Can Government Environmental Attention Improve Corporate Carbon Emission Reduction Performance?—Evidence from China A-Share Listed Companies with High-Energy-Consumption

Abstract

Government support for achieving corporate carbon emission reduction is crucial not only for sustainable business development, but it also holds strategic importance for China to achieve its “dual-carbon” goals. This article empirically explores the impact and underlying mechanisms of government environmental attention (GEA) on corporate carbon emission reduction performance (CEP), using a sample of A-share listed companies with high energy consumption from 2009 to 2020. The results show that GEA can improve corporate CEP. A heterogeneity analysis found that this effect is more pronounced in heavily polluting industries, small and medium-sized enterprises (SMEs), and companies located in the eastern regions of the country. A mechanism analysis suggested that GEA can improve corporate CEP by strengthening internal green technological innovation capabilities and attracting attention from external analysts and media. These research conclusions guide corporate carbon emission reduction practices and offer empirical evidence for the government in formulating regulatory policies for carbon reduction.

1. Introduction

Since the reform and opening-up, the government has consistently prioritized economic development [1]. China has sustained long-term high-speed economic growth, generating significant social wealth. In 2010, China surpassed Japan in GDP, becoming the second-largest economy in the world, and it continues to maintain this position at a high level. However, this economic prosperity is accompanied by environmental degradation. The extensive mode of economic growth, despite producing seemingly impressive data, imposes environmental sacrifices [2]. Greenhouse gas emissions are on the rise, and issues like high energy consumption and high carbon emissions persist [3]. This not only intensifies the burden on the ecological environment but also contradicts the ecological civilization concept that green mountains and clear waters are as valuable as mountains of gold and silver, as put forth by General Secretary Xi Jinping [4,5]. This poses a bottleneck that hampers the sustainable development of the economy and society, severely constraining the sustainable development of enterprises.

Rapid economic growth-induced environmental degradation has increased the government’s focus on the environment. China pledged at the World Climate Conference to cap its carbon peak by 2030 and achieve carbon neutrality by 2060 at the latest. The 18th Party Congress, for the first time, incorporated ecological civilization construction into the party constitution, raising the significance of the ecological environment to a new level. The enactment of stringent environmental laws and standards has propelled enterprises to adopt clean production technologies, encouraging the development of green energy to mitigate pollution and resource waste. The 20th Party Congress report reinforced the government’s responsibility in promoting green development, enhancing environmental pollution control, and persistently striving to safeguard the blue sky, clear waters, and pure land. It advocated for harmonious coexistence between humans and nature. Governments should commit to promoting sustainable development, formulating and implementing ecological protection plans, and enhancing support for environmental protection investments to facilitate the coordinated development of the economy and the environment [6]. Government-introduced environmental protection systems and policies, along with reinforced environmental education and awareness campaigns, have heightened public consciousness regarding the significance of environmental protection. This has prompted various societal sectors to actively engage in environmental protection activities, indicating the local government’s emphasis on the local environment. The signals emitted will influence enterprises’ decision-making, indirectly or directly affecting their carbon reduction initiatives.

The report of the 20th Party Congress underscores the need for the government to coordinate efforts in reducing carbon emissions, controlling pollution, expanding green initiatives, and promoting growth. It advocates for resource-efficient, environmentally friendly, and low-carbon development, actively and prudently advancing the peak and neutrality of carbon. The report calls for a planned and step-by-step implementation of actions to achieve the carbon peak and active participation in global governance to address climate change. However, the substantial emissions of greenhouse gases have resulted in irreversible global climate warming, and carbon reduction is considered an effective approach to mitigate this issue [7]. Governments worldwide are also actively pursuing their own approaches to reduce carbon emissions [7]. Being the world’s largest emitter of carbon dioxide, China is also actively engaged in this endeavor [8]. China has introduced the “dual carbon” goals, aiming to reach both peak carbon emissions and carbon neutrality. Therefore, with the growing environmental concerns of the central government, local governments must increasingly take on more environmental governance responsibilities. They assign emission reduction tasks to local high-energy-consuming enterprises, impose restrictions on large-scale enterprise production, and mandate improvements in production processes and the acceleration of green innovation. Furthermore, besides command-and-control environmental regulation policies, the government will offer financial subsidies and policy support to encourage SMEs to engage in carbon reduction activities [8]. These subsidies can optimize operational processes and improve resource utilization efficiency and environmental investment efficiency [9]. It will eventually improve corporate carbon performance with limited resources. Governmental focus on environmental issues can also encourage them to increase their awareness of environmental protection and implement corresponding measures, thereby fostering a stronger sense of environmental responsibility among these enterprises. Through these efforts, they can effectively reduce carbon emissions and enhance carbon performance despite limited resources. Despite the implementation of positive measures, such as constructing a carbon market and releasing regulations for carbon emissions trading management in recent years [10,11], the persistent rapid economic growth and continuous rise in energy demand have maintained a structure in certain industries characterized by high energy consumption and high carbon emissions. The energy consumption heavily depends on the distribution of fossil fuels, particularly coal, which presents significant challenges for carbon governance. In 2019, fossil fuels represented 85% of China’s total energy consumption. Regarding the sectoral breakdown of energy consumption, the industrial sector comprised almost 70%, with high energy consumption industries representing 80% of energy usage within this sector. This situation persists as high-energy-consuming enterprises excessively rely on carbon energy, exhibit a relatively lagging level of production technology, and possess low production process efficiency. These factors exacerbate energy over-consumption, presenting a formidable challenge for current carbon reduction efforts [12]. As major contributors to greenhouse gas emissions, high-energy-consuming enterprises should be a focal point for carbon emissions management. Consequently, researching carbon reduction issues at the enterprise level, especially for high-energy-consuming enterprises, becomes particularly crucial. Therefore, the practice of a green, low-carbon development philosophy by high-energy-consuming enterprises, enhancing carbon reduction performance, not only allows businesses to achieve a “win-win” scenario of competitiveness and environmental protection but is also a critical factor concerning the contradiction between social–economic development and environmental protection.

To address this, this study empirically investigates the impact and mechanism of GEA on the carbon reduction performance of A-share listed companies with high energy consumption from 2009 to 2020, focusing on key issues related to China’s “dual-carbon” goals. This study makes three contributions to the existing literature. Firstly, current research on GEA primarily focuses on its environmental and social impacts, along with measurement methods [8,13,14,15], while largely neglecting its effect on corporate carbon reduction performance. Therefore, this paper examines the impact of GEA on corporate carbon emission reduction performance, complementing the existing literature in the field and expanding the research perspective on CEP. Furthermore, leveraging text mining methods, we extract extensive environmental-related information from government annual work reports, providing a comprehensive insight into the carbon reduction behavior of enterprises under the influence of GEA, as opposed to the combined effect of macro-level environmental regulations on carbon mitigation. Secondly, we examine the mechanisms through which GEA impacts corporate CEP, considering internal green technology innovation and external analysts and media attention. This enriches current research on the influence of GEA on corporate carbon governance and offers a fresh perspective on how the government can aid corporate carbon reduction efforts in pursuit of dual carbon goals. Thirdly, by examining the influence of GEA on corporate CEP, we offer micro theoretical backing for the implementation of strategies like peak carbon emissions and carbon neutrality. Additionally, this exploration provides insight into the role of local governments in China’s economic growth and corporate environmental governance.

