Shared Vision in MNE Subsidiaries: The Role of Formal, Personal, and Social Control in Its Development and Its Impact on Subsidiary Learning
Abstract
Given the strategic importance of achieving global integration and worldwide learning for most multinational enterprises (MNEs), this article focuses on “shared vision” as a critical mechanism for realizing both goals. Specifically, this paper investigates the impact of three types of control mechanisms—formal (headquarters imposed rules and decisions), personal (the presence of expatriates), and social (social interaction among subunits) control—on building shared vision in subsidiaries and examines the subsequent impact of having shared vision on subsidiary learning. Overall, data from 99 subsidiaries located in the United States and headquartered either in Europe or Japan show that personal and social mechanisms of control are effective in building shared vision whereas formal control mechanisms have no impact. Results also support a significant relationship between shared vision and subsidiary learning. © 2012 Wiley Periodicals, Inc.
Introduction
One of the critical challenges facing most multinational enterprises (MNEs) is the management of the global coordination, integration, and knowledge sharing between and among their culturally and geographically dispersed subunits (Bartlett & Ghoshal, 1988; Harzing, 2000; Noorderhaven & Harzing, 2009; Rugman & Verbeke, 1992). Different conceptualizations of the MNE in the international management literature share a common thread suggesting that the drive for international competitive advantage through global integration and organizational learning is generally tempered by the diverse competitive, cultural, and institutional demands of different host markets (Bartlett & Ghoshal, 1988; Doz, Bartlett, & Prahalad, 1981; Kostova & Roth, 2003; Nohria & Ghoshal, 1997). While different strategies pursued by MNEs (i.e., transnational, global, multinational) require varying degrees of global integration and global learning to be realized throughout the organization (cf. Bartlett & Ghoshal, 1988; Kostova & Roth, 2003), building shared vision has been suggested to be one of the powerful mechanisms that may facilitate global integration as well as knowledge sharing and organizational learning in highly complex, dispersed, and differentiated organizational forms (Kostova & Roth, 2003). In this article, I examine three potential antecedents (formal, personal, and social control) of building shared vision in MNE subsidiaries and its subsequent impact on subsidiary learning.
Shared vision can be broadly defined as the collective understanding, language, culture, and norms among members of groups or organizational units governing members' actions, decisions, and behaviors (cf. Nahapiet & Ghoshal, 1998). As such, it has the potential to tie and link different parts of an organization as it facilitates a common understanding of collective goals and proper ways of acting. Kostova and Roth (2003) note that shared vision—defined in their work as the relational aspect of social capital—is one of the critical governing mechanisms in MNEs as it can serve to facilitate the global coordination of cross-border activities. Previous research on shared vision showed that it facilitates outward knowledge transfers from MNE subsidiaries to their headquarters (Li, 2005; Li, Barner-Rasmussen, & Bjorkman, 2007); leads to higher levels of trust and trustworthiness among business units (Tsai & Ghoshal, 1998); has a positive impact on new business venturing in structurally differentiated contexts (Burgers, Jansen, Van den Bosch, Frans, & Volberda, 2009); and is a pillar of learning organizations that facilitates innovation and firm performance in industrial organizations (Calantone, Cavusgil, Zhao, 2002; Senge 1990).
While the strategic importance of building shared vision in facilitating global integration and organizational learning has been established in previous studies (Burgers et al., 2009; Kostova & Roth, 2003; Li, 2005), research on mechanisms through which MNEs can develop a shared value system which can then facilitate subsidiary learning has been mostly overlooked. Therefore, both researchers and MNE managers alike can benefit from an improved understanding of these mechanisms. One notable exception to this research gap is Tsai and Ghoshal's (1998) study in which the researchers proposed relationships between social interaction among strategic business units (i.e., structural social capital) and shared vision (i.e., relational social capital) but failed to support their hypothesis, leading them to conclude that “Members of a society may share a vision or values even if they do not have specific interpersonal relationships” (Tsai & Ghoshal, 1998, p. 473). While this interpretation can hold true at the societal level or within domestic boundaries as individuals get to learn and share the same national or organizational cultural values from birth onward or upon organizational entry (Hofstede, 1980), the MNE context—defined by cultural and geographical separation—requires the organization to export its organizational goals and value system to host country units in order for a shared vision to develop and subsidiary learning to take place, as merely belonging to the MNE network would be a necessary but insufficient condition to develop shared vision when there are no means to understand that vision.
