Board Independence and Corporate Social Responsibility Disclosure: The Mediating Role of the Presence of Family Ownership

: This paper examines the impact of board independence on corporate social responsibility (CSR) disclosure and analyses the moderating effect of the presence of family ownership. Using an international sample from 29 countries from 2006 to 2014, our panel Tobit estimation shows that board independence is negatively associated with CSR disclosure practices and they present opposition to CSR disclosure practices. However, family ownership moderates the relationship and enforces the positive orientation of independent directors towards CSR disclosure. This shows that the presence of family ownership reduces independent director concern of reputation risks associated with receiving misleading information and family ﬁrms decrease the asymmetries of information between the independent director and management. The study also ﬁnds that independent directors encourage CSR disclosure in family ﬁrms more in civil law countries where investor protection is low compared to common law countries where investor protection is high.


Introduction
This paper examines the impact of board independence on corporate social responsibility (CSR) disclosure by investigating the moderating role played by family ownership.Our aim is to analyze whether the relationship between independent director and CSR disclosure is analogous for family and non-family firms.Specifically, we consider whether family business reinforces the effect of independent directors on the disclosure of information.The unique nature of agency problems between majority family shareholders and non-family minority shareholders (Chau and Gray 2002) motivated us to look into the role that CSR disclosure plays.Family firms often hold the majority of shares and important management positions, which enable them to control the management and affect the decision-making process (Chen et al. 2008).The agency problem between shareholders and managers becomes less relevant in family firms because usually family members participate in management.Nevertheless, it gives rise to an agency problem between majority and minority owners.The agency perspective suggests that family members may be opportunist and may take advantage of available information at the expense of minority shareholders (Chen et al. 2008) because family members have more information available than non-family members (Chrisman et al. 2004;Chau and Gray 2002).However, in family firms, the disclosure of information is not so necessary to decrease asymmetries because of the "information efficiency effect" (Cho et al. 2013).This effect suggests that if a family member performs insider trading based on their information advantage, less-informed investors will follow them, reducing the information asymmetry.Socio-emotional wealth theory (SEW) proposes that family firms draw utility from social wealth (Berrone et al. 2010;Gomez-Mejia et al. 2011) and as a result family members benefit more from CSR disclosure.On considering agency, from the SEW and information efficiency effect perspectives, the influence of family firms has not been clearly established in the literature as there is a possibility of both entrenchment and alignment of interests with minority shareholders.Therefore, it becomes very important to analyze, in the context of CSR disclosure, whether family firms will act opportunistically and discourage CSR disclosure or act in line with minority shareholders and encourage CSR disclosure.Hence, we analyze the relationship between independent director and CSR disclosure in the presence of this exceptional conflict between majority family and minority non-family shareholders.
CSR information is gradually gaining importance in the global economy and firms have widely adopted it as the medium of communication with all stakeholders (Cho et al. 2013).Among the voluntary disclosures, CSR information is used as the most valuable information by analysts and investors (Prince's Accounting for Sustainability Project Report 2013).CSR information is useful to different stakeholders as it covers a wide number of topics related to economic, social and the environment (Williams and Pei 1999).The investor uses CSR information as one of the important factors while making an investing decision (Cohen et al. 2011;Chau 2006) and it also reduces the information asymmetry in the firms (Dhaliwal et al. 2012).However, due to the voluntary nature of CSR information, its disclosure could be purely based on the preference and motives of the managers and directors (Healy and Palepu 2001;Meek et al. 1995).The board of directors enjoys full discretion in revealing the information and in deciding the content of the report to be disclosed (Luo et al. 2012).In such cases, the independent director plays a very important role as they are not only accountable to shareholders but also ensure the welfare of other stakeholders (Ibrahim and Angelidis 1995); they also engage in motivating firms to disclose sustainability information (Frias-Aceituno et al. 2013a).So, the independent director may encourage disclosure of CSR information in the firm.
However, for independent directors, decisions regarding CSR reporting are closely linked to their concern for career and reputation.The independent directors rely on the manager for the information related to CSR and managers always possess a risk of receiving manipulative or misleading information (Kravet and Muslu 2013;García-Sánchez and Martinez-Ferrero 2018).The CSR information, in which they usually do not have to be experts (Cramer and Hirschland 2006), supposes the risk that any misleading information can damage their reputation and may affect their future job opportunity in other firms (Fama and Jensen 1983;Fama 1980).This alleviates the concern of independent directors about the reputation associated to CSR disclosure which can minimize their efforts for CSR issues (Cheng and Courtenay 2006) and the independent director may show some opposition to CSR information and may have a negative association with CSR disclosure of the firm.However, García-Sánchez and Martinez-Ferrero (2018) find CSR performance and disclosure are positively associated, a higher performance, better information disclosure.In this sense, greater social and environmental performance reduces independent director reputation risk associated with receiving misleading information.
Given the mixed evidence in the literature, which shows independent directors can be positively or negatively associated with CSR disclosure policies, it is more evident that certain firm characteristics can moderate this relationship.One such factor may be a family influence on the firm.The presence of family ownership in the firm monitors managers and decreases management discretion (Martínez-Ferrero et al. 2016b).Therefore, we expected that the presence of family ownership will decrease the chances of an independent director receiving misleading information from the manager.Moreover, family firms will encourage more CSR disclosure as they benefit more from social wealth (SEW).In this paper, we extend the literature by analyzing the role of the family firm in reducing independent director reputation risk associated with receiving misleading information as family ownership in the firm controls the management discretion and brings down the information asymmetry for an independent director.Hence, we analyze the moderation effect of family ownership on independent directors and CSR disclosure.
We further analyze this relationship in different investor protection environments.Independent directors' decisions depend on the institutional environment in which they operate (Bebchuk and Weisbach 2010).So, we try to analyze whether independent director involvement varies with the level of investor protection environment.The countries with low investor protection will require a more robust internal corporate governance mechanism to safeguard the interest of shareholders.We study proxies of investor protection by legal tradition and perform analyses for countries with different legal traditions (common law or civil law).
The objective of the paper is to test the tendency of independent directors towards CSR disclosure and if there are differences in attitudes related to CSR disclosure between family and non-family firms.
To test this empirically we examine an international sample from 29 countries from 2006 to 2014.
The rest of the paper is structured as follows: Section 2 presents the theoretical framework and on the basis of relevant literature develops the hypothesis.Section 3 describes the sample and data selection, the definition of the variables used in the study and explains the model and technique used in the study.In Section 4 we discuss and interpret the results and finally, Section 5 concludes the paper and highlights the contribution of the study.