2. Literature Review

2.1. Government Environmental Attention

Originally rooted in psychology, attention served as a cornerstone concept, delineating the mental mechanisms by which individuals focus their consciousness on particular objects [16]. Nobel laureate Simon later integrated this concept into organizational management decision-making, anchoring it within the framework of traditional economic theory. Moreover, he argued that managers’ attention is inherently constrained compared with the abundance of external information, a limitation attributed to their bounded rationality and cognitive capabilities [17]. Therefore, he proposed that “Attention” involves managers selectively focusing on specific information while intentionally disregarding other information [17]. The distribution of decision-makers’ attention results in diverse organizational strategies and resource allocations, influencing the behavioral choices of the organization [8]. Subsequently, Jones introduced the concept of attention to the field of government management [16]. The extent of government attention to the environment mirrors the situational context of its decision-making environment. It is a conscious process where decision-makers prioritize significant information related to environmental protection, among other factors, while intentionally ignoring other behavioral information [13]. This also signifies the government’s decision information process involving perception, encoding, interpretation, and focus. Presently, research on GEA primarily focuses on its environmental and social impacts, along with measurement methods.

First, government environmental audits can significantly promote regional pollution control, improving the effectiveness of coordinated pollution reduction efforts [13,18]. GEA also affects the efficiency of land resource allocation in cities [14]. A higher level of GEA leads to a more sustainable utilization rate of land resources, contributing to local urban sustainable and high-quality development. Second, according to the market pressure mechanism, heightened government attention on the environment will strengthen media attention [19,20] and raise public awareness [15]. It stimulates the market demand for green products and investments in environmentally friendly companies, ultimately improving the quality of the social environment. Third, based on sustainable development theory [21], enhanced governmental scrutiny of environmental issues plays a vital role in the governance and supervision of ecological environments in China. This contributes to improving the quantity and quality of corporate green innovation [22,23]. It can also provide valuable insights into understanding how organizations respond to GEA and how internal factors within the organization influence CEP. This, in turn, diminishes negative impacts on public health and fosters sustainable development within enterprises [6]. Finally, in terms of the measurement approaches for GEA, initially, many scholars predominantly utilized proxy measures, such as the number of environmental protection personnel or the proportion of environmental investment, to assess GEA [8]. However, this methodology only provides a limited assessment of the government’s environmental focus and does not offer a comprehensive evaluation of its environmental governance capabilities [8,24]. With the rise of text mining technology, an increasing number of scholars have begun to extract word frequencies related to environmental protection, low-carbon, green, pollution, etc., from government annual work reports for analysis [13]. This approach aims to construct comprehensive indicators to measure GEA. Recent studies have also employed this approach to investigate the influence of GEA on air pollution and green innovation, thus affirming the validity of text analysis methodologies [13,25].

2.2. Corporate Carbon Reduction Performance

The introduction of “dual-carbon” goals has made carbon reduction a central focus in current research, garnering widespread attention and in-depth discussions from scholars globally [26,27]. In the realm of environmental protection, local government pressures [28] and environmental audits [29] contribute significantly to improving corporate carbon reduction performance. Central environmental inspections and “look-backs” can significantly reduce the emissions of pollutants like AQI and PM2.5 [30]. This, in turn, enhances the carbon reduction performance of enterprises to a certain extent [31]. Audits of natural resource assets by leading cadres, when carried out strategically, contribute to advancing green technology, upgrading industrial structures, and consequently reducing overall carbon emissions in a region [32,33]. Furthermore, public participation in environment governance, manifested through means like petitions, positively contributes to reducing carbon intensity [34,35]. Media investigation exposes corporate violations, promotes the development of green patented inventions, and effectively restrains carbon emissions from industrial enterprises [36]. In terms of implementing carbon policies, carbon emission trading has significantly improved corporate carbon reduction performance [37] and the efficiency of the carbon market [10,11,38]. The implementation of low-carbon city pilot policies can greatly enhance low-carbon technological innovation through the reinforcement of the financial market and the mitigation of financing constraints [39]. Implementing green fiscal policies can achieve carbon reduction effects by reducing energy consumption, lowering unit GDP energy consumption, promoting industrial decarbonization, and enhancing green technological innovation [40,41]. Moreover, some scholars have also confirmed the influence of green finance [42], government environmental concern [8], and corporate governance capability [43] on carbon emissions. However, they either approach their research from a provincial-level macro perspective or overlook the micro-level high-energy-consuming enterprises. Therefore, in this study, we focus on high-energy-consuming enterprises as the research sample, specifically examining the impact of GEA on corporate CEP. It can provide more specific research significance for studies in the field of carbon reduction.

In summary, previous research has produced valuable findings, offering insights and inspiration for the current study. However, there remains space for additional exploration. Through the above analysis and review, the current academic emphasis on GEA is predominantly on urban, societal, and corporate levels. However, few studies have investigated its influence on corporate CEP, failing to establish a robust connection between the two and lacking further exploration of their relationship. Therefore, this paper aims to depart from the perspective of GEA, exploring its effects and mechanisms on corporate carbon reduction performance. This contributes not only to enriching and refining related research but also to fostering a deeper comprehension of the role of local governments in environment protection and low-carbon development.

3. Research Hypothesis

3.1. Government Environmental Attention and Corporate Carbon Emission Reduction Performance

Government attention to the environment mirrors the emphasis local officials place on environmental protection and related matters [25,44]. The government’s shift in focus will directly affect decisions made by both the government and corporations, thereby impacting the carbon emission reduction performance of enterprises. First, considering government resource effects, the government’s environmental attention creates incentives, enabling companies to access government resources [45]. According to resource-based theory, it focuses on utilizing internal resources and capabilities within the enterprise [46]. When the government increases environmental oversight and implements relevant policies, enterprises are required to enhance their carbon reduction capabilities through innovation and optimized resource allocation. This leads to a competitive advantage, driving companies to improve their carbon emission behavior and thereby influencing corporate CEP. As a scarce resource, the government’s attention to the ecology sends a signal to businesses through the “invisible hand”, encouraging them to participate in environmental governance. Companies actively adhering to government environmental policies, including pollution and carbon reduction, are more likely to earn government favor. This favor could result in benefits like tax reductions, credit support, and environmental subsidies, contributing to the sustainability of carbon reduction efforts and resolving the financing dilemma for small- and medium-sized enterprises pursuing carbon reduction, which is beneficial for improving CEP.

