Does gender diversity in the board of directors influence
the firms’ dividend policy? A cultural perspective of the
influence of female board members
H.J.C.S. Roos S2807963
MSc. International Financial Management 6 - June - 2017
University of Groningen Faculty of Economics and Business
P.O. Box 800 9700 AV Groningen
Abstract
Existing literature found a relationship between gender diversity and a firm’s dividend policy but these findings are based on a single country analysis. This research contributes by extending the existing research empirically through the incorporation of a multinational setting. It is often found that females are more risk averse then males. The same applies to culture which too can be differentiated between levels of risk aversion. Some cultures might empower or oppositely weaken the influence of females on the board of directors over the dividend policy. Therefore, this research expands by hypothesising that culture moderates this relationship. This research investigates listed firms, multiple industries, and countries. T-test, univariate and multivariate regressions are used to test the main relation and whether culture does have a noticeable impact. The t-test resulted in a significant higher mean in dividend yield in favour of gender diverse board of directors. The regression resulted in a positive and significant relationship between gender diversity and the firm’s dividend policy. Furthermore, no evidence is found to support the moderating effect of culture.
Field Keywords: Gender Diversity, Agency Theory, Culture, Risk, Dividend Policy. 1st Supervisor: Silviu Ursu Phd.
1. Introduction
The board of directors of firms no longer consist of old, grey men wearing oversize suits and smoking cigars all day. Instead they have become younger and more diverse, ready to implement new ways of doing business and acting proactively. A subset of this changing diversity is a trend for females to become increasingly representative within the firm and its board of directors (Lee et al., 2015). It is agreed that females show different behaviour within the firm compared to their male counterparts in, for example, risk aversion (Desvaux and Devillard, 2008; Croson and Gneezy, 2009). Therefore, the question of how this difference translates into action is not as far-fetched as it may seem.
One of the actions of the board of directors is the determination and approval of the dividend policy. Dividend payments are important for investors because it is part of their return on investment. Mergers and acquisitions, capital budgeting, and capital structure also depend on the frequency and amount of dividends (Allen and Michaely, 1995). The agency theory argues that dividend payments lead to the management selecting better investments and using other allocations of capital (Jensen and Meckling, 1976). Furthermore, these authors argue that due to cash outflow it becomes harder for the management to use capital for their personal interest. The remaining capital, because of its reduced size, will be easier to monitor. This is done to ensure the goals of the management are in line with its shareholders (Porta et al., 2000). Logically it would seem that higher dividend payments put more pressure on the management to align their goals. Since females are on average found to be more risk averse it could well be that females prefer higher dividend payments than their male counterparts to ensure the differences between goals is minimized.
Especially with companies growing internationally they are bound to attract shareholders from various cultural backgrounds. These culturally diversified investors are left in the dark as to what the impact of the differences in nationality has on the influence of the female presence within the board of directors. Females might generally be more risk averse than males but their power and practical influence in the firm might be lower in some countries and higher in others. Therefore, this research intends to investigate the moderating effect of culture on the relationship of gender diversity and dividend policy.
whether they are independent, institutional, the percentage of shares held, and basic proportion of females. They find that gender diversity does in fact have a positive influence on the dividend policy of firms. A second paper by McGuinness et al. (2015) focusses on China. The authors investigate the CEO and the board members. Only a little difference between gender and dividend policy is found. They even challenge the notion that there is a difference in risk aversion between male and female. Their perspective is focused on cross cultural theory stating that if females are more risk diverse they should prefer to keep capital in the firm. Byoun et al. (2016) investigate the relationship without making a distinction between countries. These authors also found a positive relationship. To summarize, the results are inconclusive. Furthermore, the literature is mostly limited to a single country or does not make a distinction between countries.
Based on previous research, or rather the lack of, this research investigates if more females on the board of directors leads to a higher dividend yield and more specifically if in some countries the influence of females is stronger than others. To operationalize this, two different hypotheses are tested. The first one is: Gender Diversity is positively related with
the firm’s dividend yield. And the second is: Culture moderates the influence gender diverse boards can have on the dividend policy of the firm. An independent samples T-test finds a
significant difference in the mean between firms that have a board of directors with no female presence and a board of directors with minimal 1 female. This mean difference is a 37,45% increase in dividend yield. Even when separating the sample into separate countries 11 of them are significant. The regression analysis finds a positive and significant coefficient when regressing gender diversity on dividend yield for both circumstances with and without controls. It can be concluded that gender diversity does have an influence on dividend yield. No moderating effects are found in the regressions when the interaction term is added. Therefore, it can be concluded that cultural differences between countries does not influence the risk aversion characteristics of females within the board of directors.
shares can even exert significant influence in order push for, or alter the change to, gender diversification.
After this introduction, the second chapter contains a literature review. In this chapter literature regarding this subject is reviewed in order to derive the two hypotheses. Having derived the two hypotheses from the literature the methodology chapter provides insight into how this research is performed. Sample, timeframe, variables, and methods are all explained. After reading this chapter it is clear what is being researched and how this is executed. The results chapter highlights and interprets the results. In the following chapter these findings and interpretations are used to draw conclusions regarding the hypotheses. This research has its limitations and delimitations. The discussions chapter expresses these and reflects on the conclusions to highlight the contributions this research has made in both a theoretical and practical way.
2. Literature
The first section elaborates on the dividend policy of firms. It includes multiple theories as to whether dividend is or is not considered a good thing. The second section of this chapter explains the agency theory. This is the main theory which forms the basis for drawing the hypotheses. Where dividend is assumed to reduce agency costs, this theory enables drawing hypothesis regarding the amount of dividend that is being paid. This section is followed up by gender diversity to highlight differences between males and females. The last section explains cultural aspects which becomes important when studying multiple countries and the possible differences between them.
2.1 Dividend Policy
Dividend is the amount of capital that the firm decides to transfer to its shareholders. This transfer of capital usually comes from the firms’ earnings. Because of that, the decision is mostly how much of the profits should be retained, in order to gain or maintain future cash flows, and how much can be paid to its shareholders. If the profits are insufficient it can be that firms use their reserves.
according to the performance of the firm. Secondly, by using a stable dividend policy the firms try to pay its shareholders a quantitative equal amount each time the dividend is payed. Using this method, it provides its shareholders some form of stability. Furthermore, the firm has certainty which part of the earnings, assuming there is a positive result, is available for investments. Thirdly, by using the residual dividend policy where firms use the earnings for investments and reduction of debt if desired. The remaining earnings will be used to pays its shareholders.
The level of dividend is often determined by higher management but this decision is influenced by the advice of the board of directors who can therefore influence the determination of the amount and the frequency that is paid out to its shareholders. Minority shareholders have in practice little influence and must deal with the decision made for them. Jensen (2001) states that based on financial theory, the firm’s objective is maximising shareholder wealth and that whether cash dividend is value creating or not has been a long debate but remains inconclusive.
The important assumption in this theory is that higher dividends results in less capital within the firm which leads to unstable future cash flows.
2.2 Agency Costs
Agency costs are costs that are made to ensure the agents behave according to the wishes of the principle. The source of these costs are the differences of objectives between the two parties. Agents have a habit of acting to increase or secure their own benefits which might not be desired by the principal. An example is that the managers make investments to achieve short term profit versus long term results which often is desired by the shareholders. One method of aligning these interests is paying dividend to its shareholders (Rozeff 1982; Easterbrook, 1984; Jurkus et al. 2011). Paying out earnings instead of retaining them emphasises the need by agents to invest in riskier investments which is desired by the principal. Besides, if capital is paid out to its shareholders’ other necessary capital must be drawn in from external providers to maintain or grow the level of investments. By using external capital, the providers of this capital will form an additional way of monitoring agents. The decision to draw in capital from external sources should consider the transaction cost that accompany transactions with external providers of capital.
