Does Diversity in Top Management Teams Influence the Relationship Between Uncertainty Avoidance and Entry Mode Choice: A study of Foreign Direct Investment into the Netherlands

(1)

Does Diversity in Top Management Teams Influence the Relationship

Between Uncertainty Avoidance and Entry Mode Choice:

A study of Foreign Direct Investment into the Netherlands

Elizabeth Celine Nierop S2308002/ B5068796

MSc. Advanced International Business Management and Marketing University of Groningen, The Netherlands

Newcastle University, The United Kingdom

Mr. Drs. H.A. (Henk) Ritsema Dr Saurabh Bhattacharya

May 2017

(2)

Abstract

This paper looked at the role of uncertainty avoidance in entry mode choice of international firms. By examining the role of uncertainty avoidance specifically, it attempted to ascertain whether this cultural factor was leading the understanding that national culture influenced entry mode choice. However, this paper acknowledges previous work in the field, and therefore went further to examine the role of industry in this relationship. In addition, Upper Echelon theory was introduced to the dimension to incorporate the role managerial decisions will have on the strategic choices made. When looking at the management teams, the

research examined the role of uncertainty avoidance within the team, national diversity, gender diversity, and age, in order to provide the most detailed understanding of the entry mode drivers. This provides a paper that examines both the macro relationship between national uncertainty avoidance and entry mode choice, as well as the micro team level relationship, whilst accounting for the diverse factors that have previously been shown to play a role.

It examined 20 companies including data on the 186 managers. The companies all entered the Netherlands and originated from The United States, The United Kingdom, France and

Switzerland. These countries represented some of the highest foreign direct investors in the Netherlands and showed a range of risk avoidance values (measured through Hofstede). The findings showed an insignificant relationship between the national level of uncertainty avoidance value of the firm and its entry mode choice. In order to continue with the study, industry of the firm and the composition of the top management team was studied for its direct impact on entry mode choice. Interestingly, the uncertainty avoidance level of the team showed no direct impact on the entry choice, which this paper theorizes is due to the

increased diversity increases the objectivity of decision making. Gender makeup and percentage of diversity did not play a role, while increasing average age increased then chance of a higher investment mode chosen. The results are hampered by the relatively small sample.

Keywords: Uncertainty Avoidance, Hofstede, Entry Modes, Internationalization, Strategy, The Netherlands, France, The United Kingdom, The United States, Switzerland.

(3)

Acknowledgements

Firstly, I would like to thank my supervisors for their help. Their feedback was valuable in conducting this research. In specific, I would like to thank Mr. Ritsema for his flexibility in this process.

Secondly, I would like to thank my parents for supporting me throughout this degree and allowing me to follow my passion. Your support through the last phase of this process has been invaluable. You are the reason I believe I can follow a career in this field, and push me to be my very best. You inspire me every day.

(4)

1. Introduction

The research and development of theories on how the world works have been around for centuries. Roger Bacon is credited with establishing the scientific method in the thirteenth century (Scientific Method). Thousands, if not millions of papers and theories have since been developed and published. Over the course of time, the world has evolved as has the way in which society operates. Theories developed on management and international business may not still be applicable, or additional factors may play a role in the relationship.

Management is especially vulnerable to these trends as strategic management is decided by individuals, and thus shifts in the strategies and results of management research have been found (Scandura and Williams, 2000).

One of the crucial aspects of international business is the entry mode strategy utilized by firms. With significant time, energy and financial investments needed to ensure the correct decisions are made for secure and sustainable firm growth, the focus on entry mode strategy is always prevailing. From the introduction of the Uppsala Model (Johanson and Vahlne, 1977), to transaction cost theory (Madhok, 1997) to papers as recent as 2015 examining the role government resources and external resources can play (Shaw, 2015), the analysis of entry mode choice is ongoing and as relevant as ever. Referring back to the shift in management research (Scandura and Williams, 2000), these papers are bound to present different perspectives and approaches to management. The changes in papers over time represents an increasing knowledge base and deeper understanding.

The key influential paper by Koghut and Singh (1988) in this field introduced culture as a determining factor for the entry mode choices made by firms. A key learning from their paper that required further research was the idea to examine the singular influence of the cultural dimensions on entry mode choice, rather than culture as a whole. Specifically, the factor with the strongest correlation was Uncertainty Avoidance. This paper will pick up this thread, and delve into the role the cultural dimension of uncertainty avoidance plays in the

internationalisation choices of firms.

(6)

examining the role of uncertainty avoidance, the factor of industry will also be examined to establish potential relationships.

Delving deeper into understanding these firm decisions,

t

his paper aims to add to the changes in management research by providing an additional layer of understanding as to the reasons behind entry mode choices of multinationals. “The choice of mode of entry is essentially a decision made by a manger or a team of managers” (Kumar, 1997). As managers play the crucial decision making roles for foreign entry modes, this paper takes into account the introduction of the Upper Echelon perspective (Hambrick and Mason, 1984). This perspective describes that “organizational outcomes – strategic choices and performance levels – are partially predicted by managerial background characteristics” (pg. 193).

The managerial background characteristics can be synonymized as the managerial mental programs, referenced in the work of Hofstede. His interpretation of social systems

influencing the mental programming of humans, presented the concept of national culture influencing the actions of the individuals of that culture (Hofstede, 1984). Hofstede examined culture through four measures - later extended until 6 (Hofstede.com).

The background (and thus culture) of the managers impacts their actions and therefore the actions of the organizations they lead. Through this relationship, the culture of the Top

Management Team (TMT) is theorized to play a role in the strategic decision making of entry mode choice. Within Hofstede’s classification of culture, the variable of uncertainty

avoidance comes forward with the strongest relationship with entry mode choice (Kogut and Singh, 1988).

(7)

Therefore, the purpose of this study is twofold. Firstly, it will look to examine the specific influence of firm uncertainty avoidance on entry mode choice by looking to see if the theoretical results thus far can be reflected in the actions of real multinationals. Secondly, it will further investigate this by elaborating on the research showing the crucial role top management team composition has directly on the actions of the firm. The strong evidence proposed around Upper Echelon leads to the analysis of the role that diversity in TMTs has in this relationship. In addition, the influence of industry will be examined, to ensure it is not affecting the results.

The research questions tackled in this paper are:

1) To what extent does the national uncertainty avoidance of the firm influence their entry mode choice?

2) To what extent does the moderating variable industry influence the relationship between the firm’s uncertainty avoidance and entry mode?

3) To what extent does the composition of the TMT influence the relationship between the firm’s national uncertainty avoidance and their international entry mode choice?

