Digitization Meets Energy Transition: Shaping the Future of Environmental Sustainability

Abstract

This paper presents an empirical investigation into the effects of energy transition and digitization on carbon dioxide emissions, serving as a proxy for environmental sustainability, across 28 Chinese provinces from 2000 to 2021. Utilizing both static and dynamic regression analyses, the study reveals a significant driving effect of the energy transition on environmental sustainability, primarily through a reduction in carbon dioxide emissions. Notably, this influence is more pronounced in the eastern region of China, followed by the central and western regions, indicating regional heterogeneity in the impact of the energy transition. Furthermore, digitization is found to have a substantial moderating effect, enhancing energy conservation and emission reductions. As digitization progresses, its capability to diminish the impact of energy transitions on carbon dioxide emissions becomes more apparent, particularly in the eastern region, while this effect is less marked in central and western China. The study also delves into the nonlinear relationship between energy transitions and carbon dioxide emissions, discovering that increased levels of digitization can exacerbate the negative effects of energy transitions on emissions. These findings offer valuable insights into the dynamics of energy transition and digitalization, highlighting their crucial roles in shaping environmental sustainability in China.

1. Introduction

In recent decades, China has undergone remarkable economic growth, a trajectory that has led to a dramatic increase in both its energy consumption and carbon dioxide emissions. This situation poses a significant challenge to the country’s goals for environmental sustainability. As the foremost emitter of carbon dioxide globally, China’s strategy for managing its energy resources and implementing technological innovations holds immense importance, impacting not only its own environmental health but also having far-reaching global implications. At the heart of China’s environmental policy is a crucial shift: moving away from traditional, fossil-fuel-dependent energy sources towards greener, more sustainable alternatives. This transition, essential in the realm of energy policy, is key to China’s commitment to aligning its rapid economic expansion with the principles of environmental sustainability. Effectively navigating this energy transition is vital for reducing China’s environmental impact and contributing significantly to worldwide efforts aimed at combating climate change and enhancing ecological well-being.

Recent academic research, including the works of Guo et al. [1], Razzaq et al. [2], Woon et al. [3], Cai et al. [4], and Jiang and Khan [5], has highlighted the crucial role renewable energy plays in reducing carbon emissions. Yet, as researchers such as Chai et al. [6], Yu et al. [7], Feng et al. [8], Cheng et al. [9], and Fang [10] have observed, the adoption and effectiveness of renewable energy strategies significantly vary across the diverse regions of China. This regional disparity underscores the need for a detailed investigation into the specific ways energy transition processes are being implemented in different parts of the country. Furthermore, the role of digital technology in this context is gaining increasing importance. Cutting-edge developments like smart grid systems and AI-driven energy management techniques are becoming essential for improving energy efficiency and lowering emissions. Studies by Lin and Huang [11], Du et al. [12], Dong and Ullah [13], Tan et al. [14], Li et al. [15], and Sajjad et al. [16] explore how digitalization optimizes energy consumption and aids in integrating renewable energy sources. The synergy between digital technology and energy transition strategies, especially in the pursuit of environmental sustainability, is an emerging area of study with significant implications. In light of this, a comprehensive examination of the effects of both energy transition and digitalization on carbon dioxide emissions across 28 Chinese provinces from 2000 to 2021 is of paramount importance. This research is vital to unraveling the complex interplay between these factors and providing essential insights for policy-making that support China’s dedication to environmental sustainability. Such a study is not only crucial for deepening our understanding of these interactions but also for formulating strategies that align with the overarching goals of reducing environmental impact and promoting sustainable development.

Expanding upon the established context, this study extends the current understanding by analyzing the impacts of energy transition and digitalization on carbon dioxide emissions, a key measure of environmental sustainability, across 28 Chinese provinces from 2000 to 2021. Utilizing both static and dynamic regression techniques, the study finds that the energy transition significantly contributes to environmental sustainability, primarily by reducing carbon dioxide emissions. This reduction is most pronounced in the eastern region of China, with notable but varying effects observed in the central and western regions, highlighting regional differences in the impact of the energy transition. The research also reveals that digitalization plays a crucial role in moderating these effects, enhancing energy efficiency, and contributing to lower emissions. Notably, as digitalization advances, it becomes increasingly effective in mitigating the negative impacts of energy transitions on carbon emissions, a trend that is more evident in the eastern region compared to the central and western regions. Furthermore, the study investigates the complex, nonlinear relationships between energy transitions and carbon dioxide emissions, discovering that elevated levels of digitalization can potentially amplify the negative effects of energy transitions on emissions. These findings elucidate the intricate interplay between energy transition and digitalization, emphasizing their vital roles in promoting environmental sustainability in the Chinese context.

Building on the derived conclusions, this study makes three key contributions to the field. First, it extends the scope of existing research, such as that by Yang et al. [17], Zheng et al. [18], and Wang and Yu [19], by providing a comprehensive empirical analysis of the impacts of energy transition and digitalization on carbon dioxide emissions across 28 Chinese provinces. This regional analysis enriches our understanding of the varying impacts of the energy transition, which is crucial for creating tailored policies in different geographical areas. Second, the research underscores the significant moderating role of digitalization in the relationship between energy transitions and emissions, moving beyond the initial findings of studies like Zhao et al. [20], Wu et al. [21], and Ren et al. [22]. It illustrates how digital technologies not only enhance energy efficiency but also actively influence energy transition–emission dynamics. Finally, unlike the linear approaches commonly found in studies such as those by Dong et al. [23], Gu et al. [24], and Wang et al. [25], this study explores the nonlinear interactions between energy transitions and carbon dioxide emissions. This approach offers a more intricate view of how escalating digitalization levels might amplify the effects of energy transitions. Collectively, these contributions significantly deepen our understanding of the complex interplay between energy transition, digitalization, and environmental sustainability, paving the way for new, practical approaches to achieving sustainable development goals.

Based on the above analysis, this paper proposes the following five hypotheses: Hypothesis 1: the energy transition has a positive impact on environmental sustainability. Hypothesis 2: there is heterogeneity in the impact of the energy transition on environmental sustainability. Hypothesis 3: digitization can moderate the relationship between energy transition measures and environmental sustainability. Hypothesis 4: there is heterogeneity in the moderating effect of digitization on the relationship between energy transition measures and environmental sustainability. Hypothesis 5: there is a nonlinear relationship between energy transition and environmental sustainability.

This paper is structured as follows: the second section provides a comprehensive review of existing literature, delving into prior research relevant to this field. The third section details the research methodology employed in this study, focusing on the selection of variables and the underlying theoretical framework. The fourth section presents a detailed analysis of the empirical results obtained from the study. Finally, the fifth section summarizes the key findings, suggests policy recommendations, and highlights areas for further research within this domain.

2. Literature Review

The scholarly discourse surrounding carbon dioxide emissions is extensive and includes a significant focus on the impact of energy transitions. Fundamental research in this area, as conducted by Mostafaeipour et al. [26], Raihan et al. [27], Adams and Nsiah [28], Saidi and Omri [29], and Sovacool et al. [30], has laid the groundwork by establishing the core relationship between the shift to renewable energy sources like solar and wind power and the reduction in carbon emissions. These studies consistently suggest that such transitions inherently contribute to lower emissions. Further building on these findings, specialized research, such as that by Hoicka et al. [31], Cantarero [32], Sillak et al. [33], and Levenda et al. [34], delves into specific energy transition strategies, emphasizing the efficacy of new renewable technologies, particularly in urban settings. However, this narrative is not without its complexities, as indicated by Ansari and Holz [35], Bridge and Gailing [36], Adewuyi et al. [37], Erat et al. [38], and Gürsan and de Gooyert [39]. Their research points to significant transition challenges, including economic and infrastructural hurdles, in regions heavily dependent on fossil fuels. Additionally, the role of technological innovation in enabling energy transitions has been a prominent theme in recent literature. Studies like those by Shaqsi et al. [40], Trahey et al. [41], Olabi et al. [42], and Hassan et al. [43] explore the potential of advancements in energy storage and efficiency to accelerate emission reductions. Yet, as highlighted by Hainsch et al. [44], Mason-D’Croz et al. [45], and Perri et al. [46], technological advancement alone may not be sufficient without corresponding shifts in policy frameworks and consumer behavior, a concept further analyzed by Malinauskaite et al. [47], Economidou et al. [48], and Thomas and Rosenow [49] in the context of energy policies in Europe.

