How Does Digital Transformation Impact ESG Performance in Uncertain Environments?

Abstract

The influence of digital transformation on ESG performance has garnered considerable interest; however, previous research in this area has not adequately considered the influence of environmental uncertainty factors. This study utilized a dataset comprising Chinese A-share listed companies from 2009 to 2023 to explore how environmental uncertainty affects the correlation between digital transformation and ESG performance. Furthermore, we also examined potential pathways and heterogeneity. Our findings demonstrate that digital transformation significantly enhances ESG performance, with the positive effects persisting for up to three years post-implementation, although gradually diminishing in intensity. However, environmental uncertainty substantially reduces this positive impact across all pivotal technologies. Improvements in ESG performance are more pronounced in firms that are high-tech, technology-intensive, and capital-intensive and that do not produce heavy pollution. Quantile regression reveals that firms in the upper–middle ESG performance range benefit most. Our mediation analysis confirms that digital transformation enhances ESG performance by increasing firm value, media attention, and analyst coverage. Overall, this study contributes to the existing literature by providing empirical evidence of the impacts of environmental uncertainty. These findings provide strategic guidance for companies navigating digital transformation initiatives in turbulent business environments, while also offering concrete recommendations for regulatory authorities developing ESG disclosure frameworks and digital infrastructure investment priorities tailored to different uncertainty conditions.

1. Introduction

The concept of ESG emphasizes incorporating various dimensions into investment strategies, including environmental care, social responsibility, and governance practices [1]. In February 2024, the Shanghai, Shenzhen, and Beijing stock exchanges in China released their collaborative “Guidelines for Sustainable Development Reports of Listed Companies”. This pivotal initiative marks the beginning of China’s systematic regulations for mandatory ESG information disclosure by Chinese listed entities. In this situation, the necessity of improving corporations’ intrinsic motivation to sincerely embrace ESG principles is drawing intensified focus from both academia and the investment community.

Due to the current global digital transformation, companies face both significant growth opportunities and challenges. The digital transformation of enterprises, as delineated by Ebert and Duarte [2], involves the radical revamping and enhancement of traditional business operations through the adoption of advanced digital technologies, signifying a significant strategic shift [3]. They highlight that essential digital transformation technologies encompass foundational technologies, including artificial intelligence, blockchain, cloud computing, and big data, as well as the implementation of digital technology practices [4]. These technological innovations are set to bring about major changes in business processes and product manufacture [5], potentially influencing traditional corporate social responsibility practices.

However, the current academic investigations into how digital transformation specifically influences and boosts ESG performance remain limited, with a particular dearth of in-depth analyses examining its effects from diverse technological perspectives. For instance, research by Yang and Han [6] demonstrates how digital transformation positively impacts ESG’s environmental and social components. Wang et al. (2023) [7] observed a positive relationship between digital transformation and overall ESG performance, and the authors of [8] discussed how ESG performance is enhanced through the promotion of green technology innovation and management efficiency and reductions in information asymmetry. In summary, while preliminary research has begun to shed light on the effects of digital transformation on various aspects and functional mechanisms of ESG, yielding valuable findings, it is essential to thoroughly investigate the specific pivotal technological transformations that influence ESG performance. This will provide detailed insights for the effective utilization of digital transformation strategies to support sustainable development.

The realization of digital transformation involves complex technological reforms and significant cost investments, yet its likelihood of success remains modest. Because of this, the academic focus has grown beyond just the benefits that digital transformation brings to enterprises and now includes the problems that they face, what drives them, and what makes them successful [5]. In the aftermath of global occurrences such as the COVID-19 pandemic and US–China trade tensions, the uncertainty of the external environment that enterprises face when on a growth trajectory has escalated. This necessitates greater attention to the evolving dynamics of the business environment while devising digital transformation strategies. In this context, a pertinent inquiry emerges: as external environmental uncertainty intensifies, do enterprises striving to improve their ESG performance and using digital transformation as an internal catalyst face additional hurdles and impediments? Addressing this query holds significant theoretical and practical importance in elucidating how enterprises can achieve sustainable development in a complex and fluctuating market landscape.

Environmental uncertainty pertains to the unpredictability of the actions of stakeholders such as customers, suppliers, competitors, and regulatory agencies, along with uncertainty arising from changes in the external environment [9]. This form of uncertainty can hinder strategic planning and investment decisions within enterprises [10,11], negatively impact firm performance [12], and potentially foster opportunistic behavior among managers [13,14]. Existing research indicates that environmental uncertainty also impacts the “peer effect” during the digital transformation of enterprises [15]. From a stakeholder perspective, heightened environmental uncertainty may compel enterprises to diverge from conventional business strategies, influencing the autonomy of managerial decisions; such erroneous decision-making can precipitate significant shareholder losses. Additionally, embarking on digital transformation amidst uncertainty can escalate data security risks, including those of data breaches and cyberattacks [16,17], while also amplifying compliance risks [18]. These hazards could ultimately tarnish an enterprise’s reputation and ESG performance. Those undergoing digital transformation might focus on technological innovation at the expense of the core tenets of ESG, potentially resulting in inadequate environmental investment and diminished social responsibility. In a highly uncertain business landscape, the decision to allocate substantial resources to digital transformation for the sake of enhancing ESG performance may not be advisable. Thus, due to heightened environmental uncertainty, enterprises should carefully consider digital transformation’s impact on ESG strategies and pursue more resilient and sustainable development trajectories. Existing studies frequently neglect the influence of environmental uncertainty on the connection between digital transformation and ESG performance. Therefore, in the constantly changing market landscape, it is imperative to include new perspectives and factors to explore the non-economic performance brought about by digital transformation.

This study considers the significance of environmental uncertainty and utilizes a dataset comprising Chinese A-share listed enterprises spanning from 2009 to 2023. An empirical analysis was carried out to investigate digital transformation and its pivotal effects on ESG performance. We also investigated the role of environmental uncertainty in these effects. This study provides enterprises with a useful guide on how to balance inputs in digital transformation with improvements in ESG performance in the presence of different types of environmental uncertainty. Figure 1 shows the theoretical framework that underpins the subsequent discussions and investigations throughout this study. To summarize, this study fills a void in the extant literature, provides solid decision-making support for enterprise strategies, and offers valuable insights for regulatory authorities developing ESG disclosure policies and digital infrastructure investment strategies under varying conditions of uncertainty.

This study’s marginal contributions are as follows. First, we evaluate the overall impact of digital transformation and further investigate the critical role of pivotal technologies in shaping ESG outcomes. We analyze the effects of distinct technological changes on ESG performance, offering a unique perspective on the factors that influence ESG performance and providing a valuable reference enabling companies to plan their digital transformation pathways more accurately and effectively. Second, this study innovatively integrates digital transformation with environmental uncertainty and ESG performance within the same research framework. Through this comprehensive perspective, we can more deeply understand how digital transformation affects non-economic performance in conditions of environmental uncertainty, enriching the existing literature and offering solid support for companies to make informed decisions regarding digital initiatives and sustainability practices in a complex and volatile market environment. Third, through a lagged analysis, we demonstrate the temporal persistence of digital transformation benefits, revealing that the positive effects on ESG performance continue for up to three years post-implementation, albeit with gradually diminishing intensity. This longitudinal perspective provides crucial insights into the strategic nature of digital transformation investments, rather than offering a merely tactical perspective. Fourth, our quantile regression analysis uncovers the non-linear relationship between digital transformation and ESG performance across different performance segments, revealing that firms in the upper–middle range of ESG performance benefit the most from digital initiatives—a nuanced finding that challenges simplistic universal approaches to digital transformation. Finally, this research empirically analyzes the mediating roles of firm value, media attention, and analyst coverage and examines the varying effects of digital initiatives on ESG ratings across different industries while exploring the reasons for these differences. In summary, this research provides deeper insights into the conditions necessary to successfully boost ESG performance through digital transformation, offering more precise guidance for corporate operations.

The structure of this work is organized as follows. After thoroughly reviewing the existing research and identifying gaps in the literature, Section 2 formulates the research hypotheses. Section 3 describes the econometric models, variables, and data sources utilized in our research. Section 4 discusses the empirical findings and offers a comprehensive examination of the influence of environmental uncertainty. Section 5 further explores heterogeneity and the mechanisms of influence. Section 6 summarizes the main findings and offers policy implications.

2. Literature Review and Hypothesis Development

2.1. Literature Review

2.1.1. The Determinants of ESG Performance

Prior research has extensively examined the determinants of enterprise ESG performance, identifying both internal and external factors as critical influences. Externally, studies have focused on legislation and regulations, such as the implementation of the “Environmental Protection Tax Law”, which has been shown to significantly enhance ESG performance [19]. Government subsidies also play a vital role in improving ESG outcomes [20,21]. Additionally, investor attention acts as a monitoring mechanism, encouraging better ESG practices [22].

Internally, factors such as the ownership structure, company size, and executive characteristics are pivotal. For instance, state-owned enterprises often outperform non-state-owned enterprises in ESG achievements, as the latter are more profit-driven [23]. Larger firms tend to be more committed to sustainable development and transparency [24], while the gender and educational backgrounds of executives influence their environmental consciousness, thereby affecting ESG decisions [25,26]. These studies provide a foundation for understanding the drivers of ESG performance.

2.1.2. The Relationship Between Digital Transformation and ESG Performance

Due to the Fourth Industrial Revolution and the rise of artificial intelligence, the digital economy has emerged as a crucial driver of both economic growth and sustainability [27]. Here, we examine the relationship between digital transformation and ESG performance from both the macro and firm levels.

At the macro level, digital transformation accelerates low-carbon city transitions and fosters high-quality development. Studies show that widespread digital infrastructure and ICT significantly optimize industrial structures, enhance energy efficiency, and spur green innovation. For example, according to Li et al. (2024) [28], urban digital infrastructure in China has a significant positive effect on green innovation, mainly by facilitating talent agglomeration, increasing R&D investment, and promoting industrial upgrading. The authors of [29] found that the “Broadband China” policy enhanced the urban energy efficiency and catalyzed green technology innovation, and the authors of [30] provide empirical evidence that the digital economy boosts the carbon total factor productivity (CTFP) by transforming production methods and strengthening environmental governance. Furthermore, Zhang et al. (2023) [31] showed that digital transformation optimizes industry structures and elevates the overall ESG levels alongside high-quality development.

At the firm level, extensive research has documented the positive effects of digital transformation on ESG performance through multiple pathways. First, digital transformation enhances information transparency and the internal control efficiency. Wu et al. (2024) [32] found that, in manufacturing firms, digital transformation significantly improves ESG scores by streamlining monitoring processes and reducing information asymmetries. Chen et al. (2024) [33] note that digital technologies effectively mitigate “ESG decoupling” by reducing the gap between the reported and actual performance.

