The Effect of Non-Actual Controllers on Corporate Green Governance Performance

Abstract

With sustainable development becoming a global strategic priority, advancing corporate green governance is now pivotal for driving the systemic shift toward a green economic model. This research examines how non-actual controllers shape corporate green governance performance through a systemic lens. The results indicate that non-actual controllers exert a substantial negative impact on the performance of corporate green governance. Mechanism analysis demonstrates that non-actual controllers suppress green governance performance by reducing corporate environmental investment, increasing green agency costs, and weakening corporate social responsibility. Additionally, the inhibitory effect is found to be more pronounced in non-state-owned enterprises, companies with a lower proportion of institutional investors, firms with weaker external audit quality, and those operating under higher levels of environmental uncertainty. The results offer practical pathways for enhancing corporate green governance and deliver informed strategies for performance optimization.

1. Introduction

In contemporary society, economic development and environmental protection jointly constitute a complex, systematic project. The traditional enterprise development model often pursues economic output, neglecting resource consumption and ecological environment damage, resulting in an imbalance between the social economy and the natural system. As foundational entities in the economy, enterprises bear significant responsibility for facilitating and advancing the green transition. Green governance centers on the principle that corporate development must respect ecological boundaries. It is operationalized through strategies like technological innovation and green production to advance socio-economic and environmental sustainability [1]. Not only does green governance enhance the competitiveness of enterprises [2], but it also fosters the harmonious development of the environment and the economy. Thus, enhancing corporate green governance capabilities is a crucial avenue for achieving the integrated development of the economy, society, and environment.

In corporate governance systems, the actual controller holds ultimate control rights and has a decisive influence on shareholder meetings, board decisions, and major corporate policies [3]. In Chinese listed companies, the actual controller’s preferences are typically channeled through controlling shareholders to the strategic level, making these characteristics key to analyzing corporate governance logic and decision motives [3]. However, with completed share structure reforms and a maturing mergers and acquisitions market, listed companies now show increasingly dispersed ownership, leading to a growing number of firms without an actual controller [4]. This structural shift systematically challenges traditional controller-centered governance theories.

From a systemic perspective, an enterprise is a complex governance system formed by the joint action of internal and external factors. As a core element of the system, the existence or absence of the actual controller profoundly influences the power allocation, incentive mechanism, and strategic direction. Although existing studies have focused on various influencing factors of environmental performance, they have generally ignored the key role of the actual controller in the system [4]. Green governance generates significant positive externalities, with social and environmental benefits often exceeding corporate economic returns [5]. This leads to limited corporate motivation for its implementation. This contradiction is even more prominent in a company system where control rights are absent. Weakened accountability mechanisms and short-term orientation further marginalize green governance in such entities. Our study establishes a system-level analytical framework to examine how non-actual controllers affect corporate green governance performance and its underlying mechanisms. It addresses two core questions: How does this structural deficiency influence green governance, and through what pathways? This study seeks to generate novel insights into the governance of non-actual controller firms and to advance the theoretical foundations for corporate sustainability.

To this end, a theoretical framework integrating agency and stakeholder theories was constructed, leading to a set of testable hypotheses. Our data comprises a sample of A-share listed companies, spanning the years 2009 to 2022. Meanwhile, we used the number of enterprise mergers and acquisitions at the city level as the instrumental variable (IV) for non-actual controllers to avoid potential endogeneity problems. Our findings suggest that non-actual controllers impede corporate green governance performance. The underlying mechanism operates through three pathways: decreased environmental investments, increased green agency costs, and attenuated corporate social responsibility. The strength of the negative impact varies significantly across firms, being particularly strong among non-state-owned enterprises, those with low institutional ownership, weak external audit quality, and high environmental uncertainty.

The main contributions of our research are as follows. Firstly, starting from the systematic characteristics of the governance structure, our study for the first time incorporates the enterprise control rights allocation model of “no actual controller” into the analytical framework of green governance, revealing its influence mechanism on the green governance performance of enterprises. Unlike the existing literature, which mostly focuses on local factors, our perspective frames enterprise green governance as a complex, systematic process. In enterprises lacking an actual controller, the dispersion of equity leads to a “collective action predicament”, weakening the ability of effective supervision and strategic coordination, thereby increasing agency costs and forming a unique green governance model. This research expands the systematic understanding of corporate green governance, especially providing a new theoretical explanation for how governance structures without actual controllers affect environmental performance.

Secondly, our study constructs an integrated framework to explain how non-actual controllers inhibit corporate green governance performance at the theoretical level. By combining agency theory and stakeholder theory, we identify three mediating mechanisms: reduced environmental investment, increased green agency costs, and weakened social responsibility commitment. This framework advances the field by shifting the focus from whether dispersed ownership affects green governance to how it does so, while extending agency theory into environmental management. The study ultimately affirms that effective green governance requires a stable, authoritative governance core, repositioning corporate governance from a background condition to a key explanatory variable in green transformation. These insights provide theoretical support for both policy design and corporate practice.

The subsequent arrangements are as follows: Section 2 presents the literature review. Section 3 develops the theories and hypotheses. Section 4 outlines materials and methods. Section 5 presents the results. Section 6 discusses the conclusions.

