Exploring New Green Frontiers? How CEO Green and Technological Experience Shapes Firm Ambidextrous Green Innovation

Abstract

Green innovation has gained prominence in managerial practice and academic discourse. We build on recent findings that CEO green experience promotes firm green innovation by examining how it affects the balance between exploratory and exploitative green innovation. Recognizing that the extent to which executive attributes translate into firm-level actions and results is contingent on contextual factors, we further test the moderating roles of firm ownership, industry environmental sensitivity, and regional economic development. Utilizing a panel dataset of Chinese non-financial A-share listed companies spanning 2007 to 2023, the empirical results provide compelling evidence that CEO green experience is positively associated with exploratory green innovation. This positive association is more pronounced for non-state-owned enterprises, firms operating in environmentally nonsensitive industries, and those located in more developed regions. Our theory and findings contribute to the scholarship on CEO characteristics and green innovation. This study also delivers managerial guidance for enterprises aiming to achieve sustainable growth.

1. Introduction

Amid rising global concerns about environmental challenges and the heightened prioritization of sustainable development, green innovation has gained prominence in both managerial practice and academic discourse [1,2,3,4]. As a strategic approach, it bridges economic pursuits and environmental responsibilities, functioning as a critical driver of sustainable competitive advantage [5,6]. Building on this strategic role, organizations engage in green innovation to address societal demands for sustainability [7] and strengthen their competitive positioning by enhancing their image and reputation [8], while simultaneously improving resource utilization and reducing pollution-related expenditures to improve cost efficiency [9]. However, green innovation means substantial resource commitment and necessitates transformational shifts in a firm’s strategic orientation, core technologies, and operational processes [10]. Moreover, the inherent externalities associated with green innovation as a public good [11], coupled with the uncertainty from prolonged R&D cycles, high risks, and delayed returns on investment [6], hinder enterprises’ efforts toward green innovation.

To overcome these barriers, strategic leadership, particularly the CEO, plays a pivotal role in driving innovation activities [12,13]. According to the upper echelons theory, CEO characteristics can significantly shape strategic choices and organizational outcomes [14]. An increasing body of literature on the antecedents of green innovation acknowledges the significant role of CEO characteristics, which are typically categorized into innate traits and acquired factors. The former pertains to inherent and relatively stable individual attributes, such as gender [15] and age [16], whereas the latter reflects dynamic experiences and environmental exposures, such as hometown identity [17], technological background [18], and overseas experience [19]. Distinct from innate traits, acquired factors not only equip CEOs with specialized knowledge and capabilities but also profoundly influence their cognitive frameworks and decision-making approaches, thereby having a pronounced impact on the strategic choices and behaviors of the firms they lead [20,21,22]. Acquired through green-related education or work [8], the green experience of CEO is a particularly salient, acquired, and sustainability-oriented characteristic that enhances their awareness of environmental and sustainability issues and increases their technological knowledge of green innovation [23,24,25]. Furthermore, the attention-based view posits that organizational actions and outcomes are fundamentally shaped by where decision-makers focus their limited attention [26]. Green experiences shape CEOs’ allocation of attention toward environmental and sustainability issues, which in turn guide firms’ green innovation strategies [8,23,27]. While scholars have acknowledged the positive role of CEO green experience in promoting green innovation [25,27,28], the limited understanding of the effect on different types of green innovation constrains the development of a more nuanced view of firms’ strategic green innovation orientations. Anecdotal evidence from the business press also further illustrates the influence of green-experienced CEOs on firms’ green innovation, with examples including Lisa P. Jackson of Apple [29], a former U.S. Environmental Protection Agency administrator; Yin Zhang of Nine Dragons Paper [30], who holds early career experience in environmental issues; and Yunfeng Bai of CPCEP Group [31], who graduated with a degree in an environment-related field.

Under the ambidextrous innovation framework, balancing the effort between exploitation and exploration is vital to a firm’s sustained competitiveness [32,33]. Excessive focus on exploitative innovation may result in stagnation and obsolescence, whereas disproportionate emphasis on exploration could hinder firms’ ability to effectively capitalize on innovative ideas [34]. Therefore, pursuing innovation without achieving an appropriate equilibrium between exploitation and exploration may significantly harm firms [35]. This trade-off is equally salient in the context of green innovation, requiring firms to balance ambidextrous activities. Exploitative activities focus on improving and refining existing green products, services, technologies, and processes within existing knowledge and capabilities [36]. Exploratory efforts aim to discover and develop novel green products, services, technologies, and processes by pursuing new knowledge and resources [37]. While the importance of green ambidextrous innovation has been widely acknowledged, prior research has primarily focused on external pressures and internal organizational factors—such as environmental regulation [38], shareholder attention [11], and green entrepreneurial orientation [39]. Despite this focus, considerably less attention has been devoted to CEO characteristics, particularly their personal experiences.

To bridge these gaps, we draw on the upper echelons theory and attention-based view to examine how CEO green experience affects firms’ engagement in exploratory and exploitative green innovation. Furthermore, adopting a contingency perspective [40] and the situated attention principle [26], we posit that this relationship depends on the alignment between CEO characteristics and the contextual factors at the firm, industry, and regional levels [15]. Accordingly, we examine the moderating roles of firm ownership, industry environmental sensitivity, and regional economic development to capture these contextual contingencies. As one of the world’s largest emerging economies and a major carbon emitter, China offers a particularly salient context for studying green innovation [41]. Leveraging data spanning 2007–2023 for China’s A-share listed companies, we estimate a fractional response model [42,43] to conduct empirical analyses. The results indicate a positive (negative) association between CEO green experience and firm exploratory (exploitative) green innovation, which is more pronounced for non-state-owned enterprises (non-SOEs), firms operating in non-environmentally sensitive industries, and those situated in more developed regions.

This study enriches the literature through conceptual insights supported by empirical evidence. First, we extend the literature on the relationship between CEO characteristics and green innovation by disentangling the heterogeneous effects of CEO green experience on exploratory and exploitative green innovation. Given the importance of exploratory efforts for long-term benefits, our findings underscore the critical role of CEOs’ cognitive bases and green-related expertise in driving green innovation. Second, this study contributes to the ambidextrous green innovation literature by shedding light on CEO characteristics as critical antecedents. Previous studies have examined the influence of external pressures and organizational-level factors on ambidextrous green innovation [15,19]. However, the role of CEO characteristics and how firms navigate the inherent trade-off between exploratory and exploitative innovation remain underexplored. Our findings provide a nuanced understanding of how CEOs with green experience shape firms’ approaches to balancing ambidextrous green innovation. Third, we enrich the understanding of the contextual contingencies with which CEO green experience combines to shape ambidextrous green innovation by analyzing firm-, industry-, and regional-level factors. The empirical evidence further clarifies the boundary conditions under which CEO green experience influences ambidextrous green innovation.

The remainder of the paper is structured as follows: Section 2 develops the theoretical analysis and research hypotheses. Section 3 presents the sample selection and data collection process, variable measurement, and empirical model. Section 4 reports the empirical results, including descriptive statistics, correlation analysis, hypotheses tests, endogeneity tests, robustness tests, and supplemental analyses. Section 5 concludes with discussions of the key findings, theoretical and practical implications, and future directions.

2. Theoretical Analysis and Hypothesis Development

2.1. Review of Relevant Literature

Consistent with “environmental innovation,” “sustainable innovation,” and “eco-innovation,” green innovation can be described as innovation aiming to mitigate the adverse environmental and social impacts of business operations [3,44]. Rather than being a single, undifferentiated concept, green innovation encompasses diverse innovation activities that vary in strategic focus, resource commitment, and risk exposure [45]. Drawing on the ambidexterity innovation framework [32,46], green innovation can be conceptualized in two distinct forms: exploratory and exploitative. The former refers to substantial disruptive advancements capable of reshaping existing technological trajectories, whereas the latter emphasizes incremental improvements in or refinements to existing products, processes, or practices [47]. Although exploratory initiatives can be more effective in securing a green competitive advantage, they demand greater resource commitments and entail higher levels of risk [11].

The balance between exploratory and exploitative green innovation represents a strategic choice [11] that fundamentally reflects executives’ cognitive bases and value orientations [14]. As the pivotal figure, the CEO plays a decisive role in shaping organizational strategy and outcomes [48,49]. Consequently, the CEO’s individual characteristics—such as demographics, personality traits, and professional experience—are regarded as reliable predictors of strategic behavior and firm-level outcomes [50]. Existing research demonstrates that CEO experience influences firm green innovation, highlighting how differences in CEOs’ background characteristics shape their value orientations and cognitive capabilities, which in turn affect corporate engagement in green innovation initiatives [18,35,51]. In this study, CEO green experience refers to either environment-related education or work experience in environment-focused roles within governmental or corporate organizations [8,23,28,52].

2.2. CEO Green Experience and Firm Exploratory (Exploitative) Green Innovation

CEOs with green experience develop a distinct knowledge base characterized by heightened environmental awareness, strong sustainability-oriented values, and practical expertise in responsible environmental practices [24]. Such cognitive and value foundations shape their strategic orientation, prompting them to place greater emphasis on sustainable development and to frame external environmental challenges as opportunities for innovation and organizational transformation [8]. Furthermore, given limited managerial attention, CEOs with green experience are more likely to allocate their attention to radical sustainability transformation and shape their firms’ strategic orientation toward novel green technologies and new technological trajectories. Firms led by such CEOs tend to channel their green innovation efforts toward exploratory rather than exploitative activities. Consequently, green experience leads CEOs to prioritize substantive transformation over incremental improvements in pursuing green innovation, while their domain-specific expertise enables them to mobilize resources for exploratory initiatives and effectively manage associated risks.

