Effect of Digital Transformation in Sports Companies on Green Innovation: Evidence from Listed Companies in China

Abstract

In the context of the “dual carbon” strategic goal and sustainable development, the digital transformation of sports companies has emerged as a crucial factor in overcoming barriers to green growth and addressing institutional and efficiency challenges. This study examines the mechanism by which digital transformation drives green innovation, using a sample of Chinese-listed companies in the sports industry from 2011 to 2022. Fixed effects models were employed. The study’s findings are as follows: (1) Digital transformation has a significant positive impact on green innovation, indicating that the digitalization of companies plays a crucial role in promoting green practices. (2) A mechanism analysis revealed that digital transformation facilitates green innovation by enhancing human capital and improving internal control levels. (3) A heterogeneity analysis demonstrated that stricter environmental regulations strengthen the driving effect of digital transformation. Moreover, state-owned sports companies exhibit a stronger endogenous impetus for green innovation than non-state-owned companies, driven by their alignment with national strategic planning, thus enhancing the role of digital transformation. This study contributes to the literature by offering insights into the integration of digitization and green innovation. Furthermore, it provides practical guidance and path selection for achieving coordinated digitization and green innovation in Chinese sports companies within the framework of the “dual carbon” goal.

1. Introduction

To achieve harmonious coexistence between humans and nature, green transformation and low-carbon development are essential. In 2020, China’s dual carbon goals were declared at the United Nations General Assembly, to peak carbon emissions before 2030 and achieve carbon neutrality by 2060. Comprehensive green innovation is a vital strategy for harmonizing pollution reduction and carbon reduction efforts, as well as for advancing green and low-carbon economic and social development. The sports industry plays a significant role in economic and social development, and green innovation in this sector is a realistic requirement for high-quality development. Particularly in the context of global climate change and the pursuit of the dual carbon goal, green innovation in the sports industry is fundamental to advancing ecological civilization construction. According to data from 2022, the national sports industry had a total scale (output) of CNY 3.3 trillion in China, with an added value of CNY 1.31 trillion, accounting for 1.07% of the GDP. The number of sports venues nationwide reached 4.2268 million, covering a total area of 3.702 billion square meters, resulting in a per capita sports venue area of 2.62 square meters. By the end of 2021, there were 452,000 legal entities in the national sports industry, employing to over 7.186 million people [1]. These statistics demonstrate the significant role of the sports industry in driving economic growth, improving people’s well-being, and generating employment opportunities. However, issues such as the extensive development of the sports industry are often criticized by scholars [2,3]. The sports equipment manufacturing sector has traditionally been dominated by low-end production, such as athletic shoes and apparel, which are closely associated with high-carbon emission industries like rubber, textiles, and chemicals. In the sports service sector, large sports venues produce substantial carbon emissions during operations to meet competition requirements, while the carbon footprints of sports tourism and active sports enthusiasts have also drawn attention from international researchers [4,5]. The sports industry has been incorporated into China’s low-carbon development action plan, making its green innovation imperative [6]. However, green innovation activities in companies are typically characterized by long cycles, high investments, high risks, and strong financial constraints, which are major obstacles of green development. Additionally, the vulnerability of supply chains and the rapid evolution of information technology add to the uncertainty surrounding green innovation in companies. Therefore, a scientific exploration of how to enhance green innovation capabilities is essential for overcoming the “bottlenecks” in green development and further promoting sustainable development.

Digital technology serves as a new engine for the green innovation of sports companies, characterized by its ubiquity, openness, fluidity, and inclusiveness. It effectively addresses the management and technological challenges that sports companies encounter in their pursuit of green innovation [2,7]. Scholars have consistently focused on the relationship between digitization and greening. Firstly, some studies concentrate on the sports industry, examining the impact of new digital infrastructure and the digital economy on green development [8,9]. Research has found that the digital economy fosters the sustainable development of the sports sector. Furthermore, some studies indicate that this positive relationship is not linear; instead, the growth rate tends to decline over time [9]. Other research targets specific areas within the sports industry, exploring the applications of technologies such as edge computing, blockchain, artificial intelligence, and digital twins in fields like health and fitness [7,10], sporting events [11], and sports architecture [12], highlighting their positive contributions to low-carbon and sustainable development [12,13]. Secondly, research focusing on the manufacturing sector or companies has yielded substantial research. These studies examine the driving effects of big data, digital technology, and digital transformation on green development, green innovation, and sustainability from both macro and micro perspectives. In addition to digital transformation, some studies explore other antecedent variables influencing green innovation. Macro-level studies have examined digital government transformation [14], digital economy policy [15], industrial intelligence [16,17], environmental regulations [18,19], and technological innovation [20], while micro-level studies have examined artificial intelligence [21], digital finance [17,22,23], digital infrastructure [24], human resources management [25], and capital investment [18,26]. Despite the widespread usage of such technologies in the sector, there is a shortage of academic studies about the efficacy of scenarios like these in the sports industry [13].

Indeed, there are a limited number of studies focusing on the micro-level perspective of how digital transformation specifically affects green innovation or green transformation. However, there are some viewpoints and arguments in this area. One perspective suggests that digitalization through transforming resource integration and allocation, assisting decision making, reducing contract costs, and innovating transaction costs, can drive green transformation or innovation in companies [27]. Liu et al. [28] argue that digital transformation in companies contributes to green innovation by increasing innovation resource inputs and reducing debt costs, thereby facilitating “source reduction” and “end-cleaning” approaches. Meng et al. [29] emphasize that the embeddedness of digital technologies provides a more robust power system, greener production methods, more efficient operational efficiency, and more intelligent governance models, all of which have a positive impact on green innovation. Some studies indicate that the development of new elements (technologies, elements, platforms) in digital transformation enables the restructuring of business logic, changes in organizational form, and innovation in value models, effectively promoting efficiency improvement and structural upgrading in companies [30,31]. However, there is an opposing viewpoint suggesting that digital transformation may not entirely drive green innovation in companies. Excessive investment in digital technologies can bring high costs and is also a source of energy consumption, which can hurt green innovation. Li [32] found that digital transformation has an inhibitory effect on the environmental dimension of sustainable performance in companies.

In summary, the identified research limitations can be summarized into three aspects: (1) Limitations in research perspectives: Previous research has predominantly focused on the impact of digital technology on green innovation or green transformation in the manufacturing industry or companies. Digital technology has been widely used in the sports industry, and it has also received widespread attention for low-carbon development and green development, but the academic research is insufficient. There is a lack of research and an ongoing controversy regarding the specific impact of digital transformation on green innovation within the sports industry at the micro level. (2) Limitations in mechanism research: While some scholars have explored the impact of digital technologies on green innovation, there is a scarcity of research that delves into a comprehensive analysis of the mechanisms through which digital transformation influences green innovation within sports companies, particularly from the perspectives of “human capital” and “corporate governance”. These dimensions play a crucial role in understanding the underlying mechanisms driving green innovation. (3) Limitations in contextual analysis: Few scholars have examined whether different sources of digital technologies have varying effects on green innovation in companies. This lack of investigation into contextual factors and their influence on the relationship between digital transformation and green innovation has led to one-sided research conclusions. It is crucial to consider the influence of company property rights and environmental regulations on the relationship between digital transformation and green innovation to gain a more nuanced understanding of this dynamic.

To address these limitations, this study focuses on two key questions: First, does digital transformation effectively promote green innovation in sports companies? What are the mechanisms through which digital transformation promotes green innovation in sports companies? Second, how does the relationship between digital transformation and green innovation vary based on differences in company property rights and environmental regulations? Addressing these research questions will offer valuable support to governments and companies in implementing effective strategies for green development within the sports industry.

In light of the aforementioned considerations, the potential contributions are as follows: First, this study focuses on the impact of digital transformation on the green innovation of sports companies at the micro level. It deepens the comprehension of the consequences and controversies surrounding the impact of digital technologies, and supplements research on the interaction between digitalization and green development. By focusing on the green innovation of sports companies, this study further extends the exploration of the microeconomic implications of the digital economy. Second, this study presents a comprehensive research framework comprised of “benchmark analysis–mechanism analysis–heterogeneity analysis”. This framework effectively unveils the intricate “black box” of mechanisms underpinning the impact of digital transformation on green innovation. It captures the channels and heterogeneous effects of digital transformation driving green innovation, aiming to provide beneficial guidance for the green innovation practices in sports companies. Third, in the context of digitalization and low-carbon development, this study examines the issue of the synergy between digitalization and green innovation, thereby extending the application scope and theoretical scenarios of the theory on digital empowerment and sustainable development.

The paper is structured as follows: Section 2 discusses the theoretical background and outlines the research hypotheses; Section 3 details the research design; Section 4 presents the empirical results and provides an analysis; Section 5 concludes the research and describes practical implications; and Section 6 provides the research limitations and future research directions.

