Do chief sustainability officers and CSR committees influence CSR-related outcomes? A structured literature review based on empirical-quantitative research findings

Based on the described theoretical foundation and our deduction of three main CSR outputs, a research framework for the literature review was derived (Fig. 1). The research framework focusses on the impact of institutionalised sustainability board expertise via CSR committees and CSOs on three CSR-related activities: CSR reporting; CSRA; and CSR performance. Moreover, we are interested in moderator and mediator analyses performed in prior empirical studies on this topic. The implementation of CSR committees and CSOs can play a moderating or mediating role in the link between other corporate governance variables (e.g., board gender diversity) and CSR. Moreover, other corporate governance variables can moderate or mediate the link between CSOs and/or CSR committees and CSR. CSR measures, as dependent variables in our research framework, will be structured into three categories used mainly in prior conceptual and empirical research and which have been developed from our theories in the previous chapter:

In line with other recommendations in CSR research (Malik 2015; Velte 2017), a clear separation between performance and reporting, and between total CSR and subcategories (e.g., environmental issues), is useful to our analysis. Stakeholders should be informed explicitly about a company’s CSR activities. Most PIEs present separate CSR reports according to a variety of frameworks (e.g., Global Reporting Initiative standards), and shareholders and other stakeholder groups analyse these reports. According to our theoretical framework, firms like to gain legitimacy through their CSR reports. Moreover, over the years, financial figures, CSR reports and other information on firms are analysed by rating agencies to increase CSR ratings and scores as performance pillars. Our theoretical framework also indicates that firms intend to influence their CSR performance in a positive way. As CSR information may be connected with risks of greenwashing and information overload, it is crucial whether there will be positive stakeholder reactions. As stakeholders demand reliable and objective CSR reports, boards of directors will be active in engaging external parties to conduct CSRA. In line with our theoretical framework, we motivated the choice of these three CSR proxies with reference to the increased research activity on CSR reporting, CSRA and CSR performance (Velte and Stawinoga 2017a).

Table 2 provides a content analysis of the reviewed studies. We focussed on the basic characteristics of our sample of included literature, including time of publication, country, journal of publication, selection of independent variables (CSR committee and CSO) and dependent variables, such as CSR-related effects, theoretical frameworks and the methodology used in empirical-quantitative research. The first empirical-quantitative study was performed in 2010, and the field has grown considerably over the past two years (Fig. 2). Both cross-country studies (18) and analyses for the US-American capital market (12) were conducted (Fig. 2). Cross-country studies are relevant, as the governance environment is rather heterogeneous (e.g., shareholder rights) and should be included in research designs. The US-American setting seems to be an attractive research objective, as empirical research on top management and board committee composition has a long tradition, and high-ranking journals prefer listed US firms with regard to increased database quality. Most studies have been published in business ethics and sustainability journals (Fig. 3), especially the Journal of Business Ethics (6), Business Strategy and the Environment (4), Corporate Social Responsibility and Environmental Management (4) and the Journal of Cleaner Production (4). Prior research mainly focussed on CSR committees (41) compared with CSOs (5). The most relevant CSR effects are CSR reporting (22) and CSR performance (19). Research on CSRA (9) as a dependent variable is of little interest so far. We also noted a variety of different theories on recent empirical research on this topic. Stakeholder theory (20), legitimacy theory (13) and classical principal agent theory (11) are the most common theoretical foundations, whereas a few papers do not explicitly rely on a specific theory (5) (Fig. 2). Upper-echelons theory mostly is used in CSO studies. As CSR committee and CSO adoption is voluntary from an international perspective, most researchers use a dummy variable (presence = 1, otherwise 0). While Fig. 2 indicates that (pooled) Ordinary Least Squares (OLS) (17) and panel regressions (13) are most common, voluntary adoption of CSR committees and CSOs is linked with a self-selection (sample) bias. To address these concerns, propensity score matching (Rosenbaum and Rubin, 1983; Shipman et al. 2017), in combination with a difference-in-differences approach, is recommended in the literature. Only one study in our literature review uses propensity score matching (Kanashiro and Rivera 2019). As some studies use a dummy variable (e.g., CSRA, CDP participation), a probit or logistic regression was conducted. As the link between corporate governance variables and CSR is characterised by endogeneity concerns, the literature proposes a generalised method of moments (GMM), instrumental variables or two/three stage approaches (Wintoki et al. 2012). In our included studies, nine researchers explicitly address endogeneity in their designs.

