Diffusing MNE subsidiary initiatives across national cultural distance: The role of organizational values sharing and knowledge sharing

Williams, Christopher; Kumar, Maya

doi:https://doi.org/10.7202/1095747ar

An important source of value creation in the MNE is managers’ entrepreneurial behavior within the firm’s dispersed foreign subsidiaries. Scholars recognize such activities as subsidiary initiatives[1]: “entrepreneurial proactive behavior in organizational subunits aiming to influence strategy-making in the organization” (Strutzenberger and Ambos, 2014: 314). Subsidiary initiatives can make an impact throughout the MNE, not just in the originating subsidiary. For this to happen, the initiatives must diffuse throughout the MNE. This diffusion involves them being first understood, agreed upon and approved by headquarters, and then adopted in other parts of the organization (e.g., Birkinshaw, 1997; Dörrenbächer and Gammelgaard, 2016; Edwards et al., 2015).

While subsidiary initiative research has been plentiful, it has tended to focus on understanding the process of subsidiary initiative diffusion through the lens of the MNE’s internal organizational environment (Birkinshaw, 1999; Dörrenbächer and Gammelgaard, 2016; Schmid, Dzedek, and Lehrer, 2014; Strutzenberger and Ambos, 2014). In particular, research has focused on mechanisms within the firm that foster social integration between employees, rooted in informal, social networks and relationships (Bartlett and Ghoshal, 1998), the internal political environment (Dörrenbächer and Gammelgaard, 2016; Mudambi and Navarra, 2004), effective communications (Birkinshaw, 1997), and appropriate control mechanisms (Nohria and Ghoshal, 1994).

In contrast, there has been limited attention paid to factors in the organization’s external environment that influence subsidiary initiatives. At the country level, prior research on the external environment has considered how interconnected the host country is in the global economy (Edwards et al., 2015), the strategic importance of the host country, the relative cost of operating in the host country, and host government support (Schmid et al., 2014). While such host country factors clearly matter, distance[2] between home and host countries, an omnipresent characteristic of MNEs that impacts practices and performance (e.g., Johanson and Vahlne, 1977; Kostova and Zaheer, 1999; Xu and Shenkar, 2002), has been overlooked in the subsidiary initiative literature. In particular, national cultural distance, which continues to be at the forefront of international business research (Beugelsdijk et al., 2018; Cuypers et al., 2018) has not gained attention by subsidiary initiative scholars.[3] This is problematic because national culture creates the context in which relationships develop, social interactions occur, and subsequent knowledge sharing takes place. It shapes how new knowledge is created and distributed (De Long and Fahey, 2000). With increasing national cultural distance, information asymmetry increases, disrupting knowledge flows in the MNE (Gong, 2003; Meyer et al., 2020). In our view, this can also disrupt subsidiary initiative diffusion. Despite the potential of national cultural distance to negatively affect subsidiary initiative diffusion, there is an absence of research to-date on this topic. Moreover, while there is substantial research into the internal organizational environment and subsidiary initiatives (Dörrenbächer and Gammelgaard, 2016; Strutzenberger and Ambos, 2014), there is a lack of research on how internal organizational MNE practices and external national cultural distance interact to affect subsidiary initiative diffusion.

To address this gap, we investigate – as a primary focus - the role of national cultural distance on subsidiary initiative diffusion, as well as the direct effects of internal values sharing and knowledge sharing within the MNE, and their moderating effect in the presence of national cultural distance. Then, as a secondary post-hoc exploration, we consider the effects of sub-dimensions of national cultural distance on subsidiary initiative diffusion. Data from a survey of 111 subsidiaries shows that national cultural distance negatively impacts subsidiary initiative diffusion – an effect not previously identified in subsidiary initiative literature - and that knowledge sharing (but not values sharing) has an important moderating effect of dampening this relationship. We contribute to the literature in several ways. Firstly, we extend the literature on MNE subsidiary initiatives by demonstrating the role played by external national cultural distance. Secondly, our study contributes to the subsidiary management literature by illustrating how knowledge-related organizational practices (knowledge sharing) within the MNE - as opposed to social control-related mechanisms (values sharing) - allow MNEs to harness rejuvenating capabilities in culturally distant locations. Thirdly, our exploratory examination of sub-dimensions of national cultural distance reveals differences in power distance between MNE headquarters and subsidiaries to have the strongest effect on subsidiary initiative diffusion amongst the national cultural distance sub-dimensions. This raises new discussion points for research on power and subsidiary initiatives in MNEs (Dörrenbächer and Gammelgaard, 2016).