2. Theoretical Background and Hypothesis Development

2.1. Green Innovation of the Sports Industry

A substantial body of research defines green innovation as the adoption of environmentally friendly materials and practices that reduce negative externalities throughout the product lifecycle. This encompasses the implementation of green technologies and processes focused on conserving energy, minimizing waste, and decreasing pollution, all in the pursuit of sustainability [21]. Green innovation aims to shift from an extensive development model characterized by high pollution, high emissions, and environmental degradation to a green development model characterized by low resource consumption, low emissions, and a harmonious relationship between humans and nature, thereby promoting sustainable development [18,33]. Chen et al. [34] highlighted that green innovation encompasses both hardware and software innovation related to green products or processes, including technological innovation activities in green product design, pollution prevention and control, waste recycling, or corporate environmental management, etc. According to Dian et al. [35], green innovation refers to the enhancement of product design through the use of environmentally friendly materials or processes, aimed at minimizing the stages of the product life cycle and reducing negative impacts on the environment, ultimately contributing to sustainable development. As a pillar industry for future economic development, the sports industry is increasingly contributing to the service sector each year [9]. However, issues such as extensive development, a small scale, and low structural efficiency persist. Challenges like “having various business forms without a cohesive system”, “possessing supply chains that are not smooth”, “having elements that are not coordinated”, and “high investment with low efficiency” are particularly prominent [2,8]. Building upon the existing literature, green innovation in the sports industry refers to the development of environmentally friendly products from the supply side that include the use of green raw materials and ecological product design to reduce material usage, pollutant emissions, and the consumption of energy resources, such as water and electricity. By evaluating the extent of responsiveness to consumers’ green demands during product development and supporting cross-functional collaboration to drive green innovation in production processes and management, companies aim to enhance the flexibility and efficiency of executing their green strategies, ultimately bolstering and sustaining green performance [36,37]. This definition places significant emphasis on the greenization of the production process and promotes the shift from traditional production models to sustainable development models, ultimately leading to improvements in both economic and environmental outcomes.

Currently, research on green development in the sports industry is just emerging, with existing studies primarily consisting of qualitative descriptions. There has been some exploration of the driving mechanisms behind green development, green technological innovation, and green, low-carbon strategies in the sports industry. Additionally, a limited number of discussions have focused on antecedent variables such as digital technologies and new infrastructure. However, empirical research at the regional, industrial, and corporate levels remains relatively sparse. Research on green innovation in other industries primarily focuses on high-pollution companies, resource-intensive companies, and manufacturing firms. These studies provide a comprehensive analysis of the concepts, measurements, mechanisms, models, and pathways related to green innovation in companies. Within the realm of green innovation in companies, micro-level research primarily focuses on two key aspects: Firstly, scholars examine the driving factors behind green innovation in companies. One perspective centers on external environmental factors, such as environmental regulation [18], environmental taxes [38], and green financing [39]. Another perspective delves into internal driving factors, including green thinking [40] and technological innovation [20]. Secondly, scholars investigate the models and pathways of companies’ green innovation. Research in this area explores the green innovation models and value-added pathways of companies with high pollution levels, energy-intensive companies, and manufacturing companies, from the perspectives of sustainable development, green strategic alliances, and dynamic management [41,42,43]. These studies provide crucial theoretical and empirical foundations for a deeper comprehension of green innovation in companies.

Unlike traditional manufacturing and resource-intensive companies, sports companies emphasize user participation and interaction [44]. Digital technologies transform various aspects of sports, integrating them into the very fabric of athletic activities and promoting the spirit of sportsmanship. Digital transformation redefines fan engagement and user experience within sports companies, creating closer connections between fans, players, and teams [45]. Several important questions still require further exploration. Firstly, while leveraging digital technologies to improve production efficiency, companies must also consider the potential negative environmental impacts stemming from their production modes. It is worth noting that the digitalization process itself can be energy-intensive and contribute to environmental consequences [32]. What is the relationship between the digital transformation of sports companies and green innovation? Given that the digitalization of sports companies differs from that of traditional manufacturing companies and is continuously evolving, can digitalization in sports companies lead them toward a greener future? These include investigating the impact of policies and institutional environments on green innovation and understanding the characteristics of green innovation across different industries and company sizes. Future research endeavors can delve into these areas to advance both the practical implementation and the theoretical development of companies’ green innovation.

2.2. Digital Transformation and Green Innovation

In general, the results of corporate green innovation are manifested through aspects such as green product development, process optimization, and business model innovation [46,47] from both the supply side and demand side. Digitization transformation, driven by automation, informatization, and digitization, empowers companies to reconfigure their resources, channels, and institutional frameworks, shaping resource allocation and decision making regarding innovation, and subsequently transforming the market’s ecological environment. This study aims to analyze the influence of corporate digital transformation on green innovation.

Firstly, from the supply side, digital transformation supports green product development, driving green innovation in companies. During the research and development phase, digital technologies such as data collection and analysis, automation, standardization, and decision support systems enable management to access high-quality information in real time, including customer demands, technological advancements, and opportunities for green innovation [48], thereby alleviating management short-sightedness [49]. For example, Wei [50] emphasizes that digital management platforms can collect, analyze, and generate data in real time, aiding management in making informed decisions and enhancing the efficiency of sports equipment research and development as well as smart manufacturing processes. From the developers’ perspective, digital transformation has proven to be an effective means of reducing the ambiguity surrounding green innovation activities and improving the innovation efficiency of R&D teams. For instance, Dev [51] discovered that augmented reality and computer-aided design can model and simulate throughout the lifecycle of green products. In the sports sector, these technologies provide an intuitive and interactive design environment, enabling real-time feedback, energy efficiency assessments, intelligent optimization, and virtual physical coexistence, which enhance R&D flexibility and success rates [52]. During the production phase, data analysis tools comprehensively track information and tasks throughout the production process, monitoring production efficiency and resource usage to stimulate reforms in production processes and governance models. By deeply integrating digital technologies into processes such as material procurement and manufacturing, brand operation and promotion, product wholesale and retail, and after-sales service, traditional production organization forms are transformed. This integration continuously optimizes product distribution channels, shortens production cycles, and extends product lifecycles, achieving a fully digital and green development process. A noteworthy example is the research by Chen et al. [53], which suggests that companies can establish platforms such as digital emissions monitoring and early-warning networks. These platforms accurately track pollution sources during production, focusing on carbon capture, utilization, and storage to effectively reduce carbon emissions [54]. For instance, digital governance leverages sensors and management platforms to collect ecological chain data from the sports equipment manufacturing industry, such as waste emissions, water usage, and energy consumption. Through a controllable ecological governance platform, the entire process—from raw material procurement to production and eventual disposal-can be traced, facilitating green development in the sports equipment manufacturing sector.

Secondly, from the demand side, digital transformation brings new changes to the marketing and service models of sports companies, driving their green innovation. On one hand, digital transformation enables companies to understand and respond to the consumer demand for green products, guiding their green innovation. By collecting data on consumers’ browsing preferences, behavior patterns, and transactions related to green products, and using digital technologies for social sentiment analyses and natural language processing, companies can identify green demands and develop targeted marketing and service strategies to increase the likelihood of consumers purchasing green products [55]. For instance, during the selection of green materials, technologies such as data analysis and digital simulation can be employed to gather and analyze extensive data—such as the performance, comfort, and durability of fitness enthusiasts [7,56]. This helps predict which materials or combinations of materials will provide superior performance while balancing cost and user experience. As emphasized by Magaz-González et al. [55], digital technology enables companies to perceive and acquire various resources from consumers, allowing them to align accurately with changing preferences. On the other hand, digital channels enhance product accessibility for users. Consumers can easily obtain green products through various digital channels, lowering the barriers to purchase. These channels provide consumers with comprehensive product information, including the product’s ingredients, production processes, and environmental certifications. This transparency enhances consumer trust in green products, thereby promoting purchasing decisions [57]. Additionally, the sports industry is highly integrative, often merging with media and sporting events [57]. Companies can engage directly with consumers through digital media, promoting green products and sustainable practices during sporting events, which enhances the consumer awareness of green consumption [44]. Hypothesis 1 is put forward.

Hypothesis 1.

Digital transformation has a positive effect on green innovation in sports companies.

2.3. Mediating Roles

2.3.1. The Mediating Role of Human Capital

Experience from previous technological revolutions has demonstrated that technological progress not only has a significant impact on productivity but also brings about changes in the labor market. The widespread adoption of information technologies, such as big data, artificial intelligence, the Internet of Things, and cloud computing, has led to substantial transformations in employment and labor force structures [58]. In the context of the sports industry, intelligent development has resulted in the substitution of low-skilled labor with advanced machinery and equipment, leading to an increase in the proportion of highly skilled labor [59]. Simultaneously, as digital technologies become increasingly integrated into company production and operations, businesses expand in scale, and production techniques and machinery undergo gradual upgrades [60,61]. Consequently, there is a growing demand for high-level labor that can effectively work with these advancements [62]. This trend aligns with the “capital-skill complementarity” hypothesis, which suggests that capital and skills are complementary. Therefore, as companies undergo digital technology upgrades, there will be an increase in the utilization of highly skilled labor, promoting the enhancement of human capital within the organization [63,64].

