Organizational dynamics and global integration A perspective from subsidiary managers

Transcription

1 Journal of International Management 8 (2002) Organizational dynamics and global integration A perspective from subsidiary managers Yadong Luo* Department of Management, School of Business Administration, University of Miami, 414 Jenkins Building, Coral Gables, FL , USA Abstract This study establishes a relationship between organizational dynamics and overall integration within MNEs from the perspective of foreign subsidiaries in a dynamic environment. While previous empirical studies have emphasized that overall integration is determined by environmental and industrial pressures, we propose that integration is also affected by strategic capabilities, organizational infrastructures and strategic needs of foreign operations. Analysis of survey data including 168 foreign subsidiaries in a large emerging foreign market suggests that the resource distinctiveness committed to local operations is positively associated with overall integration, while the strength of networking with local businesses and governments is positively associated with overall responsiveness. The effectiveness of established information flow and coordination between a focal subsidiary and the rest of the network has a favorable effect on integration. Moreover, strategic needs in exploitation of factor endowment contribute to higher levels of integration. These needs also moderate the relationship between dynamic capability and overall integration. D 2002 Elsevier Science Inc. All rights reserved. Keywords: Global integration; Organizational characteristics; National adaptation 1. Introduction Global strategies are most frequently analyzed using the global integration local responsiveness (I-R) paradigm (Doz, 1976; Doz and Prahalad, 1991; Prahalad, 1975; Prahalad and Doz, 1987). A number of studies have extended various aspects of this framework including * Tel.: ; fax: address: yadong@miami.edu (Y. Luo) /02/$ see front matter D 2002 Elsevier Science Inc. All rights reserved. PII: S (02)

2 190 Y. Luo / Journal of International Management 8 (2002) structural determinants (Birkinshaw et al., 1995; Kobrin, 1991; Porter, 1986; Yip, 1995), operational flexibility (Bartlett and Ghoshal, 1989; Kogut, 1985), subsidiary mandates and initiatives (Birkinshaw, 1996, 1997; Roth and Morrison, 1992) and strategic groups (Johnson, 1995; Jarillo and Martinez, 1990; Roth and Morrison, 1990; Taggart, 1998). The I-R framework and these subsequent studies commonly assume that integration and responsiveness are primarily determined by environmental forces. 1 Pressures for integration include environmental and industrial forces that necessitate worldwide business resource deployment and global integration of dispersed businesses across national boundaries (Ghoshal, 1987; Prahalad and Doz, 1987; Roth and Morrison, 1990; Yip, 1995). Pressures for local responsiveness are environmental and industrial forces that necessitate local context-sensitive strategic decision-making and quick responses to each local market or industrial setting (Bartlett and Ghoshal, 1989; Doz, 1976; Prahalad and Doz, 1987). Compared with the above line of research, an empirical analysis of organizational dynamics in relation to global integration is inadequate. This is an important issue because an MNE s ability to globally expand and succeed is constrained by administrative heritage, that is, organizational capability and internal infrastructure (Bartlett and Ghoshal, 1988). Such a heritage, together with the strategic goals and environmental factors, jointly shape the simultaneous balance between global coordination and national flexibility (Bartlett and Ghoshal, 1989). I-R imperatives are thus influenced by organizational dynamics in addition to environmental and industrial forces. This study suggests that the optimal level of overall integration, as perceived by MNE subsidiary managers in a dynamic foreign environment, is affected by internally differentiated dynamic corporate capabilities, organizational infrastructure and strategic needs. These organizational dynamics explain global integration heterogeneity among different businesses confronting similar contextual contingencies. The primary purpose of this study is to determine specific organizational characteristics that are associated with integration. By integrating the I-R framework with resource-based theory, we address whether, and to what extent, the degree of overall integration of a foreign subsidiary in a dynamic market can be explained by an MNE s strategic capabilities (foreign experience, networking strength and resource distinctiveness), organizational infrastructure (information flow, coordination systems and resource flow) and strategic intents, as pertinent to that subsidiary. While these organizational dynamics have received some attention in global business and management literature (Birkinshaw and Hood, 1998; Egelhoff, 1988; Hedlund, 1993; Kostova, 1999), inadequate efforts have been made to link them with I-R imperatives. Because we are concerned with the I-R balance, we investigate overall integration rather than integration with respect to specific activities. Specific integration is 1 The impact of environmental and industrial forces on global integration has been well documented. For instance, Kobrin (1991) finds that technological intensity, advertising intensity and transnationality are all structural determinants, which drive global integration. Birkinshaw et al. (1995) recently observed the significant impacts of market demand, national advantage and competitive action on the degree of global integration. Yip (1995) categorizes structural determinants, which drive global integration as market, cost, governmental and competitive forces. Of these, the market structural driver appears most relevant in explaining the trade-off between integration and responsiveness.