Independent Directors and CSR Disclosure
Information asymmetry acts as the fundamental source of the agency problem between the principal (shareholders) and agent (manager) (Jensen and Meckling 1976).The manager, due to their day to day involvement in business actions, may possess information which is not available to the shareholders.This information advantage to the manager may prompt them to take actions which are beneficial to them such as unreasonable compensation and empire building at the expense of shareholders (Healy and Palepu 2001).In such a scenario, transparency and disclosure become essential to reduce the information asymmetry.The lack of disclosure may incentivize the private information and further raises the information asymmetry (Lambert et al. 2007;Diamond and Verrecchia 1991).Studies have shown that firms with better disclosure of information and transparency policies have less information asymmetry (Milgrom 1981;Grossman and Hart 1980).
One such source of information regarding economic, social and environmental issues of corporate performance is information related to CSR.Better performance in CSR leads companies to greater disclosure of aspects related to CSR (García-Sánchez and Martinez-Ferrero 2018).CSR reports are management's communication to all stakeholders about performance, which is day-by-day becoming important for market participants (Cho et al. 2013).CSR information may be of interest to stakeholders as it covers a wide range of indicators of economic, social and environmental aspects (Williams and Pei 1999).Investors in recent years are looking to CSR information as one of the key factors in making an investment decision (Cohen et al. 2011;Chau 2006).Cho et al. (2013) show it helps in bringing down the uncertainty about the firm's value and Dhaliwal et al. (2012) shows that CSR information reduces information asymmetry and decreases the analyst forecast error.
Moreover, the disclosure of CSR information depends on the selection and motives of the managers and directors (Healy and Palepu 2001;Meek et al. 1995).A board of directors has full discretionary power in deciding the information to be disclosed in the reports (Luo et al. 2012).In such a scenario, the role of an independent director becomes very crucial as they are the representative of shareholders and stakeholders in the board and responsible for safeguarding the interest of the shareholders (Fama and Jensen 1983).
Independent directors are fully independent of the management and do not enjoy personal interests in the company.Independent directors by virtue of their independent nature provide better monitoring of the management (Rosenstein and Wyatt 1990) and enhance the board's effectiveness (Michelon and Parbonetti 2012).Independent directors also help to control the agency cost (Patelli and Prencipe 2007).They are not only accountable to shareholders but also ensure the welfare of other stakeholders (Ibrahim and Angelidis 1995); they also engage in motivating firms to disclose sustainability information (Frias-Aceituno et al. 2013a).Barako and Brown (2008) and Khan (2010) show that a greater proportion of independent directors in a firm increases the focus on CSR disclosure.Khan et al. (2013) and Lattemann et al. (2009) also found similar results in developing and emerging economies.Likewise, the relationship between CSR disclosure and independents is positive when performance on CSR issues is particularly good (García-Sánchez and Martinez-Ferrero 2018).
Nevertheless, some studies report no relationship between independent directors and CSR disclosure.Michelon and Parbonetti (2012) among US and European firms and Said et al. (2009) among Malaysian firm did not find any significant relationship between independent directors and CSR disclosure practices.Moreover, Haniffa and Cooke (2005) show that independent directors discourage CSR disclosure in the firm and Prado-Lorenzo and Garcia-Sanchez (2010) report similar findings in the case of disclosure of information related to greenhouse gas emission.
The mixed results between the independent director and CSR disclosure shows that independent directors' individual preferences and interests may also have an influence on the decision-making process.Reputation and networking relations are considered to be the main benefits of directorship for independent directors (Fama and Jensen 1983;Lorsch and Young 1990).The behavior of independent directors may be motivated by their concern for career and reputation.Independent directors may avoid the risky actions of the firm that could damage their reputation (Holmström 1999).This shows their concern about reputational risk.Such a risk seems to arise in the case of CSR disclosure decisions.The sources of CSR information for independent directors are managers and there is always a risk of receiving manipulative or misleading information from the management (Kravet and Muslu 2013).They also lack in focus knowledge and specialized training about the CSR issue (Cramer and Hirschland 2006).Thus, CSR information disclosure increases the reputational concern for the independent directors and they could act in their own interest (Ravina and Sapienza 2009) and may minimize their efforts for CSR issues (Cheng and Courtenay 2006).Therefore, we expect that independent directors may show some opposition to CSR information and may have a negative influence on the CSR disclosure of a firm (García-Sánchez and Martinez-Ferrero 2018).To test this, we propose the following hypothesis.
Hypothesis 1 (H1).The independent directors on the board is negatively related to the CSR disclosure in the firms.

