Next Article in Journal
Appraisal of Sustainable Retrofitting of Historical Settlements: Less than 60% Unexpected Outcomes
Next Article in Special Issue
Research on the Configurations and Control Methods of a Hybrid System of Air-Source Heat Pumps and Gas Boilers for Space Heating: Simulation and Comparative Analysis
Previous Article in Journal
Navigating Barriers to Decarbonisation of UK’s Aviation Sector Through Green Hydrogen: A Multi-Scale Perspective
Previous Article in Special Issue
Adaptation of Existent Methods for Setting Decarbonization Targets in the Brazilian Power Sector Companies
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Driving Green Transformation Through the National Digital Economy Innovation Pilot: A Quasi-Experimental Study on Reducing Urban Energy Intensity in 282 Chinese Cities

1
School of Economics, Fujian Normal University, No. 8 Xuefu South Road, Shangjie Town, Minhou County, Fuzhou 350117, China
2
School of Economics, Management and Law, University of South China, Hengyang 421001, China
3
School of Business, University of Leicester, Brookfield, Leicester LE2 1RQ, UK
4
Group of Researchers Applying Physics in Economy and Sociology (GRAPES), Beauvallon, B-4031 Angleur, Belgium
5
Babes-Bolyai University, Str. Mihail Kogălniceanu 1, 400084 Cluj-Napoca, Romania
6
Department of Statistics and Econometrics, Bucharest University of Economic Studies, 15–17 Dorobanti Avenue, District 1, 010552 Bucharest, Romania
*
Author to whom correspondence should be addressed.
Sustainability 2025, 17(13), 5687; https://doi.org/10.3390/su17135687
Submission received: 10 May 2025 / Revised: 3 June 2025 / Accepted: 6 June 2025 / Published: 20 June 2025

Abstract

Drawing upon a quasi-natural experiment, this research investigates the influence of China’s National Digital Economy Innovation Development Pilot Policy on urban energy intensity. By examining a sample of 282 Chinese cities with the difference in differences (DID) approach, the findings provide robust empirical support for the proposition that digital economy pilot policies substantially reduce urban energy intensity. Furthermore, the policy’s effectiveness in lowering urban energy intensity differs across cities with varying administrative levels and population scales. The results suggest that the policy’s impact is more pronounced in ordinary cities (non-provincial capitals/municipalities) and in those with smaller populations. An examination of the underlying mechanisms reveals three principal pathways through which the policy affects energy consumption: (1) digital economic development, which promotes optimal resource allocation and enhanced energy intensity; (2) technological innovation, driving advances in green technologies and supporting sustainable industrial upgrades; and (3) economic agglomeration, which leverages economies of scale and industrial clustering to bolster energy efficiency. The conclusions underscore the necessity of expanding digital economy pilot zones, strengthening investments in digital infrastructure, and fostering greater technological innovation to sustain improvements in energy efficiency and environmental performance.

1. Introduction

Since the 1970s, scholars have increasingly sought to integrate environmental science and the social sciences, thus broadening our understanding of the multiple effects that environmental risks exert on societal functioning and development. This interdisciplinary endeavor ultimately gave rise to the field of environmental sociology, which has swiftly permeated various domains within the social sciences [1]. In particular, Meadows et al.’s seminal research established the fundamental necessity to examine the relationship between economic growth and ecological sustainability. Their research has coalesced around the principal perspectives of “The Limits to Growth”: the relationship between economic growth and ecological protection is characterized by negative conflict [2]. However, another proposition formulated by Grossman and Krueger suggests that an optimal development trajectory would allow a nation to pursue both economic expansion and environmental protection in parallel [3]. Of course, realizing such an ideal scenario remains challenging, and in a global context where sustainable development is increasingly emphasized, reconciling economic progress with environmental stewardship has emerged as a critical concern [4,5].
Utilizing China as a case in point, it is evident that the nation’s remarkable economic progress has been accompanied by elevated levels of energy consumption. As the largest energy consumer globally, China’s energy demands are anticipated to continue rising. According to the data released by the International Energy Agency (IEA), in 2022, China’s per capita CO2 emissions from energy consumption per capita is approximately 7.515 tons, representing merely 54.41% of the corresponding figure in the United States, thereby indicating significant potential for future growth in energy demand (data sources: https://www.iea.org/ (accessed on 4 April 2025)). Concurrently, China’s Statistical Communiqués on National Economic and Social Development in recent years signify a deceleration in the rate of urban energy conservation. This assertion is corroborated by the extant data, which demonstrate a decline in energy consumption per unit of gross domestic product (GDP) from 3.6% in 2012 to 0.5% in 2023 (data sources: https://www.stats.gov.cn/ (accessed on 4 April 2025)). This trend suggests that conventional energy-saving measures may have reached a plateau in terms of effectiveness. Against this backdrop, identifying innovative and efficacious approaches to reducing energy intensity has become pivotal to China’s sustainable development aspirations and its quest to achieve its “dual carbon” targets.
In parallel, the proliferation of digital technologies characterized by the application of big data analytics, distributed cloud architectures, and neural network-driven AI systems is redefining value chains across industries, with the digital economy now serving as a tectonic force in reshaping global production networks. Fueled by technological breakthroughs in digital innovation, the digital economy is fundamentally redefining global economic paradigms and social organization structures. Beyond driving economic growth, the digital economy facilitates digital transformation across energy supply and demand systems, demonstrating significant potential for energy conservation and emission reduction [6,7]. For instance, digital tools can be harnessed to optimize production procedures, enhance resource allocation efficiency, and reduce energy expenditure. Recognizing this potential, a growing number of nations have been proactively promoting the digital economy as a strategic pillar in their pursuit of sustainable development goals. As indicated by Jiao et al., 33 major emitting countries are leveraging the digital economy to control energy intensity [8].
China, similarly, acknowledges the pivotal role of the digital economy in fostering high-quality economic growth. In 2019, China promulgated the Implementation Plan for National Digital Economy Innovation Development Experimental Zones (hereinafter referred to as “the Plan”) (policy origin: https://www.gov.cn/xinwen/2019-10/20/content_5442574.htm (accessed on 4 April 2025)). The Plan’s initiative aimed to explore innovative paradigms and institutional frameworks for advancing the digital economy. Six regions, Hebei, Zhejiang, Fujian, Guangdong, Chongqing, and Sichuan, were chosen as national-level experimental sites under this scheme. Each local government was encouraged to harness its unique endowments and developmental trajectories to pilot a range of endeavors in areas such as the circulation mechanisms of digital economic factors; resource allocation; and the establishment of industrial clusters. The overarching objective of these experimental zones was to unleash the untapped potential of the digital economy in accelerating innovation and development.
In broad terms, location-based pilot policies are often seen as instruments through which central or local governments systematically gather and analyze the outcomes of policy experimentation, thereby informing subsequent political decisions and policy implementation. Such policy measures constitute key levers for steering the fulfillment of designated developmental goals. In this context, well-crafted policies become crucial drivers of energy efficiency enhancement and sustainable urban development [9]. Historical precedents suggest that China has rolled out a number of national pilot policies specifically designed to tackle rising urban energy intensity, including the low-carbon city pilot policy, the smart city pilot policy, and the carbon trading pilot policy. Existing research into these policy initiatives generally confirms their effectiveness.
(1)
Low-carbon city pilot policy: Empirical evidence indicates that implementing this policy has significantly reduced urban energy wastage and environmental degradation, culminating in enhanced carbon emission efficiency [10,11,12].
(2)
Smart city pilot policy: This policy is instrumental in achieving substantial energy savings in the urban production and living sectors, while concomitantly affecting a decline in per capita carbon dioxide emissions [13,14].
(3)
Carbon trading pilot policy: This policy mechanism has effectively tackled the longstanding problems of low energy efficiency and high emissions, resulting in significant reductions in energy intensity and carbon emissions of the pilot cities where it has been implemented [15].
Noticeably, the synergistic impacts of pilot policies, i.e., when multiple pilot policies operate in tandem, appear to generate additive or even synergistic reductions in energy consumption and carbon footprint, as illustrated by the integration of smart city and low-carbon frameworks [16].
Within this framework, the National Digital Economy Innovation Development Pilot emerges as another strategically vital location-based policy. Evaluating its capacity to drive urban energy-saving outcomes has clear implications for expanding its geographical scope and refining the policy measures inherited from the pilot zones.
From existing studies, since the digital economy emerged as an academic focus, research on its emission reduction effects has increased rapidly. Among these, the majority concentrate on analyzing the direct linkages between the development of the digital economy or digital technologies and urban carbon emissions. As Xiao et al. recently demonstrated, innovations in and applications of digital technologies significantly contribute to improving urban total-factor energy efficiency, thereby promoting sustainable economic development [17]. Some studies have also examined the impact of national digital economy-related policies on urban carbon emissions. For instance, as an analysis based on China’s National Big Data Comprehensive Pilot Zone demonstrated, this pilot exhibits significant carbon emission reduction effects [18]. However, these studies failed to achieve comprehensive research outcomes. Specifically, the former fails to capture the policy impacts of digital economy-related policies; as the National Big Data Comprehensive Pilot Zone is an institutional innovation and practical exploration platform centered on data elements, the latter’s policy research also focuses solely on the perspective of data elements, thereby failing to comprehensively reflect the economic activities and innovation behaviors of the digital economy. Therefore, despite the current abundance of related research efforts, significant gaps remain in understanding the policy impacts of the digital economy.
As a national policy, the National Digital Economy Innovation Development Pilot encompasses both digital economic activities and stimulates regional digital technology innovation vitality. Adopting it as the research subject allows for a more comprehensive understanding of the impacts of digital economy policies and related policy measures. However, existing studies on the National Digital Economy Innovation Development Pilot remain relatively limited, with most current research prioritizing descriptive evaluations over in-depth causal analyses. In particular, whether the national digital economy innovation development policies contribute to urban energy conservation, and if so, through what specific pathways, requires further investigation.
Against this backdrop, our paper seeks to address these lacunae by rigorously examining whether the National Digital Economy Innovation Development Pilot exerts a causal influence on reducing urban energy intensity. Drawing on panel data from 282 inland Chinese cities spanning the period from 2011 to 2021 and applying a difference in differences (DID) estimation strategy, our study endeavors to estimate the policy’s effects with greater accuracy. We further investigate variations in the policy’s effectiveness across different types of cities and attempt to analyze the potential reasons behind these discrepancies. The findings are expected to inform policy recommendations on how to leverage digital economic reforms to improve energy efficiency and reinforce sustainable development in China. The findings are also expected to have worldwide applications, mutatis mutandis.
Therefore, from a broader perspective, our study aspires to make three key contributions to the current literature. First, as noted earlier, China’s National Digital Economy Innovation and Development Policy provides a more comprehensive framework for capturing the integrated impact effects of digital economic activities and their associated policy interventions. However, empirical research in this domain remains notably scarce. Our study is among the few empirical investigations specifically examining the impact of national digital economy innovation development pilot programs on urban energy intensity, providing a novel empirical foundation for evaluating policy effectiveness. Second, by exploring both the heterogeneity of outcomes across diverse urban contexts and the transmission mechanisms underpinning any identified effects, our paper aims to deepen our understanding of how this pilot policy functions and how its influence may vary under different local conditions. Such insights can be instrumental in assisting policymakers in designing more targeted and practical measures. Finally, although the focus of our inquiry is on China, the lessons drawn from these findings may well be relevant to other countries or regions where digital economy initiatives are underway and energy-efficiency improvement remains a priority. By situating our study in the broader framework of sustainable development, we believe it can serve as a reference point for international contexts seeking to harmonize economic advancement with environmental protection.