3 Y. Luo / Journal of International Management 8 (2002) not a unidimensional construct, as it varies among different activities within the same firm. As such, MNEs using a transnational and multifocal strategy can simultaneously achieve integration and responsiveness by internally differentiating various activities and by idiosyncratically deploying resources to overseas subunits that hold different strategic importance to the firm (Bartlett and Ghoshal, 1989; Roth and Morrison, 1990). In this case, an MNE may strongly integrate some activities of a subunit while maintaining high responsiveness regarding other areas of that subunit. Unlike specific integration, overall integration can be quantitatively seen as a spectrum with very high overall integration (global strategy) and very high overall responsiveness (multidomestic strategy) serving as the endpoints and with transnational strategy standing between. This study advances our understanding of global integration in several ways. From the theoretical viewpoint, previous studies often use the information processing theory to underpin their propositions and analyses. From this theoretical perspective, different environments impose varying information costs in scanning, analyzing and interpreting environmental conditions as well as in interacting, responding and changing such environmental conditions (Hannan and Freeman, 1989). As a result, MNEs are likely to maintain different global integrations and employ different global strategies when they face different global environments (Yip, 1995). The present study extends this view by including the resource-based theory in an illustration of how global integration is influenced by organizational dynamics such as strategic capabilities and administrative heritages. As a critical component of global strategy, cross-border integration is shaped by an MNE s existing resources as well as the firm s needs for developing new resources. Resource-based view explains that MNEs in the same environment are still likely to use different strategies because of idiosyncratic resources and capabilities. This notion has been virtually missing in previous studies. From the practical viewpoint, organizational dynamics are becoming more critical than ever before to the coordination of geographically dispersed businesses and the achievement of strategic benefits from integration. This is because the competitive environment is increasingly globalized, which more vigorously necessitates capability transfer among subsidiaries in different countries and requires capability building within a globally integrated network (Luo, 2000). The integration process cannot be isolated from an MNE s capability and infrastructure such as global experience, competitive advantages and corporate systems in coordination, information flow and resource flow. In designing and maintaining global integration, MNEs should not only align appropriately with external pressures, including institutional and structural factors, but also subtly configure with internal forces such as organizational capabilities, infrastructures and needs. Because of the heterogeneity of businesses, markets and functions among MNE subsidiaries in different foreign countries, this study examines the above issue at the level of foreign subsidiaries within a major emerging foreign market. Many MNEs have established a large array of projects there, engaging in increasingly sophisticated activities at various locations using multiple entry modes. The integration of those activities that need to be managed simultaneously geographic, product, market and technological with the rest of the MNE network has already caught managerial attention.

4 192 Y. Luo / Journal of International Management 8 (2002) Theory and hypotheses 2.1. Theoretical underpinnings Over the past decade or so, the need to balance the dynamic tension between integration and adaptation has acquired increasing strategic significance towards the fulfillment of a multitude of purposes (Ghoshal, 1987). Perlmutter (1969) defines global strategy in terms of managers mind-sets, contrasting ethnocentric (home country), polycentric (host country) and geocentric (hybrid) orientations. Prahalad and Doz (1987) extends this view and offers a more systematic conceptualization for examining global strategy. Their framework, namely the global integration (I) and local responsiveness (R) paradigm, suggests that participants in global industries develop competitive postures across two dimensions (I and R). These two dimensions represent two salient imperatives that simultaneously confront a business competing internationally. MNEs can choose to emphasize one dimension over another or compete in both dimensions, resulting in three basic strategies: integrated (global), multifocal (transnational) and locally responsive (multidomestic). Globally integrated businesses link activities across nations in an attempt to minimize overall costs, avoid various taxes or maximize income. Locally responsive businesses perceive pressures to respond strategically to local needs. Lastly, multifocal businesses perceive the need to respond simultaneously to pressures both for integration and for responsiveness. The degree of overall integration increases along multidomestic, transnational and global strategies. The I-R balance is determined by multilevel factors, ranging from individual and firm levels to industry and sociopolitical levels, which can be further grouped into organizational factors and environmental factors. While the effect of environmental factors on the degree of integration have been investigated by previous research, the influence of organizational factors has not yet received adequate attention. Murtha et al. (1998) suggest that both organizational and environmental factors can enhance or inhibit competitive advantage in the global marketplace. Internationalization and international competition are cognitive processes that extend forward from a fit between strategy and structure. In such cognitive processes, organizational dynamics such as strategic capabilities and information flow infrastructure embedded in a globally integrated yet internally differentiated network are increasingly critical to the achievement of financial arbitrages and operational synergies (Birkinshaw et al., 1998; Gates and Egelhoff, 1986; Govindarajan, 1988; Kogut, 1985; Murtha et al., 1998). Organizations should make differentiated responses to diverse environments while maintaining a system of integrated actions across environments (Lawrence and Lorsch, 1967; Ghoshal and Nohria, 1989). The key aspects of international strategic capabilities derive from managers cognitive processes that balance competing country, business and functional concerns (Hedlund, 1993; Murtha et al., 1998; Prahalad and Doz, 1987). An MNE s capability in exploiting internalization and internationalization advantages depends on managerial mind-sets or what Bartlett and Ghoshal (1989) called transnational mentality that equilibrate integration and responsiveness. Transnational firms can be represented as an integrated network that entails significant cross-border flows of components, products, resources, people and information.

5 Y. Luo / Journal of International Management 8 (2002) Organizational capability, administrative heritage and managerial mentality can influence the design and outcome of global strategies. Although this study seeks to emphasize firm-level determinants and was not designed to examine individual-level variables affecting global integration, we do realize the importance of the latter in shaping the actual level of integration. As such, in diagnosing the influence of organizational dynamics on integration, we included managers networking strength with local businesses, a firm s foreign experience and its information flow system in the model, three variables (not exclusive though) being associated with managers roles in shaping overall integration. It is a particularly worthy effort in the future to investigate the extent to which managerial mentality (especially foreign subsidiary managers) affects the level and structure of global integration. Given different organizational dynamics, different firms may have idiosyncratic abilities to cope with environmental conditions by reducing their dependence on or increasing their control over external resources (Pfeffer and Salancik, 1978). This perspective is important when diagnosing the differentiation integration balance for MNEs because complex organizations are characterized by structural indeterminacy, internal differentiation, fuzzy boundaries and business multidimensionality (Doz and Prahalad, 1991). Given such complexity and heterogeneity, MNEs need to establish a strategic infrastructure, which can quickly adapt to external hazards and contextual changes. Strategic infrastructure is a multidimensional system containing strategic resources or capability and organizational infrastructure, which provide a foundation for global expansion and latent linkages within the network. 2 As a major organizational system for fulfilling global integration, strategic infrastructure helps ensure the evolutionary development of sustainable competitive advantages and generates new bundles of resources from an interplay between location-specific and competitive-specific advantages (Bartlett and Ghoshal, 1988; Kogut and Zander, 1992; Tallman, 1991). MNEs are highly heterogeneous organizations, which must make optimal trade-off for different businesses, countries, functions and tasks as a function of the whole range of economic and sociopolitical characteristics that differ between countries and affect individual businesses and tasks in varied ways (Doz and Prahalad, 1991). When an MNE s boundaries are fuzzy, a conventional organizational structure is unable to satisfy the internal need for ecological evolution within its interorganizational network (Egelhoff, 1988). In this situation, a strategic infrastructure is key to the coordination and integration of geographically dispersed businesses while maintaining internal differentiation and local responsiveness amongst individual subunits. The I-R paradigm holds that today s operating environment requires MNEs to link their strategic resources and organizational infrastructure in a way that would allow them to leverage their capabilities for achieving integration and responsiveness simultaneously (Bartlett and Ghoshal, 1988; Prahalad and Doz, 1987). Complementary to this notion, resource-based theory contends that such resources and infrastructure are both 2 Taking resources beyond their role as static sources of inimitable advantages, strategic capabilities allow a firm to diffuse, deploy, use and create organizationally embedded resources in order to attain a sustained competitive advantage. Such abilities are difficult to copy or replace (Wernerfelt, 1984). They can be bundled with a variety of hard assets and skills to encourage and permit the growth of a firm. Strategic capabilities require the capacity to learn and develop new capabilities or extract rents from current resources.

