Digital Transformation and Environmental, Social, and Governance Performance from a Human Capital Perspective

Abstract

The strategic adoption of digital technologies has increasingly been recognized as a crucial driver of cost reduction and operational efficiency in enterprises. It optimizes production processes and promotes sustainable growth. In this context, understanding the specific impact of digital transformation on enterprises’ environmental, social, and governance (ESG) performance holds significant practical value for promoting sustainable development in China’s economy and society. This study focused on Chinese A-share listed enterprises from 2010 to 2022, specifically exploring the role of digital transformation in enhancing ESG performance from the perspective of human capital. Our findings reveal that digital transformation significantly augments their ESG performance. Notably, the improvements are more pronounced in non-state-owned enterprises compared to state-owned ones. Specifically, digital transformation initiatives contribute to ESG performance enhancement by increasing the extent of high-quality labor and elevating the skill levels of the existing workforce. Furthermore, environmental regulation moderates the positive impact of corporate digital transformation on the quantity and skill level of labor, thus influencing firm-level ESG performance. The study sheds light on the transformative role of digital transformation and its implications for ESG performance improvement by elucidating the mechanisms through which digital transformation affects human capital and interacts with regulatory environments.

1. Introduction

Enterprises are increasingly realizing the importance of integrating sustainable development into their strategic frameworks as a key driver of economic and social progress. Sustainable development in enterprises serves as a crucial gauge of their overall sustainability, encompassing economic, environmental, and social dimensions. The environmental, social, and governance (ESG) ethos underscores the comprehensive consideration of environmental protection, social responsibility, and governance efficacy in both investment decisions and operational endeavors [1]. Furthermore, ESG provides enterprises with self-assessment benchmarks and serves as a pivotal metric for investors evaluating sustainable development and the fulfillment of social responsibility [2]. ESG performance significantly influences the overall value and operational efficacy of an enterprise [3]. Specifically, strong ESG performance can considerably reduce financing costs [4,5], mitigate the risks that enterprises face [6,7], and facilitate their engagement in foreign direct investment activities [8]. Nevertheless, implementing ESG practices entails additional investments that can impose financial pressure [9].

In the context of an increasingly intertwined relationship between digital technology and the real economy, enterprises must employ digital transformation to confront and effectively navigate the challenges of ESG practices [10]. Digital transformation fundamentally remodels traditional business paradigms by integrating emerging technologies such as artificial intelligence and blockchain [11]. It facilitates cost reduction, efficiency enhancement, and the creation of new opportunities for sustainable development [12,13]. Therefore, examining the impact of digital transformation on ESG performance is relevant for promoting sustainable economic and social development [14].

Research on the influences of digital transformation (DT) on ESG performance began in 2022; however, the focus has been primarily on understanding how digital transformation affects ESG performance and the underlying mechanisms [15,16].

Table 1. Existing research on the influence of digital transformation on ESG performance.

Independent Variable	Mediating Variables	Moderating Variables	Dependent Variable	Authors
DT	Green innovation	Environmental uncertainty	ESG performance	Wu, S., and Li, Y. [5]
DT	Reducing agency costs and improving goodwill	-	ESG performance	Mingyue Fang, H. Nie, Xinyi Shen [14]
DT	Green technology innovation, human capital accumulation, environmental information disclosure, and environmental governance	-	ESG performance	Xu, Xin Li, Fei Guo [17]
DT	-	Financing constraints	ESG performance	Yang, Jinmian Han [8]
DT	Managers’ short-sighted behavior, innovation capabilities, information transparency, and governance capabilities	-	ESG performance	Jingyong Wang, Zixiang Song, Lida Xue [18]
DT	Green innovation, goodwill, and agency costs	Political connections and regional institutional development	ESG performance	Liuyang Xue, Junan Dong, Yifan Zha [19]
DT	Management myopia, information, and investment	-	ESG performance	Zhong, Y., Zhao, H., and Yin, T. [10]
DT	Internal control and green innovation	The government’s supportive attitude and the degree of marketization in the region	ESG performance	Lu, Y., Xu, C., Zhu, B., and Sun, Y. [20]
DT	-	Formal and informal environmental regulations	ESG performance	Li, J., Wu, T., Liu, B., and Zhou, M. [21]

summarizes studies that have predominantly investigated the influence mechanisms from perspectives such as innovation, information transparency, and agency costs. Despite these valuable insights, exploration from the perspective of human capital has been relatively limited. Human capital, encompassing the employees’ knowledge, skills, and capabilities, is a critical driver of organizational performance and sustainability. Recognizing the importance of human capital in the context of digital transformation and ESG performance is essential for gaining a comprehensive understanding of the underlying mechanisms at play.

Our study focuses on Chinese A-share listed companies from 2010 to 2022 as the sample population. We use Python web scraping techniques to collect textual data from the Management Discussion and Analysis (MD&A) sections of the enterprises’ annual reports and quantify the extent of digital transformation. Additionally, using ESG rating data fromSino-Securities Index Information Service (Shanghai, China) Co. Ltd, we investigated the impact of corporate digital transformation on ESG performance, specifically focusing on the mediating effect of human capital. By examining how digital transformation initiatives influence the workforce composition and skill level, this study aimed to validate the pathways through which digital transformation enhances ESG performance. By incorporating the perspective of human capital dynamics, this approach enriches the understanding of the relationship between digital transformation and ESG performance. Furthermore, this study investigates the potential moderating effect of government environmental regulations on the relationships among corporate digital transformation, human capital investment, and ESG performance.

This study’s contribution lies in its holistic approach to exploring the relationships among corporate digital transformation, human capital, and ESG performance. We combine insights from human capital dynamics and government environmental regulations to provide a comprehensive understanding of the mechanisms behind digital transformation’s impact on ESG outcomes. Moreover, the findings provide valuable insights for policymakers, executives, and stakeholders interested in promoting sustainable development and responsible business practices in the context of digital transformation.

2. Research Hypotheses

2.1. Digital Transformation and ESG Performance

Digital technology plays a pivotal role in enterprises’ development and management processes, providing robust technical support for enhancing ESG practices [21]. Under the impetus of digital transformation, enterprises use digital technologies fully across various facets of production, operations, and management, thus effectively elevating operational efficiency, optimizing the quality of products and services, and fostering the achievement of novel developmental paradigms [22]. With the enhancement of economic performance, including financial conditions, enterprises not only acquire the capacity to shoulder social responsibilities but also cultivate the impetus to proactively undertake these responsibilities, thus contributing to boosting enterprise image and reputation [23].

First, the rise of the digital economy and digital finance has further streamlined resource allocation, enhanced information integration and sharing capabilities, and empowered enterprises to pursue green technology innovation more effectively [24]. Furthermore, the convenience stemming from digital technology has accelerated the transition toward a sustainable industrial structure [25], while also mitigating spatial carbon emissions and reducing carbon footprints within regions [26]. Moreover, digitalization in both business and public services contributes significantly to energy security and curbing air pollution, particularly in emerging markets [27]. Moreover, digital finance has demonstrated its effectiveness in curtailing pollutant emissions [28], collectively highlighting the pivotal role of digital technology in enhancing environmental performance across various sectors.

