The Impact of CEO Duality and Financial Performance on CSR Disclosure: Empirical Evidence from State-Owned Enterprises in China

Abstract

This paper studies the effects of a firm’s financial performance (FP) and chief executive officer’s (CEO) duality on the quality of corporate social responsibility (CSR) disclosure in the context of state-owned enterprises (SOEs) among Chinese A-share-registered companies. The results depict a negative relationship between CEO duality and CSR disclosure. Our results demonstrate that better-performing firms disclose CSR information more frequently and of higher quality compared with firms with poor financial performance. This role of financial performance in the quality of CSR disclosure is generally valuable in public enterprises; however, it is relatively sluggish in state-owned enterprises the outcomes indicate that the dual leadership structure reduces assessments and renders CEOs less liable to their stakeholders. Therefore, this study offers valuable information and details for regulators to improve corporate governance and CSR from the perspective of stakeholder theory.

1. Introduction

A noteworthy question in the arena of corporate governance is whether a person should hold both major leadership positions simultaneously, i.e., chairman of the board of directors (BOD) and CEO, or whether it would be more fitting for different people to hold each position. The reasoning revolves around the arguments for the separation of powers in contrast to that of the unity of direction in firms’ decisions. This disagreement has considerable implications, as executives play a key role in the management of corporations and their efficiency (Gul and Leung 2004; Ho and Wong 2001). Consequently, an organization’s disclosure structure and standards are shaped. The institutional context of China offers a valuable point of inquiry into the role of executives. On average, the financial performance of firms is favorably linked to CSR disclosure in Chinese listed firms (Li et al. 2013); however, some studies report the relationship as insignificant and a small number even reports a negative correlation between them. In terms of CEO duality, the research, by and large, is focused on transparency. Some studies have shown no significant relationship between CEO duality and CSR disclosure (Ahmad 2017; Cheng and Courtena 2006; Velte et al. 2017), while others have mentioned a negative and significant correlation between them (Muttakin and Subramaniam 2015). The studies with significant results reported lower CSR disclosures in relation to CEO duality compared with firms with no CEO duality problem. The mixed results of the previous studies could be mainly attributed to some missing connections with institutional factors. Furthermore, firm performance is an under-researched phenomenon for CSR disclosures in terms of ownership structure. Although some studies have attempted to explore the potential impact, they are mainly longitudinal and incorporate incomplete measures of CSR disclosure quality (Li et al. 2013).

In this paper, we explore the relationship between the financial performance of companies, CEO duality, and firms’ CSR practices from the perspective of the board’s institutional impact in terms of ownership structure. This study provides a broader understanding of the quality of CSR disclosure in two ways. Firstly, prior research indicates that the national institutional context, comprising regulatory frameworks, affects its CSR (Van der Laan Smith et al. 2005); however, less consideration has been paid to CSR disclosure quality in Chinese firms. Second, most Chinese companies are state-owned enterprises (SOEs). These SOEs have several social aims and profitability objectives, which are often in the competition with productivity goals. Consequently, when agreeing on the level of CSR disclosure, they could feature a separate priority mechanism relative to non-state-owned companies (NSOEs).

This study will add new knowledge to the body of corporate governance research in mainland China’s public corporations in terms of the duality–performance partnership. Although China has undergone rapid economic development in recent decades, Chinese business enterprises are heavily criticized for environmental emissions, human rights violations, and low-quality goods (Lin 2010; Wang and Juslin 2009). Thus, they have initiated multiple programs to support the growth of CSR since 2004. The Chinese Company Law of 2006, for example, bounds corporations not only to follow legislation and regulatory rules but also to act in line with societal and corporate ethical values. Furthermore, in 2008 the Chinese government implemented the adoption of CSR standards by firms that are to be regulated by the central government. Consequently, the number of Chinese business firms disclosing such information rose dramatically, from less than 471 in 2009 to almost 1015 in 2019 (Wu et al. 2021).

This paper also provides evidence of a new pattern of CSR disclosure quality in China. In the ownership context, our study develops the perspective of the new dimension of CSR disclosure in the relationship between business success and CSR. We contribute to stakeholder theory in terms of their control on management and influence on reporting. This study provides more insights into the degree to which a company’s CEO duality and financial performance determine (1) the decision of a company to publish its CSR operations, (2) the nature of its CSR records, and (3) the role in the ownership context. As SOEs have several social aims and profitability objectives, which are often in contrast with productivity goals, when agreeing on the level of CSR transparency, they could possess a separate prioritization mechanism relative to non-state-owned entities (NSOEs) (Li et al. 2013; Van der Laan Smith et al. 2005).

This paper is divided into six sections, including this introduction. The second section discusses the relevant research for formulating a theoretical framework and developing the hypotheses of the study. The third section highlights the research methodology employed in this study. The fourth section analyzes the sample data and highlights the analytical findings, providing an explanation and interpretation of the results. The last section provides the conclusions drawn from the findings and potential policy implications.

2. Theoretical Framework and Hypothesis Development

Through this study, we attempt to analyze the link between CEO duality and firms’ financial performance in China, along with their effect on CSR disclosure in the context of SOEs. The uncertainty over corporate transactions in Chinese firms makes it essential to resolve this issue and provide an incentive to devise a future strategy (Li and Peng 2008; Peng et al. 2009). Peng et al. (2010) and Zhou (2019) implied that cultivating a contingency viewpoint can be highly constructive when addressing the impact of CEO duality and FP on CSR disclosure. It is relevant to deepen awareness of the less discussed, collective effects on CSR disclosure, by the CEO duality, business performance and government ownership modes (Peng et al. 2010; Zhou 2019). The recent literature has concentrated on the difference in CSR disclosures across developing nations (Freedman and Stagliano 1992; Gray and Bebbington 2007; Rauf et al. 2020; Sial et al. 2018; Williams and Pei 1999).

