1
R.J.F. Kamphuis
S1686682 Supervisor: dr. B.J.W. Pennink
27-08-2012 Second supervisor: drs. H.A. Ritsema
Master’s thesis
Summary
This research investigates if diversity within the executive team of directors has an influence
on organizational performance, tested on observations from organizations located in the west
of Europe during the period 2007 to 2009. Diversity within executive team is calculated using
the equally divided average of the index of Blau on the following variables: Age, gender,
national background, education and knowledge. The outcomes indicated no rise in executive
team diversity through the chosen time frame. Specifically female and non-national directors
are still underrepresented compared to male and national directors within the Blau index
calculations. Furthermore, the women and non-nationals occupying the executive teams held
less influential positions. The impact from diversity on organizational performance is defined
by the dependent variable Tobin’s Q. The impact of the Blau index on organizational
performance measured by Tobin’s Q was a parabolic shaped graph which indicates the
relation between executive team diversity and organizational performance has an optimum.
2
Content
Introduction...3
Literature review………4
The concept of diversity………..6
The possibility of curvilinearity………..7
A range of outcomes………...9 Methodology………..15 Data……….……...17 Description………17 Variables………...19 Results...21
Development of the diverse executive team...22
Specific positions within the executive team...24
Diverse executive teams and their organizations performance...26
Conclusion...31
Discussion & Limitations...32
The optimal executive team diversity………..…...………….………..32
Limitations……….………33
References...35
Appendix 1: Codes...39
Appendix 2: Country specific executive team diversity...40
Appendix 3: Correlations...43
3
Introduction
In current management theory, attention towards diversity and the impact of on performance of organizations, is a growing issue. Diversity influences directors’ characteristics, thereby influencing decision-making. Quality in decision-making might be improved by consulting various perspectives on the same problem, which can cause better solutions to appear (Harrison & Klein, 2007), however, these various perspectives might also generate new conflicts, which decrease effectivity and rapidity in decision-making. Many organizations have introduced regulating measures, for example an intended percentage of women in the board, to match their team of managers with the image of the current society, diverse and open. Does this regulation also foster organizational performance? The question in this research is, does diversity in the executive team of directors have impact on the performance of organizations, and can we classify this impact as positive or negative?
This research is an addition to earlier studies and existing literature on the performance of
organizations in relation with diversity within the executive team of directors. A majority of earlier conducted studies in this field of interest have their focus on firms within the US, others are focused on one specific variable related to the concept of board diversity (e.g. Adams & Ferreira, 2009; Arfken, Bellar & Helms, 2004; Canella, Park & Lee, 2008; Deszo & Ross, 2008; Donaldson, 1995; Erhardt, Werbel & Sharder, 2003; Frink et al., 2003; Greening & Johnson, 1997; Jehn, Northcraft & Neale, 1999; Litz & Folker, 2002; Marimuthu & Kolandaisamy, 2009; van der Walt et al., 2006; Welbourne, Cycyota & Ferrante, 2007). Little empirical evidence is available and many studies are only conducting research within a time span of one year. In this study, however, the main research area will be Western Europe, because earlier research has mainly focused on the US. Diversity is a culture-related concept which implies major differences are possible comparing different areas. Outcomes of diversity studies considering the US may not be applicable to organizations in Western Europe. In this study we will use an alternative approach towards diversity, executive team diversity concerns various variables, including age, knowledge, education, nationality and gender. The five variables are chosen carefully, because a majority of former ‘diversity’ studies only focus on one of the diversity variables, mostly gender or age. When diversity is researched, this should cover a complete image of all the relevant differences in a team of directors because all the factors are
4 variable influences organizational performance differently, since they all have their own impact on total character and therefore decision-making tendencies. This can be concluded by comparing former research focusing on different variables. Their outcomes differ significantly in both positive and negative directions. This study will conduct research in a period of three years, performance of the concerned firms is measured through 2007-2009. This time span is chosen carefully for several
reasons. First, the main part of the literature formerly written on the diversity subject, presented further on in this paper, is based on company information of organizations within the 2007-2009 period. Therefore the comparability to this former literature can be increased by conducting this study in a comparable time span, which might be interesting for future research. Second, the diversity subject was researched frequently during this time span, so that rises in diversity are easy to detect and the influences on performance are easier to relate to the diversity concept. Finally, the information provided from the organizations in our sample is accurate and complete in our chosen time, team executives and staffing was registered carefully, thereby increasing the reliability of the research. The main question used to guide this research will therefore be:
‘What effect does diversity within the executive team have on the performance of organizations in Western Europe?
In summary, previous research into diversity has produced a variety of outcomes, which are pointing in several directions. One of the reasons this paper will add new information to existing theory, is because of this variety in former outcomes. All former research has investigated their own approach of diversity in several regions, frequently based on one factor, for example gender. This current research will apply an alternative approach of diversity research by covering a more complete concept of diversity including five variables and apply it to performance on firms located in Western Europe. This research paper starts with a summary of theory considering our chosen research question and clarification of the former findings already mentioned above. The literature review will lead to an practical and empirically testable question which points to a section of research, first a methodology part, second a description of data. Following, the outcomes will be presented and clarified, closing with the conclusion and the discussion.
Literature review
5 Various theories reflect factors not directly related to the organization, others reflect factors of
influence on the performance of firms. Several theories relating to this statement are presented by Donaldson (2005), for example the resource dependence theory, institutional theory, population-ecology theory, and the structural contingency theory. These theories have received criticism from both the empirical- and the theoretical sides during the 1980’s. With his theories, Donaldson has increased the rise of the view that managers are an important source of impact on decisions of organizations, decisional outcomes and their actions (Kiefer, 2005). This view has activated diversity research: What effect do various characteristics of directors have on the performance of an
organization. Therefore, we have developed the following supportive hypothesis: H1: The trend of composing a more diverse executive team increased during 2007-2009.
