Can the Energy Rights Trading System Become the New Engine for Corporate Carbon Reduction? Evidence from China’s Heavy-Polluting Industries

Abstract

As global climate change intensifies with unprecedented urgency, nations worldwide have increasingly adopted market-based environmental regulatory instruments to advance carbon reduction objectives. In 2017, China launched energy rights trading pilots, thereby providing a crucial policy instrument for controlling total energy consumption at its source. However, the specific impacts and transmission pathways through which this system influences corporate carbon reduction behavior remain insufficiently explored through rigorous empirical investigation. Drawing upon panel data from heavy-polluting companies listed on the Shanghai and Shenzhen A-share markets, this study employs a difference-in-differences methodology to identify the causal effects of energy rights trading systems on corporate carbon reduction. Our findings reveal that energy rights trading systems significantly reduce corporate carbon emission intensity, generating pronounced emission reduction effects. Further mechanism analysis demonstrates that this system operates through two principal pathways: first, by promoting increased green investment among enterprises, whereby short-term emission reductions are achieved through procurement of energy-saving equipment and environmental protection facilities, and second, by stimulating corporate green technological innovation, whereby long-term sustainable emission reductions are realized through the development of energy-saving technologies and clean processes. Additionally, the research reveals that enterprises with lower financing constraints and stronger supply chain bargaining power respond more actively to policy implementation, with policy effects exhibiting significant heterogeneity. This study not only enriches the theoretical understanding of market-based environmental regulatory policy effects but also provides crucial empirical evidence for improving the energy rights trading system design and enhancing policy implementation effectiveness, thereby offering important policy insights for promoting corporate green transformation and achieving “dual carbon” objectives.

1. Introduction

Global climate change has emerged as one of humanity’s most formidable challenges [1,2], prompting governments worldwide to establish carbon reduction targets in response to this global crisis [3,4]. Against this backdrop, market-based environmental regulatory instruments have garnered widespread attention due to their cost-effectiveness advantages, with carbon emissions trading systems and energy rights trading systems [5,6,7,8,9,10] being two pivotal market-based mechanisms gaining promotional application globally. China, as the world’s largest carbon emitter, initiated carbon emissions trading pilots in 2013 and established a national carbon market in 2021; simultaneously, to control total energy consumption at its source, China launched energy rights trading pilots in four provinces—Zhejiang, Fujian, Henan, and Sichuan—in 2017, with Jiangsu Province subsequently joining the pilot scope. The implementation of these market-based environmental regulatory policies provides a rare quasi-natural experimental opportunity for investigating their impacts on corporate carbon reduction behavior.

While market-based environmental regulations have demonstrated effectiveness, significant gaps remain in understanding energy rights trading systems’ mechanisms and heterogeneous effects. Existing research surrounding the influence of market-based environmental regulation on corporate environmental performance has generated extensive discussion, yielding abundant theoretical and empirical results. The theoretical foundations of market-based environmental regulation trace back to property rights and externalities theory, with subsequent developments establishing the efficiency properties of tradeable permit systems. Analysis of prices versus quantities under uncertainty and comprehensive reviews of cap-and-trade experiences provide essential frameworks for understanding how different market-based instruments may perform under varying conditions [11,12,13]. Regarding carbon emissions trading system research, scholars have discovered that this system, through constructing artificial scarcity constraints, can significantly alter corporate cost structures and decision-making logic, thereby producing substantial emission reduction effects [14]. Further research has confirmed the dual dividend effects of carbon trading policies in promoting green development efficiency and regional carbon equity [15]. Research on operational mechanisms indicates that carbon emissions trading systems primarily influence corporate carbon reduction behavior through two dimensions: technological innovation and investment adjustment. Related studies have found that this system, by increasing corporate R&D investment willingness and capital input, reduces corporate carbon emission intensity and total emissions by 4.3% and 7.5%, respectively [16]. However, other studies have identified policy effect differentiation, with analysis based on Chinese listed company data showing that carbon emissions trading pilot policies reduced overall green patent proportions by approximately 9.26%, primarily because enterprises tend to achieve emission reduction targets through production reduction rather than increased green technological innovation [17].

Compared to the rich research on carbon emissions trading systems, energy rights trading systems, as emerging market-based environmental regulatory instruments, remain relatively underexplored. Existing studies have only conducted preliminary exploration from the dual dividend perspective of corporate economic performance and carbon emissions, finding that energy rights trading policies can simultaneously improve corporate economic performance and reduce carbon emissions; however, this research lacks in-depth analysis of specific transmission mechanisms [18]. While existing research on carbon emissions trading systems provides important references for understanding market-based environmental regulation mechanisms, energy rights trading systems exhibit significant differences from carbon emissions trading systems in policy design philosophy, constraint objectives, and implementation mechanisms. By directly constraining total energy usage rather than total carbon emissions, energy rights trading systems may generate different guiding effects on corporate investment decisions and innovation directions.

Furthermore, existing research predominantly focuses on single transmission pathway analysis, with systematic research on green investment and green technological innovation as important transmission mechanisms remaining insufficient. Green investment, as a short-term strategy for enterprises to address environmental constraints, demonstrates relatively rapid effectiveness, whereas green technological innovation represents a more complex and long-term process, with both exhibiting significant differences in terms of timeliness and sustainability. Existing literature has not fully revealed how these two transmission mechanisms operate under the influence of energy rights trading systems, nor their differentiated performance under varying corporate characteristics and external environments.

Based on the aforementioned research status and deficiencies, this study addresses three specific research gaps: first, the lack of comprehensive empirical evidence on energy rights trading systems’ emission reduction effectiveness compared to well-studied carbon trading systems; second, insufficient understanding of dual transmission mechanisms through which energy constraint policies operate, particularly the complementary roles of green investment and technological innovation; and third, limited knowledge of how enterprise heterogeneity moderates policy effectiveness under different constraint mechanisms. This study proposes the following core questions: How do energy rights trading systems influence corporate carbon reduction behavior? What are their transmission mechanisms? What roles do green investment and green technological innovation play? What impacts do different corporate characteristics have on policy effectiveness?

