Academia and researchers have debated the relationship between CEO duality and improved firm performance. Two theoretical perspectives primarily present the influence of the dual roles of a CEO of a firm on business performance in terms of the agency and stewardship theories. Stewardship theorists emphasize the positive effect of the dual role of a CEO on better firm performance (Freeman and Hasnaoui, 2011; Abbas et al., 2019b). Scholars supporting the stewardship theory explained that CEO duality could enhance the productivity of business firms by maximizing shareholder interests. The dual role of a CEO encourages administrative efficiency, improves communication, and provides a flexible management arrangement, shielding the interests of a firm through a better business organization (Desai et al., 2003). Business firms with entrepreneurial networks (Abbas et al., 2019c) innovation, knowledge sharing, and social media use (Abbas et al., 2019a) can achieve better sales and profitability. These can help firms improve their business performance (Hussain et al., 2019, 2021).

According to stewardship theorists, “top business firms’ managers have desired to become outstanding agents of their business resources” (Donaldson and Davis, 1991). Senior executives try to raise the profit of their business firms by holding dual positions in a business organization and decreasing agency expenses (Beasley, 1996). In addition to business governance perception, senior management can increase the expertise of executives and the value of shareholders through the knowledge of CEOs of firms of strategic procedures and challenges (Jensen and Heckling, 1995; Bhagat and Black, 2001; Iyengar and Zampelli, 2009). Thus, opinions favoring this theory endorse a positive influence of firm CEO duality on business performance. Conversely, agency theorists highlight the adverse impacts of CEO duality on business performance as the dual position of a CEO may pursue personal benefit rather than the profitability of a firm (Fama and Jensen, 1983). Thus, logically, CEO duality can lead to agency issues between stockholders and business managers. Moreover, CEOs can take advantage of their dual role in a weak structure and assign closer individuals executive positions through CEO directives (Faleye, 2007). These findings suggest that the dual role of CEOs can cause deterioration in the business monitoring structure and lead to diversity in firm performance (Krause, 2017).

The literature presents ambiguous and contradictory arguments supporting the agency and stewardship theories. The empirical literature related to the relationship between CEO duality and business performance is primarily indecisive. Previous studies have offered mixed results, with some robust studies unable to prove a strong relationship between CEO duality effects and better business performance (Duru et al., 2016; Javeed et al., 2020). In contrast, some studies have identified a positive CEO dual role impact on firm performance, while others have reported a negative CEO dual role influence on firm performance. Besides, some studies have found a positive connection between the duality of CEOs of firms and business performance based on the stewardship theory (Donaldson and Davis, 1991; Guillet et al., 2013; Wang et al., 2014; Nekhili et al., 2018). An earlier study examined the influence of CEO duality on better firm performance in the context of European business firms (Pham and Pham, 2020). It identified a positive effect of the CEO’s dual role on firm performance, supporting the stewardship theory (Guillet et al., 2013). Agency theory supporters acknowledged the negative influence of CEO duality on firm performance as the CEO could use the resources of the firm for personal interests. A previous study has investigated this relationship in United States industries and reveals a significant and negative relationship between CEO duality and business firm performance (Tang, 2017). Another study focused on exploring the impact of CEO duality on firm performance by examining datasets from Pakistan, an emerging economy, to find adverse effects on business performance (Naseem et al., 2019). Scholars reported a negative influence of CEO duality when using different study variables to measure business performance, such as return on equity (ROE) of firms, return on investment (ROI), stockholder return, and return on assets (ROA) (Guillet et al., 2013; Yang and Zhao, 2014; Tang, 2017; Naseem et al., 2019; Naseem, 2020). Agency theory advocates argue that a single governance structure might help in decisions that increase the positive effect of CEO duality on various shareholder expenses that lead to conflicts of interests between business managers and multiple shareholders (Iyengar and Zampelli, 2009). However, some studies could not identify any significant and positive CEO duality effect on firm performance (Kang and Zardkoohi, 2005; Iyengar and Zampelli, 2009). Conversely, most of the studies in the literature provide evidence that CEO duality has a negative impact on firm performance, ultimately supporting the agency theory (Wijethilake and Ekanayake, 2019). Hence, from the study findings mentioned above, we posit the following hypothesis:

