Abstract
Carbon emission efficiency is vital for environmental performance, and the digital economy, as an emerging force, plays a critical role in enhancing carbon emission efficiency. This study employed panel data from 30 Chinese provinces (2010–2020) to measure the digital economy index and carbon emission efficiency through Principal Component Analysis and the Super-Efficiency SBM model. A double fixed effects model was applied for regression, followed by robustness tests, heterogeneity analysis, and an investigation into the mechanisms and spatial spillover effects. Five findings are elicited: (1) Spatiotemporal distribution differences exist between digital economy level and carbon emission efficiency. (2) The digital economy effectively enhances carbon emission efficiency, a result that remains robust after endogenous and robustness tests. (3) Significant spatial spillover effects are observed, where improvements in the digital economy enhance carbon emission efficiency in neighboring areas. (4) The impact of the digital economy is more pronounced in eastern and high digital development regions but weaker in the northeastern and low digital development areas. (5) The main pathways through which the digital economy influences carbon emission efficiency are industrial structure upgrading, energy structure optimization, and green technological innovation. These findings extend the environmental benefits of the digital economy, providing insights for achieving the dual carbon goals.
Similar content being viewed by others
Introduction
The industrial production model of the traditional era has primarily depended on fossil fuel consumption, driving greenhouse gas emissions to levels approaching the limits of environmental tolerance. Mitigating global warming has consequently emerged as one of the most pressing challenges of the twenty-first century. As the largest developing country, the Chinese government committed to the international community to peak carbon emissions by 2030 and achieve carbon neutrality by 2060. Enhancing carbon emission efficiency has become crucial, particularly given its urbanization’s heavy reliance on industrial development. The rise of the digital economy, represented by the internet industry, is a solution to overcoming existing challenges by transforming traditional productivity models1. The digital economy originated in the 1990s and is viewed as an emerging economic model in which digital data is the key production factor, with technological innovation as the core driving force2. In 2022, China’s digital economy reached 50.2 trillion yuan, ranking second globally, with its share of GDP rising to 41.5%. This rapid growth positions the digital economy as a crucial driver of high-quality economic development and a vital support of environmental improvements3. In fact, not only can the digital economy generate a multiplier effect on green development, but the digital economy also requires decarbonization to achieve sustainable growth4.
In the context of China’s productivity transformation strategy, the digital economy has emerged as a crucial driver of economic growth5 and a potential solution to ecological challenges6. As a key metric for balancing economic development with ecological sustainability, carbon emission efficiency plays a critical role in assessing the effectiveness of these efforts7. The digital economy, with its intelligent scheduling capabilities, promotes efficiency improvements across the industrial chain’s upstream and downstream, thereby reducing energy consumption and carbon emissions per unit of output8. In line with the new growth theory, it accelerates knowledge accumulation and dissemination, fostering the adoption of green technologies that reduce carbon emissions9. Furthermore, the digital economy’s information-sharing features enable the integration of big data to optimize resource allocation, offering more effective carbon emissions management and supporting coordinated efforts to reduce emissions across regions10. At this crucial stage of the low-carbon transition, the potential of the digital economy is immense, and investigating how it influences carbon emission efficiency can help develop strategies to leverage this potential better.
Despite its potential, there is no clear consensus on the relationship between the digital economy and carbon emissions. Ecological modernization theory posits that the information revolution driven by the digital economy is facilitating the second ecological transformation. The digital economy, driven by digital finance, has broken the resource curse and become a new engine for low-carbon growth11. Some scholars argue that industrial structure upgrading12, regional green innovation13, and geographical spatial boundary removal14 are effective strategies for the digital economy to curb carbon emissions. Conversely, others contend that the rise of the digital economy may not necessarily lead to reduced carbon emissions. The extensive use of digital devices leads to increased energy consumption, a rise that cannot be fully offset by the carbon emission reduction benefits achieved by the digital economy15. Additionally, studies have observed nonlinear relationships, such as an inverted U-shaped curve between the digital economy and carbon emissions16.
Although numerous studies have focused on measuring carbon emission intensity and efficiency, few have considered new quality productivity as the basis for selecting input–output indicators17. Moreover, evaluations of the digital economy have often been limited, focusing mainly on quantitative input–output indicators such as infrastructure and application industries18, without integrating qualitative metrics into a more comprehensive evaluation framework. Finally, the literature has overlooked the potential spatial effects of the digital economy on carbon emission efficiency and has yet to determine whether these effects are spillover or siphoning effects19.
This study contributes in three aspects. First, in contrast to prior research that focuses on traditional input–output metrics, this study applies the new quality productivity framework to select labor, capital, and energy as input variables for carbon emission efficiency. Second, this study extends existing digital economy indices by incorporating measures of digital finance and digital industrialization alongside Internet‑sector indicators, thereby capturing the contributions of manufacturing and other industries. Third, this study empirically confirms that industrial structure upgrading, energy structure optimization, and green technological innovation serve as key mediating pathways through which the digital economy enhances carbon emission efficiency and that these effects exhibit significant spatial spillovers and regional heterogeneity, offering targeted strategies for promoting coordinated regional carbon reduction.
Theoretical hypothesis
The direct impact of the digital economy on carbon emission efficiency
The digital economy, characterized by intelligent integration, data sharing, and network communication, can significantly improve carbon emission efficiency20. This emission reduction effect has evolved from solitary endeavors to a multi-stakeholder collaborative governance model involving enterprises, governments, and the public. At the organizational level, the digital economy enables enterprises to rapidly integrate production factors, optimize resource allocation through improved production processes, and transform traditional, high-pollution manufacturing industries into green, technology-driven enterprises21. At the institutional level, the digital economy, based on big data, can improve governance frameworks and establish real-time updated carbon trading platforms22, providing technical support for collaborative government regulation23. At the societal level, the digital economy expands public access to environmental information, enhances public participation in environmental governance, and promotes carbon reduction effects24.
H1
The digital economy can improve carbon emission efficiency.
The indirect impact of the digital economy on carbon emission efficiency
Industrial structure upgrading
The digital economy drives industrial upgrading through two key pathways. First, its cross-temporal and spatial data-sharing characteristics enhance information flow and reduce transaction costs, channeling resources into capital and knowledge-intensive industries25. Second, fostering cross-industry integration and optimizing resource allocation reshapes industrial structures, accelerating the transition toward a more advanced and efficient economy26.The upgrading of industrial structure indicates the redistribution of production resources, including labor and capital, across several sectors, improving production efficiency and reducing carbon emissions per output unit27. Furthermore, the transition to an increased share of the tertiary sector—propelled by knowledge, information, and other intangible production elements—diminishes dependence on conventional resource-intensive industries. This transition facilitates a more sustainable low-carbon economy and improves carbon emission efficiency28.
H2
The digital economy improves carbon emission efficiency by industrial structure upgrading.
Energy structure optimization
The digital economy optimizes the energy structure through three primary mechanisms. First, by promoting green consumption and shared mobility, the digital economy fosters sustainable lifestyles and reduces dependence on fossil fuels29. Second, cross-regional information exchange and real-time data transmission enhance the rational distribution of clean energy, fundamentally transforming the energy structure30. Third, the digital economy accelerates the development and utilization of renewable energy by stimulating market demand for hydropower and wind energy31. This transition shifts the energy framework from reliance on coal to a more diversified and sustainable structure, reducing fossil fuel consumption. Optimizing the energy structure encourages industries to adopt environmentally friendly production processes through price fluctuations in energy markets. Furthermore, this optimization increases the development and utilization of renewable energy, enhancing overall energy efficiency and reducing carbon emissions32.