Through using data of real firms that have entered the Netherlands, it aims to discover the role of uncertainty avoidance and diversity there within, in entry mode choices

The aim of this paper is relevant for businesses operating today. The connection between TMT’s and internationalization of a firm has been studied. Recently, Lin and Chang (2013) released a paper examining the “effects of the compensation level and gap between the CEO and TMT with respect to the rhythm of firm internationalization” (pg. 1380). The findings show that the compensation of CEOs and the size of the gap to the TMT both influence the regularity of foreign expansion. Therefore, it shows a direct connection between the activities of the top management in a firm and their international strategy choices. The findings of their paper can be used to establish appropriate wages and gaps, while the findings of this research will assist the diversity levels required in TMTs. This papers results, whether finding

relationships or not, add to the understanding of firms on which factors they should control for when strategy decisions occur. Understanding the role of firm level uncertainty avoidance versus team level uncertainty avoidance for example, allows real firms to both understand their past decisions and attempt to objectify their future decisions. While a firms nationality cannot be changed, the inclusion of top management teams are a relatively flexible

(8)

2. Literature Review

A thorough analysis of the existing literature will now be performed. This is to better understand the relationships between the variables and the influence they can have on each other. The literature review explains the choice of the variables shown in the conceptual model. It provides the evidence to further understand the research questions and how to study them.

In order to provide a cohesive picture of the elements, and establish connections, this review will start with the general concept of internationalization. From here, gaps and questions in the research will form the conceptual path.

Once the role of internationalization has been defined, the paper will specify its examination onto the importance of entry mode choice in real world business applications. This is

followed by the introduction of the predictor variable of culture. Once the first hypothesis has been outline, the paper shifts to examine two potential moderating variables, which would influence the relationship defined in hypothesis 1.

Firstly, industry is examined for its connection to controlling entry mode choice, without providing a proven explanatory power.

Following this, the paper dives further into the operations of an internationalizing business, to understand that internal factors of the firm exist past its ‘firm’ level. By accepting these into the paper, diversity in top management teams is introduced as the third factor. This factor is chosen due to its people level focus, which can be used alongside the variable of uncertainty avoidance on a firm level, as the culture can be directly compared.

2.1 Internationalization of firms

(9)

These international business strategies consist of a multitude of components and assets. Strategy itself cannot be compared across firms in its entirety in a model other than a case study. Due to this difficulty, the solution is to focus on a subset of strategy. Entry modes the component of strategy that will be focused on in this paper. They are the method through which a firm chooses to organize its business activities abroad (Hill et al., 1990). Specifically, the entry mode is the form through which a company undertakes “direct investment in a foreign country” (Anderson and Svensson, 1994).

The entry mode a firm chooses determines not only the resources needed, but will prescribe the infrastructure that firm uses in the new market, for a significant period. Choosing unsuccessful entry modes, can have a significant impact on the performance of the multinational firm, both in the long and short term (Anderson and Coughlan, 1987). Therefore, these choices are vital for the operations of companies, and additional understanding on the topic will always assist in firms making the most informed decision possible.

The subdivision of entry modes is done under numerous names and varying methods by researchers, divided based on control, resource commitment or risk dissemination (Hill, Hwang and Kim, 1990). For example, Woodcock et al. (1994) divided the entry modes in terms of equity, a popular methodology to rank entry modes.

This concept of varying levels of investment (equity) returns in the (updated) ‘Uppsala internationalization process model’ (Johanson and Vahlne, 2009) which is a key paper in the entry mode research field. The model introduced the series of steps that firms undertake when entering a new market, from export to creating their own subsidiary. These are due to minimizing risks and testing new markets responses to their products and services. This paper raises the fact that the research should be aware that the entry mode a firm states, could be part of a process. Often, exporting can be the initial step firms use to test the new market, after which they can move on to more developed entry modes (Forsgren, 2002). This will need to be included in any methodologies to ensure correct comparison of the choices can be made.

(10)

are publicly available information, where companies report their international expansions in the past. This data is highly reliable and can be used with a high degree of certainty.

The definitions of entry modes used in this research are based on the ‘hierarchical model of market entry modes’ founded by Pan and Tse (2000) who divided entry modes into equity and non-equity. Within these categories, they introduce wholly owned subsidiaries, joint venture, export and contractual agreements. The division into two categories and then subdivisions serves to exaggerate differences between licensing and joint ventures, which could influence the risk perception of the choices..

In this paper we use investment level combined with control as the differentiator on the new ventures (Root, 1987). Control is used to mean “authority over operational and strategic decision-making” (Hill, Hwang and Kim, 1990, pg. 118).

Now this paper can look further into the different forms of entry modes and differentiating between them. At the highest level of investment and ownership is the Greenfield entry mode. In this mode, the firm opens their own subsidiary in the new market, completely self-owned (Pan and Tse, 2000). While this allows freedom, there is limited flexibility and high risk as failure would waste a significant investment. In comparison, exporting allows high flexibility, but limited freedom, operating with the delivery infrastructure present in the host country.

2.2 Role of Culture in Entry Mode Choice

(11)

The three previously established factors to entry mode choice within a market were were control, resource commitment and risk dissemination factors (Hill, Hwang and Kim, 1990). If the entry mode choice of a firm could be solely based on a firm’s preference for level of control, risk minimization or resources it is willing to commit, we would see every firm enter every market in the same way as its previous ventures. In addition, firms that have similar resources in the same industry would enter in the identical mode. However, this is not always the case in real life.

An example of this can be seen in the expansion of Henkel, a German fast moving consumer goods company. With the recent acquisition of Sun products in North America, and the Spotless brands in Western Europe, acquisitions appear to be the preferred format of entry modes for this company (Henkel.com). Using the local advantage to enter a market by purchasing already present brands, Henkel almost eliminates the risk of unsuccessful market entry. However, Henkel growth previous to its 2014 purchase of Spotless was mainly done through exporting followed by setting up local subsidiaries that could localize the products (i.e. the laundry softener ‘Perwoll’, renamed to ‘Fleuril’ in the Netherlands). This change in strategy is not externally linked or country linked, as Henkel uses a mix of strategy at the same time. Other firms that enteed the Netherlands, set up local subsidiaries (i.e.

MacDonald’s), an yet again other firms used Joint Venture such as Liberty and Vodafone (Reuters.com, 2016).

Therefore, there must be additional factors that are influencing the choices made in this large multinational corporation.

A ground-breaking paper that found a significant factor of influence on entry modes is that of Kogut and Singh (1988). While not the first paper to claim its influence, the scale of their research (228 company entries) combined with the empirical evidence to back up their claims means they are the leading paper in showing the influence of national culture on entry mode. Culture is a multidimensional construct that has no agreed upon definition. As described by Hofstede and Hofstede (2005, pg. 3), culture “consists of the unwritten rules of the social game. It is the collective programming of the mind that distinguishes the members of a group or category of people from others”.