Second, digital transformation fosters innovation and operational improvements that directly benefit ESG outcomes. Xie et al. (2020) [27] identified multiple mediation channels, including transparency, green innovation, and internal governance improvements. Zhang and Huang (2024) [34] highlight that digital tools bolster supply chain resilience, thereby improving ESG outcomes, particularly in the social dimension. Digital transformation enables firms to optimize their structures and boost their technological innovation, thereby fostering high-quality development [35].

Third, digital transformation enhances corporate governance mechanisms. An analysis indicates that digital technology innovation further improves enterprises’ ESG performance by fostering green innovation and enhancing the corporate governance structure [36]. Wei and Zheng (2024) [37] emphasize that digital tools can improve governance structures and decision-making transparency, leading to superior ESG practices through improved board effectiveness and accountability.

However, the relationship between digital transformation and ESG appears more complex than a simple positive correlation. Several studies suggest non-linear or conditional effects that merit further investigation. The research in [7,38] documents an inverted U-shaped effect, whereby initial digital transformation improvements boost ESG performance, but, once digital transformation surpasses a threshold, managerial complexity and rising technology costs may dampen the ESG gains. This suggests that moderate levels of digital transformation may yield optimal ESG benefits.

Another concern highlighted in the literature is the potential decoupling between the ESG disclosure quality and actual environmental performance. Kotlarsky (2023) [39] found that enhanced digital reporting systems can sometimes mask underlying sustainability issues, while the authors of [40] note that rating systems enabled by digital technologies can be misaligned with real outcomes such as carbon emissions, raising the risk of greenwashing. These findings underscore the need for regulators to strengthen their ESG disclosure oversight to ensure accuracy and verifiability [41].

Implementation challenges may also limit the ESG benefits of digital transformation. The early stages often involve substantial upfront investments—upgrading infrastructure and training personnel—which can strain financial and operational resources and crowd out short-term ESG spending [41]. Despite the efficiency and productivity gains of digital transformation, emerging cybersecurity risks—including data breaches and cyberattacks [42]—may impede the effective alignment of digital strategies with ESG objectives. Additionally, delayed returns from initial digital investments can reduce firms’ willingness and capacity to sustain ESG investments, exacerbating financing constraints and impeding long-term sustainability objectives [43].

Finally, various digital technologies affect distinct dimensions of ESG performance. While artificial intelligence and big data analytics appear particularly beneficial for environmental monitoring and reporting, blockchain technologies may have stronger effects on governance transparency. Cloud computing seems to offer broader benefits across all ESG dimensions by enabling more comprehensive data integration and analysis. These heterogenous impacts warrant deeper investigation in order to help firms to optimize their technological investments for specific ESG goals.

2.1.3. Research Gaps

Despite extensive evidence of a positive relationship between digital transformation and ESG performance, key gaps persist in the literature, which our study aims to address.

First, the moderating effects of external uncertainty—whether related to the environment, policy, or the market—have received scant attention, and it is therefore unclear how firms’ digital strategies fare under volatile conditions. While studies acknowledge that the context matters, there is no systematic examination of how environmental uncertainty specifically moderates the digital transformation–ESG relationship. This gap is particularly important given the increasing volatility in global business environments.

Second, research remains fragmented across various analytical levels. Macro-level analyses explore the role of national or urban digitization in green development [28,30], while micro-level studies examine firm-level digital adoption and ESG outcomes [32,33]; however, few efforts integrate these perspectives into a unified framework that accounts for both firm capabilities and environmental conditions. This integration is essential in developing more comprehensive theories of the impact of digital transformation on sustainability.

Third, digital transformation is frequently treated as a single entity in empirical studies, obscuring the heterogeneous effects that distinct technologies (e.g., AI, big data, and cloud computing) may exert on environmental, social, and governance dimensions. Understanding these nuanced relationships would help firms to prioritize investments in specific digital technologies based on their ESG objectives and the industry context.

Fourth, although several internal mediation channels have been proposed [44], external transmission mechanisms—specifically capital market valuation, public opinion via media attention, and financial intermediation through analyst coverage—have not been systematically examined. These external mechanisms are particularly important because they indicate how digital transformation may influence ESG performance through market-based incentives and monitoring systems.

Finally, most studies employ cross-sectional approaches or limited time horizons, providing little insight into the temporal dynamics of digital transformation’s ESG benefits. The persistence of these effects over time, whether they strengthen or diminish, and how they vary across different stages of digital maturity remain underexplored.

To bridge these gaps, we introduce environmental uncertainty as a key moderator of the digital–ESG nexus; construct a triple-mediation framework incorporating firm value, media attention, and analyst coverage; and employ instrumental variable techniques and long-lag specifications to address endogeneity and explore the sustained impact of digital transformation on ESG performance. To provide a more comprehensive understanding of the relationship between digital transformation and ESG, we not only examine the concurrent effects but also investigate the long-term impacts through a lagged analysis and explore the differential effects across various ESG performance levels through quantile regression. This multi-faceted approach allows us to uncover both the temporal persistence of digital transformation’s benefits and its heterogeneous effects across firms with different ESG performance foundations, providing a more nuanced and contextually relevant understanding of the relationship between digital transformation and ESG.

2.2. Hypothesis Development

2.2.1. Digital Transformation and ESG Performance

Digital transformation is of considerable importance to businesses in the contemporary competitive landscape. Research consistently shows that it effectively improves corporate governance and internal control levels [45,46], enhances financial performance [47,48], strengthens corporate green innovation capabilities, and reduces the corporate carbon emission intensity [49,50]. These broad benefits suggest multiple pathways through which digital transformation might influence ESG performance. Previous studies have investigated how the extent of enterprise digital transformation affects ESG performance [6,7,8]. However, the drive for the total level of digital transformation is largely dependent on fundamental changes in its pivotal technologies. The impact of digital transformation on ESG performance, including these pivotal technologies, can be distilled into three primary aspects.

First, digital transformation is crucial in enhancing overall performance and promoting sustainable growth. Zhai et al. (2022) found that digital transformation markedly improves operational efficiency and the success rate of innovations, leading to a considerable enhancement in overall firm performance [48]. Guo and Xu (2021) additionally confirmed that digital transformation enhances the efficiency of core business operations, leading to increased firm performance [51]. Moreover, through extensive case studies, Lui et al. (2022) demonstrated that adopting the use of artificial intelligence in transformation strategies optimizes business tactics and increases firm value, showcasing notable commercial benefits [52]. Meanwhile, Al-Zoubi (2017) found that employing cloud computing technology not only improves firm performance but also significantly cuts down on software and hardware expenses [53]. These dual advantages of value and efficiency provide stronger financial backing for firms to meet their ESG obligations. Most importantly, digital transformation contributes to facilitating green innovation and strengthening enterprises’ sustainable development capabilities [49], offering solid technical support for ESG practices. Overall, digital transformation is crucial in enhancing the financial health of businesses, strengthening their capacity for sustainable development, and thus increasing their contributions to environmental preservation and sustainable progress.

Second, digital transformation is also essential in mitigating information asymmetry, expanding supervision mechanisms, and aligning with stakeholder expectations. Studies by Chen et al. (2023) [54] and Zhao et al. (2023) [46] show that enterprise digital transformation significantly boosts the corporate governance quality and internal control efficacy, consequently reducing information asymmetry. Additionally, the progress that has been made in pivotal digital transformation technologies, such as cloud computing and big data, has endowed companies with strong information processing capabilities that underpin their ability to disclose high-quality sustainability reports [55]. Furthermore, the media and analysts are paying greater attention to the digital transformations of enterprises, enhancing the impacts of external oversight. Such external supervisory mechanisms motivate companies to pursue sustainable development and commit to their social responsibilities [56,57]. Concurrently, aligning with modern development trends, digital transformation also secures stakeholder support and recognition more readily [4], contributing to improved enterprise ESG performance.

Third, digital transformation enhances enterprises’ inherent drive to increase their ESG investments, propelling them to take a more proactive approach to fulfilling ESG obligations. Xiao et al. (2021) [58] posit that corporate digitization enables the precise identification of societal environmental concerns through sophisticated technological means. This transformation facilitates the discovery of impactful and viable approaches to ESG initiatives, consequently boosting corporations’ willingness to voluntarily assume social responsibilities. Furthermore, digital transformation enhances the effective utilization of data assets, allowing firms to delve deeper into potential consumer demands and improve their performance with regard to social responsibility; for instance, artificial intelligence and machine learning enable firms to better respond to consumer expectations and understand their preferences [59]. Firms that are adept in using these pivotal technologies gain a competitive edge by elevating their product and service quality and customer satisfaction. This, in turn, increases their eagerness to actively engage in ESG responsibilities. Moreover, digital transformation amplifies market optimism regarding corporate prospects, setting elevated societal standards for corporate social responsibility. Faced with the need to maintain legitimacy due to external pressures, enterprises endeavor to safeguard their brand image and reputation, resulting in a heightened emphasis on their ESG performance.

Based on the above theoretical analysis and empirical evidence, we propose the following.

Hypothesis 1.

Enterprise digital transformation (artificial intelligence, blockchain, cloud computing, big data, and digital technology application, etc.) will positively affect ESG performance.

2.2.2. Digital Transformation, Environmental Uncertainty, and ESG Performance

As uncertainty factors accumulate in the global business environment, it is becoming imperative for enterprises to consider external environmental changes when planning digital transformation initiatives. These changes could affect both the strategy and effectiveness of digital transformation, extending their impact to practices concerning enterprise ESG responsibilities.

First, despite its significant role in driving corporate growth, digital transformation entails substantial expenses and potential hazards. To keep up with advancing technological trends, the digital transformation process requires significant financial investments; however, environmental uncertainty can adversely impact enterprises’ operational status and firm performance [12], escalate their financing pressures [11], and potentially lead to investment inefficiencies [11]. Given that ESG activities similarly require financial backing, resources earmarked for digital transformation initiatives could detract from investments intended for ESG activities under uncertain conditions. Amid the surge in digital transformation, some companies might excessively prioritize technological advancement, neglecting the core principles of ESG; this could lead to companies overlooking investments in environmental protection and social responsibilities in favor of technological progress. Furthermore, digital transformation may intensify data security [17] and regulatory compliance risks [18], which could be exacerbated in uncertain settings, consequently tarnishing a company’s reputation and ESG performance.