Specifically, exploratory green innovation requires substantial resource investments for breakthrough transformations. The environmental acumen of green-experienced CEOs allows them to foster stakeholder relationships in the environmental domain, thereby broadening their firms’ access to external resources [51]. Such stakeholder-oriented engagement strengthens the firm’s environmental legitimacy and increases the likelihood of obtaining government subsidies and policy support [53]. Additionally, appointing a green-experienced CEO signals the firm’s commitment to sustainable development and helps attract investment from sustainability-focused institutional investors [54]. Moreover, by evaluating the potential of green knowledge and technologies, these CEOs can leverage their environmental expertise to allocate resources more efficiently for exploratory green innovation [28].

In addition, exploratory green innovation involves substantial uncertainty, extended development horizons, and intricate technological challenges, which collectively intensify organizational risk exposure. Leveraging their professional knowledge and prior involvement, green-experienced CEOs are well equipped to mitigate these risks [55]. Such CEOs are able to recognize emerging trends in green technologies and offer strategic foresight across the innovation process, thus enhancing the firm’s resilience to technology-related uncertainties. As for regulatory and market-related risks, their familiarity with environmental regulations and heightened responsiveness to policy and market shifts allow firms to manage these challenges proactively. Accordingly, CEOs with green experience not only steer the strategic course of exploratory green innovation but also successfully navigate the complex landscape of its implementation.

Based on the preceding analysis, firms led by CEOs with green experience are expected to exhibit a higher proportion of exploratory green innovation. Thus, we propose the following hypothesis.

H1: 

CEO green experience is positively (negatively) associated with firm exploratory (exploitative) green innovation.

2.3. Contextual Moderators of the Impact of CEO Green Experience on Firm Exploratory (Exploitative) Green Innovation

Executives’ personal characteristics shape how they interpret situations, thereby influencing their decision-making processes and the organization’s strategic behavior [48]. However, contextual factors moderate the degree to which executive attributes translate into firm-level actions [56]. As Hambrick (2007) emphasized, the reflection of managerial characteristics on organizational strategy and performance is contingent on both environmental conditions and organizational factors [48]. The principle of situated attention also suggests that decision-makers’ focus and subsequent actions are shaped by the specific contexts in which they are embedded [26]. Therefore, building upon prior research on the relationship between CEO experience and green innovation [15,19,57], we introduce firm-, industry-, and regional-level moderating variables—namely, firm ownership, industry environmental sensitivity, and regional economic development—to examine the boundary conditions.

The influence of CEO green experience on green innovation varies across enterprises with different ownership structures. State-owned enterprises (SOEs) operate with a dual mandate to achieve economic returns and deliver social benefits [58]. As key actors in the national sustainability agenda [59], SOEs face greater institutional pressure to engage in green innovation [60]. To satisfy heightened legitimacy expectations, SOEs tend to redirect their attention in green innovation from radical transformation toward incremental improvement, often prioritizing outcomes that can be achieved within a shorter timeframe. Furthermore, their distinctive governance structures reinforce the preference for exploitative green innovation, as executives face unique pressures. Innovation failures can result not only in economic losses for the firm but also in potential accusations of dereliction of duty against CEOs [61]. Thus, even if CEOs are ambitious in pursuing exploratory efforts, concerns over career progression and safeguarding their reputation can still lead them to favor risk-averse strategies [62].

This dynamic is further explained by the stark differences in managerial discretion between the CEOs of SOEs and non-SOEs. CEOs in SOEs are typically appointed by the government and operate under significant institutional constraints, which reduce their managerial discretion [63]. This lack of autonomy limits the extent to which CEO green experience can shape the firm’s green innovation orientation. Unlike their counterparts in SOEs, CEOs of non-SOEs are generally selected through market-based mechanisms, which grant them greater strategic autonomy [8]. This autonomy enables them to shape the firm’s green innovation agenda in line with their environmental commitments. Furthermore, with relatively lower institutional pressures and a more flexible governance structure, CEOs are better positioned to direct their attention toward the breakthrough development of green technologies. In light of the preceding arguments, we propose the following hypothesis.

H2: 

The positive (negative) association between CEO green experience and firm exploratory (exploitative) green innovation is more pronounced for non-SOEs.

The influence of CEO green experience on a firm’s green ambidextrous innovation is contingent upon the environmental sensitivity of its industry. Firms in environmentally sensitive industries confront intense institutional pressures due to the significant environmental impact inherent in their operations [64]. These pressures manifest from two primary sources: regulatory bodies and public opinion. First, these firms are subject to rigorous government oversight, including stringent environmental standards, frequent inspections, and substantial penalties for non-compliance [63]. This regulatory environment compels firms to pursue green transformation to maintain their regulatory legitimacy. Second, any environmental incidents in these sectors are highly likely to attract widespread media attention and public concern, placing the firms under intense societal scrutiny [65,66]. Under this dual pressure, firms are strongly motivated to direct their attention toward immediate and visible environmental actions to secure both regulatory and social legitimacy. Consequently, they tend to prioritize exploitative green innovation for quick and tangible results.

Conversely, firms in non-environmentally sensitive industries operate with lower institutional pressure and greater strategic latitude. This context affords CEOs greater managerial discretion in defining the firm’s green innovation agenda [15]. Freed from the overwhelming need to demonstrate compliance through conservative measures, CEOs with green experience are better positioned to translate their environmental vision into more ambitious and exploratory forms of green innovation. Based on the previous discussion, we propose the following hypothesis.

H3: 

The positive (negative) association between CEO green experience and firm exploratory (exploitative) green innovation is more pronounced for firms in environmentally insensitive industries.

The effect of CEO green experience on green ambidextrous innovation is contingent on regional economic development. Pursuing exploratory green innovation is a resource-intensive endeavor that demands substantial financial, technological, and knowledge-based support [15,67]. Developed regions are inherently better positioned to provide these foundational resources. Moreover, the advanced institutional conditions reduce external uncertainties and enforcement risks [68], creating a stable environment for innovation activities. Firms can effectively seek intellectual property protection through patent applications [69]. Meanwhile, the prospect of securing exclusive rights provides firms with strong incentives to generate economic returns through patent licensing. These contextual conditions heighten their attentional focus on long-term sustainability challenges and novel opportunities, thereby enabling them to channel their green experience into fostering exploratory rather than exploitative green innovation.

Conversely, firms in less developed regions face a dual challenge that stifles exploratory green innovation. Financial and resource constraints create substantial barriers to funding resource-intensive initiatives. Simultaneously, an underdeveloped institutional environment, with inadequate legal and intellectual property protection, undermines the strategic rationale for these efforts [70]. For a CEO with green experience, the heightened risk of knowledge leakage and the inability to safeguard proprietary technologies critically diminish the potential returns of exploratory projects. This erosion of incentives discourages investments in novel developments and patenting, compelling CEOs to redirect focus toward safer, exploitative green innovation. Accordingly, the following hypothesis is formulated.

H4: 

The positive (negative) association between CEO green experience and firm exploratory (exploitative) green innovation is more pronounced for firms in more developed regions.

3. Research Design

3.1. Sample and Data

China serves as an appropriate and representative research context for examining firms’ green innovation. As the largest developing economy and one of the world’s major carbon emitters, China faces the dual challenge of sustaining rapid growth while addressing environmental pressures [71]. With its commitment to the “dual-carbon goals” and the implementation of a series of environmental policies, Chinese firms’ green innovation is now in a vigorous stage of development [41].

We select all A-share listed companies from 2007 to 2023, which are the years when the new accounting standard took effect, as our initial sample. Furthermore, we screen the sample by eliminating firms in the financial sector, excluding those labeled ST or PT, and dropping observations with missing values for core variables. This yields 28,768 firm-year observations for 3382 companies as the final sample. To reduce the influence of extreme values, all continuous variables are winsorized at the 5% level.

CEO green experience data are manually compiled from the China Stock Market & Accounting Research (CSMAR) Database using information on CEO personal characteristics. Measures of exploratory green innovation are obtained from the Chinese Research Data Services (CNRDS) Platform. Control variables at the CEO, board, and firm levels are sourced from the CSMAR Database.

3.2. Measurement of Variables

3.2.1. Exploitative and Exploratory Green Innovation

The existing literature widely recognizes green patents as a valid indicator of green innovation activities [19,51]. According to the International Patent Classification (IPC) Green Inventory, green patents are identified by screening patent codes [11]. Patent applications whose codes align with the IPC Green Inventory are classified as green patents. To distinguish between exploitative and exploratory types, we adopt the measurement of ambidextrous innovation based on the four-digit IPC codes [72]. Technological knowledge is assessed over a five-year rolling window [73]. A patent is classified as exploratory if at least one of its IPC codes in a given year has not appeared in the prior five years. Conversely, if all the IPC codes of a green patent appear within the past five years, the patent is categorized as a green exploitative innovation. We then assess firm-level exploratory (vs. exploitative) green innovation as the ratio of the number of exploratory green innovations to the total number of green patents filed by a firm in a given year [35]. Therefore, the variable, exploratory green innovation (Exploratory), takes values ranging from 0 to 1, with higher values indicating a greater degree of exploratory green innovation at the firm level.