In this regard, two different theories have been previously used in the literature, namely agency theory and stewardship theory. Agency theory is based on the conviction that there occurs a predictable clash between the principal (owner) and the individuals who implement (agents) (Jensen and Mecklin 1976). As supervisory activities often supersede shareholders’ returns, this provides ascent to agency issues relating to moral hazards and data irregularity. When the managers and executives focus on the maximization of their perceived efficiency, the exact information concerning different stakeholders is tempered (Fama and Jensen 1983; Hashim and Devi 2008). An agency problem is further probable to happen when the major decision maker has no ownership in shaping his/her choices. Agency theory admits that the governing board is an essential inside regulator component to regulate investors and the administration (Boyd 1995).

Pertaining to the above issues, when the offices of the chief executive and the chairperson of the board of directors are assumed by different people, the interests of stakeholders are potently addressed (Kholeif 2008). This non-duality allows the governing body’s director to efficiently observe and regulate the possible investor-value-destroying activities on part of the executives (Braun and Sharma 2007). In the light of these difficulties, agency theory endorses the segregation of the CEO and chairperson’s roles for the maximization of financial and non-financial performances.

In contrast, stewardship theory states that managers are good stewards of resources, and the merger of executive and chairperson roles in a single office reduces the opportunity of the inherent conflict of interest between owners and managerial bodies while improving the coordination and decision-making process. Promoters of this theory promulgate that CEO duality instills stronger leadership. Donaldson and Davi (1991) opined that combining the two functions would promote a firm’s efficiency through management unity and would lead to greater efficiency and quality of the CSR disclosure. Contrasting the arguments of the two perspectives under the empirical evidence presented insofar, CEO duality is instrumental in the accounting parameters while potentially decreasing the accountability and making the CEO less worried regarding their environmental and social obligations (Tschopp and Huefner 2015).

The tension between the two theories is addressed by the institutional theory which envisages the connection between the environments of an organization with its espoused goals while carving out its legitimacy (Parsons 1960). It explains the viability of agency theory and stewardship in different scenarios and is better able to explain the modes of association among CEO duality, firm’s financial performance, and CSR disclosure under the influence of SOEs.

2.1. CEO Duality and CSR Disclosure

CEO duality states that a similar individual holds the positions of the company’s CEO and chairperson simultaneously (Donaldson and Davi 1991). The impact of CEO duality on CSR performance is reported to be differentiated based on firms’ industrial affiliation and different institutional factors. In the USA, there is a general pattern of segregation between the duties of the CEO and the responsibilities of the chairperson, and empirical studies in American firms with CEO duality report poor business governance (Chen et al. 2018; Elsayed 2007; Thorne et al. 2017). However, mixed results are reported for trans-Asian studies. Said, by examining Malaysian firms, found no connection between CEO duality and the quality of CSR disclosure (Said et al. 2009). Likewise, Cheng and Courtena (2006) presented that CEO duality was not related to the degree of voluntary disclosures in the firms listed on the Singapore Stock Exchange (Cheng and Courtena 2006).

Gul and Leung (2004) confirmed the association of CEO duality with lower business voluntary disclosures in the institutional setting of Hong Kong. Their study shows that CEO duality reduces firms’ likelihood to connect with CSR widely. In the Chinese context, CEO duality is reported to have a negative impact on voluntary disclosure stages concerning listed Chinese firms (Huafang and Jianguo 2007). The research studies in the institutional settings of China, however, reveal a pattern of a negative association between sustainability-related variables and CEO duality. The low success rate of CSR procedures (Lu and Castka 2009; Michelon and Parbonetti 2012; Said et al. 2009) is attributed to CEO duality and has been stated as a double-edged blade metaphorically (Finkelstein and D’aveni 1994). Based on the above discussion, we developed the following hypothesis.

Hypothesis 1 (H1).

The duality of the CEO has a negative influence on the degree of CSR disclosure.

2.2. Firm’s Financial Performance and CSR Disclosure

The principal investor of a company is its shareholders, who are worried about the firm’s profits. From a justice viewpoint, a firm with better financial performance is more able to exhibit its social and public responsibilities and to avoid litigations; in the case of institutional regulations, they provide a more extensive report of CSR activities (Cho and Patten 2007; Garcia et al. 2018; Ng and Koh 1994). This claim recommends financial performance to be positively associated with CSR disclosure. Although some studies are in contrast to the above contention (Alnajjar 2000; Haniffa and Cooke 2005; Sial et al. 2018), on average there is a positive association between the financial performance of a firm and its CSR disclosure (Wu et al. 2021).

