SUBSIDIARY DECISION-MAKING AUTONOMY: THE STRATEGIC ROLE AND CONTEXT OF THE SUBSIDIARY AND THE MODERATING EFFECT OF TOP
MANAGEMENT TEAM HETEROGENEITY
EDWIN KLASENS (S2373696) UNIVERSITY OF GRONINGEN
MASTER’S THESIS INTERNATIONAL BUSINESS & MANAGEMENT SUPERVISOR: DR. M. H. F. RIDDER DE VAN DER SCHUEREN
CO-ASSESSOR: DR. R. W. DE VRIES DATE: 21 JUNE 2017
1. ABSTRACT
2 Key words: subsidiary decision-making autonomy, TMT nationality heterogeneity, resource dependence theory, upper echelons theory, agency theory, headquarters-subsidiary relationship.
2. INTRODUCTION
The centralization or decentralization of decisions in multinational enterprises (MNEs) has had interest from scholars since the 1980s (Garnier, 1982; Gates & Egelhoff, 1986). The literature has emerged rapidly after this topic got interest from researchers and three main factors influencing subsidiary decision-making autonomy have been found: the strategic role of the subsidiary, the MNE’s control structure, and the context in which the subsidiary operates (De Jong et al., 2015). Prior research did not study TMT heterogeneity in relationship with subsidiary decision-making autonomy yet and therefore did not include upper echelons theory (Hambrick & Mason, 1984) so far. “The central premise of upper echelons theory is that executives' experiences, values, and personalities greatly influence their interpretations of the situations they face and, in turn, affect their choices” (Hambrick, 2007, p. 334). The reason why TMT heterogeneity can be influencing subsidiary decision-making autonomy as well is because heterogeneous TMTs may be more willing to give more autonomy than homogeneous TMTs due to their diverse backgrounds and perspectives.
3 alignment in the MNE (Ambos, Asakawa, & Ambos, 2011), resulting in an agency problem between top managers and local managers.
Upper echelons theory has been widely studied. The focus has often been on the relationship between TMT heterogeneity and firm performance or strategic (internationalization) choices (Nielsen, 2010; Certo, Lester, Dalton, & Dalton, 2006; Tihanyi, Ellstrand, Daily, & Dalton, 2000). Demographic characteristics of executives can be used as valid, albeit incomplete and imprecise, proxies of executives’ cognitive frames. These demographic characteristics have often been researched in previous studies and the findings were often relevant. This implies that researchers are able to rely on the demographic indicators as executives’ functional backgrounds, industry and firm tenures, educational credentials, and affiliations to develop predictions of strategic actions (Hambrick, 2007). Using these indicators is not exactly the same as the real psychological and social processes; however, researchers have generated substantial evidence that demographic profiles of executives (both individual and TMTs) are highly related to strategy and performance outcomes (Hambrick, 2007; Boeker, 1997; D’Aveni, 1990; Eisenhardt & Schoonhoven, 1990). An important ongoing debate regarding TMTs stresses the advantages and disadvantages of TMT heterogeneity. Scholars argue that TMT heterogeneity can generate a greater set of perspectives, more solutions, and increased creativity (Barkema & Shvyrkov, 2007). However, it can also lead to decreased communication, less effective decision making, and more conflicts (Ndofor, Sirmon, & He, 2015). This paper will focus on TMT nationality heterogeneity as there is a need to decompose the construct of heterogeneity to single attributes as every aspect of heterogeneity gives different results (Nielsen & Nielsen, 2013). Multinational teams include members with different cultures. This has been found to have an enduring impact on top managers’ mindsets (Geletkanycz, 1997) and interpretation and response to strategic issues (Schneider and De Meyer, 1991).
4 logic of resource dependence theory. The main reason is because its main focus is on the control the headquarters exert over subsidiaries (De Jong et al., 2015), therefore following the logic of agency theory. In this context, resource dependence theory focuses on subsidiaries controlling important resources for the MNE, therefore having a stronger influence over the decision-making processes within the MNE (Mudambi & Pedersen, 2007). TMT heterogeneity has not been studied yet regarding the headquarters-subsidiary relationship while this relationship can be influenced by the managers having to decide about the amount of autonomy given to the subsidiary. The composition of the TMT could therefore be relevant regarding the amount of decision-making autonomy the subsidiary receives as heterogeneous TMTs are expected to make different decisions about subsidiary decision-making autonomy than homogeneous TMTs (Nadolska & Barkema, 2014).
The research question of this paper is: What factors affect subsidiary decision-making
autonomy? In this paper, subsidiary decision-making autonomy is defined as “the extent to
which the subsidiary managers are able to make decisions without headquarters’ involvement” (Ambos, Andersson, & Birkinshaw, 2010, p. 1108). To answer this research question, three sub questions are of importance. As mentioned earlier, three factors have an influence on the subsidiary decision-making autonomy, of which the strategic role of the subsidiary and the context in which the subsidiary operates will be included in this study, following the logic of resource dependence theory. This leads to the following sub-questions:
How does the strategic role of the subsidiary influence subsidiary decision-making autonomy? How does the context in which the subsidiary operates influence subsidiary decision-making autonomy? Finally, the role of TMT nationality heterogeneity will be
studied in this paper, broadening the perspective of factors that can influence subsidiary decision-making autonomy as well. Therefore, the last sub question is: What is the effect of
top management team heterogeneity on subsidiary decision-making autonomy?
5 This study is relevant for managers as TMT demographics is a concept that can be actively monitored and changed, while the factors influencing subsidiary decision-making autonomy found in prior research are rather static and cannot be changed easily. Hence, changes in the TMT can lead to differences regarding the headquarters-subsidiary relationship, including the amount of subsidiary decision-making autonomy, making the headquarters-subsidiary relationship easier to control, actively monitor, and adjust when necessary. For researchers this study is relevant because of the broad perspective it has on the factors that could be influencing subsidiary decision-making autonomy by including other factors as TMT nationality heterogeneity and Hofstede’s (2001) cultural dimensions. Therefore, this study contributes to the literature by improving the understanding about the headquarters-subsidiary relationship and setting a basis for future research.
This paper will continue as follows. First, the theoretical frame of subsidiary decision-making autonomy and the upper echelons theory will be reviewed. Following from this review, hypotheses will be draft and the conceptual framework will be illustrated. Then, the methodology will be presented, which will be followed by the results that are found after analyzing the data. Finally, the results will be discussed and conclusions will be presented, in which limitations and remarks regarding future research will be discussed.
3. LITERATURE REVIEW
3.1. Agency theory and subsidiary decision-making autonomy
Agency theory has emerged around the 1960s-1970s and the main idea of this theory was that “an agency relationship has arisen between two (or more) parties when one, designated as the agent, acts for, on behalf of, or as representative for the other, designated the principal” (Ross, 1973, p. 134). The main focus of agency theory was on corporate governance in which it is a framework addressing the contract between the owners and the agents managing the firm on their behalf (Boyd & Solarino, 2016). Agency theory is built on three assumptions: (1) all actors are narrowly self-interested, (2) all actors are bounded rational, and (3) agents are more risk averse than principals (Eisenhardt, 1989). Following this line of thought, it can be concluded that agency problems exist between agents and principals as they have different interests and agency theory focuses on reducing the agency costs arising from these problems (Bosse & Phillips, 2016).
principal-6 principal problem, in which conflicts arise between the controlling and minority shareholders. Secondly, there is the shareholder-creditor problem, whereby shareholders gain large profits, while the risks are shifted to the debt holders who only have a right on a fixed percentage (Heremans, 2007). Agency theory has been used in more fields of research than only the corporate governance of the firm as principal-agent problems arise more often in other areas such as finance or international business. Focusing not solely on the corporate governance of the firm, other principal-agent problems come in play (e.g. top managers versus middle managers).
