Board composition and gender diversity on dividend
payout in top listed firms:
evidence from civil and common law countries
Abstract
This study examines the effects of board composition and board gender diversity on dividend payout and possible differences between countries based on legal origin. Using an international sample covering data from 2007-2017, the results suggest that board composition (except CEO duality) and board gender diversity positively influence dividend payout in countries with a common law origin. This effect is not observed in countries with a civil law origin even though these firms pay higher dividends compared to firms in common law economies. Dividend payout substitutes board composition as a monitoring mechanism in civil law economies providing evidence for La Porta, Lopez-de-Silanes, Shleifer and Vishny (2000) “substitute model”.
Field keywords: dividend payout, board composition, board gender diversity, legal origin, civil law
and common law
JEL classifications: G18, G30, G35, K40, M14
R.J. Hoekstra S2381397
MSc International Financial Management Faculty of Business and Economics
University of Groningen Supervisor: dr. N. Selmane Co-assessor: dr. W. Westerman
2
Table of Content
1. Introduction ... 3
2. Literature review and hypotheses development... 5
2.1.Agency and other theories ... 5
2.2.Board composition and board gender diversity ... 7
2.3.Legal origin and corporate governance ... 11
3. Data sources, sample selection and methodology ... 12
3.1.Data sources and sample selection ... 12
3.2.Empirical specification ... 13
3.3.Descriptive analysis ... 15
4. Results ... 19
4.1.Regression results ... 19
4.2.Robustness tests ... 27
4.2.1. Industry-adjusted dividend payout ... 27
4.2.2. Firms with negative net income ... 27
4.2.3. Effects financial crisis ... 27
5. Conclusion ... 28
6. Limitations and future directions ... 30
7. References ... 31
8. Appendixes ... 36
8.1.Appendix A. Variable description ... 36
8.2.Appendix B. Correlation matrix ... 37
3
1. Introduction
Corporate boards have been investigated over multiple disciplines, there is however, little consensus among researchers regarding board composition and board gender diversity. Mixed results fuel the discussion, as Adams and Ferreira (2009) show that diversity can enhance board independence which in turn enhances board monitoring performance. Adams, de Haan, Terjesen and van Ees (2015) mention diversity could increase the likelihood of conflicts and cause higher decision-making costs within boards. Moreover, corporate scandals in recent history, like Parmalat in 2003 and WorldCom in 2002, resulted in governmental and non-governmental intervention in the practice of corporate governance. By imposing rules, regulations and certain governance codes, firms (particularly listed ones) are affected in the practice of directing.1 Whilst several guidelines exist, differences are observed amongst countries, its listed corporations and their board of directors. The rules, regulations and codes regarding corporate governance provide a framework for firms to set up their board. As firms continue to internationalize this framework is influenced by the legal origin a firm is operating in as explained in La Porta, Lopez-de-Silanes and Shleifer (2008).
Differences in board composition and board gender diversity affect firms performance and firm value (Adams and Ferreira, 2009), financial policies and risk-taking (Bernile, Bhagwat and Yonker, 2018). The policies involving dividend are an example of such financial policies. The ‘dividend puzzle’ described by Black (1976) describes the ongoing difficulty to decide why and when companies pay dividends. The work is partly based on the earlier work of Modigliani and Miller (1959, 1961) which argues that ‘the amount of dividends paid by a corporation does not affect the share value or the investor returns, because, with higher dividend payouts, the investor receives less capital appreciation, indifferent from the outcomes of business decisions. Assuming that dividends paid to investors do not influence business decisions’. As bold and essential as this theorem by Modigliani and Miller is, the dividend puzzle remains unsolved. The purpose of this research is to examine the effects of board composition and board gender diversity on the dividend payout of firms in countries with different legal origins.
A company's board of directors is crucial within a firm as the primary function is to oversee and evaluate the executive performance of management. If this performance is not satisfactory,
1 Enriques and Volpin (2007) discuss the corporate governance reforms in Continental Europe and address the
4 governing in the form of intervention is necessary (Schellenger, Wood and Tashakori, 1989). The decisions involving boards result in a wide array of possible corporate outcomes. Some more heavily investigated compared to others. Dividend payout, differing from company to company, is such a corporate outcome. In this research, the effects of board composition and board gender diversity on dividend payout are investigated using multiple dimensions of board composition. Gender diversity within responsible positions in corporations is a subject on multiple political agendas. The underlying need for equality amongst men and woman creates a trend to improve the gender-balance amongst non-executive directors of listed companies (Szydlo, 2015). Consequently, companies have responded by increasing women representation on boards. Board gender diversity at UK’ FTSE 350 and US’ S&P 500 firms grew by almost 8 and 2.5% respectively. In the EU, it increased to 23.3% in 2016 from 11.9% in 2010.2
Several articles discuss the effect of composition and gender diversity on corporate boards.3 In line with the literature regarding this subject, I am interested in the effects of board gender diversity and board composition on dividend payout. The central research question is whether board composition and board gender diversity affect dividend payout and if there are observable differences between civil and common law countries.
This research aims to contribute to the current literature by systematically analyzing the effects of firm-level board composition and board gender diversity on dividend payout in an international setting. As most studies encompass single-country studies, a multi-country study broadens the scope and could provide new insights. The developed markets’ context across the comprehensive cross-country dataset allows observing for differences between common and civil law countries while controlling for other important firm and country-specific variables. The Worldwide Governance Indicators (WGIs), reflect the perceived governance/political institutions in a country and allow for more grounded conclusions regarding the observed outcomes. The results indicate that board size, board independence, and board gender diversity indeed affect dividend policy in common law economies but fail to do so in civil law economies. This while civil law economies on average show higher dividend payout. These findings provide insights for managers and investors on dividends, board composition, and diversity within listed firms across countries with different legal origins.
2 https://www.bloomberg.com/quicktake/women-boards
3 Both Chen, Leung and Goergen (2017) and Saeed and Sameer(2017) discuss the impact of board gender diversity
5
2. Literature review and hypotheses development
This section describes several theories important in research regarding dividend payout, board composition, gender diversity and legal origin. At the firm level, common perspectives include agency, resource dependency and signaling theory. Hypotheses are formulated based on the theoretical and empirical evidence given by these studies.
2.1. Agency and other theories
Berle and Means (1932) first described the separation of ownership and control and the conflicts that arise because of this separation. Central to the analysis of modern corporations lay the conflict of interests between agents, such as controlling shareholders and managers, and principals, outside investors like minority shareholders (Jensen and Meckling, 1976). These conflicts arise when managers are empowered by cash flows that allow them to act on their own behalf and personal interest. Reducing the amount of cash available reduces the discretionary benefits available to managers. Asymmetric information between owners and managers also creates agency problems due to shareholder pessimism and resolving these asymmetries brings forth agency costs. Various mechanisms mentioned to mitigate such problems include dividend payments, debt inclusion, managerial equity compensation and board composition (Jensen, 1986; Rozeff, 1982). These mechanisms try to resolve principal-agent conflicts and align interests across groups. One common assumption within the agency theory assumes that outside directors act independently from their inside director counterparts and acts as good monitors for the interests of shareholders (Terjesen, Sealy and Singh 2009). The critical point at the center of the theory is that a failure to reduce free cash flows leads to diversion or waste which is detrimental to outsider shareholder interests. An agency approach recognizes two points and therefore move away from the Modigliani and Miller theorem assumptions. First, investment policies of a firm cannot be taken separately from its dividend policies, and dividend payout may reduce the inefficiencies of investments. Secondly, the allocation of profits to shareholders based on participation cannot be taken for granted as insiders may get preferential treatment. Dividends paid on a pro rata basis benefit outside shareholders relative to the alternative of expropriation of retained earnings (La Porta, Lopez-de-Silanes, Shleifer and Vishny, 2000). So the dividend policy is a function of many factors, and the agency problem is one of those factors shaping the policy and decision of dividend payout.
