Abstract
This paper examines the influence of skill diversity on board level. Prior studies focused predominately on demographic factors with inconclusive results. This research tried to focus more on the fundamental characteristic rather than superficial characteristics. This paper is based on panel data from 2009 till 2019 of UK listed (FTSE 250) firms. I have conducted multiple models; the fixed effect model is the most appropriate model. Based on the test results, I can conclude that that skill diversity leads to better financial performance based on the tobin’s q. However, doing the same analysis with accounting leads to insignificant
Tabe of Contents
1. Introduction ... Fout! Bladwijzer niet gedefinieerd. 2. Literature Review ... Fout! Bladwijzer niet gedefinieerd.
2.1 Underlying theories of board diversity ... Fout! Bladwijzer niet gedefinieerd. 2.1.1 Agency theory ... Fout! Bladwijzer niet gedefinieerd. 2.1.2 Resource dependence theory ... Fout! Bladwijzer niet gedefinieerd. 2.1.3 Human capital theory ... Fout! Bladwijzer niet gedefinieerd. 2.2 Empirical research on skill diversity and financial performance ... Fout! Bladwijzer niet gedefinieerd. 2.3 Hypotheses ... Fout! Bladwijzer niet gedefinieerd.
3. Conceptual Model ... Fout! Bladwijzer niet gedefinieerd. 4. Methodology ... Fout! Bladwijzer niet gedefinieerd.
4.1 Data ... Fout! Bladwijzer niet gedefinieerd. 4.2 Variable measurement ... Fout! Bladwijzer niet gedefinieerd. 4.2.1 Dependent variables - Financial performance ... Fout! Bladwijzer niet gedefinieerd. 4.2.2 Independent variables – Skill diversity and Skill-match ratio ... Fout! Bladwijzer niet gedefinieerd. 4.2.3 Control variables ... Fout! Bladwijzer niet gedefinieerd. 4.3 Empirical model ... Fout! Bladwijzer niet gedefinieerd. 4.3.1 OLS regression – Pooled ... Fout! Bladwijzer niet gedefinieerd. 4.3.2 OLS regression - Fixed effects ... Fout! Bladwijzer niet gedefinieerd. 4.3.3 OLS regression - Random effects ... Fout! Bladwijzer niet gedefinieerd. 4.3.4 Multi-collinearity ... Fout! Bladwijzer niet gedefinieerd. 4.3.5 Endogeneity ... Fout! Bladwijzer niet gedefinieerd.
5. Results & Analysis ... 24
5.1 Descriptive statistics ... 24
5.2 Financial performance and skill diversity ... 25
5.2.1 Model selection ... 29
5.3 Financial performance and committee skill-match ratio ... 29
5.3.1 Model selection ... 33
5.4 Robustness check ... 33
6. Conclusion ... 35 7. References ... Fout! Bladwijzer niet gedefinieerd. 8. Appendix ... Fout! Bladwijzer niet gedefinieerd.
1. Introduction
Diversity among board members is critical for the effective functioning of a board and the overall performance of a business. For instance, a finance-orientated board may lack expertise in sales or marketing to drive the business forward (Fedyk et al., 2017). Similarities in terms of educational background and professional experience may lead to a narrow focus of the board, resulting in a groupthink mentality (Custodio et al., 2019). A higher male to female ratio, or vice versa, in the board will disrupt the gender balance that is crucial for bringing diverse perspectives and insights (Bernile et al., 2018). On the other hand, it is also plausible that board diversity would exacerbate conflicts and interrupt board’s decision-making process, which will make consensus-attainment harder and the resulting outcomes more erratic (Bernile et al., 2018). This alternative view implies that board diversity leads to worse rather than better firm performance.
Past research on board diversity has shown several discrepancies regarding its impact on financial performance. Post and Byron (2013) tried to explain those discrepancies by conducting a meta-analysis on gender diversity and financial performance. Their purpose is to reconcile the disparate findings and address shortcomings in the existing literature. Their study not only examines whether female board representation affects firm performance, but also considers the conditions that may alter this relationship. By capitalizing on the fact that the existing literature on this topic spans firms in 35 countries and five continents, their study examines whether firms’ legal/ regulatory and socio-cultural context can explain the mixed results. The authors concluded that a more gender-diversified board that yields higher market performance depends on the socio-cultural context in which board gender diversification occurs.
goes far beyond than the demographic characteristics of board members (Oliveira et al., 2017). Firms should view diversity in a more broad sense, such as race, ethnicity, education, political preference, skill set, geography, socio-economic background, and create a high-performance environment, which reflects and represents the wide-ranging interests of stakeholders.
Oliveira and Nisbett (2017) state that cognitive diversity, such as political preference or someone’s skill set is more fundamental for board diversity than demographic factors because a demographically diverse board can still be a homogeneous group. Moreover, people who expect social groups to think differently for different types of judgments may be erroneously stereotyping, expecting variation within groups to be small and variation between groups to be large. The authors concluded that the idea that cognitive diversity improves crowd judgment is well supported, but the cognitive variation in people’s judgments or how they think is hard to directly assess. Therefore, people commonly use social diversity as a proxy for cognitive diversity and expect that people who differ externally may also differ cognitively. These differences are erroneously expected to translate into performance benefits.
2. Literature Review
This chapter examine the relationship between financial performance and skill diversity in the board of directors, this chapter elaborates on the current important concepts regarding the diversity debate. Thereafter, previous empirical research will be discussed. This chapter will be concluded with a proposed hypothesis for this research.
2.1 Theories of board diversity
2.1.1 Agency theory
Bigger firms have a separation of ownership and management, which may lead to a principal–agent problem between management (the agent) and shareholders (the principals). The management and the shareholders may have different interests. Shareholders are usually more performance-focused because their main objective is to earn profit. Management, however, has to take into account the interest of different stakeholders. For example, they are responsible for a good working environment. Management is also concerned about its reputation and relationship with its employees and other stakeholders. These interests may not lead to higher profits (Tirole, 2006 & Fama 1980). Corporate governance is needed to coordinate and align the interests of the shareholders and those of the management. Corporate governance can be seen as a collection of mechanisms, processes and relations by which corporations are controlled and operated (Adams, 2015; Coombes et al., 2004).
Corporate governance systems differ across countries. In civil law countries (also known as Anglo-Saxon countries), such as the UK or the US, a one-tier board structure is commonly used to fulfil their responsibilities of supervision. However, in common law countries, such as Germany or the Netherlands, a two-tier board structure is the norm (Doupnik & Perera 2019). Firms with a two-tier board structure consist of two separate boards, namely a supervisory and a management board (Doupnik & Perera 2019). The management board can be subdivided into executive directors and the supervisory board consists of non-executive. This implies that the tasks between supervisory and a management board are separated. The main task of the supervisory board is monitoring, appointing, and dismissing the members of the management board (Jungmann, 2006). On the other hand, management is accountable for firm’s everyday business. A separated board structure increases the independency of its board members, which is a key for better performance, monitoring, and decision making. However, more independent directors also lead to more agency conflicts because the interests of owners and managers may not always align (Doupnik & Perera 2019).