Gray (2001) advise that changes in social and organizational issues in various countries may clarify the varied outcomes of the correlation between the quality of CSR disclosure and profit from the business (Gray 2001; Khan et al. 2013). Van der Laan Smith et al. (2005) discovered proof that social and institutional components influence the quality of CSR disclosure. China has exclusive social and institutional features that influence organizational procedures regarding the triage among stakeholders’ interests, and the inducement comes out to be in the favor of shareholders (Van der Laan Smith et al. 2005). Lin (2010) also contends that even though China’s government has presented numerous initiatives to endorse CSR, Chinese firms’ are not burdened by the stakeholders to establish CSR (Lin 2010; Nelling and Webb 2009). First, while the state has launched some CSR initiatives, the constraining factors have not been addressed. Second, unlike their Western counterparts, Chinese customers lag in forcing firms to produce goods and facilities responsibly. The deficiency of customer consciousness is complemented by the lack of customer rights fortification. Third, the firms performing better in CSR are not incentivized by the government.

Fourth, while advanced nations have perceived noteworthy development in the investments considering CSR, shareholders in China are usually temporarily accommodated, with not much concern for CSR. This negatively influences decision-making discernments of the prominence of CSR and the firms opt for it only after the basic ends are met. An opinion poll of 83 Chinese companies by Zu and Song (2009) discovered that CSR activities are circumvented by the financial power of companies in China (Jamali et al. 2008; Zu and Song 2009). Due to the supervisory costs and investors’ effect in China, firms will potentially deliberate profitability while determining the level and quality of CSR disclosure. Thus, the incentive and motivation to disseminate CSR disclosure to the investors in China stems from the better performance of the firms. Given the voluntary CSR disclosure environment, the choice of whether to issue CSR disclosure or not is at the disposal of the firm, which goes for it only after addressing its financial concerns. Based on the above arguments we synthesize our second hypothesis as follows.

Hypothesis 2 (H2).

Firms with better financial performance exhibit a better quality of CSR disclosure than firms with poor financial performance.

2.3. Moderating Role of State-Owned Enterprises

We also investigate whether SOEs influence the connection between financial performance and the quality of CSR disclosure. In China, almost 60 percent of the registered firms are SOEs, providing a unique institutional atmosphere to the country. Van der Laan Smith et al. (2005) recommend that changes in shareholding design across states can influence investor–firm connections and impact the degree of CSR disclosure (Li and Zhang 2010; Van der Laan Smith et al. 2005). We theorize that state-holdings debilitate the connection between business performance and CSR disclosure for the subsequent three explanations. First, when a majority of shares are owned by the state, there arises a contrast in aims between the government and private investors. For instance, it potentially augments societal development at the cost of investors’ equity, as SOEs’ objectives contain financial and societal goals simultaneously and executives might end in a non-optimal triage among the two.

Second, compared with non-SOEs, SOEs receive economic and political help from the government (Bonardi et al. 2005; Brandt and Li 2003). Aharony et al. (2000) report that the regulators treat SOEs differently on political grounds, leveraging more financial benefits to them as compared with the non-SOEs (Aharony et al. 2000). Moreover, when in distress, SOEs are more likely to be rescued (Wang et al. 2008). Due to the administrative support, CEOs of SOEs are inclined to their authorized position and are thus more likely to reveal their CSR activities while being inattentive to profit maximization. Third, the state regularly surveys SOEs’ supervisors, and their business financial performance may be influenced by their political body. The Assets Supervision and Administration Commission (SASAC) has devised strategy rules for the assessment of directors, containing non-monitory methods in the shape of ecological and protection measures. In short, SOEs and non-SOEs prioritize different inducements to achieve societal accountability programs that advantage the public. We formulate our third hypothesis as follows.

Hypothesis 3 (H3).

SOEs moderate the relationship between CEO duality and CSR disclosure.

Hypothesis 4 (H4).

SOEs moderate the relationship between CSR disclosure and financial performance.

3. Data, Measurement, and Research Methodology

3.1. Sample

For this study, the employed data were collected from A-grade-listed firms of Shanghai and Shenzhen Stock Exchanges, China. Our database includes all CSR reports issued by approximately 1600 publicly listed Chinese firms between 2014 and 2019 from the business firm’s records and websites, with 3289 observations. We compiled the information about the CSR data subset from Chinese stock exchanges and the firms’ annual reports, along with other relevant data subsets from China Stock Market and Accounting Research (CSMAR) database. CSMAR is a wide-ranging research database, having a focus on the Chinese economy.

3.2. Dependent and Independent Variable

3.2.1. Corporate Social Responsibility (CSR) Disclosure

In this study, CSR disclosure is the dependent variable. The information has been gathered from the CSMAR data set to quantify corporate social responsibility disclosure. We built up the quality of the CSR disclosure from the items reported in the firms’ disclosures by utilizing a dichotomous strategy, assigning 1 if an item was reported and 0 if it was not (Rauf et al. 2020, 2021b; Sial et al. 2018; Tschopp and Huefner 2015). The complete-scale of CSR disclosure is shown in

Table 1. Index list of quality of CSR disclosure.

S. No	Corporate Social Responsibility Disclosure Items	Scale
(1)	Compliance with GRI sustainability disclosure guideline or not	1.0
(2)	Including shortages of firms or not	1.0
(3)	Including protection of shareholder interests or not	1.0
(4)	Including secure production or not	1.0
(5)	Including protection of creditor interests or not	1.0
(6)	Including social responsibility system construction and improvement or not	1.0
(7)	Including protection of employee interests or not	1.0
(8)	Including social and public welfare and public relations or not
(9)	Including protection of supplier interests or not	1.0
(10)	Including protection of consumers and interests of clients or not	1.0
(11)	Including sustainability and environment or not	1.0

below.