6 critical to a firm's success. As representatives of shareholders, the board of directors plays a crucial role in overseeing the creation and execution of plans by management to balance the interests of multiple stakeholders involved. Besides the overseeing role, a board of directors offers several benefits for corporations. Benefits include offering advice and counsel, access to external resources and legitimacy. To realize these benefits companies need to connect with their external environment. A board’s structure plays a pivotal role generating this connecting, gaining legitimacy and assuring access to resources. Board members are appointed in alignment with one or more of the resources required, and the company matches the resources provided by the directors with its needs (Hillman, Cannella and Peatzold, 2000). This resource dependency perspective, as described by Hillman and Dalziel (2003) examines how board capital leads to resource provision for the firm. Board capital consists of both human capital (experience, reputation and expertise) and relational capital (networks and external contingencies).
Mature and well-established firms tend to pay dividends, as young firms face abundant investments opportunities within a financial life cycle. Due to the limited resources, retention dominates distribution in firms early in the life cycle. Mature firms are more likely to pay dividends because of high profitability and the limited amount of investment opportunities available. Findings from Fama and French (2001) show that firms with high-profitability and low-growth rates tend to pay dividends. While firms with low-profit/high-growth rates tend to retain profits. DeAngelo, DeAngelo and Stulz (2006) mention non-life cycle theories like signaling theory have a hard time to explain current dividends. As signaling theory predicts that firms have higher dividend payout when outside investors need information regarding the firm’s future earnings. Early on in the life cycle, firms with low retained earnings to total equity seem ideal signaling candidates because of the lack of substantial earnings, but few of those firms pay dividends. Large, mature firms, however, with high retained earnings to total equity pay dividends even though investors have many sources of information about the firm. Denis and Osobov (2008) add to this by casting further doubt on signaling theory as a first-order determinant of dividends. They find that aggregate dividends do not decline over time and are concentrated amongst the largest, most profitable payers who are firms that are least in need of signaling their profitability.
7 possibilities. Thus high dividend payout could signal low investment possibilities and lower firm value. If low dividend payout is announced, free cash flow could be used to invest in value-enhancing projects instead of dividend payout. Therefore the announcement of high or low dividends affects shareholder wealth two ways. Either directly through the increase of wealth through dividend gains or indirectly with the increase of wealth through the rise of share prices. It is therefore difficult to decide whether or not agents are acting in the best interest of principals, as the retained earnings could be invested which favors shareholders as well.
2.2. Board composition and board gender diversity
The board of directors is responsible for two functions as suggested by Hillman and Dalziel (2003), the first is monitoring management to ensure that it is acting in the best interest of shareholders, as suggested by agency theory and the second is, facilitating access to information and other resources, as promulgated by resource dependency theory. In monitoring management, Schellenger, Wood and Tashakori (1989) hypothesize that board composition and dividend policies substitute one another as mechanisms to deal with the agency problems. They find no conclusive evidence to support this claim, however, they do find evidence that the composition of the board affects dividend policy. In particular outside directors presence, as discussed later on. Differences in the board composition may result in different choices for dividend payout. Following the work of Ruigrok, Peck and Keller (2006) several variable facets of board composition and their effects on dividend payout are examined. These variables are commonly discussed in the literature and include board size, board independence (Schellenger, Wood and Tashakori,1989; Hillman and Dalziel, 2003; de Villiers, Naiker and van Staden, 2011) and CEO duality.
8 Another finding observed through the work of Golden and Zajac (2001) is that the effective available speaking time per individual board member decreases during meetings. This results in board members acting as free riders and minimizing their contributions during these meetings. Besides the negative effects on board size there is also evidence in favor of large boards. Larger boards could be an opportunity to increase the amount of expertise and advice available to executives. Resource dependency theorists state that larger boards allow corporations to create links with business partners and deal with more environmental uncertainty. Cheng (2008) finally finds evidence that larger boards invest more conservative and the outcomes of their decisions are more moderate. Depending on the point of view this is good or bad. To conclude, on the one hand, larger boards could encounter communication and coordination problems that reduce their effectiveness. On the other hand, increasing the board size could enhance expertise, counterbalance CEO dominance and enhance effectiveness (Gonzales and André, 2014). Since a board of directors acts in the interest of shareholders and tries to counterbalance CEO dominance. I suggest that the size of the board of directors is positively related to dividend payout.
H1: The size of the board of directors positively affects dividend payout.
9 Ruigrok, Peck and Keller (2006) find that many studies analyzed the impact of board independence on corporate performance. But their results are often inconclusive and not significant. Agency theorists suggest board independence as one of the most essential requirements for board effectiveness. The agency perspective is focused on the monitoring role of the board. There are, however, several other critical board tasks. Inside directors can be valuable in the expertise-counsel role but criticized for their lack of independence. Similarly, outside directors with business or personal relations within a firm could be outstanding from a resource dependence perspective but criticized regarding their lack of independence (Dalton Daily, Johnson and Ellstrand, 1999). Related to our research, Ibrahim and Angelidis (1995) find evidence suggesting that independent directors are more likely be sensitive to social demands. These social demands are predominantly ethical and environmental however I posit that shareholders demanding dividends also pressure independent directors to act upon this request. Therefore I suggest that boards comprised of a higher amount of independent members will be more likely to serve the best interests of shareholders. Therefore hand out more dividends as this affects the power position of dominant executives.
H2: More board independence is associated with higher dividend payout.
According to Liao, Muhkerjee and Wang (2015), two factors contribute to an independent board of directors. One is the outside domination, indicating that the majority of board members are outsiders (non-executives) and second, the separation between CEO and board chairman. Because board independence, as discussed above might be compromised when the chairman of the board also serves as the CEO of the company (Fama and Jensen, 1983). Now CEO duality, which indicates a single individual serving as both CEO and board chairman is still heavily debated. Finkelstein and D’Aveni (1994) mention a “double-edged sword” referring to the trade-off between the unity of command and the oversight associated with a separate board chairman.4 Having a CEO at the top of the organization and the board results in unambiguous leadership decision making authority and sends reassuring signals to shareholders. A downside is the possibility of CEO entrenchment.5 A CEO who also is the chairperson of a firm’s board of directors dominates meetings by changing the agenda and content of this agenda. This duality also results in a conflict of interest, as the CEO who decides on the overall strategic
4 Unity of command is defined as the existence of a single manager at the top with formal authority to whom all
other managers report.
5 “CEO entrenchment occurs when managers gain so much power that they are able to use the firm to further their
10 management is also in the top position to evaluate the effectiveness of this strategy (Finkelstein and D’Aveni, 1994). Resulting from CEO dominance is the CEO entrenchment as mentioned above. Entrenched managers choose a capital structure in a manner that maximizes their ability to empire-build without compromising their ability to safeguard against hostile takeovers (Zwiebel, 1996).6 CEOs use company resources to empire build, thereby reducing the number
of earnings at hand to dispose to shareholders. I hypothesize based on the arguments mentioned above that CEO duality is negatively associated with dividend payout. As high dividend payout is often used as a mechanism to counter managerial optimism and dominance.