Besides literature reviews, some researchers also conducted meta-analyses on the link between corporate governance variables and CSR measures. Byron and Post (2016) focussed on board gender diversity’s impact on CSR, while Ortas et al. (2017) concentrated on the influence of board independence on CSR. Majumder et al. (2017) and Velte (2019a, b) included several corporate governance variables and analysed their impact on CSR performance and reporting. In line with prior literature reviews, CSR committees and CSOs were not included in the meta-analyses. Regarding the restricted number of studies and the low number of cross-country studies, a meta-analysis on this topic would not be useful yet. In a recent meta-analysis, Endrikat et al. (2020) analysed the link between several board composition variables (board size, board independence, gender diversity and CEO duality) and CSR, and included the implementation of CSR committees as a mediator. Thus, in light of increased research and practical and regulatory relevance, our literature review on the link between institutionalised sustainability board expertise via CSR committees and CSOs and CSR effects makes a useful contribution to present research on sustainable corporate governance.

Table 1 presents the main proxies that have been used as CSR committee and CSO variables and CSR measures. As previously noted, both CSR committee and CSO adoption are approximated mainly by a dummy variable (1 = yes, 0 = no) because their implementation is voluntary from an international perspective. More detailed analyses of CSO characteristics and CSR committee composition (e.g., social and environmental expertise on the CSR committee) are not common yet. In the next chapter, we differentiate our analysis between 1) CSR reporting, 2) CSRA and 3) CSR performance. If the studies relied on CSR subgroups (e.g., environmental issues), we differentiate these results from overall CSR studies. In practice, CSR reporting and CSR assurance often are classified as subcategories of CSR performance, especially due to the use of external databases (mainly MSCI/formerly Kinder, Lyndenberg and Domini (KLD), Asset4 or Bloomberg). Thus, a successful CSR management system should promote CSR reporting, CSRA and CSR performance. In total, CSR committees, CSO variables and CSR outputs are connected with a heterogeneity of included proxies in prior research. Thus, comparability within the included studies may be limited partially.

In line with the heterogeneous use of institutionalised sustainability-related board expertise and CSR outputs, we identified inconclusive results on the impact from CSR committees and CSOs on CSR reporting, performance and assurance. Based on a theoretical framework, we concluded that CSR committees and CSOs will impact CSR outputs significantly. If managers’ motivation is intrinsic and relies on stakeholder accountability perspective (‘substantive’ CSR strategies), a positive impact on CSR outputs should be realistic. If managers’ motivation mainly is extrinsic and relates to the business case perspective (‘symbolic’ CSR strategies), we conclude that an insignificant, or even negative, effect on CSR outputs exists. To sum up, using a combination of legitimacy theory, stakeholder theory and upper-echelons theory, we conclude that CSOs and CSR committees will lead to better CSR performance, CSRA and CSR reporting.

Most of our included studies in this literature review focussed on CSR committees’ impact on CSR reporting or environmental reporting (22). However, no CSO study has been conducted yet. Most found a positive impact from CSR committees on CSR reporting, using cross-country samples (Adel et al. 2019; Amran et al. 2014; Gallego-Alvarez and Pucheta-Martinez 2020; Kend 2015; Michelon 2011; Pucheta-Martinez and Gallego-Alvarez 2019) and different countries: Italy (Cucari et al. 2018; Rossi and Tarquinio 2017); the US (Hussain et al. 2018); Pakistan (Mahmood et al. 2018); and Spain (Fuente et al. 2017). Moreover, CSR committee expertise (Rossi and Tarquinio 2017) and CSR committee duration (Garcia-Torea et al. 2016) were related positively to CSR reporting. Thus, in line with legitimacy theory and stakeholder theory, there are indications that CSR committees are not used as a symbolic management tool, yet substantially increase CSR reporting quality.

With reference to environmental reporting, there are also indications of a positive impact from CSR committees, according to cross-country studies (Faisal and Achmad 2014) in Turkey (Kilic and Kuzey 2019), Italy (Jaggi et al. 2018a) and the UK (Helfaya and Moussa 2017). Berthelot and Robert (2011) found a positive impact from CSR committee size on Canadian firms’ environmental reporting.