H3
The digital economy improves carbon emission efficiency by energy structure optimization.
Green technology innovation
The digital economy fosters green technology innovation via its digital and networked characteristics through two distinct approaches. First, it reduces the expenses of developing eco-friendly products, maximizes resource efficiency, and facilitates green technology innovation for enterprises33. Furthermore, the digital economy diminishes information asymmetry and improves access to internal funding and external credit to attract increased market capital and accelerate green technology innovation34. Green technology innovation enhances carbon emission efficiency in several ways. Advanced green technologies, including resource recycling and clean purification techniques, mitigate pollutant generation and reduce carbon emissions through highly efficient treatment processes, tackling environmental challenges at source35. Furthermore, the quest for financial gains drives companies to invent green technologies, achieving dual economic and ecological benefits goals36.
H4
The digital economy improves carbon emission efficiency by green technology innovation.
Heterogeneous effects of the digital economy on carbon emission efficiency
Substantial inequalities in resource endowment, economic development, and governmental support across different regions in China lead to variations in the digital economy’s impact on carbon emission efficiency37. The eastern regions, with stronger technological innovation capabilities and greater demand for green markets, have accelerated the rise of the digital economy and the advancement of low-carbon goals. In contrast, the western regions often face constraints in technology, finance, and market access, weakening the role of the digital economy in reducing emissions38. Additionally, disparities in internet development levels, 5G network coverage, and data availability across regions affect the effectiveness of the digital economy in driving industrial transformation and technological innovation39. Regions with robust digital infrastructure are better positioned to enhance carbon efficiency, while those with insufficient infrastructure may encounter technical and economic barriers, leading to less significant emission reduction outcomes40.
H5
The impact of the digital economy on carbon emission efficiency exhibits heterogeneity based on geographic distribution and the level of digital economy development.
Spatial spillover effects of the digital economy on carbon emission efficiency
Adjacent regions often exhibit significant economic and policy interconnections. The digital economy, characterized by its ability to transcend spatial boundaries, not only enhances local carbon emission efficiency but also generates spatial spillover effects, improving the carbon emission efficiency of neighboring areas. On the one hand, driven by technologies such as the Internet of Things (IoT), the rapid development of the digital economy accelerates the cross-regional flow of information, capital, and talent, reducing resource misallocation and redundancy. This positive impact is not limited to developed regions but also radiates to neighboring areas through economic interactions and information exchange41. On the other hand, the digital economy creates positive demonstrations and learning effects in different regions. Its development promotes sustainable regional economic growth, encouraging surrounding areas to adopt similar strategies and improve their carbon emission efficiency42.
H6
The digital economy exerts spatial spillover effects on carbon emission efficiency.
Based on the above hypotheses, the research framework of this study is shown in Fig. 1.
Theoretical research framework.
Research design
Variable selection
Dependent variable
Carbon emission efficiency (CEE) has been assessed in various studies using methods such as stochastic frontier models43. This study calculates CEE for 30 Chinese provinces using a super-efficiency SBM model incorporating undesirable outputs. This model calculates super-efficiency values for decision-making units (DMUs), enabling comparative analysis among multiple efficient units. Addressing radial and angular issues overcomes the limitations of traditional models that require proportional changes in inputs and outputs, thus providing a more precise reflection of regional differences in CEE44.
Input indicators are selected to reflect the dimensions of new quality productivity, with GDP as the desirable output and carbon emissions as the undesirable output45. Labor input is measured by the number of urban employees in each region at the end of the year. While some studies assess capital input in the industrial sector using net fixed assets of large industrial enterprises, this study adopts a provincial perspective, estimating labor material input (capital) via the perpetual inventory method to derive regional capital stock46. Following established research, labor object input (energy) is represented by total regional energy consumption in each region47. Desirable output is measured by real GDP after deflation, and undesirable output is calculated as carbon emissions from eight fossil fuels—coke, coal, crude oil, gasoline, kerosene, diesel, fuel oil, and natural gas—using their respective emission factors48. The calculation formulas are as follows:
where C presents the carbon dioxide emissions, Ei denotes the energy consumption for the i-energy type, and Fi is the carbon emission factor for the i-energy type. The specific values for each type of energy are shown in Table 1.
Independent variable
Existing literature primarily measures digital economy levels using data from the telecommunications, internet, software, and other digital sectors49. However, this approach captures only part of the digital economy, as industries like manufacturing are also advancing digital economy growth. Digital finance has been introduced as a measure of industry digitalization to address this limitation50. Furthermore, recognizing the necessity for digital industries to scale up to achieve market integration has led to incorporating digital industrialization into the digital economy indicator system51. Additionally, digital infrastructure and digital innovation significantly drive digital economy growth: digital infrastructure facilitates digital economy development52, while digital innovation fosters the production of digital economy products53.
Based on the above, this paper constructs a composite index for the digital economy (DE) encompassing digital infrastructure, digital industrialization, industry digitization, and digital innovation capacity. The weights of the 12 positive indicators are calculated using the principal component analysis, and the comprehensive evaluation system is shown in Table 2.
Mediating variables
To analyze how the digital economy influences carbon emission efficiency, industrial structure upgrading, energy structure optimization, and green technological innovation are selected as mediating variables.
Industrial structure upgrading (ISU): The digital economy enables precise capture of market dynamics, driving traditional industries to transform into technology-intensive sectors that reduce carbon emissions54. ISU is measured through two dimensions—industrial structure rationalization (ISR) and industrial structure advancement (ISA).
Industrial structure rationalization (ISR) is quantified as the reciprocal of the Theil index, calculated as:
Industrial structure advancement (ISA) reflects the shift towards more advanced industries, measured by the ratio of tertiary to secondary industry output, calculated as:
Energy structure optimization (ESO): Through intelligent data analysis and cross-regional resource integration, the digital economy promotes the production and use of renewable energy, driving a transition from fossil fuels to cleaner energy sources55. ESO is represented by the proportion of coal consumption in total energy consumption across provinces, reflecting the progress towards cleaner energy sources.
Green technological innovation (GTI): The digital economy boosts data-sharing capabilities, enabling effective knowledge exchange that drives green technological innovation and improves carbon emission efficiency56. GTI is captured by the number of authorized green invention patents, indicating the level of innovation contributing to environmental sustainability in each province.
Control variables
The following control variables have been selected and incorporated into the analysis to account for other factors that may influence carbon emission efficiency.
-
Infrastructure construction (IC): The growth of digital economies relies on infrastructure. Well-developed infrastructure also attracts economic activities that support green development. This variable is measured as the logarithm of per capita road area.
-
Urbanization level (UL): Early urbanization may cause inefficient resource use, limiting carbon efficiency. As urbanization advances, better resource allocation can lead to environmental gains. This is quantified as the ratio of the urban population to the total population of each region.
-
Degree of openness (OPE): Openness to international trade introduces green technologies but can also draw resources to high-energy industries, worsening environmental impact. This is measured as the ratio of total imports and exports to regional GDP.
-
Government intervention (GOV): Government involvement can drive green technology innovation through policy incentives such as carbon reduction subsidies or tax breaks. This is measured as the ratio of fiscal expenditure to regional GDP.
-
Urban green coverage (UGC): Regions with higher green coverage often have advanced environmental technologies and awareness, providing a foundation for enhancing carbon emission efficiency. This is measured as the forest coverage rate within built-up areas.
