The effect of differences in uncertainty avoidance on
HQ-subsidiary relationships and the consequences for
knowledge sharing.
Bart van de Vliert
June 13th 2016
University of Groningen
Abstract:
In this research the idea of cultural positions is tested for generalizability and to see if it is a valuable and relevant addition to the cultural distance concept, which is widely applied in research. I empirically analyze how low or high uncertainty avoidance affects knowledge sharing within the headquarters-subsidiary relationship and investigate whether cultural positions can be distinguished with these different variables. An ANCOVA analysis was used to test these assumptions in 1714 subsidiaries with headquarters in 28 countries and located in six European countries. My findings indicate the existence and relevance of cultural positions and it is found that there are indeed differences in knowledge sharing dependent on cultural distance in the relationship between headquarters and its subsidiary. In relationships that have
low cultural distance and are similar in uncertainty avoidance I found that uncertainty avoidance contributed to less knowledge sharing within the MNC. In relationships with high
cultural distance I found that it matters if the headquarters or the subsidiary is located in an uncertainty avoidant culture since more knowledge is being shared if the knowledge sharing is
2
Introduction
This paper is an exploratory study that investigates how cultural characteristics of the national environments of two related organizational units, and the differences between them influence the units’ relationship. Culture is a phenomenon that is widely applied and used, but which is still not fully understood. This can be seen in some of the assumptions that accompany culture in research. An example of this, and what I will focus on mainly in this research is the
assumption of symmetry, in which researchers assume that cultural distance between two countries is equal and in practice this is not true. Shenkar (2001) indicated that distance is in general symmetric, but that there is no reason to assume symmetry in culture studies.
Assuming symmetry in cultural distance studies will likely lead to results suffering from one dimensionality.
Studies within the international business literature often investigate culture using the cultural distance concept measured by the Kogut and Singh index (1988). (Tihanyi, Griffith & Russell, 2005; Brouthers & Brouthers, 2001). By investigating culture with the cultural
distance concept, the actual cultural characteristics are often overlooked and therefore more recent studies have indicated the need for richer conceptualizations and metaphors of cultural distance (Shenkar, Luo & Yeheskel, 2008; Drogendijk & Zander, 2010; Yildiz, 2014). A new cultural perspective that recently emerged is the idea of cultural positions introduced by Drogendijk & Holm (2012). They define cultural positions as the actual cultural
characteristics of both parties involved in a relationship and the differences between them, position is referring to the absolute cultural characteristics of each party and to their relative content or value (adapted from Drogendijk & Holm, 2012, p383). In this research, distance is no longer seen as symmetrical, but more emphasis is placed on the actual characteristics of the different cultures. In this paper I will investigate and build upon the cultural position idea that was introduced by Drogendijk & Holm (2012). However the biggest difference is that in my study I will not focus on hierarchy and power distance to exemplify cultural positions, but I will look at a different cultural variable namely: uncertainty avoidance and its effect on HQ-subsidiary relationships on the topic of knowledge sharing.
3 present in other culture variables besides the previously tested power distance (Drogendijk & Holm, 2012). I do this mainly because uncertainty avoidance is an important determinant of the level of knowledge transfer within an MNC (Leyland, 2006).
Knowledge sharing is a concept that is well researched in the context of mergers and acquisitions and international joint ventures (Yildiz, 2014; Pak & Park, 2004; Sarala & Vaara, 2010) and there have been also studies regarding the knowledge flows that occur within multinationals and between subsidiaries, however most of these studies lack empirical evidence to back up their claims (Leyland 2006; Ciabuschi, Martin & Stahl, 2010). Then lastly there are also the studies that focus on other intra-organizational explanations for knowledge transfer like social interactions (Noorderhaven & Harzing, 2009). Despite
previous research investigating the role that national culture, or more specifically uncertainty avoidance, has on knowledge sharing, this relationship is not sufficiently empirically
supported.
The primary aim of this research is to fill this gap in the current literature and to build upon the concept of cultural positions and to examine this in the headquarters-subsidiary relationship context of multinational corporations (MNC). Therefore my research question is: “Does the concept of cultural positions still hold when examining the effect of uncertainty avoidance on knowledge sharing within a headquarters-subsidiary relationship context?” The previously mentioned model of cultural positions will be used in this exploratory empirical study to show if the concept of these cultural positions also works when considering the level of knowledge sharing that occurs in a HQ-subsidiary relationship. Afterwards I elaborate on the consequences that these four positions have on the existing view of culture and knowledge sharing.
Just like in the model of Drogendijk & Holm (2012) I assume four theoretical situations. A matrix is constructed which includes the possibilities of both the HQ and the subsidiaries’ uncertainty avoidance being low or high, resulting in a two-by-two model. The headquarters-subsidiary relationship is explored in four theoretical situations in the previously mentioned model. In two cases the uncertainty avoidance will diverge (i.e., uncertainty
avoidance of headquarters will be high and uncertainty avoidance of subsidiary will be low, or the other way around) and in two cases the uncertainty avoidance will converge (i.e.,
4 The data that is used is data collected by a survey of MNC subsidiaries located in six European countries and which are headquartered in 27 countries. This data shows the degree of knowledge sharing between subsidiaries and the usefulness of this shared knowledge. The results indicate that there are indeed four different cultural positions that can be identified and that the subsidiaries’ uncertainty avoidance is most important in achieving high knowledge sharing between subsidiaries.
1. Headquarter-subsidiary relationship and culture
To get a better understanding of the headquarter-subsidiary relationship within an MNC, it is important to have a clear understanding of the relevant factors. In this paper I define a multinational company as consisting of semi-autonomous entities (Hedlund, 1994 ; Nohria and Goshal, 1997; Doz et al., 2001; Ambos & Ambos, 2009), in which units in dispersed locations take on various missions and control heterogeneous stocks of knowledge (Foss and Pedersen, 2002; Ambos & Ambos, 2009). An important thing to mention is that because these multinational companies have subsidiaries in dispersed locations, they operate in different countries and thus have to deal with a multitude of cultures.
Cultures differ between countries and therefore it is important to know what the concept of national culture entails. I define national culture as values, beliefs and assumptions learned in early childhood that distinguish one group of people from another (Hofstede, 1991; Newman & Nollen, 1996). National culture is a concept that is deeply imbedded in individuals and in everyday life and it is to some degree susceptible to change, however this is a continuous and very slow process. Now I will mention three reasons why using the concept of cultural distance can lead to problems when doing research.
5 Singh (1988) index, in which it is portrayed that the cultural distance between Denmark and Finland scores almost the same as the cultural distance between Bangladesh and the
Philippines (1.34 and 1.35 respectively). It is likely that to calculate the difference between Denmark and Finland, other cultural elements are used than to calculate the cultural
difference between Bangladesh and the Philippines. Elements of culture are used to calculate the cultural difference between Denmark and Finland and these different cultural elements cannot be found in the cultural distance score, resulting in the one dimensionality of the cultural distance measure. One combination of countries could score lower on uncertainty avoidance while the other combination scores lower on individualism, multiple combinations can lead to the same score. Just considering the “score” of country combinations will lead to a one-dimensional approach and the non-linearity of cultural differences and how people’s relationships are affected.
