2.1 Theoretical Framework

This study's theoretical framework is based on agency theory, which is grounded on the idea that the objective of the shareholder is to maximize firm performance. To obtain this, shareholders tend to delegate the task to managers, and this separation between principal and agent in public firms causes information asymmetries and incomplete contracts between the two parties (Jensen and Meckling 1979). In this vein, corporate boards are considered a corporate governance mechanism that reduces information asymmetries between managers and shareholders (De Andres and Vallelado 2008) and ensures the alignment of their interests. Board composition is a relevant aspect of the supervisory ability of the board (Carter et al. 2010). In this sense, independent directors are non-executive board members who are not part of the company's staff and must remain free from any relationships or connections that could impair their independence, judgment, or decision-making (GGC 2015). They should also have no significant financial stake in the company beyond their responsibilities as directors.

Notwithstanding the importance of agency theory's insight, it fails in the consideration of other crucial stakeholders. Hence, we use a theoretical lens that extends agency theory in this direction, the stakeholder-agency theory (Hill and Jones 1992), which has also been called generalized agency theory (Prior et al. 2008) and multistakeholder agency theory (Chen et al. 2023). It is grounded on agency theory but fuses insights from stakeholder theory. Like the agency theory, this combined approach also views the company as a nexus of contracts between stakeholders. However, compared to traditional agency theory, the stakeholder-agency approach “broadens the scope of principals from shareholders as defined” in the former theory “to all stakeholders, assuming the agents […] as the hub that centralizes all contractual relationships with stakeholders” (Chen et al. 2023, p. 140).

Shareholders are not considered the only stakeholders who contribute to the company with critical resources and possess the power to contribute or refuse such resources. Hence, shareholders are not the only relevant stakeholders who influence managerial decisions. Many other stakeholders are depicted as being able to influence such decisions. Numerous stakeholders provide the company with critical resources and possess the power to contribute or retain such resources in an effective manner, acquiring different degrees of influence on managerial decisions. Stakeholders “supply the firm with resources on the implicit (tacit) understanding that their claims on the organizations will be recognized” (Hill and Jones 1992, p. 140). The diverse claims from stakeholders include the obvious one of return on investment (shareholders), enhanced wages (employees), and good quality products and services (consumers), among many others. Considering that managers are the company stakeholders who establish contractual relationships with the remaining stakeholders and are responsible for companies' decision-making processes, they are considered the agents of the other stakeholders. Their task is viewed as being that of balancing the remaining stakeholder interests and claims, which are often conflicting.

Contrary to agency theory, this stakeholder-agency approach accepts the existence of “short to medium-run market inefficiencies” and introduces “power differentials into the stakeholder-agent equation” (Hill and Jones 1992, p. 132). In addition, contrary to agency theory, this approach considers not only explicit contractual relationships but also implicit ones, “between all stakeholders” (ibid.). It thus presents the company as a nexus of both implicit and explicit contractual relationships between managers and all stakeholders. As Hill and Jones (1992, p. 132) put forward, “the resultant model is a generalized theory of agency.” In the words of Prior et al. (2008, p. 162), the company is viewed by the proponents of this approach as a “multilateral set of relationships amongst stakeholders” rather than a “bilateral relationship between shareholders and managers.” This approach does not look at a company's board of directors as responsible only for safeguarding the claims and interests of its stakeholders, but also as having the responsibility of ensuring a balance between striving for the fulfillment of different stakeholders' interests and being accountable regarding them. Valls et al. (2020) and Valls et al. (2022) go as far as putting forward that stakeholder-agency theory view the board of directors as acting concomitantly as the management's principal and the stakeholder's agent.

Like traditional agency theory, the stakeholder-agency approach acknowledges divergences between the interests of managers and stakeholders concerning the allocation of firm resources connected with CSR-related issues (Kock et al. 2012; Tauringana and Chithambo 2015). For example, Kock et al. (2012) argue that stakeholders are more likely to be more prone to the establishment of green practices compared to managers. As Nicolo’ and Andrades-Peña (2024, p. 4718) put it, while the former “are more prone to privilege short-term investments ensuring them personal financial benefits,” the latter “are more interested in long-term investments associated with sustainability initiatives that enhance global well-being.” Stakeholder-agency theory proponents put forward that such divergences between the CSR-related interests of stakeholders and managers are affected by both internal (such as corporate governance) and external (such as government regulations, namely those pertaining to corporate reporting) governance mechanisms (Esposito et al. 2023; Kock et al. 2012; Tauringana and Chithambo 2015). From this perspective, the active encouragement of enhanced accountability by way of ESG disclosure can be viewed as an extension of the fiduciary duties of the board toward the firm's multiple stakeholders (Nicolo' and Andrades-Peña 2024; Raimo et al. 2021; Tauringana and Chithambo 2015).