Literature has found relationships between gender diversity and agency costs. Webb (2004) argues that a higher diversity increases the board ability to understand and manage stakeholders by increasing the amount of opinions and perspectives. Furthermore, this author suggests that this is needed to create more diverse outcomes and policies. Adams and Ferreira (2009) state that females are more likely to implement control mechanisms such as dividend payments. Furthermore, it is found that females are more effective in monitoring agents (Adams and Ferreira, 2004), and that females increase the quality and quantity of information (Jurkus et al., 2011).
2.3 Gender Diversity
Gender diversity became increasingly important for organizations up to a point where legislation tries to set norms and standards. From an ethical perspective gender diversity is about what is considered fair and just. It is about giving females equal chances and equal pay. Corporate governance literature tries to determine what is considered good for firms to do and provides methods for firms to achieve this balance. Especially in the higher management and director function this distribution is still skewed toward males. But justice or fairness should not be the key driver for this balanced representation.
From an economical perspective, gender diversity is claimed to have other benefits as well. Most literature of gender diversity reflects on firm’s performance. The results often contradict (Carter et al., 2003; Gregory-Smith et al., 2014). From a negative perspective, Earley and Mosakowski (2000) state that diversity negatively influences performance of the board of directors. They argue that homogeneity leads to more communication and increases the overlap of experiences and opinions. Furthermore, Tajfel and Turner (1979) found that diversity increases the propensity for emotional conflicts. Positive effects have been found as well. Grimm et al. (2005) argues that diversity is good for the image of the firm and that this in turn leads to increased competitiveness. Carter et al. (2003) found that increased gender diversity positively influences the market valuation of the firm measured in Tobin’s Q. Furthermore, insolvency risk is found to be negatively related to gender diverse boards of directors (Wilson and Altanlar, 2009) and that the board of directors with a lower gender diversity are less likely to engage in merger and acquisition activities (Levi et al., 2014). All these findings on both the psychological and financial dimension clearly indicate that gender diversity does in fact play a role within firms.
Research finds that females are less likely to prefer risky activities (Barber and Odean, 2001; Byrnes et al., 1999). Besides their preference, other studies find that female CEO’s influence leads to better liquidity and less total risk (Faccio et al., 2016). Loukil and Yousfi (2016) finds that females in a board of directors leads to risk avoidance behaviour and that more females are related with a lower stock volatility (Adams and Ferreira, 2004). Martinez and Bel-Oms (2015) state that higher gender diversity within the board of directors will lower risk taking observed by reducing agency costs and resulting in higher dividends. Based on this research it is hypothesised that females take less risk and pay higher dividends to reduce agency costs.
Hypothesis 1: Gender Diversity positively influences the firm’s dividend yield.
2.4 Culture
national, regional and corporate. The lower the level of comparison the higher the accuracy. For example; the culture within a single country can vary greatly across regions. The existence of different cultures is not directly troublesome. When cultures collide or overlap, then issues can arise. For firms operating in multiple countries means that they need to have an increased cultural awareness to cope with the differences. Questions they need to ask themselves are: are these differences troublesome or can they be a source of synergy? Besides knowing what culture entails it is important to know how this is operationalized. This is important due to its multifaceted and complex nature. The most popular models originate from Hofstede (1980), Schwartz (1999), and House et al. (2004). All these models consist of multiple dimensions and are quite similarly constructed. When comparing two countries some dimensions might be similar whereas other are might not. Each model has its own advantages. In the case of Hofstede’s model the survey is held within a single firm. By doing so this eliminates the impact of differences between corporate cultures. Another advantage is the number of countries included. Compared to other models Hofstede includes more countries and for which there are less missing values. By using Hofstede’s cultural forces all but one country in the sample of this research have missing cultural values. Lastly, Hofstede’s model is used by many scholars and academics. Being the most used cultural model going with the flow increases comparability.
Geert Hofstede (1980) used an existing large scale survey to create his model. This firm-level survey was held within IBM. The survey included 70 countries of which he used 40 with the largest respondents. The fact that he used an existing survey aimed and measuring something else in order to generate his findings has been a criticism. Nevertheless, his model is expanded to include more countries and more dimensions. The first model included 4 different dimensions. Currently there are 6. These are: (1) Power distance; (2) Individualism; (3) Masculinity; (4) Uncertainty Avoidance; (5) Long Term Orientation; and (6) Indulgence versus Restraint. To go more in depth into Hofstede’s models each dimension’s definition is shown in Table one. All these definitions are direct quotes from the current up-to-date model.
influence the composition of the country and thus impact the culture of a nation. Lastly, this model reflects on the country as a single entity and not on the individuals within the country. As an example, a country can have two large subgroups which by themselves would score completely opposite results according the Hofstede’s dimensions. Country wise, the scores would be the average of the two and therefore might not be a good fit.
Table 1: Hofstede´s Cultural forces
This table shows the definitions of the six cultural dimensions created by Hofstede (1980) and used in this research. Dimension Definition
Power distance (PDI)
The degree to which the less powerful members of a society accept and expect that power is distributed unequally. The fundamental issue here is how a society handles inequalities among people. People in societies exhibiting a large degree of Power Distance accept a hierarchical order in which everybody has a place and which needs no further justification. In societies with low Power Distance, people strive to equalise the distribution of power and demand justification for inequalities of power.
Individualism (IDV)
Preference for a loosely-knit social framework in which individuals are expected to take care of only themselves and their immediate families. Its opposite, collectivism, represents a preference for a tightly-knit framework in society in which individuals can expect their relatives or members of a particular in-group to look after them in exchange for unquestioning loyalty. A society's position on this dimension is reflected in whether people’s self-image is defined in terms of “I” or “we.”
Masculinity (MAS)
The Masculinity side of this dimension represents a preference in society for achievement, heroism, assertiveness and material rewards for success. Society at large is more competitive. Its opposite, femininity, stands for a preference for cooperation, modesty, caring for the weak and quality of life. Society at large is more consensus-oriented. In the business context Masculinity versus Femininity is sometimes also related to as "tough versus tender" cultures.
Uncertainty Avoidance (UAV)
The Uncertainty Avoidance dimension expresses the degree to which the members of a society feel uncomfortable with uncertainty and ambiguity. The fundamental issue here is how a society deals with the fact that the future can never be known: should we try to control the future or just let it happen? Countries exhibiting strong UAI maintain rigid codes of belief and behaviour and are intolerant of unorthodox behaviour and ideas. Weak UAI societies maintain a more relaxed attitude in which practice counts more than principles.
Long Term Orientation (LTO)
Societies who score low on this dimension, for example, prefer to maintain time-honoured traditions and norms while viewing societal change with suspicion. Those with a culture which scores high, on the other hand, take a more pragmatic approach: they encourage thrift and efforts in modern education as a way to prepare for the future. In the business context this dimension is related to as "(short term) normative versus (long term) pragmatic" (PRA)
Indulgence versus Restraint (IVR)
Indulgence stands for a society that allows relatively free gratification of basic and natural human drives related to enjoying life and having fun. Restraint stands for a society that suppresses gratification of needs and regulates it by means of strict social norms.
