Board diversity and corporate performance in German
listed firms
An empirical investigation of major German Organizations
Master thesis
MSc International Business & Management
Faculty of Economics and Business
University of Groningen
23rd January 2017
By
Master thesis
MSc International Business & Management
Board diversity and corporate performance in German listed firms
An empirical investigation of major German Organizations
23/01/2017
By:
Alexander Emiljanow S2826216
Supervisor: Dr. Gjalt de Jong
Co-Assessor: P.J. Marques Morgado
Faculty of Economics and Business University of Groningen
Duisenberg Building, Nettelbosje 2, 9747 AE Groningen, The Netherlands P.O. Box 800, 9700 AV Groningen, The Netherlands
+31 (0)50 363 7098
Abstract
Over the last decades, board diversity is more and more regarded as a significant mechanism of efficient corporate governance. Hence, the question that arises is whether a homogenously or rather heterogeneously composed board contributes best to the efficiency of a firm’s management. It is generally considered that board diversity automatically leads to an increase of firm performance. For that reason, the economic impact of board diversity issues needs to be examined empirically. This thesis investi-gates the relationship between diversity within management boards and corporate per-formance for German companies. Relevant top management diversity characteristics are age, gender and nationality. Firm performance is a dependent variable in the re-gression analysis and will be measured as return on assets, return on equity and To-bin’s Q. A comprehensive literature review as well as an empirical analysis based on 91 publicly listed German MNCs for the financial years 2012, 2013, 2014, will be part of this work. There is almost no comparable analysis which is focusing on aforemen-tioned dimensions for German companies. No significant relationship between measures of the firm performance and age of directors can be found. A positive signif-icant influence can be found between ROA and percentage of women. Percentage of foreign directors has no significant relation. In order to check for robustness, Blau di-versity indices are used for age, gender and nationality. Mostly negative effects of board diversity on corporate performance can be found. There is a negative impact between Blau gender diversity and ROA as well as with ROE. Another negative relation is between Blau age diversity and ROE. Results with a German context remain unclear, if performance is positively influenced by board diversity, since the results are not reli-able enough.
Preface
The completion of this Master thesis marks the end of my academic career, and I am looking forward to my next adventures. Before a new chapter continues, I would like to acknowledge a number of people that have provided invaluable support to me during the writing process. I would like to thank my supervisor Dr. de Jong from the Faculty of Economics and Business at University of Groningen, my parents, my brother, my old friends from my first academic stage at TU Clausthal, and my new fellowship from RUG Groningen. Thanks to all the aforementioned people, my student life became a worthwhile and enjoyable experience.
Alexander Emiljanow
Table of Contents
Abstract ... III Preface... IV Table of Contents ... V List of Abbreviations ... VII List of Figures ... VIII List of Tables ... IX
1. Introduction ... 10
1.1 Background & Research Gap ... 10
2. Theoretical framework ... 12
2.1 German Two-Tier System... 12
2.2 Supportive Theories on Board diversity ... 13
2.2.1 Resource-dependency theory ... 13
2.2.2 Human capital theory ... 14
2.2.3 Agency theory ... 14
2.2.4 Upper Echelon Theory ... 15
2.3 Opposing Theories on Board Diversity ... 17
2.4 Age Diversity ... 18
2.5 Gender Diversity ... 20
2.6 National Diversity ... 21
2.7 Endogeneity and Reverse Causality ... 23
4. Findings and Interpretation ... 32 4.1 Descriptive Statistics ... 32 4.2 Correlation ... 33 4.3 Regression ... 34 4.4 Robustness Checks... 36 5. Concluding Remarks ... 39
5.1 Conclusion & Discussion ... 39
5.2 Managerial implications ... 41
5.3 Limitations & Avenues for Future Research ... 41
References ... 43
Appendices ... 57
List of Abbreviations
The following is an alphabetical list of the used abbreviations.
AG German: Aktiengesellschaft
English: Stock (limited) Corporation
AktG German: Deutsches Aktiengesetz
English: German Stock Corporation Act (GSCA)
BOD Board of Directors / Management Board
CEO Chief Executive Officer
CG Corporate Governance
FID Foreign independent director
GSCA German Stock Corporation Act
HQ Headquarter
MNC Multinational Corporation
OLS Ordinary Least Squares
ROA Return on Assets
ROE Return on Equity
SPSS Statistical Package for the Social Scientists
TMT Top Management Team
TSR Total Return to Shareholders
VIF Variance Inflation Factor
List of Figures
List of Tables
Table 1: Statistics of Variables ...33
Table 2: Cross-correlation Table ...34
Table 3: Regression Analysis ...35
1. Introduction
1.1 Background & Research Gap
The influence of Top Management Team (TMT) diversity on corporate performance and valuation has been an issue in the organizational economics as well as in the financial literature (Adams, Hermalin & Weisbach, 2010; Miliken & Martins, 1996). Most of previous studies have focused on the ideal structure and size of the board in the US sector, since it is assumed that optimal board structure can decrease agency costs caused by the separation of ownership and control (Shleifer & Vishny, 1997). In con-trast, German research has primarily focused on two different aspects of board com-position that are characteristic for its industrial and financial relations system. First, employee representation1 and, second bank representation on supervisory boards
(Frick & Lehmann, 2005). The majority of studies on employee representation include some flaws such as the fact that they are mostly based on small samples, and are based on cross-sectional data, which does not allow to monitor for unobserved firm-specific effects (FitzRoy & Kraft, 2005). Another remarkable fact is that none of these mentioned studies inspect the impact of boards on corporate performance and the market valuation of the same panel of Multinational Corporations (MNCs).
MNCs from the United States of America delivered first positive indications that board diversity leads to higher firm performance (Carter et al., 2003), hence research has tried to focus more on the relationship between diversity in board of directors and firm performance. Some studies indicate that gender and racial diversity in the TMT has a positive impact on firm performance (Carter et al., 2003; Erhardt et al., 2003). In some countries or more specifically in Europe, exist mandatory quotas for female board di-rectors. Germany has introduced a similar reform in 2015, with the main goal to include a women quota by 30% to corporate boards (Bertrand, 2014). Regardless of the posi-tive effects, other findings found negaposi-tive relations between gender diversity and firm performance (Adams & Ferreira, 2009). Despite the mixed study results from different countries, board diversity is still viewed as desirable. Previous research shows that group diversity improves group performance, abilities to exchange ideas and
ages group discussion (Schippers et al., 2003; Van Knippenberg et al., 2004). Disper-sion of age, gender diversity and board size refer to board deciDisper-sion making processes which in return influence the overall firm performance (Bøhren & Strøm, 2007). Top management team (TMT) diversity is defined as the variety inherent in the TMT’s com-position and it can be measured in various dimensions such as age, gender, national-ity, educational background, ethnicnational-ity, professional experience and organizational membership (Campbell & Mínguez-Vera, 2007). Board diversity can be differentiated into demographic (observable like gender or age) and cognitive diversity (non-observ-able like education, knowledge, values) (Milliken & Martins, 1996; Petersen, 2000).
The focus of this study is on the relationship between board diversity and financial firm performance in German companies. Consequently the relevant research question be-comes:
“Does board diversity influence financial firm performance in German firms?”