6 194 Y. Luo / Journal of International Management 8 (2002) organizationally embedded and essential determinants for generating economic rents (Prahalad and Hamel, 1990). They affect a firm s operational flexibility, resource and information flows and organizational learning (Barney, 1991; Wernerfelt, 1984). Strategic capability and organizational infrastructure have strong implications for an MNE s internalization and integration. They jointly impact a firm s path of evolution (Chang, 1995) and pattern of internal governance (Teece, 1985). Bartlett and Ghoshal (1988) view such capability and infrastructure as the prerequisite for the transnational solution, which requires both effective corporate management that does not hinder national flexibility and efficient country management that does not prevent global coordination. Teece et al. (1990) suggest that many MNEs are not lacking in strategic clarity about the need for the I-R balance but are instead short of the strategic infrastructure to implement it. Without this infrastructure, integration may increase organizational rigidity and obstruct operational flexibility (Kogut, 1985). It may further impede lateral collaboration among network subsidiaries (Egelhoff, 1988) or hinder subunit initiatives in a dynamic environment (Birkinshaw et al., 1998). For MNE subunits in a foreign market, strategic capability is the backbone of their sustained competitive advantage (Tallman, 1991) while organizational infrastructure is the ensuring device for integrating their operations with the rest of the network (Teece, 1985). Having strategic capability without appropriate organizational infrastructure leaves MNEs unable to internalize cross-border businesses. Having organizational infrastructure without necessary strategic capability makes subunits unable to fulfill their strategic goals and achieve economic benefits from international expansion. In fact, the dynamic capability perspective of the resource-based theory sees integration as the means, rather than the ends, of optimizing resource structure and allocation and maximizing gains from international expansion and evolution (Teece et al., 1990). As detailed in the hypothesis development, we propose that strategic infrastructure and organizational needs together constitute the organizational dynamics that influence the optimization of integration and responsiveness. Overseas subunits have heterogeneous goals set by headquarters. This heterogeneity helps satisfy diverse organizational needs of the entire MNE network. Ghoshal and Nohria (1989) argue that global integration is a function of organizational needs, which internally differentiate various subunits within an MNE s intraorganizational network. Similarly, Prahalad and Doz (1987) assert that organizational needs associated with the target market predict a subsidiary s dependence on local resources and its vulnerability to the hazards of that market, which in turn influence the level of required overall integration. From the resource-based view, a firm s competitive edge and market power are determined not only by its rent-generating resources (static view) but also the optimal way to allocate and utilize these resources (dynamic view) (Wernerfelt, 1984). In the equation of resource allocation, an organizational need is one of the major dynamics with which resource allocation must properly configure (Chang, 1995; Collis, 1991). Because strategic infrastructure should be built along the direction of organizational needs, such infrastructure and needs cannot be separated from each other when they are linked to integration. Collectively, the conditions of a strategic infrastructure and the underlying objectives specific to a geographic location determine what, where and how an MNE subunit should respond to indigenous needs while maintaining integration with the rest of the system

7 Y. Luo / Journal of International Management 8 (2002) in a manner consistent with the parent firm s global strategy (Bartlett and Ghoshal, 1989; Ghoshal, 1987; Roth and Morrison, 1990). Organizational dynamics embedded in the network are the basis for an evolving system by which MNE headquarters monitor complex, diversified and heterogeneous businesses located in different national markets (Govindarajan, 1988). This system spurs operational flexibility, transnational coordination, knowledge transfer and information flows among various units of the entire network (Ghoshal and Nohria, 1989; Kogut, 1985). Strategic capability, organizational structure and organizational needs each contain several components. Chang (1995) finds that resource distinctiveness (or knowledge tacitness) is a key component of strategic capability and determines organizational commitment in a global context. Johanson and Vahlne (1977), Chang (1995) and Roth et al. (1991) demonstrate that strategic capability is also manifested by foreign experience that mitigates liabilities of foreignness and newness and influences the path and pattern of international evolution. When a local environment becomes more dynamic, the criticality of such experience amplifies (Luo and Peng, 1999). Recently, Nahapiet and Ghoshal (1998), Kostova (1999) and Luo and Peng (1999) propose a new element of strategic capability in a foreign market, namely, networking skills with external stakeholders, which may have strong implications for competitive advantages and performance. As the critical part of social capital (Burt, 1997), interfirm and intrafirm networking skills are individually or organizationally embedded tacit knowledge predicting variance of global business outcomes and the desired level of global integration. According to resource-based theory, the above knowledge, experience and skills are building blocks for organizational learning and capability building and affect resource allocation and commitment to specific subunits overseas (Prahalad and Hamel, 1990; Tallman, 1991). 3 They are imperative for firms trying to reduce the liability of foreignness, mitigate vulnerability to contextual variabilities and explore market opportunities (Kogut and Zander, 1992; Prahalad and Hamel, 1990). In balancing integration and responsiveness, MNEs should appraise their strategic capabilities with respect to the needs of foreign subsidiaries. Several studies have argued that an organizational infrastructure, embedded in an internally differentiated MNE network, is composed of an information flow system, a coordination mechanism and a resource flow system (see Ghoshal, 1987; Ghoshal and Nohria, 1989; Gupta and Govindarajan, 1991). Efficient internal flows of information and resources under a wellconstructed coordination system are the foundation upon which resource misallocation can be avoided (Bartlett and Ghoshal, 1988) or dynamic capability is elevated (Teece et al., 1990). As competitors increasingly achieve parity of access to resources from various parts of the world, sources of competitiveness shift from location-specific to firm-specific factors, including the overall organizational ability to coordinate resource use in response to short-term opportunities arising worldwide (Doz and Prahalad, 1991). Information flows, both formal and 3 Although dynamic capability may include additional elements besides knowledge, experience and networking skills, this study proposes that these three elements are important organizational dynamics associated with global integration, as detailed in the hypotheses development. Those elements not associated with integration are not considered in this study.

