GOVERNANCE AND INTER NATIONALIZATION: THEORY AND EVIDENCE FROM GERMANY

GOVERNANCE AND INTER NATIONALIZATION:

THEORY AND EVIDENCE FROM GERMANY

A Master Thesis Submitted in Partial Fulfillment of the Requirements for the

Degree of Master of Science (M.Sc.)

in International Business & Management

at the University of Groningen,

Faculty of Economics and Business

Groningen, 12/06/2015

Student: Maksymilian Rojek

Abstract

iii

Content

List of Figures ... iv

List of Tables ... iv

1. Introduction ... 5

2. Literature Review and Hypotheses... 7

2.1 Internationalization ... 7

2.2 Upper Echelon Theory ... 9

2.3 Conceptual Model ... 12 3. Research Methods ... 13 3.1 Sample ... 13 3.2 Dependent Variable ... 13 3.3 Independent Variable ... 14 3.4 Control Variables ... 14 3.5 Method Assumptions ... 16 4. Empirical Results ... 21 4.1 Descriptive Statistics ... 21 4.2 Regression Results ... 22 4.1 Discussion of findings ... 25 5. Robustness Test ... 26 6. Conclusion ... 27

6.1 Added value of this study ... 27

6.2 Limitations ... 28

7. Appendix ... 29

iv

List of Figures

Figure 1: Conceptual Model ... 12

Figure 2: Scatterplot Internationalization ... 17

Figure 3: Histogram for Normality ... 19

Figure 4: Normal Probability Plot ... 20

List of Tables Table 1: Overview of Variables and Measures ... 16

Table 2: Test for Multicollinearity ... 19

Table 3: Descriptive Statistics ... 22

Table 4:Regression Results of the Impact of Board Age Diversity on Internationalization ... 24

Table 5:Robustness Test: Internationalization Ratio ... 29

Table 6:List of Companies in the Sample ... 30

5

1. Introduction

The debate on a board’s strategic contribution dates back to the decade of the 1970s (Pugliese et al, 2009). First, it was observed that boards were rather passive in the strategic involvement which often led to corporate failures and that more involvement was needed in order to restore public confidence (Clendin, 1972; Mace 1967; Vance, 1979). Later, several scholars argued that the board was in excellent position to contribute to strategy (Carpenter and Westphal, 2001). It was found that boards are becoming more active in decision making and are an important part of strategic planning (Schmidt and Bauer, 2006). Next to this, the board seemed to affect different important elements of strategy, like for example internationalization (Sanders and Carpenter, 1998; Datta, Rajagopalan and Zhang, 2003). Eventually, strategy was viewed as flow from information and decision which was detached from every person involved (Mintzberg, Raisinghani, & Théorêt, 1976). The outcomes of an organization were viewed as “reflections of the values and cognitive bases of powerful actors in the organization (Hambrick and Mason, 1984).

Also the situation within organizations in the developed countries is not the same as some decades ago. Today, many developed countries experience lower birth rates, increased prosperity, and improvements of health systems which on the one hand lead to a shrinkage in the labor pool of young people and on the other hand increase the pool of older workers (Sluiter, 2006, Tempest et al., 2002). Consequently, organizations employ workers from both extremes of the age groups (young and old), each having different experiences, skills and values. It is therefore more important than ever to understand how these different groups of people work together in combination to reach the best possible outcomes. Due to the possibility of ages varying by more than 25 years within these groups (Wegge et al., 2008), it is imperative to understand which consequences these mixtures of different age cohorts will have be it of rewarding nature or risks. This research will therefore analyze the effects of this phenomenon called age diversity.

6

which in the end will have the final effect on the internationalization decisions. In other words, the relationship between board age diversity and internationalization will be moderated by CEO tenure. For this research a sample of around 100 German top companies will be tested. Therefore, the title of this research is the following: Therefore, the title of this research is the following:

Governance and Internationalization: Theory and Evidence from Germany.

Although much research on board diversity and firm performance has already been done, the analysis of the moderating effect of the CEO tenure which could influence the final degree of internationalization is something new. One contribution of this research will be that it extends the upper echelon theory by including the board and the CEO as critical unit of analysis. Second, it contributes to the internationalization and diversity literature by analyzing the relationship of board age on internationalization, while being moderated by the tenure of the CEO and therefore adding new knowledge to the precedents of internationalization of firms which leads to a broader view of the strategic influence of the board and all parties which are involved in the strategic decision making process.

So far, very few studies relate board composition to a firm’s internationalization process. This research tries to fill this gap by giving firms an idea on the various effects of the board age diversity and the CEO tenure on internationalization. Eventually, understanding this antecedent could lead to firms understanding what needs to be done to be successful in an international environment, be it through the choice of the right board members, the CEO, or the right mix of both.

The following parts will answer these sub-questions which will be supportive in answering the main research question.

1. What is internationalization and how can this be measured? 2. What is board age diversity and how can this be measured?

3. What is the relation between board age diversity and internationalization?

4. What is the moderator effect of the CEO tenure on the relationship between board age diversity and internationalization?

7

2. Literature Review and Hypotheses 2.1 Internationalization

Before proceeding with the statistical tests, I will now first give a general overview of the topics forming the basis for understanding the research at hand: Internationalization and the Upper Echelon theory, which relates to board age diversity and its effects

Internationalization remains one of the key features for firms in our modern world economy. It has changed the nature of strategy, competition, and competitive advantage (Bartlett & Ghoshal, 1989; Melin, 1992; Porter, 1986; Prahald & Hamel, 1994). Since advances in technology and communications brought countries closer, many companies are now increasingly operating abroad. Competitive pressures from multiple sides force many firms to operate internationally, calling for many complex managerial decisions based on the coordination and integration of a firm’s geographically dispersed resources or the need to cope with different cultures and institutions (Prahald & Hamel, 1994; Gomez Mejia & Palich, 1997). Some firms are very successful in going abroad whereas others fail. Some are present in a large number of countries and others decide just to operate in a couple of countries. The aim of this research is to understand one of the antecedents of this different behaviors.

