Repositorio Institucional de Documentos

2012 69 Elisabet Garrido Martínez Technological change and internationalization in network industries : An institutional approach Departamento Director/es Economía y Dirección de Empresas Fuentelsaz Lamata, Lucio Maicas López, Juan Pablo Director/es Tesis Doctoral Autor Repositorio de la Universidad de Zaragoza – Zaguan http://zaguan.unizar.es UNIVERSIDAD DE ZARAGOZA Departamento Director/es Elisabet Garrido Martínez TECHNOLOGICAL CHANGE AND INTERNATIONALIZATION IN NETWORK INDUSTRIES : AN INSTITUTIONAL APPROACH Director/es Economía y Dirección de Empresas Fuentelsaz Lamata, Lucio Maicas López, Juan Pablo Tesis Doctoral Autor 2012 Repositorio de la Universidad de Zaragoza – Zaguan http://zaguan.unizar.es UNIVERSIDAD DE ZARAGOZA Departamento Director/es Director/es Tesis Doctoral Autor Repositorio de la Universidad de Zaragoza – Zaguan http://zaguan.unizar.es UNIVERSIDAD DE ZARAGOZA   DOCTORAL THESIS T ECHNOLOGICAL C HANGE AND I NTERNATIONALIZATION IN N ETWORK I NDUSTRIES : A N I NSTITUTIONAL A PPROACH PhD. Candidate: ELISABET GARRIDO MARTINEZ Supervised by: Dr. LUCIO FUENTELSAZ LAMATA Dr. JUAN PABLO MAICAS LOPEZ      Index iii  INDEX OF CONTENTS CHAPTER 1. INTRODUCTION .................................................................................. 1 1.1. Introduction ................................................................................................ 3 1.2. Competition in network industries ........................................................... 6 1.2.1. Network effects: concept and typology ............................................ 6 1.2.2. The extent and importance of network industries in the worldwide economy ................................................................................... 9 1.2.3. Competing in network industries: Changing the focus from product value to network value ...................... 16 1.3. The institution-based view of strategy ................................................... 18 1.3.1. An overview of strategic management research ............................ 18 1.3.2. The institution-based view of strategic management ................... .20 1.4. Structure ................................................................................................... 22 References ........................................................................................................ 27 CHAPTER 2. RESEARCH SETTING: MOBILE COMMUNICATIONS INDUSTRY ........... 33 2.1. Introduction ............................................................................................. 35 2.2. Evolution of mobile communications industry ...................................... 38 2.2.1. Evolution of mobile communications industry in Europe ............ 38 2.2.2. Evolution of mobile communications industry in the U.S ............ 41 2.2.3. Evolution of mobile communications industry in other regions .......................................................................................... 44 Index iv  2.3. Network effects in mobile communications industry ............................ 47 2.4. Mobile communication players in network competition: From national operators to international groups ................................... 48 2.5. The change from fixed to mobile communications ............................... 57 2.6. Summary ................................................................................................... 60 References ........................................................................................................ 63 CHAPTER 3. STRATEGIC CHOICES, NETWORK EFFECTS AND PERFORMANCE: A STRATEGIC APPROACH TO NETWORK VALUE IN NETWORK INDUSTRIES ............ 67 3.1. Introduction ............................................................................................. 69 3.2. Theory and hypotheses ............................................................................ 73 3.2.1. Installed base, network effects, network value and network intensity ....................................................................................... 73 3.2.2. The antecedents of network value: Expectations, Coordination and Compatibility ........................................ 76 3.2.3. Strategic choices, network value and performance ....................... 80 First-mover advantages (FMA) and network value ............................ 80 Internationalization and network value .............................................. 81 Switching costs and network value...................................................... 84 Network value and performance.......................................................... 86 Index v  3.3. Data and variables .................................................................................... 89 3.3.1. Research setting: the European mobile communications industry ...................................... 89 3.3.2. Sample ............................................................................................... 92 3.3.3. Measurement of variables ................................................................ 93 3.3.4. Descriptive Statistics ...................................................................... 103 3.4. Methods .................................................................................................. 107 3.4.1. Network value model ..................................................................... 107 3.4.2. Profitability model ......................................................................... 107 3.4.3. Estimation procedure ..................................................................... 108 3.5. Results ..................................................................................................... 109 3.5.1. Strategic choices and network value ............................................. 109 3.5.2. Performance and network value ................................................... 113 3.6. Discussion and conclusions ................................................................... 115 References ...................................................................................................... 122 CHAPTER 4. STRATEGIC CHOICES AND INSTITUTIONS IN THE FDI PROCESS: WHERE TO ENTER ............................................................................................. 135 4.1. Introduction ........................................................................................... 137 4.2. Theory and hypotheses .......................................................................... 142 4.2.1. A key decision in the internationalization process: Where to go .............................................................................................. 142 4.2.2. Institutions: cultural distance and formal institutional development ....................... 144 Index vi  4.2.3. Institutions and host market selection............................................ 146 Cultural distance and host market selection ..................................... 146 Formal institutional development and host market selection.......... 148 Cultural distance, formal institutional development and host market selection ................................................................... 149 4.3. Data, variables and methodology .......................................................... 151 4.3.1. Data ................................................................................................. 151 4.3.2. Measurement of variables .............................................................. 156 4.3.3. Methodology .................................................................................. 159 4.3.4. Descriptive statistics ....................................................................... 160 4.4. Results ..................................................................................................... 163 4.5. Discussion and conclusions ................................................................... 168 References ...................................................................................................... 174 CHAPTER 5. INSTITUTIONS AND PERFORMANCE AFTER A RADICAL TECHNOLOGICAL CHANGE: HOW THE VALUE OF SPECIALIZED COMPLEMENTARY ASSETS VARIES ACROSS MARKETS .......................................... 183 5.1. Introduction ........................................................................................... 185 5.2. Theory and hypotheses .......................................................................... 189 5.2.1. Competitive advantage, incumbents and technological change ................................................................. 189 5.2.2. The institutional context ............................................................... 193 5.2.3. Incumbency, complementary assets and performance ................ 196 5.2.4. The moderator role of the institutional context .......................... 198   Chapter 1. INTRODUCTION Chapter 1. Introduction 3  1.1. INTRODUCTION trategic management has traditionally focused on the strategic choices that firms should make in order to gain a sustainable competitive advantage. In other words, firms seek to obtain a performance that exceeds the average of the industry (Porter, 1985). To explain the sources of sustainable competitive advantage, the management literature has mainly focused on external (industry-based, Bain, 1956, 1968; Mason, 1939) and internal (firm-based, Penrose, 1959; Barney, 1991) factors. The continuous change in the focus of analysis between these two approaches has been compared to a “pendulum swing” (Hoskisson, Hitt, Wan and Yiu, 1999). The latest swing, which took place in the first decade of the 21st century, again focused on two external factors, namely, network effects and institutions, as key determinants of firm behavior and performance. This has led to the strategic consideration of variables that have traditionally been either underestimated or considered as background conditions (McIntyre and Subramaniam, 2009; Meyer, Estrin, Bhaumik and Peng, 2009; Peng, Sun, Pinkham and Chen, 2009; Suarez, 2005). First, competition under network effects has its origin in technological developments which, since the 80’s, have led to an increasing worldwide presence of information industries, such as software, hardware, telecommunications and social networks. For instance, at the end of 2011, the number of tweets per day passed the 200 million mark, there were more than 500 million Facebook users and 5,300 million mobile users. These are only some examples of industries that have been S Chapter 1. Introduction  4  recently created and have rapidly increased their presence across the world. In this type of industries, the presence of network effects means that user utility depends not only on product characteristics, such as price and quality, but also on the number of other users consuming the same product (Katz and Shapiro, 1994). Users prefer to join the network with a higher number of users and this accelerates product penetration (Gruber, 2005). Shapiro and Varian (1998) determined that network effects do not change economic laws and, thus, traditional perspectives of strategic management can be applied in this context. But it is important to note that competition under network effects introduces several particularities in strategic management analysis. In contrast to traditional industries in which product characteristics - e.g. price and quality – are key variables to determine strategy and performance, in network industries, network characteristics – e.g. size and composition – are more important in conferring competitive advantages (Arthur, 1990; McIntyre and Subramanian, 2009; Suarez, 2005). The literature has highlighted that, in network industries, value does not reside in the product itself, but in the network of users. This means that traditional strategies based on quality differentiation and cost leadership (Porter, 1985) become less important. In network industries, strategic decisions should try to influence users’ expectations about the future size of the network to increase current network value. In this way, price and quality lose importance as key strategic variables, while reputation and brand value become more prominent (Katz and Shapiro, 1994). Surprisingly, the analysis of strategic management to increase network Chapter 1. Introduction 5  value and firm performance in network industries remains underdeveloped (McIntyre and Subramaniam, 2009). Secondly, the institutional environment in which firms operate has traditionally been considered as a background condition (Peng, Wang, and Jiang, 2008). Institutions, such as laws, traditions and culture, have been relegated to being control variables in empirical analysis instead of constituting key explanatory variables. In the mid 1950’s, the neoinstitutionalist perspective emerged in sociological studies to analyze the structure and behavior of organizations (Scott, 2008). Several decades later, North (1990) incorporated the role of institutions into the economic discourse from a transaction costs perspective. However, strategic management literature has recently taken into consideration the institutional perspective of firm strategy (Peng, Sun, Pinkham and Chen, 2009). From this perspective, institutions, along with industry and resource conditions, are understood as key determinants of strategic choices and firm performance. Under the institution-based perspective of strategic management, strategic choices are understood as the result of the dynamic interaction between organizations (conditioned by industry structure and resource allocation) and institutions (both formal and informal). Thus, the behavior and performance of an organization should be analyzed within the institutional framework where it operates (Peng et al., 2005; Peng, 2002) because strategies which are successful in one institutional context can fail in others (Hoskisson et al., 1999). Surprisingly, the use of the institution-based view of strategy remains underexplored (Peng et al., 2009). Chapter 1. Introduction  6  Both network effects and institutions have traditionally been considered as background conditions and included as control variables in empirical analyses (McIntyre and Subramaniam, 2009; Peng et al., 2009). Consequently, there is still a great deal left to do in the strategic analysis of these two key elements. Not taking into account the importance of managing network value, instead of product value, and the direct impact of institutions on the outcomes of strategic choices can result in mistaken decisions which prevent firms from achieving a sustainable competitive advantage and expel them from the market. This dissertation aims to analyze the key role of network effects and institutions in strategic management analysis in depth. The following sections try to contextualize and develop the basic concepts of both research lines. Section 1.2 introduces the main concepts of competition in network industries, while Section 1.3 offers a review of the institution-based view of firm strategy. Section 1.4 describes the structure, content and contributions of the other chapters of this dissertation. 