H3: CEO duality negatively affects dividend payout.
The ratio, female to male, or male to female, at top executive and non-executive positions within big corporations is a major point of focus nowadays. Within the developed world, more and more countries commit to regulations regarding gender representation at these top positions. Important in this research is the effect of gender diversification within boards on dividend payout. Including female members onto boards has several benefits. Terjesen, Sealy and Singh (2009) studied previous work regarding board gender diversity across 400 publications on several fields of study. At firm-level several common perspectives are mentioned, these include: resource dependency, institutional and agency theory. They find that boards with three or more females are likely to ensure effective communication between board members and its stakeholders. Promote non-financial performance measures like customer satisfaction, employee satisfaction, and gender representation more actively, as well as innovation and corporate social responsibility. Women board members contribute in different ways to the governance tasks that the board of directors is entrusted with. Chen, Gramlich and Houser (2017) mention increased attendance rates at board meetings, more alternatives discussed during discussions, less political inquisitions, more equity-based payments to directors, higher and more conservative earnings quality and a stronger management oversight. Francoeur (2008) adds to this by mentioning that women often bring a fresh perspective to complex issues, which in turn helps to correct informational biases in strategy formulation and problem solving. He concludes that firms operating in complex environments, benefit from a high percentage of women officers and experience slightly larger monthly abnormal returns. However, having more women on corporate boards and top management does not generate
6 Empire building is an act of attempting to increase the size and scope of an individual’s power and influence.
11 significant excessive returns. Following his arguments, women thrive in a complex setting. Saeed and Sameer (2017) find that women tend to be more averse, however, this risk-taking propensity of female directors is context specific. They also show that board gender diversity may substitute or complement the dividend policy as disciplining mechanism. But most important is the evidence provided of a negative relationship between the presence of women on boards and dividend payments which contradicts the findings of studies in developed markets. As the sample consisted of emerging markets, the negative relationship confirms the conservative behavior of female directors in markets with high institutional uncertainty. It shows that women make decisions based on the environment (context specific). Adams and Ferreira (2009) show that gender-diverse boards are tougher monitors. Which should address the agency conflict. However, over monitoring could decrease shareholder value. Based on previous arguments I believe that female presence on the board of directors increases the amount of dividends paid to decrease the principal-agent conflict by reducing free cash flows available to agents.
H4: Board gender diversity positively affects dividend payout
2.3. Legal origin and corporate governance
12 effects a variety of institutions and governs how investors, firms and other groups act. Differences could possibly be observed and I hypothesize that in common law dividend payout is higher compared to civil law countries.
H5: Countries with a common law origin have higher dividends payout than countries with a civil law origin as the orientation is more towards shareholders.
3. Data sources, sample selection and methodology
3.1. Data sources and sample selection
For this study, data is derived from several databases. Thomson Reuters DataStream provides financial data and data regarding corporate governance. This data is merged with the Worldwide Governance Indicator(s) (WGIs) and GDP per capita in constant 2010 US$ provided by the World Bank.7 Civil and common law classifications are based on the papers of Choy, Gul and Yao (2011) and La Porta, Loped-de-Silanes, Shleifer and Vishny (1998). To test the proposed hypotheses, a sample is selected based on a constituent list of actively listed and publicly traded companies from the ASSET4 dataset. All firm-year observations with the required financial and governance data in the period of 2007-2017 in Thomas Reuters DataStream are incorporated. This sample is selected for several reasons. First, as this research is interested in dividend payout and differences amongst countries with civil and common law origin the sample consist of as many countries as possible. The final sample consists of 28 countries, 7 representing common law and 21 representing civil law origins as shown in table 1, panel A reports the sample details by country, panel B per year and panel C per industry. Secondly, increasing the number of countries allows for a better representation of the different legal origins as the sample is dominated by firms in the United States (31.28%). Japan (12.45%) is the second largest contributor to the sample. Limited observations of firms are shown in Thailand (0.65%), Belgium (0.61%) and Russian Federation (0.53%).
The initial sample consisted of 6,049 companies but after excluding financial firms (SIC code 6000-6999), utility firms (SIC code 4900-4999), firms with missing information (Liao et al., 2015) and dropping countries with below 100 firm-year observations the final unbalanced sample consists of 2,762 firms and 20,171 firm-year observations. Financial and utility firms
13 are excluded because of differences in accounting standards, regulatory requirements, and financial characteristics. These differences result in firms becoming incomparable to firms in other sectors (Saeed and Sameer, 2017).
3.2. Empirical specification
To analyze the relationship between board composition and dividend payout, an OLS regression model is applied. The model specification to test the hypotheses is:
Dividend payouti,t = α + β1Board sizei,t−1+ β2Board independence i,t−1+
β3CEO dualityi,t−1 + β4Femalediri,t−1+ 𝛾Controlsi,t−1,c+ Industryi+ Yeart+
Countryc (1)
where i denotes an index of the firm; t is an index for year and c is an index for the country. Industry, Year and Country denote industry (based on Fama-French 49), year and country fixed effects.
Dividend payout, proxied by dividend yield and calculated as the dividend per share to price per share (Chen, Leung and Goergen, 2017; Saeed and Sameer, 2017).8 The main variables of interest are board size, board independence, CEO duality and board gender diversity. Board size is measured as the number of individual members on the board. Board independence is measured as the percentage of unaffiliated members on the total board. CEO duality is a dummy variable that takes the value of one if the CEO is also the chairman of the board and a zero otherwise (Chen, Leung and Goergen, 2017; Saeed and Sameer, 2017). Board gender diversity (Femaledir) is proxied as the percentage of female directors on the board of directors, thus the amount of female directors divided by the total amount of board members (Loukil and Yousfi (2016); Saeed and Sameer (2017). Controls encompass control variables to account for the impact of firm- and country-specific characteristics on dividend payout. Fama and French (2001) show that dividend payers are often larger and have a higher profitability than non-payers. Therefore I control for firm size as the natural logarithm of market capitalization. Denis and Osobov (2008) show that dividend payout is concentrated amongst the most profitable firms requiring a control for profitability. Profitability is proxied by the return on assets and measured as net income before extra items/preferred dividends to total assets. The standard
8 The findings are robust to an alternative measure of dividend payout. The industry adjusted dividend yield
14 deviation of the return of assets over five years is used to compute the return volatility (RISK), a proxy that indicates business conditions. A control for return volatility is required as firms with high return uncertainty are less likely to pay dividends (Fama and French, 2001). Leveraged firms require more funds to pay the interest on borrowed money resulting in decreased amounts of money available for dividends. Leverage is defined as total debt to total assets (Chen, Leung and Georgen, 2017). Growth opportunities create investment possibilities, which affect the amount of retained earnings, which in turn affect the dividend payout as funds can be relocated to invest in growth opportunities. Therefore, growth is controlled for by using the market-to-book-ratio, which equals the book value of the firm divided by market capitalization (Jurkus, Park and Woodard, 2011). Civil law is an indicator variable for the legal origin of the country the company is operating in, 1 indicating a civil law origin and zero indicating a common law origin. Apart from the legal variation across countries, political and economic variation influence the results as well. To account for governance differences amongst countries the six WGIs are incorporated in the model. These are: voice and accountability, political stability and absence of violence/terrorism, government effectiveness, regulatory quality, rule of law, and control of corruption. Each of governance ranges in value from approximately -2.5 (weak) to a 2.5 (strong) governance performance. To ensure that variables as mentioned above are not just capturing the effects of the “wealth” within a country, the log of GDP per capita in constant 2010 US$ (lnGDP) is included (Bernile, Bhagwat and Yonker, 2018; Choy, Gul and Yao, 2011). The variables from the World Bank have missing values in 2017, therefore, after evaluating the data points and observing minor differences between the last values the choice is made to use the last value available if values are missing. Finally, variables are winsorized at the 1% and 99% level to minimize the possible effects of outliers and all explanatory (as well as control variables) are lagged by one year to mitigate endogeneity concerns. Appendix A provides a list with all variables, definitions and data sources.