Secondly, a single number does not enable researchers to express where cultures are positioned on a dimension or continuum (Drogendijk & Holm, 2012). Actual characteristics of cultures are ignored and the focus is on how much cultures differ but not on what the actual differences are. If a combination of low uncertainty-avoidance countries are compared with a combination of high uncertainty-avoidance countries, the cultural distance score might be the same because the distance between the countries is the same, however the culture
combinations themselves are positioned at other sides of the dimension continuum and therefore the actual characteristics differ. In figure 1a, it can be observed that even though the cultural distance between a1 and a2 is the same, they are positioned on different ends of the continuum, meaning that these cultural differences likely consists of different elements or cultural characteristics.
6
Figure 1b Equal distance-direction of relationship (From Drogendijk & Holm; 2012, p385).
Lastly, cultural differences and perceptions of cultures are not symmetrical (Shenkar, 2001; Chapman et al., 2008). Even though the distance between a headquarters and a
subsidiary can be the same, they can be positioned on different positions on the culture
continuum. Then the cultural distance is the same, but the distance is perceived differently by both the headquarters and the subsidiary. An example is that a low-scoring subsidiary will perceive their headquarters as high-scoring, as long as it is higher on the continuum than the subsidiary. In contrast, a high-scoring subsidiary is likely to view their headquarters as low-scoring as long as they score lower than the subsidiary. However, this does not mean that they actually score high or low, it is the perception that is created from the relative positions between headquarters and subsidiaries. Figure 1b shows two situations on a cultural
continuum, b1 in which the headquarters scores higher than the subsidiary and b2 in which the headquarters scores lower. Even though this has different implications for the
characteristics of the cultural difference, this is ignored because the cultural distance score would be the same. Therefore as Drogendijk & Holm (2012) already indicated, it is important for a more complete understanding of cultural distance to pay attention to both sides of the headquarters-subsidiary relationship and their cultural positions because that would reduce the one-dimensionality. This is also important in my research on knowledge sharing within the MNC, because if distance between HQ and subsidiary is perceived in a one-dimensional way, it will limit the understanding of the actual characteristics of the knowledge sharing
7
2. Knowledge sharing within the headquarters-subsidiary relationship
Knowledge in itself is a very broad definition. In this paper I adopt the following conceptualization of knowledge of Davenport & Prusak (1998: 5) also used in the paper of Noorderhaven & Harzing (2009: 721): “Knowledge is a fluid mix of framed experience, values, contextual information, and expert insight that provides a framework for evaluating and incorporating new experiences and information. It originates and is applied in the minds of knowers. In organizations, it often becomes embedded not only in documents or
repositories but also in organizational routines, processes, practices, and norms”. In our case we look at knowledge that is transferred within a multinational between headquarters and a subsidiary. The knowledge that is transferred will thus not only include documents but also its organizational routines, processes, practices and norms.
The definition already makes a clear distinction between different types of knowledge that can be transferred. In general there are two kinds of knowledge that are transferred within a multinational company: Explicit and tacit knowledge. Explicit or codifiable knowledge is more easily transferred within a multinational than tacit or non-codifiable knowledge (Kogut & Zander, 1993). In this study I look at how knowledge is spread and shared throughout a company and specifically between subsidiaries of a multinational company. I expect that the knowledge that is transferred is a combination of explicit and tacit knowledge.
Knowledge is very important for MNCs because it is often a source of a sustained competitive advantage and a superior performance (Spender and Grant, 1996). The ability to transfer knowledge within a company has been proven to be the difference between superior firm performance and a deciding factor leading to a competitive advantage (Collins and Smith, 2006; Mcevily and Chakravarthy, 2002). Because having a competitive advantage is important for firms, it is vital that they focus on creating inimitable knowledge and sharing it within the corporation (Sveiby, 2001).
8 within lateral learning relationships that are formed by the subsidiary and these learning relationships are generally not organized or imposed by the HQ. There is knowledge that is sent from the HQ to the subsidiary and knowledge generated by the subsidiary and sent to the HQ. It is also noted that sometimes the roles are reversed, in case of a subsidiary first
acquiring knowledge from the MNC and providing the MNC with generated knowledge at a later stage.
An often used model to study the intra-firm knowledge flows is the sender-receiver model. However this model is often used implicitly (Adler & Hashai, 2009). According to Szulanski (2000) the basic elements of a transfer consists of: a source, a channel, a message, a recipient and a context. Within the stream of research on the elements of a transfer,
researchers often focus on: characteristics of the sender unit, characteristics of the receiving unit, characteristics of the relationship between sender unit and receiving unit and lastly the characteristics of the knowledge that is shared. The sender-receiver model fits this research well because I also focus on the characteristics of the sending and receiving units, more specifically on their levels of uncertainty avoidance and the relationship between sender unit and receiving unit is the headquarters-subsidiary relationship.
9 control and coordination of subsidiaries (Gomez-Meija & Wiseman, 1997; Roth and
O’Donnell, 1996). Because the goals and orientations between subsidiaries differ, financial incentives for subsidiary top managers can help to ensure subsidiary compliance with headquarters goals (Gomez-Meija & Wiseman, 1997). These frameworks are essential to understand the complex mechanisms and reasoning behind knowledge sharing between a HQ and its subsidiaries.
3. National culture and the influence on knowledge sharing
A multinational company generally has to deal with the influences of multiple types of culture: organizational and societal culture. (Pothukuchi, Damanpour, Choi, Chen & Park, 2002). Organizational culture can be defined as the shared values and assumptions that guide behavior in an organization (Schein, 1990) and a common assumption is that organizational culture can provide a competitive advantage for a firm when used correctly (Byles, Aupperle & Arogyaswamy, 1991). Besides the fact that organizational culture has a significant impact on an MNC’s operations, it equally important to realize that an MNC generally operates in a large amount of countries and that the societal cultures of the countries that the MNC is operating in will influence the organizational culture of the HQ or the subsidiaries operating within that culture (Sigler & Pearson, 2000 ; Van Oudenhove, 2001) .
Societal culture and organizational culture are closely intertwined with each other. Harzing & Sorge (2003) suggest that the values and practices of subsidiaries and their HQ generally correspond with their national cultural context. This leads to a situation in which subsidiaries of an MNC operating within different national contexts will likely vary greatly in their values and practices resulting in large cultural differences being present within the MNC.
Societal culture as well as organizational culture are hard to measure because they are such broad constructs. A framework often used to operationalize the construct of societal culture is the framework of Hofstede (1980). This framework is generally accepted and applied in organization and management research (Randolph, 2000). According to my
11
3.1 Uncertainty avoidance
Uncertainty avoidance is the way that societies deal with the fact that time runs only one way and how to deal with the uncertainty that the future brings (Hofstede, 1983).