-
Human capital level (HCL): Human capital drives the technological innovation needed for digital transformation in traditional and big data sectors. This is measured as the ratio of the number of students enrolled in regular higher education institutions to the total population of the region.
Data sources
This study utilizes panel data from 30 Chinese provinces from 2011 to 2020. The data are primarily drawn from the China Statistical Yearbook, China Energy Statistical Yearbook, China Environmental Statistical Yearbook, and various provincial statistical yearbooks. GDP is adjusted for inflation using 2011 as the base year. Missing data are addressed using interpolation methods to ensure completeness. The descriptive statistics are shown in Table 3.
Research methods
Super-efficiency SBM model
To evaluate the carbon emission efficiency of 30 Chinese provinces and municipalities from 2011 to 2020, a super-efficiency slack-based measure (SBM) model is constructed, incorporating inputs such as labor, labor objects, and labor materials57. The model is written as follows:
where \({\rho }^{*}\) presents the carbon emission efficiency value of the DMU, \({\lambda }_{j}\) is a non-negative weigh vector, and \({s}_{i}^{-}\), \({s}_{r}^{+}\), \({s}_{k}^{-}\) are slack variables. The carbon emission efficiency vector representing only when \({s}_{i}^{-}, {s}_{r}^{+}\), and \({s}_{k}^{-}\) are all zero. This condition is both necessary and sufficient for a DMU to be considered DEA-efficient, otherwise classified as inefficient.
Two-way fixed effects model
To test the proposed hypotheses, a two-way fixed effects model is constructed to analyze the direct impact of the digital economy on regional carbon emission efficiency. The model is written as follows:
where \({CEE}_{it}\) represents the carbon emission efficiency of province i in year t, \({DE}_{it}\) denotes the level of the digital economy in province i in year t, and \({X}_{it}\) includes a series of control variables. \(\alpha\) represent estimated coefficients, \({\mu }_{i}\) and \({\delta }_{t}\) capture region and time fixed effects, respectively. \({\varepsilon }_{it}\) is the random error term.
Mediation effect model
The baseline regression model captures the direct impact of the digital economy on low-carbon development but does not reveal the underlying mechanisms. Therefore, this study constructs a mediation effect model with industrial structure upgrading, energy structure optimization, and green technology innovation as mediating variables to explore their indirect effects. The calculation formulas are as follows:
where \({CEE}_{it}\) and \({DE}_{it}\) represent the carbon emission efficiency and the digital economy level of province i in year t, respectively. \({M}_{it}\) refers to the mediating variables, and \({X}_{it}\) includes control variables. \({\mu }_{i}\) and \({\delta }_{t}\) account for regional and year fixed effects, \({\varepsilon }_{it}\) is the random error term. \(\alpha , \, \beta , \, \gamma\) are the estimated coefficients.
Spatial panel econometric model
Given that carbon emissions cross regional borders through air and particulate pollution, there is a strong spatial correlation in emissions across regions. This study employs the Spatial Durbin Model (SDM) to capture spatial spillover effects, primarily including the impact of neighboring regions’ carbon emission efficiency on local carbon emission efficiency, the influence of neighboring regions’ digital economy development on local carbon emission efficiency, and the effects of neighboring regions’ other control variables on local carbon emission efficiency. The model is written as follows:
where \({CEE}_{it}\) represents the carbon emission efficiency of region i at time t, \({DE}_{it}\) denotes the digital economy index of region i at time t. \({X}_{it}\) represents a series of control variables, including IC, UL, OPE, GOV, UGC and HCL. W is the economic distance spatial weight matrix, \({\mu }_{i}\) and \({\delta }_{t}\) capture region and time fixed effects, and \({\varepsilon }_{it}\) is the random error term. \({\alpha }_{1}\)、\({\beta }_{2}\)、\({\gamma }_{2}\) are the key coefficients of interest in this study, where \({\alpha }_{1}\) measures the spatial dependence of \(CEE\), \({\beta }_{2}\) indicates the spatial spillover effect of \(DE\), and \({\gamma }_{2}\) shows the spatial spillover effect of control variables.
Empirical results and analysis
Evaluation of digital economy index
The 30 provinces in China are categorized into four regions: Eastern, Central, Western, and Northeastern. The average digital economy index for each region is computed based on the digital economy performance of its provinces, as detailed in Table 4.
Temporal distribution characteristics
Table 4 demonstrates a consistent upward trend in the digital economy from 2011 to 2020, with the national average increasing from 0.262 in 2011 to 1.988 in 2020, reflecting a more than sixfold rise. Regionally, all four areas—Eastern, Central, Western, and Northeastern—exhibited growth patterns aligned with the national trend.
As illustrated in Fig. 2, the digital economy in the Eastern region expanded from 0.516 in 2011 to 2.593 in 2020, while the Central region experienced growth from 0.128 to 1.769 over the same period. Similarly, the Western region increased from 0.130 to 1.707, and the Northeastern region increased from 0.167 to 1.400. These results reflect substantial advancements in digital infrastructure across all regions, with steady progress in digital economy development. This decade of sustained growth can be attributed to governmental initiatives designed to foster the digital economy. Policies encouraging digital industrialization and facilitating digital transformation have created an environment that supports continuous and stable expansion.
Regional digital economy index in China (2011–2020).
Spatial distribution characteristics
As shown in Fig. 3, the digital economy demonstrates a regional hierarchy, with development levels ordered as “East > Central > West > Northeast”, with significant disparities between regions. Eastern provinces, such as Guangdong and Jiangsu, consistently lead, with Guangdong’s average digital economy score of 2.385 being 3.5 times that of Jilin (0.663), emphasizing the severity of the digital divide. Although the Eastern region surpasses the national average, notable internal differences persist. For instance, Tianjin’s average score of 0.899 and Hainan’s 0.851 falls significantly below the leading provinces in the East. Similarly, the Central and Western regions exhibit disparities, with provinces such as Henan and Sichuan averaging over 1.0, while Shanxi, Guangxi, Ningxia, and Qinghai lag behind other areas. The Northeast region remains the most underdeveloped, with Jilin and Heilongjiang averaging below 0.8, highlighting the urgent need for targeted policy support.
Digital economy index in China (2011–2020).
In terms of growth rate, the digital economy follows a pattern of “Central > West > Northeast > East”. While the Eastern region retains the highest overall level, its growth has been relatively slower. By contrast, the Central region has experienced substantial growth in recent years, indicating considerable development potential. However, the Western region trails behind the Central and Eastern regions and the establishment of big data centers has accelerated growth in certain provinces. For example, since Guizhou launched the first national-level big data comprehensive pilot zone in 2016, its advantages in energy, land, and other resources are being effectively leveraged to drive digital economic development.
Evaluation of carbon emission efficiency
A super-efficiency SBM model with undesirable outputs was used to analyze China’s spatiotemporal dynamics of carbon emission efficiency. The average carbon emission efficiency for each economic region is presented in Table 5.
Temporal distribution characteristics
Table 5 shows that national carbon emission efficiency generally increased with fluctuations from 2011 to 2020, rising from 0.542 to 0.712 at an average annual growth rate of 3.08%. The trend can be divided into two phases. From 2011 to 2017, efficiency increased rapidly from 0.542 to 0.706, with a 4.49% annual growth rate driven by substantial environmental investments. From 2018 to 2020, growth slowed, with efficiency rising from 0.701 to 0.712 at an annual rate of 0.79%. The deceleration in carbon emission efficiency from 2018 to 2020 was due to the implementation of stricter carbon policies, particularly the national carbon emissions trading system (ETS), which imposed cost pressures and technological demands on companies.