Furthermore, digital transformation enhances enterprise governance capabilities, allowing businesses to more effectively achieve their sustainability objectives [29]. This is achieved through the integration and exchange of vital information across different departments, reducing ambiguity and uncertainty in innovation decision making and improving the accuracy and efficiency of judgments [30]. Once the essential digital resources and capabilities are in place, businesses can leverage them to optimize the allocation of digital resources across departments [31], thus strengthening internal control and strategic management, ultimately promoting sustainable development [32]. Moreover, to better satisfy stakeholder expectations, enterprises should actively bear environmental and social responsibilities, offering comprehensive value in areas such as environmental governance and social responsibility beyond their excellent economic performance [33].

The widespread adoption of digital technology equips boards with efficient data-collecting tools and facilitates informed decision-making processes [34]. Simultaneously, digital technology mitigates information asymmetry among board members, which enhances decision-making efficiency and quality [35]. As enterprises embrace digital technologies, adjustments in board compositions become imperative for fulfilling the evolving demands. Requirements for diverse skills and experiences foster sex, age, race, and tenure diversity among board members [36]. Digital technologies enhance auditing and oversight processes, thus enabling more robust risk identification and assessment and optimizing decision making at every stage [37]. Additionally, the efficacy of digital technology in training audit and supervisory personnel and enhancing their professionalism and accountability merits further exploration. Therefore, we propose the following hypothesis:

H1. 

Digital transformation augments its ESG performance.

2.2. Human Capital’s Mediating Role

As enterprises transform through technological innovation and capital allocation, the absence of corresponding aligned adjustments in labor and other productive elements can result in reduced production efficiency or losses. The digital transformation has notably affected the demand for employees with advanced education, specifically increasing the need for those holding a bachelor’s degree or higher [38]. The advancement of industrial automation in China has precipitated a polarization within the labor force employment structure. On the one hand, individuals with junior and senior high school risk have been displaced; on the other hand, the demand for workers with both higher (college degree or above) and lower education (elementary school or below) has risen. Additionally, digital transformation facilitates the labor force structure’s optimization [39]. Technological progress tends to skew labor demands toward specific skills, rendering low-skilled jobs more vulnerable to automation. However, the increase in output and profits driven by technological progress also boosts the demand for high-skilled labor [40,41,42]. Notably, Dixon et al. offer an alternative perspective, suggesting that the adoption of robotics will decrease employment for medium-skilled workers but increase employment for both low- and high-skilled workers [43]. Moreover, specific roles, such as those involving social interaction, creative tasks, and emotional labor, challenge the full implementation of artificial intelligence technologies [44].

Here, we propose that digital transformation affects the composition of the labor force through technological upgrades and organizational structural adjustments. Notably, digital transformation involves business transformation, restructuring, and technological upgrades that require investment in complementary resources. Skilled labor can satisfy the demands of advanced technology. Therefore, during the digital transformation of an enterprise, the demand for high-skilled labor inevitably increases. Furthermore, digital transformation requires hardware upgrades and also soft strengths such as human capital. Enterprises may enhance employee skill levels by investing in human capital and expanding their recruitment of high-skilled workers to accommodate digital transformation needs [39]. In conclusion, digital transformation drives technological progress, compelling enterprises to increase investment in their workforce. This enhances human capital and optimizes the composition of their labor force.

The influx of skilled workers and the improvement in their abilities lead to the growth in the total human capital within businesses [45]. Human capital serves as the foundation of innovation and can improve overall societal well-being by fostering the creation of innovative goods and services. Different types of human capital exert varying effects on driving production and innovation processes [46], and the inefficient allocation of human resources may hinder the improvement in innovation efficiency [47]. Enterprises that successfully integrate human capital with innovative technology frequently transform resources into novel products or services effectively and nurture their competitive edge through continuous innovation. Moreover, an optimized labor force endowed with higher levels of skills and knowledge can more profoundly engage in understanding and implementing an enterprise’s sustainability strategies.

Furthermore, an optimized human capital structure plays a crucial role in enhancing ESG performance. Human capital positively influences the technological innovation capabilities of enterprises, resulting in environmental sustainability advancements [48]. Skilled and knowledgeable employees are indispensable for the development and implementation of eco-friendly technologies, thus reducing environmental impact and ensuring compliance with environmental regulations. By introducing innovative products and services that benefit society, enterprises can further enhance their societal welfare contributions. Additionally, board members with diverse expertise contribute to effective oversight, strategic planning, and risk assessment, thus enhancing corporate governance practices [49]. Human capital optimization can improve internal management efficiency and decision-making quality in an enterprise, ultimately improving its ESG performance. Accordingly, we propose the following hypotheses:

H2a. 

Digital transformation can promote ESG performance by increasing the quantity of high-quality labor.

H2b. 

Digital transformation can enhance ESG performance by elevating the existing labor force’s skill levels.

2.3. Environmental Regulation’s Moderating Role

Governments use environmental regulation as a policy tool to enforce administrative restrictions and control corporate production practices in order to address environmental pollution problems, accomplish environmental protection objectives, and improve the standard of economic development [50].

Based on the existing research, environmental regulation is likely to moderate the digital transformation impact on the structure of the workforce and, subsequently, influence ESG performance. Environmental regulations serve as a policy tool to regulate corporate behavior and promote environmental protection. They can influence the adoption of digital technologies for ESG performance improvement as they provide a clear legal framework that organizations must adhere to [51].

Environmental regulations are known to shape corporate behavior and investment decisions, including decisions related to workforce composition and skill development. Stringent environmental regulations may incentivize firms to invest in high-quality labor and enhance the skill levels of their existing workforce to comply with regulatory requirements and adopt cleaner production technologies [48,52,53]. Environmental regulations can drive firms to innovate and adopt digital technologies that improve environmental performance and influence the demand for skilled labor with digital competencies in this way [17].

The interaction between environmental regulation and digital transformation can have implications for workforce dynamics. For example, firms subject to stringent environmental regulations may prioritize investments in digital technologies to improve resource efficiency, reduce waste, and enhance environmental monitoring and reporting capabilities [54]. This move may result in changes in the quantity and skill level of the workforce as firms seek to adapt to technological advancements and regulatory requirements.

Furthermore, environmental regulation’s impact on workforce structures can cascade into ESG performance outcomes. A workforce comprising high-quality labor with advanced skills is better equipped to implement sustainable practices, innovate, and drive improvements in environmental, social, and governance performance indicators [48,55]. Therefore, environmental regulation may influence ESG performance indirectly through its effects on workforce composition and skill development. Accordingly, we propose the following hypotheses:

H3a. 

Environmental regulation moderates the positive impact of corporate digital transformation on the quantity of high-quality labor, thus influencing the firm’s ESG performance.

H3b. 

Environmental regulation moderates the positive impact of corporate digital transformation on the quantity of skilled labor, thus influencing the firm’s ESG performance.