Boards in civil law countries consist of a one-tier boards. A one-tier board less clear, because the distinction between the function of executive and non-executive directors is harder to make. However, in general it can be assumed that non-executive directors have a control function and directors are usually elected or dismissed by the shareholders of the company (Jungmann, 2006). The decision-making process in a one-tier board however, is more efficient. Because, board decisions that need approval of the non-executive directors are taken by only one-tier. This will lead to a faster and a more simplified the decision-making process. A weakness of a one-tier board structure is that non-executive directors are less independent towards the executive directors because, they are in the same team, there is no clear separation.
Moreover, committees also allow boards to divide work into manageable sections and contribute to the effectiveness and efficiency of board’s decision making-process (Doupnik & Perera 2019).
2.1.2 Resource dependence theory
Another important theoretical framework, which underpins the effects of board diversity on firm’s performance, is based on the resource dependence theory. This theory states that external resources influence the behavior of a company (Pfeffer & Salancik, 1978). Based on this principle, an organization possesses many resources, including knowledge, information, assets, organizational processes, capabilities, and firm attributes. By effectively managing these resources, a company can create a value-enhancing strategy and subsequently gains a competitive advantage. This implies that a firm which utilize an apply their resources in an efficient way, create a competitive advantage. An important take way is that the more unique these resources are, the harder it is for the competition to duplicate this (Tirole, 2006). The board of directors is an important provider of resources (known in economic theory as board capital). That can influence the firm financial performance (Hillman & Haynes, 2010).
2.1.3 Human capital theory
Another framework which is affiliated to the resource dependence theory is the human capital theory. According to the human capital theory, employees, management and board members are important factors (resources) for getting a competitive advantage (Tirole, 2006). The difference between resource dependence and human capital theory is the focus area – resource dependence theory focuses on multiple resources whereas human capital theory only focuses on information and expertise resource. Basically, the resource-based theory and the human capital theory complements each other. The resource-based theory is broader, it encompassed every resource that may influence firm financial performance. Becker (1962) states that human capital is more emphasized on personal characteristics, such as experience, habits, personality, education, and skills, which can be used to benefit the company (Terjesen et al., 2009; Carter et al., 2010). Therefore, it is critical to form the board of directors in a way that it is difficult to duplicate by its competitors (Tirole, 2006).
Human capital can be considered as the added value that everyone in the firm provide through the right utilization and composition of skill set, expertise, experience, and know-how. The combined human capital might lead to better problem solving within a firm. Skill set, expertise, experience a firm in other words, human capital leaves the firm when an employee leaves. Human capital also entails how effectively a firm uses its employees’ resources, as measured by innovation and creativity. The reputation of a firm is important, because it affects the type employer it draws to the firm and consequently also affects the human capital of a firm (Tirole, 2006).
2.2 Empirical research on skill diversity and financial
performance
homogeneous groups, since there can be a lot of cognitive diversity within homogeneous social groups or there might not much cognitive diversity within a demographic diverse group. For example, not all women think alike, not all liberals think alike, and so forth. People who expects social groups to think very differently for these types of judgments may be erroneously stereotyping (Oliveira et al. (2017). However, demographic diversity on boards, juries, and other influential decision-making teams helps ensure that the interests and values of a diverse population are fairly represented and addressed. These findings are in line with the resource-based view and human capital theory. These theories states board members with different resources (e.g. skills, experience, knowledge) complement each other. This will lead to better financial performance (Terjesen et al., 2009; Carter et al., 2010).
and are thus more effective in monitoring managers, which is important performance of a firm (Fama et al., 1983).
Despite all that evidence, some scholars concluded that skill diversity leads to worse financial performance. Adem et al. (2018) found that firms whose directors have more commonality in the skill sets usually have a better firm performance. Hambrick (1996) concluded that factors such as, different professional expertise, career and background may influence the way team members interpret information. The author also concluded that misunderstanding/ misinterpretation and disagreement can lead to a less effective decision-making processes within a team with many skills. Similarly, Garlappi et al. (2017) show that when directors have heterogeneous priors, boards may underinvest in multi-stage projects because they anticipate future disagreement. Rivas (2012) argues that having common ground among group members can overcome some of the problems of heterogeneous teams. These findings are in line with the social identity theory. Fedyk & Hodson (2017) investigated which particular skills lead to an outperformance of the market. The authors found that firms with higher percentages of directors with expertise in sales tend to outperform firms with less emphasis on this business area whereas firms with higher percentages of directors with administrative skills systematically underperform. Furthermore, the financial and manufacturing industry outperforms its competitors if a board is composed of more financial orientated people.
with higher firm values and greater innovation. Dass et al. (2013) find that the positive association between outside director experience and firm value extends to directors experience in immediately upstream and downstream industries (i.e. in supplier and customer industries). All these findings are consistent with the resource dependence and human capital theory (Terjesen et al., 2009; Carter et al., 2010).
On the contrary, some researchers found negative effects of directors with related industry experience (Rajan, Servaes, & Zingales, 2000). They concluded that outside directors with industry knowledge dispassionately advise CEOs in determining the optimal levels of investment in the industries in which the directors have experience. Alternatively, directors with experience in one of a diversified firm’s several industries may instead serve as biased for the divisions in which they have experience, and consequently, over-allocate capital to these segments. Directors with related experience also may not put as much effort into monitoring their favored segments or be unable to provide high-quality advising on the investment prospects of divisions in industries outside their expertise.
2.3 Hypotheses
Overall, diversity is still an ongoing discussion in the corporate finance literature (Campbell & Mínguez-Vera, 2008). Past research found that diversity may lead to an increase in creativity, innovation, board independence, better decision making, tougher monitoring, and higher financial performance (Baixauli-Soler et al., 2015; Nguyen, et al. 2010; Adams & Ferreira, 2009; Gul et al., 2011). However, diversity studies are inconclusive, so some studies concluded that diversity have some negative consequences. More diversity may be more time-consuming, and it may to more conflicts due to more opposing opinions. The bottom line is that there is no theory that confirms a positive relationship between diversity and firm financial performance. However, there are theories that give some arguments that higher skill diversity in the board of directors will lead to better financial performance compared to homogeneous boards. Consequently, this leads to the following hypothesis:
In addition, this study examines if boards with relatively more boards members assigned to a committee that matches their skill set influence firm’s financial performance. For example, it makes sense that a CFO will be assigned to an audit, finance, or risk committee, but this is not always the case. So, it can be hypothesized that boards with a higher percentage of members assigned to a committee in their area of expertise may lead to a higher financial performance. The principle behind board committees is to utilize a specific skillset or talent of an individual board member to advice and educate the other board members on particular areas of concern (Jungmann, 2006; Doupnik & Perera, 2019). When board members are assigned to a committee with the matching skill set, this may lead to more relevant expertise in a committee and better recommendation from committees, which in turn may increase firm’s financial performance (Doupnik & Perera, 2019). Consequently, this leads to the second hypothesis:
3. Conceptual Model
Skill Match ratio
- Skill matchSkill Diversity
- Unique skills
Financial
Performance
- ROA - Tobin’s Q+
4. Methodology
This chapter covers the method that has been used to test the relationship between skill diversity on board level and firm’s financial performance. First, a description of the data is provided. Second, the variable measurements and description for this research will be expounded. This chapter concludes with the analysis and techniques that are used to test the proposed hypotheses.