Furthermore, the upgrading of human capital is crucial for the green innovation of listed companies in the sports industry. Firstly, the level of human capital significantly influences a companies’ environmental awareness. As human capital is enhanced, companies become more cautious in their use of non-renewable energy and pay closer attention to environmentally harmful activities, thereby promoting green practices. These practices may involve conducting research and development to create new products or processes that focus on recycling and pollution reduction [65,66]. According to imprinting theory, an individual’s knowledge and education influence their actions when dealing with environmental issues in the future. Secondly, the digital literacy and operational skills of talented individuals serve as a driving force for undertaking technological innovation and accelerating the green innovation process. Endogenous growth theory emphasizes that human capital can act as a catalyst for economic growth as it serves as a fundamental input in the production process [67]. With the improvement of human capital, high-quality human and knowledge capital will be integrated into the production and operation processes, leading to a direct diffusion effect of technology [68]. More specialized sports equipment manufacturing technicians and R&D personnel are entering emerging digital fields, contributing to the accumulation of clean production and pollution control technologies within companies. This provides the human resources necessary for companies to achieve energy savings, reduce emissions, and develop a green economy. Ahmed et al. [69] focus on the role of human capital in controlling nature’s ecological footprint and argue that it contributes to the efficient extraction and utilization of natural resources. Thirdly, companies with high-level human capital can accurately identify consumers’ green preferences, understand their needs, and develop green products that meet the market demand. For example, Li Ning has attracted high-quality talent from fields such as environmental protection and materials science to drive the research and development of green products and their market promotion. Additionally, they have established specialized data analysis teams to assess the market performance of green products and adjust marketing strategies promptly. Building upon these insights, it can be speculated that digital transformation achieves green innovation by facilitating the upgrading of human capital. Hypothesis 2 is put forward.

Hypothesis 2.

Digital transformation promotes green innovation through the upgrading of human capital.

2.3.2. The Mediating Role of Internal Control

Digital transformation provides technological support for updating and iterating internal control systems. The primary purpose of internal control is to mitigate agency problems and enhance information exchange within an organization. Research suggests that integrating digital technology with internal control is an effective approach to significantly improve the quality of internal control. On one hand, digital transformation excels in enhancing information transparency and quality by incorporating precise and intelligent digital modules within the internal control framework of companies [59]. Specifically, digitization integrates all the workspaces of a company into a single information network, consolidating the collection, analysis, and access of management-related information within a unified software environment. Visualized data workflows, IT governance, and data mining greatly enhance the transparency of business decision making [70]. This integration helps reduce information asymmetry and conflicts of interest, ensuring compliance with company operations and optimizing the internal governance environment [70,71]. Studies have demonstrated that digital transformation enhances the effectiveness of internal controls by alleviating information asymmetry and reducing principal-agent costs [72]. On the other hand, digital technology significantly reduces the costs associated with supervision. By incorporating digital technology into risk assessments, control activities, and internal monitoring processes, it supports digital calculations, risk model predictions, control testing, and more, effectively reducing the supervision costs of companies [73]. For instance, the tamper-resistant and transparent nature of blockchain technology can enhance the existing regulatory mechanisms by providing more accurate and reliable financial transaction records and other relevant data [74]. In summary, digital transformation facilitates the updating and improvement of internal control systems by enhancing information transparency, improving information quality, and reducing supervision costs through the integration of digital technologies.

Limited research is available specifically on the impact of internal control on green innovation. However, scholars have extensively discussed the relationship between internal control and green innovation. On one hand, high-quality internal control can effectively standardize and supervise managerial behavior, balance managerial power, and prevent inefficient investments [75]. This is mainly because internal control effectively restrains management’s earnings manipulation and opportunistic behavior, pushing them to follow scientific decision-making processes and avoid short-sightedness and randomness in major strategic decisions. Empirical research conducted by scholars has shown that internal control significantly reduces the principal-agent problems arising from information asymmetry and insufficient oversight within sports companies, curbs inefficient investments, and strengthens information transmission among organizational levels, thereby facilitating the implementation of green innovation activities [76,77]. On the other hand, green innovation relies on a sustainable and stable support mechanism for investment. The allocation of critical resources such as funds, human resources, and technology in innovation activities is often determined by internal institutional arrangements within a company. Research conducted by Luo [78] highlights that significant flaws in internal control can lead to the misuse or misappropriation of innovation funds, resulting in strong financing constraints and other hazards that hinder increased investment in innovation. These findings indirectly demonstrate the positive impact of high-quality internal controls. In summary, strengthened internal control ensure the establishment of an effective investment mechanism by enhancing the transparency of investment decisions, improving investment compliance, and optimizing capital allocation [79]. Additionally, internal control plays a vital role in stabilizing the internal control environment and reducing the uncertainty associated with green innovation by assessing, controlling, and preventing system risks. Based on these observations, it can be speculated that digital transformation, by enhancing the level of internal control within companies, can establish a solid internal foundation for the smooth advancement of green innovation in companies.

Based on these findings, it can be speculated that digital transformation, by improving a companies’ level of internal control, contributes to the smooth progress of its green innovation, thus establishing a solid internal foundation. Therefore, Hypothesis 3 is proposed:

Hypothesis 3.

Digital transformation promotes green innovation by enhancing their level of internal control.

2.4. Moderating Roles

2.4.1. The Moderating Role of Environmental Regulation

Green innovation emphasizes the economic behavior of environmental performance, characterized by high risks, substantial investments, and dual externalities. These factors contribute to insufficient investment motivation among companies, making it difficult to effectively drive change through technological advancement and market forces alone. The “Porter Hypothesis” posits that appropriate environmental regulations can compel companies to innovate technologically, facilitating their green innovation and generating “compensatory benefits” that exceed the costs of compliance [80]. By applying digital technologies, sports companies can integrate green outcomes into their production processes, reducing dependence on traditional, polluting production methods and effectively mitigating environmental regulatory costs [19,81]. The “forcing” effect of environmental regulations manifests through extrinsic pressures from stakeholders and intrinsic motivation within companies.

On one hand, regarding extrinsic pressures, the application of digital technologies addresses the tangible demands from stakeholders for sustainable development in the sports sector. The implementation of the Environmental Protection Tax Law in China in 2018 stipulates that companies are subject to pollution discharge fees based on pollution equivalents. Henriques et al. [82] found that pressures from stakeholders compel managers to consider the consequences of environmental pollution, driving them to proactively respond to regulatory requirements. The sports industry involves a diverse array of stakeholders, including sports equipment manufacturers, sponsors, advertisers, agencies, media, sports associations, and fans, et al. [13,56]. The needs and expectations of these stakeholders are often complex and diverse, adding to the challenges of environmental regulation. The sports industry also particularly emphasizes participation and emotional engagement. Xu et al. [83] discovered that investors assign lower valuations to firms penalized for environmental issues, while those committed to green development receive higher valuations. Sports can evoke strong emotional resonance, making consumer loyalty and emotional investment crucial for the industry’s development. Consequently, the application of digital technologies in fostering green development can enhance stakeholders’ confidence in a companies’ sustainability efforts [84]. This, in turn, elevates both the product and the customer value, prompting managers to adopt green innovation strategies in response to stakeholders’ demands.

On the other hand, regarding internal incentives, although the green innovation of sports companies pays attention to the social benefits of enterprises, economic benefits are still its primary goal. Environmental regulations, such as pollution discharge fees and carbon taxes, reduce directly realizable profits [85]. However, they also encourage managers to critically evaluate the weaknesses of green development and identify feasible improvement measures [86]. The digital transformation can not only yield social benefits through energy conservation and emission reduction but also provide robust technical support for the efficient production of green products, fostering a distinctive competitive advantage in sustainability [19,87]. Currently, the application of new technologies such as digital technology, artificial intelligence, and virtual reality in the sports industry is becoming increasingly widespread, encompassing sports events, manufacturing, and venue operations [7,9]. These technologies not only redefine the performance of sports products but also enhance the user experience, achieving a balance between environmental benefits and economic gains. Given this context, the costs associated with environmental regulations incentivize sports companies to actively pursue digital transformation, thereby facilitating their green innovation. Based on this rationale, this study proposes the following research hypothesis:

Hypothesis 4.

Compared to those with weak environmental regulations, the digital transformation of companies in regions with strong environmental regulations has a stronger driving effect on green innovation.

2.4.2. The Moderating Role of the Nature of Property Rights

The nature of property rights is a significant, inherent characteristic of companies, resulting to their classification into state-owned and non-state-owned companies. The differing property rights in sports companies lead to variations across multiple dimensions, including policy frameworks, resource allocation, and operational objectives [88], all of which can influence the implementation of digital transformation.

Firstly, state-owned companies typically receive more government resources and policy support than non-state-owned companies, providing them with essential funding and technical backing for their digital transformation efforts [89,90]. This disparity arises from the Chinese government’s continued significant role in resource allocation. For instance, non-state-owned sports companies often encounter greater challenges in securing bank credit compared to their state-owned counterparts, resulting in more pressing financing constraints [91]. This situation primarily stems from information asymmetry between companies and banks, leading to issues of reluctance and caution in lending to non-state-owned companies [92]. Conversely, state-owned companies tend to receive more credit support during banks’ lending decisions due to their backing from local government credit. In addition, state-owned sports companies have greater opportunities to access advanced technologies and information resources [90]. They typically can apply for government special funds for scientific and technological research to support the development of smart equipment and sports gear. Some state-owned sports enterprises can also prioritize access to advanced sports technology and equipment through government collaboration projects, serving as models for green practices [2,9]. These advantages make state-owned sports enterprises more competitive in promoting digital and green transformation.

