Enhancing the Resilience of Resource-Based Cities: A Dual Analysis of the Driving Mechanisms and Spatial Effects of the Digital Economy

Abstract

The sustainable transformation of resource-based cities (RBCs) is a critical global challenge. The digital economy is emerging as a potential catalyst for this transition, but the precise mechanisms and spatial dynamics underlying its influence on urban resilience remain underexplored. This study addresses this gap by investigating how the digital economy impacts RBC’s resilience, with a focus on both internal mechanisms and cross-regional spatial effects. Fixed effects, mediation, exploratory spatial data analysis (ESDA) and spatial Durbin model (SDM) are used to examine the complex relationships observed in this context. The results reveal the following: (1) While the digital economy in Chinese RBCs demonstrated a stable upward trajectory, urban resilience, although it exhibited a general increase, remained fragile. (2) The digital economy significantly enhanced urban resilience (coefficient = 0.117, p < 0.05), in which context the most pronounced effects pertained to the social and economic resilience subsystems. (3) Green technological innovation (GTI) served as the core intermediary pathway (a × b = 0.017, p < 0.01). Industrial structure rationalization also served as a mediator. (4) The digital economy and urban resilience exhibited positive spatial autocorrelation (significant direct effects 0.032, p < 0.05), and the advancement of the digital economy in a focal city can enhance both the urban resilience of that city itself and that of neighboring cities indirectly.

1. Introduction

Against the backdrop of accelerating global urbanization, cities must increasingly face frequent and complex shocks such as natural disasters, climate change, and public health crises, which makes the enhancement of urban resilience a critical issue for global sustainable development [1]. In 2015, the United Nations identified “inclusive, safe, resilient, and sustainable cities” as a key agenda for the 2030 Sustainable Development Goals, while the “Making Cities Resilient 2030” initiative launched by the UN Office for Disaster Risk Reduction called on nations to strengthen their urban disaster prevention and mitigation capacities and systematically enhance their overall resilience to multiple risks by reinforcing the adaptive and recovery capacities of their infrastructures, economic systems, and social systems in the face of shocks. The European Green Deal aims to establish Europe as the globe’s initial carbon-neutral continent by 2050, focusing on separating economic advancement from resource utilization. Currently, there is an international consensus regarding the building of resilient cities, which has consequently spurred multilevel exploration, which ranges from policy to practice, across various sectors. To advance global sustainable development goals, this study investigates how the digital economy enhances resilience in resource-based cities (RBCs).

As a result of the continuous increase in global resource demand, RBCs have long relied on extensive and high-intensity extraction models, which has resulted in severe ecological environmental damage. Constrained by the nonrenewable nature of mineral resources, the contradiction between resource depletion and economic development has become increasingly prominent, slowing economic growth, decreasing demographic dividends, and even triggering phenomena such as the “resource curse” and “Dutch disease”. According to traditional “resource curse” theory, dependence on natural resources impedes long-term regional development through the crowding out of human capital and suppression of technological innovation. In the context of the digital economy, its penetrative and widely applicable nature suggests a potential pathway to overcome this developmental constraint [2]. The concept of resilience enables cities to cultivate capacities for shock resistance, adaptive transformation, and self-renewal—akin to ecological systems—promoting a virtuous cycle of economic diversification, ecological rehabilitation, and social coordination. By strengthening multidimensional adaptive and recovery capacities across economic, ecological, governance, and social spheres, RBCs bolster their role in advancing high-quality regional development and consolidate the foundation for human prosperity and sustainable progress [3].

In recent years, the digital economy, which has data as its core element and digital technology as its driving force, has demonstrated the potential to promote industrial transformation, optimize resource allocation, and enhance innovation capabilities, therefore gradually becoming a new focus in regional resilience research. Extensive research has indicated that the development of the digital economy has significantly altered human lifestyles and production modes: digital technology has become deeply integrated into daily life, with digital consumption becoming the dominant pattern of resident consumption and coming to profoundly reshape social behaviors and interaction modes [4]. Moreover, enterprises can intellectualize their production processes and enhance the flexibility of their organizational structures through the leveraging of digital technologies, thereby driving fundamental changes in production efficiency. By utilizing data as a key factor, the digital economy reshapes production, distribution, and circulation processes, thereby facilitating industrial upgrading and enhancing economic efficiency [5]. It reduces social disparities through platform sharing and intelligent governance while significantly reducing resource consumption and environmental pressure through precise energy savings, virtual substitution, and circular matching, thereby providing sustained momentum for driving urban green transformation [6].

This study systematically investigates how the digital economy enhances resilience in RBCs against systemic challenges, with a focus on its role in promoting economic diversification, ecological rehabilitation, and social governance. By integrating spatial analysis with mechanism exploration, it offers new insights into the multidimensional effects of digital transformation on urban adaptive and recovery capacities.

This study makes three principal contributions. It advances a comprehensive analytical framework integrating resilience theory, environmental economics, and urban development through examining the digital economy’s impact on RBCs. The spatial analysis further reveals significant cross-regional spillover effects of digital economic activities, elucidating synergistic mechanisms and providing theoretical foundations for regionally coordinated policy interventions. Methodologically, the research innovates by constructing an urban resilience assessment system incorporating resource resilience, thereby aligning the evaluation framework with the inherent characteristics of resource-based cities.

The subsequent sections are as follows: Section 2 provides a systematic review and analysis of the literature. Section 3 elaborates on the theoretical analysis and research hypotheses. Section 4 presents an analysis of the empirical results. Section 5 presents a discussion of the research findings. Section 6 offers conclusions and policy recommendations.

2. Literature Review

2.1. Research on the Digital Economy

The concept of the digital economy was first introduced by Tapscott (1996) and later systematically defined by the U.S. Department of Commerce (1998–1999), who clarified that its core components include digital infrastructure, e-commerce, and digital media [7]. Further research has led to the current mainstream definition, which describes the digital economy as a new economic form utilizing data as a key factor, relying on networks as its carrier, being driven by the integrated application of information technology and the digital transformation of all factors, and driving the unity of equity and efficiency [8].

At present, the use of measurement indicators for the digital economy is becoming mainstream, but a unified standard has not yet been established. Jing (2023) [9] measured from the dimensions of informatization, the internet, and digital transactions, whereas Zhang (2021) [10] focused on infrastructure, industrial strength, and integration effects. Wang (2021) [11] introduced a multidimensional indicator system that includes development carriers, industrialization and digitalization levels, and the external environment. Furthermore, the industry definition and aggregation method adopted by Dong (2022) [12], while intuitively reflecting scale, faces challenges such as blurred industry boundaries and the difficulty of fully accounting for integrated value.

As the conceptual framework has become clearer, a consensus has gradually emerged on the measurement across dimensions such as infrastructure, industrial forms, integrated applications, and the institutional environment, which often relies on comprehensive indicator systems [13]. Different studies emphasize different weighting methods, disparate levels of regional comparability, and distinct mechanism analyses. Research has been expanded to encompass various fields, including economic growth, technological innovation, environmental pollution, employment structure, industrial upgrades, and urban development, profoundly reshaping traditional industrial models through three major technological pathways: the Internet of Things enables comprehensive monitoring and precise traceability of resource flows throughout the entire chain; big data optimizes the efficiency of material cycles and energy flows and blockchain can construct a transparent and trustworthy green supply chain system [14]. This deep integration of digitalization and the circular economy provides a systematic solution for the green transformation of RBCs, enhancing resource utilization efficiency through smart technologies while reconstructing the industrial ecosystem with circular principles, achieving a fundamental shift from linear consumption to closed-loop regeneration.

Existing research still lacks an integrated multidimensional perspective [15]. While scholarly attention has extensively addressed the digital economy’s influence on urban development, its specific role in resource-based cities remains underexplored. Systematic analysis of how the digital economy shapes RBC development through the lens of urban resilience addresses this scholarly gap and offers theoretical and policy foundations for advancing regional sustainability [16].

2.2. Research on RBCs Resilience

The term “resilience” originates from the Latin “resilio” which refers to a system’s ability to absorb shocks [17]. This notion was later extended from the field of ecology to fields such as sociology and economics. After the formal proposal of “urban resilience” in 2002, three major perspectives emerged, namely, engineering resilience, ecological resilience, and evolutionary resilience. This study adopts an evolutionary resilience perspective. Building upon the theoretical progression from engineering to ecological resilience, it focuses not only on how cities recover from resource decline but also on how they leverage the digital economy to achieve adaptive transformation [18]. Urban resilience refers to a city’s ability to maintain its functions, respond to and recover from disturbances, and achieve stability in the face of internal and external pressures such as natural disasters and economic crises [19].