Theoretical framework and hypotheses

Subsidiary Initiatives in MNEs

An important stream of research in international business and management has examined how foreign subsidiaries of MNEs identify and develop entrepreneurial opportunities and start initiatives quite independently from headquarters (Birkinshaw, 1997). The literature has shown how this process allows MNEs to gain competitiveness by exploiting new opportunities throughout their global network (Andersson and Pahlberg, 1997). An aspect of subsidiary initiatives that has captivated researchers is how they eventually diffuse outward from the subsidiary and become applied in a broader market context (Birkinshaw, 1997; Dörrenbächer and Gammelgaard, 2016; Williams and Lee, 2011a). Through this process, subsidiaries gain influence and shape the overall strategy of the MNE (Ambos, Andersson, and Birkinshaw, 2010). Generating initiatives for use elsewhere in the MNE helps a subsidiary increase its knowledge output and bargaining power (Mudambi and Navarra, 2004). This diffusion typically requires headquarters’ stamp of approval and can be influenced by differing objectives between headquarters and subsidiaries (Ghoshal and Nohria, 1989) as well as historical relationships and resistance to past subsidiary initiatives (Birkinshaw, 1999). Reviews highlight how scholarly attention to subsidiary initiatives has mainly had an internal focus (Schmid et al., 2014; Strutzenberger and Ambos, 2014). These findings are unsurprising, given the importance of internal issue-selling and internal relationships between headquarters and subsidiary managers necessary for subsidiary initiative diffusion (Dörrenbächer and Gammelgaard, 2016).

National cultural distance and subsidiary initiatives in MNEs

There has been an absence of scholarly work on the role of external national cultural distance on the diffusion of subsidiary initiatives in MNEs. We find this surprising given the high likelihood that the host country context will be somewhat culturally different from that of the home country. The subsidiary managers’ cultural context, where they identify the opportunity and develop an initial response to it, will differ from that of the headquarters managers’ cultural context, who may be required to assess the initiative and sanction it for deployment in other subsidiaries.

Culture[4] is a “collective mental programming distinguishing the members of one group or category of people from others” (Hofstede, 2001: 9). While culture may be studied at different levels, it is most commonly considered at the national and organizational levels (e.g., Tian et al., 2018). In particular, national culture and national cultural distance (Beugelsdijk et al., 2018) are central factors in the international business field, focusing on the values in a country’s society (Hofstede, 1980). National culture is linked to activities associated with growth-focused initiatives in organizations such as innovation (see Tian et al., 2018 for review) and entrepreneurial activities (e.g., Bogatyreva et al., 2019). National cultural distance refers to the degree to which individuals raised and immersed in different countries differ in terms of norms, values, beliefs, and assumptions. These differences are revealed through the actions and attitudes of members of society and have major impacts on MNEs (Beugelsdijk et al., 2018). National cultural distance influences the perceptions of the value of knowledge, the context in which it is encountered, and the processes in which it is dealt with (De Long and Fahey, 2000). A larger national cultural distance between headquarters and a foreign subsidiary means greater differences in managerial and organizational practices and more uncertainty concerning these practices. Such differences create greater inefficiencies and decreased cross-border knowledge transfer effectiveness (Bhagat et al., 2002; Gong, 2003). Meyer et al.’s (2020) review of the subsidiary management literature draws attention to how cultural differences between home and host countries are obstacles to knowledge sharing within the MNE. Also, national cultural distance constitutes a contextual factor that can make relationships between managers arduous (Gong, 2003). Knowledge can be ‘sticky’ due to these challenging relationships, creating ambiguity and misinterpretation when transferring knowledge (Dellestrand and Kappen, 2012; Holden and Von Kortzfleisch, 2004). In particular, stickiness tends to have the greatest impact on knowledge with a tacit component and knowledge embedded within social networks (Kogut and Zander, 1992) - the type of knowledge underpinning subsidiary initiatives (Williams and Lee, 2011a).

Employees’ actions and attitudes also tend to reflect the countries’ cultural contexts in which the organization is located (Schneider and De Meyer, 1991). In other words, both headquarter and subsidiary manager responses to the situations and issues facing them are reinforced by their own national identity and cultural characteristics. This has implications for the way opportunities are identified and exploited, where managers identifying opportunities in one cultural context will not necessarily appreciate opportunities in another cultural context (Muzychenko, 2008). With larger national cultural distances, the combined effects of (1) different perceptions of the value of knowledge (De Long and Fahey, 2000), (2) decreased effectiveness of knowledge transfer (Bhagat et al., 2002; Meyer et al., 2020) and (3) different behavioural norms between headquarters’ and subsidiary managers in terms of opportunity identification (Muzychenko, 2008; Schneider and De Meyer, 1991) will mean the issue-selling and persuasion process that is part and parcel of subsidiary initiative diffusion (Dörrenbächer and Gammelgaard, 2016) will be less effective. There will be higher misunderstandings and misinterpretations concerning the entrepreneurial opportunity and the subsidiary’s proposed solution. Consequently, there will be a higher propensity for the subsidiary initiative to be rejected when there is a lack of equivalence in attitudes associated with the national cultural distance. Hence:

H1: The greater the national cultural distance between home and host countries, the less likely a subsidiary will generate initiatives that receive sanctioning from headquarters to be adopted elsewhere in the MNE.

