The Effects of Underlying and Observable Attributes of Board
Diversity on Firm Performance, and the Moderating Role of
Environmental Complexity and Industry Type
MANON L. HEUZEVELDT*
University of Groningen
INTRODUCTION
This study tests the relationship between board diversity and firm performance. Diversity is defined as the degree to which there are objective or subjective differences between people within a social grouping (Phillips et al, 2009). One argument for board diversity may be the fact that firms are important institutions and should reflect the different stakeholders. When there is more diversity on boards, there is a higher probability that the different stakeholders are taken into account instead of only focusing on shareholder value (Rose, 2007). Another reason is the signalling effect on job applicants because now other candidates are attracted to the firm than the ones normally recruited. It also sends a positive image to other stakeholders like customers. An alternative justification is the fact that board diversity could increase competition because women and minorities now know they also could apply for the position. Diversity in boards should lead to better corporate governance which results in a more profitable business (Rose, 2007).
Understanding the relationship between board diversity and firm performance has important implications for both public policy and the governance of firms (Carter et al., 2010). If board diversity does not influence firm performance, the desirability of diversity is primarily a public policy issue (Carter et al., 2010). However, if there are positive effects of board diversity on firm performance, the economic implications of this relationship are important. If this relationship is negative, the cost of including diverse directors becomes a factor to consider (Carter et al., 2010).
complex problems (Cox, 1993; Cox & Blake, 1991). Different perspectives lead to new insights and ideas and the consideration of alternatives (Fields & Keys, 2003). Heterogeneous groups are therefore likely to be more creative, make higher-quality decisions, and perform better than homogeneous groups (Wanous & Youtz, 1986). These effects of diversity also hold for boards of directors where these characteristics are even more critical (Erhardt et al., 2003).
A number of studies found a negative relationship between diversity and processes that affect organizational performance. Firm performance could be negatively affected by group heterogeneity because these groups are less socially integrated which hinders the exchange of information and increases turnover in top management teams (Ancona & Caldwell, 1992). Group heterogeneity has been linked to stereotyping, in- and out-group effects, affective conflict and turnover (O’Reiley et al., 1993; Pelled, 1996) which results in slower decision making and communication difficulties (O’Reilly et al., 1993). Tsui et al. (1992) also found that diversity can lead to lower psychological attachment among group members. The general conclusion of these researchers is that homogeneity increases integration, trust and communication so the advantages outweigh any disadvantage of redundancy within the group (Hambrick et al., 1998).
2006). To overcome this problem, variables should be measured independently. This research offers a more complex view of the relationship between the observable (i.e. age, gender, nationality) and underlying (i.e. functional background, educational background) variables of board diversity and firm performance moderated by the contextual variables complexity and industry.
Several contributions are made in this study. First, I integrate existing theoretical explanations like the self-categorization and value in diversity theories to explain the inconsistencies in diversity findings. A second theoretical contribution this research makes to the diversity and governance literature is by providing a better understanding of how the relationship between board diversity and firm performance operates. This will be done by differentiating between underlying and observable attributes and by incorporating moderating contextual variables. Third, the effects of diversity on organizational level are examined instead of analyzing the diversity effects at individual or workgroup level. Finally, the effects of board diversity on firm performance are investigated in Dutch firms, which has never been done before.
THEORETICAL BACKGROUND
The opposing viewpoints of diversity come from two theories. The value in diversity hypothesis and the self-categorization and social identity theory. The positive effects of diversity could be explained by the value in diversity hypothesis (Cox et al., 1991). This view states that diversity creates value and benefit for teams as well as challenges for interaction (Mannix & Neale, 2005). The basis for this view comes from the classic work from Hoffman (1959) on heterogeneity in small groups. Hoffman (1959) found that diverse groups are more effective when tasks are cognitively complex or demand multiple viewpoints. Diverse groups have a higher amount of knowledge, expertise and perspectives than homogenous groups (Mannix and Neale, 2005), leading heterogeneous groups to find higher-quality solutions for problems at hand than homogenous groups. More scholars took over Hoffman’s ideas and argued that diversity increases problem solving through the presence of cognitive conflict or a variety of viewpoints (Damon, 1991; Levine & Resnick, 1993). Diverse groups had a higher performance than homogenous groups because the different perspectives lead to new insights and solutions (Nemeth, 1986). The rationale behind this is that diversity causes exchanges between people with different contacts, skills, information and experiences which increase the opportunity for creatively solving problems (Mannix & Neale, 2005). Most support for the value in diversity hypothesis comes from studies that concentrated on functional differences like diversity in skills, information and expertise.
generates distinct group behaviour like group solidarity, conformity and discrimination against out-groups (Tajfel, 1982). When people are different in demographic characteristics like age and gender, people could engage in a self-categorization process and classify themselves and others in different groups (Turner, 1982; Hilary & Hui, 2008). This could lead to inter-group discrimination and differentiation, in-group favouritism and affective preferences for in-group over out-group (Pelled, 1996). A basic assumption of the social identity theory is that people have a need for and are therefore motivated to achieve and maintain a favorable self-image (Ely, 1994). To sustain this image, people draw intergroup comparison and prefer their own group over other groups (Ely, 1994).
Pelled (1996) states that self-categorization is more likely when people differ in visible demographics like age, gender and race than less visible characteristics. For example, a categorization based on gender would result in a person developing a psychological association with either the male or the female social group (Ali et al., 2011). Jehn et al. (1998) also stated that visible demographic attributes increase affective or relationship conflict.
Diversity is thus a complex phenomenon and appears to be a double-edged sword (Milliken & Martins, 1996) because the value in diversity hypothesis predicts positive effects of diversity and the self-categorization and social identity theory predicts negative effects. Based on these theories, competitive hypotheses can be developed and predicted for direct and linear relationships of diversity and firm performance. Competing predictions are useful when earlier research leads to two or more valid explanations (Richard et al., 2002).
Age diversity
different values and attitudes, age diversity should lead to innovation (Bantel & Jackson, 1989). Moreover, in a board context, age diversity may reduce harmful emotional conflict since similarity in age enables comparison of careers that can lead to rivalry (Pelled et al., 1999).
Because both old and young employees have their unique values, age diversity will allow groups to complement each other and positively influence firm performance. Age
diversity could increase creativity, problem solving capabilities and it could attract employees
(Li et al., 2011). Also a firm that tries to attract diversified and capable human resources
stands a much greater chance of success if it maintains balanced age diversity. This because
the firm can attain competitive advantage by being an employer of choice for both older and
younger talented workers (Li et al., 2011). Another advantage of age diversity is the fact that
it can help a firm to better understand the preferences and demands of its aging customers
which could improve marketing and financial performance (Jayne & Dipboye, 2004;
Morrison, 1992). Based on this, the following hypothesis is formulated:
Hypothesis 1a: Age diversity is positively related to firm performance.
because of decreasing interpersonal attraction and increasing cognitive biasing (Linville & Jones, 1980) which leads to less open communication and more conflict which hinders the development of team cohesiveness. With diversity, people engage in self-categorization behaviour which lowers performance because co-operation and communication is constrained (Richard, 1997). Because of the difficulty in communication, transaction costs are increased (Ostergaard et al., 2011).
