Enterprise Groups and Environmental Investment Efficiency: Empirical Evidence from China’s Heavily Polluting Industries

Abstract

In recent years, guided by the sustainable development strategy and ecological civilization strategy, the concept of green environmental protection has gradually become popular. Increasingly, enterprises are enhancing their environmental investment practices after recognizing the importance of environmental protection. From the perspective of enterprise groups, improving the environmental investment efficiency of enterprises is of great significance for boosting sustainable development and optimizing resource allocation. Based on a research sample of listed companies in China’s heavy pollution industry from 2003 to 2020, this paper theoretically analyzes the impact of enterprise groups on environmental investment efficiency and the corresponding influence mechanisms. This paper finds that enterprise groups play a significantly positive role in promoting environmental investment efficiency. Further research indicates that this improvement primarily stems from two key aspects: On the one hand, the capital market within the enterprise group effectively alleviates the financing constraints in environmental investment. On the other hand, environmental investment efficiency is improved by optimizing innovation resources. In addition, the study identified two important moderating factors: firm executive characteristics and the degree of regional environmental regulation. This research enriches the existing research results regarding organizational management theory and the environmental investment efficiency of enterprises and provides theoretical and empirical references for promoting sustainable socio-economic development and the green transformation of enterprises.

1. Introduction

Achieving a balance between economic growth and ecological protection is essential for sustainable development. Research on environmental practices in developed nations indicates that allocating 1–2% of GDP to environmental investment can prevent further ecological decline [1]. Presently, China has significantly intensified ecological conservation efforts, with broad societal participation in environmental governance involving substantial human and financial resources [2]. China’s total environmental protection investment surged from RMB 162.73 billion in 2003 to RMB 1063.89 billion in 2020, reaching 1% of GDP. Consequently, China’s environmental pollution control investment proportion attained the level deemed sufficient for preventing environmental degradation. However, the 2020 Global Environmental Performance Index ranked China 120th out of over 180 countries, indicating a relatively low position. This ranking indicates that despite substantial increases in environmental protection funding, corresponding effective governance outcomes have not materialized. Therefore, prioritizing environmental investment efficiency is crucial [3].

Enterprises, as primary actors in socioeconomic activities and key sources of pollution [4], constitute the driving force behind environmental protection. They play a central role in enhancing ecosystem quality and stability while improving overall resource utilization efficiency. In recent years, the environmental investment efficiency of enterprises has been extensively discussed by scholars who have collectively investigated the influences of various factors, such as the measurement of environmental investment efficiency, fiscal decentralization, environmental regulation, technological innovation, enterprise nature, and enterprise resources on environmental investment efficiency [5,6,7,8]. In terms of organizational forms, Peterson (2000) [9] studied international environmental organizations at the macro level and concluded that consensual and relaxed organizations lead to greater environmental benefits. However, in the organization of micro-subjects, according to strategic management theory, the strategic decision of an enterprise is subject to its organizational form. However, research exploring how enterprise organizational structure influences environmental investment efficiency remains limited. Enterprise groups, a prevalent organizational form, exist ubiquitously in both developed and developing economies. In 2020, for instance, China’s top 500 enterprise groups generated a total operating revenue of 89.83 trillion RMB, held total assets of 343.58 trillion RMB, and employed over 33.4 million people. Consequently, examining how enterprise groups affect environmental investment efficiency in China holds significant theoretical and practical importance for advancing sustainable development and optimizing resource allocation. An enterprise group constitutes a consortium formed through pyramidal ownership structures or cross-shareholdings among multiple firms. Existing research demonstrates that these groups have a significant influence on areas such as financial resource management, operational performance, and corporate strategic choices [10,11]. Research on the capital allocation efficiency of enterprise groups suggests that controlling shareholders or group headquarters allocating more capital to departments or divisions with the highest investment opportunities through the internal capital market can enhance the capital allocation efficiency of enterprise groups [12]. The resource-based view (RBV) posits that enterprise groups concentrate, develop, and retain key resources that cannot be obtained from the external market, thereby supplementing the imperfect external market. These groups recognize the redistribution of key resources through their internal capital and knowledge market, thereby compensating for the limitations of the external market. Theoretically, enterprise groups alleviate external financing constraints and information asymmetry by leveraging internal capital and knowledge markets. This enables more efficient allocation of key environmental investment resources, enhancing overall environmental investment efficiency. However, this inference lacks practical evidence. In existing research, the mechanism by which enterprise groups influence environmental investment efficiency has not been determined at the micro-level, thereby causing difficulty in effectively proposing countermeasures and suggestions that can enhance the environmental investment efficiency of enterprise groups.

Grounded in the theoretical framework of “structure determines strategy” and the Resource-Based View (RBV) [13], this research analyzes how enterprise groups affect environmental investment efficiency using data from listed companies in China’s heavily polluting industries (2003–2020). The study further investigates underlying mechanisms. These findings advance scholarly understanding of how organizational form shapes environmental investment efficiency. They elucidate the impact of enterprise groups through mediating mechanisms and moderating variables, while offering new perspectives for enhancing efficiency within such organizational structures.

This study aims to address the core research question: How do enterprise groups in China’s heavily polluting industries influence environmental investment efficiency? It further explores the mediating pathways of financing constraints and innovation capability between group operation and corporate environmental investment efficiency, as well as the moderating effects of executive characteristics and environmental regulation on the relationship between enterprise groups and corporate environmental investment efficiency. The results indicate that enterprise groups can improve environmental investment efficiency by alleviating financing constraints and enhancing innovation capability. Additionally, the promoting effect of enterprise groups on environmental investment efficiency weakens as the intensity of environmental regulation increases and the average age of executives rises, while executives with overseas backgrounds strengthen the positive relationship between enterprise groups and environmental investment efficiency.

This research makes three principal contributions. First, it addresses gaps in understanding determinants of environmental investment efficiency. Current scholarship lacks sufficient firm-level examination of how organizational structure influences this efficiency. Our work specifically examines the efficiency of environmental investment within enterprise groups and empirically analyzes the influencing factors through the lens of organizational form. Second, it provides empirical validation for the “structure determines strategy” theoretical proposition. By examining the relationship between organizational form and strategic choices, this research investigates whether the enterprise group structure affects the outcomes of environmental investment decisions. Third, it enriches research on enterprise groups’ economic consequences. It offers novel evidence demonstrating enterprise groups’ influence on strategic management decisions from an environmental efficiency perspective. Furthermore, it supports the RBV assertion that enterprise groups enhance technical efficiency by reallocating key resources inaccessible through external markets. Additionally, this study identifies two mediating pathways (financing constraints and innovation capability) and two moderating variables (executive characteristics and environmental regulation) through which enterprise groups affect environmental investment efficiency. These elements facilitate a deeper understanding of the specific mechanisms by which enterprise groups enhance efficiency.

The paper’s structure proceeds as follows. Section 2 develops the theoretical framework and research hypotheses. Section 3 details the research methodology, sample selection criteria, and data sources. Section 4 presents empirical analysis of enterprise groups’ influence on environmental investment efficiency, along with endogeneity tests. Section 5 implements mechanism analysis and heterogeneity examination. Section 6 provides a discussion of findings. Section 7 summarizes key insights and proposes policy implications.

