Digital Transformation and Corporate Carbon Emissions: The Moderating Role of Corporate Governance

Abstract

In the era of the digital modern economy, digital transformation has grown into the primary battlefield for conventional industrial competitiveness. For businesses, digital transformation is not just a future trend and requirement but also an intrinsic motivation to achieve sustainable development. The purpose of this research was to investigate the connection between digital transformation and carbon emission reduction using empirical analyses, as well as to elucidate whether the qualities of internal control, environmental disclosure, and auditing affect the connection between digital transformation and carbon emission reduction in businesses. This research used fixed-effects regression to evaluate data from China’s A-share-listed businesses from 2014 to 2023. These data suggest that corporate digital transformation may successfully reduce carbon emissions. Meanwhile, internal control quality, environmental information disclosure quality, and audit quality all have a beneficial moderating influence on corporate digital transformation and carbon emission reduction. By incorporating pertinent theories, such as digital economy theory and ecological theory, this research indicates the immediate impact of digital transformation on reducing business carbon emissions, enhances and broadens the body of knowledge on the subject, and offers methodological recommendations for reducing corporate carbon emissions and attaining rapid development. Furthermore, it provides useful recommendations on how the government, businesses, and executive teams can contribute more to digital transformation and the carbon emission reduction process. This will assist Chinese corporations in raising their own level of digital transformation and achieving ongoing improvements in the management of carbon emission reduction.

1. Introduction

With the rapid growth of society, greenhouse gas emission levels are growing, leading to concerns such as global warming, which results in severe consequences, such as glacier melting, heat waves, and excessive precipitation in metropolitan areas [1,2,3]. Lowering carbon emissions is recognized as a key strategy for mitigating environmental catastrophes, and numerous nations have implemented initiatives to slow down global warming. China, the world’s largest developing nation, accounts for around 30% of global carbon dioxide production. To comply with the Paris Agreement, the country has vowed to reach carbon neutrality by 2060 and peak carbon by 2030, focusing on low-carbon development [4] and showcasing its dedication to environmental preservation. China, a major manufacturing nation, has aggressively implemented a number of environmental governance rules to promote the transformation of old sectors in an effort to lessen the enormous strain of carbon emissions [5].

Digital technologies, including communication technologies related to mobile communications and the Internet of Things, cloud computing technologies related to high-performance computing and storage, artificial intelligence technologies related to automated inference search and machine learning, and big data technologies related to data mining and analysis, take the lead in societies and economies in the digital era [6]. Advances in science and engineering have facilitated the development of conventional sectors, and the adoption of digital technologies has grown into a strategic focus for developing competitive and sustainable economic benefits [7]. Because the manufacturing sector is the backbone of every nation’s economy and represents its richness, inventiveness, and economic might, considerable emphasis is placed on digital transformation in this industry. Digital transformation is the method of recreating a company’s or enterprise’s essential business activities using digital technology concepts and practices in order to shift from a traditional to a digital business model. Notably, digital transformation entails not only modernizing and upgrading technology but also alterations to numerous facets of business management structures. To strengthen enterprises’ digital applications, achieve breakthroughs in key common technologies for businesses to lower carbon emissions, and encourage high-quality enterprise development, China has launched a number of policy guidelines and support initiatives, as well as successive development plans for the complete embrace of enterprise digitalization and manufacturing [8].

Recently published empirical research based on corporate data has examined how digital transformation affects company carbon emissions. By reducing emissions, increasing efficiency, streamlining the production chain, and fostering green technology development, digital transformation can reduce an organization’s carbon risk level. This effect is more noticeable in businesses that prioritize the digital transformation of industrial processes, have stringent environmental requirements, and face intense industry competitiveness [9]. Clarifying the environmental demands of businesses and the general public that rely on digital technology is critical in aligning the interests of key stakeholders in the digital economy, lowering carbon emissions, and supporting green growth. This will aid in the establishment of an effective environmental system for green development, as well as ensuring its stability and long-term viability [10]. Moshood et al. demonstrated, using a spatial panel Durbin model, how the digital economy reduces carbon emissions in both direct and indirect ways [11]. Digital technology is an important tool for reaching carbon neutrality since it can promote both company economic growth and carbon emission reduction [12,13]. However, scholars have voiced differing opinions on this matter. Utilizing digital technology, exemplified by information and communications technology, can lower carbon emissions; yet, excessive investment can result in high energy consumption, and businesses will still struggle to provide a profitable return, creating the “digital transformation paradox” [14,15]. Additionally, organizations are under considerable financial pressure to increase their usage of digital technologies [16]. Businesses are increasing the extraction, manufacture, and consumption of resources since energy costs have dropped as a result of digital technology improvements. This increases energy consumption and creates a “rebound effect” [17].

To determine whether businesses undertaking digital transformation may lower carbon emissions, we selected 13,866 firms listed on China’s A-share market from 2014 and 2023. Prior research has discovered an innovation effect between lowering carbon emissions and businesses’ digital transformation. China’s National Offshore Oil Corporation has embraced the ESG sustainable development concept as its main development strategy. The company has developed an effective strategy to improve energy efficiency and lower its carbon footprint, which has culminated in a substantial decrease in pollutants, carbon monoxide, and other greenhouse gas emissions. The company has also developed the Basic Standard for Enterprise Internal Control and its promoting instructions, as well as other internal control systems and mechanisms related to finance, operations, and compliance monitoring [18]. According to the basic theory of auditing, the fundamental goal of state auditing is to ensure and promote the full and effective fulfillment of the government’s public fiduciary economic responsibilities. The Audit Office of China has issued several working opinions, proffered milestones, planned a specific action program for carbon reduction, and ensured that the “double carbon” goal is successfully reached through the process of emission reduction and carbon reduction financial fund audits, policy tracking audits, special audits, and environmental responsibility audits [19]. The European Union, through large enterprises, carries out energy audits every four years, while the German government undertakes those of medium-sized businesses to offer the requisite financial assistance, guaranteeing that carbon emissions are reduced effectively [20]. Environmental regulation’s influence on carbon emissions has been extensively studied, although a consistent conclusion has not been reached, with one perspective positing that environmental information disclosure is conducive to carbon emission reduction [21,22], and another stating otherwise (implementing dual-carbon actions to build a “beautiful China” [23]).

In addition to holding a significant place in the economy, manufacturing and other businesses are critical in establishing a new growth paradigm in China. However, there are fewer studies on how corporate digital transformation influences the reduction in carbon emissions and the moderating effects of auditing, environmental information disclosure, and internal control quality, as well as differing research viewpoints. Digital transformation has resulted in an entirely novel revolution in the growth of factories and other industries. Therefore, using data from Dibble and Bloomberg Consulting, the mechanisms and influence effects of corporate internal control quality, environmental information disclosure quality, and audit quality are examined and used as moderating elements to better understand the link between corporate digital transformation and a decrease in carbon emissions. A scientific inquiry into the link between business digital transformation and decreasing carbon emissions is also carried out. Empirical studies are conducted to determine whether the impact of moderating variables on corporate reductions in carbon emissions genuinely transfers to digital transformation and influences corporate sustainability behavior.