Given the lack of sufficient empirical evidence on mechanisms that build shared vision and its subsequent impact on subsidiary learning, the primary goal of this article is to advance our understanding of the potential antecedents and outcomes of having shared vision in MNE subsidiaries. Drawing from agency theory (Eisenhardt, 1988, 1989; Jensen & Meckling, 1976), the first objective of this study is to examine the impact of formal (headquarters imposed rules, policies, and guidelines), personal (presence of expatriates), and social (social interaction among subunits) control mechanisms in building shared vision as they are the primary means through which a common and a shared understanding of norms, goals, and practices can be exported and as a result emerge in host country units. Drawing from knowledge management and organizational learning theory and research (Lane, Salk, & Lyles, 2001; Lyles & Salk, 1996; Szulanski, 1996), the second objective of this article is to examine the impact of shared vision on facilitating subsidiary learning.
Conceptual Background and Hypotheses Development
One of the frequently cited definitions of the MNE as a unique but an increasingly ubiquitous organizational form describes it as “a group of geographically dispersed and goal-disparate organizations that include its headquarters and the different national subsidiaries” (Ghoshal & Bartlett, 1990, p. 603). One implicit assumption underlying this widely accepted definition of the MNE is that the local goals and interests of subsidiaries may not always be in perfect alignment with the goals and interests of the headquarters (Birkinshaw, Hood, & Young, 2005; Nohria & Ghoshal, 1994). Indeed, interunit power struggles; potential inconsistencies; and conflict among the goals, interests, values, practices, and routines of different subunits have been suggested to be some of the defining characteristics of most MNEs (Roth & Kostova, 2003).
These potential conflicts and inconsistencies that exist between headquarters' and subsidiaries' goals and interests mirror the traditional control problem in principal–agent relationships, which arises due to the difficulties principals face in monitoring and verifying agents' behaviors (Roth & O'Donnell, 1996). In circumstances when agents have the potential to pursue their self-interests, for example, when there is high information asymmetry between the agent and the principal, principals can rely on various mechanisms to align mutual interests such as incurring monitoring costs or designing proper incentive schemes (Eisenhardt, 1988, 1989; Jensen & Meckling, 1976).
Researchers identified several control mechanisms that have the potential to aid principals in overcoming agency problems and aligning the interests of principals and agents. Some of these control mechanisms identified by researchers include behavioral vs. output control (Ouchi, 1977), personal and bureaucratic control (Child, 1972), clan control (Ouchi, 1980), control by socialization (Edstrom & Galbraith, 1977), and cultural control (Baliga & Jaeger, 1984). In the case of MNEs, principals (i.e., headquarters) may achieve alignment of interests and consequently achieve global integration of their globally dispersed agents (i.e., host country subsidiaries) by relying on any or all of these control mechanisms as such controls are not mutually exclusive and tend to exist simultaneously in most MNEs (Nohria & Ghoshal, 1994). Therefore, the theoretical question of importance is to explain how different control mechanisms may impact shared vision in MNEs and whether they are equally or differentially effective in doing so. While a number of different control mechanisms have been identified in previous research, this study focuses on formal, personal, and social control mechanisms as this categorization is consistent with most previous research on MNE control and coordination (Andersson, Bjorkman, & Forsgren, 2005; Bjorkman, Barner-Rasmussen, & Li, 2004; Edstrom & Galbraith, 1977; Harzing, 2001a).
Formal Control
In the MNE context, formal control involves utilizing an extensive set of rules, regulations, and procedures to control subsidiary output and behavior that clearly limit the subsidiaries' role and autonomy in the host market (Andersson et al., 2005; Harzing, 2001a). It also implies direct and overt intervention by the headquarters' management in the subsidiary's ongoing operations or strategic decision making. For example, performance targets, budgets, and reports that are put in place by the headquarters control the output of subsidiaries in a bureaucratic way, while manuals, company procedures, and rules and regulations dictated and imposed by the headquarters restrict the behaviors of the subsidiaries (Baliga & Jaeger, 1984; Ouchi, 1977, 1980). While the use of such formal mechanisms that control the output and strategic behaviors of subsidiaries is inevitable and is effective in doing so, their impact on building shared vision in subsidiaries is not as clear.