1.2. COMPETITION IN NETWORK INDUSTRIES 1.2.1. Network effects: concept and typology Network industries are those which exhibit network effects. Network effects arise when user utility from consuming a product increases with the number of other users consuming the same good or service (Katz and Shapiro, 1994; Farrell and Klemperer, 2007). Chapter 1. Introduction 7  The traditional typology of network effects differentiates between direct and indirect network effects (Birke and Swann, 2006; Farrell and Klemperer, 2007; Katz and Shapiro, 1985, 1994; Suarez, 2005). Direct network effects appear when the increase in the number of users of a network directly benefits network users (Birke and Swann, 2005). Paradigmatic examples of industries with direct network effects are telephony, fax, e-mail and social networks. In this kind of industries, the use of the main product – e.g., a handset or a fax machine – does not offer any utility to a user if there are no other users consuming the same product because its utility derives from exchanging information. Thus, the intrinsic design of the product is usually a driver of direct network effects (McIntyre and Subramaniam, 2009). Indirect network effects appear when the increase in the number of users of a network indirectly benefits network users by increasing the availability of complementary and compatible products to the main product of that network (Birke and Swann, 2005). Examples such as software, hardware, video consoles and the videogame industry serve to illustrate this kind of network effects. The increase in the number of users consuming a product stimulates the interest of designers and manufacturers to develop complementary products which, ultimately, increases the utility of network users of the main product. An alternative classification of network effects differentiates, depending on the type of user under consideration, between total and marginal network effects. Total network effects refer to the increase in the utility of current users of a network when a new user is added to that network, Chapter 1. Introduction  8  whereas marginal network effects refer to the increase in the incentives of potential users to join the network when a new user is added (Farrell and Klemperer, 2007). In this classification, marginal network effects refer to another of the drivers of network effects, namely, user expectations. Users choose between networks depending on the future size that they expect the network will have. The rise in the number of current users increases potential users’ expectations and, thus, their incentives to join the network in the future (marginal network effects). Finally, it is important to note that network effects can take place at industry-level and firm-level. Industry-level network effects occur when the increase in the number of users of a firm increases the utility of users of that product of other firms in the same industry. It means that there is compatibility between products of two firms of the same industry, e.g. in the fax-machine industry. Firm-level network effects take place when the products of two firms have some degree of incompatibility which prevents users of two different networks from being able to exchange information without extra costs. An additional user only benefits users of the same firm, but not users of the other companies of the industry. For instance, firm-level network effects appear in the mobile telecommunications industry because firm networks are usually technologically compatible (users can make calls between two different networks) but are economically incompatible (users have to pay more for off-net calls than for on-net calls) (Grajek, 2010). Chapter 1. Introduction 9  1.2.2. The extent and importance of network industries in the worldwide economy One of the reasons that strategic management in network industries deserves extra attention is that there are numerous examples of businesses that exhibit network effects. However, it is important to note that the intensity of network effects varies across industries. Table 1.1, borrowed from Srinivasan, Lilien and Rangaswamy (2004), shows 45 network industries ordered according to their degree of intensity of network effects. They carry out a survey in which experts score, from 1 to 7, the degree of direct and indirect network externalities of a set of products. They build an overall index which is the sum of the scores for the intensity of direct and indirect network effects. The overall index has values between 2 (no network externalities) and 14 (high network externalities). Industries which show the most intensive network effects are software (operating system, word processing and spreadsheet), personal data assistant, fax machine, internet service provider and cellular telephone industries. As McIntyre and Subramaniam (2009) highlight, a high intensity of network effects can be explained by factors such as product design (e.g. fax machine and mobile telephone), degree of necessity of complementary products (e.g. software/hardware and video games/ consoles) and importance of social dynamics in product adoption (e.g. internet service providers). Chapter 1. Introduction  10  Table 1.1 Examples of network industries by intensity of network effects Rank Product Intensity of network effects 1 Operating system for personal computer 12.1 2 Personal data assistant 10.7 3 Fax machine 10.6 4 Word processing software 10.4 5 Spreadsheet software 10.2 6 Internet service provider 10.1 7 Cellular telephone 10 8 Database software 9.6 9 Workstation 9.6 10 Digital videodisc player 9.4 11 Home VCR 9.4 12 Videogame 9.4 13 Audiocassette player 9.3 14 CD player 9.3 15 CD-ROM drive 9.3 16 Mailframe computer 9.3 17 3.5 inch floppy disk drive 9.1 18 Personal computer 9 19 Automatic teller machine 8.9 20 Desktop publishing software 8.7 21 Notebook computer 8.7 22 Color television 8.4 23 High-definition television 8.4 24 Internet browser 7.6 25 Pager 7.4 26 Presentation Software 7.1 27 Antivirus Software 6.8 28 Computer-aided design software 6.8 29 Personal finance software 6.8 30 Flat-bed scanner 6.6 31 Digital camera 6.2 32 Ink-jet printer 6.2 33 Laser printer 6.2 34 Camcorder 6.1 35 Dot matrix printer 6 36 Home microwave ovens 5.8 37 Projection television 5.6 38 Instant photography 5.4 39 Photocopier 4.7 40 Single-use camera 4.4 41 Cordless telephone 4.3 42 Telephone answering machine 4.3 43 Food processor 4.1 44 Electric toothbrush 3.4 45 Pocket calculator 3.4 Source: Srinivasan, Lilien and Ragaswamy (2004) Chapter 1. Introduction 17  failed because they entered the market later, even though the quality was higher. One was the failure of the alternative standard to the QWERTY keyboard (David, 1985) and the light water technology for nuclear power reactors (Cowan, 1990). Whereas, in traditional industries, product quality is a strategic variable that directly influences performance (Porter, 1981), in network industries, strategic choices focused on quality improvement will not have the same effect. Second, given that an increase in network size increases the incentives of potential users to join the network ( marginal network effects), an increase in the network size in one period will lead to a higher network size in the following period. This has been called positive feedback (Arthur, 1990). In the words of Shapiro and Varian (1998), positive feedback makes the strong get stronger and the weak get weaker . If this process continued over time, it would result in markets with a monopolistic company. This case has been conceptualized as the winnertake-all situation and it would appear if all customers considered that only one firm will dominate the market in the future (McIntyre and Subramaniam, 2009). Thus, expectation management plays a key role in determining the success of firms in network industries. Strategic choices which increase user expectations about the future network size of the firm, such as brand and reputation management, will be especially important for achieving a better performance than its rivals (Katz and Shapiro, 1994; Shapiro and Varian, 1998). Finally, Katz and Shapiro (1985) determined that, when network effects exist, if consumers expect a seller to be dominant, then consumers will Chapter 1. Introduction  18  be willing to pay more for the firm’s product, and it will, in fact, be dominant . In contrast to traditional industries, in network industries users are willing to pay more to be part of a firm network even when there are competitors that offer lower prices because users want to benefit from a larger network size. The product is more valuable as more people use it (Doganoglu and Gryzbowski, 2007). Thus, strategic choices based on price management are less important in network industries. In sum, network effects require a change in the focus of attention from product value to network value. Strategic decisions based on managing users’ expectations and network size will take on a key role in achieving competitive advantage, in contrast to traditional decisions based on price and quality. Economic laws do not change but the key elements of strategic management do. Overall, strategic management analysis in network industries remains underexplored (McIntyre and Subramaniam, 2009) and this dissertation tries to shed light on this issue. 1.3. THE INSTITUTION-BASED VIEW OF STRATEGY 1.3.1. An overview of strategic management research The institutional perspective of strategic management has received increasing attention in recent years (Cuervo-Cazurra and Genc, 2008; Makino, Isobe and Chan, 2004; Meyer, Estrin, Bhaumik and Peng, 2009; Peng et al., 2009). This interest in institutions is a result of an evolution in the theories of strategic management, whose attention has varied in a “pendulum swing” between internal and external factors (Hoskisson et al., 1999). Chapter 1. Introduction 19  During the 1960’s, strategic management literature mainly consisted of case studies which focused on the role of manager skills (Andrews, 1971; Ansoff, 1965; Chandler, 1962). The next swing of the pendulum in the 1980’s highlighted the importance of industry structure in determining differences in performance between firms. The structure-conduct-results paradigm and the analysis of external competitive forces emerged in strategic management literature (Bain, 1956, 1968; Mason, 1939; Porter, 1980, 1981, 1985). However, explanations about differences in performance between firms of the same industry remained underexplored. As a consequence, a new perspective based on the internal factors of firms was developed during 1990’s. The resourcebased view proposed resource and capabilities allocation as drivers of firm performance (Barney, 1991; Dierickx and Cool, 1989; Penrose, 1959; Peteraf, 1993; Rumelt, 1984; Wernerfelt, 1984). The firm, as a set of valuable, rare, inimitable and irreplaceable resources, once again became the unit of analysis. As could be expected, the 21st century has seen the return of the environment as a key determinant of firm behavior and performance. However, strategic management analysis has not focused on specific industry environments, as it did in the 1980’s. The new perspective highlights the macroeconomic environment which is common to all industries. Formal institutions, such as laws and regulations, and informal institutions, such as customs, traditions and culture, are understood to determine both strategic decisions and performance (Peng et al., 2009). Interest in institutional variables, which are usually country specific, is mainly a consequence of the globalization process which most Chapter 1. Introduction  20  industries have undergone in recent decades (Dunning and Lundan, 2008). 1.3.2. The institution-based view of strategic management Institutions are defined as the rules of the game in a society or, more formally, as the humanly devised constraints that shape human interaction (North, 1990: 2). They have been also defined as cognitive, normative and regulative structures and activities that provide stability and meaning to social behavior (Scott, 2008: 33).They are gaining more and more relevance in the explanation of market competition and firm performance. It is argued that strategic choices do not only depend on industry structure (industry perspective) and firm resources and capabilities (resource perspective), but also on formal and informal restrictions from the institutional context (institutional perspective). It is acknowledged that firm strategies, organization structures, and governance mechanisms successfully pursued and implemented in a particular institutional context may not achieve the same outcomes in another institutional context (Hoskisson et al., 1999: 445). The interaction between institutions, organizations and strategic choices has recently become a research issue in management literature (Peng et al., 2008). The behavior and performance of an organization should be analyzed within the institutional framework in which it operates (Peng et al., 2005; Peng, 2002; Singh, 2007). According to the model of Peng (2000) (Figure 1.3), strategic choices are the outcome of a dynamic interaction between institutions and organizations which is conditioned Chapter 1. Introduction 21  by formal and informal rules, industry conditions and resource allocation. Figure 1.3. Institutions, organizations and strategic choices Source: Peng, MW (2000). Business Strategies in Transition Economies. Thousand Oaks, CA: Sage The difference between this institution-based view of strategic management and previous theories is that it integrates the industry and resource-based perspectives. The institution-based view complements previous theories and, according to Peng, Sun, Pinkham and Chen (2009), constitutes ‘a third leg for a strategy tripod’. This perspective uses the theory of transaction costs economy (TCE) to explain how institutions reduce uncertainty and asymmetric information between contractual parts (North, 1990) and how this can affect strategic choices and performance (Meyer et al., 2009). Chapter 1. Introduction  22  The introduction of the institution-based view into strategic literature is recent. Empirical analyses have mainly focused on foreign direct investment (FDI) decisions made by multinational enterprises (MNEs) (Brouthers, Brouthers and Werner, 2008; Chan, Isobe and Makino, 2008; Cuervo-Cazurra and Genc, 2008; Meyer et al., 2009). These studies have focused on the effect of macroeconomic institutions, such as property rights protection, corruption, rule of law and cultural distance, in host country selection and entry mode choice (Dunning and Lundan, 2008). In sum, although the institution-based view of strategy is considered a third key perspective in strategic management analysis (Peng et al., 2008), its integration with previous perspectives and strategic management issues requires further attention. Moreover, the influences of formal and informal institutions in strategic management have been analyzed separately, instead of being considered as complementary parts of the same puzzle (Makino and Tsang, 2011). Thus, further research which incorporates the role of context, both formal and informal, in the strategic management analysis is necessary (Bamberger, 2008; Peng, 2002). 