8

among various countries make it possible to shift the sales from low-income to high income countries. In addition, operating in more than just one country makes a firm less vulnerable to country specific shocks which has risk reducing effect (Hitt et al., 1997).

On the other hand, internationalization also has negative aspects. The engagement in international operations increases the costs which, amongst others, are for example associated with product diversification and coordination difficulties (Bobilo, Lopez-Iturriga & Tejerina-Gaite, 2010). When the degree of internationalization reaches a certain critical point, it might be that the costs will outweigh the positive effects of the international operations (Hitt et al., 1997).

There are several prominent models internationalization is associated with: For instance, there are The Uppsala Internationalization model by Johanson and Vahlne (1977), and the Innovation-Related Internationalization Models. The Uppsala model described internationalization as a process, in which the firm follows a sequence of steps starting from initial exports to the creation of final production units (Melin, 1992). Lack of knowledge of foreign markets and operations are the drivers of these small steps, in which the firms aim to acquire these skills and capabilities through successive learning as well as increased commitment. Four stages were described: No regular export activities -> Export via independent representatives (agents) -> Establishment of an overseas sales subsidiary -> Overseas production/manufacturing units (Johanson & Wiedersheim-Paul, 1975). First, firms enter countries with low psychic distance, defined as “the sum of factors preventing the flow of information from and to the market, including factors such as differences in language, culture, political systems, level of education, or level of industrial development” (Johanson & Vahlne, 1977). They associate these countries with the lowest uncertainty and rather easy to understand. As these countries are often neighboring countries, the firms then gradually proceed to countries with greater psychic distance. The authors also formulated a dynamic model, where state and change variables of internationalization are considered. The assumption is that the outcome of one cycle of events creates the input for the next cycle (Andersen, 1993).

9

2.2 Upper Echelon Theory

As companies are increasingly operating in a global environment, it becomes more and more important to understand how diversity within boards and other groups can affect strategic outcomes (Milliken & Martins, 1996). The results of research about heterogeneity in groups suggests that diversity appears to be a “double-edged sword” (Milliken & Martins, 1996), on the one hand increasing factors like creativity but on the other hand also enabling dissatisfaction of members who cannot identify with the group and thus jeopardize effective group decisions. There are several types of diversity. A common distinction used is diversity on observable or readily detectable attributes such as gender, ethnic background or age, and diversity on less visible or underlying attributes, such as functional background, tenure, etc. (Milliken & Martins, 1996). Another distinction is the one between surface- level diversity or differences in biological characteristics that are reflected in physical features and deep-level diversity, which are differences in attitudes, beliefs and values (Harrison, Price & Bell, 1998). Many types of diversity have so far been used in research. Amongst others, Nishii, Gotte & Raver (2007) for example list job-related diversity and its outcome on internationalization done by Lee & Park (2006), diversity in age and tenure and its impact on organizational innovation (Bantel & Jackson, 1989), or gender diversity and organizational culture affecting the growth orientation and performance of a company (Dwyer, Richard & Chadwick, 2003).

This research will focus on demographic characteristics of the board. More in particular, board age diversity and its effect on internationalization will be analyzed. By taking this approach, some psychological issues are bypassed “in favor of an emphasis on broad tendencies that, if empirically confirmed, can be later held up to the psychologist’s finer lens (Hambrick & Mason, 1984). According to Hambrick, (2007), demographic characteristics can be used as valid proxies of executives’ cognitive frames, although it could be incomplete and imprecise on some elements. Nonetheless, there is evidence that demographic profiles are highly related to strategy.

10

capabilities and personalities, and reflect different degrees of awareness and ambition. These differences affect their organization’s behavior and performance” (van Witteloostuijn & de Jong, 2009).

Every decision is based on the person’s cognitive base and reflect his or her own values. The cognitive base includes knowledge or assumptions about future events, alternatives and consequences of these alternatives, whereas values are principles for ordering consequences or alternatives based on preferences (Hambrick & Mason, 1984).

General research on the influence of demographic variables has found a link between demographic characteristics and specific beliefs, values, and abilities. More specifically, the works used demography as a predictor of beliefs and values (Kahalas & Groves, 1979) and viewpoints (Dearborn & Simon, 1958).

Hambrick and Snow (1977) introduced a sequential view of the perceptual process of a decision maker which can be an individual manager or a group of managers. A personal cognitive base and values are brought to the decision. The first part of the decision process is the limited field of vision, which is related to the fact that the manager or the group of people only direct their attention to a particular area. This field of vision is further limited in the next part, where each individual selectively only perceives a part of the field of vision. Then, the remaining information is interpreted through a filter shaped by ones individual cognitive base and values. The outcome is the basis of a strategic choice.

Age, in particular, is expected to influence strategic decision making perspectives and choices (It was found that flexibility decreases and resistance to change increased when people got older (Carlsson & Karlsson, 1970; Vroom & Pahl, 1971). Furthermore, Carlsson and Karlsson (1970), found a relation between age and rigidity. Therefore, older people seem to be more resistant in taking risks, whereas younger people or managers tend to be more risk oriented. Besides, older managers are more willing to change their views if they are aware of negative consequences (Taylor, 1975). As an effect they might lack the necessary conviction to drive a strategic change of a company. Age can also be called ‘tenure in life’ (Hambrick and Fukotomi, 1991), and is often referred to as a very important feature because it determines a person’s background and experience which is outside the organization.