Definitions of the sex dimensions of Hofstede’s Cultural forces (1980) Definitions are direct quotes. Source: https://geert-hofstede.com/national-culture.html
other hand it finds that higher levels are more in line with a performance oriented way of goal setting. Besides findings based on individual’s behaviour Gorodnichenko and Gerard (2011) claim this dimension also influences countries from a macroeconomic perspective and that that higher individualism relates to higher confidence and acceptance of risk. Kivetz and Simonsos (2002) found that indulgence corresponds with higher consumption and the willingness to pay for this. Laran (2010) finds that lower level of indulgence is related to a higher long term orientation (Kramer et al., 2011). To continue long term orientation Giner-Sorolla (2001) finds that the achievement of long term goals is related to pride. To complete the circle Fredrickson (2001) and Louro et al. (2007) suggest pride leads to long-term decisions in favour of forms of indulgence.
Martinez and Bel-Oms (2015) found that higher cultural diversity will lead to higher pay out of dividends in order to reduce agency costs. Since culture forms the boundaries and perception of the agency problem (Bae et al., 2012) it is unclear whether this holds within multiple cultures and listed multinationals. Existing research found relationships between culture and; dividend policy (Bae et al., 2012; Shao et al., 2010); capital structure (Chui et al., 2002); equity valuation and managerial decisions (Chuluun et al., 2014; Chang et al., 2012); and is weakened by the strength of the firm’s corporate governance levels (Li et al., 2013). To summarise, culture has the power to influence an individual’s behaviour and perception. Besides that, female representation and authority is respected in certain culture more than others. Therefore, we hypothesise that culture has the power to moderate the propensity of gender diversity to pay dividend.
Hypothesis 2: A countries culture moderates the susceptibility of the positive effect of
gender diversity on dividend yield
3. Methodology
This chapter will firstly explain the sample used in this research. This section includes which countries, what timeframe, and which limitations or additional requirements which are applied to this sample. Secondly it includes the different variables that are used. Thirdly are the descriptive statistics. Fourthly a section regarding the analyses and robustness checks which have been performed on the data.
3.1 Sample
countries that have been selected are the combined countries of the EU-15 and the G20 and shown in Table two. This resulted in including: Argentina, Australia, Austria, Belgium, Brazil, Canada, China, Denmark, Finland, France, Germany, Greece, India, Indonesia, Ireland, Italy, Luxembourg, Japan, Mexico, Netherlands, Portugal, Russia, South Africa, Spain, Sweden, Turkey, United Kingdom, United Stated of America. This selection is made by including countries that reflect most global trade. The G20 by itself reflects 90% of worldwide GDP and 80% of the world trade. Furthermore, 66% of the world population is living in these countries. With the additional countries from the EU-15 this sample reflects most global trade and inhabitants. Adjustments to the list of countries will be made to accommodate the availability of other data. For example, the countries available in Hofstede’s Dimensions. Of the combined countries of the EU 15 and the G20 only Saudi Arabia is left out due to missing cultural data. Even though these numbers could be econometrically approximated using values of surrounding countries, leaving out this country would not have a major influence on the sample countries. Furthermore, leaving out this country would be better for the accuracy of results. South Korea is removed from the sample due to the lack of financial data on both listed firms and their board members. The firm distribution is unequal distributed. The USA, Japan, India and China combined account for more than half of the total firm responses. On the contrary, the countries with the lowest amount of firm responses only account for a fraction of the total responses. Because of the large sample size, it is still possible to compare results between countries.
tendency to differ in reporting standards and regulations within their own country and between other countries. Leaving out this industry in the sample would remove potential statistical errors (Fama & French, 2002).
Table 2: Country Distribution
Table 2 lists all countries included in the initial sample. For each country, the number of firms included in the sample are shown as a number and as percentage of the total number
of firms. Name # % Argentina 66 0,36 Australia 1569 8,56 Austria 49 0,27 Belgium 91 0,50 Brazil 238 1,30 Canada 1348 7,35 China 2638 14,39 Denmark 103 0,56 Finland 121 0,66 France 601 3,28 Germany 485 2,64 United Kingdom 1149 6,27 Greece 172 0,94 India 2167 11,82 Indonesia 377 2,06 Ireland 71 0,39 Italy 203 1,11 Japan 2441 13,31 Luxembourg 44 0,24 Mexico 80 0,44 Netherlands 113 0,62 Portugal 45 0,25 Russian Federation 440 2,40 South Africa 186 1,01 Spain 129 0,70 Sweden 435 2,37 Turkey 222 1,21
United States of America 2754 15,02
Total 18337 100,00
# is the number of firms, % is the percentage of the total number of firms.
3.2 Independent variable
Table 3: Gender Diversity
This table presents an overview of the gender diversity of the firms included in the sample. These firms are separated by country. A distinction is made by firms who have a board of directors consisting of only males “0”
and firms that have a board of directors which includes at least 1 female “+1”
Name # 0 1+ % Argentina 66 40 26 39 Australia 1569 1052 517 33 Austria 49 15 34 69 Belgium 91 22 69 76 Brazil 238 154 84 35 Canada 1348 1109 239 18 China 2638 1812 826 31 Denmark 103 36 67 65 Finland 121 13 108 89 France 601 92 509 85 Germany 485 223 262 54 United Kingdom 1149 559 590 51 Greece 172 100 72 42 India 2167 286 1881 87 Indonesia 377 123 254 67 Ireland 71 29 42 59 Italy 203 40 163 80 Japan 2441 2039 402 16 Luxembourg 44 26 18 41 Mexico 80 56 24 30 Netherlands 113 45 68 60 Portugal 45 8 37 82 Russian Federation 440 123 317 72 South Africa 186 36 150 81 Spain 129 38 91 71 Sweden 435 101 334 77 Turkey 222 82 140 63
United States of America 2754 1061 1693 61
Total 18337 9320 9017 49
# is the number of firms in each country, “0” corresponds with boards with no gender diversity, “+1” are the number of firms that have a minimum of 1 female in the board of directors, % is the percentage of firms that are diverse.
3.3 Dependent variable
Dividend Policy - For the measurement of the variable dividend policy this research follows a paper of Hussainey et al. (2011). These authors used both dividend yield and pay-out ratio to investigate the dividend policy of UK firms. They define dividend yield as the dividend per share as a percentage of the share price and pay-out ratio as the ratio of dividends per share to earnings per share (Hussainey et al., 2011). This data is retrieved from DataStream. Following Farinha (2003) the variable dividend yield only contains values ranging from zero to one. Zero means that the firm has not paid any dividend. A one means that a dividend is paid that equals the share price at that time. This means that the shareholders have a 100% return on the share.
these combined assist the elimination of irregularities. Besides, using more proxies can overcome differences that exist between different industries or accounting practices.
3.4 Moderators
National Culture – A common definition of a culture originates from Hofstede (1980) and Schwartz (1999) that it is a set of norms, beliefs, shared values, and behaviour that guide human beings. This variable will be operationalised by using Hofstede’s (1980, 1991) cultural dimensions. This research follows Bae et al. (2012) for using the dimensions of uncertainty avoidance, masculinity, and long term orientation. Furthermore, this research also includes Hofstede’s dimension of individualism which is frequently researched (Fauver and McDonald, 2015). Uncertainty avoidance is whether people accept uncertainty or feel threatened by not knowing or not being able to predict. Masculinity classifies based on what extent a culture is tough or tender. Long term orientation is whether people would rather live in the present or would rather look to the future. An example is the amount of money people save. Individualism is about what extent people are tied to groups and look after each other or whether people act and look out for themselves. Table four lists all the countries that are included in the sample with the cultural scores for al dimensions.