1.4. STRUCTURE This dissertation aims to analyze the key role of network effects and institutions in strategic management analysis in depth. In the following chapters, the traditional consideration of both these elements as background conditions is replaced by their taking on a key role in explaining firm choices and performance within an industry. Figure 1.4 Chapter 1. Introduction 23  shows a summary of the structure of the dissertation, which is explained below. Chapter 2, titled “ Research Setting: Mobile Telecommunications Industry” , describes the evolution of the mobile industry in European and non-European countries and identifies the main characteristics which make this industry an adequate research setting for the three empirical analyses of the following chapters. Chapter 3, “ Strategic choices, network value and performance: a strategic approach to network value in network industries” , aims to extend previous literature by analyzing the role that firm strategy plays in markets where network effects are important. In a context of competition under direct and firm-level network effects, this chapter posits that firms can benefit from the existence of network effects through their strategic choices. It proposes a theoretical model in which strategic choices can improve network value by influencing the antecedents of network effects, i.e. expectations, coordination, and compatibility. The main contributions of Chapter 3 are the following. First, it proposes a theoretical model about how strategic choices influence network value and, thus, firm performance. Second, it empirically tests this model by considering the impact of three key strategic choices on network value, namely, timing of entry, the degree of internationalization and switching costs management. Finally, Chapter 3 proposes a new measure of network value which corrects previous measures by considering not only network size but also the intensity of network effects. Chapter 1. Introduction  24  Chapter 4, “Strategic choices and institutions in the FDI process: where to enter” , aims to analyze the influence of the institutional environment, both formal and informal, on host market selection by MNEs. This chapter takes as its context the process of internationalization, which has been the traditional focus of attention of the institution-based view. It proposes that strong formal institutions, such as laws, regulations and judicial systems which support economic exchanges by reducing contractual risks, can diminish the negative effect that cultural distance has on the likelihood of entering a country. The main contributions of Chapter 4 are the following. First, formal and informal institutions are considered complementary in explaining host market selection, whereas previous studies have tended to focus on formal and informal institutions separately. Second, this chapter offers a very rich empirical setting by covering home and host countries from five continents, while previous studies have tended to focus on entry decisions of MNEs from the same home country. Chapter 5, “Institutions and performance after a radical technological change: How the value of specialized complementary resources varies across markets” , aims to analyze the impact of formal institutions on the achievement of incumbents’ advantages after a radical technological change. The institutional perspective is used in a research topic which has traditionally been analyzed from the resource-based view. This chapter posits that, after a radical technological change, the value of specialized complementary resources – resources attained by incumbents that help them to commercialize the innovation and are difficult for Chapter 1. Introduction 25  newcomers to replicate – depends on the institutional environment in which the firms operate. The key hypothesis of this chapter is that the degree of development of formal institutions (weaker vs. stronger) moderates the relationship between the stock of specialized complementary assets and firm performance. The main contributions of Chapter 5 are threefold. First, the technology management and the institution-based view literature are integrated to understand how complementary resources can help incumbents to succeed in turbulent environments. Second, it offers additional empirical support for incumbents’ advantages in technological dynamic environments. Third, evidence based on a wide sample of countries covering the five continents is presented, which allows a higher level of generalizability for our results. Finally, “Summary and Conclusions” , includes a summary of the main findings and contributions of this dissertation. Chapter 1. Introduction  26  Figure 1.4. Dissertation structure    Chapter 2. RESEARCH SETTING: MOBILE COMMUNICATIONS INDUSTRY    Chapter 2. Research setting: mobile communications industry 35  2.1. INTRODUCTION he mobile communications industry has attracted the attention of scholars from many different disciplines (Birke and Swann, 2006; Fuentelsaz, Maicas and Polo, 2012; Gruber and Verboven, 2001, Jang, Dai and Sung, 2005; Maicas, Polo and Sese, 2010), which is not surprising given the social and economic importance of mobile communications in our society (Fuentelsaz, Maicas and Polo, 2008). Mobile telecommunications are now part of daily life. To illustrate this, Figure 2.1 shows that the mobile penetration rate grew substantially from 12% in 2000 to 78% in 2011. This means that, nowadays, almost everybody around the world possesses at least one handset. Having a mobile has become the rule when 10 years ago it was the exception. This quick diffusion of mobile technology has no precedents in other technologies. Figure 2.1 also shows that, for instance, fixed telephony and the Internet have not reached such high penetration rates and their growth rhythms are very different to that of mobile telephony. Although the worldwide presence of the Internet is increasing, its annual growth is lower than that of mobile communications. The penetration rate of fixed telephony and its comparison with mobile communications will be the focus of our attention in Section 2.5. T Chapter 2. Research setting: mobile communications industry 36  Figure 2.1. Worldwide penetration rate (2000-2010) Source: WDI (2012) It is important to note that there are remarkable differences in the mobile penetration rate across the world. For instance, in 2011, mobile telecommunications in Europe and America had a penetration rate of over 100%, whereas other regions, such as Africa (61%), Asia Pacific (78%) and the Middle East (78%), had lower penetration rates (Figure 2.2). However, if we look at the penetration rates in 2000 the situation was somewhat different. Although Western Europe and USA/Canada had penetration rates of above 40% - mainly because they were the birthplace of mobile communications -, the industry had a low diffusion rate in other regions. Only 2% of the population in Africa had a handset, 7% in Eastern Europe and Asia, 10% in the Middle East and 12% in Latin America. Only 10 years later, Eastern Europe and Latin America had penetration rates of over 100%. Chapter 2. Research setting: mobile communications industry 37  Figure 2.2. Mobile penetration rate by region (2011) Source: Own elaboration from Wireless Intelligence (2012) The rapid diffusion of mobile technology around the world is strongly linked to the existence of network effects (Doganoglu and Gryzbowski, 2007; Gruber, 2005). As Economides and Himmelberg (1995) highlight that, in network industries, it is necessary to achieve a critical mass after which network effects start to work. In the case of the mobile communications industry, this threshold in the critical mass was encouraged by the introduction of the GSM standard, initially in Europe (birthplace of the GSM technology) and subsequently in the rest of the world. In what follows, Section 2.2 analyzes the evolution of the mobile communication industry and the introduction of the GSM standard in the different regions previously defined. The next sections are devoted to highlighting the characteristics which make the mobile Chapter 2. Research setting: mobile communications industry 38  communications industry an adequate research setting in which to carry out the three empirical analyses of the following chapters. Section 2.3 describes the type of network effects, according to the classification in Chapter 1, which takes place in mobile communications industry. Section 2.4 shows that international groups which operate in very different institutional contexts have become the big players in this industry. Section 2.5, describes in depth the technological change which took place between fixed and mobile telephony. Section 2.6 closes the chapter by relating mobile industry characteristics to the research lines of this dissertation. 2.2. EVOLUTION OF MOBILE COMMUNICATIONS INDUSTRY 2.2.1. Evolution of mobile communications industry in Europe The European region is characterized by an early and homogeneous development of mobile communications in most countries. The first mobile telephone system in Europe was commercialized by Swedish Telecom in 1956. Later, mobile systems were launched in Germany (1959), the United Kingdom (1959) and other European countries in the 60’s and 70’s (Gruber, 2005). In spite of these first attempts, the industry was not really developed until the 80’s, with the introduction of analogue systems. The analogue systems were based on radio waves that varied in frequency and technology across countries (Gruber, 2005). As can be seen in Figure 2.3, the early 80’s show a substantial growth in the Chapter 2. Research setting: mobile communications industry 39  number of subscribers, probably due to the novelty of the technology. Nevertheless, the number of users was still moderate in this first stage in comparison with the following years. The literature has suggested several reasons for this low number of users, including the high prices in a monopoly regimen, the inexistence of a critical mass and the technology restrictions derived from incompatible standards between the networks of different countries. As an example of the latter, it can be mentioned that the independent development of mobile systems in each country made international roaming impossible in a European Union area that was moving towards integration (Fuentelsaz et al., 2008). Consequently, increasing concern arose about the necessity of making mobile systems compatible. As a result, the Group Special Mobile (GSM) was created in 1982 to work on the development of a compatible standard across European countries aimed at improving the quality and efficiency of phone services. Although the first agreement to implement the GSM standard was signed in September 1987 by 14 operators from 13 countries (Hillebrand, 2002), its commercial take-off occurred in 1992. This year can be considered as the beginning of the digital era of mobile technology in Europe. Chapter 2. Research setting: mobile communications industry 40  Figure 2.3. Number and growth of subscribers in European Union countries (1982-2010) Source: WDI (2012) Chapter 2. Research setting: mobile communications industry 41  As can be observed in Figure 2.3, the number of subscribers started to grow radically after 1992. The success of GSM was based on several advantages over the analogue system that Fuentelsaz et al. (2008) summarize as (i) a more efficient use of the radio spectrum, (ii) cost advantages related to microelectronic technology, (iii) the possibility of international roaming, (iv) the exploitation of scale economies by manufacturers and (v) a better distribution of the sunk costs of R&D among the European countries. The fast increase in the number of subscribers was accompanied by a growth in wireless technology penetration. The success of the GSM system is also reflected in the high intensity of network effects (Gruber, 2005; Srinivasan, Lilien and Ragaswamy, 2004), which leads to faster mobile adoption. After the introduction of the UMTS (Universal Mobile Telecommunications System) standard in 2003, the annual growth has declined, which shows that the mobile communications industry in Europe has reached an advanced maturity stage in its life cycle. 2.2.2. Evolution of mobile communications industry in the U.S. In the United States, the first mobile communication took place in 1921 (Noble, 1962) and radio communications had an important role of radio communications during the Second World War. The first true mobile telephone appeared in 1946 and it allowed an interconnection with the fixed telecommunications system (Gruber, 2005). At that time, mobile communications started to take place at city level and were extended to Chapter 2. Research setting: mobile communications industry 42  regional level, mainly under the control of the old AT&T. In the 1980s, the first analogue systems appeared in United States under the AMPS (Advance Mobile Phone System) standard. The introduction of this standard, as occurred with the GSM in Europe, allowed interstate roaming and handset compatibility (Fuentelsaz et al., 2008). After the introduction of the AMPS standard, other incompatible standards appeared in the United States in the 1990s’ as part of 2G technologies, such as TDMA (Time Division Multiple Access), CDMA (Code Division Multiple Access) and GSM. This essentially meant that users from different companies in the same market were unable to communicate with each other. The situation arose because the regulator allowed the market to decide which standards to employ in contrast to the European case (Church and Gandal, 2005). According to Gans, King and Wright (2005: 247) the failure of the U.S. to adopt a common 2G standard, with the associated benefits in terms of roaming and switching of handsets, meant the first-generation AMPS system remained the most popular mobile technology in the U.S. throughout the 1990s. The lack of technological compatibility also resulted in the penetration rate in the United States usually being lower than that of Europe (Gans et al., 2005). As can be seen in Figure 2.4, the number of subscribers gradually increased in the United States until it reached more than 275 million in 2010, which meant a penetration rate of close to 100%. As in Europe, this industry has reached an advanced maturity stage in its life cycle and, thus, annual growth in recent years has been decreasing slightly. But, in comparison to Europe, the mobile communications industry in the Chapter 2. Research setting: mobile communications industry 49  evolution of these international groups in terms of the number of subsidiaries and the distance of these subsidiaries from their home markets. The introduction of a common worldwide standard through the GSM system not only improved mobile technology diffusion, but also encouraged the internationalization of wireless operators. Gerpott and Jakopin (2005: 636) argue that the internationalization of telcos received an additional impetus with the licensing of digital mobile networks in numerous countries with most of these networks using the Global System for Mobile Communications (GSM) standard.2 As Sarkar, Cavusgil and Aulakh (1999) explained, the internationalization of telecom firms was in part driven by exploiting scope and scale economies based on factors such as, for example, the achievement of an efficient use of network capacity and a higher negotiating power over equipment suppliers. The existence of a common standard allowed multinationals to take advantage of the technology knowledge of their home countries to develop networks in other countries. It resulted in cost reductions, with the subsequent positive impact on performance. Given that the telecom industry had been a traditional regulated industry and the radio spectrum was considered a scarce resource, the authorities controlled the number of competitors in each market by  2 GSM was initially the abbreviation of Group Special Mobile but it was changed to refer to the standard, Global System for Mobile Communications, when the group was renamed Standard Mobile Group (SMG).  