Uncertainty avoidance can be either high or low and this indicates the extent to which they are uncertainty avoidant. Societies with low uncertainty avoidance tend to accept each day as it comes and take risks rather easily. Societies with high uncertainty avoidance try to beat the future and this generally results in higher levels of anxiety. These societies will also have institutions trying to create security and avoid risks (Hofstede, 1983). For managers this entails how threatened they feel when they encounter ambiguous or uncertain situations and how they perceive and view opportunities and threats (Alexander, 2012; De Luque and Javidan, 2004). Also, countries with high uncertainty avoidance are more likely to use
formalized contracts in transactions and have a strong desire to develop rules and structure in organizational processes to avoid uncertain situations while doing business (Bhardwaj, Diets & Beamish, 2007).
The national culture that the HQ and its subsidiaries are encapsulated in will affect the extent to which they avoid uncertainty and are willing to share its knowledge with other parts of the company. Uncertainty avoidance will affect the reluctance to take risks and how much risk is being taken when knowledge is being shared, because HQs or subsidiaries in such a context will be more afraid of the potential losses that could occur. In general HQs or
subsidiaries based in cultures with high uncertainty avoidance can be expected to risk averse and avoid everything that could potentially harm their own operations (Shane et al., 1995). In contrast, when uncertainty avoidance is low then companies will be more inclined to take risks because the benefits of them seem to be more appealing than the risks associated with them. Companies operating in a low uncertainty avoidance cultural context can be expected to be risk-taking and to be willing to share more, if they perceive that sharing can lead to
benefits (Shane et al., 1995).
12 shift control over to their subsidiaries (Bartlett & Goshal, 1989). Uncertainty avoidance plays a large role in this research because of the influence it has on the relationship between HQ and subsidiaries. Within this HQ-subsidiary relationship there can be different views on
uncertainty avoidance and whether or not risks should be taken. The possibility of the headquarters and the subsidiary having conflicting views within an MNC because of their respective cultures, was developed into a model by Drogendijk & Holm (2012).
Subsidiaries are expected to adapt to the wishes of the HQ and to listen to directions they are given, including instructions regarding knowledge sharing. Therefore it is to be expected that the HQ will have more influence on the subsidiary than the other way around. Both the HQ and its subsidiaries can be uncertainty avoidant and this will lead to four different conditions of knowledge sharing, which can be found in figure 2.
Figure 2: The uncertainty avoidance and knowledge sharing matrix.
13 support from the HQ, it is likely that the developed knowledge will be harder to share within the whole organization. In the headquarters driven knowledge sharing situation the subsidiary is uncertainty avoidant whereas the HQ is not, in this case it will probably lead to the HQ taking a more formal approach and developing ideas and knowledge to share with the company through hierarchical means. The HQ is not afraid to take risks, but the subsidiaries do not want to take the risk to develop new knowledge. Lastly, in the high sharing situation both HQ and the subsidiary are not uncertainty avoidant. They are both willing to take risks and it is expected that because of this willingness, this situation will result in both the HQ and the subsidiary contributing to the knowledge pool of the firm thus resulting in the optimal situation for knowledge sharing.
4. Combining knowledge sharing and culture
Within an MNC it is vital that there is a distribution of the knowledge production, but this requires transferring the produced knowledge among the organizational units (Bartlett and Ghoshal, 1989). This process facilitates global learning within an MNC but can
simultaneously have a negative effect because subsidiaries that have different interests and knowledge outflows could develop fear of losing their bargaining position compared to other subsidiaries when they share their knowledge (Mudambi & Navarra, 2004). While
subsidiaries could definitely benefit from sharing their knowledge and creating a global learning process within an MNC, it is thus not likely for a subsidiary to share their specific knowledge if they are deterred by losing their bargaining position. If a subsidiary perceives the liabilities of sharing their knowledge to be greater than the benefits, they are not likely to share their knowledge. The fact whether or not they perceive something to be worth the risk, depends on the level of uncertainty avoidance of a specific subsidiary, which is a reflection of the level of uncertainty avoidance in their national context. As explained before there are two different kinds of knowledge stream between HQ and its subsidiaries, the knowledge stream to the subsidiary and the knowledge stream from the subsidiary. In this research we will focus on the knowledge stream that goes from the subsidiary to the HQ. This knowledge is
14 on the levels of uncertainty avoidance of the subsidiaries’ country as well as the headquarters’ country. The levels of uncertainty avoidance lead to the classification of low or high
uncertainty avoidance for both headquarters and the subsidiary. These classifications are depicted in figure 2 and lead to four possible situations regarding the way that knowledge is being shared within the organization.
There are two situations in figure 2 in which the cultural distance between HQ and subsidiary is low. The first situation is the high knowledge sharing situation in which the HQ and its subsidiary are both not uncertainty avoidant. In this situation both the HQ and the subsidiary are willing to take risks, generating more knowledge and being more likely to share this knowledge with other parts of the company. Knowledge is developed by the HQ as well as the subsidiary and enough support is provided to share it. I believe this will lead to much knowledge being shared among subsidiaries within the organization. The second situation is the low knowledge sharing situation in which the HQ and its subsidiary are both uncertainty avoidant. In this situation both the HQ and the subsidiary will be reluctant to take risks, which in turn generates less knowledge and leads to a decrease in the probability of knowledge being shared throughout the organization. Overall I believe this will lead to less knowledge being shared among subsidiaries and within the organization. Therefore I hypothesize:
Hypothesis 1. The competences of a subsidiary will be shared less in the low
knowledge sharing situation compared to the high knowledge sharing situation (figure 2).
15 HQ should be stimulated because of low HQ uncertainty avoidance.
When looking at the subsidiary driven knowledge sharing situation the HQ is uncertainty avoidant but the subsidiary is not. This would indicate that the subsidiary takes risks and wants to develop new knowledge, but that sharing its competences and knowledge is not supported by the HQ. This will probably lead to subsidiaries creating local knowledge from lateral ties, which is hard to share and distribute amongst other units of the MNC because the HQ does not provide the subsidiary with sufficient assistance in diffusing and sharing this knowledge. Not having sufficient assistance in sharing this knowledge is detrimental for knowledge sharing and will therefore likely result in a situation in which knowledge is created but hardly shared, leading to localness of knowledge (Jaffe, Trajtenberg & Henderson, 1993). Knowledge creation is important but for knowledge sharing it is
essential that the knowledge is diffused throughout the organization. There is a difference between the headquarters-driven and the subsidiary-driven knowledge sharing situation and therefore I hypothesize:
Hypothesis 2. The competences of a subsidiary will be shared less in the
subsidiary-driven knowledge sharing situation compared to the headquarters subsidiary-driven knowledge sharing situation (figure 2).