3. Research Design

3.1. Data Sources and Processing

This study analyzes enterprises listed on the A-share market in China from 2010 to 2022 and utilizes data sourced from the CSMAR and Wind databases. The sample selection process adhered to the framework established by Cai et al. [15]. Specifically, enterprises facing special circumstances, such as ST (Special Treatment), *ST (Special Treatment), and those subjected to delisting during the observation period, were excluded from the sample. Additionally, observations lacking pertinent variables were omitted. Consequently, a panel dataset comprising 17,331 observations was meticulously constructed, ensuring the dataset’s robustness and relevance for subsequent analysis.

3.2. Variables’ Definition

3.2.1. ESG Performance

ESG performance is the primary explained variable. To assess the ESG performance of Chinese-listed companies, we adopt the ESG rating system developed bySino-Securities Index Information Service (Shanghai, China) Co. Ltd, per the methodology outlined by Lin et al. [56]. This rating system encompasses over 130 indicators in 14 primary themes: Environmental Management Mechanism, Environmental Business Objectives, Green Product, Environmental Accreditation, Environmental Controversies (comprising five criteria under the E dimension), Social Management Mechanism, Social Activity and Controversies, Social Contributions, Social Accreditation (comprising four criteria under the S dimension), Governance Management Mechanism, Corporate Governance, Governance Business, Operational Risk, and Governance Controversies (comprising five criteria under the G dimension).We stratify the ESG ratings into the following nine levels: C, CC, CCC, B, BB, BBB, A, AA, and AAA. In our analysis, we assign a score of 9 to companies with an AAA ESG rating, a score of 8 to those with an AA rating, and so forth, until a score of 1 is allocated to those with a C rating. Therefore, ESG scores range from 1 to 9, with higher scores indicating superior ESG performance.

3.2.2. Digital Transformation

The explanatory variable is digital transformation, and it represents the degree of adoption of digital technologies in a firm. Our primary measure of digitization, denoted as Dig, captures the level of digital transformation within firm i during year t. Drawing from the methodology proposed by Verhoef et al., we construct a metric for digitization through a three-step process [57,58]. First, we identify relevant terms associated with the digital economy based on the pertinent literature on the digitization of Chinese firms, national policies, and research reports. This forms the basis of a semantic dictionary delineating digitization in listed companies.

Second, we employ text analysis methods and machine learning techniques to develop a semantic dictionary specific to digitization within listed companies’ annual reports, particularly focusing on MD&A sections. This dictionary enables us to extract word frequency information related to digitization.

Finally, considering the variations in the length of companies’ MD&A sections across annual reports, we standardized our digitization metric by calculating the ratio of the total frequency of digitization-related words to the length of the annual discourse. The ratio offers a robust measure of digitization extent in Chinese-listed enterprises. It reflects both the frequency and prominence of digitalization references in their annual reports. For ease of interpretation, we scale this index by multiplying it by 100. Consequently, a higher value of this indicator signifies increased frequency and prominence of digitalization references in the company’s annual report, which indicates a higher digitization degree.

3.2.3. Mediating Variables

In this study, the mediating variable was bifurcated into two distinct components—namely, high-quality labor force and labor force skill level. Drawing on the framework established by Huang et al. [59], high-quality labor (highlabor) and labor skill levels (laborskills) were selected as intermediary measures.

High-quality labor is operationalized by the number of employees having a bachelor’s degree or higher. This metric was selected based on the premise of the relatively stable learning experiences of individuals in professional environments. Variations in the educational backgrounds of employees within a company’s workforce predominantly reflect adjustments in the enterprise’s recruitment strategies. Therefore, the number of employees holding a bachelor’s degree or higher serves as a critical criterion for assessing whether a business has effectively attracted and retained high-quality labor.

Concerning the evaluation of labor skill level (laborskills), this study directly utilizes the number of technical staff members as a measurement basis. Changes in the technical workforce composition within an enterprise arise from the influx of new skilled talent and the skill enhancement efforts undertaken by internal employees. Considering the dynamic nature of skill development and transformation, the number of technical staff members is deemed an important indicator—reflecting the progression of the enterprise’s current labor skill level.

3.2.4. Moderating Variable

This study uses environmental regulation (ER) as a moderating variable. Considering the absence of a standardized method for quantifying ER, we adopt a refined approach inspired by Mulatu to enhance the accuracy of measuring environmental regulation intensity [54].

To ensure the comparability of environmental regulatory intensity across different periods, we refine the measurement methodology using the ratio of investment in environmental pollution control to local gross industrial product in each province. This ratio acts as a proxy for the level of environmental regulation, reflecting the commitment of regional authorities to environmental protection efforts relative to their economic output [50,60]. Data for this measurement were obtained from the Environmental Statistics Yearbook of China.

3.2.5. Control Variables

Building on the previous research on digital transformation and ESG [15,59], our model incorporates several firm characteristics and corporate governance variables as control measures. These control variables serve to enhance comparability with prior studies and mitigate the possibility of omitted variable bias, thereby strengthening the robustness of our analysis.

Table 2. Descriptions of key variables.

Types	Names	Symbols	Descriptions
Independent variable	Digital transformation	Dig	Frequency of terms related to digital transformation divided by the length of annual report MD&A section, multiplied by 100
Dependent variables	Environmental, social responsibility, and corporate governance ratings	ESG	Adopting Sino-securities ESG rating data, assigning values from 1 to 9 in an ascending order
	Environmental ratings	Erating	Adopting Sino-securities ESG rating data, assigning values from 1 to 9 in an ascending order
	Social responsibility ratings	Srating	Adopting Sino-securities ESG rating data, assigning values from 1 to 9 in an ascending order
	Corporate governance ratings	Grating	Adopting Sino-securities ESG rating data, assigning values from 1 to 9 in an ascending order
Mediating variables	High-quality labor force	lnhighlabor	Natural logarithm of the number of employees with a bachelor’s degree or above
	Labor force skill level	lnlaborskills	Natural logarithm of the number of technical staff
Moderating variable	Environmental regulation	ER	Ratio of investment in environmental pollution control to the local gross industrial product in each province
Control variables	Company size	Size	Natural logarithm of total annual assets
	Revenue growth rate	Growth	Current year’s business revenue/last year’s business revenue-1
	Net profit margin on sales	NetProfit	Net profit/business revenue
	Leverage	Lev	Year-end total liability/year-end total assets
	Proportion of fixed assets	FIXED	Net fixed assets/total assets
	Return on assets	ROA	Net profit/average balance of total assets
	State-owned enterprises	SOE	State-owned enterprises assume the value of 1, others 0
	Number of directors	Board	Natural logarithm of the number of board directors
	Ratio of independent directors	Indep	Number of independent directors divided by the number of directors
	Cash flow ratio	Cashflow	Net cash flow from operating activities/total assets
	CEO–Chairman duality	Dual	Chairman and CEO being the same person is valued as 1, otherwise 0
	Book-to-market	BM	Book value/total market value
	Financing constraints	SA	SA index, the larger the absolute value, the more severe the financing constraints

provides the detailed descriptions of the control variables.

3.3. Model Specification

We utilize the ordinary least squares (OLS) analysis to examine the relationship between digital transformation and ESG performance from a human capital perspective.