2. Literature Review

2.1. Literature on Control Right and Non-Actual Controllers

Corporate control refers to the power to substantially influence shareholder decisions or direct corporate operations, stemming from direct or indirect equity investments [6,7]. According to China’s Company Law draft, an actual controller refers to any person or entity capable of effectively directing a company’s actions via investment relationships, agreements, or other means. In accordance with China’s Measures for the Administration of Takeovers of Listed Companies and the SZSE ChiNext Listing Rules, control over a listed company is defined as meeting any of the following conditions: (1) holding over 50% of shares as a controlling shareholder; (2) having de facto control over 30% of voting rights; (3) being able to elect more than half of the directors or significantly influence shareholder resolutions through voting rights; (4) other circumstances determined by the CSRC. The actual controller’s decisions profoundly impact corporate development, governance environment, and market confidence [8]. Evidence suggests that the influence of actual controllers varies by their background: overseas residency has been shown to constrain financing and reduce leverage [3], while conversely, a central government background can significantly moderate the R&D-financial performance relationship [9].

The emergence of companies without actual controllers arises from their ownership structures [4,6], which can be classified into two primary types. First, the dispersed ownership type, where no single shareholder holds controlling stakes and nor is there a shareholder voting agreement, resulting in no entity capable of effectively directing corporate actions [7]. A controlling shareholder typically refers to one holding over 50% of shares or, even with a smaller stake, possessing sufficient voting power to substantially influence shareholder resolutions. Second, while a direct controlling shareholder may exist, examination of its pyramid ownership structure reveals no ultimate individual capable of exercising effective control [10]. In China, securities regulations such as the Measures for the Administration of the Takeover of Listed Companies explicitly require the identification and disclosure of the actual controller, underscoring their pivotal role in corporate governance. Consequently, the absence of an actual controller in this legal context signifies a vacuum in core oversight and accountability, creating a potential void in both supervision and governance.

2.2. Literature on the Economic Effects of Non-Actual Controllers

The dispersion of equity ownership has brought increased attention to firms without actual controllers and their economic impact [11]. According to agency theory, the lack of oversight and shareholder free-riding in such firms aggravates principal-agent problems, systematically affecting corporate decisions and long-term performance. This governance structure’s economic effects appear in two main aspects. First, it promotes managerial short-termism, discouraging long-term investments like innovation. While Aghion (2013) stresses the importance of ownership concentration for innovation [12], Du and Ma (2022) empirically show that non-actual controllers significantly reduce corporate innovation [4]. Second, it undermines control stability and increases governance costs. Shleifer and Vishny (1997) emphasize that ownership concentration ensures monitoring effectiveness and strategic stability [13], whereas control voids lead to control contests and higher transaction costs [4]. Although some argue dispersed ownership might foster innovation [14] and environmental performance [15], most evidence confirms that its negative effects dominate. Beyond hindering innovation through agency conflicts, this governance structure may worsen financing constraints and harm ESG performance [16], indicating its deficiencies extend beyond financial metrics to sustainability concerns.

2.3. Literature on Corporate Green Governance

Corporate green governance and innovation rest on key drivers that depend fundamentally on stable corporate governance. Existing research categorizes these drivers into two dimensions: internal governance mechanisms and the external institutional environment. At the internal governance level, the personal characteristics and strategic cognition of management form the micro-foundation of green governance. Studies indicate that managerial traits [17] and environmental awareness [18] collectively shape the tone of a firm’s environmental strategy, translating sustainability concepts into concrete practices through top-down approaches such as green management innovation [19]. At the external level, multiple stakeholders constitute the institutional and market environment driving corporate green transformation. These include supervisory pressure from the public and media [20], market selection pressure stemming from the preferences of green investors [21], and mandatory policy pressure arising from carbon emission trading mechanisms [22], government environmental regulations [23], and enforcement intensity [24]. Additionally, competitive pressure [25] from rivals also plays a role. Although these external forces serve as important drivers, their influence is not automatically realized. Rather, they must be channeled through effective corporate governance mechanisms to reach the core of corporate decision-making and ultimately be internalized into resource allocation and strategic actions.

2.4. Literature Summary

While existing research has made progress in studying the economic effects of firms without a controlling shareholder and the drivers of green governance, a clear divide remains between these two streams. The former primarily focuses on conventional economic outcomes, while the latter emphasizes factor identification, with limited systematic integration of this specific governance structure into green governance frameworks. Furthermore, mechanistic research remains underdeveloped, particularly lacking a clear theoretical explanation of how non-actual controllers influence corporate responses to internal and external environmental pressures. To address this gap, we develop an integrated framework linking governance structure to green governance performance, systematically revealing the transmission mechanisms through which non-actual controllers affect green governance. This research not only helps address theoretical gaps in the existing literature but also provides policy insights for improving corporate governance and advancing corporate green transformation.

3. Theories and Hypotheses

Based on Jensen and Meckling’s (1976) agency theory, the separation of ownership and control inherently creates conflicts between owners and managers [7]. In firms without actual controllers, dispersed ownership exacerbates this problem, creating a “monitoring vacuum” where managers prioritize short-term results over long-term green governance objectives. Jensen’s (1986) free cash flow theory explains this tendency: unconstrained managers favor investments delivering quick personal benefits over long-term strategic goals [26]. Given that green governance investments typically involve large scales, extended cycles, and high uncertainty [27], they fundamentally conflict with managers’ utility functions prioritizing job security and short-term compensation. Therefore, rather than merely neglecting environmental projects, managers actively reallocate resources from substantive green initiatives to activities yielding faster personal returns, systematically reducing environmental investment. Thus, we hypothesize:

H1: 

Non-actual controllers curb green governance performance by reducing corporate environmental investment.