The literature also explores other dimensions of the energy transition. For instance, Balta-Ozkan et al. [50], Swilling et al. [51], Edomah et al. [52], and Pregger et al. [53] examine the socio-economic implications of these transitions, while Carley and Konisky [54], and Sayed et al. [55] assess their broader environmental impacts. Höysniemi [56], Svobodova et al. [57], and Ankrah et al. [58]’s investigation into the geopolitical aspects adds another layer of complexity to the energy transition landscape. Furthermore, studies like those by Lamnatou et al. [59], Giannelos et al. [60], and Dileep [61] highlight the integral role of digital technologies, such as smart grids, in optimizing energy use, although Ivanova et al. [62], Borle et al. [63], and Rolandi et al. [64] caution that the impacts of digitalization can vary across different socio-economic settings. The financial aspects, including investment strategies and market dynamics, are explored in depth by Qadir et al. [65], Capellán-Pérez et al. [66], and Tagliapietra et al. [67], offering insight into the economic factors influencing the adoption of renewable energy. In summary, while the existing literature solidly supports the role of energy transitions in reducing carbon emissions, it also underscores the complexity and multifaceted nature of these transitions. The interplay of technological advancements, policy support, socio-economic considerations, and geopolitical dynamics all significantly impact the effectiveness and pace of the shift to renewable energy sources. This holistic understanding is essential for policymakers and stakeholders who are tasked with developing strategies that are environmentally sound, economically feasible, and socially inclusive.

Research spearheaded by Shahbaz et al. [68], Ma et al. [69], Kwilinski et al. [70], and Wu et al. [71] delves into digitalization’s role in energy transitions, underscoring the significant impact of digital technologies in boosting energy efficiency and reducing emissions. However, Ferrari et al. [72], Małkowska et al. [73], and Wang et al. [74] highlight that the impact of digitalization varies greatly, influenced by the distinct socio-economic contexts of different regions. Furthermore, Wang et al. [75], Pingkuo and Huan [76], and Krumm et al. [77] provide an in-depth analysis of the economic aspects of energy transitions, examining the financial drivers and market dynamics that shape the adoption of renewable energy and its subsequent effects on carbon emissions. The social dimensions of energy transitions, explored by Nielsen et al. [78] and Kaffashi and Shamsudin [79], emphasize the vital role of societal acceptance and behavioral change in achieving meaningful reductions in emissions. This notion is expanded upon in studies by Beauchampet and Walsh [80], Newell et al. [81], and Chilvers et al. [82], which focus on how public engagement and community involvement can expedite the transition process. Additionally, Cambini et al. [83], Andersen et al. [84], Munro and Cairney [85], and Li and Taeihagh [86] investigate the interplay between policy and technology in energy transitions, analyzing how governmental policies can either facilitate or impede the integration of innovative energy technologies. Chen et al. [87], Erin Bass and Grøgaard [88], and Isoaho and Karhunmaa [89] shift the focus to the long-term environmental implications of energy transitions, shedding light on the sustainability of current renewable energy approaches. The importance of international cooperation in this domain is underscored by Campos and Marín-González [90] and Kabeyi and Olanrewaju [91], who point out how global collaborations can aid in sharing technologies and best practices.

Conversely, Blondeel et al. [92], Scholten et al. [93], and Bricout et al. [94] present a critical perspective on the geopolitical challenges inherent in the global energy landscape. The influence of energy transitions on industrial sectors, as investigated by Khan et al. [95], York and Bell [96], Neofytou et al. [97], Smith et al. [98], and Jebli et al. [99], highlights how industries are adapting to new energy sources and the environmental consequences of these changes. Shan et al. [100], Razmjoo et al. [101], Flores-Granobles and Saeys [102], Chien et al. [103], Godil et al. [104], and Khattak et al. [105] concentrate on the potential of emerging renewable energy technologies, examining innovative solutions and their ability to reduce emissions. Finally, comprehensive syntheses by Green and Gambhir [106], Johnson et al. [107], Bolwig et al. [108], and Meadowcroft and Rosenbloom [109] integrate these diverse viewpoints, suggesting that the success of energy transitions in mitigating carbon emissions is contingent upon a complex combination of technological advancements, economic factors, policy initiatives, and societal engagement. In summary, these varied studies collectively offer a holistic perspective on the impact of energy transitions on carbon dioxide emissions. They underscore the necessity for a multi-dimensional approach that encompasses technological innovation, economic considerations, policy formulation, and community participation, all of which are crucial for effectively reducing carbon emissions and advancing environmental sustainability.

3. Materials and Methods

3.1. Materials

Dependent variable: In the context of environmental sustainability in China, carbon dioxide emissions emerge as a pivotal indicator, reflecting the intricate interplay between rapid industrialization, technological advancement, and environmental policy. Over the period from 2000 to 2021, empirical investigations across China’s 28 provinces have consistently underscored the utility of CO₂ emissions as a barometer for gauging environmental impacts amidst diverse economic trajectories. This period, marked by substantial economic growth and technological evolution, particularly in the digital realm, has been analyzed in contemporary literature, revealing the multifaceted dimensions of China’s environmental stewardship. Ran et al. [110], Xiao [111], and Teng et al. [112] delved into the ramifications of digitalization and industrial modernization on carbon emission dynamics, underscoring the dual role of digitalization in both exacerbating and mitigating environmental impacts through improved efficiency and the transition to greener industrial practices. Concurrently, Wen and Jiang [113], Guo et al. [114], Nie et al. [115], and Zhao et al. [116] provided a better understanding of how carbon emissions are intricately linked to policy reforms, factor misallocation, and the efficacy of market-based mechanisms in steering China towards its carbon neutrality goals. These scholarly contributions illuminate the complex yet discernible patterns of China’s environmental evolution, establishing carbon dioxide emissions not merely as a metric of pollution but as a critical lens through which the broader narrative of environmental sustainability and economic transformation in China can be comprehensively understood.

Independent variable: In the landscape of contemporary China, the metric of fixed broadband subscriptions per 100 people has emerged as a robust indicator of digitization, encapsulating the nuances of technological penetration and socioeconomic transformation. The burgeoning body of scholarly work, with studies such as those by Zhang et al. [117], Su et al. [118], and He et al. [119], underscored the profound implications of broadband access in catalyzing the digital economy, highlighting its role in bridging information asymmetries and fostering market competition. These insights were complemented by research by Fu et al. [120], Wang and Cen [121], Liu et al. [122], and Wang and Li [123], which elucidated the transformative impact of widespread digital connectivity on industry, governance, and social engagement, illustrating how broadband access served as a cornerstone of digital infrastructure. In the context of China, a nation characterized by rapid urbanization and industrial modernization, the proliferation of fixed broadband subscriptions not only mirrors the digital leap but also serves as a barometer for assessing the depth and reach of digital integration across diverse provinces and socio-economic strata. This quantitative metric, thus, transcends its primary function as a measure of internet penetration, becoming a multifaceted indicator that reflects the broader contours of China’s journey towards an interconnected, digitally empowered society, where access to digital tools and resources plays a critical role in shaping economic opportunities, governance models, and societal interactions.

The proportion of renewable energy consumption within the overall energy mix stands as a critical metric in assessing China’s energy transition, a phenomenon that is deeply intertwined with global sustainability efforts and the nation’s commitment to combating climate change. In the realm of academic discourse, this indicator has been extensively scrutinized. Zhao et al. [124], Chen et al. [125], and Zhu et al. [126] provided valuable insights into the dynamics of China’s energy restructuring. They highlighted how the rising share of renewables, driven by policy initiatives and technological innovation, encapsulated the shift away from traditional fossil fuels towards cleaner energy sources. Complementing this perspective, Huang [127], Dong et al. [128], and Sun et al. [129] delved into the multifaceted implications of this energy transition, examining its impact on economic development, environmental sustainability, and social welfare. These scholarly works, in conjunction with empirical data from China’s diverse provinces, illustrated a palpable shift in the energy paradigm, marked by increasing investments in solar, wind, and hydroelectric power. The trajectory of renewable energy consumption thus emerges not only as an indicator of China’s commitment to its “peak carbon emissions by 2030 and carbon neutrality by 2060” goals but also as a reflection of broader global trends towards sustainable energy practices. This transition, pivotal for both environmental health and economic vitality, underscores the importance of renewable energy consumption as a key indicator for evaluating and understanding the depth and effectiveness of China’s energy transformation in the context of global climate objectives.