4.1 Data
For data collection, I used the FTSE 250 index which contains the 250 biggest UK companies that are listed on the London Stock Exchange. Financial services and utility firms are excluded from the sample because such firms belong to regulated industries that make their corporate decisions different from other firms (Bernile et al., 2018; Perryman et al., 2016; Sila et al., 2016). Moreover, firms that have missing or inconsistent board member information or financial data, were also excluded from the sample. The information on board of directors is gathered from the BoardEx database and the financial performance information is collected from the Orbis and Compustat database. Missing information data was retrieved from firms’ annual reports. The final sample resulted in 126 firms, including skill diversity and financial performance data for the period 2009 to 2019. The panel data has multiple advantages over cross-sectional data. For example, panel data decreases the likelihood of heterogeneity of cross-section units over time. Moreover, Panel data gives more informative information, larger sample size, more variability, less collinearity among the variables, more degree of freedom, and more efficiency (Brooks 2015). In Table 1 below, the distribution of firms per industry is given.
Table 1: Industry Composition
4.2 Variable measurement
4.2.1 Dependent variables – Firm financial performance
Firm’s financial performance is the variable that will be used in this research. Firm’s financial performance can be measured in two ways, by using accounting-based measures, such as ROA, or market-based measures, such as Tobin’s Q (Hull, 2015). The difference between these two indicators is that accounting data based on the past or short-term financial performance, which is subjected to tax rules, depreciation schedule and accounting standards. This makes the accounting data less trustworthy as a performance measure (Rose, 2007). However, accounting data as a proxy for the financial performance is accepted and it is being used in the past research (see Adem et al. 2018). The ROA will be calculated as follows:
𝑅𝑒𝑡𝑢𝑟𝑛 𝑜𝑛 𝑎𝑠𝑠𝑒𝑡𝑠 =𝑁𝑒𝑡 𝐼𝑛𝑐𝑜𝑚𝑒 Total Assets
Besides an accounting-based proxy (ROA), This research will also rely on market-based measure (Tobin’s q) for firm performance. The market-based measure (Tobin’s q) is a proxy that is based long term and future financial performance (Rose, 2007). One caveat has to be made; this proxy is based on the Efficient-market hypothesis. The hypothesis states that the stock price is the true value, all in information of a company is included in the stock price (Rancis & Kim, 2013). However, this assumption is probably not true (Tirole, 2016). For the market-based measure, I will use Tobin’s q that can be calculated as follows:
𝑇𝑜𝑏𝑖𝑛′𝑠 𝑞 =𝐸𝑞𝑢𝑖𝑡𝑦 𝑚𝑎𝑟𝑘𝑒𝑡 𝑣𝑎𝑙𝑢𝑒 + 𝑙𝑖𝑎𝑏𝑖𝑙𝑖𝑡𝑖𝑒𝑠 𝑚𝑎𝑟𝑘𝑒𝑡 𝑣𝑎𝑙𝑢𝑒 𝐸𝑞𝑢𝑖𝑡𝑦 𝑏𝑜𝑜𝑘 𝑣𝑎𝑙𝑢𝑒 + 𝑙𝑖𝑎𝑏𝑖𝑙𝑖𝑡𝑖𝑒𝑠 𝑏𝑜𝑜𝑘 𝑣𝑎𝑙𝑢𝑒
4.2.2 Independent variables – Skill diversity and Skill-match ratio
In addition to that, I will examine whether the skills simply match the committee assignments directors have. The second independent variable used in the research is the skill-match ratio. To capture the skill-skill-match ratio, I will specifically look at how many board members are assigned to a committee with a relevant skill set and divide it by the total number of board members assigned to a committee. The underlying assumption is that when board members are assigned to a committee with a matching skill, this may lead to more relevant expertise in a committee and better recommendation from committees (Adem et al. 2018).
4.2.3 Control variables
The relationship between skill diversity and firm’s financial performance may be affected by several other variables (see Adem et al. 2018; Custódio et al. 2019). Therefore, it is imminent to take into account those variables to generate unbiased results. The first variable that I will control for is the industry, since performance may vary across industries (e.g. Adem et al. 2018). I will include a dummy variable for each industry. The industry dummies are based on the industry classification benchmark (IBC), which is a framework used by the FTSE for segregating markets into sectors within the macroeconomy.
The second variable that I will control for is the board size which may also have a positive impact on firm’s financial performance (Custódio et al. 2019). Prior research showed that board size is positively correlated with financial performance. The underlying theory is that boards with more directors have access to more information, there is more knowledge on the board, the variety of knowledge on a board is more diverse. This leads to better decision-making and this may result in better firm financial performance (Custódio et al. 2019). However, Labelle (2015) have an opposing conclusion, she stated that larger boards are more prone to agency problems, which leads to negative financial performance. The total number of directors on a board will be used as board size.
The third variable I will control for is the firm size. It is suggested that the size of a firm has a positive correlation on firm financial performance. The reason behind this relation is that bigger firms have a competitive advantage relative to smaller firms, because of
that lager firms are more prone to agency problems due to opposing interests within a firm and information asymmetry. The authors observed a negative relationship in their study. The firm size will be measured by taking the natural logarithm of the total assets.
The last control variable is the debt level of a firm. Prior research showed that debt level in negatively correlated with financial performance (Dezso & Ross, 2012). The underlying theory is that the amount of debt is correlated with the riskiness of a firm. The more debt a firm have, the lower the financial performance will be. Since, a firm is more prone to bankruptcy and a firm will have less opportunities (also known as underinvestment) (Labelle et al., 2015). The debt level is calculated as follows:
𝑑𝑒𝑏𝑡 𝑙𝑒𝑣𝑒𝑙 =short term debt + long term debt
Total assets .∗ 10
4.3 Empirical model
To examine how skill diversity in a board of directors can affect financial performance, a panel data has been used for the period 2009 to 2019. Consistent with the procedures described by Hill, Griffiths, and Lim (2011), I will utilize the ordinary least squares estimation technique. Given the panel data structure, I have conducted a series of different models and statistical tests to determine the most appropriate model. Below I present the equations I used to estimated different model.
4.3.1 OLS regression – Pooled
The first model I will analyze is pooled regression, using ordinary least square (henceforth OLS). This model is commonly used when studying financial performance and board diversity (Post et al. 2015). However, the panel nature of data will be ignored with pooled least square estimates. The error is assumed to have a constant variance and is uncorrelated over time and individuals (Hill et al. 2011). The following model will be tested:
where the subscripts i and t refer to individuals and time period, respectively. The coefficients are the same for all individuals (and over time), and I assume 𝑣𝑎𝑟(𝜀"#) = 𝜎$ and 𝐸(𝜀"#𝜀%&) =
0.