3.3.1. Benchmark Regression Model

Building on the insights from Cai et al. [15], we formulate Model (1) to investigate the influence of digital transformation on their ESG performance:

$${ESG}_{i,t}=a_0+\beta {Dig}_{i,t-1}+\lambda \sum {controls}_{i,t-1}+{\delta }_i+{\gamma }_t+{\epsilon }_{i,t}$$

(1)

The subscript variables $i$ and t denote the enterprise and time, respectively. The key dependent variable, ${ESG}_{i,t}$, represents the ESG rating of enterprise i in year t. The key independent variable, ${Dig}_{i,t-1}$, denotes the degree of digitalization of enterprise $i$ in year −1. The key independent variable, ${Dig}_{i,t-1},$ is lagged by one period to account for potential time lags in the impact of digital transformation on ESG performance. This lagging enables a more accurate assessment of the causal relationship between digitalization efforts in the previous period and subsequent changes in ESG performance. By incorporating the lagged variable, we aim to capture the delayed effects of digital transformation initiatives on ESG outcomes, thereby enhancing the robustness and validity of our analysis. $Controls$ refers to a range of control variables. To manage interference from other related factors in identifying causality, we consult Huang et al. [59] and Cai et al. [15] for control selection. Moreover, ${\delta }_i$ and ${\gamma }_t$ denote the firm and year fixed effect, respectively. ${\epsilon }_{i,t}$ is the random perturbation term, and $a_0$ is the constant term.

3.3.2. Mediating Effect Test Model

Drawing on Wang et al.’s research [56], we construct a two-stage regression model based on the benchmark regression model to probe the transmission mechanism whereby digital transformation influences ESG responsibility performance through the structure of human capital. The model is as follows:

$${Mediator}_{i,t}=a_0+\beta {Dig}_{i,t-1}+\lambda \sum {controls}_{i,t-1}+{\delta }_i+{\gamma }_t+{\epsilon }_{i,t},$$

(2)

$${ESG}_{i,t}=a_0+{\beta }_1{Dig}_{i,t-1}+{\beta }_2{Mediator}_{i,t}+\lambda \sum {controls}_{i,t-1}+{\delta }_i+{\gamma }_t+{\epsilon }_{i,t}$$

(3)

Additionally, the mediator variable ${Mediator}_{i,t}$ includes two components, namely, high-quality labor (highlabor) and labor skill levels (laborskills). These components are included to assess the mediating role of workforce characteristics in the relationship between digital transformation and ESG performance. By considering both high-quality labor and labor skill levels as mediators, we aim to comprehensively examine how enhancements in digital capabilities influence workforce dynamics and, subsequently, impact ESG outcomes in enterprises.

3.3.3. Moderating Effect Test Model

Building upon the insights from Li et al. [52], Models (4) and (5) are formulated to investigate the moderating effect of ER on the relationship between digital transformation (Dig) and high-quality labor (highlabor) and labor skill levels (laborskills):

$${highlabor}_{i,t}=a_0+{\beta }_1{Dig}_{i,t-1}+{\beta }_2{ER}_{i,t}+{\beta }_3{ER}_{i,t}\times {Dig}_{i,t-1}+\lambda \sum {controls}_{i,t-1}+{\delta }_i+{\gamma }_t+{\epsilon }_{i,t},$$

(4)

$${laborskills}_{i,t}=a_0+{\beta }_1{Dig}_{i,t-1}+{\beta }_2{ER}_{i,t}+{\beta }_3{ER}_{i,t}\times {Dig}_{i,t-1}+\lambda \sum {controls}_{i,t-1}+{\delta }_i+{\gamma }_t+{\epsilon }_{i,t}$$

(5)

In Models (4) and (5), ER serves as the moderating variable, influencing how digital transformation impacts high-quality labor and labor skill levels within enterprises. By examining the interaction effects between digital transformation and environmental regulation, we aim to comprehend how regulatory environments shape the relationship between digitalization efforts and workforce characteristics related to high-quality labor and labor skill levels.

4. Empirical Result Analysis

4.1. Descriptive Statistics

Table 3. Descriptive statistics.

VarName	Obs.	Mean	SD	Min	Median	Max
Dig	17,331	1.2019	1.182	0.00	0.76	5.62
ESG	17,331	4.2010	1.066	1.00	4.00	8.00
Erating	17,331	2.0971	1.250	1.00	2.00	9.00
Srating	17,331	4.3179	1.121	1.00	4.00	8.00
Grating	17,331	5.2481	1.353	1.00	6.00	9.00
lnhighlabor	17,331	6.3942	1.363	0.00	6.30	12.16
lnlaborskills	17,331	6.1890	1.218	0.00	6.08	12.30
ER	17,331	0.0015	0.001	0.00	0.00	0.02
Size	17,331	22.4426	1.281	19.23	22.25	28.11
Growth	17,331	0.1904	1.167	−0.91	0.11	82.79
NetProfit	17,331	0.0633	0.276	−8.91	0.07	6.86
Lev	17,331	0.4245	0.194	0.01	0.42	1.96
FIXED	17,331	0.2082	0.146	0.00	0.18	0.95
ROA	17,331	0.0405	0.074	−0.89	0.04	1.28
SOE	17,331	0.3314	0.471	0.00	0.00	1.00
Board	17,331	2.1209	0.196	1.39	2.20	2.89
Indep	17,331	37.6836	5.582	14.29	36.36	80.00
Cashflow	17,331	0.0502	0.067	−0.56	0.05	0.84
Dual	17,331	0.2916	0.455	0.00	0.00	1.00
BM	17,331	0.6212	0.254	0.03	0.61	1.60
SA	17,331	−3.8517	0.256	−5.60	−3.86	−2.35

provides descriptive statistics for the key variables. The mean value of digital transformation (Dig) within enterprises is 1.2019, suggesting that, on average, 1.2019% of the terms in the MD&A section of listed enterprises’ annual reports are related to digitization. The maximum value observed for Dig is 5.62, indicating substantial variability in digitization levels across different enterprises. The distribution characteristics of the remaining variables align closely with those reported in previous studies, which makes further elaboration unnecessary.

4.2. Benchmark Regression Results

Table 4. Digital transformation and ESG performance.