Up to now, there is no similar research available, which focuses on the aforementioned determinants with a German context. The purpose of this thesis is to investigate the relationship between TMT diversity and the firm performance of German MNCs. Main goal is to explore the impact of board diversity, which is measured in age, gender and nationality, on corporate performance, which will be measured with return on assets (ROA), return on equity (ROE) and Tobin’s Q. Further, this thesis shall make a con-vincing contribution to the existing research that suffers from a certain amount of short-comings.
2. Theoretical framework
This chapter provides the necessary theoretical framework in form of literature reviews and current research findings about the effects of diversity. “Literature reviews are sys-tematic syntheses of previous work around a particular topic” (Card, 2010, p. 725) and are essential for research because: “In practice we all start our own research from the work of our predecessors, that is, we hardly ever start from scratch” (Schumpeter, 1954, p. 38).
The main reason why diversity influences firm performance is the efficient comprehsion of the marketplace (Robinson & Dechant, 1997). Whereas the firm is able to en-hance its overall diversity and adapt to consumer’s diversity, satisfied consumers and growing markets will be more likely to achieve. This idea is based on attitudes, beliefs, and cognitive functioning, which are closely related to demographic variables like age, race or gender (Robinson & Dechant, 1997). Heterogeneity leads to more effective alternative solutions and an increase variety of perspectives. At first, the German two-tier system for corporate governance will be presented. Consequently, for this thesis relevant diversity theories will be introduced. Then, empirical results of former research findings about the relationship between the management determinants age, gender and nationality, will be presented. Hypotheses will be constructed during the clarifica-tion process.
2.1 German Two-Tier System
The system of corporate governance in Germany differs fundamentally from the Anglo-American equivalent. In Anglo-Saxon countries is the one-tier2 system prevalent and
its main goal is the maximization of the returns to the company’s shareholders and to mitigate general agency issues (Shleifer & Vishny, 1997). The German system also tries to consider the views of the remaining stakeholders and to involve their disparate views in corporate decision making processes (Fauver & Fuerst, 2006). More pre-cisely, a publicly held German MNC (“Aktiengesellschaft”) is defined by a two-tier sys-tem3 and is based on two administrative bodies, which are the management board
(“Vorstand”) and the supervisory board (“Aufsichtsrat”). German law (“Aktiengesetz” /
GSCA) prescribes a dual board system for German stock corporations. A two tier-sys-tem is based on organizational separation of management and supervision. While the management board (executive directors) is leading the company under its own respon-sibility (§ 105 GSCA), the supervisory board advises, appoints and monitors the mem-bers of the board and is involved in decisions of tremendous relevance to the firm (§§ 84, 111 GSCA). The shareholders elect at the general meeting the members of the supervisory board (§ 101 GSCA). Another feature of MNCs in Germany is that compa-nies with a size of 500 or respectively 20004 employees require having one third of
employee representatives on the supervisory board (codetermination). Since 2015 ex-ist the law, which dictates that MNCs lex-isted on the German exchange market have to consider a women quota of 30% in the management board and supervisory board (§ 96 GSCA). Besides the differences between German and Anglo-Saxon governance systems (e.g. checks and balances, joint authority, etc.), similarities exist such as the duties and responsibilities of the supervisory board members (Fauver & Fuerst, 2006). Nonetheless, it is rather unclear whether findings of Anglo-Saxon studies can be trans-ferred to a German context.
2.2 Supportive Theories on Board diversity
Several other theories give sufficient evidence about the positive impact of board di-versity. A more heterogeneous board is indicated to be morally more acceptable (Broome, Conley & Krawiec, 2011). Moreover, diverse groups are more effective, per-form tasks faster, offer different perspectives, have more network connections, and presents more creativity and innovation (Broome et al., 2011). Basically, variations of firm relevant variables are the reason for also firm specific advantages. Some MNCs benefit more than other companies.
2.2.1 Resource-dependency theory
Resource-dependency theory developed by Pfeffer & Salancik (1978) is based on the mutual interaction between organizations with the main goal to support the exchange process of resources. Long-term firm success depends on the environment, the avail-ability and the controlling of critical resources (Pfeffer & Salancik, 1978). This means
4 MNCs larger than 2000 employees require an employee representation of even 50% of the supervisory
that upper echelons and especially corporate directors (who distinguish in terms of age, education, gender, nationality), with corporate administration rights are able to concentrate the diverging resources for a better output (Hillman et al., 2000). Out of that, it can be concluded that a higher effectiveness of board activates can be ex-plained with a higher information processing or even the necessity to discuss within the plenum (Carter et al., 2010, p.398). The theory explains how increasing diversity might lead to a management team, that get access to important resources (e.g. capital, customers, suppliers). A more diverse board can give better orders to lower manage-ment, because more unique information exist within the group. The market will per-ceive a more diverse board more positive (Carter et al., 2010).
2.2.2 Human capital theory
Human capital theory emphasizes that a TMT can enhance the effectiveness just by choosing a management board, which has various ages, genders, and nationalities. The result is the creation of valuable human capital. More differentiated social networks and better critical group thinking processes are some of the advantageous reasons for board diversity (Carter et al., 2010).
2.2.3 Agency theory
can increase board independence, which in turn is required to monitor the top man-agement to reduce agency issues (Randøy et al, 2006). This might indicate that the quality of monitoring has a positive impact when adding more women to the board. A homogenous board is a direct threat to an independent board, since such TMTs have negative effects like, the creation of tunnel visions or high group pressure (Lückerath-Rovers, 2013).
2.2.4 Upper Echelon Theory
Hambrick & Mason introduced the upper echelon perspective in 1984. The upper eche-lon theory is constructed on the behavioral theory of the firm5 (Cyert & March, 1963),
and proposes that choices of the management are not necessarily following rational motives but are influenced by the natural limitations of managers as human beings. Relevant are the top managers’ “cognitive maps”, which are a general representation of beliefs and concepts held by individuals (Dutton, Fahley & Narayanan, 1983; Axel-rod, 2015). Top Managers have been recognized as meaningful impact determinants on organizational decisions, actions and its outcomes or rather firm performance (Finkelstein & Hambrick, 1996; Huber & Glick, 1993). The upper echelon theory con-tributed to this view which proclaims that the organization is a pure reflection of its top manager teams. Hambrick & Mason (1984) visualize with their upper echelon model a process, where top managers make strategic choices that might have an impact on firm performance (see Figure 1). Human decisions are limited to individual interpreta-tion of certain situainterpreta-tions and top managers find soluinterpreta-tions to problems by considering two different types of characteristics. The first type is their psychological attributes and the second type is their observable demographics (e.g. age, experience, gender, na-tionality).
5 The behavioral theory of the firm strives to connect the economic theory of the firm with sociological
Figure 1: Upper-echelon theory framework (Hambrick & Mason, 1984: 198)
Marimuthu & Kolandaisamy (2009) support the theory and confirm that top manage-ment’s characteristics have an impact on the decisions that they make and succes-sively also on the actions adopted by the firm that they lead. Reasons are demographic characteristics which are connected with different values, past experiences, cognitive bases and perceptions that have direct influence on the decisions making process of top management (Talke, Salomo & Rost, 2010). With these attributes they “construct, rearrange, single out, and demolish many ‘objective’ features of their surroundings” (Weick, 1979: 164). Furthermore, the decision-making authority at senior management allows executive directors much more freedom to influence firm strategy than lower level management (Mintzberg, Ahlstrand & Lampel, 2009).