Jackson et al. (1991) found that age diversity leads to higher turnover rates. Also, people who are different in age from their group members were more absent and received a lower performance rating (Cummings et al., 1993). Bantel and Jackson (1989) found only a few, if any, positive significant effects of age diversity on innovation in top management groups (TMT). Hagendorff and Keasey (2012) found that age diversity is associated with negative shareholder returns because investors prefer experienced boards over diverse boards. Therefore:
Hypothesis 1b: Age diversity is negatively related to firm performance.
Gender diversity
women at the top altered the perceptions of lower level women about the likelihood of advancement in the firm, thereby affecting their behaviour. Thus, gender diversity sends a positive signal to other women that they are welcome in the firm (Frink et al., 2003). Companies with few women on boards are more prone to stereotyping and these women have little power. In this way it sends a negative signal to others and they lose access to the resources that female employees could have brought to the table (Frink et al., 2003). According to Frink et al. (2003) performance is maximized when woman comprised about half of a firm’s workforce. Moreover, because of the glass ceiling phenomenon, women have to demonstrate extra competence in order to reach board positions. This could lead to board of directors being more efficient because women are highly competent and hard working (Ittonen et al., 2007).
Adams and Ferreira (2009) found that women are less likely to have attendance problems than men and that more gender diverse boards devote more effort to monitoring managers. Carter et al. (2003) found a positive relationship between the amount of women on the board and market value of the firm. Adams and Ferreira (2004) found that in boards with relatively more women, more directors participate in decision-making, which could enhance their effectiveness. Adams and Ferreira (2004) found a significant association between stock market volatility and the proportion of women; firms with a lower part of women on boards have a more volatile stock price. Thus:
Hypothesis 2a: gender diversity is positively related to firm performance.
each other and people then tend to get along best with people who are similar (Frink et al., 2003). Simpson et al. (2010) stated that many women do not have the right kind of human capital to be a director of a corporation because most women do not have sufficient experience in high-level business positions. Because of those reasons group effectiveness is lowered and, ultimately, firm performance (Smither et al., 2003).
In an analysis of US firms, Adams and Ferreira (2009) found a negative relationship between the proportion of women on the board and performance. Farrell and Hersch (2005) investigated the effect of the inclusion of women to US boards but they found no evidence that adding a female to the board affects ROA or the market return to shareholders. Therefore:
Hypothesis 2b: gender diversity is negatively related to firm performance.
Nationality diversity
Nationality as a diversity measure has some benefits over ethnicity diversity; it is analytically tractable, it coincides with the implicit or explicit employee categorizations applied in many global enterprises, and it has been the basis for considerable research on individual differences (Hambrick et al., 1998). Ethnicity has the advantage of greater specificity but the categories are more ambiguous and its relevance depends on an individual’s sense of identity (Fiedler et al., 1961).
better strategic decision making (Van Veen & Marsman, 2008). Second, nationality diversity tells something about the internationalization of a firm and to what extend it is detached from its country of origin (Ruigrok & van Tulder, 1995). Also, nationality diversity can be seen as an important indicator for the transnational mindset of a company (Bartelett & Ghoshal, 1998).
Members of a multinational group will have different values. These values could affect the preference for certain task solutions and will affect their cognitions by causing them to interpret stimuli in ways that suit their value structures (Postman et al. 1948). Because of these different values, more alternatives are brought about and this increases group effectiveness because of the higher problem solving capacity of the team (Hambrick, 1998).
Hypothesis 3a: nationality diversity is positively related to firm performance.
Bochner and Hesketh (1994) found that people with different nationality perceived more discrimination in their workplace than the in-group. In a study of employees in The Netherlands, Verkuyten et al. (1993) found that individuals who were not Dutch tended to be less satisfied with their jobs than their Dutch counterparts. Watson et al. (1993) found that nationality diversity had a different influence on performance over time. In the first period homogenous groups scored higher on performance measures, but in the last time period both groups scored the same. So nationality diversity could first have a negative effect on firm performance because groups find it hard to get over their interpersonal differences. Therefore:
Functional background diversity
Board members’ professional background diversity captures how much these group members differ from each other in their experiences, competences, skills and perspectives (Huse, 2005). Different professional backgrounds represented in the boardroom should predict a greater involvement in networking, door opening and legitimating because the background diversity implies different professional categories (Huse, 2005).
An increase of members with different backgrounds may generate a variety of perspectives being considered in decision-making and it increases the likelihood of creative and innovative solutions to problems (Huse, 2005). Because of the different viewpoints, board independence and the effectiveness of the oversight function of the board could increase because people with different viewpoints ask questions that would not come from boards made up almost exclusively of executive directors (Erhardt et al., 2003). In this vein, board members’ background diversity may be associated with board effectiveness in its oversight function.
Ancona and Caldwell (1992) reported a positive indirect effect of functional diversity on innovation through an increase in frequency of communication with people outside the project group. So an advantage of functional diversity is the creation of networks with people outside the team thereby increasing access to information. Glick and colleagues (1993) found that functional diversity in top management groups is associated with more frequent communication within the team as well.
backgrounds of the team members is positively associated with innovation because this is task-related diversity (Bantel & Jackson, 1989). Bantel and Jackson (1989) also found that top management teams with different capabilities made more innovative and higher quality decisions than less diverse teams. Teams with background diversity produce a wider variety of ideas, alternatives and solutions than teams composed of people with similar demographic characteristics (Jackson, 1992).
Functional diversity can increase the chance that new knowledge will be related to knowledge already existing in the organization/team, and therefore enhancing the assimilation of that new knowledge (Daghfous, 2004). According to the cognitive resource perspective, functional diversity leads to diversity in knowledge bases and this is needed to create and combine knowledge (Webber & Donahue, 2001).
People with high functional diversity have larger and more structurally sparse networks and are therefore likely to have a good perception of where the knowledge in the firm exist and how to use it (Bunderson, 2003). Also, people with high functional diversity tend to focus on the problems that reflect their specific backgrounds and experiences (Finkelstein & Hambrick, 1996). Thus:
Hypothesis 4a: Functional background diversity is positively related to firm performance.
of process loss from slower decision making in functionally diverse teams, a negative effect on innovation may arise (Ancona & Caldwell, 1992). The alternative hypothesis concerning functional background diversity and performance is thus:
Hypothesis 4b: Functional background diversity is negatively related to firm performance.