Second, digital transformation necessitates considerable alterations to business operations and organizational structures. With the advent of novel technologies and shifts in business paradigms, corporations face challenges in fully anticipating the ramifications of these transformations. The uncertainty of the external environment may impede the ability of managers to make accurate assessments of their surroundings. Chen et al. (2021) observe that, in the context of environmental uncertainty, firms tend to emulate the digital transformation initiatives of their counterparts as a strategy to mitigate the repercussions of collective penalization mechanisms [15]. This “peer effect”, driven more by the pursuit of organizational legitimacy than operational considerations, suggests that environmental uncertainty might compromise the independence of managerial decision-making, leading them to stray from their strategic objectives and increasing the risk of errors during the journey towards digital transformation. Furthermore, the “peer effect” in digital transformation may induce managers’ decision-making errors and complicate investors’ assessment of corporate growth prospects. This situation could heighten the investment risks, undermine stakeholder interests, and adversely impact ESG performance.

Third, environmental uncertainty might amplify managerial opportunistic behaviors and diminish digital transformation’s mitigating effect on information asymmetry. Prior studies have suggested that related digital technologies can provide managers with the means to exaggerate results, resulting in increased opportunities for earnings manipulation [60] and increasing the danger of stock prices declining as a result of human factors [21]. Environmental uncertainty amplifies the pressure on management to perform, potentially increasing their tendency to engage in earnings management [14]. This can result in the early recognition of future earnings [13], exacerbating the disparity in information between managers and stockholders. Consequently, in such conditions of environmental uncertainty, management may undertake digital transformation efforts with a possible bias towards manipulating earnings rather than pursuing genuine ESG improvements.

Drawing on these theoretical mechanisms and empirical evidence, we propose the following.

Hypothesis 2.

Environmental uncertainty will reduce the positive relationship between enterprise digital transformation and ESG performance.

2.2.3. The Mediating Effects of Firm Value, Media Attention, and Analyst Coverage

To further elucidate the mechanisms through which corporate digital transformation influences ESG performance, this study introduces three pivotal mediating variables—firm value, media attention, and analyst coverage—to examine the transmission pathways from the perspectives of resource acquisition, external oversight, and the capital market response. This framework not only improves our understanding of the underlying logic of these influence pathways but also enhances the explanatory power and policy relevance of the research findings.

Digital transformation improves operational efficiency and profitability, thereby increasing firm value [48]. From a resource-based point of view, higher firm value signals greater deployable resources that support ESG investments. The authors of [24] posit that firms with abundant resources are more inclined to undertake environmental and social responsibilities. Moreover, enhanced firm value bolsters corporate reputation under legitimacy pressures, motivating firms to strengthen their ESG practices.

Digital initiatives and technological innovations tend to attract substantial media and public attention [61]. As a salient external monitoring mechanism, media attention amplifies stakeholder expectations and social scrutiny [56]. Under conditions of high visibility, firms are pressured to uphold their legitimacy and reputation, prompting more proactive ESG engagement.

By improving information transparency and data quality, digital transformation lowers the costs required for analysts to gather firm-specific information [62], thereby expanding analyst attention and reinforcing governance oversight [57]. Sustained analyst scrutiny transmits ESG signals and exerts market pressure on firms to enhance their ESG practices.

Considering the logical progression outlined above, the following third set of hypotheses is advanced.

Hypothesis 3a.

Digital transformation positively affects ESG performance through the enhancement of firm value.

Hypothesis 3b.

Digital transformation positively affects ESG performance by increasing media attention.

Hypothesis 3c.

Digital transformation positively affects ESG performance by increasing analyst coverage.

3. Methodology

3.1. Empirical Models

This study constructs Equation (1) to test the relationship between enterprise digital transformation and ESG performance:

$${\mathrm{ESG}}_{\mathrm{i},\mathrm{t}}={\alpha }_0+{\alpha }_1{\mathrm{Dig}}_{\mathrm{i},\mathrm{t}}+{\alpha }_2{\mathrm{E}\mathrm{u}}_{\mathrm{i},\mathrm{t}}+{\delta }_{\mathrm{j}}{\mathrm{Controls}}_{\mathrm{i},\mathrm{t}}+{\mu }_{\mathrm{i}}+{\theta }_{\mathrm{t}}+{\epsilon }_{\mathrm{i},\mathrm{t}}$$

(1)

where Dig represents enterprise digital transformation and its pivotal technologies’ transformation; ESG is ESG performance; and Eu represents environmental uncertainty. Control refers to the control variables; ε is the random error term; and μ and θ, respectively, represent the industry fixed effects and time fixed effects. Additionally, the regression analysis uses clustered robust standard errors at the firm level to calculate t-statistics. This study mitigates the potential endogeneity problem caused by omitted variable bias by selecting the following variables that may impact ESG performance as control variables [6,8]: firm size (Size), measured by ln(Total Assets); return on assets (Roa), calculated as net profit/total assets; listing age (Age), computed using ln(current year-listing year + 1); financial leverage (Lev), calculated as total liabilities/total assets; cash holdings (Cash), evaluated by cash equivalents/total assets; the shareholding proportion of the largest shareholder (Top), measured by the largest shareholder’s shares/total shares; board size (Bsize), calculated as ln(number of directors); the proportion of independent directors (Outr), calculated as independent directors/total directors; and property rights (Gov), which is indicated by 1 for state-owned enterprises and 0 otherwise.

To investigate environmental uncertainty’s influence on the relationship between business digital transformation and ESG performance, we construct Equation (2):

$${\mathrm{ESG}}_{\mathrm{i},\mathrm{t}}={\beta }_0+{\beta }_1{\mathrm{Dig}}_{\mathrm{i},\mathrm{t}}+{\beta }_2{\mathrm{Eu}}_{\mathrm{i},\mathrm{t}}+{\beta }_3{\mathrm{Dig}\times \mathrm{Eu}}_{\mathrm{i},\mathrm{t}}+{\delta }_{\mathrm{j}}{\mathrm{Controls}}_{\mathrm{i},\mathrm{t}}+{\mu }_{\mathrm{i}}+{\theta }_{\mathrm{t}}+{\epsilon }_{\mathrm{i},\mathrm{t}}$$

(2)

where $\mathrm{D}\mathrm{i}\mathrm{g}\times \mathrm{E}\mathrm{u}$ is the interaction of digital transformation and environmental uncertainty, including the interaction of its pivotal technologies’ transformation and environmental uncertainty.

To examine the mediating channels through which digital transformation enhances ESG performance, we employ three proxy variables: Tobin’s Q (Tq) for firm value; the natural logarithm of one plus the annual count of firm-related news reports (Media) for media attention; and the natural logarithm of one plus the number of financial analysts covering the firm (Ana) for analyst coverage. Data on media attention and analyst coverage are sourced from the CSMAR database. Following the approach of [61], Equation (3) is established to test the mediating effects of the chosen mechanism variables.

$${\mathrm{Mediator}}_{\mathrm{i},\mathrm{t}}={\gamma }_0+{\gamma }_1{\mathrm{Dig}}_{\mathrm{i},\mathrm{t}}+{{\gamma }_2{\mathrm{E}\mathrm{u}}_{\mathrm{i},\mathrm{t}}+\sum }_{\mathrm{j}}{\gamma }_{\mathrm{j}}{\mathrm{Controls}}_{\mathrm{j},\mathrm{i},\mathrm{t}}+{\mu }_{\mathrm{i}}+{\theta }_{\mathrm{t}}+{\phi }_{\mathrm{i},\mathrm{t}}$$

(3)

where Mediator denotes Tq, Media, or Ana; Dig is the enterprise’s digital transformation; Eu is environmental uncertainty; Controls is the control variables; ε is the random error term; and μ and θ, respectively, represent the industry fixed effects and time fixed effects. Apart from the mechanism variables, the rest of the variable settings remain consistent with the earlier discussion.

3.2. Measurement of Variables

3.2.1. ESG Performance

This research employs the Huazheng ESG ratings as a proxy variable to measure ESG performance. The Huazheng ESG ratings cover a comprehensive range of indicators and fully take into account the social system and market characteristics of China in their assessment. The Huazheng ESG ratings consist of nine levels, ranging from C to AAA, with each level assigned a score from 1 to 9. These levels are represented by the following symbols: C, CC, CCC, B, BB, BBB, A, AA, and AAA. The average value of the ESG quarterly scores is calculated as the proxy variable for the company’s annual ESG performance, with higher values signifying superior ESG performance.

3.2.2. Digital Transformation

The annual report, as an important channel in conveying information externally, reflects an enterprise’s financial performance and strategic planning. The digital transformation strategy is connected to enterprise development and large capital investments. The strategic planning and financial commitments related to digital transformation are likely to be clearly reflected in the annual report. Thus, the frequency of digital transformation-related words in the annual report can serve as a reliable indicator of the extent to which the enterprise has undergone digital transformation. This study uses the digital transformation-related word frequency in annual reports as a proxy variable to measure the extent of digital transformation and the extent of its pivotal technologies with the aid of Python 3.11. According to [4], the list of keywords includes five pivotal technologies—artificial intelligence, big data, cloud computing, blockchain, and digital technology application—comprising a total of 76 digital transformation-related keywords. Specifically, considering that the word frequency has typical right skewness [4], it is measured by adding 1 to the frequency and then applying the natural logarithm.

3.2.3. Environmental Uncertainty

Variations in the external environment of a company may lead to variations in its core operations and cause fluctuations in its operating income [63,64]. Ghosh and Olsen (2009) [14] use the industry-adjusted standard deviation of the operating income over the previous 5 years as a proxy variable to measure environmental uncertainty. Shen et al. (2012) [65] contend that not all fluctuations in operating income are caused by environmental uncertainty. They highlight that it is necessary to exclude fluctuations brought about by the company’s consistent sales revenue growth. Referring to [14,65], this study uses industry-adjusted sales revenue fluctuations to measure environmental uncertainty. First, this study performs a regression analysis on the operating income data of the past 5 years using Equation (4):

$$\mathrm{S}\mathrm{a}\mathrm{l}\mathrm{e}={\phi }_0+{\phi }_1\mathrm{Y}\mathrm{e}\mathrm{a}\mathrm{r}+\epsilon$$

(4)

where Sale is the sales revenue, Year is the annual variable, and ε is the abnormal sales revenue. Then, we obtain the unadjusted environmental uncertainty by dividing the standard deviation of the abnormal sales revenue in the past 5 years by the average value:

$$\mathrm{U}\mathrm{n}\mathrm{a}\mathrm{d}\mathrm{j}\_\mathrm{E}\mathrm{u}={\displaystyle \frac{{\sigma }_{\epsilon }}{\overline{\mathrm{S}\mathrm{a}\mathrm{l}\mathrm{e}}}}$$

(5)

where $\mathrm{U}\mathrm{n}\mathrm{a}\mathrm{d}\mathrm{j}\_\mathrm{E}\mathrm{u}$ is unadjusted environmental uncertainty, $\sigma \epsilon$ is the standard deviation of the sales revenue in the past 5 years, and $\overline{\mathrm{S}\mathrm{a}\mathrm{l}\mathrm{e}}$ is the average sales revenue in the past 5 years. Finally, the unadjusted environmental uncertainty divided by the median of the industry value yields the industry-adjusted environmental uncertainty:

$$\mathrm{E}\mathrm{u}={\displaystyle \frac{\mathrm{U}\mathrm{n}\mathrm{a}\mathrm{d}\mathrm{j}\_\mathrm{E}\mathrm{u}}{\mathrm{I}\mathrm{n}\mathrm{d}\mathrm{U}\mathrm{n}\mathrm{a}\mathrm{d}\mathrm{j}\_\mathrm{E}\mathrm{u}}}$$

(6)

where $\mathrm{I}\mathrm{n}\mathrm{d}\mathrm{U}\mathrm{n}\mathrm{a}\mathrm{d}\mathrm{j}\_\mathrm{E}\mathrm{u}$ is the median of the unadjusted industry environmental uncertainty, and Eu is the industry-adjusted environmental uncertainty used in this study.

