Triangle framework (Ghemawat, 2007). In order to assess what influence each of these strategies has on the scale and scope of internationalization, this research tested the relationship between a chosen international strategy and MNEs’ scale and scope of internationalization. While analyzing the multinational organizations ranked on the 2013 Global Fortune 500 list, it is found that adaptation is positively related to the scale of internationalization and all three strategies are positively related to the scope of
internationalization. Results have shown that all the relationships are linear. The moderation effect of tangible slack resources is tested and it turned out that this variable negatively moderates the relation between both the strategies of adaptation and arbitrage and the scale of internationalization, as well as the relation between arbitrage and the scope of
internationalization. This current study has proven its academic relevance by complementing extant research on AAA strategies (Ghemawat, 2007) and by testing their relationship with MNEs’ cross-border activities. The findings of this study provide managerial implications that are useful for deciding on internationalization patterns and related international strategies of multinational organizations.
Keywords: International Strategies; AAA Triangle; Multinational Organizations; Scale and Scope of Internationalization; Tangible Slack Resources
3
List of Tables
Table 1. Descriptive values independent variables……….………. 21
Table 2. Correlation matrix………... 25
Table 3. Regression results (Scale Total)………. 28
4 Table of Contents 1. Introduction ………...………... 5 2. Literature Review……….….….….….….… 7 2.1 Aggregation………..……….. 9 2.2 Adaptation………..………… 12 2.3 Arbitrage………..………….. 13 3. Theoretical Framework………. 15 4. Research Methods……….. 19 4.1 Variables………..……….. 19 4.1.1 Dependent variables………..………. 19 4.1.2 Independent variables………...……….. 20
4.1.3 Tangible slack resources……… 22
4.1.4 Control variables……… 22
4.2 Statistical analysis……….. 23
4.3 Results ………..………..………... 26
5. Discussion……….………... 32
5.1 Managerial and academic implications………... 32
5.2 Limitations and suggestions for future research……… 35
6. Conclusion……… 37
7. Acknowledgement………... 39
5
1. Introduction
Internationalization is a widely researched concept within the academic field of international business. For companies that are willing to start operating activities in cross-border countries it is necessary to undertake the right strategy. Academic research has focused on different strategies of internationalization and how the decision-making-process of multinational organizations proceeds. In order to start internationalizing, a company needs to consider first the desired locations where it would like to expand to. Deciding on the most optimal location for international operations depends on many characteristics of the firm and its business environment. Since companies are increasingly characterized by their regional
internationalization patterns, recent studies have elaborated on this subject and have introduced frameworks on global business strategies.
One important framework that conceptualizes different strategies to be adopted by internationalizing organizations is the Integration and Responsiveness (IR) framework. The theory behind this framework is defined by Prahalad and Doz (1984) in their research on managerial decision-making on international strategies. Their work highlights the inherent tension for managers in choosing between responsiveness and integration. Managers need to deal with the strategic pressures on the firm caused by the need for one of those strategies. Prahalad and Doz (1984) describe integration as the development of a network of subsidiaries in which R&D, manufacturing and distribution tasks are centrally allocated and coordinated. Pressures for integration depend on the economic, technological, and competitive conditions of the firm’s activities. Key within the concept of integration is the procurement of substantial parts of the product line from external subsidiaries instead of manufacturing them internally. Internally developed products at their turn are supplied to other subsidiaries in foreign regions (Prahalad and Doz, 1984). The study of Prahalal and Doz (1984) indicates that in the end there is need for seeking the right balance between the degree of integration and local responsiveness. Each decision requires such an extensive analysis. Changing environmental conditions might force an organization to reconsider a chosen balance between
responsiveness and international integration.
Building on this IR framework and further investigating the practice of global strategy formulation, Ghemawat (2007) introduces a framework that is helpful for thinking through the different options related to international strategy. The AAA Triangle framework (Ghemawat, 2007) explains three types of global strategies that can be adopted, known as
6
aggregation, adaptation, and arbitrage. This framework does not only cover strategies of standardization or local responsiveness. It adds on the IR framework the strategy of arbitrage as being an additional response to the challenges of cross-border integration. The three-dimensional representation provides a useful perspective on the tensions of international strategy (Ghemawat, 2008). It is suggested that global firms should focus on one or two of the AAA Triangle strategies and need to make sure new elements of a strategy do fit well enough within the organization (Ghemawat, 2007).
Internationalization has a consequence for the scale and scope of an organization. An increase in scale and scope of internationalization is considered as a highly relevant benefit of foreign expansion since it becomes the starting point for gaining specific economies of scale and scope (Hennart, 2007). Being aware of the outcome of internationalization is therefore important for the strategic guidance of a firm.
This current study is concerned with how the focus on a particular international strategy of the AAA framework (Ghemawat, 2007) affects the scale and scope of internationalization of MNEs with cross-border activities. It is tested whether such a relationship between a particular international strategy and the scale and scope of
internationalization exists. Several hypotheses are developed which hypothesize whether there is a linear or curvilinear relationship between the focus on either aggregation, adaptation or arbitrage and the augment of MNEs’ scale and scope of internationalization. Aggregation as a main international strategy is expected to lead to an augment of both the scale and scope of internationalization at an increasing rate. Further assumptions that lead to the subsequent hypotheses are related to adaptation and arbitrage. For the second hypothesis it is expected that MNEs with a focus on adaptation as an international strategy will have their scale and scope of internationalization augmented at a decreasing rate. Third, it is hypothesized that MNEs that focus on arbitrage as an international strategy will have their scale and scope of internationalization augmented at a regular rate. The relationships formulated above might be influenced by MNEs’ surplus of required resources for internationalization. A fourth
hypothesis is formulated within this study and it is tested whether there is a positively moderating effect of tangible slack resources on the relationship between the focus on an international strategy and the scale and scope of internationalization.
The following section contains the main literature that is engaged with aggregation, adaptation or arbitrage as an international strategy. Each strategic option will be defined and discussed separately. Relevant literature will be put forward that includes an analysis of either
7
aggregation or global integration, adaptation or local responsiveness, or the logic of arbitrage. Subsequently, the factors that are expected to influence the scale and scope of
internationalization are identified and hypotheses are developed. A methodology section will follow that provides more insight into the variables, the process of the data collection and the methods used to conduct this research. Furthermore, the statistical analysis that is carried out will be discussed and findings of this test will be presented and analyzed. A final discussion on the results combined with limitations and future research opportunities will conclude this study.