On one hand, formal control mechanisms may contribute to the development of shared vision in MNE subsidiaries because the use of such mechanisms communicate—in an explicit and a tangible manner—the performance expectations from the subsidiary, what the parent company values and gives importance to, its objectives, and the global strategic priorities in general. In other words, having formal control over the strategic decision-making process of the host country unit can be a means through which the host country units can understand, implement, and internalize the value system of their parent companies.
Hypothesis 1a: There will be a positive relationship between formal headquarters control and shared vision in MNE subsidiaries.
Hypothesis 1b: There will be a negative relationship between formal headquarters control and shared vision in MNE subsidiaries.
Formal control mechanisms are sufficient in governing agents in principal–agent relationships when information asymmetry between the principal and the agent is low and goal congruence between the principal and the agent is high. For a typical MNE however, cultural, geographical, and institutional separation increases information asymmetry and decreases goal congruence between the parent company and the host country subsidiaries (Ghoshal & Bartlett, 1990; Nohria & Ghoshal, 1994). Therefore, in the MNE context, the efficacy of formal control mechanisms decreases (Nohria & Ghoshal, 1994) and MNEs begin to rely on more informal, subtle, and less traditional means to govern the behaviors and output of its subunits. Less formal and more subtle control mechanisms may be achieved through lateral relations between the units of an MNE, informal communication mechanisms, use of expatriates, and the like (Baliga & Jaeger, 1984; Martinez & Jarillo, 1989).
Personal Control
Previous research established the role of expatriates in fulfilling the control function, personally, in host-country subsidiaries on behalf of the parent company (Bonache, Brewster, Suutari, De Saa, 2010; Delios & Bjorkman, 2000, Edstrom & Galbraith, 1977; Harzing, 2001a). Harzing (2001b), for example, argued that expatriates fulfill their control function either by acting as mini-headquarters in the subsidiary (bear role of expatriates) exercising direct control (i.e., expatriates holding senior or executive positions), or that they may have an indirect control role by socializing the subsidiary (bumble-bee role of expatriates) with the rest of the organization (i.e., developmental or technical expatriates). Expatriates may fulfill their direct or indirect control role by training, coaching, and mentoring local employees as part of their assignments or by conducting performance evaluations with the results having promotion and/or compensation implications for local employees and managers.
Hypothesis 2: There will be a positive relationship between personal control and shared vision in MNE subsidiaries.
Social Control
MNEs can also exercise control by weaving an informal social network of managers and employees from different subunits that regularly meet and interact with each other, share and exchange knowledge, work on global projects and processes, attend global training and development programs and as a result develop trust, trustworthiness, and personal relationships among each other (Noorderhaven & Harzing, 2009; Li et al., 2007). Some of these activities—such as global training, development, and communication initiatives—may be purposefully implemented by headquarters to instill corporate cultural values among all MNE employees. This type of control becomes the most subtle but probably the most efficient way of achieving global control and integration in highly complex, differentiated organizations, as it results in the development of a set of shared values, norms, and performance expectations (i.e., shared vision) that govern the behavior of individual subunits.
Hypothesis 3: There will be a positive relationship between social control and shared vision in MNE subsidiaries.
Shared Vision and Subsidiary Learning
While the development of a set of shared norms, values, goals, and objectives may serve the purpose of achieving control over and integration in globally dispersed operations of MNEs, a potential outcome of shared vision could be its role in facilitating worldwide organizational learning. Organizational learning in MNEs has been an important area of research in the past decade as researchers have suggested that MNEs' competitiveness in the global playing field depends, in part, on their ability to exploit firm-specific advantages and knowledge in host-country markets (Kogut & Zander, 1993) and tapping into local sources of knowledge and innovation. In this article, the focus is on the subsidiaries learning from other units in their parent companies, as the relationship between shared vision and reverse knowledge transfers (i.e., from subsidiary to headquarters) have been studied elsewhere (Li, 2005).