4.3.2 OLS regression - Fixed effects
The pooled OLS regression model does not take into account the characteristics of panel data. However, the fixed-effects model solves this problem. Therefore, I will estimate a fixed-effects model and use the fixed-effect estimator. This model allows for different intercept for each individual. The following model will be tested:
𝐹𝑖𝑛𝑎𝑛𝑐𝑖𝑎𝑙 𝑝𝑒𝑟𝑓𝑜𝑟𝑚𝑎𝑛𝑐𝑒 \ ỹ.. "#^ = 𝛽 . 1𝑠𝑘𝑖𝑙𝑙 𝑑𝑖𝑣𝑒𝑟𝑐𝑖𝑡𝑦 . . "#( . 1x` . . "#) + 𝛽 . 2𝑡𝑜𝑡𝑎𝑙 𝑎𝑠𝑠𝑒𝑡𝑠 . . "#( . 2x` . . "#) + 𝛽 . 3𝑏𝑜𝑎𝑟𝑑 𝑠𝑖𝑧𝑒 . . "#( . 3x` . . "#) + 𝛽4. 𝑑𝑒𝑏𝑡 𝑙𝑒𝑣𝑒𝑙.. "#( . 4x` . . "#) + 𝑑𝑢𝑚𝑚𝑦 𝑖𝑛𝑑𝑢𝑠𝑡𝑟𝑦( . 4x` . . "#) + ẽ . . "# where, ỹ.. "#= 𝑦 . . "#− ȳ . . "# and x̃ . . "# = 𝑥 . . "#− 𝑥̅ . .
"# are observations in terms of deviations from
individual means, ȳ..
"# and 𝑥̅
. .
"#are the individual means (average over time for each individual).
Intercept has been cancelled out (Hill et al. 2011).
4.3.3 OLS regression - Random effects
The third model I will estimate is the random-effects model. This model continues to assume that the individual differences are captured by the differences in the intercept parameter, but this model also recognize that the individuals in the sample were randomly selected, and thus I treat the individual differences as random rather than fixed.
where 𝑣.. "#= 𝑢 . "+ 𝜀 . .
"# is composed of a random individual effect 𝑢
.
" and the usual regression
random error 𝜀..
"#(Hill et al. 2011).
4.3.4 Multi-collinearity
In statistics, colinearity is a phenomenon that arises when explanatory variables in a regression model are perfect linear functions of each other. This is a violation of the OLS assumptions (Brooks, 2014). I used a correlation matrix to check for the collinearity among the independent variables. Since there is no standard criterion for collinearity, I rely on Hill, Griffiths, and Lim (2011) who suggested that when variables portray a correlation coefficient of 0.8 or higher in absolute value, collinearity between the variables will be assumed and one of the two variables will be dropped. The correlation matrix given in Table 2 shows a high correlation between Tobin's Q, ROE and ROA. This actually makes sense because Tobin’s Q, ROE, and ROA are indicators of financial performance. The criterion is below 0,8. Moreover, these variables are not explanatory variables, so he correlation between these variables doesn’t matter. The independent variable unique skill is positively correlated with the board size. This is expected because large boards are more likely to have more skills. Furthermore, unique skills are positively correlated with firm’s financial performance. This is supported by Labelle et al. (2015) findings, which suggests more diversity of knowledge leads to better decision-making and this may result in better firm financial performance.
opportunities (also known as underinvestment) The last control variable firm siz, this control variable doesn’t have a clear direction. It is suggested that the size of a firm has a positive correlation on firm financial performance. The reason behind this relation is that bigger firms have a competitive advantage relative to smaller firms, because of market power and cost advantagesdue toeconomies of scale (Tirole, 2016). Therefore, firm size is also correlated to better financial performance. However, Labelle (2015) concluded that lager firms are more prone to agency problems due to opposing interests within a firm and information asymmetry. The authors observed a negative relationship in their study. The firm size will be measured by taking the natural logarithm of the total assets.
In general, all the variables portray a correlation coefficient which is in line with the economic theory. Moreover, there appears to be no concern regarding the multi-collinearity, as shown by Variance Inflation Factor (VIF) test given in appendix D. Appendix D reports that the mean VIF values is well below the strict critical value of 4 for all different specifications of the models.
Table 2: Correlation matrix Board size Debt level Total Assets Unique skills Skill-match ratio
ROE ROA Tobin’s Q
Board size 1 Debt level 0.18* 1 Total Assets 0.399*** 0.07 1 Unique skills 0.59*** 0.09 0.14 1 Skill-match ratio 0.07 0.08 0.18** 0.07 1 ROE −0.02 −0.18** −0.01 0.01 0.01 1 ROA 0.24* 0.09 0.04 0.22** 0.04 0.39*** 1 Tobin’s Q 0.12* 0.19** −0.13* 0.22** 0.02 0.21** 0.24** 1
Note: Variable firm size (measured in total assets) is a logarithmic transformation due to concerns regarding normality. Asterisk indicate significance at the 1% (***), 5% (**) and 10% (*) level respectively.
4.3.5 Endogeneity
to omitted variable bias and reverse causality. Adem et al. (2018) provided two arguments that predict a negative correlation between performance and skill diversity due to reverse causality. The author suggested that underperforming firms may have more skill diverse boards because they needed to have more skill diversity to get advice from more perspectives. This is consistent with the findings of Tirole (2006) study, which suggested that the reputation of a firm is important for human capital because the reputation of a firm attract different employees to the company. Adem et al. (2018) also suggested that underperforming firms may engage in window-dressing, by making their directors appear more talented than they really are. On the other hand, Adem et al. (2018) also argued that the positive relationship between performance and skills diversity is due to reverse causality. The author suggested that it is possible that poorly-performing firms have other concerns and pay less attention to skill diversity as a result. Without a better understanding of how directors are matched to firms, it is difficult to sign the bias in the ordinary least squares (OLS) results. The test results suggest a positive correlation between skill diversity and financial performance. Adem et al. (2018) concluded that it is not possible to give this relationship a causal interpretation because of the potential endogeneity problems due to the reverse causality. However, Bhagat & Black (2001) concluded in their research that endogeneity is not likely, because the composition of boards changes quite slowly over time.
Endogeneity concern in the panel data models is partly taken into account, since fixed-effects model control for all the omitted variable biases. In order to test for endogeneity, the Hausman Test has been used extensively in the past for panel data models (Hill et al. 2011). The Hausman test is used to compare fixed-effects and random-effects models, but it also examines the endogeneity among the regressors and the error term. For pooled model, a Reset test can be used to examine the model specification, which check for the omitted variable bias (Hill et al. 2011). Based on these test statistics, there was no omitted variable bias found in the models.
Some studies dealt with endogeneity by using one-year lagged values for the endogenous variable to reduce reverse causality biases (Carter et al., 2010; Labelle et al., 2015). In this research, I will use one-year lagged values of the diversity variable as a check for endogeneity. First, all regressions are performed with no one-year lagged values of the diversity variable. Thereafter, I will conduct the same regression, but with one-year lagged for the diversity variable. The test statistics showed an insignificant effect, which indicates that, the likelihood of reverse causality between skill diversity and financial performance is very small.
In general, this study deals with endogeneity by including one-year lagged values, but it does not eliminate endogeneity completely. Therefore, the possibility that some of the effects could be driven by reversed causality cannot be fully rejected.
5. Results & Analysis
This section covers the empirical results of the hypotheses proposed in this study. The first hypothesis examines the impact of skill diversity on firm’s financial performance. In addition, the second hypothesis examines the impact of skill-match committee match ratio on firm’s financial performance. For each hypothesis, I examined the impact of the independent variables on Tobin’s Q and ROA. Furthermore, I analyze multiple statistical models, including pooled, fixed-effects, and random-effects model. After discussing the descriptive statistics, I present the estimated of pooled, fixed-effects, and random-effects estimates with industry dummy variables.