From 2011 to 2020, the average carbon emission efficiency for the four major regions was as follows: Eastern (0.934), Central (0.584), Western (0.471), and Northeastern (0.504). According to Fig. 4, in the Eastern region, efficiency increased sharply from 0.779 to 0.954 between 2011 and 2014, then more gradually to 1.001 by 2020, with an annual growth rate of 2.82%. The Central region exhibited similar trends, with rapid growth from 0.474 to 0.617 (2011–2014) and slower gains to 0.635 (2020), averaging 3.29% annually. Efficiency in the Western region steadily improved despite a minor decline in 2020, with a 3.09% annual growth rate. The Northeastern region experienced the highest growth, rising from 0.395 to 0.546 (2011–2015) and reaching 0.561 by 2020, with an annual growth rate of 3.98%. The regional trends align closely with the national trend, characterized by a fluctuating upward movement. These trends reflect the Eastern region benefiting from advanced resources and early reforms. In contrast, the Central and Western regions were affected by industrial shifts, particularly the relocation of energy-intensive enterprises to the West.
Regional carbon emission efficiency in China (2011–2020).
Spatial distribution characteristics
As depicted in Fig. 5, the carbon emission efficiency exhibits a regional gradient pattern of “East > Central > Northeast > West”, with a relatively stable distribution. The Eastern region stands out, with Beijing and Guangdong leading in carbon emission efficiency at 1.249 and 1.142, respectively, though internal imbalances remain. The Central region occupies an intermediate position, with Hunan achieving the highest average efficiency at 0.882, while coal-reliant provinces like Henan and Anhui face tremendous pressure to reduce emissions. In the Western region, efficiency levels are lower, although Chongqing and Sichuan perform relatively well, whereas Qinghai and Gansu confront significant challenges, with averages of only 0.374 and 0.349, respectively. As a traditional industrial hub, the Northeast suffers from resource shortages and insufficient momentum for industrial transformation, resulting in higher losses in carbon emission efficiency.
Carbon emission efficiency in China (2011–2020).
The growth trend of carbon emission efficiency follows the sequence of “Central > West > Northeast > East”. The Central region, particularly Hunan and Hubei, has made the most significant strides in green transitions, reflecting the highest growth in carbon emission efficiency. Western provinces like Chongqing and Sichuan, supported by the adoption of clean energy and industrial restructuring, also demonstrate considerable potential for improvement. In contrast, despite maintaining a leading position, the Eastern region has experienced relatively slow growth, indicating that it has reached a more mature stage of development. The Northeast lags, highlighting the need for technological innovation, such as advancements in the digital economy, to drive improvements in carbon emission efficiency.
Benchmark regression analysis
Table 6 displays the results from the benchmark regression analysis, which investigates the influence of the digital economy on carbon emission efficiency. Column (1) indicates that the digital economy has a positive and statistically significant impact on carbon emission efficiency at the 5% significance level when only the independent variable is considered. Upon the incorporation of control variables in Column (2), the coefficient for the digital economy continues to exhibit a positive value and retains statistical significance at the 1% level. This finding supports the conclusion that the digital economy enhances carbon emission efficiency, confirming H1. The digital economy improves carbon emission efficiency through enhanced resource allocation, optimized labor management, and a reduced carbon footprint in economic activities. The enhancements of the digital economy led to increased carbon emission efficiency from both input and output viewpoints58.
Infrastructure development exhibits a significant negative correlation with carbon emission efficiency among the control variables. This inverse correlation may stem from the elevated energy requirements linked to infrastructure expansion, which subsequently heighten energy consumption and diminish carbon emission efficiency. The degree of openness exhibits a notable positive impact on carbon emission efficiency. Restricting the export of energy-intensive, high-emission products while enhancing the import of low-carbon products and environmental technologies effectively improves carbon emission efficiency.
Endogeneity and robustness tests
Endogeneity test
To address potential endogeneity, an instrumental variable (IV) method is used for the endogeneity test. Before the widespread adoption of the Internet, fixed telephones served as a medium for information exchange and dissemination. Telecommunications infrastructure from this period has been shown to positively influence internet penetration rates, thereby exhibiting a strong correlation with the growth of the digital economy, satisfying the relevance criterion for an IV. Moreover, the historical number of fixed telephones per 100 people exerts a negligible effect on regional carbon emission efficiency, thus meeting the exclusion restriction for an IV. Consequently, the interaction term between the number of fixed telephones per 100 people in 1984 and the national revenue from information technology services in the prior year is utilized as the IV.
As shown in Table 7, column (1) indicates a significant correlation between the instrumental variable and the digital economy at the 1% significance level. Column (2) confirms that the positive effect of the digital economy on carbon emission efficiency remains significant, consistent with the baseline regression results. Additionally, the statistics effectively address concerns about weak instruments. In summary, the positive impact of the digital economy on carbon emission efficiency remains significant at the 1% level after incorporating instrumental variable regression, further validating the benchmark regression results.
Robustness tests
The robustness of the regression results is assessed through three approaches: substituting the dependent variable, altering the measurement of the independent variable, and excluding the four direct-controlled municipalities.
(1) Substituting the dependent variable. The ratio of regional GDP to carbon dioxide emissions is used as an alternative measure of carbon emission efficiency (CEE2). As shown in column (1) of Table 8, the estimated coefficient for the digital economy remains significantly positive at the 1% level, consistent with the benchmark regression results. This confirms that the digital economy indeed enhances carbon emission efficiency.
(2) Altering the measurement method for the independent variable. The digital economy index (DE2) is recalculated using the entropy method for each region. The regression results, shown in column (2) of Table 8, demonstrate that the digital economy continues to exhibit a significant positive correlation with carbon emission efficiency, thereby corroborating the robustness of the original regression results.
(3) Excluding the four direct-controlled municipalities. Due to their unique administrative status, the four direct-controlled municipalities may benefit from more advantageous conditions for carbon emission management, potentially overstating the impact of the digital economy. A regression analysis excludes these municipalities to correct for potential bias introduced by administrative status. Column (3) of Table 8 confirms that the research conclusions remain robust after excluding the direct-controlled municipalities.
Mechanism analysis
The mediating effect of industrial structure upgrading is examined in Table 9. Column (1) reports that the overall effect coefficient of the digital economy on regional carbon emission efficiency is significantly positive at the 1% level. Column (2) indicates a positive and significant effect of the digital economy on industrial structure upgrading at the 5% level, suggesting that the digital economy fosters advancements in industrial structure. Column (3) further shows a positive and significant effect of industrial structure upgrading on regional carbon emission efficiency at the 10% level, indicating that structural upgrades contribute to enhanced carbon efficiency. Notably, after introducing the mediating variable of industrial structure upgrading, the coefficient for the digital economy remains significantly positive, though diminished, confirming H2. The reason is that the development of the digital economy accelerates the digital transformation of traditional industries, thereby improving carbon emission efficiency by enhancing the contribution of low-carbon industries59.
The mediating effect of energy structure optimization is presented in columns (4) and (5) of Table 9. Both columns show significantly negative coefficients at the 1% level, indicating that the development of the digital economy reduces the reliance on coal in the energy structure, thereby improving carbon emission efficiency. Even after introducing the energy structure optimization variable, the coefficient for the effect of digital development on carbon emission efficiency remains significantly positive, although reduced, thereby supporting H3. The reason is that the digital economy drives energy structure optimization by reshaping energy consumption patterns and accelerating the use of renewable energy, thereby improving carbon emission efficiency through the reduction of traditional energy consumption60.