Family Ownership, Independent Directors and CSR Disclosure
Family firms play an important function in the economy (Campopiano and De Massis 2015).In family firms, family founders often remain in the top management position and are closely involved in the business processes which enables them to access more information and control managers more closely compared to non-family firms (Chen et al. 2008;Chrisman et al. 2004).Thus, the classical agency problem between shareholders (principal) and manager (agent) becomes less relevant due to the existence of a family control mechanism over managers.Nevertheless, the presence of family ownership gives rise to a unique agency problem between family and non-family shareholders (Chau and Gray 2002).The presence of family blockholding empowers them to affect the firm's decision-making processes and enables them to access more information than the non-family shareholders.This generates an information asymmetry between the family (majority) and non-family shareholders (minority).Cho et al. (2013) show an information efficiency effect, that is, if a family member tries to take information advantage and perform trading based on information, they will be followed by the less informed investors and this reduces the information asymmetry.Family firms may disclose a larger volume of information to bring down the information asymmetry with minority shareholders.One such type of information could be related to CSR.Therefore, the presence of family ownership in firms may affect the relationship between the independent directors and CSR disclosure and present differences with non-family firms.
In the above section, we have seen that independent directors always have the risk of receiving manipulative or misleading CSR information from the management which discourages them from disclosing such information.However, the presence of family can limit the management's ability to manipulate and can better monitor the manager.Martínez-Ferrero et al. (2016b) show that the presence of family blockholding in the firm decreases the management's discretion.Therefore, it is to be expected that the presence of family ownership will decrease the chances of an independent director receiving misleading information from the manager and will bring down their reputation risk.So, the presence of family ownership may increase the probability of an independent director encouraging CSR disclosure in the firm.
Further, family firms traditionally focused on reputation, family identity and trying to maintain a positive image in the business environment (Marques et al. 2014;Berrone et al. 2010;Anderson and Reeb 2003) which is supported by the socio-emotional wealth (SEW) theory.The SEW perspective shows that family firms draw utility from the social wealth which acts as the basis of a difference between family and non-family firms (Gomez-Mejia et al. 2011).The family firms will be more inclined to preserve their SEW rather than only focusing on increasing firm profitability (Marques et al. 2014;Van Gils et al. 2014).They will be more responsive towards stakeholders and will try to build long-term relationships with different stakeholders.
In order to preserve their reputation and improve the image of the firm, family firms may engage more towards the CSR disclosure demands of the shareholders and the presence of family ownership may moderate the relationship between the independent director and CSR disclosure.Cennamo et al. (2012) show that family firms are more oriented towards CSR actions and reporting to uphold their SEW.So, we expect that family firms compared to non-family firms, regarding the decision of CSR disclosure, will act differently, as they are more interested in preserving their reputation and creating more utility from socio-economic wealth.
So, based on the above argument, we expect family firms to decrease management discretion and bring down information asymmetry for the independent director.Moreover, family firms will encourage more CSR disclosure as they benefit more from social wealth (SEW).So, in this study we expect that family firms will moderate the impact of independent directors on CSR disclosure and in family firms, independent directors will be encouraging CSR disclosure compared to non-family firms.So, aiming to contribute the literature on the role of independent directors on CSR disclosure in family firms, we propose the following hypothesis: Hypothesis 2a (H2a).The independent directors on the board are positively related to the CSR disclosure in the family firms.
Independent directors' decisions depend on the institutional environment in which they operate (Bebchuk and Weisbach 2010).It is difficult to understand internal corporate governance mechanisms without considering the institutional environment.Garcia-Sanchez et al. (2015a) show that independent directors behave differently in different investor protection environments.So, we try to analyze whether independent director involvement varies with the level of investor protection environment.Related to the protection of the investor, a distinction is usually made between common law and civil law legal systems (La Porta et al. 1998).On the one hand, common law countries have a well-developed legal system that protects investors (Campbell 2006), and gives more importance to shareholders (La Porta et al. 2000).The civil law legal systems are more oriented towards the stakeholders (Frias-Aceituno et al. 2013b;Simnett et al. 2009), which may determine the type of information that should be disclosed, in this case the CSR disclosure.Companies in countries where the orientation towards stakeholders and where compliance mechanisms are more efficient (civil law) tend to reveal more CSR information than companies in those countries in which the interests of shareholders prevail (Garcia-Sanchez et al. 2015b).Moreover, La Porta et al. (1998) shows that common law countries' investors often enjoy greater protection of their interests while civil law counties have weak investor's protections.Countries with low investor protection will require a more robust internal corporate governance mechanism and disclosure to safeguard the interests of shareholders.This phenomenon becomes more crucial as family firms are more prevalent in civil law countries, characterized by a high ownership concentration that limits market control on firm, while in common law countries ownership is not concentrated and there is a strong market control.So, in civil law countries, the independent director is expected to be active and take the decision to improve the firm-level governance and encourage disclosure to offset the weak investor protections law.In this sense, we propose the following hypothesis: Hypothesis 2b (H2b).The independent directors on the board is positively related to the CSR disclosure in the family firms in civil law countries.