A more profound impact manifests in the systematic escalation of green agency costs. Green governance investments, characterized by their positive externalities and long-term returns, become a focal point for principal-agent conflicts within corporations. In environmental contexts, these costs materialize as “green agency costs” [28], comprising three core dimensions: monitoring costs incurred by shareholders to oversee management’s environmental strategies, contracting costs for designing incentives toward green innovation, and losses from reputational damage and compliance risks due to underinvestment or greenwashing [29]. In firms without actual controllers, the structural absence of oversight systematically elevates these green agency costs. Dispersed shareholders lack the incentive to conduct thorough monitoring of management’s complex environmental decisions, creating opportunities for superficially compliant but substantively ineffective greenwashing strategies. Simultaneously, to compensate for monitoring deficiencies, shareholders must implement more elaborate incentive and evaluation mechanisms, directly increasing transaction costs. Whether management opts to reduce substantive environmental investments or engage in symbolic over-investment for reputation building, their decisions likely deviate from corporate value maximization. This divergence ultimately manifests as inefficient resource allocation and performance deterioration in green governance. Thus, we hypothesize:

H2: 

Non-actual controllers curb green governance performance by increasing green agency costs

Stakeholder theory asserts that corporate responsibilities extend beyond shareholders to include employees, customers, and the state [30]. These groups exert distinct pressures on firms to adopt sustainable practices: investors push for environmental oversight to ensure long-term value [21], governments enact regulations to mandate green innovation [31], and public scrutiny incentivizes proactive environmental conduct [32]. However, the effective integration of these diverse pressures hinges on a stable and authoritative governance core. In firms without actual controllers, this core is absent. A dispersed ownership structure, characterized by fragmented shareholders with divergent interests, impedes the formation of consensus on long-term green strategies [29]. Consequently, external pressures—from both regulators and markets—fail to be internalized into a coherent environmental agenda. This coordination failure results in symbolic compliance rather than substantive transformation. To maintain legitimacy while avoiding costly investments, firms resort to a reactive governance model that merely meets minimum regulatory requirements [33] and engage in “cherry-picking” in their environmental disclosures—highlighting selective achievements while obscuring overall performance [34,35,36]. This shift from proactive governance to passive compliance signifies a dilution of social responsibility and leads to diminished green governance performance. We therefore hypothesize:

H3: 

Non-actual controllers curb green governance performance by weakening corporate social responsibility undertaking.

Based on the preceding theoretical analysis, non-actual controllers systematically inhibit corporate green governance performance through three distinct mechanisms. First, the monitoring vacuum creates strong managerial short-termism, crowding out substantive environmental investments (H1). Second, weakened oversight elevates specialized agency costs in green governance, distorting resource allocation (H2). Third, the lack of an authoritative core impedes effective coordination of diverse stakeholder demands, leading to symbolic compliance rather than substantive responsibility fulfillment (H3). Therefore, we propose our core hypothesis:

H4: 

Non-actual controllers curb corporate green governance performance.

Drawing upon the preceding discussion, we construct a conceptual framework that delineates the pathway through which non-actual controllers impact corporate green governance: Non-actual controllers → agency theory and stakeholder theory (by reducing environmental protection investment, increasing green agency costs, and weakening the assumption of social responsibility) → inhibition of green governance performance. The specific theoretical framework is shown in Figure 1:

4. Materials and Methods

4.1. Data Source

This study draws on a sample of Chinese A-share listed companies (2009–2022), with data on both non-actual controllers and green governance obtained from the CSMAR database. After obtaining the initial samples, the data is processed by these principles: (1) Companies with a listing status of “ST” or “*ST” are removed; (2) The sample excludes firms in the financial and insurance industries; (3) Companies with a debt-to-asset ratio greater than 1 are excluded; (4) Companies with severely missing key variables are removed. By applying the specified filtering criteria, a final sample of 31,186 firm-year observations is selected. To reduce the potential influence of extreme values, we perform 1% and 99% bilateral winsorization on all continuous variables. Additional data were obtained from both the CSMAR database and the China Research Data Service Platform.

4.2. Model Specification

Our research mainly investigates the effect of non-actual controllers on enterprise green governance performance. To explore the relationship between the two, the following basic regression model (1) is set:

$${\mathrm{G}\mathrm{G}\mathrm{P}}_{\mathrm{i}\mathrm{t}}=\alpha +\beta {\mathrm{N}\mathrm{A}\mathrm{C}}_{\mathrm{i}\mathrm{t}}+\gamma {\mathrm{X}}_{\mathrm{i}\mathrm{t}}+\mathrm{I}\mathrm{n}\mathrm{d}+\mathrm{Y}\mathrm{e}\mathrm{a}\mathrm{r}+{\epsilon }_{\mathrm{i}\mathrm{t}}$$

(1)

In model (1), the explained variable ${\mathrm{G}\mathrm{G}\mathrm{P}}_{\mathrm{i}\mathrm{t}}$ represents the green governance level of enterprise i in year t. The core explanatory variable ${\mathrm{N}\mathrm{A}\mathrm{C}}_{\mathrm{i}\mathrm{t}}$ is a dummy variable, representing whether enterprise i has an actual controller in year t. ${\mathrm{X}}_{\mathrm{i}\mathrm{t}}$ represents the control variables. The coefficient $\beta$ represents the estimated effect of non-actual controllers on green governance performance. If it is less than 0, it indicates that no actual controller suppresses the green governance performance. Meanwhile, we incorporate industry and year fixed effects and cluster standard errors by industry.