Second, considering government officials’ promotion, GEA constrains businesses by imposing stringent environmental regulations, restricting corporate production, thereby improving carbon reduction performance. Over an extended duration, the promotion of Chinese officials has heavily relied on local GDP growth rates. As a result, when local governments prioritize economic growth, they often neglect environmental issues [47,48]. However, the worsening severity of ecological problems has attracted the focus of the central government. In 2018, the establishment of the Ministry of Ecology and Environment at the central level, along with the creation of ecological and environmental departments in various provinces, further intensified the government’s attention on the environment [49]. To comply with environmental emission reduction directives from higher authorities, local governments must delegate these responsibilities to enterprises operating within their jurisdiction, particularly targeting those with high energy consumption, as they constitute the principal sources of local carbon emissions. Confronted with environmental penalties or incentives imposed by local authorities, enterprises may be encouraged to implement cleaner production methods, allocate resources to green technologies, integrate environmentally sustainable management strategies, and enhance CEP [8]. Previous research has demonstrated that governmental regulations regarding green credit exert an impact on the overall green productivity of enterprises [50]. This regulation can stimulate enterprises to actively participate in green research and development activities, thus improving their CEP.

Third, based on sustainable development theory, proactive responses to government environmental demands and the improvement of carbon reduction performance contribute to long-term sustainable development. This theory suggests that economic development, social progress, and environmental protection should be mutually reinforcing to ensure the sustainable utilization of resources over the long term, while also taking into account social equity and ecological equilibrium. Increased government attention to the environment directs market demand toward more eco-friendly products and services [51]. Therefore, enterprises with high energy consumption actively involve themselves in carbon reduction endeavors which not only satisfies consumer demands for eco-friendly products but also earns consumer approval, secures market acknowledgment, and enhances competitive edges. According to the theory of sustainable development, environmental protection is essential. It can motivate corporations to voluntarily decrease carbon emissions, reinforcing their sustainable development capabilities. Moreover, sustainable development theory underscores the importance of CSR [52]. Encouraged by governmental environmental initiatives, corporations integrate environmental responsibilities into their corporate culture, which promotes a culture of carbon reduction that engages all employees. This enhances employees’ environmental awareness and sense of responsibility, thereby fostering overall carbon reduction performance within the enterprise. It also emphasizes long-term interests. Driven by governmental environmental attention, businesses place greater emphasis on long-term strategic planning by integrating carbon reduction into their development strategies. This establishes a mechanism for continual improvement, ensuring enhanced long-term environmental performance. Therefore, we expect the following:

Hypothesis 1: 

GEA improves enterprises’ carbon reduction performance.

3.2. Government Environmental Attention, Green Technological Innovation, and Carbon Emission Reduction Performance

In China, enterprises closely align their actions with governmental activities, considering the government a crucial guidepost in their development decisions [15]. Governments play a crucial role in enhancing corporate green innovation [53]. On one hand, with the growing prominence of GEA, there is an increased probability of more stringent environmental regulations [54]. Markets will shift towards environmentally friendly products as well. To prevent market obsolescence, high-energy-consumption enterprises will refine resource allocation and innovate green technologies to enhance production processes, ultimately boosting, and thereby enhancing, CEP. On the other hand, governments will also boost environmental subsidies for enterprises and allocate more fiscal resources [55], which can help ease financing constraints for enterprises. These resources provided by governments can encourage green innovation within companies, leading to a compensatory effect on resources that promotes the development of green technologies [22]. It will eventually lead to increased investment in green innovation by companies, thereby improving production technologies and enhancing CEP. Based on the above analysis, we propose the following:

Hypothesis 2: 

GEA can improve enterprises’ carbon emission reduction performance by strengthening the ability of green technological innovation.

3.3. Government Environmental Attention, External Analyst Attention, and Carbon Emission Reduction Performance

Analysts, due to their professional expertise and responsibilities, play a vital role as information intermediaries in the capital market, ensuring the effective execution of their supervisory role [56]. Within the current social context promoting environmental conservation, environmental challenges stemming from climate change are increasingly recognized as threats to human survival and development, steadily gaining prominence in societal discourse [57,58]. With the increasing attention on environmental issues, external analysts are increasingly focusing on production issues within local companies [59], examining, consolidating, and analyzing their production irregularities, which impacts their CEP. Specifically, acting as crucial information channels in the market, analysts can offer the government advice on current and potential carbon reduction policies [60], assisting in policy development and improving businesses’ carbon reduction strategies. Additionally, analysts can provide accurate carbon emission data and trend analyses, aiding the government in monitoring companies’ carbon reduction progress and evaluating environmental and economic risks associated with carbon emissions, ensuring compliance with regulations and standards. They facilitate communication and collaboration among government, businesses, and stakeholders [61], collectively seeking optimal carbon reduction practices and solutions to enhance CEP. Thus, we expect the following:

Hypothesis 3: 

GEA can improve enterprises’ carbon emission reduction performance by increasing the attention of external analyst.

3.4. Government Environmental Attention, Media Attention, and Carbon Emission Reduction Performance

The Effective Supervision Mechanism suggests that the media, serving as a crucial external supervisory governance mechanism beyond legal frameworks, can expose corporate environmental violations or improper administration [62]. It can prompt companies to alter their production behaviors and improve transparency and compliance, thereby impacting CEP.

First, media attention demonstrates a focusing effect [63]. As local GEA intensifies, local media are directed to intensively report on the issue in a short time-frame, focusing societal attention. Through media analysis, interpretation, and commentary on related events, the cost for the public to access information is reduced, guiding public viewpoints and value judgments on the public event. Consequently, media attention and public opinion influence carbon-emitting behaviors of local high-energy-consuming enterprises. Faced with negative media coverage, enterprises actively respond to public concerns, adopting emission reduction measures to enhance their carbon emission reduction performance. Second, media attention exhibits an amplification effect [64]. Media supervision, by exposing a company’s non-compliant production, disseminates event information throughout society, magnifying various violations and excessive carbon emissions. Consequently, when enterprises face media exposure, to avoid negative public perception, they become more attentive to their carbon-emitting behaviors, improving carbon emission reduction performance. Third, media attention encompasses a reputation effect [65]. Media reports can influence how the public views companies [66]. If a company is portrayed negatively in the media as harming the environment or emitting carbon irresponsibly, its reputation suffers. Public opinion frequently pushes governments to act [67]. Therefore, to maintain public trust and political standing, governments may intensify the monitoring of corporate carbon emissions or implement stricter environmental policies, which impacts companies’ CEP. Therefore, we propose the following:

Hypothesis 4: 

GEA can improve enterprises’ carbon emission reduction performance by increasing the attention of media.

4. Research Method

4.1. Sample Selection and Data Processing

To validate the hypotheses, we selected listed companies with high energy consumption from the six major industries outlined in the 2010 Statistical Report on National Economic and Social Development. Additionally, the 2020 CSRC Guidelines for Industry Classification of Listed Companies were used for industry matching. This study focused on China’s A-share listed high-energy-consuming companies from 2009 to 2020. The data underwent the following treatments: (1) exclusion of ST and ^*ST samples; (2) removal of samples with missing values; (3) exclusion of samples from companies listed in the current year. Finally, a panel of data with 7730 observations was obtained. Data were sourced from the China City Statistical Yearbook and the CSMAR database. To mitigate the impact of outliers, a 1% trimmed approach was applied to all continuous variables.