The basic aims of carbon emission reduction are to reduce the greenhouse effect, halt global warming, and reduce carbon dioxide and other greenhouse gas emissions from human activity. The 1992 United Nations Conference on Environment and Development made environmental and global resource protection the top priority of national development plans for the first time. It also specifically suggested that the international community should control the rise in greenhouse gas concentrations in the atmosphere and reduce carbon dioxide emissions. China is highly suitable for championing carbon emission reduction in enterprises. During the UN General Assembly in September 2020, China said that it wants to achieve carbon neutrality by 2060 and a carbon peak by 2030, which is achievable for businesses conducting carbon reduction research. The “dual-carbon” goal not only adheres to the Paris Agreement, actively addresses climate change, and demonstrates the accountability and dedication of a great country but also has far-sighted strategic significance in accelerating China’s economic and energy transformation [24]. As the second-largest economy in the world, China has achieved significant progress in digital growth, according to the Digital China Growth Report [25]. As the world’s top manufacturing nation, it is imperative that emerging countries examine the relationship between reducing carbon emissions and the digital transformation of Chinese companies.

The following additions are made when comparing this work to the body of the current literature. First, this study illustrates the economic and environmental benefits of digital transformation in poor nations and develops the subject of digital transformation research, starting with enterprise digital transformation. Second, it confirms and expands on the positive effects of lowering carbon emissions brought about by business digital transformation. Third, it considers how the qualities of auditing, environmental disclosure, and internal control affect corporate carbon emissions and the mechanism of the regulating effect. Therefore, it is concluded that auditing, environmental data publication, and the quality of internal controls are critical supplementary assets to business digitization, which improves firm development and positively regulates the function of corporate reductions in carbon emissions. Fourth, this work contributes to the corpus of knowledge on managing corporate carbon emission reduction and digital transformation.

This paper’s remaining sections are organized as follows: Section 2 explains the theoretical background and hypothesis derivation; Section 3 describes the research model design, sample selection, data sources, and variable definitions; Section 4 reports the empirical analyses and robustness tests; and Section 5 presents the discussion, conclusions, managerial significance, and future research directions.

2. Theoretical Background and Hypotheses

Given how quickly computer information technology is developing, Tapscott’s “digital economy theory” expands the digital economy concept to the networked intelligent era. This information system is not limited to the integration of digital technology and intelligent terminals; it also refers to the technology-based environment of production and life between people [26]. The digital economy has emerged as a result of the generation and expansion of digital information in this century, which encompasses not only the conventionally fundamental electronic communications and information sector but also other digital information sectors. It is a digital information fusion of the industries and the development and extension of a new economic paradigm. Through the integration of digital information into numerous businesses, this new economic paradigm has been built and expanded.

The digital economy has digital technological innovation as its core driving force. In terms of technological innovation, digital technologies, such as the Internet of Things, big data, and artificial intelligence, can be combined with the needs of enterprises to deeply mine data and knowledge, accelerate the innovation and industrialization of low-carbon technologies, and promote the advancement of the application of green technologies. The strong permeability and wide coverage of the digital economy can break down the boundaries of industrial development and promote the upgrading and transformation of industries into high-output and low-emission sectors. Data intelligence enables enterprises to understand the real-time situation of production and sales, optimize production decisions, and reduce energy consumption in the production process; thus, organizations can then implement accurate carbon emission reduction measures to achieve the low carbon emission goals.

To provide external circumstances for the development of industrial transformation, a number of industries are increasingly focusing on the rise of the digital economy as a whole. Businesses are also paying more attention to the digital economy to advance their level of knowledge and intelligence and to provide internal motivation for their own development through digital transformation [27].

According to modern ecological philosophy, environmental issues may be solved by increasing resource efficiency and encouraging sustainable development [28]. Ecological theory emphasizes the harmonious coexistence of humans and nature, prompting companies to increase the research, development, and application of renewable energy technologies, such as solar, wind, and water energy. These clean energy sources replace traditional fossil fuel sources and can reduce carbon emissions at the source. Inspired by ecological theories, enterprises and society seek technological innovations to enhance energy use efficiency, such as developing energy-saving equipment, optimizing industrial production processes, and improving energy-saving building technologies. Based on the concept of ecological protection, the carbon dioxide emitted from industrial production can be captured and sequestered or converted into chemical products, etc., so as to achieve carbon emission reduction and recycling.

First, the corporate production process can be optimized through digital transformation brought about by technical advancements. This reduces carbon emissions through improved energy efficiency and using fewer petroleum products and other sources of energy [29]. Meanwhile, the digital revolution is eliminating the constraints of conventional industries, for example, in the role of night workers, helping to effectively reorganize and rationally allocate human resources, capital, and other resources; achieve business transformation; and drive the rationalization and structural strengthening of the manufacturing system, which is favorable for the achievement of carbon neutrality [30]. Artificial intelligence, big data, cloud computing, the IoT, and other digital technology transformations have led to the development of intelligent robots that relieve enterprise workers of manual labor, increase production and service efficiency, and lower production costs, all of which have a substantial impact on reducing carbon emissions [31]. Second, digital transformation from the perspective of management methods, convenience, and practicality considerably reduces the pressure on the environment and strongly facilitates the achievement of low-carbon development [32]. For example, the data mining capabilities of digital technology can facilitate the effective management of energy demand, improve operational efficiency and process models, and reduce energy costs [33]. Digital technology can also be used in the real-time technology tracking and monitoring of oxygen and carbon dioxide emissions to aid in the detection, collection, and organization of data concerning carbon emissions and can indirectly encourage the reduction in emissions [34]. In terms of supply chain operations management [35], such technology encourages the integration of environmental activities in both upstream and downstream industries to build a green supply chain [36] that optimizes market structure, saves energy, and reduces emissions. Ritter et al.’s qualitative study argues for the use of digital transformation to help modify company models, transform the corporate eco-environmental system, and realize corporate eco-environmental improvements [37].

With these aspects in mind, digital transformation offers a number of advantages, such as streamlining enterprise business procedures, increasing operational effectiveness, combining internal and external resources, promoting the creation of new business models, and enabling the upgrading of industrial structures, all of which are critical methods for lowering carbon emissions.

According to the evidence presented above, Hypothesis 1 is presented.

Hypothesis 1 (H1).

Digital transformation encourages companies to reduce their carbon emissions.