3.3. Sample and Data

To ensure the reliability of the test results, this study adopts data on A-share listed firms in China from 2009 to 2023, excluding financial enterprises; ST, *ST, and PT companies; and those with missing data. This selection is based on the rapid progress of digital and information technology in China. Due to the requirement of calculating environmental uncertainty with the past 5 years of operating income, companies with less than 5 years of sales revenue and those with sales revenue less than 0 are also excluded from the sample. Finally, a total of 31,223 observations are obtained. Additionally, to minimize the influence of extreme values, all continuous variables are winsorized at the upper and lower 1% percentiles. The ESG ratings data are obtained from the Wind database, financial data are obtained from the CSMAR database, and digital transformation level data are obtained from the companies’ annual reports.

4. Results

4.1. Descriptive Statistics

Table 1. Descriptive statistics.

Variable	Obs.	Mean	Std. Dev.	Min.	Max.
ESG	31,223	4.055	0.987	1.000	8.000
Dig	31,223	1.425	1.400	0.000	5.142
$\mathrm{D}\mathrm{i}\mathrm{g}\_\mathrm{a}\mathrm{i}$	31,223	0.374	0.759	0.000	3.466
$\mathrm{D}\mathrm{i}\mathrm{g}\_\mathrm{b}\mathrm{c}$	31,223	0.075	0.306	0.000	1.946
$\mathrm{D}\mathrm{i}\mathrm{g}\_\mathrm{c}\mathrm{c}$	31,223	0.505	0.918	0.000	3.932
$\mathrm{D}\mathrm{i}\mathrm{g}\_\mathrm{d}\mathrm{t}$	31,223	0.544	0.906	0.000	3.892
$\mathrm{D}\mathrm{i}\mathrm{g}\_\mathrm{a}\mathrm{d}\mathrm{t}$	31,223	0.958	1.108	0.000	4.159
Eu	31,223	1.313	1.168	0.131	6.888
Size	31,223	22.419	1.319	19.660	26.370
Roa	31,223	0.028	0.070	−0.311	0.205
Age	31,223	2.513	0.527	1.386	3.401
Lev	31,223	0.464	0.206	0.069	0.962
Cash	31,223	0.048	0.070	−0.162	0.251
Top	31,223	0.334	0.146	0.089	0.737
Bsize	31,223	2.132	0.200	1.609	2.708
Outr	31,223	0.376	0.054	0.333	0.571
Gov	31,223	0.448	0.497	0.000	1.000

Note: See Abbreviations for variable definitions.

displays the variables’ descriptive statistics. The average ESG performance (ESG) score is 4.055, with a range of 1 to 8. This suggests that most enterprises demonstrated a reasonable degree of ESG performance during the sample period. However, certain enterprises may still require improvements in their ESG performance. The digital transformation (Dig) has a range of values from 0 to 5.142, with a standard deviation of 1.400, indicating notable disparities in the extent of digital transformation among the enterprises in the sample. The mean values of artificial intelligence ($\mathrm{D}\mathrm{i}\mathrm{g}\_\mathrm{a}\mathrm{i}$), blockchain ($\mathrm{D}\mathrm{i}\mathrm{g}\_\mathrm{b}\mathrm{c}$), cloud computing ($\mathrm{D}\mathrm{i}\mathrm{g}\_\mathrm{c}\mathrm{c}$), big data ($\mathrm{D}\mathrm{i}\mathrm{g}\_\mathrm{s}\mathrm{t}$), and digital technology application ($\mathrm{D}\mathrm{i}\mathrm{g}\_\mathrm{a}\mathrm{d}\mathrm{t}$) are 0.374, 0.075, 0.505, 0.544, and 0.958, respectively. Among the five aspects of digital transformation, blockchain technology has the lowest average value, whereas digital application technology has the highest average value. These suggest that digital transformation is predominantly focused on the realm of digital application technology, with blockchain technology integration remaining comparatively minimal. The mean value of environmental uncertainty (Eu) is 1.313, with a maximum value of 6.888, indicating that some sample enterprises face significant environmental uncertainty. Additionally, other control variables’ descriptive statistics show no significant deviations from the existing literature.

4.2. Multivariate Results

4.2.1. Digital Transformation and ESG Performance (H1)

Table 2. The relationship between digital transformation and ESG performance.

Variable	ESG (1)	ESG (2)	ESG (3)	ESG (4)	ESG (5)	ESG (6)	ESG (7)
Dig	0.110 *** (10.419)	0.061 *** (7.204)
$\mathrm{D}\mathrm{i}\mathrm{g}\_\mathrm{a}\mathrm{i}$			0.102 *** (7.811)
$\mathrm{D}\mathrm{i}\mathrm{g}\_\mathrm{b}\mathrm{c}$				0.064 ** (2.227)
$\mathrm{D}\mathrm{i}\mathrm{g}\_\mathrm{c}\mathrm{c}$					0.103 *** (9.189)
$\mathrm{D}\mathrm{i}\mathrm{g}\_\mathrm{d}\mathrm{t}$						0.078 *** (6.431)
$\mathrm{D}\mathrm{i}\mathrm{g}\_\mathrm{a}\mathrm{d}\mathrm{t}$							0.040 *** (4.147)
Eu		−0.106 *** (−14.956)	−0.106 *** (−15.070)	−0.107 *** (−15.058)	−0.106 *** (−14.949)	−0.106 *** (−15.000)	−0.106 *** (−15.021)
Size		0.287 *** (29.257)	0.290 *** (29.635)	0.296 *** (30.038)	0.290 *** (29.876)	0.290 *** (29.689)	0.292 *** (29.577)
Roa		1.220 *** (10.135)	1.238 *** (10.275)	1.217 *** (10.090)	1.203 *** (10.042)	1.230 *** (10.197)	1.213 *** (10.070)
Age		−0.288 *** (−13.483)	−0.287 *** (−13.420)	−0.299 *** (−13.925)	−0.285 *** (−13.329)	−0.290 *** (−13.577)	−0.297 *** (−13.864)
Lev		−1.011 *** (−17.226)	−1.010 *** (−17.198)	−1.023 *** (−17.401)	−1.017 *** (−17.407)	−1.019 *** (−17.379)	−1.021 *** (−17.381)
Cash		0.220 ** (2.115)	0.223 ** (2.141)	0.193 * (1.851)	0.260 ** (2.509)	0.218 ** (2.100)	0.186 * (1.789)
Top		−0.034 (−0.452)	−0.029 (−0.380)	−0.052 (−0.691)	−0.023 (−0.303)	−0.036 (−0.486)	−0.051 (−0.685)
Bsize		0.044 (0.708)	0.049 (0.795)	0.043 (0.682)	0.059 (0.946)	0.045 (0.729)	0.037 (0.592)
Outr		1.202 *** (6.200)	1.231 *** (6.338)	1.246 *** (6.389)	1.228 *** (6.324)	1.229 *** (6.325)	1.224 *** (6.294)
Gov		0.212 *** (8.018)	0.204 *** (7.719)	0.199 *** (7.497)	0.202 *** (7.680)	0.203 *** (7.683)	0.208 *** (7.827)
Constant	3.472 *** (33.284)	−1.950 *** (−7.276)	−2.026 *** (−7.607)	−2.116 *** (−7.908)	−2.054 *** (−7.686)	−2.005 *** (−7.511)	−2.024 *** (−7.533)
Year	YES	YES	YES	YES	YES	YES	YES
Ind	YES	YES	YES	YES	YES	YES	YES
N	31,223	31,223	31,223	31,223	31,223	31,223	31,223
Adj. R2	0.068	0.271	0.271	0.267	0.273	0.270	0.268

Note: This table reports the relationship between digital transformation and ESG performance. *** p < 0.01, ** p < 0.05, * p < 0.1, with the t-values in parentheses. See Abbreviations for variable definitions.

reports the baseline regression results for the impact of digital transformation (Dig) on ESG performance. All models control for year and industry fixed effects. The effects of digital transformation (Dig) on ESG performance (ESG) are shown in Columns (1) and (2). Before and after the addition of the control variables, the coefficients of Dig are 0.110 and 0.061, respectively. Both values are significant at the 1% level, underscoring that digital transformation within enterprises substantially aids in enhancing ESG performance. In other words, a 1% increase in the enterprise digital transformation index is associated with a 0.061-point enhancement in the ESG score. This finding indicates that digital transformation is a key driver in bolstering ESG performance, underscoring its consistent association with ESG enhancements in practical applications. This conclusion aligns with the findings of the majority of scholars [6,7]. The coefficient on Eu is −0.106 (t = −15.070, p < 0.01), indicating that higher external uncertainty erodes ESG performance. However, additional research is needed to disentangle the specific effects of each digital technology on ESG performance, especially under varying levels of environmental uncertainty.

Columns (3) to (7) present the regression outcomes for different pivotal technologies on the ESG scores. The coefficient for artificial intelligence ($\mathrm{D}\mathrm{i}\mathrm{g}\_\mathrm{a}\mathrm{i}$) stands at 0.102; that for cloud computing ($\mathrm{D}\mathrm{i}\mathrm{g}\_\mathrm{c}\mathrm{c}$) stands at 0.103; that for big data ($\mathrm{D}\mathrm{i}\mathrm{g}\_\mathrm{d}\mathrm{t}$) stands at 0.078; and that for digital technology application ($\mathrm{D}\mathrm{i}\mathrm{g}\_\mathrm{a}\mathrm{d}\mathrm{t}$) stands at 0.040. All of these are still positive and significant at the 1% level. Our findings consistently demonstrate that these four digital transformation dimensions significantly enhance ESG performance. Blockchain technology ($\mathrm{D}\mathrm{i}\mathrm{g}\_\mathrm{b}\mathrm{c}$) shows a positive and significant coefficient of 0.064 at the 5% level, suggesting its meaningful contribution to ESG performance.