Although widely researched, agency theory has many criticasters. Aguilera & Jackson (2003, p. 448) state that “agency theory fails to sufficiently explore how corporate governance is shaped by its institutional embeddedness” and therefore is undersocialized. The assumption that principals and agents both are homogeneous and pursue their own interests is criticized as different types of shareholders can have different interests, while this argument is also valid for managers (Aguilera & Jackson, 2003). Moreover, agency theory has its focus only on the principal-agent problem and therefore other actors and their interdependent relationships are overlooked (Freeman, 1984). Finally, it is argued that the theory overlooks the institutional environment that influences the corporate governance of the firm (Lubatkin, Lane, Collin, & Very, 2001). Looking at studies conducted following the logic from agency theory, several inconsistent results have been found and therefore it can be concluded the theory should be refined to find more accurate, valid, and reliable explanations of agency relationships (Bosse & Phillips, 2016).
However, although agency theory has often been criticized, it has earned a place of prominence (Eisenhardt, 1989). In many fields of research it has been widely studied and the results have provided rich insights of principal-agent problems. In the field of subsidiary decision-making autonomy, following the agency perspective, conflicts among managers arise when managers at the headquarters are linked in an agency relationship with managers in operating divisions such as subsidiaries (Scharfstein & Stein, 2000). These subsidiaries will pursue their own interests and therefore are not solely mechanical instruments of headquarters’ will (Mudambi & Pedersen, 2007). Hence, local managers often strive for more decision-making autonomy for the subsidiary, while the headquarters’ managers will seek to maintain control in order to ensure efficiency and strategic alignment in the MNE (Ambos et al., 2011).
7 et al., 2010, p. 1108). Subsidiary decision-making autonomy has been widely researched and there are two main reasons why this has been such a popular topic. First, the headquarters-subsidiary relationship is of importance for the MNE as the headquarters-subsidiary is supposed to gain the largest possible value for the MNE as a whole. Hence, MNEs having a positive relationship with their subsidiary benefit when they are cooperating with its subsidiary. A challenging fact regarding the headquarters-subsidiary relationship is that subsidiaries are often located in a different geographical location in a different business environment (De Jong et al., 2015). In such a case, the right degree of subsidiary decision-making autonomy is crucial for the optimal performance of the subsidiary’s activities for the MNE as a whole. As De Jong et al. (2015, p. 875) state: “autonomy is a necessary (though not the only) requirement for the optimal performance of subsidiaries and their contribution to an MNE’s value chain.” Prior research focusing on subsidiaries building trust and linking up with the headquarters found that such activities are important for subsidiaries who wish to have more autonomy. It can be concluded that the headquarters-subsidiary relationship is a clear example of an agency problem in which the local manager strives for more autonomy, while the headquarters will seek to maintain control in order to ensure efficiency and strategic alignment within the MNE (Ambos et al., 2011).
The second reason why this topic gained such an interest is because it is important to know why some subsidiaries have more autonomy than others. Early studies from the 1980s-1990s had already been trying to answer this question. It was found then that factors influencing the subsidiary’s decision-making autonomy were the host market importance, the subsidiary’s assigned function, and the capabilities of the subsidiary (Doz & Prahalad, 1981; Martinez & Jarillo, 1989; Gupta & Govindarajan, 1991; Nohria & Ghoshal, 1997). These findings are in accordance with more recent studies that find that factors influencing subsidiary decision-making autonomy are the strategic role of the subsidiary (Ambos & Ambos, 2009; Rabbiosi, 2011), the context in which the subsidiary operates (Geppert & Williams, 2006; Birkinshaw & Hood, 2000), and the MNE’s control structure (Gaur & Lu, 2007; Johnston & Menguc, 2007; Maennik, Varblane, & Hannula, 2005).
8 thereby tries to bridge two fields of research. Looking from a wider perspective, this study tries to imply that factors have an influence can be found in a wider range, as many factors that could be promising have not been researched yet. It is of importance to have a broad perspective in order to find more of these factors influencing subsidiary decision-making autonomy, and therefore understanding this headquarters-subsidiary relationship better. Additionally, the effect of subsidiary decision-making autonomy on the firm’s performance is also unclear. De Jong et al. (2015) argue that subsidiary decision-making autonomy is essential for an optimal performance of the subsidiary for the MNE. However, recent studies have not empirically studied this relationship. More subsidiary decision-making autonomy is assumed to be better for the performance, but there may be negative effects as well, which are currently unknown. Moreover, the subsidiary’s decision-making autonomy’s effect on the firm’s performance may be smaller or larger than currently expected.
3.2. Resource dependence theory, the strategic role of the subsidiary, and the context of the subsidiary
Since the publication of The External Control of Organizations by Pfeffer & Salancik (1978), resource dependence theory became “one of the most influential theories in organizational theory and strategic management” (Hillman, Withers, & Collins, 2009, p. 1404). “Resource dependence theory is premised on the notion that all organizations critically depend on other organizations for the provision of vital resources, and that this dependence is often reciprocal” (Drees & Heugens, 2013, p. 1667). The amount of inter-organizational dependencies between formally independent companies influences the inter-organizational arrangement they will have, such as board interlocks, alliances, or joint ventures (Pfeffer & Salancik, 1978). Such arrangements can help organizations cope with interdependencies by bolstering their autonomy and legitimacy.
9 Barkema, 2001). Moreover, prior research has also focused on reviewing and interpreting past findings (e.g. Hillman et al., 2009).
According to Drees & Heugens (2013), the main question addressed by resource dependence theory is: Why do organizations enter into inter-organizational arrangements? It is argued that inter-organizational arrangements are instruments for reducing power imbalances and managing mutual dependencies (Casciaro & Piskorski, 2005) between a focal organization and organizations in its environment to which it depends for critical resources (Pfeffer & Salancik, 1978). When organizations implement these arrangements, it enables them to set their boundaries “at the point that maximizes strategic control over crucial external forces” (Santos & Eisenhardt, 2005, p. 495). Therefore, organizations ensure they have control over the strategic decisions that have to be made. However, when a focal organization depends on its partners resources, this control may decrease as the partner having these resources may require more control.