Chapter 2. Research setting: mobile communications industry 50  offering few GSM licenses. Thus, when firms decided to internationalize, they usually had to acquire equity of existing national operators which had obtained GSM licenses during 1990’s. In most cases, the international groups acquired a minority participation in a national operator which was later extended. For instance, Gerpott and Jakopin (2005:648) observed that based on the minority investment experiences until the mid-1990s some of the European MNO [Mobile Network Operators] began to implement new majority takeovers of foreign firms, which had already been in the MNO business, and to transform several of their minority stakes into majority holdings. The introduction of 3G standards gave a second boost to the internationalization of telecom operators through the launching of new licenses by authorities. It gave international groups the possibility of expanding into new markets through the purchase of 3G licenses (Curwen and Whalley, 2008). For instance, in 2003, the international group Hutchison (Hong Kong) acquired 3G licenses to enter Austria, Denmark, Italy, Sweden and the United Kingdom, and, in 2005, Ireland. In sum, the internationalization of telecom firms started in the 1990’s and increased during the 2000’s, especially with the introduction of 3G technology. This increasing scope of MNEs can be appreciated in Table 2.1 that shows the number of countries in which each international group is present in 2000 and 2010. As can be observed, most international groups have increased the number of countries in which they operate in this 10-year period with the only exceptions of AT&T, KPN, Millicom, TDC and Telecom Italia. Chapter 2. Research setting: mobile communications industry 51  Table 2.1. Evolution in number of subsidiaries by MNE (2000 – 2010) Group 2000 2010 Abu Dhabi - 5 Access Industries - 3 AF - 4 Altimo - 17 América Móvil 6 17 AT&T 5 3 Axiata 7 8 Batelco 1 4 Belgacom 1 2 Bharti Airtel 1 19 Bite 1 2 Cable & Wireless 24 26 Deutsche Telecom 10 18 Digicel - 32 Elisa 2 2 Etisalat 2 16 France Telecom 34 42 Hutchison 13 13 JT 1 2 KPN 8 3 Magyar 1 3 Maroc Telecom 1 5 Maxis 1 3 MegaFon 1 4 Millicom 18 14 Mobistar 1 2 MTN 6 23 MTS 1 6 NII 5 5 NTT Docomo 7 8 Oger - 2 Orascom 4 12 OTE 5 7 Portugal Telecom 3 7 Qtel 1 14 Saudi Telecom 1 8 SingTel 5 26 Sistema 1 7 Sonatel 1 4 Sudatel - 6 Chapter 2. Research setting: mobile communications industry 52  Group 2000 2010 TDC 6 1 Tele2 6 7 Telecom Argentina 2 2 Telecom Italia 8 5 Telefónica 16 20 Telekom Austria 4 8 Telekom Slovenije 1 4 Telekom Sribija 1 3 Telenor 10 21 Telia Sonera 13 25 Telstra 2 2 Trilogy - 4 Turkcell 6 8 Viettel - 3 VimpelCom 1 10 Vivendi 5 8 Vodacom 3 5 Vodafone 23 33 Wataniya 1 6 Wind - 13 Zain 1 9 Source: Wireless Intelligence Database (2012) International groups in this industry come from different regions although European MNEs are the most internationalized groups because of their longer experience in this industry. Figure 2.6 shows the international groups that were operating at the end of 2010 by region of origin. As can be observed, most international groups come from Europe and Asia, whereas Oceania and Latin America show the lowest number of international groups. Chapter 2. Research setting: mobile communications industry 53  Figure 2.6. International groups by worldwide region (2010) Source: Wireless Intelligence (2012) Although international groups tended to expand first to geographically close countries, most of them have gone one step further by entering countries in other regions which differ in terms of language, law, tradition and customs from their home countries. To illustrate this, we have selected one international group from each region that is representative of this gradual expansion process. Figure 2.7 shows a summary of regions in which Telefónica (Europe), Hutchison AFRICA Maroc Telecom, MTN, Orascom, Sonatel, Sudatel, Vodacom ASIA Abu Dhabi, AF Telecom, Altimo, Axiata, Barthi Airtel, Batelco, Etisalat, Hutchison, Maxis Communications, Megafon, MTS, NTT Docomo, Oger Telecom, Qtel, Saudi Telecom, Singtel, Sistema, Turkcell, Viettel, Vimpelcom, Wataniya, Zain EUROPE Belgacom, Cable & Wireless, Deutsche Telekom, Elisa, France Telecom, JT, KPN, Magyar Telecom, Millicom, Mobistar, OTE, Portugal Telecom, TDC, Telecom Italia, Telefónica, Telekom Austria, Telekom Slovenije, Telekom Srbija, Telenor, Tele2, Telia Sonera, Vimpelcom, Vivendi, Vodafone, Wind LATIN AMERICA América Móvil, Digicel, Telecom Argentina USA/ CANADA Access Industries, AT&T, NII, Trilogy OCEANIA Telstra Chapter 2. Research setting: mobile communications industry 54  Whampoa (Asia), Telstra (Oceania), América Móvil (Latin America), Orascom (Africa) and Trilogy (United States) have invested. Orascom started its operations in Egypt in 1998 and entered other African countries including Algeria, Ghana, Burundi, Congo, Chad, Central African Republic, Namibia, Tunisia and Zimbabwe. Orascom has also expanded into countries in the Middle East such as Israel, Iraq and Syria. This company acquired a minority participation in Hutchison in 2005 in order to be present in Hong Kong, Indonesia, Macao, Sri Lanka and Vietnam, but finally sold it in 2007. Since 2009, Orascom has operated in Canada after acquiring 65.08% of Wind Mobile. Hutchison has mainly expanded from Hong Kong to Europe (Austria, Denmark, Ireland, Italy, Sweden and the United Kingdom) by acquiring 3G licenses. It also entered Paraguay and Ghana, although it sold these participations in 2005 and 2008, respectively. Telefónica (Spain) initially expanded during the 90’s to Latin American countries, such as Argentina, Brazil, Chile, Colombia, Costa Rica, Ecuador, El Salvador, Guatemala, Mexico, Nicaragua, Panama, Peru, Uruguay and Venezuela. In the mid-2000’s, it started its European expansion to the Czech Republic, Germany, Ireland, Italy, Slovakia and the United Kingdom. Although Telefónica also entered Morocco by acquiring 31.74% of Meditel, it finally sold off this investment in 2009. Chapter 2. Research setting: mobile communications industry 55  Figure 2.7. FDI location by MNEs from all regions FDI Location Orascom (AFRICA) Hutchison (ASIA) Telefónica (EUROPE) América Móvil (LATIN AMERICA) Telstra (OCEANIA) Trilogy (USA) AFRICA ASIA EUROPE LATIN AMERICA OCEANIA USA/CANADA Source: Wireless Intelligence (2012) Chapter 2. Research setting: mobile communications industry 56  América Movil, a Mexican international group, has mainly expanded into Latin American countries such as Argentina, Brazil, Chile, Colombia, Dominican Republic, Ecuador, El Salvador, Guatemala, Honduras, Nicaragua, Panama, Paraguay, Peru and Uruguay. Thus, this international group has become Telefónica’s main competitor in the region. Although América Móvil has focused on Latin America, in the second quarter of 2012, it acquired 27.70% of the European KPN. This operation has allowed América Móvil to enter The Netherlands (home country of KPN), Germany and Belgium. Telstra is the only international group from Oceania and its international presence is limited, only operating in Australia (its home country) and Hong Kong. Trilogy is based in the United States and has entered Oceania (New Zealand) and Latin America (Bolivia, Dominican Republic and Haiti). It is important to note that US and Canadian international groups, such as Trilogy, AT&T and NII Nextel, are present in few countries because their domestic markets are so large that they have traditionally focused on regional instead of international competition. In sum, the number of international groups and the markets in which they are operating has increased greatly in the last two decades. Moreover, the examples given above illustrate that mobile MNEs have entered countries with important differences in terms of language, law, culture, etc. to their home countries. Chapter 2. Research setting: mobile communications industry 57  Chapter 4 “Strategic choices and institutions in the FDI process: where to enter” aims to analyze how host market selection by MNEs depends on the institutional context, both formal and informal, of the host countries. The mobile communications industry is an adequate research setting to empirically analyze host market selection from the institution-based view because MNEs have entered countries whose institutions greatly differ from those of their home countries. 2.5. The change from fixed to mobile communications Mobile and fixed communications have coexisted in the market in the last two decades. The evolution of the penetration rates of the two technologies seems to reflect a substitution process (Cadima and Barros, 2000; Gans et al., 2005; Horvath and Maldom, 2002). The existence of advantages in mobile services compared to fixed telephony such as, higher competition, lower prices and higher functionality, may explain this substitution effect (Gruber and Verboven, 2001; Gans et al., 2005; Rodini, Ward and Woroch, 2003). As can be seen in Figure 2.8, the number of adopters of mobile technology has continuously increased during the last decade. On the contrary, the number of users of fixed-telephony remains, roughly speaking, steady and, from 2007, starts slightly decreasing. More Chapter 2. Research setting: mobile communications industry 58  importantly, from 2002 on, the number of mobile users is substantially higher than the number of fixed users. Figure 2.8. Worldwide millions of subscribers by technology (2000-2010) Source: Wireless Intelligence Database (2012); WDI (2012) This substitution effect can be more clearly observed in Figure 2.9. If the total number of connections (fixed and mobile) is considered, mobile technology has evolved from representing 43% of connections worldwide in 2000 to 82% in 2010. This confirms the substitution of fixed by mobile telephony. Chapter 2. Research setting: mobile communications industry 65  Noble, D. 1962. The history of land mobile communications. Proceedings of the IRE , Vehicular Communications 50(5): 157-179. Wireless Intelligence Group. 2012. Wireless Intelligence Database. World Bank Group. 2012. World Development Indicators Online .  Chapter 2. Research setting: mobile communications industry 66    Chapter 3. STRATEGIC CHOICES, NETWORK VALUE AND PERFORMANCE: A Strategic Approach to Network Value in Network Industries   Chapter 3. Strategic choices, network value and performance 69  3.1. INTRODUCTION etwork industries, defined as those in which network effects are important to understand how firms compete, represent a large and growing portion of today’s economy. As shown in Chapter 1, software, mobile communications and video games are just a few examples of industries where network effects drive market competition (Shankar and Bayus, 2003; Tanriverdi and Lee, 2008). In recent years, management and economic literature have devoted increasing attention to these industries (Farrell and Klemperer, 2007; McIntyre and Subramaniam, 2009; Shankar and Bayus, 2003). This may be a reaction to evidence that network industries seem to challenge much of the thinking derived from previous models and findings (Shapiro and Varian, 1998; Suarez, 2005). However, although recent literature recognizes that the foundations of network effects have received an increasing amount of attention from researchers (Varian and Shapiro, 1998; Farrell and Klemperer, 2007), a deeper understanding of the role that firm strategy plays in leveraging network effects is needed (McIntyre and Subramaniam, 2009). One of the main premises of businesses such as software and telecommunications is that the firm’s installed customer base can be considered a key strategic asset to gain sustainable competitive advantages (Shankar and Bayus, 2003). This is because the existence of network effects implies that consumers’ utility is directly affected by the number of consumers using the same product or technology (Shy, 2011) and, thus, customers’ willingness to pay increases, with the subsequent N Chapter 3. Strategic choices, network value and performance 70  potential impact on firm performance (Shapiro and Varian, 1998; Shankar and Bayus, 2003). There is a growing body of literature that attempts to measure network effects in a variety of industries. This stream of research is mainly focused on technological standards competition (Cowan, 1990; David, 1985; Garud and Kumaraswamy, 1993), technology adoption and diffusion (Gandal, Kende and Rob, 2000; Goolsbee and Klenow, 2002; Majumdar and Venkataraman, 1998; Park, 2004) or the analysis of hedonic price functions for products exhibiting network effects (Brynjolfsson and Kemerer, 1996; Hartman and Teece, 1990, Gandal 1994). However, only a few papers have analyzed how firms’ strategic decisions may influence performance when network effects are important. These papers have paid attention to the impact of strategic dimensions such as entry timing and learning orientation (Schilling, 2002), product diversification (Tanriverdi and Lee, 2008) and pioneers’ advantages (Eisenman, 2006). One commonality of these works is that they focus their attention on specific attributes of strategic choices, without establishing a general model about how strategy helps firms to gain a competitive advantage in network industries. This chapter attempts to explain how firm-initiated strategic actions can help firms to benefit from the existence of network effects. Following McIntyre and Subramaniam (2009), this chapter aims to study the implications of strategy in network industries in greater depth. It is built on both economic and strategic literatures under the premise that understanding the drivers of network effects will allow firms to adopt a Chapter 3. Strategic choices, network value and performance 71  more proactive position and intensify the network effects to their own benefit. This chapter also extends previous research by suggesting that network value, defined as the value stemming from other consumers already using the product (McIntyre and Subramaniam, 2009:1496), is more accurate than network size for assessing a firm’s competitive position in the presence of network effects. In contrast to most of the existing empirical literature (Brynjolfsson and Kemerer, 1996; Schilling, 2002), this chapter proposes an adjusted measure of network value, based on Metcalfe’s law, that includes not only network size but also network intensity. Previous literature has identified three elements that act as antecedents of network effects (Farrell and Klemperer, 2007; Katz and Shapiro, 1994; Shapiro and Varian, 1998), namely, users’ expectations, users’ coordination and compatibility among competing networks. This chapter postulates that firms, by managing these elements through their strategic decisions, can leverage network effects and increase network value in the industries in which they operate. In particular, it is analyzed how several strategic initiatives based on the management of the installed base, such as entry timing, internationalization and switching costs, are related to users’ expectations, users’ coordination and compatibility among competing networks and, eventually, to network value. Focusing on firm-initiated actions that shape the firm’s competitive destiny in network industries, this chapter brings a strategic dimension to the research in this field by offering a theoretical model that relates Chapter 3. Strategic choices, network value and performance 72  strategic actions and the drivers of network effects. This analysis focuses on the concept of network value, which has been previously analyzed from a theoretical perspective in the literature. The main contribution of Chapter 3 lies in the proposal and analysis of an improved measure of network value that integrates the size and intensity dimensions of network effects in an empirical analysis. Finally, this chapter not only seeks to expand on prior findings by including the effect of firm strategy on network value, but also analyzes the impact of network value on firm performance. The rest of the chapter is organized as follows. The next section develops the theoretical model, paying special attention to the relationship between network effects and network value and between the latter and its main antecedents: expectations, coordination and compatibility. This section also provides a theoretical explanation of the effect of three strategic initiatives, namely, entry timing, internationalization and switching costs management, on network value. Section 3.2 also analyzes the relationship between network value and the performance of firms. The data from the European mobile communications industry and the variables used are presented in the third section 3.3, while the section 3.4 describes the estimation procedure. Following that, evidence on the impact of entry timing, internationalization and switching costs on network value and the influence of the latter on firm performance is provided. Section 3.6 closes the chapter by discussing its main findings and its managerial and policy implications. Chapter 3. Strategic choices, network value and performance 73  3.2. THEORY AND HYPOTHESES 3.2.1. Installed Base, Network Effects, Network Value and Network Intensity Previous literature has highlighted the role of the installed base as a strategic asset in network industries (Brynjolfsson and Kemerer, 1996; Chacko and Mitchell, 1998; Shankar and Bayus, 2003). The installed base can be defined as the cumulative number of users at any given time in the product’s life (McIntyre and Subramaniam, 2009:1495). This strategic consideration of the installed base in network industries is explained by the existence of network effects that are present when the utility that a user derives from consumption of the good increases with the number of other agents consuming the good (Katz and Shapiro, 1985: 424). Thus, user utility is dependent on the size of the installed base (Shapiro and Varian, 1998) and this results in interdependent demand (Rohlfs, 1974). The importance of the installed base to gain competitive advantages is clear in markets whose network effects are direct or pure,1 such as the telephone, fax and e-mail industries. Stand-alone benefit is negligible because the product or service has to be integrated into a network to obtain value from it (DePalma and Leruth, 1996; Grajek, 2010). Given  1 The literature has traditionally distinguished between direct and indirect network effects. The first refer to when adoption by different users is complementary, so that each user's adoption payoff, and his