Former CEOs chairing the board: does it matter to corporate social and environmental investments?

Former CEOs who stay on the board as Chairmen (i.e., Chair-Former-CEO or CFCEO) often play a vital role in monitoring and advising the incumbent CEOs. However, their influence on firm performance remains under-investigated. This paper aims to offer new insights into the impact that such a role can have by examining corporate investment in social and environmental responsibility. It examines the effect of CFCEOs on the firm’s social and environmental responsibility of 1,263 S&P1500 firms from 2002 to 2021. We find that firms with the presence of a CFCEO exhibit superior social and environmental performance. This finding suggests that CFCEOs can encourage long-term value creation for a broader range of stakeholders by building social capital and public trust. Additional analyses reveal that the positive association between the CFCEO and firms’ social and environmental performance was more pronounced during the COVID-19 pandemic than during the global financial crisis of 2007-9.


Introduction
This paper explores the impact of leadership turnover on corporate commitment to social and environmental responsibility during normal and turbulent times.In addition to ensuring their companies' operational and financial performance, firm leaders often face many additional challenges, such as societal inequality, climate change, financial crises, and even health crises, such as the recent pandemic (AI-Shaer et al. 2023;Cheah and Lim 2023;Chen et al. 2023;Unsal and Hassan 2023;Zhao et al. 2023).Even though such challenges may well be beyond the remit of any individual corporate leader and can be seen as defocusing CEOs from the main task at hand, contributing to addressing them can result in spillovers that can benefit the firm.Companies with superior Corporate Social Responsibility (CSR) performance are more likely to gain more favourable access to finance, lower cost of capital, higher customer loyalty, stronger competitive advantage, and hence a larger market share, even during times of crisis (see Albuquerque et al. 2020;Badía et al. 2020a;Ding et al. 2021).On the other hand, however, CSR may represent an agency problem as it may be undertaken by corporate management at the expense of shareholders' wealth (Bae et al. 2021;Chen et al. 2018;Manchiraju and Rajgopal 2017).Indeed, the prior literature has provided evidence showing that CSR can be driven by managerial narcissism to benefit executives' own interests through e.g.empire-building to increase their reputations and influence their industry (see Gul et al. 2020;Petrenko et al. 2016;Tang et al. 2018).Nevertheless, the board of directors can mitigate the CSR-related agency problem, as effective corporate governance mechanisms can reduce the possibility that executives have a self-interested CSR orientation (Aguilera et al. 2021;Badía et al. 2020b;Wang et al. 2023).
Such issues may be of higher relative importance when there is a change in leadership.Newly appointed CEOs are expected to develop a firm's future strategies on CSR (Georgakakis and Ruigrok 2017).New CEOs are self-motivated to demonstrate to the board and shareholders their capability to lead the company by aligning firm operations with industry dynamics and a constantly changing business environment (Quigley and Hambrick 2012).When it comes to voluntary CSR activities, which represent a long-term strategic investment with an unforeseeable and intangible outcome, new CEOs may cut down CSR spending to invest more in other projects that can generate immediate returns.Still, reducing CSR investment may affect the social capital that the former leader has generated via committing to long-term value creation (Lins et al. 2017).In such a case, a board can restrain new CEOs who try to execute excessive changes to existing CSR practices.The board's effectiveness in steering the new CEO's behaviours could be potentially enhanced by the presence of the former CEO, who stays on the board as Chairman (i.e., Chair-Former-CEO, or CFCEO) (Quigley and Hambrick 2012).We follow DataStream to define the CFCEO as a Chair who was CEO in previous years and CnotFCEO as a Chair who was never the CEO of the firm or is currently also the CEO.
Prior literature has provided ample evidence that the board of directors (Amin et al. 2020;de Viller et al. 2011;Oh et al. 2019) and CEO leadership (Waldman et al. 2006;McCarthy et al. 2017;Hegde and Mishra 2019) play a central role in determining the intensity of CSR activities across distinctive institutional contexts.However, such studies have not examined the impact that the presence of a CFCEO can have on its CSR commitments and strategy.CFCEOs have extensive work experience and a deep understanding of the firm's strengths and weaknesses and external contextual dynamics.In fact, according to a recent report, a CEO is more likely to be promoted to chair the board if they have achieved not only excellent records of running the company but also shown commitment to creating long-term value for shareholders, the company, and the wider society (Quigley and Hambrick 2012).Therefore, to maintain continuity, the CFCEO may monitor, advise, or encourage and collaborate with incumbent CEOs to sustain a long-term commitment to investing in CSR activities and relevant social capital creation.
Given the above gap, the primary purpose of this article is to empirically examine the association between the presence of a CFCEO and CSR.We posit that the presence of a CFCEO can encourage the successor CEO to remain committed to implementing existing CSR strategies, enhancing social capital and the trust conferred by a broader range of relevant stakeholders.CFCEOs can play a vital role in encouraging new CEOs to cherish practices that may take a long time to establish and create value for all relevant stakeholders.Based on a sample of US S&P 1500 firms characterised by efficient and established corporate governance mechanisms, our findings present strong evidence of the significantly positive role of CFCEOs in increasing social and environmental investment intensity.
We also find different effects of CFCEOs on such performance during exogenous shocks.Specifically, our analyses show that the CFCEO can maintain and increase CSR intensity by improving social and environmental performance.This effect is more pronounced during the COVID-19 pandemic, caused by an unforeseen coronavirus disruption, compared with the global financial crisis of 2007-9, which resulted from unethical business behaviours.In the case of the financial crisis, firms either maintained the same CSR levels or reduced their latitude due to the shortage of slack resources (Fehre and Weber 2016).A possible explanation is that the board had no choice but to approve cuts to social and environmental spending in response to their jeopardised operations, cash flows and threats to survival.However, the negative consequences of the global economic recession have raised public awareness of the relationship between business and society.To rebuild social connections with a broader range of relevant stakeholders, many businesses decided to invest more resources in social and environmental activities in the aftermath of the financial crisis in 2007/09.As such, we would argue that when the COVID-19 pandemic hit in early 2020 the CFCEOs and other board members may have encouraged the current leadership to sustain the firms' social capital by maintaining long-term value creation through increased CSR engagement.Also, given the unprecedented nature of the pandemic, business leaders may have underestimated how long this would last and expected only a short-term impact.As such, they did not change their long-term orientation.By doing so, companies can maintain social capital and enhance public trust in their operations.
Our study makes some important contributions and has practical implications.Firstly, our study contributes to the literature on leadership turnover by enriching our understanding of how board chairman succession can impact social and environmentally responsible activities.To this end, we empirically test whether former CEOs who are voted to chair the board of directors, due to their ongoing commitments to creating more long-term value for the focal firms, can continue to supervise and advise incumbent CEOs to maintain a superior CSR performance.We argue that the CFCEOs may supervise, advise, and collaborate with incumbent CEOs more effectively than new ones because the latter may be motivated to achieve a short-term return to show off their capability at the expense of the firm's longterm value.
Secondly, we contribute to the corporate governance and CSR literature by examining the role that CFCEOs play in influencing firms' strategic changes relating to CSR intensity.By adopting the theoretical lenses of two theories, we conceptualize and empirically test how a CFCEO could affect the performance of firms' CSR activities.As such, we contribute to CSR research (e.g., Kumar et al. 2019;Brunton et al. 2017;Miras-Rodríguez et al. 2014) by offering strong evidence that a CEO's retention as the Chairman can either maintain or increase the company's investment in social and environmental activities, benefiting all relevant stakeholders.Previous literature has offered ample evidence that firm leaders can affect their firms' CSR practices.However, it focuses on individual leaders' features, experience and personalities (e.g., CEO and other top executives' characteristics, past life and work experiences).However, limited studies (e.g., Quigley and Hambrick 2012) pay attention to the significant role played by the board chairpersons, whose main tasks include ensuring corporate compliance with corporate governance systems, setting up the agenda, leading effective communication with CEOs and other executive and independent directors, facilitating the discussion, and maintaining a healthy atmosphere in the boardroom.As such, our study contributes to CSR literature by offering robust evidence that the board chair succession and the relationship (and interactions) between the board chair and CEO can have a noticeable influence on voluntary investment decisions about social and environmental responsibility, which may result in benefits in various forms in the long run.
Thirdly, our work adds to corporate governance and CSR research in times of crisis by comparing and contrasting the impact of the global financial crisis and the COVID-19 pandemic on the link between the CFCEO and CSR performance.We provide strong evidence showing that after the 2007-2009 financial crisis, top corporate executives (including Board Chairs and CEOs) and shareholders became aware of the relevance of CSR to their businesses.Meanwhile, CEOs who focus on creating and maximising the long-term value of the firm are more likely to be selected to be the chairman in the boardroom and will continue to play a key role in monitoring, supervising, and working with incumbent CEOs to maintain superior CSR performance to enhance the connection with relevant stakeholders, preventing the adverse impact of future social and economic crises when short-termism undermines firms' value in the long run.Indeed, our findings demonstrate that after the financial crisis in 2007-2009, when former CEOs became the board chairs, their firms appeared to maintain a positive CSR investment regardless of specific purposes.Collectively, CFCEOs and other top executives seem to act ethically for intrinsic, not just instrumental profit-oriented, values when continuing their CSR investment decisions, even when they faced the recent time of crisis, i.e., the COVID-19 pandemic health crisis, which caused a devastating impact on social and economic spheres and a market crash across the world.
Finally, our empirical findings offer valuable practical implications.For example, if shareholders and board members select a former CEO to chair the board, she or he is more likely to have shown their commitment to creating long-term value for the firm.That can, in turn, lead to the synergy of CFCEOs' valuable work experience and public trust due to their persistent CSR commitment and the knowledge and capability of new CEOs, regardless of whether they are appointed internally or externally.This will increase or maintain the firm's social and environmental investment.