While resource dependence theory has been focusing on the organization-environmental relationship, this paper will follow its logic and extend it to the headquarters-subsidiary relationship. This is done because the MNE can be seen as a dispersed firm in which subsidiaries control unique and non-substitutable resources. Subsidiaries that control resources which are important for the MNE as a whole will have a stronger influence on the decisions made within the MNE (Mudambi & Pedersen, 2007). In this line of thought, subsidiaries also encounter internal competition within the MNE (Birkinshaw & Hood, 1998). When subsidiaries control important resources for the MNE, it is likely they will obtain more control over the decision-making processes, while other subsidiaries will find their control decreased or eliminated. Hence, resource dependence theory provides a basis for understanding headquarters-subsidiary and subsidiary-subsidiary relationships (Mudambi & Pedersen, 2007).
10 the fact headquarters seek control over subsidiary’s actions. This implies that subsidiaries otherwise will follow their self-interest instead of the headquarters’ interest.
The strategic role of the subsidiary can be defined as “the extent to which a subsidiary contributes to the competitive strategy of the firm” (Vereecke & Van Dierdonck, 2002, p. 15). Some subsidiaries are performing better and are more important to MNEs than others. When subsidiaries add more value to the MNE, it is more likely they will take more (or full) responsibility for the production process of their products (De Jong et al., 2015). Ambos & Ambos (2009) found that such subsidiaries generate firm-specific competences resulting in more decision-making autonomy. This line of thought was confirmed and extended by Rabbiosi (2011), who added that subsidiaries with a superior knowledge base compared to other subsidiaries are less dependent on their headquarters and the MNE network and have therefore more decision-making autonomy.
Therefore, prior research suggests that a higher subsidiary performance results in more decision-making autonomy (De Jong et al., 2015). In general, findings regarding this relationship are consistent, although there are some studies suggesting there is a decreasing marginal return of subsidiary size on subsidiary decision-making autonomy (Johnston & Menguc, 2007; Johnston, 2005). The precise form of causality of this relation is therefore not clear yet. However, most studies indicate that larger subsidiaries have more decision-making autonomy (De Jong et al., 2015).
In sum, following the logic from resource dependence theory, it can be suggested that subsidiaries contributing more to the competitive strategy of the MNE are the ones controlling important resources. As argued, such subsidiaries will obtain more control over the decision-making process than subsidiaries having less important resources and therefore contributing less to the competitive strategy of the MNE (Mudambi & Pedersen, 2007). Therefore, it can be concluded that subsidiaries contributing more to the competitive strategy of the MNE will have more autonomy to make decisions. The first hypothesis is therefore as follows:
Hypothesis 1: When a subsidiary is contributing more to the competitive strategy of a
multinational enterprise, its decision-making autonomy will likely be higher.
11 1972, p. 314). Local circumstances determine the ability of subsidiaries to develop capabilities and competences (Geppert & Williams, 2006). Prior research suggests that some less regulated markets make subsidiaries better able to add value for the MNE (De Jong et al., 2015). Therefore, subsidiaries operating in liberal market economies (LMEs) should be better able to add value to the MNE, as innovation is advancing radically in these economies. Subsidiaries operating in coordinated market economies (CMEs) will find that innovation is more incremental in these economies and therefore these economies do not allow the subsidiaries to add as much value for the MNE as in LMEs (Hall & Gingerich, 2009). Innovative contexts allow subsidiaries to be more autonomous as these subsidiaries develop competences which the MNE has not. Examples of this phenomenon are described in the paper of Birkinshaw & Hood (2000), who found that subsidiaries in leading-edge industries are more autonomous, more locally integrated, and internationally focused than other subsidiaries. Maennik et al. (2005) added to this line of thought that subsidiaries in highly technology industries need to undertake actions which require high levels of autonomy.
12 creating an understanding of the relationship between such economies and subsidiary decision-making autonomy.
In sum, following again the logic of the resource dependence theory, low power distance and uncertainty avoidance may enable subsidiaries to develop crucial capabilities and resources due to the innovative environment, making the MNE depend more on the subsidiary’s activities. This dependence from the MNE on the subsidiary could be leading to more decision-making autonomy for the subsidiary. This leads to the following hypotheses: Hypothesis 2a: When a subsidiary is operating in a country with a lower dimension of power
distance, its decision-making autonomy will likely be higher.
Hypothesis 2b: When a subsidiary is operating in a country with a lower dimension of
uncertainty avoidance, its decision-making autonomy will likely be higher.
3.3. Upper echelons theory and TMT nationality heterogeneity
Upper echelons theory was initially created by Hambrick & Mason (1984) and later adjusted again by Hambrick (2007). “The central premise of upper echelons theory is that executives' experiences, values, and personalities greatly influence their interpretations of the situations they face and, in turn, affect their choices” (Hambrick, 2007, p. 334). Upper echelons theory is therefore central when it comes to decisions of TMTs and/or individual members. Díaz-Fernández, González-Rodriguez, & Simonetti (2015) argue that strategic choices of managers are influenced by their demographic characteristics (e.g. age, gender, and nationality). Hence, the background of managers could be of importance when it comes to strategic decision making.
13 TMT demographic heterogeneity is of importance as prior studies show that it may increase firm performance through its association with cognitive heterogeneity and task conflict (Certo et al., 2006). Task conflict is seen as positive as it may generate better decisions (Pelled, Eisenhardt, & Xin, 1999). This is in accordance with the cognitive lens from Barkema & Shvyrkov (2007), who argue in their study that demographically homogeneous TMTs are less likely to make novel decisions. Task-related conflict among demographically heterogeneous TMTs is therefore positive as team members become aware of more issues, perceive these issues differently, and are more likely to propose alternative courses of action (Barkema & Shvyrkov (2007). In sum, demographically heterogeneous TMTs generate more new ideas and approaches (Amason, 1996), leading to more novel strategic decisions.
However, looking at the social lens of demographically heterogeneous TMTs, strong faultline settings may divide teams into hardened subgroups (Barkema & Shvyrkov (2007). Faultline settings arise when several team members share a certain demographic (e.g. professional background). Such a situation can lead to decreased communication between subgroups as members will identify themselves more with the subgroups rather than with the overall TMT (Li & Hambrick, 2005). Also, subgroups may add little to the overall task (Barkema & Shvyrkov (2007). In such a case, the team may not fully use the cognitive resources of the group members, leading to less novel ideas. Therefore, it can be suggested that although many positive effects of TMT heterogeneity were found, it can also result in negative effects, leading to lower performances of the TMT.
14 demographic heterogeneity is expected to strengthen the positive relationship between the strategic role of the subsidiary and its decision-making autonomy. This leads to the following hypothesis:
Hypothesis 3: A top management team with more nationality heterogeneity is likely to
positively influence the relationship between the strategic role of the subsidiary and subsidiary decision-making autonomy.
A similar positive effect is expected regarding TMT nationality heterogeneity influencing the negative relationships between the power distance and uncertainty avoidance dimensions of Hofstede (2001) and subsidiary decision-making autonomy. This is especially the case when subsidiary operate in innovative environments that require decision-making autonomy for the subsidiary as complex actions must be undertaken (Maennik et al., 2005). Nationally heterogeneous TMTs may see this need more because of the diverse set of national backgrounds than homogeneous TMTs. Therefore, the following hypotheses are as follows: Hypothesis 4a: A top management team with more nationality heterogeneity is likely to
positively influence the negative relationship between power distance and subsidiary decision-making autonomy.