The cultural distance measure has often been used to measure the distance between two cultures and it is proposed that the larger the cultural distance is between two countries, the less knowledge can be transferred effectively knowledge (Mudambi & Navarra, 2004). In the two proposed hypotheses I distinguish different situations and not only use the concept of cultural distance. Taking into consideration the actual characteristics of the cultural context and creating different cultural positions is an idea introduced by Drogendijk & Holm (2012). This method of considering both home and host country cultural context characteristics is not yet widely applied and therefore it is important to validate whether the findings of cultural positions can be replicated and used with different variables. I also assume that we can find the same cultural positions as they found in their research. Therefore I will test if the idea of cultural positions also becomes apparent in my research when using other variables. Therefore the last hypothesis is:
Hypothesis 3: Four distinct cultural positions can be identified with uncertainty
16
5. Methods
In this section I will discuss the process of acquiring the data and will define the variables and explain how the variables are operationalized in this research. I will present descriptive statistics on the data and will use an ANCOVA test just like in Drogendijk & Holm (2012) to examine whether cultural positions also play a role in the relationship of home & host country uncertainty avoidance and knowledge sharing by comparing means. The ANCOVA analysis will test whether the home & host country uncertainty avoidance differs in the four uncertainty avoidance situations. Based on the findings and if the four situations are determined, the effect that uncertainty avoidance has on the sharing of
competences can be determined and the other hypotheses can be accepted or rejected. There will also be control variables used because I expect that these control variables affect the dependent variable. The four situational groups belonging to each uncertainty avoidance situation that are tested with the ANCOVA analysis differ in size, but this is allowed for in an ANCOVA analysis and will not hamper the results.
5.1 Data Collection
In this research we make use of data collected for a large international project which was initiated in 1997 (Foss & Pedersen, 2002). Which was used to investigate the impact of subsidiary role development in MNC’s. This data was also used in the study of Drogendijk & Holm (2012) and is now used again to verify cultural positions using different variables. The dataset that was used for this research consisted of data from six countries (Austria, Denmark, Finland, Germany, Sweden and the United Kingdom). Ensuring data comparability had a great priority in designing the data collection instruments and it was constructed in such a way that the questionnaire could be used and applied the same way in all relevant countries. The aim of the questionnaire was to identify a subsidiaries’ competences, how these competences emerged within the subsidiary and how these subsidiaries having strong competences affected the competitive advantage of the multinational corporation.
17 other 20% by administrative directors, chief controllers or sales directors as explained by Drogendijk & Holm (2012).
The used sample consists of 1714 subsidiaries that were operating in the service and manufacturing industries and within the 6 aforementioned countries. The headquarters of these subsidiaries were located in 28 countries. 23,7% of the headquarters were located in North America, 6% were located in the rest of the world, and the majority of headquarters were located in Europe 70,5 %. The percentages and locations of the headquarters can be found in appendix 1. The origins of the subsidiaries is distributed in the following way: Greenfield establishment 40% and obtained through acquisitions or mergers 60%. The size of subsidiaries differed significantly and was between one and 25,460 employees with an
average of 537 employees. The sales of the subsidiaries were between 1 million USD to 11,529 million with an average of 162 million USD and the subsidiaries have been a part of the MNC for a timeframe ranging from 1 to 117 years.
5.2 Measures
In this part I will describe all variables that are relevant for this research and that will be used in the analysis. The variables consist of dependent, independent and control variables.
5.2.1 Dependent variable: Average impact of subsidiaries competences on other units in the MNC.
This measure examines the impact that a subsidiaries’ competences have on other units in the MNC. In the questionnaire we used a measure from Frost et al (2002) and
Drogendijk & Holm (2012) in which a subsidiary first had to indicate on a 7 point Likert-type scale which activity they excelled at. There were 7 categories of activities (Research,
18 developed in these seven categories all have an impact on other units within the organization. After a subsidiary indicated the extend of the impact that their competences had on other parts of the organization, these values were all combined to obtain an average to get a complete overview of the knowledge sharing variable and not of the seven individual categorical activities. In this research I assume that knowledge that is of use for other units of the MNC indicate knowledge sharing because related knowledge that can be applied easily within other units of the MNC is shown to increase knowledge sharing considerably (Hansen, 2002). Therefore I will continue with referring to the average impact of subsidiaries competences on other units in the MNC as knowledge sharing
5.2.2 Independent variable: Uncertainty avoidance
In this study I use the measure of uncertainty avoidance that was introduced by Hofstede (1980). In recent literature there has been a lot of debate on whether to use Hofstede’s constructs or the constructs from GLOBE and both measures are often used
(Venaik & Brewer, 2010). However, for the measure of uncertainty avoidance the choice does not matter too much since the values of uncertainty avoidance within globe and the
uncertainty measures from Hofstede are correlated (Leung et al., 2005), but according to the study of Venaik & Brewer (2010) if a study deals with the specific application of uncertainty avoidance then Hofstede’s measure is the better one. Therefore in this study I will apply the uncertainty measure from Hofstede (1980).
Uncertainty avoidance in this study is used to categorize the firms into the
19 uncertainty avoidant and subsidiary is high uncertainty avoidant and 4. HQ is high uncertainty avoidant and subsidiary country is low uncertainty avoidant.
5.2.3 Control variables
In this study there are numerous control variables used to control for effects that do not originate from our independent variable. The first control variable that was used is the size of the firm and this was measured by the number of employees in a subsidiary. It was
included because large subsidiaries should have more resources and therefore should find it easier to develop worthwhile competences that they can share throughout the MNC. Besides the number of employees the total sales of a subsidiary is also used as a control variable. The sales can also indicate how large a subsidiary is or how much importance it has and should therefore also find it easier to develop competences. Additional to the size and sales of a subsidiary, respondents were also asked to indicate whether the subsidiary was established as a Greenfield or was acquired through a merger or an acquisition and this was coded with a dummy variable with a “1” indicating an acquisition (Noorderhaven & Harzing, 2009). The scope of operations of a subsidiary is also an important factor influencing knowledge sharing. When a subsidiary has a wider range of in-house functions the likelihood of the subsidiary developing firm-specific and functional knowledge is higher (Björkman, Barner-Rasmussen & Li; 2004).
20
6. Analysis of the effect of Uncertainty avoidance
As mentioned at the measure section, the observations are divided into four groups based on their HQ and subsidiary country uncertainty avoidance. This results in the
previously mentioned quadrant consisting of the following four situations: high knowledge sharing, low knowledge sharing, headquarters driven knowledge sharing and subsidiary driven knowledge sharing. There are 400 observations where both HQ and subsidiary country scored high on knowledge sharing with an average score on impact of their competences of 4,00 (upper left, figure 2), 473 observations were found where both HQ and subsidiary country scored low with a knowledge sharing score of 3.74 (bottom right, figure 2), 476 where headquarters scored low and subsidiary scored high with a knowledge sharing score of 3.67 (top right, figure 2) and 365 where subsidiary scored low and headquarters scored high with a knowledge sharing score of 4.17 (bottom left, figure 2) which can be found in appendix 1.
6.1 Assumptions underlying the model
I start the analysis by explaining steps that were taken to ensure that the assumptions underlying the ANCOVA analysis were met. I will use the ANCOVA analysis to identify whether the proposed cultural positions in the case of uncertainty avoidance are present and whether there is a significant relationship between uncertainty avoidance and knowledge sharing. The assumptions for an ANCOVA analysis were not met and problems were found with assumption of normality. The dependent variable knowledge sharing is measured by Likert scale of 1 to 7 that measures the impact of a subsidiaries’ competences on other units of the firm on seven different activities. To get a complete overview of the knowledge sharing variable these scores on the seven activities are averaged, since subsidiaries only indicated the impact on other units of the firm if they believed their firm had a strong competence in the activities that they undertake (a score higher than 4) there are also missing values, these values are not included in the average. The mean of this averaged variable is 3,87 and looking at the normality plots, the variable seems to be distributed normally (appendix 3) however for a variable to be considered to be normally distributed, the normality test should not indicate any significance meaning (P>0.05). In this case we have to use a Kolmogorov-Smirnov test of normality because the amount of cases is larger than fifty. For this test of normality a
21 distribution (appendix 4).