There have been many pros and cons in using Hofstede’s model as a method of measurement. Drawbacks of using this measure is that this framework is compiled based on an internal study within IBM which was not designed to test for cultural norms. This study was executed through questionnaires which creates limitations on its own. Perception of context or socially accepted responses are just to mention a few. Furthermore, this framework can be only used to compare on a country level. It does not have power to explain individuals or subgroups within a country. Last drawback is related to the timeliness of the framework. It is unknown to what extent national cultures changes. Therefore, it is difficult to tell if this framework is outdated or not. Hofstede’s dimensions are expanded to include 111 countries and are, because of that scale, one of the largest, if not the largest cultural model currently available. Due to its scale the main advantage of this model is that it is widely accepted and commonly used in research which facilitates comparability and accessibility.
Table 4: Hofstede’s Cultural Forces
This table shows the scores of Hofstede’s cultural forces separated by country
Name PDI IDV MAS UAV LTO IVR
Argentina 49 46 56 86 20 62 Australia 38 90 61 51 21 71 Austria 11 55 79 70 60 63 Belgium 65 75 54 94 82 57 Brazil 69 38 49 76 44 59 Canada 39 80 52 48 46 68 China 80 20 66 30 87 24 Denmark 18 74 16 23 35 70 Finland 33 63 26 59 38 57 France 68 71 43 86 63 48 Germany 35 67 66 65 83 40 United Kingdom 35 89 66 35 51 69 Greece 60 35 57 112 45 50 India 77 48 56 40 51 26 Indonesia 78 14 46 48 62 38 Ireland 28 70 68 35 24 65 Italy 50 76 70 75 61 30 Japan 54 46 95 92 88 42 Luxembourg 40 60 50 70 64 56 Mexico 81 30 69 82 24 97 Netherlands 38 80 14 53 67 68 Portugal 63 27 31 104 28 33 Russian Federation 93 39 36 95 91 20 South Africa 49 65 63 49 34 63 Spain 57 51 42 86 48 44 Sweden 31 71 5 29 53 78 Turkey 66 37 45 85 46 49
United States of America 40 91 62 46 26 68
PDI = Power Distance, IDV = Individualism, MAS = Masculinity, UAV = Uncertainty Avoidance, LTO = Long Term Orientation, IVR = Indulgence versus Restraint. Each cultural dimension can range from 0 to 100.
3.5 Control Variables
This research includes both firm, and country specific control variables. The firm-specific control variables included are total Risk, firm size, debt-equity ratio, sales growth and the firms return on assets (ROA). This is in line with comparable literature (Bruer et al., 2014; Chae et al., 2009). The data for these variables is available through DataStream. The country-specific control variables in this research includes taxation, governance and the market sentiment. (Bruer, Rieger & Soypak, 2014; Bae et al., 2012)
taking the natural log of the firm’s total assets. It is expected that increased size will positively influence dividend pay-out due to increase access to capital meaning that less earnings should be retained for future investments (Bae et al., 2012). Bruer et al. (2014) argues that this relationship should be similar but used different argumentation. These authors argue that increased size makes monitoring harder resulting in shareholders demanding higher dividend as reward for higher risk and information asymmetry (Fama and French, 2001; Baker and Weigand, 2015). Bruer et al. (2014) argue that the Debt-Equity Ratio most likely negatively influences the amount of dividend due to constraints to increase corporate governance by debt holders. This variable will be measure by calculating the ratio between the book value of debt and equity. Increased firm growth rate is likely to negatively influence the firm’s dividend pay-out because these firms require more capital to sustain this growth rate (Fama and French, 2001). The growth rate is measured following Brockman and Unlu (2009) as the quotient between sales in the current and previous year. Increased
profitability is expected to positively influence dividend payments which is in line with
Bruer et al. (2014). La Porta et al. (2000) argues that this is because of profitable firms wanting to showcase performance and quality. Profitability is measured as the firm’s Return on assets This measure is frequently used in the study of this literature. The Return on Assets is defined as the earnings before interest and tax dived by total asset and measures. It indicates the riskiness of the investments that have taken place. By taking its standard deviation it is possible to measure its volatility over a certain amount of time. The underlying argument that forms the foundation on the usability of these outcome oriented measures is that a higher risk is rewarded with potentially higher returns. DataStream contains the necessary data needed to calculate the proxies of this independent variable.
2.5 and have a normal distribution. The last country specific control variable is the market status. When the market is expected to grow, it becomes easier to gain access to capital if needed and are more likely to signal profitable future results. Because of this increasing access to capital, firms are less reliant on retained earnings and therefore are more likely to pay relatively higher amounts of dividend. For the measurement of this variable this research follows Bae et al. (2012). Here the market status is measured by dividing a countries stock market capitalization by the country’s GDP.
3.6 Descriptive Statistics
The descriptive statistics are shown in Table five. This table includes the number of observations, the minimum, the maximum, mean, and finally the standard deviation. Using this information, the data can be reviewed for abnormalities and potential signals for outliers. The variable DV_DividendYield is limited to included values ranging from 0 to 100. This is in line with Farinha (2003) and done so because of lack in economic and statistical significance. Furthermore, the control variable ROA has a relative high maximum. But by checking the data it seems that this value is valid. The argument behind this is that we only exclude financial firms. It is viable that certain industries require lower amounts of assets. Lastly the control variable Growth has a maximum of 395% compared to the mean of 7%. This may be perceived as an outlier but these growth rates are not unheard of, especially new firms can experience these high levels of growth. The control variable Growth Rate is the percentage difference in sales compared to last year. Comparing growth rates only on a single year has its limitations. For example, the firm could experience a growth or decline lasting a single year due to unexpected circumstances. Using multiple year smooths outs the growth rate.
Table 5: Descriptive Statistics
This table provides descriptive statistics off the variables used in this research.
# Minimum Maximum Mean Std. Deviation
IV_GD 18337 ,00 1,00 ,11 ,144 MOD_PDI 18337 11 93 55,34 18,834 MOD_INV 18337 14 91 60,26 25,954 MOD_MAS 18337 5 95 61,55 17,859 MOD_UAV 18337 23 112 54,50 22,751 MOD_LTO 18337 20 91 56,34 24,726 MOD_IVR 18337 20 97 49,33 19,409 DV_DividendYield 17354 ,00 94,90 1,28 3,167 DV_PORatioDS 16809 ,00 150,00 ,27 1,937 DV_POratioWC 16871 ,00 76,27 ,28 1,256 CNT_TotalRiskSTDVri 17954 ,00 ,42 ,03 ,0252 CNT_ROA 16522 ,00 17,67 ,07 ,222 CNT_Size 17410 ,00 11,41 6,02 1,535 CNT_DEratio 15645 -4,99 5,97 ,56 1,036 CNT_Growth 16968 -121,00 394,89 7,11 44,165 CNT_TaxWH 18337 ,00 ,30 ,22 ,0798 CNT_TaxCOR 18337 ,13 ,35 ,25 ,063 CNT_TaxWHlog 18337 -1,00 ,00 -,68 ,193 CNT_TaxCORlog 18337 -,90 -,46 -,61 ,116 CNT_Governance 18337 -1,00 1,90 1,00 ,791 CNT_MarketStatus 18337 9,62 265,85 85,68 39,447
# is the number of firms for which the variable has values. IV is the independent variable, MOD are moderators, DV are dependent variables. CNT are control variables. IV_GD is a percentage where 1,00 equals 100%, the depended variables are percentages, 100,00 equals 100%. This difference in scale does not influence the regression.