Hypothesis 4b: A top management team with more nationality heterogeneity is likely to
positively influence the negative relationship between uncertainty avoidance and subsidiary decision-making autonomy.
3.4. Conceptual model
15 Figure 1: The conceptual model
In the next section, the methodology of this paper will be presented. The focus will be on the dataset, the measurement scales, and testing methodology. This will be followed by the results section, in which the plain results after testing the data will be presented, continued by the discussion of these results. The focus of this discussion will be on discussing the results regarding hypotheses that are presented in the theory section. Finally, the research question and sub questions will be answered and remarks regarding limitations and future research will be made.
4. METHODOLOGY
4.1. Data collection and sample
16 survey in Central and Eastern European (CEE) countries (De Jong et al., 2015; Van Dut, 2014). In both the surveys of IWH (2001) and conducted by researchers directly, certain decision topics are presented and respondents are asked to tell whether the subsidiary or the parent company decides about those topics. Other studies of subsidiary decision-making autonomy used a database like Orbis (Van Dut, 2014), which will also be used in this study. This method does not directly test subsidiary decision-making autonomy and therefore, to be able to conduct the tests, a dummy variable will be made. Although this indirect measurement of subsidiary decision-making autonomy is a disadvantage of using a database like Orbis, it has advantages as well. The database contains comprehensive firm-level data of millions of companies worldwide. In Orbis, a large amount of information is available about a substantial amount of economic activity. Therefore, this is the most appropriate database for this study. The information used in this study is from the financial reports of the subsidiaries and parent companies of the years 2015 and 2016. The hypotheses can be tested by using this information as it includes data of the key constructs (e.g. strategic role of the subsidiary). Moreover, measures regarding the control variables (e.g. size of parent company) can be conducted with this data.
The sample consists of 183 foreign subsidiaries from 19 European countries operating in the industrial sector that belong to 108 MNEs from the Netherlands. The procedure of comprising the dataset can be found in appendix A. From the original list of 184 subsidiaries, 1 subsidiary was deleted due to duplicity. Companies are seen as subsidiaries when an MNE contains 20% of their shares, therefore using the 20% definition of control (La Porta, Lopez-de-Silanes, & Shleifer, 1999). This definition states that when an owner has 20% of the shares (voting rights), it has the power to control the company. In this case, therefore, the MNE controls the subsidiary when it has at least 20% of the shares. The choice for this setting was made as the Dutch TMTs of parent companies are often more nationally heterogeneous than other countries as Germany and the United Kingdom (Van Veen & Elbertsen, 2008) and a diverse set of subsidiary context in this study is of importance for measuring the hypotheses. Moreover, the data of these large companies helps constructing datasets with complete observations (Rugman & Oh, 2010).
4.2. Variable measurement
4.2.1. Dependent variable: subsidiary decision-making autonomy
17 headquarters directly as in a case study or survey-based approach. Therefore, a proxy for subsidiary decision-making autonomy has been made using the approach of Van Dut (2014). According to this approach, there is a list of ten decision dimensions that primarily relate to business functions of subsidiaries and other potentially important management activities (Jindra, Giroud, & Scott-Kennel, 2009). Based on the dataset from Orbis, a dummy variable has been created for each of the ten dimensions, which is as follows: “R&D = 1 if the subsidiary undertakes R&D activities and 0 otherwise; Manufacturing = 1 if the subsidiary undertakes manufacturing activities and 0 otherwise; Marketing = 1 if the subsidiary undertakes marketing activities and 0 otherwise; Sales = 1 if the subsidiary undertakes sales activities in the domestic market and 0 otherwise; Market scope = 1 if the subsidiary serves foreign markets and 0 otherwise; Network = 1 if the subsidiary engages in network activities within the MNE and 0 otherwise; Outsourcing = 1 if the subsidiary engages in outsourcing activities and 0 otherwise; Cooperation = 1 if the subsidiary cooperates with external organizations and 0 otherwise; Export-import = 1 if the subsidiary engages in export and/or import activities and 0 otherwise; Subsidiary establishment = 1 if the subsidiary has its own subsidiary and 0 otherwise” (Van Dut, 2014, p. 55). The scores of the ten different dummies have been scored into one variable, ranging from 0 to 10, which has served as the proxy for subsidiary decision-making autonomy. This variable has therefore been measured on an interval level.
Although this dummy variable did not test subsidiary decision-making autonomy directly, it has been a suitable proxy because subsidiaries that perform in more business functions or activities as argued with this dummy variable will also have more decision-making autonomy as a wider range of activities leads to more complex managerial tasks and specializations (Van Dut, 2014). Moreover, the conducted correlation and exploratory factor analysis showed that no sub-categories of subsidiary decision-making autonomy exist.
18 this study has measured the relative size of the subsidiary (De Jong et al., 2015), and more specifically, the turnover a subsidiary generates relatively to the parent company’s total turnover. The data used in this study was from subsidiaries and MNE’s from 2015 and 2016. This measurement has directly been showing the relative size of the subsidiary. Moreover, the subsidiary’s turnover is related to its value-adding activities as it is the income from its operations. The reason the net profit was excluded from this study is because this variable is affected by other factors as costs, economic environment, currency exchange rates and so on. Moreover, a large company can make a loss, while a smaller company can make a profit. This would give an incorrect view as the small company would be shown as larger than the large firm. Therefore, the turnover is more stable and a more direct result of operational activities and hence, better showing the size of a certain subsidiary or headquarters. The relative subsidiary’s turnover has been measured as a percentage of the parent company’s total turnover; therefore this construct has been measured on an interval level ranging from 0 to 100.
The context in which the subsidiary operates focused on the innovativeness of the environment of the subsidiary (Birkinshaw & Hood, 2000). The choice was made in this study to focus on certain cultural dimensions of Hofstede (2001), as culture can be related with innovativeness as well, as prior research has shown. A recent study conducted by Strychalska-Rudzewicz (2016) made clear that in European countries a lower power distance and lower uncertainty avoidance in most cases results in higher innovativeness. Therefore, these dimensions have been included in this study as well. Moreover, Strychalska-Rudzewicz (2016) found that in the European context, more individualistic countries also achieve better innovative results, but this relationship is more debatable. Therefore, this dimension has been included as a control variable together with the measurement whether the subsidiary is located in an LME or CME (Hall & Gingerich, 2009). As mentioned above, the data has been derived from Hofstede’s (2001) database of cultural dimensions, which shows data on an interval level as the values range from 0 to 100.
4.2.3. Moderator: TMT nationality heterogeneity
19 the boards of companies. TMT nationality heterogeneity has been measured with a Blau index (Blau, 1977) and the formula that has been used to measure it is B = [1 - ⅀(𝑝𝑖)2], where p is the percentage of members in the ith group (Nielsen & Nielsen, 2013). Therefore, this construct was also measured on an interval level as the TMT nationality heterogeneity has been measured as a percentage ranging from 0 to 100.