2. Theoretical Foundations and Hypotheses

The establishment of the National Digital Economy Innovation Development Pilot seeks to stimulate growth in the digital sector while promoting the digitalization of traditional industries through systematic experimentation and innovative practices (policy origin: https://www.gov.cn/xinwen/2019-10/20/content_5442574.htm (accessed on 1 January 2025)). This strategic undertaking not only catalyzes the discovery of new production factors and the establishment of efficient allocation mechanisms for data-driven resources but it also fosters production relationships better aligned with the evolution of digital productivity. Likewise, it aims to develop governance structures and administrative frameworks that accommodate the exigencies of emerging business models. Within this context, advancing digital economic innovation and reducing urban energy intensity have become a pivotal issue in the quest for sustainable development. Informed by the theoretical underpinnings of extant studies and the policy objectives articulated in The Plan, our paper posits that the National Digital Economy Innovation Development Pilot Policy facilitates urban energy conservation through several channels, including digital economic growth, technological innovation, and economic agglomeration. Accordingly, the ensuing sections examine the pathways through which this pilot policy influences reductions in urban energy intensity, with particular emphasis on its roles in fostering the digital economy, spurring technological innovation, and enhancing economic agglomeration. The transmission mechanism is illustrated in Figure 1.

2.1. Digital Economy Effects

As set out in The Plan, one of the principal objectives of the National Digital Economy Innovation Development Pilot Zones is to integrate emerging technologies, such as the Internet of Things, artificial intelligence, and big data, into the real economy. This process bolsters both the digitization of existing industries and the industrialization of digital sectors, thus substantially contributing to the development of the regional digital economy. A broad array of research indicates that the digital economy has demonstrated significant structural effects in reducing urban carbon emissions and enhancing carbon efficiency, emerging as a critical force in guiding cities towards greener developmental trajectories [19,20,21,22].
From a demand-side perspective, digital economic innovations alter consumer behavior, promoting postmodern, collaborative consumption models that minimize energy waste across multiple domains [23]. From a supply-side perspective, digital economic development enhances the efficient circulation and utilization of productive resources such as labor, capital, and data [24]. Cutting-edge innovations, including cloud computing, big data analytics, and the Internet of Things, enable enterprises to streamline their production processes, rendering them increasingly automated and intelligent. Through the deep integration of live data acquisition systems and AI-enabled production protocols, both public institutions and corporate entities can establish energy system optimization frameworks to achieve systemic enhancement in production energy efficiency [25].
Taken together, the advancement of the digital economy is indispensable to the success of the National Digital Economy Innovation Development Pilot. Given that digital economic expansion effectively mitigates negative environmental externalities, alleviates pressures related to pollution and resource constraints, and contributes to lowering urban energy intensity, the following hypothesis is proposed:
H1. 
The National Digital Economy Innovation Development Pilot Policy reduces urban energy intensity by fostering the development of the digital economy.

2.2. Technological Innovation Effects

Technological innovation is widely recognized as the core driver for promoting green urban transition and reducing the energy intensity of cities [26,27]. The Plan stipulates that each pilot zone should prioritize digital core technologies to vigorously propel the development of innovative digital economies. Consequently, these zones often have technological bottlenecks, endorsing diverse supportive policies and investment initiatives to create a fertile environment for innovation-oriented actors. Existing empirical evidence demonstrates that a sound development environment fosters the quantitative expansion and qualitative improvement of green technology innovation [28]. Specifically, the availability of high-quality information and data empowers enterprises to systematically collect, analytically process, and strategically utilize actionable intelligence, thereby optimizing resource allocation efficiency and catalyzing innovation-oriented operational activities [29,30,31].
Furthermore, technology innovation is recognized as a pivotal catalyst for enhancing urban energy efficiency, demonstrating substantial spatial spillover effects that amplify decarbonization outcomes across metropolitan regions [32,33]. Firms with stronger innovative capabilities frequently invest in new technologies that elevate energy efficiency and minimize pollution, thereby reducing ecological stress [34]. Therefore, by strengthening regional technological innovation capacity, the National Digital Economy Innovation Development Pilot Policy can release greater potential for efficient energy use, ultimately lowering urban energy intensity. Thus, we posit the following hypothesis:
H2. 
The National Digital Economy Innovation Development Pilot Policy reduces urban energy intensity by unleashing regional technological innovation momentum.