3.7 Research Method
A univariate regression (1) is held to test for the strength in a linear regression. This is expanded to a multivariate regression (2) to include the control variables. Thirdly the control value by themselves are tested (3). By using a logistic regression, the purpose is to see whether there is a linear relationship and if so what the strength is of this relationship. (1) 𝐷𝑖𝑣𝑖𝑑𝑒𝑛𝑑𝑌𝑖𝑒𝑙𝑑 = 𝐺𝑒𝑛𝑑𝑒𝑟 𝐷𝑖𝑣𝑒𝑟𝑠𝑖𝑡𝑦 (2) 𝐷𝑖𝑣𝑖𝑑𝑒𝑛𝑑𝑌𝑖𝑒𝑙𝑑 = 𝐺𝑒𝑛𝑑𝑒𝑟 𝐷𝑖𝑣𝑒𝑟𝑠𝑖𝑡𝑦 + 𝑅𝑒𝑡𝑢𝑟𝑛 𝑜𝑛 𝐴𝑠𝑠𝑒𝑡𝑠 + 𝑆𝑖𝑧𝑒 + 𝐷𝑒𝑏𝑡 𝑡𝑜 𝐸𝑞𝑢𝑖𝑡𝑦 𝑟𝑎𝑡𝑖𝑜 + 𝐺𝑟𝑜𝑤𝑡ℎ 𝑟𝑎𝑡𝑒 + 𝑊𝑖𝑡ℎℎ𝑜𝑙𝑑𝑖𝑛𝑔 𝑇𝑎𝑥 + 𝐶𝑜𝑟𝑝𝑜𝑟𝑎𝑡𝑒 𝑇𝑎𝑥 + 𝑀𝑎𝑟𝑘𝑒𝑡 𝑆𝑡𝑎𝑡𝑢𝑠 (3) 𝐷𝑖𝑣𝑖𝑑𝑒𝑛𝑑𝑌𝑖𝑒𝑙𝑑 = 𝑅𝑒𝑡𝑢𝑟𝑛 𝑜𝑛 𝐴𝑠𝑠𝑒𝑡𝑠 + 𝑆𝑖𝑧𝑒 + 𝐷𝑒𝑏𝑡 𝑡𝑜 𝐸𝑞𝑢𝑖𝑡𝑦 𝑟𝑎𝑡𝑖𝑜 + 𝐺𝑟𝑜𝑤𝑡ℎ 𝑟𝑎𝑡𝑒 + 𝑊𝑖𝑡ℎℎ𝑜𝑙𝑑𝑖𝑛𝑔 𝑇𝑎𝑥 + 𝐶𝑜𝑟𝑝𝑜𝑟𝑎𝑡𝑒 𝑇𝑎𝑥 + 𝑀𝑎𝑟𝑘𝑒𝑡 𝑆𝑡𝑎𝑡𝑢𝑠
To test for a moderating effects of culture various multivariate regressions is held. First model (1) is expanded to include the hypothesised cultural variables Masculinity and Uncertainty avoidance. This results in model (4). Model (5) builds on the previous model by including the control variables. The last model includes the interaction terms to test for a moderating effect (6) (4) 𝐷𝑖𝑣𝑖𝑑𝑒𝑛𝑑𝑌𝑖𝑒𝑙𝑑 = 𝐺𝑒𝑛𝑑𝑒𝑟 𝐷𝑖𝑣𝑒𝑟𝑠𝑖𝑡𝑦 + 𝑀𝑎𝑠𝑐𝑢𝑙𝑖𝑛𝑖𝑡𝑦 + 𝑈𝑛𝑐𝑒𝑟𝑡𝑎𝑖𝑛𝑡𝑦 𝐴𝑣𝑜𝑖𝑑𝑎𝑛𝑐𝑒 (5) 𝐷𝑖𝑣𝑖𝑑𝑒𝑛𝑑𝑌𝑖𝑒𝑙𝑑 = 𝐺𝑒𝑛𝑑𝑒𝑟 𝐷𝑖𝑣𝑒𝑟𝑠𝑖𝑡𝑦 + 𝑀𝑎𝑠𝑐𝑢𝑙𝑖𝑛𝑖𝑡𝑦 + 𝑈𝑛𝑐𝑒𝑟𝑡𝑎𝑖𝑛𝑡𝑦 𝐴𝑣𝑜𝑖𝑑𝑎𝑛𝑐𝑒 + 𝑅𝑒𝑡𝑢𝑟𝑛 𝑜𝑛 𝐴𝑠𝑠𝑒𝑡𝑠 + 𝑆𝑖𝑧𝑒 + 𝐷𝑒𝑏𝑡 𝑡𝑜 𝐸𝑞𝑢𝑖𝑡𝑦 𝑟𝑎𝑡𝑖𝑜 + 𝐺𝑟𝑜𝑤𝑡ℎ 𝑟𝑎𝑡𝑒 + 𝑊𝑖𝑡ℎℎ𝑜𝑙𝑑𝑖𝑛𝑔 𝑇𝑎𝑥 + 𝐶𝑜𝑟𝑝𝑜𝑟𝑎𝑡𝑒 𝑇𝑎𝑥 + 𝑀𝑎𝑟𝑘𝑒𝑡 𝑆𝑡𝑎𝑡𝑢𝑠 (6) 𝐷𝑖𝑣𝑖𝑑𝑒𝑛𝑑𝑌𝑖𝑒𝑙𝑑 = 𝐺𝑒𝑛𝑑𝑒𝑟 𝐷𝑖𝑣𝑒𝑟𝑠𝑖𝑡𝑦 + 𝑀𝑎𝑠𝑐𝑢𝑙𝑖𝑛𝑖𝑡𝑦 + 𝑈𝑛𝑐𝑒𝑟𝑡𝑎𝑖𝑛𝑡𝑦 𝐴𝑣𝑜𝑖𝑑𝑎𝑛𝑐𝑒 + 𝑀𝑎𝑠𝑐𝑢𝑙𝑖𝑛𝑖𝑡𝑦 𝑥 𝐺𝑒𝑛𝑑𝑒𝑟 𝐷𝑖𝑣𝑒𝑟𝑠𝑖𝑡𝑦 + 𝑈𝑛𝑐𝑒𝑟𝑡𝑎𝑖𝑛𝑡𝑦 𝑥 𝐺𝑒𝑛𝑑𝑒𝑟 𝐷𝑖𝑣𝑒𝑟𝑠𝑖𝑡𝑦
4. Results
In the correlation matrix, it can be found that multiple cultural dimensions have a high correlation and should not be entered in a model at the same time. Even though only variables with a correlation above .8 ought to be dealt with it should still be checked whether other variables with lower levels of correlation show signs of multicollinearity. To achieve this the Variance Inflation factors are reflected upon. Appendix 2 shows the various Variance Inflation Factors accompanying the different variables. Different authors prescribe different levels from which multicollinearity issues are present in the data. These value range from five till ten. The Variance Inflation Factors for this research range between one and two. Therefore, no multicollinearity issues are present in this sample.