4.2.4. Control variables
Three sets of control variables have been included in this study. The first set of control variables focused on the context of the subsidiary. From the literature it became clear that there could be more than only cultural factors influencing subsidiary decision-making autonomy. Therefore, such factors have been included in this study as control variables. The first control variable in this set is whether subsidiaries are operating in an LME or CME, which will be measured on a nominal level (Hall & Gingerich, 2009). The second control variable focuses on the countries in which the subsidiaries are operating, which has been measured on an ordinal level as the scores could not be placed in a rank. Finally, the individualism versus collectivism dimension of Hofstede (2001) has been included as a control variable as it was a factor that could have an influence for the innovativeness of companies operating in European countries, therefore leading to more subsidiary decision-making autonomy, although its effect was debatable (Strychalska-Rudzewicz, 2016).
Secondly, one control variable has focused on the parent companies of the subsidiaries. Although all MNEs included in the study were relatively large, differences between them existed. Therefore, company size has been included as a control variable. Having a larger parent company may result in more decision-making autonomy for the subsidiary. This is due to the fact that larger MNEs may have a more decentralized structuring of activities, which facilitates the subsidiary’s decision-making autonomy (Van Dut, 2014). It has been measured by the amount of employees of the MNE, therefore on an interval level.
decision-20 making autonomy. Higher shareholder ownership, and especially parent companies with a majority of the voting rights, may control subsidiaries more extensively. Prior research found a negative relationship between extensive control of the MNE and subsidiary decision-making autonomy (Johnston & Menguc, 2007; Maennik et al., 2005). Therefore, shareholder ownership has also been included as a control variable. It has been measured as the percentage of shares the parent company has of the subsidiary, being measured on an interval level as well. Finally, the size of the subsidiary itself may be relevant for its decision-making autonomy. In hypothesis 1 the relative size has been measured, but in a huge MNE a large subsidiary may relatively contribute only little to the company group’s turnover. However, such a subsidiary may have tangible (capital) and intangible resources (managerial talent) creating competitive advantages for the MNE (Van Dut, 2014). Therefore, more decision-making autonomy may be the result when the subsidiary is increasing in size. It has been measured as the number of employees working at the subsidiary. Hence, this control variable has been measured on an interval level as well.
4.3. Data analysis
21 5. RESULTS
5.1.Descriptive statistics
Based on the analysis focusing on outliers in the dataset, the choice was made to exclude cases where the subsidiary was older than 80 years, when there were more than 15,000 employees working for the subsidiary, when there were more than 200,000 employees working for the MNE, and when the relative turnover of the subsidiary was more than 1,000%. These outliers were excluded as their extreme number may have such a large effect that the results would not be accurate and reliable. Also cases with missing values were excluded from the data, bringing the final amount of subsidiaries in this study to 123.
The descriptive statistics be found in table 1. Adding to these statistics, it is of importance to know in which countries subsidiaries operate. Excluding the subsidiaries with outliers or missing numbers, there were 123 subsidiaries. Of these subsidiaries, 17 were from Slovakia, 9 from Sweden, 1 from Romania, 2 from Portugal, 2 from Poland, 3 from Latvia, 1 from Lithuania, 6 from Italy, 2 from Ireland, 19 from the United Kingdom, 13 from France, 3 from Finland, 5 from Spain, 2 from Estonia, 10 from Denmark, 11 from Germany, 3 from Czech Republic, 12 from Belgium, and 2 from Austria. Also it was tested whether a subsidiary operated in an LME or CME. The statistics have showed that 21 subsidiaries operated in LMEs and 102 subsidiaries in CMEs.
Variable N Minimum Maximum Mean Std.
Deviation
Age of the subsidiary 123 1 79 24.24 14.92
Percentage of shares held by MNE 115 24.64 100 95.53 13.53 Employees of subsidiary 123 3 9,401 680.22 1486.24 Turnover of subsidiary 123 289 3,034,015 246,866.41 562,890.03 Individualistic culture 123 27 89 68.55 13.29 Power distance 123 11 100 52.13 25.17 Uncertainty avoidance 123 23 99 59.52 24.20 Turnover of MNE 123 8,670 97,242,600 11,811,700.34 18,536,545.35 Employees of MNE 123 3 114,731 12,845.33 22,047.23 Relative turnover of subsidiary 123 0.02 625.22 56.69 123.95 TMT nationality heterogeneity 123 0 85 37.50 26.15 Subsidiary decision-making autonomy 123 1 8 3.44 1.57 Valid N (listwise) 115
22 5.2. Testing the data
5.2.1. Factor analyses
The correlations for the independent variables, control variables, and dummy variables for the measurement of subsidiary decision-making autonomy are presented in table 2-4.
1 2 3 4 5 6 7
Age of subsidiary 1
Percentage of shares held by MNE .120 1
Country of subsidiary .287** -.059 1 LME/CME .099 -.142 .014 1 Individualistic culture .049 .036 .334** -.635** 1 Employees MNE -.053 -.122 .029 .150 -.027 1 Employees subsidiary .015 -.172 .025 -.019 .025 .129 1 ** P < 0.01 * p < 0.05
Table 2: Correlation control variables
1 2 3
Power distance 1
Uncertainty avoidance .444** 1
Relative turnover subsidiary .046 -.071 1
** p < 0.01 * p < 0.05
Table 3: Correlation independent variables
1 2 3 4 5 6 7 8 9 10 R&D 1 Manufacturing .251** 1 Marketing .096 .012 1 Sales -.031 .176 .099 1 Market Scope .130 .085 .186* .045 1 Network -.057 .118 .179* .049 .082 1 Outsourcing -.037 .105 -.043 .019 -.059 .034 1 Cooperation -.037 -.077 .189* .019 -,059 .034 -.008 1 Export-Import -.088 -.006 .185* -.003 ,661** .121 -.045 -.045 1 Subsidiary establishment .372** .063 .217* .065 .449** .013 -.066 .123 .112 1 ** p < 0.01 * p < 0.05
Table 4: Correlation of dummy variables of subsidiary decision-making autonomy
23 adequacy of .442, while the acceptable minimum is .6. Therefore, it was concluded that no significant underlying dimensions were present. For the independent variables, the KMO-measure showed an adequacy of .482. Hence, no underlying dimension was identified and the current set of variables was used. Finally, the same factor analysis was conducted for the dummy variables of subsidiary decision-making autonomy. This analysis also showed no significant underlying dimensions as the KMO-measure showed an adequacy of .510. Hence, the used variables for creating the dummy variables have no significant underlying dimensions and have been used to test the construct of subsidiary decision-making autonomy.
5.2.2. Testing the hypotheses and model
Variable B Std. Error Beta t Sig.
(Constant) 1.431 1.242 1.162 0.252
Age of subsidiary 0.031 0.010 0.276 3.064 0.003
Percentage shares held by MNE 0.002 0.011 0.015 0.161 0.873
Employees subsidiary 0.000 0.000 0.169 1.856 0.066
Individualism vs. collectivism 0.014 0.010 0.122 1.358 0.177
Employees MNE 4.224E-006 0.000 0.061 0.672 0.503
Table 5: control variables on subsidiary decision-making autonomy
Variable B Std. Error Beta t Sig.