3.2.2. CEO Green Experience

We measure CEO green experience using information on their educational background and professional work history [8,23,24,28,52,74,75]. Specifically, we manually review CEO resumes to identify green-related education or work experience. The former is identified by whether the CEO obtained a degree in an environment-related discipline, such as pulp and paper engineering, environmental studies, environmental engineering, or environmental science. The latter is determined by whether the CEO has held a position involving governmental institutions such as the Ministry of Environmental Protection and the Environmental Protection Committee or has served as a person responsible for pollution control within enterprises. The variable, CEO green experience (Green), is assigned a value of 1 if the CEO has green-related education or work experience, and 0 otherwise.

3.2.3. Moderators

We consider firm ownership, industry environmental sensitivity, and regional economic development as moderators representing the firm-, industry-, and regional-level contexts, respectively. Following existing studies, firm ownership (SOE) is defined as a dummy variable that takes the value of 1 if the enterprise is state-owned, and 0 otherwise [76]. Drawing on the approach of Javed et al. (2023) [15], we construct a dummy variable, environmental sensitivity (Sensitive_Industries), coded 1 for firms in environmentally sensitive industries (such as agriculture, forestry, chemical, fishing, mining, and construction), and 0 for all other industries. Consistent with the literature, regional economic development (Regional_Development) is a dummy variable equal to 1 if the National Economic Research Institute (NERI) marketization index of the firm’s province exceeds the national median in that year, and 0 otherwise [77].

3.2.4. Controls

Based on empirical studies on green and ambidextrous innovation, our model incorporates a comprehensive set of control variables encompassing CEO characteristics, board structure, and firm-level attributes. CEO characteristics include age (CEO_Age) [16], gender (CEO_Gender) [15], and overseas experience (CEO_Overseas) [19]. Board-related variables cover CEO duality (Duality) [28], board size (Board_Size) [19], board independence (Board_Independence) [15], and board diligence (Board_Diligence) [15], while firm-level controls account for firm size (Firm_Size) [11], financial leverage (Leverage) [11], profitability (ROE) [11], firm growth (Growth) [11], and equity concentration (Top5) [8].

Table 1. Measurement of the variables.

Variable	Measurements
Exploratory	The proportion of the number of exploratory green innovations in the total number of green invention applications.
Green	A dummy variable that equals 1 if the CEO has green experience, and 0 otherwise.
SOE	A dummy variable that equals 1 if the firm is state-owned, and 0 otherwise.
Sensitive_Industries	A dummy variable that equals 1 if the firm belongs to environmentally sensitive industries, such as agriculture, forestry, chemical, fishing, mining, or construction, and 0 otherwise.
Regional_Development	A dummy variable that equals 1 if the NERI marketization index of the firm’s province exceeds the national median in that year, and 0 otherwise.
CEO_Age	The age of the CEO in years.
CEO_Gender	A dummy variable that equals 1 if the CEO is female, and 0 if male.
CEO_Overseas	A dummy variable that equals 1 if the CEO has overseas experience, and 0 otherwise.
Duality	A dummy variable that equals 1 if the CEO and the chairman of the board are the same individual, and 0 otherwise.
Board_Size	The number of board directors.
Board_Independence	The proportion of independent directors on the board.
Board_Diligence	The number of board meetings held during the year.
Firm_Size	The natural logarithm of total assets.
Leverage	Total debt divided by total assets.
ROE	Net profit divided by total assets.
Growth	Revenue growth rate.
Top5	Shareholding proportion of the top 5 shareholders.

presents details of the measurement of the variables.

3.3. Empirical Model

To account for the bounded nature of the dependent variable (ranging from 0 to 1), we utilize a fractional response model in our empirical analysis. As a nonlinear extension of the generalized linear model specifically designed for proportion data, the fractional response model provides more reliable and unbiased estimates than ordinary least squares (OLS) when the outcome variable is limited to a [0, 1] interval [42]. Furthermore, in nonlinear models with unbalanced panels and a large cross-sectional dimension relative to the time span, introducing individual fixed effects may cause biased and inconsistent estimates due to the incidental parameters problem [43]. Therefore, we include year fixed effects and industry fixed effects instead of firm fixed effects in our regressions [35]. We specify the model using a probit link function, and standard errors are clustered at the firm level to account for heteroskedasticity and within-firm correlation. Equation (1) outlines the empirical model.

$${Exploratory}_{i,t}={\beta }_0+{\beta }_1{Green}_{i,t}+{\beta }_2{Controls}_{i,t}+{\mu }_i+{\omega }_t+{\epsilon }_{i,t}$$

(1)

where i denotes the company and t denotes the year. Controls are constructed using a set of control variables. $\mu$, $\omega$, and $\epsilon$ refer to the year-fixed effects, industry-fixed effects, and random error term, respectively.

4. Empirical Results

4.1. Descriptive Statistics and Correlation Analysis

Table 2. Descriptive statistics.

Variable	n	Mean	SD	Median	Min	Max
Exploratory	28,768	0.026	0.053	0.000	0.000	0.281
Green	28,768	0.055	0.228	0.000	0.000	1.000
CEO_Age	28,768	50.159	6.356	50.000	38.000	66.000
CEO_Gender	28,768	0.066	0.248	0.000	0.000	1.000
CEO_Overseas	28,768	0.085	0.279	0.000	0.000	1.000
Duality	28,768	0.218	0.413	0.000	0.000	1.000
Board_Size	28,768	10.408	2.628	10.000	7.000	19.000
Board_Independence	28,768	0.378	0.072	0.364	0.273	0.600
Board_Diligence	28,768	9.668	3.850	9.000	5.000	24.000
Firm_Size	28,768	22.493	1.268	22.320	20.544	26.385
Leverage	28,768	0.471	0.197	0.470	0.140	0.926
ROE	28,768	0.059	0.109	0.062	−0.216	0.377
Growth	28,768	0.149	0.420	0.080	−0.322	2.789
Top5	28,768	0.497	0.148	0.490	0.256	0.870

reports the descriptive statistics. The mean value of Exploratory is 0.026, indicating that, on average, exploratory green innovation accounts for only 2.6% of firms’ total green innovation activities in the sample. This relatively low level of exploratory effort underscores the importance of investigating the factors that drive such innovation. For the independent variable, Green, the mean value suggests that, on average, 5.5% of observations have CEOs with green education or work experience, indicating that CEOs with green backgrounds remain relatively rare in the corporate landscape and highlighting the potential significance of such experience in promoting exploratory green innovation.

Table 3. Correlation matrix.

	1	2	3	4	6	7	8	9	10	11	12	13	15	16
1. Exploratory	1
2. Green	0.055 ***	1
3. CEO_ Age	0.028 ***	−0.002	1
4. CEO_ Gender	−0.024 ***	−0.024 ***	−0.034 ***	1
6. CEO_ Overseas	−0.005	−0.012 **	−0.029 ***	0.034 ***	1
7. Duality	−0.016 ***	−0.029 ***	0.195 ***	−0.002	0.077 ***	1
8. Board_ Size	0.054 ***	0.068 ***	0.014 **	−0.041 ***	−0.059 ***	−0.149 ***	1
9. Board_ Independence	−0.011 *	−0.018 ***	0.028 ***	0.042 ***	0.044 ***	0.105 ***	−0.141 ***	1
10. Board_ Diligence	0.054 ***	0.036 ***	−0.050 ***	−0.006	0.008	−0.011 *	0.093 ***	−0.003	1
11. Firm_Size	0.170 ***	0.072 ***	0.122 ***	−0.034 ***	0.003	−0.111 ***	0.220 ***	−0.021 ***	0.265 ***	1
12. Leverage	0.055 ***	0.028 ***	−0.018 ***	−0.023 ***	−0.056 ***	−0.090 ***	0.122 ***	−0.041 ***	0.237 ***	0.396 ***	1
13. ROE	0.048 ***	−0.010	0.041 ***	−0.003	−0.015 **	−0.025 ***	−0.026 ***	0.001	−0.001	0.179 ***	−0.148 ***	1
15. Growth	0.053 ***	−0.011 *	−0.030 ***	−0.006	−0.003	−0.001	0.011 *	−0.020 ***	0.100 ***	0.052 ***	0.039 ***	0.256 ***	1
16. Top5	0.058 ***	0.029 ***	0.051 ***	0.002	−0.022 ***	−0.089 ***	0.087 ***	−0.017 ***	0.000	0.323 ***	0.058 ***	0.200 ***	0.079 ***	1

Note: *, **, ***: Significance at the 10%, 5%, and 1% levels, respectively.

lists the correlation matrix. The Pearson correlation coefficient between Exploratory and Green is significantly positive at the 1% level, thereby providing preliminary empirical evidence of a positive association between CEO green experience and firm exploratory green innovation. Furthermore, the correlation coefficients of all core variables fall within acceptable ranges, with the mean of the variance inflation factor (VIF) being at 1.14 and the maximum at 1.54, suggesting that multicollinearity is not a concern in this study.

4.2. Primary Results and Tests of Hypotheses

We employ a fractional response model to test our hypotheses.

Table 4. Primary regression results.