4.2. Model Design and Variable Definition

4.2.1. Model Design

This paper formulates the following model to investigate the correlation between GEA and corporate carbon emission reduction performance:

$$CE{P}_{i,t}={\alpha }_0+{\alpha }_{1}GE{A}_{i,t}+{\alpha }_{2}Control{s}_{i,t}+{{\displaystyle \sum Year+{\displaystyle \sum Industry+\epsilon }}}_{i,t}$$

(1)

In the equation above, subscripts i and t denote individual and time, respectively. CEP_i,t represents corporate carbon emission reduction performance; GEA_i,t represents government environmental attention. Controls_i,t include a set of control variables, and Year and Industry indicate fixed effects, respectively. The term ℇ represents the random disturbance. For specific variable definitions, please refer to

Table 1. Variable definition.

Variables	Definition	Calculation Method
CEP	Corporate carbon emission reduction performance	Revenue per unit of carbon emissions
GEA	GEA	Frequency ratio of ecological terms in the annual government work report multiplied by 10
GDP	Regional economic development level	Logarithm of Local GDP
Stru	Regional industrial structure	Value added from the tertiary industry divided by the regional GDP
Lev	Debt-to-Equity ratio	Total liabilities divided by total assets
Size	Firm size	Logarithm of corporate revenue
Age	Years since going public	Difference between the sample data year and the company’s year of going public
Growth	Growth capability	(End-of-period revenue−beginning-of-period revenue) divided by beginning-of-period revenue
Tobin	Tobin’s Q value	Market value of the company divided by its book value
ROA	Return on assets	Net profit divided by total assets
HHI	Herfindahl–Hirschman index	HHI = ∑(Si2), where Si is the market share of the i-th company
HHI_sq	Herfindahl–Hirschman index square	Square term of the Herfindahl–Hirschman Index
Year	Year	Dummy variable for the year
Industry	Industry	Dummy variable for the industry

.

4.2.2. Dependent Variable

The corporate carbon emission reduction performance is a dependent variable in this study. We use revenue per unit of carbon emissions as a proxy indicator for corporate carbon emission reduction performance [68] for measuring CEP. Specifically, this study employs Formula (2) to estimate the total carbon emissions of enterprises.

$$\mathrm{Corporate} \mathrm{Carbon} \mathrm{Emissions}=\frac{\mathrm{Industry} \mathrm{Carbon} \mathrm{Emissions}}{\mathrm{Industry} \mathrm{Main} \mathrm{Operating} \mathrm{Costs}}\times \mathrm{Corporate} \mathrm{Operating} \mathrm{Costs}$$

(2)

We employ Formula (2) to estimate the total carbon emissions of enterprises, subsequently utilizing Equation (3) to assess the carbon emission reduction performance of the companies. Finally, the natural logarithm of this indicator is taken as a measure of carbon emission reduction performance, where a higher numerical value indicates better CEP.

$$CEP=\frac{\mathrm{Corporate} \mathrm{Revenue}}{\mathrm{Corporate} \mathrm{Carbon} \mathrm{Emissions}}$$

(3)

4.2.3. Independent Variable

The governmental environmental attention is an independent variable in this paper. In past research, indicator substitution methods have been commonly applied to quantify GEA, with indicators such as the number of environmental protection personnel and environmental investment being used as substitutes. However, this method only considers one aspect of environmental protection and does not comprehensively assess governmental attention to and management of the environment [24]. As research advances, more scholars are using text mining methods to extract and analyze the frequency of environmentally related terms from government annual reports [8,13]. Therefore, we employ a text extraction approach to perform keyword frequency analysis on yearly government work reports. Specifically, we select keywords related to the ecological environment, pollution, emission reduction, green initiatives, and low-carbon from the government work reports. This selection mirrors the frequency of terms indicating their attention to ecology. The frequency of those keywords is calculated as a ratio to the total word frequency in the government report for that year. Then the ratio is multiplied by 10 to indicate the government’s environmental attention, with a higher value suggesting increased attention from the municipal government to the ecological environment.

4.2.4. Control Variables

Referring to the existing literature [8,69], we employ control variables from two levels: urban and enterprise. At the urban level, we consider Gross Domestic Product (GDP) and Industrial Structure (Stru) at the city level. GDP can reflect the comprehensive development strength of a region, which will exert a certain degree of influence on high-energy-consuming enterprises, thereby affecting the CEP. The composition of industries within an economy determines the level of energy intensity and carbon emissions associated with production activities, hence, it is necessary to consider the industrial structure as a control variable in the model. At the enterprise level, we select variables encompassing Leverage (Lev), Enterprise Size (Size), Time Listed (Age), Growth Capability (Growth), Tobin’s Q Ratio (Tobin), Return on Assets (ROA), Herfindahl–Hirschman Index (HHI), and their squared terms (HHI_sq). Among them, the size and age of enterprises largely influence their carbon emissions, with larger and older enterprises potentially emitting more carbon. Regarding Lev and ROA and other control variables, they, respectively, measure the financial leverage and operational capability of enterprises. Therefore, we include many enterprise-level control variables in the model to ensure the accuracy of the regression results.

5. Empirical Analysis

5.1. Descriptive Statistics

Table 2. Descriptive statistics.

Variables	N	Sd	Min	Mean	Median	Max
CEP	7730	2.174	−6.222	0.679	0.318	8.033
GEA	7730	0.013	0.009	0.035	0.034	0.071
GDP	7730	1.060	6.217	8.783	8.836	10.549
Stru	7730	0.136	0.197	0.522	0.504	0.835
Lev	7730	0.190	0.048	0.381	0.372	0.855
Size	7730	1.349	18.703	21.455	21.283	25.172
Age	7730	0.965	0.000	1.870	1.946	3.258
Growth	7730	4.075	−21.976	0.098	−0.087	21.516
Tobin	7730	1.245	0.887	2.039	1.634	8.311
ROA	7730	0.046	−0.152	0.056	0.048	0.211
HHI	7730	0.118	0.036	0.156	0.122	0.669
HHI_sq	7730	0.065	0.001	0.038	0.015	0.448

presents descriptive statistics for the main variables. As depicted, the standard deviation of CEP is 2.174, signifying volatility in the carbon emission reduction performance of current high-energy-consuming enterprises. The mean and median of CEP are 0.679 and 0.318, respectively. As the mean exceeds the median, it implies a right-skewed distribution of CEP among enterprises. The mean and median of GEA are relatively close, suggesting a uniform distribution of this variable. Other variables show values within a reasonable range, aligning with previous research.

5.2. Regression Results

Table 3. Baseline regression results.