Research background, theoretical discussions and literature review
Firms do not operate in isolation from the external world.They have relationships with many constituent groups, and these stakeholders both influence and are influenced by the firm's actions.These stakeholders can determine the success or failure of a modern business enterprise (Maharaj 2008).This is why it is incumbent on the organisation to consider all stakeholders' interests and needs.Therefore, the board of directors, as the "supra-top management team" (Finkelstein et al. 2009;p.22),should act as the chamberlain of stakeholders' interests simultaneously by overseeing management, reviewing the firm's financial performance, approving the allocation of funds, and ensuring compliance with the law and corporate responsibility matters (Main et al. 1995).
Agency theory (see Jensen and Meckling 1976;Fama and Jensen 1983) suggests a conflict between the interests of the principals (shareholders) and those of the agents (managers, as self-interested actors) running the company on their behalf.It principally supports the CEO's monetary motives to align the financial interests of shareholders and managers.As a control mechanism intended to deal with the conflict of interests, the board of directors' key role is to mitigate agency conflicts, reduce the information asymmetry between management and shareholders, and lessen agency costs through governance structures and financial incentives to influence CEO behaviour (Sajko et al. 2021).However, boards regularly fail at this fundamental task (see Gilley et al. 2019), and the agency perspective has been challenged (see McWilliams and Siegel 2001).One important aspect that has not been considered is the role of a heterogeneous board's abilities to improve its monitoring function (Hillman and Dalziel 2003).Board members bring various qualities that may impact firm decisions and performance.An effective board will have a sound balance of well-chosen, competent directors with firm-specific knowledge, experience, skills, and expertise essential for effective governance to meet the rapidly changing global marketplace (Harper 2007).Pfeffer and Salancik (1978) state that "when an organization appoints an individual to a board, it expects the individual will come to support the organization, will concern himself with its problems, will favorably present it to others, and will try to aid it" (p.163).Boards exercise independent control and serve as strategic consultants to top managers.
While agency theory emphasises managerial opportunism, agency costs, and the board's role as a control mechanism, it pays less attention to the innate and external characteristics of individual executives.To address this, one could consider the upper echelons theory, which focuses on the importance of powerful organisational actors' psychological and other observable characteristics in interpreting the external environment, forming firms' strategic decisions and organisational outcomes (Hambrick 2007).According to this theory, "organizational outcomes-strategic choices and performance levels-are partially predicted by managerial background characteristics" (Hambrick and Mason 1984;p.193).One of the main tenets of upper echelons theory is that the choices of decision-makers can vary broadly, making it possible for them to inject their unique features (e.g., leadership qualities) into such strategic decisions to impact firm performance (Waldman et al. 2004).Supporting this view, McGuire et al. (2003, p.343) argue that "managerial beliefs and discretion, rather than the constraints and incentives provided by corporate governance, are likely to be the principal drivers of exemplary social performance".In this sense, top executives play a vital role in interpreting social and environmental activities.Upper echelons theory suggests that the managerial background characteristics of top managers predict strategic decision-making processes and ultimately organizational outcomes, such as financial performance (Hambrick 2007;Finkelstein et al. 2009;Abatecola and Cristofaro 2020).According to their interpretation of reality, executives make strategic choices stemming from their cognitive base values, functional tracks, career experiences, beliefs, perceptions, personalities, educational backgrounds, and ethical conduct norms.As such, corporate governance is a multifaceted task that occurs through individual executives' attributes and their combined impact (Abatecola and Cristofaro 2020;p.117).According to prior studies, CEOs' observable background, experience, and personal characteristics are essential proxies for their underlying psychological properties (Hambrick and Mason 1984).For instance, Manner (2010) found that CEOs' educational background, career experience, and gender are positively related to exemplary CSR ratings.Slater and Dixon-Fowler (2009) conclude that CEOs with international assignment experience and an output functional background are positively associated with greater CSR.Firms led by CEOs with MBA degrees are more likely to disclose environmental information voluntarily (Lewis et al. 2014), and hence, positively impact on a firm's CSR performance (Sun et al. 2021).Similarly, career experiences, one of the observable socio-demographic variables originally suggested by Hambrick and Mason (1984), can be expected to significantly affect the types of action taken by a CEO.
Apart from the CEOs' observable traits and characteristics, a Board chair who formerly served as the CEO of the same firm might be advantageous because of the in-depth industry familiarity and knowledge gained with experience and working team relationships (Hambrick and Mason 1984).Such experience and expertise can be invaluable when it comes to externally facing activities related to corporate social responsibility.Social and environmental investment intensity is the consequence of a long-term strategic orientation since socially responsible investments have uncertain outcomes and require a long time to pay off (Oh et al. 2016).Therefore, it could be the outcome of the strategic decision-makers' characteristics associated with career experience and risk-taking ethos to bear the uncertainty.
CSR is the extent to which organisations actively engage in social activities that respond to stakeholders' needs (Mäkinen and Kourula 2012;Schwartz and Carroll 2003) and develop sustainable systems that affect corporate economic performance (Duque-Grisales et al. 2021).Growing awareness of environmental protection, product safety, and continued global economic integration has made CSR a vital initiative to gain and maintain legitimacy, cope with challenges, and minimise corporate risk (Sun et al. 2021).As such, responsiveness to CSR/ESG matters has become essential to firms' overall value-creation efforts.CSR/ ESG performance is a strategic decision underpinned by an ethical stance of firms towards stakeholders and the larger society.ESG can affect a company's ability to achieve its business strategy and create value.Companies investing in ESG activities can enhance shareholder value by reducing the stock volatility and the firm's risk (Badía et al. 2022), improving reputation and competitive advantage and reaching new markets while contributing to society's sustainable development.
Given our research objectives, we follow existing CSR literature (e.g., Servaes and Tamayo 2013) by focusing on the environmental (E) and social (S) dimensions of ESG ratings.The governance (G) dimension is not included, as it should not be considered a CSR merit (see Lins et al. 2017).Previous literature, albeit limited, has suggested that if the former CEOs stay on the board as the chair, the board of directors appear to be more effective in monitoring, supervising, and advising the succeeding CEO in leading the company to avoid aggressive strategic change and maintain a stable firm performance (Quigley and Hambrick 2012; Krause et al. 2014).Our study contributes to this body of literature by investigating the potential effects of the presence of CFCEOs, i.e., the board of directors chaired by the former CEO of the same firm, on the social and environmental performance of S&P 1,500 firms between 2002 and 2020.To mitigate the endogeneity issue, we incorporate two exogenous shocks to test the effects between CFCEO and ES scores during two times of crisis, i.e., the financial crisis of 2007-9 and the COVID-19 pandemic 2020.