The quality of CSR disclosure is calculated through the equation: ${{\displaystyle \sum }}_n^1\frac{x}{n}\times 100$. If an item is reported, it carries a value of 1, and 0 otherwise. N represents the total number of items.

3.2.2. Chief Executive Officer (CEO) Duality

In this research, we utilized chief executive officer (CEO) duality as our essential driving independent variable to investigate the CEO duality influence on CSR disclosure, also stated to be the business organization’s leadership pattern. CEO duality was calculated by utilizing dummy values. If the CEO was also assumed as the chairperson of the board of directors, 1 was reported; otherwise, 0 was reported (Alabdullah et al. 2019; Berg and Smith 1978; Giannarakis 2014).

3.2.3. Financial Performance (FP)

In this study, financial performance is an independent variable. From the recent literature, it is evident that market-based measures, such as Tobin’s Q, are used extensively (Kang et al. 2007). In the case of Chinese companies, it may not be an appropriate measure for financial performance (FP). The accounting-based measures are more likely to be consistent and stable, thus return on assets (ROA) is utilized in this study (Cosh and Hughes 1995). Some studies have also used the accounting measure of return on equity (ROE) for measuring firm performance as an additional indicator or robustness check, so we also used it in the robustness check.

3.2.4. Moderating Role of State-Owned Enterprises (SOEs)

We considered the state-owned enterprises (SOEs) as a dummy variable for corporate ownership (Batten and Vo 2009; Chen and Keefe 2020; Vo 2020). In this case, if the shareholding is solely controlled by the government, then the SOEs variable is equal to 1; otherwise, 0. Although these SOEs are classified entities, the government’s ownership is not always 100%.

3.2.5. Estimating Model

We applied the ordinary least squares (OLS) as our regression tool. Equation (1) estimates the impact of CEO duality along with the controls on CSR disclosure. The impact of financial performance on the quality of CSR disclosure is assessed through Equation (2). We additionally tested the moderating role of SOEs in CSR disclosure through Equation (3).

$$CSRDisclosur{e}_{\left(i,t\right)}=a+{\beta }_{1}CEOduality+{\displaystyle \sum }_{i=1}^N{\beta }_{n}control{s}_{\left(i,t\right)}+{\mathsf{\epsilon }}_{\left(i,t\right)}$$

(1)

$$CSRDisclosur{e}_{\left(i,t\right)}=a+{\beta }_{2}FP+{\displaystyle \sum }_{i=1}^N{\beta }_{n}control{s}_{\left(i,t\right)}+{\mathsf{\epsilon }}_{\left(i,t\right)}$$

(2)

$$CSRDisclosur{e}_{\left(i,t\right)}=a+{\beta }_{3}CEOduality \left(FP\right)+{\beta }_{4}SOEs+{\beta }_{5}CEO duality \left(FP\right)\ast SOEs+{\displaystyle \sum }_{i=1}^N{\beta }_{n}control{s}_{\left(i,t\right)}+{\mathsf{\epsilon }}_{\left(i,t\right)}$$

(3)

In the above model, the quality of CSR disclosure shows CSR information, which is used to quantify CSR disclosure, CEO duality, financial performance (FP), and state-owned enterprises (SOEs). Control variables include firm age (FA), board size (BS), independent director (ID), the book to market ratio (BTMA), board meeting (BM), financial leverage (FL), firm growth (FG), and allocation to different factors, considered in this investigation to assess their impact.

3.2.6. Control Variable

This study also defines several business firm features as control variables to check their quality of CSR disclosure effect, including firm size (FS) (Rauf et al. 2020; Ettredge et al. 2002), firm age (Cumming et al. 2016; Rauf et al. 2021a), board size (Chong et al. 2018), independent directors (Yu et al. 2018), book to mark ratio (BTMA) (Rauf et al. 2020; Sial et al. 2018), board review (Chen and Keefe 2020), financial leverage (Dalci 2018; Naveed et al. 2021) and firm growth (Lei et al. 2019). For more details, see

Table 2. Variable descriptions.

Variables	Abbreviation	Description
Corporate social responsibility disclosure	CSR disclosure	CSR disclosure index formed by corporate social responsibility disclosure’s compliance with the GRI standards.
Chief executive officer duality	CEO duality	A dummy variable concerning chief executive officer duality. Its value is equivalent to 1 if the CEO is functioning as an executive and 0 otherwise.
Financial performance (ROA)	FP (ROA)	Financial performance measured via ROA (returns on assets), the ratio of total profit and percentage of assets.
Financial performance (ROE)	FP (ROE)	Financial performance measured via ROE (returns on equity), the ratio of total profit and percentage of equity.
State-owned enterprises	SOEs	A dummy variable that is equal to 1 if the firm is an affiliate with the central government, and is otherwise 0.
Firm size	FS	Natural log of total revenue
Firm age	FA	The number of years since the firm has been listed.
Board size	BS	Total number of directors on board
Independent director	ID	The number of independent directors on board
Book to market ratio	BTMA	Variable concerning the book value over the market value of the shareholder’s capital.
Firm growth	FG	The measure of the increase or decrease in total assets.
Board review	BR	Number of meetings in a year.
Financial leverage	FL	It is the proportion of debt of an association to its assets.
Year and industry	YI	The business dummies are remembered for all regressions to control for the impacts of year and explicit firm.

.