Variables	(1) OLS	(2) OLS	(3) RE	(4) FE
	CEP	CEP	CEP	CEP
GEA	9.832 ***	3.758 **	3.749 **	3.935 *
	(5.062)	(2.106)	(2.097)	(1.702)
GDP		−0.051 *	−0.051 *	0.149
		(−1.893)	(−1.853)	(1.106)
Stru		−0.060	−0.057	−0.090
		(−0.285)	(−0.269)	(−0.211)
Lev		−0.381 **	−0.375 **	0.358
		(−2.441)	(−2.387)	(1.148)
Size		−0.114 ***	−0.116 ***	−0.373 ***
		(−4.952)	(−4.965)	(−4.826)
Age		−0.227 ***	−0.226 ***	−0.132
		(−8.438)	(−8.321)	(−1.516)
Growth		−0.002	−0.002	0.004
		(−0.414)	(−0.405)	(0.662)
Tobin		0.034	0.033	0.023
		(1.641)	(1.604)	(0.775)
ROA		2.174 ***	2.177 ***	1.717 *
		(3.678)	(3.667)	(1.945)
HHI		1.345	1.351	1.629
		(1.103)	(1.109)	(1.153)
HHI_sq		−2.061	−2.074	−2.786
		(−1.183)	(−1.192)	(−1.390)
_cons	0.339 ***	2.837 ***	2.858 ***	5.846 ***
Industry	(4.736) No	(6.028) Yes	(6.011) Yes	(3.165) Yes
Year	No	Yes	Yes	Yes
Observations	7730	7730	7730	7730
R-squared	0.003	0.322	0.296	0.306

Notes: T-statistics in parentheses, * p < 0.1, ** p < 0.05, *** p < 0.01.

displays regression results examining the link between GEA and CEP. In column (1), the regression result, excluding control variables, reveals a GEA variable coefficient of 9.832, significantly positively correlated at the 1% significance level. In column (2), when control variables are included, the GEA coefficient is 3.758, remaining significantly positive at the 5% significance level. Columns (3) and (4) depict regression results with random effects and fixed effects, respectively. The coefficients for GEA are 3.749 and 3.935, both showing significant positive effects. In summary, irrespective of the regression model used, the consistent finding is that the GEA coefficient exhibits a significant positive impact, supporting the idea that it enhances CEP and validating Hypothesis 1. The results not only offer empirical evidence supporting governmental incentives or mandates for enterprises to decrease carbon emissions via mechanisms like taxation, subsidies, regulations, and standards, but also enrich the relevant empirical research in the field of CEP.

From the perspective of control variables, the coefficients of Size and Age are all significantly negative, which is not conducive to the improvement of CEP. Specifically, larger enterprises often encounter intricate organizational structures and management hurdles, potentially leading to reduced efficiency in implementing environmental protection measures [70]. Moreover, large enterprises typically possess greater assets and resources, potentially leading them to favor traditional, high-carbon-emission production methods over transitioning to more environmentally friendly technologies and practices. On the other hand, older enterprises may trail behind in technology and management, resulting in subpar performance in carbon emission reduction. The coefficient of the HHI is positive, while its squared term HHI_sq coefficient is negative, indicating an inverted U-shaped relationship between industry competition and CEP. Moderate industry competition promotes enterprises to enhance their CEP, while excessive competition may intensify competition among enterprises, resulting in increased production by managers and reduced CEP.

5.3. Endogenous Processing

5.3.1. Propensity Score Matching Test

To improve conclusion reliability and tackle endogeneity from sample self-selection bias, we applied Propensity Score Matching methods: Nearest Neighbor 1:1 Matching, Kernel Matching, and Radius Matching. The regression results in

Table 4. PSM test.

	(1) Nearest Neighbor 1:1	(2) Kernel Matching	(3) Radius Matching
Variables	CEP	CEP	CEP
GEA	11.520 ***	3.645 **	3.640 **
Controls	(4.336) Yes	(2.040) Yes	(2.037) Yes
Industry/Year	Yes	Yes	Yes
Observations	3955	7709	7708
R-squared	0.066	0.323	0.323

** p < 0.05, *** p < 0.01.

reveal coefficients for the three matching methods: 11.520, 3.645, and 3.640, all significantly positive, ensuring consistent conclusions.

5.3.2. Instrumental Variable Test

To tackle potential issues of omitted variables and bidirectional causality endogeneity, we utilized the Instrumental Variable method. First, we used the mean of GEA in other regions of the same province and year (GEA_IV1) as an instrument. This variable underwent testing through two-stage least squares regression. The rationale is twofold: firstly, local GEA is closely related to that of other cities within the same province, satisfying the relevance principle; secondly, the environmental attention in other cities does not directly impact the local enterprises’ carbon reduction behavior, ensuring the instrument’s exogeneity. Additionally, taking inspiration from Siddique et al. [71], we selected the lagged one-period GEA (GEA_IV2) as another instrument.

The regression results in

Table 5. Instrumental variable test.

Variables	GEA	CEP	GEA	CEP
	(1) First Stage	(2) Second Stage	(3) First Stage	(4) Second Stage
GEA_IV1	0.929 ***
	(68.551)
GEA_IV2			0.549 ***
			(44.843)
GEA		6.393 **		12.072 ***
		(2.210)		(2.772)
Controls	Yes	Yes	Yes	Yes
Year/Industry	Yes	Yes	Yes	Yes
Observations	7716	7716	5070	5070
R-squared	0.499	0.322	0.312	0.116
Cragg-Donald Wald F	4699.300		2010.692

Note: In columns (1) and (3), values within parentheses represent T-statistics, while in columns (2) and (4), values within parentheses represent Z-statistics. ** p < 0.05, *** p < 0.01.

display the first-stage regression outcomes in columns (1) and (3), revealing significant positive coefficients for GEA_IV1 and GEA_IV2 at the 1% level. This demonstrates the adherence of the chosen instruments to the relevance principle, with R-squared values of 0.499 and 0.312, indicating satisfactory explanatory power of the models. Columns (2) and (4) illustrate the second-stage regression results, estimating the GEA coefficients with the fitted values of GEA. The coefficients continue to be significantly positive, ensuring the robustness of our conclusions. Moreover, both Cragg–Donald Wald F statistics for the instruments are considerably greater than the critical value (16.38) for the Stock–Yogo weak instrument test at a 10% significance level [72], rejecting the null hypothesis of weak instruments. This confirms the rationality of the instrumental variable selection, and the regression outcomes align with the baseline results, supporting the earlier conclusions.

5.4. Robustness Tests

5.4.1. Replacement of the Dependent Variable Measurement

First, we adopt the logarithm of total corporate carbon emissions (CEP_1) as a substitute indicator for the dependent variable. Second, we define corporate carbon emissions as the sum of combustion and energy fuel emissions, production process emissions, waste emissions, and emissions resulting from changes in land use (e.g., forest to industrial land). We then log-transform this sum as an alternative measure for the dependent variable (CEP_2). Higher values for both indicators denote increased carbon emissions, suggesting relatively poorer CEP. In

Table 6. Variable replacement.

Variables	(1) CEP_1	(2) CEP_2	(3) CEP	(4) CEP
GEA	−0.859 ***	−0.641 *
	(−2.775)	(−1.839)
GEA_1			1.675 ***
			(3.203)
GEA_2				0.653 ***
				(2.590)
Controls	Yes	Yes	Yes	Yes
Industry/Year	Yes	Yes	Yes	Yes
Observations	7730	7374	7681	7681
R-squared	0.971	0.920	0.066	0.066

* p < 0.1, *** p < 0.01.