Moreover, in terms of operational objectives, state-owned sports companies are significantly influenced by policy considerations. In addition to emphasizing economic efficiency and growth, they also bear social responsibilities, such as environmental protection and job creation [89,93]. As a result, these companies are inclined to pursue green goals that align with national strategies during their digital transformation [93]. Furthermore, state-owned companies foster a management culture that prioritizes social responsibility, viewing digital transformation not only as a means of enhancing efficiency but also as a crucial pathway to achieving green innovation [94]. In contrast, non-state-owned sports companies prioritize economic benefits and focus primarily on cost-effectiveness, leading to slower progress in digital transformation. Green innovation may be perceived as an additional value rather than a core goal [91]. Even when they do pursue digital transformation, their objectives often center on increasing market share and competitiveness, while their attention to long-term, high-investment green innovation may be driven by market demand and consumer preferences [94].

Hypothesis 5.

Compared to non-state-owned companies, the digital transformation of state-owned companies has a stronger driving effect on green innovation.

The theoretical analysis and logical framework of this research are shown in Figure 1.

3. Research Design

3.1. Sample and Data

According to the industry classification standards published by the China Securities Regulatory Commission, nearly 50 publicly listed companies involved in sports-related businesses were identified. Their business scope includes sports equipment, event management, sports venues, sports advertising, and football clubs. Considering the performance volatility of companies in their early stages of listing, only those listed for more than two years were selected to ensure the reliability of the data analysis. This study examines the period from 2011 to 2022, excluding ST companies during this time frame and eliminating those publicly listed companies with significant missing financial data. Ultimately, 32 companies from the sports industry were included in the final analysis.

The original sample consists of A-share listed companies in the Shanghai and Shenzhen stock exchanges from 2011 to 2022, including multiple data sources. (1) Existing research shows significant differences in the measurement indicators for green innovation. Drawing on the research by Chang [95] and Han et al. [96], the study uses the number of green patents (invention patents and utility model patents) to measure the level of green innovation. The green innovation data was manually collected by referring to green patent data published on the official website of the State Intellectual Property Office (SIPO). (2) Data on digital transformation was obtained from the annual reports of the target companies. (3) Internal control data were sourced from the DiBo database, utilizing the internal control index as a measurement indicator. (4) Other key data such as human capital and control variables were collected from the CSMAR database. Environmental regulation data were obtained from the “China Statistical Yearbook”, “China Environmental Statistical Yearbook”, and “China Urban Statistical Yearbook”. To ensure the data’s validity, the original data underwent the following processing steps: (1) The exclusion of ST and *ST companies during the sample period. (2) Control for the influence of extreme values by winsorizing the variables at the 1% and 99% levels, reducing the interference of outliers. (3) The elimination of samples with seriously abnormal observation values and companies with less than two years listed. After the aforementioned screening process, a total of 384 sets of sample data were obtained, and the data analysis was conducted by Stata 17.0 software.

3.2. Variable Definitions

3.2.1. Dependent Variable: Green Innovation (lnGI)

The three primary methods for measuring green innovation indicators are as follows: (1) A seven-item Likert scale method, which constructs a questionnaire for green innovation and which was gauged by the collected data [21]. (2) The Data Envelopment Analysis (DEA) method, which constructs green total factor productivity as a measure of green innovation [97]. (3) The alternative indicator method, which utilizes substitute indicators, such as green patents and pollution emissions, to quantify green innovation [14,91,96]. It is a common practice to use the number of corporate green patent applications as a standard to measure the results of corporate green innovation [98]. Given the data availability, this research adopted the alternative indicator method. Green patents directly reflect the output capacity of green innovation, and authorized patents are more reliable and representative than patent applications. Consequently, the total authorized number of green invention patents and green utility model patents (lnGI), or the number of green patent authorizations, served as a measurement indicator of corporate green innovation. Additionally, the number of green patent applications (lnGIA) was used as an alternative measure for evaluating green innovation (lnGIA). To collect the required data, the research retrieved the relevant page from the State Intellectual Property Office, employed the green patent IPC classification number and the name of the sports industry listed company in the “International Patent Classification Green List” released by the World Intellectual Property Organization (WIPO) in 2010 as keywords, and manually collected and organized green patent applications and patent authorization entries.

3.2.2. Independent Variable: Digital Transformation (lnDT)

Digital transformation is a complex and systematic project that involves changes in organizational structures, internal management, and business processes [99]. However, existing financial indicators are insufficient for comprehensive measurement in this context. Some scholars have adopted the proportion of intangible assets related to digital transformation as a measure of its extent [100]. Another commonly used method is the analysis of the frequency of terms related to digital transformation in annual reports [101]. Drawing on the research by Wu et al. [101] and Chen and Srinivasan [102], this study utilized text-mining methods to construct the indicators of digital transformation. Annual reports, which serve as a summary of a companies’ operations, provided insights into the degree of digital transformation. Word frequencies were obtained from the CSMAR database. Specifically, Python was used to conduct text mining on the annual reports of listed companies in the sports industry, and the digital transformation level of each company was measured based on the frequency of digital keywords in the annual reports. In general, domestic and foreign scholars define digital transformation as the digitization of production materials and processes facilitated by digital technology [103]. Building on existing research [102], this study considered artificial intelligence, big data, cloud computing, and blockchain as the four underlying core technologies of digital transformation [104]. These technologies have undergone stages of foundational embedding, pattern upgrading, and practical application in companies, ultimately enabling the construction of complex business ecological scenarios. To establish a keyword database for digital transformation, this study subdivided key terms into the four underlying digital technologies and their applications. Figure 2 details the key characteristic words associated with digital transformation.

To extract keywords from the annual reports of the sample companies, the following process was followed. Firstly, the annual reports of listed companies in the sports industry were collected, and the content of the “management discussion and analysis” section was extracted. In the text preprocessing stage, expressions containing negative words such as “none”, “no”, “not”, and “lack” before keywords were eliminated. Then, using Python software (Vesion 3.9), regular expressions were written to remove numbers, English letters, and punctuation or special characters (i.e., spaces, indents, newlines, etc.) other than periods in TXT format. Secondly, the text was split into sentences using Chinese periods as separators. Python software (Vesion 3.9) was used to invoke the word segmentation tool Jieba for word segmentation and to remove stop words such as modal particles, prepositions, conjunctions, and adverbs. Finally, the frequency of the digital keywords was calculated through manual search, screening, and a statistical analysis. The total word frequency was incremented by 1, and the natural logarithm was applied to obtain the digital transformation index.

3.2.3. Mediating Variables: Human Capital (HC) and Internal Control (IC)

The internal control index of Dibo China listed companies serves as a proxy indicator of internal control [79]. A higher index value indicates a higher level of internal control within the company. In this study, the internal control index was standardized by dividing it by 100. The measurement method for human capital aligned with the existing literature and was based on the educational level of employees. Specifically, it was calculated as the ratio of the number of employees with a bachelor’s degree or above to the total number of employees in a company [105].

3.2.4. Moderating Variables: Nature of Property Rights (NOPR) and Environmental Regulation (ER)

The nature of property rights for companies was represented as a binary variable: it took a value of 1 for state-owned companies and a value of 0 for non-state-owned companies [90]. The level of environmental regulation was measured by the proportion of the amount invested in wastewater and air pollution control in a given year to the industrial output value of that year [18], based on the location of the listed companies.

3.2.5. Control Variables

To reduce the estimation bias caused by omitted variables, variables highly correlated with green innovation were selected as control variables, including listed years, current ratio, leverage ratio, asset turnover ratio, return on equity, equity concentration, and the proportion of independent directors.

Table 1. Variable Definition.

Type	Name	Symbol	Measurement
Dependent variables	Green innovation	lnGI	ln(Number of green patents granted + 1)
		lnGIA	ln(Number of green patent applications + 1)
Independent variables	Digital transformation	lnDT	ln(Total frequency of digital transformation feature words + 1)
Mediating variables	Human capital	HC	Bachelor degree or above/headcount
	Internal control	IC	Internal control index in DIB internal control database/100
Moderating variables	Nature of property right	NOPR	The value is 1 for state-owned companies and 0 for non-state-owned companies.
	Environmental regulation	ER	Waste gas and wastewater pollution treatment investment/Industrial output value
Control variables	Listed years	Age	ln(Listed years + 1)
	Current ratio	CR	Current assets/current liabilities
	Leverage Ratio	Leverage	Total liabilities at year-end/Total Assets
	Asset Turnover Ratio	Turnover	Revenues/Total Assets
	Return on equity	ROE	Year-end net profit/Total Assets
	Equity concentration	EC	The number of shares held by the largest shareholder/the total number of shares
	Independent directors ratio	Director	Number of independent directors/Number of directors

reports the specific definitions of variables.

3.3. Model Specification

To advance empirical testing, the following mathematical model was constructed:

$${\displaystyle \frac{\ln \mathrm{G}{\mathrm{I}}_{\mathrm{i},\mathrm{t}}}{\ln \mathrm{G}\mathrm{I}{\mathrm{A}}_{\mathrm{i},\mathrm{t}}}}={\mathsf{\beta }}_0+{\mathsf{\beta }}_1\ln \mathrm{D}{\mathrm{T}}_{\mathrm{i},\mathrm{t}}+\sum \mathsf{\beta }\mathrm{C}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}+\sum \mathrm{Y}\mathrm{e}\mathrm{a}\mathrm{r}+\sum \mathrm{I}\mathrm{n}\mathrm{d}\mathrm{u}\mathrm{s}\mathrm{t}\mathrm{r}\mathrm{y}+\mathsf{\epsilon }$$

(1)

where lnDT_i,t represents the degree of digital transformation of company i in year t. lnGI_i,t represents the degree of green innovation, and lnGIA_i,t represents the number of green patent applications. Control generally refers to company-level control variables. Year-fixed effects (Year) were included to control for time variation. ε was interpreted as a random distractor. To avoid reverse causality, the independent variables were lagged by one period.