A theory contributing to the view of Kiefer (2005) is the upper echelon theory which argues
organizations to be a reflection of their team of directors. The executive team directors will use their influence in a person-related manner which associates choices of strategy and performance of firms with the tendencies of themselves (Marimuthu & Kolandaisamy, 2009). Higher-level managers are more able to exercise strategy in a personal way than lower-level managers because of the discretion related to the executive team (Mintzberg, Ahlstrand & Lampel, 2009) and the decreased control over high-level directors compared to lower-level managers. This influences the decision-making activities towards the values, attitudes and beliefs executive team directors (Talke, Salomo & Rost, 2010). Values, perceptions, beliefs, judgment, and other subjective variables are all considered as cognitive factors. The factors are not transparent and not easy to test. Therefore, the suggestion of Hambrick and Mason (1984) to use demographical factors to determine these factors is not hard to understand, based on the assumption variation in demographical factors being a trustworthy predict for outcomes in the underlying cognitive factors. We will also use this assumption to draw conclusions about the main research question, that diversity in demographical factors relate to divers cognitive outcomes which might imposes an influence on organizational performance. In this research, demographical factors are included in terms of national background of the concerned directors within the executive team. In other words, does national background as a crucial part of our chosen concept of diversity increases creativity in decision-making processes and therefore influences organizational performance? The upper echelon theory concludes executive team directors to have an major impact on the
6 The concept of ‘Diversity’
Diversity can be captured in disparity, separation and variety. In relation to financial performance, the three all have their own specific outcomes due to different advantages, and disadvantages.
Disparity
Disparity (non-equality) stands for variations in the vertical direction, for example variations in status or rank. Disparity is at the top if one of the executive team directors is more important in status than all other directors. The variation of disparity is a continuum, varying from no disparity (or a minimum) to an influential amount of disparity. Detrimental outcomes are in present when operating at maximum disparity, influencing participating, opinion and sharing of knowledge and expertise between the directors (Klein & Harrison, 2007). Disparity is handled frequently to measure diversity in terms of financial compensation, income and allowance for making decisions which decreases its importance in this research compared to separation or variety, (Harrison & Klein, 2007), in-between measuring within the team of executive directors is not a goal of this research and since we aim at performance of the organization as a whole. However, we are including disparity by researching diversity within the executive teams by examining the specific positions and looking at the individuals who occupy these positions.
Variety
Overall, variety is assumed to produce new, creative thinking, increasing the quality of decision-making and therefore affecting financial performance of firms.
Variety stands for the variation in knowledge and experience within the team of directors. The same continuum applies for variety as for diversity, running from a minimal variety (all directors falling in one type of scholarship or for example, experience) to a maximum of variety (all executive team directors are symmetrically divided over the existing categories, so all the members carry various knowledge). In the last type of variety all executive team directors are equally valuing the carried views and are not permitted to conduct coalition-forming (Harrison & Klein, 2007). Variety is often related as a positive factor considering the performance of the firm, since variety may create new ways to operate situations. Decision-making might increase in creativity and therefore effectivity can be higher due to variety (Carter, Simkins & Simpson, 2003). Variety can be accounted for new thinking and enhances creativity and the innovating-tendency (Arfken, Bellar & Helms, 2004). Variety may also produce a better fit between the executive team and the potential customers and personnel,
7 quality within decision-making (Richard & Shelor, 2002; Randoy, Thomsen & Oxelheim, 2006; Boone & Hendriks, 2009), which proves variety to be relevant and accounting for a possible outcome in our research.
Separation
Separation stands for the sum and placing of variation between executive team directors in positing or opinion-expressing on a specific issue relevant to their assignment. This variation is particularly build on beliefs, attitude and values. When all executive team directors apply the same position or opinion on a particular issue within their assignment a minimal separation is due, avoiding variations on a horizontal base. Logically, the executive team is divided between the ends in an opinion, on an equal base, separation is maximal and the executive team is split into two ‘camps’ holding explicit other beliefs, attitudes and values (Harrison & Klein, 2007). When all executive team directors are near but not precisely the same in their opinion or position, a moderate separation is due. When separation is high, decision-making might suffer from a decrease in efficiency and the ability to reach consensus. This might lead to a decrease in organizational performance and makes separation relevant in terms of a possible negative outcome in our research.
In this research, diversity will be covered as complete as possible, without focusing on only one or two aspects of diversity, which is mainly done in former research. Our chosen method to measure the entire concept of diversity considers a focus on age, knowledge, education, nationality and gender, which summarizes the focus of earlier research into diversity and therefore increases comparability towards earlier research outcomes within the diversity field. We are convinced these variables capture the characteristics of the executive team directors relevant for measuring diversity.
The possibility of curvilinearity
8 the aspects of diversity on decision-making, providing both benefits by increasing for example
creativity, as a negative impact by decreasing efficiency and speed in decision-making when executive team size increases above a certain level. In conclusion, the relationship between organizational performance and diversity might be curvilinear due to the benefits of variety versus the negative aspects of separation.
The diversity-organizational performance relation is often not sufficiently accounted for in past research (Richard, Barnett, Dwyer, Chadwick, 2004). We are researching a relation between executive team diversity and the financial performance of the related organization, expressed in a curvilinear relationship. An effect is expected according to the results of former research, although the outcomes of former diversity literature were very different. Curvilinearity is suggested by Richard et al. (2004), who argue that moderate diversity leads to categorization, implying a negative effect on social interaction by creation of in-group/out-groups and cognitive biases. In their research, however, they expect when diversity increases beyond the moderate level, categorization decreases because group-members will be more evenly diffused and in-group/out-group identities will be reduced (Alexander et al., 1995), resulting in an expected curvilinear effect of diversity. Richard and Shelor (2002) also hypothesize for curvilinearity. They explain organizational performance is not hampered by affective conflict and creativity obtained from heterogeneity is not present and therefore irrelevant for
9 them, age, gender and national background are relations-oriented features and both features have opposite effects on performance. Therefore, they argue inclusion of the task-oriented features to be crucial for capturing the effect of diversity on overall organizational performance. The importance of education and knowledge is also stressed by former research from for example Hambrick and Mason (1984), Jehn, Northcraft and Neale (1999) and Ping (2007). National background is another important diversity variable. Nielsen (2010) suggests within diversity, national background to be an important source of directors’ capabilities and their ability to cope with several complex problems, thereby affecting organizational performance. His view is supported by many authors, for example Carter et al. (2007), Carter et al. (2003) and Erhardt, Werbel and Sharder (2003). We will conclude with the factor gender. Gender is an often researched area within the field of diversity and its importance is stated by many authors, among others Ancona and Caldwell (1992), Arfken, Bellar and Helms (2004) and Campbell and Minguez Vera (2009). Obviously, gender is an important personal characteristic. The existence of stereotypes (for example, men to be tougher, and women to be more sensitive) also indicates gender to have an extensive influence on directors’ characteristics. It is important to test its influence on organizational performance within the diversity concept. Executive team diversity in this research will include age, education, knowledge, national background and gender, capturing several variables considered in former diversity literature, but applying a new approach by combining five variables in our area of research. We will measure over three years, our expectations are summarized in the following hypotheses:
H2: The relationship between executive team diversity and financial performance of the organization is curvilinear, parabolic and has an optimum.