2.3 Opposing Theories on Board Diversity
Contrary to the Upper Echelon Theory, some theories imply negative diversity effects. For instance, the Process theory indicates that diversity have influence on group pro-cesses, such as communication or conflict, and this influences strategic decisions and performance (Williams & O’Reilly, 1998). Main idea of the Process Theory is built on the Social Categorization Theory (Turner et al., 1987) and the Similar-Attraction-The-ory (Schneider, 1987). According to the Social Categorization TheSimilar-Attraction-The-ory, people favor the interaction with others who belong to the same in-group-members (‘us’ equals trust) instead of the ‘different’ out-group-members (‘they’ equals mistrust and suspicion). So-cial categorization is based on diverse characteristics (e.g. demographics) and the in-terpretation of trusting each other. Of course, it can occur within the group and affect the group interaction negatively, if others are not defined as in-group-members. In the case of group diversity and the fact that people generally prefer to cooperate with indi-viduals they perceive as similarto them, reduced effectiveness of boards is the result. (Dobbin & Jung, 2011). Similarity-attraction theory state that similarity in attitudes, be-liefs, and values will promote interpersonal attraction as well as mutual sympathy, but generate social divisions, which in turn lead to negative performance results for the whole team (Mannix & Neale, 2005). Diversity generates various contrasting views which leads to less effective and more time-consuming decision-making processes (Campbell & Mínguez-Vera, 2009). Based on the presented theories, it can be noted that increasing heterogeneity within TMTs have negative influence on group pro-cesses. Therefore, firm performance should also be affected negatively.
Group loyalty suffers if members do not identify with others. Due to age, gender and racial diversity, conflicts in small groups are increased and decision-making capacity is disturbed (Jackson at al., 2003). Heterogeneous boards are causing high costs, be-cause meetings require more time, various conflicts of interests exist, general misun-derstandings as well as decisions errors occur (Arrow, 1951). Group cohesion is neg-atively affected, whereas members of the minority have no influence so their compan-ionship has no impact on group decisions (Wetphal & Milton, 2000).
MNCs choose their directors for their characteristics.
MNCs choose board members in order to deal with the external environment (Pfeffer, 1972).
Different types of companies choose different levels of director diversity. TMTs prefer other team members who are share a certain similarity, which
af-fects team meetings and team dynamics (Zander, 1979).
Directors from minority groups perceive their outnumbered status as a barrier to their work as director and might even serve different interests than those of the shareholders (Ferreira, 2010).
From the mixed empirical findings it can be assumed that there is a dependence of various factors, such as the size of the company or the board. The following sub-chapters will continue with the analysis of the single effects of some relevant man-agement characteristics. At first, research results about the effect of age diversity of board directors will be presented. Then the effect of gender or rather women quotas of board directors will be explained in detail. Finally, the nationality aspect will be an-alyzed, in order to see if there might be some significant indications on firm profitabil-ity.
2.4 Age Diversity
One personal characteristic of top executives that is relevant for this thesis, is the as-pect of age. How does the age of top management team members affect the perfor-mance of a German MNC? In comparison to the research on gender diversity, neither the age factor has received much attention nor exist the same depth of literature for this issue.
objectives, whilst in teams that had easier tasks, heterogeneity was decreasing effec-tiveness. Age is closely connected with a higher level of experience. Older board mem-bers are better prepared for exceptional situations, have a higher sense of responsi-bility, act more conservative, and thus beneficial for improving firm performance. Vroom & Pahl (1971) emphasize that older top managers are associated with higher level of risk-aversion by increasing their preference of choosing safer options. Antia, Pantzalis & Park (2010) argue that, because of their age, older people would avoid projects with high risk, since they cannot wait for a high-risk project to pay-off. Imme-diate payback is preferred over waiting for the insecure outcome.
On the other side, younger board member input can be seen as more innovative, more open for risky projects or more energetic (Timmerman, 2000). A healthy diversity of both sides is the key to improve firm value. Another study suggests that TMT with young board members are better off than with their older members. A younger board composition tends to be more innovative and more open to monitor its managers (Rose, 2005). Other observations indicate that younger managers are more open-minded for new ideas and risky undertakings with uncertain but positive intended out-comes (Zajac & Westphal, 1994).
Since the empirical study of McIntyre et al. (2007) is one of the few research projects that explains the relationship between age diversity on the board of directors and firm performance, this thesis could contribute the understanding of age differences effect on firm performance. Besides the positive verified effects, excessive diversity is not the desired condition (McIntyre et al., 2007). A company which encourages a board of directors with different ages, benefits from the advantageous impact of an increasing social network.
In summary, older managers are normally more experienced, have strong parental responsibility senses which might result in better corporate performance, whereas young managers are more innovative, have better technical knowledge and are more venturesome. Accordingly, the following hypothesis will be tested:
2.5 Gender Diversity
Regarding the board composition, gender diversity within TMTs forms a major focus of Corporate Governance (CG) research. This focus is derived from a long-lasting politi-cal debate on gender diversity in TMTs. Main measurement method of firm perfor-mance, for this research field is Tobin’s Q (Adams & Ferreira, 2009, Fauzi & Locke, 2012). Most of previous studies have been focusing on countries, such as Canada, Scandinavia, Spain, US and UK (Francoeur et al., 2008, Torchia et al., 2011; Campbell & Mínguez-Vera, 2008; Dobbin & Jung, 2011; Brammer et al. 2009). Gender diversity is also relevant to the European Union, resulting in a directive proposal from 2012 in order to promote the implementation of a fixed women quota of 40% within TMTs. Within the European Union, a few countries like Norway, the Netherlands, France, Italy and other countries already have corporate gender quotas in place. Germany’s effort to improve female quota lags behind6. Besides the good effort of mandatory quotas,
Ahern & Dittmar indicate that the introduced mandatory 40% quota in Norway caused a significant drop in the stock price and a large decline of Tobin’s Q over a few years. Quotas are forcing firms to put women on boards, even though they may not be quali-fied enough for such a position and consequently a deterioration in operating perfor-mance might be the consequence (Matsa & Miller, 2011).
The phenomenon of under-representation of women at upper levels of hierarchy is known as “glass ceiling-problem” (Arfken et al., 2004), which is based on a predomi-nant male business culture, disparate career management, and pressure from career as well as family goals (Vinnicombe & Singh, 2003). A study from Daily et al. (1999) about the Fortune 500 MNCs suggest that female board memberships ratio have im-proved but CEO positions are still more likely to be male. Another research result indi-cates that women are more likely to tend to less hierarchical and more cooperative leadership styles, which is in favor of the communication and freedom of expression (Eagly & Johnson, 1990).
Despite the increasing relevance and usefulness, consistent empirical research results do not exist. It is rather unclear, if gender diversity has more positive or negative effects on firm performance. Positive effects on firm performance are confirmed by authors, such as Adler, (2001), Carter et al. (2003), Lückerath-Rovers (2013), Ujunwa et al.