The marginal contributions of this research are primarily manifested in three aspects based on analysis of 4842 firm-year observations from China’s heavy-polluting industries spanning 2013–2023. First, regarding research subjects, this study focuses on energy rights trading systems, contributing to research in this insufficiently explored field. Unlike carbon emissions trading systems, which constrain total emissions, energy rights trading systems directly limit energy consumption, potentially generating different corporate behavioral responses through distinct constraint mechanisms. The research identifies a 12.8% reduction in corporate carbon emission intensity, providing empirical evidence for understanding how consumption-based constraints differ from emission-based constraints in promoting corporate environmental behavior. Second, concerning mechanism analysis, this research constructs a comprehensive transmission mechanism framework encompassing green investment and green technological innovation, thoroughly analyzing the operational pathways of both mechanisms. Finally, in terms of research methodology, this study employs rigorous causal identification strategies such as difference-in-differences methods, supplemented by multiple robustness tests and heterogeneity analyses, ensuring the reliability and policy applicability of research conclusions while providing scientific evidence for policymakers to optimize environmental regulation design.

2. Literature Review and Hypothesis Development

2.1. Literature Review

Carbon emissions trading systems, as market-based environmental regulatory instruments, have attracted widespread academic attention regarding their emission reduction effects and operational mechanisms. The theoretical foundation for emissions trading can be traced back to Montgomery’s seminal work, which first established the “independence property” concept, demonstrating that regardless of initial permit allocation, the final configuration after trading would be efficient [11]. Early theoretical analyses highlighted the cost-effectiveness advantages of market-based instruments over uniform standards, providing analytical frameworks for estimating potential cost savings associated with market-based policies [19]. Subsequent research has emphasized the importance of auction design in permit allocation, arguing that auctioning permits is superior to grandfathering due to reduced tax distortions, greater flexibility in cost distribution, and enhanced innovation incentives [20,21]. Existing research indicates that carbon emissions trading systems, through constructing artificial scarcity constraints, effectively alter corporate cost structures and decision-making logic, thereby generating significant emission reduction effects [22]. Quasi-natural experimental research based on China’s regional carbon market pilots has found that carbon emissions trading systems reduce regulated enterprises’ total emissions by 16.7% and emission intensity by 9.7%, with this effect primarily achieved through energy consumption conservation and low-carbon fuel conversion [14]. Further research has confirmed the dual dividend effects of carbon trading policies in promoting green development efficiency and regional carbon equity, with these policies not only significantly improving green total factor productivity in the pilot regions but also promoting regional carbon equity through reduced investment in carbon-intensive industries [15]. Additional research has revealed the promotional effects of carbon emissions trading systems on corporate R&D investment and emission reduction, with this system, through increasing corporate R&D investment willingness and capital input, reducing corporate carbon emission intensity and total emissions by 4.3% and 7.5%, respectively [16]. However, the design and performance of emissions trading programs have revealed important lessons regarding the need for careful price moderation and program adjustments, particularly as the emerging international architecture features separate trading systems rather than the integrated global framework originally envisioned [13,23]. These studies provide solid empirical foundations for the emission reduction effectiveness of carbon emissions trading systems; however, they also demonstrate certain differences in specific emission reduction magnitudes across different studies, possibly related to variations in sample selection, time windows, and econometric methods.

Regarding the operational mechanisms through which carbon emissions trading systems influence corporate carbon reduction, existing literature primarily analyzes two dimensions: technological innovation and investment adjustment. From a strategic perspective, research has shown that when multi-jurisdictional externalities are present and uncertainties among jurisdictions are independent, the choice between price and quantity depends on the relative slopes of marginal benefits and costs, suggesting that the original nonstrategic criterion for instrument choice may have wider applicability in strategic environments [12]. At the technological innovation level, research based on Chinese provincial panel data has found that carbon emissions trading pilot policies significantly promote low-carbon technological innovation, with this promotional effect gradually strengthening over time; dynamic effect testing shows that the effect peaks in the third year following policy implementation [24]. However, research based on Chinese listed company data from 1990–2018 indicates that carbon emissions trading pilot policies overall reduced green patent proportions by approximately 9.26%, primarily because enterprises tend to achieve emission reduction targets through production reduction rather than increased green technological innovation, with enterprises more frequently choosing to cut R&D investment in response to reduced cash flows and expected revenues [17]. At the investment adjustment level, related research has found that carbon emissions trading systems promote corporate green investment through dual mechanisms of external pressure alleviation and internal incentive enhancement, with enterprises characterized by lower government intervention levels and higher internal executive compensation incentive levels demonstrating stronger policy response effects [25]. Analysis using time-varying difference-in-differences models has revealed sustained significant positive causal relationships between carbon emission prices and corporate total factor productivity, with this effect primarily achieved through mechanisms of increased R&D patents and capital-investment-enhanced R&D executive status, providing new microeconomic evidence for the Porter hypothesis [26].

Nevertheless, compared to the rich research on carbon emissions trading systems, energy rights trading systems, as emerging market-based environmental regulatory instruments, remain relatively underexplored. Existing studies have only conducted preliminary exploration from the dual dividend perspective of corporate economic performance and carbon emissions, finding that energy rights trading policies can simultaneously improve corporate economic performance and reduce carbon emissions; however, this research lacks in-depth analysis of specific transmission mechanisms [18]. While existing research on carbon emissions trading systems provides important references for understanding market-based environmental regulation mechanisms, energy rights trading systems exhibit significant differences from carbon emissions trading systems in policy design philosophy, constraint objectives, and implementation mechanisms, potentially possessing unique characteristics in their pathways and effects for influencing corporate carbon reduction. Particularly regarding the two key transmission mechanisms of green investment and green technological innovation, energy rights trading systems, through directly constraining total energy usage rather than total carbon emissions, may generate different guiding effects on corporate investment decisions and innovation directions. Moreover, existing research predominantly focuses on single transmission pathway analysis of carbon emissions trading systems, lacking systematic comparative research on the interrelationships between green investment and green technological innovation and their relative contributions to policy effects. Therefore, this study aims to fill research gaps regarding the impact of energy rights trading systems on corporate carbon reduction mechanisms by constructing more comprehensive theoretical frameworks and employing more precise empirical methods to thoroughly analyze the specific mechanisms through which energy rights trading systems influence corporate carbon reduction via green investment and green technological innovation. The aim is to provide a more solid theoretical foundation and empirical support for improving market-based environmental regulatory policy systems.