(Constant) 0.258 1.720 0.208 0.835
Relative turnover subsidiary 2.386E-005 0.001 0.002 0.021 0.983
Power distance -0.001 0.007 -0.018 -0.151 0.880
Uncertainty avoidance 0.009 0.007 0.141 1.299 0.197
Age of subsidiary 0.029 0.011 0.255 2.718 0.008
Table 6: independent variables on subsidiary decision-making autonomy
24 linear regression has been conducted of relative turnover on the amount of subsidiary decision-making autonomy. The regression was not significant: R² = 0.082, F(5.337), p = 0.933, therefore the relative turnover does not have an influence on the amount of subsidiary decision-making autonomy (B = 2.386E-005). Hence, hypothesis 1 was not supported by the data. A regression has also been conducted of uncertainty avoidance and power distance on subsidiary decision-making autonomy. Power distance was not significant: R² = 0.085, F(2, 5.667), p = 0.880. Hence, Hofstede’s (2001) dimension power distance does not have an influence on subsidiary decision-making autonomy (B = -0.001). Hypothesis 2a is therefore not supported. Uncertainty avoidance was not significant as well: R² = 0.083, F(5.396), p = 0.197. Therefore, the Hofstede’s (2001) dimension of uncertainty avoidance had no influence on subsidiary decision-making autonomy (B = 0.009). Hypothesis 2b was also not supported by the data.
Variable B Std. Error Beta t Sig.
(Constant) 2.705 0.267 10.125 0.000
TMT heterogeneity average 0.002 0.006 0.032 0.328 0.744 Relative turnover average -0.001 0.002 -0.075 -0.505 0.615 Relative turnover * TMT
heterogeneity
3.903E-005 0.000 0.066 0.451 0.653
Age of subsidiary 0.029 0.009 0.278 3.105 0.002
Table 7: tests hypothesis 3
Variable B Std. Error Beta t Sig.
(Constant) 2.990 0.205 14.559 0.000
TMT heterogeneity average 0.001 0.004 0.021 0.276 0.783 Power distance average -0.008 0.005 -0.125 -1.628 0.105 Power distance * TMT
heterogeneity
0.000 0.000 0.128 1.674 0.096
Age of subsidiary 0.019 0.006 0.232 2.990 0.003
25
Variable B Std. Error Beta t Sig.
(Constant) 2.719 0.261 10.407 0.000
TMT heterogeneity average -0.002 0.005 -0.035 -0.392 0.696 Uncertainty avoidance average 0.003 0.006 0.052 0.592 0.555 Uncertainty avoidance * TMT
heterogeneity
0.001 0.000 0.217 2.462 0.015
Age of subsidiary 0.029 0.009 0.275 3.143 0.002
Table 9: tests hypothesis 4b
To test the moderating effect of TMT nationality heterogeneity on the relationships between the independent variables and subsidiary decision-making autonomy, a regression analysis was conducted. In the case of the relative size of the subsidiary (table 7), the moderator had no significant effect: R² = 0.84, F(2.688), p = 0.653. Therefore, TMT nationality heterogeneity had no significant effect on the relationship between the subsidiary’s relative size and its decision-making autonomy. Hypothesis 3 is therefore not supported by the data. TMT nationality heterogeneity was expected to moderate on the relationships between power distance and uncertainty avoidance dimensions and subsidiary decision-making autonomy. Looking at power distance (table 8), this was not significant: R² = 0.018, F(0.732), p = 0.096. Hence, no moderating effect of TMT nationality heterogeneity exists on the relationship between power distance and subsidiary decision-making autonomy. Hypothesis 4a was not supported by the data. Finally, the moderating effect was tested for uncertainty avoidance in relation to subsidiary decision-making autonomy (table 9). This was significant at a 5% confidence level: R² = 0.127, F(4.309), p = 0.015. This means that in the relationship between uncertainty avoidance and subsidiary decision-making autonomy, more TMT nationality heterogeneity strengthens the positive effect of uncertainty avoidance on subsidiary decision-making autonomy (B = 0.001). Therefore, hypothesis 4b is supported at a 5% confidence level.
26 Variable Model 1 Explanatory variables only Model 2 Including control variables Model 3 Including moderator
B t-value B t-value B t-value
(Constant) 3.420 8.243** 0.358 0.208 0.329 0.178 Power distance -0.11 -1.725 -0.001 -0.151 -0.001 -0.156 Uncertainty Avoidance .010 1.514 0.009 1.299 0.009 1.277 Relative size subsidiary .000 -0.088 2.3860E-5 0.21 6.0880E-6 0.005 Age of subsidiary Percentage shares held by MNE Employees of subsidiary Employees of MNE Individualistic culture TMT nationality heterogeneity 0.029 0.005 0.000 2.7140E-6 0.020 2.718** 0.436 1.711 0.408 1.503 0.029 0.005 0.000 2.6110E-6 0.020 0.000 2.610* 0.437 1.682 0.371 1.484 0.047 ANOVA F(1.249), p = 0.295 F(2.280), p = 0.027 F(2.007), p = 0.045 R² 0.031 0.147 0.147 Δ R² 0.006 0.082 0.074 Observations 122 114 114 ** p < 0.01 * p < 0.05
Table 10: Linear regression models
27 6. DISCUSSION
It can be concluded that the results were different than expected. Although the tests regarding the model showed significance at a 5% confidence level, the direct and moderating effects were not significant. Only when testing the moderating effect of TMT nationality heterogeneity on the relationship between uncertainty avoidance and subsidiary decision-making autonomy directly, a significant result was found at a 5% confidence level. One interesting finding of the analyses was the significant finding of the positive relationship between the age of subsidiary, which was used as a control variable, and subsidiary decision-making autonomy. This implies that the age of the subsidiary can be an important factor regarding the amount of decision-making autonomy a subsidiary has. Having a positive, established relationship with the parent company may be crucial for the decision-making autonomy. The relational factors could therefore be of importance as well and should be included into future research.
Prior empirical research suggested a positive relationship between the strategic role of the subsidiary and its decision-making autonomy due to the availability of competences and a valuable knowledge base (Ambos & Ambos, 2009; Rabbiosi, 2011). Following the logic form resource dependence theory such a subsidiary would then gain more autonomy due to the MNE’s dependence on the subsidiary. This has not been the case in this study as no significant relationship was found between the strategic role of the subsidiary and its decision-making autonomy, therefore hypothesis 1 is not supported by the data. One explanation for this result is that even though strategically important subsidiaries could add a lot of value for the MNE, the headquarters prefer to keep control of the subsidiary’s actions and not to give it much autonomy. This can be explained by agency theory as headquarters try to keep control close to improve efficiency and effectivity, while subsidiaries may follow their own interests (Ambos et al., 2011). Strategy alignment is of importance in such a case as headquarters require all operating units to operate according to the MNE’s strategy. Therefore, having important resources or capabilities for the MNE may not influence the subsidiary’s decision-making autonomy.
28 Pedersen, 2007). Hence, decision-making autonomy may be a result of the subsidiary’s non-financial value, while this study measured its non-financial value to the MNE. Finally, subsidiary decision-making autonomy itself was measured with a dummy variable (Van Dut, 2014). Even though the measurement of the subsidiary’s activities can give an understanding of its autonomy, it does not measure it directly. Other approaches as direct surveys (IWF, 2011) could therefore be resulting in more accurate measurement of this construct.
