Does CSR committee matter? The relationship between corporate social responsibility and firm performance in the Asia–Pacific markets

In the era of global competition, sustainable development is the most important objective of firms. Besides the traditional role of firms in providing goods and services, societies expect them to act in a socially responsible manner. Therefore, the demand from firms' stakeholders for increased non-financial disclosures, in the form of corporate social responsibility (hereafter, CSR), has risen significantly over time. To this end, scholars have generally acknowledged that socially responsible investment enables firms to establish relationships with stakeholders, achieve legitimacy and enhance performance (Mathews, 1997). However, the existing evidence on the association between CSR and firm performance (hereafter, FP) is still inconclusive. One stream of studies is consistent with neoclassical theory, which claims that CSR negatively affects a firm's financial performance by diverting funds from potentially profitable investments and incurring additional costs for the organisation (Bird, Hall, Momentè, & Reggiani, 2007; Lee & Park, 2009; Lee, Singal, & Kang, 2013). In contrast, the theoretical and empirical evidence supporting stakeholder theory contends that firms should engage in the welfare of the society to build strong ties with their stakeholders, which eventually helps them achieve higher FP (Freeman, 1984, 2010; Roberts, 1992; Wang, Chen, Yu, & Hsiao, 2015; Esteban-Sanchez, de la Cuesta-Gonzalez, & Paredes-Gazquez, 2017; Kao, Yeh, Wang, & Fung, 2018).

Some studies (e.g. Yoon & Chung, 2018; Candio, 2024) have also reported mixed evidence on the relationship between CSR and FP. A possible reason for the inconsistent evidence on the CSR–FP relationship may be a lack of understanding of the underlying mechanisms that likely affect it. Moreover, the inconsistency among the studies' findings may be due to the complex nature of the CSR–FP relationship, suggesting that other firm-specific and institutional factors may drive it (Shi & Veenstra, 2021).

Recent studies suggest that several governance constituents, such as board independence, board size, CEO duality and CEO power, significantly enhance the contribution of CSR to FP (e.g. Chijoke-Mgbame, Mgbame, Akintoye, & Ohalehi, 2020; Li, Gong, Zhang, & Koh, 2018; Rossi, Chouaibi, Chouaibi, Jilani, & Chouaibi, 2021). Firms have regularly evolved their corporate governance structures and, more recently, have established a CSR committee, an essential component of their board-level subcommittees. The role of a CSR committee is to identify CSR-related issues, monitor sustainable corporate practices and ensure improved CSR performance (Hussain, Rigoni, & Orij, 2018). Firms with CSR committees are more likely to report CSR outcomes to stakeholders and strengthen corporate relationships with them (Funente et al., 2017). They follow the Global Reporting Initiative (GRI) guidelines to enhance the quality of CSR reporting (Kılıç, Uyar, Kuzey, & Karaman, 2021). These initiatives by firms signal to stakeholders that their CSR-related demands are prioritised at the strategic level (Eberhardt-Toth, Caby, Gendron, & Ramboarisata, 2019), thereby resulting in higher firm value in the long term (Gennari & Salvioni, 2019).

While there is sufficient evidence of the crucial role of corporate governance in enhancing CSR's effect on FP, the moderating role of the CSR committee, an essential constituent of corporate governance, is largely overlooked. A few exceptions include, for instance, a recent study by Kuzey, Uyar, Nizaeva, and Karaman (2021), which is limited to firms in the financial and services sectors. Likewise, Khan, Afeef, Ilyas, and Jan (2024) and Dziri and Jarboui (2024) investigate the moderating effect of the CSR committee in the CSR–FP relationship in the international and European contexts, respectively. To extend the current studies, we investigate the impact of the CSR committee on the association between CSR and FP in Asia–Pacific markets. Moreover, the current studies do not provide evidence on how the CSR committee drives the positive effect of CSR. Specifically, these studies lack investigations into the potential channels through which the CSR committee drives the CSR–FP relationship. Therefore, our second objective is to address this void by exploring the channels through which the CSR committee and CSR jointly impact FP.

The Asia–Pacific market is an excellent laboratory for examining the role of the CSR committee for two key reasons. First, compared to developed countries, firms in Asia–Pacific markets are less likely to disclose CSR information due to a weak regulatory environment and less stringent laws (Bhatia & Makkar, 2019). The role of the CSR committee in influencing CSR outcomes has not yet been explored. Second, there is considerable heterogeneity among Asia–Pacific markets concerning regulatory, political, business and social environments (Siegel, 2009; O'Connor, Vera-Muñoz, & Chan, 2011), and in such environments, various corporate governance mechanisms (e.g. the board of directors) play a vital role in CSR engagement (Al-Mamun & Seamer, 2021). Emerging economies often face issues such as poverty, inequality and limited access to education and healthcare. CSR can contribute to sustainable development by supporting initiatives that tackle these challenges, improve social welfare and promote environmental sustainability. Moreover, corporate board committees uniquely enhance firm financial performance in these markets (Singhania, Singh, & Aggrawal, 2024). Therefore, given the need for improvement in CSR performance and the role of board committees, emerging markets are an ideal laboratory for investigating the role of CSR committees in enhancing CSR performance. Thus, the central objective of this paper is to examine whether CSR committees augment the effect of CSR on FP in Asia–Pacific markets.