8 196 Y. Luo / Journal of International Management 8 (2002) informal, are key to obtaining a competitive advantage in the global market and are therefore a predominant element of an organizational infrastructure, which supports the I-R balance. Many MNEs are shifting their managerial emphasis from the physical infrastructure (e.g., prespecified organizational structure) and resource deployment to information processing networks and resource mobilization (Bartlett and Ghoshal, 1989). An advanced information flow system, together with appropriate resource flow and coordination mechanisms within a network, reduces the transaction costs of internalized businesses and bolsters a decentralized self-structuring process imperative to fuzzily bounded and latently linked MNEs. Whether or not an MNE should and can maintain high integration or high responsiveness associated with particular businesses depends on interrelated systems of information, coordination and resource flows (Roth et al., 1991) Hypotheses development Strategic capability Foreign experience. When experience in a host market increases, a foreign firm may make a greater commitment to economic activities in this market by elevating its responsiveness or reducing integration. Decentralization of experienced subsidiaries boosts their market adaptability and customer responsiveness and heightens rents yielded from localized learning (Chang, 1995). The Uppsala process model suggests that accumulated knowledge about country-specific markets, practices and environments helps firms increase their indigenous commitment, reduce operational uncertainty and enhance economic efficiency (Johanson and Vahlne, 1977). A subsidiary with greater experience is more likely to commit resources, knowledge and investment to local operations (Chang, 1995). This configuration reduces financial risks and operational uncertainty in a volatile environment such as China where lack of experience is an important obstacle to market expansion. By contrast, firms with little experience may need high integration because it can reduce a subsidiary s vulnerability to the contextual hazards precipitated by a host country s institutional and task environments (Miller, 1992). Such hazards in China, for instance, are generally beyond organizational control, thus calling for internalization. Internalization is an effective mechanism for attenuating risk propensity and economic exposure to such an environment (Root, 1988). Firms unfamiliar with a host market thus need to integrate their overseas businesses through parental contribu- 4 Organizational behavior and decisions are certainly not independent of environmental factors (Lawrence and Lorsch, 1967). Nevertheless, the organizational dynamics we examined such as strategic infrastructure and needs are something that have already existed and cannot be changed overnight or by decree. Thus, they are what Bartlett and Ghoshal (1988) called administrative heritage. The important managerial task is to make this heritage more responsive to the ever-changing environment and external demands including the integration/ responsiveness. This study controls for environmental factors when empirically assessing organizational dynamics in relation to integration.

9 Y. Luo / Journal of International Management 8 (2002) tions of tacit knowledge and internalization of financial risks and operational uncertainties. Therefore, Hypothesis 1: In a dynamic foreign market, the degree of overall integration as perceived by subsidiary managers is negatively associated with an MNE s foreign experience in this market Networking strength. Many emerging markets, most notably China, have a long tradition of doing business based on interpersonal relationships with managers in partner, supplier, buyer, distributor and competitor firms and governmental authorities. From the sociological perspective, organizational behavior is embedded in concrete, ongoing systems of social relations; one s behavior is rarely explicable without reference to the persistent effects of interactions with others and the overall patterns of group interactions (Granovetter, 1995). As a source of relational capital, managerial ties with other firms and governments counter external threats and fill in resource deficiency. These ties help foreign businesses stabilize internal operations, mitigate the external pressures of culturally specific business practices and substitute organizational weaknesses. In the absence of local ties, local responsiveness will not be able to create sustained benefits from emerging opportunities. Firms having established managerial connections with business community are more likely to be locally responsive because they will seek benefits from their already committed social capital (Burt, 1997). Established ties with governmental officials further stimulate responsiveness or curtail centralization because they allow the subunit to capitalize on its social investment by gaining access to scarce resources, regulated industries and state-owned distribution channels. When a foreign firm has already committed resources to local operations, intraorganizational activities within an MNE network may yield to interorganizational exchanges with local businesses. A strong established network with local businesses and government officials signals a foreign company s long-term commitment to host country operations and its intention to utilize indigenous factor endowments and preempt local market opportunities. When a subsidiary s operations and management becomes more institutionalized within the host environment, its interdependence with the rest of the MNE network decreases (Doz and Prahalad, 1981; Gupta, 1987). Therefore, Hypothesis 2: In a dynamic foreign market, the degree of overall integration as perceived by subsidiary managers is negatively associated with the extent to which they have established ties with managers of local businesses and government officials Resource distinctiveness. MNEs often face a dilemma when operating in an emerging market such as China: they need to contribute distinctive resources (e.g., technology, know-how, brand names, trademarks, copyrights, patents, etc.) in order to overcome liability of foreignness, but these resources lack property rights protection. Legal systems for protecting industrial or intellectual property rights in these countries are neither sufficiently established nor enforced. In this situation, a firm unnecessarily exposing its critical resources may provide local firms with an advantage in the future that will contribute