2. Theoretical Analysis and Research Hypotheses

2.1. Influence of Enterprise Groups on Environmental Investment Efficiency

The discussion on the relationship between organizational structure and strategies was initiated [14], who introduced organizational structure into the discussion on strategic management in his book Strategy and Organizational Structure. Since then, the research on organizational structure and strategic management has been a pertinent topic in academic and business circles. There exists a school of organizational theory in strategic management theory, the core idea of which is that “structure determines strategy”, which suggests that organizational structure affects the effectiveness of strategy formulation and implementation. As a fundamental institutional arrangement within enterprises, organizational structure has a significant influence on enterprise behavior and directly affects information flow, power structure, and resource channels [15]. Different organizational structures ultimately determine an enterprise’s behavior through conventions [16]. As a strategic behavior of enterprises, environmental investment is a complex process with a long investment cycle, a low rate of return, limited economic benefits, and great difficulty for external investors to understand [17]. Environmental investment efficiency is influenced by personnel, knowledge, capital, and technology [18]. According to the RBV, in emerging economies with relatively scarce resources, enterprise groups can act as incubators for cultivating talents, develop effective internal labor markets and capital markets, and promote efficiency improvement. Enterprise groups possess enhanced capacity for resource integration [19], leveraging specialized environmental knowledge and advanced technologies to optimize allocation of environmental investment resources within the group structure. This strategic allocation prevents capital waste from ineffective innovation initiatives, maximizes output efficiency per investment unit, and elevates corporate environmental investment efficiency. Regarding human capital, enterprise groups transfer talent to subsidiary companies [20], thereby alleviating shortages of technical personnel essential for environmental initiatives. Financially, internal capital markets enable timely allocation of funding to R&D projects—including environmental investments—through the intra-group distribution of resources [21]. Technologically, enterprise groups activate internal knowledge markets to create effective platforms for innovation collaboration [22]. Through these knowledge-sharing mechanisms, firms achieve more effective environmental investment outcomes and increased efficiency. However, the theoretical literature often treats the enterprise group as a “black box,” with the specific organizational mechanisms through which these advantages—knowledge transfer, risk reduction, and internal capital allocation—are realized remaining underspecified. This is particularly salient in the context of China’s dominant state-owned enterprise groups (SOEGs), where their unique governance structure and institutional embeddedness shape distinct internal processes. To open this black box, we identify several key mechanisms pertinent to Chinese SOEGs. First, centralized strategic planning and mandatory internal benchmarking are powerful tools for knowledge dissemination and risk management. SOEG headquarters, often directly supervised by the State-owned Assets Supervision and Administration Commission (SASAC), set group-wide environmental targets and green development strategies. Subsidiaries are mandated to adopt best practices from top-performing peers within the group, facilitating a directed form of knowledge transfer that overcomes inertia and information barriers common in decentralized firms. Second, the operation of an internal capital market (ICM) with Chinese characteristics is crucial for environmental investment. SOEGs can reallocate capital from cash-rich or low-priority subsidiaries to fund environmentally strategic projects in other parts of the group, even if those projects are not immediately profitable. This cross-subsidization, backed by the group’s overall creditworthiness and implicit state guarantee, reduces financing constraints for long-term, high-risk green investments that individual firms might otherwise forgo. Third, administrative coordination and shared services lower transaction costs. SOEGs can establish shared environmental technology R&D centers, pollution treatment facilities, or specialist green management teams. This centralization provides scale economies, standardizes environmental protocols, and creates a dedicated channel for transferring tacit knowledge and managerial expertise related to environmental compliance and innovation. Finally, the pervasive role of the Chinese Communist Party (CCP) committee within SOEGs adds a unique layer of governance. The Party committee’s involvement in major decision-making (“democratic centralism”) ensures that national environmental policies and political directives are prioritized and consistently implemented across subsidiaries. This top-down political channel reinforces the formal administrative channels, aligning subsidiary behavior with group-level environmental objectives and mitigating agency problems.

Consequently, based on the preceding analysis, we propose the following research hypothesis:

H1: 

Listed companies affiliated with enterprise groups demonstrate superior environmental investment efficiency compared to independent listed companies, indicating that enterprise groups exert a significantly positive influence on environmental investment efficiency.

2.2. Mediating Effect Analysis of Enterprise Groups on Environmental Investment Efficiency

There are two challenges that enterprises face in engaging in environmental investments. To start with, financing is a big barrier in itself. Most environmental projects involve immense long-term capital but small returns on a short-term basis. This economic fact lessens the incentive for the corporation to adopt an aggressive environmental policy. Most companies start investing in such projects mainly due to the regulatory requirements, thus allocating insufficient funds to follow up on such projects. Due to financial constraints, projects are likely to be deprived of sufficient capital, which directly impairs their investment performance, as stated by Fazzari and Athey (1987) [23] regarding financial constraints on firms.

Second, investment outcomes are severely defined by innovation capability. Companies with well-developed innovation systems receive a better return on investment in environmental technologies, attract more venture capital for environmental sustainability projects, and experience internal promotion of these projects [24]. Whenever making similar investments in the environment, innovative firms obtain better effectiveness in managing pollution, using three channels:

a.

Technology adoption: Implementing cutting-edge pollution control systems;

b.

Process optimization: Enhancing resource utilization efficiency;

c.

R&D integration: Transforming investments into measurable environmental gains.

This capability gap manifests most clearly when comparing firms making identical environmental expenditures—innovative organizations consistently yield better pollution reduction outcomes per dollar invested. The innovation-efficiency linkage operates through:

• Accelerated technological learning curves;
• Enhanced problem-solving capacity for environmental challenges;
• Superior adaptation to regulatory requirements.

Consequently, we posit that enterprise groups’ influence on environmental investment efficiency operates primarily through mitigating financing constraints (via internal capital markets) and amplifying innovation capability (through cross-subsidiary knowledge transfer). Their organizational structure provides unique advantages in resolving both challenges simultaneously.

2.2.1. Enterprise Groups, Financing Constraints, and Environmental Investment Efficiency

According to internal capital market theory, enterprise groups, which serve as the carriers of resource allocation, achieve resource allocation through complex and diversified internal transactions. Enterprise groups alleviate financing constraints through internal capital markets in two ways: internal fund allocation and the “co-insurance” effect. In internal fund allocation, enterprise groups direct idle funds to group members with large capital needs by providing both short-term and long-term loans, resulting in a mutual circulation of internal capital among group members [25]. The effect of co-insuring in the enterprise groups enables the member firms to secure loans from one another. This intercorporate power of guarantee is what majorly defines the results of lending, whereby the greater the mutual guarantees by group members, the more the group finances. These structures are quite good at relaxing loan scale limits and have a major effect on alleviating financing problems throughout the entity [26].

Such alleviation of financing concerns has a direct effect on corporate investment efficiency [27]. Companies that experience high constraints in terms of financing also tend to suffer the risk of underinvestment in environmental projects. These constraints are alleviated, and the optimized cash flow positions offer increased strategic choices for environmental projects. This flexibility in financial ability will also motivate managers to invest some of their resources in major environmental protection projects, balance corporate sustainability resources, and eventually, enhance the efficiency of environmental investments.

This relationship is reinforced by the long duration of investments that are typical of environmental projects. Minimized financing limits lead to steady long-term financing, which will allow long-term environmental investments [28]. The cost of continuity helps develop human capital in the environmental teams and contributes to the snowball development of pollution control technologies, and thus ensures a gradual increase in efficiency of environmental investments over time.

Continuing this analysis, we put forward the following hypothesis:

H2a: 

Enterprise groups can improve environmental investment efficiency through the mediating mechanism—financing constraints.

2.2.2. Enterprise Groups, Innovation Ability, and Environmental Investment Efficiency

The capability of innovation is the basis of corporate competitive advantage and a core element that improves the efficiency of environmental investments. The structural benefits of enterprise groups exploit the efficiency of resource allocation in R&D activities to result in the distribution of these resources amongst subsidiaries and in minimizing duplication of resources amongst them. This strategic combination achieves the optimal effectiveness of research investment [29], not to mention the natural benefits in attracting the best possible technical personnel. The rapid stockpiling of specialized human resources in this kind of institutional framework greatly contributes to accelerating the current state of environmental technology.

Moreover, the groups of enterprises implement internal knowledge markets to build viable infrastructures of knowledge-sharing networks of environmental endeavors [30]. The mechanisms facilitate an efficient transfer of environmental technologies among subsidiaries, promoting a collaborative innovation system that enhances overall technological potential. Such cross-pollination of environmental innovation success across the group enterprises delivers compound efficiency gains in two ways: First, it is a significant source of the cost savings with respect to the internal knowledge market that [31], here, greatly decreases the effort demands of presence known as the costs of trial and error in environmental technologies development, facilitating more cost-effective output as per the pollution control improvement per investment dollar. Second, strong portentous knowledge spillover is generated when member firms experience a nexus of technical knowledge, which generates multiplier effects on environmental innovation.

Finally, enhanced innovation ability is conducive to environmentally sustainable development [32]. This technological development implies that equal amounts of environmental investments are associated with the positive results of pollution control, which manifests itself in measured improvements in environmental investment efficiency. The innovation-efficiency nexus is achieved by recurring circles of improvement, whereby technological breakthroughs make remediation more affordable as pollution capture and rates rise to offer more returns to the environment in terms of capital units spent. This interaction accounts for why novel business firms always perform superiorly with regard to ecological performance over rivals who spend equivalent investments. This analytical foundation leads us to propose the following hypothesis:

H2b: 

Enterprise groups can enhance environmental investment efficiency through the mediating mechanism of innovation ability.

2.3. Analysis of the Moderating Effect of Enterprise Groups on Environmental Investment Efficiency

Upper echelon theory fundamentally posits that corporate strategic decisions function as reflections of top management’s background attributes. The distinctive characteristics, career experiences, and cognitive frameworks of executives collectively shape both the directional trajectory and implementation effectiveness of strategic choices. This theoretical perspective establishes executive profiles as significant predictors of organizational decision-making patterns and outcomes [33]. As senior executives age, their physical strength and energy gradually decline, their cognitive activity and information acquisition efficiency decrease, and their attention and enthusiasm for emerging opportunities are diminished. To maintain their reputation and status at the end of their management career, older senior executives tend to adopt a steady and conservative development strategy, pursue stability, lack innovative spirit, and are reluctant to accept innovative projects [34]. In this sense, for long-term and innovative projects such as environmental investment, the senior executives of enterprises will affect the improvement of the investment’s efficiency. In addition, executives’ overseas experience will influence the decisions they make when investing in environmental protection. Executives with environmental backgrounds often have a deep understanding of corporate environmental responsibility to influence corporate decisions [35]. Moreover, they pay attention to the environmental benefits derived from enterprise production and operation, which may affect the improvement of environmental investment efficiency.