	(1)	(2)	(3)	(4)
	ESG	Erating	Srating	Grating
Dig	0.0681 ***	0.0429 **	0.0430 ***	0.0088
	[0.015]	[0.022]	[0.016]	[0.019]
Size	0.1543 ***	0.2458 ***	0.1308 ***	0.0462
	[0.030]	[0.032]	[0.029]	[0.041]
Growth	−0.0005	−0.0031 ***	0.0002	−0.0006
	[0.001]	[0.001]	[0.001]	[0.002]
NetProfit	0.1699 ***	−0.0327	0.1608 **	0.2741 ***
	[0.065]	[0.045]	[0.066]	[0.101]
Lev	−0.6349 ***	−0.2752 **	−0.4320 ***	−1.4605 ***
	[0.104]	[0.111]	[0.105]	[0.150]
FIXED	−0.3227**	0.1401	−0.3320 **	−0.4055 **
	[0.136]	[0.157]	[0.137]	[0.189]
ROA	2.0878 ***	−0.1227	2.0732 ***	2.8821 ***
	[0.231]	[0.212]	[0.241]	[0.347]
SOE	0.0727	−0.0537	0.1359 **	0.0995
	[0.065]	[0.075]	[0.060]	[0.098]
Board	0.0387	−0.0927	0.0565	0.0666
	[0.105]	[0.107]	[0.107]	[0.131]
Indep	0.0072 **	−0.0027	0.0082 ***	0.0131 ***
	[0.003]	[0.003]	[0.003]	[0.003]
Cashflow	−0.2074	0.4148 ***	−0.2035	−0.2020
	[0.146]	[0.148]	[0.152]	[0.202]
Dual	0.0080	−0.0006	0.0151	0.0186
	[0.029]	[0.033]	[0.029]	[0.041]
BM	−0.0940	0.0148	−0.1251 *	−0.0263
	[0.067]	[0.073]	[0.066]	[0.091]
SA	1.0422 ***	1.0245 ***	1.0210 ***	1.6282 ***
	[0.202]	[0.242]	[0.200]	[0.250]
Constant	3.5261 ***	−0.4125	3.7533 ***	10.6653 ***
	[1.203]	[1.160]	[1.197]	[1.583]
Firm FE	YES	YES	YES	YES
Year FE	YES	YES	YES	YES
N	17,331	17,331	17,331	17,331
r2	0.0610	0.0693	0.1566	0.1337

Note: Parentheses contain clustered standard errors at the enterprise level. *, **, and *** denote statistical significance at the 10%, 5%, and 1% levels, respectively. The same applies below.

presents the benchmark regression results regarding the impact of digital transformation on ESG performance. Column (1) of Table 4 examines the effect of digital transformation on ESG performance. The coefficient of Dig is 0.0681, which is statistically significant at the 1% level. This indicates that digital transformation positively affects their ESG performance, thus providing support for H1. Further disaggregated analysis in Column (3) of Table 4 indicates that digital transformation significantly influences their environmental and social responsibility performance.

4.3. Mediation Effect Test

Table 5. Digital transformation, high-quality labor, and ESG performance.

	(1)	(2)	(3)	(4)	(5)
	Lnhighlabor	ESG	Erating	Srating	Grating
Dig	0.0231 **	0.0659 ***	0.0423 *	0.0410 ***	0.0080
	[0.010]	[0.015]	[0.022]	[0.016]	[0.019]
lnhighlabor		0.0934 ***	0.0267	0.0841 ***	0.0373
		[0.023]	[0.023]	[0.023]	[0.033]
Constant	−6.8903 ***	4.1698 ***	−0.2284	4.3328 ***	10.9225 ***
	[0.719]	[1.223]	[1.162]	[1.218]	[1.608]
Control variables	YES	YES	YES	YES	YES
Firm FE	YES	YES	YES	YES	YES
Year FE	YES	YES	YES	YES	YES
N	17,331	17,331	17,331	17,331	17,331
r2	0.4380	0.0627	0.0695	0.1579	0.1339

Note: Parentheses contain clustered standard errors at the enterprise level. *, **, and *** denote statistical significance at the 10%, 5%, and 1% levels, respectively. The same applies below.

presents the regression model data. It shows the influence of digital transformation on ESG performance when the quantity of high-quality labor is increased. In Column (1), the estimated parameters of digital transformation are significant at the 10% level, indicating that digital transformation significantly contributes to the enhancement of high-quality labor in the research framework.

In Column (2), regression results are reported with high-quality labor serving as the mediating variable, digital transformation as the independent variable, and ESG performance as the dependent variable. The coefficient of Dig is 0.0659, which is statistically significant at the 1% level. The coefficient of lnhighlabor is 0.0934, also statistically significant at the 1% level. These findings suggest that digital transformation can promote ESG performance by increasing the quantity of high-quality labor.

The regression coefficient of digital transformation on enterprise ESG is slightly lower than that of Table 5, Column (1), indicating that high-quality labor plays a partial mediating role between digital transformation and enterprise ESG. Consequently, H2a is supported.

Table 6. Digital transformation, labor skill levels, and ESG performance.

	(1)	(2)	(3)	(4)	(5)
	Lnlaborskills	ESG	Erating	Srating	Grating
Dig	0.0499 ***	0.0634 ***	0.0426 *	0.0385 **	0.0067
	[0.010]	[0.015]	[0.022]	[0.016]	[0.019]
lnlaborskills		0.0929 ***	0.0069	0.0890 ***	0.0427
		[0.023]	[0.023]	[0.023]	[0.033]
Constant	−6.8836 ***	4.1656 ***	−0.3652	4.3662 ***	10.9592 ***
	[0.825]	[1.218]	[1.161]	[1.217]	[1.605]
Control variables	YES	YES	YES	YES	YES
Firm FE	YES	YES	YES	YES	YES
Year FE	YES	YES	YES	YES	YES
N	17,331	17,331	17,331	17,331	17,331
r2	0.3420	0.0628	0.0694	0.1581	0.1339

Note: Parentheses contain clustered standard errors at the enterprise level. *, **, and *** denote statistical significance at the 10%, 5%, and 1% levels, respectively. The same applies below.

lists the regression model data, indicating that digital transformation can enhance ESG performance by raising the skill levels of the existing labor force. In Column (1), the estimated parameters of digital transformation are significant at the 1% level, indicating that digital transformation significantly promotes improvement in labor skill levels.

In Column (2), regression results are reported, with labor skill levels serving as the mediating variables, digital transformation as the independent variable, and ESG performance as the dependent variable. The findings reveal a significant positive relationship between the estimated coefficient of the mediating variable, labor skill level, and enterprise performance. The coefficient of Dig is 0.0634, statistically significant at the 1% level, and the coefficient of lnlaborskills is 0.0929, also statistically significant at the 1% level.

These results suggest that digital transformation enhances labor skill levels, and optimized human capital augments the fulfillment of social responsibilities by enterprises, thereby promoting improvement in their ESG performance. Thus, H2b is validated.

4.4. Moderating Effect Test

In the study, two regression models analyze the impact of Dig on the quantity of high-quality labor. In Column (1), Dig significantly positively affects the quantity of high-quality labor, with a coefficient of 0.0231 *, which behaves as expected.

Column (2) introduces an interaction term (Dig × ER) revealing a significant moderating effect of environmental regulation (ER) on the relationship between digital transformation and quantity of high-quality labor, with a coefficient of 7.5649 **, without altering the overall model fit (R-squared: 0.4383;

Table 7. Digital transformation, ER, and labor skill levels.

	(1)	(2)	(3)	(4)
	Lnlaborskills	Lnhighlabor	Lnlaborskills	Lnlaborskills
Dig	0.0231 **	0.0165	0.0499 ***	0.0447 ***
	[0.010]	[0.010]	[0.010]	[0.011]
ER		−4.2673		−7.7891
		[4.395]		[5.681]
Dig × ER		7.5649 **		6.1065 *
		[3.600]		[3.459]
Constant	−6.8903 ***	−6.9238 ***	−6.8836 ***	−6.9148 ***
	[0.719]	[0.719]	[0.825]	[0.824]
Control variables	YES	YES	YES	YES
Firm FE	YES	YES	YES	YES
Year FE	YES	YES	YES	YES
N	17,331	17,331	17,331	17,331
r2	0.4380	0.4383	0.3420	0.3423

Note: Parentheses contain clustered standard errors at the enterprise level. *, **, and *** denote statistical significance at the 10%, 5%, and 1% levels, respectively. The same applies below.