To examine the underlying channels through which non-actual controllers affect corporate green governance, we adopt the direct-testing paradigm for mechanism identification advocated by Angrist and Pischke (2009) [37] and Heckman and Pinto (2015) [38]. This approach posits that establishing a causal effect of the independent variable on the hypothesized mediator is essential for validating any theoretical mechanism. Specifically, if X affects Y through M, then X must significantly affect M. This focused examination of the X→M relationship directly reveals the inner workings of the causal black box and has been widely applied in leading empirical studies [39,40]. The following basic regression model (2) is set:

$${\mathrm{M}}_{\mathrm{i}\mathrm{t}}=\phi +\rho {\mathrm{N}\mathrm{A}\mathrm{C}}_{\mathrm{i}\mathrm{t}}+\pi {\mathrm{X}}_{\mathrm{i}\mathrm{t}}+\mathrm{I}\mathrm{n}\mathrm{d}+\mathrm{Y}\mathrm{e}\mathrm{a}\mathrm{r}+{\theta }_{\mathrm{i}\mathrm{t}}$$

(2)

In model (2), the explained variable ${\mathrm{M}}_{\mathrm{i}\mathrm{t}}$ represents the mechanism variable level of enterprise i in year t. The mechanism variables are in sequence: environmental investment, green agency cost, and social responsibility assumption. The reasons for selecting mechanism variables and the measurement methods are shown in Section 5.4. The core explanatory variable is ${\mathrm{N}\mathrm{A}\mathrm{C}}_{\mathrm{i}\mathrm{t}}$. The coefficient $\rho$ serves as the core of our mechanism analysis. A statistically significant coefficient with the expected sign provides strong empirical evidence supporting the existence of the proposed mechanism, thereby directly validating our theoretical hypotheses.

4.3. Variable Selection

4.3.1. Explained Variable

Enterprise green governance refers to a series of proactive environmental governance actions undertaken by companies, such as the green transformation of production processes, optimization of clean production technologies, and enhancement of environmental management practices. We assess green governance performance by calculating its Janis–Fadner (J-F), a method, following Jiang et al. (2021) [21], that quantifies the balance of positive and negative engagements. The formula for the calculation is presented as follows:

$$\mathrm{G}\mathrm{G}\mathrm{P}=\left\{\begin{array}{c}{(\mathrm{p}}^2-\mathrm{p}\times \left|\mathrm{q}\right|)/{\mathrm{r}}^2,\mathrm{i}\mathrm{f}\mathrm{p}>|\mathrm{q}|\\ (\mathrm{p}\times |\mathrm{q}|-{\mathrm{q}}^2)/{\mathrm{r}}^2,\mathrm{i}\mathrm{f}\mathrm{p}<\left|\mathrm{q}\right|\\ 0,\mathrm{i}\mathrm{f}\mathrm{p}=\left|\mathrm{q}\right|\end{array}\right.$$

(3)

The scoring employs a dichotomous measure: for positive governance criteria, a value of 1 is assigned for each item the company fulfills; for negative criteria, a value of −1 is assigned for each applicable item. A comprehensive green governance index is constructed from the positive and negative scores based on the J-F coefficient, bounded between −1 and 1. A higher value indicates better green governance performance. Our study extends prior work by constructing a broader framework of evaluation criteria for green governance, encompassing an expanded range of positive and negative metrics. The specific scoring criteria are shown in

Table 1. Performance scoring criteria for corporate green governance.

Positive scoring criteria for corporate green governance	Whether to pass the ISO 14000 [41] series standard certification
	Whether to receive a green reward
	Whether the pollution emissions meet the standards
	Whether the environmental governance score in the ESG rating ranks in the top 30% of the sample
	Whether the environmental management system is rated at the highest level in the ESG rating
Negative scoring criteria for corporate green governance	Whether there are any environmental violation incidents
	Whether there are any environmental petition incidents
	Whether there are any sudden environmental accidents
	Whether the pollution emissions have not met the standards
	Whether the environmental governance score in the ESG rating ranks in the bottom 30% of the sample
	Whether the environmental management system is rated at the lowest level in the ESG rating

.

4.3.2. Explanatory Variable

Following the mandatory disclosure rules for actual controllers implemented in 2003, we code NAC as 1 if a firm’s annual report indicates the absence of an actual controller, indicating that no entity holds effective control over the company. Conversely, if the annual report discloses the presence of an actual controller, the variable NAC is assigned a value of 0.

4.3.3. Control Variables

Following established practice, we incorporate a standard set of control variables, including firm Size, Age, Leverage (Lev), Return on Assets (ROA), Cash Flow (Cflow), Ownership Concentration (Top1), CEO Duality (Dual), the proportion of Independent Directors (Indep), and Institutional Ownership (Inst).

Table 2. Main variables and definitions.

Variables	Variable Definition
NAC	For a listed company without an actual controller, the value is 1; otherwise, it is 0
GGP	The J-F coefficient is calculated from the positive and negative scores of the company’s green governance.
Size	The natural logarithm of the company’s total assets
Age	The natural logarithm of the number of years the company has been established
Lev	The company’s asset–liability ratio
ROA	Company net profit/Total assets
Cflow	Net cash flow from operating activities/total assets
Top1	The shareholding ratio of the company’s largest shareholder
Dual	If the positions of general manager and chairman are combined, the value is 1; otherwise, it is 0
Indep	The number of independent directors/the number of directors
Inst	The number of shares held by institutional investors/the total number of shares

provides definitions for all main variables.