To analyse if the mean of the dividend yield is different between firms that have females in the board of directors and firms that do not have females in the board of directors an independent samples T-test has been performed. Besides including the means of the total sample, Table six also includes the means separated by all the individual countries in the sample. Regarding the total sample the independent samples T-test is significant, t(17205) = -10,774, p = 0,000 thus significant at 10%, 5%, and 1% levels. The mean dividend yield of firms with females in their board of directors (m=2,007, SD=3,217) does statistically differ from the mean dividend yield of firm with no females in their board of directors (m=1,458, SD=2,992). This is a mean increase of 37,45%.
Table 6: Independent Samples Test
This test compares the mean of two groups. The first group are firms with no females in the board of directors and the second group does have a minimum of one female in the board of directors. The results are presented for each country
separately as well as a single group.
Mean SD
# 0 Gender Diversity 0< Gender Diversity 0 Gender Diversity 0< Gender Diversity Mean ∆% t P-Value Argentina 61 1,204 1,338 2,252 3,302 11,14% -0,185 0,854 Australia 1505 0,669 1,975 3,738 5,325 194,95% -5,504 0,000*** Austria 48 1,899 2,797 1,780 2,571 47,30% -1,223 0,228 Belgium 85 0,543 2,430 1,224 9,393 347,88% -0,87 0,387 Brazil 226 3,394 4,805 4,931 6,646 41,58% -1,815 0,071 Canada 1275 0,459 1,712 2,546 3,933 273,25% -6,106 0,000*** China 2410 0,618 0,874 0,830 1,364 41,40% -5,657 0,000*** Denmark 97 3,203 1,506 12,386 1,835 -52,97% 1,062 0,291 Finland 106 0,728 3,173 1,487 3,166 335,92% -2,286 0,024** France 562 1,197 1,535 3,200 2,121 28,17% -1,207 0,288 Germany 462 1,458 2,004 2,334 3,479 37,45% -1,948 0,052* United Kingdom 1072 0,922 2,085 2,789 2,253 126,24% -7,53 0,000*** Greece 172 0,959 0,853 4,544 2,830 -11,07% 0,175 0,861 India 2084 1,447 1,016 6,423 2,519 -29,83% 2,019 0,044** Indonesia 361 1,262 1,345 2,384 2,530 6,55% -0,295 0,768 Ireland 64 0,252 1,481 0,646 1,524 487,70% -3,811 0,000*** Italy 185 2,508 1,663 6,6321 2,523 -33,60% 1,277 0,221 Japan 2325 1,475 1,482 1,059 1,077 0,43% -0,107 0,914 Luxembourg 43 1,057 1,897 1,712 2,372 79,41% -1,35 0,185 Mexico 69 1,364 1,317 2,253 1,418 -3,43% 0,091 0,928 Netherlands 100 0,891 1,469 1,713 1,823 64,93% -1,875 0,064* Portugal 45 0,031 3,919 0,088 5,938 12423,00% -1,836 0,073* Russian Federation 331 2,922 2,789 6,385 7,336 -4,52% 0,151 0,880 South Africa 182 1,529 2,250 2,269 2,373 46,86% -1,618 0,107 Spain 114 1,654 1,991 3,224 2,953 20,42% -0,543 0,588 Sweden 356 0,567 1,652 1,815 2,385 191,10% -3,622 0,000*** Turkey 214 1,397 1,766 3,004 3,081 26,43% -0,849 0,397 United States of America 2651 0,882 1,338 3,923 2,788 51,67% -3,489 0,000*** Total Average 17205 1.458 2.007 2.992 3.217 37,45% -10.774 0,000*** *** p<0.01, ** p<0.05, *p<0.1, # are the number of firms in each country with no missing values. ∆% is the change in mean between the two groups.
independent variable gender diversity leads to a 1,2704% increase in dividend yield. This finding is in line with the expectation that gender diversity leads to higher dividend.
Table 7: Regression analysis
This table provide the findings based on an OLS regression analysis. The coefficients and standard errors are provided for the various variables. Different models have been performed.
(1) Dividend Yield*** (2) Dividend Yield*** (3) Dividend Yield*** (4) Dividend Yield*** (5) Dividend Yield*** (6) Dividend Yield*** IV_GD 1,2704*** 0,5151** 1,1337*** 1,1014*** 0,3942* (0,1757) (0,2273) (0,1825) (0,1778) (0,2327) CNT_ROA 1,6500 1,6784 1,6316 (1,4571) -1,4518 (1,4638) CNT_Size 0,3013*** 0,3033*** 0,2882*** (0,0227) -0,0228 (0,0237) CNT_Deratio 0,0822** 0,0843** 0,0826* (0,0366) -0,0363 (0,0368) CNT_Growth -0,0046*** -0,0046*** -0,0046*** (0,008) (0,008) (0,008) CNT_TaxWH 1,9015*** 1,9548*** 1,4137*** (0,1789) (0,1711) (0,2099) CNT_Tax Cor 1,4660*** 1,5110*** 1,3415** (0,3506) (0,3424) (0,3512) CNT_MS -0,0052*** -0,0053*** -0,0045*** (0,0010) (0,0010) (0,0010) MOD_Mas -0,0086*** -0,0091*** -0,0098*** (0,0017) (0,0018) (0,0022) MOD_Uav 0,0153*** 0,0151*** 0,0100*** (0,0014) (0,0014) (0,0016) INT_GDxMas -0,0133 (0,0297) INT_GDxUAV -0,0398 (0,0347) Constant 1,1403*** 2,0315*** 2,1494*** 0,8568*** 0,8930*** 1,7076*** (0,0296) (0,3329) (0,3104) (0,0956) (0,1022) (0,3364) R2 0,0033 0,0446 0,0441 0,0136 0,0138 0,0493 Observations 17353 13979 13979 17353 17353 13979
*** p<0.01, ** p<0.05, *p<0.1, Parentheses contains the standards errors
country can be related to the quality of the country’s capital market, the country’s politics, or cross-country agreements such as the European union just to name a few.
To extend the basic model to encompass the country’s culture as a mean of distinguishing the included countries the remaining three models are analysed. To do so model (1) is expanded to model (4) by including the hypothesised moderating cultural variables. The results of the regression using model (4), R2 = 0,0136, F(17353) = 52,6503, p <0,01, show that there is a significant relationship of gender diversity, B = 1,337, p <0,01, on dividend yield controlled by uncertainty avoidance, B = 0,0153, p < 0,01, and masculinity, B = -0,0086, p < 0,01. A negative sign for masculinity and positive sign for uncertainty avoidance are in line with expectations. A moderating effect is present when, after adding an interaction term, a sign change or major shift in coefficient is observed. The two interaction terms are created as a product of gender diversity and each of the cultural values. Model (5) is the regression that test for the moderating effect. The results, even though significant, R = 0,0138, F(17353) = 43,3331, p < 0,01, do not show signs of a moderating effect. Both interaction terms are found to be insignificant. Model (6) is an extension of model (4) by adding the control variables. This model is significant, R2 = 0,0493, F(13986) = 56,5768, p <0,01, confirms the results from model (4) that culture does not function as a moderating variable.