For RBC’s, resilience indicates regional capacity to adapt to disruptions and undergo change despite heavy resource reliance [20]. Research on RBCs commenced gradually in the early 20th century, in studies pertaining to the transformation of RBCs, Chinese cases are numerous and representative. The principal factor behind this situation stems from China’s substantial red blood cell count and the intricate interplay between governmental, economic, social, and environmental factors. During the late 1980s, as economic globalization gained momentum and financial crises began to emerge, the landscape underwent significant transformation, the development of RBCs began to face crises, making their transformation a research hotspot. To address the crisis issues of RBCs, scholars now analyze the impact of the digital economy by deconstructing it into urban resilience subsystems such as economy, society, and ecology. The digital economy not only enhances the static resilience of cities but also improves their dynamic adaptation and transformation capabilities. Current research on urban resilience primarily focuses on three mainstream theoretical frameworks: disaster risk, urban governance, and complex adaptive systems [21].

Studies such as Sun (2022) [22] have been conducted for assessment. Urban resilience can be divided into three stages, which provides a clear structural framework for evaluating resilience across these three dimensions. Mallick (2021) [23] constructed a comprehensive evaluation indicator system based on core urban characteristics such as physical integrity, redundancy, and rapid response capability, which indicates a trend toward a diversified understanding of the connotation of resilience. Jiang (2022) [24] proposed an evaluation framework built on four dimensions of urban resilience—which emphasizes the multidimensional nature of the urban recovery process. Despite this shared theoretical foundation, standardized methods for the measurement of resilience have not yet been established.

Existing studies have built a theoretical foundation using dimensions such as adaptive governance, climate change response, and sustainable resource management, often focusing on the impact mechanisms of industrial, economic and social factors on urban development [25]. However, for RBCs, the current research is overly focused on sustainability under resource dependence, and insufficient attention has been given to the transformation and innovation processes that are driven by resource depletion and environmental pressure. Systematic research that deeply analyzes the effects on resilient development from a spatial perspective is lacking.

2.3. Research on the Digital Economy and Urban Resilience

Research indicates that data, as key production factors, drive technological, modal, and institutional innovation through a multiplier effect, providing a core impetus for the strengthening of urban resilience [26]. The digital economy restructures the modes of living and governance. It not only drives economic growth and technological upgrading, thereby enhancing a city’s ability to respond to various risks, but it also contributes to the promotion of social equity and ecological environmental protection.

Academic research on the spatial effects of the digital economy remains relatively scarce, with most discussions limited solely to the perspective of spatial spillover effects. Scholars have interpreted these spatial effects by selecting multiple dimensions and employing diverse methodologies. First, some equate spatial effects with spatial spillover effects, decomposing them into direct, indirect, and total effects. A common approach involves utilizing the spatial econometric SDM for research and decomposing the resulting spatial effects in a partial differential form. Second, others explain spatial effects from the perspectives of spatial structure and economic growth. By integrating theoretical and empirical research, they posit that spatial economic networks can drive the evolution of spatial structure, leading to the agglomeration or diffusion of factors, and subsequently promoting regional economic growth. Third, some interpret spatial effects through the lenses of both spatial agglomeration effects and spatial spillover effects [27].

RBCs commonly face challenges such as sluggish industrial development, low resource utilization efficiency, and weak green transformation. Li (2021) [28] reported that the digital economy can optimize resource allocation on the basis of information and knowledge and enhance the convenience of obtaining ecological data in real time, thereby effectively improving ecological management levels and reducing the production and diffusion of haze pollution. Wang (2022) [29] emphasized that digital technologies accelerate knowledge flow and innovation efficiency in RBCs. Pan (2022) [30] argued that the digital economy facilitates development mode transformation through three key mechanisms: accelerating the accumulation of societal knowledge capital, transforming conventional production models, and strengthening ecological resilience.

Existing studies have preliminarily revealed a positive correlation between the digital economy and urban resilience, particularly evident in the economic and social dimensions [31]. Emerging empirical research has uncovered the complex role of the digital economy in the transformation of RBCs. On one hand, the digital economy has been confirmed to significantly enhance the ecological efficiency of RBCs by promoting industrial structure upgrading and GTI. On the other hand, its impact exhibits significant heterogeneity. Some studies point out that the driving effect of the digital economy on GTI may be undermined by the industrial lock-in effect, suggesting a need for cautious analysis of its operational mechanisms [30].

Previous studies have preliminarily revealed a positive correlation between the digital economy and urban resilience, which is particularly evident in the economic and social resilience dimensions [31]. The proliferation of digital infrastructure helps cities maintain their basic functions and enhances their system stability against shocks such as pandemics and extreme weather. However, the intrinsic mechanisms through which the digital economy affects urban resilience have not yet been fully analyzed. The resilience-moderating effects of industrial structure adjustments have also not been fully explored. Most studies are constrained to assessing the city’s own dimensions and lack a systematic examination of spatial spillover effects [32]. Future research needs to focus on the cross-regional characteristics of the digital economy and conduct in-depth analyses of its spatial impact mechanisms.

2.4. Research Gaps and Contributions

With the development of digital economy, act of effectively leveraging it to promote urban resilience construction has become crucial for narrowing the global wealth gap and achieving the SDGs. The existing research is still limited as follows: First, prevailing urban resilience assessment systems are not adequately aligned with the developmental characteristics of RBCs, which indicates a significant lack of targeted applicability and applicability. Second, research on two-way interaction dynamics and the associated spatiotemporal heterogeneity remains insufficient. Third, studies on the impact of the digital economy on urban resilience are often conducted at the aggregate effect level and fail to fully elucidate the relevant internal pathways and mechanisms; in particular, the literature lacks heterogeneity tests across different resilience dimensions. Furthermore, research on the enhancement of resilience through key mediating or moderating variables is scarce. Existing studies predominantly analyze impacts from a single-city perspective, overlooking potential spatial spillover effects. Given the inherently boundary-spanning nature of digital economy development, a comprehensive understanding of its impact on RBCs necessitates further examination of spatial interactions within spatiotemporal frameworks.

This study introduces a novel inclusion of resource resilience into the urban resilience framework. Focusing on 115 Chinese RBCs, it employs the entropy method to evaluate the digital economy across infrastructure, industrial development, and technological application dimensions. The research further examines the underlying mechanisms and potential mediating effects through which digital economy development influences urban resilience. Second, to clarify the impact of the intracity digital economy on resilience, spatial exploratory analysis and spatial effects models are used to analyze the spatial spillover effects of digital economy development on the resilience of surrounding areas, thereby supplementing the existing research and indicating reference pathways for the development of resilient cities.

Accordingly, this study aimed at investigating the impact of the digital economy on the resilience of RBCs, and the mechanistic roles played by green technology and industrial restructuring in this process. The spatial effects are also decomposed. The study is meant to contribute new perspectives and strategies for the sustainable development of RBCs from both the internal urban and spatial mechanism viewpoints, enrich the research findings on their resilience, provide guidance and references for the enhancement of their risk resistance capabilities, and ultimately promote global sustainable development.

3. Study Design

3.1. Research Hypotheses

3.1.1. Impact of the Digital Economy on RBC’s Resilience

Extensive research has demonstrated that the digital economy has a positive effect on urban resilience [33,34]. For those RBCs that are characterized by singular industrial structures and high resource dependency, a thorough validation of the digital economy’s effectiveness in enhancing urban resilience is particularly crucial. Economically, it integrates advanced digital technologies with traditional industries, optimizing information flows along the industrial value chain to reduce information asymmetry and transaction costs. This allows firms to lower production costs while achieving higher labor productivity, fostering a more efficient and flexible economic system capable of rapid restoration of production following external shocks. Socially, the digital economy enhances operational stability by fostering new industries that diversify employment patterns and shorten job-matching times. Simultaneously, it ensures the continuity of essential public services, thereby maintaining critical social functions and reducing systemic vulnerability during emergencies. In terms of infrastructure, the proliferation of smart systems, including intelligent power grids and digital emergency management platforms, enhances the dynamic scheduling and responsiveness of urban assets, enabling the rapid recovery of critical functions during disruptions. Ecologically, it mitigates environmental pressures by enabling both cleaner production through intelligent resource planning and more effective end-of-pipe governance, promoting low-carbon, energy-saving operations while enhancing the ecosystem’s self-regulatory capacity. Finally, it bolsters resource resilience by applying precise analysis and intelligent control throughout the entire resource lifecycle—from extraction to recycling. This comprehensive optimization not only improves utilization rates but also helps establish robust emergency allocation systems, securing long-term supply stability against disruptions.