Control variable: In the realm of assessing the impact of digitization and energy transition on carbon dioxide emissions in China, the incorporation of a comprehensive set of control variables is paramount to yielding a nuanced and accurate analysis. Contemporary academic literature, spanning a broad spectrum of disciplines, underscores the significance of integrating variables such as provincial GDP, technology development, foreign direct investment, trade openness, urbanization, and industrialization into the analytical framework. Llorca and Meunié [130], Hung [131], Adebayo et al. [132], and Hao and Peng [133] highlighted the necessity of considering provincial GDP and foreign direct investment, as these economic indicators not only mirrored the financial health and investment landscapes of regions but also profoundly influenced their energy consumption and emission profiles. Concurrently, Li et al. [134], Pata and Caglar [135], and Ang [136] underscored the role of technological advancement and trade openness, elucidating how these factors facilitated or impeded the adoption of sustainable practices and green technologies, thereby impacting emissions. The dynamics of urbanization and industrialization, as explored in the works of Liu and Bae [137], Xu and Lin [138], Xu et al. [139], and Cheng and Hu [140], also played a pivotal role in shaping energy demand patterns and carbon footprints, reflecting the transformation of societal and industrial structures. Furthermore, the inclusion of these control variables, as advocated in studies by Qi et al. [141], Hou et al. [142], Chen et al. [143], Tan et al. [144], Han et al. [145], and Zhou et al. [146], allowed for a more refined dissection of the distinct effects of digitization and energy transition initiatives. This multivariate approach enables researchers to differentiate the specific influences of these initiatives from the broader economic and developmental trajectories, offering a clearer understanding of their respective roles in shaping China’s carbon dioxide emissions landscape amidst its ongoing economic evolution and environmental sustainability efforts.

This article sources its data from various authoritative entities, including the National Bureau of Statistics, the Statistical Bulletin on China’s Outward Foreign Direct Investment, the National Intellectual Property Office, and the annual statistical reports from each of the provinces. For an enhanced understanding of the variables used in this research, a detailed overview of their essential attributes is presented in

Table 1. Results of variable description.

Variable	Form	Definition
Carbon dioxide emissions	$\mathrm{c}\mathrm{a}\mathrm{r}$	Carbon dioxide emissions (kt) in log
Digitization	$\mathrm{d}\mathrm{i}\mathrm{g}$	Fixed broadband subscriptions per 100 people in log
Energy transition	$\mathrm{e}\mathrm{n}\mathrm{e}$	Proportion of renewable energy consumption within the overall energy mix
Gross domestic product	$\mathrm{g}\mathrm{r}\mathrm{o}$	Gross domestic product in log
Technology development	$\mathrm{t}\mathrm{e}\mathrm{c}$	Number of patents granted in log
Foreign direct investment	$\mathrm{f}\mathrm{d}\mathrm{i}$	Net inflow of foreign direct investment (% GDP)
Trade openness	$\mathrm{t}\mathrm{r}\mathrm{a}$	Ratio of total imports and exports to GDP
Urbanization	$\mathrm{u}\mathrm{r}\mathrm{b}$	Share of the urban population in the total population
Industrialization	$\mathrm{i}\mathrm{n}\mathrm{d}$	Ratio of industry to GDP

.

3.2. Econometric Methods

The first step involves using a province–year fixed effects model to analyze the impact of the energy transition on environmental sustainability. The second step confirms these results using a system GMM model. The third step explores the heterogeneity of the impact of the energy transition on environmental sustainability across China’s northeastern, central, and western regions using a province–year fixed effects model. The fourth step elucidates the role of digital technology and innovation in altering energy consumption patterns and their associated environmental impacts. The fifth step examines the differential effects produced by the interaction between digitization and energy transition in the northeastern, central, and western regions of China. The sixth step employs threshold regression to capture the nonlinear relationship between energy transition and carbon dioxide emissions.

In the analytical pursuit to delineate the influence of digitization and energy transition on carbon dioxide emissions across Chinese provinces, the employment of a year-and-province fixed effect model stands as a methodologically robust approach. This model, informed by Zhu et al. [147], Zheng and Li [148], Meng et al. [149], and Fan et al. [150], adeptly accounts for both temporal and spatial heterogeneities, thereby enhancing the validity of the analysis. By incorporating fixed effects for both time and province, the model effectively isolates the unique impact of digitization and energy transition, controlling for unobserved, time-invariant characteristics unique to each province and common temporal shocks affecting all provinces. This approach is particularly pertinent, as underscored by Yu et al. [151], Tang et al. [152], and He and Su [153], given the disparate stages of digital and energy infrastructure development across China’s provinces. The validity of this dual fixed effect model lies in its ability to provide a more accurate and nuanced estimation of the targeted effects, free from the confounding influences of inter-provincial differences and overarching temporal trends. Its appropriateness is further reinforced by its alignment with the actual situation of China’s provinces, which exhibit significant diversity in terms of economic development, technological advancement, and environmental policies. Consequently, this model is not only apt for the analysis but also vital for deriving meaningful insights into the complex dynamics of how digitization and the energy transition are shaping carbon dioxide emissions in the context of China’s environmental and economic objectives.

The baseline model is shown as follows:

$$\begin{gathered} {\mathrm{c}\mathrm{a}\mathrm{r}}_{\mathrm{i},\mathrm{t}}={\mathrm{a}}_0+{\mathrm{a}}_1{\mathrm{e}\mathrm{n}\mathrm{e}}_{\mathrm{i},\mathrm{t}}+{\mathrm{a}}_2{\mathrm{g}\mathrm{r}\mathrm{o}}_{\mathrm{i},\mathrm{t}}+{\mathrm{a}}_3{\mathrm{t}\mathrm{e}\mathrm{c}}_{\mathrm{i},\mathrm{t}}+{\mathrm{a}}_4{\mathrm{f}\mathrm{d}\mathrm{i}}_{\mathrm{i},\mathrm{t}}+{\mathrm{a}}_5{\mathrm{t}\mathrm{r}\mathrm{a}}_{\mathrm{i},\mathrm{t}}+{\mathrm{a}}_6{\mathrm{u}\mathrm{r}\mathrm{b}}_{\mathrm{i},\mathrm{t}}+ \\ {\mathrm{a}}_7{\mathrm{i}\mathrm{n}\mathrm{d}}_{\mathrm{i},\mathrm{t}}+{\mathsf{\mu }}_{\mathrm{i}}+{\mathsf{\eta }}_{\mathrm{t}}+{\mathsf{ϵ}}_{\mathrm{i},\mathrm{t}}. \end{gathered}$$

(1)

Within the framework of Equation (1), $\mathrm{i}$ represents the variable for provinces, while $\mathrm{t}$ denotes the variable for years. ${\mathrm{a}}_0$ signifies a constant value. The coefficients that require estimation are indicated by $[{\mathrm{a}}_1,{\mathrm{a}}_7]$. $\mathsf{\mu }$ is used to denote the fixed effect specific to each province; $\mathsf{\eta }$ corresponds to the fixed effect associated with each year; and $\mathsf{ϵ}$ refers to the error term, characterized as white noise. To address the potential influence of historical carbon dioxide emissions on present and future emission levels and acknowledge the propensity for the lag term to exhibit correlation with the model’s residual components, this study integrates a lagged variable of carbon dioxide emissions to construct a dynamic panel model. The inclusion of this lagged carbon dioxide emissions term within the set of explanatory variables often presents a correlation with unobserved heterogeneity effects that are specific to different cross-sections. Furthermore, the relatively limited scope of control variables incorporated into the model raises concerns about the possible exclusion of key explanatory factors. Such omissions can induce a correlation between the explanatory variables and the error terms, thereby introducing endogeneity into the model, as highlighted by Hsiao [154]. To counteract these issues and robustly estimate the model, the system generalized method of moments is employed. This methodological choice is reinforced by the utilization of the Arellano–Bond test and the Hansen test, which serve to validate the model’s integrity and the appropriateness of the instrumental variables used, ensuring that the model adequately addresses potential endogeneity concerns and accurately captures the dynamics of carbon emissions in relation to the factors under study. The formulated system generalized method of moments model is presented as:

$$\begin{gathered} {\mathrm{c}\mathrm{a}\mathrm{r}}_{\mathrm{i},\mathrm{t}}={\mathrm{b}}_0+{\mathrm{b}}_1{\mathrm{d}\mathrm{i}\mathrm{g}}_{\mathrm{i},\mathrm{t}}+{\mathrm{b}}_2{\mathrm{e}\mathrm{n}\mathrm{e}}_{\mathrm{i},\mathrm{t}}+{\mathrm{b}}_3{\mathrm{g}\mathrm{r}\mathrm{o}}_{\mathrm{i},\mathrm{t}}+{\mathrm{b}}_4{\mathrm{t}\mathrm{e}\mathrm{c}}_{\mathrm{i},\mathrm{t}}+{\mathrm{b}}_5{\mathrm{f}\mathrm{d}\mathrm{i}}_{\mathrm{i},\mathrm{t}}+{\mathrm{b}}_6{\mathrm{t}\mathrm{r}\mathrm{a}}_{\mathrm{i},\mathrm{t}}+ \\ {\mathrm{b}}_7{\mathrm{u}\mathrm{r}\mathrm{b}}_{\mathrm{i},\mathrm{t}}+{\mathrm{b}}_8{\mathrm{i}\mathrm{n}\mathrm{d}}_{\mathrm{i},\mathrm{t}}+{\mathsf{ϵ}}_{\mathrm{i},\mathrm{t}}. \end{gathered}$$