	Total (H1)	Ownership Structure (H2)		Environmentally Sensitive Industries (H3)		Regional Development (H4)
		SOEs	Non-SOEs	Sensitive Industries	Non-Sensitive Industries	More Developed Region	Less Developed Region
	Model 1	Model 2	Model 3	Model 4	Model 5	Model 6	Model 7
Green	0.066 ***	0.041	0.090 ***	0.079	0.065 **	0.073 *	0.064 **
	(2.685)	(1.224)	(2.653)	(1.129)	(2.475)	(1.880)	(2.029)
CEO_Age	−0.000	−0.001	−0.000	−0.000	−0.000	0.003	−0.002
	(−0.136)	(−0.410)	(−0.227)	(−0.055)	(−0.043)	(1.641)	(−1.497)
CEO_Gender	−0.017	−0.039	−0.001	−0.012	−0.017	−0.030	−0.010
	(−0.665)	(−0.819)	(−0.027)	(−0.149)	(−0.626)	(−0.700)	(−0.320)
CEO_Overseas	−0.027	−0.034	−0.016	0.045	−0.032	−0.016	−0.029
	(−1.225)	(−0.724)	(−0.678)	(0.739)	(−1.390)	(−0.512)	(−0.940)
Duality	0.000	−0.023	0.023	−0.000	0.000	−0.002	0.000
	(0.006)	(−0.732)	(1.277)	(−0.003)	(0.004)	(−0.075)	(0.019)
Board_Size	0.005 **	0.003	0.004	0.003	0.006 **	0.008 **	0.004
	(2.171)	(1.084)	(1.062)	(0.375)	(2.218)	(2.114)	(1.211)
Board_Independence	−0.019	0.002	−0.014	−0.138	−0.008	0.154	−0.142
	(−0.224)	(0.015)	(−0.126)	(−0.557)	(−0.086)	(1.197)	(−1.302)
Board_Diligence	0.004 **	0.005 *	0.003	0.007	0.003 *	0.006 **	0.002
	(2.283)	(1.900)	(1.460)	(1.411)	(1.925)	(2.376)	(0.837)
Firm_Size	0.108 ***	0.103 ***	0.109 ***	0.147 ***	0.103 ***	0.093 ***	0.119 ***
	(15.745)	(10.859)	(10.795)	(7.691)	(14.079)	(8.349)	(13.545)
Leverage	−0.039	−0.016	−0.068	−0.108	−0.033	−0.004	−0.059
	(−0.952)	(−0.250)	(−1.272)	(−0.900)	(−0.763)	(−0.058)	(−1.116)
ROE	0.166 ***	0.332 ***	0.055	0.166	0.167 **	0.166 *	0.167 **
	(2.646)	(3.366)	(0.671)	(0.972)	(2.493)	(1.700)	(2.048)
Growth	0.069 ***	0.095 ***	0.045 **	0.071 *	0.070 ***	0.058 ***	0.077 ***
	(5.352)	(5.196)	(2.469)	(1.942)	(4.986)	(2.728)	(4.768)
Top5	−0.011	0.001	−0.070	−0.079	−0.009	0.038	−0.057
	(−0.249)	(0.007)	(−1.105)	(−0.562)	(−0.182)	(0.542)	(−0.931)
Constant	−4.696 ***	−4.496 ***	−7.642 ***	−8.348 ***	−4.484 ***	−7.704 ***	−4.530 ***
	(−13.926)	(−11.914)	(−28.658)	(−17.922)	(−12.754)	(−18.159)	(−10.926)
Year-FE	YES	YES	YES	YES	YES	YES	YES
Industry-FE	YES	YES	YES	YES	YES	YES	YES
Observations	28,768	13,238	15,530	3613	25,155	11,917	16,851
Pseudo R-squared	0.040	0.047	0.037	0.072	0.036	0.040	0.043

Note: *, **, ***: Significance at the 10%, 5%, and 1% levels, respectively. Robust standard errors clustered at the firm level.

presents the regression results. All models are estimated based on Equation (1), with Exploratory as the dependent variable and Green as the key independent variable. H1 predicts a positive (negative) relationship between CEO green experience and firm green exploratory (exploitative) innovation. Model 1 is used to test this hypothesis. The regression results support the prediction, as the coefficient of Green is positive and statistically significant (β = 0.066, p < 0.01). This effect is economically significant, indicating that CEOs with green experience are associated with a 6.6 percentage point increase in exploratory green innovation. The sample mean of Exploratory is only 0.026, representing a substantial relative increase.

H2 proposes that the positive (negative) association is more pronounced in non-SOEs. We test this by splitting the sample and estimating separate models for SOEs and non-SOEs in Models 2 and 3, respectively. The results indicate a pronounced association between CEO green experience and exploratory green innovation in non-SOEs (β = 0.090, p < 0.01) than in SOEs (β = 0.041, p > 0.10). H3 posits that the positive (negative) association is more pronounced for firms in environmentally insensitive industries. This is supported by the comparison of results from Models 4 and 5, which show that CEO green experience has a significant effect in non-sensitive industries (β = 0.065, p < 0.05), but not in sensitive industries (β = 0.079, p > 0.10). H4 proposes that the positive (negative) relationship is more pronounced for firms in more developed regions. This argument is supported by tests comparing the effects of CEO green experience in more developed regions in Model 6 (β = 0.073, p < 0.10) to less developed regions in Model 7 (β = 0.064, p < 0.05).

4.3. Endogeneity Tests

4.3.1. Reverse Causality Test

Reverse causality may bias our regression estimates because the level of green exploratory innovation could influence the decision to appoint a CEO with green-related experience. Specifically, firms with poor green exploratory innovation may seek green-oriented leadership as a corrective strategy, whereas firms already performing well may aim to sustain or strengthen their green exploratory innovation efforts by appointing similarly aligned executives. Following the methodological framework of Harrison et al. (2025) [35], we identify all CEO appointment events within our sample period, consider a firm’s prior green exploratory innovation as the key predictor, and conduct logistic regression analyses to predict the appointment of a CEO with green experience (Green_Appointment). As

Table 5. Endogeneity test results.

	Reverse Causality Test			Two-Stage Heckman	Estimation Balancing	System GMM
	Green_ Appointment	Green_ Appointment	Green_ Appointment	Exploratory	Exploratory	Exploratory
	Model 1	Model 2	Model 3	Model 4	Model 5	Model 6
Green				0.063 **	0.056 **	0.011 *
				(2.543)	(2.289)	(1.900)
L1.Exploratory	0.710					1.015 ***
	(0.670)					(8.866)
L2.Exploratory		0.987
		(0.867)
L3.Exploratory			1.324
			(1.198)
IMR				−0.158
				(−0.940)
CEO_Age	−0.003	−0.008	−0.008	0.001	−0.001	−0.000
	(−0.293)	(−0.703)	(−0.660)	(0.434)	(−0.450)	(−1.222)
CEO_Gender	−0.115	0.026	0.029	0.009	−0.041	0.003 *
	(−0.415)	(0.095)	(0.087)	(0.195)	(−0.733)	(1.775)
CEO_Overseas	0.417 *	0.162	0.013	−0.009	−0.052	−0.000
	(1.698)	(0.639)	(0.046)	(−0.350)	(−1.110)	(−0.196)
Duality	−0.639 ***	−0.586 **	−0.640 **	0.014	0.013	0.000
	(−2.664)	(−2.265)	(−2.235)	(0.664)	(0.427)	(0.223)
Board_Size	0.080 ***	0.067 ***	0.074 ***	0.001	0.004	0.000
	(3.084)	(2.679)	(2.680)	(0.138)	(0.921)	(1.325)
Board_Independence	−0.226	0.216	−0.222	0.027	0.230	−0.004
	(−0.225)	(0.222)	(−0.208)	(0.285)	(1.498)	(−0.500)
Board_Diligence	0.046 **	0.059 ***	0.063 ***	0.003	0.006 *	−0.000
	(2.296)	(2.959)	(2.923)	(1.402)	(1.949)	(−0.985)
Firm_Size	0.111	0.094	0.134	0.100 ***	0.085 ***	−0.001
	(1.274)	(1.102)	(1.482)	(9.245)	(6.454)	(−1.379)
Leverage	−0.679	−0.839	−0.679	−0.051	−0.122	0.003
	(−1.267)	(−1.621)	(−1.280)	(−1.209)	(−1.609)	(1.536)
ROE	−1.426 *	−1.597 **	−0.900	0.174 ***	0.182	0.014 **
	(−1.799)	(−2.070)	(−1.140)	(2.613)	(1.406)	(2.396)
Growth	−0.160	−0.153	−0.201	0.079 ***	0.139 ***	0.005 ***
	(−1.378)	(−1.274)	(−1.564)	(4.829)	(4.836)	(2.834)
Top5	1.490 ***	1.567 ***	1.361 **	−0.037	−0.221 *	0.003
	(2.642)	(2.710)	(2.171)	(−0.667)	(−1.960)	(1.211)
GRI_Sustain					0.053
					(1.174)
Institutional_Investors					0.143
					(1.566)
Big_Four					−0.107 **
					(−2.489)
Constant	−7.280 ***	−6.933 ***	−3.753 *	−4.348 ***	−4.458 ***	0.027
	(−3.356)	(−3.337)	(−1.838)	(−6.979)	(−10.954)	(1.451)
Year-FE	YES	YES	YES	YES	YES	YES
Industry-FE	YES	YES	YES	YES	YES	YES
Observations	3788	3475	3127	28,768	28,768	24,009
Pseudo R-squared	0.158	0.157	0.169	0.040	0.033	-
AR(1) p	-	-	-	-	-	0.000
AR(2) p	-	-	-	-	-	0.255
Hansen p	-	-	-	-	-	0.171

Note: *, **, ***: Significance at the 10%, 5%, and 1% levels, respectively.

shows, the coefficients for lagged green exploratory innovation (one-year lag in Model 1, two-year lag in Model 2, and three-year lag in Model 3) are not statistically significant. This suggests that our results are not merely an artifact of reverse causality, whereby firms appoint green CEOs to improve or sustain green exploratory innovation.