Thus, this leads to the following hypotheses:

H1. 

The digital economy has a positive effect on RBC’s resilience.

H1a. 

The digital economy has a positive effect on economic resilience.

H1b. 

The digital economy has a positive effect on social resilience.

H1c. 

The digital economy has a positive effect on ecological resilience.

H1d. 

The digital economy has a positive effect on infrastructure resilience.

H1e. 

The digital economy has a positive effect on resource resilience.

3.1.2. Transmission Mechanisms

The digital economy strengthens RBC resilience through two principal channels. Technologically, it reduces innovation costs and enables cross-regional research collaboration, establishing a foundation for green technology investments. This dual effect boosts resource efficiency while curbing energy consumption and pollution, thereby enhancing ecological, economic and resource resilience. Structurally, digital technologies upgrade traditional resource industries through technological penetration while fostering emerging sectors like data services and platform economy. This industrial transformation toward tertiary sectors optimizes production layouts, generates employment opportunities, and ultimately improves economic shock resistance.

The digital economy further contributes to industrial structure rationalization by enhancing factor allocation efficiency [35]. Through digital platforms, information barriers are effectively dismantled, accelerating the cross-sectoral flow of production factors including labor, capital, and technology. This process mitigates structural unemployment and resource misallocation stemming from the decline of traditional industries, thereby strengthening the economic system’s overall stability and recovery capacity.

Thus, this leads to the following hypotheses:

H2a. 

GTI plays a mediating role in the relationship between the digital economy and RBC’s resilience.

H2b. 

Industrial structure advancement plays a mediating role in the relationship between the digital economy and RBC’s resilience.

H2c. 

Industrial structure rationalization plays a mediating role in the relationship between the digital economy and RBC’s resilience.

3.1.3. Spatial Impact

According to spatial economics theory, regional economic activities exhibit spatial spillover effects, meaning that development in one location can influence neighboring areas through various channels [36]. Through the leveraging of its networks and cross-regional connectivity, the digital economy strengthens the spatial interdependence of the resilience among RBCs. Those cities with higher levels of digital economic development have advantages in terms of knowledge, technology, and talent aggregation. Through technological collaboration, talent mobility, and experience diffusion, such economic development lowers the innovation barriers and transformation costs for neighboring RBCs, accelerating their digital upgrading. Simultaneously, the digital economy reshapes the regional industrial and supply chains. The digital platform economy and smart manufacturing industries in core cities drive neighboring cities toward integration into broader economic networks by supporting collaboration and promoting industrial structure diversification. Furthermore, cross-regional data flow, supported by an interconnected digital infrastructure, enables data and digital service sharing, optimizes resource allocation efficiency, improves public service levels, and enhances market responsiveness, thereby strengthening the social and infrastructure resilience of neighboring cities. Thus, this leads to the following hypothesis:

H3. 

The impact of the digital economy on RBC’s resilience has spatial spillover effects.

In summary, the theoretical analysis mechanism of this study is shown in Figure 1.

3.2. Data Sources

To ensure the reliability and completeness of the empirical analysis, we initially collected data for all 126 resource-based prefecture-level cities in China. Cities with a proportion of missing values exceeding 20% in the core indicator set over the study period were excluded from the sample. For instance, Hami, Turpan and Urumqi each exhibited missing data for key variables prior to 2017, with missing rates surpassing the threshold; such omissions could significantly affect estimation robustness.

After this screening process, our final sample comprises 115 resource-based prefecture-level cities (the distribution of the study area is shown in Figure 2), covering the period from 2010 to 2023. This time span was selected because (i) 2010 marks the year when national-level statistical reporting on several key variables of interest, including digital economy indicators, became consistent and comparable across cities, and (ii) data availability after 2023 is incomplete.

Missing data for remaining observations were supplemented using linear interpolation to preserve temporal continuity, and all variables were winsorized at the 1st and 99th percentiles to reduce the influence of extreme outliers.

The dataset for this study was compiled from multiple authoritative sources. Data concerning the digital economy and its associated metrics were sourced from the China Statistical Yearbook, China High-Tech Industry Statistical Yearbook, and the CEIC database. GTI and environmental indicators: China Environmental Statistical Yearbook. RBC’s resilience indicators, mediating variables, and control variables: relevant statistical yearbooks published by the National Bureau of Statistics of China. Digital inclusive finance index: obtained from the Peking University Digital Finance Research Center. This multi-source, systematically screened dataset ensures both robustness and comparability for the spatial econometric analysis conducted in this study.

3.3. Variable Design

3.3.1. Explained Variable

Building on existing research, this study incorporates “resource resilience” into the evaluation system for RBC’s Resilience. The framework covers five dimensions—economic, social, ecological, infrastructure, and resource—which fully align with the developmental characteristics of RBCs and comprehensively reflect their resistance, recovery, and adaptation capacities under external shocks. The level of urban resilience is measured using the entropy method, with the specific indicators selected for each dimension detailed in Table 1. In this study, the explained variable is the overall urban resilience of RBC’s (y), while the resilience scores of each dimension (y1–y5) are used as secondary explained variables to facilitate an analysis of the specific impact mechanisms of the digital economy on urban resilience.

3.3.2. Explanatory Variable

The digital economy is the key explanatory variable in our analysis. Operationalized through a three-dimensional indicator system derived from prior literature, it assesses digital infrastructure, industrial development, and technological advancement, thereby addressing the prevailing lack of a unified measurement standard [37,38]. Construct a comprehensive index using the entropy weight method. The specific indicators employed are detailed in Table 1.

The explained and explanatory variables are both measured using the entropy method, and the detailed computational procedures are presented in Appendix A, Equations (A1)–(A6).

Table 1. Evaluation index system for RBC’s resilience and digital economy.

Target	Subsystem	Indicator Measurement	Attributes	Abbreviation
Resilience of resource-based cities	Economic resilience	Per capita GDP	+	ECO
		Per capita disposable income of urban residents	+	LIV
		Trade dependence ratio	+	TRA
	Social resilience	Urbanization rate	+	UR
		Basic medical insurance coverage rate	+	INS
		Hospital beds per 10,000 population	+	HEA
		Population density	−	POP
	Ecological resilience	Per capita industrial wastewater discharge	−	WP
		Per capita industrial sulfur dioxide emissions	−	AP
		Built-up area green coverage rate	+	GC
		Harmless treatment rate of domestic waste	+	PTC
	Infrastructure resilience	Graded highway mileage per 10,000 population	+	HWM
		Proportion of highway passenger volume to total population	+	HWP
		Drainage pipeline density in built-up areas	+	DRA
	Resource resilience	Energy productivity	+	ER
		Mining industry employment/Total employment	−	RD
		Number of resource-based enterprises/Number of domestic and foreign enterprises	−	RIS
Digital economy	Digital infrastructure	Number of 4G and 5G base stations	+	STA
		Number of International Internet Users	+	NET
		Long-distance Optical Cable Density	+	CAB
		Number of mobile telephone subscriptions	+	TEL
	Digital industry development	E-commerce sales	+	ES
		Employees in information transmission, computer services and software industry	+	EIT
		Digital financial inclusion index	+	DFI
		Number of national high-tech enterprises	+	HTE
	Digital technology development	Internal R&D expenditure	+	IRE
		Number of digital patents	+	DP
		Technology contract transaction value	+	TCT

Table 1. Evaluation index system for RBC’s resilience and digital economy.