(2)

$$\begin{gathered} {\mathrm{c}\mathrm{a}\mathrm{r}}_{\mathrm{i},\mathrm{t}}-{\mathrm{c}\mathrm{a}\mathrm{r}}_{\mathrm{i},\mathrm{t}-\overline{\mathsf{\omega }}}={\mathrm{b}}_1({\mathrm{c}\mathrm{a}\mathrm{r}}_{\mathrm{i},\mathrm{t}-\overline{\mathsf{\omega }}}-{\mathrm{c}\mathrm{a}\mathrm{r}}_{\mathrm{i},\mathrm{t}-\overline{2\mathsf{\omega }}})+{\mathrm{b}}_2({\mathrm{e}\mathrm{n}\mathrm{e}}_{\mathrm{i},\mathrm{t}-\overline{\mathsf{\omega }}}-{\mathrm{e}\mathrm{n}\mathrm{e}}_{\mathrm{i},\mathrm{t}-\overline{2\mathsf{\omega }}})+ \\ {\mathrm{b}}_3({\mathrm{g}\mathrm{r}\mathrm{o}}_{\mathrm{i},\mathrm{t}-\overline{\mathsf{\omega }}}-{\mathrm{g}\mathrm{r}\mathrm{o}}_{\mathrm{i},\mathrm{t}-\overline{2\mathsf{\omega }}})+{\mathrm{b}}_4(\mathrm{l}\mathrm{o}\mathrm{g}{\mathrm{t}\mathrm{e}\mathrm{c}}_{\mathrm{i},\mathrm{t}-\mathsf{\varpi }}-\mathrm{l}\mathrm{o}\mathrm{g}{\mathrm{t}\mathrm{e}\mathrm{c}}_{\mathrm{i},\mathrm{t}-2\mathsf{\varpi }})+{\mathrm{b}}_5({\mathrm{f}\mathrm{d}\mathrm{i}}_{\mathrm{i},\mathrm{t}-\overline{\mathsf{\omega }}}- \\ {\mathrm{f}\mathrm{d}\mathrm{i}}_{\mathrm{i},\mathrm{t}-\overline{2\mathsf{\omega }}})+{\mathrm{b}}_6({\mathrm{t}\mathrm{r}\mathrm{a}}_{\mathrm{i},\mathrm{t}-\overline{\mathsf{\omega }}}-{\mathrm{t}\mathrm{r}\mathrm{a}}_{\mathrm{i},\mathrm{t}-\overline{2\mathsf{\omega }}})+{\mathrm{b}}_7({\mathrm{u}\mathrm{r}\mathrm{b}}_{\mathrm{i},\mathrm{t}-\overline{\mathsf{\omega }}}-{\mathrm{u}\mathrm{r}\mathrm{b}}_{\mathrm{i},\mathrm{t}-\overline{2\mathsf{\omega }}})+{\mathrm{b}}_8({\mathrm{i}\mathrm{n}\mathrm{d}}_{\mathrm{i},\mathrm{t}-\overline{\mathsf{\omega }}}- \\ {\mathrm{i}\mathrm{n}\mathrm{d}}_{\mathrm{i},\mathrm{t}-\overline{2\mathsf{\omega }}})+{\mathrm{b}}_9{\mathrm{d}\mathrm{i}\mathrm{g}}_{\mathrm{i},\mathrm{t}-\overline{\mathsf{\omega }}}+{\mathrm{b}}_{10}{\mathrm{e}\mathrm{n}\mathrm{e}}_{\mathrm{i},\mathrm{t}-\overline{\mathsf{\omega }}}+{\mathrm{b}}_{11}{\mathrm{g}\mathrm{r}\mathrm{o}}_{\mathrm{i},\mathrm{t}-\overline{\mathsf{\omega }}}+{\mathrm{b}}_{12}{\mathrm{t}\mathrm{e}\mathrm{c}}_{\mathrm{i},\mathrm{t}-\overline{\mathsf{\omega }}}+{\mathrm{b}}_{13}{\mathrm{f}\mathrm{d}\mathrm{i}}_{\mathrm{i},\mathrm{t}-\overline{\mathsf{\omega }}}+ \\ {\mathrm{b}}_{14}{\mathrm{t}\mathrm{r}\mathrm{a}}_{\mathrm{i},\mathrm{t}-\overline{\mathsf{\omega }}}+{\mathrm{b}}_{15}{\mathrm{u}\mathrm{r}\mathrm{b}}_{\mathrm{i},\mathrm{t}-\overline{\mathsf{\omega }}}+{\mathrm{b}}_{16}{\mathrm{i}\mathrm{n}\mathrm{d}}_{\mathrm{i},\mathrm{t}-\overline{\mathsf{\omega }}}+({\mathsf{\delta }}_{\mathrm{t}}+{\mathsf{\delta }}_{\mathrm{t}-2\mathsf{\varpi }})+({\mathsf{ϵ}}_{\mathrm{i},\mathrm{t}}-{\mathsf{ϵ}}_{\mathrm{i},\mathrm{t}-2\mathsf{\varpi }}). \end{gathered}$$

(3)

To scrutinize the potential moderating influence that digitization exerts on the relationship between energy transition and carbon dioxide emissions, this study designs a specific analytical model. Within this model, $\mathrm{d}\mathrm{i}\mathrm{g}$ is designated to represent the digitization variable. This approach allows for an in-depth examination of how advancements in digital technologies and infrastructure might alter or influence the effectiveness of energy transition strategies in reducing carbon emissions. The model aims to disentangle the direct impacts of the energy transition on emissions from the interactive effects that digitization might have on this relationship. By incorporating digitization as a key variable, the model provides insights into whether and how digitalization enhances or diminishes the efficacy of energy transition measures in the context of mitigating carbon dioxide emissions. This analysis is crucial for understanding the synergies and trade-offs between digitization and energy transition initiatives, offering valuable perspectives for policy formulation and strategic planning in the realm of environmental sustainability.

$$\begin{gathered} {\mathrm{c}\mathrm{a}\mathrm{r}}_{\mathrm{i},\mathrm{t}}={\mathrm{c}}_0+{\mathrm{c}}_1{\mathrm{e}\mathrm{n}\mathrm{e}}_{\mathrm{i},\mathrm{t}}+{\mathrm{c}}_2{\mathrm{d}\mathrm{i}\mathrm{g}}_{\mathrm{i},\mathrm{t}}+{\mathrm{c}}_3{\mathrm{e}\mathrm{n}\mathrm{e}}_{\mathrm{i},\mathrm{t}}·{\mathrm{d}\mathrm{i}\mathrm{g}}_{\mathrm{i},\mathrm{t}}+{\mathrm{c}}_4{\mathrm{g}\mathrm{r}\mathrm{o}}_{\mathrm{i},\mathrm{t}}+{\mathrm{a}}_5{\mathrm{t}\mathrm{e}\mathrm{c}}_{\mathrm{i},\mathrm{t}}+{\mathrm{a}}_6{\mathrm{f}\mathrm{d}\mathrm{i}}_{\mathrm{i},\mathrm{t}}+ \\ {\mathrm{a}}_7{\mathrm{t}\mathrm{r}\mathrm{a}}_{\mathrm{i},\mathrm{t}}+{\mathrm{a}}_8{\mathrm{u}\mathrm{r}\mathrm{b}}_{\mathrm{i},\mathrm{t}}+{\mathrm{a}}_9{\mathrm{i}\mathrm{n}\mathrm{d}}_{\mathrm{i},\mathrm{t}}+{\mathsf{ϵ}}_{\mathrm{i},\mathrm{t}}. \end{gathered}$$