3.3. Descriptive analysis
As mentioned in section 3.1., table 1 provides sample details by year, industry and country. It shows the distribution of firm-year observations based on legal origin, hence a common- versus civil law perspective.
15 Kingdom (1,897), Australia (1,194) and Canada (1,181) surpasses the total amount of observations with a civil law legal origin (8,265). As all firm-year observations with missing values are dropped and the governance variables showed the lowest amount of observations at the start I assume that the remaining distribution of the sample is partly due to reporting requirements and the number of listed firms within these markets. Common law countries tend to be more aggressive with reporting requirements, especially after the final crisis. This is supported by the yearly increasing amount of observations since 2007.
Panel B shows the distribution of firm-year observations by year and legal origin in numbers and percentages. The proportions of contribution vary over time, with a trend towards smaller differences between the two groups (based on legal origin). This while the amount of observations steadily grows. It is interesting to see that the number of companies that provide sufficient information (no missing values in variables) increase to the amount of 2,704 in 2017, this is 97.9% of all companies in the sample. Corporate information seems to be more widely available and there seems to be a decrease in opaqueness.
16 Table 1. Sample details by country, year and industry.
Final sample consists of an unbalanced panel of 21,171 firm-year obs. distributed over 2,762 non-financial firms. Table describes the destribution of observations based on legal origin accross country (panel A), year (panel B) and industry (panel C). Panel A shows the amount, percentage and cumalitive percentage of firm-years observations by country based on legal origin. Panel B describes the amount of observations by year and by legal origin (obs and percentage) and Panel C describes the distribution based on the Fama-French 12 industries by legal origin (financial and utility firms excluded).
Panel A: Country distribution
Country Name No of obs. % Cum. %
Australia 1,194 5.92 5.92 Canada 1,181 5.85 11.77 Hong Kong 735 3.64 15.42 India 316 1.57 16.98 Singapore 273 1.35 18.34 United Kingdom 1,897 9.40 27.74 United States 6,310 31.28 59.03
Common Law origin 11,906 59.03
Belgium 124 0.61 0.61 Brazil 209 1.04 1.65 China 350 1.74 3.39 Denmark 181 0.90 4.28 Finland 200 0.99 5.27 France 670 3.32 8.60 Germany 492 2.44 11.04 Italy 139 0.69 11.72 Japan 2,511 12.45 24.17 Korea, Republic of 500 2.48 26.65 Malaysia 164 0.81 27.47 Mexico 144 0.71 28.18 Netherlands 256 1.27 29.45 Norway 153 0.76 30.21 Russian Federation 107 0.53 30.74 South Africa 425 2.11 32.84 Spain 214 1.06 33.91 Sweden 327 1.62 35.53 Switzerland 338 1.68 37.20
Taiwan, Province of China 630 3.12 40.33
Thailand 131 0.65 40.97
Civil Law origin 8,265 40.97
17
Table 1. (continued)
Panel B: By year
Year No of obs. Common % Civil %
2007 987 631 63.9 356 36.1 2008 957 702 73.4 255 26.6 2009 1,198 858 71.6 340 28.4 2010 1,457 984 67.5 473 32.5 2011 1,971 1,071 54.3 900 45.7 2012 2,042 1,106 54.2 936 45.8 2013 2,096 1,123 53.6 973 46.4 2014 2,140 1,153 53.9 987 46.1 2015 2,176 1,169 53.7 1,007 46.3 2016 2,443 1,419 58.1 1,024 41.9 2017 2,704 1,690 62.5 1,014 37.5 Total 20,171 11,906 59.0% 8,265 41.0% Panel C: By industry
Industry No of obs. Common % Civil %
Non Durables 1,593 865 54.3 728 45.7 Durables 863 351 40.7 512 59.3 Manufacturing 3,139 1,486 47.3 1,653 52.7 Energy 1,458 1,096 75.2 362 24.8 Chemicals 1,096 461 42.1 635 57.9 Business Equipment 2,676 1,508 56.4 1,168 43.6 Telecom 937 454 48.5 483 51.5 Shops 2,416 1,675 69.3 741 30.7 Healthcare 1,274 822 64.5 452 35.5 Other 4,719 3,188 67.6 1,531 32.4 Total 20,171 11906 59.0% 8,265 41.0%
18 common law origin, this could be a predictor for the number of independent board members, as a dual position equals more controlling power and independent directors are supposed to decrease this controlling power. The presence of female directors is slightly higher in countries with a common law origin, which is surprising as predominantly European countries are enforcing gender quotas. However, compliance with these quotas is a multi-year project and therefore the results could not be showing yet since it is a mean over an 11 year period. Apart from that, the sample for civil law origin is dominated by Japanese firms which do not imply quotas, the government merely encourages firms to include female directors.
Regarding firm characteristics, firms in countries with a common law origin tend to be smaller but show on average more growth opportunities than firms with a civil law origin. This could indicate that the latter are more mature firms which adds to the earlier findings that overall dividend yields are higher in countries with a civil law origin. Next are the six WGIs were political stability shows no difference between the two groups (both being 0.048). All the other five variables show values particularly in favor of common law economies. As voice and accountability, government effectiveness, regulatory quality, rule of law and control of corruption are on average higher in countries with a common law origin. The WGIs show the perceptions of governance and summarize the views on the quality of governance within a country. Based on the results favoring common law economies I assume that common law countries possess a better quality of governance overall. Finally, the country level economic variable GDP per capita in constant 2010 US$ shows a significant difference of around 450 US$ between the two groups indicating that the countries with a common origin are slightly more wealthy.
19 Table 2. Summary statistics
This table reports the means and standard deviations of all variables used in this study for the whole sample and the subsamples of firms in countries with common or civil law origin. The differences between the two subsamples are reported for each variable, along with t-statistics based on the two sample t-test. (The civil law variable is excluded as this variable is used to define the groups for the two sample t-test).