We used a sample of firms from 12 Asia-Pacific markets to test our proposed moderating effect. Our study finds that the presence of a CSR committee strengthens the positive impact of CSR on FP, as measured by return on assets (ROA) and return on equity (ROE). This moderating impact is consistent across individual dimensions of CSR (e.g. environmental and social). In addition, we confirm the proposed relationship across stakeholder-oriented and shareholder-oriented countries. We found that the benefit of a CSR committee in enhancing the CSR–FP relationship is greater in stakeholder-oriented countries. The results also show that the estimated moderating effect of the CSR committee is significantly higher among firms that are less likely to have agency issues. Additionally, we test the two potential channels through which the CSR committee influences the association between CSR and FP: CSR disclosure and CSR-related controversies. First, we find that firms with CSR committees have a higher propensity to disclose CSR information and experience fewer CSR-related controversies. Second, we show that the CSR committee increases CSR contributions by promoting CSR disclosures and reducing CSR-related controversies. Our baseline results remain unchanged when we use two-stage least squares (2SLS) and generalised method of moments (GMM) specifications to address endogeneity. Moreover, we mitigate concerns about selection bias using the propensity score matching (PSM) approach. We also perform additional robustness tests. First, we use a market-based measure (Tobin's Q) as an alternative proxy for firm financial performance. Second, we re-estimate our main specification using firm and country fixed effects. All these tests validate the baseline results.

Our study offers several contributions to the existing literature. First, we contribute to the emerging literature on the moderating effect of various corporate governance mechanisms on the CSR–FP relationship. In this stream of literature, recent studies (e.g. Kuzey et al., 2021; Khan et al., 2024; Dziri & Jarboui, 2024) have documented the significance of the board's CSR committee in enhancing CSR's role in FP, specifically in developed economies. We add to this stream of literature by providing robust evidence that the CSR committee enhances CSR's positive impact on FP in Asia–Pacific markets, a context in which the legal and regulatory environment is weak, and the need for effective corporate governance is high. Second, existing studies (e.g. Kuzey et al., 2021; Khan et al., 2024; Dziri & Jarboui, 2024) have overlooked the potential channels through which a CSR committee impacts CSR performance. Our study further enriches the existing evidence by revealing that the CSR committee augments the CSR–FP nexus by promoting CSR disclosure and alleviating CSR-related controversies. Third, recent empirical studies have shown that country-level legal origins affect corporate environmental responsibility and the CSR-firm-value relationship (Kim, Park, & Ryu, 2017; Grassmann, 2021). We classify countries by legal origin and provide evidence that firms operating in stakeholder-oriented countries drive the joint effect of CSR committees and CSR engagement on FP. Lastly, we add to the literature focussing on the role of CSR in influencing agency costs. Samet and Jarboui (2017) found that the moderating effect of CSR on investment-to-cash-flow sensitivity depends on the intensity of agency costs. Similarly, Rossi and Harjoto (2020) provide evidence of a negative relationship between corporate non-financial disclosure and agency costs. Our study shows that the joint effect of the CSR committee and CSR is higher in firms with low agency costs.

This paper is structured as follows: Section 2 provides a brief literature review and presents a theoretical framework. Section 3 presents the methods, briefly discussing the data, variables and research models. Section 4 provides the empirical results, and Section 5 concludes the paper.

The association between CSR and FP can be justified by two contrasting theories: the neo-classical and stakeholder theories. The neo-classical or conventional shareholder theory (Friedman, 1970, 2007) holds that the sole responsibility of a business is to maximise profits for its owners within the basic norms of society. Consistent with this view, Aupperle, Carroll, and Hatfield (1985) suggest that investing in CSR activities will place the firm at a competitive disadvantage, resulting in a negative effect of CSR on the firm's market performance. In contrast, Freeman (1984) and Freeman and McVea (2005) advocate a stakeholder-oriented view, arguing that a firm's responsibility is to benefit all its stakeholders, e.g. shareholders, customers, communities, consumers and other related groups. Similarly, Porter and Kramer (2006) argue that CSR can be more than a cost, a constraint or a charitable deed. Instead, it can also provide opportunity, innovation and competitive advantage. Therefore, CSR has evolved from a profit-oriented perspective to a broader set of responsibilities, based on the belief that firms should generate shared value for all stakeholders (Latapí Agudelo, Jóhannsdóttir, & Davídsdóttir, 2019).

Consistent with competing theoretical views, empirical literature has also shown mixed evidence concerning the CSR–FP relationship. Positive, negative and non-linear associations between CSR and FP have been reported in the literature. Empirical literature supporting the stakeholder approach suggests that meeting stakeholders' needs and demands and building strong ties with them are beneficial for firms in the long run. This view is supported by studies that report a positive association between CSR and FP (Cochran & Wood, 1984; Chen & Wang, 2011; Wang et al., 2015; Kang, Germann, & Grewal, 2016; Esteban-Sanchez et al., 2017; Kao et al., 2018). However, empirical studies that complement the neo-classical view claim that CSR activities put extra costs on firms (Lee & Park, 2009; Lee et al., 2013), particularly when managers conduct CSR activities to pursue their own interests or stakeholders' interests at the expense of shareholders' value (Jensen, 2001). This view is generally supported by studies reporting a negative relationship between CSR and FP (e.g. Chen et al., 2015). In addition, some studies have reported insignificant and non-linear impacts of CSR on FP (Chen et al., 2015; Nollet, Filis, & Mitrokostas, 2016). The inconsistency among studies on the CSR–FP nexus is also supported by the recent review study of Kong, Antwi-Adjei, and Bawuah (2020), who documented that 34% of the papers reported a positive impact of CSR on FP, 28% reported a partial impact, while 38% of the studies reported a negative impact of CSR.

In the era of global competition, reliable and valid information for stakeholders, shareholders and competitors is necessary. In this connection, firms have established CSR committees to provide valid and reliable information on their CSR performance. A CSR committee, typically part of the board of directors, is also known as a sustainability committee (Eberhardt-Toth, 2017). It is a specialised committee of the board of directors responsible for ensuring and monitoring CSR activities and policies within a firm (Liao, Luo, & Tang, 2015). This would enable a greater understanding of business sustainability by providing stakeholders with information that allows them to assess the company's impacts and the systems implemented to manage its business in terms of economic, social and environmental consequences (García-Sánchez, Gómez-Miranda, David, & Rodríguez-Ariza, 2019). Nowadays, a critical dimension of corporate governance is corporate transparency, which positively impacts firm value (Yu, Guo, & Van Luu, 2018). In addition, the availability of reliable business information helps investors understand management decisions, boost confidence, increase foreign direct investment and reduce asymmetric information in the capital market (Bushman, Piotroski, & Smith, 2004).