Educational background diversity
Greater educational background diversity in top management teams has potential for better strategic decision-making (Erhardt et al., 2003). A person’s education has found to reflect personality, cognitive style, and values (Holland, 1976). The educational curriculum choices people make correspond to their personalities, attitudes, cognitive styles (Holland, 1976) and their job experiences throughout one’s career. So diverse teams with respect to education should benefit form the variety of perspectives when solving problems. This cognitive diversity among board members leads to different assumptions, information and interpretations in decision-making and thus overcomes groupthink. Groupthink occurs when a group makes faulty decisions because group pressures lead to a deterioration of mental efficiency, reality testing, and moral judgment (Janis, 1972). Groups affected by groupthink ignore alternatives and tend to take irrational actions that dehumanize other groups. Because educational diversity overcomes groupthink, better strategic decision-making is expected which should lead to higher firm performance.
Ginsberg (1990) claims that cognitive complexity can be inferred from educational level and that it is associated with a team’s capacity to confront the uncertainty of the environment and to make decisions to stimulate renewal and change in an organization. Cognitive heterogeneity also increases the chance that new information will be related to already existing knowledge (Cohen & Levinthal, 1990).
On the organizational level of analysis, Wiersema and Bantel (1989) found that top management team heterogeneity on educational curriculum was positively related to firms' change in diversification strategies, because educational diversity facilitates organizational adaption. Top management teams with higher cognitive heterogeneity have more disagreement and task conflicts when they have to solve complex and non-routine problems. This increases the motivation to debate and challenge the status quo (Michel & Hambrick, 1992). In this way more alternatives are mentioned, managers rethink their point of view and consider other alternatives. Therefore:
Hypothesis 5a: Educational diversity is positively related to firm performance.
Because of the different perspectives, educational diversity is linked to better strategic decision-making and an increased knowledge base; however these different viewpoints could also create strains in the decision-making process (Pfeffer, 1978) because communication may be more complicated. Simons (1995) stated that TMT educational diversity would be only beneficial when there was a process that allowed the team to surface its diversity in open debate. Bantel and Jackson (1989) found no relationship between educational heterogeneity and innovativeness of TMT’s. Hence:
MODERATING VARIABLES
Because the diversity literature is contradictive, the relationship between diversity and firm performance is not independently positive or negative but the context in which a team operates moderates this relationship (Canella et al., 2008). When studying the moderating effect of context, inconsistencies in past research could be explained and better understanding of the relationship between diversity and performance might be realized (Rosenberg, 1968). Several researchers have acknowledged that contextual considerations are critical in diversity research (Jackson et al., 2003; Martins et al., 2003). Context is defined as the situational setting in which workplace phenomena occur (Cappelli & Sherer, 1991). So context could enhance or minimize the direct effects of team diversity on firm performance (Joshi & Roh, 2009). When studying the effects of environmental context on the relationship between board diversity and firm performance it is important to first distinguish between observable and underlying attributes. Next, the theory about observable and underlying attributes will be integrated with the theory about the environmental context variables to derive hypotheses.
Observable and underlying attributes
There are a lot of ways in which people can be diverse with respect to underlying attributes. Differences in personality characteristics are one example. When people have different personality characteristics, it could create major differences in values of the members of a group. One type of diversity is especially important in organizations, namely diversity in skills or knowledge like educational background, functional background or occupational background because these attributes are more task-related than observable attributes and thus have a higher potential to impact on group task performance (Milliken & Martins, 1996). With high task-related heterogeneity, cognitive task are affected and people spend time and energy in assessing the appropriateness of strategic choices (Olsen et al, 2006). The logic behind this is that cognitive tasks demand experience and knowledge obtained through exposure to various educational and functional areas (Olsen et al., 2006). Regarding observable attributes, an opposite effect of diversity seems to occur. Specifically, Jehn et al. (1999) found that homogenous groups with respect to observable attributes have a higher group performance than heterogeneous groups. Heterogeneity in non-task related attributes may lead to conflict so board members are not able to reach consensus and may even avoid evaluating the strategic choices (Olsen et al., 2006).
There are barely any reasons for employees to reduce their stereotypical biases and feelings of in-group superiority (Brewer, 1991).
However, underlying attributes of diversity could lead the team members to complement each other because these underlying differences reflect a variety in task resources the team can apply (Harrison & Klein, 2008). When group members become aware of their underlying differences they start to discuss their individual contributions to the group and determine group actions (Jehn & Mannix, 2001; Stasser & Titus, 1985). As a consequence, employees try to use their unique experiences, insights and work strategies to become innovative (Chi et al., 2009) and to reach their performance goals (Harrison et al., 2002; Phillips et al., 2006; Phillips & Loyd, 2006).
Environmental complexity
The effects of workgroup diversity on workgroup performance are likely to be affected by structural aspects of the task (Van de Ven & Ferry, 1980). The performance of diverse groups is most prominent when they have complex tasks that require the input of different perspectives. When a task is simple and well-understood, group members have standardized work procedures so debates about the strategy of the task are not productive (Jehn, 1995). However, when a task is complex and not well-understood, discussing and debating competing perspectives and approaches is essential for group members to identify appropriate task strategies and to increase the accuracy of members' assessments of the situation (Jehn, 1995).
have many potential threats because of a highly competitive product market. Fewer rivals and highly concentrated product markets are more common in less complex environments. Careful planning and execution of a variety of strategic actions is relevant in highly complex environments. Thus more diverse boards better suit complex environments because of the different perspectives more problems and a greater range of solutions are discussed. Hambrick and Mason (1984) also stated that in turbulent, uncertain environments it is better to have a heterogeneous top management team while in stable environments a homogenous team is better suited. Low complexity environments do not require a broad range of options from which to choose (Bongjin et al., 2009).
Because of the cognitive diversity, heterogeneous groups can take advantage from the increase in creativity in complex environments (Wiersema & Bantel, 1989). Because of the differences in cognitive resources, when solving complex, non-routine problems, groups are more effective when composed of individuals having a variety of skills, knowledge, abilities, and perspectives (Shaw, 1976; Wanous & Youtz, 1986). Firms operating in complex environments have to cope with a lot of different external interest groups. Diverse boards have an advantage over less diverse boards because these boards include people with a variety of different perspectives and are able to respond to the different environmental factors (Wiersema & Bantel, 1992). The benefits of diversity will be enhanced because teams will also be less committed to the status quo (Hambrick, 1992) and are more aware of diverse information and its implications (Bunderson & Sutcliffe, 2002). When environments are dynamic and change fast it is important to adapt fast and heterogeneous teams have these capabilities more often. Also, the need for resources is higher in uncertain environments.
diversity when they face complex rather than simple environments. As observable attributes are non-task related, diversity in gender, age, and nationality has less potential or may even be a drawback for boards in dealing with more complex environments. This line of reasoning lead to the following hypotheses:
Hypothesis 6: Environmental complexity moderates the relationship between diversity in underlying attributes (functional and educational background) and firm performance such that this relationship is more positive or less negative when environmental complexity is higher rather than lower.