From a stakeholder-agency perspective, Nicolo’ and Andrades-Peña (2024, p. 4727) present the board of directors as a crucial internal governance mechanism to address “the coexistence of multiple and often ambiguous principal-agent conflicts involving a broader set of stakeholders with different financial and non-financial contingencies.” From the same perspective, Benjamin et al. (2024, p. 5) argue that board gender diversity and independence are crucial for the effective monitoring of management as well as for its alignment with stakeholders' interests. In line with this perspective, Zahra and Stanton (1988) argue that independent directors are considered a relevant control mechanism for firms since they provide points of view that are different than those of managers and shareholders, which are more centered on financial matters.

From an agency theory perspective, Agrawal and Knoeber (1996) put forward the idea that independent directors are experts which do not have a relation with the management of the firm and, thus, their opinions do not really influence board decisions. The agency approach postulates that corporate boards which include high number of independent directors are more effective in controlling and governing management (Fama and Jensen 1983; Fernández-Gago et al. 2016; Jensen and Meckling 1979; Jizi et al. 2014). Volonte (2015) notes that the inclusion of independent directors on boards mitigates agency problems because they act as a controlling mechanism for the corporate board and tend to protect the interests of shareholders. Several studies provide evidence that board independence encourages firm performance (Potharla and Amirishetty 2021; Ramdani and Witteloostuijn 2010; Uribe-Bohorquez et al. 2018) since independent directors provide greater independence and objectivity which lead to higher control and evaluation of the management. Using distinctions based on the countries' financial orientation, some studies examine the direct and positive influence of board independence on firm performance in countries with bank-oriented¹ financial systems, such as Italy (Merendino and Melville 2019), Spain (Rodriguez-Fernandez et al. 2014), and Sweden (Moursli 2020), or in market-oriented countries, such as France (Ahmadi et al. 2018) and the United Kingdom (Kyere and Ausloos 2021).

However, to the best of our knowledge, the moderating role of board independence on the impacts of ESG reporting, stakeholder engagement and board gender diversity on firm performance has not been studied. For this reason, we deem it necessary to extend this line of investigation and test the leading role of board independence on these associations and considering the financial orientation (market-or bank-oriented culture) proposed by Demirgüç-Kunt and Levine (2001).

2.2 Hypothesis Development

3.1 Sample

The data for this research is compiled from the Thomson Reuters database. The sample is composed of European sustainable companies operating in 14 countries spread across two groups depending on whether their financial orientation is directed toward banking or the market (Demirguc-Kunt and Levine 2001). Examining the European sample based on financial orientation will allow establishing more consistent governance strategies depending on whether it is oriented toward banking or the market. The two groups and the countries included in each of them are the following: (i) bank-oriented (Belgium, Germany, Ireland, Italy, Norway, Portugal, Romania, Spain, and Sweden) and (ii) market-oriented (Denmark, Finland, France, Netherlands, Switzerland, and United Kingdom).

The initial sample consisted of 120 sustainable companies included in the Euronext Vigeo 120 Index from 2012 to 2021. Following the literature, the first filtering step involved eliminating companies in the financial sector, as they follow accounting standards that differ substantially from those of other sectors. This first filtering resulted in the removal of eight companies (6.67%). The second filtering was conducted to eliminate companies with erroneous or missing data, which led to the exclusion of only 1 company (0.83%). The third and final filtering was performed to remove companies with extreme values observed solely in the indebtedness variable. Specifically, two companies were eliminated (1.67%). After this filtering process, a total of 11 companies were removed, representing 9.17% of the original sample. Therefore, the final sample is composed of 109 European companies with a total of 1090 firm-year observations for the period 2012–2021.