Internal organizational values and knowledge sharing in the MNE

When understanding behaviour in any type of organization, the internal organizational culture plays a role. Organizational culture is defined as: “the pattern of shared values and beliefs that help individuals understand organizational functioning and thus provide them with the norms for behaviour in the organization” (Deshpande and Webster, 1989: 4). It influences individuals’ behaviours and attitudes within organizations (Schein and Schein, 2016). Following management control theory (Eisenhardt, 1985; Merchant and van der Stede, 2007; Ouchi, 1979) and knowledge-based theory (Grant, 1996; Gupta and Govindarajan, 2000; Kogut and Zander, 1992; Szulanski, 1996), we argue that two principal internal organizational mechanisms intrinsic to an organization’s culture have impacts on subsidiary initiative diffusion. These internal organizational mechanisms are (1) values sharing (normative – or social - integration) as a basis of control (Hedlund, 1993; Nohria and Ghoshal, 1994) and (2) the knowledge sharing norms that underpin subsidiary initiatives (Garcia-Pont, Canales, and Noboa, 2009; Yamin and Andersson, 2011).

Firstly, instilling a common set of norms and values through values sharing is a way to control agents in an organization. Values sharing acts as a social control mechanism that is part of the corporate culture (Jaeger, 1983; Nohria and Ghoshal, 1994). It reinforces a common worldview in the MNE that allows the organization’s principles, interests, and goals to be internalized by employees (Eisenhardt, 1985; Merchant and van der Stede, 2007). The establishment of common values also encourages subsidiaries “to pursue the interests of the MNC as a whole and not just their partisan interests” (Nohria and Ghoshal, 1994: 494). It can be beneficial for different MNE units to cooperate and learn from each other to exploit the knowledge that already exists in the organization (Tsai, 2002).

Firms use corporate training programs, rotation and transfer of expatriate managers, extensive and open communication, and socialization practices to share corporate-wide values and goals and build standard corporate behaviour (Hofstede et al., 1990). These practices breed trust and motivate employees to interact and explore new ideas without fear (Lee and Williams, 2007), build mutual interdependence (Nohria and Ghoshal, 1994), and facilitate the development of a climate of collaboration and mutuality aimed at advancing common goals and generating innovative ideas (Hedlund, 1993). With subsidiary managers’ values more closely aligned to headquarters managers, the subsidiary initiatives developed are more likely to appeal to headquarters managers and the MNE at large, and subsidiary initiatives are more likely to be tolerated by headquarters’ managers. Hence:

H2: Subsidiaries whose managers share values with managers in corporate headquarters are more likely to generate initiatives that receive sanctioning by headquarters to be adopted elsewhere in the MNE.

Secondly, despite efforts by MNEs to control foreign subsidiaries using values sharing, disparate business units do not necessarily understand others’ situational contexts. Improving this understanding requires knowledge to be shared and knowledge embedded in the organization to be accessed by various dispersed individuals. Knowledge-based theory portrays the firm as a repository of knowledge, adept at creating, transferring, and internally linking together domains of knowledge that give the firm a competitive advantage (Granovetter, 1985; Grant, 1996; Kogut and Zander, 1992). Corporate socialization mechanisms have been found to influence knowledge transfer between firm units (Björkman, Barner-Rasmussen, and Li, 2004). Effective knowledge sharing in the MNE occurs when there are norms for face-to-face meetings, open communication channels, and motivation and willingness to disseminate and accept new knowledge (Gupta and Govindarajan, 2000).

The construct of knowledge sharing is different from values sharing. While values sharing emphasizes actions, processes, and inclinations to build normative integration across the enterprise, knowledge sharing is based on social relations between entities in an organizational network (Maurer, Bartsch, and Ebers, 2011). The interconnectedness of subsidiary managers with other units in the MNE reflects the subsidiary’s ability to absorb new knowledge through internal relationships (Yamin and Andersson, 2011). Knowledge sharing allows problems and opportunities faced by other MNE units to be internalized by a focal subsidiary as it pursues a new initiative. While some scholarly attention has focused on relations with external business organizations and the impact of such relationships on MNEs (Yamin and Andersson, 2011), scholars have increasingly emphasized internal embeddedness across MNE units (Garcia-Pont et al., 2009). Tight coupling between units can boost internal knowledge transfer within the MNE (Gupta and Govindarajan, 2000). Hence:

H3: Subsidiaries whose managers engage in knowledge sharing with other MNE units are more likely to generate initiatives that receive sanctioning by headquarters to be adopted elsewhere in the MNE.

The moderating role of values sharing and knowledge sharing

A growing body of literature investigates the interplay between national culture and organizational culture in organizations (e.g., Alofan, Chen, and Tan, 2020; Dastmalchian, Lee, and Ng, 2000; Webster and White, 2010). As noted previously, the growing interest in culture suggests that independently, each type of culture substantially impacts organizational behaviour, including that related to subsidiary initiative diffusion in MNEs. We further propose that together, the interactions of these cultural mechanisms (both present in cross-cultural organizational settings such as MNEs) can elicit either consistent or contradictory behaviours related to subsidiary initiatives being accepted by headquarters.

