Since the establishment of the Sustainable Development Goals (SDG) in the United Nations (2030 Agenda), Corporate Social Responsibility (CSR) has become more relevant in business management. Furthermore, the growing pressure from stakeholders for companies to disclose more information on sustainability has motivated companies to adopt actions towards environmental and social impact (Van der Zahn & Cong, 2019). Policy changes between governments and regulators place a greater emphasis on sustainable development, which has intensified the pressure on companies to improve their non-financial performance (Gavana et al., 2017).

Prior literature affirms that family and nonfamily companies are leaded differently (Miller, Minichilli, & Corbetta, 2013; Naldi et al., 2007). For instance, the family commitment in tactical and strategic policies in family business is a common practical (Dudaroğlu, Öner, & Önday, 2018). Other strand of studies shows that family businesses adopt a culture focused on a strategic direction such as sustainable practices in the domain of environmental and social concerns (Zahra, Hayton, & Salvato, 2004). In the same line, the Chief Executive Officer (CEO)’s role has been reconfigured to prioritize social and environmental issues (Walls & Berrone, 2015). According to some studies, family firms are more interested in the promotion of social and environmental actions when top leaders support is high (Dobele et al. 2014); some results denote that the existence of a family CEO increases environmental and social engagement (Marques, Presas, & Simon, 2014).

This study aims to analyze if family CEOs enhances the environmental and social performance in family listed companies from Ibero-America compared to those companies with no-family CEOs. In addition, the potential moderating effect of the structure of the board (size, independence and women as board members) in this relationship has not been examined in prior research.

This work extends the comparative literature in Ibero America, a region scarcely addressed in prior research and makes several theoretical and practical contributions. In the first place, this investigation is motivated by the behavioral agency theory, which states that if the preservation of the SEW of the family is not aligned with the interest of the manager (agent), family businesses may not invest in enhancing their CSR achievement (Meier & Schier, 2021). In the second place, it is shown that CSR in family businesses should not be studied as a homogeneous event; the type of leadership shapes the motivations and interests of the company in regard to CSR practices. First, CSR could constitute an advantage by which family CEOs can insurance a family schedule or a personal agenda, so an alignment of interests is important. From this perspective, the CSR family agenda constitutes a lever of legitimation and supports the instrumental vision of CSR in family businesses (Windsor, 2006).

The research is organized as follows. Section 2 shows a review of prior literature focused on CSR performance in family firms. Section 3 describes research methods. Section 4 details the key empirical results. The last section discusses the conclusions and implications of the research.

CSR refers to the behavior of a company towards different groups of interest, such as the community, investors, government, employees, customers and suppliers, and is represented by those multifaceted activities that include social, ethical, environmental and political actions (Yoon et al., 2018). CSR is a multidimensional construct that incorporates policies and/or activities refined by companies to enhance their social, economic, and environmental impact behind legal concerns. The company's environmental behaviour argues the company's effort to diminish resource expending and wasting disease. Social dimension emphasizes human rights, job quality, product responsibility, and community relationships (Meier & Schier, 2021). The commitment to CSR can be attributed to its perceived contribution to the competitiveness and legitimacy of companies and for normative, moral or ethical reasons (Aguilera et al., 2007). According to Hsu and Cheng (2012), the actions aimed at strengthening CSR are the result of the decisions and the management carried out by the companies’ leaders.

Companies try to encourage consistency between their value system and the value scheme followed by the society, and tend to exhibit an appearance that harmonize with the assumptions of their groups of interest (Deegan, 2007). In the context of family businesses, empirical research shows that this type of companies prioritize social interests, in terms of detriment to interest groups, than their non-family analogues. This implies that these companies are very distrubedd with corporate prestige and social legitimation (Dyer & Whetten, 2006; Gavana, Gottardo, & Moisello, 2017).

