With the increasingly severe global climate change situation, carbon emission reduction (CER) has become a focal point of international concern. As the share of the digital economy’s output continues to rise each year, the emergence of new industries such as e-commerce, mobile payments, and cloud computing has opened new avenues for CER. At its core, the digital economy (DE) leverages data, information technology, and digital tools to optimize resource allocation, enhance productivity, and foster innovation. This helps to reduce carbon dioxide emissions across both production and daily life (Abbas et al., 2022). Furthermore, the Climate Action Roadmap highlights that the application of digital technology could potentially cut global CE by around 15% (Mustajoki et al., 2024). Consequently, the CER effect of DE has become a prominent topic of interest within academic circles. As the world’s largest carbon emitter, China plays a pivotal role in global climate governance, with its success in carbon reduction exerting significant influence on the international stage (Afshan et al., 2023). The rapid development of its DE offers abundant practical insights into low-carbon transformation. Meanwhile, the stark disparities in economic development, resource endowment, and industrial structure across China’s regions provide an ideal context for examining the relationship between the DE and carbon reduction. Additionally, China’s accelerated urbanization presents both challenges in energy consumption and opportunities for integrating the DE with green development. Guided by its goals of peaking CE by 2030 and achieving carbon neutrality by 2060, China’s low-carbon policies not only support research into digital economy-driven carbon reduction pathways but also offer valuable lessons and policy references for other nations (Liu et al., 2023).

Existing research indicates that the DE can significantly influence regional CE through multiple pathways. For instance, the widespread adoption of digital technologies such as cloud computing, facial recognition, and artificial intelligence has reduced resource waste, thereby improving the efficiency of industrial production and urban management (Liu et al., 2020). The proliferation of online services has also decreased the consumption of transportation energy (Imran et al., 2023). Moreover, the DE drives industrial restructuring and technological upgrading (Chang et al., 2023). While numerous scholars have explored the CER effects of DE, few have integrated these factors into a comprehensive analytical framework. There remains a need for systematic research on the multiple pathways and conditions through which the DE affects regional CE. Additionally, studies on the regional disparities and nonlinear characteristics of the DE’s CER effects are still relatively scarce. To address the aforementioned research gaps, this study conducts a systematic analysis of the synergistic pathways through which the DE, industrial restructuring, technological innovation, and urbanization impact China’s CE. First, the study employs a System Generalized Method of Moments (SYS-GMM) model to test the direct effects of the DE on regional CE. Second, a mediation model is constructed to analyze the indirect effects of the DE on CE through industrial structure and technological innovation. The study then introduces urbanization as a threshold variable to explore the nonlinear relationship of the DE’s CER effects across different stages of urbanization. Finally, the heterogeneity of the CER effect of DE was examined from the perspectives of resource endowment and economic level.

This study provides a comprehensive framework for academics and policymakers, uncovering the multifaceted pathways through which the DE fosters regional carbon reduction. Against the backdrop of a major nation like China, the findings hold extensive practical applications and policy implications. The novelty of this research is reflected in several key aspects: first, it adopts a systematic analytical approach to holistically examine the impacts of the DE, industrial structure optimization, technological innovation, and urbanization levels on regional CE, distinguishing it from prior studies that focused solely on direct or indirect effects. Second, the study introduces urbanization levels as a threshold variable to analyze the nonlinear carbon reduction effects of the DE across different stages of urbanization, addressing a gap in the literature. Third, it explores the regional heterogeneity of the DE from the perspectives of resource endowment and economic development, revealing variations in its emission-reduction effects. Finally, by employing the SYS-GMM model to analyze direct impacts and a mediation effect model to investigate indirect mechanisms, the study provides a detailed and multidimensional perspective, enhancing both the depth and breadth of the research.

As environmental issues stemming from climate change grow increasingly severe, scholars have conducted extensive research on CER. Numerous scholars have explored the effects of factors such as energy endowment, economic openness, technological progress, industrial upgrading, and FDI on regional CE (Shaari et al., 2021; Hu et al., 2021; Wen et al., 2021; Wu et al., 2021; Rauf et al., 2023). DE, as a new economic growth point, has been validated at both macro and micro levels for its impact on regional CE. The rise of the DE has introduced new pathways for regional CER. Existing research findings primarily examine the CER the DE from the following three perspectives (Sadiq and Ali, 2024).