No normal distribution means that there is insufficient ground for an ANCOVA test, therefore the data has to be examined to determine why the variable is not normally
distributed and whether an ANCOVA test can be justified. The data is all based on answers from subsidiaries that are answered based on a Likert scale. In general Likert scales are not normally distributed because they are not a continuous variable, Likert scales can be seen as ordered categories. The answers that people can give are limited by the options given and there is not an infinite amount of answers possible, making it not normally distributed. The question whether or not an ANCOVA can be used when the data is not normally distributed is a question that arises continuously in academic discussions.
Boneau (1960) found that if the sample sizes of the individual groups (in this case the four situations) are bigger than 10 cases when performing an ANCOVA test, the variable is automatically normally distributed even if the analyses indicate otherwise and in this case an ANCOVA test can be applied. When examining the data in depth it shows that there are a lot of observations which have a value with decimals and this is because I used the average of the 7 areas (Research, Development, Production of Goods or Services, Marketing/sales,
Logistics/distribution, Purchasing and Human Resource Management). A normal Likert scale also does not include decimals, however because the previous Likert scales were averaged to obtain the total impact of a subsidiary’s competences, decimals appeared in the results. Therefore, to match the assumptions underlying the ANCOVA analysis, the results should be distributed according to Likert-scale standards. Therefore the total impact of a subsidiary’s competences should be rounded to achieve the same results of an original Likert scale again and to validate the continuation of this analysis with an ANCOVA analysis.
22
Figure 3: comparison of normality plots of rounded and non-rounded average impact subsidiary’s competences
A correlation matrix is presented in appendix 7 to depict the correlations between all variables included in this model. The independent variable of the rounded average impact of a subsidiaries’ competences on other units in the multinational company correlates with the fixed variable of the uncertainty avoidance positions. This is already a good first indication that there is a relationship between the four cultural positions and knowledge sharing. Another correlation to take note of is the relationship between subsidiary age and the four cultural positions, this means that age influences the uncertainty avoidance positions to some extent luckily there are no other variables influencing the proposed cultural positions and I will continue with the analysis of the covariates.
6.2 Analysis of Covariates
I will first start by investigating the effect that the covariates or the control variables have on uncertainty avoidance on knowledge sharing. Table 1 shows the results of the test, the F score can be found as well as the significance the variable has on the model. The model itself is significant with an F-value of 10,061 (P<0.001). It can be observed that only the amount of employees at a subsidiary influence knowledge sharing within an MNC at the 95% confidence interval. However, at the 90% significance level the sales of a subsidiary also influences the dependent variable.
23 that they simply have more knowledge to share than smaller subsidiaries. An interesting finding is that the amount of activities, age and whether a subsidiary is acquired or a Greenfield all do not significantly influence the proposed model (table 1).
The correlation table (appendix 7) shows that there is an underlying factor for the control variables acquisition and age. This means that whether or not a subsidiary is
Greenfield or an acquisition has a relationship with age. It might be that subsidiaries that have been a part of an MNC for a longer period of time have more time to optimize knowledge for spreading throughout the MNC, whereas newly acquired subsidiaries’ knowledge might not be perfectly suitable for spreading throughout the company. Another possible reason why subsidiaries are acquired is for specific knowledge that they have accumulated, which adds greatly to the already existing knowledge pool of an MNC. Therefore age can have a two-way effect meaning that age can be beneficial for the development and spreading of knowledge, but on the other hand new firms can also contribute a lot of new knowledge and competences to the MNC, which will also result in a higher knowledge sharing score.
24 Tests of Between-Subjects Effects
Dependent Variable: rounded average impact subs competences on other units in the MNC
Source
Type III Sum of Squares df Mean Square F Sig. Corrected Model 190,802a 8 23,850 10,061 ,000 Intercept 7,231 1 7,231 3,050 ,081 sub_total ,411 1 ,411 ,174 ,677 sub_emp 32,107 1 32,107 13,544 ,000 sub_sale 7,703 1 7,703 3,249 ,072 sub_acq ,500 1 ,500 ,211 ,646 sub_age 1,671 1 1,671 ,705 ,401 Mean_pos 55,985 3 18,662 7,872 ,000 Error 3769,303 1590 2,371 Total 28711,000 1599 Corrected Total 3960,105 1598
a. R Squared = ,048 (Adjusted R Squared = ,043)
Table 1: The variables and their influence on the model
6.3 Testing the hypotheses
The means of the four distinct situations of the model (figure 2) together with the number of observations of each situation can be found in table 2a and the pairwise
comparison comparing the means of these situations can be found in table 2b. The results are quite unexpected since by looking at the means, the situation with the highest mean of
25 Estimates
Dependent Variable: rounded average impact subs competences on other units in the MNC
Mean position categorized in 4 options, high, low, high-low and low-high Observations Mean Std. Error 95% Confidence Interval Lower Bound Upper Bound low knowledge sharing 473 3,804 a 0,073 3,66 3,948 high knowledge sharing 400 4,033 a 0,08 3,876 4,19 headquarters driven 476 3,747 a 0,074 3,601 3,892 subsidiary driven 365 4,232a 0,083 4,068 4,395
a. Covariates appearing in the model are evaluated at the following values: Total amount of activities performed by the subsidiary based on 13.1 till 13.7 = 5,1032, size subsidiary, number employees = 516,10, sales subsidiary = 159,41, subsidiary is acquisition, dummy = 1,61, age of subsidiary, year of establishment = 1979,86.
26 Pairwise Comparisons
Dependent Variable: rounded average impact subs competences on other units in the MNC (I) Mean position categorized in 4 options, high, low, high-low and low-high (J) Mean position categorized in 4 options, high, low, high-low and low-high Mean Difference (I-J) Std. Error Sig.b
95% Confidence Interval for Differenceb
Lower Bound Upper Bound low knowledge sharing high knowledge sharing -,229* ,109 ,035 -,442 -,016 headquarters driven ,057 ,104 ,583 -,147 ,262 subsidiary driven -,428* ,111 ,000 -,646 -,210 high knowledge sharing low knowledge sharing ,229* ,109 ,035 ,016 ,442 headquarters driven ,286* ,110 ,009 ,072 ,501 subsidiary driven -,199 ,115 ,085 -,424 ,027 headquarters driven low knowledge sharing -,057 ,104 ,583 -,262 ,147 high knowledge sharing -,286* ,110 ,009 -,501 -,072 subsidiary driven -,485* ,112 ,000 -,704 -,266 subsidiary driven low knowledge sharing ,428* ,111 ,000 ,210 ,646 high knowledge sharing ,199 ,115 ,085 -,027 ,424 headquarters driven ,485* ,112 ,000 ,266 ,704
Based on estimated marginal means
*. The mean difference is significant at the ,05 level.
b. Adjustment for multiple comparisons: Least Significant Difference (equivalent to no adjustments).