4.3.2. Heckman Two-Stage Regression

Our findings may be subject to potential bias arising from sample self-selection. Firms with certain characteristics may systematically be more inclined to appoint CEOs with green experience, which may lead to greater engagement in green exploratory innovation. To mitigate this issue, we follow existing empirical studies on CEO characteristics and corporate innovation [19,78] and adopt the Heckman two-stage method. In the first stage, we use a probit regression model to estimate the likelihood of firms having CEOs with green experience, along with the CEO characteristics, board structure, and firm-level control variables included in the main analysis. As an instrumental variable, we construct a measure of environmental regulation based on the frequency of environment-related terms identified through text analysis of provincial government work reports [79]. The second-stage analysis includes the Inverse Mills Ratio (IMR) derived from the first-stage probit regression to account for potential self-selection bias. As shown in Model 4 of Table 5, the coefficient of Green remains positive and statistically significant (β = 0.063, p < 0.05), providing robust evidence for our core findings after accounting for potential self-selection bias.

4.3.3. Entropy Balancing Estimation

An alternative interpretation of our findings is that a firm’s green exploitative innovation is primarily driven by other CEO characteristics, board structure, or firm-level attributes, rather than CEO green experience. To address this concern, we re-estimate our model via matching. Specifically, we apply entropy balancing [80] to reweight control observations by imposing moment conditions on the covariates, including the mean, variance, and skewness (first, second, and third moments). We further include GRI_Sustain, Institutional_Investors, and Big_Four as additional covariates to better account for concerns related to board preferences and governance structure, alongside the baseline control variables. Specifically, GRI_Sustain indicates whether the firm discloses sustainability information in accordance with the Global Reporting Initiative (GRI) Sustainability Reporting Guidelines. Institutional_Investors measures the proportion of institutional ownership, and Big_Four is a dummy variable indicating whether the firm is audited by one of the Big Four accounting firms. After the reweighting procedure, we conduct weighted regressions to ensure that the treatment and control groups are closely comparable in terms of covariate distributions. As reported in Model 5 of Table 5, the coefficient of Green remains positive and statistically significant (β = 0.056, p < 0.05), providing continued support for our conclusion.

4.3.4. System GMM Regression

To further address potential endogeneity issues, we employ the system generalized method of moments (system GMM) for robustness testing [81,82]. As reported in Model 6 of Table 5, the AR (1) test (p = 0.000) indicates first-order autocorrelation, whereas the AR (2) test (p = 0.255) shows no evidence of second-order autocorrelation. The Hansen test (p = 0.171) further confirms the validity of the instruments, suggesting that the model specification is appropriate. The coefficient of Green remains positive and statistically significant (β = 0.011, p < 0.10). Taken together, the baseline regression results are not dependent on a specific econometric model approach, and the main conclusions of this study remain robust and reliable even after addressing endogeneity concerns.

4.4. Robustness Tests

4.4.1. Replace Core Variables

To address potential measurement sensitivity, we conduct robustness tests by employing alternative operationalizations of the core variable. First, we replace the measurement of the dependent variable. As invention patents represent a more precise measure of technological capability than other patent types [83], we reconstruct the variable for green exploratory innovation (Exploratory_Invent) using green invention patents filed by firms, following the same construction method as the original measure based on all green patents. We re-estimate Equation (1) using a fractional response model that specifies Exploratory_Invent as the dependent variable. The result, presented in Model 1 of

Table 6. Robustness test results of replaced core variables.

	Dependent Variable	Independent Variable			Industry Environmental Sensitivity		Regional Economic Development
	Exploratory_ Invent	Exploratory	Exploratory	Exploratory	Exploratory	Exploratory	Exploratory	Exploratory
	Model 1	Model 2	Model 3	Model 4	Model 5	Model 6	Model 7	Model 8
Green	0.081 ***				0.246	0.064 ***	0.080 **	0.045
	(2.607)				(1.543)	(2.587)	(2.407)	(1.259)
Green_Education		0.190 **
		(2.511)
Green_Work			0.070 ***
			(2.746)
Green_Dual				0.327 ***
				(3.093)
CEO_Age	0.000	−0.000	−0.000	−0.000	−0.006	−0.000	−0.001	0.001
	(0.238)	(−0.139)	(−0.143)	(−0.142)	(−0.765)	(−0.151)	(−0.796)	(0.509)
CEO_Gender	−0.047	−0.019	−0.017	−0.018	0.434 **	−0.020	0.005	−0.058
	(−1.414)	(−0.716)	(−0.660)	(−0.688)	(2.542)	(−0.760)	(0.139)	(−1.396)
CEO_Overseas	−0.039	−0.027	−0.027	−0.028	−0.292	−0.023	−0.053 *	−0.004
	(−1.415)	(−1.248)	(−1.231)	(−1.278)	(−1.421)	(−1.081)	(−1.723)	(−0.145)
Duality	0.002	−0.001	0.000	−0.001	−0.093	0.002	−0.024	0.028
	(0.086)	(−0.063)	(0.006)	(−0.095)	(−0.640)	(0.106)	(−1.136)	(1.210)
Board_Size	0.007 **	0.005 **	0.005 **	0.005 **	0.002	0.004 *	0.006 *	0.004
	(2.271)	(2.268)	(2.175)	(2.280)	(0.149)	(1.763)	(1.783)	(1.175)
Board_Independence	0.050	−0.022	−0.018	−0.020	−1.737 ***	0.008	0.085	−0.160
	(0.456)	(−0.270)	(−0.213)	(−0.242)	(−2.813)	(0.092)	(0.772)	(−1.280)
Board_Diligence	0.004 *	0.004 **	0.004 **	0.004 **	0.005	0.004 **	0.003	0.003
	(1.938)	(2.228)	(2.302)	(2.245)	(0.431)	(2.278)	(1.511)	(1.353)
Firm_Size	0.118 ***	0.109 ***	0.108 ***	0.109 ***	0.007	0.111 ***	0.128 ***	0.087 ***
	(14.015)	(15.832)	(15.740)	(15.839)	(0.098)	(15.949)	(13.494)	(8.776)
Leverage	−0.043	−0.043	−0.039	−0.041	−0.502	−0.044	−0.106 *	0.023
	(−0.840)	(−1.036)	(−0.936)	(−0.995)	(−1.292)	(−1.067)	(−1.933)	(0.360)
ROE	0.215 ***	0.161 **	0.167 ***	0.161 **	−0.018	0.169 ***	0.104	0.227 **
	(2.665)	(2.566)	(2.656)	(2.575)	(−0.043)	(2.687)	(1.264)	(2.332)
Growth	0.106 ***	0.069 ***	0.069 ***	0.069 ***	0.079	0.071 ***	0.068 ***	0.074 ***
	(5.476)	(5.347)	(5.347)	(5.340)	(1.142)	(5.415)	(4.059)	(3.575)
Top5	0.031	−0.009	−0.012	−0.011	0.690 *	−0.030	−0.024	−0.005
	(0.550)	(−0.198)	(−0.262)	(−0.243)	(1.736)	(−0.636)	(−0.400)	(−0.071)
Constant	−5.632 ***	−4.710 ***	−4.696 ***	−4.710 ***	2.999 **	−4.738 ***	−4.922 ***	−4.357 ***
	(−15.004)	(−13.970)	(−13.924)	(−13.969)	(1.990)	(−14.015)	(−11.682)	(−9.950)
Year-FE	YES	YES	YES	YES	YES	YES	YES	YES
Industry-FE	YES	YES	YES	YES	YES	YES	YES	YES
Observations	28,768	28,768	28,768	28,768	393	28,375	16,369	12,399
Pseudo R-squared	0.044	0.040	0.040	0.040	0.089	0.041	0.044	0.039

Note: *, **, ***: Significance at the 10%, 5%, and 1% levels, respectively.

, shows that the coefficient of Green remains positive and statistically significant (β = 0.081, p < 0.01). Second, we replace the measurement of the independent variable. Given the potential overlap among different types of CEO green experience (Green), the variable is recoded into three distinct categories—education (Green_Education), work (Green_Work), and side positions (Green_Dual). We re-estimate Equation (1) separately and report the results in Table 6. The coefficients of Green_Education, Green_Work, and Green_Dual are all positive and statistically significant (β = 0.190, p < 0.05; β = 0.070, p < 0.01; β = 0.327, p < 0.01).

Third, we replace the measurement of the moderators. Considering the rise of the new energy industry, we adopt firms in the new energy industries as an alternative measure of environmentally sensitive industries. However, as no unified list of new energy-related listed companies is available on the official websites of major securities firms or in public databases, this study manually identifies firms whose primary business is related to the new energy industry. The classification follows the definition provided in the Strategic Emerging Industry Classification, which was issued by the National Bureau of Statistics of China in 2018, covering enterprises engaged in areas such as “new energy,” “wind power,” “solar energy,” “photovoltaics,” and “wind turbines.” The estimated results are listed in Table 6. Model 5 is the result of new energy industries, and the coefficient of Green is positive but not statistically significant (β = 0.246, p > 0.10). Model 6 is the result of non-new energy industries, and the coefficient of Green is positive and statistically significant (β = 0.064, p < 0.01). Furthermore, we replace the measurement of regional economic development by using provincial per capita GDP as a proxy for regional economic development. Consistent with our baseline approach, we assign a value of 1 if the provincial per capita GDP is above the sample average in a given year and 0 otherwise. The results are reported in Models 7 and 8 of Table 6. In more developed regions, the coefficient of Green is positive and statistically significant (β = 0.080, p < 0.05). In contrast, in less developed regions, the coefficient remains positive but fails to reach statistical significance (β = 0.045, p > 0.10). These findings are consistent with the earlier estimates, further confirming the robustness of our conclusions.