2.3. Economic Agglomeration Effects

The relationship between economic agglomeration and carbon emissions is often described as a “U-shaped” relationship, but most studies indicate that manufacturing agglomeration at the current stage can effectively promote green development performance [35,36]. Within the National Digital Economy Innovation Development Pilot Zones, policy guidance encourages close interaction and collaboration among various economic actors and innovation stakeholders whilst also promoting a high-quality investment climate capable of attracting enterprises in emerging industries, high-caliber talent, and substantial capital. These conditions can consolidate new economic growth poles in a relatively short period, thereby elevating urban economic density and unlocking broader market opportunities.
This intensified economic agglomeration yields scale effects that help reduce urban energy intensity [37]. As industrial clusters solidify, the proximity of numerous technology-intensive and innovation-oriented enterprises allows for cost-effective technology exchanges and collective access to advanced equipment. This, in turn, facilitates the adoption of technologies, such as smart grids and renewable energy solutions, thus improving energy efficiency overall [38]. Furthermore, increased population density arising from economic agglomeration augments the utilization of public infrastructure, such as mass transit systems, reducing overall energy consumption and bolstering the urban environment’s green competitiveness [39].
In sum, by promoting the clustering of industrial activities and fostering denser populations, the National Digital Economy Innovation Development Pilot Policy escalates the level of economic agglomeration and consequently contributes to lower urban energy consumption. Hence, we offer the following hypothesis:
H3. 
The National Digital Economy Innovation Development Pilot Policy reduces urban energy intensity by enhancing the level of economic agglomeration.

3. Research Design

3.1. Model Setting

3.1.1. The Baseline Regression Model

As a widely used technique of policy evaluation, the difference in differences (DiD) estimator is extensively utilized to estimate the impact of exogenous shocks in empirical research [40,41]. The National Digital Economy Innovation Development Pilot, designed as a policy-driven quasi-experiment, offers exogenous shocks for evaluating counterfactual outcomes. Given that the DiD model requires at least one year of data both prior to and following the policy implementation, along with the availability of relevant data, our study assesses the causal effects of the pilot utilizing panel data from 282 inland cities in China spanning the years 2011 to 2021. In our analysis, the 61 cities located within the six designated pilot zones are classified as the policy-exposed cohort, while the remaining 221 non-pilot cities serve as the control group. The baseline regression model is established as follows:
G D i t = α 0 + α 1 C i · D t + α i X i t + γ t + μ i + ϵ i t
In Model (1), i and t represent the prefecture-level cities and years, respectively. G D i t is the explanatory variable, representing the energy intensity of the urban i in the t year. C i · D t is the core independent variable, where C i is a regional dummy variable that takes a value of 0 or 1, with C i = 1 indicating cities in the experimental group and C i = 0 denoting cities in the control group. D t is a time dummy variable, also taking a value of 0 or 1, where D t = 1 signifies the years following the implementation of the National Digital Economy Innovation Development Pilot (2019 and onwards), while D t = 0 refers to the years prior to its implementation (before 2019).
C i · D t is the interaction term between the regional dummy variable and the time dummy variable, with the estimated coefficient α 1 reflecting the impact of the National Digital Economy Innovation Development Pilot on urban energy intensity. X i t denotes the set of control variables; γ t denotes the year fixed effects; μ i signifies the individual fixed effects; and ϵ i t is the random error term.

3.1.2. Mechanism Test Model

The Plan is expected to reduce urban energy intensity through three pathways: the digital economy effect, the technological innovation effect, and the economic agglomeration effect. To verify hypotheses H1–H3, we employ the stepwise regression method to test for mediation effects. The mediation effect model is set up in the following manner:
M E D i t = β 0 + β 1 C i · D t + β i X i t + γ t + μ i + ϵ i t
Thus
G D i t = θ 0 + θ 1 C i · D t + θ 2 M E D i t + θ i X i t + γ t + μ i + ϵ i t
In this model, M E D i t represents the mediating variable, reflecting the level of digital economy development, regional green innovation capacity, and the degree of economic agglomeration, while in this model, M E D i t represents the mediating variable, reflecting the developmental level of the digital economy, regional green innovation capacity, and the degree of economic agglomeration, while the other variables remain consistent with Model (1).

3.2. Variable Explanations

3.2.1. Dependent Variable and Independent Variable

Our study draws on existing relevant research methodologies and utilizes energy consumption per unit of GDP (GD) to quantify urban energy intensity. Energy consumption per unit of GDP (tons of standard coal per CNY ten thousand (all GDP indicators used in this study are calculated based on real GDP, with 2010 as the base year)) is defined as the ratio of total energy consumption (EC) to regional gross domestic product. The total energy consumption calculated in this paper encompasses the sum of electricity usage, gas (including both coal gas and natural gas), and liquefied petroleum gas consumption, which are then converted into tons of standard coal using the standard coal conversion coefficients for various energy types, following the methodology outlined below:
E C i t = U E i t · φ + P C i t · ω + L G i t · δ
In this model, E C i t represents the total energy consumption of city i in year t ; U E i t denotes the total electricity consumption of city i in year t (in ten thousand kilowatt-hours); P C i t indicates the total gas supply of city i in year t (in ten thousand cubic meters); L G i t signifies the total liquefied petroleum gas supply of city i in year t (in tons); φ is the conversion coefficient for electricity to standard coal (1.229 tons of standard coal per ten thousand kilowatt-hours); ω represents the conversion coefficient for gas to standard coal (13.3 tons of standard coal per ten thousand cubic meters); and δ is the conversion coefficient for liquefied petroleum gas to standard coal (1.7143 tons of standard coal per ton); the conversion coefficients are all sourced from the China Energy Statistical Yearbook. Furthermore, to eliminate the influence of price factors, the regional gross domestic product is adjusted to real terms using 2010 as the base year. A higher value of energy consumption per unit of GDP (GD) indicates that more energy is required to generate an additional unit of output.
C i · D t is constructed as the product of regional dummies and time dummies, and its coefficient estimate α 1 reflects the effect of the pilot policy on the energy intensity of the city, with the values as previously mentioned.

3.2.2. Mediating Variables

(1)
Level of Digital Economy Development (DED). Leveraging available data, our study quantifies the digital economy’s advancement across Chinese cities through two pivotal aspects: internet development and inclusive digital finance. The measurement includes five indicators: the number of internet users per hundred people, the proportion of employees engaged in computer services and software to total urban employment, per capita telecommunications volume, the number of mobile phone users per hundred people, and the China Inclusive Financial Index jointly compiled by the Peking University Digital Finance Research Center and Ant Financial Group. Drawing on the measurement method proposed by Xin et al., the entropy weight method is employed to calculate the developmental level of the digital economy for each city [42]. The weights calculated using the entropy weight method are found in Table 1.
(2)
Regional Green Innovation Capacity (RGIC). Since patent grants better reflect the true level of regional innovation, our study uses the number of green patents granted per ten thousand people to gauge the regional green innovation capacity [43].
(3)
Degree of Economic Agglomeration (DEA). Economic density is an indicator that measures the intensity of urban economic activities and the level of economic agglomeration, representing the economic value generated per unit area in a region. Drawing on methods from related research, our study employs the non-agricultural GDP per unit of administrative area as a proxy variable for economic agglomeration [44,45].

3.2.3. Control Variables

To enhance the accuracy of the causal effects of policies, our study controls for other variables that may influence regional innovation capacity based on existing research. These variables include the following:
(1)
Government Expenditure on Science and Education (GESE) is measured as the ratio of local fiscal expenditure on science and education to gross regional product.
(2)
Level of Informatization (INFO) is measured by the ratio of total postal and telecommunications volume to regional GDP.
(3)
Market activity (MA) is indicated by the ratio of the number of urban private and individual practitioners to the total number of employed individuals at year end.
(4)
Level of Financial Development (FINA) is assessed by the ratio of the year-end balance of financial institution loans and deposits to regional GDP.
(5)
Industrial structure (IS) is measured by the ratio of the tertiary industry to that of the secondary industry.

3.3. Data Description and Data Source

This research utilizes panel data from 282 inland cities in China spanning the years 2011 to 2021 as the research sample. The variable data primarily originates from the statistical yearbooks of Chinese prefecture-level cities, the “China Urban Statistical Yearbook”, the “China Urban Construction Statistical Yearbook”, and the “China Regional Economic Statistical Yearbook”. Additionally, the Chinese Research Data Services (CNRDS) database provides the green patent data employed in this study. Referring to the method of Zhang et al., the descriptive statistical results of the variables are presented in Table 2 [46].