Hypothesis 7: Environmental complexity moderates the relationship between diversity in observable attributes (gender, age, nationality) and firm performance such that this relationship is less positive or more negative when environmental complexity is higher rather than lower.
Industry type moderator
The advantages of diversity may also vary with the type of industry. Market insight is more important in services firms because it requires knowledge of the target market (Richard, 1997). Also, a lot more customer involvement is necessary in services firms than in manufacturing companies. When working in the manufacturing industry it is less important to have different perspectives and to make creative decisions than when working in the services industry because the focus in manufacturing industries is on tasks involving physical skills (Richard et al., 2007).
workforce. Because of the workforce diversity they are also able to attract a diverse set of customers which should raise sales. Manufacturing firms rely more on plant and equipment, technology, and raw materials to achieve business goals (Quinn et al., 1996). Because manufacturing relies more on physical capital and equipment and less on direct customer-based interactions, diversity is less likely to impact performance (Richard et al., 2007).
Thus, type of industry (service vs. manufacturing industry) can shape the relationship between board diversity and firm performance. Because underlying diversity attributes are task-related, board diversity in functional background and educational background may promote firm performance in particular when those boards operates in a service industry rather than a manufacturing industry. However, as observable diversity attributes are non-task related, boards that are diverse on gender, age, and nationality have less potential or even a drawback to perform well when they operate in the service industry. This leads to the following hypotheses:
Hypothesis 8: Type of industry moderates the relationship between diversity in underlying attributes (functional and educational background) and firm performance such that this relationship is more positive or less negative when the firm operates in a service industry rather than a manufacturing industry.
FIGURE 1 Conceptual Model
METHODS
Data collection and sample
The data for this research was gathered using Orbis. This database contains comprehensive information from around 100 million companies worldwide. It could be used to research individual companies, search for companies by profile and create your own analysis.
Firm Performance ROA ROE Age Difference in age of board members Functional Background Number of different categories in the board over the total
board members Nationality
Number of different categories
in the board over the total board
members Gender Proportion of women on board Environmental Complexity Educational Background Number of different categories in the board over the total
board members
Information available from Orbis varies from company financials, ratios, industry research, detailed corporate and ownership structures and stock data. Data quality is guaranteed by working with the best information providers and maintaining strong relationship with them. Their accuracy and timeliness is checked on a monthly basis. The quality and accuracy checks are done by using a combination of automated and manual testing. This is done at various stages of the process and is applied to all the datasets to guarantee consistent quality levels. Orbis has information about the board of directors from 278 Dutch companies but some of these companies were already dissolved or there was a lot of board information missing. So the effective sample was eventually 143 companies. All of these companies were categorized as very large companies based on operating revenue and both publicly quoted as well as private companies were included in the sample.
Measures
Dependent variable
Firm Performance-Firm performance is measured in two ways: return on assets (ROA) and return on equity (ROE). I use these two measures because they are among the most commonly used performance measures. ROA is measured as net income divided by total assets and ROE is measured as net income divided by shareholder’s equity.
Independent variables
Age-Age of the directors is measured as the difference in age between the youngest and oldest director (Carter et al., 2010).
Gender-Women are measured as the proportion of women to total directors (Rose, 2007). To measure gender diversity Blau’s index is used. This index is calculated as D = (1 - Σ Pi² )
where P is the proportion of individuals in a category and i is the number of categories. The index increases as the representation of men and women in the organization becomes more equal (Blau, 1997). The index range from 0 to 0.5 with 0 representing homogeneity and 0.5 representing maximum gender diversity. Index values greater than 0.25 would reflect relatively high levels of heterogeneity.
Nationality-To measure nationality diversity, all the different nationalities of the board members are counted and divided this by the total amount of board members.
member’s functional specialization. Orbis contains a large amount of information about the functional background of the board of directors. After reading this information I assigned the functional background of the director in one of the twelve groups and summed all the categories for the board of directors and divided them by the total number of board members.
Education-To measure educational background diversity I created 13 different educational categories (accounting, law, business administration, engineering, HRM, economics, mathematics, finance, science, operations, marketing, history and others). Because some directors graduated from more than one study, the average number of categories per director could be higher than one. When I summed all the different categories I divided it by the total number of board members.
Moderators
Complexity-The environmental complexity dimension is defined as the number of SIC industries in which a firm operates. The number of lines of business is an established representation of a company's complexity (Dess & Beard, 1984: Huselid & Rau, 1997). A higher score indicates more complexity.
(Richard et al. 2007), because firms in these industry groups add value to raw materials by processing them into less varied and less perishable tangible goods (Morris & Johnston 1987).
Control variables
Firm size-Firm size was measured as the logarithmic function of total assets. Firm size is used by many researchers as a control variable. Fama and French (1992) found firm size to be related to market returns. I used a logarithm to account for its skewed distribution. Adjustments for firm size reduce the incidence of heteroscedasticity (Richard & Shelor, 2002). I expect a positive relation between diversity and firm size because larger firms more often appoint woman and minorities on board.
Board size-Board size is measured as the total number of directors. I include board size as control variable because larger boards bring better information because of greater knowledge from more directors to firm decision making.
The following 6 regressions will be used in this research:
Performance i,t = β0 + β1BoardSizei,t + β2FirmSizei,t + εi,t
Performancei,t = β0 + β1BoardSizei,t + β2FirmSizei,t + β3Agei,t + β4Genderi,t +
β5Nationalityi,t + β6Educationali,t + β7Functionali,t + εi,t
Performancei,t = β0 + β1BoardSizei,t + β2FirmSizei,t + β3Agei,t + β4Genderi,t +
β5Nationalityi,t + β6Educationali,t + β7Functionali,t + β8Complexityi,t + εi,t
Performancei,t = β0 + β1BoardSizei,t + β2FirmSizei,t + β3Agei,t + β4Genderi,t +
β5Nationalityi,t + β6Educationali,t + β7Functionali,t + β8Complexityi,t + β9Complexityi,t ×
Agei,t + β10Complexityi,t × Genderi,t + β11Complexityi,t × Nationalityi,t + β12Complexityi,t ×
Performancei,t = β0 + β1BoardSizei,t + β2FirmSizei,t + β3Agei,t + β4Genderi,t +
β5Nationalityi,t + β6Educationali,t + β7Functionali,t, + β8Industryi,t + εi,t
Performancei,t = β0 + β1BoardSizei,t + β2FirmSizei,t + β3Agei,t + β4Genderi,t +
β5Nationalityi,t + β6Educationali,t + β7Functionali,t + β8Industryi,t + β9Industryi,t × Agei,t +
β10Industryi,t × Genderi,t + β11Industryi,t × Nationalityi,t + β12Industryi,t × Educationali,t +
β13Industryi,t × Functionali,t + εi,t
Data analysis
The effect of board diversity on firm performance is tested using EVIEWS. This is a useful program for making regression analysis. I first used the ordinary least squares method to test hypotheses 1 to 9. Next, I used a panel technique. The reason for doing this is that OLS has some severe limitations. Most importantly, pooling the data in this way implicitly assumes that the average values of the variables and the relationships between them are constant over time and across all of the cross-sectional units in the sample (Brooks, 2008). OLS is considered inappropriate since it does not allow for differences in average ROE and ROA at the company level. The random effects panel model is selected for the cross-sectional units. I chose to use the random effects panel model because of the fact that the board structure does not change very much each year so some independent variables were very stable over time. When the fixed effects model would be used for both cross-sectional (company) and period (year) elements, a large number of observations were lost. This is a common problem when there are relatively few observations per cross-sectional unit (Greene, 2003). So a combination of both models is used; for the cross-sectional units I used random effects and for the time period I used fixed effects.
compare them and it also reduces multicollinearity. The standardized coefficient now has a slightly different meaning because standardized coefficients tell something about how increases in the independent variables affect relative position within the group.