(12)

Culture as a concept is multidimensional, and an important concept that affects firms through multiple levels. As stated by Erez and Gati in their 2004 paper, “culture conveys the top-down-bottom-up processes where one cultural level affects changes in other levels of culture” (pg. 583-584). On the one hand, the internationalization process of firms works its way down the chain, promoting attitude changes in cultures, and adaptations. On the other hand, the culture of the of the individuals modifies the macro levels choices of the firm. Based on this, culture presents an interesting topic to delve further into for this paper, as any findings are useful for a wide variety of people.

Significant research has been done around the relationship between national culture and entry mode choice. Barkema et al. (1996) discovered that firms having invested in a country

previously are more likely to be more invested in acquisitions in those countries. The relevant aspect is that these firms followed the same pattern when entering a country from the ‘same cultural bloc’, in which case experience in a similar country would substitute for experience in that country. This points to culture presenting a significant stumbling block to firms. Once a firm has learned how to operate in the new cultural environment, they can operate in that environment again with a higher chance of success. From Barkema et. al. we can learn that the culture of the target market may exert such a strong influence on entry mode choice that it can be compare to the significant influence of legal and strategic factors. However, from this paper, the findings are merely thoughts of the authors and not the direct results from the study and thus not supported by data. These explanations are provided a stronger evidentiary

background when we examine the 1994 paper by Shane which studied the role national culture played in the choice between two entry modes (licensing and foreign direct investment). The industry level data was directly used in this case to examine this by examining the country of origin and investment to understand differences in level of trust (which can be classed within the Uncertainty Avoidance dimension of Hofstede’s

framework). With a finding showing that level of trust did influence the choice of entry mode, this line of thought is more significant and reliable than in the case of Barkema et al. (1996).

(13)

2.2.1 Understanding the Dimensions of Culture

National Culture and entry mode choice cannot be discussed without including one of the leading papers in this field by Kogut and Singh (1988). Their study followed a similar reasoning to this paper thus far, by realizing that another factor was at place when looking at the entry mode choices of international firms. They included uncertainty avoidance and cultural distance to understand the choices made. Focusing on the United States, their paper confirmed that entry mode choice was influenced by national culture of the firms.

Based on this, there is sufficient evidence to the important role culture plays in this strategic decision making. From their positive results, there is a base for this research to look forward into the parts at play.

Importantly, the Kogut and Singh paper developed its own index to study national culture distance. Using the four dimensions of Hofstede at the time, the equation of: 𝐶𝐷_# =

{(𝐼()−+,-))

/

0, }/4

4

(56 is used to measure cultural distance.

As seen from these papers, Hofstede’s cultural dimensions are the leading dimensions used in research thus far, even used in the equations of further developed indexes. But it is important to understand the weaknesses of Hofstede’s dimensions, and the alternatives out there. An acknowledged strength to Hofstede’s work is his ongoing efforts to improve the

dimensions and diversify the variable. His additional dimension of ‘Long-Term Orientation’ to measure the differences between the strategic thinking of individuals in cultures was strongly criticised by Fang (2003) for multiple reasons. Applicable to this research is the criticism about the problematic division between actions that define the two opposite ends of the spectrum (short term or long term orientation). The authors reference the ‘Chinese Yin Yang principle’ to show the overlap between Hofstede’s classifications, and interrelation between the variables. While this variable is not to be used in this paper, it brings an important point to light. The inherent flaw within measuring culture on a scale within each dimension narrows down a wide range range of behaviour into one value. Unique and

varying combinations mean that not each country scoring similar on the scale with act similar when faced with a scenario. This makes understanding the “collective mental programming” (Hofstede, 1983) more complex than the numbers show.

(14)

new way to value culture, apart from that of Hofstede. Collecting data on 62 different

countries, the questionnaire has a significant reach across the world. However, this construct faces similar challenges to that of Hofstede (Graen, 2016). A representative sample is

difficult to obtain (Globe was based on at least 300 managers per nation, Hofstede used IBM managers) to capture the intricacies of national culture. In addition, the sample focused on managers, who were willing to participate in the survey. This is arguably a relatively homogenic subset of the people in each country.

Therefore, while the Globe study may provide us with differing results, Hofstede’s

dimensions provide the clearest route for this paper to take. Its weaknesses are faced by other cultural methods, and its public publication of the results per country make it ideal for this study. In addition, by using the same dimensions as the majority of other research in the field allows for a stronger comparison of results and improves the validity of this research in the understanding of entry mode choice. The strong differentiation by Hofstede between each variable allows for understanding of the specific influence of each dimension of culture. Noted from the previous papers is the use of culture as one variable, a sum of the parts. Therefore, the use of Hofstede’s variables allows for a further deep dive into culture as a multifaceted concept.

A significant conclusion by Kogut and Singh (1988), is that they believe the role of

Uncertainty Avoidance on its own is crucial to understanding the impact of cultural distance. As a leading paper in the field, that this suggestion has not been followed up yet on the macro firm level internationalisation scale presents a significant gap in the research. Based on this finding, this paper will deep dive into uncertainty avoidance, to determine its potential singular role. The findings will add onto the discussion of the role of culture in

internationalization, and provide more tangible results than the combination of all dimensions.

Focusing on the singular role of the factor, is not unprecedented, even within the aspects of culture. A strong paper that defends the reasoning behind focusing on an individual

(15)

In addition, lower uncertainty avoidance has been found to increase entrepreneurial

orientation (Mueller and Thomas, 2000). External uncertainty has been studied in connection to international entry mode choice, such as through that language diversity between the home and host in combination with cultural uncertainty was shown to influence entry mode choice (Lopez-Duarte and Suarez, 2010). Therefore, cultural distance based only on the uncertainty avoidance values will be used in this study, to better understand that specific relationship.

In addition to the strong role uncertainty avoidance plays, it has also been shown situation other than foreign direct investment that the uncertainty avoidance is dictating the channel used by consumers rather than the choice of final decision (Money and Crots, 2001). Applying this relationship to entry mode choice, the uncertainty avoidance will directly influence the choice of entry mode, the decision to enter will remain the same.

An additional reason for the choice to focus on on only one aspect of culture is the limitation of cultural research able to be performed in the scope of this paper. As shown by the ever increasing number of dimensions introduced by Hofstede (2011), cultural research cannot be seen as all-inclusive. By focusing on one specific factors, the results are more reliable and relevant to organizations and future research through their limited scope.

When focusing on the one dimension, in order to perform a high quality analysis, a better understanding of uncertainty avoidance and the known factors that influence it must be considered.