2. Literature Review
An extensive amount of research is conducted on the multinational organization and its characteristics. In order to better understand international business and the role of
multinational organizations within global industries, academic research has analyzed MNEs and their internationalization behavior. Based on this behavior, typologies of multinational organizations have been conceptualized and used in previous studies. As operating on a global level has become important, MNEs need to evaluate their international strategy and the
pattern they want to follow when turning their business globally. Previous literature on international management tested different typologies of multinational organizations based on a variety of characteristics categorized within the headings of environment, strategy, structure, systems and processes. The most remarkable and widely accepted typologies derived from the study of Bartlett and Ghoshal (1989), which has classified MNEs into Global-,
Multidomestic-, International- and Transnational firms. The strategy of a Global
multinational organization is defined as building a cost advantage by realizing economies of scale (Harzing, 2000; Bartlett and Ghoshal, 1989). In terms of the IR framework this type of MNE has adopted an international strategy that focuses on high integration and low local responsiveness (Harzing, 2000; Bartlett and Ghoshal, 1989). The International multinational organization is characterized by exploiting parent company knowledge and capabilities through its worldwide diffusion and adaptation (Bartlett and Ghoshal, 1989). The Multidomestic multinational organization combines low integration with a high local
responsiveness strategy and the Transnational typology is characterized by combining a high integration strategy with high local responsiveness (Harzing, 2000; Bartlett and Ghoshal,
8
1989). These typologies are based on the strategic options derived from the IR framework and can help classifying and comparing MNEs and their subsidiaries.
Based on the initial concepts of global integration and local responsiveness, the AAA Triangle introduced by Ghemawat (2007) consists of three main strategies that can be adopted by global MNEs when trying to achieve competitive advantage. As indicated earlier these three basic strategies are aggregation, adaptation and arbitrage. Each of these strategies will be elaborated upon within this section. When a firm wants to compete on a global level, it needs to analyze the optional strategies and decide on what strategy to adopt in order to gain competitive advantage. Based on a predominant type of business expense, the AAA Triangle framework suggests one of its ‘A’ strategies to be adopted (Ghemawat, 2007). In his paper, Ghemawat (2008) illustrates that it would be hardly realistic to opt for each of the three strategies. Although all three strategies have to be kept in view, their strategic implications make it complicated to follow all of them. The tensions between the AAA Triangle strategies (Ghemawat, 2007) lead to an existing trade-off between those functions. Gaining insight into the differences between arbitrage, adaptation and aggregation remains inevitable for an organization to understand this kind of trade-off (Ghemawat, 2008). There is not necessarily one best global strategy at each point in time and a company needs to be prepared to change its strategy or combine several options whenever its business evolves (Ghemawat, 2007). Although a combination of all three strategies can be considered, it would hardly be possible. Such an option is highly complex and firms must not underestimate this combined
competitive strategy (Ghemawat, 2007). Companies tend to focus on either one or at most two of the alternative strategies. Comparable to seeking the right balance between integrating business operations and responding to local conditions, this framework shows a trade-off between aggregation, adaptation and arbitrage.
Recent studies use the AAA Triangle framework (Ghemawat, 2007) for analyzing businesses that leverage differences in institutional environments or organizations which are exposed to organizational uncertainty caused by today’s highly complex and changing environments (Surroca, Tribó and Zahra, 2013; Johnson, Arya and Mirchandani, 2013). It is suggested that this framework is useful for deciding on global strategies when differential comparative advantages between countries can be identified. In addition, insights in optimal global strategies from a stakeholder perspective and institutional theory point of view are presented.
9
2.1 Aggregation
The aggregation strategy is seen as the best strategy to adopt when a MNE starts to enter foreign markets. It is primarily to gain advantages by economies of scope and scale based on similarities and complementarities (Mauri and Figueiredo, 2012). According to Ghemawat (2007), adopting a global aggregation strategy would afford the most leverage for companies that practice R&D intensive operations with fixed R&D costs in order to improve economies of scale. Differences in cultural, political, geographical and economic factors between countries determine whether focusing on an aggregation strategy will be beneficial (Ghemawat, 2001). When significant economies of scale cannot be achieved, this might influence the production of sufficient levels of concentration (Zhu, Lynch and Jin, 2011). Strategists that encourage aggregation propose that companies should organize their operating units along regional lines, business lines or combinations of those two rather than just adapt their business in each host-country.
In his study, Luo (2002) precisely focuses on aggregation and global integration strategies. This study states that the optimal level of overall integration in a dynamic foreign environment is affected by several pressures and dynamics. Environmental and industrial pressures as well as organizational dynamics like internally differentiated dynamic corporate capabilities, organization infrastructure and strategic needs influence the level of integration necessary for companies to gain success. The study adds on previous research by making use of the resource-based view that allows to explain why multinational corporations operating in the same environment might use different strategies because of idiosyncratic resources and capabilities (Luo, 2002). When analyzing MNEs and their scale and scope of
internationalization, the resources and capabilities available for that organization might play a role since it will determine whether a certain strategy is preferred. The notion on using the resource-based view is critical since cross-border integration is shaped by existing resources and the need for developing new resources of multinational organizations. Results show that resource distinctiveness (especially knowledge proprietary), an established infrastructure for information flows, factor exploitation and good systems for coordination among the various units of a multinational organization favorably influence overall integration. Managerial ties with either supporting business parties or government officials are described as being negatively related to integration. Finally, Luo (2002) examined the moderating effect of strategic needs on the relationship between strategic infrastructure and overall integration. It
10
turned out that the factor of foreign experience, which in the first place had no significant relationship with overall integration in general, negatively influences integration for
multinational organizations seeking high risk diversification. When a subsidiary is managed to focus on local market expansion as a strategic need, its foreign experience turns out to be negatively influencing the overall integration (positively influencing overall responsiveness).
The strategic option of aggregation has been investigated on a regional level by
Arregle, Beamish and Hébert (2007). Their study focuses on MNEs’ localization decisions on foreign subsidiaries and how previous market entry and exit decisions influence subsequent investments. Aggregation investments are seen as actions by firms to exploit similarities and advantages across countries (Arregle et al., 2007; Ghemawat, 2003). The particular study of Arregle et al. (2007) describes aggregation as geographically concentrating subsidiaries in a specific area instead of dispersing them across areas. This logic of aggregation is useful to understand how prior market entry decisions influence subsequent entries and therefore the scale and scope of an organization. MNEs that choose for a certain geographic area and aggregation as an international strategy might do so in order to minimize liability of
foreignness (Arregle et al. 2007). This indicates that there is already a relationship between an adopted aggregation strategy and the geographic areas decided upon by MNEs to locate subsidiaries.