ROA and ROE had both some variables that were very unusual in size so I removed those outliers.
RESULTS
Descriptive statistics
Table 1. contains descriptive statistics and correlations for all variables. It is important that the explanatory variables are not correlated with one another. A small degree of association between the explanatory variables will almost always occur but it will not cause too much loss of precision. A problem occurs however when the explanatory variables are highly correlated. This is called multicollinearity. The highest correlation between explanatory variables is 0.68. According to Kennedy (1979) a correlation of 0.8 or higher may be problematic from the viewpoint of multicollinearity. So there is no sign to worry about multicollinearity.
One thing that stands out from the descriptive statistics is the low percentage of women on the board of directors with an average Blau’s index of 0.04 (SD= 0.11) while an index value of 0.5 represents maximum gender diversity. Only 17% of all the firms had at least one woman present on the board so when it comes to the occupation of firms’ board seats, these positions are almost entirely restricted to men. Another interesting thing is the fact that only gender is significantly correlated to ROA. All the other variables do not have a significant relationship with ROA. For the other performance measure this is different; ROE is positively and significantly correlated to firm size, gender and both moderators.
1.13 (SD=0.59), and of nationality diversity was 0.56 (SD=0.27). The average log value of firm size was 14.18 (SD=2.46) and the average number of board members was 4.41 (SD=3.64). The complexity of the firms had an average value of 2.34 (SD=1.52).
TABLE 1 Descriptive Statistics Variable Mean S.D N 1 2 3 4 5 6 7 8 9 10 1 ROA 2.90 8.09 139 2 ROE 6.29 16.28 138 0.85*** 3 Board Size 4.41 3.64 140 0.01 -0.02 4 Firm Size 14.18 2.46 143 0.08 0.19*** 0.03 5 Age 12.40 10.31 136 0.08 0.02 0.68*** -0.00 6 Gender 0.04 0.11 143 0.16** 0.17** 0.13* -0.02 0.25*** 7 Functional 0.71 0.35 139 -0.10 -0.04 -0.48*** -0.06 -0.51*** -0.02 8 Educational 1.13 0.59 141 0.06 -0.07 -0.31*** -0.01 -0.22*** -0.14* 0.39*** 9 Nationality 0.56 0.27 138 -0.3 0.09 -0.32*** 0.12 -0.28*** -0.20*** 0.35*** 0.34*** 10 Complexity 2.34 1.52 143 0.12 0.26*** 0.17** 0.36*** -0.15 0.09 -0.19** -0.19** -0.26*** 11 Industry 0.28 0.45 143 -0.09 -0.21*** 0.25*** 0.15* -0.16* -0.01 -0.08 -0.05 0.19*** -0.21** Note: ***p<0.001 **p<0.05 * p<0.1 Hypotheses testing
Model 1 of table 2 is a base model consisting only of the control variables. Model 2 adds the main effects of age, gender, nationality, functional and educational to test hypotheses 1 to 5. In the third model, the moderator ‘complexity’ is added. Model 4 provides the interactions of the independent variables with complexity to test hypotheses 6 and 7. In the next step, model 5, the moderator ‘industry’ is added. Model 6 consist of the interaction terms of the independent variables with industry to test hypotheses 8 and 9.
account in step 1 (See models 1 and 2 in tables 2 and 3), age diversity did not have a significant effect on ROA (beta 0.15, p>0.1) or ROE (beta -0.11, p>0.1). So hypotheses 1a and 1a were not supported. Hypothesis 2a proposed a positive effect of gender diversity on firm performance, whereas hypothesis 2b proposed a negative relationship between gender diversity and firm performance. Hypothesis 2a was supported by the data, both for ROA (beta 1.26, p<0.05) and ROE (beta 2.25, p<0.05). Hypothesis 3a proposed that nationality diversity positively influences firm performance whereas hypothesis 3b stated that nationality diversity has a negative influence on firm performance. No support was found for a relationship between nationality and ROA (beta 0.00, p>0.1) or ROE (beta -2.36, p>0.1). Furthermore, functional diversity did not influence ROA (beta 0.00, p>0.1) or ROE (beta -2.08, p>0.1); so, hypotheses 4a and 4b were not supported. The last independent variable tested was educational diversity. Hypothesis 5a proposed a positive effect on firm performance, whereas hypothesis 5b proposed a negative effect. Hypothesis 5a was supported at the 10% significance level (beta 1.12, p < .10), thereby rejecting hypothesis 5b.
The effects of the different levels of functional background diversity are plotted in the low and high complexity environment, as seen in Figure 2. At high levels of functional diversity there is a lot of difference in the effects of functional diversity in the two environments. Increasing functional diversity has a more pronounced effect on firm performance in the high complexity environment than in the low complexity environment as proposed in hypothesis 6. Therefore hypothesis 6 is partly supported.
The next tested moderator is industry. Hypothesis 8 stated that industry positively moderates underlying diversity whereas hypothesis 9 proposed a negative moderation effect for observable diversity. However, this hypothesis is not supported. None of the interaction terms were significant.