27 Hypothesis 1, stating that the competences of a subsidiary will be shared less
(perceived by subsidiary managers) in the low knowledge sharing situation compared to the high knowledge sharing situation, is supported because the mean of the high knowledge sharing situation is higher than the mean of the low knowledge sharing situation (4.033 > 3.804). Knowledge sharing is found to be higher in the high knowledge situation where both HQ and the subsidiary have low uncertainty avoidance compared to the low knowledge sharing situation where HQ and the subsidiary have high uncertainty avoidance. It thus seems likely that lower uncertainty avoidance leads to less knowledge being shared within a
company.
Hypothesis 2, stating that the competences of a subsidiary will be shared less (perceived by subsidiary managers) in the subsidiary-driven knowledge sharing situation compared to the headquarters driven knowledge sharing situation (figure 2). It is not
supported because the mean of the headquarters-driven knowledge sharing situation is lower than the mean of the subsidiary-driven knowledge sharing situation (3.747 < 4.232).
It seems that in general when a subsidiaries’ uncertainty avoidance is high, the mean of knowledge sharing will be lowest. However, within that situation where a subsidiary countries’ uncertainty avoidance is high it is better for headquarters country uncertainty avoidance to be high since that corresponds with the highest knowledge sharing mean. It seems that uncertainty avoidance of the subsidiary country is the more important factor because the mean of subsidiary driven knowledge sharing and high knowledge sharing have the highest means. However, when the subsidiary country uncertainty avoidance is low, headquarters country uncertainty avoidance should be high to achieve optimal result.
The findings of the ANCOVA analysis indicate that the subsidiary uncertainty
avoidance is more important in determining knowledge sharing than headquarters uncertainty avoidance. Where I expected to find that low uncertainty avoidance would stimulate
knowledge sharing, it seems that that is only the case for the subsidiary country variable. Therefore it seems to be beneficial for a company to have subsidiaries operating in low uncertainty avoidance countries or stimulate a low uncertainty avoidance company culture. Besides encouraging knowledge generation and sharing for subsidiaries, the headquarters should try to maintain a level of uncertainty avoidance to filter ideas to achieve the optimal result of a subsidiary driven knowledge sharing situation.
28 that the factor that enters the ANCOVA analysis (table 1). It can be observed that the
mean_pos variable in table 1 is highly significant which indicates that the four situations that were created in the model significantly relate to knowledge sharing, which is a first indication that hypothesis 3 is supported. However, subsidiary driven knowledge sharing has a higher mean (not significant) than high knowledge sharing and low knowledge sharing has a higher mean than headquarters driven knowledge sharing (not significant). These relations are not significant and therefore we look at it a bit more in depth. When the same ANCOVA is performed with a confidence of 90%, the mean difference between subsidiary-driven knowledge sharing and high knowledge sharing becomes significant, showing that when uncertainty avoidance in subsidiaries is low, there is more knowledge sharing if uncertainty avoidance of the headquarters is high. Based on the mean differences between the groups sorted by uncertainty avoidance and because of the significance of the categorical uncertainty avoidance factor (mean_pos), it can be concluded that hypothesis 3 is supported. In figure 4 the uncertainty avoidance model from figure 2 has been adapted to include the results of the ANCOVA analysis showing the confirmed significant differences between the situations.
.
29 Validity tests have been conducted to check the validity of these findings. We tested if using cultural distance in a regression analysis with the same depdent and control variables would have a more significant impact or result, but this was not the case. Cultural distance was also added as a fixed variable in the ANCOVA analysis and here again no significant impact was detected on the results. Therefore it can be concluded that cultural distance is not a better measurement than cultural positions when measuring knowledge sharing within an MNC.
7. Discussion and conclusion
Cultural differences and cultural distance play a big role in the environment of international business and it is vital for multinational companies to adapt and accept these cultural differences. A multinational’s subsidiaries often operate in culturally distant and different cultures, which also influence the operations of a multinational as a whole (Baliga & Jaeger, 1985). Besides the study of Drogendijk & Holm (2012) few studies have taken into consideration the different cultural positions that arise from that mutuality of the
HQ-subsidiary relationship context. With this study I have tried to begin confirming the existence of cultural positions with a different context and a different independent variable, because even though it is known that both HQ and subsidiary environments have different cultural contexts, they are not sufficiently taken into consideration in research (Geppert, Williams, et al, 2003; Kostova & Roth, 2002).
In this research I have tried to confirm and analyse the existence of cultural positions with the contextual variable uncertainty avoidance and the dependent variable knowledge sharing within a MNC. Based on the uncertainty levels of home and host country and the knowledge that is shared among units within the MNC, I have found four different situations: two situations in which cultural distance between HQ and subsidiary was low, one situation where both HQ and subsidiary are uncertainty avoidant in knowledge sharing (low knowledge sharing situation) and one situation where both HQ and subsidiary are not uncertainty
30 Finding the four different knowledge sharing situations indicate that
knowledge sharing within an MNC is indeed influenced by the contextual variables of countries. The findings add a nuance to the cultural distance metaphor and underline the importance of taking notion of frictions (Shenkar et al., 2008) that occur in an interaction process between HQ and subsidiary within these different cultural contexts. Furthermore the findings also indicate that the positions on the cultural dimension variable is essential and not just the distance between HQ and subsidiary. Cultural distance itself would not have been sufficient to capture the complex relationship between headquarters and subsidiaries. Especially considering the situation where cultural distance between HQ and subsidiary is large as is the case in the subsidiary driven and the headquarters driven sharing situation, the findings show the importance of considering the mutuality of culture, because had I only investigated the cultural distance I would not have found that the uncertainty avoidance of a subsidiary has more influence on the knowledge sharing process than the headquarter’s uncertainty avoidance, thus showing the asymetry of uncertainty avoidance which also was a critique on Hofstede’s cultural distance measure in the international business literature (Tung & Verbeke, 2010). The implications for these findings are that the cultural context of the HQ interacts with the cultural context of a subsidiary and that these interactions influence the way and the extent to which knowledge is being shared within the MNC. The positions of the cultural context are essential in the knowledge sharing processes and it is of great interest for MNCs to study these relationships to assess their current situation and to determine the best practices within their company.
Hypothesis 1 stating that more knowledge would be shared in the high knowledge sharing situation than in the low knowledge sharing situation was accepted. Our findings indicate found that in the situations that had low uncertainty avoidance, the most knowledge was shared. An explanation for this could be that managers are more likely to perceive opportunities and to invest in them, rather than see threats (Alexander, 2012; De Luque and Javidan, 2004). A more contradictory finding was that hypothesis 2 stating that the
31 their subsidiary (Bacharach & Lawler, 1981) and that knowledge sharing takes place between the multinational company and the subsidiary (Noorderhaven & Harzing, 2009) the subsidiary plays a more important role in the knowledge sharing process.