More recent studies on aggregation strategies integrated in internationalization patterns focus on their relationship with specific global areas, industries, performance and profitability. A recent study by Mauri and Figueiredo (2012) investigates the relationship between strategic patterns and the performance variability of MNEs. Their study finds evidence for a positive relationship between global integration and higher performance
variability. When it comes to location decisions they state that the pursuit of standardization is often motivated by a search for efficiencies resulting in patterns of geographic concentration and interdependencies. Additionally the study conducted by Johnson et al. (2013) provides support for the structural force of economies of scale that encourages a global integration strategy. Although this study is analyzing small- and medium sized multinationals, results might turn out to be applicable to larger multinational organizations as well.
Another economic pressure that supports the adoption of a global integration strategy is the standardized market demand (Johnson et al., 2013). It is indicated that this factor is specifically pervasive within the IT industry and that it would not directly be applicable to other industries. A study replicated and referred to by Johnson et al. (2013) predicts that
11
standardized market demand on the contrary has no positive relationship with business unit integration of large Western MNEs (Birkinshaw, Hulland and Morrison, 1995). These
findings suggest that differences in size of organizations should be carefully considered when relationships between international strategies and the scale and scope of internationalization are tested.
Gaining comparative advantage is as well seen as an important driver for global integration. An international subsidiary is expected to rely more on parent sources whenever pre-existing intangible or tangible resources can be leveraged from the home-country
(Johnson et al., 2013). Moreover, a high level of global competitive actions within a particular industry is negatively related to global integration of business activities. Global competitive actions have been examined in previous studies and are described as being a force that results in a shared strategy of group members overwhelming the strategy of an individual business (Birkinshaw et al. 1995; Johnson et al., 2013). This phenomenon puts great pressure on individual organizations and forces them to develop and adopt strategies that are in line with group norms. The study of Birkinshaw et al. (1995) examines large MNEs and the results indicate that global competitive actions within an industry are positively related with the global integration of business activities. Competitors that decide to globalize create increased perceived pressure for businesses to integrate operations. While this study has its focus on MNEs instead of small- and medium multinationals, the results might not be valid for the actual international business economy since they originate from a research conducted almost ten years ago.
Although the aggregation strategy is favored by strategists that do not encourage firms to adapt their business to each single country since it is not effective, in some industries aggregation is less easily accepted than in other industries. Breunig, Kvalshaugen and Hydle (2013) examine the opportunities for global integration in International Professional Service Firms (IPSF’s). Their study puts forward that industries that provide professional services rely heavily on knowledgeable individual local experts which turns their core business difficult to standardize (Breunig et al., 2013). Services are often coproduced with local clients and local responsiveness is required by local trends, laws and regulation. Thus, this study should take into account the differences in industries and operating regions of MNEs since it might influence their preferences for adopting an aggregation strategy.
12
2.2 Adaptation
Adaptation is seen as a global strategy that seeks to boost revenues and market share by maximizing a firm’s local relevance and is primarily used by companies that start expanding beyond their home markets (Ghemawat, 2007). This strategy reflects the action of adapting standard business activities, output and organizational structures to local environmental- and economic conditions and is therefore related to the local responsiveness part of the IR
framework (Prahalad and Doz, 1984). Previous literature on adaptation strategies and actions of local responsiveness includes the study of Luo (2001) on the determinants of local
responsiveness for foreign subsidiaries in emerging markets. This research examines the major factors that affect the degree of local responsiveness of foreign subsidiaries depending on the contextual contingencies and organizational dynamics of MNEs (Luo, 2001). The resulting determinants are classified into national environmental factors within the host country, industrial structural factors, and organizational factors. Depending on the degree of control given by a parent company, the decision of this strategic local responsiveness focus has to be made by either the subsidiary itself or the multinational organization (Luo, 2001). The results of his study show that environmental complexity and commercial practice specificity drive up local responsiveness, while cultural distance between home- and host-country is negatively related to the need for local responsiveness (Luo, 2001). Competitive intensity, market demand heterogeneity and component localization are forces that amplify local responsiveness. The market orientation of a subsidiary turns out to predict local responsiveness and local market orientated subsidiaries are stronger effected by
environmental complexity and structural factors for local responsiveness than subsidiaries with export oriented businesses. Finally his study shows support for the positive relationship between strong established managerial ties with local businesses and governmental
institutions and facilitated responsiveness (Luo, 2001). Necessary to point out is that his research has been conducted solely on multinational organizations situated in the dynamic environment of China. Since it is stated that determining factors of responsiveness and integration are not necessarily the same in different contexts, further research should simultaneously incorporate various factors that can drive MNEs to adaptation and local responsiveness (Luo, 2001).
The research of Martinez-Noya and Garcia-Canal (2011) gives some interesting insights into the relationship between international strategy, institutional context,
13
technological capabilities and the influence on a firm’s decision to outsource R&D activities and offshore R&D services. Their results confirm that when a firm is following an
international strategy based on achieving local responsiveness, it will have a higher
propensity to outsource offshore R&D activities and services than firms that do not have this focused strategy (Martinez-Noya and Garcia-Canal, 2011).
Other recent studies indicate that an adaptation strategy could be beneficial for MNEs when subsidiaries can utilize learning and are able to absorb and repatriate knowledge to their parent MNEs in order to support organizational renewal and corporate transformation needed for global adaptation (Koza, Tallman and Ataay, 2011). While the focus on adaptation as an international strategy started with the need to adapt to differences in demand and host-country environments forced by limitations in communication- and information technologies (Koza et al., 2011), today’s global adaptation strategy does not necessarily have to be forced by these limitations any longer. Technological advancement, similarities in demand and availability of resources and capabilities has made global homogeneous strategies more common (Koza et al., 2011). That does not mean that local responsiveness by MNEs is no longer a source of competitive advantage and success. The ability to develop variations in adapting to unique local conditions can be favorable to MNEs with high breadth of cross-border geographic market presence since they will get exposed to various unique conditions and environments to adapt to (Mauri and Figueiredo, 2012). A firm should not underestimate the costs and
complexity of adapting to a host-country environment. An adaptation strategy would result in increased complexity of coordination and control and decreased scale economies caused by the lack of leveraging standardized organizational structures, processes and output. Highly embedded organizational processes and identities can make it hard for firms to adapt to local conditions (Koza et al., 2011). Additional costs occur when there is need for change within several areas of the organization, for instance forced adjustment of information- and
communication systems and the need for a retrained labor force. Such reasons might withhold MNEs from opting for adaptation as an international strategy.