Target	Subsystem	Indicator Measurement	Attributes	Abbreviation
Resilience of resource-based cities	Economic resilience	Per capita GDP	+	ECO
		Per capita disposable income of urban residents	+	LIV
		Trade dependence ratio	+	TRA
	Social resilience	Urbanization rate	+	UR
		Basic medical insurance coverage rate	+	INS
		Hospital beds per 10,000 population	+	HEA
		Population density	−	POP
	Ecological resilience	Per capita industrial wastewater discharge	−	WP
		Per capita industrial sulfur dioxide emissions	−	AP
		Built-up area green coverage rate	+	GC
		Harmless treatment rate of domestic waste	+	PTC
	Infrastructure resilience	Graded highway mileage per 10,000 population	+	HWM
		Proportion of highway passenger volume to total population	+	HWP
		Drainage pipeline density in built-up areas	+	DRA
	Resource resilience	Energy productivity	+	ER
		Mining industry employment/Total employment	−	RD
		Number of resource-based enterprises/Number of domestic and foreign enterprises	−	RIS
Digital economy	Digital infrastructure	Number of 4G and 5G base stations	+	STA
		Number of International Internet Users	+	NET
		Long-distance Optical Cable Density	+	CAB
		Number of mobile telephone subscriptions	+	TEL
	Digital industry development	E-commerce sales	+	ES
		Employees in information transmission, computer services and software industry	+	EIT
		Digital financial inclusion index	+	DFI
		Number of national high-tech enterprises	+	HTE
	Digital technology development	Internal R&D expenditure	+	IRE
		Number of digital patents	+	DP
		Technology contract transaction value	+	TCT

3.3.3. Mediating Variables

GTI: Applying standardization processing, this study uses the number of green patent applications as a measurement.

Industrial structure advancement reflects the dynamic process through which the industrial structure evolves toward high-value-added, knowledge-intensive sectors. The development of the digital economy accelerates the “servitization” and “dematerialization” processes of the economy. Therefore, this study employs the ratio of the added value of the tertiary industry to that of the secondary industry (TS) as a measurement indicator [39].

Industrial structure rationalization is aimed at measuring the level of synergy among industries and the efficiency of resource allocation. Following existing research, this study applies the reciprocal of the Theil index (TL), which is denoted as ISR = 1/TL, to measure the level of industrial structure rationalization [40].

3.3.4. Control Variables

To analyze the relationship between the digital economy and RBC’s resilience and avoid bias in the regression results resulting from omitted variables, the following control variables are utilized: (1) Government intervention level. Government intervention impacts urban development. The ratio of the local government’s general public budget expenditure to regional GDP is used as a measure. (2) Human capital level. Talent resources serve as a crucial source of urban resilience development. This study employs the ratio of the number of higher education students to the total population as a measure. (3) Foreign investment level. FDI reflects a city’s ability to absorb advanced international experience and technology, which influences its urban economic development [41]. In this study, the ratio of the actually utilized foreign capital to GDP is used as a measure. (4) City size. City size indirectly reflects the level of urban development, where better urban development is associated with larger city size [42]. In this study, city size is measured via the logarithm of the city’s permanent population. (5) Environmental protection level. The level of environmental protection affects the quality of the urban ecology. The ratio of environmental protection fiscal expenditure to GDP is used as a measure. (6) Industrial agglomeration level. Industrial agglomeration reflects the differences in resource endowments across regions as well as the influence of industrial policies and government actions. In this study, the ratio of the number of employees to the administrative area is used as a measure.

The above variables are summarized in

Table 2. Summary of variables and their measurement scales.

Variable	Variable Name	Variable Symbol	Variable Measurement
Explanatory Variable	Digital Economy	DE	Entropy Method
Dependent Variable	Resource-based City Resilience	R-RBC’s	Entropy Method
Mediating Variables	Green Technological Innovation	GTI	Number of Green Patent Applications Received
	Industrial Structure Sophistication	TS	Ratio of Tertiary Industry Value-added to Secondary Industry Value-added
	Industrial Structure Rationalization	TL	Ln (Theil Index)
Control Variables	Government Intervention	GOV	Government General Public Budget Expenditure/GDP
	Human Capital Level	HUM	Higher Education Enrolled Students/Total Population
	Foreign Investment Level	FDI	Actual Foreign Investment Used/GDP
	City Size	US	Ln (Urban Permanent Resident Population)
	Environmental Protection Level	ENV	Environmental Protection Fiscal Expenditure/GDP
	Industrial Agglomeration Level	IGA	Number of Employed Persons/Administrative Area

.

3.4. Empirical Model

To examine the mechanism through which the digital economy affects RBC’s resilience, three categories of models are employed: the baseline regression model, the mediating effect model, and the spatial econometric model. The significance and specification tests for all models described in Section 3.4, including OLS estimation, LM tests, LR tests, and Wald tests, are fully reported in Appendix A (

Table A2. OLS Estimation and LM Test Results.

Variables	Regression Coefficients	p−Value
X	0.257 ***	0.000
GOV	−0.0028 ***	0.000
HUM	−0.0058	0.069
FDI	0.0039 ***	0.004
US	−0.062 ***	0.000
ENV	−0.0014	0.727
IGA	−0.00018 ***	0.002
cons	1.221 ***	0.000
R2	0.274
Testing Methods	Statistics	p-value
LM test no spatial error	483.171 ***	0.000
Robust LM test no spatial error	466.057 ***	0.000
LM test no spatial lag	29.288 ***	0.000
Robust LM test no spatial lag	12.174 ***	0.000

Note: *** represent significant at 1% levels.

and

Table A3. LR Test and Wald Test Results.

Testing Methods	Chi2	p-Value
LR Lag	857.76 ***	0.0000
LR Err	33.70 ***	0.0000
Wald Lag	22.04 ***	0.0025
Wald Err	22.31 ***	0.0022
LR Ind	1573.79 ***	0.0000
LR Time	1783.37 ***	0.0000

Note: *** represent significant at 1% levels.

) to ensure transparency and statistical validity.

3.4.1. Baseline Regression Model

To examine the mechanism through which the digital economy affects RBC’s resilience, the following baseline regression model is constructed:

$${\mathrm{Y}}_{\mathrm{it}}={\mathsf{\alpha }}_0+{\mathsf{\alpha }}_1{\mathrm{X}}_{\mathrm{it}}+{\mathsf{\alpha }}_2{\mathrm{Controls}}_{\mathrm{it}}+{\mathsf{\mu }}_{\mathrm{i}}{+\mathsf{\delta }}_{\mathrm{t}}+{\mathsf{\epsilon }}_{\mathrm{it}}$$

(1)

where $Y_{\mathit{it}}$ represents the explained variable. RBC’s resilience, ${\alpha }_0$ is used as the model constant, and $X_{\mathit{it}}$ is the explanatory variable. The digital economy level, ${\alpha }_1$ denotes the total effect of the digital economy on RBC’s resilience. ${\mathit{Controls}}_{\mathit{it}}$ are the control variables, ${\mu }_i$ indicates the year fixed effects, ${\delta }_t$ represents the individual fixed effects, and ${\epsilon }_{\mathit{it}}$ is the random error term.

3.4.2. Mediating Effect Model

To examine the effect of the digital economy on the mediating variables, the following formula is constructed:

$${\mathrm{M}}_{\mathrm{it}}={\mathsf{\beta }}_0+{\mathsf{\beta }}_1{\mathrm{X}}_{\mathrm{it}}+{\mathsf{\beta }}_2{\mathrm{Controls}}_{\mathrm{it}}+{\mathsf{\mu }}_{\mathrm{i}}{+\mathsf{\delta }}_{\mathrm{t}}+{\mathsf{\epsilon }}_{\mathrm{it}}$$

(2)

where ${\beta }_0$ is the model constant. $M_{\mathit{it}}$ represents the mediating variables—GTI, TS, or TL—and ${\beta }_1$ denotes the influence coefficient of the digital economy on the mediating variables.

To examine the impact of the mediating variables on RBC’s resilience, the following formula is constructed:

$${\mathrm{Y}}_{\mathrm{it}}={\mathsf{\gamma }}_0+{\mathsf{\gamma }}_1{\mathrm{M}}_{\mathrm{it}}+{\mathsf{\gamma }}_2{\mathrm{Controls}}_{\mathrm{it}}+{\mathsf{\mu }}_{\mathrm{i}}{+\mathsf{\delta }}_{\mathrm{t}}+{\mathsf{\epsilon }}_{\mathrm{it}}$$

(3)

where ${\gamma }_0$ is the model constant and ${\gamma }_1$ represents the effect of the mediating variables on RBC’s resilience.