(16)

For example, Bromiley, P (1993) discovered that past performance influenced the risk-taking behaviour of the firm, with poor past performance increasing risk taking behaviour. For this paper, it is therefore assumed that the past performance of firms influenced the perceived risk, which in turn impacted how the firm reacted in future projects, depending on their level of uncertainty avoidance.

When we look at entry mode choices, uncertainty avoidance is already strongly linked to choice in entry mode. A leading theory in entry mode choice is the Uppsala Model (Johanson and Vahlne, 1977). The Uppsala model prescribes gradual entry modes, through steps from the lowest level of investment (exporting) to the highest (Greenfield). This is due to

attempting to minimize risk for the corporations. By gradually increasing investment, the corporations can test the market gradually. The market can then be tested for its level of demand and the reactions of competitors. If it proves to be a market where the risks are too significant for the financial health of a firm, the firm can leave the market with minimal costs.

The model by Johanson and Vahlne is widely respected amongst scholars and seen as a significant paper in the understanding of entry modes. It is ultimately based on risk aversion (synonymous to uncertainty avoidance), representing the key role uncertainty avoidance plays on its own, outside of the concept of culture in general.

Based on the results of Kogut and Singh combined with the papers previously mentioned, our first hypothesis can be developed. With strong results showing higher uncertainty avoidance mean joint ventures will be chosen over wholly owned acquisitions, the effect can be used here.

H1: The countries with lower uncertainty avoidance are more likely to choose wholly owned investments such as Greenfields or Acquisitions.

H0: The level of uncertainty avoidance of the firm does not significantly influence the international entry mode choice.

2.3 Role of Industry

(17)

them (Kemp et al. 1983). Innovation focused industries may therefore be more likely to use a higher investment entry mode where they have the space to innovate.

In addition to innovation focus, the product/ service type will play a role. In standardized companies (i.e. McDonalds), exporting/ licensing minimizes risk and costs. However, these entry modes work less effectively for localized products and services who need the time and knowledge to adapt to the needs of the host market (Muller, 2007). Combining, the role of innovation and type of product/ service, the industry of the firms included in this research should be noted to see if they influence the entry modes of the firms included in this paper.

Fiet (1995) studied the differences in risk taking behaviour between investors approaching different industries, and found that they approached risk differently depending on their returns. The effect has not been included in other studies, so this would provide a small area of interest that can be received in this research. It is of specific interest in this analysis, due to the similarities between uncertainty avoidance and risk taking behaviour (low uncertainty avoidance can be related to high risk taking behaviour). The paper by Hill, Hwang and Kim (1990) argues that the the entry mode choice cannot be seen in isolation and pointed out the role industry environments can play. In addition, Yasai-Ardekani (1986) found that the “objective industry environments influence manager’s perceptions of their environments” (pg.9). The combination of these three papers show that industry is a factor that cannot be ignored in this paper.

Therefore, the second hypothesis follows the results that there will be a pattern presenting through the entry mode choice chosen by firms of different industries.

Trying to determine a trend on the industry type we can discern between manufacturing and service industries. Terpstra and Yu (1988) found no difference in the choices made between manufacturing and service firms, however Erramilli and Rao (1993) found significant differences industry types based on factors such as capital intensity and “separation of

production and consumption” (pg. 19). In an earlier paper titled ‘entry mode choice in service industries’, Erramilli (1990) found that within service industries, the form of service

determined whether the firm used a different entry mode than a manufacturing firm or not.

(18)

connect the concept of uncertainty into the relationship between industry type and entry mode choice.

However, based on the current research, no industry level trend can be determined to a reliable level, so therefore the hypothesis is based on determining if within our data set, where the relationship between uncertainty avoidance and entry modes are studied, whether industry plays a role in this relationship. If so, industry is key for real life applications, as significant differences in the industry type would impact how the managers use the findings in their companies.

H2: The industry in which the investing firms operate will influence their choice of entry mode.

H0: The industry in which the investing firms operate will not significantly influence the choice of entry mode

2.4 Upper Echelon Perspective

When we turn back to one of the key papers used thus far, that of Kogut and Singh (1988), it is interesting to see that even with the strengths of weaknesses of the research determined thus far, there is an additional layer. The following quote represents a key gap in the research that Kogut and Singh had found and believed warranted further understanding:

“But no matter how superior the current multinational corporation may be in replacing the skills of traders by international extension of organizational boundaries, the management of these firms are likely to be influenced by the dominant country culture” (p.429). Even in this 1988 paper, the role of culture within the firm as distinguished from national culture is identified as possibly playing a vital role. Thus far, this review has shown the need

(stemming both from business and scholars) for a focus on the individual role of Uncertainty Avoidance on entry mode choice. This was followed up with the role industry can play in this relationship. Both of these factors are on a macro scale looking at firm behaviour. The final part of this review will delve into the micro factors in order to perform the most insightful and valid research.

(19)

to the global market” and identified external as well as internal factors that will play a role in the choices made by firms, and how those choices are implemented. Externally, the factors consist of market, economic, environmental and competitive factors. The competitive factors can be closely linked to those of Porter and his five forces (1991). The authors theorize that the internal factors of the firm (structure, people, management process and company culture) are influences to the company’s “ability to develop and implement global strategy” (pg. 105). Therefore, in the authors opinions, these additional external factors influence the choice of entry mode as well as control, resource commitment and risk dissemination previously

introduced. However, the authors fail to consider the role the internal firm factors would have in choices themselves, limiting them only to choice implementation. As this paper was not written recently, it seems clear to represent an older school of thought that businesses were not so much influenced by culture and people factors when making the strategic choices. As Upper Echelon perspective was only introduced four years prior to the publication of the paper (Hambrick and Mason, 1984), it is probable the theory had not become widespread knowledge at that point. This paper will fill in the weakness of this paper by trying to accurately measure the direct effect of the people in the firm on its strategic choices (rather than simply implementation).

A study that connected strategy and entry mode choice (David, Desai and Francis, 2000) discovered that the strategy of a business unit does have significant influence on the entry mode choices of that unit. Following this, Harzing (2001) hypothesizes that the strategy of the firm would play an influence, but rather at the overall international strategy at a

corporate-level of the company. Using primary data collected at an international scale, 104 various headquarters were included in the results. This is a significant sample size

considering the criteria for it to be a highly multinational corporation. The results can thus be seen as relatively reliable. The results confirmed the hypothesis, showing high explanatory power of the variable. The validity of the second study represents that choices of firms are influenced at the highest levels of strategy, rather than the individuals wishes of managers in business units.

(20)

members has on the decisions made. The level of influence of individuals on decisions is key to understand the aim of this paper. Elenkov et. al. found that strategic leadership behaviour’s had a positive impact on executive influence on innovation. Therefore, the actions,

behaviours, and attributes of individuals has been shown to have a direct influence on the strategic direction of their firm.