CFCEO versus CnotFCEO roles
Figure 1 illustrates the theoretical comparison between the roles of CFCEO and CnotFCEO in the context of our study.Specifically, the CFCEO is a Chair who was previously the CEO of the firm (either promoted as soon as their CEO tenure ended or was the CEO in the past), while a CnotFCEO is either a Chair who has never been the CEO of the firm or also simultaneously serves as the CEO (DataStream; Kanadlı et al. 2020).Retaining CEOs as Chairmen could impede successors' discretion, hence controlling the incumbent CEOs' ability to make strategic changes or deliver performance that differs from pre-succession levels (Quigley and Hambrick 2012).As argued by Quigley and Hambrick (2012), "predecessor retention will tend to occur if the board welcomes the former CEO's continued influence; conversely, predecessor departure will tend to occur when the board believes there is a need for change or when the predecessor's regime has been somehow repudiated" (p.835).In addition, Oh et al. (2016) found that when CEOs get older, they tend to disengage from proactive CSR activities, due to their shorter career horizons, defined as "a psychological assessment of career security over career termination (i.e., retirement)" (p.279).In this sense, the potential for promotion and acquisition of additional decision rights within the organization can provide more performance incentives for CEOs to create firm value and maximize stakeholders' interests.

Fig. 1 Theoretical Framework
A CFCEO who has experience of being the CEO of the focal firm is expected to have accumulated human and social capital and many years of experience preceding their promotion/appointment to be the Chair of the board (Burt 1997).Investment in specific human resources (e.g., context-specific knowledge, industry-specific experience, talents, skills, reputation, and expertise embedded within a director and developed through day-to-day jobrelated experience) leads to improved performance and increased stakeholder value.They serve as intangible assets and take varying forms, including knowledge of organisational culture and operations and management practices, the capability to monitor and manage people, and the ability to adapt and innovate in the face of changing conditions.Also, boards with more human capital could better supervise and advise managers to reduce risk-taking behaviours (Zhou et al. 2019).Social or relational capital is the set of resources existing in relations between board members and senior executives, staff, other firms, and relevant stakeholders.For example, it can include the valuable information available to the board through the internal and external network of social connections and public-private relationships possessed by a director (Kroll et al. 2008;Kor and Sundaramurthy 2009).CEOs have social capital, with various resources that may be accessed for the organisation's good and contribute to better performance.They can use social capital to resolve conflicts, encourage better communication, and advocate a solid commitment to the organisation.Carpenter and Westphal (2001) conclude that boards with directors with functional backgrounds and external network ties to strategically related organisations improve the board's monitoring function and provide better advice and counsel, thereby contributing to the strategic decision-making process, including long-term CSR spending.
As argued above, human and social capital are essential "cognitive competencies" for a firm.Internal promotion made by a board with a high degree of human and social capital (i.e., a distinct set of skills and proper incentives) is likely to be based on the CEO's strong performance, is perceived as the board's vote of confidence in the CEO's ability and is highly regarded by the stock market (Jayaraman et al. 2015).It is linked to protecting the organisation's current cognitive base and sustaining its strategic stability.Delaying promotion increases the risk that the CEO will move to another firm and may also be detrimental to shareholder value (Jayaraman et al. 2015).Many CEOs continue to serve as the Chair on the board, while low-performing CEOs are less likely to hold board seats after leaving office (Brickley et al. 1999).Jayaraman et al. (2015) argue that organizations are expected to reduce risk following CEO promotion.Dedman (2016) examined the UK Corporate Governance Code's recommendation that CEOs should not become Chairmen of the same firms because this practice would harm firm performance.More specifically, she analysed a sample of 225 CEO routine departure events from 1996 to 2007, finding that making it possible for the CEO to remain the Chair did not cause any damage to accounting or stock market performance.She also concluded that firms were more likely to recruit CEOs from better-performing firms before the UK Corporate Governance Code was released.Retaining good CEOs explains why asset divestiture is less likely when the former CEO becomes the Chairman of the same company (Dedman 2016).
To our knowledge, extant research from the upper-echelon perspective has paid little attention to the relations of CFCEO to CSR performance.Career experiences and functional-track orientations shape the lenses through which executives view current strategic opportunities and problems (Hambrick 2007).As CEO characteristics can influence their cognition, interpretation, and strategic choices and ultimately affect organizational results (Hambrick 2007), we propose that the CFCEOs adds their career experiences and cognitive values into the monitoring and advising mechanisms of the board of directors, determining the development of CSR investment strategies and actions.
In this study, we conjecture that former CEOs are promoted to chair the board due to their track records of long-term value creation through building strong connections with relevant stakeholders.Therefore, we predict that organisations led by a CFCEO tend to consider higher social and environmental performance as a strategic opportunity which can add value to the company.When a CFCEO is present, their firms are likely to view CSR commitments as a responsible business activity and a strategic opportunity for their organisations to gain and improve a social and environmental reputation.This may also help maintain an economic reputation among financial shareholders and a positive image among other relevant stakeholders.
In contrast to the above, a CnotFCEO may have either little or no experience of the firm and even the industry (i.e., a Chair who was never the CEO of the firm) or excessive discretionary power, leading to a severe agency problem (i.e., a Chair who currently also serves as the CEO or Chair-CEO duality).In both cases, their influence on corporate decision-making and policies is likely to be less effective than that of the CFCEO.For example, the Chair-CEO duality is more likely to weaken the Chair's role in improving their decisions because of the disparity of the Chair and CEO roles (i.e., the case of the CFCEO or CnotFCEO) tends to encourage transparency and promote the division of duties between them.Consequently, the CFCEO or the CnotFCEO of the firm might perform better than a firm leader who is simultaneously the Chair and current CEO.When comparing the CFCEO and the CnotFCEO, we argue that the CFCEO should have more industry and company knowledge.These advantages from their prior CEO experience could strengthen the board's monitoring and advising functions.At this point, we need to consider why the former CEO was offered the current chair position.As noted earlier, such an internal promotion can be based on past superior performance and long-term sustainability orientation, which align with the shareholders' expectations.Therefore, shareholders may want to vote for them to become the Chair with an expectation that the CFCEO could maintain what they have achieved through supervision and advising responsibilities over the decision-making process of incumbent CEOs.
Although the former CEO serving as the chair is viewed as beneficial to the firm because (s)he used to appear successful and influential in managing a firm and can effectively execute policies for long-term value creation (Fahlenbrach et al. 2011), they could be detrimental to firm performance because they exercise negotiation power to be reappointed by the firm for the purposes of entrenchment.As such, Evans et al. (2010) argue that the successor CEO usually exhibits a weaker ability to initiate new policies to maintain long-term value creation for the firm, thereby continuing to fail to implement policies or consume resources in non-value-added ways.However, CSR reflects the firm's long-term value-creation activities and strategies, so it should differ from the operating/financial performance mentioned in the literature (usually measured by short-term profitability measures).In addition, if the former CEOs failed to satisfy shareholders (especially the long-term value creation) and market participants, they may be less likely to be appointed as the Chairman.One could argue that they can use their negotiation power to be reappointed.However, this agency issue may be monitored and reduced by the boards as it may send a negative signal to the US market, which is highly efficient, destroying their market value.Given that the context of our study is large US firms, the policies and regulatory environment are more established, developed, and transparent.Combining this argument with the nature of long-term and value-enhancing ES activities, it is less likely that most firms appointed the former CEOs based on agency problems (i.e., negotiation power).Instead, the successful long-term strategies of the former CEOs are more likely to be the main reasons for their reappointment and promotion.
Taken together, compared to a CnotFCEO, we hypothesize that a CFCEO may participate in more social and environmental activities, leading to higher CSR performance.We, therefore, set our hypothesis as follows: H 1 The presence of a CFCEO is associated with the firm's higher social and environmental performance.