4. Empirical Outcomes

4.1. Descriptive Statistic

Descriptive statistics are reported in

Table 3. Descriptive statistics and correlation matrix.

Variables	Mean	SD	Min	Max	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)	(12)	(13)	(14)	VIF	1/VIF
CSR disclosure	5.835	2.467	1.000	8.000	1.000
CEO duality	0.033	0.456	0.000	1.000	−0.07 *	1.000													1.05	0.954960
FP (ROA)	8.707	12.032	0.000	37.378	0.01 *	0.02 *	1.000												1.40	0.716485
FP (ROE)	1.283	0.2344	0.000	27.642	0.07 *	0.05 *	0.07 *	1.000											1.89	0.529608
SOEs	18.554	22.541	0.000	89.09	0.18 *	0.08 *	0.04 *	0.02 *	1.000										1.09	0.914932
FS	23.217	1.743	18.65	30.814	−0.02 *	−0.11 *	0.58 *	0.00	−0.03	1.000									1.12	0.895010
FA	2.646	0.445	0.000	3.891	0.08 *	−0.02 *	−0.01 *	0.05 *	0.02 *	0.01 *	1.000								1.17	0.856601
BS	9.442	2.266	4.000	22.000	−0.02 *	−0.13 *	0.21 *	−0.00	0.23 *	0.03 *	0.02 *	1.000							1.80	0.554060
ID	3.485	0.830	1.000	8.000	−0.02 *	−0.08 *	0.29 *	−0.00	0.30 *	−0.03 *	0.77 *	−0.27 *	1.000						1.16	0.832014
BTMA	1.236	1.191	0.033	10.328	−0.02 *	−0.12 *	0.41 *	−0.02 *	0.58 *	0.02 *	0.09 *	0.15 *	0.13	1.000					1.42	0.716543
FG	0.155	0.316	−0.828	10.888	−0.05 *	0.09 *	−0.04 *	−0.14 *	−0.00	−0.04 *	−0.02 *	−0.01 *	−0.04 *	−0.01 *	1.000				1.08	0.529783
BR	10.354	4.873	1.000	57.00	−0.06 *	0.00	0.17 *	−0.13 *	0.19 *	0.07 *	0.00	0.04 *	0.15 *	0.11 *	0.07 *	1.000			1.87	0.518307
FL	0.509	0.214	0.007	1.344	−0.01 *	−0.12 *	0.29 *	−0.04 *	0.51 *	0.11 *	0.11 *	0.14 *	0.57 *	−0.02 *	0.21 *	0.11 *	1.000		1.39	0.623484
FG	13.928	9.338	0.000	28.624	−0.08 *	0.01	0.17 *	−0.10 *	0.25 *	−0.07 *	0.06 *	0.08 *	0.11 *	0.34 *	0.10 *	0.08 *	0.10 *	1.000	1.83	0.574064

* Significance at the 10% level.

. The outcomes indicate that the CSR disclosure mean is 5.835, with a standard deviation (SD) of 2.467. This demonstrates that on average, Chinese firms accomplish a score of 5.835 on the composite CSR information with the highest rating of 8.00. The average firm board size is 9.442, among which 3.485 have independent directors. According to the China Securities Regulatory Commission (CSRC), the board of directors must comprise 33% independent directors, and the results coincide with the quota.

4.2. Correlation Matrix

Table 3 shows the correlation matrix among variables. The relationship between the quality of CSR disclosure and CEO duality is 0.033 and is adversely negatively significant. The connection between firms’ financial performance, SOEs, and CSR disclosure is also significant and positive. The outcomes are in accordance with the earlier studies and the mean-variance inflation factors (VIFs) is 1.89. The detailed results are illustrated in Table 3.

4.3. Regression Results

OLS regression of the impact of CEO duality and financial performance on the quality of CSR disclosure and the moderating role of SOEs and control variable is shown in

Table 4. Results of impact of CEO duality and financial performance on CSR disclosure, and moderating role of SOEs.

Variables	Model 1	Model 2	Model 3	Model 4
CEO duality	−0.417 *** (−3.39)	-	−0.604 *** (−4.04)	-
FP (ROA)	-	0.009 *** (4.00)	-	0.084 ** (2.04)
SOEs	-	-	0.069 *** (19.75)	0.067 *** (19.65)
CEO duality * SOEs	-	-	−0.200 * (−2.19)	-
FP * SOEs	-	-	-	0.050 *** (4.89)
FS	−0.236 * (−2.85)	−0.214 * (−2.41)	−0.206 * (−2.83)	−0.289 * (−2.25)
FA	0.416 *** (3.57)	0.436 *** (3.73)	0.426 *** (3.66)	0.314 *** (2.88)
BS	−0.070 * (−2.01)	−0.065 * (−1.98)	−0.067 * (−1.93)	−0.057 * (−1.77)
ID	0.678 *** (4.143)	0.389 *** (3.561)	0.787 *** (2.953)	0.788 *** (2.955)
BTMA	0.019 (0.36)	0.054 (0.97)	0.045 (0.80)	0.050 (0.96)
BR	−0.036 *** (−3.94)	−0.039 *** (−4.17)	−0.038 *** (−4.12)	−0.032 *** (−4.43)
FL	−0.067 (−0.23)	0.194 (0.64)	0.143 (0.47)	0.187 (0.66)
FG	−0.017 *** (−2.81)	−0.015 * (−2.51)	−0.015 * (−2.50)	−0.008 (−1.45)
Constant	5.736 *** (6.04)	5.054 *** (5.24)	5.328 *** (5.51)	5.200 *** (5.76)
YI	Yes	Yes	Yes	Yes
R2	0.0373	0.0355	0.0387	0.1598

Dependent variable: CSR disclosure, *, **, ***, significant at 10%, 5%, and 1%, respectively. T-statistics are reported in parentheses. For a detailed description of the variable, see Table 2.