, GEA coefficients are −0.859 and −0.641, both statistically significant, implying that GEA can effectively reduce corporate carbon emissions, thus supporting robust conclusions.

5.4.2. Replacement of the Independent Variable Measurement

Government environmental subsidies represent valuable economic resources unconditionally allocated by the government to micro-economic entities. This financial support mirrors the government’s commitment to local environmental protection [73]. Businesses excelling in carbon reduction are more likely to secure government recognition and subsidies, fostering sustainable development. Thus, government environmental subsidies can function as an alternative gauge for GEA. We utilize the ratio of government subsidies to businesses for environmental protection, divided by year-end total assets (GEA_1), and year-end total revenue (GEA_2), respectively, as surrogate indicators for regression. As shown in Table 6, both GEA_1 and GEA_2 coefficients are significantly positive, suggesting that government environmental subsidies can boost CEP, thereby reinforcing the robustness of our findings.

5.4.3. Exclusion of Tier-1 Cities and 2020 Samples

In October 2011, the National Development and Reform Commission issued a notice approving carbon emission trading pilot projects in seven provinces and cities. These cities, particularly Beijing, Shanghai, Guangzhou, and Shenzhen, the pioneers in carbon trading, exhibit significant differences in attractiveness, resource endowment, fiscal channels, and environmental regulations compared to general prefecture-level cities. Previous studies have shown that carbon trading pilot projects notably enhanced local carbon efficiency [11,74], achieving specific emission reduction and environmental performance improvements. Furthermore, in 2020, the government shifted its focus to epidemic prevention and control due to COVID-19. Nationwide lockdowns and production halts affected high-energy-consuming enterprises, leading to a substantial short-term reduction in carbon emissions [75,76]. Consequently, this paper excludes the four pilot cities and 2020 samples in the regression analysis to assess the robustness of the conclusions.

Table 7. Results of excluding samples.

Variables	(1) CEP	(2) CEP
GEA	7.669 ***	4.502 **
Controls	(3.774) Yes	(2.278) Yes
Industry/Year	Yes	Yes
Observations	5875	6749
R-squared	0.109	0.319

** p < 0.05, *** p < 0.01.

illustrates the regression results in the first column, excluding the four pilot cities. The GEA coefficient is 7.669, significant at the 1% level. In the second column, the regression results exclude the 2020 samples, revealing a GEA coefficient of 4.502, significant at the 5% level. Both findings affirm that GEA markedly improves CEP, validating the robustness of the conclusions.

5.4.4. Lagged Dependent Variable Regression

In the preceding endogeneity test, the lagged explanatory variable was regressed as an instrumental variable. Therefore, the study separately regresses the lagged dependent variable for one period (CEP_lag1) and two periods (CEP_lag2) to investigate the delayed effect of GEA.

Table 8. Lagged dependent variable and firm fixed effects test.

Variables	(1) CEP_lag1	(2) CEP_lag2	(3) CEP
GEA	4.168 *	1.484	3.935 *
Controls	(1.673) Yes	(0.532) Yes	(1.702) Yes
Industry/Year	Yes	Yes	Yes
Firm	No	No	Yes
Observations	5070	3825	7730
R-squared	0.103	0.114	0.543

* p < 0.1.

reveals that when regressing CEP lagged by one period, the coefficient for GEA is 4.168, significant at the 10% level. However, in the regression with CEP lagged by two periods, the coefficient for GEA is 1.484, which is not significant. This suggests that the enhancing role of GEA is effective only within one year.

5.4.5. Firm Fixed Effects

To mitigate endogeneity issues stemming from omitted variables at the firm level, we incorporate firm-specific fixed effects into the baseline regression model. The results presented in Table 8’s third column reveal a coefficient of 3.935 for GEA. This affirms the robustness of the conclusions.

5.5. Heterogeneity Analysis

5.5.1. Grouping by the Level of Corporate Pollution

Investigating the heterogeneity among heavily polluting enterprises, major carbon dioxide emitters, and key players in low-carbon emission reduction responsibilities is crucial for comprehending variations in emission reduction efforts across industries. Given their significant contribution to greenhouse gas emissions through concentrated energy consumption, heavily polluting enterprises play a crucial role in achieving the “dual carbon” goals. In the context of stringent government regulations, particularly targeting high-energy-consuming and highly polluting enterprises, involving substantial increases in pollution fees and stricter carbon tax enforcement, these enterprises become more sensitive to environmental regulatory costs. Consequently, to maximize profits, these enterprises may respond to penalties by reducing economic activities or cutting carbon emissions. Therefore, this study anticipates a more substantial improvement in corporate carbon-emission-reduction performance related to GEA, especially among heavily polluting enterprises compared with lightly polluting ones.

We categorize industries like non-ferrous metal smelting, rolling processing, chemical raw materials, chemical product manufacturing, petroleum processing, and coking as heavily polluting enterprises; the remaining industries are considered lightly polluting.

Table 9. Heterogeneity analysis.

	(1) Heavily Polluting Enterprises	(2) Lightly Polluting Enterprises	(3) Large Enterprises	(4) Small and Medium Enterprises	(5) Eastern Region	(6) Central and Western Regions
Variables	CEP	CEP	CEP	CEP	CEP	CEP
GEA	6.413 **	0.838	0.829	4.653 **	5.682 ***	0.868
Controls	(2.493) Yes	(0.338) Yes	(0.214) Yes	(2.293) Yes	(2.687) Yes	(0.236) Yes
Industry	Yes	Yes	Yes	Yes	Yes	Yes
Year	Yes	Yes	Yes	Yes	Yes	Yes
Observations	4043	3687	2082	5648	5505	2225
R-squared	0.329	0.317	0.253	0.694	0.302	0.286
Inter-group Coefficient Difference	p = 0.013		p = 0.051		p = 0.000

Note: The Inter-group Coefficient Differences are tested through the homoscedasticity test. ** p < 0.05, *** p < 0.01.

displays the outcomes of the regression. In column (1), the coefficient for heavily polluting enterprises surpasses that in the second column for lightly polluting enterprises. Additionally, the inter-group coefficient difference is significant at the 5% level, signifying a more pronounced impact of GEA on CEP in heavily polluting enterprises. This confirms the anticipated findings of this study.

5.5.2. Grouping by Enterprise Asset Size

Enterprise carbon emissions are intricately linked to their size, and size disparities markedly influence carbon emissions behavior [77]. Small- and medium-sized enterprises (SMEs) exhibit greater operational flexibility, enabling swift adjustments to business models and the adoption of new technologies. This agility renders them more conducive to implementing carbon reduction initiatives. SMEs commonly feature streamlined management structures, fostering efficient resource management via operational optimization and energy-saving measures. This, in turn, enables effective carbon reduction within resource constraints.