However, some researchers also found insignificant impacts from CSR committees on CSR reporting (Alshbili et al. 2019; Dilling 2010; Elsayih et al. 2018; Michelon and Parbonetti 2012) and environmental reporting (Berthelot and Robert 2011; Rankin et al. 2011). Liao et al. (2015) and Berthelot and Robert (2011) included CSR committee, size, independence and meetings, and did not find any significant impacts on environmental reporting. Thus, CSR committee implementation and composition may represent a symbolic board of directors strategy. Prior research on the impact from CSR committees on CSR reporting found heterogeneous results, while most of the included studies indicated a positive impact. According to the research questions derived from the theoretical foundation, CSR committees may be a substantive or a symbolic CSR management tool within the board of directors. As some regimes already had introduced mandatory CSR reporting rules, inconclusive results between voluntary and mandatory CSR preparers are realistic.

We identified three moderator analyses on the CSR committee-CSR reporting link. Highly polluting Italian industries (Jaggi et al. 2018a) and environmentally sensitive UK industries (Helfaya and Moussa 2017) positively moderate the relationship. Based on a cross-country sample, Adnan et al. (2018) stated that CSR committees change the negative culture (power distance)-CSR reporting link into a positive relationship.

We identified nine studies on CSR committees (7) and CSOs’ (2) impact on CSRA. Ruhnke and Gabriel (2013) found a positive impact from CSR committees on CSRA in their cross-country study. Based on civil law regimes, Martinez-Ferrero and Garcia-Sanchez (2017) found a positive impact from CSR committee meetings on both CSRA and audit firms as CSRA providers. Peters and Romi (2015) reported similar results regarding CSR committee expertise in the US. Conversely, Al-Shaer and Zaman (2018) found a negative relationship in the context of UK firms. Soh and Martinov-Bennie (2018), in examining the Australian capital market, referred to internal audit reporting’s influence with the CSR committee on annual internal work plans devoted to CSR assurance and consulting. The authors did not find any significant results. Kend’s (2015) results are also inconclusive, as CSR committees in the UK and Australia did not impact CSRA, but audit firms as CSRA providers. Thus, in line with our results on CSR reporting and according to the research questions derived from our theoretical framework, the results on the link between CSR committees and CSRA decisions led to heterogeneous results, which might be based on different regulatory settings of a CSRA engagement, which have been analysed in different empirical contexts. Considering that the number of such studies remains low, we recommend future research on this topic. Regarding legitimacy theory, we cannot clearly decide whether the implementation of CSOs and CSR committees represents a symbolic or substantive management tool.

With regard to moderator analysis, Al-Shaer and Zaman (2018) stated that the negative link between CSR committees and CSRA is changed to a positive relationship by audit committee independence and expertise. Thus, the interdependencies between CSR committees and audit committees are crucial.

In line with prior CSR committee research, CSO studies are also heterogeneous in their results and of low amount. In the US context, as examined by Peters and Romi (2015), CSO implementation and CSO expertise are linked positively with CSRA. Although CSOs, as sustainable experts, did not prefer audit firms, they preferred consultants as CSRA providers. However, CSO implementation did not impact CSRA reporting, according to Rossi and Tarquinio in Italy (2017). Thus, the research findings neither clearly support nor reject the assumptions derived from upper-echelons theory.

We identified 19 studies on CSR committees (14) and CSOs’ (5) impact on CSR and environmental performance. Thus, performance outputs represent the second highest effect in our literature review. With regard to CSR committee implementation, the majority found a positive impact on CSR performance. This results could be stated both in cross-country designs (Baraibar-Diez and Odriozola 2019; Birindelli et al. 2018) and in specific regimes in the US (Burke et al. 2019, based on CSR strengths) and in Australia (Biswas et al. 2018). Positive results also are related to environmental performance on a cross-country level (García Martín and Herrero 2019) for Turkey (Kilic and Kuzey 2019), the US (Dixon-Fowler et al. 2017; Walls et al. 2012) and the UK (Konadu 2017; Liao et al. 2015 based on carbon-emissions scope 1). Liao et al. (2015) referred to CDP participation in a UK sample and found a positive influence from CSR committee size, independence and meeting frequency. However, some researchers stated that CSR committees did not influence CSR performance (Burke et al. 2019; Lin et al. 2015 based on CSR concerns) or environmental performance (Konadu 2017 based on carbon emissions scope 2; Rodrigue et al. 2013; USA). In a cross-country design, Eberhardt-Toth (2017) found a positive impact from CSR committee independence, age and female chairs; a negative impact from CEO membership and size; and an insignificant impact from board chair membership on CSR performance. We also found heterogeneous results on the link between CSR committees and CSR performance, while most of the included studies found a positive relationship. Thus, there are indications that firms substantively use the implementation of CSR committees to strengthen their CSR activities. With regard to CSR performance, researchers include different databases with a variety of proxies. Moreover, we found country-specific differences and industry effects (e.g., environmentally sensitive industries) that could lead to inconclusive results. Thus, in line with legitimacy theory and stakeholder theory, there are indications that CSR committees are not used as a symbolic management tool, but substantially increase CSR performance.