2.3 Arbitrage
Differences between countries and markets can also be considered by international
organizations as a source of value creation. The strategic option of arbitrage is often adopted by firms that seek opportunities to exploit differences in countries and manage to locate separate parts of the supply chain in different places in order to gain benefits (Ghemawat,
14
2007). When a firm begins its foreign expansion, opportunities arise to exploit the flexibility in its international operations. This multinationality is described by Kogut (1984) as providing a firm with embedded options to respond to profit opportunities. Firms with multinational operations learn how to run their global activities and such experiences make it easier for them to engage in subsequent multinational activities (Kogut, 2002). Arbitrage as a resulting function became known as profiting from differences in costs and exploiting price
differentials (Kogut, 2002; Ghemawat, 2003). This traditional type of arbitrage can be added upon by many more forms of arbitrage that offer sustainable sources of competitive
advantage. Arbitrage has a scope as wide as the differences among countries and significant exploitation opportunities can be grounded within administrative, cultural, geographic and economic aspects (Ghemawat, 2003). Administrative arbitrage opportunities arise from the legal, institutional, and political differences among countries. Tax benefits and the possibility to produce products against less restricting rules and legislations are examples of arbitrage opportunities of administrative nature (Ghemawat, 2003). Benefits from exploiting
differences in culture among countries are part of cultural arbitrage. Culture related music, fast-food concepts, art and haute couture are subject of cultural arbitrage and is exploited by selling these products on a global scale. Even though not all countries have to face a decrease in cultural arbitrage, some other countries and product categories are only temporary
providing cultural arbitrage opportunities (Ghemawat, 2003). Geographic arbitrage is a form of arbitrage that benefits from the actual distance between countries. Organizations that gain advantage from this type of arbitrage are companies that manage supply chains and provide logistic services. The fourth type of arbitrage is known as economic arbitrage. Sources for economic arbitrage are differences in capital and cost of capital, tax differences, differences in knowledge and skilled human capital, and differences in labor costs (Ghemawat, 2003;
Ghemawat, 2008). These differences in labor costs can create opportunities for labor
arbitrage. Labor arbitrage is a well-known type of arbitrage that focuses on the cost benefits gained by labor intensive organizations when outsourcing critical parts of the production process to host-countries with low labor costs (Ghemawat, 2008). A large part of the academic literature on arbitrage focuses on exploiting cost differences and labor costs in particular.
Drivers of the international strategy of arbitrage might be related to costs savings and purchasing power of multinational organizations. Large MNEs involved with internal trade within their own network by buying and selling across borders gain great benefits from
15
arbitrage opportunities. Producing in countries with low production costs and realizing profits within countries that offer the lowest tax rates leads to cost savings as well in terms of lower total global costs (Kogut, 2002). Additionally a shift in exchange rates might support MNEs to adopt arbitrage strategies and reduce costs. Multinational organizations with facilities owned abroad can decide to move production operations towards locations that offer the least expensive operating costs as a result of the shift in exchange rates (Kogut, 2002). A more recent study on drivers and strategies of international new ventures indicates that arbitrage opportunities related to higher purchasing power as a region-specific driver contributes to early internationalization of companies in their start-up phase (Nowinski and Rialp, 2013). In terms of internationalization patterns this result implies that the scale and scope of
internationalization can be influenced by the degree of arbitrage opportunities available in certain host-countries.
3. Theoretical Framework
Most of the research conducted within the field of internationalization has its focus on one of the AAA Triangle strategies (Ghemawat, 2007) adopted by MNEs. Studies that include information on all three strategic options within the same study are more descriptive in nature and discuss the managerial implications of adopting one strategy or combining several
strategies at a certain phase of the organizational evolution. The discussed studies on
aggregation are either focusing on the different determinants that drive aggregation strategies or on its relationship with performance. The literature found on AAA Triangle strategies (Ghemawat, 2007) and their meaning within global business has mainly been limited to MNEs operating in a specific industry or a specific country. Other interesting research is conducted on small- and medium sized multinationals rather than large MNEs. The research by Arregle et al. (2007) addresses both the logic of aggregation and arbitrage and focuses on whether previous market entry and exit decisions at a regional level influence the localization decisions of firms. Their study investigates certain localization patterns of MNEs’ foreign investments, but it does not address the adaptation strategy neither examines the influence of different global strategies on the location pattern of subsidiaries. Thus, the analysis of the literature suggests the need for additional research on international strategies and the related internationalization patterns of MNEs. No research has explicitly integrated all three
16
internationalization patterns. In order to fill this gap in the literature, this study examines how different international strategies precisely affect the scale and scope of MNEs’ international activities. The research question addressed in this study is therefore:
How do different international strategies affect MNEs’ scale and scope of internationalization?
This study is particularly interested in the effect of a chosen international strategy from the AAA Triangle (Ghemawat, 2007) and will narrow its focus on aggregation, adaptation and arbitrage. For developing propositions, the type of strategy becomes the independent variable that can have influence on the scale and scope of MNEs.
As explained in early studies, economies of scale are closely linked to the advocacy of standardized products and global rationalization (Kogut, 1984; Johnson et al. 2013). Many studies on international strategies indicate that the particular strategy of aggregation or global integration entails standardization of products, centralization of technological development or vertical or horizontal integration of manufacturing (Birkinshaw, 1995; Johnson et al., 2013). It is expected that the scale of MNE activities needs to be of significant magnitude in order to successfully obtain advantages from an aggregation strategy. Since cross-border activities involve higher costs for multinational coordination, a considerable scale needs to be achieved in order to gain profitable benefits from internationalization. A focus on aggregation as an international strategy is therefore associated with an increase in the scale of
internationalization. Starting economic operations in multiple countries intends a higher scope of internationalization as well. Accordingly:
Hypothesis 1a: Ceteris paribus, as the focus on aggregation strategy is stronger, the scale of internationalization augments at an increasing rate.
Hypothesis 1b: Ceteris paribus, as the focus on aggregation strategy is stronger, the scope of internationalization augments at an increasing rate.
When MNEs adopt a strategy that is aiming for local responsiveness and adaptation, an opposite effect is expected on the outcome of the scale and scope of MNEs’
internationalization. As the research by Harzing (2000) includes a cluster analysis on three MNE typologies and four variables to measure strategy, it turned out that the organizations
17
labeled as Multidomestic showed high scores on national responsiveness and domestic competition, combined with the lowest scores on global competition and economies of scale. Since the Multidomestic typology of MNEs is characterized as being locally responsive by differentiating business operations in order to become profitable and recognized in host-country environments, its strategy is comparable to the adaptation strategy. The degree of global competition of multinational organizations is in relation with the scope of
internationalization. An increase in global presence by entering new countries goes along with an increase in the amount of global competitors that need to be dealt with. Low scores on economies of scale can obviously be the result of a lower strategic focus on increasing scale. While previous research confirms the relationship between an international adaptation strategy of Multidomestic organizations and their low scores on global competition and
economies of scale, this study expects a decreasing pace of the augment of the scale and scope of internationalization when MNEs are executing an adaptation strategy. Thus, we posit: Hypothesis 2a: Ceteris paribus, as the focus on adaptation strategy is stronger, the scale of internationalization augments at a decreasing rate.