The significant positive effect of blockchain technology on ESG performance adds an interesting dimension to our understanding of digital transformation’s impact. While blockchain technology, characterized as a digital, decentralized, and distributed ledger [66], has been theoretically recognized for its potential to bolster trust among enterprises and reduce information asymmetry, our empirical findings now provide concrete evidence of its positive influence on ESG performance. This finding challenges previous concerns about companies potentially exploiting blockchain as a speculative tool [67] and suggests that, in our sample period, firms successfully leveraged blockchain’s capabilities to enhance their ESG practices. Despite the technical security vulnerabilities and regulatory challenges associated with blockchain technology [16,17,18], our results indicate that companies have effectively managed these risks while harnessing blockchain’s benefits for ESG performance.

In summary, the results fully support H1 of this study: digital transformation can significantly improve ESG performance, with all pivotal technologies demonstrating significant positive effects on ESG performance. The findings suggest that companies are successfully leveraging pivotal technologies to enhance their ESG practices, marking a significant advancement in our understanding of how digital transformation contributes to corporate sustainability and responsibility.

4.2.2. Environmental Uncertainty, Digital Transformation, and ESG Performance (H2)

During this era of globalization and the Fourth Industrial Revolution, Chinese enterprises are undergoing a period of increased environmental uncertainty while progressing on their paths of digital transformation. Macroeconomic and policy-related challenges, such as the ongoing China–US trade war, volatile exchange rates, and fluctuating oil prices, compound this period of uncertainty. At the corporate level, the competitive landscape is undergoing significant changes due to the entry of novel internet-based startups and multinational corporations, further complicating the market dynamics. As companies shift towards adopting new digital technologies, they must confront a variety of challenges, including supply chain risks, data security vulnerabilities, and rising compliance costs, among other potential risks. For instance, while digital technologies can enhance supply chain efficiency, they also introduce vulnerabilities like reliance on digital infrastructure, which could be compromised by cyberattacks or technical failures. Additionally, the growing dependence on gathering and utilizing data intensifies the vulnerability to data breaches and cyberattacks, hence complicating the task of maintaining and safeguarding corporate information systems. As the environmental uncertainty intensifies, the complexity and potential risks linked to digital transformation efforts are escalated correspondingly. This underscores the critical necessity of examining how environmental uncertainty affects the connection between digital transformation and ESG performance. Understanding this interplay is essential in devising digital transformation and ESG strategies.

To examine the moderating effect of environmental uncertainty on the relationship between digital transformation and ESG performance, we decentralize the explanatory and moderating variables before generating interaction terms. This approach helps to minimize multicollinearity. Column (1) of

Table 3. The effect of environmental uncertainty on the relationship between digital transformation and ESG performance.

Variable	ESG (1)	ESG (2)	ESG (3)	ESG (4)	ESG (5)	ESG (6)
Dig	0.061 *** (7.276)
Eu	−0.106 *** (−15.026)	−0.110 *** (−15.198)	−0.112 *** (−15.432)	−0.109 *** (−15.221)	−0.110 *** (−15.390)	−0.108 *** (−15.262)
$\mathrm{D}\mathrm{i}\mathrm{g}\times \mathrm{E}\mathrm{u}$	−0.010 *** (−2.612)
$\mathrm{D}\mathrm{i}\mathrm{g}\_\mathrm{a}\mathrm{i}$		0.099 *** (7.407)
$\mathrm{D}\mathrm{i}\mathrm{g}\_\mathrm{a}\mathrm{i}\times \mathrm{E}\mathrm{u}$		−0.024 ** (−2.212)
$\mathrm{D}\mathrm{i}\mathrm{g}\_\mathrm{b}\mathrm{c}$			0.055 * (1.911)
$\mathrm{D}\mathrm{i}\mathrm{g}\_\mathrm{b}\mathrm{c}\times \mathrm{E}\mathrm{u}$			−0.037 *** (−3.735)
$\mathrm{D}\mathrm{i}\mathrm{g}\_\mathrm{c}\mathrm{c}$				0.099 *** (8.882)
$\mathrm{D}\mathrm{i}\mathrm{g}\_\mathrm{c}\mathrm{c}\times \mathrm{E}\mathrm{u}$				−0.024 *** (−2.614)
$\mathrm{D}\mathrm{i}\mathrm{g}\_\mathrm{d}\mathrm{t}$					0.075 *** (6.236)
$\mathrm{D}\mathrm{i}\mathrm{g}\_\mathrm{d}\mathrm{t}\times \mathrm{E}\mathrm{u}$					−0.030 *** (−3.421)
$\mathrm{D}\mathrm{i}\mathrm{g}\_\mathrm{a}\mathrm{d}\mathrm{t}$						0.038 *** (3.953)
$\mathrm{D}\mathrm{i}\mathrm{g}\_\mathrm{a}\mathrm{d}\mathrm{t}\times \mathrm{E}\mathrm{u}$						−0.022 *** (−3.266)
Size	0.287 *** (29.306)	0.290 *** (29.667)	0.296 *** (30.108)	0.290 *** (29.904)	0.290 *** (29.755)	0.293 *** (29.665)
Roa	1.204 *** (10.005)	1.221 *** (10.125)	1.194 *** (9.893)	1.189 *** (9.916)	1.209 *** (10.020)	1.191 *** (9.911)
Age	−0.289 *** (−13.542)	−0.288 *** (−13.469)	−0.300 *** (−13.969)	−0.286 *** (−13.375)	−0.291 *** (−13.627)	−0.298 *** (−13.935)
Lev	−1.015 *** (−17.315)	−1.013 *** (−17.251)	−1.028 *** (−17.493)	−1.020 *** (−17.466)	−1.023 *** (−17.474)	−1.026 *** (−17.494)
Cash	0.226 ** (2.175)	0.228 ** (2.189)	0.202 * (1.933)	0.265 ** (2.546)	0.226 ** (2.173)	0.196 * (1.885)
Top	−0.038 (−0.503)	−0.031 (−0.416)	−0.054 (−0.712)	−0.024 (−0.326)	−0.040 (−0.530)	−0.055 (−0.729)
Bsize	0.043 (0.697)	0.050 (0.800)	0.044 (0.699)	0.059 (0.951)	0.043 (0.693)	0.037 (0.590)
Outr	1.195 *** (6.158)	1.229 *** (6.323)	1.246 *** (6.394)	1.229 *** (6.328)	1.223 *** (6.295)	1.213 *** (6.234)
Gov	0.212 *** (8.022)	0.204 *** (7.735)	0.200 *** (7.516)	0.202 *** (7.684)	0.203 *** (7.681)	0.208 *** (7.815)
Constant	−1.953 *** (−7.297)	−2.020 *** (−7.588)	−2.099 *** (−7.862)	−2.044 *** (−7.654)	−1.991 *** (−7.471)	−2.025 *** (−7.548)
Year	YES	YES	YES	YES	YES	YES
Ind	YES	YES	YES	YES	YES	YES
N	31,223	31,223	31,223	31,223	31,223	31,223
Adj. R2	0.272	0.271	0.268	0.274	0.271	0.269

Note: This table reports the relationship between digital transformation and ESG performance. *** p < 0.01, ** p < 0.05, * p < 0.1, with the t-values in parentheses. See Abbreviations for variable definitions.

shows that environmental uncertainty (Eu) exhibits a significantly negative coefficient at the 1% level, illustrating that environmental uncertainty markedly detracts from enterprise ESG performance. Notably, the interaction term between digital transformation and environmental uncertainty ($\mathrm{D}\mathrm{i}\mathrm{g}\times \mathrm{E}\mathrm{u}$) registers a coefficient of −0.010, also significant at the 1% level, indicating that environmental uncertainty serves as a negative moderator in the relationship between digital transformation and ESG performance. Our findings indicate that, against the backdrop of rapidly developing global and digital economies, environmental uncertainty is indeed a crucial factor that businesses cannot overlook during their development and digital transformation processes. Factors of environmental uncertainty, such as shifts in macroeconomic policies, market volatility, and risks linked to adopting new digital technologies, could influence digital transformation’s effectiveness in enhancing ESG performance. Furthermore, our findings suggest that the strategy of undergoing digital transformation to improve ESG outcomes can be complex and challenging. Companies must consider not only the investment in relevant technology and resources but also how external environmental elements influence the success of digital transformation.

Columns (2) to (6) of the analysis further reveal estimates of how environmental uncertainty’s moderating impact varies depending on the specific digital technology utilized. Among the interaction terms between different types of digital technology and environmental uncertainty, ($\mathrm{D}\mathrm{i}\mathrm{g}\_\mathrm{a}\mathrm{i}\times \mathrm{E}\mathrm{u}$) shows significance at the 5% level, while the other interaction terms are all significantly negative at the 1% level, indicating that environmental uncertainty diminishes the positive contributions of these pivotal technologies to ESG performance.

These findings lend support to the peer effect theory posited by [15], suggesting that environmental uncertainty may compromise managerial autonomy in decision-making, prompting a propensity to emulate the digital transformation strategies of peer companies. This could weaken the beneficial impacts of digital transformation on ESG performance. Furthermore, environmental uncertainty places additional strain on business operations and financial stability, possibly increasing the motivation for earnings management among executives, which could further diminish the advantageous impacts of digital transformation on ESG outcomes. While blockchain technology’s impact on enhancing enterprise ESG performance has not been significantly demonstrated, its influence remains subject to environmental uncertainty. Consequently, H2 of this study receives partial empirical validation.

4.3. Robustness Tests

4.3.1. Alternative Measures of ESG Performance

To corroborate the reliability of our primary empirical findings, this study implements a comprehensive series of robustness checks employing multiple distinct methodological approaches. These tests address potential concerns regarding measurement, endogeneity, temporal effects, and effect heterogeneity.