Family businesses are defined by the strategic responsibility played by family representatives in organizational processes, and the degree to which family dynamics influence decision-making (Chua et al., 1999). In the family business, the CEOs’ decision-making options can be approached from the profit and loss perspective with regards to the SEW of the majority family members (Gómez-Mejía et al., 2007). That is, CSR is expected to be seen as a probable benefit by a family CEO, due to the importance of preserving the SEW of the founding family (Naldi et al., 2013), and it is possible that for a non-family CEO, it is seen as a potential loss, due to its cost and its negative influence on the prompt financial corporate performance (Wang et al., 2008).

As family businesses seek to establish their legitimacy with stakeholders, they need to adopt governance mechanisms through which they can increase their CSR compliance. For instance, nominating CEOs whose values ​​are aligned with preserving the SEW of the family business or compensating them to enhance such arrangement. It is evidenced in this work that family CEOs favor non-financial performance, which affects a greater legitimacy of companies in international markets. CSR practices can help companies gain legitimacy because it helps build business reputation (Panwar et al., 2014), and aligned with the sociocultural patterns of its institutional framework (Du & Vieira, 2012).

Family CEOs are more risk averse, and are more motivated in preserving family wealth, which could motivate them to undertake environmentally responsible actions (Lumpkin & Brigham, 2011). A recent study published by Meier and Schier (2021), concludes that family CEOs favor social performance, while non-family CEOs decrease it. Behavioral agency theory [BAT] integrates agency and socio-emotional wealth [SEW] perspectives as a point of reference for addressing possible conflicts between the agent and stakeholders (Chua et al., 2011). This theory states that family and non-family CEOs differ in their desire for the conservation of the SEW (Gómez Mejía et al., 2019). Although family CEOs have an intrinsic incentive to maintain the family SEW, such intentions are not naturally adopted by non-family CEOs (Gómez-Mejía et al., 2007). Therefore, they present differences in their preference for CSR strategies, which leads to the legitimacy of the family business in the society, a fundamental value for the preservation of the SEW.

Family businesses managed by family CEOs tend to show more concern for the local community in order for the business to succeed (Waddock & Graves, 1997). Anderson and Reeb (2003) conclude that companies run by family CEOs showed a greater commitment to the collective interests of stakeholders, including employees and community members (Deniz & Suarez, 2005). According to Choi and Wang (2009), family CEOs are more receptive to the image, corporate reputation and social legitimacy of the family, which translates into greater CSR activities and actions aimed at strengthening human capital and human rights. According to Nason et al. (2018), family businesses guided by family CEOs prioritize those aspects associated with social ties and intragroup norms. The CEO is one of the most important pillars for the implementation of a successful strategy against climate change, monitoring the participation of the company in strategic initiatives and supervising the environmental responses of companies (Peters and Romi, 2014; Walls and Berrone, 2015). Under the above discussion, we establish this hypothesis:

H1: Family CEOs favor to a greater extent the environmental and social performance of family listed companies from Ibero-America, compared to non-family CEOs.

In the same line, the agency theory affirms that governance bodies such as corporate boards should control and monitor the agents’ decisions, and enhance legitimacy through CSR practices (Uyar et al., 2021). The board of directors constitutes the main mechanism for corporate governance, being responsible for preserving the adequate interests of groups of interest of a firm through conducting its operation and by managing the decision-making process (Kaplan, 2001). The board is an important organ that induces strategies for top management and relevant decisions on corporate concerns (Kovská & Procházková, 2014). Robust good governance practices have been shown to mitigate agency problems and to strengthen managers to conduct accurately (Terjesen et al., 2015). In this context, recent literature have focused on the board of directors as the main managerial structure in a company (García-Sanchez & Martínez-Ferrero, 2018) and is frequently associated as a group of groups of interest within a firm that have comparable angles in corporate practices (Naciti, 2018). Therefore, the board composition plays a significant piece in resolving the firm’s CSR performance due to its authority in the process of the firm’s managerial decisions (Zhuang et al., 2018).