Table 3. Descriptive Statistics.

Variables	N	Mean	Sd	Min	Max
GGP	31,186	0.575	0.456	−1	1
NAC	31,186	0.0431	0.203	0	1
Size	31,186	22.22	1.295	20.02	26.33
Age	31,186	1.996	0.948	0	3.367
ROA	31,186	0.0423	0.0574	−0.195	0.201
Lev	31,186	0.408	0.202	0.0514	0.864
Cflow	31,186	0.0502	0.0683	−0.149	0.247
Top1	31,186	34.51	15.04	8.600	74.96
Indep	31,186	0.375	0.0526	0.333	0.571
Dual	31,186	0.298	0.457	0	1
Inst	31,186	0.441	0.252	0.00297	0.915

presents the descriptive statistics for the key variables. The average green governance performance of the enterprises is approximately 0.575, with a standard deviation of 0.456. There is a considerable variation in green governance performance across firms, indicating that some enterprises still have significant room for improvement. Non-actual controllers are used as the explanatory variable, with an average value of 0.0431. This implies that approximately 4.31% of the companies in the sample lack an actual controller. Descriptive statistics of other variables showed no obvious anomalies.

5. Results

5.1. Baseline Regression Result

Table 4. Baseline Regression.

	(1)	(2)	(3)	(4)
	Fixed Effect		2SLS
			First Stage	Second Stage
NAC	−0.054 ***	−0.054 ***		−1.432 **
	(0.017)	(0.014)		(0.557)
IV			0.011 ***
			(0.003)
Controls		Yes	Yes	Yes
Ind	Yes	Yes	Yes	Yes
Year	Yes	Yes	Yes	Yes
N	31,186	31,186	31,186	31,186
R2	0.047	0.064
F	9.73	51.23	16.69	47.55

Note: The standard errors are in parentheses. * p < 0.1, ** p < 0.05, *** p < 0.01. Control variables include Size, Age, Lev, ROA, Cash, Top1, Dual, Indep, and Inst. The same applies in the following tables.

presents the baseline regression results. Column (1) shows a significantly negative coefficient for NAC at the 1% level, which remains robust in column (2) even after the inclusion of control variables. The coefficient for non-actual controllers is −0.054, significant at the 1% level. Based on the sample mean of corporate green governance performance, this estimate implies an average relative decline of approximately 9.39% in green governance, indicating both statistical significance and substantial economic meaningfulness, thus providing strong empirical support for H4.

5.2. Endogeneity Test

To address endogeneity concerns, we employ the city in which the listed company is located and the number of annual mergers and acquisitions as instrumental variables for the endogeneity tests. The validity of the IV requires meeting both relevance and exclusion restriction conditions. Firstly, in terms of relevance, an active urban merger and acquisition market will create a strong atmosphere for control rights transactions, significantly increasing the possibility of the dispersion of equity in enterprises within the region and evolving into a state without an actual controller. Secondly, in terms of exclusivity, mergers and acquisitions at the city level, as external shocks at the macro level, mainly affect the control structure of enterprises by influencing regional capital operation expectations. This mechanism has no direct connection with micro-decision-making paths that affect the green governance performance of enterprises, such as investment in environmental protection technologies and fulfillment of social responsibilities. Green governance is primarily driven by internal resources, governance mechanisms, and direct regulatory pressures rather than by mergers and acquisitions activities of other firms in the same city. Furthermore, in the empirical design, by controlling a series of financial indicators such as enterprise scale, asset–liability ratio, cash flow, and profitability, we effectively excluded the potential path that the urban merger and acquisition environment might indirectly affect green governance by influencing the individual financial status of enterprises, further ensuring the exclusivity of the instrumental variables. Based on the above analysis, we believe that this instrumental variable is reasonable.

Columns (3) and (4) of Table 4 present the 2SLS regression results. Column (3) shows the first-stage results, where the instrumental variable exhibits a significantly positive coefficient at the 1% level, confirming a strong positive correlation with the no-controlling-shareholder structure. The validity of the instrumental variable is supported by multiple diagnostic tests. The first-stage F-statistic of 16.69 well exceeds the conventional threshold of 10, demonstrating strong instrument relevance. The Kleibergen–Paap Wald rk F statistic of 16.69 exceeds the Stock–Yogo critical value of 16.38 for 10% maximal bias, thereby allowing us to reject the null hypothesis of weak identification. Moreover, the Kleibergen–Paap rk LM statistic significantly rejected the null hypothesis of “insufficient identification of instrumental variables” at the 5% level. These tests collectively attest to the validity of our instrumental variable. Column (4) reports the second-stage results, demonstrating that non-actual controllers remain significantly negative. This confirms the robust inhibitory effect on corporate green governance performance after addressing endogeneity, thereby validating the reliability of our baseline results and providing further support for H4.

5.3. Robustness Tests

5.3.1. Replacing the Fixed Effect

To control the characteristic differences that do not change over time at the company or regional level, we conduct robustness checks using firm and province fixed effects. As shown in columns (1) and (2) of

Table 5. Robustness Tests.