As noted above, national cultural distance can lead to intercultural conflicts that disrupt processes (Kumar and van Dissel, 1996), preventing the effective sharing of knowledge between cultural contexts (Bhagat et al., 2002; De Long and Fahey, 2000) and creating discrepancies in how opportunities are viewed (Muzychenko, 2008). As Tylor stated, culture is complex because it encompasses “knowledge, beliefs art, morals, law, customs and any other capabilities and habits acquired by man as a member of society” (Tylor, 1924: 1). Firstly, we argue values sharing within the MNE can help overcome problems created by national cultural distance. From a management control perspective, by instilling a common set of norms across the organization, the MNE attempts to establish what has been termed a “strong [corporate] culture” (Deal and Kennedy, 2000). This culture is embedded in the values of the firm and its shared practices. Strong corporate culture can be a powerful way to guide managerial behaviour and action. Dispersed employees are encouraged to internalize the organization’s goals and values (Eisenhardt, 1985) and are influenced by these goals and values in terms of how they act. A firm’s values include the informal and unspoken institutions and rules of a firm. When employees internalize these values, they are considered a “functional member” of an organization (Jaeger, 1983), motivated and committed to the organization. Having functional members of the organization that share and accept corporate values is affected by the monitoring of organizational practices based on these values, the organization’s selection of employees, and the induction and training processes that encourage employees to internalize shared corporate values (Jaeger, 1983). Values sharing between headquarters and subsidiaries builds trust, reduces ambiguity for subsidiaries (Kostova and Roth, 2002), and develops a shared understanding and commitment to the firm’s initiatives. These aspects of values sharing within the MNE should have a potent effect on overcoming national cultural distance when it comes to subsidiary initiative diffusion. Headquarters and subsidiary managers will be able to set aside cultural differences and be more likely to work together to overcome misunderstandings and misinterpretations regarding the substance of the initiative. Both sides will appreciate how the subsidiary initiative contributes to the broader strategic interests of the MNE, rather than just that of the subsidiary. Given that values sharing enhances cultural competence within the MNE, it is more likely that agreement on identifying and accepting a subsidiary initiative would be achieved (Muzychenko, 2008). Hence:

H4: Values sharing by managers in a subsidiary with managers in corporate headquarters dampens the negative relationship between national cultural distance and the diffusion of initiatives out of the subsidiary to the wider MNE.

Secondly, from the knowledge-based perspective, by explicitly encouraging norms for knowledge sharing between managers, the MNE can enhance both disseminative and absorptive capacities of subsidiaries. This will help alleviate the negative impact of national cultural distance on subsidiary initiative diffusion. The arduous relationships that can form between sender and receiver and that contribute to knowledge sharing failure in MNEs can be overcome (Kostova, 1999).

More opportunities for managers from different cultures to gain exposure to each other will gradually improve managers’ cultural sensitivity and cultural competence. This increases their ability to overcome national cultural differences by adapting to other cultures (Johnson, Lenartowicz, and Apud, 2006). Cultural competence improves managerial success in international business. When there is a high national cultural distance, the MNE’s efforts to allow and encourage managers to interact and engage in informal knowledge flows become increasingly important. Such increased opportunities for knowledge sharing will enable managers to understand behavioural norms and viewpoints in different national cultures and appreciate how opportunities are identified. Knowledge sharing will reduce misunderstandings and inappropriate behaviours, acting as a form of “on-the-job” training, used to enable factual, conceptual, and attributional knowledge of different cultures to be shared (Leiba-O’Sullivan, 1999). This can lead to a shared understanding of the situational context and commercial issues facing other MNE units and promotes knowledge flows that underpin a subsidiary’s initiatives. Particularly with a high national cultural distance between headquarters and a subsidiary, without such efforts at enhanced knowledge sharing, knowledge of opportunities and ways of pursuing them will remain isolated in the subsidiary (Williams and Lee, 2011b). Hence:

H5: Knowledge sharing by managers in a subsidiary with managers in other MNE units dampens the negative relationship between national cultural distance and the diffusion of initiatives out of the subsidiary to the wider MNE.