Internal control is the process through which an enterprise achieves its strategic goals, realizes its business objectives, protects the safety and integrity of its assets, ensures the accuracy and reliability of its financial statements, and safeguards the efficient operation of the organization as a whole. The primary objective of internal auditing is to effectively assure the accuracy and integrity of a company’s reports and associated information, as well as to increase the level of company leadership and avoidance of risks [38]. First, effective internal control facilitates a company’s decision-making. The board of directors, supervisory board, management, and regular employees from the company’s different levels are involved in the development and execution of the internal control system, which emphasizes internal control and the separation of duties during the implementation phase. High-quality internal control can, therefore, prevent the core leader’s decision-making process and mode from being arbitrary, effectively limit the management’s freedom of choice in terms of reducing carbon emissions and protecting the environment, and improve the management’s capacity to do so. This is in addition to enhancing the management’s motivation to avoid environmental risks, promoting the legal and compliant operation of the enterprise, and reducing the possibility of the misstatement of environmental reports [39]. It is also conducive to conveying a good image of the enterprise to the outside world and understanding the enterprise environment while ensuring better risk assessment [40]. Second, good-quality internal control enables the enterprise’s environmental protection concepts to be carried out and helps in the implementation of environmental protection measures and carbon emission reductions, thus enhancing the sustainable development of the enterprise [41]. Guenther et al. demonstrated how digital transformation can foster corporate innovation and enhance internal control, both of which support an organization’s long-term growth [42]. Third, improving the quality of internal control has the potential to cultivate a superior internal control culture and control environment. The process of carbon information disclosure will be more standardized, with greater motivation toward corporate emission reduction, while effectively incentivizing corporate environmental protection and public welfare behaviors, as well as overseeing and prohibiting inappropriate corporate practices that harm the environment and public welfare. This is in the interest of business, as it increases revenue, can ensure the achievement of business objectives, and contributes to the realization of the social objectives of external stakeholders [43].

Accordingly, Hypothesis 2 is presented.

Hypothesis 2 (H2).

When it comes to corporate digital transformation, the effectiveness of internal control has a positive moderating effect on carbon emissions.

Increasing the standard for the disclosure of environmental information helps the public understand and track a company’s environmental practices, which not only promotes the further improvement of corporate environmental performance but also facilitates sustainable green development [44]. Guo and Lu found that corporate social responsibility disclosure drives enterprises to engage in green innovation by strengthening their regulatory and normative legitimacy motives [45]. On the one hand, environmental authorities are made aware of businesses’ knowledge of environmental pollution through environmental information disclosure, and regulatory pressure compels businesses to implement green innovation and reduce their carbon footprint. Using a random effects model, Guo et al. [46] experimentally showed that environmental information disclosure plays a role in corporate governance and has a considerable impact on a company’s environmental performance. The impact of reducing carbon emissions, environmental performance, and the strength of the company’s rules and regulations all contribute to the quality of environmental information disclosure [47]. The danger of failure increases when corporations publish more environmental data to improve their green innovation initiatives; the limited internal resources of enterprises necessitate a choice between high- and low-quality green emission reduction projects, while considering carbon emission reduction equipment investment, thus affecting the quality of carbon emission reduction. When the marketization process exceeds a certain value, the market and legal systems tend to be perfect, the problem of investment difficulties in the enterprise tends to be alleviated, and the enterprise needs to further disclose high-quality environmental information. For businesses to thrive in a competitive market, they must be prepared to reveal high-quality environmental information. The disclosure of environmental data improves the reduction in carbon emissions, demonstrating that it has an impact on the organization. There is a marketization threshold with a U-shaped influence, first being negative and then becoming positive [48]. However, environmental information disclosure allows investors, stakeholders, and authorities to understand the condition of business environmental emission reduction, allowing businesses to obtain incentives from both parties to reduce emissions, providing important financial support for corporate green innovation and promoting corporate green emission reduction [49]. Environmental information disclosure can encourage a company to actively participate in environmental management, establish stakeholders’ confidence and recognition of the company’s development, achieve energy-saving and emission-reduction goals, improve the credibility of the company, and establish the efficacy of the brand.

Based on the aforementioned analysis, Hypothesis 3 is proposed.

Hypothesis 3 (H3).

The quality of environmental information disclosure has a positive moderating impact on corporate digital transformation to minimize emissions of carbon dioxide.

Three degrees of national audit objectives can be distinguished: direct, realistic, and basic. Protecting people’s primary interests is the fundamental goal; promoting the rule of law, protecting people’s livelihoods, promoting reform, and facilitating development are the realistic goals; and monitoring and assessing the accuracy, legality, and efficacy of the audited units’ financial and fiscal revenues and expenditures is the direct goal [50]. First, using scientific and standardized methodologies and procedures to enhance the quality of government environmental audits, the formulation and implementation of low-carbon policy audits promote the smooth flow of governmental orders and truly reflect the situation of the audited unit, while exposing any potential problems. Moreover, carbon emission reduction can be facilitated by managing and controlling the potential for emission violations by businesses, as well as the timely disclosure of the use of environmental protection funds, enterprise allocation, and other green development issues. In addition, a high-level audit can help in the management of the use of low-carbon financial and tax funds. Jung et al. [51] posited that high-quality national audits can encourage enterprises to conduct rectification, weaken the business risks they face through national audits, and, ultimately, reduce carbon emissions [52]. Zhu and Li [53] investigated the function of stringent environmental asset exit inspections for leadership cadres in improving carbon emission reduction and capacity use, while Pan and Sun examined the spatial governance impact of government audits on carbon emissions using spatial econometric models; their findings all prove that high-quality government audits can effectively promote carbon emission reduction [54]. Local governments’ behavioral choices have an impact on the amount of carbon emissions at the local level, which also have a considerable influence on local energy governance initiatives. A national audit can be a useful tool for controlling local government behavior [55]. In addition, to strengthen the concept of scientific and technological audits, it is essential to reform the traditional project organization and management approach, conduct investigations efficiently, strengthen the whole control of the site, establish the economic responsibility audit’s “results’ push” mechanism, and enhance the standard of economic responsibility auditing. By concentrating on the application of low-carbon decision-making and deployment, audits capture the trajectory of economic power. This improves the efficiency and balance of the local government’s distribution of public resources, keeps funding for low-carbon development from being stifled by other public products with comparable long-term effects, and encourages the lowering of carbon emissions [56]. Wei et al. empirically tested the promotion of carbon emission reduction through high-quality economic responsibility audits from the perspective of carbon emissions and per capita carbon emissions [57].

Based on the analyses above, Hypothesis 4 is proposed.

Hypothesis 4 (H4).

Through the digital transformation of businesses, audit quality has a beneficial moderating effect on carbon emissions.

The study model is illustrated in Figure 1.

3. Research Design

3.1. Sample Selection and Data Sources

In 2012, early instructions for the regulation of cooperative reductions in greenhouse gas emissions were issued, signaling the start of China’s cooperative emission reduction market. Considering data availability and other criteria, in this study, Chinese A-share-listed businesses from 2014 to 2023 were used as the research object, with 13,866 research samples being collected. The sample data come from a variety of sources, including the China Research Data Service Platform database, DIB data, the Wind database, corporate websites, environmental websites, and ethical behavior and annual filings from the listed enterprises. The following procedures were used in this study to lessen the impact of extraneous variables and guarantee the validity and correctness of the data: (1) exclusion of PT (special transfer, which means pausing trade and awaiting delisting of listed businesses); ST (special treatment), including publicly traded businesses with losses in three consecutive years of operation and removing alerts; and delisted companies. (2) To account for extreme values, all the continuous variables in this study were weighted at the 1% and 99% levels. (3) To reduce the effects of variability, the key continuous variables were logarithmized. (4) In this investigation, all the continuous variables utilized in the interaction terms were pooled to avoid covariance issues.