2.2. Research Hypotheses

Energy rights trading systems, as market-based energy management instruments, alter corporate energy usage cost structures and operational logic to a certain extent through constraints on corporate energy usage quotas and the introduction of market trading mechanisms. Under traditional models, corporate energy consumption is primarily influenced by market prices and technological conditions [27,28], whereas implementation of energy rights trading systems adds quota constraints to this foundation, transforming energy usage from relatively abundant production factors into limited resources requiring rational allocation. This institutional design stimulates corporate energy conservation motivation through artificially created scarcity, prompting enterprises to focus more attention on energy usage efficiency during production processes. According to the Porter hypothesis in environmental economics, appropriately intense environmental regulation often promotes coordination between environmental protection and economic development through stimulating innovative activities [29,30,31]. Under energy rights trading systems, enterprises no longer face merely compliance pressures but rather comprehensive incentive systems encompassing cost constraints, technological choices, and market opportunities. When enterprises achieve reductions in unit product energy consumption through technological improvements or management optimization, they not only reduce direct energy usage costs but can also obtain additional revenues through selling surplus quotas, providing economic motivation for continuous enterprise improvement. Furthermore, energy rights trading systems, through transmission of market price signals, gradually transform energy conservation and emission reduction from external regulatory requirements into endogenous economic considerations for enterprises. Based on the above analysis, this research proposes the following hypothesis:

H1. 

Energy rights trading systems can significantly reduce corporate carbon emission intensity.

Green investment, as one of the important strategies for enterprises to address environmental constraints, exhibits new characteristics and motivational mechanisms under the influence of energy rights trading systems. Traditional perspectives consider corporate environmental protection investment as compliance costs—expenditures made to satisfy regulatory requirements—with such investments typically possessing certain externality characteristics where corporate private returns may be lower than social returns, thereby limiting corporate investment motivation [32,33]. However, the introduction of energy rights trading systems considerably alters this cost–benefit structure. Through marketization of energy usage rights, corporate energy conservation and emission reduction behaviors can directly translate into economic returns, thereby internalizing portions of environmental benefits as corporate private returns [34,35,36]. Specifically, when enterprises increase green investment—such as purchasing energy-saving equipment, renovating production processes, or constructing treatment facilities—these investments not only reduce corporate energy consumption and pollution emissions but, more importantly, release additional energy rights quotas for enterprises to trade in markets. The existence of this revenue mechanism transforms green investment from purely cost items into projects with certain expected investment returns, enhancing corporate investment enthusiasm to some extent. Additionally, green investment may further strengthen emission reduction effects through technological learning effects and experience accumulation, with knowledge and capabilities accumulated by enterprises during green technology application processes not only improving current environmental performance but also accumulating technological foundations for addressing stricter future environmental regulations. Based on the above analysis, this research proposes the following hypothesis:

H2. 

Energy rights trading systems reduce corporate carbon emission intensity through promoting corporate green investment.

Green technological innovation, as one of the important pathways for enterprises to achieve long-term sustainable development, may demonstrate more pronounced strategic value under the influence of energy rights trading systems. Technological innovation typically represents investment activities characterized by high uncertainty and long-term orientation, requiring enterprises to consider multiple factors, including innovation costs, success probabilities, and expected returns, when deciding on innovation investments [37,38]. Under traditional market environments, returns from green technological innovation are often difficult to accurately assess and promptly realize, with this uncertainty affecting corporate innovation willingness to some extent. According to innovation economics theory, clear market incentives and return expectations constitute important conditions for promoting corporate technological innovation. Energy rights trading systems, through establishing relatively clear energy usage cost and return mechanisms, provide relatively clear value realization pathways for green technological innovation [39,40]. When enterprises develop more efficient energy-saving technologies or clean production processes, these innovative achievements can directly translate into energy cost savings and energy rights quota surpluses, thereby providing relatively certain return expectations for innovation investments. Simultaneously, green technological innovation possesses certain cumulative and spillover characteristics, with corporate technological progress in specific areas often driving coordinated development of related technologies [41]. Under continuous incentives from energy rights trading systems, enterprises may gradually adjust their R&D strategies and innovation directions, allocating more resources to green technology fields and forming relatively sustainable innovation capabilities. This enhancement of innovation capabilities not only achieves certain emission reduction effects in current periods but may also provide support for enterprises to gain competitive advantages in future low-carbon economic transitions. Based on the above analysis, this research proposes the following hypothesis:

H3. 

Energy rights trading systems reduce corporate carbon emission intensity through promoting corporate green technological innovation.

Based on the theoretical analysis above, this study constructs a comprehensive theoretical framework to examine the mechanisms through which energy rights trading systems influence corporate carbon reduction, as illustrated in Figure 1. The framework conceptualizes energy rights trading systems as the core policy intervention that operates through two distinct but complementary transmission channels: green investment and green technological innovation. The first pathway represents a relatively immediate response mechanism where firms increase environmental protection investments to optimize energy efficiency and reduce carbon emissions in the short term, while the second pathway captures the longer-term strategic adjustments where firms enhance their green innovation capabilities to achieve sustainable emission reductions. This theoretical framework provides the conceptual foundation for our empirical investigation, and the subsequent analysis will employ real-world empirical data from China’s heavy-polluting industries to test whether these theoretical relationships hold in practice, thereby validating or refuting the proposed mechanisms through causal identification strategies.