Firstly, extensive research has been conducted on the direct impact of the DE on regional CE. Theoretical research reveals that the direct impact of the DE on CE lies in a dual dynamic: the enhancement of energy efficiency and resource utilization, coupled with the growth in energy demand. Through the widespread adoption and application of information technology, the DE significantly reduces energy waste in traditional production and daily life. Innovations such as the industrial internet, smart manufacturing, online office platforms, and e-commerce effectively lower CE. However, the development of the DE has also led to energy-intensive activities, such as data center operations, high-performance computing equipment manufacturing, and logistics distribution, contributing to increased CE. This results in a coexistence of both positive and negative direct impacts (Haita et al., 2022). Empirical research highlights both linear and nonlinear effects. Linear studies indicate that DE can significantly reduce CE. For instance, research by Karaki et al. indicates that the DE directly lowers CE in high-energy-consuming industries by driving digital transformation and enhancing production efficiency (Karaki et al., 2023). However, scholars like Salahuddin et al. hold a contrasting view, arguing that the rapid expansion of DE has led to a sharp increase in electricity consumption and the construction of new infrastructure, thereby raising regional CE (Salahuddin and Alam, 2015). Additionally, some scholars have found that the CER effect of DE is non-linear. As the DE progresses, its CER effects may exhibit threshold effects or an inverted “U”-shaped curve (Li and Wang, 2022). This relationship is similar to the Environmental Kuznets Curve, which posits that environmental degradation accompanies early stages of economic growth, but environmental quality improves with higher economic levels (Hassan et al., 2020).

Secondly, the indirect effects of the DE on CE has garnered widespread attention from scholars. Theoretical research suggests that the DE drives the growth of the tertiary sector, particularly low-energy, high-value-added industries, thereby reducing the overall carbon intensity of industrial activities. Digital technologies facilitate green innovation and its diffusion, promote the transition of energy structures toward cleaner alternatives, and provide new momentum for emission reductions. The digital platform economy optimizes resource allocation, minimizes production redundancies, and supports the widespread adoption of low-carbon production and consumption models. However, theoretical studies also highlight challenges, such as the rebound effect in consumption and the imbalance in technological diffusion, which influence the indirect effects of the DE on CE. While improving production efficiency, the DE may stimulate expanded consumption demand—manifested in activities like online shopping and instant delivery—that leads to increased CE. Moreover, regional disparities in the adoption of digital technologies may exacerbate short-term imbalances in CE across regions (Dinda, 2004). Empirical studies reveal that the DE indirectly influences CE through multiple pathways, including industrial structure upgrading, technological innovation, and green finance. Numerous studies indicate that digital technologies have facilitated the transformation of traditional industries into low-carbon and high-value-added industries. Cheng et al. note that the DE indirectly reduces CE in developed regions by promoting the intelligent and green transformation of manufacturing (Cheng et al., 2023). Furthermore, the DE enhances energy utilization efficiency by promoting technological innovation and upgrading. Adebayo’s research highlights that the DE has also fostered construction of smart transportation systems, which reduce transportation CE (Adebayo et al., 2024). In addition, digital technologies have spurred innovation in green financial instruments, such as carbon trading platforms and the digital issuance and management of green bonds (Zhang and Qian, 2023). The widespread application of these tools has facilitated financing for low-carbon projects, consequently leading to reductions in CE.

Thirdly, as research has deepened, scholars have discovered that the CER effects of the DE are influenced by a multitude of elements, including economy, policy, industry, and technology. Studies indicate that in regions with higher levels of economy, the CER effects of the DE are more pronounced. For example, Wang et al. found that in the economically advanced coastal areas of eastern China, the development of the DE, supported by superior digital infrastructure and higher technological capabilities, leads to significant reductions in CE (Wang and Zhong, 2023). The formulation of policies also plays a key role in shaping the CER effects of the DE. Yang et al. discovered that under varying intensities of environmental policies, the CER effects of the DE exhibit considerable differences. In the presence of market incentive and public participation environmental policies, the CER effects of the DE become more significant (Yang and Liang, 2023). Furthermore, industrial structure is a key player influencing the CER effects of the DE. Lyu et al. found that industrial upgrading enhances the CER of DE (Lyu et al., 2023).