4. Results and Discussion

4.1. Benchmark Regression Results

The results of the baseline regression analysis corresponding to Model (1) are reported in Table 3. The findings indicate that with the gradual inclusion of control variables, the implementation of the National Digital Economy Innovation Development Pilot Policy significantly reduced urban energy intensity. Referring to the final column, the regression coefficient for the energy intensity of the cities associated with the policy is negative, which passes the significance test at the 1% level. This implies that the implementation of the National Digital Economy Innovation Development Pilot Policy has effectively contributed to a reduction in energy intensity in the pilot cities, providing preliminary evidence of its efficacy in lowering urban energy intensity.

4.2. Parallel Trend Test

The application of the difference in differences method is predicated on the assumption that the experimental group and the control group satisfy the similar developmental trends assumption prior to the policy implementation. In our study, the parallel trends assumption signified that the energy intensity trends of the pilot cities and non-pilot cities remained parallel before the initiation of the National Digital Economy Innovation Development Pilot Policy. However, after the policy was implemented, a significant divergence in energy intensity emerged between the two groups, disrupting the previously established parallel trend. Our study refers to the approach of Alder et al. [47] and employs the event study method to analyze the parallel trends of energy consumption per unit of GDP across the sample, with the testing results illustrated in Figure 2.
As illustrated in Figure 2, the coefficients for urban energy intensity prior to the pilot policy are predominantly centered around zero, with confidence intervals that encompass zero. This indicates that before 2019, the energy intensity trends of both pilot and non-pilot cities exhibited relative stability, showing no significant differences. However, after 2019, the estimated coefficients for energy intensity approach −0.03, and the confidence intervals no longer include zero. This signifies a marked divergence in energy intensity between the experimental and control cities following the implementation of the pilot policy, thereby satisfying the parallel trends assumption.

4.3. Robustness Test

4.3.1. The Placebo Test

To exclude the influence of other omitted variables and unknown factors, our study examines the robustness of the conclusion that the implementation of the National Digital Economy Innovation Development Pilot Policy has resulted in a reduction in urban energy intensity. First, 61 cities are randomly selected from 282 to form the pseudo-experimental group, with the remaining cities forming the pseudo-control group. The pseudo-experimental group was then assumed to be the pilot cities, after which the benchmark regression was rerun to observe the results of the random sample. This process was repeated 1000 and 2000 times, respectively. Ultimately, we generated the distribution plots of the estimated coefficients for the National Digital Economy Innovation Development Pilot virtual variable C i · D t , as depicted in Figure 3 and Figure 4. From these figures, it is evident that the estimated coefficients of the policy variable C i · D t predominantly cluster around zero, indicating that the model specification has not overlooked any sufficiently significant influencing variables. Furthermore, the vast majority of random samples accepted the null hypothesis. In summary, the conclusions derived from the previous baseline regression are robust.

4.3.2. Deletion of Sample Values for the Year in Which the Policy Was Piloted

To ensure the robustness of the regression results, our study excludes sample values from policy pilot years. Since the National Digital Economy Innovation Development Pilot Policy was implemented in 2019, we removed the observations for all cities from that year and reevaluated the impact of the policy on urban energy intensity. The results are presented in column (1) of Table 4. Excluding data from the year of pilot policy implementation constitutes a counterfactual test. The results indicate that the impact coefficient of the pilot policy on urban energy intensity is statistically significantly negative at the 1% significance level. This demonstrates that the conclusions derived from baseline regression analysis remain valid under the counterfactual testing framework.

4.3.3. Replacement of Dependent Variables

Considering the close relationship between energy consumption and people’s livelihoods, this study refers to existing practices using the scale of energy consumption (the logarithm of total energy consumption (tons of standard coal)) and per capita energy consumption (tons of standard coal per ten thousand people) as the dependent variables. We re-examined the impact of the National Digital Economy Innovation Development Pilot Policy on urban energy intensity, with the results displayed in columns (2) and (3) of Table 4. The empirical results indicate that after replacing the explained variable, the impact coefficient of the pilot policy on the explained variable becomes negative and passes the significance tests at the 5% and 1% levels, respectively, demonstrating that the conclusion that the pilot policy reduces urban energy consumption remains valid.

4.3.4. PSM-DID Test (Propensity Score Matching-Based Differences in Differences Test)

As the selection of pilot cities may prioritize those with favorable development conditions, the National Digital Economy Innovation Development Pilot Policy under study could consider factors such as economic development level, informatization level, and financial support, rather than being completely randomly sampled. Consequently, using a difference in differences (DiD) model to estimate policy effects may inevitably introduce self-selection bias. Therefore, imitating the existing relevant literature, the study employs a propensity score matching-based difference in differences method (PSM-DiD) to verify the robustness of the National Digital Economy Innovation Development Pilot Policy in reducing urban energy intensity [48]. Utilizing covariates, such as government science and education expenditure, informatization level, market activity, industrial structure, and financial development level, we identified the most closely matched non-pilot cities for each pilot city. The initial sample was matched using a default kernel function and bandwidth matching method, and the double difference model was subsequently tested based on the matched sample, with the estimated results presented in column (4) of Table 4. After replacing the empirical estimation model with the PSM-DiD approach, the pilot policy exhibits an impact coefficient of −0.0132 on urban energy intensity, passing the statistical significance test at the 5% level. This mitigates endogeneity bias stemming from policymakers’ preference-based selection of pilot cities.

4.3.5. Excluding the Influence of Other Policies

Given that other national policies during the sample period may affect urban energy intensity and lead to invalid estimates, our study posits that the low-carbon city pilot policy, smart city pilot policy, and carbon trading pilot policy could impact urban energy intensity. To ensure the robustness of the estimated results, we introduced dummy variables for the low-carbon city pilot policy (LCcity), smart city pilot policy (Scity), and carbon trading pilot policy (CTcity) into the baseline regression model; the dummy variable takes a value of 1 when these policies are implemented; otherwise, it takes a value of 0. Policy origin: https://zfxxgk.ndrc.gov.cn/web/iteminfo.jsp?id=1070 (accessed on 1 April 2025) and https://www.ndrc.gov.cn/xxgk/zcfb/tz/201201/t20120113_964370.html (accessed on 1 January 2025). This allows us to investigate changes in the empirical results of the baseline regression model after accounting for and isolating the effects of these other policies, with results shown in columns (5) to (8) of Table 4. Columns (5) to (7) present results controlling for the low-carbon city pilot policy, smart city pilot policy, and carbon trading pilot policy individually, while column (8) includes results considering all three policies simultaneously. According to the regression results, even after considering the influence of other related policies, the inhibitory effect of the pilot policy on urban energy intensity is still significant at the 1% test level, indicating that the baseline regression results are robust.

4.4. Heterogeneity Test

Due to differences in administrative levels and population sizes among cities, the baseline regression results may overlook certain differential effects. Our study further examines the heterogeneity of the impact of the National Digital Economy Innovation Development Pilot Policy on reducing urban energy intensity based on city administrative levels and population sizes.
From the perspective of administrative level heterogeneity, special cities such as provincial capitals, sub-provincial cities, and municipalities enjoy inherent policy advantages. Compared to ordinary cities, these special cities possess better economic development potential, quality labor resources, and ample capital. As such, the impact of the National Digital Economy Innovation Development Pilot Policy may vary across pilot cities of different administrative levels. In our study, we categorize provincial capitals, sub-provincial cities, and municipalities as special cities, while other cities are considered ordinary cities to examine the influence of administrative level heterogeneity, with results displayed in columns (1) and (2) of Table 5.
Regarding population size heterogeneity, larger cities typically have a greater total energy consumption base, making it challenging to effectively alleviate urban energy consumption pressure in the short term. The impact of the National Digital Economy Innovation Development Pilot Policy on energy intensity may, therefore, be less pronounced in larger cities. In this analysis, cities with above-average population sizes are classified as large cities, while those with below-average populations are considered small cities. The corresponding results are reported in columns (3) and (4) of Table 5.
Table 5 indicates that, after distinguishing between special and ordinary cities, the National Digital Economy Innovation Development Pilot Policy does not exert a statistically significant effect on the energy intensity of special cities. However, for ordinary cities, the policy’s effect on energy intensity is −0.0210, which is statistically significant at the 1% level. Moreover, the policy’s effect on energy intensity is significantly negative for both large and small cities; however, the absolute value of the coefficient for small cities is greater than that for large cities. This indicates that the impact of the pilot policy on cities’ energy intensity may differ, depending on the administrative level and population size of the city in question.
This may stem from the fact that the energy structure in special and large cities is less susceptible to change, and these cities have a larger energy consumption base to meet the daily needs of residents, transportation, and production activities. Consequently, energy intensity in these cities is somewhat more resilient, implying that the policy has a lesser effect on urban energy intensity.