Different exogenous shocks and the CFCEO's CSR commitment
The effect of a CFCEO on firms' social and environmental performance may vary in times of exogenous shocks, such as the economic recession in 2007-9 and the market crash caused by the outbreak of the COVID-19 pandemic in 2020.Previous CSR studies considering the global financial crisis suggested that corporate executives tended to lessen the latitude of CSR activities, resulting in cuts in social and environmental investment due to the urgent shortage of slack resources (Fehre and Weber 2016).As the economic tsunami jeopardised the firm's operations and cash flows and endangered its survival, it was reasonable to expect the chairman and CEO to prioritise cost-saving and self-protection exercises.In this vein, investing in social and environmental activities is envisaged as an additional liability, which may result in the company's failure.Also, corporate managers and investors did not favourably value CSR when managerial short-termism was pervasive in the business domain around the early-and mid-2000s.
The unethical and greedy pursuit of profit maximisation by business entities was considered by the public as the cause of the financial crisis.Apart from stricter regulations and laws imposed by government authorities on businesses, the crisis spurred corporate leaders to amend business models, strategies, and decision-making processes by considering the interests of a wider range of stakeholders other than merely shareholders or investors.Since then, business leaders have continued to highlight the value of social capital and public trust as significant to the enduring prosperity of their company and survival, especially during times of crisis (Ding et al. 2021;Lins et al. 2017).Therefore, in the aftermath of the global economic crisis, many corporations started allocating more resources to social and environmental activities and committed to achieving long-term value instead of short-term returns (Lins et al. 2017).
The COVID-19 health crisis and subsequent national lockdowns caused a stock market crackdown and economic stagnation across the globe in 2020.However, this unexpected health emergency was not caused by corporate misconduct.When considering CSR investments, though, one argument could be that despite the different nature of the two crises, at such a challenging time, the CEOs should prioritize the firm's short-term survival by cutting down on non-essential expenditure (e.g., CSR).One could make a different argument too.Over the last decade, not only have business leaders learnt the relevance of CSR to strengthen their competitive advantage and long-term firm value, but also our society and the stock markets have gradually attached more credit to corporate socially and environmentally sustainable performance (Albuquerque et al. 2020;Azmi et al. 2021).As such, investors were likely to react to firms' CSR spending differently when COVID-19 erupted in early 2020, compared with their reaction to firm performance in the financial crisis in 2007-9.In addition, during the last decade, business leaders should have paid more attention to identifying their relevant stakeholder groups and their expectations and demands.
In this vein, we would argue that after the financial crisis in 2007-2009, firms' top executives (including Board Chairs and CEOs), and shareholders became aware of and drew more attention to the relevance of CSR to their business operations.As such, CEOs who are more concerned with the long-term value of the firm are more likely to be voted to chair the boardroom and will continue to play a key role in monitoring, supervising, and working with incumbent CEOs to maintain superior CSR performance in order to, for instance, strengthen the strong relation with their relevant stakeholders, avoiding the potential adverse impact of a further socio-economic crisis with short-termism jeopardising firms' long-term value.Our study aims to test when former CEOs become the board chair after the financial crisis, whether they tend to maintain a positive CSR investment regardless of a specific purpose -in other words, whether CFCEOs and incumbent CEO are more likely to act ethically for intrinsic, not instrumental, profit-oriented values when approving continued CSR investment decisions.
In summary, we posit that the CFCEO is likely to work with or monitor and advise the incumbent CEO to prevent potential damage to the firm's long-term value in normal times and during times of crisis, such as the market crash caused by the COVID-19 pandemic.
As such, we conjecture that the presence of a CFCEO would be more likely to lead firms to maintain or increase social and environmental investment to maintain their reputation and prolonged value creation, despite the company suffering from a loss of income during the pandemic crisis.Therefore, we propose our hypothesis as below:

Sample and data collection
Our sample comprises 1,263 companies, excluding banks, financials, insurance, and investment holding firms, and covering 14,450 firm-year observations in total.We restrict the sample to the S&P 1,500 firms for three main reasons: (1) they cover around 90% of the market capitalisation of US stocks, (2) these firms have adopted relatively efficient corporate governance mechanisms and social and environmental investment strategies, and (3) sufficient data on S&P 1,500 firms can be retrieved from various databases.Firstly, BoardEx provides information about the CFCEO, board size and compositions, executive compensation, and several other demographic characteristics of directors on the board (e.g., their gender and if they are independent).Secondly, following recent CSR studies, we use the latest Thomson Reuter Refinitive ESG Rating database as a proxy for firms' CSR performance (see Albuquerque et al. 2020;Bae et al. 2021;Lins et al. 2017).The Refinitiv ESG Rating database utilises rigorous methods to provide a firm's overall ESG rating and a comprehensive evaluation of its social, environmental and governance dimensions.Thirdly, the accounting and financial statement variables are collected from the Compustat and Refinitiv DataStream databases.We collect all variables for firms in our sample between 2002 and the end of 2021, covering the global financial crisis (2007-9) and the recent COVID-19 pandemic crisis (2020-2021), which erupted in early 2020.