To sum up, scholars have analyzed the CER effects of DE from multiple perspectives. While these studies provide valuable insights into the relationship between the DE and CE, there remain several deficiencies regarding the pathways, influencing factors, and heterogeneity of their effects. (1) The analysis of pathways through which the DE influences CER tends to be overly one-dimensional, often focusing on either direct or indirect effects without adopting a comprehensive perspective. Some studies incorporate an analysis of influencing factors, such as marketization level, technological capability, or policy environment, when discussing direct or indirect effects; however, the scope and depth of these analyses remain insufficiently broad. (2) Regarding the influencing factors of the DE on CER, existing literature predominantly focuses on aspects such as economic level, relevant policies, market demand, and technological intensity. While these elements are indeed essential in forming the CER effects of the DE, the overemphasis on them results in a relatively narrow research perspective, particularly lacking in-depth exploration of urbanization—a critical area of development. (3) The current literature’s analysis of the heterogeneity in the CER effects of the DE often centers around geographical differences, which leads to noticeable limitations in certain respects. Conditions such as varying economic development levels, policy environments, and natural resource endowments can significantly influence the CER effects of the DE.

Therefore, this study systematically analyzes the CER pathways of the DE from the perspectives of direct effects, indirect effects, threshold effects, and heterogeneity analysis. Additionally, it introduces urbanization level as a threshold variable to examine the nonlinear characteristics of the DE’s CER effects at different stages of urbanization. The study further explores the heterogeneity of the DE’s influence on regional CE based on resource endowments and economic development levels. This comprehensive approach aims to enhance the understanding of the CER effects of the DE in varied contexts, providing theoretical insights and empirical support for policymakers.

Building on the aforementioned summary of existing research, this study establishes a theoretical analysis framework encompassing four dimensions: direct effects, indirect effects, threshold effects, and heterogeneity analysis, as illustrated in Figure 1. This framework will facilitate an in-depth exploration of the CER effects of the DE in China and, based on this analysis, will propose research hypotheses.

This study systematically analyzes the mechanisms by which the DE contributes to CER in China. Through examinations of direct and indirect impacts, threshold effects, and heterogeneity analysis, it arrives at multi-layered conclusions. This section will conduct a horizontal and vertical comparative analysis of the research findings in relation to existing studies, exploring the differences between this study and other relevant research and further elucidating the underlying causes of these discrepancies.

The findings of this research indicate that the DE has a significant CER effect. Existing literature similarly confirms the potential of the DE in CER. For instance, Li et al. found that digital technologies optimize energy efficiency, thereby reducing CE (Li et al., 2022). However, Zhang et al. posited that in certain less developed regions, the catalytic effect of the DE may be constrained by inadequate infrastructure (Zhang W. et al., 2022). This observation aligns with our study’s conclusion that the CER effects of the DE are not significant in some areas. This discrepancy suggests that while the DE exhibits considerable CER effects, its efficacy is influenced by regional development levels. By further differentiating the CER effects between developed and underdeveloped regions, this study deepens this conclusion. Developed regions, leveraging advanced digital infrastructure and high levels of technological reserves, can more effectively unlock the carbon reduction potential of the DE. In contrast, in underdeveloped regions, inadequate infrastructure and lagging technological capabilities may constrain the carbon reduction effects of the DE. Therefore, we recommend tailoring strategies to local conditions to maximize the carbon reduction benefits of the DE. This includes enhancing broadband networks and data center construction in underdeveloped regions while fostering technological research and industrial clustering in developed regions (Chang et al., 2024).