The factors identified by Elenkov et. al. (2005) to influence these employee’s attributes are Strategic Leadership behaviours, Social Culture, TMT Tenure Heterogeneity and

Organizational Size & Personality Factors. The figure below represents the (moderating) roles of the factors.

Figure 1: Elenkov et al. (2005) Conceptual Model

Social Culture is defined as “a system of values, norms, attitudes, rituals, and elements of mental programming that are common for members of a social group” (Hofstede, 2001). The results of Elenkov et. al. looked at the combined effects of firm culture, the behaviour of individuals, and the strategic direction of the firm. These three variables fulfil different roles but are also reflected in this conceptual model. The connection (along with the fourth variable around TMTs) to our study shows that this is a relevant topic to analyze. The significant connection to the variables shows that relationships are present. This research is therefore aiding this to better understand a different side of the connections strategy, culture and TMTs/ individual behaviour have.

(21)

The examination of TMT diversity is important to study as there are mixed results to the research (Ancona and Caldwell, 1992; O’Reilly et al., 1993). Diversity has been shown to positively influence the abilities of TMT teams but also negative influence the abilities (Lawrence, 1997; Hambrick et al., 1996; Miller et al., 1998; Pelled et al., 1999). This means that this research will add to the various understandings of TMT diversity and perhaps provide some clarity.

Before looking further into the influences of top management team decisions, how decisions in teams are made is of interest. This allows for a better development of our hypotheses and understanding of the data. Specifically, how would diversity in these teams actually influence their decisions? An interesting paper by Amason (1996) shows how the conflict created by diversity can have effects on the quality and acceptance of decisions. The findings showed that conflict can ‘improve decision making quality’ (pg. 141).

Understanding the role TMTs can play in the decisions of a firm, theorists then looks at what factors influence this role. The factors found to have an influence on the decisions made by TMTs include demographic composition difference from competitors (Boone et al, n.d.), age and education (Chuand et al. 2009), gender (Bao et al. 2014), tenure, functional background, socio economic roots, and financial position (Hambrick and Mason, 1984). The significant presence and variability in these factors suggests that the likelihood other factors will have an influence is high.

Age is a key variable whose connection to risk has already been established by Pegels and Yang (2000). Older managers prefer financial security and thus are less likely to take risks in their strategic behaviours, while younger managers are more innovative. These ideas are not echoed in the results of Tihanyi et al. (2000) who found older managers more likely to take risks. This makes the results of this study more interesting, in order to see which theory is supported within our data set.

(22)

different papers to study the differences between gender in risk taking behaviour. Level of risk taking has been shown in this paper to be a differentiator between choices of entry mode, therefore the influence of gender to this behaviour of a firm could explain differences. While the paper acknowledges that gender gaps are growing smaller over time, there is a significant increase in risk taking behaviour for males compared to females, especially in intellectual circumstances which strategy choices would fall under. However, the inherent weakness of a meta-analysis is it relies on the validity and reliability of all the papers included within it, dating back to the 1960s. As their findings suggest, these differences have changed over time, so combining papers over a course of 40 years may not provide reliable results. Including gender in this paper may shed a light on its influence.

Therefore, in addition to examining the uncertainty avoidance diversity in the teams, this research will also look to see if patterns can be established between the gender diversity, national diversity and average age of the team

Thus, we can develop the final hypothesis of this research. This will answer our final research question on the influence of team diversity in Top Management Teams on entry mode choice. The results from the paper of Knight et al. (1999) examined the consensus reaching process in diverse management teams. The model presented in their paper “showed both direct and indirect effects of diversity on strategic consensus” (pg.445). Based on their finding that “for the most part, TMT diversity had negative effects on strategic consensus”, the expectation of this paper is that non-consensus will lead to increased discussions within teams. Therefore, the strategic choices of the firms are more heavily discussed and needed to be factually supported.

Hypothesis 3: A higher team diversity will negatively influence the impact of national culture on entry mode choice.

H0: A higher team diversity with have no influence the impact of national culture on entry mode choice.

(23)

H3a: The teams with lower uncertainty avoidance will reduce the likelihood of wholly owned investments being chosen as entry mode choice by their firm

H0: The level of uncertainty avoidance of the team does not significantly influence the international entry mode choice.

H3b: The higher the percentage of national diversity in the team will negatively influence the impact of the firm’s national uncertainty avoidance on entry mode choice. H0: The percentage of national diversity in the team will not significantly influence the the

impact of the firm’s national uncertainty avoidance on entry mode choice.

H3c: The teams with a higher percentage of females will negatively influence the impact of the firm’s national uncertainty avoidance on entry mode choice.

H0: The percentage of females on the team will not significantly influence the the impact of the firm’s national uncertainty avoidance on entry mode choice.

H3d: The teams with a lower average age will negatively influence the impact for the firm’s national uncertainty avoidance on entry mode choice.

H0: The average age on the team will not significantly influence the the impact of the firm’s national uncertainty avoidance on entry mode choice

These hypotheses can be seen represented in the conceptual model below:

Figure 2: Conceptual Model

(24)

moderating role of TMTs. Firms can use this information to specifically examine the diversity of their teams, understand the interplays, and adjust accordingly if needed.

3. Methodology

Following our development of hypothesis, the next step is to outline the exact method used to analyse and research this.

The research onion is a model developed by Saunders et al. (2007) to further outline the research strategy. Applying a positivism philosophy to the research, a deductive quantitative approach will be used. The reason for which is that this research has developed hypotheses and will test the data collected in order to see if the hypotheses are reflected in the results. It follows an experimental descriptive research design. It is a longitudinal study as the data will be used over a longer period of time in order to include the most results as possible.

3.1 Variables

As shown in the conceptual model derived from the literature review, the dependent variable of this study is the entry mode choice of the multinationals. Specifically, this variable will measure the chosen mode of entry into the target market. The five categories the entry modes will be sorted into are: wholly owned, joint venture, contractual agreement, licensing (Pan and Tse, 2000) and exporting. They can be ranked on a level of investment (labelled ownership level in the conceptual model).

When examining this variable, we need to turn back to the paper of Johanson and Vahlne (2009) to understand that entering a market is often a series of steps, and not a one-time decision. If the data leads to companies that have indeed followed a process approach to entry modes, the most resource intensive (thus often their latest approach) will be used.

To clarify, if a firm enters the Netherlands through exporting, before graduating to a joint venture with a local firm. The entry mode used in this research will be that of the joint venture. However, this may skew our results, so these companies will be marked, and if needed, removed as outliers so as not to skew our results. This is because experience in the market will inherently affect the uncertainty avoidance role we are looking at.