Whole-sample Firm-year obs. Common law origin
Firm-year obs. Civil law origin
Variable N = 21,171 N = 11,906 N = 8,265
Mean S.d. Mean S.d. Mean S.d. Diff t-stat
Dividend yield 2.050 1.905 1.894 1.950 2.274 1.814 0.380*** 14.0 Industry-adjusted DY 0.320 1.871 0.191 1.916 0.506 1.787 0.314*** 11.8 Board size 10.181 3.212 9.699 2.576 10.876 3.847 1.176*** 26.0 Board independence 0.512 0.306 0.662 0.233 0.296 0.269 -0.366*** -100.0 CEO duality 0.443 0.497 0.494 0.500 0.370 0.483 -0.125*** -17.7 Femaledir 0.120 0.114 0.133 0.103 0.102 0.125 -0.031*** -19.1 Firm size 7.148 1.043 6.703 0.791 7.790 1.026 1.088*** 84.9 Market-to-book 2.951 3.544 3.306 3.980 2.440 2.720 -0.867*** -17.2 RISK 0.010 0.038 0.013 0.046 0.005 0.022 -0.009*** -16.0 Leverage 0.240 0.172 0.241 0.176 0.238 0.167 -0.003 -1.2 Profitability 0.048 0.100 0.048 0.116 0.048 0.071 0 -0.3
Voice and acc. 1.024 0.534 1.118 0.312 0.887 0.725 0.231*** -30.9
Political stability 0.617 0.486 0.616 0.395 0.618 0.593 0.002 0.3 Government effect. 1.463 0.463 1.565 0.300 1.316 0.598 -0.249*** -39.0 Regulatory quality 1.349 0.505 1.530 0.383 1.088 0.545 -0.442*** -67.7 Rule of law 1.438 0.550 1.632 0.293 1.158 0.694 -0.473*** -66.4 Control of corrupt. 1.379 0.644 1.522 0.421 1.174 0.828 -0.348*** -39.1 ln(GDP) 10.488 0.731 10.674 0.562 10.221 0.853 -0.453*** -45.4
*** Denotes statistical significance at the 1% level
4. Results
4.1. Regression results
Table 3 presents the OLS results where the dependent variable is dividend payout, measured by dividend yield, that is dividends per share to share price in percentage. The nine regressions vary in control variables added and samples used. Column (1) to (3) show results using the whole sample created for this research. Column (4) to (6) address the common law sample and column (7) to (9) the civil law sample with regressions similar to the first three regressions. The civil law variable is blank because the legal origin is used to set up the samples.
20 the regression as firm-level control variables. Industry, year and country fixed effects are present as well. In addition to column (2), column (3) includes the country-level control variables, being the civil law indicator variable, the WGIs and the logarithm of GDP per capita (in current 2010 US$). As shortly mentioned above, regression (4) to (6) is the same as the first three regressions only focus on the common law sample with 1,717 firms (11,906 observations). Regression (7) to (9) repeats this process for the civil law sample with 1,046 firms and 8,265 observations.
Looking at regression (1), all coefficients are positive and significant at the 1% level except for CEO duality which is negative (-0.014) and not significant. To further address these results, the coefficients for board size, board independence and femaledir are 0.050, 0.587 and 1.245 respectively. Their interpretation is that a single unit increase of either board size, board independence and femaledir increases the dividend payout, as proxied by dividend yield with respectively 0.050%, 0.587% or 1.245%. It appears, looking at the whole sample, that board composition (minus CEO duality) and board gender diversity do positively affect dividend payout. These results are, apart from CEO duality in line with my expectations. Perhaps if the control variables are included the explanatory variable CEO duality becomes significant. Regression (2) shows several differences in coefficients of the explanatory variables and has an adjusted R-squared of 0.267, indicating that the explanatory power of the model increased slightly. The intercept changes from significant with a coefficient of 1.680 to insignificant with a coefficient of 0.622. Furthermore, firm size, market-to-book ratio, RISK, leverage and profitability are all statistically significant at the 1% level and CEO duality remains negative and statistically insignificant. While the coefficients of the explanatory variables slightly change they stay significant thus reinforcing earlier results.
21 earnings to invest in these growth opportunities thereby leaving less money on the table for dividend payout. The negative coefficient on riskiness is in line with predictions that more volatile returns and higher uncertainty forces firms to act more risk-averse, increasing the amount of risk could result in firms reducing the amount of dividend payout to increase firm financial buffers. Finally, more profitable firms have more money available to disperse. Civil law has a coefficient of 2.346 and is significant at the 5% level. This indicates that dividend payout is higher in countries with a civil law origin compared to countries with a common law origin. Strengthening the results of the two-sample t-test also indicating higher dividend payout in civil law economies. The other added country-level control variables are not all statistically significant. Voice and accountability and rule of law have negative coefficients, regulatory quality positive but neither are significant. In comparison, political stability, control for corruption and government effectiveness are significant at the 1% level, the first two being positive and the latter being negative. Interpreting these results it appears that political stability and control for corruption have a positive effect on dividend payout while government effectiveness negatively influences the amount of dividend payout. The ‘perceived’ governance has different influences on the amount of dividend. Thus, while including more control variables did increase the explanatory power of the model, CEO duality is still insignificant.
The results from regression (4) to (6) appear to be in line with previous results. Board size, board independence and femaledir are all positive and highly significant with coefficients of 0.040, 0.739 and 1.174 respectively while CEO duality is still negative and insignificant. The coefficients and significance of the firm-level control variables change while the direction remains the same. Firm size, market-to-book-ratio, RISK, leverage and profitability are all still significant. Civil law shows “blank” as the distribution of the sample is based on legal origin. It is interesting to mention that rule of law is the only significant WGI (coefficient is -0.927) looking at the countries with a common law origin. The negative relation indicates that an increase in the rule of law should decrease the amount of dividend payout. This is interesting as for the whole sample this coefficient was negative but insignificant. The logarithm of GDP per capita (in constant 2010 US$) is even more positive and still highly significant. If the economy increases in strength the amount of dividend payout rises.
22 to influence dividend payout in countries with a civil law origin. This is contrary to findings for the whole and common law sample. Even though dividend payout is higher (on average) in countries with a civil law origin, board composition and board gender diversity do not seem to be responsible for this difference. A potential reason for this observation could be the preferred way of financing in countries with a civil law origin. According to Da Silva, Goergen and Renneboog (2004) banks may act as substitutes to alternative governance devices like board composition and board gender diversity. Concerning the control variables, the dividend payout increases with the increase in firm size, profitability, political stability, control of corruption and the logarithm of GDP per capita. Whereas the market-to-book-ratio, government effectiveness and rule of law show a negative influence on dividend payout (all highly significant). Next to the insignificant board composition and board gender diversity variables riskiness, leverage, voice and accountability and regulatory quality do not seem to influence dividend payout as they are insignificant. In civil law four out of the six WGIs have an effect on dividend payout. For common law countries this is only one out of six as the rule of law has a negative effect on dividend payout. For civil law, an increase in the “perceived” governance by institutions has mixed effects on the dividend payout. The most notable difference is the control of corruption variable which is negative but insignificant in common law but positive and highly significant in civil law. If public power in civil law is less used for private gain the amount of dividend payout increases. This could be an indication for firms lowering financial buffers as the fear of expropriation by government decreases.
Now to address the hypotheses constructed for this research. When looking at the whole sample enough evidence has been presented that the size of the board of directors affects dividend payout. This also holds for the common law sample but not for the civil law sample. The positive and highly significant coefficient for board size confirms hypothesis 1 for countries with a common law origin. The same goes for the number of independent board members that also positively impact the dividend payout. The results are in line with earlier research from Chen, Leung and Georgen (2017) and partly with results from Saeed and Sameer (2017). They only find board independence to be significant in China which is contrary to my findings for the sample with civil law origin. Sufficient evidence is presented for the whole and common law sample to support hypothesis 1 and 2 but insufficient evidence is presented to support these claims for countries with a civil law origin.