To further examine the mediating roles of the mediating variables in the impact of the digital economy on RBC’s resilience. The following formula is constructed:

$${\mathrm{Y}}_{\mathrm{it}}={\mathsf{\lambda }}_0+{\mathsf{\lambda }}_1{\mathrm{X}}_{\mathrm{it}}+{\mathsf{\lambda }}_2{\mathrm{M}}_{\mathrm{it}}+{\mathsf{\lambda }}_3{\mathrm{Controls}}_{\mathrm{it}}+{\mathsf{\mu }}_{\mathrm{i}}{+\mathsf{\delta }}_{\mathrm{t}}+{\mathsf{\epsilon }}_{\mathrm{it}}$$

(4)

where ${\lambda }_0$ is the model constant, ${\lambda }_1$ represents the direct effect of the digital economy. ${\lambda }_2$ denotes the coefficient of the effect of the mediating variables on RBC’s resilience.

3.4.3. Spatial Econometric Model

To further explore the spatial relationship, this study employs spatial panel models to examine the spatial effects. The model is as follows:

$${\mathrm{Y}}_{\mathrm{it}}=\mathsf{\alpha }+\mathsf{\rho }{\sum }_{\mathrm{j}=1}^{\mathrm{N}}{\mathrm{W}}_{\mathrm{ij}}{\mathrm{Y}}_{\mathrm{it}}+\mathsf{\beta }{\mathrm{X}}_{\mathrm{it}}+\mathsf{\theta }{\sum }_{\mathrm{j}=1}^{\mathrm{N}}{\mathrm{W}}_{\mathrm{ij}}{\mathrm{X}}_{\mathrm{it}}+{\mathsf{\mu }}_{\mathrm{i}}+{\mathsf{\delta }}_{\mathrm{t}}+{\mathsf{\epsilon }}_{\mathrm{it}}$$

(5)

where $Y_{\mathit{it}}$ denotes the explained variable, $X_{\mathit{it}}$ represents the explanatory variable; $W_{\mathit{ij}}$ is an element of the spatial weight matrix W; t indicates the year; i and j refer to the i-th and j-th cities, respectively. $\rho$ is the spatial lag coefficient of the explained variable, which is used as a measure of the degree of spatial autocorrelation of RBC’s resilience. The regression coefficient β quantifies the extent to which the local digital economy directly influences and enhances the level of resilience in its respective urban area. $\theta$ is the regression coefficient of the spatially lagged explanatory variable, which reflects the spillover effect of neighboring regions’ digital economy level on local urban resilience. $\alpha$ is the constant term; ${\mu }_i$ and ${\delta }_t$ denote city-specific and time-period-specific fixed effects, respectively; and ${\epsilon }_{\mathrm{it}}$ is the random error term. If there is spatial autocorrelation in the error term, it can be specified as ${\epsilon }_{\mathit{it}}=\lambda {\sum }_{j=1}^N{W}_{\mathit{ij}}{\epsilon }_{\mathit{jt}}+u_{\mathit{it}}$, where $\lambda$ is the spatial autocorrelation coefficient of the error term [43].

4. Results

4.1. Measurements of the Digital Economy Development Level and Urban Resilience

From 2010 to 2022, the digital economy exhibited a continuous and robust growth trend (Figure 3a). Although there were variations in the development of the digital economy across cities, the overall growth trend remained consistent. An examination of the digital economies of the top 10 cities at various stages (Figure 3b–d) reveals that typical representatives, such as Xuzhou, Tangshan, and Zibo, are consistently concentrated, indicating that RBCs with higher levels of digital development are primarily clustered in eastern China.

In contrast, the level of urban resilience tended to increase overall, but the overall resilience index remained relatively low—more than 80% of the cities maintained a resilience index of between 0.15 and 0.40 over the long term (Figure 3e). However, certain cities performed notably well: cities such as Huzhou, Jinchang, and Dongying consistently ranked at the top, with their resilience indices remaining above 0.5 throughout the study period (Figure 3c,d), thus demonstrating strong risk resistances and recovery capabilities.

4.2. Descriptive Statistics

Table 3. Descriptive analysis.

Variable	Obs.	Mean	Std. Dev.	Min	Max
x	1610	0.109	0.125	0.011	0.644
y	1610	0.269	0.089	0.075	0.550
y1	1610	0.169	0.129	0.019	0.710
y2	1610	0.348	0.165	0.108	0.821
y3	1610	0.759	0.231	0.012	0.944
y4	1610	0.216	0.112	0.041	0.571
y5	1610	0.500	0.151	0.170	0.929
GOV	1610	22.269	10.378	7.700	60.029
HUM	1610	0.412	0.632	0.001	4.432
FDI	1610	1.310	1.537	0.003	7.213
US	1610	14.801	0.663	13.049	16.130
ENV	1610	1.166	0.485	0.245	2.388
IGA	1610	36.067	38.228	0.824	198.423
GTI	1610	154.717	210.187	3.000	1119.000
TS	1610	0.977	0.491	0.262	2.810
TL	1610	0.299	0.203	0.014	1.002

reports the results for the main variables. As shown in Appendix A 

Table A1. Correlation Analysis.

Variable	x	y	y1	y2	y3	y4	y5	GOV
x	1.000
y	0.193 ***	1.000
y1	0.313 ***	0.669 ***	1.000
y2	0.132 ***	0.589 ***	0.324 ***	1.000
y3	0.030	0.386 ***	−0.120 ***	−0.088 ***	1.000
y4	−0.246 ***	0.392 ***	0.028	0.011	0.151 ***	1.000
y5	0.261 ***	0.155 ***	0.163 ***	−0.162 ***	0.089 ***	−0.068 ***	1.000
GOV	−0.339 ***	−0.352 ***	−0.439 ***	−0.356 ***	0.018	0.119 ***	0.021	1.000
HUM	0.035	0.054	0.185 ***	−0.003	−0.088 ***	−0.035	0.020	−0.155 ***
FDI	0.224 ***	0.167 ***	0.188 ***	0.044 *	0.081 ***	−0.011	0.226 ***	−0.278 ***
US	0.562 ***	−0.205 ***	−0.149 ***	−0.223 ***	0.138 ***	−0.306 ***	0.236 ***	−0.165 ***
ENV	0.152 ***	−0.027	0.084 ***	−0.011	−0.077 ***	−0.062 **	−0.025	0.084 ***
IGA	0.415 ***	0.126 ***	0.295 ***	0.076 ***	−0.023	−0.229 ***	0.245 ***	−0.386 ***
GTI	0.724 ***	0.279 ***	0.409 ***	0.159 ***	0.037	−0.227 ***	0.285 ***	−0.329 ***
TS	−0.009	−0.244 ***	−0.170 ***	−0.221 ***	−0.084 ***	−0.044 *	0.123 ***	0.556
TL	0.724 ***	−0.064 ***	−0.082 ***	−0.016	−0.103	0.120 ***	−0.293 ***	0.108 ***
Variable	HUM	FDI	US	ENV	IGA	GTI	TS	TL
x
y
y1
y2
y3
y4
y5
GOV
HUM	1.000
FDI	−0.023	1.000
US	−0.118 ***	0.103 ***	1.000
ENV	−0.045*	−0.100 ***	0.082 ***	1.000
IGA	0.255 ***	0.327 ***	0.176 ***	0.065 ***	1.000
GTI	0.065 ***	0.237 ***	0.404 ***	0.215 ***	0.545 ***	1.000
TS	−0.067 ***	−0.174 ***	0.009	0.265 ***	−0.104 ***	0.032	1.000
TL	−0.020	−0.288 ***	−0.216 ***	−0.075 ***	−0.365 ***	−0.275 ***	−0.239 ***	1.000

Note: *, **, and *** represent significance at the 10%, 5%, and 1% levels, respectively.

, the digital economy, GTI, industrial structure advancement and industrial structure rationalization are positively correlated with the RBC’s resilience, suggesting that improvements in digitalization and industrial upgrading are generally associated with higher urban resilience. Among these, the strongest correlations are observed between the digital economy and industrial structure advancement, implying that digital development may facilitate structural upgrading. The moderate correlation magnitudes also indicate that these factors may contribute to resilience through partially distinct pathways. In addition, the mean VIF value of 1.36 (<5) confirms the absence of multicollinearity, ensuring that these variables can be jointly analyzed without estimation bias. The descriptive statistics provide preliminary evidence for the hypothesized relationships and the potential mechanisms of digital-driven industrial transformation, which will be rigorously tested in subsequent empirical modeling.

