Corporate Social Responsibility Strategies of Spanish Listed Firms and Controlling Shareholders’ Representatives

This article aims at analyzing how controlling shareholders’ representatives on boards affect corporate social responsibility (CSR) strategies (disclosing CSR matters) in Spain, a context characterized by high ownership concentration, one-tier boards, little board independence, weak legal protection for investors, and the presence of large shareholders, especially institutional shareholders. Furthermore, among controlling shareholders’ representatives, we can distinguish between those appointed by insurance companies and banks and those appointed by mutual funds, investment funds, and pension funds. The effect of these categories of directors on CSR strategies is, therefore, also analyzed. Our findings suggest that controlling shareholders’ representatives have a positive effect on CSR strategies, as do directors appointed by investment funds, pension funds, and mutual funds, while directors appointed by banks and insurance companies have no impact on CSR strategies. This analysis offers new insights into the role played by certain types of directors on CSR strategies.


Introduction
Corporate social responsibility (CSR) can be defined as "a concept whereby companies integrate social and environmental concerns in their business operations and in their interaction with their stakeholders on a voluntary basis" (European Commission, 2001, p. 7). This description underlines the idea that firms voluntarily become involved with society and the environment to enhance and improve them. As noted by Santos (2011), this definition of CSR combines diverse objectives that must be integrated into the management process and go hand in hand with the objective of achieving positive economic outcomes. expected that they differ. This investigation is performed in Spain, which gives a different setting from other countries because, among other reasons, there is an important presence of controlling shareholders in firms, mainly institutional investors. An analysis of the role performed by institutional directors on CSR reporting, in an environment where institutions are important controlling shareholders, merits attention.
It is important to remark, in line with Font, Walmsley, Cogotti, McCombes, and Häusler (2012) and Bowen and Aragon-Correa (2014), that CSR disclosure and CSR commitment do not have to go hand in hand. In this sense, the fact that firms disclose CSR information about their environmental and social programs does not mean that they have to be responsible or commit to CSR, since it may be likely that they are only hoping to improve their reputation.
This article contributes to the literature in several respects. First, our evidence reinforces the view that institutional directors protect the interests not only of the shareholders but also of all stakeholders of the firm. Our findings show that institutional directors on boards encourage CSR disclosure; it is not only the shareholders who can benefit from this but also all the other stakeholders, since CSR information is public and is accessible to both shareholders and stakeholders. Second, we fill the gap in the prior research concerning the role played by institutional directors in firms' CSR disclosure, since we show that institutional directors cannot be considered a uniform group: pressure-resistant and pressure-sensitive institutional directors behave differently with respect to CSR reporting. Third, we support the thesis that directors who are oriented toward the long term, such as pressure-resistant directors, are more likely to promote CSR disclosure, given the active monitoring role they play and because they are concerned about their reputation. Finally, our analysis provides an index that measures the level of CSR disclosure of a company, and this may be useful for NGOs, CSR observers, practitioners, and researchers in determining which firms are more proactive in disclosing information on CSR.
The article proceeds as follows: In the following section, we focus on the institutional setting; Theoretical Background and Hypotheses section describes the theoretical background and develops the hypotheses; Empirical Design section describes the sample, methodology, and variables used in the study; Results section provides the analysis of the findings; and the final section discusses the findings and conclusions, the limitations inherent to this study, and future lines of research.

Institutional Setting
In the past few decades, Spain has increased its interest in CSR policies because society is pressing companies to engage in sustainable production. For this reason, a growing number of CSR reports, guidelines, observers, studies, indicators, and indexes have exerted an influence in changing Spanish business culture.
The Spanish Conthe Code (CUBG), issued in 2006, is a key aspect of this change. The Conthe Code (CUBG, 2006) provides three specific recommendations concerning CSR. The first makes reference to the inclusion of CSR disclosure in the corporate governance rules and internal codes of conduct, and states that this is supervised by a committee. The second refers to the minimum subject matter to be included in the CSR disclosure. In particular, the CSR disclosure should include details of specific practices on issues related to shareholders, employees, customers, suppliers, social and environmental matters, diversity, fiscal responsibility, respect for human rights and prevention of illegal behavior, and monitoring mechanisms for financial risk, ethics, and business conduct. The third states that CSR issues should be reported in the management report, or in a separate report, using any of the different internationally accepted methodologies.
The most relevant features of Spanish corporate governance are the low level of board independence, the lack of active market control (the supervision of the capital markets by the government and financial authorities is weak because the most important source of finance for companies is the banking sector, and not the capital markets), the presence of dominant shareholders on boards, a high ownership concentration and family ownership, a one-tier board system (all the directors, both executive and nonexecutive, sit on the board), and a low level of legal protection for investors. In this context, dominant shareholders take important positions on the boards of Spanish listed firms, and exert a strong influence on the management because of the high ownership concentration and the weakness of investors' protection. Institutional investors are among the most important controlling shareholders in Spain and, consequently, they have an important influence on corporate governance (e.g., Crespí, García-Cestona, & Salas, 2004). Indeed, the boards of Spanish listed firms are characterized by the fact that 40% of directors are institutional directors appointed by institutional investors; this is the highest percentage among European countries (Heidrick & Struggles, 2011).
Thus, Spain offers a good setting in which to examine the role played by institutional directors, given their relevant presence on boards and given the little empirical evidence about the influence of institutional investors on CSR disclosure when they act as board directors, perhaps because the presence of institutional directors on boards appointed by institutional investors in the United States and the United Kingdom is less common.