Sample and Data
The period considered for the study is from the year 2006 to 2014.We rely primarily on the alliance of two databases to gather the data of the study.First, we use Thomson Reuters Eikon database to get the information of global stock markets from 29 stock markets across America, Europe, the Middle East, Africa and Asia.The full sample consists of almost 3927 companies after removing the duplicates companies our final sample consist of 3593 firms.Second, the data about social and environmental issues were obtained from the EIRIS database (Ethical Investment Research Service).This database provides social and environmental indicators for almost 30,000 firms.We excluded incomplete information observations and finally, we operated with a sample of 1072 firms covering all the period analyzed.The sample was not balanced because some companies were not present in all the years analyzed.

Measures
3.2.1.CSR Disclosure: The Comparability and Utility of Information as an Independent Variable CSR disclosure is usually made through the content analysis of the CSR reports.Several measures are possible.Some authors consider the volume of sustainability information disclosed (number of pages, phrases or words) (Samaha et al. 2012).The main problem is that this type of report is voluntary and usually the content of the report depends on the discretion of the managers.In addition, usually firms do not report controversies, risks or conflicts with the stakeholders and they do not provide the complete information (Adams 2002).
We create our indicator of CSR disclosure by using two variables, "Comparability" and "Utility" (García-Sánchez and Martinez-Ferrero 2018), to determine these two qualities in CSR reports (see Table 1).Both are important non-financial characteristics which analysts and investors take into account while making the financial decision (Prince's Accounting for Sustainability Project Report (2013) (A4S) and the Global Reporting Initiative (GRI)).Comparability refers to the fact that companies' information should be comparable over time and between companies.In order to make information comparable, it is necessary to use the standards and a common framework for reporting.The utility refers to the quality of the disclosed information.Information must respond to the demands and requirements of the stakeholders.So that stakeholders can make informed decisions.
CSR disclosure is a measure as to whether the firm's CSR report is following the benchmark of the Global Reporting Initiative (GRI) (Clarkson et al. 2008;García-Sánchez and Martinez-Ferrero 2018;Martínez-Ferrero et al. 2016a).The GRI guidelines include all the relevant aspects related to environmental, social and economic issues and importantly most of the indicators are reflected numerically.The GRI guidelines are widely followed all over the world.Using GRI guidelines allow us to compare the information and help us in establishing the different level of sustainability of each firm as initial, intermediate or advanced.Similar to Garcia-Sanchez et al. ( 2014) we decide the level of sustainability by measuring the number of indicators that the firm explains in the report.
So, as we mentioned we decide the CSR score of the firm by comparing the firm report with the GRI guidelines and extend to which firm follows the GRI guidelines.The first variable "Comparability" is scored between 0 and 60.The variable takes the value 0 if the firm is not disclosing CSR information, assign value 20 if firm disclosing CSR information but not according to GRI guidelines, assign value 40 for firms that disclose CSR information meets the standards of GRI (specifically the core option), and assign value 60 for firms that disclose CSR information and comply the comprehensible option of the guidelines.
The second "Utility" is measured through information obtained from the EIRIS database.They collect CSR information on employees' conditions and human rights, community, business ethics, environmental issues and establish four grades of utility.We assign 2.5 points for weak utility reporting, 5.0 points for an intermediate level, 7.5 points for a good level, and 10 points if the information is of excellent utility and if firms are not disclosing the items of CSR we assign the value of 0.
In order to represent the overall quality of CSR reports, our dependent variable is created by adding utility and comparability of information in the "CSRdisclosures" score (García-Sánchez and Martinez-Ferrero 2018).Companies that disclose CSR information adapted to the GRI guidelines.The G4 Sustainability Report Guidelines offer two options to an organization in order to prepare its sustainability report 'in accordance' with the guidelines: the Core option and the Comprehensive option.Material aspects are those that reflect the organization's significant economic, environmental and social impacts; or substantively influence the assessments and decisions of stakeholders.