3. Methodology and Data

3.1. Sample Selection

This study selects heavy-polluting listed companies on the Shanghai and Shenzhen A-share markets from 2013 to 2023 as research samples, with data spanning the complete period before and after energy rights trading system pilots, ensuring effectiveness of policy effect identification. Heavy-polluting industry identification primarily relies on the “Listed Company Industry Classification Guidelines” revised by the China Securities Regulatory Commission in 2012, the “Listed Company Environmental Protection Verification Industry Classification Management Directory” formulated by the Ministry of Environmental Protection of the People’s Republic of China in 2008, and the “Listed Company Environmental Information Disclosure Guidelines,” encompassing 16 heavy-polluting industries including coal, mining, textiles, leather manufacturing, papermaking, petrochemicals, pharmaceuticals, chemicals, metallurgy, and thermal power. These industries serve not only as key regulatory subjects for energy rights trading systems but also as crucial domains for achieving carbon reduction objectives, possessing strong policy sensitivity and representativeness. Corporate financial data is sourced from the CSMAR database, corporate carbon emission data is calculated based on environmental information disclosure content and energy consumption data in corporate annual reports, green investment data is calculated through manual collection of environmental protection-related expenditures from construction-in-progress details and management expense details in corporate annual reports, and green technological innovation data is sourced from the State Intellectual Property Office patent database and the China Research Data Service Platform (CNRDS). During data processing, this research excludes ST and *ST enterprises, performs interpolation processing for some missing values, and applies 1% WINSORIZATION to continuous variables to control extreme value effects, ultimately obtaining an unbalanced panel dataset containing 4842 enterprise-year observations.

3.2. Measurement Framework

Corporate carbon emission intensity (CI) serves as the core dependent variable in this research. Following existing research [42], this study adopts the ratio of corporate carbon emissions to operating revenue to quantify corporate carbon emission intensity, with this ratio based on industry energy consumption data conversion to derive corporate carbon emissions. Related data is sourced from the “China Energy Statistical Yearbook” and the “China Industrial Economic Statistical Yearbook,” with a CO₂ conversion coefficient of 2.493 per kilogram of standard coal according to Xiamen Energy Conservation Center standards [43]. The calculation formula is

$${\mathrm{C}\mathrm{I}}_{\mathrm{i},\mathrm{t}}={\displaystyle \frac{{\mathrm{C}\mathrm{a}\mathrm{r}\mathrm{b}\mathrm{o}\mathrm{n} \mathrm{E}\mathrm{m}\mathrm{i}\mathrm{s}\mathrm{s}\mathrm{i}\mathrm{o}\mathrm{n}\mathrm{s}}_{\mathrm{i},\mathrm{t}}}{{\mathrm{R}\mathrm{e}\mathrm{v}\mathrm{e}\mathrm{n}\mathrm{u}\mathrm{e}}_{\mathrm{i},\mathrm{t}}}}$$

(1)

where CI represents carbon emission intensity measured in tons of CO₂ per thousand CNY of revenue. This measurement provides a standardized metric for comparing emission performance across firms of different sizes and revenue scales.

The energy rights trading system (ERT) serves as the core explanatory variable in this research. According to the “Energy Rights Paid Use and Trading System Pilot Program” issued by China’s National Development and Reform Commission in 2017, Zhejiang, Fujian, Henan, and Sichuan provinces were designated as the first batch of pilot provinces, with Jiangsu Province joining the pilot in 2019. Therefore, this study constructs a treatment group dummy variable, Treat, taking value 1 if the enterprise’s registered province belongs to pilot provinces and 0 otherwise, and a time dummy variable, Post, taking value 1 for 2017 and subsequent years and 0 otherwise. The core explanatory variable for energy rights trading systems is set as

$${\mathrm{E}\mathrm{R}\mathrm{T}}_{\mathrm{i},\mathrm{t}}={\mathrm{T}\mathrm{r}\mathrm{e}\mathrm{a}\mathrm{t}}_{\mathrm{i}}\times {\mathrm{P}\mathrm{o}\mathrm{s}\mathrm{t}}_{\mathrm{t}}$$

(2)

Green investment (GIV) is measured by the ratio of corporate green investment amount to total assets. From construction-in-progress details, this study compiles capitalized environmental protection expenditures, including wastewater and waste gas treatment, energy and water conservation, desulfurization and denitrification dust removal, waste treatment, and waste heat recovery utilization; from management expense details, this study compiles expensed environmental protection expenditures, including pollution discharge fees, environmental protection fees, and vegetation restoration fees, with the sum of both constituting total corporate green investment. Green technological innovation (GTI) is measured as the proportion of corporate green patent applications to total patent applications. Green patent identification is based on the green technology list published by the World Intellectual Property Organization (WIPO) and the International Patent Classification (IPC) green list, primarily encompassing patents in technology fields such as new energy, energy conservation, environmental protection, and resource recycling. Green patents represent innovations that contribute to environmental protection, even in traditional manufacturing sectors. For companies producing electronic elements, green patents may include energy-efficient manufacturing processes, eco-friendly component designs, or waste reduction technologies. While these companies often purchase equipment from suppliers, they simultaneously develop internal process innovations and product improvements that reduce environmental impact, which are captured in their patent portfolios.

To control other factors affecting corporate carbon reduction, this research incorporates a series of corporate characteristic and external environment control variables. Enterprise size (Size) is measured by the natural logarithm of total assets, financial leverage (Lev) is represented by the asset–liability ratio, profitability (ROA) is measured by return on total assets, growth potential (Growth) is represented by the operating revenue growth rate, ownership concentration (Top1) is measured by the shareholding proportion of the largest shareholder, and board size (BS) is represented by the natural logarithm of board members. The study simultaneously controls for firm fixed effects and year fixed effects.