5. Conclusion
In recent years, there has been a major increase in research into the dividend policy of firms. Besides rewarding shareholders, are there other reasons for paying dividend? Or is a certain dividend policy the consequence of other action. Basically, is dividend policy the goal or is dividend policy the means to an end? This research follows the agency theory perspective where it presumed that higher dividend has a goal to decrease agency costs of debt by aligning the goals of the agents and the principal. One might favour reducing agency cost by increasing dividend payments whereas another would use other methods for achieving the same goal. This research focusses on the extent to which dividend payments are used to decrease agency cost.
Gender equality is a hot debate in recent years. Examples are the skewed distribution in wages and job positions between males and females. This skewness and increased awareness has led to institutions setting goals to achieve a fair distribution between gender. For example, some countries might prescribe certain equality norms through changes in legislation, or through labour unions making agreements. Changes are present and this skewed distribution is being reduced. This can be observed partly by the growing presence of females in the firm’s leadership. Previous research found multiple differences between male and females. One of these differences is the level of risk aversion between gender. Women on average favour taking less risk compared to the male counterpart. The focus in this article is analysing what influence more females in the board of directors have on the dividend policy of the firm.
potentially influence this relationship and can be the cause of the low explanatory value of the model. It can be concluded that whether a board of directors is gender diverse is more meaningful than the extent of this diversity.
Firms go to the stock market for various reasons. It can be that firms need capital to maintain their activities or that there is a desire for growth or other investments. Investors can easily go aboard to acquire shares from firms located in other countries. Besides this accessibility of shares there is an increasing trend in firms that internationalize. This can range from export or up greenfield investments. Having an interaction with people or entities from different cultures is because of the increased internationalization more likely to occur. Cultures are because of that becoming increasingly more important. Cultures influence people’s norms and values, their views and perception, one of which is risk. Different cultural models measure these differences between cultures among various dimensions. Some cultures live in the present and worry less about what the future might bring. Other cultures can be focussed on the longer term and rather avoid risk instead of embracing it. Using this knowledge to reflect upon the main hypothesis this research argues that in certain cultures the pressure by females to reduce agency cost is moderated by the countries culture. In certain countries where inhabitants rather avoid risk firms might be more open to the idea to increase dividend payments in order to decrease agency cost and follow the risk aversion preference of females. This argumentation leads to the second hypothesis.
Having found a positive significant coefficient on the main relationship the next step is to test this second hypothesis. The coefficient of gender diversity and from both cultural variables did not show a major increase in strength or experienced a sign change from a positive coefficient to a negative. Therefore, it is concluded that that there are no signs of a moderating effect and this hypothesis can be rejected. Even though culture has been proven many times to be of influence in behaviour science, there is a possibility that cultural influences are being overpowered or absorbed by organizational culture. Internationalization is the interaction between individuals and/or entities where national cultures collide. Opposed to individuals cultural, organizational culture might be better for investigating firm risk. Unfortunately, organizational culture brings forth its own unique difficulties such as measuring in a cross country and cross industry setting.
focusing on a single country of even a specific industry within a certain country. By doing so it removes noise that arises from differences between multiple countries and industries. These forms of noise are the cause of some limitation for this research. Since these findings does not prove causality there may be forces that have not been considered. Meaning that future research can benefit from focusing on the forces which also influences the change in dividend yield. Lastly, organisational culture is more specific than national culture and may provide different insights which are yet unclear. Determining a comparable method to measure organizational culture must be available.
Investors can be separated into two groups regarding expected returns. One group favours that profits are reinvested within the firm since they believe the firm has the potential to achieve higher rewards compared to the individual. Increased value will be in the form of increased share value. The other group depends on the firm to pay dividend to reward shareholders as investors of capital. These investors would rather receive cash compared to increased share value. From a practical point of view these results do indicate that gender diversity matter. A change in board composition within a firm with zero gender diversity (in this case all males) towards a composition with at least a single female has an average dividend difference of 37%. Causality might not be proven in this research but having found that the result is significant in 11 countries as well as the sample as a whole, this can still influence investor behaviour. Investors with a relatively small number of shares can decide to invest or divest in a firm that experiences a change in the structure of the board of directors according to their preference. Investors with relatively large amounts of shares can even exert significant influence in order push for, or alter the change to, gender diversification.
this research only looks at the nationality of the firm. It could have been possible to include the nationality of each board member to further investigate if that can have an influence. Data availability made it not reasonable to include this in this scope. (3) The culture is being measured at the country level. Business culture could explain a larger variance since this is being influenced by all nationalities of the stakeholders. Unfortunately, this data on this scale in not realistic. If future research is limited in size, meaning investigating a single industry or country, it might be viable to include this. Besides the delimitations which are within control of the researcher there are also limitations that are not controlled. (4) One of the limitations is that the data on the home country, which is used for the cultural hypothesis in this research, has its flaws. It is common that countries are officially located in a certain country because off some benefits, but are operationally headquartered elsewhere. In such case the culture would present an unfair image. Lastly (5), each country is represented in this research to a different extent. Capital is not equally distributed in the world market meaning that firms can be more likely to list on the stock markets in one country compared to a different one where banks and other institutions are the main provider of finance.
6. References
Adams, R.B., Ferreira, D., 2009. Women in the boardroom and their impact on governance and performance. J. financ. econ. 94, 291–309.
Adams, R.B., Ferreira, D., 2004. Gender Diversity in the Boardroom. Financ. Work. Pap. 1–35.
Allen, F., Michaely, R., 1995. Dividend Policy. Northholl. Handbooks Oper. Res. Manag. Sci. Financ. 9, 793–837.
Bae, S.C., Chang, K., Kang, E., 2012. Culture, corporate governance, and dividend policy: International evidence. J. Financ. Res. 35, 289–316.
Baker, H.K., Singleton, J.C., Veit, E.T., 2011. Survey Research in Corporate Finance: Bridging the Gap Between Theory and Practice. Oxford University Press.
Baker, H.K., Weigand, R., 2015. Corporate dividend policy revisited. Manag. Financ. 41, 126–144.
Barber, B.M., Odean, T., 2001. Boys will be boys: Gender, overconfidence, and common stock investment. Q. J. Econ. 116, 261–292.
Black, F., 1976. The pricing of commodity contracts. J. financ. econ. 3, 167–179.
Breuer, W., Rieger, M.O., Soypak, K.C., 2014. The behavioral foundations of corporate dividend policy a cross-country analysis. J. Bank. Financ. 42, 247–265.
Brockman, P., Unlu, E., 2009. Dividend policy, creditor rights, and the agency costs of debt. J. financ. econ. 92, 276–299.
Brynes, J.P., Miller, D.C., Schafer, W.D., 1999. Gender differences in risk taking: A meta-analysis. Psychol. Bull. 125, 367–383.
Carter, D.A., Simkins, B.J., Simpson, W.G., Borokhovich, K., Crutchley, C., Elson, C., Fry, M., Hersch, P., Li, J., Longhofer, S., 2003. Corporate Governance, Board Diversity, and Firm Value. Financ. Rev. 38, 33–53.
Chae, J., Kim, S., Lee, E.J., 2009. How corporate governance affects payout policy under agency problems and external financing constraints. J. Bank. Financ. 33, 2093–2101.
Chang, K., Wee, J.B., Yi, H.C., 2012. Does national culture influence the firm’s choice of debt maturity? Asia-Pacific J. Financ. Stud. 41, 424–457.
Chui, A.C.W., Lloyd, A.E., Kwok, C.C.Y., 2002. The Determination of Capital Structure: Is National Culture a Missing Piece to the Puzzle? J. Int. Bus. Stud. 33, 99–127.