points
CSR disclosure according to the comprehensive option included in the GRI guidelines.The comprehensive option includes additional indicators about the organization's strategy, governance, and ethics and the identification of specific material aspects

Independent Directors
Independent directors are the members of the board that do not have any employment relationship with the firms.Their separation from management assures a more efficient control and supervision of the corporate governance (Harjoto and Jo 2011).Usually, the measure of board independence is through the percentage of independent directors on the board (Haniffa and Cooke 2005;García-Sánchez and Martinez-Ferrero 2018).CSR disclosure may be related to the number of independent members in the board (García-Sánchez and Martinez-Ferrero 2018) and they can influence CSR disclosure if they are a higher number in the board.

Family Business as Moderating Factor
Family businesses play a crucial role in the economy.Usually, family member participates in management and governance of the firm (O' Boyle et al. 2010;Marques et al. 2014) However, an important issue to take into account is the family ownership.Their equity ownership allows them to control the firms (Berrone et al. 2010) and the percentage of ownership acts as a measure of their influence on the firm decision making (Gómez-Mejía et al. 2007).It is considered that if the family firm has more than 20% of ownership, it has a sufficient presence to influence the decisions of the company (Hambrick and Finkelstein 1995;García-Sánchez and Martinez-Ferrero 2018).In literature generally, family ownership is measured by a dichotomous variable (Kashmiri and Mahajan 2010;Berrone et al. 2010).So, the variable "Family" takes the value 1 when the family has the 20% or more of the ownership and 0 otherwise (Rodriguez-Ariza et al. 2017).We manually compile the data of family ownership by looking into the ownership structure of the firm, and later this information is combined with the Thomson database information.

Control Variables
We have included some variables that may be associated with CSR disclosure that it is necessary to control for.These control variables are "Size" measured as the logarithm of sales; "Leverage" measured as the logarithm of the proportion of total debt to total equity and to measure the innovation effort "R&D_intensity", we take the ratio of research and development expenses to total assets and, finally, we measure financial reporting quality "FRQ" as the difference between the change in accounts receivable (∆AR i,t ) and the yearly change in sales revenues for each firm in every year (∆Sales i,t ).This difference is the discretionary revenue multiplied by −1.Higher values indicate the highest quality of financial reporting (McNichols and Stubben 2008).
∆AR i,t = β 1 ∆Sales i,t + µ it "Board_Size" is measured by the number of members on the board and "Board_Activity" is measured by the number of board meetings in the year."Industry j " is a dummy variable, for the different sectors of activity; "Year n " is a dummy variable for each of the years and "Country k " is a dummy variable for the different countries in the sample.

Model of Analysis
This research aims to explore the impact of board independence on the quality of CSR reports, additionally, the moderating effect of family ownership as a control mechanism.We tested our proposition in a regression model in which our dependent variable "CSRdisclosure" is regressed on the independent directors, on the family ownership indicator and on their interaction and the control variables.The model controls the unobservable heterogeneity, η i , and the disturbance term µ it .The term "i" represents to the firm and "t" the year.

CSRdisclosures = ϕ
We applied panel data, specifically a panel Tobit model because dependent variable "CSRdisclosure" is scaled from 0 to 100.The advantage of using the Tobit model is that it follows the maximum likelihood method and takes into account censored and uncensored observations.Therefore, we obtain efficient and consistent estimates through the Tobit Panel model.The Tobit model assumes that the observable variable (called xit) explains the latent variable (called yit *).Similar to previous literature (Uribe-Bohorquez et al. ( 2018); Choi et al. (2007)) we use the variable Independent one period t − 1 which helps us avoid the problem of endogeneity.

Descriptive Statistics
Tables 2 and 3 reports the descriptive statistics and correlations for the variables used in the study.The mean value of the CSRdisclosure score is 26.81 in a possible range of 0-100 and a standard deviation of 19.97.The mean value obtained suggests that the sample companies on average disclose the CSR information, but GRI guidelines were not followed.The mean value of Independent is 59.40 indicating on an average the board has around 59.40% of independent directors.The standard deviation of Independent is 30.69,reflecting strong disparities in the number of independent directors on the board of firms.With respect to firm control factors, the mean Size of the firms is 22.80, the mean value of Leverage is 0.57, showing that level of debt in the companies is 0.57 times of their equity.The R&D_intensity of the companies on an average is 0.02.Moreover, with respect to board characteristics, on an average board size is around 11 directors.The correlation matrix shows low or moderate correlations among the variables.Some coefficients in the correlations matrix are statically significant but they are not very large which indicates that our models are not influenced by the problem of multicollinearity.CSRdisclosure represents the degree of utility and comparability of CSR report, taking values between 0 and 100.Independent is the percentage of non-executive directors.Family is a dummy variable that takes the value 1 if the largest shareholder is a family member with more than 20% of the votes, and 0 otherwise.Size is the natural logarithm of total sales.Leverage is the natural logarithm of the ratio of total debt to total equity.R&D_intensity is the the ratio of research and development expenses to total assets.FRQ represents discretionary revenues as the proxy for financial reporting quality.Board_Size is measured as the total number of directors on the board.Board_Activity is measured as the total number of meetings held in the year and is a proxy for the level of activity.N = 1072 firm-year observations.CSRdisclosure represents the degree of utility and comparability of CSR report, taking values between 0 and 100.Independent is the percentage of non-executive directors.Family is a dummy variable that takes the value 1 if the largest shareholder is a family member with more than 20% of the votes, and 0 otherwise.Size is the natural logarithm of total sales.Leverage is the natural logarithm of the ratio of total debt to total equity.R&D_intensity is the the ratio of research and development expenses to total assets.FRQ represents discretionary revenues as the proxy for financial reporting quality.Board_Size is measured as the total number of directors on the board.Board_Activity is measured as the total number of meetings held in the year and is a proxy for the level of activity.N = 1072 firm-year observations.Significance levels: * p < 0.10; ** p < 0.05; *** p < 0.01.
Table 4 reports the distribution of firms by country.In our sample, firms are distributed among 29 countries.The countries with the highest percentage of firms were USA, Japan, and Canada.Table 5 reports the descriptive statistics of family and non-family firms.In our sample, 12.5% of the entities were family firms.The tendency for CSR disclosure is greater in family firms compared to non-family firms, the average CSR disclosure score is 33.67 for family firms compared to 26.33 of non-family firms.Family firms also on average have more board members and engage in more board activity compared to non-family firms.The firm size and level of leverage in the firms is similar across the family and non-family firms.CSRdisclosure represents the degree of utility and comparability of CSR report, taking values between 0 and 100.Independent is the percentage of non-executive directors.Family is a dummy variable that takes the value 1 if the largest shareholder is a family member with more than 20% of the votes, and 0 otherwise.Size is the natural logarithm of total sales.Leverage is the natural logarithm of the ratio of total debt to total equity.R&D_intensity is the the ratio of research and development expenses to total assets.FRQ represents discretionary revenues as the proxy for financial reporting quality.Board_Size is measured as the total number of directors on the board.Board_Activity is measured as the total number of meetings held in the year and is a proxy for the level of activity.

Multivariate Results
Table 6 reports the baseline results of the study.The table reports the estimated coefficient and standard error associated with each explanatory variable.The coefficient of Independent (coef.−0.1020 ***; p < 0.001) is negative and statistically significant.This shows that independent directors tend to avoid the CSR disclosure in the firms.This result is similar to Michelon and Parbonetti (2012), Prado-Lorenzo and Garcia-Sanchez (2010) and Haniffa and Cooke (2005) which also document the negative association between the independent directors and CSR disclosures.This result indicates that independent directors tend to avoid risky actions which could adversely affect their reputation.This may be because they rely on managers regarding CSR information and as a result, the risk of receiving misleading information increases.We accept the hypothesis 1.
However, the presence of family ownership in the firm affects the relationship between independent directors and CSR disclosure.The coefficient of interaction term of Independent × Family (coef.0.1082 **; p < 0.05) is positive and statistically significant.This shows that relationship between independent directors and CSR disclosure is moderated by the family ownership and independent directors encourage CSR disclosure in family firms rather than in non-family firms (coef.Independent + coef.Independent × Family = −0.1020+ 0.1082 = 0.0062).This result indicates that presence of family ownership reduces the independent director concern of reputation risks associated with receiving misleading information.Family firm controls the management discretion, decrease the asymmetries of information between the independent director and management and they rely more on information disclosed (García-Sánchez and Martinez-Ferrero 2018).In family firms, family members are closely involved in the business process, hold key positions in the firm and tend to have information advantage than non-family firms (Chau and Gray 2010).Thus, they can monitor the managers more closely and this reduces the problem of information asymmetry that could affect independent directors.Independent directors in family firms focus more on satisfying stakeholders' demands for CSR disclosures (Martínez-Ferrero et al. 2016a;Garcia-Sanchez et al. 2015a).This also supports the SEW theory that family firms draw more utility from social wealth and will be more motivated to build a relationship with stakeholders.This finding is consistent with the socio-emotional wealth theory.We'accept Hypothesis 2a.
Moreover, regarding the coefficient obtained for the control variable, we find that firm size impacts positively on the level of CSR disclosure, which shows that bigger firms are associated with the larger volume of CSR disclosure.Financial reporting quality is positively associated with the level of CSR disclosure.With regard to board characteristics, we find that board activity impacts positively on the level of CSR disclosure, which shows that if firms hold more board meetings, it tends to increase CSR disclosure.The variable leverage, board size, and R&D_Intensity are not statistically significant in the model and we did not find any impact of these variables on CSR disclosure of firms.CSRdisclosure represents the degree of utility and comparability of CSR report, taking values between 0 and 100.Independent is the percentage of non-executive directors.Family is a dummy variable that takes the value 1 if family member has more than 20% of the votes and 0 otherwise.Size is the logarithm of total sales.Leverage is the logarithm of the ratio of total debt to total equity.R&D_intensity is the ratio of research and development expenses to total assets.FRQ represents discretionary revenues as the proxy for financial reporting quality.Board_Size is measured as the total number of directors on the board.Board_Activity is measured as the total number of meetings held in the year and is a proxy for the level of activity.N = 1072 firm-year observations.Significance levels: ** p < 0.05; *** p < 0.01.Estimated coefficients and associated standard errors are reported.

Complementary Results
In this section, we analyze the impact of independent directors on CSR disclosure in family firms for different investor protection environments.La Porta et al. 1997Porta et al. , 1998Porta et al. , 2000 show that internal corporate governance mechanisms are strongly associated with the external institutional environment and investor's rights and its effective implementation are the main factor of corporate governance.Investor protection is the level to which commercial laws protect investors.Firms in countries with better legal investor protection are valued more by the market compared to countries with low level of investor protection (La Porta et al. 2000).This indicates that countries with low investor protection will require a more robust internal corporate governance mechanism to safeguard the interests of shareholders (Garcia-Sanchez et al. 2015a, 2015b).In such an environment, firms need to orient themselves towards disclosing more information and to try to bring down the information asymmetry between the internal and external shareholders.This phenomenon becomes more crucial in family firms where family members tend to have an information advantage over other shareholders.Therefore, if family firms want to decrease the information asymmetry between the shareholders they will encourage disclosure in those countries where investor protections are weak.
La Porta et al. (1998), in their seminal work, classified legal traditions into two families-common law and civil law and finds that in common law countries investors often enjoy greater protection of their interests while civil law counties have weak investor's protections.Therefore, we proxy the investor protection by its legal tradition and perform analyses for countries with a common law and civil laws legal tradition.Table 7 shows the results.
Model 1 reports the results for civil law countries and Model 2 reports the results for common law countries.Similar to the baseline model, we find that Independent is negatively and statistically significantly related to the CSR disclosure.The coefficient of interaction term Independent × Family (coef.0.0953 **, p < 0.05) is positive and statistically significant for the civil law countries and in case of common law countries, the coefficient of interaction term is not significant.This shows that independent directors encourage CSR disclosure in family firms when the level of investor protection in low and have a higher orientation towards stakeholders (Garcia-Sanchez et al. 2015b).Family firms' decrease the information asymmetry between the shareholders in the countries were investor's protection is weaker than those countries were law is strong.This finding confirms that in civil law countries, independent directors are more active and take decision to better the firm level governance and disclosure to offset the weak investor's protections law.We accept the hypothesis H2b.CSRdisclosure represents the degree of utility and comparability of CSR report, taking values between 0 and 100.Independent is the percentage of non-executive directors.Family is a dummy variable that takes the value 1 if family member has more than 20% of the votes and 0 otherwise.Size is the logarithm of total sales.Leverage is the logarithm of the ratio of total debt to total equity.R&D_intensity is the ratio of research and development expenses to total assets.FRQ represents discretionary revenues as the proxy for financial reporting quality.Board_Size is measured as the total number of directors on the board.Board_Activity is measured as the total number of meetings held in the year and is a proxy for the level of activity.Significance levels: * p < 0.10; ** p < 0.05; *** p < 0.01.Estimated coefficients and associated standard errors are reported.

Conclusions
In this paper, we extend the literature by analyzing the role of the family firm in reducing independent director reputational risk associated with receiving misleading information from managers.Family ownership in the firm controls management discretion and helps directors to bring down information asymmetry.Therefore, we analyze the moderation effect of family ownership on the relationship between independent directors and CSR disclosure.Moreover, previous literature in the field of independent directors and CSR disclosure provides mixed evidence.The literature finds that independent directors can have positive, negative or no effect on CSR disclosure policies.This shows that certain firm characteristics can moderate this relationship.Our empirical analysis shows the negative association between independent directors and CSR disclosure.Independent directors avoid risky actions which could adversely affect their reputation.However, independent directors encourage CSR disclosure in family firms, which shows that independent directors in family firms focus more on satisfying stakeholder demands for CSR disclosure.This finding shows that monitoring of management by family reduces the chances of receiving misleading information which brings down the reputational risk for independent directors that is associated with CSR disclosure.This finding also confirms socio-emotional wealth theory which shows that family firms draw more utility from the social wealth and will be more motivated to build a relationship with stakeholders.The family focuses on building reputation, family identity and, in order to maintain the positive image in the business environment, encourages CSR disclosure.The study also analyzed whether independent director involvement varies with the level of investor protection environment.The study shows that independent directors encourage CSR disclosure in family firms more in civil law countries where investor protection is low than in common law countries where investor protection is high.This shows that in civil law countries, independent directors in family firms are more active and take the decision to improve the firm-level governance and disclosure to offset the weak investor protection laws.
This study contributes to the existing literature on corporate governance and voluntary disclosure by presenting an in-depth cross-country analysis of the effect of corporate governance on CSR disclosure in general and for family firm in particular.It focuses on the issue of the role of family firms in CSR and its reports, which was not looked at in detail in the existing literature (Benavides-Velasco et al. 2013).We provide new insight into how family firms behave differently from non-family firms in the context of CSR disclosure.The role and effect of corporate governance mechanisms also differ across family and non-family firms.
Our study extends the emerging literature on voluntary disclosure by showing how the board of director's characteristics determine CSR disclosure policies.We show how individual independent directors' interest influences CSR information in the firm.We use panel data from 29 countries which enables us to consider the country and time effects as opposed to the literature which largely focuses on cross-sectional data like Haniffa and Cooke (2005) for Malaysia, Khan (2010) for Bangladesh and Lattemann et al. (2009) and Khan et al. (2013) for developing countries.We also take recent time period for the study 2006-2014 over a longer horizon of nine years.We use CSR score which is measured by applying holistic approach including both comparability and utility which represent the level of standardization of the CSR information disclosed and social and environmental information utility.
We contribute to the mixed literature on independent directors and CSR disclosure.We show that the effect of the independent director is moderated by the firm's characteristics, particularly family ownership.Our study extends the literature on corporate governance and investor protection and shows the moderating effect of family ownership on the relationship between independent director and CSR disclosure is increased in countries with weak investor protection laws.This finding is similar to that of Kolk and Perego (2010) and Durnev and Kim (2005) and García-Sánchez and Martinez-Ferrero (2018), which show that, in the presence of weak investor protection laws, strong internal corporate governance substitutes investor protection.In this paper, we extend the literature (García-Sánchez and Martinez-Ferrero 2018) by analyzing the role of the family firm in reducing independent director reputational risk associated with receiving misleading information as family ownership in the firm controls management discretion and brings down information asymmetry for the independent director.
Our study also provides certain implications for firms, investors, policymakers, and regulators.The firms, in order to satisfy all stakeholders, should balance the proportion of executive and independent directors in the firms as we have seen in the results that board composition influences CSR disclosure.The firm should increase transparency of information related to social and environmental issues and can provide focused training for the independent directors on issues related to CSR.This will help in reducing the fear of false information and reputational damage of independent directors.Our results will help investors understand how family ownership of a firm influences board decisions.Family involvement increases the chances of CSR disclosure which will offer a number of benefits to investors and reduces information asymmetry in the firm.This study will also help policymakers and regulators in making systems more robust to make independent directors more responsible to stakeholders and to ensure they behave ethically in fulfilling the demand for CSR disclosure.
disclosure not according to the GRI guidelines 40 points CSR disclosure information according to the core option of the GRI guidelines The core option includes indicators about the main aspects of economic, environmental and social and governance performance issues

Table 1 .
Range to measure comparability and utility.

Table 4 .
Distribution of Firms by Countries.

Table 5 .
Descriptive Statistics of family and non-family firms.

Table 6 .
The Impact of Board Independence on CSR Disclosure in Family owned firms.

Table 7 .
The Impact of Board Independence on CSR Disclosure in Family owned firms for Civil and Common Law Countries.