A range of outcomes
This section will review former research in the diversity field. As explained, former research differed widely in outcomes, pointing as well in positive as in negative directions. We have divided these former studies in direction of outcomes, starting with research pointing towards positive performance outcomes.
Former research pointing towards positive performance outcomes
10 Neale (1999) displays diversity in information (knowledge- and perspective-based) is positively related to the performance of the concerned firms. The Litz and Folker-research operated on a sample size of 638 firms covered by the Fortune 1000 in 1997. The same conclusions are drawn by Deszo and Ross (2008) resulting out of panel data from 1992-2006 of Fortune 1500 organizations. Diversity in national background is also assumed to have a positive impact on the performance of firms, based on a measurement of Tobin’s Q within a sample of 1066 listed firms in the United States in 1998,
conducted by Carter, Simkins and Simpson (2003). However, these outcomes are not relying on the measure of performance since the research of Erhardt, Werbel and Shrader (2003) showed evidence for the positive effect of diversity in national background and gender diversity on performance measured in financial terms, based on return on assets and equity.
Measuring performance of firms is possible by various measuring tools. The research papers
mentioned earlier were based on market-related or accounting-related ratios. However, measuring the influence of diversity on the performance of a firm is also possible by research into the reaction of the investing side. Investing parties are assumed to appreciate diversity when a diversity-increasing management member is assigned resulting in higher returns. Kang, Ding and Charoenwong (2010) have conducted research on the responding of stock market to female high-level managers. Their research was conducted in an Asian environment including a sample of 45 organizations from Singapore, resulting in increased abnormal returns in organizations listed in public, after assigning a female high-level manager. The response was increasingly positive when the female manager was assigned for an independent direction- or executive function except for CEO’s. Francoeur, Labelle and Sinclair-Desgagné (2007) argue complex environment based firms benefit from an increased number of female direction members by producing 0,17% monthly abnormal returns, positively and relevant. Their research is based on the outcomes of 500 organizations located in Canada, using a sample within the time span 2001-2004. The influence of female management members in the executive team on both short- and long-term performance in financial measures is examined by Welbourne, Cycyota and Ferrante (2007), using a sample of 534 IPO (Initial Public Offering) firms in the United States. Their results were pointing at a positive impact on the initial pricing of the firm caused by female executive team directors, based on the outcomes of Tobin’s Q, stock-price growth over a three year period and growth based on share-related earnings. They argue for the improvement of innovation and processes concerning problem-solving by adding female executive team directors. Outcomes for complex conflicts are developed and evaluated by differences in knowledge and perspective, which is increased by diversity in gender.
11 organizations listed on the Fortune 500. They concluded their chosen variables to have a positive impact on the relation between diversity in nationality and the performance of the concerned firms. Their measurement was based on the return on sales and investment ratio’s. Also diversity based on gender was related positively to their moderating variable of innovation. Following variety, diversity-increasing activities in the management team lead to variation in perspectives, ideas and networks, affecting innovation in a positive manner. Despite these positive effects, innovation and reputation of a firm did not have a significant impact on the relation with gender.
Canella, Park and Lee (2008) conducted research on the influence of collocation of executive team directors and uncertainty in the environment in relation to diversity within the executive team and the performance in financial terms. Their research was based on information from over 200 organizations from the United States. The relation turned out to increase in strength (positively) when more directors are located in the same area or when uncertainty increases. They concluded an executive team is capable of getting advantages from diversity on the condition of sharing the specific and significant knowledge that the executive team possesses. If executive team directors are located in one location it is less difficult to apply to this condition. Next to these outcomes, when executive team directors are demographically spread, the relation is negatively influenced by a decreased number of meetings. Identity of the group is therefore less or not at all developed causing functioning as a group to be impossible. In this situation, the executive team is conducted out of various individuals with various responsibilities rather than a team. Educational background and its impact on innovation is concluded to be positive according to Camelo-Ordaz, Hernandez-Lara and Valle-Cabrera (2005). Their research over 100 organizations located in Spain led to an argument that increased creativity in problem-solving and idea-developing is increased by a more evolved education. Next to this conclusion they researched the influence of diversity in background of function on innovation, which, however
variations in background are often related with strategy-developing of innovating and changing, turned out to be non-related.
In summary, several research papers investigating the impact of diversity have presented significantly positive outcomes. They used different samples in different regions aiming at different variables. There is not a clear conclusion yet in variables representing diversity pointing towards a positive effect on organizational performance. In the next section, we will focus on former research producing the opposite in outcomes, a decreased organizational performance.
Former conducted research producing decreased performance outcomes
12 Despite the outcomes of United States investigations considering the relation diversity-financial performance are dominantly positive, outcomes of studies investigating Europe are covering both sides. The influence of gender and educational diversity within executive team of directors on the performance of firms is researched by Smith, Smith & Verner (2005), covering 2500 firms from Denmark in the period 1993-2001. In this study, the positive effects on performance following from former research papers, are mostly applying only to female directors who had studied at the university. The remaining female directors within this sample did not influenced performance in financial terms significantly. To compare this result with the foregoing section and set a clear example for the contradiction in outcomes of former studies, Wilson and Altanlar (2009) had researched the same issue in the United Kingdom and their outcomes reflected the attendance of a minimum of one female director within the executive team decreased the risk of failure of the whole organization in terms of financial assets. Nielsen (2010) researched the influence of nationality diversity on entry in a foreign market and performance in financial terms. The research area covered 165 companies located in Switzerland. Her outcomes stated that the capability of an executive team to conduct foreign
operations is positively related with nationality diversity. Nationality diversity of the team of directors fosters entry to a foreign market, influencing performance of the organization in financial terms positively. Several studies concluded decreased diversity within the team of directors in gender (Shrader, Blackburn & Iles 1997; Bohren & Strom, 2007) or tenure (O’Reilly, Snyder & Boothe, 1993) concern increased firm performance. Shrader, Blackburn and Iles researched 200 organizations of Fortune 500. The Bohren and Strom-investigation considered around 200 organizations located in Norway in the 1982-2002 period. The researchers investigated the influence of diversity on
performance in financial terms in various manners. Shader et al. choose return on equity and assets as a measuring tool, while Bohren and Strom used Tobin’s Q. The influence of age differences and gender on the results were investigated by Eklund, Palmberg and Wiberg (2009). Their sample of organizations based in Sweden led to the conclusion that age differences were not influential. If diversity in gender led to a larger management team, the impact of this source of diversity decreased the performance of the organization. The outcomes were less significant if the size of the management team remained stable. Hambrick and D’Aveni (1992), investigating major organization failure,
concluded that the unsolved issues due to conflicting characteristics within the executive team were an significant issue of the resulting breakdown. Ancona and Caldwell (1992) researched the influence of the demographics of a team on performance in organizations especially focused on the technological field. The outcome can be summarized in diversity fostering performance, although coordinating these outcomes were of high cost for the organization concerned. They concluded, all variables counting in their decision, the overall impact of diversity decreased performance of the considered firm. In summary, compared to the former section these results display opposite (negative) outcomes within the diversity-organizational performance relationship. This might be a result of geographical
13 focus of investigation within the diversity concept, for example age versus gender. In the following section we will conclude our review with results from former studies without significant results on organizational performance.