Theoretical Background and Hypotheses
Stakeholder theory posits that a company should consider all the stakeholders that may affect or be affected by its activities (Freeman, 2010). This theory supports a link between business ethics and organizational management, as a result of which firms have to take into account not only their shareholders but also a wider range of stakeholders, such as customers, the local community, and their employees, among others. Therefore, firms have to pay attention not only to their economic responsibilities but also to the stakeholders' interests.
Most sustainable management research is based on the stakeholder perspective (e.g., Montiel & Delgado-Ceballos, 2014). To the extent that boards are responsible for setting the values and standards within their organizations, companies interested in integrating CSR policies and addressing the interests of different stakeholders may have to adapt their board composition and functioning to this new role. Authors such as Arjoon (2005) support the notion that directors have to demonstrate greater ethical responsibility and a stronger ethical attitude when conducting business than when they are not directors, because they represent not only the shareholders but also the other stakeholders of the firm, and because society is putting pressure on them following the numerous scandals that came to light in the wake of the economic crisis. Walls, Berrone, and Phan (2012) also stress the role of boards in relation to environmental performance, measured in two ways-as the addition of the values of the items of strength or as the addition of the values of the items of concern. In particular, they report that environmental performance is better if the board is more diverse, less independent, and smaller. Thus, boards meeting these conditions are effective monitoring tools for environmental performance. In the same vein, Walls and Hoffman (2013) highlight the importance of the board of directors in supervising managers and affecting the firm's strategic orientation. These authors show that boards with greater environmental experience are more likely to diverge in a positive direction from institutional norms. Jain and Jamali (2016) argue that the board of directors is a "rich source of knowledge and guidance" and can, therefore, challenge managers to implement certain actions promoting the firm's CSR. Galbreath (2017) posits that directors on boards are supposed to move toward the promotion of actions associated with CSR. Concretely, Galbreath (2017) proposes that board composition is a very important issue when a firm's CSR is explored, since board members are a heterogeneous group because of their differing incentives and temporal orientations. The author finds that insider directors show a short-term orientation that is contrary to the long-term orientation needed for CSR and, thus, that insiders have a negative impact on CSR reporting.
In continental European countries, particularly Spain, where listed firms are characterized by high ownership concentration, two types of directors are distinguished in their boards of directors: insiders (executive) and outsiders or nonexecutive. Executive directors are vested with wide managerial powers and guide the firm's management team, while nonexecutive directors are mainly tasked with monitoring functions. The boards tend to be numerous, mainly in big companies, and among nonexecutive directors, two kinds of directors can be considered: those representing the owners or controlling shareholders (institutional directors) and independent directors (a more detailed discussion of this matter can be found in Ferrarini and Filippelli [2013]). Whereas executive directors are insiders and are directly involved in the management of the firm, both independent and institutional directors are considered outsiders, with different agendas and incentives in terms of controlling managers. In this sense, independent directors are members who are "free of any business, family or other relationship with the company, its controlling shareholder or the management of either, that creates a conflict of interest such as to impair its judgment" (European Commission, 2005). Furthermore, independent directors are board members appointed in view of their recognized professional expertise, experience, and prestige and are neither executive nor institutional directors, while institutional directors are those members who represent dominant institutional investors in the company that are involved somehow with the company and can influence the management of the firm.
Institutional directors have received attention from previous empirical research (e.g., Bhojraj & Sengupta, 2003;López-Iturriaga et al., 2015). Authors such as Hartzell and Starks (2003) suggest that institutional investors perform a relevant monitoring role and Parrino, Sias, and Starks (2003) emphasize their capacity to discipline managers and their active behavior in response to the corporate governance problems of firms. These characteristics involve an activism that is known as institutional shareholder activism. In this regard, Eesley, DeCelles, and Lenox (2016) show that the activism of investors, such as institutional investors, is an important matter to take into account when examining social changes in companies, since these investors may press firms to report CSR information. Thus, it is logical to assume that firms with institutional directors on their boards are more likely to be transparent and, hence, to encourage CSR reporting, as the costs associated with the monitoring role of institutional directors will be lower in companies that disclose CSR issues. Furthermore, institutional directors tend to be more aware of social and environmental demands, taking responsibility in these areas (Harjoto & Jo, 2011). In other words, these directors take an active role and, thus, may influence CSR reporting as they are interested in their reputation.
Furthermore, continental European base research, especially that conducted in Spain (e.g., García-Meca & Sánchez-Ballesta, 2009;Pucheta-Martínez & García-Meca, 2014), shows that independent directors on boards, in contrast to the United States and United Kingdom, appear to be less effective in carrying out the theoretical role of monitoring and, therefore, the supervising role on boards is not played by independent directors but by directors who represent controlling shareholders such as institutional investors.
Agency theory argues that institutional directors have incentives to act independently of managers and, therefore, will protect the interests of all the stakeholders, particularly shareholders, to mitigate agency costs (Colpan & Yoshikawa, 2012). Institutional directors should react to the expectations of the stakeholders, and institutional directors would therefore be expected to be more committed to CSR reporting demands, given that it might reinforce the internal control of companies and might decrease opportunistic behaviors due to asymmetric information, as Frias-Aceituno, Rodríguez-Ariza, and García-Sánchez (2013) suggest. Furthermore, CSR disclosure is a long-term responsibility of a firm and, accordingly, companies with institutional directors, who are usually associated with a long-term orientation, will have an interest in the long-term sustainability of the company and will be more likely to disclose CSR issues because they are necessary for long-term value creation and sustainable firm performance (Mahapatra, 1984). In this vein, the prior literature (e.g., Zattoni, 2011) shows that institutional directors exert a positive impact on CSR disclosure.