Studies have suggested that the values and objectives of executives influence the CSR practices of a firm to attain better firm performance. CSR practices affect the decision-making processes of business managers. CSR practices of firms refer to their business activities from ethical and economic perspectives and reflect their economic profit and public welfare approach (Freeman and Hasnaoui, 2011). CSR describes activities in society and is defined as an obligation of groups and firms to decrease or eliminate unsafe corporate practices done in the wrong ways (Tarrow, 2001). The relationship between CSR and corporate governance has been examined in numerous studies. The CSR practices of firms establish business accountability internally and externally and reflect their profits (Chen and Hung-Baesecke, 2014). Investing in CSR activities can help firms to enhance their reputation and relationship with stakeholders (Chen and Hung-Baesecke, 2014; Tang et al., 2021).

Stakeholders can be classified into two groups: investing stakeholders (shareholders) and non-investing stakeholders (environment, local community, employees, customers, and suppliers). Additionally, CSR is said to increase firm value by considering all stakeholders equally and protecting their interests. A firm can decrease its risks and costs by establishing a proper relationship with its stakeholders. Competitive benefits may arise, but a worthy association with all shareholders always positively affects firm values (Harjoto and Jo, 2011). Overall, firm performance is positively related to CSR practices (Rehman et al., 2015). A firm with CSR activities that enhance stakeholder confidence will ultimately lead to higher firm performance (Adams, 2002). The reason for supporting CSR policies is that it causes firms to be accountable to the public and demonstrate their loyalty to society using CSR as a valuable tool (Adams, 2002). Second, shareholders usually assume that optimistic firms engage in CSR activities (Freeman and Hasnaoui, 2011). Some studies support the argument that enhancing environmental activities can improve firm performance (Adams, 2002; Casado-Díaz et al., 2014; Abbas et al., 2019b).

Chief executive officers are highly involved with organizational strategies concerning CSR activities, and this can lead to higher profits. Additionally, CSR revolves around solving issues between the CEO and shareholders in diverse matters (Wang and Chen, 2017). CEOs can use CSR to measure their association with shareholders and maintain a positive relationship (Chen and Hung-Baesecke, 2014; Shah et al., 2021). Thus, managers may want to maximize their long-term profits through improved connection with stakeholders, enhanced CSR practices, and better societal trust (Godos-Díez et al., 2011). However, CSR can be considered a shareholder expense that leads to lower firm performance, because executives invest in CSR activities for personal benefit (Barnett, 2007). Studies have shown that agency costs can be decreased through CSR practices and released to stakeholders (Jensen, 2002). Stakeholder representatives have also stated that CSR generates balanced wealth for shareholders and motivates other investors to participate in CSR firms. Studies have shown that firm performance is positively related to CSR. Furthermore, the CEO of a firm can use CSR to solve the issues of stakeholders and increase their wealth and firm profits (Jensen, 2002; Sen et al., 2006). We propose the following hypothesis by summarizing the above studies:

Empirical and experimental studies related to the influence of CEO duality on Chinese enterprises/firms have produced mixed results (Tang, 2017; Singh et al., 2018; Naseem et al., 2020). However, not much evidence is available on why these results differ in the literature. A previous study has shown that some organizational factors may not allow firms to benefit from CEO duality (Naseem et al., 2020). It suggests that the effect of CEO duality on business performance might depend on some particular government environment; the contingency theory identified firm size as an organizational factor (Li and Chen, 2018). The literature has documented and recognized some moderating mechanisms that might help or limit business activities to achieve their desired objectives, such as interests of managers, firm improvement, and decision-making (Damanpour, 2010; Zona et al., 2013; Li and Chen, 2018). Even though scholars consider firm size (small and large) a critical study variable related to corporate finance studies, several studies consider it a control variable solely for exploring the connection between the duality of enterprise CEOs and firm performance (Li and Chen, 2018; Naseem et al., 2019). This study tries to identify whether firm size (small and large) moderates the relationship between CEO duality and business performance. This study also aimed to observe whether CEO duality improves or creates constraints for better outcomes. Hence, our model uses firm size (small and large) as a regulatory factor to explain why the empirical results on the relationship between CEO duality and company performance have been reported to be contradictory. From a theoretical perspective, small and large firms have different organizational structures (Li and Chen, 2018). For instance, large businesses have better financial resources than small organizations, as big organizations have extra financing opportunities to develop their business. A study has argued that large organizations would obtain more support because of their progress (Ezeoha, 2008). From another point of view, banks are always more eager to provide loans to more creditworthy business firms. The agency theory supporters claim that an organization would grow and try to cross state borders and become a significant business leader as its business develops. It would gain the capability to fulfill the goals of its shareholders and reduce its agency costs (Jensen, 2002; Maladjian and El Khoury, 2014; Mui and Mustapha, 2016; Raza and Karim, 2016; Hashmi et al., 2020). Besides, firm size plays a dynamic role in introducing new products or services to a firm. Large firms can quickly expand their business owing to their better resources as compared with smaller organizations. Small companies may be proficient in the construction and operative area where they are working (Hashmi et al., 2020). Accordingly, large business firms and corporations are more prominent in business activities. Their business operations may also reflect more noticeable and significant processes to produce more products, ensuring greater volumes of sales and profitability (Lee et al., 2014). The higher revenue level of business companies can lead to more income and an increase in profitability. It will enable them to produce more profits from the investment of stakeholders and their equity.

However, larger firms have more resources and a fair market reputation than smaller ones. They are more skilled to introduce new products and achieve desired goals (Rajan and Zingales, 1995; Moeller et al., 2004). In general, larger firms are more advanced and well-organized to respond to market changes. They can react quickly to shareholder pressures and work for shareholder interests rather than the private benefits of CEOs (Chandran and Rasiah, 2013). These firms have more assets and a well-organized structure and can compete in a competitive market and increase their profits (Jensen and Murphy, 1990). These arguments indicate that CEO duality helps larger firms perform better because of their organizational structure, good market reputation, and high market share (Weyzig, 2014).

Scholars have tried to study the effect of firm size on the profitability of business companies with different results ranging from negative, positive, or weak in terms of the selected variables. Previous research has reported a positive association between firm size and profits (Ibhagui and Olokoyo, 2018). One study analyzed the business performance of Greek firms over the period 1995–1999 (Papadogonas, 2007). This study divided Greek business firms into classes and employed regression analysis to draw the desired results. The results indicated that firm size affected the productivity and profitability of all categories of business companies. Studies have also examined the role of firm size by employing a fixed-effects model on a sample of over 7,000 publicly held United States firms (Lee, 2009). The research findings indicated that the size of the business firms played an indispensable role in enhancing sales and profitability (Amato and Burson, 2007). One study tested the relationship between profitability and firm size of companies operating in the financial services sector. Another study examined the nature of the relationship between firm size and profitability by employing financial and economic data (Andries and Faems, 2013). A study recognized a significant dissimilarity in profitability among medium, small, and large corporations (Ibhagui and Olokoyo, 2018). The investigation of firm size on the relationship between CEO duality and the performance of business firms assumes that the activities of executives and firm performance vary by firm size (Moeller et al., 2004; Dang et al., 2018; Sarfraz et al., 2020). Even though theoretical recommendations described the role of a CEO as changing over time with the size of other firms (Dang et al., 2018), these suggestions lead to possible pressure between firm size and CEO duality and are linked to firm performance (McWilliams et al., 2006). From the above debate, our model theorizes that firm size is a powerful mechanism to examine how CEO duality can increase or improve the performance of business firms (Yang and Zhao, 2014). The results of this study indicate why and how a selected variable affects other variables (Baron and Kenny, 1986). From this debate, we posit the following hypotheses:

This present proposed study model primarily focused on examining the relationship between CEO duality role and sustainable business performance of Chinese firms. This study explored how firm size and CSR practices moderate the relationship between CEO duality and the performance of Chinese business firms. This proposed study has incorporated four factors into this designed model. In this proposed research model, the duality of CEO is the independent variable, and this study incorporates the performance of Chinese firms as the dependent variable. CSR practices and firms’ size (small and large) are the moderating variables. The study applied a generalized method of moments (GMM) model to investigate the effects of CEO duality on business performance by taking advantage of previous research (Singh et al., 2018). Figure 1 demonstrates the theoretical framework of our proposed model after a detailed review of the scientific literature. See Figure 1, which describes the study model explained below.