First, we address potential concerns regarding the measurement of the dependent variable by employing alternative ESG metrics. We substitute our primary ESG measure (Huazheng ESG ratings) with an alternative metric, the Wind ESG ratings (Wind ESG). Since the Wind ESG rating system was introduced in 2018, the sample period for this robustness check spans from 2018 to 2023. The Wind ESG rating system classifies firms into seven tiers (CCC, B, BB, BBB, A, AA, AAA, from lowest to highest), which we convert into a numeric scale from 1 to 7. Wind’s methodology incorporates international ESG standards while being tailored to the Chinese corporate context, making it an appropriate alternative measure. Re-estimating our baseline models using the Wind ESG scores yields results that are qualitatively consistent with our main findings, as shown in

Table 4. The results of robustness checks.

Variable	ESG (1)	ESG (2)	ESG (3)
Dig	0.031 *** (3.641)	0.064 *** (7.544)	0.033 *** (5.202)
Eu	−0.055 *** (−6.913)	−0.103 *** (−14.264)	−0.055 *** (−11.704)
Size	0.221 *** (19.653)	0.288 *** (29.420)	0.275 *** (26.306)
Roa	0.272 ** (2.323)	1.252 *** (10.210)	0.082 (0.934)
Age	−0.231 *** (−10.332)	−0.284 *** (−13.253)	0.004 (0.110)
Lev	−0.569 *** (−9.299)	−1.022 *** (−17.313)	−0.748 *** (−17.314)
Cash	0.179 (1.461)	0.220 ** (2.074)	−0.231 *** (−3.183)
Top	0.055 (0.666)	−0.045 (−0.598)	0.087 (1.196)
Bsize	0.156 ** (2.243)	0.041 (0.659)	0.063 (1.359)
Outr	0.657 *** (2.843)	1.211 *** (6.232)	1.027 *** (7.326)
Gov	0.129 *** (4.624)	0.209 *** (7.941)	0.063 ** (2.287)
Constant	−1.801 *** (−6.069)	−1.961 *** (−7.336)	−2.325 *** (−8.626)
Year	YES	YES	YES
Ind	YES	YES	YES
Firm	NO	NO	YES
N	16,824	31,223	31,022
Adj. R2	0.190	0.253	0.549

Note: This table reports the results of the mechanism test. *** p < 0.01, ** p < 0.05, with the t-values in parentheses. See Abbreviations for variable definitions.

, Column (1). The coefficient of Dig is 0.031, which is positive and significant at the 1% level.

Additionally, to address potential concerns about the aggregation method, we employ the annual median of the ESG scores as an alternative proxy to evaluate enterprise ESG performance, substituting the initial annual average of the ESG scores [68]. This approach helps to mitigate the influence of outliers in the quarterly ESG ratings. As shown in Column (2) of Table 4, the coefficient on Dig remains positive and significant. These findings confirm that the relationship between digital transformation and ESG performance is robust to variations in the ESG measurement methodologies.

4.3.2. Controlling for Firm Fixed Effects

Next, we control for firm fixed effects. Enterprises exhibiting superior ESG performance may inherently be larger and technologically more advanced, thus possessing inherent advantages in digital transformation. To mitigate biases stemming from such unobserved firm-specific characteristics, we estimate models that include firm fixed effects.

Column (3) of Table 4 presents these results, showing a significantly positive Dig coefficient of 0.033. The persistence of a significant positive effect after controlling for firm-specific time-invariant characteristics suggests that our main findings are not driven by unobserved heterogeneity between firms.

4.3.3. Temporal Persistence Analysis

Third, we examine the long-term effects of digital transformation on ESG performance through a lagged analysis. As digital transformation often requires time to fully materialize and impact corporate operations, we introduce lagged digital transformation variables (L.Dig, L2.Dig, L3.Dig) to assess potential time-delayed effects. This approach also helps to address reverse causality concerns by establishing a clear temporal sequence.

Table 5. The results of the long-term effects of digital transformation on ESG performance.

Variable	ESG (1)	ESG (2)	ESG (3)
L.Dig	0.063 *** (7.099)
L2.Dig		0.055 *** (5.801)
L3.Dig			0.047 *** (4.678)
Eu	−0.105 *** (−13.882)	−0.108 *** (−13.582)	−0.112 *** (−13.335)
Size	0.293 *** (28.766)	0.302 *** (28.242)	0.307 *** (27.225)
Roa	1.225 *** (9.600)	1.201 *** (9.079)	1.208 *** (8.624)
Age	−0.297 *** (−12.199)	−0.306 *** (−10.949)	−0.314 *** (−9.500)
Lev	−1.048 *** (−16.882)	−1.084 *** (−16.519)	−1.112 *** (−16.002)
Cash	0.253 ** (2.242)	0.267 ** (2.205)	0.252 * (1.919)
Top	−0.028 (−0.348)	−0.033 (−0.379)	−0.012 (−0.132)
Bsize	0.044 (0.668)	0.029 (0.410)	0.020 (0.262)
Outr	1.187 *** (5.832)	1.162 *** (5.357)	1.138 *** (4.860)
Gov	0.222 *** (8.062)	0.230 *** (7.970)	0.230 *** (7.592)
Constant	−2.141 *** (−7.586)	−2.180 *** (−7.230)	−2.124 *** (−6.580)
Year	YES	YES	YES
Ind	YES	YES	YES
N	27,744	24,534	21,444
Adj. R2	0.276	0.279	0.282

Note: This table reports the results of the mechanism test. *** p < 0.01, ** p < 0.05, * p < 0.1, with the t-values in parentheses. See Abbreviations for variable definitions.

presents these results, where Columns (1), (2), and (3) incorporate one-year, two-year, and three-year lagged digital transformation measures, respectively. The coefficients of L.Dig (0.063, t = 7.099), L2.Dig (0.055, t = 5.801), and L3.Dig (0.047, t = 4.678) all maintain positive significance at the 1% level, suggesting that the impact of digital transformation on ESG performance persists over multiple years. Notably, the effect size gradually diminishes from 0.063 in the first year to 0.047 by the third year, indicating a sustained but slightly decreasing impact over time. This pattern is consistent with the notion that digital transformation yields immediate benefits that partially diminish as the initial competitive advantages become industry standards.

The declining coefficient magnitude across the three-year period follows a clear pattern, with each subsequent year showing an approximately 0.008 reduction in the effect size. This finding suggests that, while digital transformation provides long-lasting ESG benefits, firms need to continually refresh their digital capabilities to maintain the maximum ESG performance advantages. These results confirm that the beneficial influence of digital transformation on ESG performance extends beyond immediate implementation, demonstrating long-term sustainability and the strategic rather than merely tactical nature of digital transformation investments.

4.3.4. Instrumental Variable Analysis

Fourth, to more rigorously address endogeneity concerns arising from omitted variables or reverse causality, we employ two-stage least squares (2SLS) estimation using two distinct instrumental variables.

The first instrumental variable (IV1) is based on an external policy shock. Following the approach of [69], IV1 is the “Broadband China” pilot policy. This policy represents an exogenous shock to the local digital infrastructure. Specifically, we create a dummy variable that equals one if a firm’s registered city was selected as a “Broadband China” pilot city during the sample period and zero otherwise. The “Broadband China” policy serves as an appropriate instrumental variable for several compelling reasons. First, the selection of pilot cities was determined primarily by national strategic considerations rather than firm-specific characteristics, ensuring exogeneity to individual corporate ESG activities. Second, this policy significantly enhanced the local digital infrastructure and internet penetration, creating a favorable environment for corporate digital transformation without directly influencing ESG performance through other channels. Third, the staggered implementation of this policy across different regions creates quasi-experimental variation in digital transformation opportunities, helping to isolate its causal impact on ESG outcomes. Fourth, this policy shock satisfies the relevance criterion by directly affecting firms’ digital transformation capabilities and strategies, while remaining uncorrelated with the error term in the ESG performance equation.

The second instrumental variable (IV2), drawing from the methodology of [70], is constructed as the interaction between a city’s fixed-line telephone density (number of fixed-line telephones per 10,000 people in 1984) and the one-year lagged national internet penetration rate. The rationale for IV2 is that historical telecommunications infrastructure strongly predicts a region’s subsequent digital adoption capacity, thus satisfying the relevance condition. Simultaneously, these historical city-level data from 1984 predate contemporary ESG concerns and are plausibly exogenous to the current firm-specific ESG performance, fulfilling the exclusion restriction.

Table 6. The results of 2SLS.

Variable	DIG (1)	ESG (2)	DIG (3)	ESG (4)
Dig		0.337 *** (6.696)		0.372 *** (6.703)
IV(1)	0.271 *** (15.924)
IV(2)			0.090 *** (13.678)
Eu	−0.036 *** (−5.954)	−0.094 *** (−17.613)	−0.037 *** (−5.945)	−0.095 *** (−16.764)
Size	0.168 *** (26.152)	0.240 *** (23.905)	0.176 *** (26.634)	0.238 *** (21.780)
Roa	−0.173 (−1.452)	1.249 *** (12.035)	−0.156 (−1.253)	1.236 *** (11.447)
Age	−0.199 *** (−13.422)	−0.228 *** (−14.523)	−0.193 *** (−12.533)	−0.215 *** (−12.661)
Lev	−0.214 *** (−5.263)	−0.971 *** (−26.900)	−0.265 *** (−6.270)	−0.970 *** (−25.463)
Cash	−0.643 *** (−6.148)	0.437 *** (4.662)	−0.687 *** (−6.393)	0.419 *** (4.298)
Top	−0.435 *** (−8.999)	0.087 * (1.914)	−0.437 *** (−8.756)	0.095 ** (1.990)
Bsize	−0.073 * (−1.758)	0.070 ** (2.012)	−0.139 *** (−3.252)	0.065 * (1.775)
Outr	0.551 *** (3.811)	1.101 *** (8.972)	0.646 *** (4.339)	1.025 *** (7.905)
Gov	−0.222 *** (−13.481)	0.278 *** (15.950)	−0.193 *** (−11.235)	0.282 *** (15.058)
Constant	−2.812 *** (−17.673)	−1.272 *** (−6.356)	−2.926 *** (−17.904)	−1.246 *** (−5.767)
Year	YES	YES	YES	YES
Ind	YES	YES	YES	YES
N	28,309	28,309	26,854	26,854
Adj. R2	0.395	0.181	0.396	0.183
$\mathrm{Kleibergen–Paap} \mathrm{r}\mathrm{k} \mathrm{L}\mathrm{M} \mathrm{s}\mathrm{t}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{s}\mathrm{t}\mathrm{i}\mathrm{c}$		249.523 ***		176.854 ***
$\mathrm{Kleibergen–Paap} \mathrm{r}\mathrm{k} \mathrm{W}\mathrm{a}\mathrm{l}\mathrm{d} \mathrm{F} \mathrm{s}\mathrm{t}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{s}\mathrm{t}\mathrm{i}\mathrm{c}$		253.562		187.094
		{16.38}		{16.38}

Note: This table reports the results of the mechanism test. *** p < 0.01, ** p < 0.05, * p < 0.1, with the t-values in parentheses. See Abbreviations for variable definitions.

presents the two-stage least-squares (2SLS) regression results. Columns (1) and (3) display the first-stage regressions, confirming the instruments’ relevance, with significant coefficients for IV(1) (0.271, t = 15.924) and IV(2) (0.090, t = 13.678). The strong significance of these coefficients suggests that both instruments strongly predict firms’ digital transformation levels.