First, having a shared vision can facilitate subsidiary learning from the parent company by enhancing the interunit homophily (i.e., the extent to which two parties are similar in their attributes, attitudes, and values) that exists between a focal subsidiary and other units within the MNE (Gupta & Govindarajan, 2000). When different units share common meanings, language, attributes, and values, the communication and exchange of ideas have greater impact on knowledge gains and learning (Gupta & Govindarajan, 2000). As such, when employees of a subsidiary share similar goals, ambitions, values, and perceptions about how to interact with other MNE employees, possible misunderstandings and miscommunication is less likely to occur and individuals have more opportunities to exchange ideas, share knowledge, and learn from each other.
Second, having a shared vision may help overcome a major motivational barrier to organizational learning—identified as the not-invented-here syndrome in previous research (Szulanski, 1996) and consequently may increase a subsidiary's motivational disposition to acquire knowledge from other units in the MNE (Gupta & Govindarajan, 2000). Not-invented-here syndrome can arise either from ego-defense mechanisms that may lead subsidiary managers to resist any information that suggests that other parent-company employees are more competent than they are; or from existing power struggles within the MNE, which may lead some subsidiary managers to downgrade the potential power of peer units by pretending that their knowledge is not valuable (Gupta & Govindarajan, 2000). As a result, lack of motivation on the part of the recipient unit (i.e., the subsidiary) may lead to foot dragging or outright rejection in the acquisition, implementation, and use of knowledge (Szulanksi, 1996; Zaltman, Duncan, & Holbeck, 1973).
Hypothesis 4: There will be a positive relationship between shared vision and subsidiary learning from the MNE network.
Methods
Data Collection Procedure
Using a survey methodology, data for this study were collected from the US-based subsidiaries of MNEs headquartered in Europe or Japan. Data collection took place between June 2008 and November 2008. Prior to collecting data, a contact list including names of the highest-ranking executives, addresses, and telephone numbers of the US subsidiaries of 869 MNEs were compiled through Dun & Bradstreet. Since sectoral differences are not central to the research questions posed in this article, such differences were controlled by design through selecting companies operating only in the manufacturing sector.
Because research shows a sharp decline in the response rates achieved from the upper echelons of firms (Cycyota & Harrison, 2006; Cycyota, Harrison, & Stahl, 2002), I followed Dillman's (2007) tailored design approach in the collection of data. Invitation letters to subsidiary executives were followed, approximately two weeks later, by a wave of reminders, which were then followed by two e-mail reminders. There were four undelivered mails, two cases where the respondents reported they were no longer foreign-owned, and one company which reported to have a policy of not responding to surveys of this kind. Among the completed responses, there was one organization that was not foreign-owned and one subsidiary that was a start-up and therefore excluded from the study. These instances reduced the number of subsidiaries that were eligible for participation from 869 to 860. After a listwise deletion of cases with missing values, 99 cases (i.e., 99 US subsidiaries from 99 different parent companies) were used in analyses yielding an effective response rate of 11.5%. Hambrick, Geletkanycz, and Fredrickson (1993) report that 10 to 12% response rates are typical for mailed surveys to top executives in the United States. Harzing (1997), in an international study of headquarters–subsidiary relationships, reports that the United States has the second-lowest response rate, with 11%, among 22 parent countries. Therefore, the response rate achieved in this study from this population can be considered comparable to previous studies.
Sample and Participants
While there was only one respondent per subsidiary that answered the survey-based questions, the responses were from key informants in the subsidiary who had in-depth knowledge, understanding, and organizational experience about the issues under investigation in this study based on their tenures and titles. Respondents had an average tenure of 14.9 (SD = 9.3) years working for the US subsidiary and held titles such as president; general manager; chairman; director; chief executive officer; chief financial officer; and vice president, human resources. Because the research questions focus on the subsidiary's exposure to formal, personal, and social control by the parent company and their assessment of shared vision and subsidiary learning, subsidiary executives were in the best position to know about these subjects.