Variables and measures
Dependent variable.The environmental and social Rating, i.e., the Refinitiv ES Rating, has been used in several recent studies as a proxy for CSR performance (see Albuquerque et al. 2020;Bae et al. 2021;Lins et al. 2017).The Refinitiv ESG database provides a detailed evaluation of the environmental, social and governance aspects.Each aspect includes multiple assessments of, for example, resource use, emissions and innovation, workplace, human rights, community and product responsibility, management, shareholders, and CSR strategies.Following Albuquerque et al. (2020), Bae et al. (2021), andDyck et al. (2019), our main CSR proxy is based on a firm's 'E' (Environmental) and 'S' (Social) scores (i.e., ES_ Refinitiv).The 'E' dimension is associated with a firm's performance regarding resource use, emissions, and innovation, while the 'S' relates to a firm's performance concerning workplace, human rights, community, and product responsibility.We exclude 'G' (Governance) as this is not directly linked to a firm's CSR activities (Lins et al. 2017).The 'G', which is scored in respect of management, shareholders, and CSR strategies, will be used as a governance control variable in our empirical models1 .
Our sensitivity tests use alternative proxies for firms' CSR intensity of firms.We separately measure a firm's CSR performance by its environmental performance (Environment_ Refinitiv) and social commitments (Social_Refinitiv).
Independent variable.Chair-Former-CEO (CFCEO), our main independent variable, is a dummy variable, which has the value of one if the Chairman has previously served as the CEO of the same firm (or, the "Chair held CEO position in the firm prior to becoming Chair" -Veprauskaite and Adams 2013, p.233) and zero otherwise.The term Chair-Former-CEO is used in the study of Veprauskaite and Adams (2013), and its definition and measurement are in line with those provided in the DataStream.

Empirical model
We construct our empirical model as below: ), we control for several corporate governance characteristics and firm-level financial indicators that could potentially affect CEOs' decision-making about social and environmental engagement.The control variables include the "G" (Governance)2 mentioned above and specific measures for board independence, such as the board size and the proportions of independent and female directors on the board.Several studies have confirmed that improved board independence and diverse perspectives among board members can encourage executives to engage with and intensify social and environmental activities, thus fostering sound relations with a wider range of stakeholders, who confer social capital and trust on the firm (Amin et al. 2020).We also control for Chair-CEO duality and senior executive compensation, enabling the board to effectively monitor and advise executives to invest in social and environmental activities for the firm's long-term value creation (Amin et al. 2020).
Prior CSR research has shown that a firm's financial ability or slack resources are influential factors driving its CEO to enhance social and environmental performance.As such, we control related financial measures, including firms' total assets, leverage, return on assets, ratios of cash and cash dividends to total assets, and Tobin's Q (Amin et al. 2020).Table 1 presents more details about all the variables used in our study.

Descriptive statistics and correlation matrix
As per Table 2, the descriptive statistics show that 45.8% of the total firm-year observations are featured with the CFCEO, and the ES ratings for these companies have been distinctive over the past 18 years.The three versions of ES ratings representing the firms' CSR intensity, the mean (median; maximum) maximum values of ES rating (ES_Refinitive), S rating (S_Refinitive), and E rating (E_Refinitive) are as follows: 0.336 (0.273; 0.877), 0.413 (0.384; 0.027), and 0.260 (0.164; 0.901), respectively.This shows that, on average, firms engage more in social than environmental activities.
Regarding the control variables, the mean of board gender diversity (%Female) is 0.168, and the highest proportion of female directors on a board is 50%, while the lowest ratio is 0. Regarding the number of independent directors (%Ind), on average, the sample firms have 73.4% independent directors on the board, but the highest ratio is 93.3%, and the lowest is 0. Regarding the CEO duality (Chair-CEO Duality), 60.4% of the total firm-year observa- Board size, estimated by the natural logarithm of the size of the board of directors.

Amin et al. (2020) %Ind
The number of independent directors divided by total number of directors.

Li et al. (2023)
Chair-CEO Duality Dummy variable, which has a value of 1 if the chair and CEO are the same person, and 0 otherwise.

Amin et al. (2020) %Female
The number of female directors divided by total number of directors.

Li et al. (2023)
ExComp/TA The ratio of senior executive compensation to a firm's total assets.

Li et al. (2023) Leverage
The sum of long-term debts and debts in current liabilities, scaled by the book value of total assets.

Amin et al. (2020) ROA
Return on assets.The ratio of earnings to the book value of total assets.

Amin et al. (2020) Cash/TA
The ratio of cash to a firm's total assets.tions have their CEOs also serving as the chairman of the board of directors.On average, executive compensation (ExComp/TA) accounts for 0.7% of firms' total assets, and its values range from 0 to 13.9%.For conventional accounting and financial variables, the value of LnTA spans from 0 to 12.324.The ratios of Cash (Cash/TA) and Cash dividends (Div/TA) in relation to total assets range from 0 to 0.445 and 0.116, respectively.The return on assets (ROA) is 4.2% on average, whereas the minimum value is negative (i.e., -19.1%).Finally, the minimum and maximum leverage are 0 and 76.8%.
As discussed above, a high heterogeneity exists among our sample firms in terms of their CSR performance measured through alternative ES ratings over the past two decades, covering the crises of 2007-9 and 2020-2021.Our test results on the presence of multicollinearity in the sample are presented in Table 3, the correlation matrix.The vast majority of correlations are far below 0.8, which indicates that multicollinearity is not an issue among the variables used in the analysis.Low unreported VIF values (provided upon request) also validate this result.

Determinants of the chair-Former-CEO
Not all firms employ a former CEO as a chair; hence, we first explore how firms with a CFCEO compare with those without a CFCEO.We use logistics models to examine the  determinants of the CFCEO of a firm.Our dependent factor is CFCEO, and our explanatory variables include the firm's performance and other characteristics.We consider that the presence of the CFCEO can vary across firms and industries and over time by having the firm, industry and year-fixed effects.In unreported models excluding industry-fixed effects, the standard errors become smaller.Table 4 reports the results.We find that firms with superior accounting-based performance (reflected by ROA) and market-based performance (reflected by Tobin's Q), higher levels of dividend payouts, lower levels of financial leverage and agency issues (reflected by the better governance quality scores and lower levels of cash flow) are more likely to have a CFCEO.Not surprisingly, these outcomes may satisfy shareholders, so the appointment of a former CEO to serve as the Chair is more likely.