(25)

The first independent variable of this paper is the level of uncertainty avoidance of the firm, based on the national level in the home country of the firm. Hofstede’s value will be used for this and is viewed as reliable to its extensive use across previous studies (Kogut and Singh, 1988).

However, operating in any country involves interacting with individuals or corporations from that country, to some degree (Dox et al, 1981).. The value used in this research therefore is the difference in uncertainty avoidance. This measure is the difference between the home country uncertainty avoidance value and the local country uncertainty avoidance value. A positive value shows that the uncertainty avoidance of the host country is higher than that of the home country. A negative value shows the uncertainty avoidance in the host country is lower than that of the home country. This variable is classified as an interval variable.

The first co-variable (moderator variable) is industry, upon which our second hypothesis is based. Due to the high variety in the industry each company operates in, this category has been broadly divided into three categories: Service, Physical product, and other. The division of the companies into each category can be seen in Appendix B. This is included in the appendix for those wishing to repeat the study, but the specific firms are not needed to understand this research. Due to the categorization, this variable is a nominal variable. The use of physical product and other allows for further division than the previous papers which divided companies into simply service and manufacturing firms.

(26)

The second variable within the influence of the top management team is the level of diversity. As described in the literature review, the diversity can influence discussions and thus influence the choices made (Amason, 1996). As the focus of this paper is on the cultural value of uncertainty avoidance, it is incorporated in this paper through measuring diversity of nationalities. The variable is therefore represented by the percentage of the team whose nationality is not the same as the home country of the firm. For example, in the case of a firm from the United States, all members who are not American would therefore count towards the diversity percentage. This is classified as an interval variable.

The final two variable included in our third level of hypotheses are the average age and gender composition of the team. As these two factors were shown to influence the choices of the TMT’s, they are included here. The average age is calculated through the average age of the board members at the time of the annual report. The gender composition is measured through the percentage of the team that is female (and thus not male). Both variables can be classified as interval variables.

A significant weakness of this research is the inability collected date on all of the potential variables that could influence uncertainty avoidance as presented in the literature review. For example, accurately measuring for past performance cannot be included, as this would significantly reduce the sample of data included in this paper, and the measurement would be subjectively influenced by the researcher’s opinions on what ‘poor’ performance constitutes of (Bromiley, 1993).

In addition, while every effort was made to collect data on a 5th_{team diversity variable}

(executive experience), it cannot be included in the research. This is due to the lack of clarity around what this variable includes. As there is no set career path or steps that every executive has followed to reach their position, the data collected on this variable is highly subjective and due to the personal opinion of the researcher about the point in time where executive experience begins. The variable can therefore not be objectively collected within the realm of this study.

(27)

3.2 Sample

The Netherlands was chosen as the base country for this study as it experiences an exceptionally high level of foreign direct investment, especially in relation to its size.

Historically, the Netherlands was an important player in world trade due to its ports and trade roots. This trend has stayed until present. Taking into account only trade following the financial crisis (end 2008-end 2015), the Netherlands ranked 14th of the world economies with level of investments with a value of 561,400 USD (CIA Factbook, 2015). As of 2012, the total value of Dutch FDI assets was almost 4,580 billion USD. As only the 67th most populous country in the world, this is a stark difference.

The current reason attributed to this high level of investment is the lower tax planning levels, as the Netherlands hold a ‘bilateral tax treaty’ with many countries, lowering withholding taxes on treaty payments (TaxJustice.net).

In regards to the uncertainty avoidance value used in this study, this number is therefore to be based off of the difference with the Netherlands. As the Netherlands scores almost neutral on uncertainty avoidance at a value of 53 (Hofstede, 2010), it is an ideal market to enter.

However, its slight amount of risk aversion must be accounted for, and therefore the difference in Uncertainty Avoidance is used rather than Uncertainty Avoidance itself. In order to be able to measure the impact of the firm’s uncertainty avoidance level on entry mode choice, investments stemming from multiple countries must be chosen.

The specific influence of the Netherlands on entering firms can be found when looking into their uncertainty avoidance influences. The Netherland have a slight preference for rules, with precision and punctuality, and a small resistance to innovation (geert-hofstede.com, 2017). These attributes are clearly positioned to play a role in contract negotiations and social interactions. These interactions are therefore what are being accounted for by calculating the difference.

(28)

Using data from 2012, the following table shows the 7 top investing countries into the Netherlands. This table was chosen as it includes intra-European Union trade at a country level.

Main Investing Countries 2012 (%)

United states 13.4 Luxembourg 12.4 UK 10.5 Germany 10.4 Belgium 9.7 France 9.3 Switzerland 6.9

Table 1: Investing Countries in the Netherlands (Santander Trade)

Relevant from the table, we can see that a country with the population size of Luxembourg (584,953 in 2016) ranks above the United Kingdom (65,280,2014 in 2016) in level of investments in 2012 (Countrymeters.info, 2016). While the geographical proximity and cultural similarity of Luxembourg may play a role, this data is only based on a snapshot of time. Therefore, this should be not seen as a ranking of importance or overall value over time. In addition, Luxembourg is often used a tax haven for large companies, so the data based in this country will be scrutinized to determine its level of inclusion (TaxJustice.net). The same principles will apply to Switzerland, another tax saving country attractive to international firms.

Given the limited time and significant time cost of manual data collection per firm, the data is limited to only collecting data from 4 countries. This is sufficient as the variety between the countries can be captured by including more than 2 countries.

The choice for the four countries included is based on the seven top countries discussed previously. Excluded are Luxembourg, Germany and Belgium.

(29)

plays in entry mode choice. An additional reason to exclude Luxembourg was the high degree of firms registered there for tax purposes. To ensure no incorrect firm is included, it was removed.

Within the 4 countries remaining (The United States, The United Kingdom, France, & Switzerland), 5 firms per country are used. Given the size of management teams within companies and individual data collection needed per person, this allows for enough detailed data to be collected per firm whilst retaining enough firms per country.

However, while this study is collecting enough data to hopefully find trends, it cannot follow to scope of larger research papers such as 300 firms per country used in the Kogut and Singh (1988) research. But this paper will still contribute to this field by showing its study whether a trend starts to emerge or not. This will signal to further researchers if there is an influence in this sector that needs to be further studied.

As this research is using publicly available information about the companies and its

management, no ethical permissions will be crossed. However, to ensure ethicality is met in all opinions, the raw data set containing management names as listed on annual reports will be destroyed following the research. The management teams will rather be identified through the statistics such as size, gender breakdown and nationality breakdown.

3.3 Data Collection

The firms included must have been launched in the countries they are registered, thus avoiding the firms who have moved their headquarters for tax exemption reasons.