(4)

Evaluating the role of digitization as a moderating factor in the influence of energy transition strategies on carbon dioxide emissions hinges on the analysis of the interaction coefficient’s significance within the model. This interaction coefficient, derived from the interplay between energy transition and digitization, is critical in determining the extent to which digital advancements can affect the impact of energy transition efforts on emissions levels. However, when the model includes both the interaction term and the primary variables of energy transition and digitization concurrently, there is a potential risk of encountering collinearity issues, which can skew the regression outcomes and lead to inaccurate interpretations. To mitigate this risk and enhance the reliability of the model’s findings, the interaction terms in model (4) undergo a process of centralization. This methodological step is essential for reducing multicollinearity concerns, ensuring that the model accurately captures the distinct effects of the interaction between energy transition and digitization on carbon dioxide emissions and thus providing a more precise and trustworthy understanding of how these two critical factors interrelate within the broader context of environmental policy and sustainability efforts.

The analysis of regression outcomes pertaining to the moderating impact of digitization on the relationship between energy transition and carbon dioxide emissions reveals regional variations in this effect. This observation leads to the hypothesis that the influence of digitization on this relationship may be nonlinear. In other words, as the level of digitization progresses, its interplay with the energy transition and consequent effects on carbon dioxide emissions do not follow a consistent, linear pattern. Instead, these interactions might manifest differently across various stages of digital development. Drawing inspiration from the works of Wang et al. [155], Sung [156], and Xu et al. [157], which explored similar complex dynamics, this study proposes the formulation of model (5). This model is designed to capture and quantify the potentially nonlinear dynamics between digitization, energy transition, and carbon emissions, factoring in the varying levels of digital advancement. By doing so, it aims to provide a more nuanced understanding of how digitization’s role evolves and influences the efficacy of energy transition strategies in reducing carbon emissions, particularly in varying regional contexts and stages of digital maturity. This approach is crucial for tailoring effective environmental policies that are responsive to the specific digital landscapes and energy transition phases of different regions. The model is shown as follows:

$$\begin{gathered} {\mathrm{c}\mathrm{a}\mathrm{r}}_{\mathrm{i},\mathrm{t}}={\mathrm{d}}_0+{\mathrm{d}}_1{\mathrm{e}\mathrm{n}\mathrm{e}}_{\mathrm{i},\mathrm{t}}·\mathrm{F}({\mathrm{d}\mathrm{i}\mathrm{g}}_{\mathrm{i},\mathrm{t}}\le \mathrm{q})+{\mathrm{d}}_2{\mathrm{e}\mathrm{n}\mathrm{e}}_{\mathrm{i},\mathrm{t}}·\mathrm{F}({\mathrm{d}\mathrm{i}\mathrm{g}}_{\mathrm{i},\mathrm{t}}\ge \mathrm{q})+{\mathrm{d}}_3{\mathrm{g}\mathrm{r}\mathrm{o}}_{\mathrm{i},\mathrm{t}}+ \\ {\mathrm{d}}_4{\mathrm{t}\mathrm{e}\mathrm{c}}_{\mathrm{i},\mathrm{t}}+{\mathrm{d}}_5{\mathrm{f}\mathrm{d}\mathrm{i}}_{\mathrm{i},\mathrm{t}}+{\mathrm{d}}_6{\mathrm{t}\mathrm{r}\mathrm{a}}_{\mathrm{i},\mathrm{t}}+{\mathrm{d}}_7{\mathrm{u}\mathrm{r}\mathrm{b}}_{\mathrm{i},\mathrm{t}}+{\mathrm{d}}_8{\mathrm{i}\mathrm{n}\mathrm{d}}_{\mathrm{i},\mathrm{t}}+{\mathsf{\mu }}_{\mathrm{i}}+{\mathsf{ϵ}}_{\mathrm{i},\mathrm{t}}. \end{gathered}$$

(5)

In Equation (5), $\mathrm{F}(·)$ represents an indicator function, where $\mathrm{q}$ denotes a predetermined threshold value. The roles of these elements are integral to the model’s framework. The indicator function is employed to demarcate specific conditions or states within the model based on the threshold value. This threshold acts as a critical pivot, defining the point at which certain effects or interactions within the model become significant or change in nature. All other variables within this setup retain their definitions and roles as previously established in the model. The use of such an indicator function, particularly in conjunction with a defined threshold, allows for a more nuanced analysis, especially in scenarios where relationships or impacts within the data are not linear or straightforward. It enables the model to adaptively respond to varying levels of the variables under study, thereby providing a more dynamic and accurate representation of the underlying phenomena being investigated.

4. Results and Discussion

4.1. The Effect of Energy Transition on Environmental Sustainability

In this subsection, our analysis focuses on evaluating the impact of energy transitions on carbon dioxide emissions, which serve as a key indicator for assessing environmental sustainability. We have incorporated both regional (province-level) variations and temporal dynamics into our model to ensure a better understanding of the trend in carbon dioxide emissions. By controlling for these provincial and temporal factors, we aim to isolate the specific influence of the energy transition on carbon dioxide emissions. Moreover, in our research, three distinct econometric models have been developed to assess the effects of energy transitions on carbon dioxide emissions. The first model, referred to as Model 1, offers a baseline estimation by analyzing the relationship between energy transition and carbon dioxide emissions without the inclusion of any control variables. This provides an initial, unadjusted view of the impact. Model 2, on the other hand, enriches this analysis by incorporating a range of control variables. These variables account for other factors that might influence carbon dioxide emissions, thus allowing for a more nuanced understanding of the energy transition’s effects. Finally, Model 3 employs the system-generalized method of moments approach. This advanced econometric technique is particularly adept at handling potential endogeneity issues, providing a more robust and dynamic perspective on the impact of energy transitions on carbon dioxide emissions. Each model progressively builds on the complexity and depth of the analysis, offering a comprehensive view of the effects of energy transition policies on environmental outcomes. The findings of our study are detailed in

Table 2. Results of the effect of energy transition on environmental sustainability.

Variable and Model	Model (1)	Model (2)	Model (3)
$\mathrm{e}\mathrm{n}\mathrm{e}$	−0.204 *** (−5.523)	−0.117 *** (−5.907)	−0.176 *** (−5.019)
$\mathrm{g}\mathrm{r}\mathrm{o}$		0.784 *** (3.423)	0.847 *** (3.338)
$\mathrm{t}\mathrm{e}\mathrm{c}$		−0.018 * (−1.971)	−0.019 ** (−2.042)
$\mathrm{f}\mathrm{d}\mathrm{i}$		−0.068 ** (−2.065)	−0.073 ** (−2.119)
$\mathrm{t}\mathrm{r}\mathrm{a}$		−0.135 * (−1.639)	−0.122 * (−1.787)
$\mathrm{u}\mathrm{r}\mathrm{b}$		0.345 *** (2.552)	0.311 ** (2.119)
$\mathrm{i}\mathrm{n}\mathrm{d}$		0.433 * (1.828)	0.406 ** (2.044)
${\mathrm{c}\mathrm{a}\mathrm{r}}_{-1}$			0.094 *** (7.095)
$\mathrm{c}$	3.345 * (1.679)	4.045 (1.268)	5.608 (1.129)
Province-fixed effect	Yes	Yes	No
Year-fixed effect	Yes	Yes	Yes
F-test	72.883 ***	64.505 ***
AR(2)			0.875
Hansen test			12.711
${\mathrm{R}}^2$	0.314	0.312

Note: t-value in the parentheses; * a 10% significant level; ** a 5% significant level; *** a 1% significant level; Model (3) system-GMM; c constant.

.