García-Sánchez et al. (2019) show that the CSR committee promotes sustainable management policies and systems and improves stakeholder communication. Furthermore, Baraibar-Diez and Odriozola (2019) reveal that firms with a CSR committee perform better in CSR disclosure than those without. Kubo and Sasaki (2024) show that CSR committees provide structural and dedicated efforts towards sustainability, which leads to a positive effect on CSR performance. Thus, the existence of a CSR committee facilitates and promotes sustainable management policies and systems and enhances communication with stakeholders. In addition, these specialised committees often commission assurance for sustainability reports to reinforce strategies to improve corporate transparency (García-Sánchez et al., 2019). The board-level CSR committee is established to strengthen CSR by addressing stakeholders' interests and to deliver sufficient value to the firm (Burke, Hoitash, & Hoitash, 2019). Altogether, these studies suggest that a CSR committee enhances corporate social performance by strengthening the firm's relationship with its stakeholders, potentially leading to increased FP. Hence, we established the following hypothesis.

According to the stakeholder-agency theory, CSR disclosure is viewed as a tool to align managers' and stakeholders' interests. Therefore, sustainability reporting must contain information on CSR activities that assists managers not only in detecting weaknesses but also in formulating specific policies and mechanisms to enhance the firm's internal capabilities to apply CSR-based strategies to satisfy its stakeholders (International Finance Corporation, 2010). However, the quality of information disseminated in sustainability reports has been questioned regarding its relationship to actual performance (Pellegrino & Lodhia, 2012), and a complex reporting structure may mislead users and create hurdles to meeting users' demands (O'Dwyer & Owen, 2005). Therefore, there is a need to report CSR activities in a structured way that is easy to understand and offers greater practical value to the relevant stakeholders. As a result, many firms have begun disclosing CSR information in line with the GRI guidelines to improve transparency in CSR reporting.

From a theoretical perspective, establishing a CSR committee helps a firm gain CSR-oriented legitimacy (Patten, 2019) by encouraging voluntary CSR reporting. Thus, the CSR committee promotes the firm's ethical culture, creates management awareness of CSR issues and transparently discloses CSR information. Following the GRI-International Finance Corporation (IFC) strategy, CSR disclosure may help the firm strengthen stakeholder engagement and improve internal monitoring of CSR-related activities. As GRI reporting aims to promote transparency in sustainability reporting, adopting GRI guidelines conveys a clear presentation of firms' social and environmental disclosures (Einwiller, Ruppel, & Schnauber, 2016). GRI reporting effectively responds to stakeholders' demands for CSR information (Nikolaeva & Bicho, 2011). Empirical evidence indicates that the CSR committee advises the board to implement the GRI-IFC strategy for disclosing CSR activities (García-Sánchez et al., 2019). Baraibar-Diez and Odriozola (2019) also report similar results. Consistent with Velte and Stawinoga's (2020) findings, a positive impact of CSR committees on CSR disclosure enhances FP. We also believe that the relationship between CSR and FP would be stronger in a firm where the CSR committee discloses a GRI-based CSR report. Hence, we formulated the following hypothesis.

In addition to direct participation in the formulation and revision of CSR policies (Ricart, Rodrı́guez, & Sanchez, 2005), the CSR committee not only plays an active role in the strategic implementation of the firms' CSR programs but also is responsible for following up on such programs and making a timely adjustments to carry them forward in their true spirit. This helps create a balance that satisfies both management and stakeholders, whereby firms' investment in CSR activities contributes to achieving both financial and non-financial goals. Therefore, the monitoring role of the CSR committee becomes crucial. For example, the CSR committee's responsibility is to enhance CSR activities to maximise utility for internal and external stakeholders and to disclose transparent information about these activities. On the other hand, the CSR committee is responsible for mitigating CSR-related financial and non-financial controversies by fostering more robust communication about CSR strategies and their implications, which helps the firm meet stakeholders' demands.

Generally, controversial stories receive extensive media coverage, attracting more attention and triggering public scepticism. Evidence suggests that firms lose their credibility (Godfrey, Merrill, & Hansen, 2009) due to stakeholder scepticism in response to aggressive media coverage of corporate controversies (Du, Bhattacharya, & Sen, 2010). Similarly, investors may pay closer attention to a firm under media scrutiny because of its involvement in CSR-related controversies (Cai, Jo, & Pan, 2012). CSR-related controversies are negative news stories concerning CSR activities, such as questionable social behaviours and product-related irresponsibility that draw media attention to the firm (Cai et al., 2012). Therefore, suspicious CSR activities may hurt a firm's value by increasing uncertainty about its prospects and reputation. The following literature (Kang & Kim, 2014; Oikonomou, Brooks, & Pavelin, 2012) concludes that CSR-related controversies may increase systemic risk. Subsequently, such firms may perform poorly and lose significant market share.

The insurance mechanism view of CSR performance suggests that the poor performance of controversial CSR firms can be mitigated by building CSR-based moral capital (Godfrey, 2005), which can temper stakeholders' negative judgments and sanctions when adverse developments materialise, as confirmed by Godfrey et al. (2009). Koh, Qian, and Wang (2014) report variations in CSR performance and insurance value, with high-risk firms benefiting more. Financially sound firms and socially contested industry firms can benefit from CSR performance. Recently, Aouadi and Marsat (2018) reported that CSR-related controversies are harmless for firms with higher CSR performance. Such firms have better financial performance, particularly if they are large, high-attention firms and are located in countries with higher press freedom.