. Model 1 confirms the association between CEO duality and the quality of CSR disclosure. Through our results, we find that there is a significant and negative coefficient (t = −3.39, p < 0.001) of CEO duality, which suggests that CEO duality influences CSR disclosure accordingly, supporting hypothesis H1. This outcome is consistent with Michelon and Parbonetti (2012) and Said et al. (2009). In Model 2, we check the connection between financial performance and the quality of CSR disclosure. We find a positive and significant coefficient (t = 4.00, p < 0.000) of financial performance which suggests that financial performance affects the quality of CSR disclosure, in this way supporting hypothesis (H2). This is in accordance with previous studies (Garcia et al. 2018; Nelling and Webb 2009; Salehi et al. 2018; Sial et al. 2018; Zu and Song 2009).

By using Model 3, the moderation has been checked as being proposed in hypothesis H3. The results (see Table 4) suggest that CEO duality played a negative interaction role among SOEs in the quality of CSR disclosure (t = −2.19, p < 0.029). By applying model 4, we checked hypothesis H4, which indicates that SOEs have a central (positive) interaction role between FP and the quality of CSR disclosure (t = 4.89, p < 0.000) (Córdoba-Pachón et al. 2014). Control variables mostly had an insignificant correlation with the quality of CSR disclosure, as shown in Table 4.

Similarly, the coefficient of firm size, firm age, board size, and number of board meetings are negative and are statistically significant. This demonstrates the common observation of businesses with higher growth potential in terms of increased consideration for social responsibility (Asmar et al. 2018; Rauf et al. 2020; Yu et al. 2018). The values of the control variables are also in accordance with the previous studies.

4.3.1. Robustness Tests with Fixed and Random Effects

To check the robustness, we applied the fixed and random effects to the regressions. In contrast to OLS, where the intercepts are assumed to be the same, the time-based and individual-based entities are calculated with distinguished intercepts. As per the Levin et al. (2002) test, the panel data for the current study are stationary (Levin et al. 2002). The results of the Hausman test for the random effect models are reported at the ends of

Table 5. Results of impact of CEO duality and financial performance (ROE) on CSR disclosure, and moderating role of SOEs (panel data analysis with fixed effects and random effects robustness test).

	Fixed Effects				Random Effects
Variables	Model 1	Model 2	Model 3	Model 4	Model 5	Model 6	Model 7	Model 8
CSR disclosure	CEO duality	FP (ROE)	CEO duality * SOEs	FP(ROE) * SOEs	CEO duality	FP (ROE)	CEO duality * SOEs	FP(ROE) * SOEs
CEO duality	−5.199 *** (−5.76)	-	−6.792 *** (−6.14)	-	−0.034 *** (−8.07)	-	−0.024 *** (−3.85)	-
FP (ROE)	-	0.455 *** (6.38)	-	1.437 *** (2.49)	-	1.112 *** (2.79)	-	1.013 *** (2.90)
SOEs	-	-	3.513 *** (5.36)	5.591 *** (3.60)	-	-	0.218 *** (3.12)	0.063 *** (19.60)
CEO duality * SOEs	-	-	−0.014 *** (−3.28)	-	-	-	−0.058 *** (−5.47)	-
FP(ROE) * SOEs	-	-	-	1.495 *** (2.58)	-	-	-	0.294 *** (3.87)
FS	−0.016 (−0.38)	−0.199 * (−2.20)	−0.193 * (−2.14)	−0.193 *** (−2.14)	−0.019 (−0.43)	−0.321 *** (−3.23)	−0.301 *** (−3.04)	−0.128 (−1.51)
FA	0.422 *** (3.48)	0.443 *** (3.64)	0.436 *** (3.59)	0.436 *** (3.59)	0.419 *** (3.46)	0.452 *** (3.73)	0.441 *** (3.63)	0.330 *** (2.90)
BS	−0.061 * (−1.74)	−0.046 (−1.32)	−0.059 * (−1.67)	−0.059 * (−1.67)	−0.068 (−1.92)	−0.045 (−1.28)	−0.064 * (−1.80)	−0.043 (−1.30)
ID	0.784 *** (2.951)	0.476 *** (3.871)	0.453 *** (3.631)	0.426 *** (3.843)	0.471 *** (3.251)	0.514 *** (3.572)	0.623 *** (3.686)	0.627 *** (3.684)
BTMA	−0.000 (−0.01)	0.034 (0.62)	0.026 (0.47)	0.026 (0.47)	−0.004 (−0.07)	0.053 (0.94)	0.038 (0.68)	0.042 (0.80)
BM	−0.035 *** (−3.79)	−0.037 *** (−4.02)	−0.037 *** (−3.97)	−0.037 *** (−3.97)	−0.035 *** (−3.77)	−0.039 *** (−4.16)	−0.038 *** (−4.07)	−0.012 (−1.42)
FL	0.027 (0.09)	0.291 *** (0.94)	0.243 (0.78)	0.243 (0.78)	0.008 (0.03)	0.440 (1.40)	0.356 (1.13)	0.331 (1.13)
FG	−0.017 ** (−2.76)	−0.015 * (−2.45)	−0.015 (−2.43)	−0.015 (−2.43)	−0.017 ** (−2.75)	−0.014 * (−2.24)	−0.013 * (−2.22)	−0.008 (−1.51)
Constant	5.960 *** (6.16)	5.277 *** (5.37)	5.533 *** (5.61)	5.533 *** (16.70)	6.085 *** (6.28)	4.986 *** (5.06)	5.392 *** (5.42)	5.392 *** (5.79)
YI	Included	Included	Included	Included	Included	Included	Included	Included
R2 or pseudo R2	20.64%	20.97%	17.86%	20.73%	0.0305	0.0292	0.0315	0.1512
Hausman test (Chi2)	-	-	-	-	10.79 *** (0.0045)	10.87 *** (0.0056)	10.93 *** (0.0059)	10.98 *** (0.0062)