Former conducted research producing no significant results
Considering these diverse outcomes, there are also researches conducted with outcomes pointing at a lack of significance between the performance of an organization and diversity. Hiring a female in the executive team does not have any creating or destructive effects, according to Farrel and Hersch (2005). Their measuring tool were abnormal return of stocks, which was not influenced significantly when a female entered the management team, and they investigated 300 organizations represented in the Fortune 1000 in the period 1990-1999. However, choosing a new management team member was not neutrally conducted concerning gender, because the female representation in the organization was negatively related with the option to search for another female in their management team.
Furthermore, Marinova, Plantenga and Remery (2010) investigated about 200 organizations from the Netherlands and Denmark on gender diversity, which did not produce any significant effects on the performance of the organization. No relevant link was also found by Rose (2007), considering gender, national background and diversity in education in relation to the financial performance of an
organization. The study was conducted in Denmark in 1998-2001, using Tobin’s Q as a measurement tool. Van der Walt, Ingley, Shergill and Townsend (2006) have conducted research relating diversity in the executive team interior and the organizations environment concerning strategy with the performance of the considered organization. Around 60 organizations from New Zealand were investigated during 1997-2001, while expecting a non-heterogeneous management team performed better in an risk-free environment, and heterogeneous boards would perform better in an environment with risk, due to the more complex strategy needed. Their reasoning was based on the assumption that heterogeneous management teams possessed more different knowledge and necessary look at
problems to organize and review problem-solving strategies. However, there was not found any influence of diversity within the management teams on performance and the related environmental strategy also was not influential. Identical outcomes were found by Kiefer (2005) who conducted a similar research in Germany. Some over 1000 organizations in the US were researched by Richard and Shelor (2002), who concluded no significant impact on executive team age diversity-performance of the organizations considered.
14 diversity, thereby influencing variety or separation. Including these five factors and thereby using an alternative approach towards diversity may define the course of influence diversity has on
organizational performance, perhaps thereby summarizing the range of former results. This research will further elaborate to investigate the relation between a more complete concept of diversity and its influence on the performance of firms in Western-Europe. This will hopefully provide us with a clear conclusion between the range of different outcomes of former diversity studies, and extend the field of diversity research towards Western Europe. We will use the following conceptual model:
Table 1: Conceptual model
In this model, executive team diversity is influenced by five variables (together they represent
15 Methodology
In this section we will present the methodology we will use in our research to investigate our main research question and provide clarification on the hypotheses.
A significant score on the index of Blau for heterogeneity and inclusion of educational background and knowledge categories might reflect variety since diversity in these categories might foster decision-making and creativity of executive team members. Variety might co-occur with separation because our expectations include a curvilinear relationship. Separation might be reflected when the score on the index of Blau becomes inefficient, hampering decision-making and decreasing efficiency (Harrison & Klein, 2007; Campbell & Minguez Vera, 2008; Miller & Triana, 2009). The index of Blau as a measuring tool is recommended to investigate diversity within clusters of subjects (Miller & Triana, 2009). The index of Blau for heterogeneity is calculated using the formula:
1-∑(p/k²)
k: Variety: k=1, k=2, etc. (possible categories in K)
p: Share of higher management staff in the category k
If diversity within the executive team increases, the value calculated by the formula above also increases. The outcome can never be below zero and the highest possible outcome is related to the specific variable and its related categories. In theory, the highest possible outcome can be found using the equation (Biemann & Kearney, 2009):
(K-1)/K
K: categories related to a specific variable
16 Table 2: Chosen variables, possible number of categories (K) and their theoretical maximum. As our variables show, the highest possible index outcome for the education variable is higher than the same outcome for the gender variable with a difference of 0,33. The outcomes are only possible to compare across the various variables concerning diversity if they are adapted in their differences concerning the related category-number. A correction-method for this measure is designed by Agresti & Agresti (1987), named the Qualitative Variation Index. They multiply the index of Blau with K/(K-1) in order to standardize the k. Equality is arranged and the categories related to the specific variables are comparable although they may vary in possible range. This is necessary and therefore integrated within this research because we have divided the concept of diversity in several smaller variables (age, gender, national- & educational background, knowledge). For example, gender has only two possible outcomes and age is divided in multiple levels.
We will investigate the development of diversity during our chosen time frame to check for a possible trend in diversity. When the development in organizational performance during the time frame is relating to the development of diversity, there is an indication the two are related, on which we will build upon. Furthermore, we will also examine every position within each executive team for the chosen organizations within our research, for each member. Within this list of positions, we will summarize them to the chief executive officer, the chief financial officer, the chief operating officer and a list of remaining member positions. Within our classification, an investigation will be done focusing on two of our variables, national background and whether the member was either male or female. This examination includes disparity by indicating whether there is diversity within all power-levels concerning the executive positions.
17 Hausman-Wu to exclude the possibility of endogeneity. If the relation is endogenous, we start
applying a 2-stage-least-squares test on the data of the panel we investigate (Wooldridge, 2006). If the explanatory variables turn out not to be endogenous, a 2-stage-least-squares test brings the risk for a lower efficiency compared to the ordinary-least-squares. Therefore, when Hausman-Wu indicates no endogeneity, we will conduct an ordinary-least-squares test. We will proceed with an investigation to check if executive team diversity is actually affecting the performance of the firm instead of firms who have a better performance also choose an executive team which is of increased diversity (Marinova, Plantenga & Remery, 2010). To determine this direction we will use a causality test. We will conduct our research estimation by using the following formula’s in statistical software:
Performance of the specific firm: α0+(α1 x executive team diversity)+∑αx+ε Executive team diversity: β0+(β1 x performance of the specific firm)+∑βz+ε x, z: control variable vector
ε: factor of error
Executive team diversity influences the performance of the firm if significance is found in the estimation of the coefficient of α1. To confirm this causal influence, β1 shouldn’t produce the same effect, which excludes the possibility of causal bias.