It is evident that higher CSR performance may help mitigate the negative impact of controversies. Therefore, regarding the monitoring role of the CSR committee, we believe that firms with a CSR committee significantly reduce CSR-related controversies, improve their CSR reporting and enhance firm value through CSR performance. Therefore, consistent with the above discussion, we propose H2b as follows:

Initially, we selected a sample of 1,862 publicly listed firms across 12 Asia–Pacific countries for the period from 2002 to 2018. Data on CSR, including its sub-dimensions, were retrieved from the ASSET4 database of Thomson Reuters, while financial data were retrieved from Worldscope. We only use firms for which data were available in ASSET4. Firm-year observations with missing CSR scores are removed from the sample. In addition, firms included in the financial sector (SIC codes 6000–6900) and utility sector (SIC codes 4900–4999) were removed from the sample because these firms follow a different set of environmental and social regulations, such as the “Equator Principles” ^[1] (Macve & Chen, 2010). After selection, the final sample comprises 1,544 firms and 12,231 firm-year observations. All financial variables were winsorised at the 1^st and 99^th percentiles to alleviate the potential effect of outliers ^[2]. Further details about the sample selection process are provided in online Appendix A2.

Relying on prior literature, we use ROA (net income divided by total assets) and ROE (net income divided by shareholders' equity) as proxies for FP in the baseline analysis, following Shi and Veenstra (2021). Alternatively, we employ Tobin's Q, a market-based measure of FP, calculated as the sum of a firm's total market capitalisation and total debt, divided by total assets (Pucheta-Martı́nez & Gallego-Álvarez, 2020).

CSR is the independent variable proxied by a firm's aggregate percentile score in environmental, social and governance activities. The environmental component reflects a firm's engagement in various environmental activities, such as reducing pollution through environmentally friendly production, efficient resource utilisation and the introduction of new technologies that facilitate green production. The social component reflects a firm's engagement in various social activities, such as supporting the workforce by providing a better work environment, promoting human rights, maintaining good citizenship and ethical standards and fulfilling product-related responsibilities. Finally, the governance component reflects a firm's commitment to upholding good governance practices, protecting shareholders' rights and integrating its CSR strategy into daily decision-making. In addition to the composite CSR score, we consider the environmental and social component scores as proxies for CSR, following Rjiba, Jahmane, and Abid (2020).

The CSR committee is a subcommittee of the corporate board whose role is to enhance the monitoring and transparency of CSR activities (Burke et al., 2019). It is an indicator variable coded as 1 if a firm has a CSR committee and 0 otherwise.

There is a concern that firm-level factors may capture the effect of CSR on FP. We rely on prior literature that uses several firm-level financial variables associated with FP to control for these factors. We control for the effects of firm leverage, firm size, sales growth and research and development (R&D) expenditures, following Bennouri, Chtioui, Nagati, and Nekhili (2018). The leverage ratio measures a firm's debt as the ratio of total debt to total assets. Studies on the association between debt levels (e.g. leverage ratios) and FP yield mixed evidence. One stream of research (e.g. Abor, 2005; Robb & Robinson, 2014) suggests that high leverage is beneficial because the net gain from debt exceeds its interest expense. In addition, a firm with a high debt level enjoys a tax advantage, which translates into a higher market value. On the contrary, Ibhagui and Olokoyo (2018) show that an increase in leverage negatively affects FP. Similarly, the effect of firm size (the natural log of total assets) is also controlled for. It is widely used to determine FP in the prior literature. Sales growth (the percentage change in sales from the previous year) is controlled for as it has been found to have a positive relationship with FP (King & Santor, 2008). Consistent with Padgett and Galan (2010), we also include R&D expenditures as a control variable to enhance CSR activities and FP through the competitive advantage.

Furthermore, we control for corporate board characteristics, following recent studies (e.g. Pucheta-Martı́nez & Gallego-Álvarez, 2020). Three board characteristics are considered as control variables: board size and board independence. Board size is proxied by the natural log of the total number of board members. A large board size provides more diverse and valuable resources for firms to continue their CSR activities (Giannarakis, 2014). Moreover, from an agency theory perspective, boards with more independent directors play a more effective role in monitoring management (Jensen & Meckling, 1976; Jizi, Salama, Dixon, & Stratling, 2014). Board independence is calculated as the percentage of independent directors on the board. In line with the agency and resource dependence theories, Pucheta-Martı́nez and Gallego-Álvarez (2020) find that larger and more independent boards positively impact FP. We also control for country-level governance and institutional quality by including the institutional quality index (IQ) from the Political Risk Services (PRS) database of the International Country Risk Guide (ICRG). IQ is the aggregate index of country-specific political risk rating, covering institutional and governance dimensions. We also include the gross domestic product (GDP) growth rate to control for macroeconomic differences across countries since GDP growth is strongly linked to corporate performance (Al-Najjar, 2014). An additional explanation of the variable measurements is provided in Appendix A1.

We follow the empirical specifications of Kuzey et al. (2021) and Khan et al. (2024) to test the effect of a CSR committee on the association between CSR and FP.

where FP represents the two dependent variables (ROA and ROE), while the explanatory variable is CSR, measured by the aggregate environmental, social and governance score and its individual environmental (ENV) and social (SOC) pillars. Our interest is in the moderating effect of the CSR committee on the CSR–FP relationship, captured by the interaction between CSR and the CSR committee (CSR×CSR_COM). The moderating variable is the CSR committee (CSR_COM) defined in Section 3.1.2. Control variables include the leverage ratio (LEVERAGE), firm size (SIZE), sales growth (GROWTH), R&D expenditures (RD), firm board size (B_SIZE), board independence (B_IND), country-level institutional quality (IQ) and GDP growth. Several empirical studies have documented that unobserved firm-related factors can affect our predicted relationship, which we control for by including year and firm fixed effects in the model. In line with Petersen (2009), standard errors are adjusted for heteroskedasticity and serial correlation using firm-level clustering.