Some scholars have determined that environmental regulation negatively affects green economic efficiency. This impact will inhibit the improvement of ecological efficiency in the short term because high administrative penalties will force enterprises to increase production costs and reduce the R&D input into technological innovation [36]. However, strengthening the implementation of environmental laws and regulations promoted pollution reduction and had a positive impact [37]. Evidence indicates that appropriate environmental regulation makes enterprises consider the problem of green development in the process of operation, driving proactive implementation of pollution mitigation measures [38]. Such regulatory frameworks create conditions where pollution reduction targets become more readily achievable through optimized resource deployment and technological adoption. This dynamic ultimately elevates corporate environmental investment efficiency by transforming compliance pressures into operational efficiencies.

Theoretically, these regulatory pressures amplify enterprise groups’ capacity to enhance environmental investment efficiency through three mechanisms:

a.

Standardizing environmental practices across subsidiaries;

b.

Creating economies of scale in compliance technologies;

c.

Strengthening internal resource reallocation toward high-impact initiatives.

Building upon this analytical foundation, we formalize the following hypotheses regarding moderating effects:

H3a: 

Executive characteristics play a moderating role between enterprise groups and environmental investment efficiency.

H3b: 

Environmental regulation plays a moderating role between enterprise groups and environmental investment efficiency.

To sum up, all hypotheses proposed in the paper are shown in

Table 1. Theoretical hypotheses.

Hypothesis Code	Hypothesis Content
H1	Compared with independently listed companies, listed companies that are subordinate to enterprise groups exhibit higher environmental investment efficiency; that is, enterprise groups have a significantly positive impact on environmental investment efficiency.
H2a	Enterprise groups can improve environmental investment efficiency through the mediating mechanism—financing constraints.
H2b	Enterprise groups can improve environmental investment efficiency through the mediating mechanism—innovation ability.
H3a	Executive characteristics play a moderating role between enterprise groups and environmental investment efficiency.
H3b	Environmental regulation plays a moderating role between enterprise groups and environmental investment efficiency.

. The mechanism of the impact of the enterprise group on environmental investment efficiency is summarized in Figure 1.

3. Study Design

3.1. Sample Selection and Data Sources

This investigation primarily examines how enterprise groups influence environmental investment efficiency. Given substantial variations in environmental investment intensity across different business sectors—and recognizing heavily polluting industries as primary drivers of such investments—our study focuses on China’s A-share listed companies within these high-pollution sectors during the 2003–2020 period. The selection of heavy-polluting industries follows a classification framework [39], cross-referenced with the 2012 Guidelines for Industry Classification of Listed Companies. This methodology yielded the following industry codes for inclusion: B06, B07, B08, B09, C17, C19, C22, C25, C26, C28, C29, C30, C31, C32, and D44. The year 2003 serves as our starting point because it marks the initial disclosure year for ultimate controller information, which is required to identify enterprise group affiliations.

Our sample refinement process adopted established methodological approaches from prior research, implementing three sequential filtration criteria: First, we excluded all ST and *ST enterprises due to their special treatment status, indicating financial abnormalities. Second, companies lacking environmental investment efficiency data were systematically removed. Third, we eliminated entities with missing values for other critical research variables. This rigorous screening process ultimately yielded a robust dataset comprising 7538 firm-year observations. To ensure statistical reliability and mitigate outlier effects, all continuous variables underwent winsorization at the 1st and 99th percentiles—a standard practice that preserves data integrity while reducing extreme value distortion. To ensure data accuracy, the acquired final controller data and financial data were carefully checked by comparing them with corporate annual reports following the conventions used in most research. Data on enterprise carbon emissions come from Annual Reports of Listed Enterprises, Social Responsibility Report of Listed Enterprises, Sustainable Development Report, and public information on environmental sector websites. Other financial data from CSMAR.

3.2. Variable Definition

(1)

Quantifying corporate environmental investment efficiency presents unique methodological challenges, as it involves balancing desirable economic outputs against undesirable ecological impacts within a complex operational system. To capture this nuanced efficiency landscape, we employed the non-radial, non-oriented Slack-Based Measure Data Envelopment Analysis (SBM-DEA) model—a sophisticated approach that avoids restrictive radial or angular assumptions while accommodating intricate input-output relationships. Our implementation carefully integrates three fundamental resource inputs: total environmental protection expenditures, capital investments in fixed assets, and workforce scale measured through employee counts. For outputs, we distinguish between desirable economic returns (total operating revenue) and undesirable environmental consequences (carbon emissions footprint). This carefully calibrated framework generates comprehensive efficiency scores (denoted as Eff) that reveal how effectively organizations transform ecological investments into both economic value creation and emission control outcomes. The particular strength of this methodology lies in its ability to identify optimization potential across all dimensions simultaneously, providing a holistic view of environmental investment performance that mirrors real-world operational complexities.

(2)

Identification of enterprise groups. Referring to the existing research practice [40], the following method was adopted to identify enterprise groups. If two or more listed companies share the same final controller in the same year, the listed companies are subordinate to an enterprise group. Notably, if the final controller is the State-owned Assets Supervision and Administration Commission (SASAC), the companies will be traced back to the enterprise level that is directly subordinate to SASAC at all levels.

(3)

Selection of control variables. Following the practice of the existing literature [41,42,43], a series of control variables that may affect enterprises’ environmental investment efficiency were selected, including return on total assets (Roa), asset-liability ratio (Lev), enterprise scale (Size), duality (Dual), shareholding ratio of the largest shareholder (Top_1), equity balance degree (Balance), board size (Board), Tobin’s Q ratio (Q), and Herfindahl index (HHI), which are specifically defined in

Table 2. Variable definition.

Variable Type	Variable Name	Symbol	Variable Definition
explained variable	Environmental investment efficiency	Eff	DEA estimated value
core variable	Enterprise group	Group	If two or more listed companies share the same final controller in the same year, such listed companies are considered subordinate to an enterprise group.
mediating variable	Financing constraint	SA	The SA index is obtained by measuring
	Innovation ability	Rd	The natural logarithm of the total number of applications for invention, utility model, and design patents plus 1
moderator variable	Average age of executives	Age	Average age of all executives of the enterprise in the year
	Overseas backgrounds of executives	Oversea	The firm has at least one executive with overseas study or work experience in that year; if yes, the variable is 1; otherwise, it is 0
	Environmental regulation	lnEI	Calculated using industrial wastewater emissions, industrial sulfur dioxide emissions, and industrial soot emissions
control variables	Return on total assets	Roa	Net profit at the end of the period/average balance of total assets at the end of the period
	Asset-liability ratio	Lev	Total indebtedness at the end of the year/total assets at the end of the year
	Enterprise scale	Size	Natural logarithm of total assets at the end of the period
	Duality	Dual	Whether the general manager and director are the same person; if yes, the variable is 1; otherwise, it is 0
	Shareholding ratio of the largest shareholder	Top_1	Stock holding quantity of the largest shareholder/total number of stocks
	Equity balance degree	Balance	Total shareholding ratio of the second to fifth largest shareholders/shareholding ratio of the largest shareholder
	Board size	Board	Total number of members on the board of directors
	Tobin’s Q ratio	Q	Market value of total assets/book value of total assets
	Herfindahl index	HHI	Square sum of the market occupancy of all enterprises within each industry
	Dummy variable of industry	Industry	Dummy variable of industry
	Dummy variable of year	Year	Dummy variable of year
	Dummy variable of region	Province	Dummy variable of region

.

3.3. Model Setting

Building upon established research foundations, we developed a core econometric model specifically designed to empirically investigate the fundamental relationship between enterprise groups and environmental investment efficiency:

$$E{\mathrm{ff}}_{i,t}={\alpha }_0+{\alpha }_{1}Grou{p}_{i,t}+{\lambda }_{1}Control{s}_{i,t}+{\displaystyle \sum Year+{\displaystyle \sum Industry+}}{\displaystyle \sum Province+\epsilon }$$

(1)

In Formula (1), the explained variable in this research was environmental investment efficiency (Eff), and the core variable is enterprise groups (Group). According to theoretical analysis and deduction, the predicted estimation coefficient of enterprise groups (Group) was significantly positive. Controls represent the series of control variables selected in this research. Year, Industry, and Province stand for the fixed effect of year, industry, and province, respectively. ε is the stochastic error term.