23 not support this hypothesis as the coefficient is negative and not significant. CEO duality could be insignificant due to a limitation in the number of CEOs with dual positions. After looking at the data, it is clear that this is not the case since 44% of CEOs are also chairing their board. It seems to be that CEO duality has no effect on dividend payout, Chen, Leung and Goergen (2017) find similar results when rerunning their OLS regression for six different proxies of management entrenchment. CEO duality does not affect dividend payout firms in tech industries, firms with CEOs with short tenures and in firms where board independence is high. The positive and highly significant coefficients for femaldir in column (1) to (6) support my claim at hypothesis 4 for the whole and common law sample. However, in column (7) to (9) the coefficient is insignificant indicating that hypothesis 4 does not hold for the civil law sample. For the common law sample, the coefficient (1.174) is quite high compared to the other board composition variables indicating that the amount of female board members substantially affects dividend payout. It appears that the presence of female directors increases dividend payout for countries with a common law origin but not in countries with a civil law origin. Perhaps, an increase in the amount of female board members does not affect dividend payout in countries with a civil law origin because the amount of female directors within an individual firm is to low compared to firms in common law countries. Table 2 showed that the average amount of female board members is higher in common law compared to civil law economies and Terjesen, Sealy and Singh (2009) find positive effects when there are three or more females on the board. It is also possible that board gender diversity has no effect on dividend payout in civil law countries because the high(er) dividend payout (compared to common law) acts as a substitute for board composition to deal with agency conflicts.
25 Table 3. Board composition, board gender diversity and dividend payout
This table reports the results of the OLS regressions for the relationship between board composition, board gender diversity and dividend payout. The dependent variable is dividend yield, dividends per share over share price. Independent variables are board size, total amount of board members, board independence, is the amount fraction of independent board members to total board size, CEO duality, whether the CEO is also the chairman of the board, a 1 if so and a 0 otherwise. Female directors (femaledir) is the fraction of female directors on the board. Firm size is the log of market capitalization, market-to-book ratio is the market value of common equity to the balance sheet value. Riskiness is the 5 year adjusted ROA, leverage is total debt to total assets and profitability the return on assets. Civil law is a dummy variable indicating legal origin. Others are country controls. Between brackets is the standard error.
OLS regressions
Dependent variable: Dividend Yield
(1) (2) (3)
Variable Whole sample (n = 2,762 firms)
Intercept 1.680*** 0.622 -18.347*** [0.424] [0.518] [4.507] Board size 0.050*** 0.024*** 0.024*** [0.009] [0.009] [0.009] Board independence 0.587*** 0.430*** 0.425*** [0.131] [0.128] [0.128] CEO duality -0.014 -0.035 -0.043 [0.051] [0.049] [0.050] Femaledir 1.245*** 0.987*** 1.026*** [0.277] [0.276] [0.278] Firm size 0.214*** 0.220*** [0.053] [0.053] Market-to-book ratio -0.017*** -0.017*** [0.006] [0.006] RISK -2.840*** -2.824*** [0.526] [0.527] Leverage 0.530*** 0.511*** [0.166] [0.166] Profitability 2.958*** 2.968*** [0.256] [0.256] Civil law 2.346** [0.912]
Voice and accountability -0.373
[0.253] Political stability 0.326*** [0.123] Government effectiveness -0.654*** [0.179] Regulatory quality 0.122 [0.173] Rule of law -0.139 [0.226] Control of corruption 0.405** [0.169] ln(GDP) 1.794*** [0.418] Adjusted R² 0.231 0.267 0.268 N 20,171 20,171 20,171
Industry effects Yes Yes Yes
Year effects Yes Yes Yes
Country effects Yes Yes Yes
26 Table 3. (continued)
OLS regressions
Dependent variable: Dividend Yield
(4) (5) (6) (7) (8) (9)
Variable Common law (n=1,717) Civil law (n= 1,047)
Intercept 0.649 0.097 -21.480*** 1.212** -0.584 -18.924*** [0.599] [0.785] [8.183] [0.511] [0.678] [5.040] Board size 0.075*** 0.039** 0.040** 0.027*** 0.014 0.013 [0.015] [0.016] [0.016] [0.010] [0.010] [0.010] Board independence 0.924*** 0.722*** 0.739*** 0.084 0.053 0.025 [0.205] [0.204] [0.205] [0.165] [0.152] [0.153] CEO duality -0.005 -0.036 -0.036 -0.032 -0.002 -0.003 [0.071] [0.070] [0.070] [0.068] [0.064] [0.066] Femaledir 1.305*** 1.152*** 1.174*** 0.47 0.199 0.363 [0.366] [0.370] [0.372] [0.416] [0.394] [0.407] Firm size 0.161** 0.157** 0.237*** 0.248*** [0.073] [0.073] [0.078] [0.078] Market-to-book ratio -0.012* -0.012* -0.056*** -0.055*** [0.006] [0.006] [0.018] [0.018] RISK -3.088*** -3.090*** -1.467 -1.354 [0.590] [0.590] [1.177] [1.191] Leverage 0.853*** 0.845*** 0.161 0.142 [0.222] [0.222] [0.242] [0.241] Profitability 2.147*** 2.164*** 6.228*** 6.236*** [0.292] [0.292] [0.617] [0.618] Civil law
Voice and account. -0.455 -0.379
[0.404] [0.358] Political stability 0.009 0.533** [0.191] [0.209] Government effectiveness -0.094 -0.547** [0.399] [0.221] Regulatory quality -0.081 0.112 [0.299] [0.246] Rule of law -0.927** -0.704** [0.424] [0.333] Control of corruption -0.485 1.009*** [0.307] [0.243] ln(GDP) 2.313*** 1.733*** [0.772] [0.469] Adjusted R² 0.232 0.262 0.263 0.258 0.314 0.319 N 11,906 11,906 11,906 8,265 8,265 8,265
Industry effects Yes Yes Yes Yes Yes Yes
Year effects Yes Yes Yes Yes Yes Yes
Country effects Yes Yes Yes Yes Yes Yes
27 4.2. Robustness tests
4.2.1. Industry-adjusted dividend payout
Several robustness test results are shown in Appendix C that strengthens earlier findings regarding board composition, board gender diversity and dividend payout. To ensure that our findings are not driven by the chosen proxy for dividend payout an alternative measure is used. This measure is a worldwide industry-adjusted dividend yield in line with the work of La Porta, Lopez-de-Silanes, Shleifer and Vishny (2000). Each firm within a given industry is adjusted by the worldwide median for that industry based on the Fama-French 12 industry classification. For each firm, the difference between the worldwide industry dividend yield median and dividend yield is computed and used as measure. The rationale behind this is the different stages of maturity and growth within industries that determine dividend payout.
4.2.2. Firms with negative net income
The ability to pay dividends is undeniably associated with the amount of income a firm has. Firms generating losses are less inclined to pay dividends compared to firms generating profits (confirmed in table 3 see profitability coefficient). It is therefore that I exclude all firms reporting negative or zero net incomes during the period of 2007-2017, this reduces the sample size with 2,824 firm-year observations. For the robustness test I re-run the first three regressions and see if results have changed. Results are in line with earlier findings, except for the political variable voice and accountability which is now statistically significant. This could indicate that the perceived freedom of citizens influences the amount of dividend payout. As better governance reduces the amount of dividend payout perhaps because firms see no need to reduce the amounts of excess cash at hand.