Panel A of Table 1 presents country-wise averages of CSR and its two sub-components (ENV and SOC) across 12 countries. The sample comprises 1,544 firms, with the largest share coming from Japanese firms at about 23.77%, while the Philippines is represented by 16 firms, accounting for only 1.04%. The range of CSR disclosure scores is between 40.62% in China and 54.39% in Thailand. Japan has the highest ENV score, at 65.99%, while Singapore has the lowest, at 39.54%, amongst the 12 countries. Regarding the SOC score, Indonesia ranks highest with a 63.63% contribution, while China ranks lowest at 38.04%. Finally, India (70.2%) and Thailand (71.8%) have the highest representation of CSR committees on boards, while Hong Kong and the Philippines firms are at the bottom, with only 27% of firms having CSR committees.

Panel B of Table 1 presents the descriptive statistics of the variables over the sample period from 2004 to 2018. The mean and median values of ROA are 0.039 and 0.038, respectively. The average ROE was 0.082 with a median value of 0.092. The mean (median) of the variable of interest (CSR) was 48.24% (47.84%). The mean (median) values for the subcomponent ENV are 53.28% (53.32%), whereas the mean (median) for the subcomponent SOC is 48.81% (46.99%). The sample shows that almost 47.4% of firm-year observations include CSR committee (CSR_COM) observations. The sample firms use an average of 22.7% debt financing, and the average firm size is almost 15.32. The sample firms experience an average growth rate of 9.8% over the sample period, and the average firm spends 13.9% of its revenue on capital expenditures (CAPEX). The average firm's investment in R&D is 0.006% of sales, with a maximum investment of about 0.4%. On average, there are 10 members on a board. The independent directors represent about 35.24% of the board members. The average IQ across our sample countries was 78.16. The average GDP growth across our sample countries is 2.743%.

In Table 2, a pairwise correlation analysis is conducted to assess the bivariate relationships between each pair of variables. The correlation coefficients of CSR, ENV and SOC with the dependent variables ROA (ROE) are 0.039 (−0.028) (p-value <0.05), 0.013 (−0.136) and 0.045 (−0.037) (p-value <0.05), respectively. These coefficients suggest that CSR and SOC have positive and significant effects on FP. The correlation coefficient between CSR_COM and ROA is −0.018, showing a negative but insignificant association between the CSR committee and FP. The correlation coefficient between CSR_COM and ROE is −0.081, indicating a significant negative association between the CSR committee and FP. Moreover, the correlation coefficients representing the association of CSR_COM with CSR, ENV and SOC are 0.576 (P-value <0.05), 0.581 (P-value <0.05) and 0.567 (P-value < 0.05), respectively. Overall, the pairwise correlation matrix suggests that the correlation scores among the variables of interest and the control variables are below 0.6. This implies that the multivariate analysis may not suffer from multicollinearity. Moreover, the variance inflation factor (VIF) values (not reported) for all variables are below 5, consistent with the correlation results.

In this section, we present the baseline results on the moderating role of the CSR committee in the nexus between CSR and FP, using ordinary least squares (OLS) regression with year and firm fixed effects. First, we used stepwise regression with backward elimination to select significant predictors of FP. Initially, we considered several control variables based on the prior literature. We used a 10% significance level (p-value) as the significance level for variable selection. The model excludes two control variables (CEO duality and firm-specific anti-takeover protection) because they have higher p-values. The stepwise regression results suggest that CSR_COM positively influences the relationship between CSR, ENV, SOC and FP ^[3].

Table 3 reports the results of the OLS model using the variable combination suggested by the stepwise regression. Models 1–3 report the results of the CSR committee's impact on the CSR (SOC and ENV) and FP (ROA) nexus, respectively. In models 1 and 3, the stand-alone effects of CSR and SOC on ROA are insignificant. Moreover, in Model 2, the stand-alone effect of ENV is negative and statistically significant. Economically, a one-standard-deviation increase in the ENV score corresponds to about a 7% ^[4] decrease in ROA. These results suggest that CSR activities alone may not contribute to FP in firms without a CSR committee. Likewise, the stand-alone effect of CSR_COM is negative and statistically significant. With respect to the CSR committee, a one-standard-deviation increase corresponds to about 43.5% of a standard deviation decrease in FP, indicating a substantial negative effect. This suggests that although the CSR committee is an important governance constituent, it may adversely affect FP when it does not integrate with CSR investment. Interestingly, the interaction term (CSR×CSR_COM) is positively and significantly associated with ROA (coefficient 0.031, P < 0.1), indicating a positive moderating effect of CSR_COM on the association between CSR and FP. This evidence suggests that CSR needs to be integrated with the CSR committee to enhance its effect on FP. Economically, a one-standard-deviation increase in CSR×CSR_COM leads to about a 10.5% increase in ROA. In models 2 and 3, as expected, the interaction between CSR_COM and individual components of CSR (e.g. ENV and SOC) remains significantly positive. Economically, a one-standard-deviation increase in the interaction term with the ENV component leads to an 11.9% increase in ROA, while the interaction term with the SOC component leads to a 16.8% increase in ROA.

Models 4–6 report the results of the CSR committee's impact on the CSR (SOC and ENV) and ROE nexus, respectively. Overall, the stand-alone effects of CSR, ENV and SOC on ROE are similar to those on ROA. Only the stand-alone effect of ENV is negatively significant in model 5. Economically, a one-standard-deviation increase in the ENV score corresponds to about an 8.4% decrease in ROE. The effects of the interactive variables (such as CSR×CSR_COM, ENV × CSR_COM and SOC×CSR_COM) on ROE are positively significant at the 10%, 5% and 1% levels in models 4–6, respectively. In terms of economic significance, a one-standard-deviation increase in CSR×CSR_COM, ENV×CSR_COM and SOC×CSR_COM leads to 5.6%, 5.6% and 7.9% increases in ROE in models 4–6, respectively.