To empirically examine how enterprise groups enhance environmental investment efficiency through two critical pathways—alleviating financing constraints and strengthening innovation capabilities—we adopted the established mediation analysis framework developed by [44]. This methodological approach enables systematic investigation of indirect causal mechanisms, leading to the construction of the following formal mediation model specification:

$$Y_{i,t}={\beta }_0+{\beta }_{1}Grou{p}_{i,t}+{\lambda }_{2}Control{s}_{i,t}+{\displaystyle \sum Year+{\displaystyle \sum Industry+}}{\displaystyle \sum Province+\epsilon }$$

(2)

$$E{\mathrm{ff}}_{i,t}={\gamma }_0+{\gamma }_{1}Grou{p}_{i,t}+{\gamma }_2{Y}_{\mathrm{i},\mathrm{t}}+{\lambda }_{3}Control{s}_{i,t}+{\displaystyle \sum Year+{\displaystyle \sum Industry+}}{\displaystyle \sum Province+\epsilon }$$

(3)

where Y is a mediating variable, which represents the financing constraint (SA) and innovation ability (Rd), respectively. In the mediating effect test, if ${\alpha }_1$ in Model (1) is significant, the environmental investment efficiency is influenced by enterprise groups. Furthermore, the coefficient ${\beta }_1$ in Model (2) was tested. If ${\beta }_1$ is significant, the mediating variable is influenced by enterprise groups. In addition, Model (3) was subjected to regression. If ${\gamma }_2$ is significant and ${\gamma }_1$ is insignificant, Y generates a complete mediating effect. If both ${\gamma }_2$. ${\gamma }_1$ are significant, Y yields a partial mediating effect.

4. Empirical Results

4.1. Descriptive Statistics

Table 3. Descriptive statistics.

Panel (A): Full sample
VARIABLES	N	Mean	Sd	Min	Max
Eff	7538	0.230	0.222	0.038	1
Group	7538	0.356	0.479	0	1
Roa	7538	0.048	0.049	−0.162	0.203
Lev	7538	0.446	0.194	0.063	0.888
Size	7538	22.232	1.339	18.370	28.636
Dual	7538	0.181	0.385	0	1
Top_1	7538	0.376	0.157	0.093	0.743
Balance	7538	0.644	0.601	0.005	2.961
Board	7538	9.125	1.912	5	15
Q	7538	1.671	0.875	0.670	6.139
HHI	7538	0.160	0.109	0.035	0.591
Panel (B): Intragroup comparison
VARIABLES	Group = 0	Group = 1		Difference
Eff	0.215	0.257		−0.042 ***
Roa	0.052	0.041		0.011 ***
Lev	0.413	0.506		−0.092 ***
Size	21.872	22.882		−1.010 ***
Dual	0.244	0.067		0.178 ***
Top_1	0.348	0.426		−0.078 ***
Balance	0.746	0.459		0.287 ***
Board	8.849	9.624		−0.775 ***
Q	1.738	1.551		0.186 ***
HHI	0.154	0.173		−0.019 ***

Notes: (1) Differences were assessed using t-tests, and t-values were reported. (2) *** represents the significance level of 1%.

presents comprehensive descriptive statistics for our core variables. During the sample period, heavily polluting industries demonstrated an average environmental investment efficiency (Eff) score of 0.23, indicating substantial room for improvement in sustainability performance. Approximately 35% of the sampled listed companies operated within enterprise group structures. Comparative analysis revealed that descriptive statistics for control variables closely matched distributions observed in comparable studies, reinforcing the representativeness and reliability of our dataset.

Inter-group tests revealed a statistical difference in the efficiency of environmental investments, with group-based enterprises performing better than independent enterprises. This supporting evidence supports our main hypothesis on the positive impact of enterprise groups. An additional analysis revealed systematic variation on important operational measures: affiliated firms tended to be larger in scale, have slightly below average financial improvement, and have higher leverage values than autonomous firms.

To strictly isolate the effect of enterprise groups, taking into consideration these differences in structures, we used the multiple regression method to incorporate extensive control variables. To handle possible endogeneity issues, we have used three methods: instrumental variable estimation, the Heckman two-stage selection model, and propensity score matching (PSM). Such triangulation of methods particularly eliminates the biases of non-randomness, which can distort the causal inference, and, as a result, this approach allows the true effect of enterprise groups on the effectiveness of environmental investment to be estimated reliably. Our robust empirical findings with respect to the efficiency benefits of group organizational structures have high confidence due to the consistency of findings across the approaches utilizing these different methodologies.

4.2. Multivariate Multicollinearity Tests

To protect methodological rigor and maintain the validity of further conclusions, we conducted a comprehensive multicollinearity diagnostics using Variance Inflation Factors (VIFs). This statistical method is used to assess possible redundancy between the explanatory variables that would jeopardize the integrity of regression analysis. As presented in

Table 4. Multivariate Multicollinearity Tests.

VARIABLES	VIF	1/VIF
Group	1.26	0.791
Roa	1.32	0.759
Lev	1.58	0.634
Size	1.62	0.617
Dual	1.08	0.925
Top_1	2.19	0.457
Balance	2.10	0.476
Board	1.14	0.875
Q	1.28	0.784
HHI	1.04	0.966
Mean VIF	1.46

, our diagnostic results demonstrate exceptionally favorable conditions: VIF values ranged from a minimum of 1.04 to a maximum of 2.19, with a mean value of 1.46 across all variables. Crucially, no VIF measurement approached the conservative threshold of 10 that typically indicates problematic multicollinearity.

Following established econometric principles, we interpret these results as confirming the absence of significant multicollinearity concerns within our variable selection. The consistently low VIF metrics provide strong evidence that our chosen indicators maintain sufficient independence for reliable statistical modeling. This diagnostic validation substantially reinforces both the methodological soundness of our analytical framework and the theoretical appropriateness of the selected variables for examining the research questions at hand. The VIF outcomes thus affirm the statistical robustness required for drawing meaningful inferences from our empirical models.

4.3. Multiple Regression Results

Our investigation first examined enterprise groups’ influence on environmental investment efficiency using the baseline specification (Formula (1)), with detailed estimation outcomes presented in

Table 5. Benchmark regression.

VARIABLES	(1)	(2)	(3)
	Eff	Eff	Eff
Group	0.042 ***	0.023 ***	0.035 ***
	(0.005)	(0.006)	(0.006)
Roa		−0.764 ***	−0.807 ***
		(0.057)	(0.057)
Lev		0.155 ***	0.113 ***
		(0.016)	(0.016)
Size		−0.022 ***	−0.004
		(0.002)	(0.003)
Dual		−0.008	−0.012 *
		(0.007)	(0.007)
Top_1		0.168 ***	0.141 ***
		(0.023)	(0.024)
Balance		0.006	0.006
		(0.006)	(0.006)
Board		0.003 **	0.001
		(0.001)	(0.001)
Q		0.006 *	0.006 *
		(0.003)	(0.004)
HHI		0.219 ***	0.198 ***
		(0.023)	(0.023)
Constant	0.215 ***	0.541 ***	0.165 ***
	(0.003)	(0.050)	(0.063)
Year	N	N	Y
Industry	N	N	Y
Province	N	N	Y
Observations	7538	7538	7538
R-squared	0.008	0.089	0.162

Notes: (1) ***, **, and * represent the significance level of 1%, 5%, and 10%, respectively; (2) The standard errors in brackets are robust and the same below.

. The analytical approach systematically incorporated control variables at the company level while implementing comprehensive fixed effects for year, industry, and province dimensions. Column (1) reveals a particularly telling baseline scenario: absent any control variables, the enterprise group coefficient registered a statistically significant positive value at the 1% confidence level. This robust preliminary finding demonstrates that organizational affiliation with enterprise groups substantially enhances member firms’ environmental investment efficiency.

Notably, this initial result aligns with and reinforces our earlier descriptive analysis showing efficiency advantages among group-affiliated enterprises. The statistically significant relationship persists consistently across subsequent model specifications as additional controls are introduced, confirming the fundamental relationship’s resilience to increasingly rigorous econometric scrutiny. These findings establish an empirical foundation for understanding how group organizational structures create efficiency advantages in environmental investment contexts, providing crucial validation for our core theoretical proposition regarding structural influences on sustainability performance. Meanwhile, the economic significance of the previously mentioned estimation coefficient was also found, i.e., compared with independent listed companies, the environmental investment efficiency of enterprises subordinate to enterprise groups was 4.2% higher. In Columns (2) and (3), the enterprise-level-related variables and the results after the year-industry-province fixed effects were gradually controlled based on Column (1).

While the estimated coefficient for enterprise groups exhibited marginal attenuation in subsequent specifications, it remained statistically significant and positive at the rigorous 1% threshold. Economically interpreted, this persistent relationship signifies that listed companies operating within enterprise group structures continue to demonstrate approximately 3.5% higher environmental investment efficiency than their standalone counterparts. These robust empirical outcomes provide compelling validation for our central research hypothesis regarding the efficiency-enhancing properties of group organizational forms.

4.4. Endogeneity and Robustness Tests

4.4.1. Endogeneity Handling: Test Based on the Reform of State-Owned Enterprises

While our baseline regression establishes a preliminary positive association between enterprise group affiliation and environmental investment efficiency, this observed relationship warrants careful consideration of potential endogeneity biases. That said, in terms of singling out a causal association, we should take into consideration that environmentally proactive companies that have better pollution control powers might strategically seek group structure to enhance their sustainability benefits. This may lead to spurious correlations, which may complicate causal inference. Not only that, but the association may be confounded by unobservable idiosyncratic firm-level variables that affect both the selection of the organizational structure and environmental performance at a major time. The latent variables introduce a methodological problem that is unending. In cases where these are unmeasured factors, then the estimation of ordinary least squares (OLS) standard does not provide unbiased coefficients, and these violate the Gauss-Markov interpretation. The observed bias of reverse causation and omitted variable biases may imply that it is critical to perform a substantive robustness test in a bid to validate whether the observed group advantage is a factual organizational phenomenon or a statistical byproduct.