3.3. Empirical Strategy

This research employs difference-in-differences (DID) methodology to identify causal effects of energy rights trading systems on corporate carbon reduction. The benchmark regression model is specified as

$${\mathrm{C}\mathrm{I}}_{\mathrm{i},\mathrm{t}} = {\mathsf{\alpha }}_0+{\mathsf{\alpha }}_1{\mathrm{ERT}}_{\mathrm{i},\mathrm{t}}+\sum _{\mathrm{k}=2}^7{\mathsf{\gamma }}_{\mathrm{k}}{\mathrm{C}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}\mathrm{s}}_{\mathrm{i},\mathrm{t}}+{\mathsf{\mu }}_{\mathrm{i}}+{\mathsf{\lambda }}_{\mathrm{t}}+{\mathsf{\epsilon }}_{\mathrm{i},\mathrm{t}}$$

(3)

where ${\mathrm{C}\mathrm{I}}_{\mathrm{i},\mathrm{t}}$ represents the corporate carbon emission intensity of enterprise i in year t, ${\mathrm{E}\mathrm{R}\mathrm{T}}_{\mathrm{i},\mathrm{t}}$ is the interaction term for energy rights trading systems, and ${\mathsf{\alpha }}_1$ is the core coefficient of primary concern in this research, reflecting the net effect of energy rights trading systems on corporate carbon emission intensity. ${\mathrm{C}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}\mathrm{s}}_{\mathrm{i},\mathrm{t}}$ represents the control variable vector set, including corporate characteristic variables such as enterprise size, financial leverage, profitability, growth potential, ownership concentration, and board size, and ${\mathsf{\gamma }}_{\mathrm{k}}$ denotes the coefficients for these control variables. ${\mathsf{\mu }}_{\mathrm{i}}$ represents firm fixed effects, used to control time-invariant enterprise heterogeneity characteristics; ${\mathsf{\lambda }}_{\mathrm{t}}$ represents time fixed effects, used to control influences of time-varying factors such as macroeconomic environments; and ${\mathsf{\epsilon }}_{\mathrm{i},\mathrm{t}}$ is the random disturbance term. If the coefficient ${\mathsf{\alpha }}_1$ in the model is significantly negative, this indicates that energy rights trading systems significantly reduce corporate carbon emission intensity, thereby promoting corporate carbon reduction.

4. Results and Findings

4.1. Descriptive Statistics

Table 1. Descriptive statistics.

Variable	Observations	Mean	S.D.	Min	Max
CI	4842	0.286	0.218	0.008	1.467
ERT	4842	0.247	0.431	0.000	1.000
GIV	4842	0.019	0.022	0.000	0.134
GTI	4842	0.127	0.169	0.000	0.768
Size	4842	22.340	1.280	20.120	25.890
Lev	4842	0.453	0.162	0.134	0.826
ROA	4842	0.042	0.051	−0.187	0.234
Growth	4842	0.089	0.263	−0.418	1.523
Top1	4842	0.346	0.142	0.098	0.687
BS	4842	2.178	0.183	1.792	2.639

Note: Sample includes 4842 firm-year observations from heavy-polluting listed companies (2013–2023). CI = carbon emission intensity (tons of CO₂/thousand CNY revenue); ERT = energy rights trading system dummy; GIV = green investment ratio; GTI = green innovation ratio; Size = ln(total assets); Lev = leverage ratio; ROA = return on assets; Growth = revenue growth rate; Top1 = largest shareholder proportion; BS = ln(board size).

presents the descriptive statistics for all variables used in our empirical analysis. The sample comprises 4842 firm-year observations from heavy-polluting listed companies across China’s Shanghai and Shenzhen A-share markets from 2013 to 2023. The dependent variable, corporate carbon emission intensity (CI), exhibits substantial variation with a mean value of 0.286 tons of CO₂ per thousand CNY of revenue and a standard deviation of 0.218, indicating considerable heterogeneity in carbon performance across firms and over time, which reflects the diverse technological capabilities, production processes, and environmental management practices within China’s heavy-polluting industries, ranging from relatively cleaner pharmaceutical manufacturing to energy-intensive thermal power generation and steel production. The energy rights trading system dummy variable (ERT) shows a mean of 0.247, indicating that approximately 24.7% of the sample observations are affected by policy implementation, which aligns with the geographical coverage of the pilot provinces and the gradual rollout timeline beginning in 2017. The green investment variable (GIV) demonstrates a mean ratio of 0.0186 relative to total assets, reflecting the moderate scale of environmental protection investments and the cautious approach typically adopted by enterprises when allocating resources to environmental protection activities, while green technological innovation (GTI) shows a mean proportion of 0.127, indicating that approximately 12.7% of total patent applications relate to green technologies, which reflects the emerging focus on environmental innovation within traditional heavy-polluting industries. Among the control variables, firm size (Size) exhibits a mean natural logarithm of total assets of 22.34, which is consistent with the large-scale capital-intensive nature of heavy-polluting industries, while financial leverage (Lev) shows a mean asset–liability ratio of 0.453 and profitability (ROA) demonstrates a mean return on assets of 0.042, reflecting moderate debt levels and modest profitability typical of heavy-polluting industries facing environmental compliance costs and competitive pressures, thereby providing important context for understanding the institutional and economic environment within which energy rights trading systems operate.

4.2. Main Effects

Table 2. Primary estimation results.

Variable	(1)	(2)
	CI	CI
ERT	−0.142 **	−0.128 **
	(−2.29)	(−2.08)
Size		−0.076 ***
		(−3.42)
Lev		0.089 *
		(1.84)
ROA		−0.234 ***
		(−4.17)
Growth		−0.015
		(−0.67)
Top1		0.038
		(1.23)
BS		0.024
		(0.91)
Firm	Yes	Yes
Year	Yes	Yes
Observations	4842	4842
R2	0.736	0.748

Note: ***, **, * indicate significance at 1%, 5%, and 10% levels, respectively, with t-values in parentheses. The dependent variable is CI (tons of CO₂/thousand CNY revenue). ERT = energy rights trading system dummy (1 for pilot provinces after 2017). All regressions include firm and year fixed effects with clustered standard errors.

reports the benchmark regression results regarding the impact of energy rights trading systems on corporate carbon emission intensity. Column (1) presents regression results without control variables, where the coefficient of the energy rights trading system (ERT) is −0.142, significant at the 5% level, indicating that implementation of energy rights trading systems significantly reduces corporate carbon emission intensity. In Column (2), after controlling for relevant corporate characteristic variables, the coefficient of energy rights trading systems remains −0.128, still significantly negative at the 5% level, with the absolute value of the coefficient slightly decreased but maintaining statistical significance, demonstrating that after controlling for relevant corporate characteristics, energy rights trading systems can still effectively promote corporate carbon reduction.