3.2 Variables

The dependent variable is firm performance (FIRMPER), measured using Tobin's Q (Erena et al. 2021; Shan 2019). The first independent variable is ESGREP and is defined as the proportion of activities carried out by the company related to ESG issues, included in its ESG reports (Birindelli et al. 2018; Govindan et al. 2021). The second independent variable is STAKENG and provides information on the level of engagement of a company with their stakeholders, as well as to what extent it makes them participate in the corporate decision-making process, as reported by the company (Amor-Esteban et al. 2018). The third independent variable is GENDIV and is calculated as the ratio between the proportion of female directors on the board of directors divided by the total number of its members (Grau and Bel 2022; Khuonget al. 2022). Furthermore, the interaction variable is measured as the number of independent directors on the board of directors divided by the total number of directors that compose the board (García-Ramos and Olalla 2014; Lepore et al. 2022). Previous studies, also in the European context, have used this same measure to quantify the level of independence of the board (García-Ramos and Olalla 2014; Lepore et al. 2022). Other studies conducted in non-European settings have also used this measure (Hu et al. 2022; Lu et al. 2022; Shan 2019).

We also consider several factors potentially affecting firm performance. First, we control profitability, measured by ROA and calculated as the operating income before interests and taxes over total assets (Valls et al. 2022). Other control variables are: firm leverage (INDEB), measured as debt over total assets (Valls et al. 2020); GRI Report Guidelines (GRIREP), a dummy variable that takes the value 1 if the company's CSR report is published in accordance with the GRI guidelines/standards and 0 otherwise (Dicuonzo et al. 2022); and the board size (BOARSIZE), measured as the total number of board members at the end of the fiscal year (Pucheta-Martínez et al. 2021). We also consider firm size (LFIRMSIZE), calculated as the logarithm of total assets (Valls et al. 2020), and the number of board meetings (NUMEET), measured as the number of board meetings during the year (Pucheta-Martínez and Gallego-Álvarez 2020). The summary of the variables is presented in Table 1.

3.3 Estimation Models

Additionally, $$$ {\upeta}_{\mathrm{j}} $$$ represents the unobservable heterogeneity related to the specific qualities of each company that are assumed to be constant, the management capacity and skills of managers, the barriers to entry into the industry, etc. $$$ {\uplambda}_{\mathrm{t}} $$$ are dummy variables that change over time and are identical for all firms. Thus, it is intended to model the economic variables not controlled by companies, but that influence their financial decisions. $$$ {\upvarepsilon}_{\mathrm{jt}} $$$ are the random perturbations.

The parameters and their relevance were estimated using the Generalized Method of Moments (GMM) applied to the first-differenced equation, incorporating instrumental variables—specifically, the model's explanatory variables lagged by one period. This approach accounts for unobserved heterogeneity and mitigates potential endogeneity issues, as instrumental variables help address the influence of random shocks that may simultaneously affect ESG performance decisions and other firm attributes. The method was originally proposed by Arellano and Bond (1991), based on the nature of the random disturbances. Given the data frequency and the presence of homoscedastic residuals, we applied the procedure in a single step. This methodology effectively tackles endogeneity concerns, recognizing that random disturbances influencing value creation decisions may also impact additional company characteristics (Ullah et al. 2018).

It is important to note that the coefficients of the cross effects measure the impact of the variations of the main variables on firm performance based on the independence of the board. In this way, if its coefficient has the same sign as that the main variable, the effect that said variable has on the firm's performance increases proportionally to the increase in the level of independence of the board. On the other hand, if the coefficient of the crossed variable is of the opposite sign to that of the main variable, the direct effect of said variable is reduced in proportion to the increase in the degree of board independence.

Additionally, four econometric variants are used. Model 1 only includes the control variables. Model 2 adds the moderating variable of board independence. Model 3 includes all the main variables together with the control variables. Finally, Model 4 includes all the variables in equation (1). This last model will allow us to examine the effect that board independence has on the value of sustainable European companies, not directly but indirectly through its interaction with the other main variables. This is the way in which the degree of compliance with the proposed hypotheses will be corroborated.

The parameters and their statistical significance have been estimated incorporating instrumental variables through the Generalized Method of Moments (GMM) to the first difference equation. This methodology makes it possible to control endogeneity problems quite successfully.

At the bottom of Tables 4 and 5, two additional contrasts appear to reinforce the validity of the procedure. First, the Arellano and Bond (1991) second-order serial correlation test is performed to check the consistency of the estimates. Second, the restriction overidentification test was also used to verify the absence of correlation between the instruments and the error term (Sargan 1958).