5.1 Descriptive statistics
In Table 3, the descriptive statistics of the variables used in this study are presented. The descriptive statistics include the mean, median, standard deviation, minimum and the maximum values of discussed variables. Both accounting-based and market-based indicators for financial performance, I observe positive mean values, which implies that the companies in my dataset generate profits from their assets/investments. However, I observe large variation in the accounting data. For example, the return on investments (ROI) ranges from -1,914 to 1,077. Thus, there is a mixture of companies that operate efficiently and those that operate poorly. Similar patterns have been for market data. Tobin’s Q has a positive value of 1,64 on average, which implies that the market value of these firms surpasses its replacement costs of assets. However, the difference in Tobin’s Q values among companies is relatively large, ranging from 0,033 to 13,003. Such huge differences in minimum and maximum values may be an indicator for outliers in my dataset.
are still financial or marketing- or sales-orientated. Other boards have fewer skills, but the skill distribution is more balanced. See appendix B and C for more detailed descriptive statistics.
Table 3: Descriptive statistics
Variables Mean Median Std. Dev. Min Max
Firm characteristics
Total Assets 10687,854 3605,3 21550,486 28,817 156985
Earnings per share 0,662 0,354 1,07 -4,363 7,932
Tobin’s Q 1,640 1,400 1,010 0,033 13,003 ROI 0,101 0,093 0,149 -1,914 1,077 ROA 0,055 0,050 0,069 -0,606 1,049 ROE 0,105 0,064 0,433 -3,818 4,100 Firm Age 69,32 22 34,56 4 504 Board characteristics Director tenure 8,463 5 8,295 0 57 Board size 9,738 9 2,203 5 15 Female directors 0,31 0,418 0,34 0 0.632 Director age 59,2 56 11,39 32 89 Number of unique skills on board 10,299 10,000 1,958 5,000 15,000 Number of skills board member 2,42 2 1,29 1 7 Board committees 2,57 3 0,492 1 5
5.2 Financial performance and skill diversity
Table 4: Regression results of unique skills on tobin’s q
Dependent Variable: Tobin’s Q
Pooled Fixed-Effects Random-Effects OLS
Independent Variables 1 2 3 4 Constant 1.009*** (0.178) 0.501 (0.375) 0.831*** (0.274) Unique skills 0.069** 0.031) 0.234*** (0.054) 0.139*** (0.049) 0.071*** (0.015) log (total assets) (0.054) -0.046 (0.060) 0.054 (0.048) 0.017 -0.067** (0.028)
debt 0.865** (0.395) -0.862* (0.484) (0.393) -0.376 0.936*** (0.201) board size (0.030) 0.002 (0.030) 0.052 -0.048 0.035) (0.017) 0.013 Consumer goods 0.402*** (0.126) Consumeer services (0.141) 0.157 Healthcare 0.274* 0.166) Industrial 0.155 (0.130) Real Estate (0.141) 0.092 Telecommunication (0.176) 0.225 Technology 0.104 (0.238) Observations 624 624 624 624 Adjusted R2 0.110 0.108 0.086 0.180
Note: Variable firm size (measured in total assets) is a logarithmic transformation due to concerns regarding normality. The standard errors are between brackets. Standard errors are adjusted for heteroskedasticity. Asterisk indicate significance at the 1% (***), 5% (**) and 10% (*) level respectively.
underinvestment) (Labelle et al., 2015).Board size and firm size are statistically insignificant, and the coefficients do not have a clear direction.
When doing the OLS regression with the panel data, the residuals may be correlated across observations due to the changes that happened within a firm or within a year. This makes the standard errors biased and result in either the over or underestimation of the true variability of the coefficient estimates. To account for this problem, firm- and year-fixed effects are needed to be included in the model – fixed-effects model (see column 2). This model does not have an intercept, since this model have different intercepts for each individual. Industry dummies cannot be included in the fixed- effects model because these dummies are invariant. However, fixed-effects model controls for all those time-invariant characteristics. The independent variable unique skill has a statistically significant positive coefficient of 0.234 at the 5% significant level. This means that skill diversity leads to a higher firm’s financial performance. For the control variables, debt level has a negative coefficient and is significant at a 10% level. The coefficient for board size and total assets is positive, but insignificant. The sign of these coefficients are in line with economic theory.
Column 3 presents the estimates for the random-effects model, which is a special case of the fixed-effects model. This model assumes that there are no fixed effects. Looking at the test statistics, only the unique skills variables is significant at a 1% level. All the remaining variables are insignificant. Despite the insignificant coefficients, the sign of each coefficient is in line with what economic theory predicts, except for board size. Board size should be positive, since it is assumed that boards with more directors have access to more information, there is more knowledge on the board, the variety of knowledge on a board is more diverse. This leads to better decision-making and this may result in better firm financial performance (Custódio et al. 2019). However, Labelle (2015) have an opposing conclusion, she stated that larger boards are more prone to agency problems, which leads to negative financial performance. The total number of directors on a board will be used as board size
significant at a 1% level. The control variables firm size and debt level are also significant whereas the debt level has an insignificant positive coefficient. This means that the debt level of a company contributes to better financial performance, which is not in line with economic theory.
In addition, the same regression analyses have been conducted, but the dependent variable Tobin’s Q (market-based performance) has been replaced with ROA (accounting-based performance).
Table 5: Regression results of unique skills on roa
Dependent Variable: ROA
OLS Fixed-Effects Random-Effects OLS
Independent Variables 1 2 3 4
Constant -1.802*** (0.287) -2.004*** (0.612) -1.031** (0.426)
Unique skills 0.110*** (0.023) (0.049) 0.021 0.090** (0.042) 0.085*** (0.023)
log (total assets) 0.215*** (0.036) 0.369** (0.151) 0.263*** (0.082) 0.233*** (0.043)
debt (0.030) -0.047 -0.213*** (0.040) -1.945*** (0.684) -1.054*** (0.314) board size (0.026) -0.019 (0.030)0.049 (0.054) -0.009 (0.026) -0.021 Consumer goods -0.954*** (0.196) Consumeer services -0.624*** (0.219) Healthcare (0.259) 0.027 Industrial -0.896*** (0.202) Real Estate -0.465** (0.219) Telecommunication -1.655*** (0.274) Technology -0.765** (0.371) Observations 624 624 624 624 Adjusted R2 0.142 0.125 0.106 0.226
variable has a positive coefficient (0.110) that is significant at a 1% level. The control variables, debt level and board size in particular, are insignificant. The second model is a fixed-effects model. The unique skills variable is now insignificant. However, the debt level and firm size are significant at a 1% and 5% level, respectively. The third model is a random-effects model, where the unique skills variable has a positive coefficient and is significant at a 5% level. The last model is an OLS model that controls for industry effects. This model has a statistically significant coefficient for unique skills at 1% level. Most control variables are significant, but the board size is not significant. This was also the case with other models. The industry dummies are in general significant, but health care industry is not significant.
5.2.1 Model selection
After examining the different models, it can be concluded that there is a positive link between skill diversity and financial performance in most of the models. However, some models include coefficients which are against the economic theory. To determine the most appropriate model, the Hausman test and the redundant fixed-effects test have been conducted. The null hypothesis of Hausman test states that there is no correlation between the explanatory variables. Based on this test it can be concluded that fixed-effects most is most appropriate. Thereafter, the redundant fixed-effects test has been conducted. This test examines if the pooled model is preferred over the fixed-effects model. The null hypothesis states that there is no fixed effect, just one intercept. Based on the F-test, it can be concluded that the fixed-effects model is the most appropriate model (Hill et al. 2011). To sum up, random-effects model is not required whereas fixed-effects model is preferred over a simple pooled model.