Columns (2) and (4) report the second-stage results, showing that the instrumented digital transformation variable maintains a positive and significant effect on ESG performance. The Kleibergen–Paap rk LM statistics (249.523 and 176.854, p < 0.01) strongly reject the null hypothesis of under-identification, indicating that our excluded instruments are relevant. The Kleibergen–Paap rk Wald F statistics (253.562 and 187.094) substantially exceed the critical threshold of 16.38, as suggested by the Stock–Yogo weak IV test at the 10% significance level, confirming the absence of a weak instrumental variable issue. These findings substantiate the causal relationship between digital transformation and enhanced ESG performance, strengthening the causal interpretation of our results.

4.3.5. Quantile Regression Analysis

Finally, to investigate whether the impact of digital transformation varies across different levels of ESG performance, we employ quantile regression techniques, which allow us to examine the effect of digital transformation across the entire distribution of ESG performance, rather than just at the conditional mean. This approach helps us to understand whether firms with different ESG performance foundations benefit differently from digital initiatives.

Table 7. The results of quantile regression.

Variable	Q10 (1)	Q25 (2)	Q50 (3)	Q75 (4)	Q90 (5)
Dig	0.059 *** (0.007)	0.068 *** (0.005)	0.057 *** (0.005)	0.073 *** (0.005)	0.060 *** (0.005)
Eu	−0.164 *** (0.011)	−0.136 *** (0.008)	−0.100 *** (0.005)	−0.096 *** (0.007)	−0.079 *** (0.006)
Size	0.249 *** (0.008)	0.250 *** (0.006)	0.247 *** (0.006)	0.283 *** (0.007)	0.297 *** (0.007)
Roa	1.864 *** (0.148)	1.720 *** (0.148)	1.493 *** (0.100)	1.274 *** (0.129)	1.272 *** (0.125)
Age	−0.257 *** (0.018)	−0.272 *** (0.016)	−0.227 *** (0.011)	−0.218 *** (0.014)	−0.198 *** (0.016)
Lev	−0.983 *** (0.051)	−0.927 *** (0.044)	−0.727 *** (0.040)	−0.738 *** (0.044)	−0.675 *** (0.042)
Cash	−0.461 *** (0.155)	−0.293 *** (0.110)	−0.119 (0.088)	0.063 (0.114)	0.090 (0.105)
Top	0.033 (0.072)	0.019 (0.046)	−0.001 (0.046)	0.042 (0.056)	−0.011 (0.044)
Bsize	0.059 (0.061)	−0.004 (0.039)	−0.013 (0.038)	−0.024 (0.044)	0.085 * (0.047)
Outr	1.001 *** (0.204)	1.013 *** (0.128)	1.048 *** (0.122)	1.182 *** (0.174)	1.515 *** (0.148)
Gov	0.260 *** (0.022)	0.204 *** (0.018)	0.180 *** (0.014)	0.162 *** (0.016)	0.143 *** (0.015)
Constant	−2.050 *** (0.224)	−1.364 *** (0.181)	−0.975 *** (0.162)	−1.293 *** (0.173)	−1.576 *** (0.196)
N	31,223	31,223	31,223	31,223	31,223
Adj. R	0.110	0.126	0.080	0.120	0.129

Note: This table reports the relationship between digital transformation and ESG performance. *** p < 0.01, * p < 0.1, with the t-values in parentheses. See Abbreviations for variable definitions.

shows the quantile regression results. The coefficient of Dig remains positive and significant across all quantiles, ranging from 0.057 to 0.073, which indicates that digital transformation benefits firms across the entire ESG performance distribution. This consistency across quantiles reinforces the robustness of our main findings.

Interestingly, the effect follows a non-monotonic pattern, with the strongest impact observed at the 75th percentile (0.073, p < 0.01) and the weakest at the median (0.057, p < 0.01). This suggests that firms in the upper–middle range of ESG performance benefit most from digital transformation, followed by those in the lower–middle range (25th percentile, 0.068, p < 0.01). The slightly reduced effect at the highest performance level (90th percentile, 0.060, p < 0.01) indicates potential diminishing returns for already high-performing ESG companies. These findings highlight the complex, non-linear relationship between digital transformation and ESG performance across different performance segments, with the greatest marginal benefits accruing in firms that have already established ESG foundations but still have room for substantial improvement. This non-linearity challenges simplistic universal approaches to digital transformation and suggests that firms should calibrate their digital strategies according to their current ESG performance levels to maximize the benefits.

Overall, these comprehensive robustness tests—employing alternative ESG measurements, controlling for firm fixed effects, examining lagged effects, addressing endogeneity concerns through instrumental variables, and assessing heterogeneous effects across quantiles—collectively affirm the robustness of our primary finding that digital transformation positively impacts corporate ESG performance. The consistency of the results across multiple methodological approaches strengthens the confidence in our conclusion that digital transformation serves as a significant driver of improved ESG performance, even when accounting for potential measurement issues, endogeneity concerns, and effect heterogeneity.

5. Further Discussion of Heterogeneity and Mechanism

5.1. Analysis of Heterogeneity Across Industry Types

Different industries have distinct technology adoption rates and capabilities, which can impact digital transformation’s influence on ESG performance. High-tech companies are typically at the forefront in adopting new technologies and thus may integrate digital transformation more seamlessly into their operations, potentially leading to greater improvements in ESG performance. Heavy-pollution companies confront deep-rooted environmental issues and face high environmental remediation costs, making it challenging to effect fundamental changes solely through digital technology. Additionally, digital transformation’s benefits for enterprises are primarily focused on enhancing efficiency and reducing costs [51,53], rather than directly tackling environmental pollution. Therefore, the direct effect of digital transformation on ESG performance in heavy-pollution companies may be limited. Labor-intensive, capital-intensive, and technology-intensive classifications reflect differences in resource allocation, which affect the effectiveness of digital transformation strategies. Thus, the influence of digital transformation on ESG outcomes may differ across industries.

To reveal the specific effects of digital transformation on ESG performance across various industry sectors and understand how digital strategies can be utilized effectively to boost ESG performance in varying industrial contexts, this study classifies enterprises into high-tech versus non-high-tech companies, heavy-pollution versus non-heavy-pollution companies, and labor-intensive, capital-intensive, and technology-intensive companies to conduct group analyses. The regression results for high-tech versus non-high-tech companies are presented in Columns (1) and (2) of

Table 8. The results of the heterogeneity test.

Industry Type	High-Tech	Non-High-Tech	Heavy Pollution	Non-Heavy Pollution	Labor-Intensive	Capital-Intensive	Technology-Intensive
Variable	ESG (1)	ESG (2)	ESG (3)	ESG (4)	ESG (5)	ESG (6)	ESG (7)
Dig	0.085 *** (8.008)	0.016 (1.165)	0.032 (1.424)	0.065 *** (7.089)	0.007 (0.398)	0.041 ** (2.319)	0.085 *** (7.384)
Eu	−0.118 *** (−12.375)	−0.089 *** (−8.814)	−0.099 *** (−6.374)	−0.105 *** (−13.243)	−0.084 *** (−6.636)	−0.114 *** (−8.330)	−0.117 *** (−10.859)
Size	0.287 *** (22.047)	0.295 *** (20.168)	0.279 *** (13.516)	0.292 *** (26.273)	0.295 *** (16.017)	0.271 *** (15.039)	0.301 *** (20.306)
Roa	1.242 *** (7.959)	1.028 *** (5.508)	0.918 *** (3.100)	1.317 *** (9.993)	0.666 *** (2.777)	1.522 *** (6.361)	1.236 *** (7.277)
Age	−0.317 *** (−11.400)	−0.247 *** (−7.532)	−0.344 *** (−6.613)	−0.266 *** (−11.452)	−0.228 *** (−5.125)	−0.356 *** (−9.164)	−0.276 *** (−9.017)
Lev	−1.007 *** (−12.754)	−0.993 *** (−11.546)	−1.090 *** (−8.423)	−0.998 *** (−15.199)	−1.008 *** (−8.878)	−1.040 *** (−9.523)	−0.962 *** (−11.095)
Cash	0.352 ** (2.417)	0.097 (0.653)	0.093 (0.392)	0.266 ** (2.319)	0.097 (0.505)	−0.032 (−0.180)	0.615 *** (3.774)
Top	−0.089 (−0.867)	0.048 (0.449)	−0.193 (−1.146)	0.037 (0.443)	−0.072 (−0.516)	0.060 (0.429)	−0.031 (−0.269)
Bsize	0.103 (1.244)	−0.011 (−0.123)	0.078 (0.579)	0.033 (0.471)	−0.053 (−0.464)	0.177 (1.476)	0.049 (0.537)
Outr	0.922 *** (3.536)	1.530 *** (5.457)	1.109 ** (2.427)	1.202 *** (5.646)	1.476 *** (4.119)	1.288 *** (3.442)	0.982 *** (3.446)
Gov	0.188 *** (5.285)	0.247 *** (6.321)	0.250 *** (4.026)	0.197 *** (6.848)	0.255 *** (4.804)	0.257 *** (5.543)	0.145 *** (3.676)
Constant	−1.696 *** (−5.120)	−2.254 *** (−6.186)	−1.622 *** (−3.070)	−2.128 *** (−7.181)	−2.132 *** (−4.775)	−1.527 *** (−3.265)	−2.075 *** (−5.688)
Year	YES	YES	YES	YES	YES	YES	YES
Ind	YES	YES	YES	YES	YES	YES	YES
N	17,469	13,754	7440	23,783	8737	9054	13,432
Adj. R	0.243	0.317	0.209	0.296	0.272	0.304	0.267

Note: This table reports the results of the heterogeneity test. *** p < 0.01, ** p < 0.05, with the t-values in parentheses. See Abbreviations for variable definitions.