Method

We tested these hypotheses on data collected from a global questionnaire survey of middle managers in wholly-owned MNE subsidiaries. This data was complemented with secondary data on national dimensions for each host country within the sample. Middle managers were selected as they are vital in progressing entrepreneurial initiatives in MNEs, have capabilities to connect and work with various levels in the organization, and can combine macro information of the organization with micro information (Williams, Colovic and Zhu, 2017). We conducted a pilot questionnaire by post, followed by a full survey by purposive networking, drawing on personal business and professional networks to recruit respondents. This approach allowed us to create a heterogeneous dataset in terms of home and host country, as well as primary industry. A total of 150 questionnaires were returned. Due to missing values against our variables of interest and cases where the subsidiary was located in the same country as headquarters, 39 were omitted from the analysis. The final sample of 111 subsidiary managers in 111 subsidiaries of 84 MNEs covered 15 countries of origin, 17 host countries, and 21 industries. There was a mix of well-known diversified MNEs and less-known specialized MNEs. In terms of employees, at the time of the questionnaire response, 24 cases (21.6%) had more than 100,000 employees, and 33 cases (29.7%) had less than 5,000 employees. Respondents had various functions: country managers, marketing directors, sales directors, project managers, training managers, and internal strategy consultants. Table 1 shows headquarters (home) and subsidiary (host) countries of respondents.

Measures

The dependent variable was measured through four items and reflects the impact of new ideas originating in the subsidiary. The items include: the subsidiary starting initiatives which tend to be adopted by a regional or global HQ, the subsidiary generating ideas for new projects which eventually get the go-ahead and investment from HQ, the subsidiary starting initiatives which tend to be adopted by other subsidiaries, and managers in the subsidiary developing new initiatives that become credible projects and are approved by an HQ (Birkinshaw, 1997, 1999; Williams and Lee, 2011a) (Cronbach’s α = 0.79) (Table 2).

Table 1

**Location of headquarters and subsidiaries in the sample**

The measures for national culture for both home and host county were taken from the Hofstede index. Hofstede’s measures continue to be used in the field of international business study (Bogatyreva et al., 2019; Dellestrand and Kappen, 2012) and have received validation against alternative indices (Drogendijk and Slangen, 2006). National cultural distance between home and host countries was measured with Kogut and Singh’s formula (1988) using the four initial Hofstede sub-dimensions: power distance, individualism, masculinity and uncertainty avoidance. We used this composite national cultural distance index in line with other studies of overall cultural distance (Beugelsjik et al., 2017). Though formally defined by Kogut and Singh over three decades ago (1988), it continues to be “of pivotal importance for future studies in international business and beyond” (Cuypers et al., 2018: 1139) and has gained wide acceptance in the international business literature (Shenkar, 2012).

Values sharing was constructed from four items: control through normative integration (Ghoshal and Nohria, 1989), reducing the divergence of preferences and interests among members of the organization (Nohria and Ghoshal, 1994), creating trust bonds between managers in headquarters and subsidiaries, and supporting headquarters’ initiatives as a consequence of the shared relational context in the MNE (Kostova and Roth, 2002) (Cronbach’s α = 0.77) (Table 2).

Knowledge sharing was constructed from four items that relate to the process of knowledge exchange between parties through qualitative coordination (Granovetter, 1985). This includes promoting informal communication between subsidiaries, face-to-face meetings between subsidiaries, and readily accepting and sharing new ideas with other subsidiaries without putting up barriers (Cronbach’s α = 0.61) (Table 2).

We included a number of control variables. MNE size was controlled for using the natural logarithm of the number of employees. Size may be associated with resources and capabilities available for entrepreneurial ventures. Size has also been found to affect reverse knowledge transfer (Yang, Mudambi, and Meyer, 2008). We also controlled for the service industry. Services have been globalizing much more rapidly than manufacturing. A dichotomous variable was used (non-services = 0, services = 1). We controlled for a range of home country features forming the national context for the headquarters decision to approve subsidiary initiatives. Given the U.S. has a high entrepreneurial nature, we controlled for the U.S. as home country with a dichotomous variable. We also included home country GDP per capita, and home country openness to foreign markets using three items: extent of foreign technology absorption and extent of foreign labour in the workforce taken from the World Economic Forum’s Global Competitiveness Report, as well as the country’s attitude towards globalization taken from the Institute of Management Development World Competitiveness Yearbook (IMD, 2005).

At the host country level, we accounted for economic development; subsidiaries operating in more economically developed countries are exposed to a broader range of technological and business opportunities (Birkinshaw, 1997). We also controlled for the host country’s strategic importance to the MNE as this may be a determinant of the subsidiary’s likelihood of generating initiatives (Bartlett and Ghoshal, 1998) and breadth of subsidiary capabilities. The latter is an indicator of knowledge generation and bargaining power for a subsidiary (Mudambi and Navarra, 2004).

Data quality and analysis

We tested and found no significant differences when comparing late responses to early responses, respondents with and without international placement experience, or respondents with low and high employment tenure. To detect potential common method bias, we performed Harman’s single-factor test for all dependent and independent variables (Podsakoff et al., 2003). The result showed less than 50% of the variance explained by the first factor. Secondly, the questionnaire items used for the scales were placed in non-sequential positions to limit the impact of question context and item embeddedness. Thirdly, other variables were collected from secondary sources or calculated as a distance, thus reducing the likelihood of common method bias. A Confirmatory Factor Analysis (CFA) was conducted on the survey scales to test for model fit. We first assessed unidimensionality, which conveys the degree to which items reflect a given construct and only that construct. The complete measurement model comprised a total of 12 items (χ² =67.67; df=51; p=0.06; RMR = 0.06; RMSEA = 0.06; pclose = 0.39; CFI = 0.96; TLI = 0.95; CD = 0.97, Hu and Bentler, 1999). Table 2 shows the questionnaire items used to produce the scales.