The empirical analysis results show that DE indirectly reduces China’s CE by optimizing industrial structure and accelerating technological innovation. And the mediating effect of accelerating technological innovation is greater than that of optimizing industrial structure. This conclusion is consistent with many related literature. Such as, Wang et al. indicate that the DE promotes technological advancement and application, thus accelerating the CER process (Wang et al., 2022). Similarly, Cheng et al. suggest that the DE fosters industrial transformation, contributing to the acceleration of CER (Cheng et al., 2023). However, unlike existing research, this study reveals that the mediating effect of technological innovation surpasses that of industrial structure optimization. The investigation indicates that technological innovation, driven by the rapid spread of technology, can yield rapid CER effects. In contrast, optimizing the industrial structure typically requires a longer adjustment period. Furthermore, technological innovation possesses cross-industry spillover effects, making its CER impact more pronounced. Industrial structure adjustments are constrained by factors such as regional economic structure and resource endowments. This is particularly evident in China’s central and western regions, where traditional heavy industries dominate, leading to a lag in the CER effects of industrial structure optimization. Thus, while both pathways contribute to reducing CE, the immediacy and breadth of technological innovation render it a more critical factor in driving CER efforts in the context of the DE. Technological innovation, through the widespread application of digital technologies and cross-sectoral spillover effects, can rapidly achieve carbon reductions with more pronounced overall outcomes. In contrast, industrial structure optimization, involving deep adjustments to the economic framework, is constrained by factors such as regional resource endowments and the proportion of traditional heavy industries. This process requires longer adjustment cycles, leading to relatively delayed effects. Therefore, we recommend that governments intensify support for technological innovation, encouraging the application of digital technologies in energy conservation and environmental protection. Simultaneously, regional coordinated development should be promoted by optimizing industrial structures, with a focus on supporting green transitions in central and western regions to reduce reliance on traditional heavy industries (Shi et al., 2023).

This study finds that the relationship between the DE and CE presents an inverted “U” shape influenced by urbanization factors. With the continuous expansion of regional urban areas, the CER effect of the DE shifts from promoting regional CE to suppressing them. This conclusion is similar to the findings of Musah et al., who noted that in the early stages of urbanization, increased urban construction and energy demand lead to higher CE (Musah et al., 2021). However, as urbanization deepens, the gradual improvement of digital infrastructure enhances energy efficiency, resulting in a decline in CE. The research conclusion of this study further validates the findings of scholars such as Musah. It underscores the importance of reaching a certain level of urbanization for the DE to fully leverage its potential for CER. This relationship emphasizes the need for targeted policies that foster digital infrastructure development alongside urbanization to ensure sustainable environmental outcomes. The impact of the DE on CE exhibits an inverted “U”-shaped pattern, highlighting the phased characteristics of digital economic development during urbanization. Its carbon reduction potential can only be fully realized after reaching a certain level of urbanization. Therefore, we recommend implementing phased, differentiated policies. In the early stages of urbanization, efforts should focus on guiding low-carbon city construction and enhancing the application of clean energy and green building technologies. In the mid-to-late stages, investments in digital infrastructure should be increased to improve energy efficiency and promote the application of digital technologies in energy management, traffic optimization, and other domains (Wu et al., 2023).

This study found that DE exhibits CER effects on both resource-based and non-resource-based provinces, with a greater effect on non-resource-based provinces. The analysis of economic level heterogeneity indicates that the CER effect of the DE is significantly higher in economically developed regions compared to less developed ones. The conclusion regarding the heterogeneity of resource endowments is supported by numerous scholars. For instance, research by Chen et al. shows that resource-based provinces rely heavily on traditional energy sources, which limits the development of the DE due to the difficulties associated with transforming their industrial structure (Chen S. et al., 2023). In contrast, non-resource-based provinces, characterized by more diversified industrial structures, are more amenable to the adoption of digital technologies, thus exhibiting stronger CER effects. In comparison to existing heterogeneity studies related to economic levels, this research further emphasizes that, despite the challenges posed by weaker economic foundations in less developed areas, increasing investment in digital infrastructure can still enhance CER effects. This is particularly true when driven by technological innovation, highlighting the potential for digital transformation to promote sustainable development even in economically disadvantaged regions. Resource-based provinces, constrained by their reliance on traditional energy and the challenges of industrial transformation, face limitations in realizing the carbon reduction potential of the DE. In contrast, economically developed regions, with robust infrastructure and abundant technological resources, excel in fostering synergy between the DE and carbon reduction. Therefore, we recommend that resource-based provinces accelerate industrial transformation, reduce dependence on traditional energy, and promote the deep integration of digital technologies with energy management. For economically underdeveloped regions, we propose increasing fiscal support and targeted investment to prioritize digital infrastructure development. Establishing technological innovation platforms is essential to narrowing regional disparities in the development of the DE and its carbon reduction effects (Zhang J. et al., 2022).

In summary, this study not only establishes a more comprehensive connection between the DE and CER but also compares different pathways and mechanisms involved. It indicates that the CER effect of DE is constrained by many factors. This further highlights the necessity for differentiated policies tailored to various regions and stages of development, ensuring that CER strategies effectively leverage the unique characteristics and challenges of each area.