4. Analysis and Results

4.1. Descriptive Statistics

Table 2. Descriptive statistics of variables.

Variables	Obs	Mean	Std. Dev	Min	Max
lnGI	384	0.489	0.771	0.000	3.045
lnGIA	384	0.706	0.977	0.000	4.025
lnDT	384	1.439	1.257	0.000	4.820
HC	384	0.287	0.225	0.000	1.000
IC	384	6.119	1.564	0.000	9.214
NOPR	384	0.466	0.500	0.000	1.000
ER	384	0.002	0.002	0.000	0.014
Age	384	14.880	7.814	0.000	30.000
EP	384	0.002	0.002	0.000	0.014
PRO	384	0.466	0.500	0.000	1.000
CR	384	2.061	1.785	0.175	16.020
Leverage	384	0.447	0.212	0.054	0.895
EC	384	0.330	0.135	0.078	0.659
Director	384	0.374	0.055	0.200	0.600
Turnover	384	0.644	0.424	0.015	2.018
ROE	384	0.021	0.079	−0.554	0.482

reports the statistical characteristics of each variable. Among them, the mean value of lnGI is 0.489, with a standard deviation of 0.771. This suggests that the overall level of green innovation in the sports industry’s listed companies is not high, and the data has a significant degree of dispersion. The average value of lnDT is 1.439, indicating that the digital transformation level of the sample companies is generally low.

4.2. Correlation Analysis and Multicollinearity Test

Table 3. Correlation analysis and multicollinearity test.

Variables	1	2	3	4	5	6	7	8	9	10	11	12	13	14
1. lnGIA	1
2. lnGI	0.736 ***	1
3. lnDT	0.022 *	0.020 *	1
4. HC	−0.080	−0.113 **	0.247 ***	1
5. IC	0.043	0.019	0.066	−0.015	1
6. NOPR	−0.003	−0.056	−0.079	0.163 **	−0.016	1
7. ER	0.102 *	0.156 **	−0.154 **	−0.088	0.020	0.125 *	1
8. Age	0.025	0.062	−0.071	0.232 ***	−0.086 *	0.503 **	0.188 **	1
9. CR	−0.153 ***	−0.146 ***	0.133 ***	0.085 *	0.051	−0.179 **	−0.184 **	−0.458 ***	1
10. Leverage	0.181 ***	0.205 ***	−0.144 ***	0.011	−0.023	0.266 **	0.185 **	0.403 ***	−0.617 ***	1
11. EC	−0.085 *	−0.062	−0.135 ***	−0.089 *	0.100 **	0.055	−0.086	−0.127 **	0.059	−0.106 **	1
12. Director	−0.014	−0.011	0.104 **	0.083	0.054	−0.120 *	−0.073	−0.229 ***	−0.019	−0.047	0.010	1
13. Turnover	−0.152 ***	−0.094 *	−0.076	−0.222 ***	0.007	−0.132 **	0.059	−0.070	−0.150 ***	0.120 **	0.192 ***	0.026	1
14. ROA	−0.003	−0.041	0.097 *	0.021	0.153 ***	0.017	0.029	−0.155 ***	0.129 **	−0.260 ***	0.173 ***	0.011	0.105 **	1
VIF	—	1.090	1.140	1.290	1.040	1.470	1.135	1.610	1.930	1.850	1.110	1.120	1.170	1.150

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

lists the variables’ pairwise correlations and variance inflation factors (VIF). lnGI and lnGIA are significantly positively correlated with lnDT, providing initial support for the proposed research Hypotheses 1. The model does not suffer from serious multicollinearity issues because all VIF values are less than the critical value of 10.

4.3. Benchmark Regression

4.3.1. Digital Transformation and Green Innovation

The results in

Table 4. Effect of digital transformation on green innovation.

Variables	Model (1)	Model (2)	Model (3)	Model (4)
	lnGI	lnGI	lnGIA	lnGIA
lnDT	0.156 ***	0.095 **	0.155 ***	0.103 **
	(4.52)	(2.45)	(3.68)	(2.17)
Age		0.311 ***		0.187
		(3.13)		(1.53)
CR		0.020		0.003
		(0.81)		(0.12)
Leverage		0.063		−0.082
		(0.20)		(−0.22)
EC		−0.227		−1.050
		(−0.36)		(−1.34)
Director		0.267		0.290
		(0.41)		(0.36)
Turnover		−0.034		−0.034
		(−0.29)		(−0.23)
ROA		−0.204		−0.343
		(−0.52)		(−0.70)
_cons	0.265 ***	−0.519	0.483 ***	0.371
	(4.68)	(−1.08)	(6.99)	(0.63)
N	384	384	384	384
R2	0.055	0.100	0.037	0.065
F	20.453	4.757	13.573	2.976

Note: ** p < 0.05, *** p < 0.01. The data in parentheses are T values, the same as below.

support Hypothesis 1. Before regression, the Hausman test rejected the null hypothesis, therefore, a fixed effects model was adopted. In Model (2), the coefficient of lnDT is 0.095, which is significant at the 5% level, indicating that an increase of one standard deviation in digital transformation leads to a 9.5% increase in the logarithm of green patent authorizations. The R-squared value for Model (2) is 10%. In Model (4), lnDT is positively correlated with lnGIA, and the coefficient of lnDT is 0.103, which means that with the support of digital transformation, the logarithmic increase in the number of green patent applications is relatively large, but the R-squared value for Model (4) is only 6.5%. Taking comprehensive considerations into account, lnGI was mainly used as the dependent variable in the subsequent tests.

4.3.2. Robustness Test of Main Effects and Endogeneity Handling

(1)

Quantile regression

Considering the robustness of the results, a quantile analysis was conducted based on the level of green innovation (quantile type was 0.1 to 0.9, interval 0.1), and the results are shown in

Table 5. Quantile regression results.

Variables	lnGI	lnGI	lnGI	lnGI	lnGI	lnGI	lnGI	lnGI	lnGI
Quantile	0.10	0.20	0.30	0.40	0.50	0.60	0.70	0.80	0.90
lnDT	0.000	0.000	0.000	0.000	0.000	0.000	0.270 *	0.282 **	0.293 **
	(0.000)	(0.000)	(0.000)	(0.000)	(0.000)	(0.000)	(2.560)	(4.685)	(4.915)
cons	0.000	0.000	0.000	0.000	0.000	0.000	0.000	0.372 **	1.216 **
	(0.000)	(0.000)	(0.000)	(0.000)	(0.000)	(0.000)	(0.000)	(3.722)	(4.720)
R2	−0.000	−0.000	−0.000	−0.000	−0.000	−0.000	0.052	0.030	0.012

Note: * p < 0.1, ** p < 0.05. The data in parentheses are T values.

. Through plotting, the regression coefficients and their confidence intervals at different quantiles were obtained (Table 5). It can be observed that with the changes in digital transformation, digital transformation maintains a positive impact on green innovation, and the coefficient continues to increase. This supports Hypothesis 1.

(2)

Difference-in-Differences

Considering the gradual digital transformation in the sports industry as a quasi-natural experiment, a multi-period difference-in-differences model was used to address endogeneity issues [102]. The following models were constructed:

$$\ln \mathrm{G}{\mathrm{I}}_{\mathrm{i},\mathrm{t}}=\mathsf{\alpha }+{\mathsf{\rho }}_1\mathrm{D}{\mathrm{u}}_{\mathrm{i},\mathrm{t}}+{\mathsf{\rho }}_2\left(\mathrm{D}{\mathrm{u}}_{\mathrm{i},\mathrm{t}}\times \mathrm{D}{\mathrm{t}}_{\mathrm{i},\mathrm{t}}\right)+{\mathsf{\rho }}_3\mathrm{D}{\mathrm{t}}_{\mathrm{i},\mathrm{t}}+\sum \mathsf{\varphi }\mathrm{C}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}+\mathsf{\epsilon }$$

(2)

$$\ln \mathrm{G}{\mathrm{I}}_{\mathrm{i},\mathrm{t}}={\mathsf{\alpha }}^{\prime }+{{\mathsf{\rho }}_1}^{\prime }\left(\mathrm{D}{\mathrm{u}}_{\mathrm{i},\mathrm{t}}\times \mathrm{D}{\mathrm{t}}_{\mathrm{i},\mathrm{t}}\right)+\sum {\mathsf{\varphi }}^{\prime }\mathrm{C}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}+\sum \mathrm{Y}\mathrm{e}\mathrm{a}\mathrm{r}+\sum \mathrm{I}\mathrm{n}\mathrm{d}\mathrm{u}\mathrm{s}\mathrm{t}\mathrm{r}\mathrm{y}+\mathsf{\epsilon }$$

(3)

where Du represents individual dummy variables. Du was assigned a value of 1 for companies that had undergone digital transformation during the statistical period; otherwise, it was assigned a value of 0. It should be pointed out that two procedures were implemented to ensure an adequate observation period before and after digital transformation for the double-difference sample: (1) Eliminate samples in the Du = 1 group that have implemented digital transformation for less than two years; (2) Delete Samples of digital transformation keywords that appear every year. Dt represents the time dummy variable. If a sample company undergoes digital transformation in the current year and every year thereafter, then Dt = 1; otherwise, Dt = 0. The magnitude and direction of ${\mathsf{\rho }}_1$ reflect the changes in green innovation before and after digital transformation. Equation (3) adds year dummy variables based on Equation (2). Since Equation (2) ignores the impact of digital transformation intensity, a difference-in-difference model with a moderating effect was further constructed to estimate this impact:

$$\ln \mathrm{G}{\mathrm{I}}_{\mathrm{i},\mathrm{t}}={\mathsf{\alpha }}^{″}+{{\mathsf{\rho }}_1}^{″}\left(\mathrm{D}{\mathrm{u}}_{\mathrm{i},\mathrm{t}}\times \mathrm{D}{\mathrm{t}}_{\mathrm{i},\mathrm{t}}\times \ln \mathrm{D}{\mathrm{T}}_{\mathrm{i},\mathrm{t}}\right)+\sum {\mathsf{\varphi }}^{″}\mathrm{C}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}+\sum \mathrm{Y}\mathrm{e}\mathrm{a}\mathrm{r}+\mathsf{\epsilon }$$

(4)

In the given context, lnDT represents the intensity of the digital transformation, while “${\mathsf{\rho }}_1$” is the main parameter to be estimated and reflects the impact of the transformation intensity after implementing digital transformation on green innovation.