Table 2

**Scale construction and evaluation of the measurement model**

The χ² value of the overall model is not significant at the 95% level. The p-value should ideally be statistically insignificant (p>0.05) to claim unidimensionality. The root mean residual (RMR) falls within an acceptable range for the complete model (less than 0.07) (Bagozzi and Yi, 1988). The RMSEA considers the error of approximation in the population and is sensitive to the number of estimated parameters in the model. Our values, between 0.05 and 0.08, are acceptable. The comparative fit index (CFI) and Tucker-Lewis index (TLI) are both greater than 0.9. This analysis shows we have an acceptable fit, and the model fits the empirical data well. We tested the measurement model on various measures of validity and reliability. Values sharing had adequate individual item reliability (squared multiple correlation coefficients greater than 0.30). Two of the knowledge sharing items and one of the subsidiary initiative diffusion items did not meet this criterion. These items, however, still made theoretical sense to include. Practices of informal communication and face-to-face meetings both promote knowledge transfer and we retained them in the construct (Bagozzi and Yi, 1988). Cronbach’s α scores were good for subsidiary initiative diffusion and values sharing (α_SID= 0.79; α_VS= 0.77) and acceptable as a developmental scale for knowledge sharing (α_KS= 0.61). We proceeded with this construct as previous literature in the field of knowledge transfer in MNEs has reported Cronbach α’s between 0.6 and 0.7 (Minbaeva, 2007).

Convergent validity was assessed through average variance extracted (AVE). This considers the degree to which two or more distinct items converge on the same construct and is recommended to be above 0.5 (Bagozzi and Yi, 1988). Two of our latent variables are acceptable as they are above 0.5 or very close (values sharing, AVE = 0.52; subsidiary initiative diffusion AVE = 0.49), while one is markedly lower (knowledge sharing, AVE = 0.30). Discriminant validity can be demonstrated if the square root of AVE is larger than the correlation with items belonging to other constructs. In this situation, each construct has more internally extracted variance than shared variance with another construct (Fornell and Larcker, 1981). The square root of AVE values for values sharing (0.72) and subsidiary initiative diffusion (0.70) are larger than the corresponding correlation coefficients, demonstrating discriminant validity. However, this is borderline for knowledge sharing where the square root of AVE is 0.55 and the correlations with values sharing and subsidiary initiative diffusion are each 0.56. To confirm our model’s robustness, we reran the complete model, adapting the knowledge sharing construct to include only the two items with the strongest loadings. Model fit and convergent and discriminant validities were acceptable; the same pattern of coefficients was achieved.

We tested the model by examining the correlation statistics and linear regressions using robust standard errors. We calculated the variance inflation factor (VIF) in the direct effects models to check for potential multicollinearity issues. The highest VIF value in our main direct effects models was 3.63, within an acceptable range (Hair et al., 2006). We do not expect multicollinearity to hinder our interpretation of the regression results. We examined the interaction effects for H4 and H5 by graphically plotting them (reported below). Finally, as an exploratory extension, we examined the individual effects of differences between home and host country for each of the four sub-dimensions of national culture. These distances were calculated as the absolute difference (δ) between Hofstede scores for home and host country. These were examined in regression models as well as graphically. While our baseline hypothesis (H1) concerns overall cultural distance and not any specific sub-dimension, this exploratory analysis aids in interpretation and leads to avenues for future work (reported below).

Results

Table 3 shows descriptive statistics. Table 4 shows the correlations between the variables of interest. Table 5 shows the regression analysis.

The first model in Table 5 contains control variables. The second model tests the direct effect of national cultural distance (H1). The third model tests the direct effects of national cultural distance with values and knowledge sharing (H2 and H3). The fourth model includes interaction terms (H4 and H5). The best-fitting model (R² = 0.53) is Model 4, with all variables entered. In terms of control variables, we note some significance on (1) the number of employees, (2) the home country being the U.S., (3) home country foreign labour, (4) home country attitude towards globalization, and (5) subsidiary range of capabilities. Home country attitude towards globalization is the only control variable that is significant in all models. Cultural distance (H1) is consistently negative and statistically significant in Models 3 (p<0.01) and 4 (p<0.05). In terms of direct effects, values sharing (H2) is positive (p<0.05), and knowledge sharing (H3) is positive (p<0.01 and p<0.1) in Models 3 and 4. This provides support for H1, H2, and H3. Unexpectedly, the interaction effect between national cultural distance and values sharing is not significant, providing no support to H4. The interaction effect between national cultural distance and knowledge sharing is positive (p<0.05), supporting H5 (Figure 1).