10 198 Y. Luo / Journal of International Management 8 (2002) to losing its own sustained competitive edge (Collis, 1991). High integration will better protect a firm s tacit knowledge and strategic resources. Internalization protects ownershipspecific assets and creates the maximum payoff from ownership-specific advantages through the deployment of strategic assets without giving rise to uncompensated leakage to the partner or other local businesses. It also improves expertise in generating the economic rents gained by allocating assets in a new territory (Baliga and Jaeger, 1984; Dunning, 1981; Hennart, 1989). The greater the tacitness of proprietary knowledge or the perceived economic rents to be gained from such knowledge, the greater the need to maintain global integration. Therefore, Hypothesis 3: In a dynamic foreign market, the degree of overall global integration as perceived by subsidiary managers is positively associated with the distinctiveness of strategic resources that an MNE has committed to local operations Organizational infrastructure Information flow. When operating in a volatile environment such as China, information flow with the rest of the MNE network is fundamental because it leverages the extent of organizational control over subunits while facilitating the dispersal of existing resources and the creation of new bundles of knowledge. An information flow system is an important part of the organizational infrastructure. It ensures that global integration is maintained, monitored and adjusted within a portfolio network (Bartlett and Ghoshal, 1989). Information system transparency forces conflicts to be resolved on the basis of improved problem definition rather than having them be smoothed over or decided upon on the basis of leadership skills, intellectual acumen, personal savvy or hierarchical position (Prahalad, 1975). The flow of information can be structured with sufficient asymmetry that different subsidiary managers will be encouraged to identify either with responsiveness or with integration (Prahalad and Doz, 1987). A wellconstructed information flow system enables the MNE to oversee and adjust integration dynamically by manipulating these characteristics (Cray, 1984; Doz and Prahalad, 1981, 1984). This dynamic adjustment is important given that diversified MNEs use unstructured, evolving decision-making processes in order to handle their multidimensional, heterogeneous global businesses. Better information flow thus facilitates overall integration. Therefore, Hypothesis 4: In a dynamic foreign market, the degree of overall integration as perceived by subsidiary managers is positively associated with the efficiency of an established information flow system between each subsidiary and the rest of the network Coordination system. While formal coordination mechanisms such as centralization, formalization, planning and output controls may not fit with fuzzy business boundaries within a highly complex MNE network (Prahalad and Doz, 1987) or with subsidiaries in a highly uncertain market (Ghoshal and Nohria, 1989), informal or subtle coordination

11 mechanisms such as lateral relations, informal communication and organizational culture can fundamentally facilitate or impede global integration, depending upon the extent to which these mechanisms have been established and responsive to changing environments (Doz and Prahalad, 1984). The breadth of coordination (the number of units in the coordination network) and the diversity of coordination (the number of functions coordinated) are both important aspects of the degree to which a subsidiary should be integrated into the MNE network (Baliga and Jaeger, 1984). An advanced coordination mechanism that aligns well with both system needs and critical externalities fosters global integration. It enhances the efficiency and effectiveness of information flow, resource sharing, operational flexibility and intraorganizational transactions between each subsidiary and the rest of the system (Kobrin, 1991; Kogut, 1985). These discussions suggest that coordination infrastructure favorably affects global integration. Therefore, Hypothesis 5: In a dynamic foreign market, the degree of overall integration as perceived by subsidiary managers is positively associated with the efficiency of an established coordination system between each subsidiary and the rest of the network Resource flow. Resource flows buffer operational difficulty and instability in a disturbed environment. It reduces a firm s dependence on critical external resources, which, in China, are mostly controlled by the government. In Gupta and Govindarajan s (1991) scheme, the key factor underlying global integration is resource or knowledge flow, which can be defined as the transfer of both visible (e.g., tangible assets and production factors) and invisible (e.g., intangible assets, capital and knowledge) resources amongst geographically dispersed units. Any MNE subsidiary can be arrayed along two dimensions of knowledge flow, that is, the extent to which a subsidiary receives resource inflow from the rest of the system and the extent to which the subsidiary provides resource outflow to the rest of the corporation. A well-constructed resource flow infrastructure improves input processes, production processes and output processes within a system. It drives up overall integration because it enhances transaction cost economies for the system as a whole (Dunning, 1981), promotes taxation avoidance (Kogut, 1985) and yields more revenues from the cross-border value chain (Porter, 1986). Bartlett and Ghoshal (1989) maintain that such an infrastructure helps an MNE shift from conventional, rigid controls to using strategic orientations to monitor foreign operations, the latter being more flexible and effective (Golden, 1992). This line of reasoning suggests that well-run resource flows facilitate integration. Therefore, Hypothesis 6: In a dynamic foreign market, the degree of overall integration as perceived by subsidiary managers is positively associated with the efficiency of an established resource flow system between each subsidiary and the rest of the network Strategic needs Y. Luo / Journal of International Management 8 (2002) Risk diversification. Globally dispersed subunits should be structured in such a way as to help fulfill an MNE s organizational needs without taking excessive risks (Ghoshal,

12 200 Y. Luo / Journal of International Management 8 (2002) ). This is often achieved by setting up collectively integrated yet individually differentiated strategic goals for specific foreign subsidiaries. When firms are operating in a volatile context such as China, interactions with indigenous contingencies and institutions result in greater operational uncertainties and financial risks for foreign businesses. High responsiveness in such an uncertain environment entails more switching and exit costs should undesirable events occur (Teece, 1985). When firm-specific assets are exposed to a highly unpredictable environment, the risks embodied in high responsiveness become difficult, if not impossible, to neutralize or mitigate (Root, 1988). In an effort to reduce risks in a volatile market, a foreign subsidiary should maintain low responsiveness to the host environment and high integration with the rest of the network. Therefore, Hypothesis 7: In a dynamic foreign market, the degree of overall integration as perceived by subsidiary managers is positively associated with the extent to which a subsidiary is designed to achieve risk diversification Market expansion. Despite uncertainty, a dynamic market such as China also implies opportunities, which make market expansion attractive. Firms seeking a market share and competitive position in a host market require high responsiveness. Continuous commitment to local operations is necessary to attain and maintain a competitive edge and strong market power compared with rivals (Porter, 1986). Product differentiation, customer responsiveness and market segmentation are often essential to the achievement of a competitive advantage and abnormal returns. These business strategies necessitate high overall responsiveness (Morrison and Roth, 1992). High integration and centralization are likely to hamper the capability building and localized learning needed for firm growth and business expansion. In a dynamic foreign market, local market expansion often requires a long haul investment by an MNE as well as innovation in host country operations. In order to preempt emerging opportunities, foreign firms must not only commit resources and capabilities to host operations but also cultivate long-term relationships within the business community and with governmental authorities. Responsiveness and adaptation are prerequisites for establishing connections with local stakeholders (Luo and Peng, 1999). Therefore, Hypothesis 8: In a dynamic foreign market, the degree of overall integration as perceived by subsidiary managers is negatively associated with the extent to which a subsidiary is designed to achieve local market expansion Factor exploitation. Emerging markets such as China are characterized by not only pent-up demand but also cheaper production factors, especially labor costs. While market expansion is mainly intended to maximize on benefits gained from host market demand or industrial structural discrepancies between host and home countries, factor exploitation aims largely at gaining access to indigenous resources and benefiting from the factor endowments of a host country. In Dunning s (1981) eclectic paradigm and Porter s (1990) diamond model, factor conditions and natural endowments are major components