The average size of the participating subsidiaries—measured as the number of US employees—was 2,968 (SD = 6,774), and the average age of the subsidiaries was 20.18 years (SD = 16.05). The percentage of MNEs that entered the US market through acquisition was 76%, and the rest were greenfield operations. Eighty percent of the sample had research and development (R&D) facilities in the United States, and 94% had manufacturing facilities, with 75% having both R&D and manufacturing facilities. The parent companies of these MNEs are located in 11 different countries, including Denmark (1.9%), Finland (5.7%), France (9%), Germany (27%), Ireland (1.9%), Italy (1.9%), Japan (18.9%), the Netherlands (4.7%), Sweden (4.7%), Switzerland (4.7%), and the United Kingdom (19.8%).
Measures
Formal Control
Formal control was assessed by adapting a five-item measure from Taggart (1997). Respondents were asked to indicate the degree to which the foreign parent company has control over the US subsidiary's strategic decision making in areas of marketing and sales, supply-chain management, organizational management, product and packaging, manufacturing, and technology-related decisions. A seven-point Likert-type scale was used, with 1 = no control at all and 7 = total control. The coefficient alpha for this measure was .88, indicating good interitem reliability.
Personal Control
Personal control was assessed by a single item that inquired about the total number of expatriates in the US workforce. Since it was already established in the theory section that expatriates—regardless of their assignment types—can fulfill their control roles either directly or indirectly, this single-item measure was deemed appropriate. The reported number of expatriates for the US operations ranged from 0 to 400, with a mean value of 23.43 (SD = 53.88).
Social Control
Social control was assessed by four items adapted from previous studies in MNE knowledge flows (Gupta & Govindarajan, 2000; Bjorkman et al., 2004) using a seven-point Likert scale ranging from 1 = not at all to 7 = very frequently. Respondents were asked to indicate the extent to which the key employees of the subsidiary acquire knowledge through international trips and visits to other units of the foreign parent company; by joining international committees, teams, and task forces within the foreign parent company; by attending international training programs or meetings involving participants from other units of the foreign parent company; and through phone calls, conference calls, and e-mails. The coefficient alpha for this measure was .78.
Shared Vision
Shared vision was measured by two items from Tsai and Ghoshal (1998) on a seven-point Likert scale ranging from 1 = strongly disagree to 7 = strongly agree and asked respondents whether the US subsidiary shares the same ambitions and vision with other units in the foreign parent and whether the employees of the US subsidiary are enthusiastic about pursuing the collective goals and missions of the whole foreign parent organization. The reliability score for this measure was .83.
Organizational Learning
Learning was measured by five items from Lyles and Salk (1996) and Lane, Salk, and Lyles (2001) on a seven-point Likert scale, with anchors ranging from 1 = not at all to 7 = to a great extent. The question stem asks respondents the extent to which the US subsidiary has learned from its foreign parent marketing and sales know-how, management techniques and practices, manufacturing processes, product development, and know-how on new technologies. The coefficient alpha for this measure was .80.
Control Variables
Several subsidiary-level variables that may be associated with the outcome variables of this study were used as control variables. Subsidiary size was measured as the total number of employees of the US operation. Subsidiary age was also controlled for. For determining a subsidiary's age, if the mode of entry was acquisition, the year in which the acquisition was realized (rather than the original establishment date) was used, as the former one is more relevant for both shared vision and subsidiary learning. Upstream competence of the subsidiary was also controlled for. Presence of an upstream value activity was measured by combining two items that asked about the presence of R&D facilities and the presence of manufacturing facilities in the United States, with a resulting score ranging from 0 to 2. The mode of entry—whether the US subsidiary was acquired or started as a greenfield—was also controlled for, with the value of one for an acquisition and zero for a greenfield subsidiary. To control for regional/cultural effects, a Japanese dummy variable was used in the analyses, with the value of one for a Japanese parent company and zero for a European parent company.1 Finally, to control for value chain integration of the subsidiary, two items on the percentage of sales to and purchases from the MNE network were averaged.
Results
Preliminary Analyses
Sampling Bias
To check for potential sampling bias in the data, I performed a series of independent sample t-tests to compare the means of respondent and nonrespondent firms on several organizational attributes. Using information from Dun & Bradstreet as the data source, firms were compared with respect to their total US sales, number of US employees, and their experience in the US market (i.e., age). There were no significant differences in the profile of respondent and nonrespondent firms on any of these organizational characteristics. Thus, sampling bias should not be a major concern for this study.