4.5. Mechanism Regression Model Robustness Test

To further examine the digital economic effects, technological innovation effects, and economic agglomeration effects of the National Digital Economy Innovation Development Pilot Policy on reducing urban energy intensity, our study employs the digital economic development level (DED), regional green innovation capability (RGIC), and economic agglomeration degree (DEA) as mediating variables to conduct stepwise regression tests on Models (2) and (3), with the regression results presented in Table 6.
From the perspective of the digital economic effect, as shown in columns (2) and (3) of Table 6, the regression coefficient of the National Digital Economy Innovation Development Pilot Policy on digital economic development level is 0.0057, which passes the significance test at the 5% level. Furthermore, the regression coefficient of the digital economic development level on urban energy intensity is −0.1649, passing the significance test at the 1% level. This indicates that the National Digital Economy Innovation Development Pilot Policy effectively promotes the development of the digital economy, thereby reducing urban energy intensity, thus validating H1. The underlying reason is that the National Digital Economy Innovation Development Pilot Policy is a strategic program designed to explore new approaches to developing China’s digital economy. Consequently, the policy prioritizes unlocking the digital economy development potential of the pilot cities, thereby facilitating their digital economic development. The development of the digital economy signifies the digital transformation of factors and enterprises, as well as industrial upgrading, which will significantly enhance the efficiency of production, consumption, exchange, and distribution in real space, thereby reducing energy loss in intermediate processes.
Regarding the technological innovation effect, columns (4) and (5) of Table 6 reveal that the regression coefficient of the National Digital Economy Innovation Development Pilot Policy on regional green innovation capability is positive at the 1% significance level. Additionally, the regression coefficient of regional green innovation capability on urban energy intensity is −0.0036, also passing the significance test at the 1% level. This demonstrates that the implementation of the National Digital Economy Innovation Development Pilot Policy enhances the green innovation capability of pilot cities, which in turn effectively reduces urban energy intensity, thus validating H2. The results indicate that the National Digital Economy Innovation Development Pilot Policy’s emphasis on “innovative development” logically leads to an improvement in regional green innovation capability. The enhancement of regional green innovation capability supports better integration of renewable energy, improves the utilization and stability of clean energy, and reduces urban reliance on resource-based development and high-energy consumption industrial chains, thereby lowering overall energy intensity.
From the perspective of the economic agglomeration effect, columns (6) and (7) of Table 6 show that the regression coefficient of the National Digital Economy Innovation Development Pilot Policy on economic agglomeration degree is 0.1128, passing the significance test at the 1% level. Moreover, the regression coefficient of economic agglomeration degree on urban energy intensity is significantly negative at the 5% significance level. This suggests that the implementation of the National Digital Economy Innovation Development Pilot Policy intensifies economic agglomeration and, through the economies of scale associated with agglomeration, reduces urban energy intensity, thus validating H3. The requirement for a substantial amount of talent and capital to achieve innovative development in the digital economy makes the promotion of economic agglomeration by the National Digital Economy Innovation Development Pilot Policy unsurprising. However, as previously noted, there is controversy regarding the relationship between economic agglomeration and urban energy consumption. When density becomes excessively high, economic agglomeration may result in diseconomies of scale, leading to a sharp increase in urban energy pressure. Nevertheless, based on the findings of our study, it is evident that improving economic agglomeration degree has a positive effect on alleviating urban energy pressure. A possible explanation is that most pilot cities have not yet reached a saturated state of economic agglomeration, thereby achieving economies of scale.

5. Conclusions and Recommendations

5.1. Conclusions

Building upon a quasi-natural experimental setting, our study investigates the National Digital Economy Innovation Development Pilot Policy by drawing on panel data for 282 cities from 2011 to 2021, using 2019 as the baseline year for policy implementation. A difference in differences model is employed to analyze the policy’s effect on urban energy intensity, as well as to explore heterogeneity and potential transmission mechanisms. The empirical findings are summarized as follows:
(1)
Significant reduction in urban energy intensity: The establishment of National Digital Economy Innovation Development Pilot Zones yields a pronounced decrease in urban energy intensity. These results are robust across diverse estimation strategies, which lends credence to the argument that the trial regions could be expanded and that policy lessons gleaned from pilot cities may be beneficially adopted elsewhere.
(2)
Heterogeneous effects by administrative level and urban scale: The influence of the National Digital Economy Innovation Development Pilot on urban energy intensity varies according to the administrative hierarchy and population size. Specifically, the policy exerts a stronger effect in ordinary (non-provincial) municipalities and urban areas characterized by smaller populations.
(3)
Multiple pathways in reducing energy intensity: Further investigation reveals that the pilot policy reduces urban energy intensity via three channels: (1) digital economic effects, (2) technological innovation effects, and (3) economic agglomeration effects.

5.2. Recommendations

In light of the regression results and the conclusions of our study, the following four suggestions are offered to optimize the National Digital Economy Innovation Development Pilot Policy for reducing urban energy intensity in China.

5.2.1. Strengthen Top-Level Design and Clarify Strategic Positioning

Given the substantial energy-saving benefits identified in our study, it is recommended that the establishment of a comprehensive, tiered implementation system that goes beyond the pilot districts be expedited. Such a system should fully harness the capacity of the policy to elevate energy efficiency and restructure energy usage, thereby extending and amplifying the achievements realized in the pilot zones. This conclusion also offers valuable insights for other countries: similar digital economy development policies can create policy-driven comparative advantages. By enhancing the development level of the digital economy, promoting green technology innovation activities, and strengthening economic agglomeration, such policies could effectively reduce urban energy intensity, thereby accelerating the global timeline for achieving carbon peaking and carbon neutrality goals. Notably, the heterogeneity analysis demonstrates that digital economy development policies exhibit more pronounced carbon emission reduction effects in more populous cities. This finding holds positive implications for populous countries, such as India.

5.2.2. Enhance Policy Support for Digital Economy Development

A renewed emphasis on strengthening the digital economy’s role within the pilot zones is crucial. As the digitalization of industrial processes promotes efficient interconnectivity and spillovers of productive resources, it supports urban collaboration, raises energy efficiency, and ultimately curbs overall urban energy intensity. Consequently, the rollout of digital economy innovation development initiatives should facilitate a deep integration of energy systems with cutting-edge digital technologies, enabling a more precise equilibrium of energy supply and demand. It is equally important to leverage the foundational influence of the digital economy in urban energy planning and administration, guiding the digital transformation of production and mitigating energy wastage in these sectors.

5.2.3. Prioritize Core Technological Breakthroughs and Cultivate an Innovation-Friendly Environment

Technological breakthroughs in key areas should be fostered through a supportive ecosystem within the National Digital Economy Innovation Development Pilot Zones, thereby driving more robust innovation outcomes. As one of the most dynamic and wide-ranging sectors of scientific research, new technologies should be proactively deployed in urban planning, transport networks, and construction projects to substantially reduce urban energy intensity. Establishing a robust innovation system with well-equipped research facilities can facilitate breakthroughs in pivotal technologies and provide favorable support for both innovation-focused entities and specialized talent. Moreover, encouraging continuous advancements in new energy technologies will further promote the adoption of renewable energy within cities via intelligent, digitized frameworks, thereby increasing local rates of clean energy utilization.

5.2.4. Optimize Resource Allocation and Stimulate Economic Agglomeration

To amplify the energy-saving effects identified in the pilot zones, policymakers should strategically promote the convergence of diverse production factors, including human capital, investment, and technological resources. This necessitates introducing and refining institutional measures within the pilot zones that create appealing conditions to attract high-caliber talent and capital, thus undergirding the broader advancement of the digital economy and reductions in urban energy intensity. By making the most of digital resource advantages, the pilot zones can enhance the efficiency of resource distribution and expand the pathways for resource mobility, thereby spurring inter-regional or international cooperation. Streamlining the “arteries” for the flow of production factors fosters larger-scale industrial clustering, reinforcing the positive economies of scale inherent in economic agglomeration and paving the way for further reductions in urban energy consumption.