4.4.2. Alternative Estimation Method

To further assess the robustness of our findings, we replace the fractional response model with a fixed-effects OLS regression. Specifically, instead of using the proportion of green exploratory patents as the dependent variable, we construct new outcome variables by taking the natural logarithm of the number of total, exploratory, and exploitative green patents filed plus one, denoted as Total_Quantity, Exploratory_Quantity, and Exploitative_Quantity, respectively. The results, presented in Models 1 to 3 of

Table 7. Other robustness test results.

	Alternative Estimation Method			Lagged Independent Variable			Crossed Fixed Effects	Excluding Alternative Explanations
	Total_ Quantity	Exploratory_ Quantity	Exploitative_ Quantity	Exploratory	Exploratory	Exploratory	Exploratory	Exploratory	Exploratory
	Model 1	Model 2	Model 3	Model 4	Model 5	Model 6	Model 7	Model 8	Model 9
Green	0.115 ***	0.119 ***	0.109 ***				0.060 ***	0.066 ***	0.066 ***
	(4.827)	(5.159)	(4.970)				(2.581)	(2.685)	(2.694)
L1.Green				0.062 **
				(2.400)
L2.Green					0.043
					(1.617)
L3.Green						0.076 **
						(2.542)
CEO_Age	0.000	0.000	−0.000	−0.001	−0.001	−0.001	−0.001	−0.000	−0.000
	(0.075)	(0.420)	(−0.266)	(−0.608)	(−0.890)	(−1.066)	(−0.492)	(−0.100)	(−0.092)
CEO_Gender	−0.029	−0.016	−0.032 **	−0.014	0.003	−0.000	−0.006	−0.017	−0.018
	(−1.428)	(−0.995)	(−2.106)	(−0.525)	(0.095)	(−0.000)	(−0.235)	(−0.660)	(−0.677)
CEO_Overseas	−0.039 **	−0.042 ***	−0.033 **	−0.020	−0.015	−0.027	−0.034	−0.026	−0.027
	(−2.185)	(−2.837)	(−1.968)	(−0.845)	(−0.619)	(−1.032)	(−1.612)	(−1.218)	(−1.255)
Duality	−0.002	−0.005	−0.005	−0.018	−0.023	−0.047 **	0.003	−0.000	−0.001
	(−0.184)	(−0.453)	(−0.477)	(−1.072)	(−1.254)	(−2.451)	(0.166)	(−0.002)	(−0.083)
Board_Size	0.002	0.003 **	0.002	0.005 *	0.005 **	0.004	0.005 **	0.005 **	0.005 **
	(0.942)	(1.966)	(0.855)	(1.797)	(2.036)	(1.379)	(2.240)	(2.148)	(2.192)
Board_ Independence	0.077	0.032	0.041	0.025	−0.017	−0.001	−0.004	−0.017	−0.021
	(1.190)	(0.565)	(0.762)	(0.273)	(−0.176)	(−0.012)	(−0.050)	(−0.208)	(−0.256)
Board_Diligence	0.003 **	0.003 ***	0.003 **	0.005 **	0.004 **	0.005 **	0.004 **	0.004 **	0.004 **
	(2.347)	(2.635)	(2.526)	(2.568)	(2.031)	(2.434)	(2.483)	(2.280)	(2.177)
Firm_Size	0.207 ***	0.148 ***	0.184 ***	0.107 ***	0.104 ***	0.101 ***	0.105 ***	0.108 ***	0.107 ***
	(22.362)	(27.457)	(25.294)	(14.336)	(13.407)	(12.356)	(15.264)	(15.720)	(15.473)
Leverage	−0.060 *	−0.052 *	−0.042	−0.018	−0.025	−0.004	−0.028	−0.041	−0.037
	(−1.951)	(−1.884)	(−1.480)	(−0.400)	(−0.532)	(−0.079)	(−0.678)	(−0.998)	(−0.908)
ROE	0.067	0.082 **	0.074 *	0.186 ***	0.161 **	0.185 **	0.184 ***	0.164 ***	0.165 ***
	(1.536)	(2.003)	(1.945)	(2.715)	(2.227)	(2.366)	(2.907)	(2.620)	(2.628)
Growth	−0.008	0.038 ***	−0.017 ***	0.063 ***	0.065 ***	0.075 ***	0.078 ***	0.070 ***	0.069 ***
	(−1.162)	(3.992)	(−2.973)	(4.195)	(4.062)	(4.608)	(5.937)	(5.380)	(5.351)
Top5	−0.031	0.026	−0.049	−0.027	0.019	0.011	−0.041	−0.012	−0.010
	(−0.693)	(0.754)	(−1.276)	(−0.534)	(0.355)	(0.200)	(−0.872)	(−0.250)	(−0.222)
Carbon_Policy								0.086 **
								(2.120)
Digital_ Transformation									0.008
									(1.238)
Constant	−1.542 ***	−3.055 ***	−1.012 ***	−4.333 ***	−4.364 ***	−4.650 ***	−7.786	−4.719 ***	−4.678 ***
	(−7.363)	(−25.767)	(−6.346)	(−16.081)	(−14.417)	(−18.074)	(.)	(−13.996)	(−13.858)
Year-FE	YES	YES	YES	YES	YES	YES	YES	YES	YES
Industry-FE	YES	YES	YES	YES	YES	YES	YES	YES	YES
Year-Industry-FE	NO	NO	NO	NO	NO	NO	YES	NO	NO
Observations	28,768	28,768	28,768	24,042	20,813	18,337	28,768	28,768	28,768
Pseudo R-squared	-	-	-	0.040	0.039	0.039	0.040	0.040	0.040
Adjusted R-squared	0.397	0.174	0.400	-	-	-	-	-	-

Note: *, **, ***: Significance at the 10%, 5%, and 1% levels, respectively. “(.)” indicates that the variable is omitted due to absorption by the interaction fixed effects.

, show that Green is positively significant and associated with Total_Quantity (β = 0.115, p < 0.01), Exploratory_Quantity (β = 0.119, p < 0.01), and Exploitative_Quantity (β = 0.109, p < 0.01). This indicates that CEOs with green experience contribute not only to overall green innovation but also to both exploratory and exploitative forms of ambidextrous green innovation. In terms of economic significance, a change in CEO green experience from 0 to 1 is associated with approximately a 12.2% ($\approx {\mathrm{e}}^{0.115}$ − 1) increase in total green patents, a 12.6% ($\approx {\mathrm{e}}^{0.119}$ − 1) increase in green exploratory patents, and an 11.5% ($\approx {\mathrm{e}}^{0.109}$ − 1) increase in green exploitative patents, with no need for standardization, given the binary nature of the independent variable. The effect of CEO green experience on green exploratory innovation is the strongest for exploratory green innovation, followed by total green innovation, and the weakest for exploitative green innovation. The findings indicate that the increase in exploratory innovation exceeds the average increase in green innovation, aligning with our baseline findings that CEOs with green experience are more inclined toward green exploratory innovation.

4.4.3. Lagged Treatment

Recognizing the potential temporal lag between CEO green experience (Green) and green exploratory innovation (Exploratory), we incorporate dynamic specifications into the fractional response model to capture the delayed effects. We conduct a distributed lag analysis by incorporating one-, two-, and three-period lagged terms of Green (denoted as L1.Green, L2.Green, and L3.Green, respectively) into our regression models. The results are reported in Models 4–6 of Table 7. Model 4 reports that L.Green has a positive and statistically significant effect on green exploratory innovation (β = 0.062, p < 0.05). Model 5 uses L2.Green, which remains positive but not statistically significant (β = 0.043, p > 0.10), possibly indicating that the effect weakens or takes longer than two periods to become pronounced. Model 6 employs L3.Green, yielding a positive and statistically significant coefficient (β = 0.076, p < 0.05). These findings suggest that the influence of CEO green experience may take time to materialize and can persist over multiple periods, providing robust evidence for our conclusion.

4.4.4. Crossed Fixed Effects

To further account for the unobserved macroeconomic and sectoral trends that may affect firms’ green innovation outcomes, we include crossed fixed effects by interacting year and industry dummies. Model 7 in Table 7 reports that the coefficient of Green continues to be positive and statistically significant (β = 0.060, p < 0.01), reinforcing the robustness of the main conclusions after controlling for year–industry fixed effects.

4.4.5. Excluding Alternative Explanations

Considering the emerging contexts that may influence green innovation, we incorporate the potential confounding effects of digital transformation [84] and the Dual-Carbon Policy [41] into the analysis. We employ the logarithm of the frequency of digital-related keywords in firms’ annual reports as a proxy for the degree of corporate digital transformation (Digital_Transformation) [85]. A difference-in-differences (DID) variable, Carbon_Policy, is constructed to capture the effect of the Dual-Carbon Policy. In September 2020, the Chinese government officially announced its “dual-carbon” goals—peaking carbon emissions by 2030 and achieving carbon neutrality by 2060. Specifically, Carbon_Policy takes the value of 1 if the firm operates in an environmentally sensitive industry and the year is later than 2020, and 0 otherwise [41]. We separately include these two variables in the regression models, corresponding to Models 8 and 9 in Table 7. The estimated coefficients of Green remain positive and statistically significant in both cases (β = 0.066, p < 0.01; β = 0.066, p < 0.01), indicating that the main findings are robust after considering the potential confounding effects of digital transformation and the Dual-Carbon Policy.