	(1)	(2)	(3)	(4)	(5)	(6)
NAC	−0.035 *	−0.048 ***	−0.045 *	−0.052 ***		−0.056 ***
	(0.018)	(0.015)	(0.021)	(0.014)		(0.014)
RealNAC					−0.058 ***
					(0.017)
_cons	0.080	−0.299 **	−0.150	−0.226	−0.317 ***	−0.293 ***
	(0.182)	(0.112)	(0.353)	(0.158)	(0.090)	(0.099)
Controls	Yes	Yes	Yes	Yes	Yes	Yes
Year	Yes	Yes	Yes	Yes	Yes	Yes
Ind			Yes	Yes	Yes	Yes
Firm	Yes
Province		Yes
N	30,625	31,157	2419	4477	31,186	31,186
R2	0.375	0.054	0.075	0.071	0.064	0.064

Note: The standard errors are in parentheses. * p < 0.1, ** p < 0.05, *** p < 0.01. Columns (3)–(4) present the estimates of 1:1 and 1:3 nearest neighbor matching, respectively. Column (6) further controls ownership concentration, adding variables: TopTenHoldersRate; Shrhfd; Shrhfd3; Shrhfd5.

, the coefficient on non-actual controllers remains significantly negative under both specifications.

5.3.2. Propensity Score Matching Method

We employ the propensity score matching (PSM) method to conduct robustness checks. The characteristic variables selected for matching include enterprise size, enterprise age, financial leverage, return on net assets, proportion of independent directors, equity concentration, and concurrent positions. We utilize both 1:1 and 1:3 nearest neighbor matching methods. As presented in columns (3)–(4) of Table 5, the negative coefficient for non-actual controllers continues to hold after addressing sample selection concerns, underscoring the reliability of our findings.

5.3.3. Replacing the Core Explanatory Variable

Some companies claim to have no definitive controller; however, this may be a tactic to hide the actual ownership structure, thereby creating the illusion of lacking a true controller, which ultimately serves the interests of major shareholders or management. This practice may cause an incorrect specification of the core explanatory variable. To address this issue, we introduce a new explanatory variable, “Real Non-Actual Controllers” (RealNAC). We define RealNAC as 1 if a firm’s annual report confirms no actual controller and no shareholder holds >30% of shares, and 0 otherwise. The persistently significant results in column (5) of Table 5 confirm robustness using this alternative definition.

5.3.4. Excluding the Competitive Explanation

The influence of non-actual controllers can be attributed to equity dispersion, which weakens oversight and leads to managerial short-termism [4]. To eliminate this potential effect, the following control variables are added: The Top Ten Shareholders Rate, Shrhfd (the sum of the squares of the shareholding ratios of the largest shareholder), Shrhfd3, and Shrhfd5, all of which serve to measure the degree of shareholding dispersion. Column (6) of Table 5 suggests the results remain robust.

5.4. Mechanism Inspection

5.4.1. Environmental Investment

Environmental Investment (Green_Invest) is selected to capture the strategic-dimension consequence of agency problems. Rooted in agency theory, this variable directly measures the firm’s long-term strategic commitment to green initiatives. Environmental capital expenditures are inherently long-term and asset-specific, offering no immediate financial returns. In the absence of effective monitoring from an actual controller, managers with short-term performance horizons are incentivized to systematically underinvest in such projects. A significant negative effect of non-actual controllers on Green_Invest would thus provide direct evidence of a strategic contraction in the firm’s environmental agenda, testing the channel proposed in H1.

To examine the environmental investment mechanism, we follow Zhang et al. (2019) [42] and manually collect environmental-related items from the “construction in progress” account in corporate financial statements. The environmental investment intensity variable is constructed as the ratio of the annual increase in environmental investment (the sum of the relevant items) to the total assets at the beginning of the year. As shown in column (1) of

Table 6. Mechanism Analysis.

	(1)	(2)	(3)
	Green_Invest	Green_Cost	ESG_Score
NAC	−0.011 ***	0.123 **	−0.699 ***
	(0.002)	(0.055)	(0.183)
_cons	−0.066	1.140 ***	43.533 ***
	(0.052)	(0.397)	(1.241)
Controls	Yes	Yes	Yes
Ind	Yes	Yes	Yes
Year	Yes	Yes	Yes
N	28,061	30,059	31,186
R2	0.061	0.792	0.174

Note: The standard errors are in parentheses. * p < 0.1, ** p < 0.05, *** p < 0.01.

, the coefficient for non-actual controllers is −0.011, significant at the 1% level, pointing to a substantial inhibitory effect. This reduction weakens the resource base for green governance, thereby inhibiting green governance performance, thus supporting H1.

5.4.2. Green Agency Cost

Green Agency Costs (Green_Cost) quantifies operational inefficiencies stemming from agency issues. This variable shifts the focus from strategic decisions to the day-to-day execution of environmental management, capturing resource waste from the misalignment between managerial actions and shareholder interests. Weak monitoring can lead to two types of inefficiency: costly but superficial “greenwashing” initiatives, or plain managerial slack in environmental oversight. Both forms result in unnecessary or inflated administrative expenses. Consequently, a positive relationship between non-actual controllers and Green_Cost would empirically validate the operational inefficiency channel posited in H2.