3.2. Variable Definition

3.2.1. Dependent Variable: Corporate Carbon Emissions

The dependent variable in this study is business carbon emissions, which reflect the quantity of carbon emissions generated by the firm during a given time period. The combustion of fossil fuels is the main source of carbon monoxide and carbon dioxide emissions, and carbon emissions are a widely used indicator of corporate carbon performance. This paper cites Ping et al.’s study, along with information from the Intergovernmental Panel on Climate Change, which uses the natural logarithm as a basis for summarization and considers that direct and indirect emissions of carbon from petroleum and coal, as well as carbon greenhouse gases during the manufacturing process, account for the majority of total carbon emissions [58]. The total quantity of carbon emissions is a combination of burning and emissions that escape, waste, manufacturing processes, and environmental changes (for example, converting forests to commercial areas). Carbon emissions were calculated using data from multiple statistics calendars produced by the National Bureau of Statistics, social responsibility reports, annual reports from publicly traded corporations, and environmental websites.

3.2.2. Independent Variable: Digital Transformation

Studies on digital transformation measurements include textual analyses [59] and questionnaires [60]. However, questionnaires often produce erroneous results due to methodological or subjective biases. Listed businesses’ annual reports can clearly and successfully convey their strategic posture, and they also include phrases associated with digital transformation [61]. Therefore, this study employed textual analysis to statistically quantify firms’ digital transformation based on annual report data, as follows.

1. In this study, a body of research on using content analysis to assess digital transformation was examined with the goal of developing a vocabulary of terms relevant to this process [62]. The results included two major keywords related to digital transformation: fundamentals of digital technology and its applications [63]. Moreover, we analyzed digital-transformation-related phrases from the literature and compared them to those found in China’s Stock Market and Accounting Research database. Lastly, we assembled 76 keywords related to digital transformation, including blockchain, machine learning, and the use of cloud computing.

2. Python(3.12.0) software was employed to produce annual financial statements of China’s A-share-listed industrial enterprises from 2014 to 2023. The Java PDF box 3.0.x was used to extract the text content from each company’s annual report. Interestingly, one of the most beneficial disclosures in financial reporting is management discussion and analysis (MD&A) [64], and it includes more precise and proactive business data [65]. Based on existing research [46], this investigation centered on textual evaluations of the MD&A. Annual reports were used to provide textual master data that can be searched with keywords linked to digital transformation.

3. To determine the total word frequency of digital transformation, terms from the MD&A text database were searched, matched, tallied, and combined. Finally, an assessment measure for the organization’s level of digital transformation was created by multiplying the total number of times the incidence of the gathered terms occurred through the text size of the most recent MD&A by 100. The organization’s level of digital transformation increases with a higher score [62,66].

3.2.3. Moderating Variables

Internal control quality: In this research, we measured the internal control quality of the listed corporations using the “internal control index/100”, which is derived from the Dibbo database. Dibbo invented the Dibbo Internal Control Index to assess the efficacy of an organization’s internal controls. By assessing every facet of a company’s control system, the index represents the general caliber and degree of such control.

Environmental disclosure quality: The Environmental Disclosure Index evaluates the quality of a company’s environmental information disclosure. It is based on a company’s publicly available environmental information and is scored according to certain standards and methods that can help investors, governments, companies, and other users understand an organization’s sustainability record, allowing them to make better decisions. Environmental management, environmental regulation and certification, environmental performance and governance, and 25 other indicators are scored in seven different areas in this study. The logarithm of nature was determined through the addition of one to the total score.

Audit quality: Drawing on Gallego-Álvarez et al.’s [67] method, the quality of an audit was computed by first assessing the chance of the auditor issuing a standard opinion on the audit using a model, subtracting that likelihood from the actual audit opinion issued, and then calculating the negative absolute value of the difference. Poor audit quality is indicated if there is a significant discrepancy between the auditor’s actual audit opinion and the anticipated likelihood of giving an unqualified opinion. This is because the positive deviation represents the auditor’s degree of aggressiveness: the greater the level of aggressiveness, the more likely the auditor is to mislead investors. A negative deviation represents the auditor’s degree of conservatism: the greater the level of conservatism, the more likely the information in the financial statements will be reduced in value. The convenience of this study was assessed using the actual audit opinion and the negative absolute value of the degree of divergence from the expected probability of providing an unqualified opinion audit quality; the larger the audit quality’s worth, the higher the audit quality.

3.2.4. Control Variables

This study referenced Chen and Mao et al.’s work in order to lessen the impact of additional potential causes [68]. A total of seven factors for control were chosen from several perspectives [57]: enterprise size (scale), operational revenue growth rate, leveraging proportion, return on assets, quick ratio, equity concentration, and percentage of unbiased members. The year and industry were also controlled.

Table 1. Variable definitions.

Variable Type	Variable Name	Variable Code	Variable Definitions
Dependent Variable	Carbon Emission Intensity	CEI	Ln (total annual corporate carbon emissions)
Independent Variable	Digital Transformation	DT	(Sum of the related word frequency in the annual report/text length of the annual report MD&A) × 100
Moderating Variables	Internal Control Quality	ICQ	Internal control index/100, internal control index of listed companies from Dibbo database
	Environmental Information Disclosure Quality	EIDQ	Score 25 indicators in seven areas, including environmental management, environmental regulation and certification, and environmental performance and governance, and take the natural logarithm by adding 1 to the total score
	Audit Quality	AQ	The negative absolute value of the difference between the actual auditor’s opinion and the expected probability of an unqualified opinion is taken.
Control Variables	Enterprise Size	Size	Total assets are taken as the logarithm.
	Revenue Growth Rate	Growth	Business liabilities/business assets
	Asset and Liability Ratio	Lev	Total liabilities/total assets
	Return on Assets	Roa	Net profit after tax/total assets
	Quick Ratio	Quick	Quick assets/current liabilities
	Ratio of Independent Directors	Indep	Number of independent directors/number of board of directors
	Shareholding Concentration	Top10	Shareholding ratio of top ten shareholders

details the particular variables and their descriptions.

3.3. Model Construction

A Hausman test was conducted to ensure the fit of the research model. The results demonstrate that this study was more applicable to fixed-effects models; therefore, this study used a fixed-effects model that incorporated fixed effects for both industry and year. The framework examines the relationship between company carbon emissions and digital transformation. H1 is supported if β₁ is negative and satisfies the importance test, demonstrating that digital transformation lowers corporate carbon emissions.

$${CEI}_{i,t}={\beta }_0+{\beta }_1{DT}_{i,t}+{\beta }_k{Control}_{i,t}+{Industry}_i+{Year}_t+{\epsilon }_{i,t}$$

(1)

The subsequent models were created by including the term for interaction and moderating variables of digital transformation to the regression model so as to confirm the regulating mechanisms of internal control quality, environmental information disclosure quality, and audit quality on the effect of digital transformation on business carbon emissions:

$${CEI}_{i,t}={\beta }_0+{\beta }_1{DT}_{i,t}+{{\beta }_2{ICQ}_{i,t}+{\beta }_3{DT}_{i,t}{ICQ}_{i,t}+\beta }_k{Control}_{i,t}+{Industry}_i+{year}_t+{\epsilon }_{i,t}$$

(2)

$${CEI}_{i,t}={\beta }_0+{\beta }_1{DT}_{i,t}+{{\beta }_2{EIDQ}_{i,t}+{\beta }_3{DT}_{i,t}\ast {EIDQ}_{i,t}+\beta }_k{Control}_{i,t}+{Industry}_i+{year}_t+{\epsilon }_{i,t}$$

(3)

$${CEI}_{i,t}={\beta }_0+{\beta }_1{DT}_{i,t}+{{\beta }_2{AQ}_{i,t}+{\beta }_3{DT}_{i,t}\ast {AQ}_{i,t}+\beta }_k{Control}_{i,t}+{Industry}_i+{year}_t+{\epsilon }_{i,t}$$

(4)

In Model (2), if the coefficient of the interaction term β₃ is negative and mathematically significant and β₁ is negative and passes the test of significance, it indicates that the greater the quality of internal control, the more powerful the inhibition of digital transformation on corporate carbon emissions, and H2 holds. Models (3) and (4) share a similar meaning to Model (2) and are not duplicated.