6. Study Limitations and Future Research

This study utilizes the difference in differences (DiD) method to examine the relationship between China’s National Digital Economy Innovation and Development Pilot Policy and urban energy intensity. The findings demonstrate a significant downward trend in energy intensity in pilot cities after the policy implementation. However, the study has limitations that require improvement. First, the sample coverage is limited. The dataset includes 282 Chinese cities from 2011 to 2021. Although the sample size is sufficient, the experimental group has fewer valid post-policy years (as the policy was initiated in 2019), which may affect the robustness of the conclusions. Second, the methodology needs refinement. The study employs a traditional DiD model, and future research could adopt more advanced approaches, such as double machine learning, to more accurately capture the policy effects. Therefore, subsequent studies should extend the sample timeframe based on data availability or introduce more innovative empirical methods to enhance analytical depth.

Author Contributions

Conceptualization, Q.L.; Methodology, Q.L.; Validation, Q.W. and M.A.; Investigation, C.S.; Resources, S.L.; Writing—original draft, Q.L.; Writing—review & editing, C.S. and M.A.; Visualization, Q.W.; Supervision, S.L. and M.A.; Project administration, Q.W. All authors have read and agreed to the published version of the manuscript.

Funding

Authors acknowledge support as follows: this research was supported by multiple funding sources. The Youth Talent Support Project for Publicity, Ideology and Culture of China (2021QNYC062) provided financial support specifically for the article publication fee. Additional funding and resource support throughout the research and publication process were generously provided by the following projects: the Social Science Foundation of Fujian (FJ2025B020), the Project of the Hunan Province Department of Education (24B0416), and the Social Science Foundation of Hunan Province (24YBQ035). These funds supported various stages of the article’s development, including data collection, analysis, and academic collaboration. S.L., Q.L., Q.W., C-Y. S. gratefully acknowledge the support provided by these foundations. Moreover, M.A.’s work was partially supported by the project “A better understanding of socio-economic systems using Quantitative Methods from Physics”, funded by the European Union---Next generation EU and the Romanian Government under the National Recovery and Resilience Plan for Romania, contract no.760034/23.05.2023, code PNRR-C9-I8-CF 255/29.11.2022, through the Romanian Ministry of Research, Innovation and Digitalization, within Component 9, “Investment I8”.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