(2012). Negative effects are proven by Adams & Ferreira (2009), Dobbin & Jung (2011), Fauzi & Locke (2012), Shrader et al. (1997) and Krishnan & Parsons (2008). Missing correlations also have been detected (Carter et al., 2010; Erhardt et al., 2003; Jhunjhunwala & Mishra, 2012; Sun et al., 2011).
According to Ibrahim and Angelidis (1994), differences among female and male board members are quite significant. They argue that male members tend to be more con-cerned about economic performance, whereas female board members are more ori-ented toward philanthropic aspects and corporate responsibility. A MNC can benefit from both genders. Female board members can contribute with various new perspec-tives, contacts or information to the board. Taking into account the existing empirical findings about gender diversity, the hypothesis on this issue is the following:
H2: Female board members have positive impact on the companies’ profitability.
2.6 National Diversity
The last determinant which will be tested is the influence of manager’s nationality on firm performance. There is no adequate amount of research available, so consequently this thesis will try to contribute to the existing research.
The study from Masulis et al. (2011) focused on the effect of foreign directors on cor-porate governance and firm performance. Scientific insights were that MNC with for-eign independent directors (FIDs) make much better cross-border acquisitions when the targets are from the home countries. Another finding is that FIDs do not attend board meetings constantly. In addition, they state that MNCs with FIDs are more likely to commit misreporting and that they are generally associated with poorer perfor-mance. Board of directors have two major roles, which are advising and monitoring (Masulis et al., 2011). This means that besides the FID’s capability to contribute valu-able advisory and international expertise to their parent company, they are mostly less effective in monitoring, which can deteriorate the whole boards’ monitoring of manage-ment. There are various reasons for the foreign directors’ ineffectiveness. One obvious reason is the geographical distance to headquarter. It is not only costly but also time-consuming trying to attend all meetings. This observation is supported by Lerner (1995) who emphasizes that venture capitalists are unwilling to be on boards of MNCs that are geographical distant. Directors who are not located at the companies’ HQ, do not have the required amount of information about the current situation in the company, because they are literally cutoff from the resource network7 (Coval & Moskowitz, 1999).
Another research view indicates that national diversity can result in negative effects and outcomes (Milliken & Martins, 1996).
Adams, Hermalin & Weisbach (2010) prove that the advisory role of foreign directors is beneficial for boards. With foreign directors, MNCs are able to establish networks with foreign contacts, they can use their valuable advice for expanding efforts (e.g. knowledge about regional regulations, cultural & social norms, market preferences, political situation, etc.) and gain new business insights (Maznevski, 1994). Ruigrok et al. (2007) argue that the effectiveness of the boards will increase, due to the pure presence of foreigners. The authors justify their finding that the advantages will out-weigh the negative effects when contrasting norms, values, different perspectives and knowledge will be combined in order to profit from national diversity. Adams et al. (2010) emphasize that it is not easy to measure the effect of board diversity on firm
performance, because most of variables of interest are commonly endogenous and there are no specific instruments that can exclude all of the endogeneity.
Previous research studies did not find significant links between board diversity and performance, but all of them agree that foreign directors are beneficial in form of giving knowledge, information and helping MNCs to get familiar with foreign markets (e.g. Adams et al., 2010, Carter et al, 2010). The question that arises is, if the positive side is convincing enough to compensate for other inefficiencies caused by the foreign di-rectors. In theory, foreign directors are advantageous for MNCs, but evidence is miss-ing from research about the positive relationship between firm performance and the influence of foreign directors on the board. This empirical research will try to get an-swers whether foreign directors in German MNCs have direct impact on the financial performance. Therefore, the following hypothesis will be tested:
H3: National diversity of TMTs has a negative impact on the companies’ profitability.
2.7 Endogeneity and Reverse Causality
This kind of study generates endogeneity and reverse causality problems. Character-istics of TMT members and firm performance are jointly endogenous (Hermalin & Weisbach, 2003). Board diversity can positively influence performance, but due to the causality, it is also possible that a negative outcome can occur. Previous results indi-cate that omitted unobservable MNC characteristics can trigger endogeneity. More progressive companies have better governance and a higher number of female direc-tors (Adams & Ferreira, 2009). Through the assumption that corporate culture is time-invariant, fixed effects that control for culture will help to show if biased results exist. Better performing MNCs have more resources in order to enhance board diversity. They have increased willingness and ability to employ women, to attract more women, and this may be the reason why these companies have more diverse boards (Smith et al., 2005). The instrumental variables method is useful to address reverse causality issues and results highlight negative relationship between the increasing amounts of women in firms that are well-governed (Adams & Ferreira, 2009).
lagged variables model and add firm and time fixed effects. Lagged variables are help-ful since they assume the condition that board diversity effects on firm performance will appear over time, and not immediately. Hence, lagged values of dependent varia-bles will be used within this research model, too. In order to address endogeneity, firm performance will be measured as the average of ROA, ROE and Tobin’s Q and all of the variables are lagged by one period. The time-lagged structure, which will be adopted for measuring the three dependent performance variables, allows to reduce potential biases of single year outliers (Johnson & Greening, 1999), and also to en-hance causality in the studied relationship (Hambrick, 2007). To confront the reverse causality issue, a 3SLS estimation is mostly used but no significant relationship be-tween board diversity and firm performance exist (Carter et al., 2010).
3. Methodology
3.1 Data Collection
In order to analyze the relationship between board diversity and corporate perfor-mance, information is gathered for the years 2012, 2013, 2014 on the largest 107 pub-licly traded German companies, which are listed in the blue-chip indices DAX308,
MDAX9, TecDAX1011. Overall, 321 observations are considered for this research. The
107 largest listed MNCs were selected due to the availability of data and the case that these three indices blue chips, mid-caps and small caps of different industries. In ac-cordance with the German Stock Corporation Act (AktG) (Wirth et al., 2004), which states that a German TMT is specified by the company’s executive board (sec. 76 §1 AktG), which is elected by the supervisory board (sec. 84 AktG), both boards will be included. The data about the characteristics (age, gender & nationality) of the directors as well as the accounting and stock data is obtained from the database of ORBIS,
8 The DAX30 is frequently defined as the benchmark index for the German equity market. The DAX
index measures the share performance of the 30 largest German companies in term of exchange turn-over and market capitalization. Hence, it is an established indicator for the performance of the German company as a whole (Deutsche Boerse Group, 2012).
9 The MDAX contains the 50 largest companies (also known as mid-caps) from classic sectors in Prime
Standard ranking directly below the DAX30 shares. The index primarily includes share from the sectors chemistry, finance, pharmaceuticals and machinery (Deutsche Boerse Group, 2012).
10 The TecDAX includes the price development of the 30 largest technology share in Prime Standard
below DAX30 (Deutsche Boerse Group, 2012).
provided by Bureau van Dijk. ORBIS is a database which contains information about directors and their biographies of listed companies and financial data for over 79 million private and public MNCs worldwide. Where data on a board member are missing, an-nual reports of companies and Internet sources12 are used to fill the gaps. For
meas-urement of firm performance, ROA, ROE and Tobin’s Q are used. In order to increase normality distribution, some variables are corrected by using a log function. Regression analysis is performed by using IBM SPSS13.