FIGURE 2
Interaction between functional diversity and complexity
-5 0 5 10 15 20
Low Functional Diversity High Functional Diversity
R
O
E Low Complexity
TABLE 2
Results of OLS regression analysis
Dependent variable = ROA
Variables Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Step 1 - Controls
Board Size 0.09 -0.26 -0.38 -0.52 -0.08 -0.81 Firm Size 0.40 0.57 0.33 0.35 0.61 0.45 Step 2 - Independent Variables
Age 0.15 0.37 0.53 0.13 0.97 Gender 1.26** 1.20* 1.13* 1.28** 1.43* Nationality 0.00 0.00 -0.22 -0.02 -0.33 Functional 0.00 -0.81 -0.80 -0.92 -1.72** Educational 1.12* 1.12* 1.30* 1.09* 1.52 Step 3 - Moderators Complexity 0.65 0.67 Industry -0.79 -0.70
Step 4 - Interaction Terms
Complexity x age 0.19 Complexity x gender -0.23 Complexity x nationality -0.41 Complexity x functional 0.77 Complexity x educational -0.38 Industry x age 1.17 Industry x gender 0.53 Industry x nationality -0.37 Industry x functional -1.86 Industry x educational 0.38 Change in R² 0.07 -0.03 0.04 0.00 0.06 F-Value 0.54 1.74* 1.70* 1.14 1.62 1.76**
TABLE 3
Results of OLS regression analysis
Dependent variable = ROE
Variables Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Step 1 - Controls
Board Size -0.25 -1.15 -1.91 -1.72 -0.46 -1.19 Firm Size 3.41** 3.83** 2.57 2.45 4.06*** 3.88*** Step 2 - Independent Variables
Age -0.11 1.00 1.36 -0.05 2.07 Gender 2.52** 2.22 1.65 2.49** 1.53 Nationality -2.36 -1.49 -2.09 -1.69 -1.37 Functional -2.08 -1.75 -0.77 -2.06 -3.07 Educational 2.45 2.63 2.64 2.26 3.25 Step 3 - Moderator Complexity 3.18 3.49*** Industry -3.54 -4.13
Step 4 - Interaction Terms
Complexity x age -0.14 Complexity x gender 0.32 Complexity x nationality -2.33 Complexity x functional 3.98** Complexity x educational 0.51 Industry x age 3.53 Industry x gender -2.03 Industry x nationality 1.51 Industry x functional -3.36 Industry x educational 1.46 Change in R² 0.07 0.03 0.04 0.02 0.03 F-Value 4.25** 2.61** 2.97*** 2.35*** 2.63*** 2.02**
The results of the random panel effects model are shown in tables 3 and 4. After adding the controls in step 1, the independent variables are added. Only gender had a positive significant effect on ROA (beta 1.22) and ROE (beta 2.42) at the 10% level. Thus hypothesis 2a was supported. The rest of these variables did not have a significant effect on ROA or ROE so hypotheses 1 and 3-5 are not supported. The next step is to test for moderation effects. Complexity had a positive significant effect on ROE (beta 3.43, p<0.05). However, only the interaction term complexity × functional had a significant influence on ROE. The rest of the interaction terms did not affect ROA or ROE. The variable industry did not have an influence on firm performance and also none of the interaction terms were significant. In sum, hypotheses 7 to 9 are not supported.
Figure 3 illustrates the interactions between functional background diversity and complexity. When functional diversity is high, there is a large difference between low and high complexity environments. The influence of functional diversity on firm performance is most positive when functional diversity occurred in combination with a complex environment. This is consistent with hypothesis 6. So hypothesis 6 is partially supported.
FIGURE 3
Interaction between functional diversity and complexity
-5 0 5 10 15 20
Low Functional Diversity High Functional Diversity
R
O
E Low Complexity
TABLE 4
Results of random effects regression analysis
Dependent variable = ROA
Variables Model 1 Model 2 Model 3 Model 4 Model 5 Model 6
Step 1 - Controls
Board Size 0.13 -0.31 -0.42 -0.54 -0.13 -0.81 Firm Size 0.22 0.44 0.21 0.25 0.46 0.37 Step 2 - Independent Variables
Age 0.20 0.42 0.57 0.19 1.07 Gender 1.22* 1.16* 1.09* 1.23** 1.31* Nationality -0.25 -0.07 -0.30 -0.09 -0.34 Functional -0.83 -0.72 -0.70 -0.83 -1.58* Educational 0.91 0.94 1.10 0.88 1.30 Step 3 - Moderators Complexity 0.67 0.69 Industry -0.81 -0.76
Step 4 - Interaction Terms
Complexity x age 0.23 Complexity x gender -0.30 Complexity x nationality -0.44 Complexity x functional 0.82 Complexity x educational -0.34 Industry x age 1.38 Industry x gender 0.37 Industry x nationality -0.31 Industry x functional -1.72 Industry x educational 0.28 Change in R² 0.05 0.01 0.01 0.00 0.05 F-Value 0.12 1.05 1.07 0.77 1.00 1.27
TABLE 5
Results of random effects regression analysis
Dependent variable = ROE
Variables Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Step 1 - Controls
Board Size 0.04 -1.34 -2.06 -1.75 -0.67 -1.28 Firm Size 3.06** 3.64** 2.32 2.21 3.88** 3.70** Step 2 - Independent Variables
Age 0.31 1.38 1.60 0.39 2.33 Gender 2.42* 2.11* 1.69* 2.39** 1.55 Nationality -2.44 -1.57 -2.04 -1.82 -1.55 Functional -1.60 -1.22 -0.43 -1.56 -2.46 Educational 1.20 1.42 1.46 1.03 2.03 Step 3 - Moderators Complexity 3.43** 3.71*** Industry -3.55 -3.99
Step 4 - Interaction Terms
Complexity x age -0.19 Complexity x gender 0.28 Complexity x nationality -2.15 Complexity x functional 3.65** Complexity x educational 0.77 Industry x age 3.32 Industry x gender -1.94 Industry x nationality 1.60 Industry x functional -3.15 Industry x educational 1.38 Change in R² 0.05 0.03 0.03 0.01 0.02 F-Value 1.67 1.56 1.89* 1.54* 1.61 1.29
CONCLUSION
The objective of this study was to help reconcile the mixed findings of earlier research on the relationship between board diversity and firm performance by both expanding the diversity theory and examining environmental context. Because of the inconsistencies in theories about diversity, first competitive hypotheses were developed. Next, the moderating effects of context were researched since the relationship between board diversity and firm performance is not independently positive or negative. To eventually arrive at an answer on how board diversity influences firm performance, the literature about observable and underlying attributes of diversity were integrated with the theory about the contextual variables.
A couple of inferences could be drawn from this research. First, I found no relationship between age, nationality and functional diversity and firm performance and I found only a small positive effect of educational diversity on ROA when using OLS. An explanation for this could be the fact that observable and underlying attributes have different time dependence in their implications for team functioning (Harrison et al., 1998). Observable attributes are readily observable and the diversity effects decrease with time because people get to know each other. With underlying attributes it takes more times for the diversity effects to become prominent. When differences in values and attitudes eventually become apparent, conflicts may arise. This could explain why some underlying diversity measures have a negative effect while most observable attributes have a positive effect.
than one board position. So the argument from the old boys network could possibly be applied.
It is also possible that board diversity does not influence performance because of all the other organizational and environmental factors like the economy that overrule the influence of board diversity. Additionally, there are a lot of variables that mediate the relationship between board diversity and firm performance and these mediators also have their own moderators. Some environmental and organizational effects in this research were captured using panel least squared regression but there are still a lot effects left.
Educational diversity only had a small significant effect on ROA using OLS but when using a stricter test like panel least squares, this effect disappeared. The lack of a highly significant relationship could be due to the fact that the task board members have to carry out, do not require very specific knowledge. As long as people have experience in a board function or have a long business life they are able to understand the information provided and no specific educational background is needed.