Another interesting note is the fact that even though the mean differences were not significant at the 95% level, at the 90% significance level there is a mean difference between subsidiary driven knowledge sharing and high knowledge sharing, meaning that subsidiary driven seems to be the ideal situation and that it would be better for an HQ to be located in an uncertainty avoidant country. This can be logical because an HQ has to filter all the ideas to ensure that sharing goes easily. Hansen (2002) has found that in knowledge sharing within an MNC, subsidiaries have difficulty absorbing non-codified knowledge. A high uncertainty avoidance of a firm might aliveate this problem by filtering the generated ideas of
subsidiaries, codifying them and diffusing them throughout the organization. Understanding this relationship, MNCs can utilize this to better facilitate knowledge sharing within their organization.
This paper tried to test whether or not knowledge sharing was was influenced by uncertainty avoidance while at the same time trying to confirm the existence of the proposed cultural positions by Drogendijk & Holm (2012). I tried to confirm the fact that the actual cultural characteristics of a country mattered and that cultural distance cannot be simply used to accurately depict the differences that exist between two countries. In practice there will be much more differences in culture and values that will be encountered by companies and this is not accurately covered by the cultural distance variables.
This study specifically looked at the implications that the actual cultural differences had on the knowledge that was being shared in a HQ-subsidiary relationship. It was
specifically focussed on one cultural variable because it is easier to define and interpret the differences that show in the four different cultural positions that were found and to isolate the effect. However, it has to be taken into account that culture is a very broad concept, which is almost impossible to cover completely, which has to be taken into consideration when considering appropriate behavior within inter-cultural interactions. These inter-cultural interactions consist not only of uncertainty avoidance, but include many other cultural variables that have an effect on knowledge sharing.
32 share their own accumulated knowledge throughout the MNC. Another example that
influences knowledge sharing is power distance because each subsidiary has a specific task within an MNC that fits into the MNC’s strategy (Taggart, 1998). When power distance is high among subsidiaries there will be difference between them resulting in a lower level of interdependence. However, the amount of interdepence should be high for a subsidiary to be compelled to explore and share locally acumulated knowledge (Gupta & Govindarajan, 1986). Another limitation to this research is the measurement of the knowledge sharing variable which was done by measuring the extent to which a subsidiaries’ competences were of use for other units in the MNC. It might be the case that even though this indicates
knowledge sharing, but that we actually measure if it can be shared and not if it is actually shared. There might be a big difference between competences that are of use and competences that are actually shared among other units in the MNC. Furthermore we provided possible options of activities that these competences could be build on, but because we provided options we might have excluded others. If a subsidiary was really good at customer service, this could not be included as an answer. Not being able to indicate all of a subsidiaries competences might also explain the missing values among the availability of competences among subsidiares.
Even though this study confirms the existence of cultural positions, the effect of uncertainty avoidance on knowledge sharing is still mainly unclear. The R-square is relatevely small (0.05) and thus it seems that more research is needed on this topic, perhaps this is due to the limitation of the measurement of knowledge sharing . It could be that because of the difference in phrasing, merely the use for other MNCs was captured instead of the actual knowledge that is shared. For upcoming research these questions could be better specified to avoid unclearity for the respondents. More options should be given to respondents to not limit the areas they possibly have competences in. Researchers could also think of a way to
33
References
Adler, N., & Hashai, N. (2007). Knowledge flows and the modelling of the multinational enterprise. Journal of International Business Studies, 38(4), 639-657.
Alexander, E. A. (2012). The Effects of Legal, Normative, and Cultural-Cognitive Institutions on Innovation in Technology Alliances. Management International Review, 52(6), 791-815. Ambos, T. C., & Ambos, B. (2009). The impact of distance on knowledge transfer
effectiveness in multinational corporations. Journal of International Management, 15(1), 1-14. Anderson, J. (1983). The architecture of cognition. Cambridge, MA: Harvard University Press.
Bacharach, S. B., & Lawler, E. J. (1981). Power and politics in organizations. San Francisco: Jossey-Bass.
Baliga, B. R., & Jaeger, A. M. (1985). Multinational corporations: Control systems and delegation issues. Journal of International Business Studies, 16(3), 25–40
Bartlett, C. A., & Ghoshal, S. (1998). Managing across borders: The transnational solution. Harvard Business School Press.
Bhardwaj, A., Dietz, J., & Beamish, P. W. (2007). Host Country Cultural Influences on Foreign Direct Investment. Management International Review, 47(1), 29-50.
Björkman, I., Barner-Rasmussen, W., & Li, L. (2004) Managing Knowledge Tranfer in MNCs: The Impact of Headquarters Control Mechanisms. Journal of International Business Studies, 35(5), 443-455.
Boneau, C. A. (1960). The effects of violations of assumptions underlying the t test. Psychological Bulletin, 57, 49–64.
34 Journal of International Business Studies, 32(1), 177-189.
Ciabuschi, F., Martin, O. M., & Stahl, B. (2010). How Do MNC HeadQuarters Add Value?. Management International Review, 50(4), 471-491.
Collins, C. J., Smith, K. G. (2006). Knowledge exchange and combination: the role of human resource practices in the performance of high-technology firms. Academy of Management Journal, 49(3), 544-560.
Davenport, T. H., & Prusak, L. (1998). What do we talk about when we talk about knowledge? Boston, MA: Harvard Business School Press.
De Luque, M. S., & Javidan, M. (2004).Uncertainty avoidance. In R.J. House et al. (Eds.), Culture, leadership, and organizations: The GLOBE study of 62 societies, 602-653. Thousand Oaks, California: Sage
Doz, Y., Santos, J. F . P., & Williamson, P. J. (1997). From Global to Metanational. Harvard Business School Press, Boston.
Drogendijk, R., & Holm, U. (2012). Cultural distance or cultural positions? Analysing the effect of culture on the HQ-subsidiary relationship. International Business Review, 21, 383-396.
Drogendijk, R., & Zander, L. (2010). Walking the cultural distance: In search of direction beyond friction. In Devinney, T., Pedersen, T., & Tihanyi, I., (Eds.), The past, present and future of international business and management. Advances of International Management, 23, 189-212 Bingley, UK: Emerald.
Forsgren, M., Holm, U., & Johanson, J. (2005). Managing the embedded multinational: A business network view. Cheltenham: Edward Elgar.
Foss, N. J., & Pedersen, T. (2002). Transferring knowledge in MNCs: the role of sources of subsidiary knowledge in organizational context. Journal of International Management, 8(1), 49-67.
35 Geppert, M., Williams, K., & Matten, D. (2003). The social construction of contextual
rationalities in MNCs: An Anglo-German comparison of subsidiary choice. Journal of Management Studies, 40, 617–641.
Gomez-Mejia, L. R., Wiseman, R. M. (1997). Refraining executive compensation: an assessment and outlook. Journal of Management, 23, 291-374.
Gupta, A. K., & Govindarajan, V. (1986). Resource Sharing Among SBUs: Strategic
Antecedents and Administrative Implications, Academy of Management Journal, 29, 714-895.
Gupta, A. K., & Govindarajan, V. (2000). Knowledge Flows Within Multinational Corporations. Strategic Management Journal, 21(4), 473-496.