In conclusion, we will measure diversity on five factors within the executive team of directors using the index of Blau. Furthermore, we will measure the performance of the firm using Tobin’s Q. If an increase in diversity leads to an increase in organizational performance, diversity has a positive influence on organizational performance. Counter wise, if an increase in diversity leads to a decrease in organizational performance, diversity has a negative influence on organizational performance, following reasoning in former research of for example Adams & Ferreira (2009), Carter, Simkins & Simpson (2003) and Welbourne, Cycyota & Ferrante (2007). We will also check for linearity. Comparing all the organizations in our sample during the time span 2007-2009 and considering the five variables within the diversity concept, we can find the effect of diversity on organizational performance within an executive team of directors.
Data
Description18 increase congruity with former research, completeness of data and include the trend period of
diversity. The sample countries chosen to clearly represent this part of Europe are the Netherlands, Belgium, Germany, France and Spain. These countries are chosen specifically because they cover the largest part of Western Europe and they also include most cultural characteristics present in Western Europe, for example the time-oriented workplace in the Netherlands as opposed to the temperature-oriented planning in Spain. As we have introduced in the literature review, these characteristics might influence organizational performance. Furthermore, they provide the most accurate and complete panel of data to increase reliability of the research. We will focus on main stock exchange of the concerned countries. In the following table the exchange of stock is displayed as a function of the country (in short) related.
Table 3: Sample countries and related stock exchange.
Our selected sample of companies includes 126 organizations in the larger segment, which totals our amount of data to 378, taken into account that our research is aimed at a time frame of 3 annual reports (2007, 2008 & 2009). In the following section we will clarify the exclusions.
Annual reports of organizations are seen as the greatest source of data. When executives or their specific positions, for example, were missing, Bureau van Dijks’ and Thomson Reuters’ datasets are used. The first considers financial data on organizations located in Europe and their executive team. The latter is required to include significant data pointed at finance and accountancy.
Since only executive directors who are incorporated in organizational decision-making and developing directions in strategy are relevant in this research, the members without executive positions will not be included in this research. The impact of members with executive positions is larger than non-executive members looking at the decisions of the organization (Kiefer, 2005; Ping, 2007; Cannella, Park & Lee, 2008; Marimuthu & Kolandaisamy, 2009; Talke, Salomo & Rost, 2010).
19 double listed and therefore overrepresented or their executive team composition information was incomplete in for example characteristics, which could cause bias. Both reasons would cause bias in our results. The companies chosen were filtered on outliers in performance of their organization. Resulting out of an investigation for the minimal required amount of information for significant research an analyses for sensitivity was conducted. The outcome reflected 0.8 to be a criteria of a minimum of information provided regarding executive team diversity. Organizations that scored above 0.2 for missing factors relevant for our diversity investigation were excluded. Example: investigation in an executive team consisting of five directors is conducted, in which for one director the enjoyed education information is missing, providing us with an 0.8 factor of information concerning
educational background. The actual index of Blau based on a situation with all required information provided is a 0.16 difference with the former recalled factor. The measurement of diversity in total is divided across all of the five variables on diversity in terms of the average index on Blau. The deviation of 0.16 is therefore a minimal impact which is not significant enough to exclude this example from our research. This decreases possible bias. Due to deviating statistics, one firm from Germany and one from the Netherlands were not taken into account in this investigation. Both were extreme outliers. They had extreme values in firm performance due to factors not related to diversity, resulting in a Tobin’s Q value above ten while the other observations remained beneath the value of ten (Appendix 4).
Variables
We will proceed this part with a clarification of our chosen variables. Variables measuring the performance of the organization
20 (Equity market-value+Debt book-value) /Total assets book-value
The outcome will be above 1 if the organization is in possession of specific assets which are intangible or specific chances to increase the organizations value. The common prospect of investing parties is the organization to be competent enough to use the assets properly to increase the value of the organization. The opposite is due when the outcome is beneath one, which is due when the
organization is valued beneath the book-value and therefore accused of insufficient creation of value out of the assets provided (Sudarsanam, 2003).
In summary, we will use Tobin’s Q to measure organizational performance. We will compare the value of Tobin’s Q with the index of Blau to conclude what effect diversity within the executive team has on organizational performance.
Variables measuring diversity
This research is mainly measuring diversity in terms of variety and separation. Disparity is included by investigating the occupation of the positions within the executive team. If the data provided is divided in different sections variety is able to present diversity within the several categories (Nuesch, 2009; Boone & Hendriks, 2009). The information regarding the executive teams is transformed into an index to define the diversity within the group of the executive teams individually (average of all individual variables concerning diversity divided equal) and to define the index of Blau. The function of the directors is also included, translated into an index. Clarification in the form of a code-table of indexes is included in the appendix.
Other variables
Finally, a sufficient research demands an extensive check on possible other variables which might have an influence on performance of the organization and show a correlation with the variables concerning our measurement of diversity. These specific variables are: organizational age, organizational industry, size of the organization, size of the executive team, location of the
21 Furthermore, executive team size is of no less importance because of its positive relation with an increase in diversity and a higher level of creativity, due to the larger input. Biemann and Kearney (2009) have showed that excluding the control variable of executive team size results in a consequent underestimation of our variety-factor in executive teams decreased in size. However, Kiefer (2009) showed us there is also a downside of large executive teams, tending to decrease inertness and effectiveness. Therefore, we have included the control variables organizational age, organizational industry, size of the organization, size of the executive team and location of the organization. The location of the organization is included to define our estimation regarding a possible trend in diversity during 2007-2009.
The following table displays the variables and their related explanation.