11

described as an innovative and complex process, involving one’s willingness to take risks. Taken this together with the information above, a first logical assumption would be that a young board, which consists of many people with high risk taking propensities, will lead to higher degrees of internationalization. However, many authors argue that that risk taking propensity is not enough. For instance, Rivas (2012), states that this risk taking propensity needs to be counterbalanced with things like with experience, wisdom and the resources that older members usually have. Only this will finally benefit a process as complex as internationalization. This is in accordance with the findings of Gellner and Veen (2013) who expect growing benefits with increasing age diversity as well as these benefits to be more pronounced in companies that engage in innovative activities and creative tasks, to which internationalization can be counted as previously mentioned. Furthermore, the different attributes of young and old directors are said to complement each other and these differences can be leveraged by the organizations to improve their strategic decision making (Ali, Lu Ng & Kulik, 2014). Finally, by expanding the diversity of the board, the aggregated human and social capital can be maximized, which also is underlined by Carter et al. (2010,p.398), who stated that “diversity holds the potential to improve the information provided by the board to managers due to the unique information held by diverse directors.” Taking everything together, this leads me to my first hypothesis, which is:

H1) There is a positive relationship between board age diversity and a firm’s degree of

internationalization.

12

less likely to initiate changes in their organizations and added that “the longer the tenure of an individual, the more rigid his or her cognitive structures and the less likely he or she is to promote or champion change” (Finkelstein & Hambrick, 1996).

Thus, it is hypothesized that the CEO tenure will have a negative impact on the relationship between board age diversity and internationalization of a firm.

H2) CEO tenure has a negative moderating effect on the relationship between board age

diversity and internationalization.

2.3 Conceptual Model

Figure 1 shows my conceptual model consisting of the relationship between the dependent variable “Internationalization of the Firm” and the independent variable “Board Age Diversity”, and the moderator effect of the variable “CEO tenure”. Each effect stands for one of my hypotheses, which will be tested in the following sections.

Board Age Diversity Internationalization of the Firm

CEO Tenure

H1

H2

13

3. Research Methods

In this section I will introduce the sample which is used to test my previously mentioned hypotheses. Then, the dependent, independent as well as the control variables and their measures will be explained. Eventually, the method assumptions will be tested and reported.

3.1 Sample

The study builds on a database collected from Bureau van Dijk's (BvD, 2015) Orbis database as well as multiple corporation websites. Orbis is the most appropriate single-source firm-level database for this research because it is one of the most comprehensive and inter-temporal pan-European databases, containing detailed information about many public and private companies in virtually all European countries, including information on the CEO and executive boards of these firms. However, during the initial data collection process it was found that much data was incomplete or outdated. This is why corporation websites have been added as a source of recent data.

The sample consists of the top 90 largest German firms. The top 90 was found by using the Fortune Global 500 study from the year 2014 (Fortune, 2014), ranking the companies by revenues as well as a recent study from PricewaterhouseCoopers, which ranked them by market capitalization (PwC, 2013). After checking for outliers, a few values have been left out.

3.2 Dependent Variable

14

international entanglement of a corporation at a certain time and are related to the foreign activities of MNCs.

3.3 Independent Variable

The independent variable board age diversity will be measured with the coefficient of variation. According to Williams and O’Reilly (1998), it is one of the most commonly used indices for team level demographic diversity.

𝐶𝑉 = √1_{𝑁 ∑}𝑛 (𝑥_𝑖 𝑖=1 − 𝜇)² 1 𝑁 ∑𝑛𝑖=1 𝑥𝑖 =𝜎 𝜇 = 𝑆𝑡𝑎𝑛𝑑𝑎𝑟𝑑 𝐷𝑒𝑣𝑖𝑎𝑡𝑖𝑜𝑛 𝑀𝑒𝑎𝑛

The formula is in essence the within-group Standard Deviation (SD) divided by the group mean. More precisely, I first needed to collect the ages of all the top 90 German board members, then calculate the means as well as the standard deviations, and finally, divide the standard deviations by the means, to get the values of the coefficients of variation (CV). In general, the CV value can range from zero to positive infinity, with the minimum value of 0 showing zero variability in the ages of the group (meaning that everyone has the same age), and positive values standing for increasing age variations. In summary, as the variability in the group gets larger and/or the mean smaller, CV will increase (Harrison & Sin, 2006).

3.4 Control Variables

In order to test the validity of this study, control variables are included. These are variables that are held constant in order to assess or clarify the relationship between the dependent and the independent variable. The control variables in this research are firm size and firm age. Further, to also capture the possible influence of categorical variables, a dummy variables were created based on a firm’s industry.

15

Another control variable is firm age. In their paper, Andersson, Gabrielsson and Wictor (2004) stated that “Both firm size and firm age have traditionally been used as the main predictors of a firm’s international activities”. It will be measured accordingly as the number of years that have passed since the firm was founded until now (year 2015).

The dummy variables were created by taking the five industries with the most observations. In my sample, the top 90 companies were operating in 15 different industries. However, out of 90 firms, some industries only appeared once or twice, which should not have a bigger influence on my results. Therefore, the industries were summarized to five most observed ones1, namely

 Machinery/Equipment/Furniture/Recycling  Chemicals/Rubber/Plastics

 Gas,Water,Electricity  Services

 Rest

The moderator variable CEO tenure will be measured by the CEO’s years in office of the focal company. An overview of all variables, definitions and measures can be found in table 1.