4.2.3. Effects financial crisis
28 After excluding the years of the financial crisis political stability becomes less important compared to perceived freedom (voice and accountability).
5. Conclusion
In this research paper, the purpose is to examine the effects of board composition and board gender diversity on dividend payout of firms in countries with different legal origins. The results indicate differences amongst the whole sample, the common law sample and the civil law sample.
The first four hypotheses cover the composition and gender diversity facets of a board of directors. Hypothesis 1,2 and 4 hold for the whole and common law sample as board size, board independence and the percentage of female directors (femaledir) are all positively significant and thus impact the amount of dividend payout. The results of the analysis suggest that each one-point increase of board size, board independence or board gender diversity increases the dividend payout in countries with a common law origin with respectively 0.040%, 0.739% and 1.174%. On the contrary, no effects on dividend payout show in countries with a civil law origin. These differences are observable as board size, board independence and board gender diversity become insignificant and do not affect dividend payout anymore. Hypothesis 1, 2 and 4 are all accepted for the whole and common law sample but rejected for the civil law sample. No evidence to support my claim at hypothesis 3 has been presented. A CEO also chairing the board does not significantly influence dividend payout while looking at the whole sample, the common law sample and the civil law sample. Chen, Leung and Goergen (2017) found similar results while looking at different types of managerial entrenchment. Furthermore, the average dividend yield of companies with a CEO as chairman is lower than companies without a CEO in dual positions (1.783 to 2.263). From an agency perspective, it seems that CEOs in dual positions retain more earnings however it is not clear how the firms perform and therefore it is merely an assumption.
29 La Porta, Lopez-de-Silanes and Shleifer (2000), who find that more investor protection leads to higher dividends as minority shareholders are able to extract dividends. Instead, the “substitute” by La Porta, Lopez-de-Silanes and Shleifer (2000) is more suited because in order to raise external capital firms need to establish a good reputation towards shareholders. Therefore, in countries with weak(er) legal protection of minority shareholders, a higher dividend payout is required to establish a reputation for good treatment towards shareholders. Similar, in countries with stronger shareholder protection, the need for dividend payout is weaker as the reputational mechanism is less strong.’ It is possible that the “perceptions” behind the WGIs are off and therefore do not fully reflect true values. Regarding governance indicators, their influence is substantial in civil law countries and barely influential in common law countries. Results point out that in a stronger institutional setting dividend payout and board composition do not substitute each other as monitoring mechanisms. This is the case for common law countries but not for civil law countries as the board composition/board gender diversity variables are not significant but several WGIs are. So in a weaker “perceived” institutional setting dividends and board composition do substitute each other.
30
6. Limitations and future directions
31
7. References
Adams, R.B., de Haan, J., Terjesen, S., van Ees, H., 2015. Board diversity: moving the field forward. Corporate Governance: An International Review 23, 77-82.
Adams, R.B., Ferreira, D., 2009. Women in the boardroom and their impact on governance and performance. Journal of Financial Economics 94, 291-309.
Berle, A.A., Means, G.C., 1932. The modern corporation and private property. Macmillan, New York.
Bernile, G., Bhagwat, V., Yonker, S., 2018. Board diversity, firm risk and corporate policies. Journal of Financial Economics (aip), 1-25.
Black, F., 1976. The dividend puzzle. Journal of Portfolio Management 2, 5-8.
Booth, L., Zhou, J., 2017. Dividend policy: a selective review of results from around the world. Global Finance Journal 34, 1-15.
Canada, J., Kuhn, J.R., Sutton, S.G., 2008. Accidentally in the public interest: the perfect storm that yielded the Sarbanes-Oxley act. Critical Perspectives on Accounting 19, 987-1003.
Chen, J., Leung, W.S., Goergen, M., 2017. The impact of board gender composition on dividend payouts. Journal of Corporate Finance 43, 86-105.
Chen, L.H., Gramlich, J., Houser, K.A., 2017. The effects of board gender diversity on a firm’s risk strategies. Accounting and Finance , 1-41.
Cheng, S., 2008. Board size and the variability of corporate performance. Journal of Financial Economics 87, 157-176.
Choy, H.L., Gul, F.A., Yao, J., 2011. Does political economy reduce agency costs? Some evidence from policies around the world. Journal of Empirical Finance, 18, 16-35.
32 Da Silva, L.C., Goergen, M., Renneboog, L., 2004. Dividend policy and corporate
governance. Oxford University Press on Demand.
Dalton, D.R., Daily, C.M., Johnson, J.L. Ellstrand, A.E., 1999. Number of directors and financial performance: a meta-analysis. Academy of Management Journal 42, 674-686.
DeAngelo, H., L, DeAngelo., Stulz, R.M., 2006. Dividend policy and the earned/contributed capital mix: a test of the life-cycle theory. Journal of Financial Economics 81, 227-254.
Denis, D.J., Osobov, I., 2008. Why do firms pay dividends? International evidence on the determinants of dividend policy. Journal of Financial Economics 89, 62-82.
De Villiers, C., Naiker, V., van Staden, C.J., 2011. The effect of board characteristics on firm environmental performance. Journal of Management 37, 1636-1663.
Enriques, L., Volpin, P., 2007. Corporate governance reforms in continental Europe. Journal of Economic Perspectives 21(1), 117-140.
Fama, E.F., French, K., 2001. Disappearing dividends: changing firm characteristics or lower propensity to pay? Journal of Financial Economics 60, 3–43.
Fama, E.F., Jensen, M.C., 1983. Separation of ownership and control. The Journal of Law & Economics 26, 301-325.
Finkelstein, S., D’Aveni, R.A., 1994. CEO duality as a double-edged sword: how do boards of directors balance entrenchment avoidance and unity of command. The Academy of Management Journal 37, 1079-1108.
Francoeur, C., Labelle, R., Sinclair-Desgagné, B., 2008. Gender diversity in corporate governance and top management. Journal of Business Ethics 81, 83-95.
33 Gonzalez, A., André, P., 2014. Board effectiveness and short termism. Journal of Business Finance & Accounting 41, 185-209.
Harjeto, M., Laksmana, I., Lee, R., 2015. Board diversity and corporate social responsibility. Journal of Business Ethics 132, 641-660.
Hillman, A.J., Cannella, A.A., Peatzold, R.L., 2000. The resources dependence role of corporate directors: strategic adaptation of board composition in response to environmental change. Journal of Management Studies 37, 235-255.
Hillman, A.J., Dalziel, T., 2003. Board of directors and firm performance: integrating agency and resource dependency perspectives. Academy of Management Review 28, 383-396.
Ibrahim, N.A., Angelidis, J.P., 1995. The corporate social responsiveness orientation of bord members: are there differences between inside and outside directors? Journal of Business Ethics 14, 405-410.
Jensen, M.C., 1986. Agency cost of free cash flow, corporate finance, and takeovers. American Economic Review 76, 323-329.
Jensen, M.C., 1993. The modern industrial revolution, exit and the failure of internal control systems. The Journal of Finance 48, 831-880.
Jensen, M.C., Meckling, W.H., 1976. Theory of the firm: managerial behavior, agency costs and ownership structure. Journal of Financial Economics 3, 305-360.
Jurkus, A.F., Park, J.C., Woodard, L.S., 2011. Women in top management and agency costs. Journal of Business Research 64, 180-186.
Kaufman, D., Kraay, A., Mastruzzi, M., 2010. The worldwide governance indicators. Policy research working paper. Development Research Group, The World Bank.