Hypothesis 2b: Ceteris paribus, as the focus on adaptation strategy is stronger, the scope of internationalization augments at a decreasing rate.
When looking at the arbitrage strategy of internationalization, extant research confirms that MNEs opt for this strategy when they are aware of certain exploitation opportunities offered by host-countries. Arbitrage as an international strategy can lead to great competitive
advantages when country differences can be exploited (Ghemawat, 2003). Therefore this strategy requires specific country analysis to identify possible exploitation opportunities based on low production costs, low tax rates, better developed knowledge or favorable regulations to increase profitability. Organizations willing to benefit from the differences between countries need to look for specific areas that offer such advantages. As some regions offer better pull effects for MNEs to attract cross-border activities (Kogut and Chang, 1991), MNEs might prefer one country above another as becoming their host-country. At the same time a specific arbitrage opportunity might be temporary and decreasing over time caused by the narrowing of differences between countries (Ghemawat, 2003). Labor arbitrage that is applied to countries with low salaries is sometimes argued to be unsustainable since labor costs may rise until they meet a level that will not any longer lead to cost savings. From this last point of
18
view it can be expected that companies continue to exploit low labor costs in a country that offers this arbitrage opportunity, as long as it stays profitable. When sources of arbitrage are held constant and are expected to be stable for a longer period of time, MNEs are relatively less pulled by other regions and will less likely consider to expand to other countries.
Companies do not have any incentive to move to other countries if their current host-country still offers the optimal exploitation circumstances. Therefore it is expected that, all other factors held equal, MNEs will augment their scale and scope of internationalization at a regular pace when they adopt an arbitrage strategy for internationalization. Only when current exploitation activities lose value and other regions offer better arbitrage opportunities, a MNE might consider to increase its scale and scope of internationalization. Therefore, we
hypothesize that:
Hypothesis 3a: Ceteris Paribus, as the focus on arbitrage strategy is stronger, the scale of internationalization augments at a regular rate.
Hypothesis 3b: Ceteris paribus, as the focus on arbitrage strategy is stronger, the scope of internationalization augments at a regular rate.
In order to be able to expand and enter new markets, an organization needs to make sure its potential slack is of sufficient size. This potential slack refers to the resources a company owns in excess of the required resources and can help organizations to survive a strategic move (Chang and Rhee, 2011; Cyert & March, 1963; Bourgeois, 1981). Potential slack serves as a buffer and needs to be higher when an organization wants to successfully operate across borders (Chang and Rhee, 2011). A large tangible slack makes entering new foreign markets less risky and would make foreign investment strategies more feasible. Accordingly, we expect that tangible slack resources positively moderate the relationship between a chosen international strategy and the scale and scope of internationalization. The final hypothesis tested within this study includes tangible slack resources as a moderator:
Hypothesis 4: MNEs’ tangible slack resources positively moderate the relationship between a chosen international strategy and the scale and scope of internationalization.
19
4. Research Methods
This study contains a quantitative research based on econometric modeling and it answers the research question by using cross-sectional data gathered from the 500 multinational
organizations ranked on the 2013 Global Fortune 500 list. Bureau van Dijk’s Orbis database is used for the data collection since it contains comprehensive information on important financials, ratios, and strength indicators of companies worldwide. Relevant data gathered for this research contains figures from the fiscal year 2012 on (international) sales, (international) assets, expenses, leverage, number of employees and foreign subsidiaries. Necessary figures that are not included in the formats downloaded from the Orbis database are retrieved from the company’s annual report of 2012. The multinational organizations included in the sample operate in different industries and are classified according to the Standard Industrial
Classification (SIC) codes. A total sample of 287 multinational organizations turned out to be useful for further analysis after correcting the data for potential outliers and missing values. The outliers and missing values are excluded from the whole analysis by means of listwise deletion. Since this study is interested in the relationship between international strategies and the scale and scope of a multinational organization, an OLS regression analysis is executed to find results. For this analysis the statistical program SPSS is used.
4.1 Variables
4.1.1 Dependent variables
The dependent variables of this study include the scale and scope of internationalization of multinational organizations. The scale and scope of MNEs’ international operations are measured separately by two methods used in previous research (Oh, 2009; Jong, de and van Houten, 2014). The scale of MNEs’ international activities is measured according to a valid method demonstrated in the paper of Oh (2009) that includes the geographic dispersion of MNEs’ sales and assets. Sales and assets are perceived as good proxies for the performance and structure of international activities (Oh, 2009). The geographic dispersion of MNEs’ sales and assets is calculated by two separate metrics. The first metric is the ratio of foreign-to-total sales and the second metric is the ratio of foreign-to-total assets. Then, the calculation of geographic dispersion is operationalized as follows:
20
FSTS stands for the ratio of foreign-to-total sales and FATA is the ratio of foreign-to-total assets. Those ratios are divided by the maximum observed ratio on foreign-to-total sales (max{FSTS}) and the maximum observed ratio on foreign-to-total assets (max{FATA}) respectively.
The scope of MNEs’ international activities is measured by using the
multidimensional metrics of international diversification which has been validly used in previous empirical research to calculate geographical scope (Jong, de and van Houten, 2014; Chao and Kumar, 2010; Sanders and Carpenter, 1998). This international diversification can be calculated by using the following formula:
First the number of foreign equity affiliates (N) and the number of foreign countries (K) is required. Subsequently both measures of foreign equity affiliates (N) and foreign countries (K) are divided by the maximum observed number of affiliates (max{N}) and maximum observed number of countries (max{K}) respectively. Eventually the two ratios are summed up and the average is calculated which lead to the final figure representing the total scope of a company’s internationalization. These figures range between 0 and 1 and a higher range indicates a higher scope of internationalization. The above used formula makes sure that the counts of the variables are changed into ratios. Since the scale and scope of international activities will automatically increase when a multinational organization engages in additional cross-border activities, it is of interest what the magnitude of this increase will be like. This magnitude will be tested by verifying the coefficients that result from the analyses.
4.1.2 Independent variables
The independent variables that will be tested refer to the three international strategies of aggregation, adaptation and arbitrage. In order to test the relationship between a chosen international strategy and the scale and scope of the surveyed companies, the three strategies of the AAA Triangle are measured by different expenditure categories as a percentage of total sales (Ghemawat, 2007). The ratio of R&D expenses to total sales is a proxy for aggregation as a main strategy. The operationalization of the adaptation strategy is done using the ratio of advertising expenses over total sales. The strategy of arbitrage is measured by calculating labor expenses as a percentage of total sales. Standardized scores of the independent variables
21
are used to run the regressions. These z-scores are calculated by the figures on the mean and standard deviation shown in Table 1.