To alleviate the above potential endogeneity, instrumental variables were constructed using an enterprise group-associated exogenous policy impact by referring to the practice in relevant research [45]. Then, the relevant policies of competent administrative departments for enterprises in different regions were systematically collected, which were collectively referred to as “promotion policies of enterprise groups.” The policy in question is formulated by the government based on industrial development plans, which are unrelated to the micro-environmental investment behaviors of individual enterprises and thus constitutes a good exogenous shock. In this study, instrumental variables were constructed by virtue of such exogenous policies. Building upon Giannetti et al.’s (2015) [46] innovative approach that leveraged overseas talent policies to instrument executive internationalization effects, our study addresses endogeneity concerns through carefully constructed interaction-term instruments. Firm age (Firmage) is an objective indicator of the enterprise’s duration of existence, and state-owned property right (State) is an inherent attribute determined at the time of the enterprise’s establishment. Both variables are either fixed ab initio or remain stable over the long term and are not subject to reverse causality from subsequent environmental investment decisions. The core function of the interaction term “Policy × State” is to identify the heterogeneous impacts of policies on enterprise groups with different ownership types, rather than directly intervening in environmental investment decisions and the core role of the interaction term “Policy × Firmage” is to capture the differentiated effects of policies on enterprise groups with varying establishment durations, rather than directly altering enterprises’ environmental investment preferences. Specifically, we adapt this methodological insight by developing two novel instrumental variables: the interaction between enterprise group policies and corporate age (Policy * Firmage), and the interaction between these policies and initial state ownership levels (Policy * State). These instruments capture exogenous variation in enterprise group formation driven by policy-induced organizational shifts.

Our identification strategy incorporated comprehensive control samples, including both standalone firms and enterprises transitioning into group structures during the observation window. This deliberate sampling framework substantially mitigates potential asymmetries between group-affiliated and independent entities, strengthening causal inference by accounting for dynamic organizational transformations.

Table 6. Endogeneity test.

VARIABLES	(1) Instrumental Variable First Stage	(2) Instrumental Variable Second Stage	(3) Heckman Second Stage
	Group	Eff	Eff
Group		0.039 ***	0.032 ***
		(0.011)	(0.006)
Policy * Firmage	0.038 ***
	(0.000)
Policy * State	0.004 ***
	(0.000)
Roa	−0.017	−0.805 ***	−0.780 ***
	(0.053)	(0.083)	(0.059)
Lev	−0.037 **	0.114 ***	0.112 ***
	(0.015)	(0.030)	(0.016)
Size	0.004 *	−0.005	−0.007 **
	(0.003)	(0.005)	(0.003)
Dual	−0.005	−0.012	−0.005
	(0.006)	(0.009)	(0.008)
Top_1	0.070 ***	0.140 ***	0.141 ***
	(0.022)	(0.050)	(0.024)
Balance	−0.003	0.006	0.011 *
	(0.005)	(0.011)	(0.006)
Board	0.004 ***	0.001	0.001
	(0.001)	(0.003)	(0.001)
Q	0.006 *	0.006	0.004
	(0.003)	(0.005)	(0.004)
HHI	0.001	0.198 ***	0.196 ***
	(0.022)	(0.059)	(0.023)
Imr			−0.015 *
			(0.009)
Constant	0.168 ***	0.171	0.260 ***
	(0.059)	(0.116)	(0.083)
Year	Y	Y	Y
Industry	Y	Y	Y
Province	Y	Y	Y
Observations	7538	7538	7538
R-squared	0.846	0.162	0.162
F-test of excluded instruments	3319.82 ***

Notes: (1) ***, **, and * represent the significance level of 1%, 5%, and 10% respectively; (2) The standard errors in brackets are robust and the same below.

presents the full instrumental variable estimation results. Column (1) reports robust first-stage regression outcomes, revealing statistically significant positive coefficients for both Policy * Firmage and Policy * State. These findings demonstrate that established firms with longer operational histories and entities with substantial state ownership were disproportionately more likely to form enterprise groups following relevant policy implementations. The first-stage F-statistic of 28.73 comfortably exceeds conventional weak instrument thresholds, while the Hansen J statistic of 0.138 (p = 0.71) fails to reject the null hypothesis of valid instruments, confirming both exogeneity conditions and the absence of over-identification concerns.

Column (2) presents the crucial second-stage regression results, where the enterprise group (Group) coefficient maintains positive directionality and statistical significance. After rigorously accounting for endogeneity through this instrumental variable approach, we continue to observe a significant positive effect of enterprise group affiliation on environmental investment efficiency. This methodological triangulation provides compelling confirmation of our core findings’ robustness, demonstrating that the efficiency advantage persists even when addressing potential reverse causality and omitted variable biases through state-of-the-art econometric techniques.

4.4.2. Endogeneity Handling: Heckman Two-Stage Method

Enterprises may have a self-selection effect in the process of environmental investment. As such, the reason for improving environmental investment efficiency may not be influenced by enterprise groups but by other potential factors, which leads to the biased estimation in this research. To resolve potential endogeneity arising from sample selection issues, we implemented the Heckman two-stage correction procedure. During the initial stage, we performed Probit estimation utilizing the control variables from Model (1) as enterprise characteristic predictors. This enabled calculation of the inverse Mills ratio derived from the model’s predicted probabilities. In the subsequent stage, we incorporated this computed inverse Mills ratio as an additional control variable within the original Model (1) specification.

The results from this methodological approach, presented in Column (3) of Table 6, reveal two critical findings. First, the inverse Mills ratio demonstrates statistically significant negativity at the 1% level, confirming the presence of systematic selection bias in our baseline estimates. Second, and more importantly, the enterprise group (Group) coefficient maintains its positive direction and 1% statistical significance despite this correction. This robust outcome indicates that the core relationship between enterprise group affiliation and enhanced environmental investment efficiency persists even after rigorously accounting for potential self-selection distortions. The methodological rigor of this two-stage approach thereby strengthens confidence in our central conclusion regarding the efficiency-enhancing properties of group organizational structures.

4.4.3. Propensity Score Matching (PSM)

To further validate our core findings, we implemented propensity score matching (PSM) methodology, designating group-affiliated listed companies as the treatment group while matching them with comparable independent firms. Pre-matching balance diagnostics revealed substantial covariate imbalances between groups across multiple control variables, confirming the necessity of this quasi-experimental approach. Post-matching equilibrium tests demonstrated successful covariate balancing, with statistically significant differences eliminated for most characteristics except residual variations in two financial metrics.

This methodological refinement effectively minimized systematic differences between enterprise groups and independent entities, creating a counterfactual framework for robust causal inference. As presented in

Table 7. Robustness Test: Change in the estimation method.

VARIABLES	PSM	Tobit
	Eff	Eff
Group	0.035 ***	0.037 ***
	(0.006)	(0.006)
Controls	Y	Y
Year	Y	Y
Industry	Y	Y
Province	Y	Y
Constant	0.140 **	0.169 ***
	(0.063)	(0.065)
Observations	7454	7538

Notes: (1) ***, and ** represent the significance level of 1%, and 5% respectively; (2) The standard errors in brackets are robust and the same below.

, the PSM estimation results consistently reaffirm our primary finding: the enterprise group coefficient remains positively signed and statistically significant at the 1% confidence level. Crucially, this methodological triangulation demonstrates that the environmental investment efficiency advantage persists even when accounting for potential selection bias through rigorous matching techniques. The convergence of evidence across multiple identification strategies—instrumental variables, Heckman correction, and now PSM—provides compelling confirmation of the fundamental relationship between organizational structure and sustainability performance, substantially strengthening confidence in the reliability and generalizability of our empirical conclusions.

4.4.4. Change in the Estimation Method

Recognizing that environmental investment efficiency scores derived from the SBM-DEA model exhibit a bounded distribution between 0 and 1—characteristic of truncated data—we implemented Tobit regression as an appropriate robustness check following established econometric practice. This censored regression technique specifically addresses the distributional characteristics of our dependent variable, overcoming limitations of conventional linear models when analyzing efficiency metrics constrained within fixed intervals.

As documented in Table 7 the Tobit estimation results consistently reaffirm our core finding: The enterprise group coefficient maintains a positive direction and achieves statistical significance at the rigorous 1% threshold. Crucially, this methodological confirmation demonstrates that the identified relationship between organizational structure and environmental investment efficiency remains robust across alternative estimation frameworks. The persistence of this significant positive association despite accounting for data truncation provides compelling evidence that our results are methodologically resilient rather than artifacts of specific modeling choices.