Dependent variables: CSR disclosure, *, **, ***, significant at 10%, 5%, and 1%, respectively. T-statistics are reported in parentheses. For a detailed description of the variable, see Table 2.

, which indicate that the fixed effect is needed to ensure more robust results as compared with the random effect model. Similarly, to ensure further robustness, ROE has been used as an alternative measure of the financial performance of firms (Rothenberg et al. 2017). The results of fixed effects and random effects models 1–8 are reported in Table 5. In models 1 and 5, according to the results, CEO duality was found to have a significant negative effect on CSR disclosure (t= −5.76; p < 0.000); (t= −8.07; p < 0.001); thus, H1 is accepted. This finding is compatible with prior studies (Alabdullah et al. 2019; Fahad and Rahman 2020). In models 2 and 6, similarly, FP had a positive significant effect on the dependent variable (t = 6.38; p < 0.001); (t = 2.79; p < 0.000), implying that H2 is accepted. This result confirms the studies by Pham and Tran (2020) and Thuy et al. (2021). In addition, by using models 3 and 7, the same fixed effects and random effects revealed that there was a significant negative effect between the role of CEO duality interaction among SOEs and CSR disclosure (−3.28; p < 0.000); (−5.47; p < 0.002); hence, H3 is accepted, in line with (Chen and Keefe 2020). Finally, in models 4 and 8, SOEs have a significant (positive) interaction role between FP and the quality of CSR disclosure (t = 2.58, p < 0.000); (t = 3.87, p < 0.000); thus, H4 is accepted. This result confirms several studies in the prior literature (Jain et al. 2017; Wang et al. 2016). Table 5 depicts that the results of the random effect and fixed effect models have no drastic difference as compared to the OLS regression, therefore confirming the robustness of the study.

4.3.2. Endogeneity Control

To manage the endogeneity issue, we used two distinct models. Initially, we used one-year lagged CEO duality, financial performance and the quality of CSR disclosure, and the moderating role of SOEs in

Table 6. Results of impact of CEO duality and financial performance (ROA) on CSR disclosure, and moderating role of SOEs (robustness two-stage least square (2SLS)).

Lagged Measure					2SLS
Variables	Model 1	Model 2	Model 3	Model 4	Model 5	Model 6	Model 7	Model 8
CEO duality	−0.394 *** (−3.15)	-	−0.588 *** (−3.86)	-	−0.601 *** (−3.13)	-	−0.900 *** (−4.26)	-
FP (ROA)	-	1.305 *** (5.65)	-	2.588 *** (3.86)	-	1.363 *** (5.67)	-	2.784 *** (3.93)
SOEs	-	-	0.068 *** (19.73)	0.064 *** (19.62)	-	-	0.067 *** (19.72)	0.063 *** (19.60)
CEO duality * SOEs	-	-	−0.207 *** (−2.23)	-	-	-	−0.332 ** (−3.16)	-
FP(ROA) * SOEs	-	-	-	0.381 *** (3.05)	-	-	-	0.085 *** (5.48)
FS	−0.016 (−0.38)	−0.199 * (−2.20)	−0.193 * (−2.14)	−0.193 *** (−2.14)	−0.019 (−0.43)	−0.321 *** (−3.23)	−0.301 *** (−3.04)	−0.128 (−1.51)
FA	0.422 *** (3.48)	0.443 *** (3.64)	0.436 *** (3.59)	0.436 *** (3.59)	0.419 *** (3.46)	0.452 *** (3.73)	0.441 *** (3.63)	0.330 *** (2.90)
BS	−0.061 * (−1.74)	−0.046 (−1.32)	−0.059 * (−1.67)	−0.059 * (−1.67)	−0.068 (−1.92)	−0.045 (−1.28)	−0.064 * (−1.80)	−0.043 (−1.30)
ID	0.784 *** (2.951)	0.476 *** (3.871)	0.453 *** (3.631)	0.426 *** (3.843)	0.471 *** (3.251)	0.514 *** (3.572)	0.623 *** (3.686)	0.627 *** (3.684)
BTMA	−0.000 (−0.01)	0.034 (0.62)	0.026 (0.47)	0.026 (0.47)	−0.004 (−0.07)	0.053 (0.94)	0.038 (0.68)	0.042 (0.80)
BM	−0.035 *** (−3.79)	−0.037 *** (−4.02)	−0.037 *** (−3.97)	−0.037 *** (−3.97)	−0.035 *** (−3.77)	−0.039 *** (−4.16)	−0.038 *** (−4.07)	−0.012 (−1.42)
FL	0.027 (0.09)	0.291 *** (0.94)	0.243 (0.78)	0.243 (0.78)	0.008 (0.03)	0.440 (1.40)	0.356 (1.13)	0.331 (1.13)
FG	−0.017 ** (−2.76)	−0.015 * (−2.45)	−0.015 (−2.43)	−0.015 (−2.43)	−0.017 ** (−2.75)	−0.014 * (−2.24)	−0.013 * (−2.22)	−0.008 (−1.51)
Constant	5.960 *** (6.16)	5.277 *** (5.37)	5.533 *** (5.61)	5.533 *** (16.70)	6.085 *** (6.28)	4.986 *** (5.06)	5.392 *** (5.42)	5.392 *** (5.79)
YI	Included	Included	Included	Included	Included	Included	Included	Included
R2	0.0313	0.0300	0.0328	0.0276	0.0305	0.0292	0.0315	0.1512