The primary datasets used in this research study are from the China Stock Market and Accounting Research (CSMAR) database, which contains the data of A-share listed companies and offers extensive details of financial data of Chinese firms (Mahmood et al., 2019). The initial sample consisted of listed Chinese firms in the manufacturing sector. This study employed a sample based on the 2012–2017 panel data because information related to the influence of CEO duality on business performance was not available and omitted. The datasets in the study are summarized as follows: (1) all companies with missing data were included, and then businesses with missing CEO duality and CSR information were removed. (2) The data of organizations that altered their administrative management structure were excluded, such as CEO single and dual positions (CEO and chair for the board of director positions). (3) Finally, a sample of 417 companies with 2,502 observations was selected for 6 years. As the biggest developing economy, China shows fast economic progress and great success in the emerging economy (Mamirkulova et al., 2020). Moreover, market participants and government authorities developed the governance structure of Chinese firms. For instance, an increasing number of firms in China have been considering CSR activities as a social good (Mubeen et al., 2020). A distinctive governance structure and quick stakeholder response make China an exciting research setting. Many scholars examine how CEO duality influences the performance of business firms in China.

This study employed the GMM based on the support of existing studies that are based on some logical factors (Singh et al., 2018). First, we incorporated a panel dataset to examine the relationship between CEO duality and firm performance through the moderating role of firm size and CSR. Previous studies have shown that these endogeneity problems exist typically in panel datasets to produce biased and unreliable results (Javeed and Lefen, 2019). According to econometrics, if a variable has endogenous issues in the proposed model, scholars will suggest helpful ways to deal with them (Tien et al., 2013; Javeed et al., 2020).

Second, in a regression model, when there is a correlation between error terms, variables face endogeneity issues. In addition, when CEO decisions are correlated to error terms because of potential factors, these problems may arise from automatic regressions with missing variables, autocorrelation errors, and measurement errors (Adams et al., 2010). Therefore, these endogenous issues need to be controlled in CEO duality analysis.

Third, studies related to CEO duality positions advocate that CEO duality exists in potential endogeneity issues (Guillet et al., 2013). For instance, the appointments of CEOs and CEO chairs are endogenous (Nekhili et al., 2018). The probable effect of this on the performance of the CEO and CEO chair might determine the characteristics of firms that influence the CEO role and firm performance. Studies have also specified that the single leadership structure of a company is primarily endogenous and that the CEO dual position and CEO chairman role are related to some unobserved company characteristics, which could lead to endogenous problems (Kang and Zardkoohi, 2005; Li and Chen, 2018). Nevertheless, this study finds a positive relationship between endogeneity and unobservable firm characteristics.

Fourth, forming a panel dataset places limitations on the ordinary least squares (OLS) model, because it leads to biased estimations and could be inefficient because of unobserved heterogeneity (Wooldridge, 2002; Duru et al., 2016; Bae et al., 2018; Fralich and Fan, 2018; Javeed et al., 2020). Accordingly, it is particularly essential to control for endogenous problems in CEO duality analysis. Consequently, this study applied the GMM approach to explore the relationship between CEO duality and company performance to solve this endogenous problem. The findings of this study are in line with the literature (Singh et al., 2018).