Here, CEI represents total corporate carbon emissions; DT reflects the degree of corporate digitization; ICQ represents internal control quality; EIDQ represents disclosure of environmental data excellence; and AQ represents the moderating variable of audit quality. Control represents the control variables, ${\beta }_1-{\beta }_3$ represent the coefficients of each variable, t represents the year of study, i represents the industry, and ε is the random disturbance term.

4. Empirical Analysis Results

4.1. Descriptive Statistics

According to the company descriptive data presented in

Table 2. Descriptive statistics.

Variable	N	Mean	Median	SD	Max	Min
CEI	13,866	11.48	11.34	1.289	16.50	8.057
DT	13,866	2.981	2.813	0.993	5.920	0
ICQ	13,866	6.485	6.502	0.117	6.849	5.291
EIDQ	13,866	8.429	7.895	4.981	52.48	0
AQ	13,866	−0.0225	−0.0115	0.0573	−0.00230	−0.970
Size	13,866	22.27	22.07	1.266	26.61	19.93
Lev	13,866	0.406	0.399	0.193	0.901	0.0338
Roa	13,866	0.0495	0.0432	0.0542	0.250	−0.250
Quick	13,866	1.941	1.271	2.053	21.11	0.149
Growth	13,866	0.175	0.119	0.333	2.591	−0.542
Indep	13,866	37.58	33.33	5.386	60	30
Top10	13,866	59.36	60.26	14.57	91.70	21.77

, carbon emissions vary significantly among organizations. The mean value of total corporate carbon emissions is 11.48, ranging from 8.057 to 16.5. The sample firms’ digital transformation index, which serves as the independent variable, has an average of 2.981 and a standard deviation of 2.813. The variance is 0.117, the mean is 6.485, and the minimum, median, and maximum internal control quality scores are 5.291, 6.502, and 6.849, respectively. With the lowest and maximum values of 0 and 52.48, respectively, the average score for disclosure of environmental data quality is 8.429, with a standard deviation of 4.981. The sample firms’ overall audit quality is fair, as indicated by the mean value of −0.0225 and the median of −0.0115. This finding suggests that firms vary considerably in the quality of their green disclosures. Furthermore, the variables are chosen from a reasonable range and do not contain any noteworthy indicators or outliers, which goes against the regression concept. Overall, the listed Chinese firms differ substantially in their level of digitization, which is consistent with reports in the literature thus far [69]. The data utilized in this investigation meet these requirements.

4.2. Correlation Analysis

As shown in

Table 3. Correlation analysis.

	CEI	DT	ICQ	EIDQ	AQ	Size	Lev	Roa	Quick	Growth	Indep	Top 10
CEI	1
DT	−0.026 ***	1
ICQ	0.164 ***	−0.017 **	1
EIDQ	0.219 ***	−0.150 ***	0.035 ***	1
AQ	0.016 *	0	0.229 ***	0.026 ***	1
Size	0.936 ***	−0.00800	0.144 ***	0.234 ***	0.00600	1
Lev	0.584 ***	−0.056 ***	−0.00100	0.087 ***	−0.212 ***	0.557 ***	1
Roa	0.009	0.027 ***	0.285 ***	0.017 **	0.357 ***	−0.070 ***	−0.397 ***	1
Quick	−0.426 ***	0.038 ***	0.00200	−0.102 ***	0.119 ***	−0.356 ***	−0.661 ***	0.276 ***	1
Growth	0.099 ***	0.051 ***	0.173 ***	−0.063 ***	0.091 ***	0.022 ***	0.042 ***	0.256 ***	−0.044 ***	1
Indep	0	0.079 ***	0.014 *	−0.054 ***	−0.00300	0	0.00100	−0.019 **	0.014 *	−0.00100	1
Top 10	0.117 ***	0.042 ***	0.140 ***	0.008	0.130 ***	0.115 ***	−0.081 ***	0.207 ***	0.111 ***	0.055 ***	0.066 ***	1

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

, before analyzing the effect of enterprise digital transformation on corporate carbon emissions, a Pearson correlation test was performed to test for multicollinearity and to understand the correlation between the variables. The table demonstrates a substantial relationship between digital transformation and corporate carbon emissions. Through a correlation value of −0.026 (1%), the data show a significant and negative relationship between the independent variable (enterprise digital transformation) and the dependent variable (corporate carbon emissions), which lends support to the hypothesis that digitization reduces carbon emissions. There is a significant negative correlation between the dependent variable (CEI) and independent variable (corporate digital transformation). The factors that influence variance inflation all have values less than 3, with an average of 1.49. Thus, convergence is minimal in terms of the study’s primary findings.

4.3. Regression Results and Analysis

According to the Hausman test, the p-value is less than 0.05. Therefore, a two-way fixed-effects model controlling for year and industry was chosen for the empirical analysis. The results of the regression analysis are presented in

Table 4. Regression results.

Variables	(1)	(2)	(3)	(4)
	CEI	CEI	CEI	CEI
DT	−0.0269 ***	−0.0334 ***	−0.0259 ***	−0.0297 ***
	(−4.99)	(−3.83)	(−3.31)	(−5.38)
ICQ		−0.00561
		(−1.05)
DT × ICQ		−0.0563 **
		(−2.20)
EIDQ			−0.00109
			(−0.08)
DT × EIDQ			−0.0190 **
			(−2.43)
AQ				0.252
				(1.48)
DT × AQ				−0.126 **
				(−2.34)
Size	0.820 ***	0.820 ***	0.822 ***	0.823 ***
	(87.20)	(86.21)	(86.23)	(87.13)
Lev	0.326 ***	0.326 ***	0.325 ***	0.317 ***
	(8.38)	(8.32)	(8.30)	(8.09)
Roa	1.995 ***	2.013 ***	1.983 ***	2.047 ***
	(26.29)	(25.96)	(26.02)	(25.82)
Quick	−0.0375 ***	−0.0375 ***	−0.0373 ***	−0.0378 ***
	(−14.92)	(−14.89)	(−14.82)	(−15.02)
Growth	0.292 ***	0.296 ***	0.292 ***	0.291 ***
	(32.39)	(32.47)	(32.23)	(32.31)
Indep	−0.000145	−0.000128	−0.0000841	−0.000159
	(−0.18)	(−0.16)	(−0.10)	(−0.20)
Top 10	−0.000901 **	−0.000871 **	−0.000921 **	−0.000910 **
	(−2.14)	(−2.05)	(−2.16)	(−2.16)
_cons	−6.834 ***	−6.432 ***	−6.845 ***	−6.886 ***
	(−26.72)	(−21.21)	(−26.61)	(−26.85)
Industry FE	Yes	Yes	Yes	Yes
Year FE	Yes	Yes	Yes	Yes
N	13,866	13,866	13,866	13,866
R2	0.698	0.697	0.697	0.699

Note: t statistics in parentheses, ** p < 0.01, *** p < 0.001.