Likewise, from a legitimacy perspective, firms need to be legitimated by society (Patten & Crampton, 2004) in order to successfully run the business. If society notices that firms behave in an inappropriate manner, such legitimacy can be threatened, so to guarantee it, firms carry out legitimation strategies such as CSR reporting. In this regards, CSR disclosure becomes a mechanism for acquiring legitimacy and for protecting a firm's identity and reputation (Panwar, Paul, Nybakk, Hansen, & Thompson, 2014). As a result, given that institutional directors are concerned with their reputation, they will be more likely to disclose CSR information. Additionally, CSR disclosure is viewed by institutional directors as crucial to attain competitive advantage and sustainability, as it can help companies achieve legitimacy through support from different stakeholders (Neubaum & Zahra, 2006).
In line with agency and legitimacy approaches, the signaling perspective posits that in situations of asymmetric information, firms signal efforts to disclose sustainability matters in order to reduce this asymmetry and to assure legitimacy in society (Connelly, Certo, Ireland, & Reutzel, 2010). CSR and sustainability issues may be viewed as such asymmetric information because not all stakeholders have access to this type of information. Hence, institutional directors will support decisions such as CSR disclosure in order to send signals to stakeholders. In this way, institutional directors may send a signal to third parties communicating companies' unobservable features, such as their commitment to environmental and social matters, thus preserving their reputation. Additionally, if stakeholders value these unobservable characteristics, they might reward firms (Ramchander, Schwebach, & Staking, 2012).
Based on the above arguments, we posit the following hypothesis: Hypothesis 1: CSR disclosure is positively affected by institutional directors.
As stressed above, institutional directors (those who represent institutional investors on boards) are considered to be a heterogeneous group since their interests and incentives are dissimilar. In line with this, earlier research (e.g., Ramalingegowda & Yu, 2012) shows that institutional directors do not behave in a monolithic way, as they represent institutional investors who may or may not maintain business ties with the firm in which they invest and, thus, they may have different attitudes with regard to antitakeover amendments, CEO compensation and CSR disclosure, inter alia. Such directors can be separated into two groups: pressure-sensitive institutional directors (those representing banks and insurance firms, who both invest in the firm and maintain commercial ties with it) and pressure-resistant institutional directors (those representing mutual funds, investment funds, pension funds, and venture capital firms, who only invest in the firm). Huse (2007) also argues that the impact of institutional directors on boards depends on the type of ownership that they represent in companies. In this sense, while pressure-resistant institutional investors are oriented toward the long term, pressure-sensitive institutional investors are oriented toward the short term. The long-term orientation of pressure-resistant directors may explain why they are concerned about their reputation and, consequently, why they are more likely to reduce agency problems and to encourage social issues such as CSR disclosure. On the other hand, maximizing corporate performance and the expansion of their own businesses to achieve greater benefit are the results of the short-term orientation of pressure-sensitive directors. Their commercial ties with the firm may cause them conflicts of interest and, therefore, they will be less likely to encourage changes or to mitigate agency costs. Hence, the temporal horizon will determine the behavior of both types of institutional directors in their supervision of the management team in relation to reducing agency problems and disclosing CSR information. This idea is supported by Jain and Jamali (2016), who argue that pressure-sensitive investors such as banks have a shortterm investment perspective and will be less interested in supporting CSR activities because these practices show their benefits in a long-term horizon; thus, pressure-sensitive investors consider the costs of CSR activities unnecessary. Meanwhile, pension funds (pressure-resistant investors) will be more likely to support CSR firms' disclosure because they have a long-term investment approach (more detail about this point can be found in Jain and Jamali's [2016] article).
Pressure-sensitive institutional investors, as underlined above, tend to maintain both business and investment links with companies and, in order to preserve their business ties, may not be disposed to take issue with management decisions. Thus, because their profitability depends on it, pressure-sensitive institutional investors will tend to maintain a solid and excellent relationship with the firm. This view is supported by Almazan et al. (2005), who show that pressure-resistant directors can provide more intense managerial monitoring than pressure-sensitive investors because they are more active. According to Huse (2007), when pressure-sensitive institutional directors dominate the board, it will tend to align with managers rather than to control them.
The prior literature (e.g., Almazan et al., 2005;David, Kochhar, & Levitas, 1998) emphasizes the hypothesis that banks and insurance companies seek their own profit because of the design of their incentive systems. In this vein, Johnson and Greening (1999) find that pressure-sensitive institutional directors will try to put pressure on the company to adopt a short-term orientation, given their short-term profitability perspective. Hence, these authors suggest that such directors face more conflicts of interest in their relationship with the firm; thus, they will be less concerned about monitoring managers or CSR matters, for example. This idea is consistent with those of Eng (1999), who reports that insurance companies have a short-term horizon and, consequently, may not be interested in safeguarding the interests of stakeholders.
Within the category of pressure-sensitive institutional investors, banks are the most common, and pressure-sensitive institutional investors may, thus, become both shareholders and creditors. They may perceive that this dual role gives them access to more information than other types of shareholder, and they may use this information for their own interests. This may encourage the formation of controlling coalitions between such types of institutional investor and managers or other stakeholders, creating corporate groups to extract private benefits (Ali, Chen, & Radhakrishnan, 2007). Pressure-sensitive institutional directors will be less effective in monitoring managers, since they can be expected to align themselves with managers as they do not want to lose their business or potential business. Therefore, pressure-sensitive institutional directors may behave in the interests of those whom they represent, and not in the interests of the company's stakeholders. As a result, they will be less likely to require the disclosure of CSR information. This suggests a negative association between pressure-sensitive institutional directors and CSR disclosure, consistent with the collusion hypothesis, which argues that some directors may have incentives to collude with managers. This negative association would mean that pressuresensitive institutional directors would not press managers to be involved with CSR reporting.