Table 3

**Descriptive statistics (n = 111)**

Table 5

**Subsidiary initiative diffusion regression models**

Table 6 shows the exploratory post-hoc analysis to examine the effects of national cultural distance sub-dimensions on subsidiary initiative diffusion. Model 1 includes all control variables along with values sharing and knowledge sharing. In Models 2–5, all four distances have negative signs when entered alone after accounting for all control variables, values sharing and knowledge sharing. However, only differences in power distance (δPDI) and uncertainty avoidance (δUAV) are significant (p<0.01 and p<0.1, respectively). δPDI accounts for an additional 6% of variance explained, while for δUAV, it is only 2%. While all sub-dimension distances are entered in Model 6, three of them retain their negative sign; however, only δPDI is significant (p<0.01). This pattern is supported by the two-way linear fit plot for the four sub-dimension distances (Figure 2), showing that the strongest negative effect to be the sub-dimension for the difference in power distance.

Figure 1

Moderating effects of values and knowledge sharing on the relationship between national cultural distance (CDIST) and subsidiary initiative diffusion

(a)

**Moderating effect of values sharing (vs)**

(b)

**Moderating effect of knowledge sharing (ks)**

Note: Values sharing was not significant in Table 5

-> Voir la liste des figures

Figure 2

**Two-way linear fit plots of subsidiary initiative diffusion with sub-dimensions of national cultural distance**

Table 6

**Effects of sub-dimensions of national cultural distance on subsidiary initiative diffusion**

Discussion

Our primary findings demonstrate how external national cultural and internal organization-level factors interact to influence the diffusion of subsidiary initiatives in MNEs. To our knowledge, no previous study has taken this perspective. While there has been some scholarly attention to how contextual, national-level factors influence the phenomenon of subsidiary initiative diffusion in MNEs (Edwards et al., 2015; Schmid et al., 2014), the bulk of the emphasis in previous literature has been on the internal organizational context (e.g., Ambos et al., 2010; Bartlett and Ghoshal, 1998; Birkinshaw, 1999; Dörrenbächer and Gammelgaard, 2016), such as internal knowledge transfer (Ambos and Ambos, 2009). Our findings provide strong support for an extended perspective on entrepreneurial dynamics in MNEs that accounts for the interaction between external national and internal organizational variables.

We combine elements of management control logic (Eisenhardt, 1985; Merchant and van der Stede, 2007), knowledge-based logic (Kogut and Zander, 1992; Gupta and Govindarajan, 2000), and external national culture. First, our approach supports the argument that differences in national culture negatively impact the knowledge transfer that supports international initiatives. This finding is in line with research suggesting larger national cultural distance increases complexity and communication and coordination costs (Kostova, Marano, and Tallman, 2016). Secondly, it supports how organizational control through values sharing and knowledge sharing mechanisms within the MNE may moderate this negative relationship. This finding is in line with recent research by Alofan, Chen and Tan (2020), showing how national cultural distance effects can be compensated for by organizational culture practices. Our study extends this work by demonstrating how this is grounded in knowledge sharing practices and not values sharing.

Scholars have shown how values sharing influences employee interaction and how trust is generated and reflected in decisions. We show how this effect also applies when we consider subsidiary initiatives in an MNE. However, the lack of support for H4 suggests that using values sharing for culturally distant subsidiaries is not beneficial for harnessing initiative in those subsidiaries. Indeed, the plot shows that value sharing could dampen the likelihood of culturally-distant subsidiary managers sharing initiatives in the wider MNE (Figure 1). One reason for this is corporate values and norms may not be appropriate in the subsidiary host country, creating tension between internal corporate culture (represented through values sharing) and external national cultural distance (Adler and Jelinek, 1986). This tension could lead to cultural clashes such that subsidiary managers question the necessity to seek acceptance of their own initiatives from their headquarters, resulting in the subsidiary being less likely to seek to transfer its initiatives beyond its national boundaries. Given that MNEs acquire and transfer knowledge through social processes, the cultural make-up of dispersed individuals across the MNE will matter. On the other hand, the support for H5 suggests the use of knowledge-based theory to understand subsidiary initiatives in the MNE needs to adopt a more explicit acknowledgment of how internal integrative mechanisms mitigate hazards from national cultural distance –ambiguity, interference, and lack of equivalence (Holden and Von Kortzfleisch, 2004). Developing social relations clearly facilitates this, allowing MNE knowledge resources to be exploited across the countries and cultures in which they operate.

In the post-hoc analysis, we explored the effects of differences in the sub-dimensions of national culture (Table 6 and Figure 2). The analysis reveals differences in power distance (as opposed to individualism, masculinity and uncertainty avoidance) between headquarters and subsidiaries to be the main driver behind the negative effect on subsidiary initiative diffusion. This finding is particularly interesting because of the relationship between power and subsidiary initiative taking in MNEs. Dörrenbächer and Gammelgaard (2016), for instance, reveal power to be vital to the issue-selling process that subsidiaries go through when attempting to convince headquarters managers to accept their proposals. When subsidiaries have more power (e.g., because of their market position or capabilities), they can develop initiatives more fully and have credibility in headquarters managers’ eyes. Power enhancement is also seen as a consequence of subsidiary initiatives; subsidiaries become more powerful when their initiatives are diffused across the MNE (Mudambi and Navarra, 2004). Diffusing subsidiary initiatives challenges existing power structures (Strutzenberger and Ambos, 2014).