incentive to adopt, increases as more others adopt . The second arise through improved opportunities to trade with the other side of a market (Farrell and Klemperer, 2007: 1974). This Chapter focuses its attention on direct network effects, although most of the arguments offered would also stand for indirect network effects. Chapter 3. Strategic choices, network value and performance 74  the existence of network effects, the main competitive advantage of the firm is based on creating a higher network value than its rivals, and not exclusively on generating a higher network-independent value based on quality issues (McIntyre and Subramaniam, 2009).2 Network value has been defined as the value stemming from other consumers already using the product and it is the reflection of the benefits associated with a large cohort of fellow adopters (installed base) for the product (McIntyre and Subramaniam, 2009:1496). As a consequence, network value directly depends on the size of the installed base. The higher the number of users of a network, the higher the interaction possibilities between its members and, thus, the greater the utility they receive from belonging to that network. It is necessary to note that network value is not merely the size of the installed base. Network value must also take into account the existence of network effects, which make it important for users to consume the product within a community. McIntyre and Subramaniam (2009) recognize that the relationship between the installed base and network value is not linear but depends on the strength of network effects or network intensity, which can be defined as the relative value generated  2 This chapter focuses on the network value that is directly dependent on the existence of other users consuming the product, that is, the value that comes from the existence of network effects. McIntyre and Subramaniam (2009) also identify a part of network value that can be network-independent. This network-independent value captures quality characteristics of the product that “ are under the full control of the producer ” (Bental and Spiegel, 1995:197), such as, in our industry, network coverage or network reliability. Accordingly to McIntyre (2011), companies with higher network value also tend to offer, from the organizational learning perspective, greater network-independent value since they have accumulated more experience and capabilities in the industry. Chapter 3. Strategic choices, network value and performance 81  longer time in the market has a larger network value because it has had more time to make efforts in the management of users’ expectations through the achievement of an early installed base before the entry of rivals. It is also important to note that the inertia that has been discussed before will lead late users to choose the firms with a larger installed base. If a pioneer is able to convince early users about its dominance, late consumers will prefer to follow them into the same network and the pioneers’ product will become the standard in the industry (Schmalensee, 1982; Carpenter and Nakamoto, 1986; Farrell and Klemperer, 2007). Having achieved a leading position, the pioneers’ installed base will persist because of the difficulty of modifying users’ preferences (Lieberman and Montgomery, 1988). This is the main idea of the bandwagon effects which have been previously referred to. Accordingly, it is expected that time in the market increases the firms’ opportunities to influence user expectations about their networks. As a result of inertia, the network value of a firm that has been in the market a long time will be higher. H1. The time that a firm has been operating in the market has a positive effect on its network value Internationalization and network value. The literature has tended to study markets with indirect network effects in which diversification in complementary products plays an important role (Hill, 1992; Schilling, 2002; Tanriverdi and Lee, 2008). However, less attention has been paid to other growth strategies in markets with direct network effects such as Chapter 3. Strategic choices, network value and performance 82  international diversification, especially when international network effects operate (Gruber and Verboven, 2001). Internationalization is, nowadays, an important topic of discussion because many firms are trying to compete globally (Barkema and Drogendijk, 2007; Grant, 2005). As a result, not only are firms present in various countries, but customers also “think” globally. National and regional preferences are disappearing as a consequence of a process of homogenization derived from technology, communication and travel (Grant, 2005). This means that customers are becoming more and more familiar with international firms and their brands. The internationalization of firms could be a means of attracting the interest of users in different countries since users value established brands (Lane and Jacobson, 1995). It would be expected the internationalization of a firm to influence its network value through its impact on expectations, coordination and perceived compatibility. First, internationalization can be understood as a signal that influences users’ expectations about future network dominance. There is an advantage for a firm entering a new local market when it has a wide international scope. It will have a larger perceived installed base compared to new domestic firms. Accordingly, the literature has highlighted the existence of international network effects through which the utility of each consumer rises with the increase in the number of consumers who use the same brand regardless of whether they live in their own country or abroad (Shy, 2001: 92). Thus, an international firm will reinforce the positive expectations of users about its future survival Chapter 3. Strategic choices, network value and performance 83  on the basis of being present in other countries and the familiarity of domestic users with its brand through the leverage of international network effects. Consequently, I also expect that internationalization will facilitate coordination through international bandwagon effects. If users know that a firm has been chosen by users in other countries, inertia could lead them to make the same choice in their home market. Users will have more incentives to choose the international firm, replicating the choices of foreign users, since they want to imitate global trends (Grant, 2005). Firms with an international presence try to create interdependences among different countries, which result in a close relation between the competitive position in one national market and the competitive position in others (Ghoshal, 1987: 425). Finally, it is also important to note that compatibility among intercountry networks is necessary to influence users’ decisions. In the case of mobile telecommunications, Gruber and Verboven (2001) suggest that, with GSM wide-ranging international roaming, users may have greater incentives to adopt mobile communications since they benefit from international network effects. The firms that offer comparable, seamless and compatible services across international markets will obtain the commitment of users that exchange information internationally (Sarkar, Cavusgil and Aulakh, 1999). As a consequence, it is expected that the presence of the firm in various countries will create a larger network value through its influence on expectations and coordination as firms try to compete globally in order Chapter 3. Strategic choices, network value and performance 84  to attract users across countries. Compatibility will reinforce the influence of internationalization on network value by allowing international network effects. H2. The level of internationalization of a firm has a positive effect on its network value Switching costs and network value. Switching costs are present in all network markets and their management has a strategic dimension (Gomez and Maicas, 2011; Shapiro and Varian, 1998). Consumer switching costs appear when consumers who have previously purchased from one firm have (or perceive) costs of switching to a competitor’s product, even when the two firms’ products are functionally identical (Klemperer, 1995: 515). The literature has highlighted how switching costs can increase the market power of a firm, allowing it to create entry barriers (Karakaya and Stahl, 1989; Kerin, Varadarajan and Peterson, 1992) and obtain abnormal returns that allow the firm to achieve sustainable competitive advantages (Amit and Zott, 2001; Klemperer, 1987; Lieberman and Montgomery, 1988; Schmalensee, 1982). However, the effectiveness of this mechanism as a basis for sustainable competitive advantages in information markets has been questioned (Mata, Fuerst and Barney, 1995). The effect of high switching costs may result in the loss of network value through their impact on expectations and coordination, as argued below. As mentioned before, network value depends on the installed base and users’ utility in the presence of network effects. While switching costs have been used as an instrument to maintain the installed base by Chapter 3. Strategic choices, network value and performance 85  reducing customers’ desire to leave their current provider (Burnham, Frels and Mahajan, 2003), these costs reduce users’ utility (Maicas, Polo and Sese, 2009) not only because switching from one provider to another is costly but also because users perceive the threat of opportunistic firm behavior that could lead to future price increases in a bargain-then-ripoff pricing strategy (Farrell and Klemperer, 2007). It is not surprising that this expected opportunism leads users to form a negative image of the firm (Mata et al., 1995). Since potential users tend to form expectations about the future survival of the firm not only with quantitative signals such as the installed base, but also with qualitative signals like brand image or reputation (Katz and Shapiro, 1994), they will be reluctant to choose a firm with high switching costs. Frels, Shervani and Srivastava (2003) comment that a network of previous adopters is believed to influence adoption among non-adopters by providing opinions by word of mouth and observation. The negative experience of the current installed base will result in the formation of negative expectations about a firm network with higher switching costs and will prevent user coordination with this network, leading to a negative impact on network value. Mata et al. (1995: 490) explain that the value of opportunities lost because of a reputation for exploiting captured customers can be much larger than the value extracted from those captured customers . Switching costs are especially high when networks are incompatible. In particular, technological incompatibility is one of the main drivers of consumer switching costs (Garcia-Mariñoso, 2001). It is costly to abandon a network because of learning costs or loss of communication Chapter 3. Strategic choices, network value and performance 86  possibilities with current users. Economic or artificial incompatibility also arises when the costs of communication among users are cheaper if they belong to the same network (Grajek, 2010). In this case, economic incompatibility increases the pecuniary switching costs derived from the higher costs of communicating with users of the previous network. Thus, incompatibility will reinforce the negative effect of switching costs on utility and, consequently, on network value. H3. Switching costs have a negative effect on firm network value . Network value and performance. In network industries, current performance is strongly dependent on past events (Farrell and Klemperer, 2007; McIntyre and Subramaniam, 2009). This is the socalled positive feedback that reinforces that which gains success or aggravates that which suffers loss (Arthur, 1996: 100). The literature has suggested that a continuous increase in network value is followed by an increase in the willingness to pay to have access to that network (Doganoglu and Grzybowski, 2007) and the subsequent decrease of the marginal costs of each information interchange (Arthur, 1990). This is because the value does not lie in the product itself, but in the size and intensity of the network (De Palma and Leruth, 1996; Grajek, 2010). The product is more valuable as more people use it (Doganoglu and Gryzbowski, 2007). While a greater network value permits a higher price, marginal costs decrease as more and more information ties take place. In spite of a large initial investment, the marginal costs of producing an additional exchange are relatively cheap Chapter 3. Strategic choices, network value and performance 87  (Shapiro and Varian, 1998) because information markets are knowledgebased (Arthur, 1990). It is expected that a firm with a larger network value will also obtain a higher marginal net income from each information exchange derived from a higher price and lower marginal costs. Thus, performance will be positively related to network value. H4. Network value has a positive effect on firm performance. According to previous arguments, I build a model to test in the following sections as it is shown in Figure 3.1. Chapter 3. Strategic choices, network value and performance 88  Figure 3.1. Strategic Choices, Network Value and Performance (Hypotheses) Chapter 3. Strategic choices, network value and performance 89  3.3. DATA AND VARIABLES 3.3.1. Research Setting: the European Mobile Communications Industry6 As told in Chapter 2, mobile technology has reached the highest penetration rate in European countries given the consecution of a common standard at the beginning of 1990s’. The European mobile communications industry represents a large, fruitful and growing portion of Europe’s economy. This industry has become an important source of wealth in Europe. For instance, the telecommunications industry made up 2.83% of the GDP at the end of 2007, whereas, for example, agriculture constituted 1.82% (World Bank Group, 2010). The Financial Times Global 500 Index (2011) shows that 11 of the 50 largest firms in the world belong to network industries, five of them being mobile operators of which two, moreover, are European (Vodafone in the United Kingdom and Telefónica in Spain). The literature emphasizes the role of expectations and users’ coordination on users’ choice of mobile network (Doganoglu and Grazybowski, 2007; Gandal, 2002; Church and Gandal, 2005). It has been  6Given that Chapter 3 does not introduce yet the institution-based view of firm strategy and focuses on strategic management on network industries, it has been preferred to focus on the mobile telecommunications in European countries. Using this mature industry in Europe assures a higher homogeneity in terms of competition, users’ preferences and regulation. It will allow that sample is not biased by institutional differences between countries in terms of technology acceptance and regulation. Moreover, taking European countries assures the existence of a common standard and, thus, roaming availability between countries (in order to test Hypothesis 2). Chapter 3. Strategic choices, network value and performance 90  shown that, among other factors, the total installed base of an operator plays an important role in users’ expectations and coordination (Birke and Swann, 2006). Because of this, small operators in European markets may fail if they do not achieve a minimum critical mass to influence users’ expectations and coordination (Economides and Himmelberg, 1995). Incompatibility issues have been especially remarkable in the European context in determining the scope of networks and understanding the existence of tariff-mediated or artificial network effects. As previously mentioned, the scope of networks is dependent on technological and economic compatibility. With regard to technological compatibility, in 1984, the European Commission, through the Group Special Mobile (GSM), encouraged the development of a common technological standard which allowed mobile services within national and international networks. As a consequence, a user can employ his/her handset to make calls to the mobile phones of any firm in the country without technological restrictions and can