The mediating effect of green technology innovation is demonstrated in columns (6) and (7) of Table 9. Both columns show significantly positive coefficients at the 1% level, suggesting that the digital economy promotes green technology innovation, which positively influences carbon emission efficiency. After incorporating green technology innovation into the model, the effect of the digital economy on carbon emission efficiency weakens but remains significant. This allows the rejection of the null hypothesis at the 5% level, thus supporting H4. The reason is that the digital economy promotes the development of green products and innovative technologies, reducing pollutant emissions at the source and improving carbon emission efficiency61.
The Bootstrap method was applied to test mediation effects to enhance the robustness of the digital economy’s impact on carbon emission efficiency62. The mediating effects of industrial structure upgrading, energy structure optimization, and green technology innovation were further decomposed into direct and indirect effects, as shown in Table 10. The results reveal that the direct and indirect effects of industrial structure upgrading are 0.725 and 0.553, both significant at the 1% level. Similarly, the direct and indirect effects of energy structure optimization are − 0.396 and − 0.236, and for green technology innovation, 0.057 and 0.050, all significant at the 1% level. Notably, the 95% confidence intervals for the indirect effects of these three mediators exclude 0, indicating that the mediation effect exists. The mediation effect proportions are 0.433 for industrial structure upgrading, 0.373 for energy structure optimization, and 0.465 for green technology innovation.
Further analysis
Heterogeneity analysis
The vast geography of China, combined with significant disparities in resource endowments, economic development, and policy support across regions, may contribute to the digital divide. This section investigates the heterogeneous effects of the digital economy on carbon emission efficiency, focusing on the geographic distribution and the development level of the digital economy.
Geographic distribution
As shown in columns (1)–(4) of Table 11, the digital economy’s impact on carbon emission efficiency varies significantly across regions, strongest in the East, moderate in the Central and Western regions, and insignificant in the Northeast. In the eastern region, rapid economic growth has attracted digitally innovative enterprises. Strong policy support has also accelerated the digitization of high-carbon industries such as manufacturing, leading to the strongest impact of the digital economy on carbon emissions63. The Central region has promoted the development of a digital economy due to its advantageous natural resource endowment. Although the West shows a stronger effect than the Central region, it has primarily absorbed high-carbon, mid- to low-end industries driven by policy and cost factors64. These transferred industries, relative to the West’s original sectors, represent a form of industrial upgrading. However, while such upgrading could theoretically lower emissions, the additional carbon emissions from these relocated industries partially offset these gains65. The northeastern region is in the early stages of digital development, with insufficient digital infrastructure and policy support66. Its reliance on high-carbon heating systems has also increased fossil fuel consumption, reducing the digital economy’s impact on carbon emissions.
Level of digital economy development
To explore whether effects differ by digital economy development level, regions are divided into high and low digital economy development groups using the 50th percentile as a threshold: those above the 50th percentile are designated as high digital economy regions, while those below are considered low digital economy regions. Columns (5) and (6) in Table 11 reveal that high digital economy regions have significantly positive coefficients, while low digital economy regions do not exhibit significant effects. Regions with higher digital economy development levels benefit from more advanced infrastructure and mature digital technology applications, allowing for a more effective realization of the carbon reduction potential associated with the digital economy67. In contrast, regions with lower digital economy levels, still in the early stages of expanding digital infrastructure, have yet to fully leverage the emission reduction potential of the digital economy68. Additionally, the installation of new digital infrastructure in these regions tends to increase energy consumption and carbon emissions.
Spatial effect analysis
The global Moran’s I index was applied to assess spatial autocorrelation in carbon emission efficiency across 30 provinces in China from 2011 to 2020. As displayed in Table 12, Moran’s I values were consistently positive and significant at the 1% level, indicating strong spatial clustering of carbon emission efficiency.
A series of tests were conducted to identify the appropriate spatial econometric model. As shown in Table 13, the LM lag and LM error tests revealed significant spatial lag and error effects, indicating that a simple panel regression model would be inadequate. Subsequently, the Hausman test was employed to distinguish between fixed and random effects models. The results confirmed the suitability of the fixed effects model. Finally, the LR and Wald tests supported the SDM as the superior model compared to the Spatial Autoregressive Model (SAR) and Spatial Error Model (SEM).
While Hausman and LM tests were used to select the spatial model, additional robustness checks were conducted by adding alternative spatial models, such as the SAR and the SEM, as provided in Table 14. A comparison of the three models suggests that spatial dependence in the model stems from the dependent variable, independent variables, and error terms, validating the choice of the SDM as it cannot be reduced to SAR or SEM. In the SDM model, the coefficient of the digital economy is 0.031 and is positive at the 1% significance level. To minimize the estimation errors resulting from point estimation, this paper applies a partial derivative approach69 to decompose the total effect into direct and indirect effects, as shown in Table 15.
Table 15 reveals that the coefficient of the digital economy on carbon emission efficiency is 0.031 for direct effects, 0.071 for indirect effects, and 0.102 for total effects, all statistically significant at the 1% level. These results suggest that the digital economy directly improves regional carbon emission efficiency and induces significant spillover effects across neighboring areas.
Specifically, the direct effect arises because the digital economy facilitates the low-carbon transformation of local enterprises, such as adopting advanced low-carbon technologies to improve production processes and enhance carbon emission efficiency. The indirect effects reveal how the digital economy boosts carbon emission efficiency across regional borders, thereby supporting H5. The spillover effect of the digital economy may stem from technology sharing. Enterprises in regions with advanced digital economies possess cutting-edge low-carbon technologies, and the spread of these technological advancements to neighboring regions can enhance their carbon emission efficiency70. Furthermore, capital’s pursuit of profit leads it toward regions with advanced digital economies. Through interconnected supply chains, this capital mobility can facilitate industrial restructuring in neighboring areas, indirectly improving their carbon emission efficiency71.
Regarding control variables, infrastructure development suppresses carbon emission efficiency within the region and neighboring areas. Conversely, increased openness enhances carbon emission efficiency in both contexts. While improvements in human capital do not significantly impact local carbon emission efficiency, they positively influence efficiency in neighboring regions due to labor mobility across borders. However, urbanization inhibits improvements in carbon emission efficiency in neighboring regions due to industrial migration and increased transportation demands associated with urban expansion. Government interventions generally exhibit localized effects with limited cross-regional impact, while the environmental benefits of urban greening are primarily confined to the areas where green spaces are established. Consequently, no spatial spillover effects are observed for these two factors.
Conclusions and policy recommendation
Conclusions
First, the digital economy in Chinese provinces exhibited consistent enhancement from 2011 to 2020. The digital economy index ranks the regions as follows: Eastern, Central, Western, and Northeastern. The Central and Western areas saw the highest growth rates, followed by the Northeastern and Eastern regions. Second, the development of the digital economy has markedly improved carbon emission efficiency, with the results successfully undergoing robustness tests. The effect of the digital economy on carbon emission efficiency exhibits regional heterogeneity. Carbon emission efficiency is markedly enhanced in the Eastern, Central, and Western areas but not in the Northeastern region. Regions exhibiting advanced digital economy development discover much more significant positive impacts. Third, the enhancement in carbon efficiency is mainly influenced by upgrading industrial structure, optimizing energy structure, and innovation in green technology, with mediation impact proportions of 0.433, 0.373, and 0.465, respectively. Furthermore, spatial regression reveals that the advancement of the digital economy in one area positively influences the carbon efficiency of adjacent regions, indicating the existence of space spillover effects.