4.3. Regression Analysis and Hypothesis Testing

4.3.1. Regression Analysis

Based on the Hausman test findings (Appendix A Table A2), a regression analysis examining the link between the digital economy and RBC’s resilience was performed using an individual fixed-effects model. The outcome displayed in column y of

Table 4. Baseline regression results.

Variable	y	y1	y2	y3	y4	y5
x	0.117 **	0.179 ***	0.392 ***	0.085 **	0.090 **	0.095 ***
	(2.38)	(4.9)	(−10.31)	(2.41)	(−2.25)	(−2.79)
GOV	−0.020	−0.001	−0.006 ***	0.143 ***	0.078	0.032
	(−0.47)	(−0.05)	(−16.14)	(3.49)	(−1.2)	(−0.71)
HUM	−0.147 ***	0.011	−0.029 ***	−0.052	−0.144 **	0.109
	(−3.49)	(0.22)	(−4.85)	(−1.24)	(−2.24)	(−1.12)
FDI	0.109 **	−0.019	−0.007 ***	0.113 **	0.063	−0.068
	(2.13)	(−0.53)	(−2.74)	(2.35)	(−0.88)	(−1.07)
US	1.090 ***	0.713 ***	−0.112 ***	1.388 ***	0.766 **	−0.624
	(2.83)	(4.2)	(−17.17)	(3.04)	(−2.36)	(−1.56)
ENV	−0.111 ***	0.062***	0.002	−0.186 ***	−0.098 **	0.075 ***
	(−4.44)	(3.37)	(−0.32)	(−8.11)	(−2.47)	(−2.68)
IGA	0.018	0.182 ***	−0.001 ***	−0.219 ***	−0.119 **	0.051
	(0.39)	(7.42)	(−2.73)	(−4.45)	(−2.23)	(−0.81)
Constant	0.000 ***	0.000 ***	2.127 ***	0.000 **	0.000	0.000
	(−3.1)	(−4.5)	(−22.17)	(−2.08)	(−1.28)	−0.87
Observations	1610	1610	1610	1610	1610	1610
Number of groups	115	115	115	115	115	115
R-squared	0.071	0.198	0.111	0.082	0.069	0.067
City FE	Yes	Yes	Yes	Yes	Yes	Yes
F	15.98	20.93	12.28	52.83	7.28	4.35

Note: **, and *** represent significance at the 5%, and 1% levels, respectively; t-values are displayed in brackets.

indicates a statistically significant positive influence of the digital economy on resilience (β = 0.117, p < 0.05), which validates Hypothesis H1. Furthermore, the regression outcomes for the five distinct dimensions of RBC’s resilience are separately shown in columns y1 through y5. The results displayed in columns y1 to y5 indicate that the digital economy can increase the levels of RBC’s resilience in economic, social, ecological, infrastructure, and resource aspects (β₁ = 0.179, p < 0.01; β₂ = 0.392, p < 0.01; β₃ = 0.085, p < 0.05; β₄ = 0.090, p < 0.05; β₅ = 0.095, p < 0.01), thereby verifying Hypotheses H1a to H1e. These results demonstrate that the digital economy provides a crucial development pathway for the transformational development and risk resistance capacity building of traditional resource-dependent cities through mechanisms such as industrial structure optimization, innovation-driven strategies, and the improvement of resource allocation efficiency.

4.3.2. Mediating Effects of GTI, TS, and TL

This study first conducted regression tests on the relationships between the digital economy, GTI, industrial structure advancement (TS), and industrial structure rationalization (TL). The results, presented in

Table 5. Regression results of the digital economy on GTI, TS, and TL.

Variable	GTI	TS	TL
x	943.824 ***	0.578 ***	−0.161 ***
	(26.61)	(5.55)	(−3.28)
Controls	Yes	Yes	Yes
Constant	−128.136	−0.016	0.917 ***
	(−1.43)	(−0.06)	(7.42)
Observations	1610	1610	1610
Number of groups	115	115	115
R-squared	0.609	0.384	0.202
City FE	Yes	Yes	Yes

Note: *** represent significance at the 1% levels; t-values are displayed in brackets.

, indicate that the digital economy has a significant positive effect on GTI (coefficient = 943.824, p < 0.01) and TS (coefficient = 0.578, p < 0.01), and a significant negative effect on TL (coefficient = −0.161, p < 0.01).

Building on these regression results, we applied the bootstrap method, following established literature, to examine multiple mediating effects. As shown in

Table 6. Bootstrap test results for the mediating effects.

	a	b	a × b (Boot SE)	a × b (z Value)	a × b (p Value)	a × b (95% Boot CI)
x→GTI→y	943.8237 ***	0.000 ***	0.017	7.79	0.000	0.098~0.163
x→TS→y	0.5767 ***	−0.0162 ***	0.003	−2.88	0.004	−0.016~−0.003
x→TL→y	−0.1616 ***	−0.0189	0.002	1.47	0.140	−0.001~0.007

Note: *** represent significance at the 1% levels.

, the mediating effect of GTI is positive and significant (a ×b = 0.017, p < 0.01), with the 95% confidence interval excluding zero, indicating that GTI functions as an important positive mediator. Thus, the digital economy significantly enhances RBC resilience by promoting GTI, supporting Hypothesis H2a.

For industrial structure advancement (TS), although the regression path coefficient was significant, its mediating effect was negative (a × b = −0.0162, p < 0.01, 95% Boot CI: [−0.016, −0.003]), suggesting a suppression effect in the transmission mechanism that partially offsets the positive impact of the digital economy. In other words, while TS plays a statistically significant mediating role, the direction of its effect contradicts the hypothesized positive mediation. Therefore, H2b is not supported in its hypothesized form, but rather manifests as a significant suppression-type mediation.

Regarding industrial structure rationalization (TL), the mediating effect was positive (a × b = 0.003), although relatively small and statistically insignificant at conventional levels, indicating a weak partial mediation pathway. Consequently, H2c is considered supported, albeit with limited mediating strength.

As summarized in

Table 7. Summary of the sizes of the mediating effects.

	Test Conclusion	c Total Effect	a × b Indirect Effect	c′ Direct Effect	Calculation Formula	Effect Proportion
x→GTI→y	Partial Mediation	0.257	0.130	0.127	a × b/c	50.58%
x→TS→y	Suppression Effect	0.257	−0.009	0.267	a × b/c	−3.64%
x→TL→y	Partial Mediation	0.257	0.003	0.254	a × b/c	1.18%

, GTI accounts for 50.58% of the total effect, TS contributes −3.64%, and TL contributes 1.18%. Overall, these findings demonstrate that the digital economy primarily enhances RBC resilience through the promotion of GTI; industrial structure advancement exerts a statistically significant but inhibitory influence, and industrial structure rationalization plays a minor, supportive role.

4.3.3. Spatial Effect Analysis

Prior to conducting the spatial econometric analysis, spatial autocorrelation tests were performed on RBC’s resilience and the digital economy. To comprehensively capture potential spatial dependence from different perspectives, we employed three types of spatial weight matrices: (1) adjacency matrix (w1), which reflects direct geographical proximity between cities; (2) economic distance matrix (w2), which measures spatial relationships based on the similarity in economic development levels, thereby capturing connections driven by economic conditions rather than purely geographic proximity; and (3) geographical distance matrix (w3), which accounts for the inverse great-circle distances between cities, thus incorporating broader spatial interaction effects beyond immediate adjacency. Using multiple matrices ensures the robustness of the spatial autocorrelation results by evaluating consistency across alternative spatial structures, as spatial spillover effects in economic and resilience-related variables may arise from both physical proximity and economic similarity. Based on our results, all three matrices exhibit positive and significant global Moran’s I indices, confirming the presence of spatial clustering. However, the adjacency matrix (w1) produced relatively higher and more stable Moran’s I values across the study period, suggesting that direct geographical proximity is the most influential spatial link in the observed resilience–digital economy relationship within the study area.

The results (Figure 4) show that the global Moran’s I indices for each of the years under each weight matrix were significantly greater than zero and passed the significance test at the 1% level, indicating a positive spatial autocorrelation for both X and Y. This positive correlation suggests that higher levels cities tend to cluster in regions with higher RBC’s resilience, and vice versa, which reflects a significant agglomeration effect in their spatial distribution. Furthermore, although the Moran’s I indices fluctuated to a certain extent across different weight matrices, the overall trend remained consistent, verifying a stable spatial correlation between the digital economy and the selected resilience indicators within the selected study area. By combining the results of the local correlation analysis (Figure 5), distinct high–high and low–low clustering effects can be observed, indicating that some regions exhibit both high digital economy levels and high RBC’s resilience, whereas others simultaneously exhibit low levels of both factors.