Hence, the combined effect of the CSR committee and the pillars of CSR on FP remains consistent. Thus, our hypothesis that the “CSR committee strengthens the association between CSR and FP” is confirmed. Notably, the CSR committee enhances the firm's performance by moderating the adverse effects of environmental performance and positively driving the stand-alone effects of CSR and social performance. Additionally, the coefficients of the control variables are generally consistent with prior literature. The positive (negative) relationship between firm size and ROA (ROE) is consistent with arguments about economies of scale (and the lack of innovation). Moreover, the positive significance of sales growth, board size, board independence and GDP suggests that fast-growing firms with larger boards are better monitored and that firms operating in countries with higher GDP growth enjoy higher performance.

These findings can be explained by instrumental stakeholder theory (Donaldson & Preston, 1995; Wood & Jones, 1995), which suggests that a firm can improve its reputation and reduce operating losses through effective CSR practices. Therefore, stakeholders view the CSR committee as an effective monitor of the firm's CSR investments, thereby bringing transparency to CSR activities and limiting the firm's spending on relevant, policy-based CSR initiatives. Consequently, reducing the wastage of firms' resources on irrelevant CSR activities improves firm profitability. In addition, socially responsible CSR activities and timely disclosure help firms earn a market premium. Our results align with prior findings that stakeholders obtain insurance through CSR committees to meet their CSR-related demands (Hussain et al., 2018). Furthermore, in a recent study, Baraibar-Diez and Odriozola (2019) reported differences in CSR activities between firms with CSR committees and those without. Therefore, we may explain our findings by noting that the CSR committee may align the firm's management with sustainable activities that are relevant to the firm's CSR policies and the expectations of its stakeholders.

Following Xiao and Xi (2023), this study uses different methodologies to address endogeneity: 2SLS regression, the GMM model and the PSM approach. First, the study uses instrumental 2SLS estimation to address endogeneity concerns arising from firm-specific variables that may determine CSR performance. Therefore, FP may be determined jointly by firm-specific variables and CSR performance. Furthermore, a firm's CSR performance may be influenced by external variables that could change in the relationship between CSR performance and FP. The level and direction of a firm's CSR activities may be influenced by the CSR practices of the industry in which it operates. Thus, we select the industry-year mean of CSR as an instrumental variable, following Xu, Wu, and Dao (2020), and apply 2SLS regression to address endogeneity.

Panel A of Table 4 reports the findings using 2SLS regression. Models 1–3 in panel A of Table 4 report the findings from the first-step regression that use the industry-year means of CSR, ENV and SOC scores as instruments, along with the firm-specific control variables used in the baseline regressions. The instrumental variables (Ind_csr, Ind_env and Ind_soc) are positively associated with CSR, ENV and SOC in models 1–3. Table 4 also shows the findings from the second-stage regressions using predicted values from the first-stage regressions, with ROA as the dependent variable in models 4–6 and ROE in models 7–9. Overall, the second-stage regression findings are consistent with the baseline results. Our variables of interest (CSR×CSR_COM, ENV×CSR_COM and CSR×CSR_COM) are positively significant with the coefficients (0.085, 0.094 and 0.052) at P < 0.01 in models 4–6, respectively, while the coefficients (0.108, 0.109 and 0.056) are significant at p < 0.05 in models 7–9, respectively. The coefficients of CSR_COM are significantly negatively related at P < 0.01 in all models, while the coefficients of ENV_P are only significant at p < 0.01 (p < 0.05) in models 5 and 8. These findings support our hypothesis that monitoring the CSR committee's CSR performance enhances the firm's overall performance.

We performed several post-estimation tests to further support our choice of instruments. First, we performed an endogeneity test to confirm that our treatment of CSR as an endogenous variable was correct. The p-value reported by the Wu–Hausman test for all three CSR components is below 1%, suggesting that the null hypothesis of the exogenous variable is rejected for all three components of CSR. Hence, treating CSR as an endogenous variable is justified. In addition, we performed the weak-instrument test at the first stage. Again, the p-values of all three instruments are below 1%, suggesting that the null hypothesis of a weak instrument is rejected. Furthermore, the partial R-squares (unreported) show that the correlation between each pair of endogenous and instrumental variables is close to 10% in all cases, suggesting that the instruments are unlikely to be correlated with the error terms.

The 2SLS regressions address the selection of CSR activities in response to the firm's industry. However, they ignore endogeneity arising from other factors, such as omitted variable bias, reverse causality and time-invariant variables. The literature provides evidence that the two-step system GMM is a superior estimator (Antoniou, Guney, & Paudyal, 2008) in the presence of endogeneity, heteroskedasticity and autocorrelation (Baum, Schaffer, & Stillman, 2003) among dynamic panel data methods. Therefore, we also use the two-step system GMM as we cannot assume that our set of variables is free of omitted variable bias. Furthermore, CSR performance requires resources; therefore, firms are more likely to engage in higher levels of CSR activities if they have better financial performance. If this is true, it is purely a case of reverse causality. Finally, our variable of interest (CSR Committee) is time-invariant. For example, it is more likely that a firm may carry a CSR committee in all the following years after incorporating such a committee.