Our analysis, as detailed in Table 2, employs three econometric models to explore the impact of energy transitions on carbon dioxide emissions. In Model 1, a foundational model without control variables, the coefficient for energy transition is −0.204. This implies a significant inverse relationship: a 1% increase in renewable energy consumption correlates with a 0.204% decrease in carbon dioxide emissions. Model 2, incorporating control variables to consider other influential factors, reveals a reduced coefficient of −0.117. Finally, Model 3 utilizes the system-GMM method, resulting in a coefficient of −0.176. Each coefficient, significant at the 1% level, strongly suggests that energy transitions inversely affect carbon dioxide emissions. The situation in China, as documented in recent studies by Wang et al. [158], Qi et al. [159], Shen et al. [160], Su and Tan [161], and Wang and He [162], reflects an aggressive move towards renewable energy, aligning with broader environmental sustainability objectives. These studies underscore China’s renewable energy expansion strategies, the role of government incentives, and the inherent challenges and opportunities inherent to this shift. This transition is central to China’s strategy for reducing its carbon footprint and addressing global climate change. Comparing our findings with those in Li and Taeihagh [163], Zhang and Chen [164], and Shen et al.’s [165] research, we find general agreement in the negative correlation between renewable energy use and carbon dioxide emissions. However, discrepancies in the magnitude of this correlation can be traced to variations in methodology, regional focus, and the timeframe of these studies. For example, works by Cheng et al. [166], Ding and Liu [167], and Zhang and Du [168] concentrate on the short-term effects of energy policies, while others assess long-term impacts. Moreover, our results highlight the vital role of renewable energy in promoting environmental sustainability. By diminishing dependence on fossil fuels and thereby reducing carbon dioxide emissions, renewable energy is pivotal in mitigating climate change and achieving a sustainable ecological equilibrium. This is in line with the broader goals of environmental sustainability, which stress the necessity of sustainable energy solutions to safeguard ecological systems, ensure the welfare of future generations, and preserve the health of our planet.

Moreover, our analysis, as detailed in Table 2, also reveals a multifaceted relationship between various economic growth indicators and carbon dioxide emissions. The study establishes a pronounced positive correlation between factors such as GDP growth, urbanization, industrialization, and the level of carbon dioxide emissions. This trend indicates that economic expansion, coupled with increased urban and industrial activities, often leads to a rise in carbon emissions. Additionally, a significant finding is the influence of historical carbon dioxide emission levels on current emissions, suggesting that past emissions have a persistent and accumulating effect on the current environmental state. In the Chinese context, this pattern has been extensively documented. Wang and Su [169], Li and Haneklaus [170], Sikder et al. [171], Bai et al. [172], and Kongkuah et al. [173] have observed that China’s rapid economic growth and urbanization have escalated energy demands, primarily met through fossil fuels, thus increasing carbon dioxide emissions. This trend is reflective of a larger pattern observed in emerging economies, where the initial stages of economic development tend to increase emissions. In contrast, our study also identifies a negative correlation between technological advancement, foreign direct investment, trade openness, and carbon dioxide emissions. This suggests that embracing advanced technologies, increasing international investments, and fostering global market integration are instrumental in reducing emissions. Recent Chinese studies, such as those by Hassan et al. [174], Liu et al. [175], Zhou et al. [176], and Xie and Teo [177], highlight how China’s investment in cleaner technologies and international trade participation have led to a shift towards less carbon-intensive industrial practices. Comparing these results with similar research, like that of Xie et al. [178], Li et al. [179], Zhang et al. [180], Chi and Meng [181], and Boamah et al. [182], we observe consensus in the positive correlation between economic growth and emissions. However, the extent of this relationship and its mitigation through technology and foreign direct investment vary. This variation could be attributed to differences in research methodologies, regional focuses, and the specific dynamics of foreign direct investment and technological advancements considered in these studies. From an environmental sustainability perspective, our findings highlight the twin challenges of sustaining economic growth and reducing environmental impacts. The positive correlation between economic growth indicators and emissions underscores the environmental cost associated with traditional development models. Conversely, the negative correlation with technological progress, foreign direct investment, and trade openness indicates potential strategies for delinking economic growth from environmental harm. These insights are crucial in devising policies that strike a balance between economic development and the pressing need for environmental sustainability, particularly in rapidly developing economies such as China. More importantly, Hypothesis 1 has been confirmed.

4.2. Heterogeneity Test Regarding the Effect of Energy Transition on Environmental Sustainability

Given the diverse stages of economic and industrial development across China, categorizing the country into eastern, central, and western regions provides a valuable framework to understand regional distinctions. This three-part division, reflective of the unique attributes of each region, is consistent with strategic objectives for regional development, a topic extensively covered in academic discussions, including those by Feng et al. [183], Zou and Pan [184], Ren et al. [185], Wu et al. [186], Fan et al. [187], Zhao et al. [188], Xie et al. [189], and Xu and Yi [190]. The relevance of this regional breakdown is particularly crucial in our examination of how energy transitions impact carbon dioxide emissions. To this end, our study employs distinct regression models for each region. The premise is that regional variations may significantly shape the relationship between energy transitions and carbon emissions. Our analyses, conducted independently for the eastern, central, and western regions, are designed to reveal the specific dynamics of energy transitions and their impact on carbon emission trends within each region. This methodology not only deepens our comprehension of the regional nuances in China’s energy transition landscape but also provides critical insights into the ongoing discourse on sustainable energy practices and their environmental ramifications in varied economic settings. The findings of this comprehensive analysis are presented in

Table 3. Results of heterogeneity test regarding the effect of energy transition on environmental sustainability.

Variable and Model	Model (4)	Model (5)	Model (6)
$\mathrm{e}\mathrm{n}\mathrm{e}$	−0.198 *** (−4.453)	−0.139 *** (−4.825)	−0.096 *** (−4.756)
$\mathrm{c}\mathrm{v}$	Yes	Yes	Yes
$\mathrm{c}$	4.033 (1.444)	4.536* (1.723)	4.402 (1.036)
Province-fixed effect	Yes	Yes	Yes
Year-fixed effect	Yes	Yes	Yes
F-test	88.725 ***	83.848 ***	78.045 ***
${\mathrm{R}}^2$	0.326	0.339	0.299
$\mathrm{Religion}$	Eastern religion	Central religion	Western religion

Note: t-value in the parentheses; * a 10% significant level; *** a 1% significant level; c constant; $\mathrm{c}\mathrm{v}$ control variable.

.

According to the findings presented in Table 3, the impact of the energy transition on carbon dioxide emissions is statistically significant and varies distinctly across the different regions of China. The data demonstrate a gradient of influence, with the most substantial impact in the eastern region, followed by the central region, and the least in the western region. This regional disparity in the effectiveness of renewable energy consumption on reducing CO₂ emissions can be linked to several key factors. Primarily, the eastern region of China, characterized by its advanced economic development and high level of urbanization, exhibits a more pronounced adoption of renewable energy technologies. This trend is evident in the significantly greater reduction in carbon emissions in this area. Research by scholars such as Wang et al. [191], Xu et al. [192], Qamruzzaman and Karim [193], Bai et al. [194], and Sun [195] emphasizes the eastern region’s sophisticated infrastructure and the robust investment, both public and private, in renewable energy. These investments have led to more efficient energy use and, consequently, a notable decrease in emissions, reflecting the effectiveness of renewable energy strategies in more developed urban settings.

In contrast to the eastern region, the central region of China, though experiencing significant development, still trails the eastern region in economic prosperity and the adoption of technological advancements. This disparity is reflected in the moderate level of renewable energy consumption’s impact on carbon emissions reduction. Studies by Yu et al. [196], Ahmad et al. [197], Zhang et al. [198], and Chen et al. [199] have documented this intermediate stage of development and its corresponding effects on environmental efforts. Meanwhile, the western region, with its relatively underdeveloped economy and continued reliance on traditional energy sources, exhibits the least impact from the renewable energy transition. Research by scholars such as Harlan [200], Wang and You [201], Li et al. [202], Xu et al. [203], and Li et al. [204] supports this observation, indicating that the slower integration of renewable energy in the western region leads to less significant reductions in carbon emissions. These regional variations in the effectiveness of energy transition strategies highlight the necessity for customized approaches to environmental sustainability in China. It is essential to implement policies tailored to the specific economic, infrastructural, and technological contexts of each region. Such regionally nuanced strategies are vital for maximizing the environmental benefits of renewable energy adoption, effectively contributing to the overarching goal of sustainable development and nationwide reductions in carbon emissions. Significantly, Hypothesis 2 has been substantiated.