Data are contained within the article.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Hannigan, J. Environmental Sociology; Routledge: London, UK; New York, NY, USA, 2006. [Google Scholar]
  2. Meadows, D.; Meadows, D.; Randers, J.; Behrens, W.W., III. The Limits to Growth; Universe Books: New York, NY, USA, 1972. [Google Scholar]
  3. Grossman, G.M.; Krueger, A.B. Economic growth and the environment. Q. J. Econ. 1995, 110, 353–377. [Google Scholar] [CrossRef]
  4. Kenny, J. Economic conditions and support for the prioritisation of environmental protection during the great recession. Environ. Polit. 2020, 29, 937–958. [Google Scholar] [CrossRef]
  5. McNeil, A.; Barnes, L. The environment-economic growth trade-off: Does support for environmental protection depend on its economic consequences? Ecol. Econ. 2025, 230, 108522. [Google Scholar] [CrossRef]
  6. Xue, Y.; Tang, C.; Wu, H.; Liu, J.; Hao, Y. The emerging driving force of energy consumption in China: Does digital economy development matter? Energy Policy 2022, 165, 112997. [Google Scholar] [CrossRef]
  7. Huang, J.; Wang, Y.; Luan, B.; Zou, H.; Wang, J. The energy intensity reduction effect of developing digital economy: Theory and empirical evidence from China. Energy Econ. 2023, 128, 107193. [Google Scholar] [CrossRef]
  8. Jiao, J.L.; Song, J.F.; Ding, T. The impact of synergistic development of renewable energy and digital economy on energy intensity: Evidence from 33 countries. Energy 2024, 295, 130997. [Google Scholar] [CrossRef]
  9. Usman, A.; Ozturk, I.; Hassan, A.; Zafar, S.M.; Ullah, S. The effect of ICT on energy consumption and economic growth in South Asian economies: An empirical analysis. Telemat. Inform. 2021, 58, 101537. [Google Scholar] [CrossRef]
  10. Song, M.L.; Zhao, X.; Shang, Y.P. The impact of low-carbon city construction on ecological efficiency: Empirical evidence from quasi-natural experiments. Resour. Conserv. Recycl. 2020, 157, 104777. [Google Scholar] [CrossRef]
  11. Liu, X.; Li, Y.; Chen, X.; Liu, J. Evaluation of low carbon city pilot policy effect on carbon abatement in China: An empirical evidence based on time-varying DID model. Cities 2022, 123, 103582. [Google Scholar] [CrossRef]
  12. Yu, Y.; Zhang, N. Low-carbon city pilot and carbon emission efficiency: Quasi-experimental evidence from China. Energy Econ. 2021, 96, 105125. [Google Scholar] [CrossRef]
  13. Guo, Q.; Wang, Y.; Dong, X. Effects of smart city construction on energy saving and CO2 emission reduction: Evidence from China. Appl. Energy 2022, 313, 118879. [Google Scholar] [CrossRef]
  14. Liu, K.; Meng, C.; Tan, J.; Zhang, G. Do smart cities promote a green economy? Evidence from a quasi-experiment of 253 cities in China. Environ. Impact Assess. 2023, 99, 107009. [Google Scholar] [CrossRef]
  15. Hong, Q.; Cui, L.; Hong, P. The impact of carbon emissions trading on energy efficiency: Evidence from quasi-experiment in China’s carbon emissions trading pilot. Energy Econ. 2022, 110, 106025. [Google Scholar] [CrossRef]
  16. Zhang, X.F.; Fan, D.C. Collaborative emission reduction research on dual-pilot policies of the low-carbon city and smart city from the perspective of multiple innovations. Urban. Clim. 2022, 47, 101364. [Google Scholar]
  17. Xiao, Y.; Duan, Y.; Zhou, H.; Han, X. Has digital technology innovation improved urban total factor energy efficiency?—Evidence from 282 prefecture-level cities in China. J. Environ. Manag. 2025, 378, 124784. [Google Scholar] [CrossRef]
  18. Pan, L.; Zhang, Y.; Li, J.; Lin, F. Digital Economy and Urban Carbon Emissions: A Quasi-Natural Experiment of the National Big Data Comprehensive Pilot Zone in China. Energy J. 2024, 45, 107–134. [Google Scholar] [CrossRef]
  19. Zhang, J.; Lyu, Y.; Li, Y.; Geng, Y. Digital economy: An innovation driving factor for low-carbon development. Environ. Impact Assess. 2022, 96, 106821. [Google Scholar] [CrossRef]
  20. 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. 2022, 68, 101910. [Google Scholar] [CrossRef]
  21. Shang, Y.; 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. 2023, 86, 1–13. [Google Scholar] [CrossRef]
  22. Liu, X.; Chen, L.; Lu, Y.; Chang, M.; Xiao, Y.; Yang, H.; Kong, D.; Zhang, L. Research on the impact of the digital economy on carbon emissions based on the dual perspectives of carbon emission reduction and carbon efficiency. Sci. Rep. 2025, 15, 3416. [Google Scholar] [CrossRef] [PubMed]
  23. Hamari, J.; Sjklint, M.; Ukkonen, A. The sharing economy: Why people participate in collaborative consumption. J. Assoc. Inf. Sci. Technol. 2016, 67, 2047–2059. [Google Scholar] [CrossRef]
  24. Lee, C.C.; Wang, F.H. How does digital inclusive finance affect carbon intensity? Econ. Anal. Policy 2022, 75, 174–190. [Google Scholar] [CrossRef]
  25. Ukoba, K.; Olatunji, K.O.; Adeoye, E.; Jen, T.C.; Madyira, D.M. Optimizing renewable energy systems through artificial intelligence: Review and future prospects. Energy Environ. 2024, 35, 3833–3879. [Google Scholar] [CrossRef]
  26. Zhang, F.; Deng, X.; Phillips, F.; Fang, C.; Wang, C. Impacts of industrial structure and technical progress on carbon emission intensity: Evidence from 281 cities in China. Technol. Forecast. Soc. 2020, 154, 119949. [Google Scholar] [CrossRef]
  27. Cheng, M.; Wang, J.; Yang, S.; Li, Q. The driving effect of technological innovation on green development: From the perspective of efficiency. Energy Policy 2024, 188, 114089. [Google Scholar] [CrossRef]
  28. Song, S.; Wen, J.; Li, Y.; Li, L. How does digital economy affect green technological innovation in China? New evidence from the “Broadband China” policy. Econ. Anal. Policy 2024, 81, 1093–1112. [Google Scholar] [CrossRef]
  29. Paunov, C.; Rollo, V. Has the internet fostered inclusive innovation in the developing world? World Dev. 2016, 78, 587–609. [Google Scholar] [CrossRef]
  30. Goldfarb, A.; Tucker, C. Digital Economics. J. Econ. Lit. 2019, 57, 3–43. [Google Scholar] [CrossRef]
  31. Ge, Y.Q.; Xia, Y.F.; Wang, T.Y. Digital economy, data resources and enterprise green technology innovation: Evidence from A-listed Chinese Firms. Resour. Policy 2024, 92, 105035. [Google Scholar] [CrossRef]
  32. Birol, F.; Keppler, J.H. Prices, technology development and the rebound effect. Energy Policy 2000, 28, 457–469. [Google Scholar] [CrossRef]
  33. Lu, X.H.; Lu, Z.R. How does green technology innovation affect urban carbon emissions? Evidence from Chinese cities. Energy Build. 2024, 325, 115025. [Google Scholar] [CrossRef]
  34. Wang, J.; Wang, S.; Li, S.; Cai, Q.; Gao, S. Evaluating the energy-environment efficiency and its determinants in Guangdong using a slack-based measure with environmental undesirable outputs and panel data model. Sci. Total Environ. 2019, 663, 878–888. [Google Scholar] [CrossRef]
  35. Yuan, H.; Feng, Y.; Lee, C.C.; Cen, Y. How does manufacturing agglomeration affect green economic efficiency? Energy Econ. 2020, 92, 104944. [Google Scholar] [CrossRef]
  36. Zheng, J.; Yuan, B.B.; Wu, J.L.; Chen, S. The impact of manufacturing agglomeration on green development performance: Evidence from the Yangtze river economic belt in China. J. Clean. Prod. 2024, 471, 143407. [Google Scholar] [CrossRef]
  37. Qu, C.; Shao, J.; Shi, Z. Does financial agglomeration promote the increase of energy efficiency in China? Energy Policy 2020, 146, 111810. [Google Scholar] [CrossRef]
  38. Wang, Y.; Yin, S.; Fang, X.; Chen, W. Interaction of economic agglomeration, energy conservation and emission reduction: Evidence from three major urban agglomerations in China. Energy 2021, 241, 122519. [Google Scholar] [CrossRef]
  39. Yu, Q.; Li, M.; Li, Q.; Wang, Y.; Chen, W. Economic agglomeration and emissions reduction: Does high agglomeration in China’s urban clusters lead to higher carbon intensity? Urban Clim. 2022, 43, 101174. [Google Scholar] [CrossRef]
  40. Hausman, J.; Kuersteiner, G.M. Difference in difference meets generalized least squares: Higher order properties of hypotheses tests. J. Econom. 2008, 144, 371–391. [Google Scholar] [CrossRef]
  41. de Chaisemartin, C.; D’Haultfoeuille, X. Fuzzy Differences-in-Differences. Rev. Econ. Stud. 2018, 85, 999–1028. [Google Scholar] [CrossRef]
  42. Xin, Y.; Song, H.; Shen, Z.; Wang, J. Measurement of the integration level between the digital economy and industry and its impact on energy consumption. Energy Econ. 2023, 126, 106988. [Google Scholar] [CrossRef]
  43. Fang, L.T.; Li, Z.H. Corporate digitalization and green innovation: Evidence from textual analysis of firm annual reports and corporate green patent data in China. Bus. Strategy Environ. 2024, 33, 3936–3964. [Google Scholar] [CrossRef]
  44. Sun, G.Z.; Gao, B.; Zhang, L.G. High-Speed Rail Opening, Economic Agglomeration, and Urban Innovation. Sage Open 2024, 14, 21582440241303255. [Google Scholar] [CrossRef]
  45. Wu, T. How can open public data promote efficient and equitable green production? Evidence from eco-efficiency in China. Environ. Sci. Pollut. Res. 2024, 31, 35173–35193. [Google Scholar] [CrossRef]