).

As Table 7 reveals, Column (3) demonstrates a significant positive relationship between Dig and the skill levels of the existing labor force, with a coefficient of 0.0499 ***, indicating a strong digital transformation impact.

Column (4) includes the interaction term Dig × ER, which reveals a substantial positive effect with a coefficient of 6.1065 *, suggesting that the digital transformation on the skill levels of the existing labor force is significantly enhanced in the context of ER. The overall model fit improves slightly with the inclusion of the interaction term (R-squared: 0.3423). Thus, H3a and H3b are supported.

4.5. Endogeneity Test

The causal relationship between digital transformation and ESG performance may entail mutual influence, thereby introducing endogeneity concerns. To address this issue, this study adopts Cai et al.’s [15] method, employing the average level of digital transformation among other enterprises in the same region and industry (ivdigital_mean) as an instrumental variable. Theoretically, the level of digital development within the industry of an enterprise’s region can influence its degree of digitalization, thus fulfilling the relevance condition. Simultaneously, the digital development levels of other enterprises in the same industry and region do not directly affect the enterprise’s revenue distribution decisions, thus satisfying the exogeneity condition. As

Table 8. Endogeneity test.

	First Stage	Second Stage
	(1)	(2)	(3)	(4)	(5)	(6)	(7)
	Dig	ESG	Erating	Srating	Grating	Lnhighlabor	Lnlaborskills
ivdigital_mean	0.0028 ***	0.1500 ***	0.1044 ***	0.2219 ***	0.0532 **	0.2442 ***	0.4357 ***
	[0.024]	[0.024]	[0.033]	[0.025]	[0.026]	[0.026]	[0.025]
Constant	−2.098	−0.2698	−3.2845 ***	−1.4036 ***	3.4508 ***	−12.2582 ***	−8.2613 ***
	[0.850]	[0.422]	[0.561]	[0.431]	[0.460]	[0.424]	[0.435]
Control variables	YES	YES	YES	YES	YES	YES	YES
Firm FE	YES	YES	YES	YES	YES	YES	YES
Year FE	YES	YES	YES	YES	YES	YES	YES
N	16,434	16,434	16,434	16,434	16,434	16,434	16,434
r2	0.3254	0.1293	0.0674	0.1162	0.2015	0.6193	0.5203

Note: Parentheses contain clustered standard errors at the enterprise level. **, and *** denote statistical significance at the 5%, and 1% levels, respectively. The same applies below.

indicates, the coefficient of ivdigital_mean in the first-stage regression is 0.0028, which is significantly positive at the 1% level, and the first-stage F-test is 706.995, supporting the feasibility of the selected instrumental variable. Column (2) of Table 8 demonstrates that this study’s main conclusions remain robust and reliable after employing the instrumental variable.

4.6. Robustness Test

To further validate our empirical investigation’s outcomes, we conducted robustness checks on our model employing a variable substitution technique.

Table 9. Robustness test’s results.

	Changed Explained Variable	Improving Sample Quality	Eliminate Big Events	Incorporating Province-Year Fixed Effects
	(1)	(2)	(3)	(4)
	ESG	ESG	ESG	ESG
Dig	0.0048 ***	0.0732 ***	0.0558 **	0.0583 ***
	[0.001]	[0.017]	[0.023]	[0.014]
Constant	4.2665 ***	4.0857 ***	5.7537 ***	1.4106 **
	[0.086]	[1.274]	[1.430]	[0.567]
Control variables	YES	YES	YES	YES
Firm FE	YES	YES	YES	YES
Province FE	NO	NO	NO	YES
Year FE	YES	YES	YES	YES
N	17,331	15,203	12,531	17,331
r2	0.0722	0.0635	0.0750	0.0525

Note: Parentheses contain clustered standard errors at the enterprise level. **, and *** denote statistical significance at the 5%, and 1% levels, respectively. The same applies below.

presents these assessments. In Column (1), we replace the dependent variable by directly taking the logarithm of the score results to replace the rating results of the Huazhong ESG score. The analysis yields a result consistent with our original findings: digital transformation has a significantly positive impact on the ESG metric, thus supporting our initial results’ reliability at the 1% significance level.

To prevent the inflation of the overall digitization level caused by specific companies’ crucial digital attributes, we exclude companies with industry C39 codes in the computer, communication, and other electronic equipment manufacturing sectors. The regression outcomes after this sample refinement are displayed in Column (2) of Table 9. The magnitude and significance level of digital transformation for ESG remain unaltered.

This study does not consider the influence of major global events. In 2020, China halted production activities in response to the coronavirus pandemic, potentially causing deviations in environmental performance. To address this issue, samples from 2010 and 2020 are excluded from re-estimation. Compared to the baseline regression results, Column (3) of Table 9 indicates that the digital transformation coefficient and significance level of ESG remain consistent, underscoring the reliability of the research conclusions.

Incorporating province-year fixed effects into the model, the outcomes in Column (4) of Table 9 indicate that, with the inclusion of fixed effects, digital transformation estimated impacts on ESG performance remain robust.

4.7. Heterogeneity Exploration

After explaining the intrinsic mechanism whereby digital transformation influences ESG performance through the quantity of high-quality labor and labor skill levels, this section investigates the effects of digital transformation on enhancing ESG performance, as some properties of the enterprises may vary. Consequently, a group regression was conducted based on ownership and industry types, with the estimated results presented in

Table 10. Exploration of heterogeneity in enterprise ownership type (high-quality labor force).

	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
	ESG		Erating		Srating		Grating
	SOE	Non-SOE	SOE	Non-SOE	SOE	Non-SOE	SOE	Non-SOE
Dig	0.0563 **	0.0703 ***	0.0325	0.0459 *	0.0287	0.0475 **	−0.0041	0.0168
	[0.028]	[0.018]	[0.040]	[0.026]	[0.028]	[0.019]	[0.029]	[0.023]
lnhighlabor	0.0327	0.1285 ***	−0.0144	0.0477 *	0.0356	0.1171 ***	0.0353	0.0533
	[0.038]	[0.029]	[0.042]	[0.028]	[0.036]	[0.029]	[0.046]	[0.042]
Constant	1.6179	3.9735 ***	1.0434	−0.8898	0.8979	4.3643 ***	4.9192 *	12.6373 ***
	[2.189]	[1.403]	[2.192]	[1.448]	[2.245]	[1.369]	[2.875]	[1.796]
Control variables	YES	YES	YES	YES	YES	YES	YES	YES
Firm FE	YES	YES	YES	YES	YES	YES	YES	YES
Year FE	YES	YES	YES	YES	YES	YES	YES	YES
N	5744	11,587	5744	11,587	5744	11,587	5744	11,587
r2	0.0713	0.0731	0.0622	0.0771	0.1320	0.1826	0.0888	0.1678

Note: Parentheses contain clustered standard errors at the enterprise level. *, **, and *** denote statistical significance at the 10%, 5%, and 1% levels, respectively. The same applies below.