Common Method Bias
Because the research methodology relies on cross-sectional data collected from a single respondent, I took both procedural precautions in the design of the survey and performed post-hoc analysis to reduce concerns about common method bias (Podsakoff, McKenize, Lee, & Podsakoff, 2003). Procedural remedies included using different scale formats for different variables (five-point and seven-point Likert scales) and using different scale anchors for different measures (strongly agree to strongly disagree; not at all to very frequently). Also, different constructs were separated from each other by placing them in different pages to reduce artificial covariance among them (Podsakoff et al., 2003). Since the survey included more items on different aspects of subsidiary structure and strategy than the ones used for the purpose of this research, it would be hard for the respondents to guess the specific research questions posed in this study and respond accordingly. In order to further reduce concerns of common method bias, I performed a post-hoc analysis on all survey items using Harman's single-factor test, which resulted in multiple factors with eigenvalues greater than one and none of the factors accounting for the majority of the variance in the data. Therefore, concerns about this potential problem affecting the results of this study were reduced.
Validity of Measures
To establish the discriminant validity of the measures used in this study, I performed a principal components analysis with varimax rotation. All items had high loadings on their respective constructs with no cross-loadings above .40 (Table 1). The results suggest that social control, formal control, shared vision, and subsidiary learning variables are sufficiently distinct to be treated as separate constructs in hypotheses testing and regression analyses.
Hypotheses Testing
Descriptive statistics and zero-order correlations for all study variables are provided in Table 2. Significant correlations among key study variables provide preliminary evidence for the existence of proposed relationships. To assess the effects of the hypothesized variables on shared vision and the effect of shared vision on subsidiary learning, a series of OLS regression analyses were performed. Tables 3 and 4 present the results of the regression analyses used to test the hypotheses. When testing hypotheses, control variables were entered first in order to partial out variance due to the impact of these variables on shared vision and subsidiary learning. In the second step of the regression analyses, independent variables were entered to see whether they explain variance over and beyond control variables.
Hypotheses 1, 2, and 3 investigated the impact of formal, personal, and social control mechanisms on shared vision, respectively. Multicollinearity checks showed that variance inflation factors for all variables had values less than two (1.18–1.70) and their tolerances did not approach zero (0.56–0.86), suggesting that multicollinearity did not threaten parameter estimates. Model 1 in Table 3 shows that control variables as a set explain 9% of the variance in the dependent variable, with value-chain integration being significantly related to the level of shared vision (β = .24, p < .05). When independent variables were entered in Model 2, the overall model became highly significant (F = 3.02, p < .01), suggesting support for the theoretical model (the impact of control mechanism on shared vision). Explained variance in shared vision also increased to 24% in the full model. Value-chain integration was no longer significant in the presence of control mechanisms but subsidiary size was negative and significant as a predictor of shared vision (β = −.25, p < .05).
There was no support for the competing set of hypothesis of 1a and 1b as formal control was not a significant predictor of shared vision. Hypothesis 2 examined the relationship between personal control (presence of expatriates) and shared vision. Number of expatriates had a positive and significant coefficient providing support for Hypothesis 2 (β = .26, p < .05). Hypothesis 3 predicted a positive relationship between social control mechanisms and shared vision, which was also supported (β = .25, p < .05).
Hypothesis 4 investigated the relationship between shared vision and subsidiary learning. Multicollinearity checks showed that variance inflation factors for all variables had values less than two (1.08–1.70) and their tolerances did not approach zero (0.63–0.92), suggesting that multicollinearity did not threaten parameter estimates. In this model, independent variables from the first set of hypotheses (formal, personal, and social control) were also controlled for due to their significant correlations between subsidiary learning reported in Table 2. Model 1 in Table 4 reports that control variables as a set explain 36% of the variance in subsidiary learning (F = 5.33, p < .001). Results suggest that both formal control (β = .26, p < .05) and social control (β = .26, p < .05) are significant predictors of subsidiary learning, whereas mode of entry (acquisition) is marginally and negatively significant as a predictor (β = −.19, p < .10). The addition of shared vision to the regression equation in Model 2 increased explained variance to 39% and the overall model was highly significant (F = 5.35, p < .001). Shared vision was a positive and significant predictor of subsidiary learning (β = .19, p < .05) supporting Hypothesis 4. In addition, both formal control (β = .24, p < .05) and social control (β = .21, p < .10) remained as significant predictors of subsidiary learning. In sum, while formal control has no impact on building shared vision, it is as important as shared vision and social control in facilitating subsidiary learning.