5.3 Financial performance and committee skill-match ratio
Table 6: Regression results of committee skill-match ratio on tobin’s q
Dependent Variable: Tobin’s Q
OLS Fixed-Effects Random-Effects OLS
Independent Variables 1 2 3 4 Constant 1.367*** (0.026) 0.851** (0.361) 0.863*** (0.281) Committee skill-match ratio (0.028) 0.010 (0.903) 1.523 (0.709) 0.929 (0.170) -0.084 log (total assets) -0.052*** (0.004) (0.057) 0.059 (0.049) 0.012 -0.067** (0.028)
debt 0.847*** (0.030) (0.532) -0.809 (0.420) -0.366 0.933*** (0.202) board size 0.043*** (0.003) (0.029) 0.046 (0.032) 0.030 (0.017) 0.012 Consumer goods 0.403*** (0.126) Consumeer services (0.141) 0.153 Healthcare (0.167) 0.270 Industrial 0.157 (0.130) Real Estate (0.239) 0.094 Telecommunication (0.141) 0.087 Technology 0.244 (0.181) Observations 624 624 624 624 Adjusted R2 0.142 0.087 0.036 0.209
Note: Variable firm size (measured in total assets) is a logarithmic transformation due to concerns regarding normality. The standard errors are between brackets. Standard errors are adjusted for heteroskedasticity. Asterisk indicate significance at the 1% (***), 5% (**) and 10% (*) level respectively.
financial performance. Moreover, all the control variables are insignificant. However, the signs of the coefficients are in line with economic theory. Another model that accounts for the characteristics of the panel data is the random-effects model. Besides the intercept term, no variable in this model is significant. The last model is an OLS model that also controls for industry effects. The coefficients for each industry are shown in table 6. Even after controlling for the industry effects, the variable skill-match ratio is insignificant. The control variables, firm size and debt level, are significant at 5% and 1% level, respectively. However, debt level has a positive coefficient.
Table 7: Regression results of committe skill-match ratio on roa
Dependent Variable: ROA
OLS Fixed-Effects Random-Effects OLS
Independent Variables 1 2 3 4 Constant -1.085*** (0.026) -1.463** (0.613) -1.038** (0.438) Committee skill-match ratio (0.328) -0.468 (1.025) -0.256 (0.528) -0.303 (0.265) 0.018 log (total assets) 0.216*** (0.004) 0.369** (0.152) 0.265*** (0.086) 0.233*** (0.043)
debt -1.534*** (0.030) -2.821*** (1.030) -1.971*** (0.698) -1.054*** (0.314) board size 0.045*** (0.003) (0.035) 0.045 (0.044) 0.045 (0.026) -0.021 Consumer goods -0.954*** (0.196) Consumeer services -0.624*** (0.219) Healthcare (0.259) 0.028 Industrial -0.896*** (0.203) Real Estate -0.763** (0.373) Telecommunication -0.464** (0.220) Technology -1.659*** (0.281) Observations 624 624 624 624 Adjusted R2 0.165 0.093 0.090 0.245
Note: Variable firm size (measured in total assets) is a logarithmic transformation due to concerns regarding normality. The standard errors are between brackets. Standard errors are adjusted for heteroskedasticity. Asterisk indicate significance at the 1% (***), 5% (**) and 10% (*) level respectively.
level are highly significant, likewise the fixed-effects model. On top of that, the sign of each coefficient in both models are in line with the literature. The last model is an OLS model that is controlled for industry effect. This statistic of this model differs from the first 3 models. Interestingly, the independent variable committee skill-match ratio is now positive (0.018), but the skill-match ratio is still insignificant. The control variables, firm size and debt level, are significant at 1% level. The industry dummies are also significant, except the dummy for the health care industry.
5.3.1 Model selection
After examining the different models, it can be concluded that committee skill-match ratio has a positive but insignificant relationship with Tobin’s Q. For the ROA, the opposite effects are found where the skill-match ratio has an insignificant and negative relationship with ROA. Similar to the unique skills variable, some coefficients are against the economic theory. To determine the most appropriate model, the same diagnostic tests have been conducted as for the first hypothesis in the previous section. Based on the diagnostics, the fixed-effects model appears to be the most appropriate model (Hill et al. 2011). Therefore, the random-effects model is not required whereas fixed-effects model is preferred over a simple pooled model.
5.4 Robustness check
As mentioned earlier, I re-examined the previous models after replacing the independent variables (i.e. unique skills and committee skill-match ratio) with their lagged values, in order to control for endogeneity. Replacing the endogenous explanatory variables with their lagged values is a very common approach to address endogeneity in applied economics (Hill et al., 2011). The underlying logic is that the lagged values of the endogenous variables are typically not related to the error term. Therefore, I also utilize this technique as a robustness check to examine the impact of skills diversity on firm’s financial performance. The regression results of the models are reported in appendix E.
all the models (i.e. 0.057, 0.140, 0.086, and 0.057), which suggests a positive impact of lagged unique skills on firm’s financial performance (i.e. tobin’s q). Similar results are found in Table 12 where I replace the dependent variable with ROA (accounting-based measure). The lagged unique skills variable has a statistically significant positive effect on firm’s ROA. Regarding the control variables, I do not observe any major changes with respect the sign and magnitude and signs of the coefficients. However, most of the significant industry effects disappear, when I include the lagged unique skill variables. Therefore, it can be concluded that more unique skills indeed lead to higher firm’s performance.
Table 13 and 14 report the regression results of lagged skill-match ratio variable on Tobin’s q and ROA. Regarding the impact on Tobin’s Q, the coefficient of the lagged skill-match ratio is statistically significant and positive for the fixed- and random-effects model (see column 2 and 3 in Table 13). This implies a significant positive impact of lagged skill-match ratio on firm’s financial performance (market-based measure). This finding is in contrast to what we saw when we use the current values of skill-match ratio as an independent variable. Based on this finding, it seems reasonable to argue that the skill-match ratio indeed positively influences firm’s financial performance. With respect to the impact on ROA, I also found a significant impact of the lagged skill-match ratio variable on firm’s ROA in the random-effects model and regression with industry dummies (see column 3 and 4 in Table 14), but the effect is insignificant in the fixed-effects model. The Hausman test reveals that the fixed-effects model is the most the appropriate model. Therefore, I will use the estimates from the fixed-effects model for interpretation.
6. Conclusion
This research examines the effects of skill diversity on firm’s financial performance. This will be measured with market-based (Tobin’s Q) and accounting-based measures (ROA). It is assumed that more diversity in the board is better for financial performance. This positive relationship is supported by different theoretical concepts. The most important theoretical frameworks include resource-based, human capital, agency theory. Based on these theoretical frameworks, scholars found that diversity leads to an increase in creativity (supported by the resource-based and human capital theory), more board independency (agency theory), innovation (human capital theory), better decision making (resource-based, human capital, and agency theory), and better monitoring (agency theory), which subsequently leads to better financial performance (Ferreira, 2009; Gul et al., 2011). Despite all this, some studies concluded that diversity will more time-consuming, and it may generate more conflicts due to more/opposing opinions (social identity theory; see Terjesen et al., 2009).