1_{Initial industries:}  Machinery,Equipment,Furniture,Recycling  Gas,Water,Electricity  Insurance  Chemicals,rubber,plastics  Wholesale & Retail trade  Telecommunications & Post  Banking  Cargo services  Publishing,Printing  Textiles,Apparel  Other Services  Food,beverages,tobacco  Transport  Construction

16

Table 1: Overview of Variables and Measures

3.5 Method Assumptions

This research will be based on an OLS regression. OLS makes four crucial assumptions which have to be satisfied in order to provide the best linear unbiased estimates. The assumptions that are tested for are homoscedasticity, endogeneity, multicollinearity and normality. A violation of these assumption might create problems with the reliability of the results.

Homoscedasticity

Homoscedasticity assumes that the variance of the error term is constant and the same for all observations (i) (var(ei) = σ²). If this assumption is violated and the error variance for all observations is not identical then heteroscedasticity is present. Heteroscedasticity results in biased estimates of the standard errors which leads in turn to bias in the test statistics and thus to incorrect hypotheses tests, p-values and confidence intervals. Heteroscedasticity itself does not lead to biased coefficients but makes the estimator no longer the best with the smallest variance (Hill, Griffiths & Lim, 2009). The problem with heteroscedasticity is that more weight is given to observations with potentially larger error terms. In that case observations furthest away from the true regression line provide us with the least information about the true regression line. In case heteroscedasticity is detected, there are methodological ways to correct for the weight of larger error variances to get estimates with the smallest sum of squared errors. To test this assumption, I plotted the standardized residuals against the standardized predicted values, getting the following result:

Type Variable Measure

Dependent Internationalization

The number of different countries in which a firm has subsidiaries

Independent Board age diversity Coefficient of variation

Control Firm size Number of employees

Control Firm age Number of years since foundation (until 2015)

Control

Machinery/Equipment/Furnitur e/Recycling

Dummy (1=Machinery/Equipment/furniture, recycling; 0=Other)

Control Chemicals/Rubber/Plastics Dummy (1=Chemicals/Rubber/Plastics;0=Other) Control Gas,Water,Electricity Dummy (1=Gas,Water,Electricity;0=Other)

Control Services Dummy (1= Services;0=Other)

17

Figure 2: Scatterplot Internationalization

As the points are randomly and evenly dispersed throughout the plot and show a random array of dots evenly dispersed around zero, it can be said that the assumption of homoscedasticity has been met.

Endogeneity

18

and observed by management. If endogeneity exists, then it is more likely that indicators that are based on respondent’s judgment are correlated with the error term and could thus be influenced by unobserved variables. However, such indicators are not present as independent variables in the sample. If concepts cannot be observed directly, instrumental variables are commonly used to measure those latent constructs. Since the common tests for endogeneity rely on instrumental variables, it is difficult to prepare such a test in the present sample as such additional variables are not available. The exogenous nature of the independent variable has thus to be assumed.

Multicollinearity

Another crucial assumption of OLS is that the independent variables are not perfectly correlated such that “the values of 𝑥_𝑖𝑘 are not exact linear functions of other explanatory variables” (Hill et al., 2009: 154). If this assumption is violated, variables are said to be collinear, meaning that they are very strongly correlated which makes it difficult to isolate the relationship between variables (Field, 2009). OLS estimates will not be biased and still be the best linear unbiased estimates.The primary concern is that when the multicollinearity degree increases, the estimates of the regression model become unstable and the standard errors of the coefficients can get inflated. More specific, we would get relatively imprecise information about our unknown parameters due to large standard errors. It might then be have difficult to predict the true parameters, an issue which might even become more problematic when there is little variation in the explanatory variables. There are several ways to test for

multicollinearity. One way it to check the correlation matrix of the predictor variables and scan for very high correlations (around .80 or .90) (Field, 2009). However, this method might miss some forms of multicollinearity.That is why I used the variation inflation factor (VIF), as part of collinearity diagnostics to test for the presence of multicollinearity. This index

19

Table 2: Test for Multicollinearity

Variable Tolerance VIF

Firm Size .950 1.053

Firm Age .722 1.384

Machinery Equipment Furniture Recycling .543 1.842

Gas Water Electricity .728 1.374

Services .489 2.047

Rest .529 1.890

Board Age Diversity .827 1.210

Diversity_tenure .855 1.170

VIF Mean 1.509

Normality

For the final test, OLS assumes that the values of the error term are normally distributed about their mean. A violation of the normality assumption might lead to biased p-values and thus affects the test for statistical significance (Hill et al. 2009). To test this, I had a look at the histogram and the normal probability plot in the figures 4 and 5 respectively. Since the histogram looks like a bell-shaped curve, normality is assumed. Also the probability plot shows the same result, as the points all lie close to the straight line which represents a normal distribution.

20

21

4. Empirical Results

Due to the fact that all method assumptions have been satisfied meaning that there are no severe problems with normality, heteroscedasticity, endogeneity and multicollinearity, I will now continue with reporting my results. Before reporting the actual regression results, I will talk about descriptive statistics.

4.1 Descriptive Statistics

22

Table 3: Descriptive Statistics

Mean Std.Dev Min Max

Dependent Variable

Internationalization 32.831 25.379 1 97

Independent Variable

Board Age Diversity 0.090 0.397 0.01 0.222

Control Variables Firm Size 46,964.71 65,315.72 1,029 279,972 Firm Age 76.133 54.822 3 180 Machinery/Equipment/Furniture/Recycling 0.253 0.437 0 1 Chemicals/Rubber/Plastics 0.176 0.383 0 1 Gas/Water/Electricity 0.066 0.250 0 1 Services 0.187 0.392 0 1 Rest 0.308 0.464 0 1 4.2 Regression Results

Table 4 below shows the OLS regression results. The results comprise three different models. Next to standardized beta values, standard errors, the number of observations, R², Adj. R², the significance levels and the F-values are listed.