Chair-Former-CEO and firms' social and environmental performance
The results from the Refinitiv ES ratings (Table 5, models 1 to 9) show that the presence of a CFCEO (CFCEO) is positively and significantly (at 1% level) associated with firms' social and environmental performance (ES_Refinitiv).We also examine the moderating effects of external shocks (i.e., COVID-19 and the Global Financial Crisis) on the direct relationship between the CFCEO and social and environmental intensity.To achieve this, we added the interaction terms between the CFCEO and the Covid Crisis dummy variable (CFCEO*Covid Crisis) to models 4-6 and then continued adding the interaction terms between the CFCEO and the Financial Crisis dummy variable (CFCEO*Financial Crisis) to models 7-9.Our results show that the positive effect of the presence of the CFCEO on social and environmental ratings is more likely to intensify throughout the COVID-19 crisis than the global financial crisis.This is evidenced by the positive and significant coefficients on the CFCEO*Covid Crisis and negative and significant coefficients on CFCEO*Financial Crisis.
Although not the focus of our research, we still report the results of relevant control variables as their coefficients may provide us with some insights into firms' social and environmental performance.We generally find a positive relationship between the Covid Crisis dummy and a firm's social and environmental performance via the ES rating measure.However, the Financial Crisis variable coefficients are significant and negative across all models, suggesting that social and environmental investments are reduced throughout the global financial turmoil.We also confirm results reported in prior studies about the positive effects of board independence, Chair-CEO duality, financial leverage, dividend payouts, and Tobin's Q on a firm's social and environmental performance (score) in all or most of the model specifications (e.g., Amin et al. 2020;Dyck et al. 2019;Liang and Renneboog 2017).

Testing for firm governance quality
Table 6 reports the regression results for the association between the presence of a CFCEO and the ES ratings between two subsamples: Better-governed firms (i.e., the value of Governance is equal to or higher than its median) and Worse-governed firms (i.e., the value of Governance is lower than its median).We find consistent findings across the two sam-ples, implying that the positive effects of CFCEO on CSR performance and the moderating impact of the Covid-19 shock on such associations remain the same regardless of the firm's governance quality.However, the adverse and significant effects of the global financial crisis are more pronounced in the worse-governed firms.The result on better-governed peers becomes insignificant, indicating that firms with better governance quality can avoid the negative influences of the global crisis on the presence of CFCEO.

Testing for mergers and acquisitions (M&A)
Table 7 presents the regression results regarding the effects of M&A on the association between the presence of a CFCEO and the ES rating through the interaction terms and two subsamples: Firms with M&A (i.e., the value of M&A is equal to 1) and Firms without M&A (i.e., the value of M&A is equal to 0).The dependent variable is CSR performance estimated by ES_Refinitiv.The independent variables include the CFCEO, its interactions with M&A, and other controls.In model 1, when we employ the interaction term, we find that the positive impact of the CFCEO on ES performance is not affected by the M&A.Model 2 and model 3 (two subsamples) show no difference in the results.As such, we conclude that the effect of CFCEO holds even if the firms are experiencing M&A.period 2002-2004, the Global crisis period 2007-2009, and , and the Growing period 2017-2019.The classification of the growing and crisis periods is based on the US gross domestic product (GDP) growth3 .We find that the positive impact of CFCEO on ES performance is more pronounced during the recent growth period of 2017-2019, when the US GDP growth seems to have peaked in mid-2018, boosted by several prior events such as the 2017 Tax Cuts and Job Act, as well as the US federal budget agreement for the fiscal year 2018.Not surprisingly, although the growing periods 2002-2004 and 2017-2019 experienced a higher level of GDP growth, the awareness of CSR was only improved after the global crisis of 2007-2009.This is possibly why the positive effect of CFCEO on ES rating during 2002-2004 is insignificant.Even if the former CEO is appointed as the chair, they seem still not to be interested in CSR investments.

Chair-Former-CEO and CSR components
In this study, our main CSR performance proxy (ES_Refinitiv) is estimated based on two important components: Social and Environmental.Therefore, we next check which of the two components drives the above results.To do so, we alternatively use two proxies for CSR performance: social (Social_Refinitiv) and environmental scores (Environmental_Refinitiv). Our results are reported in Table 9 (Panel A and B), which confirms the positive association between the CFCEO and CSR performance and the more pronounced effects during COVID-19 than that of global financial turmoil.This is consistent across all models (1-6).As such, we can conclude that our main findings are not driven by a specific component (Social or Environment) but are affected by all of them.

Chair-Former-CEO and relative measure of CSR performance
Next, we follow the study of Cornett et al. (2016) to measure CSR performance by relative proxy.This is estimated by the difference between the CSR score and its min value, scaled by the difference between the max and min values of the CSR score.We repeat this method for three alternative CSR measures to create three relative proxies: Relative "ES" score (R_ES), Relative "E" score (R_Environment), and Relative "S" score (R_Social).The rationale for this approach is that CSR activities should also follow certain norms in all economic trends.Table 10 reports results for the effects of the CFCEO on CSR performance measured by the three new relative proxies.Across all models 1-6, all our main findings for single variables and interaction terms remained unchanged, further confirming the robustness of our results.

Endogeneity concerns and treatment
So far, using a pooled OLS model with alternative proxies for CSR performance, our results show a positive association between the CFCEO and CSR, and such effects are more pronounced during the crisis time of COVID-19 than the global financial crisis.However, it is argued that any governance-related research could suffer from endogeneity, which describes three cases: omitted variable bias, measurement errors and reverse causality issues (Hermalin and Weisbach 1988).The first two causes of endogeneity can be addressed by (1) a panel data analysis, (2) a comprehensive set of control variables and year/firm/industry fixed effects, and (3) accurate variable measurements.However, the third cause, i.e., reverse causality, can happen if firms with superior CSR performance prefer to employ a CFCEO to benefit from their knowledge, information, experience, network and other resources.We therefore use different techniques to address or minimise the presence of endogeneity.
Instrumental variable (IV) approach.We first employ the IV approach via the two-step system generalised method of moments (GMM).The GMM technique utilises internal IVs as lagged values of endogenous factors because those variables observed in previous years are unlikely to be affected by CSR scores in subsequent years.It also captures the unobserved influences by transforming all variables into first differences (see Arellano and Bover 1995;Blundell and Bond 1998), which can alleviate unobserved heterogeneity and omitted variable biases.Results from the GMM (those after capturing the unobserved heterogeneity, simultaneity and dynamic endogeneity) are reported in Table 11 (Panel A).They confirm the positive association between the presence of a CFCEO and CSR ratings, and such an impact is more intensified during the COVID-19 shock than during the global financial crisis.The diagnostic results of AR (1), AR (2) and Hansen/Sargan tests are satisfactory.
Two-stage OLS regression models.We use the one-year lagged and two-stage OLS regression models to address the endogeneity problem.Specifically, we retest our main model by using a two-step OLS method.In the first stage, we run an OLS regression of CSR performance (measured by ES Rating) on all independent variables (except the main independent variables: CFCEO, Covid Crisis, Financial Crisis and their interactions, including CFCEO*Covid Crisis and CFCEO*Financial Crisis).We then extract the residuals from this regression.In the second stage, we run OLS robust standard errors with the dependent variable using the extracted residuals (obtained in the first stage) as a function of the five variables above, which we had excluded earlier.We report the results of these two stages in Panel B of Table 11, which indicates that the main findings are relatively unchanged.
Propensity Score Matching (PSM) approach.We finally follow the research design of Rosenbaum and Rubin (1983) to test the research question using the propensity score matching (PSM) approach.This estimation is considered a treatment test for sample selection bias and possible endogeneity for the CFCEO variable.Such endogeneity usually arises in corporate governance research.As the CFCEO is a dummy factor denoted one, if the firm's current Chairman had previously served as the CEO, and zero otherwise, we were able to conduct a three-step PSM estimation as follows.
In the first step, using the probit approach, we estimate the propensity scores for firms with a CFCEO (treatment group) and those that do not have a CFCEO (control group).In 0.000*** 0.000*** 0.000*** 0.000*** 0.000*** 0.000*** The table presents and compares regression results for the association between the presence of the CFCEO and the CSR performance between two subsamples: Better governed firms (the value of Governance is equal to or higher than its median) and Worse governed firms (the value of Governance is lower than its median).The dependent variable is CSR performance estimated by ES_Refinitiv.The independent variables include the CFCEO, its interactions with Covid Crisis and Financial Crisis, and other controls.the second step, we match the propensity scores of the treatment and control groups in the first step.To do so, we use four various matching techniques: (i) 1-to-1 nearest neighbour matching with a replacement, (ii) 1-to-1 nearest neighbour matching without a replacement, (iii) the nearest neighbour matching with n = 2 with replacement, and (iv) the nearest neighbour matching with n = 3 with replacement.Using all these four techniques confirms the robustness of our results and the quality of sample matching (shown in the Appendix).