Table 1. Descriptive values independent variables
Minimum Maximum Mean
Standard Deviation 1) Aggregation_Raw 0,0001 0,2754 0,0321 0,0474 2) Adaptation_Raw 0,0000 0,3016 0,0379 0,0549 3) Arbitrage_Raw 0,0000 0,5597 0,1244 0,1220 Valid N=287
Not all of the companies from the 2013 Global Fortune 500 list publish figures on advertising-, R&D- and labor expenses. In order to exclude all the outliers and missing values, a listwise deletion is applied to get to the final sample. The final sample valid and useful to gather the descriptive statistics from counts 287 companies in total. The descriptive values of the independent variables show a maximum percentage of R&D expenses to total sales of 27.54%. The mean of the ratios on aggregation is low and counts for 0.0321. On average a company has spent 3.21% of its sales on R&D expenses. The relatively low standard deviation of the aggregation ratio indicates that companies do not vary in great extent when it comes to percentages spent on R&D to their total sales. The second group of ratios related to adaptation gives a higher value of its maximum. The company with the highest ratio allocates 30.16% of its total sales to advertising expenses. On average the population of MNEs spent 3.79% of its total sales on advertising. The third row of Table 1 shows the final ratios of the observed companies related to the percentage spent on labor expenses. The highest maximum score of 55.97% of total sales is spent on labor expenses. Although it cannot be asserted that all the observed companies are exclusively focusing on one international strategy, it is shown in Table 1 that there are companies spending a minimum of 0.00% of their total sales on either R&D-, advertising- or labor expenses.
22
4.1.3 Tangible slack resources
Tangible slack resources as a moderating factor within this study is measured by MNE financial leverage. This leverage variable is used decently in a previous study conducted by Chang and Rhee (2011) and is therefore considered as being reliable. The analyzed indicator used to operationalize this moderating factor is the gearing ratio that measures the degree to which a firm’s activities are funded by owners’ funds or creditor funds. In other words, it indicates what proportion of equity and debt the company uses to finance its assets. The lower the gearing ratio, the easier a company can acquire additional resources necessary to expand and the lower its likelihood to become bankrupt. The gearing ratio corresponds to the ratio of non-current liabilities and loans over shareholder funds. The outcome times 100% will lead to the final ratio.
4.1.4 Control variables
Since other factors next to the independent variables and moderating variables might
influence the scale and the scope of internationalization, control variables are included in the research. The three variables that are controlled for are the size of the firm, the age of the firm and the industry a company is in. The size of the firm and its relationship with the scale and scope of internationalization has been analyzed in previous studies that come up with evidence that larger firms are more likely to enter a foreign market in an early stage than small firms (Gaba, Pan and Ungson, 2002). Larger firms often have more resources to invest abroad as well as that they can more easily compete with economies of scale and scope (Gaba et al., 2002; Cohen, 1996; Chandler, 1962). Given that the size of the firm influences the scale and scope of internationalization, this study controls for this variable by including the number of employees the firm accounted for during the fiscal year of 2012 as done in previous
research (Larsen, Manning and Pedersen, 2013).
The age of the company might also influence the scale and scope of
internationalization. The longer a company exists, the more chances it had to increase its scale and scope and to gain experience by starting operations abroad. As it is concluded in other studies on foreign direct investments, a company with previous experience in
internationalization and with an existing subsidiary in a foreign market is more likely to make subsequent investments into that foreign market (Benito and Gripsrud, 1991; Davidson, 1980). This eventually affects the scale of internationalization. Previous experience might
23
also affect the scope of international activities since it reduces the influence of cultural distance which would make companies invest in other countries more easily (Jong, de et al., 2014). Age of a firm is controlled for by subtracting the year of incorporation from the sample year 2012.
A final control variable added to this research is the type of industry a company is in. The variation between industries in terms of profitability might lead to differences in the scale and scope of internationalization (Jong, de et al., 2014). An industry that offers more
opportunities to increase profit can make companies invest abroad more easily. Next to that, industries differ in need for global activities. One industry might be more supportive towards cross-border activities and foreign investments than the other. An industry that has to deal with highly sensitive data might not want to have international activities in specific countries. All industries the observed companies are active in are divided into the three main categories of primary-, secondary- and tertiary industries. The group of primary industries consists of companies that deal with natural resources. The secondary industries are selected on the base of being active in production manufacturing. Service industries are assigned to the tertiary industries.
4.2 Statistical analysis
To test whether there is a significant relationship between the independent variables of international strategies and the dependent variables of the scale and scope of
internationalization, a regression analysis is applied. Since the two dependent variables have continuous values, the model of Ordinary Least Squares (OLS) seems appropriate. The first two hypotheses suggest a non-linear relationship that has a decreasing or increasing effect on the augment of the output. To check whether the relationship is rather non-linear than linear, a polynomial regression is run with different powers for the independent variables. Both linear and curvilinear regressions are analyzed. To find out the right shape of the curve, the second degree polynomial is checked by running a regression with quadratic terms of the
independent variables related to aggregation and adaptation. Before running the regressions, the reliability and multicollinearity of the used variables is checked. All independent variables that are part of the regression analyses are tested for multicollinearity by running single linear regressions with collinearity diagnostics. The VIF-scores of the output of these regressions are analyzed since they indicate whether a predictor has a strong linear relationship with other predictors. A definite significant correlation is indicated by a VIF-score that is higher than 10
24
(Field, 2013). After analyzing the diagnostics, no multicollinearity problems are found since all the values of the VIF-scores are below 10. Reliability of the data is checked by testing for correlations between variables. The correlation matrix indicates whether a correlation is found between the variables and if this correlation is significant. For this matrix a 99% confidence interval is used and a correlation is significant whenever its p-value is lower than 0.01. Table 2 captures descriptive data and correlation coefficients of all the used variables within this study. To see whether a significant correlation between variables becomes problematic, the squared correlation coefficient is analyzed. This squared correlation coefficient of
determination measures the amount of variability in one variable that is shared by the other (Field, 2013). A squared correlation coefficient greater than 0.5 (correlation coefficient greater than 0.7) means that the largest part of the variability of one variable is shared by the other. Such a case is problematic since it does not leave any share of the variability to be accounted for by other variables (Field, 2013). The correlation coefficients of this study do not encounter any values greater than 0.7 and therefore all variables can be retained.