The economic magnitude of these effects merits detailed discussion. The coefficient of −0.128 in Column (2) indicates that energy rights trading systems reduce carbon emission intensity by 0.128 tons of CO₂ per thousand CNY of revenue compared to the control group. Given that the sample mean of carbon emission intensity is 0.286 tons of CO₂ per thousand CNY of revenue, this absolute reduction represents approximately 44.8% of the baseline mean. This substantial magnitude demonstrates that energy rights trading systems generate economically meaningful environmental benefits while maintaining statistical robustness across different model specifications.

4.3. Sensitivity Analysis

A fundamental assumption of the difference-in-differences methodology is that treatment and control groups would have followed parallel trends in the absence of policy intervention. Figure 2 displays the dynamic changes in policy effects across years before and after the energy rights trading system implementation. The results demonstrate that prior to policy implementation (2013–2016), the trends in carbon emission intensity changes between treatment and control group enterprises remain essentially parallel, with coefficients across all pre-treatment years being statistically insignificant and fluctuating around zero. This provides strong evidence that the parallel trends assumption is satisfied. Following policy implementation (beginning in 2017), coefficients gradually become negative, and statistical significance strengthens over time. The coefficients represent estimated treatment effects measured in tons of CO₂ per thousand CNY of revenue reduction relative to the control group baseline. For instance, the coefficient of approximately −0.125 at t + 1 indicates that firms in pilot provinces experienced a reduction in carbon emission intensity of 0.125 tons of CO₂ per thousand CNY one year after policy implementation compared to non-pilot firms. The temporal pattern reveals that emission reduction effects strengthen over time, reaching approximately −0.388 tons of CO₂ per thousand CNY by the final observation period, suggesting cumulative policy learning and adaptation effects as enterprises require time to adjust their production processes, investment strategies, and technological capabilities.

To further eliminate potential interference from other unobservable factors and validate that our results are not driven by spurious correlations, this study conducts comprehensive placebo tests using random treatment assignment. Figure 3 displays the results of 2000 random treatment group allocation tests, where we randomly reassign treatment status to provinces and re-estimate our baseline model. The actual estimated coefficient (indicated by the red vertical line at approximately −0.128 tons of CO₂ per thousand CNY) is located in the left tail portion of the random allocation coefficient distribution, with the vast majority of randomly allocated coefficients distributed around zero values. Specifically, less than 5% of the randomly generated coefficients are as negative as our actual estimate, providing strong evidence against the null hypothesis that our results are due to chance. The kernel density distribution of placebo coefficients follows a normal distribution centered around zero, with our actual coefficient falling well outside the 95% confidence interval of the placebo distribution, confirming that the observed emission reduction effects represent genuine policy impacts rather than spurious correlations or unobserved confounding factors.

Additionally,

Table 3. Sensitivity check outcomes.

Variable	(1)	(2)	(3)
	CI	CI	CI
ERT	−0.115 *	−0.122 **	−0.113 *
	(−1.89)	(−2.15)	(−1.84)
CETS			−0.102 **
			(−2.31)
Controls	Yes	Yes	Yes
Firm	Yes	Yes	Yes
Year	Yes	Yes	Yes
Observations	3476	4120	4842
R2	0.752	0.745	0.741

Note: **, * indicate significance at 5%, and 10% levels, respectively, with t-values in parentheses. Robustness tests: (1) PSM-DID; (2) rolling window 2014–2022; (3) controlling for carbon trading systems. CI = carbon emission intensity (tons of CO₂/thousand CNY revenue); ERT = energy rights trading system dummy; CETS = carbon trading system dummy. All regressions include controls for firm and year fixed effects.

reports results from other robustness tests. Column (1) employs propensity score matching-difference-in-differences (PSM-DID) methodology to address sample selection bias issues, with the coefficient of energy rights trading systems in the matched sample being −0.115, significant at the 10% level, and consistent with benchmark result directions. Column (2) adopts a rolling window approach, adjusting the sample time window from 2013–2023 to 2014–2022 for testing, with results showing that policy effects remain robust under the shortened time window.

Considering potential influences from other contemporaneous environmental policies, this study pays particular attention to the effects of carbon emissions trading systems. As another important market-based environmental regulatory instrument, carbon emissions trading systems similarly aim to promote corporate carbon emission reduction through market mechanisms, potentially generating synergistic or substitutive effects with energy rights trading systems. Accordingly, this study constructs a carbon emissions trading system dummy variable (CETS), taking value 1 for enterprises in pilot provinces implementing carbon emissions trading systems in corresponding years and 0 otherwise. As shown in Column (3) of Table 3, after controlling for carbon emissions trading system influences, the coefficient of energy rights trading systems is −0.113, significant at the 10% level, indicating that energy rights trading systems possess independent emission reduction effects and remain robust after excluding other contemporaneous environmental policies.

4.4. Cross-Sectional Variations

To deeply understand the differentiated impacts of energy rights trading systems on enterprises with varying characteristics, this study conducts heterogeneity analysis across two dimensions: financing constraints and supply chain concentration. Financing constraints directly affect enterprise capabilities to respond to energy policy shocks, as green transformation and low-carbon technology investments often require substantial financial support; consequently, enterprises with different levels of financing constraints may adopt different response strategies when facing energy rights trading systems. This study employs the KZ index to measure enterprise financing constraint levels, dividing samples into low financing constraint and high financing constraint groups based on annual industry medians. As shown in Columns (1) and (2) of

Table 4. Subgroup analysis findings.

Variable	(1)	(2)	(3)	(4)
	Low Financing Constraints	High Financing Constraints	High Supply Chain Concentration	Low Supply Chain Concentration
ERT	−0.167 **	−0.058	−0.145 **	−0.089
	(−2.28)	(−0.84)	(−2.19)	(−1.33)
Controls	Yes	Yes	Yes	Yes
Firm	Yes	Yes	Yes	Yes
Year	Yes	Yes	Yes	Yes
Observations	2156	2686	2178	2664
R2	0.753	0.744	0.751	0.746
Inter-Group Coefficient Difference Test p-Value	0.084		0.092

Note: ** indicates significance at 5% level, respectively, with t-values in parentheses. Heterogeneity analysis by financing constraints (KZ index) and supply chain concentration. CI = carbon emission intensity (tons of CO₂/thousand CNY revenue); ERT = energy rights trading system dummy. Groups split by annual industry medians. All regressions include controls for firm and year fixed effects.