Policy recommendation
Specific measures are required to eliminate regional disparities in the advancement of the digital economy. The eastern regions should capitalize on its advantages to promote advanced technology and exchange experiences with adjacent areas. The central and western regions should incorporate digital transformation into industry and services, utilizing technology such as big data to facilitate low-carbon enhancements. The northeastern region must enhance its digital infrastructure, absorb insights from more developed areas, and leverage its climatic advantages.
The government should prioritize the progress of the digital economy, enhance the integration of digital and physical economies, and expedite the transition to smart manufacturing. Efforts must concentrate on optimizing the energy structure, refining resource allocation, increasing the share of renewable energy, and enhancing energy efficiency. Moreover, the distribution of advanced knowledge and interdisciplinary cooperation can reduce information asymmetry and promote innovation in green technology. The mediating variables mentioned above can significantly enhance the positive impact of the digital economy on carbon emission efficiency.
The government should foster cross-regional coordination to enhance emission reduction effects and prevent resource waste due to policy redundancies, emphasizing the spatial spillover effect of the digital economy on carbon emission efficiency. A digital regulatory platform must be created to enhance transparency in cross-domain environmental governance. The government should collaborate with enterprises, providing policy support for digital transformation and promoting voluntary engagement in emission reduction coalitions. Moreover, enhancing online channels for public participation will increase engagement in environmental monitoring.
Limitations and future work
This study examines the relationship between the digital economy and carbon emission efficiency, although it possesses several limitations. The selection of indicators might involve potential exclusions, indicating that evaluation measures and methodologies should be improved. Secondly, while this study is primarily empirical, subsequent research could incorporate conventional production functions and investigate multidisciplinary links with economics. Ultimately, additional potential mediating factors may modify this relationship, underscoring the necessity to perpetually refine the theoretical framework to encompass the various ways the digital economy impacts carbon emission efficiency.
Data availability
The data supporting the findings of this study are available in the China Statistical Yearbook at “https://www.stats.gov.cn/sj/ndsj/index.html”. These data were derived from publicly available resources at “https://data.stats.gov.cn/easyquery.htm?cn=E0103”.
References
Hosan, S., Karmaker, S. C., Rahman, M. M., Chapman, A. J. & Saha, B. B. Dynamic links among the demographic dividend, digitalization, energy intensity and sustainable economic growth: Empirical evidence from emerging economies. J. Clean. Prod. 330, 12. https://doi.org/10.1016/j.jclepro.2021.129858 (2022).
Pan, W. R., Xie, T., Wang, Z. W. & Ma, L. S. Digital economy: An innovation driver for total factor productivity. J. Bus. Res. 139, 303–311. https://doi.org/10.1016/j.jbusres.2021.09.061 (2022).
Dwivedi, Y. K. et al. Climate change and COP26: Are digital technologies and information management part of the problem or the solution? An editorial reflection and call to action. Int. J. Inf. Manage 63, 39. https://doi.org/10.1016/j.ijinfomgt.2021.102456 (2022).
Xing, Z. Y., Huang, J. & Wang, J. Unleashing the potential: Exploring the nexus between low-carbon digital economy and regional economic-social development in China. J. Clean. Prod. 413, 17. https://doi.org/10.1016/j.jclepro.2023.137552 (2023).
Yu, G. & Zhou, X. B. The influence and countermeasures of digital economy on cultivating new driving force of high-quality economic development in Henan Province under the background of “double circulation”. Ann. Oper. Res. 326, 31–31. https://doi.org/10.1007/s10479-021-04325-4 (2023).
Eiteneyer, N., Bendig, D. & Brettel, M. Social capital and the digital crowd: Involving backers to promote new product innovativeness. Res. Policy 48, 15. https://doi.org/10.1016/j.respol.2019.01.017 (2019).
Wei, Q. et al. Spatiotemporal analysis of carbon emission efficiency across economic development stages and synergistic emission reduction in the Beijing-Tianjin-Hebei region. J. Environ. Manage 377, 124609. https://doi.org/10.1016/j.jenvman.2025.124609 (2025).
Li, Y. J., Li, N. & Li, Z. B. Evolution of carbon emissions in China’s digital economy: An empirical analysis from an entire industry chain perspective. J. Clean. Product. 414, 137419. https://doi.org/10.1016/j.jclepro.2023.137419 (2023).
Xu, Y., Li, X. Y., Tao, C. Q. & Zhou, X. Connected knowledge spillovers, technological cluster innovation and efficient industrial structure. J. Innov. Knowl. 7, 11. https://doi.org/10.1016/j.jik.2022.100195 (2022).
Li, Z. H., Yuan, B. B., Wang, Y., Qian, J. W. & Wu, H. T. The role of digital finance on the synergistic governance of pollution & carbon: Evidence from Chinese cities. Sustain. Cities Soc. 115, 18. https://doi.org/10.1016/j.scs.2024.105812 (2024).
Dong, Z. M. & Yao, S. J. Digital finance reduces urban carbon footprint pressure in 277 Chinese cities. Sci. Rep. 14, 16. https://doi.org/10.1038/s41598-024-67315-z (2024).
Lin, B. Q. & Zhou, Y. C. Does the Internet development affect energy and carbon emission performance?. Sustain. Prod. Consump. 28, 1–10. https://doi.org/10.1016/j.spc.2021.03.016 (2021).
Wu, H. T., Hao, Y., Ren, S. Y., Yang, X. D. & Xie, G. Does internet development improve green total factor energy efficiency? Evidence from China. Energy Policy 153, 13. https://doi.org/10.1016/j.enpol.2021.112247 (2021).
Ren, S. Y., Hao, Y., Xu, L., Wu, H. T. & Ba, N. Digitalization and energy: How does internet development affect China’s energy consumption?. Energy Econ. 98, 20. https://doi.org/10.1016/j.eneco.2021.105220 (2021).
Zhou, X. Y., Zhou, D. Q., Wang, Q. W. & Su, B. How information and communication technology drives carbon emissions: A sector-level analysis for China. Energy Econ. 81, 380–392. https://doi.org/10.1016/j.eneco.2019.04.014 (2019).
Bai, L. et al. Effects of digital economy on carbon emission intensity in Chinese cities: A life-cycle theory and the application of non-linear spatial panel smooth transition threshold model. Energy Policy 183, 13. https://doi.org/10.1016/j.enpol.2023.113792 (2023).
Yu, K. & Li, Z. Coupling coordination and spatial network characteristics of carbon emission efficiency and urban green innovation in the Yellow River Basin, China. Sci. Rep. 14, 27690. https://doi.org/10.1038/s41598-024-78099-7 (2024).
Yang, Z. et al. Digitalization and carbon emissions: How does digital city construction affect china’s carbon emission reduction?. Sustain. Cities Soc. 87, 14. https://doi.org/10.1016/j.scs.2022.104201 (2022).
Liu, J. L., Yu, Q. H., Chen, Y. Y. & Liu, J. G. The impact of digital technology development on carbon emissions: A spatial effect analysis for China. Resour. Conserv. Recycl. 185, 15. https://doi.org/10.1016/j.resconrec.2022.106445 (2022).