To test the green agency cost mechanism, our study draws on Wang et al. (2021) [28] and Zhang Yun (2024) [43], manually collecting environmental governance expenses such as greening and sanitation fees from the “management expenses” account in corporate financial statements. The logarithm of these costs is used to measure green agency costs, with higher values indicating more severe agency problems. As presented in column (2) of Table 6, the coefficient for non-actual controllers is a significant 0.123, at the 5% level, suggesting that non-actual controllers exacerbate interest misalignment between management and shareholders regarding environmental objectives, thereby increasing agency costs. This elevated agency cost not only consumes resources but also discourages long-term environmental investments, ultimately undermining green governance. Thus, H2 is validated.

5.4.3. Social Responsibility Undertaking

Environmental Responsibility Assumption (ESG_Score) is employed to capture the firm’s legitimacy-based response to stakeholder pressures. Grounded in stakeholder theory, this metric reflects a firm’s overall commitment to maintaining its social license and reputation. Firms lacking an actual controller often have no strategic core focused on long-term reputation building. As a result, they tend to adopt a passive compliance strategy, doing only the minimum required by regulation rather than proactively improving their comprehensive ESG performance. A significant negative effect on ESG_Score would thus confirm that a weakened legitimacy commitment is a key mechanism, as stated in H3, that impairs the firm’s response to stakeholder demands.

From an external standpoint, the assumption of social responsibility by enterprises is shaped by the evaluation and recognition of external stakeholders. Corporate social responsibility awareness is operationalized through the ESG score. ESG can reflect the corporate commitment to sustainable development and social responsibility. A higher ESG score signifies stronger performance in environmental protection, as well as a greater commitment to social responsibility. The results in column (3) of Table 6 reveal that the coefficient for the non-actual controller is −0.699, significant at the 1% level, suggesting that this ownership structure undermines corporate social responsibility. The weakening of corporate social responsibility undermines both the internal drive for deep green governance and external legitimacy, thereby inhibiting corporate green governance performance. Consequently, H3 is supported.

5.5. Heterogeneity Analysis

5.5.1. Property Rights Attribute

From a political standpoint, state-owned enterprises are under strict government oversight and bear enhanced social responsibilities, demonstrating both economic and political characteristics. Compared to private enterprises, state-owned enterprises face higher pressure from environmental regulations and market supervision, necessitating active engagement in green governance. With government backing, state-owned enterprises have better access to public funds, creating favorable conditions for green technology and product innovation, thus enhancing green governance performance. However, non-state-owned enterprises generally lack government financial support and policy guidance and are constrained by limited resources, which can result in weaker green governance. A stark contrast is observed in columns (1)–(2) of

Table 7. Heterogeneity Analysis.

	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
	Property rights attribute		Institutional investor shareholding ratio		Audit quality		Environmental uncertainty
	State-owned	Non-state-owned	High	Low	High	Low	High	Low
NAC	−0.033	−0.059 ***	−0.038	−0.065 ***	0.013	−0.062 ***	−0.067 ***	−0.038
	(0.029)	(0.017)	(0.023)	(0.019)	(0.023)	(0.017)	(0.013)	(0.030)
_cons	−0.371 **	−0.303 **	−0.363 ***	−0.235	−0.249	−0.316 ***	−0.197 *	−0.416 ***
	(0.147)	(0.127)	(0.114)	(0.142)	(0.365)	(0.097)	(0.103)	(0.115)
Controls	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Ind	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Year	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
N	10,855	20,330	15,593	15,593	2057	29,129	19,927	11,259
R2	0.080	0.060	0.075	0.058	0.085	0.063	0.068	0.063

Note: The standard errors are in parentheses. * p < 0.1, ** p < 0.05, *** p < 0.01.

, where non-actual controllers significantly reduce green governance in non-state-owned enterprises (coefficient = −0.059, p < 0.01) but show no significant effect in state-owned enterprises. The evidence implies that the detrimental impact of non-actual controllers on green governance is substantially accentuated in the non-state firms.

5.5.2. Institutional Investor Shareholding

As significant external monitors, institutional investors exert considerable influence on corporate green governance [44]. Through improved green investment efficiency, promoted green innovation, and mitigated agency conflicts, institutional investors play a vital role in bolstering corporate environmental performance [45]. Meanwhile, their exit threat helps deter corporate greenwashing and incentivizes other stakeholders to strengthen environmental oversight. However, the shareholding ratios and strategic objectives of institutional investors differ, leading to varying levels of investment and supervision in green governance. Investors with higher shareholding ratios tend to prioritize long-term stability and sustainability, thereby promoting better disclosure and improved social responsibility performance [46]. In contrast, investors with smaller stakes may lack sufficient oversight, resulting in lower green governance performance. We employ a median split on institutional ownership to categorize the sample into high and low groups. Columns (3)–(4) of Table 7 suggest that in firms with a lower proportion of institutional investors, the non-actual controller shows a more pronounced inhibitory effect on green governance. The negative effect is smaller and statistically insignificant in firms with higher institutional ownership. These findings suggest that higher institutional ownership serves as an effective external monitoring mechanism, compensating for the weakened internal governance caused by dispersed ownership. Conversely, lower institutional ownership amplifies this monitoring gap, leading to more constrained green governance performance.