Chuluun, T., Prevost, A., Puthenpurackal, J., 2014. Board ties and the cost of corporate debt. Financ. Manag. 43, 533– 568.
Croson, R., Gneezy, U., 2009. Gender Differences in Preferences. J. Econ. Lit. 47, 448–474.
De Mooij, M., 2004. Consumer Behavior and Culture: Consequences for Global Marketing and Advertising, Journal of Consumer Policy.
Desvaux, G., Devillard, S., 2008. Female Leadership, a Competitive Edge for the Future. Women Matter 1, 1–23.
Earley, P.C., Mosakowski, E., 2000. Creating hybrid team cultures: An empirical test of transnational team functioning. Acad. Manag. J. 43, 26–49.
Easterbrook, F.H., 1984. Two agency-cost explanations of dividends. Am. Econ. Rev. 74, 650.
Faccio, M., Marchica, M.T., Mura, R., 2016. CEO gender, corporate risk-taking, and the efficiency of capital allocation. J. Corp. Financ. 39, 193–209.
Fama, E.F., French, K.R., 2001. Disappearing dividends: Changing firm characteristics or lower propensity to pay? J. financ. econ. 60, 3–43.
Fama, E. F., & French, K. R., 2002. Testing trade‐off and pecking order predictions about dividends and debt. Rev. of finance. Studies. 15(1), 1-33 .
Farinha, J., 2003. Dividend policy, corporate governance and the managerial entrenchment hypothesis: An empirical analysis. J. Bus. Financ. Account. 30, 1173–1209.
Fauver, L., McDonald, M.B., 2015. Culture, agency costs, and governance: International evidence on capital structure. Pacific-Basin Financ. J. 34, 1–23.
Fredrickson, B.L., 2001. The Role of Positive Emotions in Positive Psychology. Am. Psychol. 56, 218–226.
Giner-Sorolla, R., 2001. Guilty pleasures and grim necessities: affective attitudes in dilemmas of self-control. J. Pers. Soc. Psychol. 80, 206–221.
Gorodnichenko, Y., Roland, G., 2011. Which dimensions of culture matter for long term growth? Am. Econ. Rev. 101, 492–498.
Gregory-Smith, I., Main, B.G.M., O’Reilly, C.A., 2014. Appointments, pay and performance in UK boardrooms by gender. Econ. J. 124.
Grimm, C.M., Lee, H., Smith, K.G., 2005. Strategy as action: Competitive dynamics and competitive advantage, Strategic Management. Oxfort University Press.
Hofstede, 1991. Cultures and Organizations, Cultures and Organizations.
House, R.J., Hanges, P.J., Javidan, M., Dorfman, P.W., Gupta, V., 2004. Book Review: Culture, Leadership, and Organizations: The Globe Study of 62 Societies. J. CrossCultural Psychol. 36, 628–630.
Hussainey, K., Mgbame, C.O., Chijoke-Mgbame, A.M., 2011. Dividend policy and share price volatility: UK evidence. J. Risk Financ. 12, 57–68.
Jensen, M., 2001. Value Maximisation, Stakeholder Theory, and the Corporate Objective Function. Eur. Financ. Manag. 7, 297–317.
Jensen, M.C., Meckling, W.H., 1976. Theory of the Firm : Managerial Behavior , Agency Costs and Ownership Structure. J. financ. econ. 3, 305–360.
Jurkus, A.F., Park, J.C., Woodard, L.S., 2011. Women in top management and agency costs. J. Bus. Res. 64, 180–186. Kaufmann, D., Kraay, A., 2017. Worldwide governance indicators [WWW Document]. URL
http://info.worldbank.org/governance/wgi/index.aspx#home (accessed 5.5.17).
Kivetz, R., Simonson, I., 2002. Self-Control for the Righteous: Toward a Theory of Precommitment to Indulgence. J. Consum. Res. 29, 199–217.
La Porta, R., Lopez-De-Silanes, F., Shleifer, A., Vishny, R.W., 2000. Agency problems and dividend policies around the world. J. Finance 55, 1–33.
Laran, J., 2010. Choosing Your Future: Temporal Distance and the Balance between Self Control and Indulgence. J. Consum. Res. 36, 1002–1015.
Lee, L.-E., Marshall, R., Rallis, D., Matt, M., 2015. Global trends in gender diversity on corporate boards. Levi, M., Li, K., Zhang, F., 2014. Director gender and mergers and acquisitions. J. Corp. Financ. 28, 185–200.
Li, K., Griffin, D., Yue, H., Zhao, L., 2013. How does culture influence corporate risk-taking? J. Corp. Financ. 23, 1–22. Lintner, J., 1964. Optimal dividends and corporate growth under uncertainty. Q. J. Econ. 78, 49–95.
Loukil, N., Yousfi, O., 2016. Does gender diversity on corporate boards increase risk-taking?, in: Canadian Journal of Administrative Sciences. pp. 66–81.
Louro, M.J., Pieters, R., Zeelenberg, M., 2007. Dynamics of multiple-goal pursuit. J. Pers. Soc. Psychol. 93, 174–193. Luna, D., Gupta, S.F., 2001. An integrative framework for cross-cultural consumer behavior. Int. Mark. Rev. 18, 45–69.
McGuinness, P.B., Lam, K.C.K., Vieito, J.P., 2015. Gender and other major board characteristics in China: Explaining corporate dividend policy and governance. Asia Pacific J. Manag. 32, 989–1038.
Miller, M.H., Modiglini, F., 1961. Dividend Policy, Growth, and the Valuation of Shares. Source J. Bus. 34, 411–433.
Pucheta-Martínez, M.C., Bel-Oms, I., 2015. The board of directors and dividend policy: the effect of gender diversity. Ind. Corp. Chang. First publ, 1–25.
Rozeff, M.S., 1982. Growth, Beta and Agency Costs as Determinants of Dividend Payout Ratios. J. Financ. Res. 5, 249– 259.
Schwartz, S., 1999. A Theory of Cultural Values and Some Implications for Work. Appl. Psychol. 48, 23.
Shao, L., Kwok, C.C., Guedhami, O., 2010. National culture and dividend policy. J. Int. Bus. Stud. 41, 1391–1414.
Soares, A.M., Farhangmehr, M., Shoham, A., 2007. Hofstede’s dimensions of culture in international marketing studies. J. Bus. Res. 60, 277–284.
Upadhyay, A.D., Bhargava, R., Faircloth, S., Zeng, H., 2017. Inside directors, risk aversion, and firm performance. Rev. Financ. Econ. 32, 64–74.
Webb, E., 2004. An examination of social responsible firms’ board structure. J. Manag. Gov. 8, 255–277.
Wilcox, K., Kramer, T., Sen, S., 2011. Indulgence or Self-Control: A Dual Process Model of the Effect of Incidental Pride on Indulgent Choice. J. Consum. Res. 38, 151–163.
Wilson, N., Altanlar, A., 2009. Director characteristics, gender balance and insolvency risk: an empirical study. Gend. Balanc. Insolv. Risk An Empir. Study 54.
Appendix
Appendix 1: Breusch-Pagan result
Test for heteroskedasticity
Mean F Sig.
Regression 888572,655 52,043 ,000b
P value is < 0,05. This means heteroskedasticity is present and robust standard errors should be used.
Appendix 2: Collinearity Statistics
Appendix 3: Collinearity Statistics
Scatterplot of Gender Diversity and Dividend Yield. Added is a fitted line