3.2 Excluded Firms
Banks, insurance- and investment MNCs are completely excluded from the sample since they feature quite specific accounting, which cause scientifically research issues for the data collection and especially for the calculation of Tobin’s Q ratio. This ap-proach is similar to previous studies (e.g. Marinova et al., 2010). In total, 14 firms from the aforementioned industries are excluded from this sample. Moreover, two other firms are excluded as well14, since no sufficient data was available about past
mana-gerial information for the required time period. The new thinned sample consists of 9115
German MNCs.
3.3 Variables
3.3.1 Dependent Variables
The dependent variable of interest is firm performance, which is defined as a firm’s financial performance as illustrated in annual financial reports (Certo et al., 2006). It should be noted that there is no ultimate set of financial ratios that can be used to determine firm performance (Otley, 2007). For this thesis, performance is measured by using three different financial ratios to get a better credibility: Return on Assets (ROA), Return on Equity (ROE) and Tobin’s Q. These chosen financial ratios are usu-ally used as a measure of financial performance but all of them deliver different results.
12 e.g. companies’ websites
Return on Assets (ROA) is a ratio of net income to book value of total assets at the end of the year (Adams, Almeida & Ferreira, 2005). It is an indication about the profit-ability of a firm and efficiency of the management compared to its total assets. ROA can be determined by annual net income divided by the book value of total assets at the end of the year (Erhardt et al., 2003). Megginson et al. (2007) argue that profitability ratios are among the most carefully observed and broadly quoted financial ratios. Moreover, the authors add that ROA measures the general effectiveness of the man-agement by using firm assets in order to generate returns for common stockholders.
Return on Equity (ROE) is “the amount of net income returned as a percentage of shareholders equity” (Khatab et al., 2011, p.40). It measures a firm’s profitability by revealing the generated profit, which is based on the invested money by the share-holders. ROE can be calculated as profit after tax divided by total equity shares in issue.
Tobin’s Q is the predominant measuring method used in corporate governance re-search (Rose, 2007). This method is a forward-looking measure of expected long-run firm performance (Bozec, 2010). It is the ratio of the stock market value of a company to the replacement value of its assets, which is recognized as the best measure of a company’s market value (Berger & Ofek, 1995). Tobin’s Q is determined through the sum of market value of equity plus book value of total debt divided by the book value of the assets (Chung & Pruitt, 1994, p.71). Whereas the calculated result is greater than one, investors can expect that the company is capable of creating value by using available resources in an effective way. Consequently, MNCs with a Tobin’s Q ratio of less than one are regarded as ineffective resource users (Campbell & Mínguez-Vera, 2008). In order to reduce skewness, a natural logarithm will be used.
The Tobin’s Q formula is the following:
Tobin’s Q = (Market value of equity + Book value of assets – Book value of equity) / Book value of assets
3.3.2 Independent Variables
being observable and in comparison to other diversity characteristics, are easy to col-lect. This explains why the majority of studies included them, too (Carter et al., 2003). Age diversity will be measured as the average age of all board members. Gender di-versity will be measured by the natural logarithm of the percentage of female board members, which can be acquired by dividing the total amount of women on the boards by the total amount of directors. National board diversity will be measured by the nat-ural logarithm of percentage of foreign board directors, which can be gathered by di-viding the total amount of foreign (non-German) board members by the total amount of directors. The application of the Blau Index for age, gender and national diversity will help to check for the robustness of the findings. For the purpose of avoiding en-dogeneity and reverse causality, lagged independent variables (one lag = one year) are used for this analysis.
3.3.3 Control Variables
The control variables of this thesis are based on prior studies that should be included, due to the fact that they have significant relationship with the dependent variables.
Firm size is a substantial variable for market returns (French & Fama, 1992) and is mainly used in order to investigate the relationship between board diversity and finan-cial firm performance (Carter et al., 2003; Campbell & Mínguez-Vera, 2008; Koch & McGrath, 1996). It can be expected that firm size will have a positive impact on financial firm performance as well as board diversity, because large firms also have a larger scale of operations and organizational settings (Koch & McGrath, 1996). Further rea-sons are that larger companies are more in the spotlight of the public’s attention, have to act as role models, and have to comply with the societal pressure for board diversity (Marinova et al., 2010). Similar to the research measures of Baysinger & Hoskisson (1989), this thesis will measure firm size by firm’s total assets. In order to minimize the effect of extreme values, a natural logarithm of total assets will be used (Carpenter and Fredrickson, 2001).
(Webber & Donahue, 2001). According to Blau (1977), board size has a tremendous influence on the team’s level of diversity, because larger groups are more likely to feature dissimilarity (Wiersema & Bantel, 1992). A positive relationship between board size and Tobin’s Q is proven by Jackling & Johl (2009), whereas larger boards also have a greater pool of knowledge and excelling information. In contrast, a negative relationship between board size and Tobin’s Q is found out by Yermack (1996). An-other negative effect could be that larger groups have an increased possibility to suffer from problems of communication, coordination, control, and lack of social integration with the outcome of declined performance (Smith et al., 1994). Board size will be meas-ured by taking the natural logarithm of the number of directors.
Firm age measures the total number of years since the MNC’s incorporation. It is proven that firm age is an influential determinant in examining the relationship between TMT diversity and financial performance (Low et. al, 2015). Previous studies indicate that older MNCs are much more stable as well as familiar with misfortunes. Their in-creased experience-pool allows MNCs to deal with uncertain circumstances (Jackling & Johl, 2009; Dezsö & Ross, 2012). Thus, firm age can enhance firm performance.
Industry dummies will be used as a control variable, too. For the regression analysis, it is necessary to make distinctions between industries, because industries differ in market volume and dynamic (Adams, Almeida & Ferreira, 2005). Two major industry categories are created (derived from the SIC16 codes: OSHA, 2010). Industry dummy
1 is the largest, since it includes firms from the agriculture, mining, construction or manufacturing industry. The second industry dummy includes firms from sectors, such as transportation, communications, electric, gas, wholesale- or retail trade industry. A third industry dummy was created to include the finance, insurance, services or public administration, but it was excluded due to the decision in chapter 3.2. Industry dummy 1 is the first group and will be omitted, because it is the reference or default category against which the industry effects of the other categories are assessed. The results can be interpreted as the difference between each category and this omitted category.
For a better overview, Appendix 1 contains the summary of all variable definitions.
3.4 Method
The regression method which is used for this study, is the Ordinary Least Squares (OLS) regression analysis. This is in line with previous studies (e.g. Dezsö & Ross, 2012). To confirm the use of OLS, it is necessary to control for the autocorrelation of the error terms between the independent and the dependent variables, by using the Durbin-Watson test. The Durbin-Watson statistic include range values from 0 to 4. Val-ues near 2 emphasize that no autocorrelation exists, valVal-ues towards 0 show positive autocorrelation, and values near 4 imply negative autocorrelation (Montgomery et al., 2015). By using the Durbin-Watson statistic, it turned out that the error terms are not correlating with each other, since the statistics are around the value of 2 (see Appendix 4). Therefore, only OLS will be used, instead of 2SLS.