As can be seen from

Table 6. Endogenetic analysis.

Variables	Model (1)	Model (2)	Model (3)
	lnGI	lnGI	lnGI
Du	−0.024
	(−1.322)
Dt	−0.274
	(−1.281)
Du × Dt	0.391 *	0.124 *
	(1.82)	(2.029)
Du × Dt × lnDT			0.132 *
			(2.154)
Control	Yes	Yes	Yes
Year	Yes	Yes	Yes
R2	0.068	0.062	0.064
F	2.014	2.309	2.378

Note: * p < 0.1. The data in parentheses are T values.

, in both Model (1) and Model (2), Du × Dt is significantly and positively correlated with lnGI, indicating that the level of green innovation significantly improves after undergoing digital transformation. In Model (3), considering the differences in the impact of green innovation under different digital transformation intensities, the research results show that the interaction term Du × Dt × lnDT still has a significant positive impact on lnGI. It can be seen that the endogeneity analysis results are highly consistent with the conclusions of the benchmark regression.

(3)

Poisson regression

Due to the dependent variable, i.e., the number of patent grants, being a non-negative integer, it exhibits the characteristics of count data. This study employed a Poisson regression model for robustness checks. The assumption of Poisson regression is that the variance of the independent variable is approximately equal to its mean. In our case, the mean of lnDT is 1.439 and the variance is 1.579, indicating a close match and satisfying the assumptions of Poisson regression.

The empirical results are shown in

Table 7. Poisson regression.

Variables	lnGI
lnDT	0.133 *
	(2.161)
Age	0.227
	(1.608)
CR	−0.288 **
	(−3.073)
Leverage	0.680 *
	(2.527)
EC	−0.277
	(−0.385)
Director	0.721
	(0.794)
Turnover	−0.738 **
	(−2.969)
ROA	0.175
	(0.165)
Cons	−0.735
	(−0.889)
alpha	−0.210 **
	(−4.828)
N	384
Likelihood ratio test	31.500
AIC	706.345
BIC	745.860
McFadden R2	0.044

Note: * p < 0.1, ** p < 0.05. The data in parentheses are T values.

. The likelihood ratio test for the Poisson regression model was used to evaluate the overall model’s validity. The results indicate a p-value less than 0.05, confirming that the model is effective. The regression results reveal that the coefficient for digital transformation on green innovation is 0.133, which is significant at the 10% level. This finding is consistent with our baseline regression results, alleviating concerns about changes in the sequential characteristics of the dependent variable after logarithmic transformation. The Poisson regression test further corroborates the positive impact of digital transformation on green innovation.

(4)

Panel instrumental variable estimation

In the digital transformation of sports companies, endogeneity is a significant issue. Different sports companies have varying demands for digital transformation based on their development goals. For instance, sports companies that prioritize green development as a strategic objective may have a stronger preference for the development and application of digital technologies. This endogeneity problem, arising from reverse causality, can lead to biased model estimation results.

To address the endogeneity caused by reverse causation, this study employs instrumental variable (IV) methods to mitigate the selection bias. Given that digital transformation in companies exhibits demonstration effects, the level of digital transformation among firms in the same province and industry during the same year is closely related to the individual digital transformation levels of sports companies, while not having a direct causal relationship with their green innovation performance. Consequently, the average digital transformation level for the province–industry–year combination was selected as the instrumental variable. We utilized the two-stage least squares (2SLS) estimation method for the analysis, with the panel instrumental variable estimation results presented in

Table 8. Panel Instrumental Variable Estimation.

Dependent Variables	lnDT	lnGI
OLS	Stage 1	Stage 2
lnDT		0.161 **
		(2.891)
Instrument	0.762 **
	(13.918)
Control	Yes	Yes
year	Yes	Yes
N	384	384
R2	0.384	0.012
F	29.256
Cragg-Daniels statistic	98.821	98.821
Wald χ2		31.996
Wu-Hausman test		13.641

Note: ** p < 0.05, The data in parentheses are T values.

.

First, the Hausman–Wu endogeneity test was conducted to determine if endogeneity is present. The p-value is less than 0.05, indicating that endogeneity is indeed an issue in the model. Thus, using instrumental variable estimation was necessary. The first-stage results show that the estimated coefficient for the instrumental variable is 0.762 and is significant at the 5% level. This suggests that a higher level of digital transformation among companies in the same region, industry, and year positively influences the digital transformation level of sports companies due to spillover effects.

The Cragg–Donald statistic F-value is significant, with the first-stage F-value exceeding 10, which strongly rejects the null hypothesis of weak instruments, indicating that the model does not suffer from weak instrument issues. The second-stage results reveal that the estimated coefficient for the impact of digital transformation on green innovation is 0.161 and is significant at the 5% level. The panel instrumental variable estimation further confirms that digital transformation has a positive effect on the green innovation of companies.

4.4. Mechanism Analysis

Using the approach proposed by [106], the following models can be constructed to elucidate the pathways through which digital transformation affects green innovation:

$$\ln \mathrm{G}{\mathrm{I}}_{\mathrm{i},\mathrm{t}}={\mathsf{\alpha }}_0+{\mathsf{\alpha }}_1\ln \mathrm{D}{\mathrm{T}}_{\mathrm{i},\mathrm{t}}+\sum \mathsf{\alpha }\mathrm{C}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}+\sum \mathrm{Y}\mathrm{e}\mathrm{a}\mathrm{r}+\sum \mathrm{I}\mathrm{n}\mathrm{d}\mathrm{u}\mathrm{s}\mathrm{t}\mathrm{r}\mathrm{y}+\mathsf{\epsilon }$$

(5)

$$\mathrm{M}\mathrm{e}\mathrm{d}\mathrm{i}\mathrm{a}\mathrm{t}\mathrm{o}{\mathrm{r}}_{\mathrm{i},\mathrm{t}}={\mathsf{\theta }}_0+{\mathsf{\theta }}_1\ln \mathrm{D}{\mathrm{T}}_{\mathrm{i},\mathrm{t}}+\sum \mathsf{\theta }\mathrm{C}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}+\sum \mathrm{Y}\mathrm{e}\mathrm{a}\mathrm{r}+\sum \mathrm{I}\mathrm{n}\mathrm{d}\mathrm{u}\mathrm{s}\mathrm{t}\mathrm{r}\mathrm{y}+\mathsf{\epsilon }$$

(6)

$$\ln \mathrm{G}{\mathrm{I}}_{\mathrm{i},\mathrm{t}}={\mathsf{\varphi }}_0+{\mathsf{\varphi }}_1\ln \mathrm{D}{\mathrm{T}}_{\mathrm{i},\mathrm{t}}+{\mathsf{\varphi }}_2\mathrm{M}\mathrm{e}\mathrm{d}\mathrm{i}\mathrm{a}\mathrm{t}\mathrm{o}{\mathrm{r}}_{\mathrm{i},\mathrm{t}}+\sum \mathsf{\varphi }\mathrm{C}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}+\sum \mathrm{Y}\mathrm{e}\mathrm{a}\mathrm{r}+\sum \mathrm{I}\mathrm{n}\mathrm{d}\mathrm{u}\mathrm{s}\mathrm{t}\mathrm{r}\mathrm{y}+\mathsf{\epsilon }$$

(7)

In the given context, the Mediator_i,t refers to the mediating variables, which include firm human capital and internal control. The model focuses on the magnitude and direction of the coefficient ${\mathsf{\varphi }}_1$.

Table 9. The mediating role of human capital and internal control.