23

according to table 3), internationalization increases by 0.379 standard deviations. Since the standard deviation for internationalization is, according to table 3, 25.379, it means that this constitutes a change of 9.619 (0.379*25.379) in internationalization, or more precisely, around 9 countries will be added in which a firm has subsidiaries. However, this effect is only true if the effects of the other variables are held constant. Accordingly, if firm age increases by one standard deviation (54.822), internationalization increases by 0.338 standard deviations, leading to a change of 8.578 in internationalization (holding other effects constant). The industry dummies all have negative effects on internationalization. The dummy “Rest” has the highest absolute value (-0.433) suggesting a high degree of importance in this model. If a firm is member of an industry other than the other four dummies suggest, internationalization will decrease by -0.433 standard deviations. This would mean an effect of -10.989 (-0.433*25.379) on internationalization. Besides the dummy “Rest”, the industries machinery, equipment, furniture, recycling, chemicals, rubber, plastics, gas, water, electricity and services indeed have less significant effects than the previously mentioned one, but still influence internationalization negatively.

Model 2 adds the main independent variable board age diversity to the regression. There is actually no significant change in R², suggesting that board age diversity does not account for any more variance in the outcome than the controls. In addition, there is a negative but not significant effect of board age diversity on internationalization (-0.015). This means that hypothesis one needs to be rejected. There is no significant positive effect of board age diversity on internationalization. There was only a slight change for the other variables. The most mentionable one is the decrease in significance of firm age, now only having a p-value of p<0.01.

Model 3 finally tests my interaction variable diversity*tenure (CEO tenure), which analyses the moderating effect of hypothesis 2. First, it needs to be said that R² increased from 0.398 to 0.405, or more precisely, from 39.8% to 40.5%, meaning that the interaction variable added 0.7% (40.5%-39.8%) to the total amount explaining the variation in internationalization. Then, it can be said that except for the dummy services whose effect lost significance, not much has changed in comparison to the previous model.

Unfortunately, although there is a negative effect of CEO tenure on internationalization with a value of (-0.092), it is not significant. Therefore, I also need to reject hypothesis 2.

24

The biggest difference can be found in model 3, where it is 0.062 (0.405-0.343). This shrinkage means, that if the model was derived from the population rather than a sample, it would account for approximately 6.2% less variance in the outcome. To summarize the findings, it was found that both hypotheses needed to be rejected. Except for the unexpected negative direction of the effect of board age diversity on internationalization, no major surprises could be found. The findings will be discussed in the following part.

Table 4: Regression Results of the Impact of Board Age Diversity on

Internationalization

2 _{Intercept term}

Model 1 Model 2 Model 3

Controls Dependent Moderator

Controls Firm Size 0.379*** 0.378*** 0.374*** (0.000) (0.000) (0.000) Firm Age 0.338*** 0.333** 0.344** (0.045) (0.048) (0.048) Machinery,Equipment,Furniture,Recycling -0.298* -0.301* -0.286* (6.909) (7.068) (7.128) Gas,Water,Electricity -0.284** -0.286** -0.283** (10.029) (10.185) (10.195) Services -0.266* -0.267* -0.236 (7.822) (7.882) (8.154) Rest -0.433*** -0.435*** -0.442*** (6.492) (6.564) (6.579) Dependent variable

Zscore: Board Age Diversity -0.015 -0.003

(2.552) (2.574) Moderator diversity_tenure -0.092 (2.397) (Constant)2 _{30.730 31.038 30.598} (7.150) (7.452) (7.469) Observations 85 85 85 R² 0.398 0.398 0.405 Adj. R² 0.352 0.343 0.343 F- value 8.593*** 7.277*** 6.478***

25

4.1 Discussion of findings

It has been found that firm size and firm age both had significant positive effect on internationalization. This is in line with the theory mentioned in the previous parts. What I want to focus on here is the surprising negative and not significant effect of board age diversity on internationalization. Due to several different possibilities for this result which will be discussed in the limitations part, I would like to point out one of them here. So far, I have been mostly underlining the positive aspects of high diverse boards. However, although literature has often mentioned that diversity means more different views, experiences and knowledge which then enhances variety and creativity, diversity can also have many downsides. The reason for this often lies in two different theories: Social categorization and attraction/similarity. In this thesis, age cohorts play the most important role as explanation for the various effects of diversity. A cohort is a group of people that have a relevant date in common, be it year of birth, year of marriage, entry into the job market etc. (Hambrick & Mason, 1984). Here, the focus is on people of the same age. In an organization, different groups of people often form so called sub-groups within bigger groups, consisting of people with which they can identify. While forming these groups, people compare themselves with others on different salient characteristics like in this case age (Williams & O’Reilly, 1998). Persons of similar age can develop similar outlooks in life and shared experiences. People of one sub-group are referred to in-group members. The problem here is that these people can start to perceive other persons who are not member of this sub-group, also called out-group members, as less trustworthy, honest and cooperative as their own members (Williams & O’Reilly, 1998). As these two groups can have different normative and cultural attitudes, it is possible that conflict will increase which will have a negative effect on productivity within the group. Furthermore, since members of different cohorts find it difficult to communicate with each other, communication in general could become increasingly strained which in extreme cases can lead to the inability of the group to take actions or make decisions (Wiersema & Bantel, 1992). Therefore, organizations could have a hard time to keep their creative groups together.

26

effect. It is therefore all about finding the right mix. Further research could maybe analyze a possible curvilinear relationship in this case.