Hansen, M. T. (2002). Knowledge Networks: Explaining Effective Knowledge Sharing in Multiunit Companies. Organization Science, 13(3), 232-248.
Hedlund, G. (1994). A model of knowledge management and the N-form corporation. Strategic Management Journal, 15, 73-90.
Hofstede, G. (1980). Culture’s consequences: International differences in work-related values . Beverly Hills: Sage Publications.
Hofstede, G. (1983). The Cultural Relativity of Organizational Practices and Theories. Journal of International Business Studies, 14(2), 75-89.
Hofstede, G. (1991). Cultures and organizations. London: McGraw-Hill.
Jaffe, A., Trajtenberg, M., & Henderson, R. (1993). Geographic localization of knowledge spillovers as evidenced by patent citations. Quarterly Journal of Economics, 108, 577-598.
Kogut B., Zander U. (1992). Knowledge of the Firm, Combinative Capabilities, and the Replication of Technology. Organization Science 3(3), 383-397.
36 Kostova, T., & Roth, K. (2002). Adoption of organizational practice by subsidiaries of
multinational corporations: Institutional and relational effects. Academy of Management Journal, 45, 215–233.
Leung, K. R., Bhagat, N., Buchan, M., Erez, & C. Gibson. (2005). Culture and international business: recent advances and their implications for future research. Journal of International Business Studies, 36(4), 357-78.
Leyland, M. L. (2006). The Role of Culture on Knowledge Transfer: The Case of the Multinational Corporation. The Learning Organization, 13(3), 257-275.
Mahnke, V., Venzin, M. (2003). Governing knowledge teams in the MNC. Management International Review, 43(3), 47-67.
McEvily, S. K., & Chakravarthy, B. (2002). The persistence of knowledge-based advantage: an empirical test for product performance and technological knowledge. Strategic
Management Journal, 23(4), 285-305.
Mudambi, R., & Navarra, P. (2004). Is Knowledge Power? Knowledge Flows, Subsidiary Power and rent-seeking within MNCs. Journal of International Business Studies, 35(5), 385-406.
Newman, K. L., & Nollen, S. D. (1996). Culture and Congruence: The Fit between
Management Practices and National Culture. Journal of International Business Studies, 27(4), 753-779.
Nohria, N., & Ghoshal, S. (1997). The differentiated Network: Organizing Multinational Corporations for Value Creation. Jossey Bass, San Francisco (1997).
Noorderhaven, N. & Harzing, A. W. (2009) Knowledge-sharing and social interaction within MNCs. Journal of International Business studies, 40(5), 719-741.
Pak, Y., & Park, Y. (2004). A Framework of Knowledge Transfer in cross-border Joint Ventures: An Empirical Test of the Korean Context. Management International Review, 44, 417-434.
37 Podolny, J. M., & Page, K. L. (1998). Networks Forms of Organization. Annual Review of Sociology, 24, 57-76.
Pothukuchi, V., Damanpour, F., Choi, J., Chen, C. C., & Park, S. H. (2002). National and Organizational Culture Differences and International Joint Venture Performance. Journal of International Business Studies, 33(2), 243-265.
Randolph, W. A. (2000). Re-thinking empowerment: Why is it so hard to achieve? Organizational Dynamics, 29(2), 94-107.
Rues, T. H., & Lamont, B. T. (2009). The double-edged sword of cultural distance in international business. Journal of International Business Studies, 40, 1298-1316.
Salara, R. M., & Vaara, E. (2010). Cultural differences, convergence, and crossvergence as explanations of knowledge transfer in international acquisitions. Journal of International Business Studies, 41(8), 1365-1390.
Schein, E. H. (1990). Organizational culture. American Psychologist, 45(2), 109-119. Shane, S., Venkataraman, S., & MacMillan, I. (1995). Cultural differences in innovation championing strategies. Journal of Management, 21(5), 931-952.
Shankar, O., Luo, Y., & Yeheskel, O. (2008). From ‘distance’ to ‘friction’: Substituting metaphors and redirecting intercultural research. Academy of Management Review, 33, 905-923.
Shenkar, O. (2001). Cultural distance revisited: Towards a more rigorous conceptualization and measurement of cultural differences. Journal of International Business Studies, 32, 519-536.
Sigler, T. H., & Pearson, C. M. (2000). Creating an Empowering Culture: Examining the Relationship between Organizational Culture and Perceptions of Empowerment. Journal of Quality Management, 5, 27-57
38 Sveiby, K. E. (2001). A knowledge-based theory of the firm to guide in strategy formulation. Journal of Intellectual Capital, 2(4), 344-358.
Szulanski, G. (2000). The process of knowledge transfer: A diachronic analysis of stickiness. Organizational Human Decision Processes, 82(1), 9-27.
Taggart, J. H. (1998). Strategy Shifts in MNC Subsidiaries. Strategic Management Journal, 19, 663-681.
Tihanyi, L., Griffith, D. A., & Russell, C. J. (2005). The Effect of Cultural Distance on Entry Mode Choice, International Diversification, and MNC Performance: A Meta-Analysis. Journal of International Business Studies, 36(3), 270-283.
Tung, R. L., & Verbeke, A. (2010). Beyond Hofstede and GLOBE: Improving the quality of cross-cultural research. Journal of International Business Studies, 41, 1259-1274.
Van Oudenhoven, J. P. (2001). Do organizations reflect national cultures? A 10-nation study. International Journal of Intercultural Relations, 25, 89–107.
Venaik, S. & Brewer, P. (2010). Avoiding uncertainty in Hofstede and GLOBE. Journal of International Business Studies, 41(8), 1294-1315.
Yildiz, H. E. (2014). Not All Differences Are the Same: Dual Roles of Status and Cultural Distance in Sociocultural Integration in Cross-border M&As. Journal of International Management, 20, 25-37.
39
Appendix 1:
Case Processing Summary Mean position categorized in 4 options, high, low, high-low and low-high Cases
Valid Missing Total
N Percent N Percent N Percent rounded average impact subs competences on other units in the MNC low knowledge sharing 473 100,0% 0 0,0% 473 100,0% high knowledge sharing 400 100,0% 0 0,0% 400 100,0% headquarters driven 476 100,0% 0 0,0% 476 100,0% subsidiary driven 365 100,0% 0 0,0% 365 100,0% Descriptives
Mean position categorized in 4 options, high, low,
high-low and low-high Statistic
Std. Error
Average impact subs competences on other units in the MNC
low knowledge sharing Mean 3,7382 ,07028
Median 3,8000
high knowledge sharing Mean 3,9960 ,08102
Median 4,0000
40
Median 3,8000
subsidiary driven Mean 4,1750 ,08049
42
Appendix 4:
Tests of Normality
Kolmogorov-Smirnova Shapiro-Wilk
Statistic df Sig. Statistic df Sig.
Average impact subs competences on other units in the MNC
,055 1714 ,000 ,987 1714 ,000
a. Lilliefors Significance Correction
43
Appendix 6:
Tests of Normality
Kolmogorov-Smirnova Shapiro-Wilk
Statistic df Sig. Statistic df Sig.
rounded average impact subs competences on other units in the MNC
,135 1714 ,000 ,958 1714 ,000