Yu, Z. P., Liu, Y., Yan, T. H. & Zhang, M. Carbon emission efficiency in the age of digital economy: New insights on green technology progress and industrial structure distortion. Bus. Strateg. Environ. 33, 4039–4057. https://doi.org/10.1002/bse.3683 (2024).
Shang, Y. P., Raza, S. A., Huo, Z., Shahzad, U. & Zhao, X. Does enterprise digital transformation contribute to the carbon emission reduction? Micro-level evidence from China. Int. Rev. Econ. Financ. 86, 1–13. https://doi.org/10.1016/j.iref.2023.02.019 (2023).
Medaglia, R., Rukanova, B. & Zhang, Z. Y. Digital government and the circular economy transition: An analytical framework and a research agenda. Gov. Inf. Q. 41, 13. https://doi.org/10.1016/j.giq.2023.101904 (2024).
Zhang, W., Liu, X. M., Wang, D. & Zhou, J. P. Digital economy and carbon emission performance: Evidence at China’s city level. Energy Policy 165, 22. https://doi.org/10.1016/j.enpol.2022.112927 (2022).
Wei, T. Y., Chen, M., Wang, F. R. & Cheng, S. L. The impact of public demands on local environmental governance performance: Evidence from civil environmental complaints placed on leaders at different government levels in China. J. Environ. Manage. 360, 14. https://doi.org/10.1016/j.jenvman.2024.121216 (2024).
Zhang, Q. M. & Zhao, X. Can the digital economy facilitate the optimization of industrial structure in resource-based cities?. Struct. Chang. Econ. Dyn. 71, 405–416. https://doi.org/10.1016/j.strueco.2024.08.010 (2024).
Chang, H., Ding, Q. Y., Zhao, W. Z., Hou, N. & Liu, W. W. The digital economy, industrial structure upgrading, and carbon emission intensity—empirical evidence from China’s provinces. Energy Strateg. Rev. 50, 101218. https://doi.org/10.1016/j.esr.2023.101218 (2023).
Wang, S. H., Wang, X. Q. & Chen, S. S. Global value chains and carbon emission reduction in developing countries: Does industrial upgrading matter?. Environ. Impact Assess. Rev. 97, 14. https://doi.org/10.1016/j.eiar.2022.106895 (2022).
Pan, S. L., Carter, L., Tim, Y. & Sandeep, M. S. Digital sustainability, climate change, and information systems solutions: Opportunities for future research. Int. J. Inf. Manage. 63, 5. https://doi.org/10.1016/j.ijinfomgt.2021.102444 (2022).
Wang, J., Dong, K., Dong, X. & Taghizadeh-Hesary, F. Assessing the digital economy and its carbon-mitigation effects: The case of China. Energy Econ. 113, 14. https://doi.org/10.1016/j.eneco.2022.106198 (2022).
Haldar, A. & Sethi, N. Environmental effects of Information and Communication Technology - Exploring the roles of renewable energy, innovation, trade and financial development. Renew. Sust. Energ. Rev. 153, 14. https://doi.org/10.1016/j.rser.2021.111754 (2022).
Shahbaz, M., Wang, J. D., Dong, K. Y. & Zhao, J. The impact of digital economy on energy transition across the globe: The mediating role of government governance. Renew. Sust. Energ. Rev. 166, 14. https://doi.org/10.1016/j.rser.2022.112620 (2022).
Li, Z. G. & Wang, J. The dynamic impact of digital economy on carbon emission reduction: Evidence city-level empirical data in China. J. Clean. Prod. 351, 12. https://doi.org/10.1016/j.jclepro.2022.131570 (2022).
Yang, R., Tang, W. S. & Zhang, J. X. Technology improvement strategy for green products under competition: The role of government subsidy. Eur. J. Oper. Res. 289, 553–568. https://doi.org/10.1016/j.ejor.2020.07.030 (2021).
Zhang, H. K., Wu, J. C., Mei, Y. & Hong, X. Y. Exploring the relationship between digital transformation and green innovation: The mediating role of financing modes. J. Environ. Manage 356, 18. https://doi.org/10.1016/j.jenvman.2024.120558 (2024).
Yang, T. L., Li, F. M., Du, M., Huang, M. & Li, Y. N. Impacts of alternative energy production innovation on reducing CO2 emissions: Evidence from China. Energy 268, 10. https://doi.org/10.1016/j.energy.2023.126684 (2023).
Zhao, Z. Y. et al. Green innovation and carbon emission performance: The role of digital economy. Energy Policy 195, 27. https://doi.org/10.1016/j.enpol.2024.114344 (2024).
Balland, P. A. et al. The new paradigm of economic complexity. Res. Policy 51, 11. https://doi.org/10.1016/j.respol.2021.104450 (2022).
Eslami, H., Najem, S., Ghanem, D. A. & Ahmad, A. The potential of urban distributed solar energy in transition economies: The case of Beirut city. J. Environ. Manage 285, 10. https://doi.org/10.1016/j.jenvman.2021.112121 (2021).
Feng, Y. J. et al. Spatially-explicit modeling and intensity analysis of China’s land use change 2000–2050. J. Environ. Manage 263, 12. https://doi.org/10.1016/j.jenvman.2020.110407 (2020).
Wang, X. L., Zhang, T. Y., Nathwani, J., Yang, F. M. & Shao, Q. L. Environmental regulation, technology innovation, and low carbon development: Revisiting the EKC Hypothesis, Porter Hypothesis, and Jevons’ Paradox in China’s iron & steel industry. Technol. Forecast. Soc. Chang. 176, 12. https://doi.org/10.1016/j.techfore.2022.121471 (2022).
Hu, G. G. Is knowledge spillover from human capital investment a catalyst for technological innovation? The curious case of fourth industrial revolution in BRIGS economies. Technol. Forecast. Soc. Chang. 162, 7. https://doi.org/10.1016/j.techfore.2020.120327 (2021).
Shen, X. B. & Wang, Z. C. Can digital industrialization promote energy conservation development in China? Empirical evidence based on national big data comprehensive pilot zone policy. J. Environ. Manage 368, 17. https://doi.org/10.1016/j.jenvman.2024.122125 (2024).
Sun, W. & Huang, C. C. How does urbanization affect carbon emission efficiency? Evidence from China. J. Clean. Prod. 272, 9. https://doi.org/10.1016/j.jclepro.2020.122828 (2020).
Huang, L., Zhang, H. F., Si, H. Y. & Wang, H. Y. Can the digital economy promote urban green economic efficiency? Evidence from 273 cities in China. Ecol. Indic. 155, 12. https://doi.org/10.1016/j.ecolind.2023.110977 (2023).
Wang, H., Chen, Z. P., Wu, X. Y. & Niea, X. Can a carbon trading system promote the transformation of a low-carbon economy under the framework of the porter hypothesis? -Empirical analysis based on the PSM-DID method. Energy Policy 129, 930–938. https://doi.org/10.1016/j.enpol.2019.03.007 (2019).
Cui, S. T., Wang, R. L., Li, X. Y. & Bai, X. G. Policy-driven analysis of carbon emission efficiency under uncertainty and its application in Chinese industry: Hybrid delta-slacks-based model and ordinal priority approach. Energy 324, 135832. https://doi.org/10.1016/j.energy.2025.135832 (2025).
Peng, B. H. & Gao, F. Crafting the perfect policy combination: Exploring the synergistic effects of dual-pilot energy policies on urban carbon emission efficiency. Urban Clim. 59, 102260. https://doi.org/10.1016/j.uclim.2024.102260 (2025).