4.1 Descriptive Statistics

Table 2 presents the evolution of descriptive statistics for firm performance and board independence. Taking the 14 countries analyzed as a whole, the board independence has increased until 2018 (Table 2: Panel A). From that moment on and until the end of the period examined in this study (2021), this variable decreased and then stagnated. When reproducing this same analysis for the groups, the pattern of behavior obtained is very similar. The average value for the complete sample is 61.78%, whereas for the subsamples of countries with bank-oriented or market-oriented countries this value is 55.48% and 66.14%, respectively. Therefore, there is an impasse in the level of independence within the board from 2018 onward, with market-oriented countries showing, on average, the highest level of independence.

The descriptive statistics (Table 3: Panel A) provide evidence that, for the complete sample, the level of ESG reporting and stakeholder engagement is very good. Sustainable European companies comply, on average, with 92% of the activities related to ESG principles. 90% of companies in our sample explain how they engage with stakeholders and increase firm performance, and the proportion of female directors is, on average, 31.26%. When comparing these data at the group level (Table 3: Panels B and C), the results show that the countries whose financial orientation is to the market, in general, have much better positions (higher level of compliance with ESG principles, greater representation of gender diversity on the board, higher profitability, lower level of indebtedness, etc.).

To identify the best way to carry out the groupings, three groups of dummy variables were previously tested: (i) depending on the governance structure of the council (DONETIER, DTWOTIER, or DMIXED) in line with La Porta et al. (1998) and Choudhuri (2017); (ii) depending on the legal determinants (DGOVERN); and (iii) depending on the financial orientation (DORIENTED) Demirgüç-Kunt and Levine (2001). The Variance Inflation Factor (VIF) that appears at the bottom of Table 3 for the complete sample (EUROPE) suggests the absence of multicollinearity problems. Values for VIF range from 1.1087 to 5.6295, with a mean value of 2.2534. These results are very similar for the groups according to financial orientation.

Table 4 presents the estimation and contrast results, which show that there are only significant differences for the countries grouped by two-tier with low significance (10%) and depending on the orientation being aimed at banking or the market with high significance (5%). Consequently, financial orientation is the grouping chosen to carry out to extend past research.

4.2 Regression Results

In Table 4, Model 4 provides evidence that the variable ESG reporting presents a negative sign and is statistically significant. Thus, we have to partially accept the first hypothesis and conclude that the association between ESG reporting and firm performance is negatively moderated by board independence. Thus, the likelihood of increasing firm performance is lower when a firm has both ESG reporting and board independence. Both mechanisms (board independence and ESG reporting) are more substitutive than complementary. This finding is consistent with García-Sánchez and Martínez-Ferrero's (2018) findings of some level of opposition of independent directors to ESG disclosure unless such disclosure is consistent with the underlying ESG performance. As these researchers explain, independent directors are aware that managers may provide deceptive ESG information as well was of the risk of such information jeopardizing their reputation and imperiling future professional opportunities. Thus, they are likely to try to avoid practices that lead to such negative effects, such as ESG disclosure they do not view as credible. This is a behavior that may explain our finding. If this is the case, stakeholder-agency proponents may be right in viewing the board of directors as a crucial internal governance mechanism (Nicolo' and Andrades-Peña 2024) in the sense that independent directors act here as a mechanism that hinder the pursuance of self-serving interests and promote transparency and accountability.

Moreover, the results also show that the cross variable STAKENG*DINDBO is positive and not statistically significant. Thus, we must reject the second hypothesis and conclude that the association between firm performance and stakeholder engagement is not moderated by board independence. In this regard, our evidence shows that although board independence has a relevant influence on corporate board decisions, they do not affect the relationship between stakeholder engagement and firm performance, in line with the findings of Karim et al. Consistent with the agency approach, companies where higher independence representation is the prospective cause of conflict of interest between shareholders and management, which leads to no influence on firm performance. Another possible reason could be that the moderating effect of board independence may change depending on the dialogue and channel used by the stakeholder group and consequently, the influence on firm performance may be changed, in line with García-Sánchez et al. (2022).

Finally, the interaction between board independence and board gender diversity (GENDIV*DINDBO) exhibits a positive sign and is not statistically significant. This leads us to reject hypothesis 3 and conclude that the association between board gender diversity and firm performance is not moderated by board independence. In contrast with agency theory, this evidence reveals that the opinions of experts who do not have relations with the companies may not really influence board decisions, in line with Agrawal and Knoeber (1996). Furthermore, this result also suggests that excessive involvement of board independence in the daily activities of the companies may restrict the managers from performing their functions liberally, and thus negatively moderate the relationship, in line with Wu and Wu (2014). Although it may be expected that companies where board gender diversity would supervise the companies' decisions, board independence does not seem to enhance firm performance. Thus, the initial premise that board independence monitors the managers, in line with agency theory, and that their effect could increase when the companies have female directors on corporate boards does not seem to apply in European companies.