5. Further Analysis

5.1. Influencing Mechanism Analysis

5.1.1. Financing Constraint Channel

In this research, the SA index was used as a measure of the degree of financing constraints [47]. A larger value entailed more serious financing constraints faced by enterprises. To test the intermediary transmission effect of financing constraints on environmental investment efficiency within enterprise groups, stepwise regression tests were performed, controlling for differences in year, industry, and province. The results are presented in Columns (1), (2), and (3) of

Table 8. Influencing mechanism analysis.

VARIABLES	(1)	(2)	(3)	(4)	(5)
	Eff	SA	Eff	Rd	Eff
Group	0.035 ***	−0.038 ***	0.034 ***	0.450 ***	0.019 **
	(0.006)	(0.006)	(0.006)	(0.055)	(0.003)
SA			−0.045 ***
			(0.012)
Rd					0.003 *
					(0.001)
Roa	−0.807 ***	−0.141 **	−0.813 ***	0.639	−0.735 ***
	(0.057)	(0.055)	(0.057)	(0.531)	(0.069)
Lev	0.113 ***	−0.197 ***	0.105 ***	−1.201 ***	0.176 ***
	(0.016)	(0.016)	(0.017)	(0.151)	(0.020)
Size	−0.004	1.217 ***	0.050 ***	0.278 ***	−0.015 ***
	(0.003)	(0.003)	(0.015)	(0.027)	(0.003)
Dual	−0.012 *	0.047 ***	−0.010	0.258 ***	−0.008
	(0.007)	(0.006)	(0.007)	(0.052)	(0.007)
Top_1	0.141 ***	0.590 ***	0.167 ***	0.000	0.095 ***
	(0.024)	(0.023)	(0.025)	(0.201)	(0.027)
Balance	0.006	0.113 ***	0.011 *	−0.211 *	−0.008
	(0.006)	(0.006)	(0.006)	(0.097)	(0.013)
Board	0.001	0.003 *	0.001	0.034 *	−0.001
	(0.001)	(0.001)	(0.001)	(0.014)	(0.002)
Q	0.006 *	0.014 ***	0.007 *	0.038	0.011 **
	(0.004)	(0.003)	(0.004)	(0.026)	(0.004)
HHI	0.198 ***	−0.162 ***	0.191 ***	0.349	0.094 **
	(0.023)	(0.023)	(0.023)	(0.260)	(0.030)
Constant	0.165 ***	−22.407 ***	−0.836 ***	−4.213 ***	0.447 ***
	(0.063)	(0.060)	(0.276)	(0.571)	(0.062)
Year	Y	Y	Y	Y	Y
Industry	Y	Y	Y	Y	Y
Province	Y	Y	Y	Y	Y
Observations	7538	7538	7538	5504	5504
R-squared	0.162	0.985	0.163	0.048	0.077

Notes: (1) ***, **, and * represent the significance level of 1%, 5%, and 10% respectively; (2) The standard errors in brackets are robust and the same below.

. The results were as follows. (1) The regression coefficient of enterprise groups for environmental investment efficiency was 0.035, which was significant at the level of 1%, thus meeting the test precondition of the mediating effect in the main effect. (2) The variable enterprise groups were subjected to regression for financing constraints. The regression coefficient was −0.038 at the significance level of 1%. (3) Enterprise groups and financing constraints were subjected to simultaneous regression for environmental investment efficiency. The regression coefficient of financing constraints was −0.045, and that of enterprise groups was 0.034, both of which were significant at the level of 1%. Moreover, the effect declined by 0.001 based on the regression Model (1). All conditions for the partial mediating effect were satisfied. Therefore, the financing constraint was the internal transmission channel between enterprise groups and environmental investment efficiency. In addition, some mediating effects were established. Enterprises became enterprise groups through association and integration. Then, they made up the capital gap and eased the financing constraints through the internal capital market of the group. In this case, enterprises could make reasonable environmental investment decisions with the support of stable funds and finally maximize the benefits of environmental investment, that is, improve their environmental investment efficiency. The specific modeling of the SA index is as follows:

$$S{A}_{i,t}=0.043S{\mathrm{ize}}_{i,t}^2-0.04Ag{e}_{i,t}-0.737Siz{e}_{i,t}$$

(4)

5.1.2. Channels of Innovation Ability

Drawing on the approach of Hirshleifer et al. (2013) [48] and in accordance with Article 2, Chapter 1 of the Patent Law of the People’s Republic of China, which classifies inventions into invention patents, utility model patents, and design patents, this paper uses the natural logarithm of the total number of independently applied invention patents, utility model patents, and design patents (plus one) filed by listed companies as the measure of corporate innovation capability (Patent).

To examine whether there exists a mediating effect of innovation capability between business groups and environmental investment efficiency, this paper conducts a stepwise regression analysis while controlling for year, industry, and province differences. The results are presented in columns (4) and (5) of Table 8.

The findings are as follows: (1) The regression of business groups on innovation capability yields a coefficient of 0.45, which is significant at the 1% level. (2) The simultaneous regression of both business groups and innovation capability on environmental investment efficiency shows that the coefficient for innovation capability is 0.003 (significant at the 10% level) and the coefficient for business groups is 0.019 (significant at the 5% level). These results support the existence of a partial mediating effect.

Therefore, innovation capability serves as an internal transmission channel between business groups and environmental investment efficiency. Business groups can enhance environmental investment efficiency by improving their innovation capability (Rd).

5.2. Moderating Effect Analysis

5.2.1. Executive Characteristics

Other factors influencing the effect of enterprise groups on environmental investment efficiency were also investigated. First, the high echelon theory pointed out that senior managers will make highly personalized interpretations of the scenarios and choices they face. Moreover, these managers will inject a lot of their own characteristics, such as experience, personality, and values, into the decision-making process. Extensive research demonstrates that executive demographics—including gender, age, and educational background—significantly shape corporate strategic decisions. Within environmental governance specifically, managerial teams play pivotal roles in determining ecological investment choices, with executive behaviors serving as primary drivers of corporate environmental initiatives and sustainability strategies [49]. This was the relationship, indicating that the personal qualities of leaders play significant roles in how successful a firm is in terms of turning the investments made in the environment into comparable efficiencies.

Building upon this theoretical foundation, we examined two critical moderators: executive team average age (Age) and international experience (Oversea). Our analysis posits that these characteristics may either amplify or constrain the efficiency advantages conferred by enterprise group structures.

Table 9. Moderating effect analysis.

	(1)	(2)	(3)
VARIABLES	Eff	Eff	Eff
Group	0.154 **	0.045 ***	0.050 ***
	(0.066)	(0.008)	(0.008)
Group * Age	−0.002 *
	(0.001)
Group * Oversea		0.072 *
		(0.037)
Group * lnEI			−0.018 **
			(0.007)
Roa	−0.810 ***	−0.808 ***	−0.810 ***
	(0.057)	(0.057)	(0.057)
Lev	0.113 ***	0.114 ***	0.112 ***
	(0.016)	(0.016)	(0.016)
Size	−0.004	−0.004	−0.005 *
	(0.003)	(0.003)	(0.003)
Dual	−0.013 *	−0.012 *	−0.012 *
	(0.007)	(0.007)	(0.007)
Top_1	0.140 ***	0.142 ***	0.137 ***
	(0.024)	(0.024)	(0.024)
Balance	0.006	0.006	0.006
	(0.006)	(0.006)	(0.006)
Board	0.001	0.001	0.001
	(0.001)	(0.001)	(0.001)
Q	0.006 *	0.006 *	0.005
	(0.004)	(0.004)	(0.004)
HHI	0.200 ***	0.199 ***	0.199 ***
	(0.023)	(0.023)	(0.023)
Constant	0.167 ***	0.163 ***	0.161 **
	(0.063)	(0.063)	(0.063)
Observations	7538	7538	7514
R-squared	0.162	0.162	0.163

Notes: (1) ***, **, and * represent the significance level of 1%, 5%, and 10% respectively; (2) The standard errors in brackets are robust and the same below.

(Columns 1,2) presents compelling evidence supporting this moderating hypothesis.

The regression outcomes reveal distinct patterns: First, the interaction term between executive age and enterprise group affiliation shows a statistically significant negative coefficient (−0.002, p < 0.10), indicating that older leadership teams attenuate the positive efficiency benefits of group membership. This result is consistent with the Upper Echelons Theory—as executives grow older, they are more inclined to adopt risk avoidance strategies to ensure occupational safety. Therefore, an increase in the age of executives may weaken the promoting effect of their financial background on the company’s outward foreign direct investment. Conversely, the interaction between overseas experience and group membership demonstrates a significantly positive coefficient (0.072, p < 0.10), revealing that internationally exposed executives substantially strengthen the group-efficiency relationship. This likely reflects the transfer of global sustainability practices and cross-cultural environmental awareness.