Furthermore, SMEs, characterized by fewer facilities, production equipment, and employees, frequently encounter significant financing constraints. With the growing government emphasis on the environment, SMEs can mitigate these constraints by reducing carbon emissions to access government resources. These resources not only confer implicit credibility but also cultivate consumer trust and garner favor with banks and external investors. This, in turn, alleviates financing limitations, addressing societal development needs. Consequently, SMEs are anticipated to be more responsive to the incentives stemming from GEA, demonstrating a heightened proactive stance toward carbon reduction in comparison with larger enterprises.

On the contrary, larger enterprises usually possess more intricate supply chains and organizational structures [70], resulting in slower and more cumbersome decision-making processes. This complexity poses challenges for swift adjustments and changes during the implementation of carbon reduction strategies. Additionally, larger enterprises encounter greater hurdles in technology upgrades and transformation due to prolonged timelines and the higher costs associated with updating and replacing existing production equipment and processes. This difficulty hinders their ability to meet immediate carbon reduction demands. Given these considerations, this study expects that the influence of GEA on enhancing carbon emission reduction performance will be more noticeable in SMEs compared with larger enterprises.

We define the largest 30% of enterprises assets as large and the remaining 70% are classified as SMEs; the grouped regression results in Table 9 reveal a notably positive coefficient for SMEs in column (4). This exceeds the regression coefficient for large enterprises in column (3). Additionally, the inter-group coefficient difference yields a p-value of 0.051, suggesting significance at the 10% level. These findings affirm the greater impact of GEA on CEP in SMEs. Moreover, in the baseline regression section, the coefficient for Size is significantly negative, suggesting an inverse correlation between enterprise size and carbon emission reduction performance, thereby aligning with the study’s expectations to a certain extent.

5.5.3. Grouping by Enterprise Regions

China’s vast territory displays substantial economic development disparities across regions. In economically advanced areas like the east, multiple factors contribute to the notable influence of government environmental attention on corporate carbon emission reduction performance.

Compared with the central and western regions, the eastern region of China boasts a more developed economic structure and stronger capabilities in technological innovation and R&D [8]. Therefore, high-energy-consuming enterprises in this region have easier access to green innovation technologies and can employ advanced low-carbon production techniques, thereby improving their CEP. Moreover, the industrial structure in the eastern region is gradually transitioning towards green industries [78], prompting local enterprises to adopt cleaner and more environmentally friendly production methods. Furthermore, the mature environmental governance system in the eastern region encourages local enterprises to adhere more closely to carbon reduction regulations, contributing to the enhancement of their carbon reduction performance. Finally, the mature carbon market in the eastern region enables high-energy-consuming enterprises to actively engage in carbon trading markets [79], facilitating carbon emission reductions and potential economic benefits, thus further enhancing their CEP.

This study divides the companies into eastern and central and western enterprises for grouped regression. The findings in Table 9, column (5), reveal that the regression coefficient of GEA with CEP surpasses that in column (6). The inter-group coefficient difference is highly significant at the 1% level, validating that the influence of GEA on CEP is more pronounced in the eastern region, consistent with the study’s expectations.

6. Mechanism Analysis

6.1. Green Technological Innovation

To validate the mechanism of green technological innovation, we employ a stepwise regression method to build the models for examination.

$$GT{I}_{i,t}={\beta }_0+{\beta }_{1}GE{A}_{i,t}+{\beta }_{2}Control{s}_{i,t}+{\displaystyle \sum Year}+{\displaystyle \sum Industry}+{\epsilon }_{i,t}$$

(4)

$$CE{P}_{i,t}={\gamma }_0+{\gamma }_{1}GE{A}_{i,t}+{\gamma }_{2}GT{I}_{i,t}+{\gamma }_{3}Control{s}_{i,t}+{\displaystyle \sum Year+{\displaystyle \sum Industry}}+{\epsilon }_{i,t}$$

(5)

Our study employs the total count of three patent types—green invention patents, green utility model patents, and green design patents—applied to enterprises as an indicator of their green technological innovation (GTI_1). The data are retrieved from the Chinese Research Data Services (CNRDS) database. In line with established literature practices, a logarithmic transformation is applied, adding one to the patent count to address errors arising from right-skewed distributions.

The regression results in

Table 10. Green technological innovation mechanism test.

	(1)	(2)	(3)	(4)
Variables	GTI_1	CEP	GTI_2	CEP
GEA	3.111 ***	3.761 **	3.651 **	3.825 **
	(2.917)	(2.113)	(1.991)	(2.150)
GTI_1		0.131 ***
		(6.351)
GTI_2				0.077 ***
Controls	Yes	Yes	Yes	(6.930) Yes
Industry/Year	Yes	Yes	Yes	Yes
Observations	7730	7730	7730	7730
R-squared	0.036	0.326	0.025	0.327

** p < 0.05, *** p < 0.01.

reveal that, in column (1), the coefficient for GEA is 3.111, significant at the 1% level. This indicates that GEA indeed fosters enterprise green technological innovation. Simultaneously incorporating GTI_1 and GEA into the model, the coefficient for GTI_1 remains significantly positive at the 1% level. Notably, the coefficient for GEA also remains significantly positive. This suggests that GEA enhances enterprise CEP by strengthening internal green technological innovation capability. Green technological innovation plays a partial mediating role.

Additionally, to enhance the robustness of our conclusions, we modify the measure of green technological innovation to encompass the sum of three types of patents applied for independently and collectively by enterprises. Subsequently, we take the logarithm of the sum, adding one as an indicator for their green technological innovation capability (GTI_2). Regression results in columns (3) and (4) validate the persistent presence of the mediating role of green technological innovation, thereby upholding the robustness of our conclusions.

6.2. External Analyst Attention

Analysts, extending beyond their supervisory role, impose short-term performance pressure on management, as evidenced in previous studies [80]. This pressure compels management to prioritize short-term goals. To promptly boost short-term performance, garner government favor, access fiscal subsidies, and gain trust from downstream customers, management enhances the standardization of operational practices, diminishes environmental pollution, and enhances carbon emission reduction performance in production. Subsequently, production performance contributes to improved financial outcomes, augmenting the enterprise’s competitiveness. To investigate the mediating role of external analysts, a model is developed through stepwise regression.

$$AL{A}_{i,t}={\eta }_0+{\eta }_{1}GE{A}_{i,t}+{\eta }_{2}Control{s}_{i,t}+{\displaystyle \sum Year+{\displaystyle \sum Industry}}+{\epsilon }_{i,t}$$

(6)

$$CE{P}_{i,t}={\lambda }_0+{\lambda }_{1}GE{A}_{i,t}+{\lambda }_{2}AL{A}_{i,t}+{\lambda }_{3}Control{s}_{i,t}+{\displaystyle \sum Year+{\displaystyle \sum Industry+{\epsilon }_{i,t}}}$$

(7)

Employing established research methodologies, this study utilizes the logarithmically transformed analyst attention index (ALA_1) as a metric for external analyst attention to the enterprise. The regression outcomes are outlined in

Table 11. Analyst attention mechanism test.