Definitions of the variables are reported in
In the third step, we conduct three things.First, we report the average treatment effects (ATE) with the four alternative nearest neighbour matching techniques in Panel A of Table 12.Results for both unmatched and matched samples suggest positive and significant differences between the treated and control groups.This implies that firms with the presence of a CFCEO exhibit a significantly higher CSR performance than their peers without such a Chairman.Second, in Panel B of Table 12, we report the average treatment effect on the treated (ATT) with 1-to-1 nearest neighbour matching and bootstrapping of standard errors (i.e., 100, 1000, 10,000 replications).The number of treated observations is 6170.The observed difference and T-statistics reveal a positive and significant result, implying the same findings.Finally, we run regression tests on the matched samples and report findings in Panel C of Table 12.Across all models 1-4 using four alternative matching techniques, we consistently find that the CFCEO is positively and significantly associated with the CSR performance measured by ES scores.Our PSM results provide an additional robustness check for our main findings.

Discussion and conclusion
Social and environmental investment has become an essential pillar of organisational strategies in enhancing corporate reputation and firm value (e.g., Chintrakarn et al. 2021;Boone et al. 2020).Prior research has found that CSR activity is driven by financial (e.g., profitability, capital structure) and non-financial factors, such as corporate governance.In this study, we focused on the latter.More specifically, given that prior literature has provided evidence on the board of directors (Amin et al. 2020;de Viller et al. 2011;Oh et al. 2019) and CEO leadership (Waldman et al. 2006;McCarthy et al. 2017;Hegde and Mishra 2019) when it comes to the intensity of CSR activities, we focused on leadership continuity and the potential impact this could have on CSR activities.Our objective was to examine if the presence of a former CEO becoming the Chair of the Board of Directors can encourage the incumbent CEO to remain committed to CSR.The presence of a CFCEO may encourage  table presents the effects of M&A on the association between the presence of the CFCEO and the CSR performance through interaction terms and two subsamples: Firms with M&A (the value of M&A is equal to 1) and Firms without M&A (the value of M&A is equal to 0).The dependent variable is CSR performance estimated by ES_Refinitiv.The independent variables include the CFCEO, its interactions with M&A, and other controls.Definitions of the variables are reported in Table 1 Table 7 (continued) the incumbent CEO to sustain long-term investment in CSR activities.The influence that such a presence can exert could be important in ensuring the continuity of a strategy and associated practices that took time to establish and create value for stakeholders.While the Chair/CEO succession is critical for strategic (re)alignment (Pfeffer and Salancik 1978), the monitoring effectiveness by the Chair in relation to the new CEO's behaviour can be indispensable.New CEOs are likely to take their institutions in new directions to the extent they have the discretion to do so (Hambrick and Finkelstein, 1987).Our study suggests that the continuing presence in a firm of the predecessor CEO as board chair could encourage stability, which can help an organisation's long-term sustainability goal.While considering a set   (2012, p. 853), who stated that "as long as the predecessor remains as chair, company per-  formance tends to adhere to pre-succession levels.New CEOs who are restricted in their actions are correspondingly restricted in the degree to which they can alter performance".A former CEO is more likely to be selected as the chairman of the board of directors if they have a strong record in running the company, a deeper understanding of the firm's culture and the industry dynamics, and a pursuit of long-term value for not only the firm's shareholders, but also other relevant stakeholders through committed social and environmental activities.With this assumption, we find that if a former CEO stays on the board as the Chair, they effectively monitor and advise the new CEO to maintain the firm's focus on long-term value creation via sustained social and environmental performance instead of short-term returns that newly appointed CEOs are likely to showcase.Our results on such CSR-relevant activity, therefore, extend the findings of Quigley and Hambrick (2012), who focus on the restrictive influences of predecessors retained as board chairs on (i) successor discretion, (ii) strategic change and (iii) overall financial performance.
Our findings support Hypothesis 1 that a CFCEO has a significant role in increasing social and environmental ratings.Based on upper echelons theory, we argue that CFCEOs may rely on prior work experience and knowledge about the firm and its industry to monitor, advise or persuade incumbent executives to remain committed to the firm's long-term value by continually investing in social and environmental activities.We cannot precisely esti-  1 mate the result's economic significance because social and environmental performance is measured by social and environmental scores that reflect point values rather than actual dollars spent.However, we can contextualise the economic effects by suggesting that the firms with the presence of a CFCEO exhibit higher social and environmental performance ratings by approximately 0.021 to 0.028 points than their peers without a CFCEO.Similarly, our empirical evidence also supports Hypothesis 2. Our findings suggest that there are differential effects of CFCEOs on social and environmental performance during times of crisis, such as the economic recession in 2007-9 and the COVID-19 pandemic crisis in 2020.Beyond the contextual differences that such crises may have, both cases were global and have had an immense impact on all aspects of our societies beyond firm financial performances.As such, understanding the role that CRS can play can offer valuable insights when it comes to preparing for the next crisis.Unfortunately, it is not a matter of "if" a crisis will take place, but "when".To this end, and to explain why we found differences between the two cases, one may refer to how, over the last decade, social and environmental performance has been valued more highly by market participants.The positive association between a CFCEO and the firm's social and environmental performance was found to be more pronounced during the COVID-19 pandemic crisis than during the economic recession back in 2007-9.Extending existing knowledge of governance-CSR associations (e.g., Jo and Harjoto 2011;Oh et al. 2019;Chintrakarn et al. 2021), our study suggests that the presence of a CFCEO is an important driver for corporate executives to commit to fostering continuity and in turn a resilient relationship with their stakeholders.Such resilience could be critical when the next crisis threatens the firm's survival.