13 Y. Luo / Journal of International Management 8 (2002) for determining a nation s competitive or locational advantage. When a foreign subsidiary is designed to exploit factor endowments, which is normally the case of upstream or backward foreign direct investment (Dunning, 1981), the host project is often a platform for assembly-export or serves as a sourcing-supply base for the MNE network. Thus, the MNE benefits from both the comparative advantages of the host country and the internalization advantages of an integrated network (Dunning, 1981; Ghoshal, 1987). Pressures for overall integration are stronger when firms attempt to pursue factor endowments. This is because firms need to use integration-based approaches such as transfer pricing, intrafirm financing, vertical integration and value chain coordination to attain these benefits. Therefore, Hypothesis 9: In a dynamic foreign market, the degree of overall integration as perceived by subsidiary managers is positively associated with the extent to which a subsidiary is designed to achieve factor endowment benefits. 3. Research methods 3.1. Data collection A nationwide mail survey of senior foreign managers (general or deputy general managers) in MNE manufacturing subsidiaries in the People s Republic of China was undertaken during Our sample list was drawn from information in the Directory of Foreign- Invested Industrial Enterprises compiled by MOFTEC, China in 1996 and the Almanac of China s Foreign Economic Relations and Trade ( ). We sent both English and Chinese versions of the questionnaire to 500 foreign subsidiaries in China through an independent contractor. The geographical focus was on investments in the Yangtze River Delta (Shanghai, Jiangsu and Zhejiang) and the Pearl River Delta (Guandong and Fujian). After three rounds of reminders, 168 complete responses were received, 5 including 76 wholly owned subsidiaries and 92 foreign-dominated equity joint ventures. Foreign minority joint ventures were excluded because the dominant position of local partners often changes an MNE s economic rationale, strategic ability and resource commitment to subsidiary operations overseas. We made several efforts to check the threat of common method variance. Firstly, we followed the post hoc procedural method suggested by Podsakoff and Organ (1986). In early 1998, we sent the same questionnaires to 29 randomly selected senior managers 5 The industries involved in this study include electronics, telecommunications, garments, fiber products, food processing and beverages, leather, rubber and plastic products, chemical products, medical equipment, pharmaceuticals, electric equipment and machinery, among others. Major countries from which investment originates include the US, Hong Kong, Japan, Germany, France, the UK, Italy, Australia, Taiwan, Canada, Singapore and Korea.

14 202 Y. Luo / Journal of International Management 8 (2002) who had responded to the early round. Correlation analysis of 22 responses exhibited strong consistency amongst the survey items between the two different periods (all at P <.0001). Secondly, after completing the post hoc procedural methods, a statistical remedy was employed. We conducted Harman s One-Factor Test where all variables revealed in the questionnaire were entered into a factor analysis. The results uncovered that neither a single factor emerged nor there was a general factor that could account for the majority of the covariance in these variables. Thirdly, we computed the proxy index (%) for local responsiveness (total R&D, marketing and after-sale service expenses incurred for a host market relative to total sales including export) for 33 sample firms in Jiangsu, the information obtained from our field visit in The correlation of this index with the perceived integration reported in the initial questionnaire for these 33 firms was significantly negative (b = 0.85, P <.0001). All of the above analyses suggest that there are no noticeable threats of common method variance. Lastly, our use of archival data to measure a number of control variables helps dispel the incidence of such variance Variables and measurement Overall integration is measured as the average of responses to three questions on a seven-point Likert scale, as indicated in Appendix A. These questions cover how well integrated a subsidiary is with the rest of the MNE network in terms of internalization, coordination and interdependence, respectively. While Cronbach s a (.78) suggests internal consistency for this dependent variable, communality estimates ( ) revealed by the principal component analysis indicate that all dimensions of this construct are important. Murtha et al. (1998) have validated that global integration is a distinct construct compared with coordination as well as other organizational variables in managers minds. Of the dynamic capability variables, foreign experience in the host country was measured by the number of years that each corporation has been present and operating in China. Networking strength was operationalized as a multidimensional construct including personal ties with managers of other firms (buyers, suppliers, distributors and competitors) and officials from the political government, industrial bureaus or regulatory or support organizations. Each subdimension was first arithmetically averaged to obtain a score for managerial ties with other firms and with government officials, respectively. These were then further averaged to get a score for business networking. The reliability of the two dimensions was confirmed by a high level of Cronbach s a ( >.72). The communality estimate for each dimension (0.85 and 0.87) suggested the appropriateness of including both the business community and the government authorities in defining managerial ties in a dynamic environment such as China. Lastly, resource distinctiveness was measured as the mean of resource rareness/specificity, inimitability and sustainability. Each of these dimensions was computed based on the arithmetic average of three strategic resources: product and process technology, industrial and intellectual property rights and managerial and organizational skills. Both the internal consistency and the dimensionality