, the coefficient of energy rights trading systems in low financing constraint enterprises is −0.167, significant at the 5% level, whereas in high financing constraint enterprises, the coefficient is −0.058, which is statistically insignificant. The p-value for inter-group coefficient difference tests is 0.084, indicating significant differences between groups at the 10% level. This result aligns with theoretical expectations: low financing constraint enterprises, possessing relatively abundant capital, can actively invest in energy conservation and emission reduction technologies and green production equipment, achieving carbon reduction through technological upgrades and process improvements; conversely, high financing constraint enterprises, due to capital constraints, are more likely to adopt passive strategies such as production reduction or investment delays, making effective carbon reduction difficult to achieve.

Supply chain concentration reflects enterprise bargaining power and cost-shifting capabilities within supply chains, characteristics of significant importance under energy policy shocks. High supply chain concentration enterprises typically establish stable cooperative relationships with a small number of core suppliers or customers, possessing strong bargaining power and the ability to shift portions of environmental protection costs to upstream and downstream enterprises, thereby maintaining operational stability while sustaining environmental protection investments. This study measures supply chain concentration using the average of procurement proportions from the top five suppliers and sales proportions to the top five customers, similarly grouping based on annual industry medians. Results shown in Columns (3) and (4) of Table 4 indicate that in high supply chain concentration enterprises, the coefficient of energy rights trading systems is −0.145, significant at the 5% level; in low supply chain concentration enterprises, the coefficient is −0.089, which is statistically insignificant. The p-value for inter-group difference tests is 0.092, significant at the 10% level. This demonstrates that enterprises with stronger supply chain bargaining capabilities can better cope with cost pressures from energy rights trading systems, achieving effective carbon reduction through cost shifting and resource integration, while enterprises with weaker bargaining capabilities face greater policy adjustment difficulties.

4.5. Mechanism Analysis

To reveal the specific mechanisms through which energy rights trading systems influence corporate carbon reduction, this study further examines the transmission effects of green investment and green technological innovation. From a theoretical mechanism perspective, energy rights trading systems, by establishing upper limits on energy usage quotas, essentially construct a form of “energy scarcity” constraint; this constraint mechanism compels enterprises to reexamine their energy usage efficiency and production methods. Under quota constraints, enterprises face two choices: either pay higher energy costs to maintain the status quo or improve energy usage efficiency through technological upgrades and investment renovations. Obviously, the latter represents a more sustainable and economically viable long-term strategy, which constitutes the core logic of policy design.

The empirical results in

Table 5. Mechanism test results.

Variable	(1)	(2)	(3)	(4)
	GIV	CI	GTI	CI
ERT	0.077 *	−0.116 *	0.045 *	−0.119 *
	(1.76)	(−1.89)	(1.68)	(−1.93)
GIV		−0.156 **
		(−2.34)
GTI				−0.198 **
				(−2.41)
Controls	Yes	Yes	Yes	Yes
Firm	Yes	Yes	Yes	Yes
Year	Yes	Yes	Yes	Yes
Observations	4842	4842	4842	4842
R2	0.623	0.752	0.587	0.754

Note: **, * indicate significance at 5%, and 10% levels, respectively, with t-values in parentheses. Mechanism analysis via green investment and innovation. GIV = green investment ratio; GTI = green innovation ratio; CI = carbon emission intensity (tons of CO₂/thousand CNY revenue); ERT = energy rights trading system dummy. All regressions include controls for firm and year fixed effects.

validate this theoretical inference. Regarding the transmission pathway of green investment, energy rights trading systems significantly promote corporate green investment behavior, reflecting rational choices by enterprises when facing rising energy cost pressures. Corporate increases in green investment primarily manifest in two aspects: first, capitalized environmental protection equipment investments, such as fixed asset investments in waste gas treatment facilities and energy-saving equipment, and second, expensed environmental protection expenditures, such as daily operational costs, including pollution discharge fees and environmental treatment expenses. This increase in investment behavior not only directly improves enterprise energy usage structures but, more importantly, establishes capability foundations for enterprises to address future stricter environmental regulations. When enterprises simultaneously face energy rights trading system constraints and engage in green investment, their carbon reduction effects are further strengthened, indicating that green investment indeed plays an important transmission role.

The empirical evidence supporting altered innovation orientations comes from our regression results. Table 5 Column (3) shows that energy rights trading systems significantly increase the proportion of green patents (coefficient 0.045, p < 0.10), indicating a shift in R&D focus toward environmental technologies rather than simply increasing total innovation. The temporal patterns in Figure 2 demonstrate that this effect strengthens over time, consistent with strategic reorientation rather than immediate compliance responses. Energy rights trading systems not only promote current-period green technological innovation activities but, more importantly, alter enterprise innovation orientations and R&D strategies. Under traditional production models, corporate technological innovation often emphasizes improving production efficiency or reducing production costs, with environmental friendliness not being a primary consideration. However, implementation of energy rights trading systems makes energy efficiency an important component of enterprise competitiveness, prompting enterprises to allocate more R&D resources toward green technology fields. This transformation in innovation orientation is reflected not only in changes to patent application structures but also in shifts in enterprise development philosophies. From a long-term perspective, improvements in green technological innovation capabilities will provide sustained competitive advantages for enterprises in future low-carbon economic transitions, representing the profound value of policy effects.

Notably, while both green investment and green technological innovation transmission mechanisms play important roles, they differ in operational methods and temporal effectiveness. Green investment manifests more as direct short-term investment with rapid results, enabling enterprises to achieve emission reduction targets in relatively short periods through purchasing advanced equipment or improving production processes. Conversely, green technological innovation represents a more complex and long-term process, requiring longer time cycles from R&D investment to technological breakthroughs to industrial applications; however, once breakthroughs are achieved, they generate more lasting emission reduction effects. This mechanism difference also explains why emission reduction effects of energy rights trading systems are relatively moderate in early policy implementation stages but gradually strengthen over time, precisely because green technological innovation effects require longer periods to fully manifest.