use the same handset in any European country thanks to international roaming agreements. Nevertheless, in spite of this technological compatibility guided by supranational authorities, an economic incompatibility between firms’ networks comes from the price discrimination between on-net and offnet calls. It generates what the literature has called tariff-mediated network effects, which appear at firm-level (Grajek, 2010; Laffont, Rey and Tirole, 1998). Users prefer to belong to a larger network to reduce Chapter 3. Strategic choices, network value and performance 97  there are no price differences between on-net and off-net calls and accepting that the calls from one network to another are proportional to the sizes of the installed bases, the expected call probability among users of different networks is given by the product of their respective market shares as shown in the following matrix (Table 3.2). Table 3.2. Likelihood Matrix of Calls Across Networks To Network 1 2 3 4 Calls from Network 1 m1 m1 m1 m2 m1 m3 m1 m4 2 m2 m1 m2 m2 m2 m3 m2 m4 3 m3 m1 m3 m2 m3 m3 m3 m4 4 m4 m1 m4 m2 m4 m3 m4 m4 Source: Birke and Swann (2006). The probability of making on-net calls (probon-net) is given by the elements of the matrix diagonal (), whereas the off-diagonal elements () refer to off-net call probability (proboff-net) between networks for each firm. Thus, the probability of making on-net calls over off-net calls for each firm i in a market with M companies is given by the ratio:  ∑      [i≠j] (3.1) Chapter 3. Strategic choices, network value and performance 98  By modifying Zipf’s Law with this ratio, the adjusted network value ( NETWORK VALUE’ ) is expressed as: (3.2) As a consequence, the adjusted network value will be higher when: a) there is a larger installed base that allows greater communications possibilities among current users of the network (network size dimension of network value); b) there is a larger difference between the network sizes of the reference firm and its rivals, which gives it a competitive advantage to leverage more intensive network effects and make its network more attractive to potential users (network intensity dimension of network value). Performance ( PERFORMANCE ). Firm profitability is measured through EBITDA (Earnings Before Interest, Taxes, Depreciation and Amortization) divided by the total revenues of the firm. Both EBITDA and revenues are calculated for each firm in each national market. Time in the market ( TIME ). Different concepts of pioneering have been used when modeling first-mover advantages. Srinivasan et al. (2004) consider the pioneer to be the first firm to commercialize a new product. Lieberman and Montgomery (1988) suggest some alternative measures such as the numerical order of entry, rates of company survival, duration of advantages and time from pioneer entry. Brown and Lattin (1994) suggest time in the market as an adequate measure of FMA. i i im m nn i1 )log( Chapter 3. Strategic choices, network value and performance 99  Finally, this variable counts the number of months that a firm has been operating in digital wireless technology (GSM). The decision to take GSM as the starting point of the market responds to the scarce acceptance of analogical technology. For example, in the ten years between 1980 and 1990 when analogical technology was available, the rate of penetration only grew from 0.0% to 0.92%. Accordingly, it is assumed that the market was almost non-existent before the introduction of the digital generation. International presence ( INTERNATIONALIZATION ). The literature has traditionally measured international diversification through variables such as international sales over total sales (Strike, Gao and Bansal, 2006), number of workers abroad (Brock, Yaffe and Dembovsky, 2006), sales in a country weighted by the importance of this market (Hitt et al., 1997), number of international subsidiaries (Barkema and Drogendijk, 2007; Strike et al., 2006) and the number of countries in which the firm operates (Brock et al., 2006). In this chapter, it has been chosen the number of countries in which the firm is present with an ownership of above 50%. The theoretical rationale is that to influence network value, the level of firm internationalization has to be in the users’ minds. Therefore, the main reason to choose the number of countries in which the firm is operating is that this information is known by the user, while other alternatives previously mentioned – number of workers abroad, international sales… – are not easy for the user to identify. More importantly, the criteria of 50% of ownership has been selected to assure that the international group considers the national operator as Chapter 3. Strategic choices, network value and performance 100  part of the core organization and that international network effects can develop. After reviewing annual reports of international groups in Europe, it is observed that there has been a gradual acquisition of the ownership of national operators, from minority to majority, by international groups. Only after acquiring more than 50% of the ownership, have international groups included the national companies as part of their organizational chart. Moreover, for international network effects to exist, users must be able to recognize the same firm operating in different markets (Shy, 2001), so the international groups in Europe have started to build global brands. The rebranding of acquired operators by international groups has only taken place after the acquisition of an ownership above 50%. Switching costs ( SWITCHING COSTS ). According to the existing literature, there is an important gap between the theoretical and the empirical research on switching costs (Stango, 2002; Grzybowski, 2007; Chen and Hitt, 2007; Viard, 2007). Only a few articles have tried to properly measure their magnitude. I closely follow the model proposed by Shy (2002). This author develops a method for estimating switching costs among firms in a context where it is only needed to have information about prices and market shares. It is important to note that Shy’s method has been previously used in the literature with very similar purposes to those of this chapter (Carlsson and Löfgren, 2006; Gomez and Maicas, 2011; Krafft and Salies, 2008). Shy (2002) considers a market with two firms (A and B). Consumers are assumed to be distributed between the firms so that, initially, NA Chapter 3. Strategic choices, network value and performance 101  consumers have already purchased brand A (type a consumers) and NB consumers have already purchased brand B (type b consumers). pA and pB represent firm A and B prices, respectively, and s is the cost of switching brands. The utility UA ( UB ) for a user who is now buying from A (B), can be written as: The number of subscribers for A (B), nA ( nB ) in the following period is given by, If it is assumed that the firm’s production costs are zero, the profit, A (B), of each firm is: Shy (2002) postulates that the pair of prices that solve the problem for firms A and B and constitute a Nash-Bertrand equilibrium are: Shy (2002) extends the model to a multi-firm industry. He considers the possibility of more than two firms, each indexed by i, i = 1,…, M (firms staying with brand A switching to brand B A def A B p Ups       0 if if if AB AA B AB AB AB pps nN pspps NN pps            , A AB AA pp pn            22 22 22 and ABA B AB AB A B AABB AABB NNN Ns NN NNs pp NNNN NNNN       Chapter 3. Strategic choices, network value and performance 102  in order of higher to lower market share). The expressions for switching costs in a multi-firm industry are: , if I {1, …, M-1} and In this model, it is important to have a precise measure of sizes and prices. Sizes are incorporated into the switching costs function through the market shares of the firms. A more controversial issue is to define prices in mobile communications. Prices usually vary depending on the characteristics of the user, the receiver of the phone call (on-net vs. offnet calls) or the time of the day. To solve this problem, Shy (2002) derives prices from the Average Revenue per User (ARpU) in his calculation of switching costs in mobile communications in Israel. Furthermore, the use of ARpU as a proxy of prices is also motivated by its widespread use in industry and regulatory circles (McCloughan and Lyons, 2006:523). An additional advantage of ARpU is that it makes comparisons among countries possible. Control variables. Besides the variables described to test the proposed hypotheses, models also control for additional covariates. First, there is a control for the population in each national market ( POPULATION ), which is expected to have a positive relationship with network value and performance because the communication possibilities in each national market will be higher. Given that population can be considered as a proxy of the potential size of the industry, the introduction of this variable also allows us to control for the existence of industry-level network effects. Country-specific rivalry is also controlled by taking into Mi MM ii NN PN ps   M MM NN PN ps   1 11 Chapter 3. Strategic choices, network value and performance 103  account the number of firms operating in each market ( FIRMS ). This variable is expected to negatively affect firm performance. However, the relationship between the number of firms and network value is not so clear. A higher number of firms would probably result in smaller networks, decreasing network value. But the increase in the number of firms could also constitute an improvement in the competitiveness of the market and price reductions. It might enhance users’ utility and technology adoption, with a subsequent increase of network value. Finally, the model also includes year dummies to control for timespecific influences ( YEAR ). 3.3.4. Descriptive Statistics Descriptive statistics are shown in Tables 3.3 and 3.4. The first includes the determinants of the network value model and the second those of the profitability model. The existence of missing values in dependent variables implies that the final sample has 2,032 observations for the network value model and 1,991 for the profitability model. As can be seen in Table 3.3, the average value of the first measure of network value ( NETWORK VALUE ) is 15.28, while it is 9.25 for the adjusted network value ( NETWORK VALUE’ ). Moreover, the average European firm has been operating in the market for nine years (107.5 months) at the end of the study range, has established a presence in 8 countries around the world and has positive switching costs of around 17 euros per user. The average number of firms per market is 3. When the correlation matrix is analyzed, it can be observed that both network value and adjusted network value are highly correlated with population Chapter 3. Strategic choices, network value and performance 104  and with time in the market. Nevertheless, the correlation among the independent variables is moderate. Table 3.4 shows that the performance is better than the performance in the previous period, exhibiting a positive relationship with network value but a negative one with population and number of firms. Chapter 3. Strategic choices, network value and performance 105  Table 3.3. Descriptive Statistics Model 1 (n= 2,032) Variable Mean Std. Dev. Min Max 1 2 3 4 5 6 7 1. NETWORK VALUE 15.28 24.09 -0.37 140.08 - 2. NETWORK VALUE’ 9.25 17.14 -0.13 105.38 0.92* - 3. TIME 107.48 44.36 3.00 258.00 0.37* 0.33* - 4. INTERNATIONALIZATION 7.95 7.04 1.00 28.00 0.28* 0.21* 0.31* - 5. SWITCHING COSTS 17.22 11.19 -18.28 56.51 0.07* 0.15* 0.24* 0.13* - 6. FIRMS 3.27 0.65 2.00 5.00 0.14* 0.03 -0.10* 0.15* 0.02 - 7. POPULATION 27.12 25.64 3.87 82.541 0.76* 0.56* 0.08* 0.13* -0.02 0.31* - *p < 0.01 Chapter 3. Strategic choices, network value and performance 106  Table 3.4. Descriptive Statistics Model 2 (N=1,991) Variable Mean Std. Dev. Min Max 1 2 3 4 5 6 1. PERFORMANCE t 0.29 0.24 -3.57 .58 - 2. PERFORMANCE t-1 0.28 0.38 -9.17 0.58 0.90* - 3. NETWORK VALUE 15.59 24.29 -0.37 140.08 0.16* 0.17* - 4. NETWORK VALUE’ 9.54 17.49 -0.07 105.38 0.18* 0.18* 0.91* - 5. FIRMS 3.28 0.66 2.00 5.00 -0.14* -0.15* 0.14* 0.03 - 6. POPULATION 26.48 25.57 3.87 82.54 -0.01 0.01 0.73* 0.57* 0.31* - *p < 0.01 Chapter 3. Strategic choices, network value and performance 113  3.5.2. Performance and Network Value The results of the estimations of the performance model are shown in Table 3.6. Model C.1 introduces the control variables and the lagged performance, whereas models C.2 and C.3 add network value and adjusted network value, respectively. The specification choice is based on a System GMM with first differences, a one-step estimation that is robust to heteroskedasticity and takes into account the potential endogeneity of the explanatory variables. To assess the validity of the System GMM estimators, I run the Arellano-Bond test for first-order and second-order serial correlation. Table 3.6 reports the significant m1 and insignificant m2 serial correlation statistics. This indicates that there is no second-order correlation in the level of residuals. The Hansen test is also reported and its non-significance validates the robustness of estimations. Lagged performance has a positive and significant influence on performance with a coefficient that is highly stable in the three estimations. This means that performance in the previous period positively influences current performance. This result justifies the use of the GMM estimator in this part of the analysis. Firm network value has, as expected, a positive and significant impact on performance (models C.2 and C.3), which supports Hypothesis 4. The variable firms has a negative and significant influence on firm performance as a result of increasing rivalry and year dummies are also statistically significant. Population does not seem to influence performance, except for model C.2 in which the influence is marginally negative. Chapter 3. Strategic choices, network value and performance 114  TABLE 3.6. PERFORMANCE AND NETWORK VALUE (SYSTEM GMM) (C.1) (C.2) (C.3) PERFORMANCEt PERFORMANCEt PERFORMANCEt NETWORK VALUE 0.002*** (3.99) NETWORK VALUE’ 0.003*** (2.97) PERFORMANCE t-1 0.467*** 0.467*** 0.464*** (20.56) (19.60) (19.75) FIRMS -0.047*** -0.041*** -0.036** (-3.04) (-2.92) (-2.46) POPULATION 0.001 -0.001* -0.001 (1.24) (-1.75) (-1.21) YEAR Dummies YES*** YES*** YES*** Constant 0.316*** 0.286*** 0.274*** (7.26) (7.48) (6.94) Number of observations 1,991 1,991 1,991 m1 -2.92*** -2.95*** -2.94*** m2 -0.04 -0.04 -0.04 Hansen Test 37.64 53.76 48.16 F-Test vs. 1 15.88*** 8.85*** t -statistics in parentheses * p < 0.10 ** p < 0.05 *** p < 0.01 Chapter 3. Strategic choices, network value and performance 115  3.6. DISCUSSION AND CONCLUSIONS This chapter contributes to the study of markets with network effects from a strategic perspective by introducing network value as a key concept. I have empirically tested a conceptual model in which the firm’s strategy may condition network effects and firm profitability through the three main elements that the literature has highlighted in network markets, i.e.: expectations, coordination and compatibility. This chapter, by focusing on firm-initiated actions to leverage network effects, has led to a greater understanding of firm-level strategy in network industries. Results reveal the importance of entry timing in markets with network effects. This result is highly consistent with previous findings (Gomez and Maicas, 2011; Usero and Fernández, 2009). Switching costs also appear as a key strategic tool that influences network value. High switching costs have been shown to dissuade the selection of a firm network by potential users with the subsequent negative effect on network value. Users distrust firms with high switching costs because they suspect that these firms will behave opportunistically (Mata et al., 1995), thus decreasing the effectiveness of network effects. Consequently, firms have to find a trade-off between creating high switching costs to retain their customers and being less aggressive so as to be perceived by potential customers as an appealing and trustworthy alternative. Contrary to what was expected, operating in various international markets