4.3 Regression Results According to Financial Orientation

The coefficients of the dummy variables and their significance are presented in Table 4 and, according to Model 4, indicate the statistical relevance (at 1%) of the grouping that suggests separating the countries according to whether they are oriented toward banking or toward the market (DORIENT). Consequently, this research will be extended by reproducing the analysis of the previous section but now separating the countries into these two groups. In this way, it will be possible to verify if the fulfillment of the hypotheses that are exposed for Europe can be generalized or not for all the countries that comprise it.

Table 5 presents the results of the regression when the sample is divided depending on financial orientation; Panel A includes bank-oriented, and Panel B presents market-oriented.

In Panel A, the interaction between board independence and ESG reporting presents a negative sign and is not statistically significant. This leads us to reject the hypothesis 1 and conclude that ESG reporting does not affect firm performance in bank-oriented countries, when it interacts with board independence, firm performance is lower, namely, board independence moderates negatively the association between ESG reporting and firm performance in bank-oriented countries. However, for Panel B, the interaction presents a negative sign and is statistically significant. Thus, we have to partially accept hypothesis 1 and conclude that the association between ESG reporting and firm performance is negatively moderated by board independence in market-oriented countries. These results are supported by the European sample. Therefore, it is expected that, in market-oriented countries (Table 5, Panel B), to the extent that activities related to ESG principles are led by board independence and when this level of independence increases, sustainable firms performance will grow, but at a slower pace. This is because of the interaction effect that occurs between board independence, the preparation of ESG reports and firm performance. Therefore, this would imply that firm performance would be affected, this effect being lower for market-oriented countries (coefficient of −1.1291 at 10%) compared to the group that includes all the countries in the sample (coefficient of −1.5009 at 1%). This is consistent with Kim et al. (2022) findings of a less positive association between ESG reporting and financial performance in market-oriented countries since ESG reporting is focused on long-term performance while the market-oriented culture is based on short-term performance.

Furthermore, in Table 5, Panel A, we analyze the moderating effect of board independence between stakeholder engagement and firm performance in bank-oriented countries. The effect is positive and significant. Accordingly, in this case, hypothesis 2 is accepted. These findings suggest that board independence positively moderates the impact of stakeholder engagement on firm performance in bank-oriented countries. This indicates that companies with stakeholder engagement should establish independent boards because they emphasize the encouragement of firm performance by stakeholder engagement. It can be said that board independence and stakeholder engagement are substitute mechanisms. This result supports the thesis that firms with higher board independence tend to engage more with stakeholders in bank-oriented countries, in line with the arguments of agency and stakeholder theories. Having stakeholder engagement with the development of a more efficient corporate governance culture signals the trust of managers in the employees, which leads to reduced information asymmetries and ultimately improves firm performance. Moreover, bank-orientation is the main way to obtain corporate funding, and these relevant decisions may enhance firm performance. However, the findings obtained cannot be extended to the group of market-oriented countries since the interaction between board independence and stakeholder engagement is positive but statistically insignificant (Table 5, Panel B). For this reason, board independence does not moderate the positive impact of stakeholder engagement on firm performance in market-oriented countries. Although it may be expected that firms where board independence would monitor more effectively by overseeing that stakeholder decisions are not risked, it is shown that stakeholder engagement may mitigate the potential control played by board independence.

Finally, in Table 5, Panel A and Panel B, we analyze the moderating effect of board independence on the impact of board gender diversity on firm performance in bank-oriented and market-oriented countries, respectively. In both cases, the interaction between board independence and board gender diversity is positive but is statistically insignificant. Accordingly, in these cases, hypothesis 3 is also rejected. Then, board independence does not moderate the positive influence of board gender diversity on firm performance in bank-oriented and market-oriented countries, in line with the results obtained in the European sample.

Given the results presented so far, it is worth noting the role played by the grouping based on financial orientation, since it has allowed a more precise description of the moderating effect of the board independence for European sustainable companies. This is one of the most relevant findings that emerges from this research after exploring the entire sample and separating it based on the financial orientation of their countries.