Chen, W., Wang, G., Xu, N., Ji, M. & Zeng, J. Promoting or inhibiting? New-type urbanization and urban carbon emissions efficiency in China. Cities 140, 104429. https://doi.org/10.1016/j.cities.2023.104429 (2023).
Ma, Q., Tariq, M., Mahmood, H. & Khan, Z. The nexus between digital economy and carbon dioxide emissions in China: The moderating role of investments in research and development. Technol. Soc. 68, 10. https://doi.org/10.1016/j.techsoc.2022.101910 (2022).
Jiang, H. Y., Elahi, E., Gao, M. T., Huang, Y. M. & Liu, X. R. Digital economy to encourage sustainable consumption and reduce carbon emissions. J. Clean. Prod. 443, 14. https://doi.org/10.1016/j.jclepro.2024.140867 (2024).
Zhang, J. N., Lyu, Y. W., Li, Y. T. & Geng, Y. Digital economy: An innovation driving factor for low-carbon development. Environ. Impact Assess. Rev. 96, 15. https://doi.org/10.1016/j.eiar.2022.106821 (2022).
Afolabi, J. A. Advancing digital economy in Africa: The role of critical enablers. Technol. Soc. 75, 11. https://doi.org/10.1016/j.techsoc.2023.102367 (2023).
Hu, J. J., Huo, D. & Wu, D. L. Digital economy development and venture capital networks: Empirical evidence from China. Technol. Forecast. Soc. Chang. 203, 15. https://doi.org/10.1016/j.techfore.2024.123338 (2024).
Paschou, T., Rapaccini, M., Adrodegari, F. & Saccani, N. Digital servitization in manufacturing: A systematic literature review and research agenda. Ind. Mark. Manage 89, 278–292. https://doi.org/10.1016/j.indmarman.2020.02.012 (2020).
Xue, Y., Tang, C., Wu, H. T., Liu, J. M. & Hao, Y. The emerging driving force of energy consumption in China: Does digital economy development matter?. Energy Policy 165, 18. https://doi.org/10.1016/j.enpol.2022.112997 (2022).
Song, C. Z., Liu, Q. F., Song, J. P. & Ma, W. Impact path of digital economy on carbon emission efficiency: Mediating effect based on technological innovation. J. Environ. Manage 358, 14. https://doi.org/10.1016/j.jenvman.2024.120940 (2024).
Du, J., Liang, L. & Zhu, J. A slacks-based measure of super-efficiency in data envelopment analysis: A comment. Eur. J. Oper. Res. 204, 694–697. https://doi.org/10.1016/j.ejor.2009.12.007 (2010).
Wang, J. D., Dong, K. Y., Sha, Y. Z. & Yan, C. Envisaging the carbon emissions efficiency of digitalization: The case of the internet economy for China. Technol. Forecast. Soc. Chang. 184, 14. https://doi.org/10.1016/j.techfore.2022.121965 (2022).
Tan, L. S. et al. Toward low-carbon sustainable development: Exploring the impact of digital economy development and industrial restructuring. Bus. Strateg. Environ. 33, 2159–2172. https://doi.org/10.1002/bse.3584 (2024).
Lange, S., Pohl, J. & Santarius, T. Digitalization and energy consumption. Does ICT reduce energy demand?. Ecol. Econ. 176, 14. https://doi.org/10.1016/j.ecolecon.2020.106760 (2020).
Zhang, M. L. & Liu, Y. Influence of digital finance and green technology innovation on China’s carbon emission efficiency: Empirical analysis based on spatial metrology. Sci. Total Environ. 838, 15. https://doi.org/10.1016/j.scitotenv.2022.156463 (2022).
Preacher, K. J. & Hayes, A. F. SPSS and SAS procedures for estimating indirect effects in simple mediation models. Behav. Res. Methods Instrum. Comput. J. Psychonomic Soc. Inc 36, 717–731. https://doi.org/10.3758/bf03206553 (2004).
Ran, Q. Y., Yang, X. D., Yan, H. C., Xu, Y. & Cao, J. H. Natural resource consumption and industrial green transformation: Does the digital economy matter?. Resour. Policy 81, 13. https://doi.org/10.1016/j.resourpol.2023.103396 (2023).
Bai, T. T., Qi, Y., Li, Z. H. & Xu, D. Digital economy, industrial transformation and upgrading, and spatial transfer of carbon emissions: The paths for low-carbon transformation of Chinese cities. J. Environ. Manage 344, 12. https://doi.org/10.1016/j.jenvman.2023.118528 (2023).
Li, Z. H., Bai, T. T., Qian, J. W. & Wu, H. T. The digital revolution’s environmental paradox: Exploring the synergistic effects of pollution and carbon reduction via industrial metamorphosis and displacement. Technol. Forecast. Soc. Chang. 206, 14. https://doi.org/10.1016/j.techfore.2024.123528 (2024).
Song, Z. Y., Wang, C. & Bergmann, L. China’s prefectural digital divide: Spatial analysis and multivariate determinants of ICT diffusion. Int. J. Inf. Manage. 52, 12. https://doi.org/10.1016/j.ijinfomgt.2020.102072 (2020).
Chang, H., Ding, Q. Y., Zhao, W. Z., Hou, N. & Liu, W. W. The digital economy, industrial structure upgrading, and carbon emission intensity—Empirical evidence from China’s provinces. Energy Strateg. Rev. 50, 14. https://doi.org/10.1016/j.esr.2023.101218 (2023).
Sun, H. Y. & Kim, G. The composite impact of ICT industry on lowering carbon intensity: From the perspective of regional heterogeneity*. Technol. Soc. 66, 14. https://doi.org/10.1016/j.techsoc.2021.101661 (2021).
LeSage, J. & Pace, R. K. Introduction to spatial econometrics (Chapman and Hall/CRC, 2009).
Ma, Z. Y., Xiao, H., Li, J., Chen, H. T. & Chen, W. H. Study on how the digital economy affects urban carbon emissions. Renew. Sust. Energ. Rev. 207, 13. https://doi.org/10.1016/j.rser.2024.114910 (2025).
Wang, L. & Chen, L. Impacts of digital economy agglomeration on carbon emission: A two-tier stochastic frontier and spatial decomposition analysis of China. Sustain. Cities Soc. 95, 104624 (2023).
Funding
This research was funded by the Social Science Planning Project of Jiangxi Province: Evolution of Subject Influence and Government Action Mechanism in Network Public Opinion (Project No. 22GL39).
Author information
Authors and Affiliations
Contributions
Conceptualization, Yifen Xia.; Methodology, Yilin Qin.; Software, Yifen Xia.; Validation, Yuanzhuo Wu.; Formal analysis, Yifen Xia.; Investigation, Yuanzhuo Wu.; Resources, Chun Fu.; Data curation, Yilin Qin.; Writing—original draft preparation, Yifen Xia.; Writing—review and editing, Yifen Xia.; Visualization, Yifen Xia.; Supervision, Yuanzhuo Wu.; Project administration, Chun Fu.; Funding acquisition, Yuanzhuo Wu.
Corresponding author
Ethics declarations
Competing interest
The authors declare no competing interests.
Additional information
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
About this article
Cite this article
Xia, Y., Wu, Y., Qin, Y. et al. Mechanism and spatial spillover effect of the digital economy on carbon emission efficiency in Chinese provinces. Sci Rep 15, 19025 (2025). https://doi.org/10.1038/s41598-025-02184-8
Received:
Accepted:
Published:
Version of record:
DOI: https://doi.org/10.1038/s41598-025-02184-8