On the other hand, pressure-resistant directors, who represent institutional investors with no business links but only investment ties to the firm and may be mutual funds, investment funds, pension funds, or venture capital firms, play a monitoring and disciplinary role on boards (López-Iturriaga et al., 2015). In contrast to pressure-sensitive institutional directors, pressure-resistant directors do not face a conflict of interests, as those they represent do not maintain any business relationship with the firm. Accordingly, these directors are able to discipline managers. Furthermore, Finkelstein (1992) argues that public pension funds and mutual funds, among others, have no "fear of retribution" from the firm's managers because they can be involved in corporate governance matters without being influenced by them.
Pressure-resistant institutional directors will prefer sustainable firms because the costs associated with monitoring the managers of such firms are lower. Furthermore, Johnson and Greening (1999) find that pressure-resistant institutional directors have a positive effect on CSR issues. More concretely, these authors provide evidence that pension funds have a positive impact on product quality (which includes the environment), the community, human talent matters, gender diversity, and minorities. Given that the benefits related to CSR disclosure are derived in the long term, and pressure-resistant directors on boards are characterized by a long-term horizon, the integration between these directors and CSR disclosure is relevant. This argument is also supported by Sethi (2005), who shows that pension funds perform a very relevant role in corporate governance, as they try to ensure that firms have a long-term perspective, which may encourage more CSR firms' disclosure.
Furthermore, Cotter and Najah (2012) provide evidence that large institutional investors, such as pension funds, press firms on climate change issues. According to this evidence, the pressure exerted by their stakeholders means that firms must be concerned about CSR matters if they are to be competitive. Consequently, we expect that pressure-resistant institutional directors will have a positive impact on CSR disclosure, in line with the supervision (contest or monitoring) hypothesis, which argues that some directors have the motivation to supervise the management team. Such directors will be concerned for their reputation and, accordingly, will persuade the management to report on CSR issues.
Consequently, we posit the following two hypotheses: Hypothesis 2: CSR disclosure is negatively affected by pressure-sensitive institutional directors. Hypothesis 3: CSR disclosure is positively affected by pressure-resistant institutional directors.

Sample
Our sample consists of listed Spanish nonfinancial firms from 2004 to 2013. Financial entities have been excluded from the sample for several reasons: (a) because they are under special scrutiny by financial authorities that constrain the role of their board of directors and (b) due to the special accounting principles and rules that they have to comply with when preparing their financial statements, which make it more difficult to compare them with those of nonfinancial firms since they are not homogenous. Our final sample, after excluding financial entities, is composed of an unbalanced panel data set of 1,018 firm-year observations. Our panel is unbalanced because during the period of the study some companies were acquired or merged and, consequently, were delisted, while other companies went bankrupt or became public companies. According to Arellano (2003), the findings provided by an unbalanced panel are as reliable as those reported for a balanced panel and, therefore, there is no concern regarding bias derived from missing data. The information used to construct the CSR index was collected from web pages of firms (annual sustainability reports) and the Global Reporting Initiative website, on which the annual Global Reporting Initiative report of companies that have voluntarily disclosed CSR information can be found. Financial data (total assets, total debts, and earnings before interest and taxation for measuring the variables leverage, return on assets, and firm size) were collected from the financial statements, and corporate governance data (the proportion of institutional, pressure-sensitive, and pressure-resistant directors, the proportion of stocks owned by insiders, board independence, and board size) were obtained from the annual corporate governance reports. Sector data were also collected from the financial reports. Both financial and corporate governance reports must be annually disclosed by listed firms in the Public Register of the Spanish Securities Market Commission (CNMV).

Variables
Our dependent variable, CSR disclosure or reporting, is defined as the CSR_Index and is created using social, environmental, and stakeholder information, since CSR is a multidimensional concept (Peloza & Shang, 2011). We base our CSR_Index on Cuadrado-Ballestero et al. (2015), who focus on the three types of information mentioned above to construct their CSR disclosure measure, which is also consistent with the recent articles of Jaggi, Allini, Macchioni, and Zagaria (2018) and Liu and Zhang (2017), who analyze the voluntary disclosure of carbon information among Italian firms and of CSR information among Chinese companies, respectively. In line with these articles, we also employ content analysis to capture CSR information.
We examine the three matters mentioned above as follows. First, we consider the social information reported by Spanish listed companies. Thus, we analyze information in sustainability reports regarding social labor performed by firms, using their resources (such as financial or human resources) to alleviate social problems or to help underprivileged social groups. In this vein, some companies may, for example, allocate a percentage of their profits to social activities or allow employees to dedicate part of their time to cooperating with NGOs without reducing their salaries. Second, we analyze the environmental information disclosed by the listed firms, taking into account whether they disclose information in their sustainability reports about the impact of their activities on the environment and how they alter their activities if they do affect the environment, their environmental management system, and their policies and environmental commitments. Finally, regarding stakeholder information disclosure, we bear in mind the level of commitment undertaken by the firm. In relation to this, we explore issues such as management systems for general interest groups, gender diversity, development of employees, human rights matters, and relationships with customers and suppliers. To the best of our knowledge, no prior research conducted with Spanish listed firms has, in contrast to our analysis, considered all economic sectors in order to construct the CSR disclosure measure.