There is plenty of literature that uses different methods for measuring TMT diversity. Either panel data or cross-sectional data is used. For this thesis, panel data and con-sequently a panel regression of variables is used. The time period is for 3 years, from 2012-2014. A linear combination of independent variables that best forecast a depend-ent variables is called regression model (Sekaran & Bougie, 2010; Zikmund, 2013). A linear multiple regression model / equation is as follows:
Y = b0 + b1X1 + b2X2 + b3X3 + b4X4 + b5X5 + e Y = dependent variable
X = independent variable
b0= constant, which equals to the mean if slope coefficients are zero b = slope coefficient associated with each independent variable e = random error or residual
Based on the equation, the regression equation for this study can be formulated as follows:
For the purpose of robustness tests, the frequently used measure for heterogeneity of variables, Blau’s Index will be applied. Blau’s Index is recognized as an ideal measure of diversity to include variations within a group (Miller & Triana, 2009). The Equation which will be used it the following:
Variety can take k = 1, …, K possible categories and p stands for the proportion of board members in the kth category. An increase of board diversity equals an increase
of Blau’s Index17.
The equation including the Blau index is the following:
Firm Performance = Constant + b1*BlauIndexAge + b2*BlauIndexGender + b3* BlauIn-dexNationality + b4*FirmSize + b5*BoardSize + b6*FirmAge + b7*IndustryDummy + e
For this thesis, a random effects method is used as the regression model in order to test the 3 hypotheses. A fixed-effects model is not considered to check for fixed unob-servable characteristics such as specific firm culture, because the independent varia-bles (age, gender, nationality) feature very little variation over 3 years. First, directors normally do not change over the considered 3 year period. Second, even if they do, new elected directors tend to have homogenous characteristics. For instance, if the former director had a German nationality, it is very probable that the new director might have the same nationality, too. Accordingly, fixed effects model is not appropriate to analyze time-invariant causes of the dependent variables, which is the case in this thesis. Instead, random-effects model is used, because the rationale behind this model is that some differences across entities might have some impact on the dependent variables. Random effects model is based on the main assumption that independent variables are uncorrelated with the error term (Bartels, 2008). This research approach allows to get insights if firms with higher diversity have better performance ratio or not. For the estimates all observations are used in a pooled cross-sectional database.
The three dependent variables, ROA, ROE and Tobin’s Q are used, because this will lead to an increase of more credible findings. Lagged independent variables will be
included in this model in order to avoid endogeneity problems, which are based on reverse causality. This approach is consistent with the research of Carter et al. (2010). They will be used, since the main assumption is that board diversity has an impact on firm performance after certain time period. According to the literature, there is no pre-determined length of time, which is required for effects. Thus, for this thesis it is deter-mined to check the findings for lags of one year. This allows to understand if board diversity might change the firm performance after the lag, or it might be the case that the change is caused the other way around.
For the purpose to check for robustness of the findings, diversity indices are con-structed. To measure diversity, Blau index will be applied (Blau, 1977). It is a commonly used measure in the field of demographic research. With the help of the Blau index, it will be possible to determine the variation in categorical data. The regression equation is almost the same, besides that the average age of board members, percentage of women and foreign directors will be replaced with age diversity, gender diversity and national diversity, measured by the Blau index formula.
3.5 Multicollinearity
4. Findings and Interpretation
4.1 Descriptive Statistics
Table 2 provides the descriptive statistics of all the variables which are used in this study. From the 91 German listed firms, there are 1656 directors in the sample. The number of observations is 273, confirming that there is no missing data.
The performance measure ROA has a relatively high average of 5,2366%, indicating that the average German listed firms perform well. The average value for ROE is equal to 13,7092%, which means that shareholders earn around 13,71% on their invest-ments in firms. Mean value for Tobin’s Q is 0,92, slightly below 1, which means that the market value of the firm is lower than the book value of the assets (‘undervalued’). Mean board age is 57 with a range from the youngest with 26 years to the oldest with 87 years. Average percentage of female board members is 18% with a standard devi-ation of 0,08%, 6 firms have no female board members at all, and one firm has a max-imum of 35%. The Blau index diversity indicates that a score of zero stands for a totally homogenous team and if the value reaches one a total heterogeneous team exist. Age diversity has a score of 0,52, female diversity 0,29 and nationality diversity 0,21.
Firm size has a range from € 104 million18 to € 328 billion19 and an average of € 20
billion. Average number of directors in the boardroom is 18 with a minimum of 6 board members and a maximum boardroom of 35 directors. The results of firm age point out that the oldest firm in the sample is 348 years20 old and the youngest only 1 year21 old.
While calculating the natural logarithm for some of the variables, a certain peculiarity occurred. The majority of the Tobin’s Q results are less than one but greater than zero, which explains the negative tendency of the natural logarithm results. That also applies to the calculation of the natural logarithm for gender and foreigners on the board, since the values are indeed very small.
Variable
Observa-tions Mean Std. Deviation Min Max
Dependent Variables ROA 273 5,2366 6,0084 -17,0500 28,5900 ROE 273 13,7092 15,2028 -38,1200 74,1800 Ln Tobin's Q 273 -0,1843 0,3732 -1,2800 0,8000 Tobin's Q 273 (0,9200) (0,9752) (0) (6,2777) Independent Variables Board Age 273 56,8267 8,7159 26,0000 87,0000 Ln Women 273 -0,6763 0,2303 -1,08 0 % Women 273 (0,1862) (0,0792) (0) (0,3500) Ln Foreign Directors 273 -0,6234 0,4531 -1,5100 0 % Foreign Directors 273 (0,1582) (0,1638) (0) (0,7143)
Blau Index Age Diversity 273 0,5211 0,0905 0,3000 0,6800
Blau Index Gender Diversity 273 0,2906 0,1066 0 0,4527
Blau Index National Diversity 273 0,2133 0,1681 0 0,5000
Control Variables Ln Total Assets 273 6,6354 0,8101 5,0200 8,5200 Total Assets 273 (20.990.853) (46.775.318,74) (104.171,00) (328.395.333,00) Ln Board Size 273 1,2139 0,2093 0,7800 1,5400 Board Size 273 (18,1978) (7,9404) (6,0000) (35,0000) Firm Age 273 73,5055 60,2988 1,0000 348,0000 Industry Dummy 1 273 0,6264 0,4865 0 1,0000 Industry Dummy 2 273 0,3736 0,4865 0 1,0000
Table 1: Statistics of Variables
4.2 Correlation
firm age seems to influence board size too. Other correlations between the different independent variables are low or there is even no effect. None of the correlations ex-ceed 0,9 which is the harmful level of correlation (Tabachnick & Fidell, 1996).