4.5. Supplemental Analyses

4.5.1. Mechanism Test

To further examine the mechanisms, we test whether CEOs with green experience shape the firm’s green strategic orientation and facilitate resource acquisition for green innovation. Green strategic orientation (Green_Cognition) is measured using a text analysis approach. Specifically, we extract keywords representing three dimensions: external environmental pressure, green competitive advantage, and awareness of social responsibility. The frequency of these keywords in the listed companies’ annual reports in relevant corporate documents is then calculated to serve as a proxy [86]. As for resource acquisition, we construct a proxy variable defined as the ratio of green subsidies to total assets. Specifically, Green_Subsidies are identified from firms’ annual reports by screening subsidy items containing the keywords “green,” “environmental protection,” “clean,” “environmental treatment,” and “comprehensive management” [87]. The monetary value of these subsidies is then scaled by total assets to serve as the proxy.

We build on the traditional causal steps approach proposed by Baron and Kenny (1986) [88] and further employ the bootstrap method [89] to test the significance of mediation effects. Equations (2) and (3) correspond to the first stage, whereas Equations (4) and (5) pertain to the second stage. The empirical models are presented as follows:

$${Green}_{i,t}={\beta }_0+{\beta }_1{Green\_Cogniton}_{i,t}+{\beta }_2{Controls}_{i,t}+{\mu }_i+{\omega }_t+{\epsilon }_{i,t}$$

(2)

$${Green}_{i,t}={\beta }_0+{\beta }_1{Green\_Subsidies}_{i,t}+{\beta }_2{Controls}_{i,t}+{\mu }_i+{\omega }_t+{\epsilon }_{i,t}$$

(3)

$${Exploratory}_{i,t}={\beta }_0+{\beta }_1{Green}_{i,t}+{\beta }_2{Green\_Cogniton}_{i,t}+{\beta }_3{Controls}_{i,t}+{\mu }_i+{\omega }_t+{\epsilon }_{i,t}$$

(4)

$${Exploratory}_{i,t}={\beta }_0+{\beta }_1{Green}_{i,t}+{\beta }_2{Green\_\mathit{Subsidies}}_{i,t}+{\beta }_3{Controls}_{i,t}+{\mu }_i+{\omega }_t+{\epsilon }_{i,t}$$

(5)

Table 8. Two-stage test results.

	First Stage		Second Stage
	Green_ Cognition	Green_ Subsidies	Exploratory	Exploratory
	Model 1	Model 2	Model 3	Model 4
Green	1.095 ***	0.061 ***	0.056 **	0.065 ***
	(5.591)	(3.201)	(2.249)	(2.629)
Green_Cognition			0.009 ***
			(5.120)
Green_Subsidies				1.287 ***
				(3.632)
CEO_Age	0.013 **	0.000	−0.000	−0.000
	(2.263)	(0.597)	(−0.276)	(−0.181)
CEO_Gender	−0.202	0.014	−0.016	−0.018
	(−1.470)	(0.644)	(−0.606)	(−0.698)
CEO_Overseas	−0.103	−0.071 ***	−0.026	−0.026
	(−0.812)	(−3.397)	(−1.197)	(−1.190)
Duality	−0.347 ***	−0.053 ***	0.004	0.002
	(−4.168)	(−3.839)	(0.235)	(0.107)
Board_Size	0.006	0.005 ***	0.005 **	0.005 **
	(0.489)	(2.671)	(2.153)	(2.107)
Board_Independence	−0.344	0.037	−0.016	−0.019
	(−0.853)	(0.503)	(−0.196)	(−0.229)
Board_Diligence	−0.006	0.003 *	0.004 **	0.004 **
	(−0.712)	(1.869)	(2.302)	(2.263)
Firm_Size	0.465 ***	−0.015 ***	0.104 ***	0.109 ***
	(10.587)	(−2.854)	(15.082)	(15.870)
Leverage	0.064	0.187 ***	−0.041	−0.044
	(0.276)	(5.972)	(−0.995)	(−1.061)
ROE	0.598 **	−0.021	0.159 **	0.167 ***
	(2.022)	(−0.419)	(2.537)	(2.669)
Growth	−0.071	−0.054 ***	0.070 ***	0.071 ***
	(−1.596)	(−3.748)	(5.419)	(5.463)
Top5	0.303	−0.048	−0.014	−0.010
	(1.054)	(2.589)	(−0.303)	(−0.176)
Constant	−11.414 ***	−5.238 ***	−4.595 ***	−4.699 ***
	(−10.561)	(−38.107)	(−13.524)	(−13.904)
Year-FE	YES	YES	YES	YES
Industry-FE	YES	YES	YES	YES
Observations	28,768	28,768	28,768	28,768
Pseudo R-squared	-	0.061	0.041	0.400
Adjusted R-squared	0.301	-	-	-

Note: *, **, ***: Significance at the 10%, 5%, and 1% levels, respectively.

lists the results of the two-stage test. In the first stage, the coefficients of Green of Model 1 and Model 2 are both positive and statistically significant (β = 1.095, p < 0.01; β = 0.061, p < 0.01), indicating that CEO green experience can promote the firm’s green strategic orientation and help the firm facilitate resource acquisition for green innovation. In the second stage, the coefficients of Green_Cognition and Green_Subsidies are also both positive and statistically significant (β = 0.009, p < 0.01; β = 1.287, p < 0.01), and the absolute value of the coefficients of Green is lower than 0.066 in Model 1 of Table 4, representing that CEO green experience promotes exploratory green innovation through promoting green strategic orientation and facilitating resource acquisition for green innovation. The bootstrap test results are reported in

Table 9. Bootstrap test results.

Dependent Variables	Mediator Variable	Effect Type	Coefficient	Std. Error	p Value	95% Confidence Interval
Exploratory	Green_Cognition	Indirect	0.003109	0.000296	0.000	[0.002530, 0.003696]
		Direct	0.006848	0.001608	0.000	[0.003696, 0.010001]
	Green_Subsidies	Indirect	0.000633	0.000123	0.000	[0.000393, 0.000874]
		Direct	0.009324	0.001529	0.000	[0.006328, 0.012320]

. The indirect effect of Green_Cognition on Exploratory is 0.003109 (p < 0.01), with a 95% confidence interval of [0.002530, 0.003696], not including zero. The direct effect of that is 0.006848 (p < 0.01), suggesting that Green_Cognition has a partial mediation effect. Similarly, the indirect effect of Green_Subsidies on Exploratory is 0.000633 (p < 0.01), with a 95% confidence interval of [0.000393, 0.000874], excluding zero. The direct effect of that is 0.009324 (p < 0.01), confirming that the partial mediation effect of Green_Subsidies is supported.

4.5.2. Heterogeneity Test

To gain deeper insights into the heterogeneous effects of CEO green experience on firms’ ambidextrous green innovation, we conduct subgroup analyses across CEO age, gender, and tenure. CEO age and tenure are dichotomized relative to the annual sample mean, distinguishing between older vs. younger and longer-tenured vs. shorter-tenured CEOs. The heterogeneity test results are listed in

Table 10. Heterogeneity test results.

	CEO Age		CEO Gender		CEO Tenure
	Older	Younger	Female	Male	Long-Tenured	Short-Tenured
	Model 1	Model 2	Model 3	Model 4	Model 5	Model 6
Green	0.052	0.076 *	0.023	0.069 ***	0.051	0.076 **
	(1.540)	(1.907)	(0.213)	(2.744)	(1.242)	(2.572)
CEO_Age			−0.002	0.000	−0.001	−0.000
			(−0.410)	(0.146)	(−0.354)	(−0.052)
CEO_Gender	−0.040	0.015			−0.002	−0.036
	(−1.149)	(0.407)			(−0.054)	(−1.101)
CEO_Overseas	−0.021	−0.038	0.109	−0.038 *	−0.033	−0.029
	(−0.707)	(−1.268)	(1.610)	(−1.677)	(−1.050)	(−1.021)
Duality	−0.000	0.011	−0.055	0.001	0.037 *	−0.048 **
	(−0.022)	(0.434)	(−0.970)	(0.038)	(1.748)	(−2.143)
Board_Size	0.006 *	0.005	0.003	0.005 **	0.008 *	0.004
	(1.853)	(1.469)	(0.323)	(2.110)	(1.956)	(1.305)
Board_Independence	−0.032	−0.018	0.353	−0.039	0.014	−0.070
	(−0.285)	(−0.157)	(1.202)	(−0.446)	(0.113)	(−0.662)
Board_Diligence	0.004 *	0.003	0.008	0.004 **	0.004	0.004 *
	(1.780)	(1.385)	(1.339)	(2.109)	(1.403)	(1.953)
Firm_Size	0.100 ***	0.121 ***	0.102 ***	0.110 ***	0.127 ***	0.096 ***
	(11.325)	(12.397)	(4.762)	(15.454)	(12.084)	(11.280)
Leverage	−0.058	−0.021	0.039	−0.048	−0.142 **	0.017
	(−1.071)	(−0.357)	(0.270)	(−1.137)	(−2.211)	(0.331)
ROE	0.044	0.284 ***	−0.124	0.182 ***	0.129	0.173 **
	(0.499)	(3.339)	(−0.551)	(2.803)	(1.229)	(2.290)
Growth	0.077 ***	0.060 ***	−0.115	0.076 ***	0.037	0.081 ***
	(4.273)	(3.218)	(−1.633)	(5.797)	(1.463)	(5.297)
Top5	0.020	−0.052	−0.372 **	−0.002	−0.107	0.054
	(0.329)	(−0.765)	(−2.174)	(−0.048)	(−1.489)	(0.913)
Constant	−4.255 ***	−8.212 ***	−2.658 ***	−5.227 ***	−5.203 ***	−4.362 ***
	(−11.245)	(−32.141)	(−5.276)	(−16.924)	(−13.061)	(−9.442)
Year-FE	YES	YES	YES	YES	YES	YES
Industry-FE	YES	YES	YES	YES	YES	YES
Observations	14841	13927	1893	26875	11702	17066
Pseudo R-squared	0.042	0.041	0.076	0.040	0.042	0.042

Note: *, **, ***: Significance at the 10%, 5%, and 1% levels, respectively.