Discussion
Theoretical Implications and Future Research Directions
Previous international management research suggests that building shared vision plays a key role in the integration of complex, globally distributed, and differentiated MNEs as well as in facilitating worldwide learning. Despite being a variable of strategic importance for most MNEs, few attempts have been made to understand the mechanisms that may enhance shared vision in these organizations and the extent to which it facilitates subsidiary learning from the parent. By integrating ideas from theories of MNEs, agency theory, and organizational learning, the main findings of this study show that, for building shared vision, the use of social and personal control mechanisms matter and that shared vision consequently facilitates subsidiary learning.
This study operationalized personal control as the number of expatriates present in the subsidiary and found that the number of expatriates was positively related to subsidiaries' shared vision. While this study did not explicitly consider the nature of their international assignments, the results confirm that expatriates are successful in instilling the values, culture, policies, and best practices of the parent company to the host country subsidiaries regardless of their positions. As the community of expatriates present in a subsidiary gets larger, there are greater opportunities for local personnel to interact, mingle, communicate, and work with them and to learn about the headquarters' values, mission, and culture. This could be achieved in an intentional and purposeful manner when expatriates have to train, mentor, and evaluate the performance of subsidiary employees as part of their assignments; or as an unintended consequence resulting from increased and frequent formal and informal interaction between the expatriate pool and subsidiary employees. As a result, as the findings of this study indicate, host-country subsidiaries will be better socialized into the culture of the parent company and come to share its set of values and belief system.
The finding that the number of expatriates is positively related to shared vision can also be interpreted within the light of recent research on the subsidiary-level outcomes of utilizing expatriates (Colakoglu & Caligiuri, 2008; Colakoglu, Tarique, & Caligiuri, 2009). There is some empirical research that points to the negative impact of utilizing expatriates in key positions on subsidiary performance (Colakoglu & Caligiuri, 2008), whereas other research suggests that this relationship may depend on cultural or strategic variables (Colakoglu et al., 2009; Gong, 2003). The use of expats was also shown to be negatively related to subsidiary local embeddedness in previous research (Andersson, Bjorkman, & Forsgren, 2005). The results of this study coupled with previous empirical findings may indicate that while expatriates are a potential source of liability for achieving local market outcomes, they are valuable assets for achieving desirable global outcomes such as worldwide knowledge sharing and global integration, making the “expatriate” construct the quintessential manifestation of the strategic dilemma between global integration and local differentiation faced by the majority of MNEs.
On a related note, MNEs that are increasingly faced with cost issues and rising staff immobility are substituting or complementing traditional expatriate assignments with alternative assignments such as short-term, commuter, rotational, contractual, and virtual assignments (Tahvanainen, Welch, & Worm, 2005), blurring the line between personal and social control. Yet the use of such alternative assignments that involve limited and shorter-term exposure to and interaction with subsidiaries may be better classified under social control. Therefore, in the future, we may expect MNEs to rely on personal control by traditional expatriates to a lesser degree and on social control by frequent commuters, short-term assignees, and the like to a larger degree.
This study also found that social mechanisms of control are instrumental in building shared vision in MNE subsidiaries. The more subsidiary employees communicate with other MNE employees through joint global projects, international training programs, and the like, the more they come to develop a shared set of values and internalize the mission and values of the parent company over time. While such social interaction between subsidiary employees and other MNE employees was previously reported to be related to knowledge inflows to and outflows from subsidiaries (Gupta & Govindarajan, 2000; Noorderhaven & Harzing, 2009), this study is the first to empirically demonstrate the relationship between social interaction and shared vision—a relationship that Tsai and Ghoshal (1998) proposed but failed to empirically support. This finding also provides partial support to Kostova and Roth's (2003) theory on the formation of social capital in MNEs. According to their theory, boundary spanners—those employees who have direct interactions with other MNE employees—are crucial in the formation of MNE social capital and it is the amount, intensity, and the efficacy of the boundary spanners' interaction with other MNE employees that builds individual social capital, which is then transformed into public social capital through social information processing. While this research did not explicitly consider the intensity and the efficacy of social interaction but rather focused on the amount, future researchers may investigate other dimensions of social interaction (i.e., intensity and efficacy) and extend on the findings of this study.
While both personal and social control can be utilized by MNEs in an intentional and purposeful way to govern subsidiary operations, these control mechanisms are more informal and subtle than the use of overt and formal control that involves direct intervention of the headquarters on the strategic processes of subsidiaries. Based on the finding of this study, which showed that formal control did not have a positive or a negative impact on building shared vision, it seems that less formal and subtler ways of governing subsidiaries tend to be more effective in building shared vision than implementing formal control in a traditional way. This result, coupled with the theoretical explanations provided for either a knowledge-based positive or a power/conflict-based negative impact, may suggest that the real impact of formal control mechanisms on building shared vision may depend on other organizational, strategic, or cultural factors. For example, formal control may have a negative impact on shared vision if an MNE is pursuing a transnational strategy but may have a positive impact if the MNE is pursuing a global strategy. Alternatively, the relationship may depend on the cultural values of the host country or cultural distance between the home and the host countries. Therefore, future research is needed to uncover the potential impact of formal control, if any, on shared vision.
The results of this study confirmed that shared vision facilitates subsidiary learning and complements previous research which reported shared vision to be related to knowledge outflows (Li, 2005). That is, shared vision does not only support global integration efforts in MNEs but also is a critical mechanism that facilitates multidirectional knowledge flows, organizational learning, and worldwide knowledge sharing.
Finally, another finding of this study that may deserve further attention was documentation of the positive impact of formal control on subsidiary learning, even though this was not a hypothesized relationship. Formal control—defined in this research as formally imposed decisions, policies, and actions—may indeed be instrumental for the subsidiary to learn explicit and codified firm-specific knowledge from the MNE. Since this study did not differentiate between the learning of tacit vs. explicit knowledge, future researchers should attempt to examine whether different types of control mechanisms are differentially instrumental in learning tacit vs. explicit MNE know-how.
Limitations
As with any research, this study is not without its limitations and therefore, its results should be interpreted with caution. First, given the use of cross-sectional data, no causal inference can be made regarding the relationships in this study, although the relationships proposed are based on prior theorizing on MNEs, principal–agent relationships, and knowledge management. Furthermore, despite precautions taken to avoid common method bias and the diagnostic test showing no major signs, the possibility of such a bias cannot be completely eliminated. Thus, future researchers should try to use a longitudinal design to uncover the influence of control mechanism on shared vision and of shared vision on subsidiary learning. Another limitation of this study can be its sampling frame. This study was conducted using data from a sample of foreign firms that have subsidiaries in the United States operating within the manufacturing sector. Thus, its generalizability is limited to such companies. While the current research design inherently controls for host country effects on the studied phenomena, future research can sample MNE subsidiaries operating in multiple host countries to explore whether the results of this study hold in multiple host country contexts. Finally, one of the core constructs in this study—shared vision—was operationalized by a two-item measure used previously in other research (Li, 2005; Tsai & Ghoshal, 1998). While the reliability estimate of this measure as well as its significant correlations with other variables in its nomological network (i.e., value-chain integration and the three forms of control) provide evidence of its psychometric properties, further research is needed to develop and validate multi-item measures of shared vision in order to enhance the rigor of future empirical findings.
Despite some of its limitations, this study established some important but unexamined antecedents and outcomes of building shared vision and will hopefully open up new research avenues on the study of shared vision in global organizations.
Biographical Information
Saba Colakoglu (PhD, Rutgers University, Human Resources and Industrial Relations) is an assistant professor of management at Berry College, Campbell School of Business. She researches and publishes in areas related to strategic management of human resources in a global context, with particular emphasis on the use of expatriates and knowledge management and transfer practices in MNC subsidiaries. Her research on these topics has appeared in the International Journal of Human Resource Management, International Journal of Cross-Cultural Management, Human Resource Management Journal, Human Resource Management Review, and Journal of Management Education. Dr. Colakoglu also serves on the editorial review boards of the Journal of International Business Studies and Journal of World Business.