An alternative explanation why the diversity measures did not influence firm performance may be because of the positive effects of creativity and problem solving that are nullified by the coordination problems diverse groups have to cope with. This could be due to the fact that diverse groups have members with different experiences and as a consequence have different perspectives on key issues or problems (Jackson et al., 1991). This could lead to discomfort for the group members and lower integration within the group (Jackson et al., 1991). Group diversity thus could have a negative impact on people’s feeling of satisfaction through decreasing individuals' sense of identification or social integration within the group (Ancona & Caldwell, 1992). The more similar people are in background variables like attitudes the more they initially are attracted to each other (Kanter, 1977).
Gender diversity was the only diversity measure that was significantly related to ROA and ROE for both the OLS and Panel model. This effect could be due to the positive message it sends to stakeholders. It shows that women are welcome which influence the perceptions of lower-level woman and potential applicants. Another explanation is that women have different perceptions than men which influence strategic decision-making and creativity. This finding is remarkable because it contradicts the study of Frink et al. (2003) in which they stated that boards should be comprised about half of women to benefit from diversity. This research shows that for gender diversity to have a positive effect on performance it is not necessary to have at least 50% of women on board since in this sample the average Blau’s index is only 0.04.
Theoretical and practical implications
research made a distinction between underlying and observable attributes and integrated this theory with the theory about environmental context variables. Third, this study tested the effects of board diversity on organizational level. The result is a more complex, contextual understanding of the impact of diversity on firm performance.
The findings of this study have also some practical implications. First, the results show that the financial performance of a firm can be increased by including women on the board of directors. This might pose a challenge because most boards were very homogenous when it comes to gender diversity. Firms will therefore have to reconsider their recruitment procedures. Second, complexity moderated the relationship between functional diversity and firm performance. So when selecting members with different functional backgrounds, it is important to make sure they have the context that will support functional diversity. Organizations that operate in a complex environment will realize the benefits from different functional backgrounds.
Limitations
diversity and firm performance. Aspects of organizational context where could be thought of are: climate, culture and leadership (Joshi & Roh, 2009).
Future research should also look at board actions instead of focusing on the direct link between board diversity and firm performance. Adams and Ferreira (2009) found that in firms with more women in boards, the relationship between executive officer turnover and performance is more sensitive. So when looking at actual behaviour of board members, more could be explained about the relationship between board composition and performance. Other important mediators where could be thought of are the quality of decision-making, quality of leadership and the quality of personnel.
The sample consists of very large firms. When measuring board diversity, very detailed information is needed and this is not readily available for smaller firms. This non-availability of data limits the generalizability of this research. Future research should also investigate the hypotheses in smaller firms.
REFERENCES
Adams, R. and Ferreira, D., 2004. Gender Diversity in the Boardroom. Finance Working Paper from Stockholm School of Economics.
Adams, R. and Ferreira, D., 2004. Diversity and Incentives in Teams: Evidence from corporate boards. Finance Working paper from Stockholm School of Economics.
Ali, M., Kulik, C.T. and Metz, I., 2011. The gender diversity-performance relationship in services and manufacturing organizations. International Journal of Human Resource Management, vol. 22, issue 7, p1464-1485.
Ancona, D.G. and Caldwell, D. F., 1992. Demography and design: Predictors of new product team performance. Organization Science,. vol. 3, Issue 32, p1-34.
Bantel, K.A. and Jackson, S.E., 1989. Top Management and Innovations in Banking: Doe the Composition of the top team make a difference? Strategic Management Journal, vol. 10, p107-124.
Barney, J.B., 1991. Firm Resources and Sustained Competitive Advantage. Journal of Management. vol. 17, p99–120.
Black, S.E. and Lynch, L.M., 2004. What is Driving the New Economy? The Benefits of Workplace Innovation. Economic Journal, vol. 114, p97–116.
Bongjin, K., Burns, M. and Prescott, J., 2009. The Strategic Role of the Board: The Impact of Board Structure on Top Management Team Strategic Action Capability. Corporate Governance: An International Review, vol. 17, p728-743.
Bowen, D.E. and Schneider. B., 1988. Services Marketing and Management: Implications for Organizational Behavior. Organizational Behavior, vol. 10, p43–80.
Blau, P. M., 1977. Inequality and heterogeneity. (NewYork: Free Press).
Brewer, M. B., 1991. The social self: On being the same and different at the same time. Personality and Social Psychology Bulletin, vol. 17, p475-482.
Brooks, C., 2008. Introductory Economics for Finance. (Cambridge: University Press).
Bunderson, J. S., 2003. Team member functional background and involvement in management teams: Direct effects and the moderating role of power centralization. Academy of Management Journal, vol. 46, p458–473.
Carter, D.A., Simkins, B.J., Simpson, W.G., 2003. Corporate Governance, Board Diversity, and Firm Value. Financial Review, vol. 38, issue 1, p33-53.
Carter, D.A., D'Souza, F. and Simkins, J., 2010. The Gender and Ethnic Diversity of US Boards and Board Committees and Firm Financial Performance. Corporate Governance: An International Review, vol. 18, issue 5, p396-414.
Cappelli, P. and Sherer, P., 1991. The missing role of context in OB: The need for a meso-level approach. Research in organizational behaviour, vol.13, p55–110.
Caves, B.E. and Ghemawat, P., 1992. Identifying mobility barriers. Strategic Management Journal, vol. 13, p1–12.
Certo, S. T., Lester, R. H., Dalton, C.M. and Dalton, D. R., 2006. Top management teams, strategy and financial performance: a meta-analytic examination. Journal of Management Studies, vol. 43, p813–39.
Chi, N-W., Huang, Y-M. and Lin, S-C., 2009. A double-edged sword? Exploring the curvilinear relationship between organizational tenure diversity and team innovation: The moderating role of team-oriented HR practices. Group & Organizational Management, vol. 34, p698-726.
Cohen, W.M. and Levinthal, D.A., 1990. Absorptive capacity: a new perspective of learning and innovation. Administrative Science Quarterly, vol. 35, issue 1, p128–152.
Cox, T.H., 1993. Cultural diversity in organizations: Theory. research. and practice. (San Francisco: Berrett- Koehler).
Cox, T., Lobel, S. and McLeod, P., 1991. Effects of ethnic group cultural differences on cooperative and competitive behaviour on a group task. Academy of Management Journal, vol. 34, p827–847.
Cummings, A., Zhou, J. and Oldman, G., 1993. Demographic Differences and Employee Work Outcomes: Effects of Multiple Comparison Groups, International Journal
of Conflict Management
Daghfous, A., 2004. Absorptive Capacity and the Implementation of Knowledge-Intensive Practices. SAM Advanced Management Journal, p21–27.
Damon, W., 1991. Problems of direction in socially shared cognition. (Washington, DC: APA Books).