is not a strategy that greatly influences users’ expectations and, thus, its impact on network effects is not significant. Chapter 3. Strategic choices, network value and performance 116  The explanation I can provide for this unexpected finding in the industry is threefold. First, while it is true that a number of mobile service providers are competing globally, users are restricted in their choices to companies operating in their local markets. In mobile telecommunications, users take into account only the network of the country where they live whereas, in other information industries such as software, hardware and online auctions, users do not perceive national boundaries in their decisions. Second, the internationalization of mobile operators could have become a strategic necessity. This seems to be clear from an analysis of the recent evolution of the industry in which the international diversification of the main operators has been quite similar. Finally, the availability of roaming services in all European countries, the similarity of roaming coverage and charges within operators, and the lack of complete information for users about roaming charges within the operators of the same international group (Salsas and Koboldt, 2004) may limit the existence of international network effects. Summarizing, although international network effects could exist in the industry, current market conditions do not favor them. This chapter also analyzes how network value is an element that is positively related to firm performance. The main premise is that users are willing to pay more for being part of a network with a larger installed base since the product does not provide any value by itself. The value comes from the communication ties that the network offers to users and this allows firms to increase the price of their product or service. Chapter 3. Strategic choices, network value and performance 117  Through the analysis of the above relationships, this research makes a contribution by offering a more accurate measurement of network value. Traditionally, network value has been considered to be proportional to network size. Although this can be reasonable, this chapter has added the intensity dimension to the traditional approach. I have adjusted previous measures by considering not only the firm’s own network, but also its rivals’ networks, that is, market competition is introduced into the assessment of network intensity and, thus, network value. Although the main findings do not substantially change, the adjusted measure I use shows a lower network value, which is perfectly understandable as it is considered the existence of other firms’ networks that reduce users’ utility since the probability of making off-net information exchanges with higher costs increases. Chapter 3 has several managerial implications. It recommends paying special attention to entry timing strategies in network industries. Firms should try to attract users to their network as soon as possible to gain competitive advantage. Because of this, it is not surprising to observe that bargain-then-rip-off strategies are very common in the first stages of market evolution as an adequate mechanism to attract users that will be exploited at a later stage. Thus, entry timing and price strategy have to be considered simultaneously when network effects are important. However, firms in these markets should be aware of not overexploiting their customers when lock-in is a likely market outcome. The perception of high switching costs may lead users to suspect that firms will behave opportunistically, which could result in fewer incentives to enter into a relationship with the firm. This chapter also has implications for Chapter 3. Strategic choices, network value and performance 118  managers about the international diversification of mobile operators. Apparently, international presence has no impact on network value, which, in my view, does not mean that firms need not pay attention to their international strategy, but rather that it may have become a strategic necessity to survive in the industry. It is important to note that the research setting refers to an industry in which the regulator plays a key role. For this reason, several policy implications can also be derived. Importantly, the effectiveness of FMA in the mobile communications industry depends on the winning of a license that is granted by national authorities and that is compulsory to compete for. Governments should be aware of the direct impact that their decisions have on competition in each local market. A reduced number of licenses or restrictive criteria to start an activity could reduce the number of competitors. This initial restriction could constitute an entry barrier in the future because a firm that cannot obtain a license at the first stage of competition will lose time in the market, which has been revealed as a valuable resource. Additionally, results show the important effect of switching costs in reducing network value and consumers’ welfare in network markets. Thus, the regulator should bear in mind that switching costs are a prevailing feature in the industry that can be harmful to customers’ interests. Indeed, in the context of mobile communications, the regulator has already recognized the importance of this dimension, reducing switching barriers and developing several measures to make switching easier and less costly. Mobile number portability is, perhaps, the most noteworthy effort in this direction and Chapter 3. Strategic choices, network value and performance 119  it has had, according to the literature, the desired effects (Lee, Kim, Lee and Park, 2006). To our knowledge, this chapter is one of the first attempts to empirically integrate network size and network intensity as part of network value into firm strategy. However, several issues deserve further attention. First, I use an adjusted measure of network value, which does not confer the same importance to all users and takes into account the market position of each firm as a source of different network intensities. However, while it is true that it is made an effort to incorporate several dimensions into my network value approach, the way in which I consider the tendency to make on-net communication only includes market shares and not price differences. Future research should try to improve the measure of network value with detailed data that reflects a more accurate dimension of the probability of making on-net over offnet connections by incorporating an explicit quantification of price discrimination. Although I take the existence of price discrimination as an issue, the inclusion of the degree of price discrimination as a source of network intensity and its evolution over time would improve the measure of network value. In the same vein, another possible extension would be to incorporate the existence of social network effects that reinforce network value. Users do not only select a firm because they believe it will be bigger than the others. Consumer behavior is also influenced by the previous decisions of the people who are socially related to them. Chapter 3. Strategic choices, network value and performance 120  Second, this chapter has taken a theoretical approach to refer to the three antecedents of network effects and network value, i.e. user expectations and coordination, and compatibility. Although they have been useful to build the theoretical foundations of the impact of strategic choices on network value, a deeper understanding and quantification of these elements would constitute a promising avenue for further research. Third, it has been shown that time in the market is an important determinant of network value. However, it would be interesting to analyze how this expectation of dominance of the first mover can be counteracted by late entrants and diminished over time. Although this chapter has focused on the network-dependent value of a firm, further analysis should study how the improvement of network-independent value by late entrants can reduce the network-dependent advantages of early movers. Finally, international presence has been shown not to have any significant impact on network value. Although some explanations have been put forward, a better understanding of how the internationalization process has influenced firm performance in these markets and become a strategic necessity is needed. The fact that various operators are competing simultaneously in the same markets would suggest the use of institutional or multimarket contact theories. Moreover, I have adopted a measure of the degree of internationalization that theoretically fits the mobile telecommunications industry. This measure assumes the existence of international network effects, but does not quantify them. With the aim of overcoming this limitation, further Chapter 3. Strategic choices, network value and performance 121  studies should try to develop additional measures of international diversification to the specific context of network industries with international network effects. Chapter 3. Strategic choices, network value and performance 122  REFERENCES  Amit R, Zott C. 2001. Value creation in e-business. Strategic Management Journal 22(6/7): 493 – 520. Arellano M, Bover O. 1995. Another look at the instrumental variables estimation of error-components models. Journal of Econometrics 68(1): 29-51. Arthur WB. 1989. Competing Technologies, Increasing Returns, and Lock-In by Historical Events. The Economic Journal 99(394): 116-131. Arthur WB. 1990. Positive Feedbacks in the Economy. Scientific American , 262(2): 92-99 Arthur WB. 1996. Increasing Returns and the New World of Business. Harvard Business Review 74(4): 100-108. Barkema HG, Drogendijk R. 2007. Internationalizing in Small, Incremental or Larger Steps? Journal of International Business Studies 38(7): 1132-1148. Bijwaard GE, Janssen MCW, Maasland E. 2008. Early mover advantages: An empirical analysis of European mobile phone markets. Telecommunications Policy 32(3-4): 246-261. Birke D, Swann P. 2006. Network Effects and the Choice of Mobile Operator. Journal of Evolutionary Economics 16(1/2): 65-84. Birke D, Swann P. 2010. Network effects, network structure and consumer interaction in mobile telecommunications in Europe and Asia. Journal of Economic Behavior & Organization 76(2): 153-157. Chapter 3. Strategic choices, network value and performance 129  Katz M, Shapiro C. 1994. Systems Competition and Network Effects. Journal of Economic Perspectives 8(2): 93-115. Kerin R, Varadarajan PR, Peterson RA. 1992. First-Mover Advantage: A Synthesis, Conceptual Framework, and Research Propositions. Journal of Marketing 56(4): 33-52. Kim H, Kwon N. 2003. The Advantage of Network Size in Acquiring New Subscribers: A conditional logit analysis of the Korean telephony market. Information Economics and Policy 15(1): 17–33. Klemperer P. 1995. Competition When Consumers Have Switching Costs: An Overview With Applications to Industrial Organization, Macroeconomics and International Trade. Review of Economic Studies 62(213): 515-539. Krafft J, Salies E. 2008. The diffusion of ADSL and costs of switching Internet providers in the broadband industry: Evidence from the French case. Research Policy 37(4): 706-719. Laffont JJ, Rey P, Tirole J. 1998. Network Competition: II. Price Discrimination. Rand Journal of Economics 29(1): 38-56. Lambkin M. 1988. Order of Entry and Performance in New Markets. Strategic Management Journal 9(5): 127-140. Lane V, Jacobson R. 1995. Stock Market Reactions to Brand Extension Announcement: The Effects of Brand Attitude and Familiarity. Journal of Marketing 59(1): 63-77. Lee D, Mendelson H. 2007. Adoption of Information Technology Under Network Effects. Information Systems Research 18(4): 395-413. Chapter 3. Strategic choices, network value and performance 130  Lee J, Kim Y, Lee JD, Park Y. 2006. Estimating the Extent of Potential Competition in the Korean Mobile Telecommunications Market: Switching Costs and Number Portability. International Journal of Industrial Organization 24(1): 107-124. Lieberman MB, Montgomery DB. 1988. First-mover Advantages. Strategic Management Journal 9(S1): 41-58. Liebenstein H. 1950. Bandwagon, Snob and Veblen Effects in the Theory of Consumers’ Demand. The Quarterly Journal of Economics 64(2): 183207. Maicas JP, Polo Y, Sese J. 2009. The role of (personal) network effects and switching costs in determining mobile users’ choice. Journal of Information Technology 24(2): 160-171. Majumdar SK, Venkataraman S. 1998. Network effects and the adoption of new technology: Evidence from the U.S. telecommunications industry. Strategic Management Journal 19(11): 1045-1062. Mata F, Fuerst W, Barney JB. 1995. Information Technology and Sustained Firm-Level Value: a Resource-Based Analysis. MIS Quarterly 19(4): 487505. McCloughan P, Lyons S. 2006. Accounting for ARPU: New evidence from international panel data. Telecommunications Policy 30(10-11): 521-532. McIntyre DP. 2011. Where there’s a way, is there a will? Installed base and product quality in a network industry. Journal of High Technology Management Research 22(1): 59-66. Chapter 3. Strategic choices, network value and performance 131  McIntyre DP, Subramaniam M. 2009. Strategy in Network Industries: A Review and Research Agenda. Journal of Management 35(6): 1494-1517. Merrill Lynch. 2010. Global Wireless Matrix 2009 . Ofcom. 2007. The Communications Market , United Kingdom. Ofcom. 2009. Review of mobile tariffs in the UK and US, Annex 9 in Wholesale Mobile Voice Call Termination , United Kingdom. Ofcom. 2011. Three response to mobile call termination supplemental consultation: the relationship between mobile termination rates, market share and competition . United Kingdom. Park S. 2004. Quantitative Analysis of Network Externalities in Competing Technologies: The VCR Case. The Review of Economics and Statistics 86(4): 937-945. Reed DP. 1999. That sneaky exponential: Beyond Metcalfe’s Law to the Power of Community Building. Context , Spring Issue available at: http://www.contextmag.com/archives/199903/digitalstrategyreedslaw.asp Rohlfs JH. 1974. Theory of Interdependent Demand for a Communications Service. The Bell Journal of Economics and Management Science 5(1): 16-37. Rohlfs JH. 2001. Bandwagon Effects in High Technology Industries . Cambridge, MA: MIT Publishing. Roodman D. 2006. How to Do xtabond2: An introduction to Difference and System GMM in Stata. Working Paper 103, Center for Global Development, Washington. Chapter 3. Strategic choices, network value and performance 132  Ross PE. 2003. 5 Commandments: Technology Laws and Rules of Thumb. IEEE Spectrum 40(12): 30-35. Salsas R, Koboldt C. 2004. Roaming free? Roaming network selection and inter-operator tariffs. Information Economics and Policy 16(4): 497-517. Sarkar MB, Cavusgil ST, Aulakh PS. 1999. International Expansion of Telecommunications Carriers: The Influence of Market Structure, Network Characteristics, and Entry Imperfections. Journal of International Business Studies 30(2): 361-382. Schilling MA. 2002. Technology Success and Failure in Winner-Take-All Markets: the Impact of Learning Orientation, Timing, and Network Externalities. Academy of Management Journal 45(2): 387-398. Shankar V, Bayus BL. 2003. Network Effects and Competition: an Empirical Analysis of the Home Video Game Industry. Strategic Management Journal 24(4): 375-384. Schmalensee R. 1982. Product Differentiation Advantages of Pioneering Brands. The American Economic Review 72(3): 349-365. Shapiro C, Varian HR. 1998. Information Rules: a Strategic Guide to the Network Economy. Harvard Business School, Boston. Shy O. 2001. The Economics of Network Industries . Cambridge University Press. Cambridge, United Kingdom. Shy O. 2002. A Quick-and-Easy Method for Estimating Switching Cost. International Journal of Industrial Organization 20(1): 71-81. Shy O. 2011. A Short Survey of Network Economics. Review of Industrial Organization 38(2): 119-149. Chapter 3. Strategic choices, network value and performance 133  Srinivasan R, Lilien GL, Rangaswamy A. 2004. First in, First out? The Effects of Network Externalities on Pioneer Survival. Journal of Marketing 68(1): 41-58. Stabell CB, Fjeldstad ØD. 1998. Configuring value for competitive advantage: on chains, shops, and networks. Strategic Management Journal 19(5): 413-437. Stango V. 2002. Pricing with Consumer Switching Costs: Evidence from the Credit Card Market. Journal of Industrial Economics 50(4): 475-492. Strike V, Gao J, Bansal P. 2006. Being Good while Being Bad: Social Responsibility and the International Diversification of US Firms. Journal of International Business Studies 37(6): 850-862. Suarez F. 2005. Network Effects Revisited: The role of ties in technology selection. Academy of Management Journal 48(4): 710-720. Swann P. 2002. The Functional Form of Network Effects. Information Economics and Policy 14(3): 417-429. Tanriverdi H, Lee CH. 2008. Within-industry Diversification and Firm Performance in the Presence of Network Externalities: Evidence from the Software Industry. Academy of Management Journal 51(2): 381-397. Usero B, Fernández Z. 2009. First come, first served: How market and nonmarket actions influence pioneer market share. Journal of Business Research 62(11): 1139-1145. Viard B. 2007. Do Switching Costs Make Markets More or Less Competitive? The Case of 800-Number Portability. Rand Journal of Economics 38(1): 146-164. Chapter 3. Strategic choices, network value and performance 134  World Bank Group. 