4.3. Environmental Sustainability Implications of the Digitization-Led Energy Transition

The accelerated digital transformation in China is having a profound impact across various societal sectors, particularly in the realms of energy and environmental sustainability. Given this backdrop, the main aim of this subsection is to conduct a quantitative analysis of how digitalization moderates the relationship between energy transition initiatives and carbon dioxide emissions. This investigation seeks to elucidate the role of digital technologies and innovations in altering patterns of energy consumption and their associated environmental impacts. The findings of this in-depth analysis, which are crucial for informing future policy directions in sustainable energy and digital infrastructure, are detailed in

Table 4. Results of environmental implications of the digitization-led energy transition.

Variable and Model	Model (7)	Model (8)	Model (9)
$\mathrm{e}\mathrm{n}\mathrm{e}$	−0.275 *** (−3.286)	−0.225 *** (−3.653)	−0.268 *** (−3.745)
$\mathrm{d}\mathrm{i}\mathrm{g}$	0.032 (1.453)	0.029 * (1.794)	0.047 (1.302)
$\mathrm{e}\mathrm{n}\mathrm{e}\times \mathrm{d}\mathrm{i}\mathrm{g}$	−0.102 *** (−2.534)	−0.113 *** (−2.808)	−0.108 *** (−2.618)
$\mathrm{c}\mathrm{v}$	No	Yes	Yes
$\mathrm{c}$	1.074 * (1.182)	1.275 (1.494)	1.659 (1.106)
F-test	75.874 ***	71.724 ***
AR(2)			0.673
Hansen test			13.363
${\mathrm{R}}^2$	0.213	0.331
Province-fixed effect	Yes	Yes	Yes
Year-fixed effect	Yes	Yes	No

Note: t-value in the parentheses; * a 10% significant level; *** a 1% significant level; Model (9) system-GMM; c constant; $\mathrm{c}\mathrm{v}$ control variable.

.

The analysis derived from Table 4 sheds light on the dynamic relationship between energy transition and digitalization within the context of China. The interaction terms between these two variables display significant negative regression coefficients, indicating a substantial moderating effect of digitalization on the relationship between energy transition initiatives and carbon dioxide emissions. This effect is a testament to the transformative role of digital technologies in China’s energy sector, as outlined in recent scholarly publications by Li et al. [205], Lyu and Liu [206], and Zhang et al. [207]. The advancements in digital technologies, including big data, artificial intelligence, and the Internet of Things (IoT), have been instrumental in revolutionizing China’s approach to energy consumption and emission management. For example, the implementation of IoT in smart grid systems allows for real-time monitoring and optimization of energy distribution, which enhances the efficiency of renewable energy sources and reduces waste. Artificial intelligence also plays a pivotal role by accurately forecasting energy demand, leading to a more sustainable and balanced energy mix. The enhanced level of digitalization mitigates some of the potential drawbacks associated with the shift to renewable energy sources, thereby aiding in the reduction in carbon dioxide emissions. Digital technologies bolster the efficacy of renewable energy systems, streamline consumption patterns, and contribute significantly to the reduction in carbon emissions. From the viewpoint of environmental sustainability, these findings hold considerable importance. They suggest that the role of digitalization extends beyond fostering economic and social advancement; it is also crucial for realizing environmental objectives. The integration of digital technologies into energy transition strategies is a key driver in propelling China towards achieving its carbon reduction and sustainability goals. This not only aligns with China’s commitment to environmental stewardship but also exemplifies the global potential of digital transformation as a catalyst for sustainable development. Crucially, Hypothesis 3 has received validation.

4.4. Heterogeneity Test in Terms of the Environmental Sustainability Implications of the Digitization-Led Energy Transition

Considering the varied stages of digital progress and energy transition across the provinces of China, it is anticipated that the moderating role of digitalization in energy transition is not uniform regionally. This analysis seeks to explore these potential regional variations. Detailed in

Table 5. Results of heterogeneity test in terms of the environmental sustainability implications of the digitization-led energy transition.

Variable and Model	Model (10)	Model (11)	Model (12)
$\mathrm{e}\mathrm{n}\mathrm{e}$	−2.897 *** (−5.341)	−2.275 *** (−4.461)	−2.009 *** (−4.707)
$\mathrm{d}\mathrm{i}\mathrm{g}$	0.069 (1.397)	0.038 (1.401)	0.021 * (1.711)
$\mathrm{e}\mathrm{n}\mathrm{e}\times \mathrm{d}\mathrm{i}\mathrm{g}$	−0.143 * (−1.716)	−0.128 (−1.184)	−0.107 (−1.734)
$\mathrm{c}\mathrm{v}$	Yes	Yes	Yes
$\mathrm{c}$	2.308 (1.502)	2.181 (1.474)	2.197 * (1.637)
Province-fixed effect	Yes	Yes	Yes
Year-fixed effect	Yes	Yes	Yes
F-test	87.294 ***	90.108 ***	89.476 ***
${\mathrm{R}}^2$	0.265	0.279	0.301
$\mathrm{Religion}$	Eastern religion	Central religion	Western religion

Note: t-value in the parentheses; * a 10% significant level; *** a 1% significant level; c constant; $\mathrm{c}\mathrm{v}$ control variable.

, our study examines the differential effects arising from the interaction between digitalization and energy transition across the diverse provinces. This examination is essential to understanding how regional disparities in digital and energy infrastructure influence the broader energy transformation landscape.

Table 5 presents data that elucidate the regional variations in the impact of digitalization on the energy transition in China. In the eastern region, the interaction coefficients between energy transition and digitalization are significantly negative, exceeding 10%. This sharply contrasts with the central and western regions, where these coefficients do not show statistical significance, indicating a lack of a moderating effect of digitalization in these areas. These results imply that digitalization significantly influences the way energy transitions affect carbon dioxide emissions in the eastern region, a trend not evident in the central and western regions. The disparity in these outcomes is rooted in the different levels of digital and economic development across China’s provinces. The eastern region, with its advanced economic and technological development, has integrated digitalization more comprehensively across various sectors, including energy. Research by Yi et al. [208], Du et al. [209], and Wang et al. [210] highlights the region’s sophisticated digital infrastructure, which supports the efficient implementation of renewable energy technologies and effective carbon dioxide emissions management. In contrast, the central and western regions are at different stages of digital development. As indicated by Zhang et al. [211], Liu et al. [212], and Wang et al. [213], these regions are experiencing a slower digital progression, affecting their ability to fully utilize digitalization in their energy transition and carbon dioxide emissions reduction efforts. These insights demonstrate the critical role of digitalization in enhancing the environmental benefits of energy transitions, especially in regions with higher technological advancement. The eastern region’s experience shows that improving digital development can lessen the negative impact of energy transitions on carbon emissions, aiding in achieving both regional and national environmental sustainability goals. However, for the central and western regions, the findings suggest a need for increased investment in digital infrastructure to garner similar benefits. This underscores the importance of region-specific strategies in China’s pursuit of sustainable energy and digitalization, ensuring that all regions can effectively leverage digital technologies to achieve their environmental sustainability objectives. Significantly, the validation of Hypothesis 4 has been established.

4.5. The Nonlinear Effect of Energy Transition on Environmental Sustainability

As China experiences rapid economic development and an increasing demand for energy, the energy transition has become a key strategy for reducing greenhouse gas emissions and achieving environmental sustainability. China is currently transitioning from traditional fossil fuels to cleaner energy sources. This process exhibits complex nonlinear characteristics in energy consumption structure, technological advancement, and policy intervention across different regions and stages of development. Threshold regression analysis allows us to capture the nonlinear relationship between energy transition and carbon dioxide emissions within this dynamic context and identify how this relationship varies under different energy policies, economic levels, or patterns of energy consumption. Moreover, this method can reveal the differences in energy transition processes among various regions, providing a basis for formulating more precise and effective regional energy policies and environmental strategies. Therefore, threshold regression analysis is not only suitable for analyzing the complexity and dynamics of China’s energy transition but also serves as an important tool in understanding and promoting China’s efforts towards environmental sustainability. Then, we conducted the threshold effect test. The results are shown in

Table 6. Results of threshold effect test.

Number of Thresholds	Threshold Value	F-Statistic Value
Single threshold	0.302	83.456 ***
Double threshold	0.245	74.603 ***
Triple threshold	0.179	16.459

Note: *** a 1% significant level.

.