5. Robustness Test

To make valid causal inferences, robustness is a necessary characteristic of the underlying method. Robustness checks are frequently used to examine how certain empirical coefficient estimates are. When coefficient estimates do not change sign and magnitude significantly following modifications, it is commonly taken to be evidence that estimates are robust. Generally, we speak of structural validity when coefficients are plausible and robust (White & Lu, 2010).

27

6. Conclusion

The objective of this study was to analyze the antecedents for successful internationalization. More in particular, the influence of board age diversity on internationalization was tested. The findings in this thesis are for the most part in line with the theory. However, there are still a lot of limitations and possibilities for further research in order to improve the results. One hundred German top companies have been tested on the effect of their board age diversities on the internationalization level. Although theory on the one hand suggests increasing creativity and innovation levels due to increasingly diverse boards which then leads to high levels of internationalization, on the other hand there are also lists of also authors emphasizing the possible negative results of diversity, leading to conflict and unsatisfying results which then decreases internationalization. In the end, a negative but not significant result was found while running the regression, which rejected hypothesis one. As a moderator effect the influence of CEO tenure on the whole relationship was analyzed. Also this hypothesis needed to be rejected, showing negative but not significant results. The list of control variables all gave satisfying and significant results, implying that firm size, firm age as well as the industry a firm operates in all play an important role in determining the final level of internationalization.

6.1 Added value of this study

Despite the vast amount of literature about internationalization, this study is among the few which takes internationalization antecedents as stand-alone variables. So far, most studies were aiming to discuss the internationalization-performance relationship. Besides, board diversity was mostly analyzed as a general phenomenon consisting of many diverse characteristics like tenure, age, educational background, size etc. This study is new in terms of placing the focus on board age diversity alone. The findings in this study also add to the upper echelon perspective, in that organizational outcome was shown to be associated with the demographic characteristics of the board.

28

an easier time to decide how the composition of their boards should look like, especially when focusing on a successful internationalization.

6.2 Limitations

Like any other study, also this study does not come without limitations. Although focusing on just one country may have its benefits, it also has its downsides. First of all, a sample size of 85 might be too small to come up with generalizable results. Further research could check for more clear results when running the same regression with a larger number of firms. Second, by focusing only on German companies, it will be hard to actually use the findings for firms whose headquarters are in countries other than Germany. To test for the possibility of similar results, the same study would have to be made with samples from different countries. Besides, only top companies have been used in the study. Therefore, the results cannot be perfectly used for smaller less successful companies.

In addition, a topic as complex as internationalization and board diversity clearly has a large number of influencing control variables. Since in this study only few control variables were added, further investigations should aim for increasing their number. Due to incomplete data in the Orbis database, other measures for internationalization next to the number of countries in which a firm has subsidiaries were not possible to be used. This resulted in a rather incomplete and maybe even ineffective measurement of the dependent variable internationalization. Although the coefficient of variation is widely used in the literature as a measure of age diversity, it is hard to interpret the actual results, as it is rather impossible to get an actual idea board composition by just looking at the value of the coefficient. Is it unknown whether the board consists of more younger or older people. It only says something about the variation of the mean. So for a more clear interpretation, the mean should also be taken into consideration. Furthermore, it could be that observable demographic aspects like age are not suitable to predict an outcome like internationalization. Even if people share the same age, it does not mean that they are not diverse in many other ways. A human is way too complex to be analyzed from just this one perspective. Therefore, diversity has to be studies in a more complex way.

29

7. Appendix

Table 5:Robustness Test: Internationalization Ratio

3 _{Intercept term}

Model 1 Model 2 Model 3

Controls Dependent Moderator

Controls Firm Size 0.379*** 0.378*** 0.374*** (0.000) (0.000) (0.000) Firm Age 0.338*** 0.333** 0.344** (0.000) (0.000) (0.000) Machinery,Equipment,Furniture,Recycling -0.298* -0.301* -0.286* (0.071) (0.073) (0.073) Gas,Water,Electricity -0.284** -0.286** -0.283** (0.103) (0.105) (0.105) Services -0.266* -0.267* -0.236 (0.081) (0.081) (0.084) Rest -0.433*** -0.435*** -0.442*** (0.067) (0.068) (0.068) Dependent variable

Zscore: Board Age Diversity -0.015 -0.003

(0.026) (0.027) Moderator diversity_tenure -0.092 (0.025) (Constant)3 0.317 0.320 0.315 (0.074) (0.077) (0.077) Observations 85 85 85 R² 0.398 0.398 0.405 Adj. R² 0.352 0.343 0.343 F- value 8.593*** 7.277*** 6.478***

30

Table 6:List of Companies in the Sample

Name of companies used in the sample

1 VW 46 Symrise AG

2 Siemens 47 Wacker Chemie

3 E.ON 48 Fucks Petrolub SE

4 Daimler 49 MTU Aero Engines AG

5 Allianz 50 K+S AG

6 BMW 51 Bilfinger SE

7 BASF 52 Fielmann AG

8 Metro 53 Puma SE

9 Telekom 54 Osram Licht AG

10 Post 55 Celesio AG

11 RWE 56 Wirecard AG

12 Audi 57 TUI AG

13 Münchner RE 58 BP Europe SE

14 Bayer 59 Robert Bosch GmbH

15 Continental 60 Edeka Zentrale AG

16 Deutsche Bank 61 Bertelsmann SE

17 Lufthansa 62 Freenet AG

18 SAP 63 Carl Zeiss Meditec

19 Henkel 64 Mainova 20 Merck 65 Krones AG 21 Linde AG 66 Gelsenwasser AG 22 Porsche 67 Aurubis AG 23 Adidas 68 Elringklinger AG 24 Beiersdorf 69 SGL Carbon SE