We stress only one mediator analysis in our sample, in which CSR strategy mediates the relationship between CSR committees and CSR performance in a UK setting (Orazalin 2020).

Over the years, three studies also have included CSR committees as moderators of other corporate governance drivers of CSR performance. Benjamin et al.’s (2019) US study indicated that CSR committees negatively moderate the board independence-CSR performance link (sustainable supply chain responsibility). According to Baraibar-Diez et al. (2019), CSR committees in Europe positively moderate the link between CSR-related compensation and CSR performance. Furthermore, CSR committees in the US weaken the negative link between less-able CEOs and CSR performance (Garcia-Sanchez et al. 2019). Moreover, environmental team and board gender diversity (Konadu 2017) and firm performance (Lin et al. 2015) positively moderate the relationship between CSR committees and CSR/environmental performance.

In line with other effect variables, CSO studies are very scant, and their US-related results are inconclusive. While Fu et al. (2020) noted a significant positive impact from CSO on CSR performance, Peters et al. (2019) did not find any significant link. However, CSO non-expertise decreased CSR performance. According to Kanashiro and Rivera (2019), CSO implementation and environmental performance are negatively linked. Thus, in line with legitimacy theory, it remains unclear whether the implementation of a CSO represents a symbolic or substantive management tool.

We also stress three studies that included moderator analyses of CSO implementation. Kanashiro and Rivera (2019) found that environmental regulatory pressure positively moderates the CSO-environmental performance link. According to Peters et al. (2019), prior CSR performance represents a moderator of the positive impact from CSOs on future CSR performance. Finally, Dixon-Fowler et al. (2017)’s study is exclusive, as both CSO and CSR committees are integrated into one research design. The authors found that CSOs moderate the positive influence on environmental performance from environmental stakeholder representation on CSR committees.

Table 2 contains a detailed description of all included studies.

We see a variety of future research opportunities. First, we note that research on CSOs and the impact of sustainability-related board expertise on CSRA remains scant. To gain a deeper understanding of CSR effects, future studies should not only focus on environmental issues (e.g., carbon performance), but also analyse their influence on social performance and disclosure (e.g., diversity reporting), triple bottom line (TBL) reporting and integrated reporting (Velte and Stawinoga 2017b). Similar to how we noted that CSOs and CSR committees may be used as symbolic, sustainable corporate governance tools, more attention should be paid to the legitimacy approach towards specific CSO characteristics and CSR committee composition (e.g., independence, expertise, education, size, meetings, gender and cooperation with other corporate governance institutions, such as internal auditors, audit committees and external auditors). With regard to CSO research based on the upper-echelons approach, behavioural aspects also should be integrated (e.g., CSO overconfidence, narcissism, values or materialism) in line with recent CEO research (Tang et al. 2018). Furthermore, the connections between CSOs, CEOs and CFOs should be analysed. CSR strategies have gained much interest among sustainable investors over the years from an international perspective. Thus, the impact from CSOs and CSR committees on ownership structure and other stakeholder reactions (e.g., media pressure) represents a fruitful research pathway. In addition, effects other than CSR may be affected by sustainability-related board composition, e.g., reduced earnings, lower taxes or other financial consequences (e.g., increased financial performance, reduced cost of capital). Prior research mainly relies on the consequences of CSR committees and CSOs, with only a few addressing possible determinants (Gennari 2019; Gennari and Salvioni 2019).

While we identified the dominance of selected theories (legitimacy theory, stakeholder theory with regard to CSR committees and upper-echelons theory related to CSOs), some studies in our included sample did not rely on a specific theory (e.g., Dilling 2010). In addition, stakeholder theory and principal agent theory are used simultaneously to contrast the heterogeneous effects from sustainability-related board expertise on CSR. However, the combination of these two theories as stakeholder agency theory (Hill and Jones 1992) can be a useful theoretical contribution.