Dearborn, D. C. and Simon, H.A., 1959. Selective perception: A note on the department identifications of executives. Sociometry, vol. 21, p140-144.
Dess, G. and Beard, D.W., 1984. Dimension.s of Organizational Task Environments. Administrative Science Quarterly, vol. 19, issue 52.
Earley, P. C. and Mosakowski, E., 2000. Creating hybrid team cultures: An empirical test of transnational team functioning. Academy of Management Journal, vol. 43, p26–49.
Ely, R. J., 1994. The effects of organizational demographics and social identity on relationships among professional women. Administrative Science Quarterly, vol. 39, p203-238.
Erhardt, N. L., Werbel. J. D. and Shrader. C. B., 2003. Board of director diversity and firm financial performance. Corporate Governance, vol. 11, p102–11.
Fama, E. R. and French. K., 1992. The cross-section of expected stock returns. Journal of Finance, vol. 47, p427–465.
Farrell, K. A. and Hersch, P.L., 2005. Additions to corporate boards: the effect of gender.
Journal of Corporate Finance, vol. 11, issue 1/2, p85-106.
Fenwick, G.D and Neal, D.J., 2001. Effect of Gender Composition on Group Performance. Gender, Work & Organization, vol. 8, issue 2, p205-226.
Fiedler, F.W.M. and Oonk, S., 1961. Performance on laboratory tasks requiring group creativity. Acta Psychologica. vol. 8, p100-119.
Fine, C., 1998. Clockspeed: Winning industry control in the age of temporary advantage. (Perseus Books).
Frink, D.D., Robinson, R.K., Reithel, B., Arthur, M.M., Ammeter, A.P., Ferris, G.R., Kaplan, D.M., Morrisette, H,S., 2003. Group & Organization Management, vol. 28, issue 1, p127-147.
Finkelstein, S. and Hambrick, D. C., 1996. Strategic leadership: Top executives and their effects on organizations. (Minneapolis/St. Paul: West).
Ginsberg, A., 1990. Connecting diversification to performance: a sociocognitive approach. Academy of Management Review, vol. 15, p514-35.
Glick, W. H., Miller, C. C. and Huber, G. P., 1993. The impact of upper echelon diversity on organizational performance. Organizational change and re-design: Ideas and insights for improving performance, p176-124.
Greene, W. H., 2003. Econometric analysis. (Upper Saddle River, NJ: Prentice Hall).
Gujarati, D. N. and Porter, D. C., 2009. Basic econometrics. (New York: The McGraw-Hill Companies Inc).
Haleblian, J. and Finkelstein, S., 1993. Top management team size, CEO dominance, and firm performance: the moderating roles of environmental turbulence and discretion, Academy of Management Journal, vol. 36, p844–63.
Hambrick, D. C., 1994. Top management groups: a conceptual integration and reconsideration of the team label. Research in Organizational Behaviour, vol. 16, p171–213.
Hambrick, D.C., Davison, S.C., Snell, S.A. and Snow, C.C., 1998. When Groups Consist of Multiple Nationalities: Towards a New Understanding of the implications. Organization studies, vol. 19, p181-205.
Hambrick, D.C. and Mason, P.A., 1984. Upper echelons: the organization as a reflection of its top managers. Academy of Management Review, vol. 9, p193-206.
Hambrick, D.C. and MacMillan., 1985. Efficiency of Product R & D in Business Units: The Role of Strategic Context. Academy of Management Journal, vol. 28, p527-47.
Harrison, D.A. and Klein, K.J., 2007. What’s the difference? Diversity constructs as separation, variety, or disparity in organizations. Academy of Management Review, vol. 32, issue 4, p1199-1228.
Hermalin, B.E. and Weisbach, M.S., 2003. Boards of directors as an endogenously determined institution: A survey of the economic literature. Economic Policy Review, vol. 9, p7–26.
Hilary, G. and Hui, K.W., 2008. Does religion matter in corporate decision making in America? Journal of Financial Economics, vol. 93, issue 3, p455-473.
Hitt, M.A., Hoskisson, R. E. and Kim, H., 1997. International diversification: effects on innovation and firm performance in product-diversified firms. Academy of Management Journal, vol. 40, p767–99.
Hoffman, L., 1959. Homogeneity and member personality and its effect on group problem solving. Journal of Abnormal and Social Psychology, vol. 58, p27–32.
Holland, J.L., 1976. Vocational preferences. Handbook of industrial/organizational psychology, p521-570.
Huse, M., 2005. Accountability and creating accountability: a framework for exploring behavioural perspectives of corporate governance. British Journal of Management, vol.16, p65–79.
Jackson, S., 1992. Consequences of Group Composition for the Interpersonal Dynamics of Strategic Issue Processing. Advances in Strategic Management, vol. 8, p345-82.
Jackson, S. E., Joshi, A. and Erhardt, N. L., 2003. Recent research on team and organizational diversity: SWOT analysis and implications. Journal of Management, vol. 29, p801–830.
Janis, I. L., 1972. Victims of Groupthink: a Psychological Study of Foreign-Policy Decisions and Fiascos. (Boston: Houghton Mifflin).
Jayne, M.E.A. and Dipboye, R. L., 2004. Leveraging diversity to improve business
performance: Research findings and recommendations for organizations. Human Resource
Management, vol. 43, issue 4, p409–424.
Jehn, K.A., 1995. Multimethod examination of the benefits and detriments of intragroup conflict. Administrative Science Quarterly, vol. 40, p256-282.
Jehn, K., Chadwick and Bushnell, 1998. To Agree or Not to Agree: The Effects of Value Congruence, Member Diversity, and Conflict on Workgroup Outcomes. International Journal of Conflict Management.
Jehn, K.A., Mannix, E. A., 2001. The dynamic nature of conflict: a longitudinal study of intragroup conflict and group performance. Academy of Management Journal, vol. 44, issue 2, p238-251.
field study of diversity, conflict, and performance in workgroups. Administrative Science Quarterly, vol. 44, p741-763.
Jensen, M. C. and Meckling, W.H., 1976. Theory of the firm: managerial behaviour, agency costs and ownership structure. Journal of Financial Economics, vol. 3, p305–360.
Joshi, A. and Roh, H., 2009. The role of context in work team diversity research: a meta-analytic review. Academy of Management Journal, vol. 52, issue 3, p599-627.
Kanter, R.M., 1977. Some Effects of Proportions on Group Life: Skewed Sex Ratios and Responses to Token Women. The American Journal of Sociology, vol. 82, p965–990.
Keck, S. L., 1997. Top management team structure: Differential effects by environmental context. Organization Science, vol. 8, p143–156.
Kennedy, P., 1979. A guide to econometrics (Cambridge Mass.: MIT Press)
Kravitz, D.A., 2003. More Women in the Workplace: Is There a Payoff in Firm Performance? Academy of Management Executive, vol. 17, issue 3, p148-149.