	(1)	(2)	(3)	(4)
Variables	ALA_1	CEP	ALA_2	CEP
GEA	2.187 **	3.430 *	2.140 **	3.463 *
	(2.473)	(1.919)	(1.969)	(1.938)
ALA_1		0.095 ***
		(4.123)
ALA_2				0.082 ***
Controls	Yes	Yes	Yes	(4.356) Yes
Industry/Year	Yes	Yes	Yes	Yes
Observations	7690	7690	7690	7690
R-squared	0.430	0.324	0.425	0.324

* p < 0.1, ** p < 0.05, *** p < 0.01.

. In the initial column, the GEA coefficient is significantly positive at the 5% significance level, indicating that GEA indeed amplifies external analyst attention on the enterprise, assuming a supervisory role in carbon reduction. The subsequent column presents the regression outcomes incorporating both external analyst attention and GEA into the model. The coefficient for ALA_1 is notably positive, and importantly, the GEA coefficient remains significantly positive. This suggests that GEA augments enterprise carbon emission reduction performance by elevating external analyst attention to the enterprise. External analyst attention manifests a partial mediating effect.

Moreover, to enhance robustness, this study modifies the measurement of analyst attention by substituting it with the logarithmically transformed analyst coverage in research reports (ALA_2). The regression outcomes in columns (3) and (4) illustrate the persistent mediating effect of analyst attention, ensuring the stability of the conclusions.

6.3. Media Attention

Based on the above analysis, this study utilizes a stepwise regression model for examination.

$$M{\mathrm{edia}}_{\mathrm{i},\mathrm{t}}={\varphi }_0+{\varphi }_{1}GE{A}_{i,t}+{\varphi }_{2}Control{s}_{i,t}+{\displaystyle \sum Year+{\displaystyle \sum Industry+{\epsilon }_{i,t}}}$$

(8)

$$CE{P}_{i,t}={\mu }_0+{\mu }_{1}GE{A}_{i,t}+{\mu }_{2}Medi{a}_{i,t}+{\mu }_{3}Control{s}_{i,t}+{\displaystyle \sum Year+{\displaystyle \sum Industry+{\epsilon }_{i,t}}}$$

(9)

We use the natural logarithm of the total number of times Company i is covered by online media plus 1 (Media_1) as an indicator for media attention, obtaining data from the CNRDS database.

The regression results are outlined in

Table 12. Media attention mechanism test.

	(1)	(2)	(3)	(4)
Variables	Media_1	CEP	Media_2	CEP
GEA	5.230 ***	3.927 **	7.147 ***	3.821 **
	(6.848)	(2.179)	(7.529)	(2.114)
Media_1		0.068 **
		(2.502)
Media_2				0.050 **
Controls	Yes	Yes	Yes	(2.366) Yes
Industry/Year	Yes	Yes	Yes	Yes
Observations	7633	7633	7569	7569
R-squared	0.367	0.323	0.366	0.324

** p < 0.05, *** p < 0.01.

. In the initial column, the GEA coefficient is 5.230 and significantly positive, indicating that GEA genuinely amplifies media attention, playing a supervisory role in improving CEP. In the subsequent column, integrating both media attention and GEA into the regression model, the Media_1 coefficient is 0.068 and significant. Crucially, the GEA coefficient is 3.927, suggesting that GEA can improve CEP by heightening media attention, with media attention partially mediating the effect. Additionally, for result robustness, this study replaces the metric for media attention with the natural logarithm of the total number of times Company i is featured in newspapers plus 1 (Media_2). The regression outcomes in columns (3) and (4) confirm the persisting mediating role of media attention, reinforcing robust conclusions.

7. Conclusions and Recommendations

7.1. Conclusions

Concentrating on A-share high-energy-consumption enterprises from 2009 to 2020, this study empirically explores the relationship between GEA and CEP. The results indicate that GEA can significantly improve corporate CEP. After utilizing PSM and instrumental variable tests to address endogeneity, our research findings demonstrate consistency. The robustness test further confirms the reliability of these results. Additionally, heterogeneity analysis reveals that this effect is more pronounced in heavily polluting industries, SMEs, and companies located in the eastern regions of China. The mechanism analysis suggests that internal green technological innovation capabilities, external analysts’ attention, and media attention are important ways to achieve the emission reduction effect of GEA. The findings provide valuable insights into the intricate dynamics between government initiatives and corporate carbon reduction outcomes, and they also offer valuable guidance for enterprises seeking to reduce carbon emissions.

7.2. Recommendations

First, governments should enhance their attention on environmental matters and formulate effective environmental protection policies to improve corporate CEP. Considering government attention to the environment as a limited resource signaling environmental priorities, it is crucial for the government to allocate attention efficiently, enhance decision-making efficiency, and reinforce ecological awareness in key public forums and official reports. Therefore, it is significant for governments to promote environmental awareness among the public and corporations, cultivating a strong environmental market system by enhancing environmental attention. Encouraging enterprises to explore sustainable development models that align green mountains and waters with economic prosperity is crucial for promoting a green development ethos.

Second, governments should tailor the intensity of their environmental attention to local contexts, avoiding a one-size-fits-all approach that might constrain enterprises and impede carbon reduction efficiency. Considering differences in emission levels, economic development, regional characteristics, and individual corporate traits, the marginal impact of GEA on CEP may vary across regions. Therefore, governments should fine-tune their environmental focus based on local circumstances, directing attention to highly polluting and resource-intensive enterprises. Emphasizing regional carbon reduction issues in line with the “dual-carbon” vision and the demand for high-quality green economic development will effectively empower GEA to improve corporate carbon emission reduction performance.

Third, enterprises should proactively embrace carbon reduction policies, injecting momentum into the pursuit of “dual-carbon” goals. It is essential for enterprises to actively pursue carbon reduction as a strategic imperative. Therefore, they should develop a robust sense of responsibility for carbon reduction. They should implement proactive carbon reduction measures, such as technological upgrades, the development and utilization of new energy sources, and comprehensive waste utilization, to garner support from market stakeholders. This also fosters positive collaborative relationships, enhances competitive advantages, and facilitates enterprises in realizing green and low-carbon development.

7.3. Limitations and Future Prospects

This study has several limitations that should be considered in future research. First, constrained by sample data, this research focuses solely on high-energy-consuming listed companies in the A-share market from 2009 to 2020. Therefore, it may not fully reflect recent policy changes and market dynamics, especially given the frequent updates to carbon reduction policies. Future research could include a broader range of industries, and enterprises with different ownership structures to improve the generalizability and representativeness of the findings.

Second, this study primarily examines the impact of GEA on corporate CEP, without considering other external factors such as international market demand changes and the global carbon trading system. Therefore, future research could include external factors like international market demand, the global carbon trading system, and economic cycles in the research model to comprehensively assess their impact on corporate CEP.

Third, this study uses Chinese firms as the research sample. However, due to varying institutional frameworks, industrial structures, and economic development levels across countries, additional research samples from other countries are needed to assess the comprehensive impact of government environmental attention on corporate carbon reduction performance.

Fourth, while research suggests that GEA enhances corporate CEP via internal green technology innovation capabilities and external analyst and media attention, the possible presence of mediating variables in government decision-making remains uncertain. Thus, future studies could investigate this matter from the standpoint of government environmental regulatory decisions.