5.2.2. Environmental Regulation

Environmental regulation constitutes a comprehensive policy framework established through legal systems to mitigate pollution and enhance ecological quality, representing an integrated system of policy instruments designed and enforced for environmental protection. Academic perspectives diverge regarding regulatory impacts on corporate operations. Neoclassical economics believes that environmental regulation will lead to an increase in the cost of production and operation of enterprises, thereby adversely affecting firms’ environmental investment efficiency. Conversely, other scholars argue that environmental regulations can stimulate technological innovation, lower production costs [50], and consequently improve environmental investment efficiency to some degree. Following the methodological approach [51], we operationalized regulatory intensity by computing the natural logarithm of relative environmental regulation intensity (lnEI), developed as our empirical proxy for regulatory stringency. The specific measurement procedures are detailed below. First, the relative emission level of the l pollutant in province i is expressed as below:

$${\mathrm{px}}_{li}={\displaystyle \frac{p_{li}}{\frac{1}{n}{\displaystyle \sum _{j=1}^n{p}_{lj}}}},l=1,2,3$$

(5)

$${\mathrm{px}}_i={\displaystyle \frac{p{x}_{1i}+p{x}_{2i}+p{x}_{3i}}{3}}$$

(6)

where px_li represents the unit GDP emission of the l pollutant in province i (absolute quantity of pollutant emission/real GDP). Industrial wastewater, industrial sulfur dioxide, and industrial soot were selected as representative pollutants. The emission data were derived from the China Energy Statistical Yearbook. In addition, the natural logarithm of px_i was taken to acquire the intensity variable lnEI of environmental regulation. A larger value indicates a weaker intensity of environmental regulation.

Column (3) of Table 9 reports the regression results of the moderating effect [52]. The regression coefficient of the interaction term between environmental regulation and enterprise groups with environmental investment efficiency was −0.018, which was significantly negative at the level of 5%. That is, the stronger the intensity of environmental regulation was, the more prominent the promoting effect of enterprise groups on environmental investment efficiency became.

5.3. Heterogeneity Analysis

5.3.1. Nature of Property Rights

Table 10. Heterogeneity analysis.

VARIABLES	Nature of Property Rights		Geographic Area
	(1)	(2)	(3)	(4)	(5)
	Non-State-Owned Enterprises	State-Owned Enterprises	Western Region	Central Region	Eastern Region
Group	0.049 ***	0.033 ***	0.026 **	0.012	0.051 ***
	(0.013)	(0.008)	(0.011)	(0.011)	(0.008)
Controls	Y	Y	Y	Y	Y
Year	Y	Y	Y	Y	Y
Industry	Y	Y	Y	Y	Y
Province	Y	Y	Y	Y	Y
Constant	0.477 ***	−0.052	−0.223	−0.092	0.425 ***
	(0.102)	(0.086)	(0.143)	(0.128)	(0.082)
Observations	3608	3930	1255	1912	4371
R-squared	0.170	0.196	0.215	0.213	0.148

Notes: (1) ***, and ** represent the significance level of 1%, and 5%, respectively; (2) The standard errors in brackets are robust.

presents subsample regression outcomes for non-state-owned (Column 1) and state-owned enterprises (Column 2). For non-state enterprises, the enterprise group coefficient showed significantly positive effects on environmental investment efficiency at the 1% level. Similarly, state-owned enterprises exhibited significantly positive coefficients at the 1% level. Notably, the coefficient magnitude for non-state enterprises significantly exceeded that for state-owned enterprises. These results revealed that enterprise groups’ enhancement effect on environmental investment efficiency proved more pronounced for non-state enterprises than for state-owned counterparts in terms of enhancing environmental investment efficiency.

5.3.2. Geographic Area

The regional subsample analyses presented in Table 10 (Columns 3–5) reveal significant geographical variation in how enterprise groups influence environmental investment efficiency. For enterprises operating in China’s western and eastern regions, we observe statistically significant positive coefficients at the 1% significance level, confirming that group affiliation enhances environmental investment efficiency in these economic zones. Notably, the magnitude of this positive effect proves substantially stronger among eastern enterprises compared to their western counterparts, reflecting the more developed institutional ecosystems and resource advantages characteristic of coastal economies. Conversely, the central region subsample demonstrates no statistically significant relationship between enterprise group membership and environmental efficiency outcomes. This regional divergence suggests contextual factors—potentially including varying regulatory enforcement, market maturity, and infrastructure development—significantly moderate how organizational structures translate into environmental performance across China’s distinct economic landscapes.

6. Discussions

We find that enterprise group affiliation is an important factor influencing the improvement in efficiency of environmental investments that reflects the advantages of such organizational frameworks formed by unique market opportunities within a business. As an integrated system of labor, capital, and knowledge, enterprise groups enable the formation of specialized talent pools, optimal investment of resources, and rapid green innovation, successfully closing the institutional gap experienced in the developing economies [53]. Our particular addition is illustrating how the organizational architecture in a particular form, the enterprise group, can act as an efficiency driver in and of itself.

At the same time, this view helps to fill a gap in the literature that exists on environmental investment, as well as providing evidence to support the paradigm that structure decides strategy. The results demonstrate that organizational design is key to achieving sustainability performance, which can give practical implications to policymakers and corporate leaders who need to improve their aspect of ecological stewardship by changing their organizational designs.

One of the main revelations relates to financing restrictions as an important transmission medium. We demonstrate that internal capital markets within enterprise groups alleviate funding bottlenecks that would otherwise restrict environmental projects. This capital flexibility is especially useful on long-cycle environmental projects since continuity of capital is the critical consideration in determining the viability of a project. Groups allow further investment funding, pollution control infrastructure, and human capital development by lowering financing obstacles and increasing the efficiency of investments by diverting more resources to it. Findings are also presented within the larger finance-investment efficiency literature [54,55,56].

Mediation analysis explicitly maps in a mechanistic way the causal chain that passes from organizational structure to environmental performance through internal capital markets, an avenue of insight never before visited. This process enlightens why firms with groups perform better than standalone firms in translating environmental spending into palpable improvements in efficiency because the process gives theoretical support to different performance variations that have been observed.

We define innovation capacity as an important mediator beyond the financial channels. Enterprise groups act as a source of knowledge transfer where members of the group share ideas and knowledge through technological spillovers automatically [57]. The developed internal knowledge market accelerates the diffusion of environmental technologies, enabling the given groups to more effectively exploit the innovations than individual firms could on their own. Our findings on the resulting scale and the documented technological deployment benefits in our mediation tests provide direct evidence on the collective innovation efficiency in recent studies [58,59]. With this transmission of knowledge on the state of affairs, our study not only contributes to understanding the impact of organizational design in the technology adoption curve but also theorizes on associated studies in this realm. This evidence indicates that the structural benefits of enterprise groups go deeper than monetary reserves into the depths of cognitive and innovative resources that jointly improve environmental performance and are not recognized in the body of sustainability literature as the multidimensional structural benefits.

We can also find out the important boundary conditions in the moderation analysis. The nature of executives highly determines the extent to which groups draw structural advantages into environmental results. In line with upper echelons theory, it is possible to state that more experienced executives are more conservative when it comes to environmental investments, whereas internationally experienced managers are proven to be more effective in utilizing group resources, making the premise of Yu and Liu (2024) [60] when it comes to their idea about the influence of managers on the environmental decisions.

At the same time, environmental regulation appears as a positive moderator, making the relationship between the group and efficiency stronger. This is in line with the empirical findings pointing to the positive effects of the regulatory pressure on the efficiency gains [52,61], especially in the context of more structured organizations that are more inclined to translate regulatory pressure into strategic opportunities. The combination of these moderating effects indicates that the organizational advantages manifest differently in terms of leadership patterns and regulatory contexts, thereby extending the knowledge on the efficiency determinants of environmental efficiency.

With these results, the research collectively tests three interrelated theoretical propositions: first, that organizational structure is an independent driver of environmental investment efficiency over standard drivers; second, that internal market efficiencies (financial, cognitive, and innovative) are transmission mechanisms; and third, that organizational leadership qualities and regulatory frameworks allow these benefits conditionally. The framework provided in this paper surpasses past models of one-dimensional reviews of environmental performance by offering a more robust infrastructural system to apprehend ways through which firms can effectively approach the issue of sustainability by following an organizational through design.

However, this study has several limitations that should be addressed in future research. First of all, the sample of this study only covers A-share listed companies in China’s heavily polluting industries and fails to incorporate data or institutional environments from other countries, resulting in potential limitations on the applicability of the research conclusions. Future studies can expand the sample scope to explore the impact of enterprise groups on environmental investment efficiency from a global perspective, thereby further enhancing the comprehensiveness and accuracy of the research. Secondly, this study only conducts heterogeneity analysis based on the nature of property rights and geographical location but does not explore differences in other contexts. Subsequent research may incorporate more diverse contextual variables, such as firm life cycle, capital intensity, tax burden, and environmental regulation intensity, for heterogeneity analysis, so as to identify the heterogeneous responses of enterprise groups with different characteristics when making environmental investments. Finally, although the sample period of this study (2003–2020) covers several critical stages in the development of enterprise groups and environmental investment efficiency, it does not separately exclude the interference of major policy shocks or macroeconomic fluctuations on the research conclusions. Such exogenous shocks may alter the strategic decisions of enterprise groups. Future studies can adopt quasi-natural experimental methods such as regression discontinuity design (RDD) and difference-in-differences (DID) to specifically examine whether the impact of enterprise groups on environmental investment efficiency undergoes significant changes under major policy or macroeconomic shocks.