  46. Zhang, Y.J.; Wang, W. How does China’s carbon emissions trading (CET) policy affect the investment of CET-covered enterprises? Energy Econ. 2021, 98, 105224. [Google Scholar] [CrossRef]
  47. Alder, S.; Shao, L.; Zilibotti, F. Economic reforms and industrial policy in a panel of Chinese cities. J. Econ. Growth 2016, 21, 305–349. [Google Scholar] [CrossRef]
  48. Ouyang, X.; Fang, X.; Cao, Y.; Sun, C. Factors behind CO2 emission reduction in Chinese heavy industries: Do environmental regulations matter? Energy Policy 2020, 145, 111765. [Google Scholar] [CrossRef]
Figure 1. Sketch of the mechanisms by which the National Digital Economy Innovation Development Pilot Policy reduces urban energy intensity.
Figure 1. Sketch of the mechanisms by which the National Digital Economy Innovation Development Pilot Policy reduces urban energy intensity.
Sustainability 17 05687 g001
Figure 2. Parallel trend test estimated coefficient year variation.
Figure 2. Parallel trend test estimated coefficient year variation.
Sustainability 17 05687 g002
Figure 3. Nuclear density distribution plot of 1000 random samples; data: bold dots; fit: continuous line.
Figure 3. Nuclear density distribution plot of 1000 random samples; data: bold dots; fit: continuous line.
Sustainability 17 05687 g003
Figure 4. Nuclear density distribution plot of 2000 random samples; data: bold dots; fit: continuous line.
Figure 4. Nuclear density distribution plot of 2000 random samples; data: bold dots; fit: continuous line.
Sustainability 17 05687 g004
Table 1. Digital economy development index system.
Table 1. Digital economy development index system.
Primary
Indicator
Secondary
Indicators
DefinitionWeight
DEDInternet developmentThe number of broadband Internet users per 100 people0.1401
The ratio of the number of people employed in information transmission, computer services, and software to the number of people employed in regional societies0.1030
Per capita total telecom business0.2699
Mobile phone users per 100 people0.2726
Digital financial
inclusion
China’s Digital Inclusive Finance Index0.2144
Table 2. Descriptive statistics.
Table 2. Descriptive statistics.
VariableVariable Name (Units)MeanStd.Min.Q25MedianQ75Max.
GDEnergy intensity/energy
consumption per unit of GDP (tons of standard coal/CNY ten thousand)
0.10470.10300.00490.00840.08690.96582.1319
DEDThe digital economy development level0.18690.08580.00100.01230.17860.78190.8649
RGICRegional green innovation
capacity (items/ten thousand people)
1.03732.2624000.325324.691937.0361
DEAThe degree of economic
agglomeration (billion CNY per 100 square kilometers)
0.28960.65410.00230.00260.11438.921210.4773
GESEExpenditure on financial
science and education (%)
0.03760.01760.00790.00900.03340.14750.1508
INFOInformation level (%) 0.02380.01920.00230.00320.02000.24840.2943
MAMarket activity (%)1.30210.87220.00220.04461.11567.033217.1414
FINAFinancial development level (%)1.47760.70380.37110.39251.34775.181220.1002
ISIndustrial structure (%)1.05540.59370.11360..18880.91545.07215.3482
Table 3. Benchmark regression results corresponding to Model (1).
Table 3. Benchmark regression results corresponding to Model (1).
VariableGD
C i · D t −0.0196 ***−0.0195 ***−0.0191 **−0.0191 ***−0.0193 ***0.0058−0.0189 ***
(0.0058)(0.0058)(0.0058)(0.0058)(0.0058)(0.0057)(0.0059)
GESE −0.8079 ***−0.8889 **−0.8971 ***−0.8768 ***−1.3727 ***−0.9020 ***
(0.2193)(0.2201)(0.2202)(0.2283)(0.2319)(0.2329)
INFO 0.2479 ***0.2474 ***0.2483 ***0.2517 ***0.2466 ***
(0.0720)(0.0720)(0.0721)(0.0762)(0.0722)
MA 0.00210.00200.0118 ***0.0021
(0.0019)(0.0019)(0.0019)(0.0019)
FINA −0.00100.0035−0.0012
(0.0029)(0.0030)(0.0029)
IS 0.0537 ***0.0029
(0.0041)(0.0053)
cons0.0547 ***0.0832 ***0.0807 ***0.0784 ***0.0792 ***0.0467 **0.0784 ***
(0.0183)(0.0198)(0.0198)(0.0199)(0.0201)(0.0210)(0.0201)
Year fixedYesYesYesYesYesNoYes
City fixedYesYesYesYesYesYesYes
Obs3102310231023102310231023102
R20.69600.69750.69880.69890.69890.66040.6989
Note: figures in parentheses indicate a steady standard error; ** and *** are significant at 5% and 1%, respectively, which is the same as other below Tables.
Table 4. Results of the robustness test.
Table 4. Results of the robustness test.
VariableGD
(1)(2)(3)(4)(5)(6)(7)(8)
C i · D t −0.0193 ***−0.0848 **−0.1620 ***−0.0132 ** −0.0190 ***−0.0194 ***−0.0141 ** −0.0146 **
(0.0069)(0.0371)(0.0460)(0.0058)(0.059)(0.0059)(0.0062)(0.0062)
LCcity −0.0017 −0.0011
(0.0066) (0.0067)
Scity −0.0109 ** −0.0111 **
(0.0052) (0.0052)
CTcity −0.0144**−0.0147 **
(0.0059)(0.0059)
GESE−0.9215 ***−7.7279 ***4.4546 **−0.4449 *−0.9018 ***−0.9165 ***−0.9001 ***−0.9148 ***
(0.2601)(1.4693)(1.8235)(0.2656)(0.2329)(0.2328)(0.2327)(0.2327)
INFO0.2196 **0.8653 *1.4063 **0.1935 **0.2461 ***0.2446 ***0.2457 ***0.2434 ***
(0.0859)(0.4554)(0.5651)(0.0834)(0.0722)(0.0721)(0.0721)(0.0721)
MA0.0024−0.00260.01330.00210.00210.00210.00160.0016
(0.0019)(0.0117)(0.0146)(0.0023)(0.0019)(0.0019)(0.0019)(0.0019)
FINA−0.0071−0.0327 *−0.01270.0003−0.0012−0.0013−0.0016−0.0018
(0.0058)(0.0183)(0.0228)(0.0028)(0.0029)(0.0029)(0.0029)(0.0029)
IS0.00320.0179−0.2658 ***0.00320.00290.00260.00220.0019
(0.0061)(0.0335)(0.0416)(0.0054)(0.0053)(0.0053)(0.0053)(0.0053)
cons0.0931 *** 12.3980 ***0.13280.0587 ***0.0783 ***0.0746 ***0.0782 ***0.0742 ***
(0.0217)(0.1270)(0.1576)(0.0208)(0.0201)(0.0202)(0.0201)(0.0202)
Year fixedYesYesYesYesYesYesYesYes
City fixedYesYesYesYesYesYesYesYes
Obs28203102310227913102310231023102
R20.69130.90750.84580.72340.69890.69940.69960.7001
Note: figures in parentheses indicate a steady standard error; *, **, and *** are significant at 10%, 5%, and 1%, respectively.
Table 5. Results of the heterogeneity test.
Table 5. Results of the heterogeneity test.
Variable(1)(2)(3)(4)
Special CityOrdinary CitiesLarge CitiesSmall Cities
C i · D t −0.0001−0.0210 ***−0.0145 ***−0.0206 **
(0.0084)(0.0066)(0.0045)(0.0096)
GESE0.7063−0.9137 ***−0.3073−1.2143 ***
(0.5334)(0.2507)(0.2222)(0.3437)
INFO0.8896 ***0.1595 **−0.01750.4094 ***
(0.0973)(0.0812)(0.0569)(0.1150)
MA0.00390.00220.00060.0063
(0.0047)(0.0020)(0.0011)(0.0040)
FINA−0.0055−0.00310.0059−0.0026
(0.0054)(0.0032)(0.0058)(0.0037)
IS0.01170.0064−0.0096 *0.0080
(0.0078)(0.0060)(0.0059)(0.0074)
cons0.0516 *0.0792 ***0.1043 ***0.0769 ***
(0.0278)(0.0211)(0.0195)(0.0261)
Year fixedYesYesYesYes
City fixedYesYesYesYes
Obs385271712761826
R20.89600.68600.61740.6948
Note: figures in parentheses indicate a steady standard error; *, **, and *** are significant at 10%, 5%, and 1%, respectively.
Table 6. Results of the mechanism test.
Table 6. Results of the mechanism test.
Variable(1)(2)(3)(4)(5)(6)(7)
GDDEDGDRGICGDDEAGD
C i · D t −0.0189 ***0.0057 **−0.0180 ***0.3437 ***−0.0177 ***0.1128 ***−0.0136 **
(0.0059)(0.0025)(0.0059)(0.1294)(0.0059)(0.0137)(0.0059)
DED −0.1649 ***
(0.0436)
RGIC −0.0036 ***
(0.0009)
DEA −0.0476 ***
(0.0081)
GESE−0.9020 ***0.0398−0.8954 ***8.4198−0.8716 ***3.0257 ***−0.7578 ***
(0.2329)(0.1008)(0.2323)(5.1318)(0.2323)(0.5427)(0.2328)
INFO0.2466 ***−0.00020.2465 ***0.05950.2467 ***−0.03840.2447 ***
(0.0722)(0.0312)(0.0720)(1.5904)(0.0720)(0.1682)(0.0717)
MA0.0021−0.00030.0021−0.05460.0019−0.00690.0018
(0.0019)(0.0019)(0.0019)(0.0410)(0.0019)(0.0043)(0.0019)
FINA−0.00120.0016−0.0010−0.0624−0.0015−0.0035−0.0014
(0.0029)(0.0013)(0.0029)(0.0640)(0.0029)(0.0068)(0.0029)
IS0.00290.0089 ***0.0044−0.5558 ***0.0009−0.0692 ***−0.0004
(0.0053)(0.0023)(0.0053)(0.1169)(0.0053)(0.0124)(0.0053)
Cons0.0784 ***0.1386 ***0.1013 ***−0.24000.0776 ***−0.06470.0754 ***
(0.0201)(0.0087)(0.0210)(0.4435)(0.0201)(0.0469)(0.0200)
Year fixedYesYesYesYesYesYesYes
City fixedYesYesYesYesYesYesYes
Obs3102310231023102310231023102
R20.69890.91870.70050.69690.70080.95950.7026
Note: figures in parentheses indicate a steady standard error; ** and *** are significant at 5% and 1%, respectively.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Lin, S.; Lin, Q.; Wang, Q.; Shi, C.; Ausloos, M. Driving Green Transformation Through the National Digital Economy Innovation Pilot: A Quasi-Experimental Study on Reducing Urban Energy Intensity in 282 Chinese Cities. Sustainability 2025, 17, 5687. https://doi.org/10.3390/su17135687

AMA Style

Lin S, Lin Q, Wang Q, Shi C, Ausloos M. Driving Green Transformation Through the National Digital Economy Innovation Pilot: A Quasi-Experimental Study on Reducing Urban Energy Intensity in 282 Chinese Cities. Sustainability. 2025; 17(13):5687. https://doi.org/10.3390/su17135687

Chicago/Turabian Style

Lin, Shoufu, Quan Lin, Qian Wang, Chenyong Shi, and Marcel Ausloos. 2025. "Driving Green Transformation Through the National Digital Economy Innovation Pilot: A Quasi-Experimental Study on Reducing Urban Energy Intensity in 282 Chinese Cities" Sustainability 17, no. 13: 5687. https://doi.org/10.3390/su17135687

APA Style

Lin, S., Lin, Q., Wang, Q., Shi, C., & Ausloos, M. (2025). Driving Green Transformation Through the National Digital Economy Innovation Pilot: A Quasi-Experimental Study on Reducing Urban Energy Intensity in 282 Chinese Cities. Sustainability, 17(13), 5687. https://doi.org/10.3390/su17135687

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop