“How does subsidiary autonomy affect intra-firm
knowledge flows?”
Student: Davide Venier Student number: 11186631
Supervisor: Dr. Lori DiVito Study: Msc Business Administration
Second reader: Dr. Michelle Westermann-Behaylo Track: International Management
STATEMENT OF ORIGINALITY
This document is written by Student Davide Venier who declares to take full
responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no
source other than those mentioned in the text and its references have been used in
creating it.
The Faculty of Economics and Business is responsible solely for the supervision of
TABLE OF CONTENTS ABSTRACT ... 1 1. INTRODUCTION ... 2 2. LITERATURE REVIEW ... 6 3. THEORETICAL FRAMEWORK ... 16 4. METHOD ... 19 5. SAMPLE ... 19 6. MEASURES ... 21 6.1KNOWLEDGE TRANSFER ... 21 6.2SUBSIDIARY AUTONOMY... 21 6.3TRUST IN MANAGEMENT ... 21 6.4EMPLOYEES’ ABILITY ... 22 6.5CONTROL VARIABLES ... 22 7. ANALYSIS ... 23
7.1REGRESSION ANALYSIS – HYPOTHESES TESTING ... 28
7.2MODERATION AND MEDIATION EFFECT ANALYSIS ... 29
8. FINDINGS AND DISCUSSION ... 30
9. THEORETICAL, PRACTICAL IMPLICATIONS AND FUTURE RESEARCH ... 33
10. LIMITATIONS ... 35
11. CONCLUSIONS ... 35
ACKNOWLEDGMENTS ... 37
REFERENCE LIST ... 38
1
Abstract
The process of intra-firm knowledge transfer, as part of the knowledge creation and integration cycle, is widely recognized by the literature as one of the most relevant factors for any firm’s survival. In fact, both the ability to innovate and to proper communicate this newly created knowledge within the organization, are fundamental for every MNE, in order to adapt and take advantage of local environments and, ultimately, make the most of dispersed operations. First of all, this study highlights how the increasing importance of the knowledge creation and integration process, reflects on the way MNEs shape their structure to be able to tap into resources pools and profitable opportunities scattered all over the world. Secondly, with a review of the literature, the degree of autonomous behavior accorded to subsidiaries, is described as a tool that managers can use to improve knowledge creation and to a certain extent control knowledge flows. Lastly, with a statistical analysis on empirical data, gathered from several foreign subsidiaries of multinational enterprises, this study’s hypotheses are tested to control for the theorized positive effect of Subsidiary Autonomy over Intra-firm Knowledge Transfer. In addition, it is argued that two more elements can potentially influence the main relationship, and therefore, they are included and controlled in the model: the moderating effect of employees’ trust in their firm’s management and the mediating influence of employees’ abilities and skills. Even though the results do not support the main hypotheses, this research sheds light on the firm- and country- specific factors that affect intra-firm knowledge transfer, and attracts attention to the different tools that companies can use to manage the knowledge creation and integration process. Finally, it may serve as basis and direction for future research.
2
1. Introduction
Following World War II, the general economic development that many countries experienced, in particular the U.S. where the modern type of enterprise was born, fostered technological developments, international competition and institutional and policy changes towards a more integrated and globalized market (Holtbrugge & Berg, 2004). These economic and environmental changes deeply affected the way to think about business with a consequent influence on the development of Multinational Enterprises (MNEs), namely how their structure is shaped and organized in order to manage cross-border activities characterized by increasing complexity and uncertainty.
In face of these stimuli, MNEs needed to adapt and adjust their strategies if they intended to survive. Traditionally in fact, companies’ internationalization, was seen as the most obvious way to exploit ownership advantages through their foreign subsidiaries in order to access new markets, either to gather different and low cost resources, or to adapt the existing offer to local demand and conditions (Dunning 1993, Piscitello & Rabbiosi, 2006). On the contrary, beginning from the late 60s, and more evidently in the 80s, a debate started with the main idea that MNEs can undertake Foreign Direct Investment (FDI) for “competence creating” reasons (Cantwell and Mudambi, 2005), thanks to the access of foreign subsidiaries to local skills and competences which are potentially complementary with the skill base of the organization. With this perspective, foreign subsidiaries aren’t limited anymore to merely receive technologies and know-how from the parent company and successively exploit it in the host-country, but they are seen as a firm’s active unit that creates competences and capabilities that can be deployed in different parts of the company or in different markets and environments (Bartlett and Ghoshal, 1988; Gupta and Govindarajan, 1991, Piscitello & Rabbiosi, 2006). Accordingly, it is possible to say that, in order to survive to this changing environment, and to be ready to tap into profitable opportunities, MNEs had to adapt their organizational mechanisms from a more static, formal and bureaucratic kind to a more informal and networked one (Kostova et
3
al., 2016). Despite facing complexity and interdependencies in this “growing and increasingly globalized international marketplace” (Kostova et al., 2016, p. 176), MNEs have been able to “take advantage of improvements in information and communication technology, lowering barriers to trade and investment, increasing organizational experience with international operations” (Kostova et al., 2016, p.176). Moreover, thanks to more sophisticated, competent and knowledgeable managers and employees, firms have been able to better delegate and distribute responsibilities and tasks among multiple organizational units “in the name of greater organizational efficiency, motivation, and innovation.” (Kostova et al., 2015, p. 176).
In this context, foreign subsidiaries and their employees have gained primary importance for the organizations as they can access and exploit new settings characterized by different types of resources, cultural beliefs, mechanisms and limitations. In fact, thanks to their proximity to heterogeneous stimuli and resources of the host countries, foreign subsidiaries are able to generate innovation, for instance by creating a tailored solution to a local problem, and contribute to develop the firm’s competitive advantage (Frost, 2001; Birkinshaw et al., 1998; Ambos et al., 2006; Bartlett and Ghoshal, 1988). Foreign subsidiaries have increasingly done so, by tapping into “location-specific knowledge and assimilate it advantageously into global knowledge available throughout the corporation (Bartlett, Doz, & Hedlund, 1990)” (Lagerstrom & Andersson, 2003, p. 85). Therefore, knowledge, thanks to its relative inimitability and complexity, has become a critic resource for every organization and source of subsidiary- and firm- specific competitive advantage when combined with previous organizational knowledge (Gupta & Govindarajan, 1994; Mudambi & Navarra, 2004; Mu, Gnyawali and Hatfield, 2007). Within this perspective MNEs are seen as superior entities, relative to the market, for the transfer and recombination of knowledge, but this doesn’t imply that a company’s knowledge management happens regularly and efficiently (Bartlett and Ghoshal, 1988). In fact, organizations, have to set up proper means and tools to facilitate interaction and the creation of new knowledge stemming from many different sources. Nonetheless, following the absorptive capacity
4
argumentations of Cohen & Levinthal in their article of 1990, organizations, also “need prior related knowledge to assimilate and use new knowledge” (Cohen & Levinthal, 1990, p. 129). This presents companies with a trade-off, between the correct degree of overlapping knowledge, to provide a shared language which ensures effective communication, and labor-division which, by assigning specialized tasks, allows to maintain diversity within the organization and therefore the possibility to create “novel linkages and associations” (Cohen & Levinthal, 1990, p. 135), and ultimately innovate. In fact, only when the new knowledge is properly represented in a common language, it can be efficiently combined with existing knowledge, codified and recorded (Cohen & Levinthal, 1990; Grant, 1996). If MNEs cannot properly control internal knowledge creation and knowledge flows, they will lack a common vision and strategy and most of the knowledge created will be unrelated to the firm’s core business or too embedded in the host environment and in its relationships. This, results, first of all, in valueless knowledge from a corporate perspective, secondly in missed opportunities, and, lastly, such a company will rarely be the best of its industry since it is not innovating and will most likely follow the best technology present among its competitors (Mu, Gnyawali & Hatfield, 2007; Yamin & Forsgren 2005, Levitt & March, 1988).
This paper aims to contribute to the literature by connecting the knowledge theory of the firm (Nonaka, 1994; Grant, 1996) to organizational theory, and, in particular, to the internal control dilemma and division of labor. Therefore, it will mostly focus on the new enhanced role of foreign subsidiaries and its employees, on subsidiary autonomy, on organizational innovation and knowledge flows. In fact, MNEs internationalization follows different motivations that can be market-, efficiency-, resource- or strategic asset- seeking (Dunning, 1993), but it always starts by setting up subsidiaries in the host countries with enough resources, technologies and know-how to operate. Consequently, foreign subsidiaries who prove to be profitable and contribute to the firm as a whole, by successfully engaging with the host environment, will gain more resources, trust and strategic relevance in the company’s network (Ghoshal & Bartlett, 1988; Frost, 2001). This further development of the subsidiary’s capabilities and resources usually results in less dependence on the
5
headquarters which in turn increases the subsidiary’s decisional power and embeddedness in the host country (Birkinshaw & Hood, 2001; Ghoshal & Bartlett, 1988; Frost et al. 2002; Mu, Gnyawali & Hatfield, 2007). Such a subsidiary is more likely to have more “room for manoeuvre”, explore its environment and develop external relationships, therefore innovating by creating specialized knowledge which, to be of any use for the whole organization, has to be understood, properly transferred, recombined, crystallized in the company’s knowledge base and subsequently re-deployed (March, 1991; Edwards & Tempel, 2010; Nonaka, 1994).
Nonetheless, when a subsidiary holds valuable resources, as knowledge, that are not present in the rest of the organization, it gains strategic relevance and power (Mu, Gnyawali & Hatfield, 2007). In an agency perspective, this subsidiary empowerment, allows it more chances to behave opportunistically, by following its own objectives and increasing its rents to the expenses of the company (Ciabuschi, Dellestrand & Kappen, 2012). Moreover, increased subsidiary autonomy can hinder knowledge flows in the attempt to defend the subsidiary’s source of authority within the MNE’s network, or it can produce knowledge which is too specialized and embedded in the host country and of no use for the rest of the company (Mudambi & Navarra, 2004; Mu, Gnyawali & Hatfield, 2007).
However, there are many other factors that differently affect innovation, knowledge integration and transfer such as, for example, the employees’ and managers’ willingness to share information but the focus of this paper is on subsidiary autonomy since it has always represented a fundamental dilemma, which can be traced up to the choice between local responsiveness and global integration (Prahalad & Doz, 1997), and because there is no common agreement on its effect on knowledge management. Since knowledge is gaining primary importance in this increasingly developed and globalized world, and organizations adopt different structures to be able to better acquire it and exploit it through different allocations of decisional power and resources, and by setting up proper mechanisms for knowledge transfer, the research question of this paper follows:
6
RQ: “How does subsidiary autonomy affect intra-firm knowledge transfer?”
As stated earlier, subsidiary autonomy can provide room for exploration, innovation and the creation of new knowledge, but at the same time can also affect negatively the performance of the whole company, for instance by objective misalignment with the Headquarters or by utilizing an external market instead of an internal one (Ciabuschi, Dellestrand & Kappen, 2012). Moreover, this research will take into consideration other potential mediators or contributors of the subsidiary’s autonomy-knowledge flows relationship, broadening this perspective. First, with the literature review I will highlight the links between subsidiary autonomy and the knowledge-based theory of the firm, then with the theoretical framework the hypotheses for the analysis will be introduced and, after the research design will be explained, they will be tested and the results reported.
2. Literature Review
The aim of this review of the literature is to highlight the links between the evolution of organizational forms and their structure, with a particular focus on the new enhanced role of the subsidiaries in the MNEs’ internal value network, and the recent emphasis of knowledge exploitation and creation as main source of competitive advantage.
Innovation, described by Schumpeter as “new combination of means of production” (Schumpeter 1934, p. 47), or in other words the combination of new and old resources, can take different forms like for example a new product or a new process (Schumpeter, 1934). Moreover, it is possible to enrich this perspective by thinking about knowledge as one of the most valuable resources for its complexity and inimitability, thus picturing innovation as combination of new specialized knowledge with previous organizational knowledge (Mu, Gnyawali & Hatfield, 2007). From this definition, it is clear how knowledge creation, in particular, and the innovation process, in general, are fundamental for every organization that wants to survive to this fast-paced and ever changing business
7
environment; therefore, these processes have to be properly managed. Within this perspective the issue for MNEs is twofold: first of all, organizations need an internal structure that allows them to tap into different opportunities or knowledge pools thanks to internationally scattered units with proper resources and capabilities, secondly they need to set up mechanisms to foster and manage knowledge creation and knowledge flows.
The increasing attention of research and practitioners to the importance of knowledge for MNEs, is fairly recent and only from the 90s we can find a knowledge-based theory of the firm (Nonaka, 1991,1994,2000; Nonaka & Takeuchi, 1995). The main arguments of this theory are described by Nonaka et al. in their article of 2000, as the authors view the firm as a “knowledge-creating entity, and argue that knowledge, and the capability to create and utilize such knowledge, are the most important source of a firm’s sustainable competitive advantage. Knowledge and skills give a firm a competitive advantage because it is through this set of knowledge and skills that a firm is able to innovate new product/processes/services, or improve existing ones more efficiently and/or effectively.” (Nonaka et al., 2000, p. 1). Moreover, knowledge can be described as flow of information which is “anchored in the beliefs and commitment of its holder” (Nonaka and Takeuchi, 1995) and therefore “essentially related to human action” (Nonaka and Takeuchi, 1995). For these reasons, it is useful to further distinguish between explicit and tacit knowledge: the first characterized by being easily codified and transferable, and the latter by being more complex, “stickier” and harder to codify and transfer. Therefore, Nonaka (1994) further argued that knowledge is created through “conversion between tacit and explicit knowledge” (Nonaka, 1994, p. 18) which allows “to postulate four different "modes" of knowledge conversion: (1) from tacit knowledge to tacit knowledge, (2) from explicit knowledge to explicit knowledge, (3) from tacit knowledge to explicit knowledge, and (4) from explicit knowledge to tacit knowledge.” (Nonaka, 1994, p. 18-19). Further, the author argued that for organizational knowledge creation to happen, all four modes of knowledge conversion have to be “organizationally managed to form a continual cycle” (Nonaka, 1994, p. 20). This cycle is
8
shaped by shifts among the four modes and is configured as follows: first, in the socialization mode a “team/field of interaction” is built to help its members sharing experiences and perspectives. Secondly, in the externalization mode, the team members articulate their perspectives and reveal hidden tacit knowledge through the use of “metaphors”. Further, in the combination mode these new concepts are combined with previous existing knowledge through team coordination and data documentation. Lastly, in the experimentation mode, thanks to an iterative process of trial and error and “learning by doing”, concepts are developed and shared in the form of explicit knowledge that is gradually translated into different aspects of tacit knowledge (Nonaka, 1994).
This view of the organizational knowledge creation process is consistent through the literature as other authors proposed similar circular patterns (Nonaka & Takeuchi, 1995; Osterloh & Frey, 2000; Carlile, 2003; Carlile & Rebentisch, 2004). Figure 1 represents this process as a spiral as seen by Nonaka.
Figure 1
The increasing importance of knowledge for MNEs’ innovation, adaptation and survival is reflected in the evolution of the Headquarters-Subsidiary relationship and the adjustment of organizational
9
strategy and structure, which are mainly linked to “contingencies related to context (home and host country context), the role of knowledge (creation, sharing, and utilization), and the role of people (especially expatriates and global HRM).” (Kostova et al., 2016, p. 177). In fact, it is possible to witness in the literature a shift from “formal structures and design” (Kostova et al., 2016, p. 177), characteristic of the 60s and 70s and mainly focused on the primary role of the headquarters, to “an expanded and more diverse (theoretically and methodologically) set of individual- and organizational-level factors examined in tandem with contextual contingencies in which HQS relationships take place” (Kostova et al., 2016, p.177). This shift is underlined by the globalization and market integration trend, which, by increasing operations’ scope and complexity, and giving more importance to emerging markets and to institutional and cultural differences between countries, assigned more responsibilities and a new enhanced strategic role to foreign subsidiaries (Kostova et al., 2016). This perspective is supported by other authors such as Birkinshaw & Hood, who, in their article of 1998, argued that “traditionally, in academic models researchers assumed that ownership-specific advantages were developed at the corporate headquarters and leveraged overseas through the transfer of technology to a network of foreign subsidiaries (Dunning; 1981; Vernon, 1966). As these overseas subsidiaries grew in size and developed their own unique resources, however, it became apparent to many researchers that corporate headquarters was no longer the sole source of competitive advantage for the MNC” (Birkinshaw et al., 1998, p. 773).
The importance of foreign subsidiaries for their ability to access local stimuli and experiences, and from these, create unique, complex and inimitable knowledge, which can produce innovation and contribute to the storage of organizational knowledge, is widely recognized through the literature. In fact, researchers have been investigating subsidiaries through their primary importance in a local responsiveness type of strategy, thanks to their proximity to the local needs of the customers (Doz & Prahalad, 1981 – 1984, Gates & Egelhoff, 1986) and through their recognized ability to create and maintain FSAs as a consequence of their initiatives, as explained by Birkinshaw et al. in their article of 1998, who argued that capabilities “develop over time as a result of past experiences and are
10
subsequently applied to new or related areas of business. To some extent, new capabilities are always being developed, but they typically emerge at the margin of existing capabilities in response to competitive demands” (Birkinshaw et al., 1998, p. 781).
Moreover, subsidiaries have been studied through their evolution process which happens “typically through the accumulation of resources and through the development of specialized capabilities (Hedlund, 1986; Prahalad & Doz, 1981)” (Birkinshaw et al., 1998, p. 774) and is determined by HO assignments, by subsidiary initiative, aimed to define their own role in the company, and by the “local environment determinism” (Birkinshaw et al., 1998, p. 775) or in other words “the constraints and opportunities in the local market” (Birkinshaw et al., 1998, p. 775). To summarize Birkinshaw and Hood argue that “the particular geographical setting and history of the subsidiary are responsible for defining a development path that is absolutely unique to that subsidiary, which, in turn, results in a profile of capabilities that is unique (Teece et al., 1997)” (Birkinshaw et al., 1998, p. 781) therefore “subsidiary evolution is defined in terms of (1) the enhancement/atrophy of capabilities in the subsidiary and (2) the establishment/loss of the commensurate charter” (Birkinshaw et al., 1998, p. 783).
It appears clear how subsidiaries are “increasingly depicted as partially autonomous entities with power to shape strategy in the MNC network […] (and) capable of driving proactive, autonomous, and risk-taking initiatives for local and global applications (Schmid, Dzedek, & Lehrer, 2014).” (Kostova et al., 2016, p. 180).
Nonetheless, as most companies base the biggest part of their revenues and assets abroad, the HQ are presented with a control issue over the subsidiaries (Doz & Prahalad 1981), which is initially exerted through resource or technology dependence, but, as subsidiaries mature and grow in size, they become more autonomous and independent from strategic resources, leaving HQ to find a substitute control mechanism (Doz & Prahalad, 1981). The typical hierarchical control “is an efficient mechanism for coordinating a complex system comprising multiple specialized units” (Grant, 1996, p. 117), but it
11
fails in settings characterized by the need for knowledge creation and integration, the major part of which, is tacit (Grant, 1996).
Therefore, foreign subsidiaries, on one hand, to be source of competitive advantage, need to have a proactive attitude towards the host environment and room to take initiatives, but on the other hand this autonomy puts them in an agency perspective where they are both profit- and rent- seeking, and they aim to increase their bargaining power in function of their contributory role and their embeddedness within a firm’s network (Mudambi & Navarra, 2004). Based on the degree of power they manage to achieve, they can determine the performance of the whole company, as well as become isolated, and therefore a problem (Monteiro et al., 2008). For these reasons, it is important to understand how individual, country- and firm- specific factors affect knowledge creation and transfer and which mechanisms and tools an organization can set up to facilitate it, in a context characterized by many different sources, information asymmetry and potential objective misalignment.
The literature, once again, is a useful starting point: Schmidt & Sofka (2009) found that subsidiaries involved in knowledge sourcing activities suffer from liability of foreignness compared to the local competitors and they highlight the contingencies that are at its base; Miller et al. (2006) added to March’s model direct interpersonal learning, the location of individuals, which differs for local and distant search for sources of learning, and highlighted the tacit dimension of knowledge. As stated earlier, this type of knowledge is probably the most important for every firm as main source of competitive advantage, but, in order to properly exploit it, “the organization has to ensure that individuals and groups who need to interact and work together have similar knowledge capacities” (Goh, 2002, p. 27) since “tacit knowledge may be best transferred through more interpersonal means and using processes that are less structured. Some examples are mentoring, teamwork, chat rooms, personal intranets, and opportunities for face-to-face conversations such as group dialogue or personal reflections on experiences and lessons learned.”(Goh, 2002, p. 27). Lastly in their article of 2010 Edwards & Tempel “examine the phenomenon of reverse diffusion” (Edwards & Tempel, 2010) and
12
its variations and constraints through which foreign operations can diffuse back to the home country newly acquired best practices.
Other authors explored some of the mechanisms that facilitate or mediate knowledge creation and transfer: Foss & Pedersen (2002) found that for a successful transfer, different organizational means are needed for knowledge stemming from diverse sources (Foss & Pedersen, 2002), while Najafi-Tavani et al. in their article of 2014 examined how “the possession of strategic resources (knowledge or embedded relations) increase subsidiary influence only when they are transferred back to the headquarters” (Najavi-Tavani et al., 2014, p. 122). In addition to this, Goh (2002) argued that technology can help knowledge transfer, especially for highly dispersed organizations, since it facilitates information sharing and learning, but it is not sufficient alone because “best practices networks can fall victim to organizational silos and the dominant culture of competitiveness between subunits and work in teams in an organization, especially when resources are scarce.” (Goh, 2002, p. 25). For these reasons the author highlights the need to develop and promote good leadership and a culture of cooperation and collaboration where trust is fundamental to willingly share information between individuals and groups (Goh, 2002) and to access valuable resources, know-how and institutions presented by the host environment (Gammelgaard et al., 2012). Moreover, the author, suggests other tools and mechanisms which can be adopted to improve knowledge transfer, such as: a culture of problem seeking and problem solving, which can be viewed as “an experimenting and innovative culture” that “encourages employees to look for problems as a way to improve the organization” and where “failures in experimentation should be expected and tolerated, and treated as learning lesson by employees and the organization”( Goh, 2002, p. 25-26), and the development of support infrastructures like a reward system or cross-functional teams (Goh, 2002). Moreover, Gammelgaard et al. in their article of 2012, argued that the literature “indicates that the development of intra-organizational networks provides performance boosting effects connected to improved resource development, enhancing learning and innovation, as well as developing entrepreneurial
13
skills and increasing the importance of the subsidiary with the MNC, which enhances investment by parent companies.” (Gammelgaard et al., 2012, p. 1160). Nonetheless these tools can present some downsides too, and that’s the case for trust, as explained by Szulanski et al. (2004), who argue that the “intricate effect of trustworthiness reflects the fact that trustworthiness promotes both functional and dysfunctional processes, fostering receptivity on the one hand and lessening the perceived need for vigilance on the other” (Szulanski et al., 2004, p. 608) therefore, “when casual ambiguity is high, trustworthiness may prove counterproductive” (Szulanski et al., 2004, p. 608).
Lastly, Hedlund, in his article of 1986, proposed the “heterarchical MNC” as the most effective organizational form to “exploit competitive advantages derived from a home country base on the one hand, and actively seeking advantages originating in the global spread of the firm on the other” (Hedlund, 1986). The “heterarchical MNC” is characterized by a “geographically diffused pattern of expertise” which is the “reflection of dissimilarities between countries”, and a more autonomous and strategic role assigned to every subsidiary (Hedlund, 1986).
While there is a tendency to agree upon the effectiveness of socialization mechanisms for the combination and transfer of knowledge, in particular of its tacit dimension, there is still no common agreement on the effect, that an action over the degree of autonomy granted to a foreign subsidiary, has on its ability to create and diffuse knowledge.
In fact, “from an input standpoint, subsidiary autonomy was seen as one factor in a collection of forces that drives the MNC's evolution (Birkinshaw and Hood, 1998). More specifically, it was seen as influencing subsidiary capabilities (Bartlett and Ghoshal, 1986), subsidiary strategy (Bartlett and Ghoshal, 1986; Young et al., 1988), the adoption of a world product mandate (Poynter and Rugman, 1982; Rugman and Bennett, 1982; White and Poynter, 1984), the firm- specific advantages of the MNC (Birkinshaw et al., 1998), strategic evolution (Taggart, 1996), and the creation of local innovations (Bartlett and Ghoshal, 1989).” (Johnston & Menguc, 2007, p. 789).
14
Moreover, as the industry and the subsidiaries mature they become less and less dependent over technologies and strategic resources closing the technology gap with the HQs and becoming more autonomous (Doz & Prahalad, 1981). Similarly, Johnston & Menguc (2007), argued that “increasing subsidiary size will correlate with increasing resources in the subsidiary and a consequent increase in subsidiary autonomy. This positive linear relationship persists until an inflection point is reached and subsidiary autonomy begins to decline. This is due to increasing subsidiary size bringing increasing coordination complexity, a need for greater inputs of managerial experience and expertise, and growing interdependence between the subsidiary and the rest of the corporation.” (Johnston &
Menguc, 2007, p. 787).
Increased autonomy is desired since it fosters initiative and exploration, which in turn leads to new products, processes and services (March, 1991), further it “has been linked to a subsidiary's better access to resources and to greater knowledge creation in the subsidiary (Andersson et al., 2002; Gupta and Govindarajan, 1991; Venaik et al., 2005)” (Rabbiosi, 2011, p. 109). Moreover, increased autonomy was found to be related to improved subsidiary performance, increased inter- and intra- network relationships and knowledge flows (Gammelgaard et al., 2012; Schulz, 2001), increased complexity in R&D and development of new products for local markets (Taggart and Hood, 1999). Further, a higher degree of subsidiary autonomy, was also thought to be “necessary to meet the condition of feasibility” of the innovation process, and to spur new technologies (Ghoshal & Bartlett, 1988). Lastly, Gupta and Govindarajan, in their article of 1991, argued that the degree of subsidiary autonomy is inevitably linked to subsidiary’s role which in turn is defined by “(a) the extent to which the subsidiary engages in knowledge inflows from the rest of the corporation and (b) the extent to which the subsidiary engages in knowledge outflows to the rest of the corporation” (Gupta & Govindarajan, 1991, p. 783). The authors subsequently define four subsidiary roles characterized by increasing strategic importance and by increasing subsidiary autonomy: “Implementor (low outflow,
15
high inflow)”, “Local Innovator (low outflow, low inflow)”, “Integrated Player (high outflow, high inflow)”, and “Global Innovator (high outflow, low inflow)” (Gupta & Govindarajan, 1991, p. 773) Nonetheless, “on one hand, subsidiaries with greater R&D capabilities (like WPMs) may be less technologically dependent on the HQ and hence display higher levels of autonomy (Pearce, 1999; Taggart & Hood, 1999); on the other hand, a counter-argument exists, on the grounds that the strategic sensitiveness of knowledge-related activities leads to tighter control by HQ (Bartlett & Ghoshal, 1989; Martı´nez& Jarillo, 1991)” (Tavares & Young, 2004, p. 221). Further, subsidiary autonomy is usually related to a more “contributory role” (Birkinshaw et al., 1988) of the subsidiary with consequent gains in bargaining power which may restrain it from willingly share information and know-how with the other parts of the MNE in fear of losing their privileged position (Mudambi & Navarra, 2004). For instance, subsidiaries which are competing “for internal resources and external market share” may try “to make private gains in an attempt to outperform the partners (Khanna et al. 1998)” (Asakawa, 1996). In these situations, forms of hierarchical control can be even more detrimental in terms of knowledge transfer (Asakawa, 1996,).
Moreover, as argued by Ciabuschi et al. in their article of 2012 “subsidiaries have objectives and motives not necessarily in congruence with their headquarters or sister subsidiaries” and if they control critical resources they have some degree of power which may “not in itself the objective of subsidiaries, but it does provide a means to pursue opportunities of self-interest.” (Ciabuschi et al., 2012, p. 664).
In addition to this, another issue related to subsidiary autonomy other than rent-seeking attitudes and opportunistic behaviors, is the creation of unrelated or too specialized knowledge as “radical competences may lead to excessive differentiation and take the subsidiary too far from the mainstream evolution of the multinational. The needed ‘organizational isolation’ (Yamin, 2000), appropriate to stimulate subsidiaries to develop distinctive capabilities, may lead to empire building
16
(Birkinshaw & Ridderstra˚le, 1999) and entropic or disintegrative tendencies (Tavares, 2001).” (Tavares & Young, 2004, p. 224).
Lastly, many studies failed to find any direct effect of subsidiary autonomy on the communication with the Headquarters or other subsidiaries (Ghoshal, Korine & Zulanski, 1994; Schulz, 2001; Rabbiosi, 2011). Therefore, it is not exactly clear which degree of subsidiary autonomy is desired in order to efficiently exploit and explore knowledge, even because this effect may be mediated by different elements such as trust in the management, organizational culture and routines HR practices, incentives, etc.
3. Theoretical framework
In this paper, I highlighted the importance of the knowledge-based theory of the firm, of knowledge as a critical resource for competitive advantage, for the innovation process but also as a source of power that can create tension within an organization (Nonaka, 1994, 2000; March 1991, Kostova, 2015; Carlile et al., 2003; Carlile, 2004; etc.). Moreover, I connected this new perspective to organizational theory by showing how knowledge creation and exploitation are among the main drivers of the shift from a hierarchical and bureaucratic structure to a more networked and autonomous one, where foreign subsidiaries cover a fundamental role (see Birkinshaw et al., 1998; Kostova, 2015; Najafi-Tavani, 2014; Doz & Prahalad, 1981,1984; Mudambi & Navarra, 2004; Foss & Pedersen, 2002; Edwards & Tempel, 2010; etc.).
These themes are further supported by Carlile & Rebentisch who argued, in their article of 2003, that MNEs are everyday more involved in complex technological activities and settings, and everyday more in contact with other actors such as suppliers, customers or even allies. Therefore, the common theories on knowledge transfer are no more sufficient since they cannot properly explain the integration of path dependent contributions coming from many different specialty areas in order to
17
create new knowledge (Carlile & Rebentisch 2003).In this context, subsidiaries represent the link between location-specific stimula, like a particular technological setting or a crucial business relationship in a foreign country, and the creation of new knowledge that can be exploited in the whole organization. Therefore, the authors proposed a cyclical framework for the repeated process of knowledge transformation, similar to the one proposed by Nonaka and explained in the literature review, which only through a proper knowledge representation, can provide opportunities for innovations, and it can be considered the way organizations adapt, develop and evolve. It follows that firms need to encourage every part of their structure to interact and to take part in this process while facilitating its understanding and diffusion with any mean possible. In fact, in their research of 2010 which included fieldwork, Edwards & Tempel argued how proper organizational and human resources practices, such as organizational conduits or a network based system, foster reversed diffusion of knowledge and how it increases performance and influence. As it has already been explained throughout this research, there are many arguments supporting the fact that the transfer of knowledge flows from subsidiary to the headquarters has become even more important of the opposite. To corroborate this, Miller et al., in their article of 2006, building upon the work of March (1991), argued that organizations achieve the highest long-run knowledge when there is “fast learning by the code and slow learning from the code” (Miller et al., 2006, p. 714); with the code reflecting the best organizational practices shaping and shaped by the members’ beliefs (Miller et al., 2006). Within this perspective, Headquarters are presented with a dilemma: on the one hand allowing a higher degree of subsidiary autonomy creates room for initiatives and exploration, for instance, an autonomous subsidiary’s response to local competitive pressures, and also fosters the access to local resources and the creation of new knowledge. On the other hand, this new knowledge to be developed and to be of any use for the whole organization has to be transferred, combined with existing knowledge, stored and successively redeployed; such a complex process can be hindered in a context characterized by highly autonomous subsidiaries, since it can trigger competition and opportunistic
18
behaviors impeding knowledge creation and knowledge flows. For these reasons, I hypothesize a positive linear relationship between subsidiary autonomy and knowledge flows, in the sense that, at first, when autonomy is granted, intra-firm knowledge flows are likely to increase until too much autonomy is allowed and knowledge flows are hindered:
H1: When a subsidiary is granted with a higher degree of autonomy, intra-firm knowledge flows are
likely to increase. Past a certain degree of autonomy, intra-firm knowledge flows are likely to stabilize if not diminish.
Moreover, this relationship is expected to be affected by abilities and skills of the focus susbsidiaries’ employees and the trust they have in the management of the company. In fact, in order to have meaningful and successful knowledge flows, the individuals involved in the transfer of knowledge should have similar skills, capabilities and absorptive capacity seen, as previously mentioned, as the ability deriving from prior knowledge to “recognize the value of new information, assimilate it, and apply it to commercial ends” (Cohen & Levinthal, 1990, p. 128). Lastly, a higher degree of trust in the actions of the management team is likely to strengthen the relationship under analysis. Therefore:
H2: The positive relationship between Subsidiary Autonomy and Knowledge transfer is moderated
by Trust in Management, so that the relationship is stronger for higher values of Trust in Management.
H3: The positive relationship between Subsidiary Autonomy and Intra-firm knowledge transfer is
mediated by Employees’ ability, so that when a subsidiary is granted with a higher degree of autonomy, it develops more skilled employees who facilitate the intra-firm knowledge transfer.
19 Figure 2 – Theoretical framework
H2 + H3 + +
H1
4. Method
This study takes a deductive and testing approach and it’s structured as a quantitative research carried out through cross-sectional survey design. This design was chosen to easily test the hypotheses and to be able to generalize the findings to some extent. It consists of one questionnaire, administered online, and sent to MNEs with foreign subsidiaries. The biggest part of the responses were collected from Company A, a forwarding and logistic multinational company with 67 subsidiaries and more than 200 offices all over the world, while smaller samples of responses came from other two MNEs, namely company B and company C. Lastly, after testing the different items and variables, a hierarchical linear regression analysis will be performed and the results analyzed.
5. Sample
The population for this study are employees of foreign subsidiaries belonging to multinational-enterprises. Because of the magnitude of the population and the difficulty in accessing companies, the research is conducted with a non-probability convenience sampling technique. This research aims to collect as many responses as possible. Table 1 reports some relevant information about the sample.
TRUST IN
MANAGEMENT EMPLOYEES’ ABILITY
SUBSIDIARY AUTONOMY
INTRAFIRM KNOWLEDGE
20 Table 1 - Sample Industry Size (N. Employees) Global HQ Location Regional HQ Location N. of Subsidiaries N. of valid Responses Subsidiary location Avg. Age of Respondents Avg. Gender of Respondents Company A Forwarding and Logistics 3.200+ Florence, Italy New York, Shanghai, Tokyo, Singapore, Sao Paulo 67 137 London, Mumbai,
Buenos Aires, Sao Paulo, Mexico City, Panamà, Bangkok, Madrid, Valencia, St. Petersburg, Santo Domingo, Shenzen – Poland, Texas, Illinois, New Jersey, Georgia, Mexico, Japan. 30-50 years old Male Company B Cloud Computing 15.000+ San Francisco, California
Santa Fe, Sao Paulo, London, Singapore 30+ 18 Dublin 30-50 years old Male Company C Medical Technologies 50.000+ Washington, D.C. - 80+ 34 Milan, Verona, Padua 30-50 years old Male
21
6. Measures
6.1 Knowledge transfer
Building upon the work of Minbaeva et al. (2003), intra-firm knowledge transfer, as the dependent variable, will be measured by asking to assess the degree of inbound and outbound knowledge flows. Respondents will be asked: regarding inbound knowledge flows, “To what extent does your subsidiary utilize firm-specific capabilities (e.g. technologies, operational know-how, etc.) from the Head Office and from other subsidiary units?”, while to assess outbound knowledge flows, stemming from the focus subsidiary, they will be asked “To what extent are strategic resources and capabilities (e.g. knowledge, strategic partnerships, local networks, technologies), from your subsidiary, utilized by the Head Office and other subsidiary units?”. Respondents will answer on a five-point Likert scale, where 1 is “No use of resources and capabilities” and 5 is “Substantial use of resources and capabilities”. (See appendix)
6.2 Subsidiary autonomy
Building upon the work of Johnston and Menguc (2007) subsidiary autonomy, as the independent variable, will be captured by asking respondents to assess the influence of the Headquarters and of the subsidiaries on the decision-making process relative to some example scenarios such as the introduction of a product in a local market or hiring operational personnel. Respondents will be asked to indicate on a scale from 1 to 5 whether the decision was taken at the Headquarter level, with or without the influence of the subsidiary, at the subsidiary level, with or without the influence of the Headquarters, or with equal influence by both. (See appendix)
6.3 Trust in management
Trust in management is expected to be a moderator of the main relationships under analysis; in the sense that for higher levels of trust the relationship will be stronger.
22
Trust in management, as the degree of “employee faith in an organization’s goal attainment and organizational leaders and the belief that organizational action will be beneficial for employees” (Renzl, 2008, p. 208), will be measured adapting the Cook & Wall’s scale and the work of Chen, Chen & Xin (2004) by using a seven-point Likert scale to assess the employees’ faith in peers and in management and the employees’ confidence in the actions of peers and of the management, where 1 means firmly disagreeing with the statement and 7 means firmly agreeing with the statement. (See appendix)
6.4 Employees’ ability
Lastly, building upon Minbaeva et al. 2003, employees’ ability, as a mediator of the main relationship, is defined as the potential and the capabilities of a subsidiary. It will be measured by asking employees to assess their colleagues’ ability on three items: interpersonal skills (e.g. communication, teamwork, etc.), job-related skills (required skills specific to certain tasks, e.g. accounting, business development, etc.) and educational level. Respondents will indicate it “on a seven-point Likert scale going from 1="Far below average" to 7="Far above average" for all three items.” (Minabaeva et al., 2003). (See appendix)
6.5 Control variables
In order to control for structural elements of the subsidiary that can affect knowledge transfer some control variables will be included: the gender and age of the respondents, number of subsidiary’s employees as a proxy for the size of the subsidiary and age of the subsidiary, which can relate to the experience which has been accumulated throughout the years.
23
7. Analysis
In order to conduct this analysis, the data, collected through the online survey, have been exported into SPSS and adjusted for the tests. First of all, a constant value was assigned to missing responses and cases were excluded listwise. Secondly, frequencies were run for all the variables, resulting in no error in data entry. The mean of Subsidiary Autonomy is 3,28 meaning that, for the subsidiaries under analysis, decisions are mostly taken with influence of both the Subsidiary and the Headquarters, while the mean of Knowledge Transfer is 3,23 which indicates a somewhat higher than average rate of knowledge inflows and outflows. The two highest mean scores are reported by Trust in Management (3,59) and Employees’ Ability (3,48) which indicate that the employees on average trust their Management and value the skills of their colleagues.
Further, multiple tests were run to check for normality in the variables’ distributions: looking at the graphic tests, which are reported in Table 2 both the variables’ histograms and Q-Q plots seem to be normally distributed, but to get more powerful insights, the descriptives statistics were run on SPSS. Almost all the variables present Skewness and Kurtosis between -.5 and +.5 which indicates a normal distribution. Only Trust in Management presents a Skewness of -.8 and Employees’ Ability a Kurtosis of +.7, which is still in an acceptable range especially because “with reasonably large samples, skewness will not make a substantive difference in the analysis” (Tabachnick & Fidell, 2001, p. 74), and “kurtosis can result in an underestimate of the variance, but this risk is also reduced with a large sample” (Tabachnick & Fidell, 2001, p. 75). For this research more than 200 responses were collected, therefore reducing the possibility that the two statistics have to alter the analysis.
24 Table 2 – Graphic tests
25
Further, the only counter-indicative item, rTrust5, was recoded into Trust5 to have the same scale as the other items. The control variables were manipulated too: Gender was recoded to have only 0 and 1 values while Subsidiary Size and Age were divided in three categories identified by 1,2,3 values (respectively 1= Small subsidiary – 0 to 20 employees, 2= Medium Subsidiary – 21 to 160 employees, 3= Big Subsidiary – 161 to 6000 employees; 1= Newly Established Subsidiary – 0 to 33 years, 2= Established Subsidiary – 34 to 66 years, 3= Mature Subsidiary – 67 to 100 years).
Moreover, a reliability test was run for all the items of the main variables and all reported a Cronbach’s alpha above 0.7 (Subsidiary Autonomy= 0.702, Knowledge transfer=0.737, Trust in Management= 0.857, Employees’ ability= 0.826). Once it was assured that it is a reliable scale, scale means were computed for the different variables and checked for potential outliers and multicollinearity problems. To rule out the latter, the VIF (Variance Inflation Factor) were controlled, referring to Knowledge Transfer as the dependent variable. All tolerance values are close to 1, which is acceptable, and all the VIF present score in the range between 1 and 1.5 which are well below the threshold of 3. Table 3 reports the results of the test.
Table 3 – Multicollinearity statistics for the dependent variable Knowledge Transfer
Model Collinearity Statistics Tolerance VIF Subsidiary Autonomy ,896 1,116 Trust in Management ,686 1,459 Employees’ ability ,732 1,366
a. Dependent Variable: KnwlTOT
Lastly, after filtering out the cases with outliers, a correlation matrix was run to preliminarily check for the effect of interaction between the different variables. From Table 4, which reports the correlation results, it is possible to see that the control variable Subsidiary Age has a significant positive correlation with Subsidiary Size and Employees’ Ability (respectively 0,346 P<0.01 and 0,270 P<0.01) meaning, as expected, that a mature subsidiary tends to be bigger and have more skilled
26
employees. The latter, is also negatively correlated to the control variable Gender (-0,177 P<0.01) and presents positive correlation with Trust in Management (0,487 P<0.01). Lastly, Trust in Management appears to be negatively correlated to both Subsidiary Size (-0,216 P<0.01) and Subsidiary Autonomy (-0,246 P<0.01).
27 Table 4 – Descriptive statistics and Correlation
Variables Mean SD Gender Age Subsidiary
Size Subsidiary Age Subsidiary Autonomy Trust in Management Employees’ ability Knowledge transfer Gender 1,32 0,47 - Age 2,74 1,1 -,209** - Subsidiary Size 1,69 0,70 ,119 ,078 - Subidiary Age 1,12 0,40 -,143 ,061 ,346** - Subsidiary Autonomy 3,29 0,87 ,046 -,046 -,093 -,029 (0,7) Trust in Management 3,59 0,70 -,055 ,007 -,216** -,044 -,246** (0,86) Employees ability 3,48 0,77 -,004 -,177* -,066 ,270** ,037 ,487** (0,83) Knowledge transfer 3,23 0,81 -,016 -,084 ,026 -,031 -,022 ,110 ,135 (0,74)
** Correlation is significant at the 0.01 level (2-tailed) * Correlation is significant at the 0.05 level (2-tailed)
28
7.1 Regression analysis – hypotheses testing
After cleaning the data in preparation for the test of the hypotheses, and in order to study the relationship between Subsidiary Autonomy and Knowledge Transfer a hierarchical regression analysis was performed: in the first step of the model the control variables are introduced to check to what extent they are able to predict the dependent variable; in the second step the independent variable is introduced jointly with the others, in order to control its ability to predict Knowledge Transfer on its own. In the first phase, the model is statistically not significant and explains a very small portion of the variance of Knowledge Transfer (2.8%), and although the model improves considerably by adding Subsidiary Autonomy in the second phase, it remains statistically not significant for p<.05.
Table 5 - Regression analysis and Moderation effect
Dependent Variable – Knowledge transfer
H1 H2
Model 1 Model 2 Model 3
Control Variables Beta Sig. Beta Sig. Beta Sig.
Gender -0.127 0.233 -0.142 0.182 -0.237 0.215 Age -0.135 0.192 -0.167 0.113 -0.116 0.141 Subsidiary Size 0.020 0.852 0.016 0.878 0.092 0.956 Subsidiary Age -0.114 0.274 -0.109 0.296 -0.163 0.439
Independent Variable – Subsidiary Autonomy
Moderator - Trust in Management
-0.148 0.117 -0.127 0.310 Interaction effect -0.156 0.349 zTrust in Management*SubAut Constant 3.873 0.000 4.439 0.000 3.869 0.000 R 0.17 0.22 0.26 R2 0.028 0.049 0.066 Adjusted R2 -0.006 0.007
29
Table 5 reports the results of the regression analysis with an additional third step, which will be described in the following section, where the moderation effect was tested aswell.
Further, to analyze the possibility of a non-linear relationship, another hierarchical regression was performed. In the first step, only the independent variable was introduced and in the second step the quadratic function of the independent variable was added, but this test also resulted in no statistical significance.
7.2 Moderation and mediation effect analysis
Even though the previous analysis proved non-significant, the moderation and mediation effects over the relationship under study, were controlled to see if there is any meaningful influence of Employees Ability and Trust in Management over the main relationship. Both effects were tested through PROCESS by Andrew F. Hayes in SPSS.
Table 6 and 7 report the results of the model when the Mediation variable is introduced and, although it presents a positive effect of Subsidiary Autonomy on Employees’ ability and of the latter on Knowledge Transfer, as theorized, it is statistically not significant for p<.001
Table 6 – Mediation analysis
Employees’ Ability (M) Knowledge Transfer (Y) Coeff. SE p Coeff. SE p Subsidiary Autonomy (X) 0.039 0.076 0.613 -0.160 0.093 0.087 Employees’ Ability (M) ---- ---- ---- 0.277 0.113 0.016 Constant 3.014 0.046 0.000 3.602 0.065 0.000 R2 0.117 0.096
30 Table 7 – Total, Direct and Indirect Effect
Effect SE p
Direct Effect -0.160 0.093 0.087
Total Effect -0.149 0.095 0.117
Boot SE Boot LLCI Boot ULCI
Indirect Effect 0.010 0.023 -0.017 0.083
Table 5 reports the results of the Moderation analysis and once again after introducing Trust in Management as the Moderator variable, the coefficient of interaction presents a positive sign, meaning that the main relationship between Subsidiary Autonomy and Knowledge transfer might be strengthened for higher values of Trust in Management. Nonetheless, the model, after introducing the Moderator, explains less of the variance and it is statistically not significant for p<.001.
8. Findings and discussion
This study examines the relationship between Subsidiary Autonomy and Intra-firm Knowledge transfer which is measured by the inflows and outflows of technology, information, capabilities and know-how. It is argued, in the first hypothesis, that a more operationally and strategically autonomous subsidiary will more likely engage in both knowledge inflows and outflows from and towards the Headquarters and the other subsidiaries of the company. In fact, it is further argued that more room for subsidiaries’ autonomous behavior will likely result in the creation of specialized knowledge which can potentially be a source of competitive advantage for the whole company, and, therefore, a source of power for the subsidiary that control that kind of resources. Once a subsidiary has a relevant position in the MNE’s internal network, it will try to retain the power previously gained by defending its resources and therefore hindering knowledge flows. Nonetheless, through the hierarchical regression analysis it is possible to see that the sign of the effect is opposite to what expected, and,
31
although the introduction of Subsidiary Autonomy generally improves the model, it is not significant, thus lacking explanatory power. Therefore, Hypothesis 1 is not supported and this could be caused by a number of different reasons. First of all, as previously mentioned, in order to create new knowledge, a company might not only need to allow to its subsidiaries more operational and strategic autonomy, but it also needs to possess prior related knowledge (Cohen & Levinthal, 1990). Moreover, it also has to invest in nurturing its organizational and individual absorptive capacity which in turn, generates a path-dependent development of knowledge and capabilities, that is specific to every firm and can influence, among the others, the type of internal control and the transfer of information (Cohen & Levinthal, 1990). In fact, communications between subsidiary units with different absorptive capacity will not be as effective and efficient, and may result in subsidiary isolation or the creation of unrelated knowledge. Therefore, the effect of absorptive capacity on knowledge creation, transfer and integration might explain why the analysis yielded non-significant results and Subsidiary Autonomy cannot properly explain Intra-Firm Knowledge Transfer on its own.
Secondly, is not easy to capture both the dependent and independent variable since they are both very wide concepts that every MNE can manage differently. In fact, many firm- and country- specific factors, not controlled or included in this study, can affect both dimensions in many different ways. For istance, as argued by Rabbiosi in her article of 2011, different degrees of subsidiary autonomy, may depend on the role that the Headquarters decide to assign to every subsidiary, which is also closely related to decision over which is the best internal control instrument (Rabbiosi, 2011). Moreover, authors Najafi and Tavani, argued in their article of 2004, that subsidiaries can successfully increase their relevance in their company’s internal network only when they actively engage in the transfer of strategic resources, such as knowledge or embedded relations, towards the Headquarters; therefore, contrasting with the agency perspective previously mentioned (Najafi & Tavani, 2004). Moreover, the possession of critical resources, and the recognition of a subsidiary as a center of value creation, might result in even tighter control by the Headquarters and less subsidiaries’ independence in the attempt to integrate these units in the company’s “global network of innovation, production
32
and distribution” (Frost et al., 2002), which is in contrast with this study assumptions (Frost et al., 2002; Bartlett & Ghoshal, 1989; Martı´nez & Jarillo, 1991; Tavares & Young, 2004). Further, there is support for the influence of the type of knowledge to be transferred, explicit or tacit, and the mean of transfer on the effectiveness of the knowledge transfer itself (Goh, 2002; Nonaka & Takeuchi, 1995). In fact, even though there is strong support for the beneficial effect of autonomy on knowledge creation (March, 1991; Rabbiosi, 2011; Ghoshal & Bartlett, 1988), this new knowledge might prove too difficult to transfer or useless for the company. Accordingly, in the literature, there is still no agreement over the real effect of a higher degree of subsidiary autonomy on the subsidiaries’ abilities to transfer the new specialized knowledge, and many studies support both sides and also the absence of any significant effect. For instance, the authors Goshal, Korine and Szulanski, in their article of 1994, found that “while subsidiary autonomy has almost no effect on either subsidiary-headquarters or inter-subsidiary communication, interpersonal networking has significant positive effects on the ongoing communication of subsidiary managers, both with their counterparts in the headquarters and with managers in other subsidiaries” (Goshal, Korine & Szulanski, 1994, p. 106). Furthermore, even though the regression analysis with the quadratic function of Subsidiary Autonomy proved non-significant, the main relationship under study is potentially much more complex than hypothesized and might be non-linear. Lastly, it is very difficult to gather enough relevant and non-biased data through a survey, since companies usually do not wish to disclose such information and employees might feel pressured to respond in a way they are “expected to”.
Hypotheses 2 and 3 argued that the main relationship, between Subsidiary Autonomy and Knowledge Transfer, is mediated by the ability and skills of the subsidiaries’ employees and moderated by their trust in the firm’s management. In fact, it is argued that a more autonomous subsidiary would develop more skilled employees, corroborated by the literature, that supports the positive effect of subsidiary autonomy on efficiency and performance (Gammelgaard et al., 2012; Schulz, 2001); and in turn more skilled employees are thought to have a better internal communication and engage more in intra-firm knowledge transfer (Kostova et al., 2015; Gammelgaard et al., 2012 ; Goh, 2002). Moreover, it is
33
hypothesized that when employees trust their management’s decisions and actions, the effect of autonomy on knowledge inflows and outflows is stronger.
Both the mediator and moderator analysis present a p value far above .05 so that both hypotheses are not supported, but the effect’s sign is the same as theorized. Again, similarly to the previous case there might be several reasons for the statistical non-significance of the model and first of all the fact that the analysis was run over the regression model that already presented non-significant results. Moreover, context-specific factors such as the industry, the market, the geographical area, the companies’ culture, but also the firms’ cumulative absorptive capacity, may come into play influencing the relationship under study in specific cases. Lastly, the sample size isn’t optimal for this kind of analysis and it is difficult to gather non-biased data.
9. Theoretical, practical implications and future research
So far, the literature, has widely researched the importance of knowledge for MNEs in this evolving, fast-paced and globalized environment; but it often overlooked how knowledge is transferred in order to be integrated, combined and re-deployed. In fact, many studies took for granted the act of transferring information, technologies, know-how, etc., without exploring its contingencies and constraints. Also, the Headquarters-subsidiary relationship has been widely studied, and it is one of the basis of organizational theory, but the role of individuals and the role of knowledge, which is shaping modern companies’ structures, has not been deeply analyzed yet. This study connected these two perspectives, with the intention to fill some of these gaps and give more insights on the various mechanisms and factors that affect at the same time the internal knowledge integration process of modern firms and their organizational structure. In fact, it is argued that the creation of new knowledge, and its use throughout organizations, is source of sustainable competitive advantage and, in order to control it, MNEs are reshaping their internal structures and redefining the role of subsidiaries. These subsidiaries, for their privileged access to strategic resources and relationships,
34
are able to create location-specific knowledge that, with the proper means and tools, can be recombined and find useful generic applications. Nonetheless, it is not clear which means and tools can actually improve knowledge creation, transfer and recombination. Therefore, even though all the hypotheses didn’t find support in the data, this study, explored some of the most important mechanisms that can be used to transfer internally different types of knowledge such as socialization and IT systems. Moreover, it can serve as basis for future research which may include more of the firm- and country-specific factors which influence the different steps of the knowledge integration process. In fact, it would be interesting to understand exactly how managers can take fully advantage of the dynamics internal to their companies, and successfully build sustainable competitive advantages which are initially created at the subsidiaries’ level.
Based on this literature review and the results of the analysis it is also possible to draw some practical implications for MNEs and MNEs’ management. In fact, this study highlighted the importance for firms to shape their internal organization towards a more bureaucratic structure, which allows subsidiaries to better and more freely engage in exploration of the local environment, and tap into pools of strategic resources which consequently permit to develop new capabilities and technologies. Nonetheless, since the new knowledge is created at the companies’ margins, having a proper organizational structure alone it is not sufficient. In fact, MNEs also need to develop mechanisms which allow managers to effectively improve, handle and control intra-firm knowledge flows, so that potential competitive advantages don’t get overlooked. This study, proposed Subsidiary Autonomy as a potential tool to encourage individuals to create and exchange more often strategic information and capabilities. Again, none of the hypotheses were statistically supported, which may indicate that a higher degree of autonomy granted to a subsidiary doesn’t necessarily affect the above-mentioned transfer of knowledge, or that, it has to be studied together with more, if not all, of the other firm- and location-specific factors that may influence the relationship. This could also be the basis for future studies, which aim to find all the leverages that managers can set up in order to improve knowledge
35
flows, firm competitive advantages and ultimately to exploit innovations at the subsidiary level, which would otherwise find very limited application. Lastly, comparisons between countries and companies would prove of great interest for similar studies.
10. Limitations
The limitations of this study are mainly related to its quantitative nature and to the method of data gathering. First of all, as mentioned before, companies usually do not wish to disclose data over their subsidiaries and subsidiaries’ employees so that it was really difficult to gather enough respondents for this study and therefore the sample might not be optimal. Secondly, this study, because of the chosen sampling technique, is able to collect a large amount of data in a short period of time; but it might present some issues of representativeness and generalizability. Also, the fact that all the responses came mainly from one company and just a few others, hinders the study’s external validity. Moreover, its cross-sectional design doesn’t allow to measure changes over time and all the questionnaires to some extent can be biased because of social desirability type of pressures. Lastly, this research didn’t include other potentially important variables for the relationship under study, such as the mean used for the knowledge transfer and the type of knowledge to be transferred, which ultimately might be the reason for the low explanatory power of the model.
11. Conclusions
Many authors have argued over the effect of an higher degree of subsidiary autonomy over a range of dimensions such as subsidiary performance (Gammelgaard et al., 2012; Schulz, 2001), subsidiary capabilities (Bartlett and Ghoshal, 1986), subsidiary strategy (Bartlett and Ghoshal, 1986; Young et al., 1988) and subsidiary innovation (Ghoshal & Bartlett, 1988). Nonetheless, little study has been conducted on the effect of subsidiary autonomy on intra-firm knowledge transfer and the few arguments are usually contrasting. In fact, Edwards and Tempel, in their article of 2010, similarly to Najafi -Tavani et al. (2012), argued that, only when the new knowledge created at the subsidiary level