25 Fresenius 70 Deutsche Euroshop AG

26 MAN AG 71 Salzgitter AG

27 Evonik 72 Elster Group SE

28 Heidelberg Cement AG 73 KWS Saat SE

29 ENBW 74 CTS Eventim AG&Co.KGAA

30 Kabel Deutschland Holding AG 75 Wincor Nixdorf AG

31 Commerzbank AG 76 Gerry Weber International AG

32 Infineon AG 77 Sartorius AG

33 Prosieben Sat1 78 Douglas Holding AG

34 Brenntag 79 Deutsche Wohnen AG

35 Hugo Boss AG 80 Drägerwerk AG&Co.KGAA

36 GEA Group AG 81 MVV Energie AG

37 Telefonica Deutschland 82 Rheinmetall AG

38 Sky Deutschland 83 Gerresheimer AG

39 Südzucker AG 84 GFK SE

40 United Internet AG 85 Hamburger Hafen und Logistik AG

41 Fraport AG 86 Deutsche Börse AG

42 Hochtief AG 87 Thyssen Krupp AG

43 Deutsche Annington Immobilien 88 Rational AG

44 Lanxess AG 89 Kion Group AG

31 Table 7: Correlations 1 2 3 4 5 6 7 8 9 10 1. Internationalization 1 2. Firm Size .355 ₁ 3. Firm Age .382 -.034 ₁ 4. Machinery,Equipment,Furniture, Recycling .010 .035 .047 ₁

5. Chemicals, Rubber, Plastics .387 -.027 .226 -.267 ₁

6. Gas,Water,Electricty -.196 -.096 -.110 -.153 -.133 ₁

7.Services -.169 -.110 -.359 -.267 -.232 -.133 ₁

8. Rest -.085 .135 .129 -.378 -.329 -.188 -.329 ₁

9. Zscore:Board Age Diversity -.132 -.079 -.338 -.156 .056 -.031 .188 -.047 ₁

32

8. References

Ali, Muhammad, Yin Lu Ng, and Carol T. Kulik. "Board age and gender diversity: A test of competing linear and curvilinear predictions." Journal of Business Ethics 125.3 (2014): 497-512.

Andersson, S., Gabrielsson, J. and Wictor, I. (2004), International Activities in Small Firms: Examining Factors Influencing the Internationalization and Export Growth of Small Firms. CAN J ADM SCI, 21: 22–34

Andersen, O. (1993). On the Internationalization Process of Firms: A Critical Analysis. Journal of International Business Studies (1993) 24, 209–231.

Bantel, K.A. & Jackson, S.E. (1989). Top management and innovations in banking: Does the Composition of the top team make a difference? Strategic Management Journal, 10, 107-124.

Bilkey, Warren J. & George Tesar. 1977. The export behavior of smaller Wisconsin manufacturing firms. Journal of International Businiess Studies, 9:93 98.

Bobillo, A.M., Lopez-Iturriaga, F., & Tejerina-Gaite, F. (2010). Firm performance and international diversification: The internal and external competitive advantage. International Business Review, 19, 607-618.

Cavusgil, S., & John R. Nevin. 1981. Internal determinants of export marketing behavior: An empirical investigation. Journal of Marketing Research. 1: 14-19.

Calof, J.L. and Beamish, P.W. (1994) Succeeding internationally with a geocentric attitude. Business Quarterly, Autumn, pp. 2-7.

Calof, J.L.(1994) The Relationship between Firm Size and Export Behavior Revisited. Journal of International Business Studies,Vol. 25, No. 2 (2nd Qtr., 1994), pp. 367-387

Carlson, R., & Karlsson, K. 1970. Age, cohorts, and the generation of generations. American Sociological Review, 35: 710-718.

33

Child, J. 1974. Managerial and organizational factors associated with company performance. Journal of Management Studies, 11: 13-27.

Czinkota, Michael R. 1982. Export development strategies: US promotion policies. New York: Praeger Publishers.

De Jong, G., and J. van Houten (2014).The impact of MNE cultural diversity on the internationalization -performance relationship. Theory and evidence from European multinational enterprises. International Business Review, 23, 313- 326.

Dwyer, S., Richard, O.C., & Chadwick, K. (2003). Gender diversity in management and firm Performance: The influence of growth orientation and organizational culture. Journal of Business Research, 56(12), 1009-1019.

Gellner, U., & Veen, S. (2013). Positive effects of ageing and age diversity in innovative companies–large‐scale empirical evidence on company productivity. Human Resource

Management Journal, 23(3), 279-295.

Ghoshal, S., & Nohria, N. 1989. Internal differentiation within multinational corporations. Strategic Management Journal, 10: 323-337.

Gomez-Mejia, L., & Palich, L. 1997. Cultural diversity and the performance of multinational firms. Journal of International Business Studies, 28: 309-336.

Hambrick, D.C. and G. Fukutomi, "The Seasons of a CEO's Tenure," Academy of Management Review, October 1991.

Hambrick, D. C., & Snow, C. C. A contextual model of strategic decision making in

organizations. In R. L. Taylor, M. J. O'Connell, R. A. Zawacki, & D. D. Warrick (Eds.), Academy of Management Proceedings, 1977, 109-112.

Harrison, D. A., & Sin, H. (2006). What is diversity and how should it be measured. Handbook of workplace diversity, 191-216.

Henderson, A. D., Miller, D. and Hambrick, D. C. (2006), How quickly do CEOs become obsolete? Industry dynamism, CEO tenure, and company performance. Strat. Mgmt. J., 27: 447–460. doi: 10.1002/smj.524.