Theoretical and practical implications
We make several tangible theoretical and practical contributions by addressing our research objectives.Our study offers new evidence of the role that leadership continuity, as manifested by the presence of a CFCEO, can play.It could have been possible that a former CEO holding the Chair title does so merely in a symbolic manner.Such an appointment may acknowledge their past contributions but have little future influence on setting future strategies.Our results point to the contrary, highlighting the importance that continuity can play, especially in an area driven by external engagement and relationships such as CSR.Our findings extend those of Quigley and Hambrick (2012) by showing that CFCEO influences the CEO's strategies related to social and environmental activity, not just financial matters.We also contribute to the stream of literature on CEO succession (e.g., Lewis et al. 2014;Hegde and Mishra 2019), which considers this to be an adaptive event, and that on managerial discretion (Hambrick 2007), which examines the general effects of managers and a wide range array of factors relating to organisational outcomes and discretion.Additionally, our findings add to the attempts of Quigley and Hambrick (2012) and other Chair-CEO duality research in reframing the debate about the merits of separating the board chair and CEO roles, but differently.Specifically, we consider this separation as one of the two possible CFCEO cases and relate it to sustainability performance to explore the longterm benefits of the disparity between the two top senior positions rather than short-term performance.By doing so, we are among the proponents of separation and contribute to the findings on the efficiency of such a separation.
As a result of the above, our study could also contribute to opening up promising new avenues for how key corporate leadership roles and their interplays can impact performance.Future studies may opt to distinguish among (i) a Chair who was promoted as soon as their CEO tenure ended (ii) a Chair who has experience of being the CEO of the firm; (iii) a Chair who was never the CEO of the firm and (iv) a Chair who currently also serves as the CEO (Duality).Such an explicit delineation may offer valuable insights into how power relationships operate in a firm and provide a richer backdrop to interpret future findings.Our empirical evidence could also stimulate future work examining the performance effects of the above four main Chair types concerning different financial and social indicators.Our paper also adds to prior studies on US firms, which show mixed findings on the duality but fail to acknowledge the CFCEO case (i.e., the separation).We suggest separating the Chair and CEO may benefit the institutional social and environmental activity, especially when the Chair had previously served as the CEO.
In addition to the above, in the supplementary analyses, our findings provide strong evidence when explaining how crises could affect the relationship between the CFCEO and social and environmental investment intensity.It is crucial to understand how the presence of a CFCEO can influence the firm's social and environmental performance during turbulent times.Therefore, our results contribute to the topical research efforts invested in understanding the impact of the COVID-19 pandemic by comparing the effects of such health crises to the context of the 2007-9 financial crisis, which are different in nature (e.g., causes, responses and consequences).
From a practical and managerial perspective, our research could inform future appointments for the Chairman of the Board and the impact that succession and internal promotion can have.Our findings suggest that there can be benefits in separating the Chair and CEO roles, noting that such a separation should be more beneficial to the firm's long-term sustainability performance when the Chair has prior experience being the former CEO of the firm.Also, our study's extended findings related to the external shocks offer additional empirical evidence on the beneficial impact of the presence of a CFCEO on sustaining external engagement via social and environmental activity in times of crisis, such as the recent pandemic.

Limitations and future research
Our study is subject to limitations that pave the way for future research avenues.Firstly, it was not possible to investigate the personal relationships between the CFCEO and the newly appointed CEO due to personal data limitations and availability.A qualitative analysis that captures such interactions in more detail may shed light on how power and influence operate between the two sides.Similarly, future studies could test the hypotheses put forward in different contexts, both during normal times and also during different local or even global crises.In addition, future work could examine different organizations' contexts (e.g., culture, norms, financial distress, financial constraints).Extending the empirical evidence on the above two fronts could contribute to the literature on corporate governance, finance, and strategic management.Finally, from a methodological point of view, future studies can advance our study by considering such contexts to minimise endogeneity issues.
8 Appendix: matching sample quality γCovid Crisis i,t + δGlobal Crisis i,t + ϑCF CEO i,t Covid Crisis i,t + ϑCF CEO i,t * Global Crisis i,t + Year FE + Industry FE + Firm FE + θ + ε (1) where CSR i,t represents the CSR score of firm i at year t, measured by ES Rating in the main tests and other alternative ones (e.g., Social_Refinitiv and Environment_Refinitiv). CF CEO i,t represents the presence of a Chairman who had previously served as the CEO in the same firm.CovidCrisis i,t and GlobalCrisis i,t represent the COVID-19 pandemic crisis and global financial crisis and are binary variables.The COVID-19 dummy variable takes a value of 1 if the evaluated year is 2020-2021 and 0 otherwise (Elnahass et al. 2021).The Global financial crisis dummy variable takes the value of 1 if the observed year is 2007-2009 and 0 otherwise (Hagendorff et al. 2018).In addition, CF CEO i,t * CovidCrisis i,t and CF CEO i,t * GlobalCrisis i,t are the interaction terms between the CFCEO and Covid Crisis and Financial Crisis, respectively.θ represents the control variables.Control variables.Following previous literature (e.g., Amin et al. 2020

Table 1
Variable Definitions Variable Definition References Panel A: ESG Rating ES_Refinitiv Refinitiv ES Rating, estimated by the average of the Environment and Social scores, but the Governance score is excluded.It is estimated by a firm's Environmental € performance in terms of resource use, emissions, and innovations; and Social (S) performance in terms of workplace, human rights, community, and product responsibility.

Firm-level control variables
GovernanceThe G element of the Refinitive ES Rating, estimated by the Social performance evaluated in terms of management, shareholders, and corporate social responsibility strategy.
This table reports measurements, definitions and references of all the variables which are employed throughout our research

Table 2
The table presents the descriptive statistics of all the main variables used in this research.All of them are winsorised at the 1% level.Definitions of the variables are reported in Table1 The table presents the Pearson correlation matrix among all the independent variables employed in this research.The * denotes a significance level of 1% Table8presents the OLS regression results on the association between the presence of a CFCEO and the CSR performance through three economic growth periods: the growing

Table 5
Chair-Former-CEO and firms' CSR Performance: The Effects of Crisis ShocksThe table presents OLS regression results for the association between the presence of the CFCEO and CSR performance, and the effects of crisis shocks including the Covid-19 pandemic and the global financial crisis 2007-9.The dependent variable is CSR performance estimated by ES_Refinitiv.The independent variables include the CFCEO, its interactions with Covid Crisis and Financial Crisis, and other controls.Definitions of the variables are reported in Table1

Table 6
Testing for Firm Governance Quality

Table 7
Testing for the Mergers and Acquisitions (M&A)

Table 9
Chair-Former-CEO and CSR Components of comprehensive control variables that can affect firms' social and environmental activity, we found that CFCEOs can encourage incumbent CEOs to remain committed to social and environmental activities.This is consistent with the findings of Quigley and Hambrick The table presents OLS regression results for the association between the presence of a CFCEO and the relative measures of CSR performance, and the effects of crisis shocks including the Covid-19 pandemic and the global financial crisis 2007-9.The dependent variables are CSR performance estimated by alternative proxies including R_ES, R_Environment, and R_Social.The independent variables include the CFCEO, its interactions with Covid Crisis and Financial Crisis, and other controls.Definitions of the variables are reported in Table1The table presents OLS regression results for the association between the presence of the CFCEO and the components of CSR performance, and the effects of crisis shocks including the Covid-19 pandemic and the global financial crisis 2007-9.The dependent variables are CSR performance estimated by two alternative proxies including Social_Refinitiv and Environmental_Refinitiv.The independent variables include the CFCEO, its interactions with Covid Crisis and Financial Crisis, and other controls.Definitions of the variables are reported in Table1

Table 12
Propensity Score Matching estimation: Chair-Former-CEO and CSR Performance Dependent variable: ES_Refinitiv Independent variables: CFCEO Panel A: Average treatment effects (ATE) with nearest neighbour matching techniques Average treatment effect on the treated (ATT) with 1:1 nearest neighbour matching and bootstrapping of standard errors No of treated obs.