To proceed with analyzing the data, two groups of regressions are run with either having the scale or the scope of internationalization as a dependent variable. First, a linear regression is performed with the control variables and the scale of internationalization as a dependent variable. This first model with control variables is used as a benchmark model for all subsequent analyses. The second group of regressions examines each standardized score of the independent variable separately by adding it to the first model with the scale of
internationalization as a dependent variable. The output of these tests are shown in Model 2, Model 3 and Model 4. Additionally, a curvilinear relationship is tested by adding the
quadratic termed scores of aggregation and adaptation to the previous regressions run with their standardized scores. No quadratic term of the independent variable of arbitrage is added to the model since hypothesis 3a and b do not indicate a curvilinear relationship. The output of these latter tests is analyzed in order to compare the linear regressions to the curvilinear regressions. The values of the R² and the R²-change are assessed to see whether the models of the linear regression and curvilinear regression can be used for interpretation. Output data of these regressions can be found in Model 5 and Model 6 included in Table 3. A final number of regressions is performed to test for a moderating effect of tangible slack resources on the relationship between the independent and the dependent variable. First, the mean-centered product terms are created between the gearing variable and each standardized independent
Table 2. Correlation matrix (N=287) 1 2 3 4 5 6 7 8 9 1) Scale_Total 1 2) Scope_Total 0.470** 1 3) Aggregation_Standardized 0.512** 0.402** 1 4) Adaptation_Standardized 0.518** 0.175 0.418** 1 5) Arbitrage_Standardized 0.197 0.356* 0.210 0.408** 1 6) Gearing -0.083 0.235 -0.051 0.069 0.097 1 7) Size -0.044 0.196 0.007 -0.107 0.072 0.052 1 8) Age 0.448** 0.350* 0.479** 0.331* 0.296 -0.057 -0.107 1 9) Industry -0.164 0.004 -0.104 -0.108 0.216 -0.090 0.122 -0.148 1 Mean 0.3704 0.0987 0.000 0.000 0.000 142,322 120430.91 63.260 2,270 Std. dev. 0.2891 0.1223 1.000 1.000 1.000 123.383 152277.71 51.587 0,674 Min. values 0.000 0.000 -0.675 -0.691 -0.969 0.000 189 3 1 Max. values 0.99 0.76 5.132 4.806 6.322 736.32 2200000 349 3 a N=287
** Correlation at 0.01 level (2-tailed)
variable. These mean-centered product terms are used separately to test for a significant interaction effect. Three regressions are run with the particular standardized independent variable, the quadratic termed independent variable in case of aggregation and adaptation, the centered gearing and the product termed interaction variable added to the first model. To see whether there is a moderating effect, the significance of the model with the product termed interaction is checked by its R-value and correlation coefficients. The output of these final regressions is shown in Model 7, Model 8 and Model 9 of Table 3. All of the regressions explained above are repeated for the second group that takes the scope of internationalization as a dependent variable. Output of these regressions on the scope of internationalization can be found in Table 4. The results of both groups of regressions are further explained in the following section.
4.3 Results
The results of the examined regression analyses described in previous section are presented in Table 3 and Table 4. All tested models show significance in their model fit and output figures can therefore be considered reliable. For describing the results of the first three hypotheses, Model 2 up to and including Model 6 of both Table 3 and 4 are useful. When looking at the R-square values of these models a difference in significance is noticed. The results of the first model are based on the linear regression analyses with the control variables and the scale and scope of internationalization. The first model of both Table 3 and 4 shows significant p-values and a reasonable R-square value of 0.033 and 0.093 respectively on the scale and scope of internationalization. More important are the figures shown in Model 2 up to and including Model 6 which include the standardized independent variables and two quadratic terms of aggregation and adaptation as an independent variable. While performing the regressions with the standardized independent variables, it turned out that adding the two quadratic terms of aggregation and adaptation in Model 5 and Model 6 results in a sufficiently increase in the square value. Model 5 of Table 3 comes up with a square value of 0.078 towards the R-square value of 0.073 related to Model 2. This change in R-R-square value of 0.005 indicates that adding the quadratic term of the aggregation variable to the model with the standardized aggregation variable increases the explaining character of the model by 0.5%. This increase in R-square value is found for the quadratic term of the adaptation variable as well, that rises from 0.158 to 0.160. Adding the quadratic termed independent variables of aggregation and adaptation to the models on scope also leads to an increase in R-square values. In Table 4 the
27
R-square values on aggregation show an increase of 0.008 from 0.147 to 0.155. Model 3 and Model 6 on adaptation come up with an increase in R-square value of 0.006 from 0.090 to 0.096. The quadratic termed independent variables of aggregation and adaptation fit the model of this research and a curvilinear relationship is therefore expected. In order to look whether the hypotheses are supported or rejected, the ß-value of each model and its
significance is analyzed and discussed.
Starting with the control variables, age of the firm has got a positive significant effect on the scale of internationalization in the first two models and Model 5. The age of a firm is stronger related with a firm’s scope of internationalization and is significantly related to this dependent variable in each model. Remarkable with the type of industry as a control variable is that it has a negative relationship with the scale of internationalization except for Model 9. This negative relationship is only significant in Model 4. The type of industry of a firm is significant and positively controlling for the relationship between each standardized score of aggregation, adaptation and arbitrage, and the scope of internationalization. It additionally shows a significant correlation in Model 6 with the quadratic termed score of adaptation and the scope of internationalization as well as for the moderating effect of the gearing on the relationship between arbitrage and the scope of internationalization. The size of the firm only shows a significant positive correlation with the scale of internationalization when
aggregation is used as a standardized-, quadratic-, and product termed independent variable. For the scope of internationalization, the firm size is again significant and controls positively for the models with aggregation as a standardized-, quadratic- and product termed
independent variable.
When looking at Model 2 of Table 3, no significant relationship is found between aggregation as a standardized independent variable and the scale of internationalization. Although the coefficient of Model 5 shows a negative score for the curvilinear relationship between aggregation and the scale of internationalization, this coefficient is not significant and there is no support for hypothesis 1a. Model 2 of Table 4 indicates that a positive and significant linear relationship between aggregation and the scope of internationalization is found with a ß-value of 0.168. For the curvilinear relationship between aggregation and the scope of internationalization a negative and non-significant coefficient is shown in Model 5. Therefore it cannot be concluded that the focus on aggregation as an international strategy leads to an augment of the scope of internationalization at an increasing rate. Hypothesis 1b is rejected as well.