Moreover, we should note that behavioural agency theory has not been used in prior research on CSR committees and CSOs so far (Pepper and Gore 2015). As a result, further top management team actions depend on their perceived current positions. Bosse and Phillips (2016) further explained perceptions of fairness. Executives are bounded, self-interested actors; thus, they are driven not only by self-interest, but also are influenced by social norms, such as fairness or justification. A fair treatment by an agent will lead to increased CSR activities, but if bounded, self-interested agents perceive their treatment as unfair, they demand additional compensation for the sake of (perceived) injustice (Bosse et al. 2009) and to engage in CSR activities.

Another theoretical approach that is highly relevant in gender diversity research is critical mass theory (Kanter 1977), which also can be useful in future studies on sustainability-related board expertise via CSR committees and CSOs. In light of substantive CSR strategies, this theory can lead to further discussions as to whether having one person on the top management team as CSO is enough to implement a successful sustainable management system. Moreover, critical mass theory may contribute to future analyses on CSR committees’ composition, e.g., how many sustainability experts need to be included on the committee, or deciding whether the CSR committee should comprise executive or non-executive directors.

Prior research is related to many methodological restrictions that should be addressed in this section. First, sustainability-related board composition often is used as a dummy variable by indicating whether a CSR committee or CSO exists (yes = 1, no = 0). There should be more concentration on committee composition, CSO characteristics and incentives, especially to differentiate symbolic from substantive use of these sustainable corporate governance mechanisms. The recognition of dummy variables, such as whether firms have implemented CSR committees or CSOs, is linked with limited explanatory power, as the researcher cannot evaluate symbolic or substantive use of institutionalised board expertise. In line with prior corporate governance research, a positive connection exists among board efficiency variables, e.g., expertise, its members’ independence or activity and monitoring quality in line with stakeholders’ interests (Dienes et al. 2016). In particular, sustainability expertise and board members’ independence will lead to a strict focus on CSR-related monitoring activities and should lead to more substantive CSR activities and increased CSR outputs. These composition proxies lower the risk of greenwashing behaviour and information overload as symbolic uses of CSR institutions. Moreover, symbolic or substantive use of CSR committees and CSOs may be better analysed by including CSR assurance as CSR outputs. Symbolic use of CSR committees and CSOs is linked with a reduced management incentive to engage an external party for CSRA.

Second, we noticed few moderator analyses and only one mediator analysis in our included studies. There should be more research attention paid to CSR committees and CSOs as moderators and mediators of other corporate governance variables. Moreover, the interactions between CSR committees and CSOs should be addressed in future research designs. Third, researchers should conduct cross-country studies and include country-related governance items (e.g., enforcement strength, shareholder rights, case vs. code law regimes, and one- vs. two-tiered systems) and cultural aspects (Hofstede and Bond 1984). Fourth, as greenwashing policies and information overload may be linked to symbolic use of CSR committees and CSOs, moderators or subsamples of firms that include environmental management systems or ISO 14,000 certification are most useful (Chen 2005; Melnyk et al. 2003). Fifth, many of our included studies do not explicitly address endogeneity concerns, e.g., omitted variable bias, simultaneity or measurement error. Thus, future research should analyse bidirectional relationships between sustainability-related board composition and CSR, and implement GMM and instrumental variables, the Heckman approach or propensity-score matching (Wintoki et al. 2012). As the implementation of CSR committees and CSOs is voluntary in most regimes, a selection bias could be realistic.

Propensity score matching (Rosenbaum and Rubin 1983; Shipman et al. 2017), sometimes in combination with a difference-in-differences approach, was proposed to increase the validity of empirical-quantitative research. Propensity score matching addressed potential nonlinearities in the control variables and composed a control group that was similar to the treatment companies, but did not implement CSR committees and CSOs during the sample period. The difference-in-differences method differentiates the effect from adopting CSR committees and CSOs from the effect of firm characteristics, which normally is associated with the implementation of these board variables. Reversed causality and omitted variables also are problematic in empirical-quantitative research on sustainable corporate governance. While valid instrumental variables should induce changes in the proxies related to CSR committees and CSOs, they do not independently affect CSR outputs, allowing the researcher to uncover these proxies’ causal effect (Wintoki et al. 2012). With regard to CSR committee composition, prior studies included linear regression models, but a non-linear relationship ((inverted) U-shape) may be more realistic. Sixth, most of our included studies relied on external CSR and corporate governance databases. A mixed model approach is recommended to analyse corporate governance and sustainability reports, as well as to increase the validity of CSO, CSR committee and CSR variables. Empirical-qualitative research via interviews, surveys and case studies also is important as complementary research designs to analyse organisational and processual aspects and challenges from internal and external stakeholders’ perspectives (investors, customers, suppliers).