The Orbis database will be tool used for this. The database contains firm level information on over 200 million companies worldwide (Orbis, 2016). The database can be filtered to finding a list of companies who are registered in our Sample countries, that subsequently have performed business activities in the Netherlands. The database cannot differentiate between firms originating in a country or simply partly registered there, so further manual data

(30)

‘Vodafone Investments Luxembourg SARL’ as the third largest firm (2014 operating turnover: 7,417,969 USD) who has entered the Netherlands. Therefore, these adjusted number of raw companies do not accurately represent the data. While this is a weakness, the case companies used in this study will still provide accurate data through manually ensuring of not including incorrectly registered companies.

Based on the sample, the company website is used to use their self-reported historical data to ascertain entry mode used. If the company does not report it, and no trustworthy external source can be found, the company could not be used in the analysis.

Once the entry mode is chosen, the annual reports (AnnualReport.com, 2016) and Bloomberg (Bloomberg.com, 2016) will be used to determine date entered and the management team. Key to this analysis is the use of team data from the annual report of the year prior to the foreign market (the Netherlands) entry. year before entering the Netherlands (t-1). Strategic decisions will be made a minimum of one year prior to actual reported entry so this team is the one that will have had influence over decisions (if any).

The nationality, gender and age of the board members will also be collected from these two sources, in addition to Linked In (LinkedIn.com, 2016). The nationality is self-reported, in the case of multiple nationalities, an average will be taken.

A weakness of using their self-reported nationality and a large limitation of this paper is the assumption that their individual culture is reflected in their national culture. This does not account for the variability in culture or the other factors such as international experience that can change the individual’s cultural values.

(31)

The data collected consists of data on the 54 executives in the United States firms, the 48 executives in the Swiss Firms, the 44 executives within the British firms, and the 43 executives within the French firms. Therefore, the sample consists of 20 firms over 4 countries, and 189 executives.

3.4 Data Analysis Methodology

In order to analyse the data collected, the variables need to be fully defined in a way that optimizes the analytical opportunities.

Entry modes are the consistent dependent variable, and require a further explanation of the process of ranking them. It has been decided to use the following five entry modes in the following order: 1. Exporting 2. Licensing 3. Acquisition 4. Joint Venture 5. New Venture

The choice of these five modes is based on the research done by Woodcock et al. (1994) and Johansson and Vahlne (2009). Woodcock et al. ranked entry modes (levels 2-5) based on level of ownership, and the paper by Johansson and Vahlne was used to classify exporting as the lowest level of entry mode (1.). The connection between level of ownership and

(32)

This ranking does not represent an even difference between the entry modes. The levels of risk are not even split between the entry modes. Therefore, this ranking cannot be used to calculate averages for example.

In order to analyse the data, the statistical program of SPSS is used. The data will be analysed using a Spearman’s Rank Test (Hypothesis 1), Kruskal-Wallis Test (Hypothesis 2) and for the final Hypotheses we will test for multicollinearity between the variables and then perform Spearman Rank Tests if no multicollinearity is found.

The hypotheses that are being tested were outline in the literature review and are repeated in the following chapter.

4 Findings

Now that the data has been collected, the variables can be calculated. This data can be found in Appendix C. With every variable, causality cannot be determined directly through data analysis. This is because data is collected at a singular point of time, and not in a longitudinal fashion. However, due to the form of the data (such as the team diversity data collected in the years previous to internationalisation), causality can be deduced to a point, which will be included in the discussion session. During this analysis, the focus remains on determining the significance of a relationship between variables.

4.1 The influence of a firm’s national uncertainty avoidance on entry mode choice.

The first hypothesis was:

H1: The countries with lower uncertainty avoidance are more likely to choose wholly owned investments such as Greenfields or Acquisitions.

This hypothesis requires an analysis to be carried out to determine if a relationship exists between the dependent variable of entry mode choice, and independent variable of

(33)

In order to have a numerical value for national uncertainty avoidance, prior to analysis, all uncertainty avoidance scores were calculated and transferred to difference scores. This was calculated through determining the difference in value between the national uncertainty avoidance score of each firm’s home country, and the score of the Netherlands. For example, the United States firms (i.e. Netflix) have a home level of Uncertainty Avoidance of 46. The Netherlands value is 53, therefore the score used for analysing is the positive difference of 7. The context if the entry mode is key (and therefore the difference). Especially in the case of the United States with a relatively ‘average’ level of uncertainty avoidance, the

environmental context will play a key role in their actions, which is why the influence of the Dutch environment must be established.

In order to test the relationship between the variables, a method is to discover if the the means of the two variables are statistically significant. However, our dependent variable is not a continuous variable, so the mean of the variable holds no value. The average entry mode choice holds limited value, if the extremes are chosen. As the difference between the various modes is not equal, the average numerical value does not accurately describe the average equity investment into entry mode choice. Therefore, due to the nature of the variables, it does not meet the assumptions to conduct an Anova test.

The best fitting test is the Spearmann rank test as it allows this research to measures the association between the two variables in a way that does not use incorrect average information (Laerd.com, 2016).

In order to conduct the rank test, both variables need to be a ranked variable. Entry mode is already once, so the transformation must occur in our uncertainty avoidance difference value into an ordinal (ranked) variable. As this variable is based on the national country that is founded, there are only four levels of uncertainty avoidance. These are ranked from 1 to 4, with 1 representing the highest uncertainty avoidance (France), and 4 representing the lowest uncertainty avoidance (United Kingdom). This can be seen in the table below:

Firm Label UA Difference Ranking of UA

Difference

Ranking Entry Mode

United States Firm 1 7 3 1

(34)

Switzerland Firm 1 -5 2 3

United Kingdom Firm 1 18 4 3

France Firm 1 -33 1 3

Table 2: Translation of Uncertainty Avoidance Difference to ranking

A level 4 ranking of the independent variable means that the United Kingdom has the strongest positive difference with the Uncertainty avoidance of the Netherlands, aka the lowest uncertainty avoidance of the four sample countries. A level 1 ranking represents that France has the strongest negative difference with the uncertainty avoidance of the

Netherlands, so is the most uncertainty adverse.

Unfortunately, this methodology does mean that both variables are treated as ordinal. Through doing this, we lose the specifics of the data we collected, as the size of differences between the uncertainty avoidance is not captured. Therefore, the fact that the difference between our two least risk averse countries is 28, whereas between our most risk averse is only 11 is lost. However, this is a limitation of the data collected. It does allow us to perform a test which uses the individual data into multiple grouped variables, without needing to create binomial variables that would lose the detailed trend data (for example creating binomial entry mode groupings of service and manufacturing firms).

(35)

relationship cannot be identified within our sample between the national level of uncertainty avoidance and the entry mode choice. Uncertainty avoidance is not strong enough in this case to explain the choice of entry mode.