The analysis conducted in this study examines the impact of skill diversity in boards on firm’s financial performance. Furthermore, this research also addresses the endogeneity concerns, which are common in the past literature, by applying the fixed-effects model. One-year lagged values for the endogenous variable has also been used to reduce the potential reverse causality biases (Carter et al., 2010; Labelle et al., 2015). This thesis uses a panel dataset of the largest UK firms listed on the FTSE 250 for the period 2009 to 2019. A number of regression analyses were used to empirically test the effects of skill diversity on financial performance measured by Tobin’s Q and ROA. After the careful examination of the different models, it can be concluded that most models suggest a positive link between skill diversity and financial performance. However, some models have variable coefficients that are against the economic theory. After performing several the test diagnostics, the fixed-effects model is found to be the most appropriate model.
research regarding the relationship between board diversity and the financial performance also used different proxies and provided inconclusive results. Some researchers found positive results, while others found negative and or no results at all. The discrepancy between the test results may be partly explained due to the use of different performance measures. Therefore, an important take away is that the effect of skill diversity differs between the two measurements for financial performance, which indicates that the choice of the performance measure is important when conducting research with the purpose to compare the empirical results.
The second hypothesis states that boards with a higher skill committee match ratio positively influences the financial performance of firms. The intuition behind this is when board members are assigned to a committee with a matching skill this may lead to more relevant expertise in the committee and better recommendation and advice from committees (Adem et al., 2018). After examining the different models, it can be concluded that most models have a positive coefficient for the skill-match ratio, but the effect is insignificant if measured with the market-based measure Tobin’s Q. similarly, test results based on the accounting-based measure (i.e. ROA) also portrays a negative and insignificant effect of skill-match ratio. Similar to hypothesis 1, the fixed-effects model is found to be the most appropriate model for the second hypothesis based on the test diagnostics. The regression results of the fixed-effects model show all control variables are in line with economic theory. Since the variable of interest i.e. skill-match ratio is insignificant, the second hypothesis is not supported by the analysis. This implies that boards with a higher skill committee match ratio will not influence firm’s financial performance. An important caveat is that there is a huge variation in the disclosure between firms. Some firms disclose exactly why someone is assigned to a committee whereas other firms disclose nothing.
7. References
Adams, R, B., Akyol, C, A, Verwijmeren, P. (2018). Director skill sets. Journal of Financial Economics Volume. 130, Issue 3, 641-662.
Adams, R., Ragunathan, V., Tumarkin, R., 2015. The changing nature of corporate board activity. Unpublished working paper. University of New South Wales.
Adams, R.B., Ferreira, D., 2009. Women in the boardroom and their impact on governance and performance. Journal of Financial Economics. 94 (2), 291–309.
Ahmed, A, S., Duellman, S., 2013. Managerial overconfidence and accounting conservatism. Journal of Accounting research 51, 1-30.
Banerjee, S., Humphery-Jenner, M., Nanda, K, V. 2015. Restraining Overconfident CEOs through Improved Governance: Evidence from the Sarbanes-Oxley Act. Review of Financial Studies 28, 2812-2858.
Barber, B. M., & Odean, T. (2001). Boys will be boys: Gender, overconfidence, and common stock investment. The Quarterly Journal of Economics, 116(1), 261-292.
Barney, J. 1991. Firm Resources and Sustained Competitive Advantage. Journal of Management 17, 99–120.
Bhagat, S., & Black, B. (2001). The Non-Correlation between Board Independence and Long-Term Firm Performance. Journal of Corporation Law, 27, 231–274.
Baixauli-Soler, J. S., Belda-Ruiz, M., & Sanchez-Marin, G. (2015). Executive stock options, gender diversity in the top management team, and firm risk taking. Journal of Business Research, 68(2), 451-463.
Becker, G., 1962. Investment in human capital: A theoretical analysis. J. Political Econom 70, 9–49.
Bernile, G., Bhagwat, V.,Yonker., S.,2018. Board Diversity, Firm Risk, and Corporate Policies. Unpublished working paper. University of Miami.Brooks, C., 2015.Introductory Econometrics for Finance, Cambridge University Press, Cambridge, 4th edition.
Benadou, R.,Tirole, J., 2003. Intrinsic and extrinsic motivation. Review of economic studies: 489 – 520.
Carter, D. A., Simkins, B. J., & Simpson, W. G. 2003. Corporate governance, board diversity, and firm value. Financial Review 38, 33-53.
Carter, D. A., D’Souza, F., Simkins, B. J., & Simpson, W. G. 2010. The Gender and Ethnic Diversity of US Boards and Board Committees and Firm Financial Performance. Corporate Governance: An International Review, 18(5), 396–414.
Carrillo, j.,Gromb, D., 1999. On the strength of corporate culture. European economic review. Coombes, p and S. Wong. 2004 Why coeds of governance work. In a new era in governance,
pp. 48-53 McKinsey.
Cox, Taylor H., & Blake, S. (1991). Managing Cultural Diversity: Implications for Organizational Competitiveness. Academy of Management Executive 5, 45–56.
Custódio, Ferreira, and Matos., 2019. Do General Managerial Skills Spur Innovation? 476 Management Science65, 459–476.
Dah, M. and Jizi, M. 2017. Board independence and the efficacy of social reporting, Journal of International Accounting Research 17. 25-45.
Dass. N, O. Kini, V. Nanda, B. Onal, Wang J. 2013 Board expertise: do directors from related industries help bridge the information gap?Financial Review, 27.
Dezso, C. L., & Ross, D. G. 2012. Does Female Representation in Top Management Improve Firm Performance? A Panel Data Investigation. Strategic Management Journal 33, 1072–1089.
Doupnik, T., H. Perera. International Accounting, 2019, 5th edition.
Ellis, A., Edward, C., Thomas, S., 2018. Playing Favorites? Industry Expert Directors in Diversified Firms Jesse. Journal of financial and quantitative analysis 53, 1679–1714. Erhardt, N. L., Werbel, J. D., & Shrader, C. B. (2003). Board of director diversity and firm
financial performance. Corporate governance: An international review, 11(2), 102-111. Fama, E. F. (1980). Agency problems and the theory of the firm. Journal of Political Economy,
88(2), 288-307.
Fama, E.F. & M.C. Jensen (1983), “Separation of Ownership and Control”, Journal of Law & Economics 26, pp. 301-325.
Faleye, O., Hoitash, R., Hoitash, U., 2018. Industry expertise on corporate boards. review of quantitative finance and accounting 50, 441-479.
Fich, M.E., 2005. Are some outside directors better than others? evidence from director appointments by fortune 1000 Firms. The Journal of Business 78, 1943-1972.
Garlappi, L., Giammarino, R.,Lazrak, A., 2017. Ambiguity and the corporation: Group disagreement and underinvestment. Journal of Financial Economics125. 417-43. Gul, F.A., Srinidhi, B., Ng, A.C., 2011. Does board gender diversity improve the informativeness
of stock prices? J. Account. Econ. 51, 314–338.