Based on the spatial autocorrelation results, LM, LR, and Wald tests were conducted to determine an appropriate form for the spatial econometric model. The test results (Appendix A Table A3) show that both the LR and Wald tests reject the null hypothesis that the SDM degenerates into a simpler model. Furthermore, the LR test statistics for two-way (time-individual) fixed effects were significant at the 1% level, indicating that a SDM with two-way fixed effects is the most effective specification. Consequently, a spatiotemporal doubly fixed SDM was ultimately utilized for empirical analysis. The regression results are presented in

Table 8. Results of the SDM.

	w1	w2	w3
Spa-rho	0.285 ***	−0.002	−1.094 ***
	(12.11)	(−0.50)	(−3.42)
x	0.032 **	0.061 ***	0.025
	(2.04)	(3.66)	(1.52)
Wx	0.104 ***	0.005	−1.396 ***
	(5.08)	(0.30)	(−3.44)
Direct Effect	0.046 ***	0.062 ***	0.036 ***
	(2.87)	(3.60)	(2.10)
Indirect Effect	0.133 ***	0.004	−0.707 ***
	(5.87)	(0.28)	(−3.06)
Total Effect	0.179 ***	0.066 ***	−0.671 ***
	(5.86)	(2.92)	(−2.87)
Controls	Yes	Yes	Yes
City FE	Yes	Yes	Yes
Year FE	Yes	Yes	Yes
Observations	1610	1610	1610
R-squared	0.233	0.176	0.196
Log-L	3285.613	3167.973	3232.342

Note: **, and *** represent significance at the 5%, and 1% levels, respectively; t-values are displayed in brackets.

.

In this study, the direct and spatial spillover effects were examined using an adjacency matrix (w1), an economic distance matrix (w2), and a geographical distance matrix (w3) (Table 8).

The SDM results demonstrated that the digital economy significantly enhances urban resilience in RBC’s. Under the adjacency matrix, the digital economy exhibits both significant direct effects (0.032, p < 0.05) and strong positive spillover effects through spatially lagged terms (Wx = 0.104, p < 0.001), resulting in a substantial total effect of 0.179 (p < 0.001), thereby supporting hypothesis H3. This indicates that digital economy development not only improves local urban resilience but also benefits neighboring regions through interregional industrial linkages and technology diffusion, reflecting pronounced spatial spillover effects. In contrast, alternative spatial weight matrices based on economic and geographical distances yield weaker or different spillover patterns, highlighting how various spatial association mechanisms moderate the impact pathways of the digital economy. Overall, coordinated digital economy development among adjacent regions emerges as a critical pathway for enhancing resource-based city resilience, emphasizing the importance of regional collaborative governance and digital infrastructure interconnectivity.

The decomposition of these spatial effects not only reveals the complexity of the spatial transmission mechanisms but also provides an empirical basis for the coordination of regional development policies, which underscores the need to fully consider the interregional interactions and competitive-cooperative mechanisms at play in the formulation of strategies.

4.4. Regional Heterogeneity Analysis

Owing to the numerous differences in area conditions, resource endowments, and policy orientations among RBCs, both the digital economy and urban resilience exhibit significant heterogeneous characteristics across regions. Therefore, the impact may also vary regionally, thus necessitating more granular research. In accordance with China’s regional classification standards, the RBCs are divided into three major regions: the Eastern, Central, and Western regions. The results of the heterogeneity analysis are shown in

Table 9. Regression results for regional heterogeneity.

Variable	(1) Eastern	(2) Central	(3) Western
x	0.210 ***	0.218 ***	0.264 ***
	(2.61)	(5.22)	(2.68)
Controls	Yes	Yes	Yes
Constant	1.221	0.829 ***	1.218 ***
	(0.86)	(4.21)	(9.44)
City FE	Yes	Yes	Yes
Year FE	Yes	Yes	Yes
Observations	1610	1610	1610
R-squared	0.912	0.310	0.605

Note: *** represent significance at the 1% levels; t-values are displayed in brackets.

.

The estimated coefficients for the eastern, central, and western regions are 0.210, 0.218 and 0.264, respectively. Confirming that the digital economy exerts a significantly positive effect on RBC’s resilience across all these areas. The analysis confirms that the positive role of the digital economy in strengthening RBC’s resilience remains robust, even when accounting for significant inter-regional differences in location, natural resources, and policy directions. Notably, there are certain differences in the effects across regions. The explanatory power is higher in the eastern region (R² = 0.912) and lower in the central region (R² = 0.310). This difference may reflect the more mature economic, technological, and institutional environment of the eastern region, which facilitates the realization of the digital economy’s effects, whereas the higher coefficient of the western region suggests that although infrastructure is relatively lacking, the digital economy still has significant potential for improving RBC’s resilience in this region.

4.5. Robustness Tests

To ensure the robustness of the regression results, the following three methods were employed for robustness testing: First, data from specific years were excluded. Considering the severity of the global COVID-19 pandemic, which may have affected the resilience building of resource-based regions and, consequently, the regression results, data from these two years were excluded and the regression analysis was rerun. Second, one-period lagged variables were used for regression analysis. To acknowledge the potential lagged effect of the digital economy’s enabling role on RBC’s resilience, a one-period lag of the independent variable was applied for regression. Third, the regression method was changed by employing ordinary least squares (OLS) for estimation. The results obtained using the above methods are shown in

Table 10. Results of the robustness tests.

Variable	(4) y	(5) y	(6) y
x	0.071 ***	0.092 ***	0.257 ***
	(3.72)	(3.23)	(12.58)
Controls	Yes	Yes	Yes
Constant	0.525 *	1.114 ***	1.221 ***
	(1.72)	(2.74)	(23.63)
Observations	1380	1150	1610
R-squared	0.864	0.747	0.277
Number of groups	115	115	-
City FE	Yes	Yes	No
Year FE	Yes	Yes	No

Note: *, and *** represent significance at the 10%, and 1% levels, respectively; t-values are displayed in brackets.

. The results indicate that the regression coefficients shown in columns (4) to (6) are 0.071, 0.092, and 0.257, respectively, and they are significant at the 1% level, demonstrating that the digital economy still positively promotes RBC’s resilience under these three testing methods. Overall, these findings are consistent with the previous hypothesis testing results. Therefore, the conclusions of this study are robust.

5. Discussion

5.1. Effectiveness of Our Framework and Implications for Others

This study investigates the specific impact mechanisms of the digital economy on urban resilience among Chinese RBCs by measuring the relationships and its mediating effects.

The results show that the digital economy can positively promote the enhancement of resilience among Chinese RBCs. Its positive effect on economic and social resilience is particularly strong, but it also positively influences ecological, infrastructure, and resource resilience. This observed effect holds particular relevance for RBCs, which frequently demonstrate structural rigidities and heightened susceptibility to external shocks due to their specialized industrial composition and entrenched path dependence [44]. The digital economy effectively counteracts these constraints by enhancing allocative efficiency, reducing information asymmetries, and fostering innovative business models-thereby circumventing traditional geographical and industrial limitations. These mechanisms provide substantive developmental impetus for cities confronting resource curse challenges, aligning with established scholarly findings [45]. During the advancement of digital transformation, vigilance is required against access-based and skills-based digital divides arising from uneven infrastructure coverage, insufficient digital skills among industrial workers, and significant energy consumption by data centers. Meanwhile, the concentration of substantial talent and capital in regional central cities also exposes RBCs to the risk of hollowing-out caused by brain drain and capital flight. Confronted with these structural challenges, it is imperative to address them through systematic governance. Only by achieving an effective balance between efficiency and equity, development and sustainability, and innovation and stability can RBCs fully unleash the potential of the digital economy while effectively mitigating its negative effects, ultimately realizing resilient and inclusive high-quality development.