Correlations Correlation Coefficient Significance (2-tailed) N Ranking Difference Uncertainty Avoidance -0.93 0.698 20

Table 3: Results Spearman’s Rank Test (SPSS). Translation of Uncertainty Avoidance Difference to ranking

Due to the lack of significant relationship found here, the remaining data will have to be applied directly on entry mode choice. As the dependent variable is ordinal, an Anova including coefficients is not possible. With the finding of no relationship within this data set, the influence of the remaining variables is tested instead directly on entry mode choice. This allows the research to still further examine the role these variables play in the entry mode choices of firms. Therefore, the remaining data is still relevant and valid.

4.2 The influence of industry on entry mode choice

The second hypothesis looked to see if the industry of the firm influenced their choice of entry mode:

H2: The industry in which the investing firms operate will influence their choice of entry

mode.

Given the lack of correlation found in the first hypothesis, this theory may provide an explanation to understand the entry mode choice of the firms included in this analysis. To still test the role of industry, it will be tested for its direct correlation to entry the entry mode choices. The papers included in the literature review tested based on the binomial industry variables of service or manufacturing industry. Our industries are ordered into three groups, and the analysis includes industries from multiple countries, which provides a new

(36)

country, so this research would potentially provide new insights on a more international scale.

As the order of the three industry categories provide no information to their value, and cannot be seen as ranking, meaning the industry is treated as a categorical variable. Therefore, we are establishing if there is influence between the categories of industry, and the categories of entry mode choice.

The Kruskal-Wallis test can be used to determine if the entry mode choice differed based on industry category of the firms. This test allows a categorical and an ordinal variable to compared and accurately reflect the data.

The assumption of the test is regarding the independence of observations (Laerd.com, 2016). This is incorporated in our study design by ensuring none of the companies included are owned by the same parent organizations, or interconnected through joint venture

internationally. No company has been assigned to multiple entry modes or industries, so they can be considered independent. Our data therefore meets this assumption.

The other factor to consider before looking at the test relates to the the interpretation of the results, with the distribution variability shape influencing the meaning that can be recognized behind the results. In order to determine the distribution of each group, we can run an

analysis. The Shapiro Wilk test of normality is relevant to this data set as the sets consist of less than 2000 parts.

What we can see from the results in the following table, is that the the significance for our first two industry segments (service & physical product) the significant is below 0.05, showing a non-normal distribution for these industries. While the distribution is normal within the ‘other’ industry group, the overall results show that our data is not normally distributed. Therefore, the test can only be used to compare the mean ranks of our data and the medians cannot be referenced (SPSS-Tutorials, 2016).

Normality Tests: Shapiro-Wilk

Industry Statistic Degrees of Freedom Significance

Service 0.600 7 0.000

Physical Product 0.810 8 0.037

(37)

Kruskal Wallis Statistics

Chi-Square Degrees of Freedom Asymp. Significance

Influence of Industry 1.608 2 0.448

Table 4: Influence of Industry on Entry Mode Choice

The results of the Kruskal-Wallis test show that the entry mode choice does not show any effect from the industry in which the firms operate, 𝑥8 _{2 = 1.608, 𝑝 = 0.448 (larger than} 0.05 threshold for a significant finding). Therefore, our second null hypothesis cannot be rejected. This represents a lack of relationship between the industry type and entry mode choice. The impact of this will be looked at in the discussion.

4.3 Role of Top Management Teams in Entry Mode Choice A significant portion of focus on this paper looked at our third hypothesis:

Hypothesis 3: A higher team diversity will negatively influence the impact of national culture on entry mode choice.

Within this hypothesis, this paper will examine the impact of the average national uncertainty, diversity in nationality, gender diversity and age diversity within the teams. Importantly, this hypothesis looked at the influence of top management teams on the relationship between national uncertainty avoidance of the firm and entry mode choice. While this paper has shown that within our data set there is no relationship between the two variables on a macro level, this hypothesis can still be tested on an individual/ micro level. However, it will be tested to determine if there is a relationship between each variable (i.e. the team level of uncertainty avoidance) and the entry mode choice. If a relationship is found, this would illustrate the importance of incorporating team composition into research, as without it no direct relationship could be understood.

(38)

factors to determine if the rankings show a relationship between the variables which allows us to partly answer our hypotheses about the influence of each team factor on entry choice.

However, first, we need to examine if our four factors of the top management team are interrelated (multi collinear) which could not only influence the validity of the results but would also provide an insight into how the composition of teams works. Interrelated data would not be covered in the aim of this papers research, but the insights could inspire other research that focuses on the composition of top management teams and the causes for it. From the multicollinearity coefficients test results below, we can see that all collinearity statistics (VIF) are below 3 so we do not have an issue of multicollinearity. For example, the average age of the board is not related to the ratio of the genders.

Average Age Percentage of Females Team Uncertainty Avoidance Percentage of national diversity Average Age - 1.030 1.010 1.037 Percentage of Females 1.006 - 1.015 1.010 Team Uncertainty Avoidance Difference (to the Netherlands).

1.005 1.034 - 1.029

Percentage of national diversity

1.011 1.007 1.007 -

Table 5: Checking for Multicollinearity (VIF)

(39)

Rank Uncertainty Avoidance Team Rank Percentage of Females Rank Average Age Rank Percentage of National Diversity France 1.00 7.00 8.00 1.00 3.00 1.00 1.00 11.00 5.00 3.00 12.00 12.00 1.00 1.00 10.00 1.00 2.00 1.00 13.00 5.00 Switzerland 8.00 4.00 12.00 13.00 8.00 2.00 2.00 6.00 6.00 1.00 5.00 10.00 4.00 1.00 11.00 15.00 7.00 1.00 6.00 7.00 United States 12.00 6.00 6.00 1.00 10.00 1.00 3.00 2.00 13.00 4.00 4.00 8.00 11.00 8.00 9.00 5.00 12.00 1.00 3.00 1.00 United Kingdom 9.00 1.00 5.00 14.00 13.00 1.00 7.00 9.00 14.00 1.00 14.00 4.00 11.00 5.00 15.00 9.00 15.00 1.00 15.00 3.00

Table 6: Transformation Independent Variables Management Team Composition

The Spearman Rank Test can now be conducted. Based on the correlations presented, we can only reject one null hypothesis.

Rank Uncertainty Avoidance Team Rank Percentage of Females Rank Average Age Rank Percentage National Diversity Correlation Coefficient -0.171 -0.319 0.456 -0.160 Significance (2-tailed) 0.472 0.170 0.043 0.501

Does Diversity in Top Management Teams Influence the Relationship Between Uncertainty Avoidance and Entry Mode Choice: A study of Foreign Direct Investment into the Netherlands