Hambrick, D., Cho, T., & Chen, M. (1996). The Influence of Top Management Team Heterogeneity on Firms’ Competitive Moves. Administrative Science Quarterly 41, 659-685.
Han, J., Han, J., Brass, J, D., 2014. Human capital diversity in the creation of social capital for team creativity. Journal of Organizational Behavior 35, 54–71.
Haynes, K. T., & Hillman, A. (2010). The effect of board capital and CEO power on strategic change. Strategic Management Journal (John Wiley & Sons, Inc. ), 31(11), 1145–1163. Hill, R., Griffiths, W. and Lim, G. (2011). Principles of econometrics. Principles of econometrics.
Hoboken, NJ: Wiley.
Hillman A. J., Cannella A. A., Paetzold R. L. 2000. The resource dependence role of corporate directors: Strategic adaptation of board composition in response to environmental change. Journal of Management studies 37, 235-256.
Hillman, A. J., Dalziel, T. 2003. Boards of Directors and Firm Performance: Integrating Agency and Resource Dependence Perspectives. Academy of Management Review 28, 383– 396.
Hull, J.C., 2015, Options, Futures, and Other Derivatives, 9th edition, Pearson.
Jungmann, C. (2006). The effectiveness of corporate governance in one-tier and two-tier board Systems–Evidence from the UK and Germany–. European Company and Financial Law Review3, 426-474.
Kanter, R. M. 1977b, 'Some Effects of Proportions on Group Life', American Journal of Sociology 82, 965- 990.
Masulis, R.,Ruzzier, C., Xiao, S., Zhao, S., 2012. Do Independent Expert Directors Matter? Unpublished working paper. University library of Munich,Germany.
Labelle, R., Francoeur, C.,Lakhal, F., 2015. To Regulate Or Not To Regulate? Early Evidence on the Means Used Around the World to Promote Gender Diversity in the Boardroom. Gender, Work & Organization 22, 339–363.
Oliveiraa, S. de and Nisbetta R.E. (2018). Demographically diverse crowds are typically not much wiser than homogeneous crowds. Department of Psychology, University of Michigan.
Perryman, A, A., Fernando, G, D., Tripathy, A., 2016. Do gender differences persist? An examination of gender diversity on firm performance, risk, and executive. Journal of Business Research 69 579-586.
Pfeffer, J. and G. R. Salancik (1978). The External Control of Organizations: A Resource Dependence Perspective. New York, NY, Harper and Row.
Post, C & Byron, K. (2015). Woman on boards and firm financial performance: A meta-analysis, Academy of Management Journal58, 1546–1571.
Rancis, J.C., D. Kim. (2013) Modern Portfolio Theory: Foundation, Analysis, and New Developments*, Wiley Edition, John Wiley & Sons.
Rajan, R., Servaes, H., Zingales, L., 2000 The cost of diversity: the diversification discount and inefficient investment. Journal of finance 55, 35 – 80.
Rivas, J. L. 2012. Diversity & Internationalization: The Case of Boards and TMT’s. International Business Review 21, 1–12.
Rose, C., 2007. Does female board representation influence firm performance? The Danish evidence. Corp. Gov. 15 (2), 404–413.
Sila, V., Gonzalez, A., Hagendorff, J., 2016. Women on board: Does boardroom gender diversity affect firm risk. Journal of corporate finance 36, 26-53.
Simons, T., &Pelled, L. 1999. Understanding Executive Diversity: More than Meets the Eye. Human Resource Planning 22, 49-51.
Tirole, Jean. (2006). The theory of corporate finance. Princeton, N.J.: Princeton University Press. Terjesen, S., Sealy, R., & Singh, V. 2009. Women directors on corporate boards: A review and
research agenda. Corporate Governance: An International Review 17, 320-337.
Wang, C., Xie, F., Zhu, M., 2015 Industry Expertise of Independent Directors and Board Monitoring. Journal of Financial and Quantitative Analysis 50, 929-962.
Williams, K., & O’Reilly, C. A. 1998. Demography and Diversity in Organizations: A Review of 40 Years of Research. Research in Organizational Behavior, 20, 77–140.
Yusoff, W, F, W., Alhaji, I, A., 2012. Corporate governance and firm performance of listed companies in Malaysia. Trends and Development in Management Studies 1, 43-65.
8. Appendix
8.1 Appendix A
Table 8: Definition of each skill, based on skill diversity research of Admans 2018
Variables Description
Academic Director with academia or has a higher degree (such as a Ph.D.). corporate governance experience in governmental, policy, or regulatory.
Company Experience within the company.
Financial Experience in banking or financial related jobs.
HRM Experience in human resource management.
International International experience.
Industry Industry experience.
Leadership leadership skills/experience.
Legal Experience in law.
Marketing Marketing and sales experience.
Multiple industries Experience more industries.
Operational Experience in daily operations of a business.
Risk management Risk management experience.
strategy Strategy skills or strategy planning experience.
Sustainability Experience in environmental and sustainability related issues.
8.2 Appendix B
Figure 1: Director skills and frequency. This figure shows how many skills each board possesses
Figure 2: Director and board skills. This figure shows the percentage of directors with specific skills on a board. For instance, all boards have at least one board member with the skill financial.
0 50 100 150 200 250 300 5 6 7 8 9 10 11 12 13 14 15 Fre qu en ty Number of skills
Director skills
Totaal 0 20 40 60 80 100 120 Acad emic corpo rate g overn ance Comp any Finan cial HRM Intern ation al Indus try Leade rship Lega l Marke ting Multip le ind ustri es Opera tiona l Risk m anag emen t strate gy Susta inabil ity Tech nolog y pe rce nt ag e skills8.3 Appendix C
Table 9:
Descriptive statistics: skill distribution board of directors
Skill Mean Median Std. dev. Min Max
8.4 Appendix D
Table 10a:
Variable VIF 1/VIF
Total assets 1.54 0.648654 Unique skills 1.53 0.653452 Board size 1.01 0.986158 Debt 1.00 0.998911 Mean VIF 1.27 Table 10b:
Variable VIF 1/VIF
Total assets 1.01 0.986385
Skillmatch ratio 1.01 0.987514
Board size 1.01 0.987594
Debt 1.00 0,998649
8.4 Appendix E
Table 13: Regression results of lagged committee skill-match ratio on tobin’s q
Dependent Variable: Tobin’s Q
OLS Fixed-Effects Random-Effects OLS
Independent Variables 1 2 3 4
Constant 1.495*** 1.130*** 1.339***
(0.354) (0.356) (0.301)
Lag committee skill-match ratio 0.022 1.271** 0.775*** -0.022
(0.342) (0.548) (0.294) (0.183)
log (total assets) -0.06 0.031 -0.017 -0.080**
(0.059) (0.055) (0.048) (0.033) debt 0.895** -0.541 -0.071 1.066*** (0.418) (0.610) (0.456) (0.234) board size 0.039 0.028 0.053*** (0.032) (0.032) (0.015) Consumer goods 0.365*** (0.131) Consumeer services 0.088 (0.154) Healthcare 0.209 (0.133) Industrial 0.096 (0.130) Real Estate 0.073 (0.167) Telecommunication 0.101 (0.146) Technology 0.06 (0.153) Observations 546 546 546 546 Adjusted R2 0.046 -0.131 0.031 0.072