. Column (1) shows a negative and significant correlation between the level of CEI and digital transformation (DT) with a correlation coefficient of −0.0269 (1% level). This indicates that the digital transformation of the firms reduces the total carbon emissions. Reducing carbon emissions is made easier by digital transformation, thus bolstering digital transformation; advancing product intelligence; facilitating the incorporation of the Web, big data, and machine learning in organizations; decreasing energy usage; increasing energy use efficiency; and fostering the peaceful coexistence of humans and nature. This is consistent with the Sustainable Development Goals put forth by the UN, which state that new energy materials and technological advancements created by businesses utilizing digital technology can increase the sustainability and efficiency of energy systems [70,71]. Therefore, hypothesis 1 is further supported.

In column (2), the dependent variable, corporate carbon emissions, is significantly negatively correlated with the independent variable, digital transformation (DT), with a correlation coefficient of −0.0334 (1% level). Meanwhile, the interaction term of digital transformation (DT) and internal control quality (ICQ) is significantly negatively correlated with total corporate carbon emissions at the 5% level, with a regression coefficient of −0.0563. This suggests that internal control quality (ICQ) has a positive moderating effect on digital transformation (DT) to reduce corporate carbon emissions. Good internal oversight is beneficial to business decision-making, thereby shaping and cultivating a superior internal control culture and control environment. Thus, the enterprise’s environmental protection concepts have been implemented to reduce carbon emissions. Therefore, hypothesis 2 is supported.

In column (3), the dependent variable, corporate carbon emissions, is significantly negatively correlated with the independent variable, digital transformation (DT), with a correlation coefficient of −0.0259 (1% level). Meanwhile, the interaction term of digital transformation (DT) and environmental information disclosure quality (EIDQ) is significantly negatively related to carbon emissions at the 5% level. The regression coefficient is −0.019, which indicates that environmental information disclosure quality (EIDQ) positively moderates digital transformation (DT) to reduce corporate carbon emissions. Environmental information disclosure helps investors, stakeholders, and lawmakers understand the current condition of a business’s ecological emission reductions. It also encourages companies to actively engage in environmental management efforts to reduce emissions and save energy. Therefore, hypothesis 3 is supported.

In column (4), the dependent variable, corporate carbon emissions, is significantly negatively correlated with the independent variable, digital transformation (DT), with a correlation coefficient of −0.0297 (1% level). Meanwhile, the interaction term of digital transformation (DT) and audit quality (AQ) is significantly negatively correlated with corporate carbon emissions at the 5% level, with a regression coefficient of −0.126. This indicates that audit quality plays a positive moderating role in reducing carbon emissions through digital transformation (DT). Through scientific and standardized methods and approaches that enhance the standard of government environmental audits, in addition to establishing environmentally friendly regulations to expose any potential problems, the timely disclosure of enterprises’ use of environmental protection funds, configurations, and other green development issues promotes a decrease in carbon emissions by focusing on the implementation of environmentally friendly decisions. Therefore, hypothesis 4 is supported.

4.4. Robustness Test

Three methods were employed for robustness testing. In addition to re-running the regression analysis after removing the samples for the pandemic year, considering that the degree of digital transformation varies significantly between industries, a regression analysis was carried out by studying a single industry sample to test the reliability of the results. To further verify these findings and solve endogenous problems, we used instrumental variables and conducted a two-stage least squares (2SLS) regression to study the robustness of the main effect regression.

4.4.1. Removing Samples for the 2020 Pandemic Year for Robustness Testing

In late 2019 and early 2020, the COVID-19 pandemic overtook the globe. In response, all countries adopted blockade and quarantine policies. During this period, many businesses ceased production, and travel activities were restricted, leading to significant changes in carbon and other GHG emissions worldwide. According to the Global Energy Review: CO2 Emissions 2020 report issued by the International Energy Agency, worldwide energy-related CO₂ emissions decreased by 5.8% as a result of COVID-19, marking the highest annual fall since World War II. Owing to decreased energy demand, worldwide energy-related CO₂ emissions in 2020 were approximately 2 billion tons lower than those in the previous year. Of these, CO₂ emissions from the transportation sector owing to oil use alone were reduced by 1.1 billion tons.

To ensure the accuracy of the results, the decision was made to remove the samples from 2020, a year with a sudden force majeure event, and rerun the regression test. The results shown in

Table 5. Robustness test: removing 2020 samples.

Variables	(1)	(2)	(3)	(4)
	CEI	CEI	CEI	CEI
DT	−0.026 ***	−0.027 **	−0.026 **	−0.037 ***
	(−3.23)	(−2.02)	(−2.11)	(−3.91)
ICQ		−0.002
		(−0.25)
DT × ICQ		−0.072 **
		(−1.98)
EIDQ			−0.006
			(−0.29)
DT × EIDQ			−0.029 **
			(−2.45)
AQ				0.597
				(0.73)
DT × AQ				−0.684 **
				(−2.19)
Size	0.752 ***	0.755 ***	0.759 ***	0.756 ***
	(45.72)	(45.52)	(45.78)	(45.60)
Lev	0.393 ***	0.395 ***	0.397 ***	0.380 ***
	(6.24)	(6.25)	(6.29)	(5.95)
Roa	2.040 ***	2.072 ***	2.027 ***	2.128 ***
	(16.30)	(16.29)	(16.15)	(14.99)
Quick	−0.035 ***	−0.035 ***	−0.035 ***	−0.036 ***
	(−8.29)	(−8.21)	(−8.26)	(−8.38)
Growth	0.263 ***	0.267 ***	0.262 ***	0.262 ***
	(21.51)	(21.55)	(21.32)	(21.37)
Indep	0.002 *	0.002	0.002 *	0.002 *
	(1.67)	(1.64)	(1.75)	(1.71)
Top10	−0.001	−0.001	−0.001	−0.001
	(−1.32)	(−1.35)	(−1.57)	(−1.46)
_cons	−5.407 ***	−4.994 ***	−5.493***	−5.463 ***
	(−15.35)	(−11.87)	(−15.56)	(−15.40)
Industry FE	Yes	Yes	Yes	Yes
Year FE	Yes	Yes	Yes	Yes
N	8199	8199	8199	8199
R2	0.583	0.584	0.585	0.583

Note: t statistics in parentheses, * p < 0.05, ** p < 0.01, *** p < 0.001.

indicate that the regression coefficient of the independent variable digital transformation on firms’ carbon emissions, shown in column (1) of Table 5, is −0.026 at the 1% level of significance, which is consistent with the results of the previous regression analyses, suggesting that digital transformation reduces firms’ carbon emissions. With the inclusion of the moderating variables in columns (2) to (4), the three interaction terms are significant at the 5%, 5%, and 1% levels. These findings are consistent with the results of the regression analysis in this study. With the involvement of the moderating variables above, the contribution of corporate digital transformation to reducing corporate carbon emissions is further enhanced.