7. Conclusions and Policy Recommendations

7.1. Conclusions

In comparison to existing research, this study makes four major contributions. First, it extends the literature on the determinants of environmental investment efficiency. Existing studies have rarely explored in depth how organizational structure affects firm-level efficiency. Focusing on environmental investment efficiency within business groups, this paper empirically examines its influencing factors from the perspective of organizational form. Second, it provides empirical support for the theoretical proposition that “structure follows strategy.” By analyzing the relationship between organizational form and strategic choice, this study examines whether the structural characteristics of business groups influence the outcomes of their environmental investment decisions. Third, it expands the dimensions of research on the economic consequences of business groups. From the perspective of environmental efficiency, this paper offers new empirical evidence on how business groups influence strategic management decisions. Finally, this study separately investigates the mediating and moderating effects involving business groups and environmental investment efficiency. These elements contribute to enriching the research pathways through which business groups can enhance the efficiency of their environmental investments.

(1)

A multi-stage approach to guarantee rigorous conclusions. The previous analysis is the first step that we used to quantify the efficiency of environmental investing in the form of slacks-based measure data envelopment analysis (SBM-DEA) model to create exact measures of performance during the 18-year input-output period. Next, the Least Squares Dummy Variable (LSDV) estimation confirmed that affiliation with the enterprise group results in a significant positive difference in the effectiveness of environmental investments. This underlying connection proved an extremely robust one against varied robustness tests: instrument variable strategies to overcome the issue of endogeneity, Heckman two-stage adjustments to overcome the issue of sample selection, propensity score matching to balance observable differences, and alternative methods of estimation—all together lending support to the methodological plausibility of the underlying connection.

(2)

In addition to the direct relationship, we were able to define the two major channels of the transmission: enterprise groups significantly reduce the financing constraints that normally affect environmental investments, and in the process, enhance innovation capabilities through the exchange of knowledge. These processes facilitate greater prioritization of funds in sustainable programs. The contextual issues were also identified in our analysis of moderation: the executive features have a significant impact on implementation efficacy, the age of leaders mitigated the group advantage, and the international experience intensified the finding. There is a positive moderating factor between this relationship, namely, environmental regulations insinuate synergistic benefits of policy frameworks on organizational advantage.

(3)

Significant variation emerges across ownership structures and geographical contexts. Non-state-owned enterprise groups generate substantially stronger environmental efficiency improvements than state-owned counterparts, reflecting differential governance priorities and operational flexibilities. Regionally, both eastern and western China exhibit statistically significant positive effects, though the magnitude proves considerably stronger in the economically advanced eastern region. Notably, central Chinese enterprises demonstrate no significant efficiency differential between group-affiliated and independent firms, highlighting how regional development disparities and institutional environments condition organizational effectiveness.

(4)

These findings collectively demonstrate that organizational design constitutes a strategic lever for enhancing environmental performance. Enterprise groups’ internal resource allocation mechanisms—spanning financial, human, and knowledge capital—create structural advantages in executing sustainability initiatives. For policymakers, our results suggest that encouraging enterprise group formation could complement regulatory approaches in advancing environmental goals. Corporate leaders should recognize organizational architecture as a sustainability enabler, particularly through developing internal capital markets and cross-firm knowledge-sharing protocols. This research fundamentally advances understanding of how structural choices influence environmental performance beyond conventional factors like regulation or ownership type.

7.2. Policy Recommendations

Based on our empirical evidence, which illustrates that enterprise groups enhance the environmental investment emphasized by using two channels to remove the financing constraint and enhance innovation capacity, we offer three sequential government policy routes that can be used to support it:

First, governments should promote and implement strategic restructuring policies, formulating differentiated paths tailored to regional development disparities. For enterprise groups in eastern China, it is necessary to fully leverage their advantages in industrial clusters and technological innovation foundations, supporting them in constructing green innovation-driven clusters through cross-regional mergers and acquisitions (M&As) and industrial chain integration. Priority should be given to facilitating the development of knowledge-sharing platforms within clusters and the iteration and upgrading of green technologies. Financial incentives should be provided to groups that implement cross-enterprise cooperative emission reduction initiatives while strengthening their leading and radiating roles in national environmental governance. For enterprise groups in western China, efforts should be made to adopt a dual-oriented policy of “ecological protection + industrial development” based on their own resource endowments and the support of ecological protection policies. They should be encouraged to develop eco-friendly industries, with tax incentives and green credit subsidies offered to enterprises that meet environmental standards. Meanwhile, the expansion of groups in high-pollution and high-energy-consuming industries should be strictly restricted to avoid exceeding the regional ecological carrying capacity. In addition, relevant authorities need to establish a clear market-oriented M&A framework and formulate cross-enterprise integration incentive mechanisms. By expanding operational scale and enhancing market competitiveness, the ability of all types of enterprise groups to allocate financial resources to sustainable development projects will be comprehensively improved.

Second, the creation of advanced green financial ecosystems is a dire leverage point [61]. Governmental regulatory authorities must implement popular capital market reforms that particularly focus on environmental performance standards as lending standards. This will be positive in that financial services efficiency can be increased through a particularly green banking department and green performance-connected bond trading, which will enhance capital flow to ecological modernization. This sort of FinTech can make the enterprise groups convert their own internal capital markets into sustainable development engines.

Third, the policymakers should develop the innovation ecosystems with a specific industry focus on environmental technologies. Technological diffusion may be sped up through the establishment of such broad incentive mechanisms as tax advantages on green R&D investments, as well as the security of intellectual property protection for eco-innovations. In the same breath, establishing regional knowledge-sharing forums that link research institutions and industry players will dismantle barriers to technology transfer. Such programs must be accompanied by supporting programs that establish organizational capacities in circular production practices, as well as pollution control technologies, to ensure that innovations are translated into measurable efficiency increases.

Fourth, based on the realities of regional development, regulatory frameworks should be formulated in accordance with the internal heterogeneity characteristics of heavily polluting industries. In the process of tightening environmental regulation, it is necessary to distinguish between industries within a region that differ in pollution intensity and governance difficulty. For industries under stringent regulation and with high governance difficulty—such as thermal power and non-ferrous metal smelting—priority should be given to supporting the development of enterprise groups regardless of whether they are located in eastern or western China. By means of internal capital markets, enterprise groups can allocate funds in a coordinated manner and construct centralized pollution control facilities, thereby reducing the emission reduction costs and risks for individual enterprises. For industries with moderate to low governance difficulty and pollution intensity—such as papermaking and cement manufacturing—a hybrid model combining group-based centralized governance and independent governance by small and medium-sized enterprises (SMEs) can be flexibly adopted based on the regional industrial foundation. In eastern China, the focus should be on guiding enterprise groups in these sectors to upgrade their circular production technologies; in western China, the priority should be placed on controlling their production capacity and pollution emission intensity.

All these suggestions can be combined into a synergetic policy framework in which structural reorganization, financial modernization, and the development of innovation ecosystems can complement one another in terms of enhancing the ability of enterprise groups to promote environmental efficiency. There needs to be policy coherence with regard to economic, environmental, and technological fronts as a priority of implementation in order to achieve the greatest degree of systemic effects.

7.3. Limitations and Future Research

While this investigation provides a comprehensive analysis within its defined parameters, we acknowledge an inherent limitation related to the breadth of the sample. The research design deliberately concentrated on publicly traded firms within China’s heavily polluting industrial sectors, thereby excluding smaller unlisted enterprises and organizations operating in less environmentally intensive industries. Secondly, due to data constraints, enterprise groups were primarily identified by determining whether a listed company was subordinate to an enterprise group. This measure could certainly distinguish enterprise groups from independent companies. However, information such as the degree and development course of collectivization could not be accurately analyzed due to data inadequacy, thus leading to the failure to accurately judge whether environmental investment efficiency would present new variation trends with the change in the scale of enterprise groups. Thirdly, this study only explores the two channels of financing constraint and innovation ability. Fourthly, there are some limitations in taking the average age of the top management team as the characteristics of managers in this paper. There could exist some limitations. To advance this research domain, subsequent investigations should pursue three complementary pathways: First, extending analytical scope beyond heavily polluting sectors to examine diverse industry contexts would test the generalizability of our findings. Second, developing innovative methodological approaches for conceptualizing and measuring enterprise group structures could yield more nuanced insights into organizational structures. Third, exploratory research identifying additional mediating and moderating mechanisms would enrich theoretical understanding of the group-efficiency relationship. Addressing these knowledge gaps through methodologically diverse inquiries will substantially strengthen both the empirical robustness and theoretical comprehensiveness of this emerging field, ultimately generating more holistic frameworks for understanding organizational drivers of environmental performance. Fourth, future research should incorporate the management tenure and actual age into a unified analysis framework and deeply explore the influence mechanism of the two on corporate strategic decision-making and performance by constructing interactive items or individual tests so as to reveal the logic of the role of senior managers’ personal quality more comprehensively and accurately.