,

Table 11. Exploration of heterogeneity in enterprise ownership type (labor force skill level).

	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
	ESG		Erating		Srating		Grating
	SOE	Non-SOE	SOE	Non-SOE	SOE	Non-SOE	SOE	Non-SOE
Dig	0.0558 **	0.0665 ***	0.0311	0.0465 *	0.0276	0.0440 **	−0.0041	0.0150
	[0.028]	[0.018]	[0.040]	[0.026]	[0.028]	[0.019]	[0.029]	[0.023]
lnlaborskills	0.0215	0.1357 ***	0.0187	0.0084	0.0333	0.1261 ***	0.0124	0.0623
	[0.036]	[0.028]	[0.040]	[0.030]	[0.037]	[0.029]	[0.043]	[0.045]
Constant	1.6012	3.8636***	1.3162	−1.1722	0.9725	4.2784 ***	4.7992 *	12.6262 ***
	[2.192]	[1.387]	[2.178]	[1.455]	[2.254]	[1.364]	[2.881]	[1.777]
Control variables	YES	YES	YES	YES	YES	YES	YES	YES
Firm FE	YES	YES	YES	YES	YES	YES	YES	YES
Year FE	YES	YES	YES	YES	YES	YES	YES	YES
N	5744	11,587	5744	11,587	5744	11,587	5744	11,587
r2	0.0712	0.0736	0.0622	0.0767	0.1320	0.1830	0.0886	0.1679

Note: Parentheses contain clustered standard errors at the enterprise level. *, **, and *** denote statistical significance at the 10%, 5%, and 1% levels, respectively. The same applies below.

,

Table 12. Exploration of heterogeneity in enterprise industry type (high-quality labor force).

	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
	ESG		Erating		Srating		Grating
	Manufacturing	Non-Manufacturing	Manufacturing	Non-Manufacturing	Manufacturing	Non-Manufacturing	Manufacturing	Non-Manufacturing
Dig	0.0589 ***	0.0968 ***	0.0949 ***	−0.0614	0.0375 *	0.0717 **	−0.0082	0.0589
	[0.019]	[0.028]	[0.023]	[0.050]	[0.019]	[0.029]	[0.022]	[0.038]
lnhighlabor	0.0714 ***	0.2298 ***	0.0207	0.0064	0.0700 ***	0.2074 ***	0.0117	0.1987 **
	[0.025]	[0.054]	[0.026]	[0.049]	[0.025]	[0.055]	[0.035]	[0.082]
Constant	3.7379 ***	6.7628 ***	−0.6327	1.8835	4.1176 ***	6.5262 ***	11.2805 ***	9.3974 ***
	[1.384]	[2.204]	[1.452]	[2.237]	[1.340]	[2.249]	[1.848]	[2.849]
Control variables	YES	YES	YES	YES	YES	YES	YES	YES
Firm FE	YES	YES	YES	YES	YES	YES	YES	YES
Year FE	YES	YES	YES	YES	YES	YES	YES	YES
N	13,125	4206	13,125	4206	13,125	4206	13,125	4206
r2	0.0572	0.0927	0.0791	0.0639	0.1721	0.1299	0.1342	0.1392

Note: Parentheses contain clustered standard errors at the enterprise level. *, **, and *** denote statistical significance at the 10%, 5%, and 1% levels, respectively. The same applies below.

and

Table 13. Exploration of heterogeneity in enterprise industry type (labor force skill level).

	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
	ESG		Erating		Srating		Grating
	Manufacturing	Non-Manufacturing	Manufacturing	Non-Manufacturing	Manufacturing	Non-Manufacturing	Manufacturing	Non-Manufacturing
Dig	0.0564 ***	0.0912 ***	0.0949 ***	−0.0605	0.0350 *	0.0658 **	−0.0077	0.0513
	[0.019]	[0.028]	[0.023]	[0.050]	[0.019]	[0.029]	[0.022]	[0.038]
lnlaborskills	0.0896 ***	0.1328 ***	0.0122	−0.0190	0.0889 ***	0.1364 ***	−0.0030	0.1703 ***
	[0.028]	[0.044]	[0.028]	[0.045]	[0.028]	[0.045]	[0.040]	[0.059]
Constant	3.7426 ***	7.1425 ***	−0.7176	1.7456	4.1291 ***	6.9768 ***	11.1708 ***	10.0877 ***
	[1.373]	[2.237]	[1.449]	[2.245]	[1.334]	[2.281]	[1.836]	[2.903]
Control variables	YES	YES	YES	YES	YES	YES	YES	YES
Firm FE	YES	YES	YES	YES	YES	YES	YES	YES
Year FE	YES	YES	YES	YES	YES	YES	YES	YES
N	13,125	4206	13,125	4206	13,125	4206	13,125	4206
r2	0.0577	0.0878	0.0791	0.0639	0.1725	0.1275	0.1341	0.1396

Note: Parentheses contain clustered standard errors at the enterprise level. *, **, and *** denote statistical significance at the 10%, 5%, and 1% levels, respectively. The same applies below.

.

This study classifies the sample into two categories based on the type of enterprise ownership—state-owned and non-state-owned enterprises. As Table 10 details, the findings indicate significantly enhanced ESG performance through the quantity of high-quality labor generated by digital transformation in non-state-owned enterprises compared to state-owned enterprises.

As Table 11 indicates, the research findings imply significantly enhanced ESG performance through labor skill level by digital transformation in non-state-owned enterprises compared to state-owned enterprises.

This study divides the sample into two groups based on the disparate industry types of the enterprises—specifically, manufacturing and non-manufacturing. As Table 12 and Table 13 indicate, the industry type does not significantly influence the outcomes. Both manufacturing and non-manufacturing sectors demonstrate comparable impacts of digital transformation on ESG performance metrics, suggesting that the benefits of digital transformation in fostering sustainable practices and enhancing ESG performance extend across diverse industry sectors.

5. Discussion

Our findings contribute to the ongoing discourse on the relationship between digital transformation and ESG performance, elucidating both expected outcomes and nuanced disparities. Consistent with the existing research, our study supports the assertion that digital transformation initiatives generally affect ESG performance positively. However, a deeper examination reveals intriguing variations across different ESG dimensions. Environmental and social dimensions are particularly sensitive to the influence of digital transformation, with discernible improvements in scores. These findings resonate with prior studies, affirming the narrative that technological advancements and process optimizations frequently translate into reduced environmental footprints and enhanced social responsibility within organizations. Fang’s study echoes this sentiment, illustrating how digitization efforts correlate with enhanced social metrics and, to some extent, governance standards [14]. This study diverges notably in its observation of insignificant improvements in governance scores, departing from some previous conclusions.

The differences in findings between studies can be attributed to various factors, such as variations in sample characteristics, methodological approaches, and the complex nature of digital transformation and ESG. Fang’s emphasis on non-state-owned listed companies highlights distinct dynamics that may not be fully reflected in our study, which includes a wider range of enterprises. Moreover, the specific aspects of governance evaluated in our analysis warrant scrutiny. Governance encompasses a multifaceted array of factors, ranging from board composition to risk management practices, and the efficacy of digital transformation initiatives in addressing these facets may vary. One plausible explanation for the lack of significant improvements in governance scores in our study lies in the focus on digital transformation initiatives within the sampled enterprises. While efforts may have prioritized operational efficiency and social responsibility, governance practices might not have been explicitly targeted. Furthermore, our study’s temporal scope might not capture the gradual evolution of governance frameworks, which can manifest over an extended period.