Variables	Model (1)	Model (2)	Model (3)	Model (4)	Model (5)
	lnGI	HC	lnGI	IC	lnGI
lnDT	0.095 **	0.013 *	0.090 **	−0.004	0.095 **
	(2.45)	(1.75)	(2.32)	(−0.05)	(2.46)
HC			0.392 *
			(1.84)
IC					0.032 *
					(1.72)
Age	0.311 ***	0.070 ***	0.284 ***	−0.132	0.315 ***
	(3.13)	(2.82)	(2.83)	(−0.57)	(3.17)
CR	0.020	0.014 **	0.014	0.047	0.018
	(0.81)	(2.33)	(0.58)	(0.67)	(0.75)
Leverage	0.063	0.092	0.026	0.845	0.036
	(0.20)	(1.19)	(0.09)	(1.19)	(0.12)
EC	−0.227	−0.778 ***	0.078	0.241	−0.235
	(−0.36)	(−4.85)	(0.12)	(0.17)	(−0.37)
Director	0.267	−0.212	0.350	1.277	0.235
	(0.41)	(−1.30)	(0.54)	(0.85)	(0.36)
Turnover	−0.034	0.078 **	−0.065	−0.526 *	−0.018
	(−0.29)	(2.57)	(−0.54)	(−1.93)	(−0.15)
ROA	−0.204	0.002	−0.205	2.049 *	−0.261
	(−0.52)	(0.02)	(−0.52)	(1.84)	(−0.66)
_cons	−0.519	0.301 **	−0.637	5.801 ***	−0.704
	(−1.08)	(2.50)	(−1.32)	(5.08)	(−1.41)
N	384	384	384	384	384
R2	0.100	0.189	0.109	0.024	0.104
F	4.757	10.005	4.631	1.069	4.431

Note: * p < 0.1, ** p < 0.05, *** p < 0.01. The data in parentheses are T values.

reports the results of the mediation effect. Following the steps of testing mediation effects, the coefficients of lnDT and HC were significantly positive in Model (1) and Model (2). In Model (3), the coefficients of the corresponding variables changed. After including HC as a mediating variable in Model (3), the coefficient of lnDT decreased to 0.090, while HC showed a significant positive correlation with lnGI. This confirms that firm human capital partially mediates the impact of digital transformation on green innovation. The results of Model (5) indicate that internal control plays a partial mediating role in the relationship between digital transformation and green innovation. Based on this, Hypotheses 2 and 3 are supported.

Model 4, in the process of SPSS, was used to test the mediating effect again.

Table 10. Mediating effect test: based on the Bootstrap test program.

Influence Mechanism	Indirect Effect	Boot SE	95% Confidence Interval
lnDT—HC—lnGI	0.0213	0.0081	[−0.0515, −0.0100]
lnDT—IC—lnGI	0.0021	0.0022	[−0.0068, −0.0021]

shows the results of the Bootstrap test procedure with 1000 iterations. As can be seen from the table, digital transformation has a significant indirect effect on green innovation through both human capital and internal control. Furthermore, human capital shows a stronger indirect effect.

4.5. Extension Analysis

Firstly, the analysis of company heterogeneity focuses on the nature of property rights. This is because state-owned companies (SOEs) and non-state-owned companies (non-SOEs) have distinct development goals, political functions, and market competition pursuits. The differences in resource allocation intensity, including technology and other resources, make firms vulnerable to the influence of unobserved characteristics. The results in

Table 11. Heterogeneity analysis: property rights and environmental regulations.

Variables	Model (1)	Model (2)	Model (3)	Model (4)
	lnGI	lnGI	lnGI	lnGI
	SOEs	Non-SOEs	Strong Environmental Regulation	Weak Environmental Regulation
lnDT	0.156 **	0.069 *	0.123 **	0.053
	(2.34)	1.66	(2.09)	(1.05)
Age	0.207 **	0.208 **	0.336 ***	0.285 ***
	(2.51)	2.04	(4.03)	(3.47)
CR	−0.058	−0.028	−0.039	0.001
	(−0.84)	−0.89	(−0.66)	(0.02)
Leverage	−0.637	−0.153	−0.649	−0.404
	(−0.89)	−0.41	(−1.40)	(−0.78)
EC	−3.083 *	0.871	0.432	−0.287
	(−1.87)	1.44	(0.48)	(−0.31)
Director	1.711	−0.166	1.112	0.505
	(1.54)	−0.21	(1.09)	(0.50)
Turnover	0.110	−0.110	−0.200	0.124
	(0.54)	−0.77	(−1.19)	(0.67)
ROA	−0.635	−0.631	0.028	−0.935
	(−0.66)	−1.35	(0.04)	(−1.39)
_cons	0.112	−0.221	−0.983	−0.628
	(0.12)	−0.39	(−1.46)	(−1.04)
N	179	205	181	203
R2	0.217	0.1016	0.195	0.118
F	5.239	2.47	4.243	2.704

Note: * p < 0.1, ** p < 0.05, *** p < 0.01. The data in parentheses are T values.

indicate that digital transformation has a significant positive impact on the intensity of green innovation in firms with different property rights. However, the positive effect on the green innovation of state-owned companies is significantly greater than that of non-state-owned companies.

In driving green innovation, state-owned sports companies are more significantly influenced by digital transformation. A deeper understanding reveals that, compared to non-state-owned companies, state-owned companies not only pursue economic profits but also bear greater social responsibilities. They often face stronger regulatory constraints and place a higher emphasis on environmental performance to avoid excessive exploitation, extensive resource utilization, and uncontrolled consumption. In contrast, the development of non-state-owned companies relies heavily on technological innovation. In the context of disruptions caused by information technology, these companies are more inclined to autonomously integrate resources and to adapt to turbulent external environments. Consequently, non-state-owned companies possess a strong capacity to explore, learn, and leverage cross-border knowledge to facilitate green innovation. This ability, to some extent, diminishes the driving effect of digital transformation on their green initiatives.

Second, the micro-paradigm of digital transformation-green innovation is embedded in external macro-contextual conditions, such as environmental regulations. This study attempts to explore the changes in the impact of digital transformation on green innovation under external environment fluctuations. The samples were classified and regressed according to the mean value of environmental regulations. The results are shown in Table 9. Under the conditions of strict environmental regulations, the coefficient of lnDT is significantly positive. On the contrary, under the conditions of weak environmental regulations, the coefficient of lnDT is not significant.

The reasons for the above results may be as follows: Although environmental regulation represents constraints on corporate activities that have evolved from pollution control to ecological footprint reduction, the incentive effect of environmental regulation on green innovation in sports companies can be constrained by crowding-out effects and an opaque information environment. Therefore, in regions with strict environmental regulation, sports companies undergoing digital transformation can improve the transparency of environmental behavior while reducing environmental costs. Through green product design, process control, and business model innovation, they can reduce energy consumption, achieve energy conservation and emissions reduction. This logic aligns with the ideas of [85], who suggest that the impact of information and communication technology on green total-factor energy efficiency varies depending on the intensity of environmental regulation.

4.6. Robustness Analysis

To promote robustness testing, the main variables were replaced with alternative indicators. Firstly, considering the differences in the length of annual reports, the product of the total frequency of digital transformation and the ratio to the total word count of the annual report, multiplied by 100, was used as a substitute indicator for lnDT. Secondly, the number of patent applications was used as a proxy variable for green innovation. The robustness testing results for the main effects and mediating effects are shown in

Table 12. Robustness checks.

Variables	Model(1)	Model(2)	Model(3)	Model(4)	Model(5)
	lnGI	HC	lnGI	IC	lnGI
lnDT	0.233 *	0.016	0.223 *	0.032	0.232 *
	(1.91)	(0.63)	(1.84)	(0.14)	(1.90)
HC			0.658 **
			(2.51)
IC					0.032 *
					(1.66)
Age	0.244 **	0.079 ***	0.192	−0.123	0.248 **
	(2.09)	(3.32)	(1.63)	(−0.58)	(2.12)
CR	0.009	0.015 **	−0.001	0.043	0.008
	(0.31)	(2.44)	(−0.02)	(0.80)	(0.26)
Leverage	−0.027	0.099	−0.092	0.851	−0.054
	(−0.07)	(1.28)	(−0.25)	(1.24)	(−0.14)
EC	−1.107	−0.791 ***	−0.586	0.289	−1.116
	(−1.41)	(−4.94)	(−0.73)	(0.20)	(−1.42)
Director	0.209	−0.235	0.363	1.084	0.175
	(0.26)	(−1.44)	(0.46)	(0.75)	(0.22)
Turnover	−0.046	0.077 **	−0.096	−0.520 *	−0.030
	(−0.31)	(2.53)	(−0.65)	(−1.93)	(−0.20)
ROA	−0.275	0.014	−0.284	1.746 **	−0.330
	(−0.57)	(0.14)	(−0.59)	(1.98)	(−0.68)
_cons	0.334	0.303 **	0.135	5.827 ***	0.150
	(0.56)	(2.50)	(0.23)	(5.40)	(0.24)
N	384	384	384	384	384
R2	0.062	0.186	0.079	0.026	0.065
F	2.837	9.786	3.259	1.132	2.648

Note: * p < 0.1, ** p < 0.05, *** p < 0.01. The data in parenthesis are T values.

. The coefficients and directions of the core variables, such as lnDT, HC, and IC, are generally consistent, indicating the robustness of the previous research conclusions.

5. Discussions

Based on the logical framework of digitalization driving greenization in companies, this study empirically investigates the impact of digital transformation on green innovation, explores the underlying mechanisms, and examines the consequences of heterogeneous effects. The analysis focuses on Chinese A-share-listed companies in the sports industry from 2011 to 2022. The main findings can be summarized in the following three aspects:

Firstly, this study finds that digital transformation has a significant positive impact on sports companies’ green innovation. This highlights the importance of strengthening digital technologies as a key driver in promoting sustainable practices within companies. Secondly, the examination of underlying mechanisms reveals that human capital and internal control partially mediate the relationship between digital transformation and green innovation. Human capital plays a more prominent mediating role compared to internal control. Thirdly, the study uncovers heterogeneous effects, indicating that corporate property rights and environmental regulation asymmetrically affect the process of digital transformation driving green innovation. Specifically, digital transformation has a stronger positive effect on green innovation in state-owned companies compared to non-state-owned companies. Moreover, under stringent environmental regulation, digital transformation significantly promotes green innovation, while the relationship between the two is not significant under lax environmental regulation. These findings provide valuable insights for governments to develop customized green development policies based on local conditions.