15 of this construct were confirmed by a high level of Cronbach s a (.76) and communality ( ). The organizational infrastructure variables are all multidimensional. The efficiency of an established information flow system between a focal subsidiary and the rest of the network was defined as the mean of its outflow efficiency, inflow efficiency and transparency. The efficiency of an established coordination system was measured as the mean of appropriateness of coordination breadth/diversity, efficiency of the coordination mechanism and effectiveness of communication. Similarly, the efficiency of an established resource flow system was measured as the mean efficiency of production factor flow, tangible asset and technology transfer and intangible asset and knowledge flow within the established system. Internal consistency amongst each of these infrastructure constructs was validated by high coefficients of Cronbach s a (.68.79). The appropriateness of dimensionality for each construct was supported by high communality estimates ( ). The three variables of strategic needs were measured by each respondent s assessment of the extent to which a subsidiary was designed to pursue risk diversification, local market expansion and factor endowment exploitation Control variables Y. Luo / Journal of International Management 8 (2002) In assessing the organizational effects of integration, we controlled for several industrial and environmental factors that may influence integration. These variables include (1) Industrial Structural Uncertainty, defined as the geometric average of standard deviations of the sales growth and net profit growth of the industry in which a subsidiary operated during , (2) Industrial Sales Growth, measured by the compound average of industry sales during , (3) Industrial Asset Intensity, measured by the average net value of total fixed assets ( ) in an industry divided by the average total sales of an industry during the same period and (4) Institutional Deterrence over MNE operations, defined as a categorical construct (5 if a subsidiary s industry lies in Restricted Category B, 4 if in Restricted Category A, 3 if in the Newly Opened Category, 2 if in the Encouraged Nontechnological Category and 1 if in the Encouraged Technological Category). Data on the first three industrial variables were collected from the consecutive editions of the China Statistical Yearbook ( ). The full list of projects and industries under each of these categories was obtained from The Orientation Directory of Industries for FDI, compiled by the State Council, China. Three organizational-level variables are also controlled for, including (1) Cultural Distance, computed according to the composite index used by Kogut and Singh (1988) and data contained in Hofstede (1984) and Huo and Randall (1991), (2) Entry Mode, defined as a dummy variable (1 if wholly owned, 0 otherwise) and (3) Subsidiary Size, defined as the number of full time employees in an MNE subsidiary. Information about country of origin, entry mode and number of employees was obtained from the aforementioned Directory of Foreign-Invested Industrial Enterprises, Almanac of China s Foreign Economic Relations and Trade and 22,000 Businesses in PRC. Table 1 exhibits the descriptive statistics and Pearson correlation coefficients.

16 Table 1 Descriptive statistics and Pearson correlation (n = 168) Variables Mean S.D Overall integration 2 Foreign * experience 3 Networking **.20** strength 4 Resource ***.22**.03 distinctiveness 5 Information *** * flow 6 Coordination *** ***.36*** system 7 Resource flow **.23**.12.21**.32***.51*** 8 Risk ** *.09 diversification 9 Local market ***.17*.11.28*** **.07 expansion 10 Factor ** **.33***.23**.03 exploitation 11 Cultural ***.17*.25***.19**.04.25***.37***.03.35***.14 distance 12 Firm size Entry mode **.26*** * Structural *** **.15* *** ** uncertainty 15 Institutional ** * **.18* **.25**.11 deterrence 16 Industrial *** *.27*** **.03 growth 17 Industrial * ** * *.02.30*** *.08 asset intensity * P <.05. ** P <.01. *** P < Y. Luo / Journal of International Management 8 (2002)

17 4. Analysis and results Y. Luo / Journal of International Management 8 (2002) A standardized regression analysis was conducted to investigate the effect of organizational capability, infrastructure and needs on overall integration (Table 2). The values of the variance inflation factors ( in Models 1 5) suggest an absence of serious multicollinearity among predictor and control variables. A modified Kolmogorov Smirnov test was performed to check the univariate normality of each dependent, independent and control variables. Except for networking strength, the results for all other variables ( , P >.10) justified the validity of the normality of these variables. Networking strength (0.11, P <.01) was transformed by taking its logarithm because its distribution was somewhat positively skewed. It demonstrated normality after this transformation. Table 2 Influence of organizational dynamics on overall integration: standardized regression (n = 168) Dependent variable: overall integration Independent variables Model 1 Model 2 Model 3 Model 4 Model 5 Dynamic capability Foreign experience Networking strength 0.15 y 0.19* Resource distinctiveness 0.35*** 0.21** Organizational infrastructure Information flow 0.37*** 0.34*** Coordination system 0.28*** 0.15 y Resource flow Strategic needs Risk diversification Local market expansion 0.20* 0.14 Factor exploitation 0.23** 0.25*** Control variables Cultural distance 0.19* 0.22** 0.18* 0.23** 0.14* Firm size Entry mode 0.15 y 0.19* 0.18* y Structural uncertainty 0.25*** 0.21** 0.22** 0.26*** 0.30*** Institutional deterrence 0.18* 0.17 y 0.19* 0.18* 0.19* Industrial growth 0.13 y 0.16* 0.14 y ** Industrial asset intensity Model F P < Adjusted R The entries in the table are the standardized b s. y P <.10. * P <.05. ** P <.01. *** P <.001.