2010. World Development Indicators Online .   Chapter 4. STRATEGIC CHOICES AND INSTITUTIONS IN THE FDI PROCESS: Where to enter  Resumen y conclusiones  238  La aproximación a las instituciones informales y formales se realiza sobre la base de dos conceptos claves. El primero de ellos se refiere a la distancia cultural , considerada como el conjunto de factores, como la diferencia en lenguaje, tradiciones, educación y códigos de conducta, que dificulta el flujo de información entre el mercado de origen y destino de la inversión (Johanson y Vahlne, 1977). El segundo concepto clave se refiere al desarrollo de las instituciones formales , es decir, al grado en el que las normas explícitas de un mercado – tales como la ley, los contratos, las sentencias, etc. – reducen las asimetrías de información y ayudan a una mayor protección de los derechos de propiedad entre las partes contratantes en una transacción económica (Meyer et al., 2008). Las hipótesis del Capítulo 4 plantean en qué medida la distancia cultural entre el país de origen y destino de la inversión y el grado de desarrollo de las instituciones formales del país de destino influyen sobre la probabilidad de que un grupo internacional entre en ese país. Así, la primera hipótesis del Capítulo 4 es que una mayor distancia cultural entre un país y el país de origen de un grupo internacional reduce la probabilidad de que dicho grupo entre en ese país. La segunda hipótesis del Capítulo 4 es que un mayor desarrollo de las instituciones formales de un mercado incrementa la probabilidad de que un grupo internacional entre en él. Finalmente, y como hipótesis clave en este Capítulo, se propone la consideración conjunta de las instituciones formales e informales a la hora de decidir la entrada de un grupo en un país. Cuando el país de destino de la inversión tiene instituciones formales más desarrolladas, el impacto negativo de la distancia cultural sobre la probabilidad de entrada del grupo en ese país pasa a ser menor. Resumen y conclusiones 239  Esto es así porque cuando las instituciones formales de un país apoyan la realización de intercambios económicos protegiendo los derechos de propiedad de las partes, el desconocimiento de las reglas informales derivado de la distancia cultural pierde importancia de cara a dificultar la consecución de las ventajas de internacionalización. El apartado empírico del Capítulo 4 incluye la totalidad de decisiones de entrada que todos grupos internacionales de telefonía móvil realizaron entre 2000 y 2010 en todos los países del mundo para los que existen índices institucionales formales e informales. Se confirma que la distancia cultural influye negativamente en la probabilidad de que una multinacional entre en un mercado. Asimismo, se observa que el desarrollo de las instituciones formales no tiene un efecto directo significativo. Pero sí que lo tiene de forma indirecta, en la medida que un mayor desarrollo de las instituciones formales incrementa la probabilidad de que un grupo internacional entre en un país que guarda una mayor distancia cultural con su país de origen. Se confirma, por tanto, ese efecto moderador de las instituciones formales sobre la relación entre las instituciones informales y la decisión de entrada. El análisis conjunto de las instituciones formales e informales es una de las principales contribuciones del Capítulo 4 puesto que la literatura anterior ha tendido a considerar separadamente el papel de las instituciones formales e informales sobre las decisiones de entrada de los grupos. Como segunda contribución del capítulo 4 cabría destacar el propio análisis empírico que se realiza, al tener en consideración la totalidad de países a nivel mundial en los que se han producido entradas Resumen y conclusiones  240  de todos los grupos internacionales de una misma industria durante un periodo de 11 años. El Capítulo 4 ha tomado en consideración la perspectiva institucional de la estrategia empresarial en su contexto de aplicación tradicional, como es el del proceso de internacionalización. Sin embargo, la perspectiva institucional de la estrategia empresarial aboga por determinar el impacto del marco institucional en otros ámbitos de investigación, y no sólo en el análisis del proceso de internacionalización por parte de los grupos internacionales. Por esa razón, el Capítulo 5, “Instituciones y resultados después de un cambio tecnológico radical: Cómo el valor de los recursos complementarios especializados varía entre mercados” , ha tratado de integrar la literatura sobre la perspectiva institucional de la estrategia con la referente a discontinuidades tecnológicas. El Capítulo 5 se sitúa en el contexto posterior a un cambio tecnológico radical en el que las empresas establecidas en la tecnología anterior han invertido en la nueva tecnología. Bajo este contexto, se analiza cómo los recursos de la cadena de valor de las empresas establecidas que sirven para la comercialización de la tecnología anterior continúan siendo valiosos para la comercialización de la nueva tecnología. Recursos tales como la reputación, los canales de ventas y distribución, el servicio postventa, los vínculos con los usuarios o el conocimiento de la industria conservan su valor y, en la medida que resultan difícilmente imitables por los nuevos entrantes, son capaces de conferir un mayor beneficio a las empresas establecidas sobre los nuevos competidores (Mitchell, 1991; Tripsas, 1997; Rothaermel, 2001). Esa relación positiva entre la posesión Resumen y conclusiones 241  de estos recursos complementarios y el beneficio de la empresa tras un cambio tecnológico radical es la primera hipótesis de este capítulo. La contribución principal de este Capítulo 5 descansa en la segunda hipótesis que propone el efecto moderador del marco institucional sobre la intensidad de la ventaja de las establecidas. Dado que las empresas establecidas mantienen el valor de los activos complementarios en la nueva etapa tecnológica, también mantienen las relaciones de confianza establecidas con los agentes del mercado con los que han de interactuar para hacer uso de esos activos, tales como usuarios, proveedores, inversores, empleados o autoridades públicas. El capítulo propone que el valor de los recursos complementarios, al garantizar el mantenimiento de la red informal de relaciones de la empresa, es mayor en contextos de menor desarrollo de las instituciones formales. Esta afirmación se sustenta sobre la base de que, a falta de organismos que faciliten el intercambio de información y garanticen los derechos de propiedad derivados de la ley y los contratos, los agentes del mercado preferirán realizar intercambios económicos con empresas que ya conocen con anterioridad. Así podrán reducir las asimetrías de información y los posibles comportamientos oportunistas. Los nuevos entrantes, por tanto, se encontrarán en una mayor desventaja con respecto a las empresas establecidas en estos entornos de instituciones formales débiles al carecer de esa red previa de relaciones informales. El análisis empírico desarrollado sobre 46 mercados de todas las regiones mundiales confirma el efecto positivo que los recursos complementarios tienen sobre el beneficio empresarial. Asimismo, se observa que a mayor Resumen y conclusiones  242  desarrollo de las instituciones formales de un mercado, menor es el impacto positivo de esos recursos sobre la obtención de resultados. Por tanto, los mismos recursos estratégicos en dos contextos institucionales diferentes no conllevan la consecución de una misma ventaja competitiva. El Capítulo 5 tiene tres contribuciones clave. En primer lugar, aplica la perspectiva institucional de la estrategia en un contexto de investigación novedoso, como es el de las discontinuidades tecnológicas. En segundo lugar, ofrece una evidencia adicional acerca de la importancia de los recursos complementarios en la consecución de una ventaja competitiva por parte de las empresas establecidas tras un cambio tecnológico radical. Finalmente, tal y como ocurría en el Capítulo 4, se ofrece un análisis empírico basado en una muestra de mercados que cubre todas las regiones mundiales, lo que permite un mayor grado de generalización de los resultados. Resumen y conclusiones 243  CONCLUSIONES A continuación se detallan las principales conclusiones de los capítulos que abordan los objetivos de investigación propuestos en el Capítulo 1 y que analizan empíricamente las hipótesis planteadas en el sector de la telefonía móvil, descrito en el Capítulo 2. Conclusiones del Capítulo 3: El papel de la estrategia en mercados con efectos de red La estrategia en industrias con efectos de red debe de estar enfocada a la gestión del valor de la red, que se convierte en un determinante clave del beneficio frente a otro tipo de industrias en las que el valor reside en el producto. Las empresas pueden intensificar los efectos de red en su propio beneficio para así generar un mayor valor de red a través de decisiones estratégicas que influyan sobre las expectativas y coordinación de los usuarios y la compatibilidad percibida de su red con la de empresas rivales. En concreto, se observa que el tiempo que una empresa lleva operando en el mercado incrementa el valor de su red. Por el contrario, altos costes de cambio reducen el valor de la red al desincentivar la selección de la red por parte de los usuarios potenciales. La dimensión estratégica del valor de la red se plasma en la medición alternativa que propone la presente tesis doctoral. Mientras medidas anteriores equiparaban el valor de la red al tamaño de la misma, la medida alternativa que se propone en la tesis resulta una función no sólo Resumen y conclusiones  244  del tamaño de la red sino también de la intensidad de los efectos de red. Dicha intensidad se hace depender del tamaño relativo de la red de la empresa con respecto al de sus rivales en un mercado específico. La consideración de la existencia de competidores en la medida del valor de la red permite otorgar a ésta un marcado carácter estratégico. La tesis avanza en el análisis del papel de la estrategia en industrias de red. Primero, a través de la construcción de un modelo que explica el impacto de cualquier decisión estratégica sobre el valor de la red y el beneficio. Segundo, mediante la construcción de una medida alternativa a la tradicional del valor de la red que incorpora la dimensión estratégica del mismo. Conclusiones del Capítulo 4: El carácter complementario de las instituciones formales e informales Las decisiones estratégicas de las empresas son entendidas como el resultado de la interacción dinámica entre la organización y su entorno institucional. En el marco del proceso de internacionalización, las multinacionales tienen que tomar una decisión estratégica clave como es la selección del país destino de la inversión extranjera directa. La tesis observa que la distancia cultural y el desarrollo de las instituciones formales determinan, de forma simultánea, la probabilidad de que un grupo internacional entre en un país o no. Las diferencias en cultura, religión, códigos de conducta o lenguaje entre el mercado de origen de una multinacional y otro mercado dificultan la Resumen y conclusiones 245  obtención de las ventajas esperadas de la internacionalización. Por ejemplo, la transferencia de capacidades y know-how de la matriz a la subsidiara es más difícil cuando descansan en rutinas que no son comprendidas por los trabajadores de la subsidiaria. La multinacional también tiene problemas para entrar en la red de negocios de otro país dada la dificultad para desarrollar vínculos con los inversores, trabajadores o proveedores de ese mercado. Ello dificulta la adquisición de recursos financieros y humanos valiosos. Por tanto, los grupos internacionales prefieren entrar en países cuyas instituciones informales son más semejantes a las de su país de origen. No obstante, las instituciones formales del país de destino de la inversión (p.ej., sistema legal, administrativo y judicial) pueden garantizar el cumplimiento de los contratos entre la multinacional y los otros agentes del mercado (trabajadores, proveedores, inversores, etc.) y facilitar la entrada de la multinacional en la red de negocios del mercado (por ejemplo, a través de la creación de intermediarios tales como bancos, cámaras de comercio, etc.). En tal caso, la distancia cultural pierde importancia como elemento restrictivo de la decisión de entrada en un país por parte de las empresas multinacionales. Por tanto, el regulador que desee fomentar la entrada de capital extranjero para incrementar la competitividad del mercado debe de reforzar los sistemas legales, administrativos y judiciales para que sean capaces de garantizar el cumplimiento de los contratos y la apertura del mercado a la competencia. Resumen y conclusiones  246  La tesis avanza en la aplicación de la perspectiva institucional de la estrategia a través de la consideración conjunta del papel de las instituciones formales e informales en una decisión estratégica clave, tal y como es la selección del país de destino de la inversión por parte de una multinacional. Conclusiones del Capítulo 5: La perspectiva institucional aplicada a la dirección estratégica de innovaciones radicales El desarrollo de las instituciones formales facilita la entrada de nuevos competidores en un mercado que ha sufrido un cambio tecnológico radical puesto que proporciona una mayor comprensión del funcionamiento del mercado, aumenta la información disponible y garantiza la protección de los derechos de propiedad derivados de los contratos. Se observa que cuando las instituciones formales fallan a la hora de reducir los riesgos contractuales, los nuevos entrantes se encuentran en desventaja con respecto a las empresas que proceden de la etapa tecnológica anterior y deciden invertir en la innovación radical. La posesión de recursos complementarios que ayudan a la comercialización de la innovación radical se convierte en una fuente de ventaja competitiva para las empresas establecidas en la tecnología anterior. Dichos recursos, situados en la parte inferior de la cadena de valor, ayudan a estas empresas a conservar las relaciones con los agentes del entorno tales como usuarios, proveedores, inversores, trabajadores y Resumen y conclusiones 247  gobiernos. Estos lazos informales son capaces de reducir las asimetrías de información y el riesgo contractual a falta de instituciones formales que se encarguen de ello. En consecuencia, los recursos de las empresas establecidas en la tecnología anterior tienen un mayor valor en entornos con bajo desarrollo de las instituciones formales en la medida que ayudan a la comercialización de la innovación y a mantener las relaciones con los agentes del entorno que reducen los riesgos contractuales. La tesis avanza en la aplicación de la perspectiva institucional de la estrategia empresarial considerando el carácter moderador de las instituciones formales de un mercado sobre la intensidad de la ventaja competitiva de las empresas establecidas.