5.1. Theoretical Significance

The potential theoretical contributions are mainly in the following three aspects: First, the digital transformation of Chinese sports companies has a positive impact on green innovation. Within the dual context of digitalization and green development, this study confirms the driving effect of digital transformation on green innovation, which is consistent with previous research findings. Although the results align with those of Xu et al. [90] and Han et al. [96], who found that digital transformation may positively predict green innovation, there are significant differences in the research effect and research perspective. On one hand, some scholars have previously proposed that digital transformation has positive effects on economic activity, such as transforming resource integration and allocation, assisting decision making, and reducing innovation transaction costs [27], thereby facilitating “source reduction” and “end-cleaning” approaches [28]. Meng et al. [29] emphasize that digital technology provides a more robust power system, greener production methods, efficient operational efficiency, and intelligent governance models. However, some studies found the negative effects of digital transformation, such as Li [32], who identified issues such as resource overconsumption and the digital divide. How does digital transformation perform in effecting green innovation in sports companies? This study provides evidence of the positive impact of digital transformation on green innovation in sports companies. A possible explanation is that the sports industry, being a green sector, may have inherent environmental advantages that offset the potential negative impacts of digital transformation. Additionally, many sports companies are still in the exploratory stage of their digital transformation, meaning that any negative effects have not yet become apparent. On the other hand, within the dual context of digitalization and low-carbon development, this study examines the driving effect of digital transformation on green innovation at the micro level of sports companies, complementing existing research at the industry and regional levels [24,29]. In the previous studies, the research has primarily focused on macro-level analyses, examining their effects on industrial green innovation and the quality of economic development [15,107]. For example, Wei et al. [9] studied the impact of digitalization on sustainable innovation in the sports industry, and Wang et al. [15] unraveled the positive impact of the digital economy on green innovation in heavily polluting industries. However, few researchers have studied the impact of digital transformation on green innovation in sports companies. In this study, by focusing on green innovation in these companies, the research further extends the exploration of the microeconomic implications of the digital economy. In summary, this research responds to Ratten et al.’s call for more empirical studies on sports companies in the new context [13], enriching the theoretical support for advancing sustainable development strategies from the perspective of digitalization in the sports sector. It deepens the comprehension of the consequences and controversies surrounding the impact of digital technologies. Consequently, this research provides robust theoretical support for the promotion of sustainable development strategies from a digital standpoint.

Second, the research sheds light on the “black box” of the impact of digital transformation on green innovation and enriches the research on the channel effect of digital transformation. This study presents a comprehensive research framework comprised of a “benchmark analysis–mechanism analysis–heterogeneity analysis”. This framework effectively unveils the intricate “black box” of mechanisms underpinning the impact of digital transformation on green innovation. The mechanism analysis indicates that digital transformation in sports companies can promote green innovation by enhancing human capital levels and internal control. Consistent with the “capital–skill complementarity” hypothesis, this digital transformation leads to the replacement of low-level human capital and an increased demand for high-level skills [63,64,105], thereby facilitating the upgrading of human capital. Furthermore, this upgrading influences green innovation by enhancing the capacity for green ideas and practices [69]. Regarding internal control, similar to findings in manufacturing companies [72,79], the digital transformation of sports companies provides technological support that improves information transparency and quality, reducing regulatory costs [66]. This enriches the existing research on the integration of corporate governance and green development within a digital context. Importantly, our study finds that the mediating role of human capital in sports companies is stronger than that of internal control. This may be attributed to the fact that human capital serves as a primary participant in green development, possessing the agency to better integrate new technologies and respond swiftly to the demands of green innovation during the digital transformation [18,105]. In summary, this research explores the intrinsic mechanisms through which digital transformation impacts corporate green innovation, thereby deepening the theoretical understanding of corporate green innovation and providing a foundation for developing effective talent cultivation and internal control systems for sustainable development.

Third, this study examines the heterogeneous effects of digital transformation on driving green innovation. The heterogeneity analysis was conducted under the varying conditions related to company property rights and environmental regulations. The findings indicate that the positive impact of digital transformation on green innovation is greater for state-owned companies than non-state-owned companies. This aligns with the perspective of Li et al. [108], who believe that compared to non-state-owned companies, state-owned companies not only pursue economic profits, but also undertake more social responsibilities. They often encounter stricter regulatory constraints and pay more attention to environmental performance. Additionally, this study explores how external environmental fluctuations affect the impact of digital transformation on green innovation. This study finds that the positive impact of companies’ digital transformation on green innovation is only significant in the context of strong environmental regulation. As a complement to existing research in the sports sector, this study aligns with findings from manufacturing and heavily polluting industries. Sports companies facing stringent environmental regulations are often more proactive in pursuing digital transformation to achieve green development, driven by external pressures from stakeholders and internal motivations to comply with environmental constraints. This logic is consistent with the view of Hao et al. [85], who believe that the impact of information and communication technology on green total factor energy efficiency depends on the strength of environmental regulation. However, it contrasts with the research by Chen and Wang [109], which highlights the negative effects of environmental regulations on the efficiency of the sports ecosystem. The performance of digital transformation on green innovation may differ for companies with different ownership characteristics and that are located in different regions (with varying environmental regulations). The study enhances our understanding of the contextual factors influencing the relationship between digital transformation and green innovation, along with its contingency mechanism. It aims to provide clear guidance for the green innovation practices of sports companies and offers valuable insights for the formulation of environmental policies aimed at promoting corporate green innovation in the context of the “dual carbon” goals.

5.2. Practical Implications

The research conclusions provide important management enlightenment for sports companies. First, sports companies should prioritize strengthening digital transformation as a driver of green innovation. Given the positive correlation between digital transformation and green innovation, sports companies should capitalize on the opportunities presented by digitalization and foster the seamless integration of digital technologies such as blockchain, cloud computing, and big data across sports product manufacturing, organizational management, and business models. This will establish a solid foundation for sustainable development. For government departments, it is crucial to enhance support for digital transformation. This can be achieved through initiatives such as bolstering digital infrastructure development, including the expansion of 5G base stations and high-speed broadband networks, as well as the establishment of comprehensive big data centers. Furthermore, efforts should be made to enhance talent support and provide multi-level investment and financing services to facilitate digital transformation in the sports industry. Governments should also leverage policy guidance and allocate resources strategically to support and incentivize digital transformation initiatives.

Second, sports companies should harness the transmission effects of human capital and internal control within their organizations. The findings regarding the mediating effects indicate that sports companies should establish a dynamic transmission mechanism that encompasses “digital transformation–human capital/internal control–green innovation”. Sports companies should focus on recruiting top-level talent who possess a strong understanding of and practical expertise in digital technologies. These individuals can drive the integration of digital tools and solutions into the companies’ operations, fostering green innovation. Additionally, sports companies should develop and implement feasible plans for green governance, ensuring that the organization has effective mechanisms in place to monitor and regulate environmental performance. By establishing intelligent production systems and management systems based on digital technologies, sports companies can stimulate and control the systematic and sustainable adoption of green development strategies.

Third, sports companies should tailor strategies based on their unique characteristics. The findings regarding heterogeneous effects suggest that when implementing digital strategies, sports companies should take into account their differentiation and uniqueness. It is crucial to allocate resources wisely, considering both the policy environment and the specific characteristics of each sports company. On one hand, there should be a focus on guiding the digital transformation of state-owned sports companies. This entails fully harnessing the potential of digital technologies to drive green innovation and leveraging the “demonstration effect” that state-owned companies can have. On the other hand, the government should create an enabling external environment that promotes digital benefits. This can be achieved by formulating scientifically sound policies and regulations related to the environment. The government should also actively encourage digital transformation entities to embrace green innovation concepts, formulate green development strategies, actively participate in environmental governance, and fulfill their social responsibilities.

6. Limitations and Future Research

Although the hypotheses proposed herein are supported, this study has certain research limitations. The first limitation pertains to the research subject. While the research sample includes Chinese companies in the sports industry, it is limited to those listed on the Shanghai and Shenzhen exchanges. The study does not devote attention to companies in other countries. With the advancement of a new generation of digital technologies, the impact of digital transformation on companies will be further enhanced. In the future, research on the impact of digital transformation on company transformation could sample companies from other countries or industries to understand the impact of digital transformation on green innovation more comprehensively. The second limitation pertains to the research scope. The impact of digital transformation on companies is far-reaching and multi-faceted, and this study focuses only on the positive impact of digital transformation on company development. Future research can conduct more comprehensive theoretical studies and data mining to explore the potential negative effects of companies’ digital transformation, in order to gain a more complete and clear understanding of the overall impact of digital transformation. The third limitation pertains to the mechanisms. This study discusses only the human capital and internal control mechanisms. In the future, research on the impact of digital transformation on companies could be extended to other aspects such as management attention and corporate responsibility, expanding our understanding of how digital transformation affects company development.