(1) ROA (2) ROE (3) LnTQ (4) LnWo (5) LnFD (6) BoAg (7) LnFS (8) LnBS (9) FiAg (10) InD1 (11) InD2 (1) ROA 1,000 (2) ROE 0,718 1,000 (3) LnTQ 0,462 0,246 1,000 (4) LnWo 0,226 0,116 0,055 1,000 (5) LnFD 0,101 0,116 0,126 0,116 1,000 (6) BoAg -0,035 0,008 -0,029 -0,156 -0,165 1,000 (7) LnFS -0,239 -0,103 -0,464 -0,490 -0,301 0,124 1,000 (8) LnBS -0,138 -0,079 -0,371 -0,109 -0,412 -0,020 0,776 1,000 (9) FiAg -0,047 -0,148 -0,198 -0,056 -0,278 -0,470 0,413 0,388 1,000 (10) InD1 -0,078 -0,040 -0,048 -0,086 -0,195 0,004 0,246 0,398 0,295 1,000 (11) InD2 0,078 0,040 0,048 0,086 0,195 -0,004 -0,246 -0,398 -0,295 -1,000 1,000
Table 2: Cross-correlation Table
4.3 Regression
In order to test the hypotheses, an ordinary least squares regression (OLS) analysis is performed and presented in Table 4. In column 1 are the results of regression done with ROA as the dependent variable. Consequently, columns 2 includes ROE and col-umn 3 Tobin’s Q as dependent variable. Standard errors are reported in parentheses. Industry dummy 1 is a reference category. The first 3 variables are the independent variables board age, percentage of women and percentage of foreign directors. As control variables, firm size, board size, firm age and industry dummies are used.
Legend:
Variables (1.) (2.) (3.)
ROA ROE Tobin's Q
Board Age 0,052 0,089 0,001 (0,078) (0,207) (0,005) Ln % of Women 6,109** 7,449 0,064 (2,747) (7,270) (0,161) Ln % of Foreign Directors 0,905 2,974 -0,017 (1,551) (4,104) (0,091)
Natural Log of Total Assets -3,083** -1,965 -0,200***
(1,293) (3,422) (0,076)
Natural Log of Board Size 6,597 6,594 -0,117
(5,338) (14,129) (0,312) Firm Age 0,009 -0,027 0 (0,012) (0,031) (0,001) Industry Dummy 2 0,743 -0,275 -0,069 (1,419) (3,757) (0,083) Constant 18,458*** 22,638 1,289*** (6,743) (17,848) (0,394) Observations 273 273 273 R² 0,125 0,43 0,225 Adjusted R² 0,052 -0,038 0,159 F 1,701 0,534 3,437***
***. Correlation is significant at the 0.01 level (2-tailed). **. Correlation is significant at the 0.05 level (2-tailed).
*. Correlation is significant at the 0.10 level (2-tailed).
Table 3: Regression Analysis
better financial performance than bigger and older MNCS during 2012-2014. Hypoth-esis 2 is therefore confirmed. Model 1 has an insignificant F-value, which indicates that the results need to be considered with this consideration.
ROE does not have any significant coefficient. R-squared of this regression is 0,430 which shows that 43% of variation in ROE is explained by independent variables. The relationship is not significant at any level, thus the null hypothesis cannot be rejected. None of the 3 hypotheses can be confirmed either. Board diversity does not have any kind of effect on the ROE of a firm. This is confirming the findings of Shrader et al. (1997), where they argue that women on boards do not have much impact on the firm, since the “critical mass” of women is missing. The average quota of 18% of female board members in German boards supports this, too. Upper Echelon theory, on which the constructed 3 hypotheses are based, is not confirmed, since no impact can be noticed. Model 2 has an insignificant F-value, so it has to be considered with this con-sideration, too.
Last performance measure is the market-based Tobin’s Q. The link between age, gen-der or national diversity and Tobin’s Q is not significant at any level. This is not in line with McIntyre et al. 2007, who did a study about the impact of age diversity in Canadian firms on firm performance. Their findings highlight a significant positive relationship and is very much a contrast to the results of this thesis, since only insignificant links between age diversity and firm performance can be reported. An explanation could be that they used data for only 1 year (2001) and that in this year, the stock market prices were declining. A one year period is no sufficient to obtain reliable results. In addition, they did not use any other dependent variables but Tobin’s Q, which decreases the reliability of results. According to the F-value, model 3 is significant.
4.4 Robustness Checks
Variables (1.) (2.) (3.)
ROA ROE Tobin's Q
Blau Index Age -9,761 -36,496** -0,620
(7,189) (18,017) (0,412)
Blau Index Gender -16,272** -54,961*** 0,273
(6,734) (16,875) (0,385)
Blau Index Nationality -1,19 0,765 0,291
(3,777) (9,465) (0,216) NaturalLogofTotalAssets -2,996** -2,856 -0,235*** (1,287) (3,226) (0,074) NaturalLogofBoardSize 9,729* 21,668 -0,098 (5,436) (13,623) (0,311) Firm Age 0,007 -0,029 0 (0,012) (0,029) (0,001) Industry Dummy 2 1,444 2,200 -0,034 (1,458) (3,655) (0,083) Constant 22,309*** 42,470** 1,681*** (6,770) (16,965) (0,388) Observations 273 273 273 R² 0,14 0,156 0,269 Adjusted R² 0,067 0,085 0,208 F 1,926* 2,193** 4,370***
***. Correlation is significant at the 0.01 level (2-tailed). **. Correlation is significant at the 0.05 level (2-tailed).
*. Correlation is significant at the 0.10 level (2-tailed).
Table 4: Test of Robustness
support findings of Shrader et al. (1997), since they could not confirm that higher per-centages of female managers on the TMT are associated with a higher ROA. Their explanation is that their study includes only 4,5% of women on boards and that there are no female CEOs. In comparison, the overall percentage of women in German boards is 18% and therefore also low. Additionally, no female CEOs can be confirmed either. Bilimoria & Piderit (1994) suggest that female directors are slightly disadvan-taged by the type of board committee assignments that have less instrumental influ-ence on the firm. It can be assumed that women have not been in TMTs long enough to have significant impact. Model 1 has a significant fit, which can be observed in the F-statistic (p: < 0,01).
ROE has a negative significant (5%) coefficient with Blau Index Age (-36,496). This result is in line with the findings of previous studies indicating a negative relation be-tween age diversity and corporate performance (Richard & Shelor, 2002; Simons et al., 1999). An increase of age diversity results in negative ROE outcome in German boards. This could imply that older directors are more experienced and lead to a more trustworthy interaction with strategic choices. Thus, H1 is rejected as it is not possible to identify a positive influence of age diversity on corporate performance. Another neg-ative significant (1%) coefficient (-54,961) can be observed with Blau Index Gender. This can be interpreted that financial weak firms have decreased gender diversity. Op-posite to the negative finding, Catalyst (2004) found a positive impact on ROE and Total Return to Shareholders (TSR). Hence, H2 is not confirmed, since it has negative influence. Model 2 shows a significant F-statistic (p: < 0,05).
According to column 3, there is no significant relationship between board diversity and Tobin’s Q. All hypotheses cannot be confirmed. Results are in line with studies with a relation focus on Tobin’s Q and gender (Rose, 2007; Carter et al., 2010). Only one negative significant relation can be observed with the firm size. The model 3 is signifi-cant as indicated by the signifisignifi-cant F-value (p: < 0,01).
Q. This implies that hypothesis 1 and 2, stating that age and gender diversity are pos-itively related to financial performance are not supported in this sample. Also hypothe-sis 3, stating that nationality diversity is negatively related to performance, cannot con-firmed. The three alternative hypotheses cannot be accepted since the null hypothesis is not rejected.