. Model 1 and Model 2 indicate a pronounced association between CEO green experience and exploratory green innovation for younger CEOs (β = 0.076, p < 0.10) than for older CEOs (β = 0.052, p > 0.10). This result may be attributed to the fact that younger CEOs typically possess a stronger risk-taking propensity and a greater willingness to pursue innovative activities [90]. Model 3 and Model 4 suggest a pronounced association between CEO green experience and exploratory green innovation for male CEOs (β = 0.069, p < 0.01) than for female CEOs (β = 0.023, p > 0.10). This finding could be explained by the tendency of female CEOs to exhibit greater risk aversion and lower confidence in strategic decision-making compared to their male counterparts [91]. Model 5 and Model 6 demonstrate a pronounced association between CEO green experience and exploratory green innovation for short-tenured CEOs (β = 0.076, p < 0.05) than for long-tenured CEOs (β = 0.051, p > 0.10). One reasonable reason is that long-tenured CEOs are more likely to rely on established innovation routines, thereby perpetuating existing inertia rather than initiating new exploratory efforts [92].

5. Conclusions and Discussion

5.1. Discussion of Key Findings

This study investigates the nexus between CEO green experience and firm ambidextrous green innovation by considering the moderating roles of firm ownership, industry environmental sensitivity, and regional economic development. With China’s commitment to achieving “dual-carbon goals” and the continued implementation of comprehensive environmental policies, corporate green innovation has entered a vigorous stage of development. Therefore, we use the panel data from Chinese non-financial A-share listed companies between 2007 and 2023 as the research sample.

The empirical results indicate that CEO green experience is positively associated with exploratory, rather than exploitative, green innovation. Drawing on their sustainability cognition and relevant expertise, CEOs with green experience can shape firms’ strategic orientation and broaden access to resources for green innovation. As a result, firms not only demonstrate the willingness but also possess the capability to engage in exploratory green innovation. The mechanism test provides further evidence that a firm’s green strategic orientation and resource acquisition for green innovation are the key mechanisms through which CEO green experience promotes exploratory green innovation.

The positive association is particularly evident in non-SOEs, firms operating in less environmentally sensitive industries, and those based in more developed regions. First, the pronounced institutional pressure, distinctive governance structures, and limited managerial discretion of SOEs tend to dampen the influence of CEO green experience and lead firms’ green innovation activities to become more risk-averse. Second, strong regulatory requirements and pressures for social legitimacy in environmentally sensitive industries compel CEOs to prioritize exploitative green innovation to achieve quick and tangible outcomes. Third, abundant resource endowments and mature intellectual property protection systems in more developed regions amplify the influence of CEO green experience and offer greater assurance for firm exploratory green innovation activities. Furthermore, the heterogeneity test reveals that the association between CEO green experience and exploratory green innovation varies significantly with CEO age, gender, and tenure. Specifically, younger, male, and short-tenured CEOs tend to exhibit stronger risk-taking propensities, thereby enabling their green experience to exert a greater effect on firms’ pursuit of exploratory green innovation.

After alleviating endogeneity concerns through reverse causality tests, Heckman two-stage regression, entropy balancing estimation, and system GMM regression, our main conclusions remain robust. Further robustness checks—including replacing the core variables, employing alternative estimation methods, introducing lagged treatments, controlling for crossed fixed effects, and excluding alternative explanations—also produce consistent results.

5.2. Theoretical Contributions

Our theory and findings contribute to a more nuanced and comprehensive understanding of how CEO green experience shapes firm green innovation. In highlighting the effect of CEO green experience on green innovation, the existing literature has implicitly suggested that green CEOs have similar positive effects, without differentiating between different types [25,27]. Even when distinctions are drawn, for example, between technological and managerial innovations, their effects are presumed to be consistently positive [93]. We challenge and extend this perspective by introducing the lens of ambidextrous innovation into the green innovation context. Our findings suggest that CEOs with green experience, owing to heightened environmental awareness and accumulated green-related expertise, are more inclined to foster exploratory, rather than exploitative, green innovation.

Our study also contributes to the literature on the antecedents of ambidextrous green innovation. On the one hand, existing research has largely focused on external pressures [38,47] and internal organizational factors [39,94], overlooking the critical influence of CEO characteristics. We address this gap by emphasizing the role of CEOs’ personal experiences in shaping green ambidextrous innovation. On the other hand, most studies on this topic have concentrated on the drivers of ambidexterity [36,95] or the simultaneous pursuit of exploratory and exploitative green innovation [47]. However, the inherent contradictions between exploratory and exploitative innovation often constrain firms from pursuing both simultaneously, thus forcing a strategic trade-off [96]. We advance the literature by demonstrating that CEOs with green experience play a significant role in shaping how firms manage the tradeoff involved in green innovation.

Another intriguing implication concerns how certain contextual factors moderate the extent to which CEO green experience translates into a firm’s propensity toward exploratory green innovation. Scholars acknowledge that the positive relationship between CEO characteristics and green innovation is amplified under stringent environmental regulations or strong legitimacy pressures [15,19]. However, our analysis suggests that when green innovation becomes an urgent priority, firms tend to favor an exploitative rather than an exploratory approach. Although goals of regulatory compliance and legitimacy can be promptly met through the replication and utilization of existing green innovation knowledge and technology [97], such an approach undermines long-term sustainability and erodes firms’ competitive advantage [98]. Beyond regulatory and legitimacy pressures, the presence of mature markets and well-developed institutional environments further enhances the impact of CEO green experience on a firm’s commitment to exploratory green innovation by providing critical resources and intellectual property protection.

5.3. Practical Implications

This study offers valuable practical guidance for advancing green innovation. Senior executives, particularly CEOs, play a pivotal role in driving exploratory green innovation. Firms should attach greater importance to attracting and recruiting green-experienced CEOs who are more likely to guide firms toward technological breakthroughs and sustainable growth. For incumbent CEOs, expanding their green expertise and strengthening managerial capabilities through targeted training and professional development in environmental fields is also a valuable approach.

A flexible and adaptive approach should be adopted to balance exploratory and exploitative green innovation. In mature market environments, firms should actively pursue exploratory green innovation to achieve substantive green transformation. When facing substantial legitimacy pressures, firms can focus on exploitative green innovation to gradually build knowledge and technological capabilities, thereby facilitating an incremental transition toward sustainable development.

Policymakers should prioritize the development of favorable institutional conditions, including effective intellectual property protection, accessible green financing, and targeted policy incentives. Especially in less developed regions, policy priorities should address financial and institutional constraints. This can be achieved by easing access to external funding, fostering supportive knowledge and technology networks, and strengthening the enforcement of intellectual property rights, thereby creating a more enabling environment for green exploratory efforts. Additionally, for SOEs and firms in environmentally sensitive industries, regulatory frameworks should be more nuanced and differentiated to help ease excessive legitimacy pressures. Reducing these pressures can encourage a shift from short-term, compliance-driven innovation toward longer-term strategies focused on exploratory green innovation.

5.4. Limitations and Future Directions

This study has some limitations, which may provide directions for future research. First, we measure CEO green experience using green-related education or work background due to data availability constraints. However, this measure may not fully capture the heterogeneity in the sources and developmental processes of CEO green experience, such as variations in the duration or depth of their exposure. Future research should explore these aspects to provide a more nuanced understanding.

Second, the empirical analysis draws on data from A-share listed companies in China. Although Chinese firms have actively advanced green innovation [59], this context also has distinctive institutional and cultural features. Examining other settings in future research would help assess the extent to which these findings can be generalized. Additionally, this study does not capture the green innovation practices of small and medium-sized enterprises (SMEs). These firms often operate under greater resource constraints and exhibit lower risk tolerance in pursuing green transformation, making the CEO’s role potentially even more critical. Future research could therefore examine SMEs to provide a fuller understanding of how CEO green experience influences ambidextrous green innovation. It is also important to recognize that non-technological forms of green innovation play a vital role in advancing firms’ sustainability objectives.

Third, green patent applications are adopted as a measurement for firms’ green innovation. While this approach is widely accepted in the literature on green innovation, it confines the sample to firms that filed patents during the study period. However, non-technological forms of green innovation are also critical to firms’ sustainability efforts [11]. With advancements in firm-level innovation measurement methods, such as text-based analysis [99], future research could extend to enterprises without patent filings, thereby enabling a more accurate capture of their green innovation activities.

Finally, while our study highlights the positive impact of CEO green experience, future research should also pay attention to its potential “dark side.” An excessive focus on green initiatives may unintentionally crowd out other strategic priorities, lead to overinvestment in uncertain technologies, or create tensions with short-term financial goals. In addition, our analysis is limited to the influence of CEOs on firm green innovation and does not incorporate cross-level mechanisms (e.g., CEO–board interactions). Future research could examine such cross-level dynamics to provide a more comprehensive understanding.