Hitt, M.A., Hoskisson, R.e., & Hicheon, K. (1997). International diversification: Antecedents, outcomes, and firm performance in product-diversified firms. The Academy of

34

Buckley, P.J., & Ghauri P. (1999). The internationalization of the firm. London: International Thomson Business Press.

Click, R.W., & Harrison, P. (2000). Does multinationality matter? Evidence of value destruction in U.S. multinational corporations. FEDS Working paper no. 2000-21. Clendenin, W. D. (1972) Company Presidents look at the Board of Directors, California

Management Review, Spring, 60-66.

Dearborn, D. C., & Simon, H. A. 1958. Selective perception: A note on the department identi-fications of executives. Sociometry, 21: 140-144.

Dörrenbächer, Christoph (2000): Measuring corporate internationalisation: A review of measurement concepts and their use, WZB Discussion Paper, No. FS I 00-101 Finkelstein, S., & Hambrick, D. C. (1996). Strategic leadership: Top executives and their

effects on organizations. Minneapolis: West Publishing.

Fortune (2015)- Global 500. Retrieved from: http://fortune.com/global500/

Hambrick, D. C. (2007). Upper echelon theory: An update. Academy of Management, (32) 2, 334-343.

Hambrick, D. E., & Mason, P. A. 1984. Upper echelons: The organization as a reflection of its top managers. Academy of management review 9: 193-206.

Hill, R. C., Griffiths, W. E., & Lim, G. C. 2009. Principles of Econometrics 4th Edition. Jackson S. E., & Alvarez, E. B. 1992. Working through diversity as a strategic imperative. In

S. E. Jackson (Ed.), Diversity in the workplace: Human resources initiatives, 13-29. New York: Guilford Press.

Johansen, J., & Vahlne, J. 1977. The internationalization process of the firm-A model of knowledge development and increasing foreign market commitments. Journal of International Business Studies, 8: 23-32.

35

Lee, H.J., & Park, J.H. (2006). Top team diversity, internationalization and the mediating effect of international alliances. British Journal of Management, 17(3), 195-213.

Mace, M. (1976) Attracting new directors, Harvard Business Review, September-October, 46-51.

Melin, L. 1992. Internationalization as a strategy process. Strategic Management Journal, 13:99-118.

Menard, S. (1995). Applied logistic regression analysis. Sage university paper series on quantitative applications in the social sciences, 07-106. Thousand Oaks, CA: Sage. Milliken, F. J. and Martins, L. L. (1996). ‘Searching for common threads: understanding the

multiple effects of diversity in organizational groups’. Academy of Management Review, 21, 402–33

Myers, R. (1990). Classical and modern regression with applications (2nd ed.). Boston, MA: Duxbury.

Neter, J., Wasserman, W., Nachtsheim, C., & Kutner, M. 1985. Applied Linear Statistical Models. McGraw-Hill/Irwin

Nishii, L.H., Gotte, A. and Raver, J.L. (2007). Upper echelon theory revisited: The

relationship between Upper echelon diversity, the adoption of diversity practices and organizational performance. CAHRS Working Paper Series.Cornell University. Porter, M. 1986. Competition in global industries. Boston: Harvard Business School Press. Prahalad, C., & Hamel, G. 1994. Strategy as a field of study: Why search for a new paradigm?

Strategic Management Journal, 5: 5-16.

PwC (2013). Top 100 companies Europe and Germany. Retrieved from:

http://www.pwc.de/de/kapitalmarktorientierte-unternehmen/assets/top-100-companies-europe-and-germany.pdf

Rivas, J. L., Hamori, M., & Mayo, M. (2009, August). Board Composition And Firm

Internationalization. Academy of Management Proceedings (Vol. 2009, No. 1, pp. 1-6). Academy of Management.

36

Sanders, W. M. G., & Carpenter, M. A. (1998).Internationalization and firm governance: The roles of CEO compensation, top team composition, and board structure. Academy of Management Journal, 41(2), 158-178.

Sluiter, J.K. (2006), “High-demand jobs: age-related diversity in work ability?”, Applied Ergonomics, Vol. 37 No. 4, pp. 429–440.

Tempest, S., Barnatt, C. and Coupland, C. (2002), “Grey advantage”, Long Range Planning, Vol. 35 No. 5, pp. 475–492.

Turnbull, Peter W. 1987. A challenge to the stages theory of the internationalization process. In Stan D. Reid & Philip J. Rosson, editors, Managing export entry and expansion. NewYork: Praeger.

Vance, J. O. (1979) The Care and Feeding of the Board of Directors, California Management Review, Summer, 29-34.

Van Witteloostuijn, A., and G. de Jong (2010). An ecology of national rule birth. A longitudinal study of Dutch higher education law, 1960-2004. Journal of Public Administration Research and Theory, 20(1), 187-213.

Vroom, V., & Pahl, B. 1971. Relationship between age and risk-taking among managers. Journal of Applied Psychology, 55: 399-405.

Welch, Lawrence S. & Reijo Luostarinen. 1988. Internationalization: Evolution of a concept. Journal of General Management, 14(2): 34-55.

Wegge, J., Roth, C., Neubach, B., Schmidt, K.-H. and Kanfer, R. (2008), “Age and gender diversity as determinants of performance and health in a public organization: The role of task complexity and group size”, Journal of Applied Psychology, Vol. 93 No. 6, pp. 1301–1313.

White, H., & Lu, X. 2010. Robustness Checks and Robustness Tests in Applied Economics. UCSD Department of Economics discussion paper, 2010. Retrieved from:

http://econ.ucsb.edu/~doug/245a/Papers/Robustness%20Checks.pdf

37