4.4.2. Robustness Test: Taking a Single Industry from the Total Sample as a Mechanism Sample

Considering that the degree of digital transformation varies considerably between industries, we divided the total sample according to industry. Among all industries, manufacturing pays particular attention to digital transformation, so these enterprises in the total sample were regressed to verify the results. According to the regression results in

Table 6. Robustness test: manufacturing sample.

Variables	(1)	(2)	(3)	(4)
	CEI	CEI	CEI	CEI
DT	−0.024 ***	−0.030 ***	−0.025 ***	−0.027 ***
	(−4.09)	(−2.02)	(−2.11)	(−4.51)
ICQ		−0.005
		(−0.91)
DT × ICQ		−0.068 **
		(−2.53)
EIDQ			−0.005
			(−0.34)
DT × EIDQ			−0.019 **
			(−2.27)
AQ				0.133
				(0.71)
DT × AQ				−0.190 **
				(−2.09)
Size	0.820 ***	0.820 ***	0.823 ***	0.822 ***
	(81.36)	(80.61)	(80.74)	(81.21)
Lev	0.361 ***	0.363 ***	0.362 ***	0.360 ***
	(8.69)	(8.67)	(8.66)	(8.62)
Roa	1.969 ***	1.993 ***	1.955 ***	1.982 ***
	(24.56)	(24.36)	(24.28)	(23.33)
Quick	−0.035 ***	−0.035 ***	−0.034 ***	−0.036 ***
	(−14.06)	(−14.03)	(−13.97)	(−14.11)
Growth	0.286 ***	0.289 ***	0.285 ***	0.285 ***
	(28.46)	(28.58)	(28.29)	(28.41)
Indep	−0.0003	−0.0002	−0.0002	−0.0003
	(−0.31)	(−0.28)	(−0.26)	(−0.30)
Top 10	−0.001 **	−0.001 **	−0.001 **	−0.001 **
	(−2.09)	(−1.99)	(−2.16)	(−2.11)
_cons	−6.830 ***	−6.359 ***	−6.843 ***	−6.848 ***
	(−31.71)	(−31.67)	(−31.55)	(−31.70)
Year FE	Yes	Yes	Yes	Yes
N	5766	5766	5766	5766
R2	0.713	0.715	0.713	0.711

Note: t statistics in parentheses, ** p < 0.01, *** p < 0.001.

, they are all significant at the 1% level, and the interaction terms are also all significant, which confirms the previous regression results and further validates their stability.

4.4.3. Robustness Test Based on the Two-Stage Least Squares Method

This study may have the endogeneity problem of reverse causality, which means that corporate carbon emissions may inhibit digital transformation. On the contrary, digital transformation may have an inhibitory effect on corporate carbon emissions. To overcome the bias of the empirical results caused by this potential endogeneity problem, we referred to Gao and Jin [72], selected the one-period lag of digital transformation as the instrumental variable, and used the two-stage least squares (2SLS) method to test the robustness of the results.

The 2SLS approach’s first- and second-stage regression models are represented by Equations (5) and (6), respectively.

$${DT}_{i,t}={\beta }_0+{\beta }_1{LDT}_{i,t}+{\beta }_k{Control}_{i,t}+{Industry}_i+{Year}_t+{\epsilon }_{i,t}$$

(5)

$${CEI}_{i,t}={\beta }_0+{\beta }_1{DT}_{i,t}+{\beta }_k{Control}_{i,t}+{Industry}_i+{Year}_t+{\epsilon }_{i,t}$$

(6)

where LDT is one period behind DT.

Table 7. Robustness test: results of 2SLS.

	First Stage DT	Second Stage CEI
LDT	0.338 ***
	(33.06)
DT		−0.0598 ***
		(−3.05)
Size	0.076 ***	0.800 ***
	(3.88)	(62.39)
Lev	−0.092	0.316 ***
	(−1.16)	(6.14)
Roa	0.491 ***	1.906 ***
	(3.43)	(20.41)
Quick	−0.0002	−0.0346 ***
	(−0.03)	(−9.27)
Growth	0.054 ***	0.311 ***
	(2.89)	(25.59)
Indep	−0.002	−0.00115
	(−1.53)	(−1.16)
Top10	0.003 ***	−0.000737
	(3.75)	(−1.31)
Constant	−0.423	−6.212 ***
	(−0.88)	(−19.85)
Industry FE	Yes	Yes
Year FE	Yes	Yes
N	10280	9940
R2	0.442	0.654
Underidentification test p-value	0.000
Cragg–Donald Wald F statistic	1093.042

Note: t statistics in parentheses, *** p < 0.001.

displays the 2SLS regression findings. After modeling DT and LDT in the initial stage, the coefficient of regression among digital transformation and enterprise carbon emissions is −0.0598 (1% level) in the subsequent phase. In the initial stage, the correlation between DT and LDT is 0.338 (1% level). Furthermore, the Cragg–Donald Wald F-statistic value of 1093.042 with a p-value of 0.000 corroborated the suitability of the instrumental variables and demonstrated their statistical significance, satisfying the criteria of weak indirect factors and non-identifiability. These data confirm H1 once more, revealing that, despite controlling for endogenous variables, digitization remains significantly and adversely connected with corporate carbon emissions.

5. Discussion and Conclusions

5.1. Discussion

This analysis clarified the confusing relationship between technological change and carbon emissions found in prior studies [12,13,14,15], confirming that technological transition is critical to reducing carbon emissions. This study addressed two major global issues, digital transformation and carbon emission reduction, clearly demonstrating a large and beneficial connection between corporate digital transformation and carbon emission reduction at the enterprise level [73]. The capacity to develop energy conservation and emission reduction programs based on national conditions via digital transformation is tremendously useful for countries all over the world.

According to this report, digital transformation can successfully lower carbon emissions at the corporate level, assisting Chinese businesses in resolving energy saving and emission reduction concerns. Many scholars have studied the implications of digital transformation on corporate profitability and environmental responsibility from many perspectives. Moreover, different conclusions have been drawn. For example, Yisa and Taiwen [74] argued that digital transformation significantly promotes corporate sustainable development, while Xu et al. [75] hold the opposite view. This study focused on the environmental protection and security of organizations, as well as worldwide long-term development, rather than utilizing metrics to assess the short-term success of the enterprises under consideration. This study used carbon emission reduction as the variable of interest and investigated the influence of digitization on firms, which is more consistent with having long-term corporate and global aims with broad intentions.

Government environmental subsidies [76] and executives’ environmental perceptions [77] were used as independent variables to investigate their influence on corporate carbon emissions. We argue that company digitization may significantly improve corporate carbon emission reduction by including moderating variables from ecological theories and stakeholder perspectives. Notably, further research into the mechanisms of corporate digital transformation and lowering carbon emissions can offer theoretical and practical recommendations to businesses. This will assist regulators in creating environmental protection policies that are appropriate for local conditions; help clarify the crucial role that the government, media, and senior management play in encouraging corporate reductions in carbon emissions; and motivate businesses to aim for low-carbon growth, practice energy efficiency and emission decrease, and understand green sustainable growth.