Dependent variables: CSR disclosure, *, **, ***, significant at 10%, 5%, and 1%, respectively. T-statistics are reported in parentheses. For a detailed description of the variable, see Table 2.

because the firm’s efficiency required time before they affected CSR disclosure results. Sets 1, 2, 3, and 4 of Table 6 show the lagged measures’ outcomes.

The results of one-year lagged and 2SLS models 1–8 are reported in Table 6. The coefficient of CEO duality remains negatively significant in models 1 and 5 (t = −3.15 p < 0.001); (t = −3.13 p < 0.000). Thus, H1 is accepted. This finding is compatible with prior studies (Alabdullah et al. 2019; Fahad and Rahman 2020). The coefficient of FP remains positively significant in models 2 and 6 (t = 5.65 p < 0.000); (t = 5.67 p < 0.000), implying that H2 is accepted. This result confirms those of the studies by Pham and Tran (2020) and Thuy et al. (2021). The coefficient of interaction term (CEO duality × SOEs) remains negative and significant in models 3 and 7 (t = −2.23 p < 0.000); (t = −3.16 p < 0.005), implying that H3 is accepted. This result confirms those of the studies by (Chen and Keefe 2020). The coefficient of moderation role FP (ROA) × SOEs remains positive and significant in models 4 and 8 (t = 3.0523 p < 0.000); (t = 5.48 p < 0.005); thus, H4 is accepted. This result confirms several studies in the prior literature (Jain et al. 2017; Wang et al. 2016).

According to Table 6, the lagged proportions of CEO duality, financial performance, CSR disclosure, and the moderating role of SOEs variable are constant with our preliminary results. Second, the endogeneity problem has been addressed through the use of two-stage least squares (2SLS). The instrumental variable in this regard is a one-year lagged firm efficiency. We also used the average of the firm’s CEO duality, financial performance, CSR disclosure, and SOEs’ moderating role as instrumental variables. The results are shown in Table 6 and the findings are in accordance with our main results.

5. Conclusions and Policy Implications

This research analyzed the outcomes of CEO duality and financial performance on the quality of CSR disclosure and the moderating role of state ownership in China. The study’s findings show a negative relationship between CEO duality and CSR disclosure. Even though Chinese regulators have attempted to promote CSR procedures, a low success rate prevails (Lu and Castka 2009). The results of this study sustain the inference of agency theory regarding CEO duality. CEO duality is a double-edged blade metaphorically (Finkelstein and D’aveni 1994). Even if it can reinforce the central command role, this dominant figure can negatively affect the firm. It can hinder the board’s monitoring function as the power is intensely centralized. The CEO will maintain control, so accountability will be placed at risk, causing a limitation of good corporate governance.

Consequently, the shareholders lose confidence in the firm’s management as the CEO duality merger drastically reduces CEO liability towards the stakeholders. The China Securities Regulatory Commission (CSRC) decided to recommend separating the two top management roles, as CEO duality costs outweighed the benefits. While many corporate governance characteristics justify an organization’s propensity to report CSR efficiently, this analysis focuses on the CEO’s duality. Further research may provide better insights, especially when board independence and monitoring procedures are contingent upon independent directors.

Based on stakeholder management perceptions in China, we examined stakeholder characteristics and speculated that shareholders’ predominance governs Chinese companies’ judgement in regard to report CSR operations. Firm success showed a positive impact on CSR transparency. Our findings also indicate that organizations are interested in the disclosure of their CSR operations, and their level of CSR disclosure is higher when they perform well. State ownership has a regulatory influence on the relationship between CEO duality, company success, and CSR disclosure. The relationship between the company’s success and the possibility of CSR disclosure and its quality is stronger among NSOEs than SOEs.

Limitations and Future Research

The foremost limitation of the study is its emphasis on the general index of CSR. There should be a more thorough review of various aspects of CSR and its impacts on stakeholders. The relevant quality measures of CSR disclosure, such as environmental and recycling disclosure, disclosure on the protection of the interests of staff and customers, and other CSR tools (word limit, content review, phrases, passages, etc.) can also be used in future research. Second, the CEO may wear three crowns. The third crown (position) is acting as the communist party secretary, and it can be worked out whether this promotes or obstructs the financial and non-financial performances of firms.

To assess the firm’s financial performance, this analysis is limited to financial performance metrics, such as ROA. Additional business success metrics, i.e., Tobin’s Q, the market value of total assets divided by asset value, should be included in a potential analysis.