5.5.3. Audit Quality

External auditing serves as a critical oversight mechanism in corporate governance [7], playing a key role in green governance by enhancing information transparency and curbing managerial opportunism. High-quality audits not only ensure the authenticity and reliability of environmental disclosures [47] but also mitigate principal-agent problems, thereby aligning investor and managerial interests in green objectives. We measure audit quality based on whether an enterprise engages a “Big Four” accounting firm, with the presence of such firms indicating high audit quality. Columns (5)–(6) of Table 7 indicate that non-actual controllers have a significantly negative impact on green governance performance in firms with weaker audit quality, while this effect is insignificant in the higher audit quality firms. High-quality external auditing offsets weakened internal monitoring from dispersed ownership, mitigating the negative impact of non-controlling structures on green governance. In contrast, weak audit oversight converts internal governance deficiencies into actual green governance losses.

5.5.4. Environmental Uncertainty

As the external environment becomes increasingly complex, environmental uncertainty has emerged as a key factor influencing corporate strategic decisions. Environmental uncertainty breeds managerial conservatism in investment decisions, particularly in firms without actual controllers [48]. This volatility also undermines earnings stability and long-term sustainability [49], creating an environment detrimental to green innovation. Referring to Shen et al. (2012) [50], we measure environmental uncertainty through performance fluctuations and classify the samples into two groups based on the median of environmental uncertainty. Columns (7)–(8) of Table 7 indicate that in the high-uncertainty group, the non-actual controller has a significant negative effect on green governance performance, while the effect is insignificant in the low-uncertainty group. These findings indicate that heightened environmental uncertainty amplifies the inhibitory effect of firms without an actual controller on green governance, as external volatility and internal oversight deficiencies jointly constrain both green investment and implementation effectiveness.

6. Conclusions

Centered on the critical challenge of corporate green governance, this paper analyzes the consequences of the growing phenomenon of non-actual controllers for governance performance. Existing literature has not systematically addressed this relationship, and our research provides new theoretical and empirical evidence. The findings reveal that non-actual controllers significantly reduce the level of corporate green governance performance. The underlying mechanism appears to be that non-actual controllers significantly reduce environmental protection investments, increase green agency costs, and weaken corporate assumption of social responsibility. Heterogeneity tests show that in non-state-owned enterprises, enterprises with a low shareholding ratio of institutional investors and low audit quality, and those in a highly uncertain environment, the negative correlation is more significant. Overall, this study reveals that non-actual controllers systematically undermine green governance by suppressing substantive environmental investment and promoting symbolic governance. This is not merely an issue of corporate social responsibility, but also a core strategic and corporate governance problem. In the era of green economy, true environmental performance is the key driving force for long-term value creation and the construction of sustainable competitive advantages. Thus, what appears as a governance vacancy is, in effect, a governance failure that directly compromises a firm’s capacity for sustainable value creation.

Compared with existing literature, our study provides robust evidence that non-actual controllers intensify agency problems. Our mechanism analysis reveals three pathways through which governance deficiencies impair green performance: reduced environmental investment, higher green agency costs, and weakened social responsibility. Moreover, our research broadens the scope of green governance antecedents. Beyond external policies or managerial traits, we establish that the stability of a firm’s power structure is a fundamental driver of green governance, highlighting internal governance optimization as a prerequisite for corporate green transformation. The theoretical contribution of our study lies in moving beyond ownership concentration to demonstrate how the non-actual controller structure systematically curbs green transformation. This offers a new analytical framework for understanding corporate green governance disparities, thus aiding regulators in precise risk identification and differentiated policy-making.

Our research offers clear policy and practical implications. First, its findings hold universal relevance for mainstream corporate governance models worldwide. In market-oriented systems, measures should enhance board checks and balances, strengthen the role of independent director-led ESG committees, and activate institutional monitoring to counter managerial short-termism. Under the stakeholder-cooperative model, green objectives should be embedded into governance cores—integrating environmental responsibilities into supervisory boards and guiding banks to incorporate green performance into credit assessments. Second, corporations should improve internal governance by establishing sustainability committees chaired by independent directors and incorporating environmental experts. Executive incentives should be aligned with environmental performance to reduce agency costs and foster long-term focus. Finally, regulators should adopt targeted policies: prioritizing non-state firms and companies with weak institutional ownership or audit quality, strengthening environmental disclosure and inspections, and linking tax incentives and subsidies to environmental improvements. Encouraging institutional investors to support green initiatives in decision-making can also enhance external oversight. A synergistic approach that aligns internal governance refinement with external policy guidance can systematically counteract the adverse effects of non-actual controllers and orchestrate corporate green transformation.

Our study has several limitations. Firstly, the research conclusion is based on the Chinese context, and its universality needs further verification. Countries differ in legal environments, market supervision, and ESG development. Strict environmental regulations or mature ESG investment markets may mitigate agency issues associated with the non-actual controller structure. Future cross-national comparative studies should examine the moderating effects of national-level factors. Secondly, the discussion on the particularity of the industry is insufficient. The environmental regulations and competitive landscapes faced by different industries vary greatly. Future research could compare high-pollution and clean-technology industries to analyze how the non-actual controller structure interacts with sector-specific green strategies, supporting more targeted regulation and investment. Third, our discussion of symbolic strategies like greenwashing and cherry-picking, as well as the impact of major crises, remains limited. Future research should directly investigate these strategies to provide a more nuanced understanding of corporate responses to stakeholder pressure under specific governance structures. Finally, the revelation of the influencing mechanism is still relatively preliminary. Large-sample analysis makes it difficult to depict the actual formation process of mechanisms such as green agency costs within enterprises. To unpack the corporate governance “black box”, future research should employ qualitative methods like case studies, which can offer richer causal inference at the micro level.