Table 4: Chosen control variables and explanation SIC codes (OSHA, 2010)
Results
The outcomes of the research will be presented in the following part. We will start with a diversity-related overview, from 2007 to 2009. Here we will present the level of diversity during the time span. The evolvement of diversity will be explained in the following section. So we will first enlighten the current diversity levels within the time span 2007-2009, proceeding with an overview of how this level of diversity is applied within our sample of organizations. These results will answer our first
22 Development of the diverse executive team
Here we will enlighten the change in diversity rate within executive team composition. This evolvement is showed by the adapted edition of the index of Blau over the period we have chose to focus our research on, from 2007 to 2009. Analyzing the development of a diverse executive team of directors is useful for several reasons. First, a trend suggests a possible outcome for our research, because it imply increased performance of the organization due to diverse executive directors. Second, a possible trend in diversity has an effect on our research outcomes because if organizational
performance is influenced by the extent of diversity within the executive team, the results will be in proportion with this trend. In most cases the trend of the diverse executive team has become somewhat higher during the time frame. This is represented by the following percentages which summarizes the extent of diversity based on the variables age, gender, national- & educational background and knowledge:
Table 5: Development of the diverse executive team during the time span.
23 Table 5 represents the percentage of firms in our investigation having an adapted index of Blau above zero. A conclusion which can be drawn from our research (and the next table): within the five different variables concerning diversity in the executive teams, age, education and knowledge are more used to diversify in, compared to the remaining variables, gender and national background. This result may be due to a trend within the composition of existing executive teams. The operationalized constructs of our variables are reflected in Appendix 1.
Table 6: Development of the diverse executive team during the time span, specified for our chosen variables.
We will shortly focus on the development of diversity for the five various countries, however we should take in consideration the course of a trend might take longer than three years, however it might indicate a possible rise in the following period. The Netherlands started with an initial increase in 2008 and ended with a relatively smaller decrease in 2009; Belgium started with an initial decrease in 2008 and ended with a relatively smaller increase in 2009; Germany started with an initial increase in 2008 and proceeded with a comparable increase in 2009, France started with an initial large increase in 2008 and proceeded with a smaller increase in 2009. Finally, changes in Spain were relatively small with a minimal increase in 2008 and a somewhat bigger decrease in 2009. In all of the countries except Germany and France an increase might be expected in relation with debates in public and political actions, although this seemed a faulty expectation. In NL and BE no rise of diversity could be
detected. The rate of diversity within the executive team turned out to be decreasing in Spain regarding our chosen variables, which is interesting. A possible reason was subscribed to the countries culture, for example, females are relatively underrepresented in the total workforce in Spain compared to the other sample countries (Heidrick & Struggles, 2007). Campbell and Minguez Vera (2009) subscribed
0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1
Age Knowledge Educational
24 the latter as a result of their traditional place in society bound to their culture. According to their research we might assume a relative decreased outcome for a diverse executive team in terms of gender in Spain compared to the other countries, but this assumption also turned out to be faulty. The relatively low value for diverse executive teams in Spain is for a large part caused by the national background variable. In all of the organizations from Spain we have examined, only 16% employed a (minimum of one) member having a foreign background. This information is relevant because it turns out, inclusion of our chosen five variables does provide us with an alternative outcome regarding the decreased diversity rate in Spain, since we are now able to subscribe the cause of the decrease to the national background variable.
Most of our outcomes in empirical testing so far indicated a trend in overall executive team diversity which because it might cause a positive effect on organizational performance, which is our main research question, but, more importantly, verifies our hypothesis 1. This measure included all executive team directors. We will further investigate this in the following sections.
Specific positions within the executive team
The executive team of an organization is composed out of several positions. One of the most well-known positions is, for example, the chief executive officer, the CEO. Furthermore, most
organizations have a chief financial officer, taking care of the financial part of the organization, and a chief operating officer, taking care for the organizations core money-making activities. The
performance of the organizations is partly dependent on decision-making in strategy and therefore every member of the team is able to make influential decisions. Not every member, however, has the same influence. Not every function is positioned at the same power level and therefore, some positions are more influential regarding the performance of the organization. For example, the chief executive officer, is most important within the team, because he takes the ultimate responsibility. The chief financial officer and the chief operating officer are obliged to inform the chief executive officer about the current state and progress of their work. There are more positions possible within every executive team of every organization.
We found most executive teams to be composed out of males and nationals. A small insight in the known distribution is already given by our former table, showing us there is an higher percentage male members occupying the most crucial positions over female members. Almost 18% of our male
25 Table 7: Percentile differences in occupation per position in terms of gender.
Table 8: Percentile differences in occupation per position in terms of national background.
Note that a 1% difference might seem small, but in the sample of managers some positions are relatively rarely occupied, like chief operating officer. In summary, the number of male/female members and national/non-national members is even, although the female/non-national groups are mainly occupying positions lower in influence on the organizations strategic decision-making. Positions with a higher power-level are mainly occupied by males and nationals, which represents a relatively low diversity rate within these specific high-power positions.
We have identified the relevance of national background and gender within our diversity framework. We have examined every position within each executive team for the chosen organizations within our research, for each member. Within this list of positions, we have summarized them to the chief
executive officer, the chief financial officer, the chief operating officer and a list of remaining member positions. Within our classification, an investigation is done focusing on two of our variables, national background and whether the member was either male or female. The outcomes are displayed in the following tables in percentages and cover all participants in the executive teams we have reviewed within 126 organizations during 2007-2009.
These results are important to state that female and non-national directors are not only
underrepresented, but also cover less important positions in the executive team, which might indicate a low level of diversity. We will therefore include the position of the executive member in our
26 Table 9: Gender division within total sample of executive directors and their function.
Table 10: National background division within total sample of executive directors and their function.
Diverse executive teams and their organizations performance
We proceed to the most important part of our research. What effect does diversity have on
organizational performance? This section is based on the following descriptive values of our main variables:
Table 11: Descriptive values of the main variables, considering 378 observations, reflecting the mean, standard deviation, minimum- and maximum values of each variable, starting with the values of all the variables summarized.
27 This table presents an average outcome of 0,45 on total executive team diversity in our chosen regions. The highest- and lowest scores were presented in the table. In calculating Tobin’s Q we have excluded two extreme scores, providing us with the results presented above. The ‘standard’ executive team has six managers, four above the samples lowest score and nine beneath the samples highest score. Finally, the ‘standard’ organization is 78,84 years old since their foundation. The ‘standard’ outcome of size reflects 16,40. Firm size is measured by the natural logarithm of the total sales of the concerned organization.
As described in the methodology, a one-stage-least-squares test is more reliable (over a two-stage-least-squares test) when there is no endogeneity present in the explanatory variables. The Hausman-Wu test showed us endogenous variables were most likely not present because the coefficient scores on the leftovers were not significantly present. Wooldridge (2006) suggested for this situation
28 Table 12: Outcomes of the pooled one-stage-least-squares test, including standard error values and mentioning significance at the 1%, 5% and 10% levels by ***, ** and *. Indicators are dummy variables. The scores between ‘()’ reflect negative values. The indicator for FR is categorized as a reference, just as the first industry indicator and the 2009 indicator.