Power distance relates to the acceptance of inequalities of power in a society. As noted by Daniels and Greguras (2014), where there is high power distance, people “believe that authority figures should be respected and shown deference” while with low power distance people “do not perceive many distinctions based on social strata, power, or hierarchical position” (Daniels and Greguras, 2014: 1203). Recent work shows low levels of power distance to be associated with a preference for greater transparency (Jain and Jain, 2018). It is important to note that our main analysis concerns distance, not the absolute levels of power distance. When subsidiary and headquarters’ national cultures are similar in terms of how power inequalities are viewed, there will be a greater level of congruence regarding norms for authority, transparency and decision-making. In this scenario, both sides accept the power and authority residing in the headquarters (because both sides are high power distance) or accept that subsidiaries can assume greater power (because both sides are low power distance). This congruence will mean a lower chance of internal confrontation and conflict as the subsidiary exerts itself.

On the other hand, issues arise with a power distance mismatch: when low power distance human resource management practices are used in a high power distance country, there can be distrust or fear of managers as they appear weak (Newman and Nollen, 1996). Following Jain and Jain (2018), one side prefers greater transparency regarding the subsidiary initiative, while the other does not. This suggests that when subsidiary and headquarters’ managers are culturally distant in terms of their acceptance of power inequality, there is a higher chance that at least one side will distrust the other side (Daniels and Greguras, 2014: 1219), leading to a problematic issue-selling process and a lower propensity for subsidiary initiatives to be accepted. While we focus on a composite national cultural distance index, this interesting finding from our post-hoc analysis can be explored more deeply in future work with research designs that capture both internal and external constructs of power as they affect subsidiary initiatives.

The results have implications for MNE managers. At the headquarters level, managers should be aware that increased national cultural distance can impact a subsidiary’s ability to drive initiatives for wider adoption. Suppose a subsidiary is to fulfil an entrepreneurial mandate given by headquarters but is in a culturally distant country. The MNE will need to install comprehensive knowledge sharing capabilities to compensate. In this respect, headquarters managers need to be ‘self-aware.’ Managers can also consider these factors as they make location and mandate decisions on subsidiaries. Forcing corporate-wide values on a subsidiary could backfire when national cultural distance is high, leading to tension and clashes between corporate and national cultures. At the subsidiary level, organizational practices should be carefully selected to promote communication and knowledge sharing to overcome national differences. The subsidiary can lead the selection and implementation of these practices as part of an evolutionary process of becoming more influential (Mudambi and Navarra, 2004).

The present study has limitations. Firstly, we used a purposive sample, aiming to achieve heterogeneity across countries and industries. One must be careful not to generalize the implications of this study to specific industries or countries. Secondly, the relationships under consideration may be far more complex than the combination of variables assessed in this study. In particular, for national-level variables, we only considered national cultural distance and specific institutional measures from secondary data sources as control variables. Many other national external environment variables can be explored in the future to complement this analysis. It will be useful to investigate the effect of other forms of distance. Thirdly, while we focus only on subsidiaries’ interactions with headquarters, headquarters’ responses to subsidiary initiatives are also influenced by other subsidiaries. We are limited in having only data from subsidiary managers’ perceptions and not headquarters as well. Finally, we do not have data on subsidiary-to-subsidiary interaction that may have contributed to initiatives or any regional or cluster effects at play. Future research can address these issues and may also examine whether sharing specific organizational values within an MNE will affect subsidiary initiative diffusion. Future case studies and phenomenological approaches can be used to capture subsidiary and headquarter manager entrepreneurial behaviour in the context of different levels and types of cultural distance. They can also be used to assess the long-term performance of diffused subsidiary initiatives and contrast those that arose from culturally-distant subsidiaries with those arising from culturally-close subsidiaries. We hope such future research will develop insight into the interaction between national cultural distance and organizational factors when it comes to entrepreneurial dynamics and rejuvenation in MNEs.

Conclusion

We find, as hypothesized, that national cultural distance between host and home countries has a material and negative role to play in how subsidiary initiatives are diffused within MNEs. Consistent with the knowledge-based view of the MNE, increases in knowledge sharing within the MNE help to overturn this negative effect. The same, however, cannot be said for values sharing, a finding that counters the management control perspective as it applies to internalization of values across remote units of MNEs. Exploratory post-hoc analyses reveal it is differences in power distance between host and home countries that underpin the negative effect of national cultural distance, raising new areas for research on the power-based view of subsidiary initiatives within MNEs.

Abstract

Résumé

Resumen