Building on the conclusions drawn from this study, we assert that digital transformation serves as a catalyst for enhancing ESG performance through its profound impact on human capital. The study suggests that a primary mechanism through which digital transformation fosters ESG outcomes is by augmenting the quality of labor within organizations while concurrently elevating the skill levels of the existing workforce. This restructuring of human capital enhances the technological innovation capacity of firms, thereby contributing to environmental sustainability. Skilled and knowledgeable employees are essential for developing and implementing environmentally friendly technologies that minimize environmental impact and ensure compliance with environmental regulations. Additionally, by launching innovative products and services that benefit society, companies can further increase their contributions to the social good. However, our analysis revealed that, while digital transformation significantly improves environmental and social responsibility metrics, its impact on governance dimensions is less pronounced. One potential reason for the limited impact on corporate governance in our study is that digital transformation in businesses tends to bring about operational advantages more quickly, while improvements in corporate governance may take longer to become apparent. Governance includes various factors such as board composition, transparency, accountability, and risk management practices, which may require more time to be influenced by digital transformation efforts.

This study highlights that environmental regulation moderates the positive impact of corporate digital transformation on the quantity of high-quality labor and the skill levels of the existing labor force, thereby influencing firm ESG performance. This finding is consistent with Li et al. [52], highlighting the significant role of both formal and informal environmental regulations in this context, with informal regulations exhibiting a stronger influence. Informal regulations, encompassing social norms and industry standards, complement formal regulatory frameworks, compelling organizations to integrate ESG considerations into their operations. Environmental regulation incentivizes firms to adopt digital transformation solutions that enhance resource efficiency and environmental performance. Additionally, it promotes talent development initiatives to fulfill regulatory requirements and address sustainability challenges, thereby augmenting digital transformation’s positive impact on labor quality and productivity.

The study concludes that digital transformation significantly enhances ESG performance through improvements in both the quantity of high-quality labor and labor skill levels. Importantly, these enhancements are more pronounced in non-state-owned enterprises. Our findings suggest that digital transformation initiatives within non-state-owned enterprises play a pivotal role in augmenting ESG performance metrics, particularly in terms of labor quality and skill development. By employing digital technologies and innovation, non-state-owned enterprises are better positioned to attract and retain high-quality talent, thereby fostering a more skilled and productive workforce. This, in turn, contributes to improved operational efficiency, innovation capacity, and overall ESG performance outcomes. The differential impact of digital transformation on ESG performance across ownership structures underscores the importance of organizational context in shaping the outcomes of digital transformation initiatives. Non-state-owned enterprises—characterized by greater flexibility, agility, and market orientation—are better equipped to capitalize on the transformative potential of digital transformation and drive sustainable value creation across environmental, social, and governance dimensions.

From the perspective of industry and industrial attributes, the impact of digital transformation varies. The manufacturing industry, characterized by heavy assets and high carbon footprints due to its reliance on high inputs of energy and resources, faces significant historical burdens and institutional pressures. These factors make digital transformation and ESG practices more challenging yet necessary for manufacturing firms. In contrast, the impact of enterprise digital transformation on non-manufacturing industries appears more significant. Non-manufacturing sectors may be more agile and capable of integrating digital technologies into their business models, thereby achieving better ESG outcomes more quickly.

6. Conclusions

This study contributed to the scholarly discourse on the relationship between digital transformation and ESG performance, focusing specifically on the role of human capital. Our findings affirm the prevailing notion that digital transformation initiatives generally have positive outcomes for ESG performance, though the impact’s magnitude varies across dimensions. Additionally, this study identified human capital as the primary mechanism through which digital transformation fosters ESG performance enhancement. By augmenting the quality of labor within organizations and elevating the skill levels of the existing workforce, digital transformation initiatives contribute to improved environmental stewardship, social responsibility, and operational governance. This study underscored environmental regulation’s moderating role in shaping the impact of digital transformation on labor quality and skill development, thereby influencing ESG performance at the firm level.

To fully utilize the potential of digital transformation for enhancing enterprise performance across ESG facets, enterprises must adopt a comprehensive and integrated suite of management strategies. Foremost among these is the reinforcement of employee skill development. Considering the rapid pace of technological advancement, continuous investment in employee education and training is essential for ensuring that their skills evolve with digital transformation. Cultivating a culture of lifelong learning empowers the workforce to effectively utilize new technologies and accelerates the achievement of ESG goals. Optimizing human resource management and motivational mechanisms is equally crucial. Enterprises should integrate ESG performance metrics into employees’ evaluation and reward systems to instill enthusiasm and accountability toward ESG objectives. Recognizing and rewarding staff members who champion ESG agendas further reinforces the organization’s commitment to sustainability. Additionally, enhancing data governance standards and transparency is vital for fostering stakeholder trust. Enterprises must ensure precise, reliable, and transparent data management throughout the digital transformation process, which enables stakeholders to clearly grasp advancements and challenges in ESG domains. Open and clear communication strengthens relationships with stakeholders and enhances the brand’s reputation. Furthermore, integrating ESG goals into strategic initiatives is pivotal for synergizing digital transformation and ESG performance. Enterprises should align ESG objectives with overarching strategic ambitions and foster cross-departmental collaboration to ensure concerted efforts to advance the ESG agenda. Implementing effective monitoring and feedback mechanisms enables the continuous tracking and assessment of digital transformation’s impact on ESG performance, facilitating necessary adjustments and improvements.

For policymakers and regulatory bodies, acknowledging the pivotal role of environmental regulations in incentivizing sustainable practices in digitally transformed enterprises is crucial. Stringent environmental standards enforced by governments can stimulate enterprises to prioritize sustainability alongside digital innovation, thus contributing to broader societal and environmental objectives. First, governments should incessantly refine the macro-policy framework for enterprise digital transformation, which can entail increasing specialized funding support, offering credit incentives, facilitating financing, and implementing tax benefits to encourage digital transformation initiatives. Second, central governments should focus on strengthening environmental regulations across regions, which includes intensifying environmental regulatory measures to increase environmental risks and regulatory costs for enterprises. Through environmental regulations, governments can guide and supervise the direction, focus, and scale of ESG activities conducted by enterprises. This approach aims to deepen environmental protection concepts within enterprises, encourage proactive social responsibility, enhance sustainable development governance mechanisms, and comprehensively improve enterprise ESG performance.

Our study has limitations. We only included Chinese A-share listed companies in our sample, which limited our findings’ generalizability. Additionally, we focused solely on the role of human capital in the relationship between digital transformation and ESG performance, neglecting other factors. Future studies should diversify samples and consider multiple perspectives to further clarify this relationship. Furthermore, our study’s cross-sectional nature prevented us from establishing causality or long-term effects. Future research should employ longitudinal studies to track digital transformation’s impact on ESG performance over time.