Prior research has found that more independent directors the voluntary disclosure of sustainability, while Hussain et al. (2018) argue that independent members in the board might be decisive in supervising and governing sustainable results. In the context of board diversity, Zhang et al. (2012) and Reyes-Bastidas and Briano-Turrent (2018) demonstrated that a high CSR performance is related with a significant representation of women on the board. By contrast, the separation of the Chair of the Board (COB) and the Chief Executive Officer (CEO) responsibilities increases the independence of the board from management, and therefore the CSR performance is increased (Michelon & Parbonetti, 2012). Therefore, we hypothesize:

H2: There is a significant moderator effect between family CEOs and the board composition that increases the environmental and social performance of family listed companies in Ibero-America.

Our sample consists of those family companies with the highest stock market listing that belong to the Stock Exchanges of Colombia (total companies listed on the Stock Exchange: 50 companies), Chile (IPSA: 33 companies), Spain (IBEX 35: 28 companies) and Mexico (IPyC: 31 companies) from 2011-2016. Table 1 shows the total observations per industry and per country during the period of analysis. Those companies listened in the banking and insurance sector were excluded because they are subject to greater regulatory supervision compared to other companies and their corporate information structure differs from that of other companies.

Table 1

Industrial sector/Country	Colombia	Chile	Spain	Mexico	Total
Oil and gas	24	6	30	4	64
Industrial	54	43	18	84	199
Consumer goods	78	36	12	66	192
Health care	12	0	6	6	24
Consumer goods	29	17	30	6	82
Telecommunications	6	6	8	6	26
Utilities	48	36	6	0	90
Financial (excluding banking and insurance)	24	36	15	0	75
Technology	0	0	12	0	12
Basic Materials	18	18	24	12	72
Total companies	50	33	28	31	142
Total Observations	293	198	161	184	836

This table describes the study sample, analyzing by industry type and country. The sample is composed by 836 firm-year observations for family listed companies in Colombia, Chile, Spain and Mexico over the period 2011-2016. We have adopted the Industry Classification Benchmark (ICB).

To measure CSR performance, an index based on two of the three dimensions of the Global Reporting Initiative (GRI) version G-4 is proposed: 1) Environmental dimension, which is related to the use of materials, water, energy, biodiversity, emissions, effluents and waste, products and services, transportation, environmental evaluation of suppliers, claim mechanisms in environmental matters, and 2) social dimension, referring to labor practices and decent work, human rights, society and product liability (Global Reporting Initiative, 2015).

The information on the study variables was compiled manually from each of the annual documents (financial and sustainability) of the companies that make up the study sample. Through the content analysis technique, keywords were searched to integrate the study variables (Samaha & Dahawy, 2011). To validate the reliability of the proposed CSR index, we have performed a correlation analysis between the elements that make up the qualitative dimensions of the index (social and environmental), through the standardized Cronbach's alpha, yielding an average greater than 0.80 in each dimension, which guarantees the reliability of the scale (Cronbach, 1951). Outliers or extreme values ​​were also treated in the case of financial variables, truncating the extreme values ​​to percentile 2.

The Generalized Method of Moments (GMM) is used as the main methodology to address possible problems of endogeneity and serial correlation of the study variables. The system-GMM proposed by Arellano and Bover (1995) and by Blundell and Bond (1998) control for simultaneity concerns, omitted variable bias and the dynamic relationship of family CEOs and CSR elements (Wintoki et al, 2012). Other firm specific characteristics that are unobservable or unmeasured in our model (e.g. the appointment of a new CEO) could result in an estimation bias. The GMM method is adjusted for data panels that have fewer time observations than individuals (firms). This methodology allows the lag of the dependent variable, which becomes an independent variable and serves as an instrument in the regression analysis (Poletti-Hughes & Briano-Turrent, 2019).