5.2. Conclusions

Empirical research was conducted in this study to determine how digitization influences corporate reductions in carbon emissions. The following findings are reached when the moderating variables of audit quality, environmental information disclosure quality, and internal control quality are included. (1) Digital transformation can significantly reduce the carbon emissions of enterprises. The reliability of the results is enhanced by three robustness tests. Reducing carbon emissions is made easier by digital transformation. Through bolstering digital transformation, advancing product intelligence, fostering the comprehensive incorporation of the web, big data, and machine learning in organizations, and decreasing energy usage, efficiency in energy use is increased and the peaceful coexistence of humans and nature is fostered. To promote enterprise carbon emission reduction, social development is steadily improving. (2) The standard of internal procedures has a positive moderating effect on company carbon emissions reduction. Good internal oversight is beneficial to business decision-making, while emphasizing the division of roles and internal checks in the execution process, which can avoid the arbitrary action of senior leaders in the decision-making mode and process and enhance the management’s motivation to avoid environmental risks, thereby shaping and cultivating a superior internal control culture and control environment. (3) Reducing carbon emissions is positively regulated by the quality of environmental information dissemination. Environmental regulators will be better able to comprehend the information provided by businesses on environmental pollution thanks to environmental information disclosure, and their regulatory pressure will compel businesses to implement green innovation and lower carbon emissions. (4) Audit quality has a favorable regulatory impact on business carbon emissions reduction. The improvement of audit quality helps to capture the trajectory of economic power operation, improve the balance and efficiency of local governments’ allocation of public resources, prevent low-carbon development funds from being inhibited by other public products with similar long-term effects, and promote carbon emission reduction.

5.3. Implications

Theoretically, economic development in the age of digitization means prioritizing low-carbon and green development while striving for maximum economic efficiency. Organizational digital transformation and carbon emission reduction are two contemporary research areas of interest. In this study, we used actual corporate data to conduct an empirical investigation of the link between digital transformation and reductions in carbon emissions in enterprises. There is a scarcity of theoretical research that contributes to relevant hypotheses regarding the link between technological change and carbon emission reduction in businesses. In this study, stakeholder, ecological, and digital economy theories were used to investigate the moderating roles of auditing, ecological publication, and internal control quality in corporate digitization and carbon emission reduction. We achieved this by incorporating relevant theories and research literature and contributing to the body of knowledge already available on the topic.

Practically speaking, the digital transformation of Chinese businesses began later than in developed nations, and while the growth trend has been noticeable in recent years, the transformation process has been rather unstable, with many businesses still encountering challenges related to the transformation. An overview of the suggestions drawn from the present study is provided below.

(1) Enterprises: Enterprises should establish an internal control system for the integration of digital transformation and carbon emission reduction and clarify the responsibilities of each department, such as stipulating that the information technology department is responsible for the maintenance of the digital system and the environmental protection department takes the lead in the implementation of carbon emission reduction strategies. Enterprises should carry out internal control experience exchange activities, share details of effective internal control cases, learn from advanced experience, continuously improve their own internal control system, and enhance the management of digital transformation and carbon emission reduction integration. A regulatory mechanism for environmental information disclosure should be established to improve the authenticity and completeness of the information disclosed by enterprises. Moreover, introducing advanced digital audit software and using big data analysis technology to conduct a comprehensive review of financial data and business process data will ensure compliance and maximize the benefits of the use of funds for projects related to digital transformation and carbon emission reduction. For example, data analysis can be used to quickly screen out abnormal energy procurement costs and pinpoint audit suspects. A special audit plan could be formulated for digital transformation and carbon emission reduction, specifying the scope, focus, and frequency of the audit, and regular assessments should be made to determine whether the digital transformation project is progressing as planned and whether the carbon emission reduction measures are being effectively implemented. This can be achieved using methods such as reviewing the progress of the carbon emission reduction technological transformation project in terms of its deviation from the expected results. Strengthening the training of auditors in terms of digital technology and knowledge of carbon emission reduction is also important, with auditors’ skills being upgraded through specialized training courses and seminars.

(2) Government: The government should increase the amount of technical and financial help for corporate digitization activities. An internal control quality supervision mechanism should be established to regularly assess the implementation of the internal control system. Through internal audits and special inspections, the timely rectification of problems and serious violations of accountability can be carried out to ensure the effective operation of the internal control system. Unified environmental information disclosure standards should be formulated, requiring enterprises to disclose the specific contributions made by digital transformation to carbon emission reduction, such as digital energy-saving technologies, the amount of carbon emission reduction achieved, and the accounting method. The reporting format should be unified to improve the comparability of information. In addition, the content of enterprises’ environmental information disclosure should be audited to ensure that the data are true, accurate, and complete. Penalties, such as warnings and fines, can be imposed on enterprises that falsely disclose or conceal important information. This way, social supervision is strengthened, and public reporting is encouraged. Environmental information disclosure training for enterprises should be developed to raise awareness of the importance of disclosure. The government has set up an audit team dedicated to the integration of digital transformation and carbon reduction projects for enterprises and conducts regular audits of enterprise projects to verify compliance in the use of funds, such as tracking the flow of government subsidies for the purchase of digital energy-saving equipment. The application of big data and artificial intelligence in auditing should be promoted, along with establishing audit databases, collecting relevant data from enterprises, and using data analysis and mining technology to analyze trends in carbon emissions and energy consumption indicators before and after the digital transformation of enterprises to accurately identify suspicious activity, such as the possible misrepresentation of carbon emission reduction results. Audit reports should be open and transparent, with enterprises urged to rectify any problems found within a certain period of time. Audit results should be linked to the subsequent policy support of the enterprises. For enterprises that fail to rectify the problems, subsidies and preferential policies for digital transformation and carbon emission reduction projects can be reduced or even suspended.

Public Chinese firms were chosen as the topic of this study since innovations and digital transformations in China have received significant interest in the past few years [78,79,80]. With the listed Chinese companies serving as the research backdrop, the findings are more instructive and have greater research value. China’s innovation and digital transformation are also quite typical and indicative.

5.4. Limitations and Future Research

The ongoing research suffers from various limitations. First, this study only involved companies listed in China’s A-shares as a study sample and does not fully discuss the institutional characteristics of the Chinese market and the impact of the policy environment on the results. In addition, the lack of international comparisons made it difficult to generalize these organizations as global companies. In order to provide an in-depth analysis of the mechanisms exploring the impact of digital transformation on carbon emissions through technology upgrades and management innovations, or to increase the applicability of the research, data from private companies could be added in future research, or comparative studies between countries could be conducted. Another limitation is that the mediating effect has not been fully explored, with only the moderating effect considered. This may cause other key mechanisms to be missed, such as how digital technology can improve energy efficiency and optimize supply chain management. In future studies, it is necessary to study the mechanisms underlying these mediating effects. Finally, this study used a single metric—relevant digital transformation word frequencies—to gauge how much an organization has transformed digitally. In the future, more diverse measurement methods could be used to increase the reliability of this study’s findings.