34 Kock, C.J., Min, B.S., 2016. Legal origins, corporate governance, and environmental outcomes. Journal of Business Ethics 138, 507-524.
La Porta, R., Lopez-de-Silanes, F., Shleifer, A., Vishny, R.W., 1998. Law and finance. Journal of Political Economy 106, 1113-1155.
La Porta, R., Lopez-de-Silanes, F., Shleifer, A., Vishny, R.W., 2000. Agency problems and dividend policies around the world. Journal of Finance 55, 1-33.
Liao, L.K.C., Muhkerjee, T., Wang, W., 2015. Corporate governance and capital structure dynamics: an empirical study. Journal of Financial Research 39, 169-191.
Loukil, N., Yousfi, O., 2016. Does gender diversity on corporate board increase risk-taking? Canadian Journal of Administrative Sciences 33, 66-81.
Modigliani, F., Miller, H., 1958. Corporate finance, and the theory of investment. American Economic Review 48, 261-297.
Modigliani, F., Miller, H., 1961. Dividend policy, growth and the valuation of shares. Journal of Business 34, 411-433.
Mueller, G.C., Barker III, V.L., 1997. Upper echelons and board characteristics of turnaround and non-turnaround declining firms. Journal of Business Research 39, 119-134.
Rozeff, M.S., 1982. Growth, beta and agency costs as determinants of dividend payout ratios. Journal of Financial Research 5, 249-259.
Ruigrok, W., Peck, S.I., Keller, H., 2006. Board characteristics and involvement in strategic decision making: evidence from Swiss companies. Journal of Management Studies 43, 1201-1226.
Saeed, A., Sameer, M., 2017. Impact of board gender diversity on dividend payments: evidence from some emerging economies. International Business Review 26, 1100-1113.
35 Szydlo, M., 2015. Gender equality on the boards of the EU companies: between economic efficiency, fundamental rights and democratic legitimization of economic governance. European Law Journal 21, 97-115.
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.
Van den Berghe, L.A.A., Baelden, T., 2005. The complex relation between director independence and board effectiveness. Corporate Governance: The International Journal of Business in Society 5, 58-83.
Weisbach, M.S., 1988. Outside directors and CEO turnover. Journal of Financial Economics 20, 431-460.
Zwiebel, J., 1996. Dynamic capital structure under managerial entrenchment. The American Economic Review 86, 1197-1215.
36 Appendix A. Variable definitions
Variables Definition Source
Dividend payout
Dividend yield Expresses the dividend per share as a percentage of the share price. DataStream Industry adjusted DY Worldwide industry median adjusted dividend yield, difference between dividend yield
and wordwide industry median.
La Porta et al. (2000) Board composition
Board size The total number of board members at the end of the fiscal year. DataStream
Board independence Percentage of independent board members as reported by the company. DataStream CEO duality Indicator variable which takes the value of 1 if CEO = chairman, 0 otherwise. DataStream
Femaledir Percentage of women on the board of directors. DataStream
Firm specific
Firm size Logarithm of market capitalization (market price- year end * common shares). DataStream Market-to-book Market value of the ordinary (common) equity/ balance sheet value. DataStream RISK Riskiness computed as average return of assets by company. Difference with overall
average. Square root of difference with average roa and sum within company divided by (t-1). t equals 5 years.
DataStream
Leverage Total debt divided by total assets. DataStream
Profitability Net income before extrordinarty items/preferred dividends divided by total asset. DataStream Political governance Reflects perceptions of:
Voice and accountability The extent to which a country's citizens are able to participate in selecting their
government, as well as freedom of expression, freedom of association, and a free media. WorldBank
Political stability and Absence of
Violence/Terrorism
Political Stability and Absence of Violence/Terrorism measures perceptions of the likelihood of political instability and/or politically-motivated violence, including terrorism.
WorldBank
Government effectiveness The quality of public services, the quality of the civil service and the degree of its independence from political pressures, the quality of policy formulation and
implementation, and the credibility of the government's commitment to such policies.
WorldBank
Regulatory quality The ability of the government to formulate and implement sound policies and regulations
that permit and promote private sector development. WorldBank
Rule of law The extent to which agents have confidence in and abide by the rules of society, and in particular the quality of contract enforcement, property rights, the police, and the courts, as well as the likelihood of crime and violence.
WorldBank
Control of corruption The extent to which public power is exercised for private gain, including both petty and
grand forms of corruption, as well as "capture" of the state by elites and private interests. WorldBank Country specific
Civillaw Indicator variable which takes the value of 1 if country is civil law, 0 otherwise.
37
Appendix B. Correlation matrix
Panel A: Full Sample 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
1. Dividend yield 1 2. Board size 0.0637* 1 3. Board independence -0.0840* -0.1625* 1 4. CEO duality -0.1252* 0.0713* 0.1495* 1 5. Femaledir 0.1018* 0.1070* 0.3019* 0.0065 1 6. Firm size 0.0113 0.3108* -0.3635* 0.0631* -0.2049* 1 7. Market-to-book-ratio -0.0004 -0.0221* 0.1279* 0.0361* 0.1250* -0.0128 1 8. RISK -0.1303* -0.1493* 0.0512* -0.0498* -0.0408* -0.2138* -0.0008 1 9. Leverage 0.0368* 0.1299* 0.0414* 0.0173 0.0623* -0.0404* 0.0101 -0.0228* 1 10. Profitability 0.1672* 0.0502* 0.0235* 0.0401* 0.0493* 0.1868* 0.2182* -0.3238* -0.2023* 1 11. Civillaw 0.0981* 0.1801* -0.5871* -0.1234* -0.1333* 0.5130* -0.1203* -0.1120* -0.0086 -0.002 1
12. Voice and acc. 0.0057 -0.1478* 0.2100* 0.0193* 0.1770* -0.3301* 0.0032 0.0559* -0.0384* -0.0496* -0.2124* 1
13. Political stability 0.0659* -0.1732* -0.0632* -0.0303* -0.0243* -0.0985* -0.0979* 0.0424* -0.0880* -0.0816* 0.0022 0.5733* 1 14. Government eff. 0.0081 -0.1805* 0.1223* 0.0122 0.0689* -0.2321* -0.0332* 0.0308* -0.0770* -0.0350* -0.2646* 0.6394* 0.8423* 1 15. Regulatory quality 0.0571* -0.2097* 0.1954* -0.0466* 0.1318* -0.4218* -0.0072 0.0830* -0.0616* -0.0495* -0.4301* 0.6716* 0.7348* 0.9018* 1 16. Rule of law -0.0107 -0.2110* 0.2625* 0.0398* 0.1550* -0.3779* -0.0001 0.0555* -0.0508* -0.0469* -0.4233* 0.7813* 0.7532* 0.9372* 0.9231* 1 17. Control of corrupt. 0.0571* -0.1891* 0.0996* -0.0505* 0.1066* -0.3296* -0.0367* 0.0537* -0.0766* -0.0542* -0.2654* 0.7160* 0.8328* 0.9368* 0.8920* 0.9304* 1 18. ln(GDP) -0.0853* -0.1368* 0.2665* 0.0993* 0.1238* -0.3155* -0.009 0.0519* -0.0094 -0.0712* -0.3047* 0.6457* 0.6573* 0.8028* 0.7511* 0.8387* 0.8126* 1 * Denotes statistical significance at the 1% level