The results presented in Table 6 provide an insight into the threshold effect of digitalization on the relationship between energy transition and carbon dioxide emissions. The analysis, utilizing single-, double-, and triple-threshold tests, reveals notable findings. The single- and double-threshold tests indicate that the threshold value of digitalization is statistically significant at the 1% level, suggesting a pronounced impact at these levels. However, the triple-threshold test does not achieve significance at the 10% level, implying that a third threshold level does not further clarify the relationship. These results collectively indicate that the influence of energy transitions on carbon dioxide emissions exhibits a significant dual-threshold characteristic in the context of digitalization. This dual-threshold phenomenon remains robust even when control variables are introduced into the model. Importantly, the panel threshold model consistently rejects the null hypothesis of there being no threshold effect, affirming the existence of a significant threshold effect in the energy transition’s impact on carbon emissions. This analysis underscores the importance of considering regional digitalization levels in understanding the energy transition’s impact on carbon emissions. The variability in digitalization across regions appears to play a crucial role in modulating this relationship. Consequently, the incorporation of a panel threshold model is essential for a comprehensive analysis as it allows for the exploration of how different levels of digitalization across regions influence the effectiveness of the energy transition in reducing carbon dioxide emissions. The results of threshold effect regression are shown in

Table 7. Results of the nonlinear effect of energy transition on environmental sustainability.

Variable and Model	Model (13)
$\mathrm{e}\mathrm{n}\mathrm{e}·\mathrm{F}(\mathrm{d}\mathrm{i}\mathrm{g}\le 0.245)$	−0.195 *** (−4.814)
$\mathrm{e}\mathrm{n}\mathrm{e}·\mathrm{F}(0.245\le \mathrm{d}\mathrm{i}\mathrm{g}\le 0.302)$	−0.296 *** (−4.355)
$\mathrm{e}\mathrm{n}\mathrm{e}·\mathrm{F}(\mathrm{d}\mathrm{i}\mathrm{g}\ge 0.302)$	−0.387 *** (−4.133)
$\mathrm{c}\mathrm{v}$	Yes
$\mathrm{c}$	1.228 * (1.646)
${\mathrm{R}}^2$	0.546

Note: t-value in the parentheses; * a 10% significant level; *** a 1% significant level; c constant; $\mathrm{c}\mathrm{v}$ control variable.

.

The results presented in Table 7 illustrate a nonlinear relationship between energy transition and carbon dioxide emissions, which is significantly influenced by the level of digitalization. When digitalization is below a threshold of 0.245, the coefficient for the energy transition’s impact on carbon dioxide emissions is −0.195, which is significant at the 1% level. This coefficient is increased to −0.296 when digitalization exceeds the threshold of 0.245 and remains statistically significant at the 1% level. Furthermore, surpassing a second threshold of 0.302, the coefficient further declines to −0.387, still maintaining significance at the 1% level. These results indicate that as digitalization levels increase, the adverse impact of the energy transition on carbon dioxide emissions increases. The underlying reasons for these findings are rooted in the advancements in China’s digital landscape, as delineated in recent scholarly works by Ma et al. [214] and Zhang et al. [215]. Increased digitalization in China has been associated with more efficient energy usage, improved integration of renewable energy sources, and more effective emissions management. As digitalization continues to evolve, it drives innovation and optimization in energy consumption, thereby reducing the carbon intensity of energy production and utilization. These trends are in line with the goals of environmental sustainability. Higher digitalization facilitates enhanced monitoring and management of energy resources, helps in reducing waste, and supports the transition to cleaner, renewable energy sources. The findings from Table 7 highlight the essential role of digitalization in aiding China’s shift towards a more sustainable energy infrastructure. Our results suggest that advancements in digital technology are crucial to diminishing the environmental impacts of energy consumption. In summary, the research not only demonstrates the nonlinear dynamics between energy transition and carbon emissions but also underscores the vital importance of digitalization in China’s path to environmental sustainability. It implies that bolstering digital capabilities is key to significantly reducing the carbon footprint of the energy sector in China, contributing to global efforts to address climate change. According to the study, the validation of Hypothesis 5 has been successfully achieved.

5. Conclusions

This research presents an empirical analysis of how energy transition and digitalization affect carbon dioxide emissions, a key measure of environmental sustainability, in 28 provinces of China, spanning the years 2000 to 2021. The study also investigates the differential impacts of these variables across various regions. Employing both static and dynamic regression methodologies, the findings reveal a substantial contribution of energy transition to environmental sustainability, primarily through the reduction in carbon dioxide emissions. This impact is most significant in the eastern region of China, with the central and western regions following suit. Moreover, the study underscores the vital role of digitalization in moderating these effects. It is observed that digital technologies enhance energy efficiency and contribute to emission reductions. The influence of digitalization on lessening the negative impacts of energy transitions on carbon dioxide emissions becomes more pronounced with its progressive development. This trend is notably more evident in eastern China compared to the central and western regions. Additionally, the research explores the nonlinear relationship between energy transitions and carbon dioxide emissions. It identifies that increasing levels of digitalization can, in fact, intensify the adverse impacts of energy transitions on carbon dioxide emissions. This insight is critical for understanding the intricate balance between technological advancement, energy consumption, and environmental outcomes. The results of this study offer valuable guidance for policymakers and stakeholders in developing strategies that leverage the positive aspects of energy transition and digitalization while mitigating their negative environmental consequences, thus advancing toward achieving sustainable development objectives.

This paper derives several key insights based on the preceding analysis. First, the study underscores the considerable regional variations in how energy transition affects carbon dioxide emissions, with the eastern region of China exhibiting the most significant reductions. It is imperative for policymakers to craft energy policies tailored to the unique economic, technological, and environmental contexts of each region. For instance, while the eastern region might focus on pioneering advanced clean energy technologies, the central and western regions may need to prioritize foundational shifts in energy infrastructure and incentives to encourage the uptake of renewable energy. Second, the crucial moderating role of digitalization in China’s energy strategy cannot be overstated. Policy initiatives should foster the development and integration of digital technologies, such as smart grids, AI-driven energy analytics, and IoT solutions, to enhance energy management efficiency. This approach is anticipated to bolster energy efficiency and contribute significantly to reducing overall carbon emissions. Third, the research reveals a complex, nonlinear relationship between digitalization and its impact on the outcomes of energy transitions. Policymakers must acknowledge that the benefits of increasing digitalization for environmental sustainability may not always be straightforward. Continuous monitoring and evaluation of the environmental effects of digitalization within the energy sector are essential. Adjustments to strategies should be made to ensure that technological progress is consistent with sustainability objectives. Finally, addressing the potential negative consequences of energy transitions, particularly in areas with lower levels of digitalization, necessitates robust environmental regulations and incentives. Policy measures might include implementing stricter emissions standards, offering tax benefits for adopting clean energy, and providing subsidies to businesses investing in green technology. Furthermore, investing in educational initiatives and public awareness campaigns to highlight the advantages of energy efficiency and digitalization can cultivate wider community support for these efforts.

Based on this study’s analysis and findings, several research limitations are evident, leading to opportunities for future exploration. Firstly, the research’s focus on 28 provinces within China may not fully represent the diverse economic, technological, and environmental landscapes found in other regions or countries. Future research could broaden this scope by comparing the effects of energy transition and digitalization on environmental sustainability in a variety of international contexts. Secondly, the study’s timeframe, spanning from 2000 to 2021, excludes developments and changes that have occurred post-2021, including recent advancements in technology and policy. Given the fast-paced nature of technological and policy evolution, extending the research timeline beyond 2021 could provide more current and relevant insights into emerging trends and their impacts. Thirdly, while the study concentrates on energy transitions and digitalization, it does not comprehensively address other influential factors such as policy dynamics, economic fluctuations, and societal shifts. Future research could incorporate a wider array of variables, including political elements, economic cycles, social attitudes, and specific technological innovations, to offer a more holistic understanding of the subject. Fourthly, the current study does not include energy intensity (total energy consumption/GDP) as a variable. This omission limits the comprehensiveness of the model, particularly in analyzing the impact of digitalization on greenhouse gas emissions and environmental sustainability. Therefore, future studies should incorporate energy intensity as a key variable. This addition will enable a more holistic understanding of the relationship between energy use, economic activity, and environmental impacts. It will also enhance the model’s capability to assess the effectiveness of digitalization and energy transition strategies in reducing greenhouse gas emissions. Lastly, the application of static and dynamic regression methods in this study may not adequately capture the intricate and evolving interactions between the various factors. Future studies could benefit from applying more advanced statistical techniques or models, such as machine learning algorithms or system dynamics modeling. These methods could provide deeper insights into the nonlinear relationships and causal mechanisms at play among the study variables.