. For the high-tech company group, the Dig coefficient is 0.085, demonstrating significant positivity at the 1% level. In contrast, the coefficient for the non-high-tech company group does not reach statistical significance. Digital transformation is often associated with complex, technically demanding, and forward-looking changes. For non-high-tech companies, undergoing digital transformation entails substantial resource and time investments to acquire and implement sophisticated new technologies, encountering numerous barriers, such as the high costs associated with network infrastructure and the limitations of operational structures for technological innovation [71]. The process of leveraging digital technological advancements to enhance ESG performance may thus be more prolonged for non-high-tech companies. On the other hand, high-tech companies inherently possess extensive technological expertise and resources for transformation, facilitating a more efficient transition process. Consequently, high-tech companies are better positioned to capitalize on digital transformation initiatives to bolster their ESG performance.

Columns (3) and (4) show the regression results for groups divided into heavy-pollution versus non-heavy-pollution companies. The coefficient for digital transformation (Dig) among the heavy-pollution company group is 0.032, which is not statistically significant. In contrast, for the non-heavy-pollution company group, the Dig coefficient is 0.065, significantly positive at the 1% level. Due to the inherent nature of their industries, heavy-pollution companies face challenges whereby superficial digital transformation efforts do not address the fundamental issue of significant pollution emissions during production processes. Research [38] indicates that the relationship between digital transformation and environmental performance in heavy-pollution companies follows a significant U shape, suggesting that digital transformation in such companies needs to reach a certain threshold before it can impact environmental performance positively. However, a deep digital transformation could consume substantial resources and impose considerable financial strain, affecting the financial condition and operational efficiency of these companies. Even though, in the long term, thorough digital technology has the potential to aid in energy conservation and emission reduction in heavy-pollution companies, the immediate availability of funds to enhance environmental performance and ESG-related infrastructure may be insufficient in the short term. In contrast, non-heavy-pollution companies do not face the fundamental environmental issues of severe pollution. In their operations, it is easier to monitor and improve their carbon emissions through digital transformation, thus mitigating the negative impact on the environment and enhancing their ESG scores. Therefore, the contribution of digital transformation to ESG performance in heavy-pollution companies may not be directly reflected in their current year’s ESG outcomes, while non-heavy-pollution companies can more readily enhance their ESG performance through digital transformation.

Columns (5), (6), and (7) report the results for labor-, capital-, and technology-intensive firms, respectively. Labor-intensive firms exhibit a negligible and insignificant Dig coefficient of 0.007. In contrast, capital-intensive firms show a statistically significant Dig coefficient of 0.041 (t = 2.319, p < 0.05), and technology-intensive firms display a stronger and highly significant coefficient of 0.085 (t = 7.384, p < 0.01). These patterns align with the existing literature demonstrating that industries with greater capital and technological endowments are better equipped to leverage digital transformation—through investments in advanced equipment, software platforms, and skilled personnel—to drive ESG performance, whereas labor-centric sectors may lack complementary assets or scale efficiencies needed to translate digital efforts into sustainability outcomes [72]. Therefore, technology-intensive and capital-intensive companies, as opposed to labor-intensive ones, appear more adept in undergoing digital transformation and capitalizing on its benefits, leading to more substantial positive effects on ESG performance.

5.2. Mechanism Test

Table 9. The results of the mechanism test.

Variable	Tq (1)	Media (2)	Ana (3)
Dig	0.041 *** (3.265)	0.042 *** (8.652)	0.061 *** (7.546)
Eu	0.078 *** (5.701)	0.047 *** (11.037)	−0.033 *** (−4.829)
Size	−0.526 *** (−22.966)	0.199 *** (30.030)	0.395 *** (36.579)
Roa	3.000 *** (11.261)	0.033 (0.424)	4.411 *** (25.160)
Age	0.171 *** (4.653)	−0.036 *** (−2.819)	−0.269 *** (−11.929)
Lev	0.134 (1.106)	0.110 *** (3.297)	−0.367 *** (−5.951)
Cash	1.273 *** (6.577)	0.550 *** (8.416)	1.085 *** (9.519)
Top	0.163 (1.510)	−0.506 *** (−11.782)	−0.395 *** (−5.279)
Bsize	0.115 (1.247)	0.004 (0.115)	0.083 (1.353)
Outr	1.608 *** (5.182)	0.365 *** (2.821)	0.344 (1.599)
Gov	−0.078 ** (−1.975)	−0.192 *** (−12.996)	−0.152 *** (−5.525)
Constant	12.456 *** (27.258)	0.529 *** (3.124)	−5.897 *** (−22.471)
Year	YES	YES	YES
Ind	YES	YES	YES
N	30,745	31,149	20,502
Adj. R2	0.326	0.612	0.356

Note: This table reports the results of the mechanism test. *** p < 0.01, ** p < 0.05, with the t-values in parentheses. See Abbreviations for variable definitions.

presents the mechanism test results. In Column (1), the coefficient of Dig is 0.041, significantly positive at the 1% level, supporting H3a that digital transformation significantly enhances firm value. In Column (2), the Dig coefficient on media attention is 0.042 (t = 8.652, p < 0.01), confirming H3b that digital transformation increases media coverage. In Column (3), the Dig coefficient on analyst coverage is 0.061 (t = 7.546, p < 0.01), validating H3c that digital transformation attracts more analyst coverage. These findings collectively demonstrate that digital transformation indirectly improves ESG performance through three key channels: enhancing firm value, amplifying media attention, and expanding analyst oversight.

6. Conclusions and Policy Implications

6.1. Conclusions

Given the escalating external environmental uncertainty, this study employs a sample of China’s A-share listed companies from 2009 to 2023 to empirically test the impact of digital transformation and its pivotal digital technological changes on ESG performance. Our key findings are as follows. First, this study confirms digital transformation’s role in boosting corporate ESG metrics, indicating that digital transformation and all pivotal technologies positively influence ESG performance. Our longitudinal analysis further reveals that these positive effects persist for up to three years after implementation, although with gradually diminishing intensity. Second, it is noteworthy that environmental uncertainty weakens the positive impact that digital initiatives have on ESG performance. This finding underscores the importance of incorporating environmental uncertainty into digital transformation plans. Furthermore, heterogeneous effects from digital transformation are observed across various ESG levels and industry types. Within high-tech or technology-intensive companies, and those that do not produce heavy pollution, digital transformation more effectively enhances ESG scores. Our quantile regression results add nuance to this understanding, showing that firms in the upper–middle range of ESG performance benefit most from digital transformation, while those at the median experience the smallest improvements, indicating a complex non-linear relationship across the ESG performance spectrum. Finally, we confirm the mechanisms through which digital transformation impacts ESG performance: it can indirectly improve ESG performance by elevating firm value and drawing more media and analyst coverage. Overall, based on our confirmation of the positive effect that the digital transformation of Chinese enterprises has on ESG performance, this study emphasizes the significant impact of environmental uncertainty, providing a reference enabling enterprises to make optimal decisions in turbulent business environments.

6.2. Policy Implications

Drawing from the conclusions above, we recommend the following countermeasures for different stakeholders. For enterprises, digital transformation is not only a method of pursuing value enhancement and strengthening core competitiveness but also a key pathway to achieving sustainable development and improved ESG performance. Enterprises can integrate digital transformation strategies with goals to elevate ESG performance. On one hand, they can influence green innovation using digital technologies like artificial intelligence, swiftly tackling environmental concerns that are of public interest and thus offering technical backing for the optimization of environmental performance. On the other hand, they can improve the transparency of accounting information and internal controls through applications such as cloud computing, big data, or other technologies, which can improve their social and corporate governance performance. While taking advantage of digital transformation opportunities to advance sustainable development, companies should ensure that external environmental uncertainty is fully accounted for. In highly uncertain situations, companies can establish resilient digital strategies, reserving space to cope with uncertainties, allowing for quick strategic adjustments when facing external changes, protecting stakeholders’ interests, and integrating digital transformation with ESG as a positive, dynamic development model.

For investors, continuously monitoring corporate ESG performance and information disclosure and establishing an ESG investment philosophy are key to long-term value investment. On this basis, enhancing the understanding and application of digital technologies helps investors to more effectively assess enterprises’ strengths and risks. Additionally, investors can utilize digital information to reduce the costs of identifying companies, thereby making wiser investment decisions. Considering the impact of environmental uncertainty, investors should enhance their understanding of uncertainty factors when evaluating companies’ digital transformation and ESG performance. This includes assessing companies’ capabilities to adapt to external environmental changes, evaluating whether their digital strategies are overly aggressive, determining if they can withstand the impact of environmental uncertainty, and considering whether companies can continuously enhance their ESG performance in the face of uncertainties.

Ultimately, the government has a vital role in facilitating digital transformation and the advancement of ESG performance. Many countries, including China, are actively preparing to formulate policies related to ESG information disclosure. While strengthening ESG disclosure standards and requirements, the government can also provide policy support for corporate digital transformation, fully harnessing the intrinsic motivation for enterprises to improve their ESG performance and encouraging the use of digital technologies to optimize their environmental and social responsibility performance. This includes offering related tax incentives, financial subsidies, and technical support measures. It is crucial for governments to maintain stable and continuous policies, mitigating the uncertainty that enterprises face due to policy shifts and fostering a macroenvironment that supports the digital economy and sustainable development.

Our findings suggest several important considerations specifically for regulatory authorities. First, they should develop differentiated ESG disclosure requirements based on industry characteristics and environmental uncertainty levels, acknowledging that the impact of digital transformation on ESG performance varies significantly across sectors. Second, during periods of heightened environmental uncertainty (such as economic downturns or policy transitions), temporary regulatory flexibility might be warranted to prevent overly burdensome compliance costs that could hinder firms’ sustainability efforts. Third, regulators should design targeted digital infrastructure investment strategies, prioritizing sectors where digital technologies demonstrate the strongest ESG enhancement potential—particularly those that were identified in our research within the high-tech industries and those that do not produce heavy pollution. Finally, establishing comprehensive ESG evaluation frameworks that consider both the direct effects of digital transformation and the moderating influence of environmental uncertainty would enable more accurate performance assessment and better-informed policy formulation.

6.3. Research Limitations and Future Directions

Despite this study’s in-depth investigation of the effect of environmental uncertainty on digital transformation and ESG performance, some limitations remain. First, our measurement of environmental uncertainty mainly relied on enterprises’ responses to the external environment, without directly incorporating external factors. Future research could achieve breakthroughs by constructing more comprehensive indicators of environmental uncertainty. Second, we explored environmental uncertainty’s impact on the overall listed enterprises in China’s A-shares. However, considering that different industries may face varying external environments, future research could focus on comparing the impacts of environmental uncertainty across industries. Additionally, while our findings offer meaningful insights for the Chinese context, we acknowledge that the generalizability of these results may be constrained by China’s unique institutional environment, market structure, and policy landscape. Future studies could expand this research through cross-country comparisons to determine whether the observed relationships are maintained in different economic and regulatory contexts.