18 206 Y. Luo / Journal of International Management 8 (2002) According to Table 2, each group of factors (Models 1 4) is collectively important in influencing overall integration ( P <.001). When incorporating all four groups into a test model (Model 5), the model presents a significantly higher statistical power to explain variations in overall integration, as evidenced by a substantial increase in adjusted R 2.As shown in Model 5, in which all relevant variables are controlled, foreign experience is not associated with overall integration. This rejects Hypothesis 1, which predicted a negative relationship between the two. During our post hoc interviews with 29 respondents (13 from subsidiaries exhibiting overall integration above the mean and 16 below), we asked them to rate the importance of host country experience to the success of global operations on a fivepoint scale from not important (1) to very important (5). Both groups rated experience as highly important (81% rated 4 in the high integration group and 86% rated 4 in the high responsiveness group). Combined with this field interview, the above result implies that foreign experience is important to global operations irrespective of the degree of overall integration. Model 5 shows that networking strength is significantly and negatively associated with overall integration ( P <.05). Hypothesis 2 is thus supported. This indicates that the established strength of managerial ties with other business people and government officials is related to increased overall responsiveness. Future research could further assess if a causal relationship between networking strength and responsiveness exists longitudinally. Model 5 also shows that resource distinctiveness has a significant and favorable influence on overall integration ( P <.01). In the perception of managers in foreign subunits, knowledge proprietariness is linked to greater integration. This finding lends support to Hypothesis 3. Of the three organizational infrastructure variables, information flow and coordination systems are found to be positively and significantly related to overall integration after controlling for other related variables (Model 5). Establishing an infrastructure for information flow and coordination among various units of an MNE network is necessary to ensure effective and efficient integration in a way that corroborates with an MNE s intentions. Hypotheses 4 and 5 are supported. In contrast to our prediction, however, resource flow infrastructure does not affect overall integration either before or after partialling out the effect of control variables (Models 2 and 5). This finding rejects Hypothesis 6. During our post hoc interviews with 29 respondents, 76% in the high integration group and 74% of the high responsiveness group rated at 4 on a five-point Likert scale. Combined with our interview results, this implies that resource flow systems are important regardless of the degree of integration and do not necessarily diminish as overall integration decreases. Locally responsive subsidiaries also need this system to boost and preserve their competitive edge in a host market. Amongst organizational needs, only factor exploitation reveals a fundamental effect on overall integration ( P <.001) when controlling for other factors (Model 5). Integration is an important organizational mechanism, which ensures the exploitation of factor endowments in a host country and the realization of internalization benefits. Regression analysis suggests that local market-pursuing objectives are not necessarily negatively associated with integration (Models 3 and 5). Seeking benefits such as market share and competitive power in a dynamic, emerging market such as China s does not eliminate the importance of integration. Finally,

19 Y. Luo / Journal of International Management 8 (2002) risk diversification is not significant in relation to overall integration. Risk diversification in a highly volatile and complex environment such as China may be so complex that it is difficult to attenuate risk through organizational forces. The above results support Hypothesis 9 but reject Hypotheses 7 and 8. Table 3 presents the regression results of the moderating effect of strategic needs on the relationship between strategic infrastructure and overall integration based on the subgroup analysis. We address this issue because the function of organizational capability and infrastructure in overall integration is likely to be affected by strategic goals. For instance, it is possible that the effect of capability and infrastructure on integration is stronger when a foreign subsidiary emphasizes internalization synergies. As the literature Table 3 Moderating effect of strategic needs on integration: subgroup regression analysis (n = 168) Dependent variable: overall integration Model 1: risk diversification Model 2: market expansion Model 3: factor exploration Independent variables Low group (n = 121) High group (n = 47) Low group (n = 112) High group (n = 56) Low group (n = 92) Dynamic capability Foreign experience ** * 0.19* 0.09 Networking strength 0.29*** ** 0.29*** 0.10 Resource distinctiveness 0.22** ** 0.25** 0.07 Organizational infrastructure Information flow 0.46*** 0.35*** 0.27** 0.39*** *** Coordination system 0.18* 0.21* 0.34*** ** Resource flow Control variables Cultural distance Firm size * 0.08 Entry mode * ** 0.18* 0.09 Structural uncertainty 0.26** ** 0.26** 0.12 Institutional deterrence * * Industrial growth y * 0.16 y 0.12 Industrial asset intensity Model F P < Adjusted R The entries in the table are the standardized b s in relation to overall integration. y P <.10. * P <.05. ** P <.01. *** P <.001. High group (n = 76)

20 208 Y. Luo / Journal of International Management 8 (2002) lacks theoretical and empirical support for this issue, we treated it as a research question. Each subgroup is differentiated by the mean of that strategic need (e.g., m < 5.89 for a subgroup with low incentives to expand in the market while m 5.89 for high incentives to expand). According to Model 1 in Table 3, foreign experience remains insignificant to integration for the low risk diversification group but negatively influences integration for the high risk diversification group. Knowing the host environment is therefore critical for those seeking high risk diversification. Networking strength and resource distinctiveness are important to overall integration or responsiveness for those not emphasizing risk diversification. Interestingly, the effect of organizational infrastructure variables on overall integration remain unchanged in the subgroup analysis. Whether or not firms are concerned with risk diversification, information and coordination systems are important to overall integration. As shown in Model 2, when a subsidiary is designed to focus on local market expansion, foreign experience and networking strength are positively related to overall responsiveness while resource distinctiveness is positively linked with overall integration. When a subunit does not focus on this goal, these capability variables become insignificant to integration. An information flow infrastructure is crucial to integration in both high and low subgroups, suggesting the importance of information flow regardless of the degree of indigenous participation. A coordination system does not influence integration for the group pursuing a strong local market position but is significant for the other group. This implies that coordination is more important to integration for firms focusing on internalization than those targeting host market power or share. A market expansion objective does not moderate the effect of resource flow on integration as the result remains unchanged compared with Table 2 and is insignificant in both groups. Model 3 in Table 3 suggests that the relationship between the dynamic capability variables and the overall integration is moderated by having a factor exploitation objective. Experience and networking strength are insignificant in relation to integration in the high group but have a strong negative impact on integration in the low group. When a foreign firm does not emphasize exploitation of factor endowments in the host country, host country experience and networking ability become more important to overall responsiveness. Resource distinctiveness also becomes significant in influencing integration when a firm shifts its strategic focus from factor exploitation to other emphases such as demand exploitation. Information flow and coordination systems are positively linked to integration for the high exploitation group while they are not significant for the low exploitation group, suggesting that the importance of these systems is stronger for firms focusing on factor exploitation. 5. Discussion and conclusion This study examines the relationship between organizational dynamics and overall integration for MNEs in a dynamic environment. While previous studies suggest that overall global integration is largely determined by environmental and structural contingencies, this