The context in which the subsidiary operates was also expected to result in more subsidiary decision-making autonomy as subsidiaries operating in innovative environments would be able to develop important resources and capabilities for the MNE (Birkinshaw & Hood, 2000). Following the logic from resource dependence theory, this could lead to more decision-making autonomy for the subsidiary due to its valuable resources and capabilities. Moreover, operating in such environments may require more autonomy due to the complex managerial actions that need to be undertaken (Maennik et al., 2005). The focus of this study was on Hofstede’s (2001) cultural dimensions, and specifically on the dimensions power distance and uncertainty avoidance. Countries with low scores on these dimensions have been found to be more innovative (Strychalska-Rudzewicz, 2016), Hence, subsidiaries operating in such countries could develop important resources and capabilities for the MNE.
The results, however, showed no significant relationships between the cultural dimensions and subsidiary decision-making autonomy. Hence, hypotheses 2a and 2b were not supported by the data. One possible explanation is again that even though subsidiaries develop important resources and/or capabilities for the MNE, this may not result in decision-making autonomy. This can happen when the subsidiary historically had no decision-decision-making autonomy already, therefore having an established relationship with the headquarters that could be difficult to change. However, it can also be the case of headquarters refusing to give decision-making autonomy, controlling the strategy alignment in the MNE group. Hence, agency theory could also explain this non-significant relationship by arguing that subsidiaries, whether or not they may develop important resources and capabilities for the MNE due to its innovative context, will not receive much decision-making autonomy as the headquarters continue to decide about strategic decisions themselves to ensure efficiency, effectiveness, and strategy alignment within the MNE (Ambos et al., 2011).
29 (Strychalska-Rudzewicz, 2016), therefore leading to more decision-making autonomy. It could be possible this assumption is incorrect and that such a direct relationship does not exist. Also, to be able to better understand the relationship between Hofstede’s (2001) cultural dimensions and a country’s innovativeness, more empirical work in this field is necessary. Hence, future studies focusing on this relationship should be conducted to get a better understanding of the phenomenon.
TMT nationality heterogeneity was expected to moderate the previously discussed relationship as more nationally heterogeneous TMT may be able to give more decision-making autonomy to the subsidiary. Following the cognitive lens from Barkema & Shvyrkov (2007), this study focused on the positive effects from MNEs having heterogeneous TMT. Strategically important subsidiaries and subsidiaries operating in innovative environments may have or develop important resources for the MNE. These effects could be strengthened when team members of the TMT are nationally heterogeneous, as they may see the need of foreign local managers to have decision-making autonomy for adding as much as possible value for the MNE (De Jong et al., 2015).
The moderating effects did not show the expected results either. Hypothesis 3 was not supported by the data as there was no significant moderating effect of TMT nationality heterogeneity on the relationship between the strategic role of the subsidiary and its decision-making autonomy. One argument for not finding a significant moderating effect is due to the lack of a significant direct effect of the strategic role of the subsidiary and its decision-making autonomy in the first place. It could also be the case that, as mentioned before, the strategic role of the subsidiary is influencing subsidiary decision-making autonomy due to non-financial resources. This relationship could therefore be moderated by TMT nationality heterogeneity, which should be a research interest for future studies.
30 subsidiary decision-making autonomy as predicted. Therefore, nationally heterogeneous TMTs may take the local environment of the subsidiary more into consideration due to the team members’ backgrounds than nationally homogeneous TMTs.
7. CONCLUSION
In this study, subsidiary decision-making autonomy was researched from a broad perspective. The existing literature tends to focus only on a few factors influencing subsidiary decision-making autonomy, while this study tried to gain a better understanding and provide a basis for future research by including more factors from other fields of research. In doing so, this study tried to answer the research question: What factors affect subsidiary decision-making
autonomy? To answer this research question, three sub-questions were formulated, which
were as follows: How does the strategic role of the subsidiary influence subsidiary
decision-making autonomy? How does the context in which the subsidiary operates influence subsidiary decision-making autonomy? What is the effect of top management team heterogeneity on subsidiary decision-making autonomy? To answer these questions, this
study built on the existing literature by following the resource dependence theory, and specifically the earlier found factors influencing subsidiary decision-making autonomy. Moreover, upper echelons theory was included, therefore combining these two fields of research with the aim of bridging this gap in the existing literature. Afterwards, data from the Orbis database was derived and tested using SPSS. Finally, these results were discussed and it is now time to conclude this study by answering these questions.
From the results it can be concluded that the strategic role of the subsidiary does not have a significant effect on its decision-making autonomy. The same conclusion can be made for the second sub question, the role of the context in which the subsidiary operates in relation to its decision-making autonomy. Also, no significant relationship was found for this factor regarding subsidiary decision-making autonomy. For the moderating effect of TMT nationality heterogeneity a significant effect was found on the relationship between Hofstede’s (2001) dimension uncertainty avoidance and subsidiary decision-making autonomy was found, implying that a nationally heterogeneous TMT will strengthen this relationship, while a nationally homogeneous TMT will weaken this relationship.
31 autonomy in certain situations as nationally heterogeneous TMTs may consider the local environment of the subsidiary more than nationally homogeneous TMTs. Moreover, a positive effect of the age of the subsidiary was found in relation to its decision-making autonomy. In sum, it can be concluded that TMT nationality heterogeneity as a moderator and the age of the subsidiary seem to have a positive influence on subsidiary decision-making autonomy.
The implications of this study for researchers are that subsidiary decision-making autonomy is a wide construct and more factors can have an influence than the factors that have been studied in this field of research so far. An established, positive relationship between the subsidiary and the headquarters may be more important than the importance of the subsidiary in terms of value-adding activities. Relational factors and theories could have an influence on subsidiary decision-making autonomy and should therefore be further studied in future research. Moreover, the TMT of the parent company can be a crucial factor as eventually the TMT decides about the amount of decision-making autonomy for the subsidiaries. Hence, the backgrounds of the team members are relevant to study more in the future as these backgrounds determine the team members’ decisions (Hambrick, 2007). For managers, the results are relevant for the subsidiaries itself when they look for more decision-making autonomy. Working on a close, positive relationship with the headquarters could increase the subsidiary’s decision-making autonomy. For the MNEs, having nationally heterogeneous TMTs may lead to an increase of the subsidiaries’ decision-making autonomy, giving these MNEs an active way of being able to influence the subsidiary’s decision-making autonomy by hiring a nationally heterogeneous TMT.
32 Although the results have certain contributions to the literature, several limitations exist. Firstly, this study was conducted using the database Orbis. Although comprehensive firm-level information is available on Orbis, it is data from a secondary source. The main problem with using a database like Orbis is that most of the factors are studied with a proxy or indicator. In this study, a dummy variable was used to study subsidiary decision-making autonomy as the sum of business activities of the subsidiaries. Therefore, this was not measured directly as would have been possible when using surveys at companies. Also, the strategic role of the subsidiary was measured in terms of financial value for the MNE, while this factor may be more relevant in terms of knowledge base, resources, and capabilities of the subsidiary (Rabbiosi, 2011; Ambos & Ambos, 2009). Finally, not all companies included in this study were having available data. Hence, there were missing values in the dataset. Using a survey directly at companies may have resulted in a smaller amount of missing values and more direct measures of the constructs.