Table 3. Regression results (Scale Total)
Dependent variable
ScaleTotal Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7 Model 8 Model 9
Control variables Firm Size 0.102** 0.185** 0.045 0.119 0.175** 0.047 0.165** 0.054 0.121 Firm Age 0.116** 0.164** 0.125 0.079 0.149** 0.117 0.134 0.113 0.036 Operating Industry -0.096 -0.025 -0.156 -0.137** -0.035 -0.153 -0.018 -0.103 0.028 Independent variables Aggregation_Standardized 0.085 0.190 0.207 Adaptation_Standardized 0.345** 0.423** 0.487** Arbitrage_Standardized 0.095 0.255** Aggregation_Quadratic -0.128 -0.049 Adaptation_Quadratic -0.093 -0.147 Moderator variable Gearing _centered -0.034 -0.076 -0.119 Aggregation_Standardized×Gearing 0.159 Adaptation_Standardized×Gearing -0.210** Arbitrage_Standardized×Gearing -0.224** Model fit N 374 208 133 237 208 133 191 107 176 R² 0.033 0.073 0.158 0.047 0.078 0.160 0.090 0.199 0.104 Adj. R² 0.025 0.055 0.131 0.031 0.056 0.127 0.056 0.142 0.072 F-stat 4,147 4,013 5,993 2,883 3,434 4,843 2,598 3,511 3,276 P-value 0.007** 0.004** 0.000** 0.023** 0.005** 0.000** 0.014** 0.002** 0.004**
Table 4. Regression results (Scope Total)
Dependent variable
Scope Total Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7 Model 8 Model 9
Control variables Firm Size 0.084 0.203** 0.056 0.081 0.191** 0.054 0.199** 0.051 0.128 Firm Age 0.241** 0.216** 0.205** 0.170** 0,197** 0.205** 0.224** 0.212** 0.166** Operating Industry 0.151** 0.125** 0.159** 0.149** 0.114 0.159** 0.082 0.097 0.169** Independent variables Aggregation_Standardized 0.168** 0.298** 0.374** Adaptation_Standardized 0.189** 0.065 0.136 Arbitrage_Standardized 0.239** 0.280** Aggregation_Quadratic -0.155 -0.212 Adaptation_Quadratic 0.148 0.142 Moderator variable Gearing _centered -0.018 -0.026 -0.064 Aggregation_Standardized×Gearing -0.007 Adaptation_Standardized×Gearing -0.014 Arbitrage_Standardized×Gearing -0.138** Model fit N 477 235 150 280 235 150 217 115 200 R² 0.093 0.147 0.090 0.145 0.155 0.096 0.197 0.121 0.176 Adj. R² 0.087 0.133 0.065 0.133 0.136 0.065 0.171 0.063 0.151 F-stat 16,131 9,949 3,570 11,657 8,376 3,067 7,345 2,095 6,890 P-value 0.000** 0.000** 0.008** 0.000** 0,000** 0.012** 0.000** 0.050 0.000**
Turning back to Table 3 and looking at Model 3 and Model 6 provides insight in the results on Hypothesis 2a. A significant linear correlation is found between adaptation as a
standardized independent variable and the scale of internationalization. This is represented in Model 3 by the ß-value of 0.345. Model 6 does not give a significant correlation between adaptation as a quadratic independent variable and the scale of internationalization. Although the related ß-value of -0.093 has a negative slope and indicates a curvilinear relation, the coefficient is not significant and hypothesis 2a is rejected. Model 3 of Table 4 gives a significant correlation coefficient of 0.189 between the standardized adaptation variable and the scope of internationalization. Again there is no significant coefficient found in Model 6 for the relationship between the quadratic independent variable of adaptation and the scope of internationalization. The related ß-value of 0.148 has no negative slope which indicates that the relationship is not curvilinear and no U-shaped curve can be found. Since this result is not in line with hypothesis 2b, this hypothesis is rejected as well.
For the third hypothesis there is no significant relationship found between arbitrage and the scale of internationalization. Model 4 of Table 3 shows a non-significant coefficient for the standardized arbitrage variable with a ß-value of 0.095. There is not enough support to conclude that when a MNE focuses on arbitrage as an international strategy its scale of internationalization will augment at a regular rate. Hypothesis 3a is therefore rejected. When looking at Table 4 there is a significant correlation between arbitrage and the scope of internationalization. The ß-value of 0.239 in Model 4 is significant and the positive slope indicates that a focus on arbitrage as an international strategy leads to an increase of the scope of internationalization. Since the ß-value of 0.239 is relatively high compared to the scores of the other independent variables, it is questionable whether the scope augments at a regular rate. While the correlation is not that strong with the coefficient still below 0.5, hypothesis 3b is accepted.
When continuing to the fourth hypothesis it is necessary to assess Model 7, Model 8 and Model 9 of both tables. These models show the results of a tested moderation effect of tangible slack resources on the relationship between the independent and dependent variables. When looking at the outcome of the tested interaction variables on scale as a dependent variable, two of the three interaction variables show a significant coefficient. The results indicate that whenever there is a significant relationship between adaptation or arbitrage as an international strategy and the scale of internationalization, the gearing of the company has a moderating effect on this relationship. Model 7 of Table 3 gives a non-significant positive
ß-31
value of 0.159 that explains that there is no effect of an increase in tangible slack resources on the relationship between aggregation as a main strategy and the scale of internationalization. This result is not in line with the fourth hypothesis. For adaptation and arbitrage there is a significant but negative correlation found in Model 8 and Model 9 with a ß-value of -0.210 and -0.224 respectively. When a company has its focus on either an adaptation or arbitrage strategy and it has a high level of tangible slack resources, this interaction effect will
negatively influence the scale of internationalization. The negative slope of these results are not in line with hypothesis 4. Model 8 furthermore shows a significant ß-value of 0.487 for adaptation as a main international strategy. A positive relationship is found for adaptation and the increase in the scale of internationalization when tangible slack resources are added as a moderating variable. Remarkable is that Model 9 shows a significant ß-value of 0.255 for arbitrage as a focus strategy and its relationship with the scale of internationalization when tangible slack resources are added as a moderating variable. Although no initial significant relationship is found in Model 4 between arbitrage as an independent variable and the scale of internationalization, a significant positive relationship between this independent variable and the scale of internationalization is found when the moderating variable is added to the model. Table 4 provides the results of the interaction effect of a company’s tangible slack resources on the relation between an international strategy and the scope of internationalization as a dependent variable. A negative moderating effect of a company’s tangible slack resources on the relationship between arbitrage as a focus strategy and the scope of internationalization is found and represented by the -0.138 ß-value in Model 9. Furthermore, it is found that adaptation as an independent variable has no significant relationship with the scope of
internationalization when the tangible slack resources are added to the model as a moderating variable. The ß-value of 0.136 is not significant while the initial relationship found in Model 3 between adaptation as an independent variable and the scope of internationalization is
significant. According to the significant results, hypothesis 4 is not supported when it concerns adaptation and arbitrage as an independent variable on the scale of
internationalization and arbitrage as an independent variable on the scope of
internationalization. Actually, an opposite negatively moderating effect of tangible slack resources is found when analyzing the significant ß-values.