29 does not influence the level of bias within these estimators. No significant impact was found for the control variable executive team size on organizational performance. Therefore, the effect of executive team size in its function as control variable is irrelevant to our research since there is no influence in our test results. The test value of executive team size is reflected in Appendix 3. The indicator variable for Belgian organizations turned out to be the only significant country indicator which implies our result is country-specific. Tobin’s Q showed a difference of 1,02 in advantage for organizations from France in comparison with organizations from Belgium. This is subscribed to internal country characteristics. Organizational performance was influenced importantly by both the industrial indicators. Industrial firms from category two (organizations classified in communication, electric, gas, retail, transport or wholesale industries according to OSHA, 2010) outperformed organizations from category one (organizations classified in agriculture, construction, manufacturing or mining industries according to OSHA, 2010) by a score difference at Tobin’s Q from 0,58. Moreover, the rest of the firms, from category three (organizations classified in administration, finance, insurance or service industries according to OSHA, 2010) scored lowest, with another 0.33 decrease in Tobin’s Q. Finally, we cover our year indicator variables, showing a decrease in organizational performing during the years, displaying a decrease of 0.79 in Tobin’s Q. We used scores including all five variables because in practice, within the executive directors, it is also not possible to isolate them.
In conclusion, a significant influence is found from executive team diversity on organizational performance. This is caused by the diversity variables and not the other way around, because organizational performance turned out not to have a significant impact on the rate of diversity. Organizational size and age also had a significant impact on organizational performance, suggesting young and small organizations to perform better.
We will build further on the relationship between the executive team and organizational performance. Following from our analysis, linearity in executive team diversity-organizational performance
30 Table 13: The difference between total diversity and the squared version of total diversity The values above reflect the relation between organizational performance and diversity to be non-linear, which is relevant to investigate hypothesis 2. The final relation between diversity within executive teams and organizational performance in financial terms can be calculated by:
Table 14: Relation between diversity within the executive team and organizational performance, following from our foregoing analysis and table 10. The optimum in executive team diversity
according to this formula reflects 0.52. The calculation represents the effect we have found from diversity on performance of the organization.
This calculation has an highest possible score of 0,52 due to the positive impact of the diversity variable and the opposite impact for the diversity² variable. For every increase in diversity before the 0,52 score, organizational performance is increased, displaying variety. After the value of 0.52 is reached, organizational performance will decrease, representing separation. Diversities’ and
31 creativity in decision-making. However, above the 0,52 boundary in the index of Blau diversity causes a negative influence on organizational performance, possibly due to waste of time in making decisions and lack of effective reasoning. We can conclude so far diversity does affect organizational
performance positively, until it reaches its optimum and from there negatively affects organizational performance. In summary, 0.52 represents an effective diversity score within an executive team, stimulating organizational performance. An executive team is having most benefit from diversity when approaching the 0.52 score, reflecting variety. Beyond 0.52 however, the value of diversity decreases and performance also decreases, reflecting separation. Therefore we can conclude no former research was wrong, providing both positive and negative conclusions regarding the rate of diversity within an executive team. The diversity rate is actually able to influence performance in a positive and a negative way. Contrary to former research, we are able to provide a clear image of the course of the rate of diversity within an executive team. Looking at Harrison and Klein (2007) both variety and separation have their specific effects on our investigated relationship in such a shape that hypothesis 2 turns out to be valid.
Conclusion
Within our research we selected a set of data of 378 observations of 126 large-size organizations covering each industry as classified by OSHA, 2010. The sample organizations were located in the Netherlands, Germany, Belgium, France and Spain and our sample comprised the 2007-2009 period. We measured our executive team diversity using the index of Blau plus organizational performance in terms of Tobin’s Q. Our outcomes reflect an important impact of executive team diversity on the performance of our selected organizations. This impact showed no linearity and was shaped as a parabolic line, showing us the existence of a perfect optimum in a diverse executive team. Up to our score of 0,52 a higher diversity rate is responsible for a higher organizational performance although this change appears slower. Beneath the 0,52 score appointing an executive team director causing an increase in the diversity rate can be viewed as a good evolvement. Hypothesis 2 turned out to be valid. For example nationality, appointing a non-national member when there were only national members before increases organizational performance in terms of the index of Blau. This increase is caused by a higher level of creativity in defining strategy and increased quality in decision-making. Although, if the score on the index of Blau passes 0,52 appointing a member increasing the index further implies a decrease in organizational performing. This decrease might be caused by waste of time and lower efficiency in making decisions. The advantages of diversity, creativity in defining strategy and increased quality in decision-making might be overruled by conflict in this situation.
32 regarding a rise in diversity caused by political actions and attention in public, this trend is only present measuring our selected countries as a whole. The last conclusion verifies our hypothesis 1, which concerned our total sample. This emphasizes the importance of measuring our selected
countries separately, showing no trend in Belgian and Dutch organizations. The Spanish organizations gave us a slight decrease in diversity and the German and French organizations were the only ones showing us a higher diversity rate at the end of our measuring period compared with their score in the first year. Regarding our five diversity-related variables, their distribution within the organizations can be seen as even. Overall, the Blau index of our male/female and national background issues were decreased compared to the variables of education, age and knowledge. We can conclude women and non-national members of the team to hold less positions in total, and next to this, also occupy positions decreased in power compared to their male and national counterpart
Discussion & Limitations
First, we will expand our theory concerning the optimum we found in diversity within the executive team. Thereafter, we will discuss the rest of our research and review limitations and options for future research.
The optimal executive team diversity
The outcomes of these results lead us to another issue, since we suggest there is an optimum in diversifying the executive team in actual organizations. We have based this research on an alternative approach of diversity, including five variables. This approach provides us with the opportunity to develop a model of an executive team using an optimal rate of diversity. What should organizations do to reach the most effective rate of diversity within their executive team? Following, we included a description showing us the perfect executive team within the ‘standard’ picture of six members, composed within the variables central to our research: age, gender, knowledge, national- and
educational background. It turned out it is impossible to compose a team ending at the exact score of 0,52. In our description, the most optimal situation is derived. In the ‘most effective’ executive team regarding all of our chosen diversity variables, division of gender is one to five, in both directions. Specifically for the gender factor former research have supported the parabolic form and the authors suggested a perfect balance of men and women would compose an ideal executive team (Frink et al., 2003). However, this study was composed including non-executive members which makes
33 Regarding national background, national members and members from one other nationality should be balanced, or four national members should be combined with two other members with both different foreign nationalities. Furthermore, regarding age, four members of one age should be combined with two members from one other age, noting that an age categories are divided by a five-year-period. The similar education of four members should be compensated with one other form of education of the other two members. Finally, the same balance should be perfect regarding knowledge. These findings are summarized in the table below.