For each area (social, environmental, and stakeholder information), a company is awarded one point if it gives complete information (meaning that the company reports information on every item we have considered for that particular area). If a company does not disclose any information about the areas we take into account, it receives zero points. A company receives half a point if it partially discloses information. For instance, if in the environmental area we take into account four items and the firm discloses information about one, two, or three of them, this is considered partial disclosure.
Thus, the CSR_Index is calculated as follows: CSR Index score social area score environmental area score sta As can be seen, we divide the total score for the firm by 3, which is the maximum score that a firm can attain.
Then, using the CSR_Index, we classify the firms in the following way: If a firm has a CSR_ Index of 0, the firm's CSR disclosure is null; a CSR_Index between 0.1 and 0.5 means that the firm's CSR disclosure is moderate; a CSR_Index between 0.6 and 0.9 represents an acceptable level for the firm's CSR disclosure; and a CSR_Index equal to 1 indicates that the firm's CSR disclosure is complete in respect of the issues we take into account.
Some authors, such as Kim and Lyon (2011), Maxwell (2011), andMarquis, Toffel, andZhou (2016), base their investigations on selective environmental disclosure or, in other words, on how companies can create a false good image by divulging their positive actions, while hiding their negative impacts on the environment. Toffel (2005) explores why firms voluntarily report regulatory environmental violations, while Reid and Toffel (2009) examine how companies behave under shareholder activism that respects the disclosure of corporate social actions, specifically considering the disclosure of information focused on climate change strategies. Kayser, Toffel, and Maxwell (2014) analyze whether firms voluntarily join environmental and social programs with lax access conditions. More precisely, they analyze whether stakeholder scrutiny encourages companies to participate in the Global Compact program when their prior practices are consistent with the program's principles. In comparison with these articles based on environmental issues, our measure is more complete since it not only takes into account the disclosure of environmental information but also considers social and stakeholder information reported by the firms. In this way, our index provides a new insight into all the information disclosed by Spanish listed firms on CSR matters.
Several independent variables are used. The variable for the percentage of institutional directors, defined as INST, is calculated as the number of institutional directors on the board of directors as a percentage of the total number of board members. We also consider the representatives of pressure-sensitive and pressure-resistant institutional investors, that is, pressure-sensitive and pressure-resistant directors, respectively. We label the pressure-sensitive directors as SENSIT, and pressure-resistant directors as RESIST, and these variables are calculated as the total number of pressure-sensitive directors on the board as a percentage of the total number of board members, and the total number of pressure-resistant directors on the board as a percentage of the total number of board members, respectively.
Regarding control variables, we consider several factors that may affect the CSR_Index. First, we control for board independence (BINDEP), which is the total number of independent directors on the board as a percentage of the total number of board members. Hence, we expect a positive sign between BINDEP and CSR reporting. Second, leverage (LEV) is measured as the ratio of the book value of debt to the total assets. We predict a positive sign for this variable. Third, we control for profitability or performance, defined as the return on assets (ROA), and calculate this as the ratio of earnings before interest and taxation (EBIT) to book assets. We expect a positive sign for ROA. Fourth, we control for firm size, labeled as SIZE, and measure as the log of the total assets (expressed in thousands of Euros). Consistent with Aragon-Correa, Martin-Tapia, and Hurtado-Torres (2013), we predict a positive association between SIZE and CSR reporting. Fifth, we consider board size (BSIZE), measured as the total number of board members. de Villiers, Naiker, and van Staden (2011) show a positive relationship between board size and CSR disclosure. Another control variable considered is insider ownership (INSOWN), calculated as the proportion of stocks held by inside directors. In line with Harjoto and Jo (2011), we predict a negative association between insider ownership and CSR disclosure. We control for the sector, calculated as a dummy variable that takes the value 1 if the firm belongs to the particular sector analyzed and 0, otherwise. Here, we draw on the CNMV classification, which considers the following sectors (labels given in parentheses): transport (TR; the reference category); cement, glass, and construction materials (CGCM); commerce and other services (COMER); construction (CONST); energy and water (EW); financial and insurance (FININS); chemical industry (CHIN); real estate (RE); mass media (MASSM); basic metal industries (BMI); other processing industries (OPI); metal processing industries (MPI); and, finally, the remaining sectors not included in any of the above categories and that consist of few firms (OTHR). According to García-Ayuso and Larrinaga (2003), some sectors are more likely to report CSR information than others, as their activities have a greater impact on society. Finally, we also control for firm and year fixed effects. Table 1 reports the descriptive statistics for all the variables. The data show that the CSR_Index is, on average, 0.382 out of 1 and, therefore, that the firms in our sample are moderate disclosers of CSR information. Furthermore, the statistics show that, on average, 32.38% of the directors on a board are institutional, with pressure-sensitive and pressure-resistant institutional directors accounting for 6.23% and 26.15%, respectively. In addition, 31.91% of the directors are Note. CSR_Index = the dependent variable, measured as the sum of the score of the three areas analyzed divided by 3; INST = the proportion of institutional directors on board; SENSIT = the proportion of the board directors who are representative of pressure-sensitive institutional investors (i.e., banks and insurance companies); RESIST = the proportion of the board directors who are representative of pressure-resistant institutional investors (i.e., investment funds); BINDEP = the proportion of independent directors on the board; LEV = the debt over total assets; ROA = the operate income before interests and taxes over total assets; SIZE = the log of total assets; BSIZE = the number of member on boards; INSOWN = the percentage of shares held by insiders (directors) on board and from TR to OTHR represent the sectors of the firms, measured as dummy variables that take the value 1 if the company belongs to the sector analyzed and 0, otherwise; TR = transport; CGCM = cement, glass, and construction materials; COMER = commerce and other services; CONST = construction; EW = energy and water; FININS = financial and insurance; CHIN = chemical industry; RE = real estate; MASSM = mass media; BMI = basic metal industries; OPI = other processing industries; MPI = metal processing industries; OTHR = the remaining sectors which not included in any of the above categories and that consist of few firms. Panels A and B show the continuous and dummy variables, respectively.

Descriptive Statistics
independent directors, on average, and the mean for insider ownership is 8.72%. Furthermore, the average firm size (the log of the total assets in thousands of Euros) is 13.25, the average leverage is 52.34%, the average board size is 10.77 members, and the average ROA is 3.55%. Finally, the transport sector accounts for 3.34% of firms; cement, glass, and construction materials 2.94%; commerce and other services 10.61%; construction 7.95%; energy and water 7.76%; the financial and insurance sector 2.85%; the chemical industry 3.14%; real estate 11.59%; mass media 3.73%; basic metal industries 4.32%; other processing industries 2.41%; metal processing industries 9.82%; and other sectors 7.76%.

Multivariate Analysis
The correlation matrix has been calculated to test for multicollinearity. The findings, unreported for the sake of brevity, show that all the coefficients are lower than 0.7, except for the pairs INST and RESIST, but this is not a problem because these variables do not interact together in any model. However, Hair, Black, Babin, Anderson, and Tatham (2006) suggest that the correlation matrix is not sufficient to check the lack of high correlation values among variables to evaluate the presence of multicollinearity. Accordingly, as Hair et al. (2006) indicate, we have also calculated the variance inflation factor (VIF) and the tolerance (1/VIF), provided in Table 2, where the regression findings are presented. According to Haan (2002), VIFs higher than 10 show multicollinearity concerns. In this study, the highest VIF is 7.21 and, therefore multicollinearity is not a problem. For estimating the models, we use a Tobit regression because this is designed to estimate linear relationships between variables when there is either left or right censoring in the dependent variable. In this article, the dependent variable, CSR_Index, is either 0 or a positive number and ranges from 0 to 1. Thus, given that the variable is censored on two sides [0-1], a Tobit regression is most suitable for handling the data. All the assumptions for a Tobit regression are fulfilled. According to Wooldridge (2010), the Tobit model with panel data can be estimated as a pooled or a random effects regression. The likelihood-ratio test of rho = 0 allows us to discriminate between pooled or random effects (Flanagan, 2014). Our findings of the likelihood-ratio test show that a random effects estimator is more appropriate than a pooled one. Furthermore, we have performed the likelihood-ratio test for heteroskedasticity and the Wooldridge test for autocorrelation in panel data models. While the likelihood-ratio test shows that heteroskedasticity is not a concern in our model, the Wooldridge test finds that autocorrelation is. Thus, autocorrelation problems have been corrected by using the bootstrap standard errors method, which clusters residual standard errors (e.g., Gaynor, Li, & Vogt, 2006). Table 2 shows the results for Model 1, in which the effect of institutional directors on the CSR_Index is analyzed. As expected, the proportion of institutional directors exhibits a positive sign, and this is statistically significant. Thus, the first hypothesis cannot be rejected. Our results suggest that institutional directors on boards have a positive impact on CSR disclosure; this is consistent with the findings of authors such as Cox, Brammer, and Millington (2004), Neubaum and Zahra (2006), and Zattoni (2011), who find a positive relationship between institutional directors and CSR matters. In this vein, our findings reinforce the argument that institutional directors have a long-term horizon, play an effective monitoring role in relation to managers, and are concerned about their reputation; for these reasons, they press managers to disclose CSR information. This evidence supports the supervision hypothesis, because institutional directors will prefer companies that are more involved with CSR disclosure to monitor the managers more efficiently. Therefore, firms with more institutional directors on their boards are more likely to report CSR information. Table 2 also reports the findings for Models 2 and 3, in which we analyze how pressure-sensitive institutional directors and pressure-resistant institutional directors, respectively, affect CSR reporting.
Model 2 in Table 2 shows that the percentage of pressure-sensitive institutional directors, as predicted, shows a negative sign; however, this result is insignificant. Thus, we reject the second hypothesis. Consequently, this finding suggests that pressure-sensitive institutional directors do not have an impact on CSR reporting, and this result is in line with the findings of Johnson and   Note. VIF = variance inflation factor; INST = the proportion of institutional directors on board; SENSIT = the proportion of the board directors who are representative of pressuresensitive institutional investors (i.e., banks and insurance companies); RESIST = the proportion of the board directors who are representative of pressure-resistant institutional investors (i.e., investment funds); BINDEP = the proportion of independent directors on the board; LEV = the debt over total assets; ROA = the operate income before interests and taxes over total assets; SIZE = the log of total assets; BSIZE = the number of member on boards; INSOWN = the percentage of shares held by insiders (directors) on board and from TR to OTHR represent the sectors of the firms, measured as dummy variables that take the value 1 if the company belongs to the sector analyzed and 0, otherwise; TR = transport; CGCM = cement, glass, and construction materials; COMER = commerce and other services; CONST = construction; EW = energy and water; FININS = financial and insurance; CHIN = chemical industry; RE = real estate; MASSM = mass media; BMI = basic metal industries; OPI = other processing industries; MPI = metal processing industries; OTHR = the remaining sectors which not included in any of the above categories and that consist of few firms. *Significant at 90% confidence level. **Significant at 95% confidence level. ***Significant at 99% confidence level.
Greening (1999), who found that pressure-sensitive institutional directors did not have an effect on CSR disclosure. However, our results, contrary to our expectations, do not support the collusion hypothesis for this type of director. According to Johnson and Greening (1999), the fact that these directors frequently change their portfolios, and that the long-term profits derived from firms' CSR activities are not of interest to them, may explain the lack of impact of pressuresensitive institutional directors on CSR disclosure. Model 3 in Table 2 reports that the proportion of pressure-resistant institutional directors exhibits a positive sign, as predicted. Moreover, this result is statistically significant. Hence, the third hypothesis cannot be rejected. Accordingly, our findings suggest that pressure-resistant institutional directors have a positive effect on CSR disclosure, which is in line with the findings of Johnson and Greening (1999) suggesting that pressure-resistant institutional directors are more likely to encourage the disclosure of social, stakeholder, and environmental information. This evidence supports the idea that pressure-resistant institutional directors prefer firms that are more committed to CSR reporting because the costs associated with monitoring managers in such firms are lower. In addition, pressure-resistant directors are concerned with their reputation, have a long-term perspective, and prefer firms that are more involved with CSR matters. This result supports the thesis that pressure-resistant institutional directors have incentives to encourage managers to disclose information about CSR activities, which is consistent with the results of Sethi (2005) and with the monitoring hypothesis.
Concerning the control variables, board independence (BINDEP) and firm size (SIZE) are positive and statistically associated with CSR disclosure in all models displayed in Table 2. The control variable insider ownership (INSOWN) shows a negative sign, according to our expectations, and this is statistically significant in all models. Finally, our findings demonstrate which sectors disclose more or less CSR information compared with the transport sector (TR-the reference category). In particular, compared with TR, the construction (CONST) and basic metal industries (BMI) sectors show a positive sign on CSR disclosure, and the results are statistically significant. The commerce and other services (COMER), the energy and water (EW), the mass media (MASSM), and other processing industries (OPI) sectors also present a positive sign in comparison with the transport sector, but the findings show that the COMER, MASSM, and OPI sectors are only statistically significant in Models 1 and 3 and for the EW sector in Model 1. Therefore, the aforementioned sectors positively affect CSR reporting, which is a logical consequence of the high environmental and social impact of their activities. Thus, compared with the transport sector, all these sectors disclose more CSR information. By contrast, compared with TR, the real estate (RE) sector is negatively associated with CSR reporting in a statistically significant way, which implies that the real sector reports less CSR information than the transport sector. The remaining variables are insignificant.
We also address the issue of possible endogeneity between our independent variables (INST, SENSIT, and RESIST) and CSR disclosure. The direction of causality between the independent variables and CSR reporting is more likely to go from the directors analyzed in the article to CSR disclosure, although it is also likely that CSR disclosure has an effect on board composition. This matter is addressed by lagging the independent variables in our models. For the sake of brevity, these findings are not shown, but they are in line with the core results of our analysis.

Sensitivity Analysis
We check the robustness of our models using, as the dependent variable, the ratio between the CSR_Index and the firm size (SIZE), since firm size is a factor that has a positive effect on CSR reporting, as our results show: Bigger companies are more likely to disclose CSR information than smaller ones. The results, which for the sake of brevity are not provided, demonstrate that the proportions of institutional directors (INST) and pressure-resistant institutional directors (RESIST) exhibit a positive sign, as expected, and are statistically significant. Regarding pressure-sensitive institutional directors (SENSIT), the results are insignificant. Hence, this evidence corroborates our results and, accordingly, we can affirm that our findings do not depend on the measure of CSR reporting.

Discussion and Conclusions
Our article tries to fill the gap in the literature concerning the role played by institutional directors in firms' decision-making processes in relation to CSR disclosure. To deepen the analysis of institutional directors and their impact on CSR disclosure, we also distinguish between institutional directors who represent institutional investors with business ties to the company and institutional directors who represent institutional investors that do not maintain such business links.
Our results reveal that institutional and pressure-resistant directors have a positive effect on CSR disclosure. This finding supports the monitoring hypothesis, stressing the importance of reputation for these directors, in line with their long-term perspective. Their preferences for longterm investments and for mitigating agency costs support the view that institutional and pressureresistant directors may have sufficient power to supervise managers and activities, to reduce opportunistic behavior, to influence company decisions in favor of stakeholders' interests, and to foster changes in companies such as the widening of CSR reporting to include not only economic matters but also social and environmental ones. In this vein, it will be more complicated for managers to collude with institutional and pressure-resistant directors than with other directors since these institutional directors might be interested in sharing supervision costs, resulting in a more active and efficient supervisory role. Furthermore, institutional and pressure-resistant directors on boards may avoid the temptation to join other institutional and pressure-sensitive directors in taking part in tunneling activities (i.e., expropriating wealth from minority owners), since managers might influence them to gain their support and, in return, these directors might meet their aims. The more active supervision of managers' activities by these directors might prevent tunneling activities and might challenge the boards of directors and managers to disclose CSR issues. This is because institutional and pressure-resistant directors may link the competitive advantages with the benefits provided by CSR; this will lead to the long-term sustainability of the company and an improvement in company performance. Provided that pressure-resistant directors represent institutional investors who do not maintain commercial ties with the firms, they do not face any conflicts of interest arising from such business relationships and, consequently, they can behave in a more independent way. The firms in which the investors have invested can exert less pressure on such directors.
Contrary to our expectations, pressure-sensitive institutional directors do not have an impact on CSR disclosure, suggesting that banks and insurance companies hold short-term positions and, thus, simply seek to enhance their own profit. This result may be explained by arguing that supervision costs are high and that pressure-sensitive directors would have to bear these costs if they engaged in supervision, while sharing the benefits of the supervision with all the other stakeholders. This will discourage pressure-sensitive directors from becoming involved with monitoring any activities that could increase the CSR reporting. Moreover, pressure-sensitive directors might influence the decision-making process for their own benefit, and in order to achieve their own aims they might prefer to align themselves with the company. Thus, it is reasonable to suggest that pressure-sensitive directors may be interested in taking part in tunneling activities rather than supervision activities. Consequently, pressure-sensitive directors might not be interested in promoting CSR reporting, even though CSR disclosure may increase firm value. Following this line of argument, pressure-sensitive directors will be more interested in obtaining private benefits for the investors they represent than in enhancing firm performance, and more interested in monitoring other pressure-sensitive directors to prevent them from achieving their own purposes.
Pressure-sensitive directors will align themselves with managers and will be less willing to confront them with CSR reporting; this is a result of the conflicts of interest faced by pressure-sensitive institutional directors because of the business ties between the firm and the pressure-sensitive institutional investors they represent. If these directors do not align themselves with managers, they might compromise these commercial links. Hence, pressure-sensitive directors may collude with managers, aligning themselves with their views and decisions, such as decisions not to disclose CSR information, and in this way, pressure-sensitive directors may meet their own aims and preserve the commercial relationships. Additionally, pressure-sensitive directors tend to be interested in short-term earnings and, consequently, may support activities that increase shortterm profitability; CSR reporting is not one of these activities because its benefits are displayed across a long-term horizon. Thus, the supervision costs, the conflict of interest felt by pressuresensitive directors and these directors' short-term orientation support the absence of influence of these directors on CSR disclosure. This research has significant implications. Our results offer a new insight into the positive role played by institutional directors, particularly pressure-resistant directors, on CSR disclosure and may, therefore, help policy makers promote CSR reporting in the Spanish context, where almost 40% of board members are institutional directors. Hence, this result has important implications for both public policy and the governance of firms. When there are differences between institutional directors, especially pressure-resistant directors, and other experienced directors, then the interests of the institutional directors would mainly become a public policy matter. Regulatory bodies may recommend board structures with institutional directors, particularly pressure-resistant directors, not only for listed firms but also for nonlisted firms, since institutional directorship might encourage firms to disclose CSR information, which may be of benefit to all the stakeholders. Noncompliance with these recommendations may be perceived negatively by the capital markets and other funding sources, which may encourage companies to disclose CSR information. Second, our results suggest that institutional directors and pressure-resistant directors may challenge boards and managers to report on CSR issues, in line with the monitoring hypothesis and, consequently, these directors focus not only on economic performance but also on environmental and social matters. Accordingly, this evidence supports the view that institutional directors and pressure-resistant directors might actively supervise managers and might abandon their conventional inactive role in the supervision of managers, which would have an effect on CSR reporting. Third, our findings offer important evidence in relation to earlier literature by examining whether institutional investors, when they appoint board directors, have an effect on CSR reporting. As a result, when academics examine the role played by institutional investors, they should address not only the number of shares held by institutional investors but also their contribution to other methods of corporate control such as board membership. Fourth, our evidence demonstrates that institutional directors should be considered as a heterogeneous group, because pressure-resistant directors have a positive impact on CSR reporting, while pressure-sensitive directors do not have any effect. Fifth, this article provides new insights for NGOs, which may consider our evidence when they help companies get a better understanding of the importance of institutional directors in firms' decisions to disclose CSR information; such disclosures are in the interests of the institutional investors and of society, and they are of special benefit to all stakeholders. Finally, further investigations into the role of institutional directors become essential to give a better understanding of how they engage in efficient corporate governance mechanisms. Although earlier and more recent evidence has shed some light on these matters, academics must go even more deeply into the implications and incentives of institutional directors in the governance of firms. This research has several limitations. First, despite taking into account as many factors as possible on the basis of theory and the prior empirical evidence, it is likely that there are unknown factors affecting CSR disclosure that have not been considered. Second, our research is based on Spanish listed firms from 2004 to 2013 and, therefore, our findings should not be extended to other periods. Third, although a significant percentage of directors on Spanish boards are institutional directors, within this category, there are fewer pressure-sensitive directors than pressureresistant directors. Finally, our CSR_Index, based on the prior literature, is composed of three areas and is measured using many factors, but it is possible that other characteristics having an impact on the three areas of our analysis have not been addressed.
Finally, this investigation leads to suggestions for future research. First, it may be interesting to analyze the role played by institutional directors in relation to CSR disclosure for Spanish financial entities, and for small-to medium-sized enterprises. Second, it would also be interesting to explore whether there are institutional directors appointed by NGOs on the boards of listed and nonlisted firms, and the role that these directors perform in relation to CSR reporting, given the work in society undertaken by NGOs. Finally, the number of female directors of listed firms is increasing, particularly in Spain, and consequently, the role played by female institutional directors in relation to CSR disclosure is also a relevant topic.