Panel B of Table 4 reports the findings of the two-step GMM regressions using ROA and ROE as the dependent variables. The significant coefficients of the lagged-dependent variables confirm the dynamic nature of the models. Moreover, the significance of the AR1 and the insignificance of AR2 and Hensen J. statistics validate the fitness of the two-step system GMM regressions in models from M1-M3. Overall, the findings of panel B in Table 4 are consistent with baseline findings. Our variables of interest (CSR×CSR_COM, ENV×CSR_COM and SOC×CSR_COM) are significantly positive with coefficients (0.508, 0.172 and 0.204) and (1.731, 0.701 and 0.657) at P < 0.05, P < 0.01 (Models 1–3) and P < 0.05, P < 0.05, P < 0.01 (Models 4–6), respectively. The coefficients of CSR, ENV and SOC are −0.648 (P < 0.01), −0.246 (P < 0.01) and −0.317 (P < 0.1) in models 5 and 6, respectively, suggesting that all dimensions of CSR are negatively associated with firms without a CSR committee.

We also performed a PSM-based analysis to address the potential selection bias in the sample. The existence of a CSR committee within a firm may depend on the firm's observable characteristics, potentially leading to selection bias. The PSM approach was first introduced by Rosenbaum and Rubin (1983) to mitigate selection bias by randomising the sample. To apply this approach, we first create treatment groups (firms with a CSR committee) and the control group (firms without a CSR committee), as indicated by the variable (TREAT), with a value of 1 indicating firms in the treatment group and 0 otherwise. Second, we estimate a propensity score using a logit model by regressing TREAT on all the control variables, i.e. leverage, size, sales growth, R&D, board size, board independence, CEO duality and GDP. Third, we randomise the sample by matching firms with CSR committees to firms without CSR committees using nearest-neighbour PLS (one-to-one). Finally, the influence of the CSR committee on the association between CSR and FP is re-estimated using the PSM sample. To confirm the quality of the matching, the variables for the treatment and control groups are compared using t-statistics for differences, as reported in Panel C of Table 4. The insignificant t-statistics for almost all variables indicate that the treated and control groups are qualitatively similar. The sample is consistent with the requirements of PSM regression. Panel D of Table 4 reports the PSM regression results, showing that the coefficients (e.g. CSE×CSR_COM, ENV×CSR_COM and SOC×CSR_COM) are positive and significant across all models with ROA and ROE as dependent variables, indicating that the findings from the PSM sample are qualitatively similar to the baseline results.

Previous research suggests that firms in countries with stakeholder-oriented policies need to legitimise their socially responsible activities to their stakeholders to a greater extent. Conversely, firms located in shareholder-oriented countries prioritise value creation for shareholders. Moreover, in shareholder-oriented countries, the influence of various stakeholder groups on corporate CSR-related actions is minimal, shifting firms' attention towards shareholders rather than stakeholders (Simnett, Vanstraelen, & Chua, 2009).

Given this difference, we re-estimate Equation (1) by examining the effect of CSR committees in stakeholder- and shareholder-oriented countries to address potential differences between the two samples. Building on Grassmann (2021), the sample is segmented into stakeholder- and shareholder-oriented countries using the common law and civil law (code law) classification. We classify each country as a common law or civil law, following Ball, Kothari, and Robin (2000) and Djankov, La Porta, Lopez-de-Silanes, and Shleifer (2008). Using ROA as the dependent variable, Column 1 of Panel A Table 5 shows that the interaction between CSR_COM and CSR (CSR×CSR_COM) is positive and significant in the stakeholder-oriented sample. At the same time, this effect is insignificant in shareholder-oriented countries. We also tested the impact of the CSR committee on the CSR's individual pillars. We found that the estimated impact of the CSR committee in the stakeholder-oriented sample is primarily captured by the social pillar (SOC) of CSR, suggesting that stakeholder-oriented countries prioritise social aspects of CSR. In contrast, the moderating role of the CSR committee in the association between the environmental pillar and FP is more pronounced in shareholder-oriented countries.

Panel B of Table 5 reports similar results to those in Panel A, using ROE as the dependent variable. The results generally reinforce the view that firms with a CSR committee benefit from CSR activities in stakeholder-oriented countries.

In this section, we assess the role of the CSR committee in enhancing the benefits of CSR activities across firms with high and low agency costs. Relying on the free cash flow hypothesis (Jensen, 1986), we divide our sample into firms with high and low agency costs, following Choi, Chung, Kim, and Lee (2021). Firms with free cash flow above the sample median are placed in the high agency cost group, while firms with free cash flow equal to or below the sample median are placed in the low agency cost group. Panel A of Table 6 reports the results using ROA as the dependent variable. Results reported in columns 1 and 2 of Table 6 show that the interaction between CSR_COM and CSR load positively and significantly affects FP in both high- and low-agency firms. However, the estimated coefficient of CSR_COM is approximately three times larger in low-agency firms than in high-agency firms. This difference is statistically significant, as indicated by the P-value for the difference (P-value <0.01). Similarly, the coefficients on ENV×CSR_COM and SOC×CSR_COM are significantly more pronounced in low-agency firms than in high-agency firms.

The findings in panel B, using ROE as the dependent variable, are qualitatively similar to those in panel A of Table 6, where the coefficients on the variables of interest are stronger in low-agency-cost firms across all models. These findings indicate that CSR committees are more effective in strengthening the CSR–FP relationship when the risk of free cash flow expropriation is lower. Low-agency-cost firms are typically more financially disciplined and face tighter constraints on discretionary spending, necessitating more accurate and proactive monitoring by CSR committees to ensure that CSR resources are allocated to value-enhancing initiatives. Conversely, in firms with high agency costs, CSR initiatives may be more financially feasible due to available slack, but weaker governance and oversight can undermine their strategic alignment with FP objectives.