To analyze the relationship between the influence of family CEOs on the environmental and social dimensions of the CSR, the study variables have been grouped into two groups; the first group refers to the CSR performance, which is made up of 20 elements, of which 4 belong to the environmental dimension (20%) and 16 items for the social dimension (80%). The proposed index is based on the GRI-G4, so it is made up of a greater number of elements for the social dimension [labor legislation and decent work, human rights, health and safety problems, society and product responsibility] (Reyes-Bastidas et al., 2020). The elements of each index are dichotomous, that is, they take the value of 1 if the company adopts this practice and 0 if otherwise (see table 2). The control variables are related to the board composition (size, independence and female on the board); the structure of the board decreases the uncertainty of the company when CEOs associate the company with its external environment (Hillman et al., 2000). Another group of control variables are those related to profitability (RoA), level of indebtedness, size and age of the company (Bhagat & Bolton, 2008). According to the above, the following econometric models are established:

C S R ( E n v i , t )     =     α i   +   Σ β 1 ( C S R E n v i , t - 1 )   +   Σ β 2 F a m C E O   +   Σ β 3 B o a r d C o m p   +   Σ β 4 ( C o n t r o l i , t )   + µ i , t (1)

C S R ( S o c i a l i , t )   =   α 0   +   Σ β 1 ( C S R S o c i a l i , t - 1 )   +   Σ β 2 F a m C E O   *   B o a r d C o m p   +   Σ β 3 ( C o n t r o l i , t )   +   µ i , t (2)

Where i represents i-nth cross-sectional unit (company); t for the nth period of time (year), The dependent variable CSR is measured by the social and environmental dimensions. The primary independent variable is FamCEO, a dummy variable that takes a value of 1 if a firm is defined as a family-CEO firm and 0 otherwise (Briano-Turrent et al., 2020). The moderating variables (BoardComp) are focused on the board composition (size, independence and female on the board). Control is a vector for the control variables related to profitability, indebtedness, company size and board structure; µ _i,t is the error term.

Table 2

Source: Own elaboration based on the GRI-G4 elements (2015) and Reyes Bastidas & Briano Turrent (2018).

This research contributes to advance both the family firms and CSR literature. First, based on the BAT, the objective was to analyze the relationship between the participation of family CEOs and the environmental and social performance of listed family companies in four countries from Ibero-America (Colombia, Chile, Spain and Mexico), integrating the moderator effect of the board composition on this relationship. Second, our findings suggest that family CEOs adopt a higher CSR behavior to fulfill their objectives in terms of new legitimation needs and to prioritize the SEW of the family business. The role of the CEO has been transformed towards a sustainable vision, emphasizing more and more the implementation of strategies that respond to the social and environmental demands of the stakeholders. Hence, the nature of the CEO (family vs. non-family) and their interaction with corporate governance practices influence non-financial decision-making.

From a practical viewpoint, this research could be potentially interesting for policymakers and practitioners. First, given the high proportion of family firms in Latin America and the increasing trend of family firms engaging with CSR, this paper support that family CEOs in companies with strong corporate governance practices (for instance, independence on the board) are positively related with corporate social performance. The BAT affirms that a family CEO enhances the influence of family ownership on the firm’s CSR performance (Mariani et al., 2021).

This work has some limitations that could be addressed in future studies. First, the sample is limited to those companies with the highest stock market prices in four Ibero-American countries, so the results could not be generalized to other types of family businesses such as small or medium-sized (SMEs) ones. Second, the study variables are limited to the board composition, the nature of CEOs (family vs non-family) and social and environmental dimensions of CSR, excluding other corporate governance variables such as board support committees, ownership patterns, or CEOs or board members attributes. These limitations motivate more research on the change of non-economic family priorities. Additional research is required to understand how other institutional factors (number of generations, participation of new generations, ownership patterns, etc.) can motivate the formulation of non-economic objectives in family businesses. One of the great challenges would be to design CSR indices for these companies, given that the current CSR ratings proposed by several agencies tend to prioritize large public companies.