Several econometric procedures are available to manage endogeneity issues. For instance, random effects and fixed-effect models, control variables, instrumental and lagged dependent variables, and the GMM model help control such issues (Li, 2016). Among all the analytical panel data techniques, the GMM model is the best approach with the highest power to deal with endogenous issues (Li, 2016). Because of its higher ability to deal with endogeneity, the preferred technique for this study is the GMM model. Previous studies have recommended the use of this model (Arellano and Bond, 1991). Within a company, the GMM model allows for autocorrelation and heteroscedasticity. The reliability of this model depends on instrumental validity. This study has described the details of the GMM model used, with suitable citations to support it. This model applied GMM, which allows testing for heteroscedasticity and autocorrelation of the chosen framework. The consistency of the GMM model depends on instrumental validity. Moreover, the GMM model is a more convenient and reliable approach than other techniques, as it has extreme effects to examine coefficient correction (Zou et al., 2015; Lin et al., 2019; Javeed et al., 2020, 2021; Banto and Monsia, 2021; Mittal and Garg, 2021). This study has incorporated the statistical software STATA for analysis purposes. It has tested a direct and moderating impact using the STATA software to confirm the significance of a test significance and satisfactory level. Tables 4, 5 present test results that show appropriate levels. Here, “*”indicates the significance level in this study model [significance levels: ^***p < 0.01, ^**p < 0.05, and *p < 0.1]. Accordingly, the GMM model was employed in this study, as recommended by numerous past studies, and our results are consistent with the literature findings (Li and Chen, 2018; Matemilola et al., 2018; Singh et al., 2018; Madaleno and Vieira, 2020; Boakye et al., 2021; Mittal and Garg, 2021).

This study examined the robustness of the tests and used ROI and ROE to measure firm performance. It investigated the ROI performance measure and used ROI as the proxy in the first test. In the second phase, the ROE was applied as a proxy to measure performance to find a satisfactory level, as indicated in Tables 4, 5. These findings provided adequate results based on the two variables to measure firm performance. We applied the STATA software for analysis and to draw the results. Before incorporating the GMM, we performed some tests to analyze the validity of the instruments, that is, whether they are suitable for this analysis or not. First, we tested the variance inflation factor (VIF) approach to examine the multicollinearity issues among the chosen variables. The VIF test results indicated no multicollinearity problem in the variables of the selected model. The Wald test for heteroscedasticity was applied, and the results confirmed that no heteroscedasticity variables existed among the chosen variables. Third, a Sargan analysis was performed on the model to check the validity of the selected instruments and over-identifying restrictions. The Sargan analysis results showed that the instrument variables were valid for providing support for the chosen variables. Finally, we used a GMM method to solve the endogeneity issues between the error terms and variables. The outcomes of all the study tests revealed that the data provided adequate results and did not suffer from weak instruments.

Here,

Table 3 demonstrates the empirical analysis of the descriptive statistics of this study along with the VIF (variance inflation factor) of the dependent and independent variables of the study model based on “Panel A.” This study used two dependent variables, ROI and ROE. Panel A specifies that the average ROI value is 0.34 and that the standard deviation value is 0.605. The ROE value is 0.461, and the standard deviation is 0.985. Firm size and CEO duality have an average output of 10.09 and 0.258, and standard deviation values are 0.57 and 0.437. The capital growth indicates a mean value (M = 0.088) and a standard deviation (SD = 0.094). The firm advantage reveals an average value of 0.088. The asset tangibility shows mean value (M = 3), with a standard deviation (SD = 5.87), and the CSR an average value (CSR = 1.95), with a standard deviation (SD = 1.76).

“Panel A” specifies the descriptive statistics analysis. In a regression model, multicollinearity can result in higher standard error outcomes and make the inference difficult and biased. Accordingly, this study employed VIF analysis (variance inflation factor) to trace the multicollinearity issues. The study applied the VIF to confirm the absence of multicollinearity problems in this sample. Panel B examines the VIF analysis. The average values of the chosen variables are lower than 10, which confirms that the data are free from multicollinearity issues. Higher values of VIF might indicate that data suffer from multicollinearity problems (Hair et al., 2016, 2017). Table 3 shows the detailed VIF values related to Panel B. See Table 3 below.