With respect to the results of the mediating effect tests, GTI was shown to play a dominant mediating role (accounting for 50.58% of the mediating effect). The economic growth of RBCs is highly coupled with environmental pollution, forming a vicious cycle that is difficult to break [46]. The digital economy provides strong support for R&D and the application of green technologies, thereby enabling the precise identification of pollution sources, the optimization of energy use efficiency, and the acceleration of the iteration of clean production processes [47]. This indicates that the digital economy does not enhance resilience in a general sense, but rather facilitates a qualitative increase in urban resilience by helping RBCs resolve the core contradiction between the economy and the environment through a greening pathway. In contrast, although industrial structure rationalization has a positive mediating effect, its contribution to the mediating effect of the digital economy on urban resilience is only 1.18%. The structural transformation of RBCs demonstrates a protracted nature, wherein the digital economy’s capacity to optimize inter-firm resource allocation manifests delayed effects. The fundamental reconfiguration of these cities’ deeply embedded resource-dependent industrial frameworks constitutes a gradual evolutionary process.

Furthermore, this study reveals a suppression effect that is associated with industrial structure advancement. Although the digital economy promotes industrial structure advancement (a shift from secondary to tertiary industry), this advancement itself has a negative effect on urban resilience. This finding may be due to the fact that in many RBCs, this industrial advancement might represent a low-quality, premature form of deindustrialization [48]. The development of the service sector, lacking the driving force of high-tech manufacturing, tends to gravitate toward lower-end activities and fails to establish effective industrial synergy and technological linkages with it. This shift from a real to a virtual economy, far from enhancing economic efficiency, may instead crowd out the real economy, triggering industrial hollowing-out and structural unemployment, thereby undermining the city’s economic and social resilience [49]. These findings emphasize that in RBCs, the quality of industrial transformation is more important than its speed.

With respect to the spatial effect analysis, the digital economy is shown to exert a spatial influence on RBC’s resilience, with spillover effects between cities. Unlike traditional economic factors, digital technology and information can cross administrative boundaries at a very low cost, thereby facilitating regional knowledge diffusion and technological collaboration [50]. Consequently, the digital economy can positively influence the resilience of other cities in spatial terms. This finding indicates that the resilience building of RBCs should not be isolated; the enhanced resilience of a single city can benefit surrounding areas through digital networks, making the construction of cross-regional digital economy collaboration zones and green innovation corridors crucial for achieving holistic and sustainable regional transformation. The transformation of RBCs should transcend a single-city perspective and actively integrate into the national “regional coordinated development” strategy.

5.2. Recommendations

First, the government should recognize the critical role of GTI in enhancing RBC’s resilience. It should secure funding for new energy technologies, environmental governance innovation, and GTI, while optimizing the structure of scientific expenditure to prioritize research and application in digital innovation technologies and GTI. Targeted implementation strategies should be adopted for different types of RBCs. For regenerative RBCs, where industrial transformation has achieved initial success, the policy focus should shift towards supporting the high-end integrated application of digital and GTI, and fostering advanced green manufacturing clusters and high-end producer services. For growing RBCs, which are still in a phase of rising resource extraction, industries should be guided towards green and digital transformation. For declining RBCs facing resource depletion and economic contraction, efforts should continue to develop digital solutions and GTI that address pressing social and environmental issues, thereby laying the foundation for economic recovery.

Second, in response to the development status of RBCs industries and environmental issues, cities themselves should adopt differentiated strategies to develop their digital economies. For declining RBCs, which lack endogenous drivers of development, it is crucial to vigorously promote their own digital economy, actively integrate it with industrial, economic and environmental, rely on the digital economy as a key pillar, cultivate new digital industries, and revitalize their urban economic development. For growing RBCs, digital technologies should be leveraged to boost the sustainable development of core industries, promote the digital transformation of traditional manufacturing and mining sectors, and enhance production efficiency, supply chain resilience, and environmental monitoring capabilities through digital tools, thereby strengthening the competitiveness and sustainability of their dominant industries. Regenerative RBCs need to proactively align with the national “regional coordinated development” strategy, commit to building cross-regional digital economy collaboration zones and green innovation corridors, and utilize their first-mover advantages to drive synergistic development across broader areas.

Third, governments should take full advantage of the spatial radiation effect of the digital economy to create a favorable pattern for its coordinated development. Innovation is the key to digital technology development. Local governments should, on the basis of their own resource endowments and guided by the importance of the national development positioning that has been granted to RBCs. It is essential to strengthen digital governance coordination within urban agglomerations. It is recommended that under the framework of national-level urban agglomerations such as Beijing-Tianjin-Hebei and Yangtze River Delta, RBCs and central cities should be promoted to co-construct digital infrastructure and share data resources to avoid redundant construction. Establishing GTI and industrial transformation alliances across administrative boundaries is advised. RBCs with geographical proximity and complementary industrial endowments should be encouraged to establish regular collaboration mechanisms, jointly plan and develop digital economy belts and GTI corridors, and institutionalize and materialize the spatial spillover effects of the digital economy through policy coordination, park co-construction, and talent sharing, ultimately achieving overall regional sustainable development.

Fourth, the sustainable transformation capacity of RBCs has become a core topic in global sustainable development research. Existing research indicates that Europe’s development philosophy has evolved from emphasizing engineering resilience, which focuses on returning to the original state, to evolutionary resilience, which prioritizes adaptation and transformation. The successful transformation of Germany’s Ruhr Region demonstrates that traditional industries should not be simply phased out, but rather, deep integration between advanced manufacturing and knowledge-intensive services should be achieved through building regional innovation networks. This aligns with the findings of our study regarding resilience weakening caused by low-quality deindustrialization. Rotterdam in the Netherlands has organically integrated climate resilience building with circular economy development, illustrating that ecological governance can also serve as an opportunity to cultivate new economic growth points, providing a reference for Chinese RBCs in advancing ecological governance and industrial upgrading [51]. The European Union’s achievement of regional synergistic development through its Cohesion Policy strongly corroborates the necessity and feasibility of constructing cross-regional collaboration zones as proposed in this study.

In summary, the policy recommendations for enhancing RBC’s resilience that have been proposed in this study are applicable not only to Chinese RBCs but also to other international RBCs, such as Germany’s Ruhr Region, Rotterdam in the Netherlands, and Houston in the United States. These regions can flexibly adapt and apply these recommendations based on their specific resource endowments, current economic development stage, and regional development positioning to formulate tailored resilience development strategies. Therefore, the policy recommendations presented in this study not only provide guidance for the RBCs in China but also offer a scientific reference for the sustainable development of other resource-based regions globally.

6. Conclusions

This study systematically reveals the evolutionary patterns, impact mechanisms, and spatial effects of the digital economy and its impact on urban resilience in Chinese RBCs during the period of 2003 to 2022. The main conclusions are as follows.

First, both the digital economy and the urban resilience levels of RBCs tended to increase. However, 80% of RBCs remain at a low level of resilience, with their resilience indices ranging from 0.15 to 0.40. Moreover, RBCs that perform well in both aspects are primarily concentrated in the eastern region of China.

Second, the digital economy of RBCs has a significant positive effect on their urban resilience, with an impact coefficient of 0.117. Furthermore, to interpret the more detailed impact of the digital economy on RBC’s resilience, fixed-effects regressions between the digital economy and each dimension of resilience were conducted. The results indicate that the digital economy has the strongest positive impact on social resilience, with a correlation coefficient of 0.392, followed by economic resilience (0.179), resource resilience (0.095), infrastructure resilience (0.090), and ecological resilience (0.085).

Third, this study reveals that GTI plays a core mediating role in the impact of the digital economy on urban resilience, accounting for 50.58% of the mediating effect. Industrial structure rationalization also has a partial mediating effect, but its contribution is only 1.18%. Industrial structure advancement has a “suppression effect” and its mediating role is nonsignificant. The positive impact of the digital economy on RBC’s resilience exhibits spatial spillover effects. Both the digital economy and the urban resilience of RBCs demonstrate spatial transmission; the development of the local digital economy can drive the growth of cities, and the enhancement of local resilience can also radiate to surrounding cities. Spatial effect decomposition confirms that the development of the digital economy in RBCs can itself indirectly increase the resilience level of adjacent cities through spatial mechanisms.

Although this study achieves certain results in analyzing the impact of China’s digital economy on RBCs, several limitations remain. Future research can be improved and expanded in the following aspects. First, the research scope of the current study is limited. This study is solely focused on Chinese RBCs and lacks comparisons with global resource-based regions. Future research can broaden the study scope and conduct comparative analyses with other resource-based regions worldwide to better understand the differences among impacts and their underlying implications. Second, this study has data limitations. The research data focused on the prefecture-level city scale, thus making it difficult to capture the detailed characteristics of county-level units. Future studies can integrate multisource data and incorporate more granular data units to assess the impact more precisely.