Table 15: The outcome of our results regarding the ‘most effective’ composition of the executive team, considering organizational performance.
Group A and group B are for every variable different categories within their sector. Group A and B within gender for example, represent males and females, not per definition in exact order. However, we should add this composition and diversity rate provides the executive team with the most complete picture of characteristics. The executive team may use this benefit to improve their decision-making and other processes influencing organizational performance. It is not said that they actually will use this benefit, so composing an executive team approaching the 0.52 diversity rate will not guarantee organizational performance.
Limitations
There is no research without limitations. In the literature review we have showed diversity in several concepts, separation, disparity and variety. By selecting our measuring instrument, the index of Blau, we view diversity in terms of variety and separation. We also include disparity in our research by
0 1 2 3 4 5 6 Gender National background
Age Education Knowledge
34 investigating diversity in terms of occupation of positions within the executive team. The index of Blau is a measuring tool providing us with diversity within the units of data within categories. Variety can coexist with separation. Nuesh (2009), Boone and Hendriks (2009) viewed the index of Blau as the most reliable available measuring instrument, although it is not perfect. Our observations are acquired from five countries, regarding Dutch, Belgian, German, French, and Spanish organizations representing the west of Europe. The chosen regions do form a major percentage of the west of Europe, however a sample always imposes limitations and opportunities for research in future perspectives. The region can be expanded, replaced by other countries or specified, using the same alternative approach of diversity. Specific country differences are suggested in our results which imply a reason for future research. We have included our own limitations per location by including the criteria of organizations being listed in exchange of stock. We have chosen to acquire observations from larger organizations located in the west of Europe. This choice is made in regard of the need for specific data about members of the executive team and their related characteristics, which is hard to obtain from organizations smaller in size. Increased organizational size imposes a limitation on the appliance of the outcomes in general because it might be a specific relation bound to the size of the organization. However often applied in studies regarding diversity issues, use of measuring
instruments, in this study the index of Blau and Tobin’s Q, might impose limitations and opportunities. The main question answered in this research can be repeated using other instruments, to check for reliability of the outcomes. Furthermore, we have not found highly diverse executive teams in this paper. Improved testing of the relation between diversity and organizational performance will need with executive teams covering all possible diversity levels. It is difficult to organize such a test since ‘perfect’ diverse executive teams are nonexistent.
35 References
Adams, R.B., Ferreira, D. (2009). Women in the boardroom and their impact on governance and performance, Journal of Financial Economics, 94, 291-309.
Agresti, A., Agresti, B.F. (1978). Statistical analysis of qualitative variation, Sociological Methodology, 9, 204-237.
Albright, S.C., Winston, W., Zappe, C. (2009) Data Analysis and Decision Making with Microsoft Excel, Canada: Nelson Education Ltd.
Alexander, J., Nuchols, B., Bloom, J. & Lee, S. (1995) Organizational demography and turnover: An examination of multiform and nonlinear heterogeneity, Human Relations, 48: 1455-1480.
Ancona, D.G., Caldwell, D.F. (1992). Demography and design: predictors of new product team performance, Organization Science, 3:3, 321-341.
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, 177-186.
Biemann, T., Kearney, E. (2009). Size Does Matter: How Varying Group Sizes in a Sample Affect the Most Common Measures of Group Diversity, Organizational Research Methods, 13:3, 582-599. Bohren, O., Strom, R.O. (2007). Aligned, Informed, and Decisive: Characteristics of Value-Creating Boards, EFA 2007 Ljubljana Meetings Paper, available at: http://ssrn.com/abstract=966407.
Boone, C., Hendriks, W. (2009). Top Management Team Diversity and Firm Performance:
Moderators of Functional-Background and Locus-of-Control Diversity, Management Science, 55:2, 165-180.
Brealey, R.A., Myers, S.C., Marcus, A.J. (2007). Fundamentals of corporate finance, United States: McGraw-Hill Irwin.
36 Campbell, K., Minguez Vera, A. (2009). Female board appointments and firm valuation: short and long-term effects, Journal of Management and Governance, 14, 37-59.
Cannella Jr, A.A., Park, J.H., Lee, H.U. (2008). Top management team functional background diversity and firm performance: examining the roles of team member collocation and environmental uncertainty, Academy of Management Journal, 51:4, 768-784.
Carter, D.A., Simkins, B.J., D’Souza, F., Simpson, W.G. (2007). The diversity of corporate board committees and financial performance (working paper), available at:
http://ssrn.com/abstract=1106698.
Carter, D.A., Simkins, B.J., Simpson, W.G., (2003). Corporate governance, board diversity, and firm value, Financial Review, 38, 33-53.
Dezso, C.L., Ross, D.G. (2008). Girl power: female participation in top management and firm performance, Robert H. Smith School, Research Paper No. RHS 06-104.
Donaldson, L. (1995). American anti-management theories of organization: a critique of paradigm proliferation, United Kingdom: Cambridge university press.
Eklund, J.E., Palmberg, J., Wiberg, D. (2009). Ownership structure, board composition and investment performance, Centre of Excellence for Science and Innovation Studies, working paper 172.
Erhardt, N.L., Werbel, J., Sharder, C.B., (2003). Board of director diversity and firm financial performance, Corporate Governance: An International Perspective, 11, 102-111.
Farrell, K.A., Hersch, P.L. (2005). Additions to corporate boards: the effect of gender, Journal of Corporate Finance, 11, 85-106.
Francoeur, C., Labelle, R., Sinclair-Desgagné, B. (2008). Gender diversity in corporate governance and top management, Journal of Business Ethics, 81, 83-95.
Frink, D.D., Robinson, R.K., Reithel, B., Arthur, M.M., Ammeter, A.P., Ferris, G.R., Kaplan, D.M., Morrisette, H.S. (2003). Gender demography and organization performance: a two-study investigation with convergence, Group and Organization Management, 28:1, 127-147.
Greening, D.W., Johnson, R.A. (1997). Managing industrial and environmental crises: the role of heterogeneous top management teams, Business Society, 36:4, 334-361.