5. Discussion

Through difference-in-differences methodology, this study empirically examines the impacts of energy rights trading systems on corporate carbon reduction and their transmission mechanisms, finding that these systems can significantly reduce corporate carbon emission intensity while operating through two pathways: green investment and green technological innovation. This finding establishes beneficial dialogue and complementarity with existing carbon emissions trading system research. Our study confirms the emission reduction effectiveness of market-based environmental regulatory instruments, similar to research conclusions based on China’s regional carbon market pilots, though with certain differences in specific emission reduction magnitudes, possibly reflecting the uniqueness of different policy instruments in constraint objectives and implementation mechanisms [14]. Notably, this study finds that emission reduction effects of energy rights trading systems are relatively moderate, forming an interesting contrast with perspectives that carbon emissions trading systems may initially inhibit green innovation, suggesting that energy rights trading systems may better balance relationships between short-term compliance costs and long-term innovation incentives in policy design [17]. Furthermore, heterogeneity analysis results indicate that enterprises with lower financing constraints and stronger supply chain bargaining capabilities indeed respond better to energy rights trading system implementation, highly consistent with existing findings regarding enterprise characteristics influencing policy effects, further validating the crucial role of enterprise resource endowments in environmental policy implementation [26].

From mechanism transmission pathway perspectives, the dual transmission mechanisms of green investment and green technological innovation identified in this study enrich existing theoretical understanding regarding operational principles of market-based environmental regulation. This study confirms that energy rights trading systems can stimulate corporate green technological innovation activities, echoing research on carbon emissions trading systems promoting low-carbon technological innovation, though both systems exhibit differentiated characteristics in innovation incentive temporality and intensity [24]. More importantly, this study finds that green investment, as a relatively rapid-effect transmission channel, forms complementary relationships with green technological innovation as a long-term mechanism, filling gaps in existing research lacking in-depth analysis of specific transmission mechanisms while providing more complete explanatory frameworks for understanding how energy rights trading systems achieve coordinated development between economic and environmental performance [18]. Simultaneously, this study’s findings regarding differences in temporality and sustainability between the two transmission mechanisms provide micro-level mechanism validation for theoretical perspectives on promoting corporate green investment through alleviating external pressures and internal incentives, indicating that full policy effect release requires consideration of temporal differences and synergistic effects across different transmission pathways [25].

6. Conclusions

Based on panel data from heavy-polluting listed companies on Shanghai and Shenzhen A-share markets, this study employs difference-in-differences methodology to thoroughly explore the impact effects and operational mechanisms of energy rights trading systems on corporate carbon reduction behavior. Empirical results demonstrate that energy rights trading systems significantly reduce corporate carbon emission intensity by 12.8%, effectively stimulating enterprise emission reduction motivation through constructing energy scarcity constraints. Mechanism analysis reveals that these systems primarily operate through two transmission pathways: green investment and green technological innovation, with green investment serving as a short-term strategy capable of rapidly achieving emission reduction objectives, while green technological innovation provides enterprises with long-term sustainable emission reduction capabilities. The study further discovers significant heterogeneity in policy effects across different enterprises, with enterprises possessing lower financing constraints and stronger supply chain bargaining capabilities demonstrating superior policy response capabilities, reflecting the important moderating role of enterprise resource endowments in environmental policy implementation. These findings not only enrich theoretical understanding of market-based environmental regulatory mechanisms but also provide empirical evidence that energy rights trading systems, as consumption-based constraint instruments, generate distinct pathways and effects compared to emission-based carbon trading systems.

Based on these research findings, this paper proposes differentiated policy recommendations addressing the identified mechanisms and heterogeneity patterns. First, policymakers should optimize the energy rights trading system design by refining quota allocation methods and enhancing price discovery mechanisms to provide enterprises with clearer emission reduction incentive signals, thereby strengthening the direct effects documented in our analysis. Second, given the significant role of financing constraints in moderating policy effectiveness, governments should establish differentiated support systems, including green development funds and preferential financing arrangements for enterprises with higher financing constraints, enabling broader participation in emission reduction activities. Third, recognizing the complementary roles of green investment and technological innovation transmission mechanisms, policy frameworks should encourage short-term corporate green investment while simultaneously strengthening long-term support for green technological innovation through tax incentives and R&D subsidies, ensuring sustained release of emission reduction effects. Finally, considering the heterogeneous responses based on supply chain characteristics, regulatory authorities should develop tailored implementation strategies that account for enterprise bargaining power differences, potentially offering enhanced technical assistance and coordination support for enterprises with weaker supply chain positions to ensure equitable policy outcomes across different enterprise types.

Despite these contributions, this study acknowledges several limitations that present opportunities for future research. First, the analysis focuses exclusively on listed companies in heavy-polluting industries, which may limit the generalizability of findings to small and medium enterprises or other sectors that could also benefit from energy rights trading systems. Second, while the study examines an 11-year period spanning policy implementation, the relatively short post-treatment observation window constrains our ability to assess long-term equilibrium effects and potential policy adaptation behaviors. Third, the measurement of green investment and technological innovation, while comprehensive, relies primarily on disclosed financial data and patent applications, which may not fully capture all forms of environmental improvements, informal innovation activities, or non-patentable process optimizations that enterprises undertake in response to energy constraints. Additionally, the green patent classification, though based on internationally recognized WIPO standards, may not perfectly capture all environmentally beneficial innovations, particularly those involving incremental improvements or cross-sector applications. Finally, although the difference-in-differences methodology provides robust causal identification, the analysis cannot completely rule out potential spillover effects between pilot and non-pilot regions or account for anticipatory behaviors that enterprises might have adopted prior to formal policy announcement. Future research could address these limitations by extending the analytical framework to broader enterprise samples, incorporating longer time horizons, developing more comprehensive measures of environmental performance, and exploring dynamic policy interactions across different regulatory jurisdictions.