Table 4 indicates the results related to the stated hypotheses (H 1 and H 2) of this study and the coefficient estimates of variables to explore the direct relationship between CEO duality and the performance of Chinese firms with the moderating role of CSR in this relationship. Models 1 and 2 reveal the direct impact of CEO duality on firm performance with two different performance measurements, as ROI and ROE. Model 1 suggests that CEO duality impact is statistically significant and negative and that the coefficient value is ROI = −0.534. Besides, Model 2 displays that CEO duality is negative and substantial, with an ROE = −1.21 coefficient value. These results support hypothesis 1, which describes that CEO duality has a negative influence on firm performance. The study results designated that the effect of CEO duality exhibits a significant adverse relationship to both dependent variables ROI and ROE. The study results specify that CEO duality has a negative impact on Chinese firm performance. Models 3 and 4 exhibited that CSR activities moderate the relationship between the duality of CEOs and Chinese firm performance connection. This study created models 3 and 4 to explore how CSR practice plays a moderating role in the relationship between CEO duality and Chinese firm performance. Models 3 and 4 results show that the coefficient designated a fair value of 1.521 and 1.01 with performance measurements ROI and ROE, respectively. These outcomes specified that the CEO duality position exhibited a significant and positive influence on the performance of Chinese firms by adding CSR as a moderating variable. The positive impact of CSR on CEO duality and firm performance association confirmed the proposed hypothesis 2. Hypothesis 2 of this study stated that CSR practices moderate the relationship between the duality of CEOs and the performance of Chinese firms. The results advocate that CEO duality effects proved a substantial and positive relationship with Chinese performance measurements (ROI and ROE) through CSR moderating effects. The argument supported that this positive association is noticeable in Chinese firms with CSR activities. These findings confirm that the influence of CEO duality on business firm performance has a positive association with the effects of CSR and that the results of the study agreed with the past literature. As shown in Table 4 for further information.

Table 5 presents the findings of this study. The results of this study indicate that CEO duality influences the performance of Chinese firms through the moderating role of firm size. Models 5 and 6 specify that CEO duality affects the performance of Chinese firms through the moderating role of small firms. Besides, model 5 designates that the coefficient value of CEO duality is ROI = −1.58 at a 1% significance level. Similarly, model 6 specifies that the coefficient value of CEO duality is −2.11, statistically significant at a 1% significance level. The study findings indicate that small firms have a significant and negative relationship with CEO duality and its impact on the performance of business firms, with performance measurements evidenced with ROI and ROE.

Models 7 and 8 display the connection between CEO duality and its impact on Chinese firm performance with the moderating effects of large firms. Model 7 presents the coefficient value of CEO duality to be 0.155 and stipulates a positive and statistically significant influence on business performance. Similarly, model 8 indicates a positive and considerable power of CEO duality coefficient with a value of 1.027. Both study models offer a positive coefficient and specify that large firms positively impact CEO duality influence on the performance of Chinese business firms (Raith, 2003). Refer to Table 5 below.

The findings of this study recommend some future implications to describe the benefits of the dual role of CEO on Chinese firm performance with the moderating influence of firm size (small/large) and CSR practices. First, executives need to consider the demands of their stakeholders and develop a robust approach to solving their problems. For instance, a CEO can increase the profits of a firm by improving CSR activities and engaging employees and societies. Second, this study encourages investors to play an active role in investing in large business firms to exploit more profits. Besides, the findings of this research study acknowledged that firm size moderated the impact on the relationship between CEO duality and Chinese firm performance. It might lead key shareholders to invest in large firms to take advantage of CEO dual positions. Third, the most critical two-leadership positions are not necessary/must for healthier business performance, even though the joint leadership position is beneficial for the firm and helps develop firm values due to closer relationships among managers and shareholders. CEO duality is not necessarily for adverse effects on firm performance. This study also added to the theoretical discussions of CEO duality theories, which proposed that CEO dual role leads to superior firm performance.

This study has some limitations like any empirical research. First, the study datasets reported on a sample of Chinese business firms (listed companies), and significant limitations refer to a unique organized business environment for Chinese firms. However, it is the second-biggest global market. For instance, publicly available information influences the market, and it will manifest in share prices. Thus, the findings of this research model might not offer generalizability for other regions and countries. However, studies from multiple contexts are relatively rare. Conversely, the limitations suggest a potential for future studies and might help understand the connection of CEO duality and firm performance through double moderating effects with different variables. Besides, this research only discussed the moderating role of firm size (small and large) and CSR on the relationship of CEO dual roles and firm performance. It did not comprise other relevant firm strategies (ownership structure) to describe the relationship between CEO duality and firm performance. Hence, future studies can test these strategies and add some other mediators or moderator variables. This research model recommends other potential moderators, such as ownership structure, firm age, and financing configuration, by selecting some other economy for future studies to explore compelling results for business managers, CEOs, and other stakeholders.