Table 7 reports the results for the effect of the CSR committee across firms with CSR weaknesses and strengths. According to the Refinitiv rating, a CSR score below the 30th percentile is considered weak, and a CSR score equal to or above the 50^th percentile is considered strong. Based on this category, we further segregate our sample into firms with strong CSR and weak CSR. The results are reported in Panels A and B of Table 7, using ROA and ROE as the dependent variables. Overall, our cross-sectional investigations reveal that the CSR committee enhances the CSR–FP relationship in firms with weak CSR than in firms with strong CSR. These results imply that the CSR committee is more beneficial for firms that underperform in CSR than for their peer firms. This evidence aligns with Velte and Stawinoga (2020), who state that the CSR committee effectively responds to CSR-related problems from the various stakeholders, leading to increased CSR performance. Interestingly, the role of the CSR committee is insignificant in firms with strong CSR, indicating that the positive effect of CSR on FP does not primarily rely on the presence of the CSR committee.

Panel A of Table 8 presents the findings of three models, similar to those in Table 3, using the market-based performance measure Tobin's Q. The CSR, ENV and SOC variables were insignificant in all three models. The coefficients of CSR_COM are −0.417, −0.304 and −0.219, with P < 0.01, P < 0.05 and P < 0.01 in models 1–3, respectively. The coefficients (significance level) of interaction terms CSR×CSR_COM, ENV×CSR_COM and SOC×CSR_COM are 0.858 (P < 0.01), 0.542 (P < 0.01) and 0.462 (P < 0.01) in models 1–3, respectively. The overall findings in Table 8, Panel A, suggest that the baseline results in Table 3 hold when using an alternative proxy for the FP.

Panel B (C) of Table 8 reports the findings of fixed-effect regressions after controlling for industry (country) fixed effects. Overall, these results support our baseline findings. In addition, the directions and signs of the variable of interests (CSR×CSR_COM, ENV×CSR_COM and SOC×CSR_COM), CSR performance (CSR, ENV and SOC) and CSR_COM are similar to the baseline findings using both ROA and ROE as the dependent variables. These results further strengthen our hypothesis.

Panel D of Table 8 reports the results for the effect of the previous year's CSR and CSR committee (lagged) on FP. We capture the previous year's joint effect of CSR and the CSR committee by re-estimating our baseline model with lagged CSR, the CSR committee, their interaction terms (CSR×CSR_COM, ENV×CSR_COM and SOC×CSR_COM) and all control variables. The lag-variable approach is essential for addressing reverse causality and is widely acknowledged in the literature. Our results remain aligned with baseline results.

Thus far, we have shown that the incremental value of CSRs depends on the existence of a CSR committee as it is associated with building stakeholders' trust through effective planning and the initiation and communication of CSR activities. However, the question of how a CSR committee enhances the CSRs' contributions to FP needs to be explored. In this additional set of analyses, we explore the channels through which the CSR committee is likely to influence CSR and its contribution to FP. First, from a theoretical perspective, establishing a CSR committee is useful for gaining CSR-oriented legitimacy (Patten, 2019) as it encourages firms to prepare voluntary CSR reporting. Consistent with these arguments, Velte and Stawinoga (2020) reveal a positive impact of CSR committees on CSR disclosure and performance. Moreover, the CSR committee also facilitates firms in making effective CSR strategies, enhances CSR performance and reduces CSR-related controversies (Elmaghrabi, 2021).

Building on these findings, we examine whether the CSR committee enhances CSRs' contribution to FP by increasing CSR disclosures and mitigating CSR-related controversies. We specifically test this conjecture in two different ways. First, we test whether the CSR committee potentially promotes CSR disclosures and mitigates CSR-related controversies. This investigation is conducted to corroborate evidence in the literature on the role of CSR committees in CSR disclosures during controversies. Second, regarding channel analysis, we examine whether the CSR committee strengthens the CSR–FP nexus by increasing (decreasing) CSR disclosure (CSR controversies). We proxy for the CSR reporting using the GRIs, an indicator obtained from the ASSET4 database coded as 1 for firms that publish their CSR-related information according to the GRIs and 0 otherwise. GRI reporting aims to promote transparency in sustainability reporting; therefore, adopting GRI reporting guidelines presents firms' social and environmental disclosures (Einwiller et al., 2016).

The GRI reporting is a reasonable attempt to respond to the stakeholders' demands for CSR information (Nikolaeva & Bicho, 2011). Cai et al. (2012) defined CSR-related controversies as negative news stories concerning CSR activities, such as sceptical social behaviours and product-related irresponsibility that draw media attention to the firm. The variable CSR-related controversies (CSR_CONT) is measured by its percentile score from the ASSET4 database, which accounts for 23 CSR controversy themes, such as “business ethics controversies “and “controversies privacy” (for details, see Thomson Reuters, 2019). Consistent with our expectations and the existing literature, Table 9 shows that firms with CSR committees are more likely to increase CSR reporting (Column 1) and experience fewer CSR-related controversies (Column 2), as reflected by their respective coefficients.

In Table 10, we report results for the triple interaction effects (e.g. CSR_COM×DISC×CSR and CSR_COM×CSR_CONT×CSR) to further explain the potential impact of CSR_COM on the CSR–FP relationship via the CSR disclosures (DISC) and CSR-related controversies (CSR_CONT) channels. In Panel A, using ROA as a proxy of FP (columns 1–3), we check for the interaction effect of CSR with DISC without accounting for the impact of CSR_COM. The interaction coefficients (e.g. CSR × DISC, ENV × DISC and SOC × DISC) suggest that CSR disclosures alone cannot enhance CSR's contribution to FP in firms without a CSR committee. However, when accounting for the effect of CSR_COM, as indicated by the coefficients of the triple interaction in columns 4–6, the impact of the CSR disclosure and CSR on FP turns positively significant in all models. For ROE in Panel B of Table 10, we found only a marginally positive impact of the triple interaction with SOC, but the results are not significant for CSR and ENV. The results show that the CSR committee and CSR disclosure jointly augment the positive effect of CSR on FP only with ROA. Therefore, the collective benefits of CSR disclosure and CSR can be obtained when firms have a sustainability committee. Thus, it confirms our hypothesis (H2a) that the CSR committee influences the association between CSR and FP through CSR disclosure channel”.