5. Concluding Remarks
5.1 Conclusion & Discussion
Previous empirical CG research implies diverging results regarding the impact of board diversity on corporate performance. This research examined the relationship between board diversity, in terms of age dispersion, gender composition, nationality diversity, and the financial firm performance measures return on assets, return on equity, and Tobin’s Q. The empirical analysis was conducted by 91 listed firms (DAX, TecDAX, and MDAX) for the business years, 2012, 2013, and 2014, while controlling for factors which may influence the independent or dependent variables. Two different measures were applied for every independent variable. Age, gender and nationality were meas-ured by the percentage share and with the Blau diversity index. Mostly negative effects of board diversity characteristics can be identified, especially regarding age and gen-der diversity. Robustness checks showed that the results are not reliable, consequently the findings have to be treated with caution. The following three hypotheses were tested:
H1: Age diversity of top management teams has a positive impact on profitability.
H2: Female board members have positive impact on the companies’ profitability. H3: National diversity of TMTs has a negative impact on the companies’ profitability. Hypothesis 2 is the only alternative hypothesis which could get confirmed during the analysis process, but had to be rejected after the robustness test. Gender diversity in relation to ROA is the only significant result with a positive effect. After the robustness test, by using the Blau-index calculation, the results changed drastically. Significant relationship was still given, but now with a negative effect on ROA exist. Apart from that, negative but significant effects can be observed for Blau Index Age and Blau
model results, the robustness test findings also indicate the lack of reliability. Hence, the three alternative hypotheses cannot be accepted, because the null hypothesis was not rejected.
A negative effect of Blau Index Age diversity of board directors on corporate perfor-mance could be observed through this research. In contrast, McIntyre et al. (2007) found a positive relation in the age of board members and firm performance. Their study contains data for only one year, which is neither sufficient for long-term effects nor statistically reliable either. Reason could be that older TMTs provide counsel and wisdom to the executive managers but do not have the young spirit, have no fresh ideas, and are not dynamic.
Furthermore, the results are consistent with the study of Adams & Ferreira (2009), because they do not find significant evidence between gender diversity and MNCs’ performance, measured by ROA and Tobin’s Q. Also Farrel & Hersch (2005) did not get significant results when analyzing the effect of gender on ROA.
National diversity, measured by the percentage of foreign directors on TMTs, has no significant diversity effect on any of the chosen performance indicators. This underlines the findings of Randøy et al., (2006.). When calculated with the Blau index, national diversity presents no significant relationship to firm performance. This is consistent with similar studies (Adams et al., 2009; Carter et al., 2010). Also Carpenter (2002) slightly supports the findings, since the author finds only a weak relationship between TMT nationality and performance.
summary, it can therefore be said concerning the question of the influence of board diversity on financial firm performance in German firms, that this work cannot confirm a positive impact. The question remains ambiguous.
5.2 Managerial implications
In order to give some advice in form of managerial implications, MNCs should retain a high level of expertise and skill. Moreover, companies should set a target of TMT di-versity in regards to whatever is realistic of their requirements. Common sense and good judgement should help to focus on a balance between women and men, rather than the simple fact of the presence of female board members. This also applies to dispersion of age, nationalities, expertise or skills. Required qualities have to be deter-mined before appointing a younger director. While it is crucial to have diversity of ex-periences, skills and talent, it is even more essential to actually apply those (Wan & Ong, 2005).
5.3 Limitations & Avenues for Future Research
This thesis is not without limitations. First, this study analyzes only the demographic diversity of board directors. Characteristics such as education, diversity of languages, knowledge, skills, and international experience can be relevant for corporate govern-ance too, but are left out due to the fact, that these factors are not easy to obtain. Second, the analyzed 3-year period might be too short to consider medium or long-term outcomes. Board directors normally stay on the board for at least 4 years, which means that changes of directors and its resulting impact on performance could not be measured. Third, the sample used in this thesis consists of only large German compa-nies. To measure diversity, it is necessary to have access to detailed information about top managers. It is easier to gather the desired information from larger firms than for smaller companies. However, the generalizability of the results is limited, due to the lack of knowledge about the relationship between diversity and firm performance in smaller firms.
References
Adams, R. B., Almeida, H., Ferreira, D. (2005). Powerful CEOs and Their Impact on Corporate Performance. The Review of Financial Studies 18, 1403-1432.
Adams, R. B., Almeida, H., Ferreira, D. (2009). Understanding the relationship be-tween founder-CEO's and firm performance. Journal of Empirical Finance 16, 136-150.
Adams, R.B., D. Ferreira (2009). Women in the boardroom and their impact on gov-ernance and performance, Journal of Financial Economics, 94, pp. 291-309.
Adams, R. B., Hermalin, B. E., & Weisbach, M. S. (2010). The role of boards of direc-tors in corporate governance: A conceptual framework and survey. Journal of
Eco-nomic Literature, 48(1), 58-107.
Adler, R.D. (2001). Women in the executive suite correlate to high profits, Harvard
Business Review, 79 (10), pp. 1-10.
Antia, M., Pantzalis, C., Park, J. (2010). CEO decision horizon and firm performance: An empirical investigation, Journal of Corporate Finance 16, 288–301.
Arrow, K. J., 1951, Social choice and individual values, New York: Wiley.
Arfken, D. E., Bellar, S. L., & Helms, M. M. (2004). The ultimate glass ceiling revisited: The presence of women on corporate boards. Journal of Business ethics, 50(2), 177-186.
Ahern, K. R., & Dittmar, A. K. (2012). The changing of the boards: The impact on firm valuation of mandated female board representation. Quarterly Journal of Economics,
127(1), 137-197.
Bartels, B. (2008). Beyond "fixed versus random effects": a framework for improving substantive and statistical analysis of panel, time-series cross-sectional, and multilevel data. The Society for Political Methodology, 1-43.
Baysinger, B., & Hoskisson, R. E. (1989). Diversification strategy and R&D intensity in multiproduct firms. Academy of Management journal, 32(2), 310-332.
Berger, P. G., & Ofek, E. (1995). Diversification's effect on firm value. Journal of
finan-cial economics, 37(1), 39-65.
Berle, A.A., Means, G.C. (1932). The Modern Corporation and Private Property, New York: McMillian.
Bertrand, M., Black, S. E., Jensen, S., & Lleras-Muney, A. (2014). Breaking the glass
ceiling? The effect of board quotas on female labor market outcomes in Norway (No.
w20256). National Bureau of Economic Research.
Bilimoria, D., & Piderit, S. K. (1994). Board committee membership: Effects of sex-based bias. Academy of management journal, 37(6), 1453-1477.
Blau, P. M. (1977). Inequality and heterogeneity: A primitive theory of social structure (Vol. 7). New York: Free Press.
Bohren, O., & Strom, R. (2007). Aligned, Informed, and Decisive: Characteristics of Value-Creating Boards. EFA Ljubljana Meetings Paper.
Bøhren, Ø., & Strøm, R. Ø. (2007). Aligned, informed, and decisive: Characteristics of value-creating boards. In EFA 2007 Ljubljana Meetings Paper.
Bozec, R., Dia, M., & Bozec, Y. (2010). Governance–performance relationship: a re‐ examination using technical efficiency measures. British Journal of Management,