Secondly, this study had a setting of European industrial subsidiaries of Dutch MNEs. The results may therefore be different when testing in different settings as the Americas, Asia, and so on. Moreover, the study focused on foreign subsidiaries for the Dutch MNEs, therefore excluding domestic subsidiaries. Also, taking only one country for the MNE’s location makes it more difficult to generalize the results to other settings, as MNEs from other countries may be dealing differently in terms of subsidiary decision-making autonomy.
Thirdly, this study measured the data from one point in time. Therefore, this study was not longitudinal, making causation of the results not proven. Only correlation can be concluded with regard to the found factor in relation to subsidiary decision-making autonomy. The choice for not measuring at a longitudinal basis was made because of two reasons. Firstly, not all studied subsidiaries and MNEs had data in Orbis from before 2014, therefore making it difficult to create a representative dataset. Finding this data may have been possible if more time was available for conducting this study, which is the second reason for not doing a longitudinal study.
33 strategic role of the subsidiary and the context in which it operates. However, more research is needed to understand the effects these factors have on subsidiary decision-making autonomy.
Moreover, the role of agency theory should be studied more extensively. The results suggested that the logic of resource dependence theory was not applying in the context of subsidiary decision-making autonomy. Agency theory however, would explain not finding significant results as agents will try to control subsidiaries anyways to improve efficiency, effectiveness, and strategy alignment (Mudambi & Pedersen, 2007). Moreover, the third factor influencing subsidiary decision-making autonomy, the MNE’s control structure (De Jong et al., 2015), can also be a result of agency problems between the headquarters and subsidiary as top manager may try to keep the subsidiaries close and therefore control it extensively due to the different interests of the headquarters and subsidiaries.
Finally, future research should study alternative theories and factors in relation to subsidiary decision-making autonomy. This study found significant effects for the age of the subsidiary and the moderator TMT nationality heterogeneity, which are factors that have not been included in prior studies regarding subsidiary decision-making autonomy before. Therefore, it seems that upper echelons theory and relational theories could be relevant theories in this field of research and should be studied further. Moreover, other theories and factors which this study did not include either could be influencing subsidiary decision-making autonomy. Examples could be transaction cost economics, resource-based view, the earlier mentioned agency theory, and so on.
34 8. REFEENCES
Aguilera, R. V., & Jackson, G. (2003). The cross-national diversity of corporate governance: Dimensions and determinants. Academy of Management Review, 28(3), 447-465. Amason, A. C. (1996). Distinguishing the effects of functional and dysfunctional conflict on strategic decision making: Resolving a paradox for top management teams. Academy of management journal, 39(1), 123-148.
Ambos, B., Asakawa, K., & Ambos, T. C. (2011). A dynamic perspective on subsidiary autonomy. Global Strategy Journal, 1(3‐4), 301-316.
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.
Ambos, T. C., Andersson, U., & Birkinshaw, J. (2010). What are the consequences of initiative-taking in multinational subsidiaries? Journal of International Business Studies, 41(7), 1099-1118.
Andersson, U., Björkman, I., & Forsgren, M. (2005). Managing subsidiary knowledge creation: The effect of control mechanisms on subsidiary local embeddedness. International
Business Review, 14(5), 521-538.
Barkema, H. G., & Shvyrkov, O. (2007). Does top management team diversity promote or hamper foreign expansion?. Strategic Management Journal, 28(7), 663-680.
Baskerville, R. F. (2003). Hofstede never studied culture. Accounting, Organizations
and Society, 28(1), 1-14.
Berman, S. L., Wicks, A. C., Kotha, S., & Jones, T. M. (1999). Does stakeholder orientation matter? The relationship between stakeholder management models and firm financial performance. Academy of Management journal, 42(5), 488-506.
35 Birkinshaw, J., Hood, N., & Jonsson, S. (1998). Building firm-specific advantages in multinational corporations: The role of subsidiary initiative. Strategic Management Journal, 221-241.
Blau, P. M. (1977). Inequality and heterogeneity: A primitive theory of social
structure (Vol. 7). New York; Free Press.
Boeker, W. (1997). Strategic change: The influence of managerial characteristics and organizational growth. Academy of Management Journal, 40(1), 152-170.
Bosse, D. A., & Phillips, R. A. (2016). Agency theory and bounded self-interest. Academy of Management Review, 41(2), 276-297.
Boyd, B. K., & Solarino, A. M. (2016). Ownership of Corporations A Review, Synthesis, and Research Agenda. Journal of Management, 42(5), 1282-1314.
Cantwell, J., & Mudambi, R. (2005). MNE competence‐creating subsidiary mandates. Strategic Management Journal, 26(12), 1109-1128.
Carpenter, M. A. (2002). The implications of strategy and social context for the relationship between top management team heterogeneity and firm performance. Strategic
Management Journal, 23(3), 275–284.
Carpenter, M. A. and Fredrickson, J. W. (2001). Top management teams, global strategic posture, and the moderating role of uncertainty. Academy of Management Journal,
44(3), 533–545.
Casciaro, T., & Piskorski, M. J. (2005). Power imbalance, mutual dependence, and constraint absorption: A closer look at resource dependence theory. Administrative Science
Quarterly, 50(2), 167-199.
Certo, S. T., Lester, R. H., Dalton C. M., & Dalton, D. R. (2006). Top management teams, strategy and financial performance: A meta-analytic examination. Journal of
Management Studies, 43(4), 813-839.
36 De Jong, G., Van Dut, V., Jindra, B., & Marek, P. (2015). Does country context distance determine subsidiary decision-making autonomy? Theory and evidence from European transition economies. International Business Review, 24(5), 874-889.
Díaz-Fernández, M. C., González-Rodriguez, M. R., & Simonetti, B. (2015). Top management teams' demographic characteristics and their influence on strategic change.
Quality & Quantity, 49(3), 1305-1322.
Doz, Y. L., & Prahalad, C. K. (1981). Headquarters influence and strategic control in MNCs. Sloan Management Review, 23(1), 15.
Drees, J. M., & Heugens, P. P. (2013). Synthesizing and extending resource dependence theory: A meta-analysis. Journal of Management, 39(6), 1666-1698.
Duncan, R. B. (1972). Characteristics of organizational environments and perceived environmental uncertainty. Administrative Science Quarterly, 313-327.
Dussauge, P., Garrette, B., & Mitchell, W. (2000). Learning from competing partners: outcomes and durations of scale and link alliances in Europe, North America and Asia. Strategic Management Journal, 21, 99-126.
Eisenhardt, K. M. (1989). Agency theory: An assessment and review. Academy of
Management Review, 14(1), 57-74.
Eisenhardt, K. M., & Schoonhoven, C. B. (1990). Organizational growth: Linking founding team, strategy, environment, and growth among US semiconductor ventures, 1978-1988. Administrative Science Quarterly, 504-529.
Fang, T. (2003). A critique of Hofstede’s fifth national culture dimension. International Journal of Cross Cultural Management, 3(3), 347-368.
Freeman, R. E. (1984). Strategic management: A stakeholder approach. Cambridge University Press.
